^^
United States
Environmental Protection
Agency
Air and Radiation EPA420-P-04-020
December 2004
MOVES2004 Highway
Vehicle Population and
Activity Data
Draft
-------
EPA420-P-04-020
December 2004
Draft
Assessment and Standards Division
Office of Transportation and Air Quality
U.S. Environmental Protection Agency
Megan Beardsley
Dave Brzezinski
Bob Gianelli
John Koupal
Sujan Srivastava
-------
Table of Contents
1. Introduction 6
2. Data Sources 9
2.1. VIUS97 9
2.2. Polk NVPP® and TIP® 9
2.3. FHWAHighway Statistics 9
2.4. FTA National Transit Database 9
2.5. School Bus Fleet Fact Book 9
2.6. MOBILE6 9
2.7. Annual Energy Outlook & National Energy Modeling System 10
2.8. Transportation Energy Data Book 10
2.9. Oak Ridge National Laboratory Light-duty Vehicle Database 10
3. SourceTypeYear 11
3.1. SourceTypePopulation 11
3.2. SalesGrowthFactor 14
3.3. MigrationRate 16
4. SourceTypeModelYear 17
5. SourceTypeAge 18
5.1. SurvivalRate 18
5.2. Relative MAR 19
5.3. FunctioningACFraction 21
6. SourceTypeAgeDistribution 23
6.1. Motorcycles 23
6.2. Passenger Cars 23
6.3. Trucks 23
6.4. Intercity Buses 24
6.5. School Buses and Motor Homes 24
6.6. Transit Buses 24
7. SourceBinDistribution 26
7.1. Motorcycles 27
7.2. Passenger Cars 27
7.3. Trucks 29
7.4. Buses 36
7.5. Refuse Trucks 41
7.6. Motor Homes 42
7.7. SourceBinDistributions for 2000-and-later 45
8. SourceUseType 55
8.1. SourceMass 55
8.2. Road Load Coefficients 56
9. RoadTypeDistribution 59
10. Average Speed Distribution 61
ll.HPMSVTypeYear 63
H.l.HPMSBaseYearVMT 63
11.2. BaseYearOffNetVMT 63
11.3. VMTGrowthFactor 63
-------
12. Temporal Distributions of VMT 67
12.1. MonthVMTFraction 67
12.2. DayVMTFraction 67
12.3.HourVMTFraction 68
IS.DriveSchedule 70
14. Drive Schedule Association 72
15. SourceTypeHour 74
IS.l.StartsPerSHO 74
15.2. IdleSHOFactor 78
16. ZoneYearRoadType 79
17. ZoneYear 80
17.1. StartAllocFactor 80
17.2. IdleAllocFactor 80
18. SCCVTypeDistribution 82
19. MonthGroupHour 83
20. ZoneMonthHour 84
21. Fuel Types 86
21.1.FuelSubType 86
21.2. FuelSupply 87
22. Peer Review 89
23. References 94
-------
List of Tables and Figures
Table 1-1. MOVES SourceTypes 7
Table 1-2. MOVES Database Elements Covered in This Report 8
Table 3-1. Vehicle Population Comparisons 1999 12
Table 3-2. Adjusted Vehicle Populations 12
Table 3-3. VIUS97 Codes Used for Distinguishing Truck SourceTypes 13
Table 3-4. 1999 Truck SourceType Distribution and Populations 13
Table 3-5. Bus Population Comparisons 1999 14
Table 3-6. 1999 SourceType Populations for MOVES 15
Table 3-7. SalesGrowthFactor by Calendar Year and Use Type 16
Table 5-1. SurvivalRate by Age and SourceType 19
Table 5-2. Equations for Calculating Annual Mileage Accumulation Rates 21
Table 5 -3. FunctioningACFraction by Age (All Use Types Except Motorcycles) 22
Table 6-1. 1999 Age Fractions for MOVES Source Types 25
Table 7-1. Data Tables Used by SourceBinGenerator 26
Table 7-2. Motorcycle Engine Size and Average Weight Distributions for Selected Model Years 27
Table 7-3. Mapping Polk Fuel Codes to MOVES 28
Table 7-4. Mapping VIUS97 ENGTYP to MOVES FuelTypelD 29
Table 7-5. Diesel Fractions for Trucks 30
Table 7-6. Mapping VIUS97 Engine Size Categories to MOVES EngSizelD 31
Table 7-7 Fraction of Light-Duty Trucks among Gasoline-Fueled Trucks 35
Table 7-8. Fraction of Light-Duty Trucks among Diesel-fueled Trucks 36
Table 7-9. Mapping National Transit Database Fuel Types to MOVES Fuel Types 37
Table 7-10. Fuel Fractions for Transit Buses 38
Table 7-11. Fuel Fractions for School Buses 39
Table 7-12. FTA Estimate of Bus Weights 39
Table 7-13. California School Buses 40
Table 7-14. Weight Distributions for Buses by Fuel Type 41
Table 7-15. Fuel Fractions for Refuse Trucks by Model Year 41
Table 7-16. Refuse Truck SizeWeight Fractions by Fuel Type 42
Table 7-17. Diesel Fractions for Motor Homes 43
Table 7-18. Weight Fractions for Diesel Motor Homes by Model Year 44
Table 7-19. Weight Fractions for Gasoline Motor Homes by Model Year 45
Table 7-20. Supported Fuels and Technologies for 2000-and-later Model Years 46
Table 7.21. Fuel and Engine Technology Fractions for 2000-and-later Passenger Cars 49
Table 7.22. Fuel and Engine Technology Fractions for 2000-and-later Light Trucks 50
Table 7.23. Fuel and Engine Technology Fractions for 2000-and-later Buses 51
Table 7.24. Fuel and Engine Technology Fractions for 2000-and-later Motor Homes and Single-Unit
Short-haul and Long-haul Trucks 52
Table 7.25. Fuel and Engine Technology Fractions for Refuse Trucks and Short-haul and Long-haul
Combination Trucks 54
Table 8-1. MOVES Weight Classes 56
Table 8-2. Road Load Coefficients for Heavy-Duty Trucks, Buses, and Motor Homes 57
Table 8-3. SourceUseType Characteristics 58
Table 9-1. Road Type Codes in MOVES 59
Table 9-2. Road Type Fractions by HPMS Vehicle Type 60
Table 10-1. Mapping of MOVES Road Types to MOBILE6 Road Types 61
Table 10-2. MOVES Speed Bin Categories 62
Table 11-1. BaseYearVMT and VMTGrowthFactor by HPMS Vehicle Class 64
-------
Table 11-2.2003 and later VMTGrowthFactors for Medium-Duty & Heavy-Duty Trucks 65
Table 11-3.2003 and later VMTGrowthFactor Calculation for Passenger Cars and Trucks 66
Table 12-1. MonthVMTFraction 67
Table 12-2. DayVMTFractions 68
Table 12-3. HourVMTFractions 69
Table 13-1. Default MOVES Drive Schedules 71
Table 14-1. Drive Schedule Mapping 72
Table 15-1. Data Sources for Trip Starts Per Source Hour of Operation (SHO) 75
Table 15-2. MOBILE6 Vehicle Classifications 76
Table 15-3. MOBILE6 Starts Per Day, Miles Driven Per Day and Average Speed And Calculated Starts
Per Source Hour Operating 77
Table 18-1. SCC Mappings for Selected SourceTypes 82
Figure 19-1: Air Conditioning Activity Demand as a Function of Heat Index 83
Table 21-1. Fuel Types 86
Table 21-2. Fuel SubTypes 87
Table 21-3. FuelSupply Table Description 87
-------
1. Introduction
This report documents the default "fleet" and "activity" data used by MOVES2004 in
order to estimate energy consumption and emissions of methane (CH/j) and nitrous oxide (N2O)
for all on-road sources from calendar years 1999 through 2050, for each county in the U.S. Fleet
data refers to information characterizing the vehicle fleet such as population estimates, age
distributions, survival rates, sales growth rates, and distribution across "source bins" used to
estimate energy and emissions. Activity data refers to information characterizing how the fleet
operates, such as: vehicle miles traveled (VMT), VMT growth, average speed distributions, and
driving patterns.
The report focuses on the data sources for fleet and activity data and methodology used to
produce the default estimates. The base year for MOVES2004 is 1999, so most of the data is
anchored to this year; sales and VMT growth rates which allow projection through 2050 are also
documented as well. All of the fleet and activity data discussed in this report are contained in a
series of data tables in the MOVES Default database. Where space allows, the resulting default
data are also presented in this report; otherwise the reader is directed to the database itself to
view the data. The report is structured so that each section (for Sections 3 through 19) is
centered on a different database table (entity); and the subsections are the data fields (attributes)
within that table. This report focuses just on the data and methods used to populate fleet and
activity data - it does not document the structure of the database itself, or how the data is used in
the MOVES2004 calculations. This information is contained in the separate document,
"MOVES2004 Software Design Reference Manual"; the reader is encouraged to first read this
manual in order to fully understand the context of the data presented in this report.
While many of the fleet and activity data concepts will be familiar to users of MOBILE
(e.g. VMT), MOVES2004 does introduce several new concepts with regard to vehicle
classification and activity characterization. There are two primary reasons for this: first, the
MOVES design is substantially different from MOBILE in order to support multi-scale analysis,
and second MOVES is designed to reconcile internally fundamental differences between how
activity data is collected and characterized, and how emission data is collected and characterized.
With regard to multi-scale analysis, MOVES uses a "modal" approach to estimating energy and
emissions based on discrete vehicle power bins, and characterizes energy rates on a time basis
(e.g. grams per hour) instead of the traditional mile basis (e.g. grams per mile). This approach
requires activity data to generate the distribution of activity in modal bins, and for conversions of
mile-based activity data (VMT) to time-based activity data (e.g. source hours operating, or
SHO); the process for this is discussed in detail in the Software Design Reference Manual.
With regard to reconciling differences between activity and emission data, a long-standing
challenge in the generation of on-road mobile source emission inventories is the disconnect
between how vehicle activity data sources characterize vehicles and how emission or fuel
economy regulations characterize vehicles. An example of this is how vehicles are characterized
by the Highway Performance Monitoring System (HPMS) - by a combination of the number of
tires and axles - and EPA's weight-based emission classifications such as LDV, LDT1, LDT2
etc.
-------
Reconciling activity and emissions data generally requires "mapping" between the two.
The MOBILE series of models have traditionally grouped vehicles according to the EPA
emission classifications, and provided external guidance on mapping these categories to the
sources of activity data, such as HPMS. MOVES is designed to take these mappings into
account internally, so that the casual user of MOVES will not have to deal with external
mapping. Doing this, however, requires some complexity in the design. Vehicles are
characterized both according to activity patterns and energy/emission performance, and are
mapped internal to the model. Thus the model uses data for both the activity and
energy/emission methods of characterization. On the activity side, vehicles are grouped into
"Source Use Types", or use types, defined as groups expected to have unique activity patterns.
Because HPMS data is a fundamental source of activity, the MOVES use types are defined as
subsets of HPMS vehicles classifications. These use types are shown in Table 1-1. The majority
of activity data presented in this document are based on these classifications.
To characterize factors important for energy consumption and emissions, the MOVES
design has implemented the concept of "Source Bins". Unique source bins are defined by those
characteristics with the largest influence on fuel (energy) consumption and emissions. Source
bins are defined completely separate from use types, but are mapped to source use types internal
to MOVES by the Source Bin Distribution Generator, discussed in the Software Design
Reference Manual. The distributions of source bin attributes (e.g. fuel type, vehicle weight and
engine size) used to generate the overall mapping of source bins to source use types are also
included in Section 7 of this document. The energy and emission rates themselves are
documented in a separate report, "MOVES2004 Energy and Emission Inputs".
The data tables and fields discussed in this report are shown in Table 1-2.
Table 1-1. MOVES SourceTypes
SourceType ID
11
21
31
32
41
42
43
51
52
53
54
61
62
SourceType
Motorcycles
Passenger Cars
Passenger Trucks (primarily personal use)
Light Commercial Trucks (other use)
Intercity Buses (non-school, non-transit)
Transit Buses
School Buses
Refuse Trucks
Single Unit Short-haul Trucks
Single Unit Long-haul Trucks
Motor Homes
Combination Short-haul Trucks
Combination Long-haul Trucks
HPMS Vehicle Class
Motorcycles
Passenger Cars
Other Two-Axle/Four Tire, Single Unit
Other Two-Axle/Four Tire, Single Unit
Buses
Buses
Buses
Single Unit
Single Unit
Single Unit
Single Unit
Combination
Combination
"Long-haul" trucks are defined as trucks for which most trips are 200 miles or more.
7
-------
Table 1-2. MOVES Database Elements Covered in This Report
Database Table Name*
SourceTypeYear
SourceTypeModelYear
SourceTypeAge
SourceTypeAgeDistribution
SourceBinDistribution*
SourceUseType
RoadTypeDistribution
AvgSpeedDistribution
HPMSVtypeYear
Month VMTFraction
DayVMTFraction
HourVMTFraction
Drive Schedule
Drive Schedule Second
DriveScheduleAssociation
SourceTypeHour
ZoneYearRoadType
ZoneYear
SCCVTypeDistribution
MonthGroupHour
Fields
SourceTypePopulation
SalesGrowthFactor
MigrationRate
ACPenetrationFraction
SurvivalRate
RelativeMAR
FunctioningACFraction
AgeFraction
SourceBinActivityFraction
RollingTerm
RotatingTerm
DragTerm
SourceMass
RoadType VMTFraction
AvgSpeedFraction
HPMSBaseYearVMT
BaseYearOffNetVMT
VMTGrowthFactor
Month VMTFraction
DayVMTFraction
HourVMTFraction
Average Speed
Speed
SourceTypelD
RoadTypelD
Drive SchedulelD
IsRamp
StartsPerSHO
IdleSHOFactor
SHOAllocFactor
IdleAllocFactor
StartAllocFactor
SCCVTypeFraction
AC Activity Terms (A, B & C)
*See also Table 7-1, listing tables and fields used by the SourceBinGenerator.
8
-------
2. Data Sources
A number of organizations collect data relevant to this report. The most important
sources used to populate the vehicle population and activity portions of MOVES database are
described here. These sources are referred to throughout this document by the abbreviated name
given in this description, but the reference citation is only given here.
2.1. VIUS97
Every five years the U.S. Census Bureau conducts the Vehicle Inventory and Use Survey
(VIUS)1 to collect data on the physical characteristics and activity of U.S. trucks. The 1997
survey is a sample of private and commercial trucks that were registered in the U.S. on July 1,
1997. The survey excludes automobiles, motorcycles, government-owed vehicles, ambulances,
buses, motor homes and nonroad equipment. For MOVES, VIUS97 provides information to
characterize trucks by SourceType and to estimate age distributions.
2.2. Polk NVPP® and TIP®
R.L. Polk & Co. is a private company providing automotive information services. The
company maintains two databases relevant for MOVES: the National Vehicle Population Profile
(NVPP®)2 and the Trucking Industry Profile (TIP®Net) Vehicles in Operation database.3 The
first focuses on light-duty cars and trucks, the second focuses on medium and heavy-duty trucks.
Both compile data from state vehicle registration lists. For MOVES2004, EPA is using the 1999
NVPP® and TIP®.
2.3. FHWA Highway Statistics
Each year the Federal Highway Administration's (FHWA) Office of Highway Policy
Information publishes Highway Statistics. This volume summarizes a vast amount of roadway
and vehicle data from the states and other sources. For MOVES, we will use data on vehicle
registrations and vehicle miles traveled (VMT), summarized in four tables. 4 5 6 7. Hereafter,
references will be to FHWA MV-1, MV-10, VM-1, and VM-2. For the 1999 base year, we used
Highway Statistics 1999. For 2000-2002 VMT growth rates, we used the 2000, 2001 and 2002
versions.
2.4. FTA National Transit Database
The Federal Transit Administration (FTA) summarizes financial and operating data from
U.S. mass transit agencies in the National Transit Database (NTD).8 For MOVES2004, we used
1999 data from the report, "Age Distribution of Active Revenue Vehicle Inventory: Details by
Transit Agency."
2.5. School Bus Fleet Fact Book
The School Bus Fleet 1999 Fact Book includes estimates, by state, of number of school
buses and total miles traveled.9 The Fact Book is published by Bobit Publications.
2.6. MOBILE6
In some cases, we have been able to use data from MOBILE6 with only minor
adaptation. The MOBILE6 data is documented in technical reports, particularly M6.FLT.002
-------
"Update of Fleet Characterization Data for Use in MOBILE6 - Final Report."10 Additional
MOBILE6 documentation is available on the web at http://www.epa.gov/otaq/m6.htm
2.7. Annual Energy Outlook & National Energy Modeling System
The Annual Energy Outlook (AEO) u describes Department of Energy forecasts for
future energy consumption. The National Energy Modeling System (NEMS) is used to generate
these projections based on economic and demographic projections. We used AEO2004 to
forecast VMT growth and vehicle sales growth.
2.8. Transportation Energy Data Book
Each year, Oak Ridge National Laboratory produces the DOE Transportation Energy
Data Book (TEDB). This book summarizes transportation and energy data from a variety of
sources. For MOVES2004, we relied on Edition 22, published in September 200212 and Edition
23, published in October 2003.13
2.9. Oak Ridge National Laboratory Light-duty Vehicle Database
Oak Ridge National Laboratory Center for Transportation Analysis has compiled a
database of light-duty vehicle information which combines EPA Test vehicle data and Ward's
Automotive Inc. data spanning 1976 - 2001.14 We used this database to determine weight
distributions for light trucks by model year.
10
-------
3. SourceTypeYear
SourceTypeYear consist of three fields—SourceTypePopulation, SalesGrowthFactor,
and Migration Rate. Each field is described in terms of what information it contains, the
sources of the data used for the field, and, when applicable, tables providing certain data points
used in determining the field parameters.
3.1. SourceTypePopulation
The SourceTypePopulation field stores the total population of vehicles by SourceType
for a given base year and domain: in this case, the entire United States in 1999. Some of the
values are taken directly from the indicated sources; other values needed to be derived from
available data and are not found explicitly in any of the data sources.
SourceTypePopulation provides base year populations and provides the basis for Total
Activity Generator calculation of populations in calendar years after the base year. These
populations are, in turn, used to generate travel fractions by age and SourceType and to allow
allocation of VMT by age.
The primary sources for calendar year 1999 vehicle population data are the FHWA
Highway Statistics Tables MV-1 and MV-10 and the Polk NVPP® and TIP® databases. The
Transportation Energy Data Book (TEDB22) explains three factors that account for differences
between the two sources:
1. Polk data includes only vehicles that were registered on July 1 of 1999. FHWA data
includes all vehicles that have been registered at any time throughout the year and
thus may include vehicles that were retired during the year or may double count
vehicles registered in two or more states.
2. Polk and FHWA may differ in how they classify some minivans and SUVs as trucks
or automobiles. (This difference is less important since 1990).
3. FHWA includes all non-military Federal vehicles. Polk data includes only those
Federal vehicles that are registered within a state.
Also, FHWA data is available for Puerto Rico, but Puerto Rico does not appear to be
included in our Polk data set. MOVES will cover Puerto Rico and the Virgin Islands. In
addition, Polk collects data on Gross Vehicle Weight (GVW) class 3 vehicles in both the
NVPP® and TIP® databases, but the values are not the same. Polk staff recommended using the
TIP® values.15 Finally, our Polk data set includes school buses and motor homes (which can be
counted separately), but does not include "non-school buses."
The Department of Transportation's National Household Transportation Survey (NHTS)
combines the previous National Personal Transportation Survey and the American Travel Survey
to collect data on personal travel patterns and includes data on both personal trucks and
automobiles.16 This data is included in Table 3-1, but is not used in MOVES because it is two
years newer than the FHWA and Polk data, and it excludes non-household vehicles. Values
from the three sources are compared in Table 3-1.
11
-------
Table 3-1. Vehicle Population Comparisons 1999
Data Source
FHWA (w Puerto Rico
and Federal vehicles)
FHWA (w/o Puerto Rico
and Federal vehicles)
Polk NVPP® & TIP®
NETS (2001)
Motorcycles
4,173,869
na
4,951,747
Automobiles
134,480,432
131,076,551
126,868,744
115,914,908
Trucks (total)
83,178,092
81,060,369
80,323,528*
80,499,939
Buses (total)
732,189
na
Motor
Homes
na
902,949
1,446,469
* Excluding motor homes and NVPP® GVW3 trucks.
For automobiles and trucks, it is possible to do a direct comparison of Polk and FHWA
data. To estimate the MOVES population, we adjust the FHWA data to account for double-
counting by multiplying the total FHWA population by the ratio of the Polk population to the
FHWA population without Federal vehicles and Puerto Rican vehicles.
Adjusted Population = FHWA w federal & PR * (Polk/FHWA w/0 federal & PR)
This leads to the values in Table 3-2.
Table 3-2. Adjusted Vehicle Populations
Automobiles
Trucks (total)
Population
130,163,354
83,007,993
For MOVES, total trucks are sub-classified into seven SourceTypes. The proportion of
total trucks in each subtype was estimated using VIUS97 responses for Axle Arrangement,
Primary Area of Operation, Body Type and Major Use as detailed in Table 3-3.
With these definitions and with vehicles that lack AREAOP codes assigned
proportionally to the corresponding SourceTypes, we computed the distributions in Table 3-4.
These distributions were multiplied by the total truck population from Table 3-2 to derive
population values for MOVES.
12
-------
Table 3-3. VIUS97 Codes Used for Distinguishing Truck SourceTypes.
SourceType
Passenger Trucks
Light Commercial
Trucks
Refuse Trucks
Single Unit Short-
haul Trucks
Single Unit Long-
haul Trucks
Combination Short-
haul Trucks
Combination Long-
haul Trucks
Axle Arrangement
2 axle/4 tire (AXLRE=
1,5,6,7)
2 axle/4 tire (AXLRE=
1,5,6,7)
Single Unit (AXLRE =
2-4, 8-16)
Single Unit (AXLRE =
2-4, 8-16)
Single Unit (AXLRE =
2-4, 8-16)
Combination (AXLRE
>=17)
Combination (AXLRE
>=17)
Primary Area of
Operation
any
any
off-road, local or short-
range (AREAOP <=4)
off-road, local or short-
range (AREAOP <=4)
long-range (AREAOP
>=5)
off-road, local or
medium (AREAOP <=4)
long-range (AREAOP
>=5)
Body Type
any
any
garbage hauler
(BODTYP=30)
any except
garbage hauler
any except
garbage hauler
any
any
Major Use
personal
transportation
(MAJUSE=20)
any but personal
transportation
any
any
any
any
any
Table 3-4. 1999 Truck SourceType Distribution and Populations
SourceType
Passenger Trucks
Light Commercial Trucks
Refuse Trucks
Single Unit Short-haul Trucks
Single Unit Long-haul Trucks
Combination Short-haul Trucks
Combination Long-haul Trucks
Total
Percent
68.90%
23.02%
0.11%
5.39%
0.32%
1.31%
.97%
100.00%
Population
57,190,192
19,106,257
88,607
4,470,798
264,435
1,084,366
803,337
83,007,993
For buses, we needed to distribute the total buses from FHWA to the three MOVES
classes. Additional information on bus numbers was available from the FTA NTD, Polk TIP®,
and the School Bus Fleet Fact Book, and the American Bus Association "Motorcoach Census
17
2000". The FTA NTD provides population numbers for a variety of transit options. To
determine the number of transit buses, we summed their counts for Articulated Motor Buses,
Motor Bus Class A, B & C, and Double Decked buses.
-------
Table 3-5. Bus Population Comparisons 1999
Data Source
FHWAMV-1
FHWAMV-10
(excludes PR)
FTA NTD
Polk TIP®
School Bus Fleet Fact
Book
Motorcoach Census**
Total Buses
732,189
728,777
Intercity Buses
44,200
Transit Buses
55,706
School Buses
592,029*
460,178
429,086
* Includes some church & industrial buses.
** Includes Canada.
As Table 3-5 shows, estimates of school bus numbers vary. We chose to use the FHWA
value because it includes church and industrial buses that we believe have activity patterns more
similar to school buses than to intercity buses. To calculate the number of buses for the
categories needed for MOVES, we used the FHWA school bus value and the FTA transit bus
value. We assigned the remaining total FHWA buses (732,189-592,029-55,706 = 84,454) to the
intercity category. Note this value substantially exceeds the estimate of intercity buses provided
by the Motorcoach Census.
For the remaining categories, motorcycles and motor homes, we used the only available
data. For motorcycles we used the FHWA value from table MV-1. For motor homes we used the
population from the Polk TIP® database.
Table 3-6 summarizes the 1999 vehicle populations proposed for use in MOVES2004.
3.2. SalesGrowthFactor
The SalesGrowthFactor field stores a multiplicative factor indicating changes in sales by
SourceType for calendar years after the base year. It determines the number of new vehicles
added to the vehicle population each year, and is expressed relative to the previous year's sales.
For example, 1 means no change from previous year sales levels, 1.02 means a two percent
increase in sales, and 0.98 means a two percent decrease in sales. SalesGrowthFactor is used in
the Total Activity Generator calculation of source type populations for calendar years after the
base year, meaning calendar years 2000 through 2050 in MOVES2004.
14
-------
Table 3-6. 1999 SourceType Populations for MOVES
SourceType ID
11
21
31
32
41
42
43
51
52
53
54
61
62
SourceType
Motorcycles
Passenger Cars
Passenger Trucks
Light Commercial Trucks
Intercity Buses
Transit Buses
School Buses
Refuse Trucks
Single Unit Short-haul Trucks
Single Unit Long-haul Trucks
Motor Homes
Combination Short-haul Trucks
Combination Long-haul Trucks
1999 Population
4,173,869
130,163,354
57,190,192
19,106,257
84,454
55,706
592,029
88,607
4,470,798
264,435
902,949
1,084,366
803,337
SalesGrowthFactor estimates were derived from actual sales data from TEDB23 (2003),
whose primary source is Ward's Motor Vehicle Facts and Figures, and from sales projections
from AEO2004. Beyond 2025, SalesGrowthFactor was set to 1, indicating no growth in sales.
The data sources and methodology by source use type are detailed following:
• Passenger Cars: SalesGrowthFactors for calendar year 2000 and 2001 were derived
from total sales numbers reported in the TEDB23 Table 4.5. Factors for calendar
years 2002 through 2025 were derived from new car sales estimates of presented in
AEO2004 Supplemental Table 45, generated by NEMS.
• Motorcycles: SalesGrowthFactors for calendar year 2000 and 2001 were computed
1 &
from sales values in the Motorcycle Industry Council Statistical Annual.
SalesGrowthFactors for years 2002 through 2025 were set equal to those for
passenger cars.
• Passenger Trucks/Commercial Trucks: SalesGrowthFactor for calendar year 2000 and
2001 were derived from total sales numbers reported in the TEDB23 Table 4.6.
Factors for Calendar year 2002 through 2025 were derived from new light truck sales
estimates presented in AEO2004 Supplemental Table 45, generated by NEMS.
• Buses, Single Unit Trucks, Motor Homes: Calendar years 2000-2001 based on sales
as reported in TEDB23 Table 5.3 (gross weight range 10,000-33,000 Ibs). Years
2002 through 2025 calculated from medium-duty truck sales projections from
AEO2004 Supplemental Table 55.
• Combination Trucks, Refuse Trucks: Calendar years 2000-2001 based on sales as
reported in TEDB23 Table 5.3 (gross weight range 33,001 and greater pounds).
Years 2002 through 2025 calculated from heavy-duty truck sales projections found in
AEO2003 Supplemental Table 55.
-------
The resulting SalesGrowthFactors by use type are shown in Table 3-7:
Table 3-7. SalesGrowthFactor by Calendar Year and Use Type
Calendar Year
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
Passenger Cars,
Motorcycles
1.017*
0.952*
0.962
0.980
1.001
1.017
1.010
1.002
0.996
0.995
0.997
1.001
0.994
1.001
0.996
0.997
1.006
1.011
1.010
1.005
1.003
0.992
0.999
1.003
1.004
1.005
Passenger Trucks
Light Comm.
Trucks
1.039
1.037
1.001
0.966
1.067
1.028
1.024
1.014
1.011
1.010
1.016
1.018
1.015
1.022
1.013
1.013
1.022
1.026
1.027
1.021
1.020
1.007
1.014
1.020
1.022
1.025
All Buses,
Single-Unit Trucks,
Motor Homes
0.968
0.850
0.821
0.981
.050
.107
.085
.031
.017
.006
.002
.000
.005
.018
.010
.003
.004
.017
.020
.003
.007
0.996
.008
.016
.018
.023
Refuse Trucks,
Combination
Trucks
0.809
0.660
1.043
0.925
.050
.162
.121
.025
.020
.006
0.993
0.987
1.002
1.020
1.013
0.996
0.989
1.011
1.018
0.989
0.997
0.988
1.005
1.017
1.018
1.027
* The table values for2000&2001 apply only to cars. Motorcycle values are 1.317 and 1.197,
respectively.
MOBILE6 also projected vehicle sales in order to calculate vehicle counts46. MOVES
SalesGrowthF actors are based on more recent information.
3.3. MigrationRate
The MigrationRate field stores a yearly multiplicative factor used to estimate how many
vehicles join or leave the population of a SourceType in the given domain in a given year. We
expect this field will be most useful when modeling emissions on relatively small geographic
scale.
For the initial MOVES release, the domain is the entire U.S. and we are using a
simplifying assumption of no migration: that is, a migration rate of 1.
16
-------
4. SourceTypeModelYear
SourceTypeModelYear stores the field ACPenerationFraction, which is the fraction of
vehicles equipped with air conditioning, by source type and model year. ACPenetrationRate is
used in the calculation of the A/C adjustment in the MOVES emission rate calculators.
Default data used for MOVES2004 is taken directly from MOBILE6. 19 Base market
penetration data by model year were gathered from Ward's Automotive Handbook for light-duty
vehicles and light-duty trucks through the 1995 Model Year. This information was available
from 1972 for cars and from 1975 for trucks. Year-to-year rates are more variable in the first
few years of available data, so values for earlier model years will be estimated by applying the
1972 and 1975 rates for cars and trucks, respectively. In the later years, the rate of increase
becomes more steady. Projections beyond 1995 were developed by taking the average yearly
rate of increase from the last five years of available data and applying them to each subsequent
year until a predetermined cap was reached. A cap of 98% was placed on cars and 95% on
trucks under the assumption that there will always be vehicles sold without air conditioning
systems, more likely on trucks than cars. The caps are in place by the 1999 model year and will
remain for subsequent years. For MOVES, the light-duty vehicle rates were applied to passenger
cars, and the light-duty truck rates were applied to all other use types (except motorcycles, for
which AC penetration is assumed to be zero).
-------
5. SourceTypeAge
Three fields comprise SourceTypeAge in MOVES2004—SurvivalRate, Relative MAR,
and FunctioningACFraction. Each one is described below, including data sources and some
relevant data points used in the model.
5.1. SurvivalRate
The SurvivalRate field describes the fraction of vehicles of a given SourceType and Age
(relative to the total number originally sold) that remain on the road one year to the next.
SurvivalRate is used in the Total Activity Generator in the calculation of source type populations
by age in calendar years after the base year.
The data for all SourceTypes except motorcycles came from the Transportation Energy
Data Book (TEDB22, unchanged for version 23). For Passenger Cars we used survival rates for
the 1990 model year (TEDB22, Table 6.9). For Passenger Trucks and Light Commercial Trucks
we used survival rates for the 1990 model year (TEDB22, Table 6.10).
SurvivalRate for all other SourceTypes were from the Heavy-Duty rates for the 1980
model year (TEDB22, Table 6.11). The 1990 model year rates were not used because they were
significantly higher than the other model years in the analysis (e.g. 45 percent survival rate for 30
year-old trucks), and seemed unrealistically high. While limited data exists to confirm this
judgment, a snapshot of 5-year survival rates can be derived from VIUS 1992 and 1997 results
for comparison. According to VIUS, the average survival rate for model years 1988-1991
between the 1992 and 1997 surveys was 88 percent. The comparable survival rate for 1990
model year Heavy-Duty vehicles from TEDB was 96 percent, while the rate for 1980 model year
trucks was 91 percent. This comparison lends credence to the decision that the 1980 model year
survival rates are more in line with available data.
TEDB22 does not include scrappage rates for GVWR 10,000-26,000 vehicles, so it was
necessary to apply the Heavy-Duty rates to predominantly Medium-Duty use types.
SurvivalRates for motorcycles were calculated based on regression of data provided by
the Motorcycle Industry Council (MIC).20
SurvivalRates are shown in Table 5-1.
The concept of SurvivalRates as used in MOVES differs from that used in the MOBILE6
model46. In MOBILE6, survival rates were applied to the each vehicle class fleet as a whole.
Different survival rates were used for different ranges of calendar years in developing vehicle
counts for MOBILE6. In MOVES, a separate SurvivalRate is applied to each age in each
SourceType fleet. These SurvivalRates by age are based on the observed scrappage of a single
model year (1980 or 1990) over time. These SurvivalRates in MOVES are used for all model
years in a SourceType in all calendar years.
18
-------
Table 5-1. SurvivalRate by Age and SourceType
Age
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Motorcycles
0.99
0.99
0.98
0.97
0.96
0.96
0.95
0.94
0.93
0.92
0.92
0.91
0.90
0.89
0.89
0.88
0.87
0.86
0.85
0.85
0.84
0.83
0.82
0.82
0.81
0.80
0.79
0.78
0.78
0.77
0.76
Passenger Cars
.00
.00
.00
.00
.00
.00
0.99
0.96
0.93
0.89
0.84
0.80
0.75
0.70
0.65
0.60
0.55
0.50
0.45
0.40
0.35
0.31
0.27
0.24
0.20
0.17
0.15
0.12
0.10
0.08
0.07
Passenger Trucks
Light Comm. Trucks
1.00
1.00
1.00
1.00
0.99
0.97
0.94
0.91
0.87
0.83
0.78
0.73
0.68
0.63
0.58
0.53
0.48
0.43
0.38
0.33
0.29
0.25
0.21
0.18
0.15
0.13
0.10
0.08
0.07
0.05
0.04
All Other
SourceTypes
1.00
1.00
1.00
1.00
0.99
0.97
0.95
0.92
0.89
0.86
0.83
0.79
0.75
0.72
0.68
0.64
0.60
0.56
0.52
0.48
0.44
0.41
0.37
0.34
0.31
0.28
0.25
0.22
0.20
0.18
0.16
5.2. Relative MAR
The Relative Mileage Accumulation Rate (Relative MAR) is listed for each MOVES
SourceType and Age. The Relative MAR is computed as the annual MAR divided by the highest
MAR within the HPMS vehicle class. Table 1-2 lists the groupings of the MOVES SourceTypes
within the six HPMS Vehicle Classes. The following discussion refers to the Source Type ID
numbers (often in parentheses) found in this table.
For most SourceTypes, the annual MARs were derived from the MARs developed for
MOBILE6. These were mapped from the MOBILE6 Vehicle Classes to the MOVES
SourceTypes. We then used regression to smooth these initial MARs and to extend the MARs
from 25 to 30 ages.
19
-------
5.2.1. Motorcycles and Passenger Cars
The initial MARs for passenger cars (category 21) and motorcycles (category 11) were
set to equal those in MOBILE6.
5.2.2. Trucks
The initial MARs for truck categories 31, 32, 51, 52, 53, 61, and 62 in MOVES were
calculated based on weighting fractions assigned to the MOBILE6 truck classes. We used
VIUS97 values for Gross Vehicle Weight (PKGVW) to determine weighting fractions by model
year. To separate Light-Duty Trucks 1 and Light-Duty Trucks 2, which are distinguished by
Loaded Vehicle Weights, we used information from the Oak Ridge National Lab Light Duty
Vehicle database. To separate Class 2a and 2b trucks, we used information from the Oak Ridge
National Laboratory Report by Davis and Truitt.21 The initial MARs for the MOVES truck
categories were then calculated as the product of the weighting fractions and the MARs from
MOBILE6.
5.2.3. Buses
For the School Buses (category 43) the initial MARs were taken from the MOBILE6
value for diesel school buses (HDDBS). As in MOBILE6, the same annual MAR was used for
each age. The MOBILE6 value of 9,939 miles per year came from the 1997 School Bus Fleet
Fact Book.
For Transit Buses (category 42), the initial MARs were taken from the MOBILE6 values
for diesel transit buses (UDDBT). This mileage data was obtained from the 1994 Federal
Transportation Administration survey of transit agencies. 22
For Intercity Buses (category 41), the initial MARs were taken from Motorcoach Census
2000.23 The data did not distinguish vehicle age, so the same MAR was used for each age.
5.2.4. Motor Homes
For motor homes (category 54), the initial MARs were taken from an independent
research study24 conducted in October 2000 among members of the Good Sam Club. The
members are active recreation vehicle (RV) enthusiasts who own motor homes, trailers and
trucks. The average annual mileage was estimated to be 4,566 miles. The data did not distinguish
vehicle age, so the same MAR was used for each age.
5.2.5. Calculating Relative MARs
In order to smooth the data and to extend the MARs from the 25 ages in MOBILE6 to the
30 ages in MOVES, we used statistical regression to determine the curves that best fit the data
for years starting in 1997 and going back to 1973 (ages 1 to 25). Table 5-2 presents the resulting
regression equations for each MOVES category. Note, as in MOBILE6, the motorcycle values
were estimated as a linear function to age 12. Ages 13 through 30 are then estimated as a
constant.
20
-------
Table 5-2. Equations for Calculating Annual Mileage Accumulation Rates
MOVES Source Type
Motorcycles
Passenger Cars
Passenger Trucks
Light Commercial Trucks
Refuse Trucks
Single Unit Short-haul Trucks
Single Unit Long-haul Trucks
Motor Homes
Intercity Buses
Transit Buses
School Buses
Combination Short-haul Trucks
Combination Long-haul Trucks
Source
Type ID
11
21
31
32
51
52
53
54
41
42
43
61
62
Regression Equation
na
y=0.1568e-° 0506x
y=0.0002x2 -0.01 18x + 0.2096
y=0.0002x2-0.0129x+0.2196
y=0.8674e-°-1148x
y=0.4289e-° 0990x
y=0.3339e-°-0762x
y=0.0457
y=0.6000
y=0.46659e-°-0324x
y=0.0994
y=0.0016x2-0.0762x +0.9655
y=0.0021x2-0.0887x+1.0496
R2 from
Regression
na*
1.0
0.998
0.998
0.904
0.990
0.864
na
na
na*
na
0.977
0.879
* For Motorcycles and Transit Buses, the equations from MOBILE6 were used
The values calculated from the equations were then used to calculate the relative MARs
by computing the ratio of the value for each SourceType and age to the highest value within the
HPMS class.
5.3. FunctioningACFraction
The FunctioningACFraction field indicates the fraction of the air-conditioning equipped
fleet with fully functional A/C systems, by source type and vehicle age. A value of 1 means all
systems are functional. This is used in the calculation of total energy to account for vehicles
without functioning A/C systems. Default estimates were developed for all source types using
the "unrepaired malfunction" rates used for 1992-and-later model years in MOBILE6.25 These
were applied to all source use types except motorcycles, which were assigned a value of zero for
all years.
21
-------
Table 5-3. FunctioningACFraction by Age (All Use Types Except Motorcycles)
Age
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
FunctioningAC
Fraction
1
1
1
1
0.99
0.99
0.99
0.99
0.98
0.98
0.98
0.98
0.98
0.96
0.96
0.96
0.96
0.96
0.95
0.95
0.95
0.95
0.95
0.95
0.95
0.95
0.95
0.95
0.95
0.95
0.95
22
-------
6. SourceTypeAgeDistribution
This element of the MOVES2004 contains the field AgeFraction, which stores values
that describe the age distribution of a SourceType in the base year. AgeFractions are determined
differently for various SourceTypes, as the following describes.
6.1. Motorcycles
To determine age fractions for motorcycles, we began with Motorcycle Industry Council
estimates of the number of motorcycles in use by model year in 1998. We used the estimates of
sales growth and survival rates to grow these population estimates to 1999, then computed age
fractions. These fractions are summarized in Table 6-1.
6.2. Passenger Cars
To determine age fractions for passenger cars, we began with Polk NVPP® 1999 data on
car registration by model year. However, this data presents a snapshot of registrations on July 1,
1999, and we needed age fractions as of December 31, 1999. To adjust the values, we used
monthly data from the Polk new car database to estimate the number of new cars registered in
the months July through December 1999. Model Year 1998 cars were added to the previous
estimate of "Age 1" cars and Model Year 1999 and 2000 cars were added to the "Age 0" cars.
We then computed fractions by age. These fractions are summarized in Table 6-1.
6.3. Trucks
To determine age fractions for passenger trucks, light commercial trucks, refuse trucks,
short-haul and long-haul single unit trucks and short-haul and long-haul combination trucks, we
used data from the VIUS97 database. Vehicles in the VIUS97 database were assigned to
MOVES source types as summarized in Table 3-3.
VIUS97 does not include a model year field and records ages as 0 through 10 and 11-
and-greater. Because we needed greater detail on the older vehicles, we followed the practice
used for MOBILE6 and determined the model year for some of the older vehicles by using the
responses to the VIUS97 questions "How did you obtain this vehicle?" (VIUS field "OBTAIN")
and "When did you obtain this vehicle?" (VIUS field "ACQYR") to derive the model year of the
vehicles that were obtained new. These derived model years also were used for much of the
source bin distribution work described later in this report.
To calculate age fractions, it was important to account for the inconsistent methodologies
used for the older and newer vehicles. Thus, for each source type, we adjusted the age 11-and-
older vehicle counts by dividing the original count by model year by the fraction of the older
vehicles that were coded as "obtained new." This created an array of adjusted vehicle counts by
model year for calendar year 1997. This 1997 array may overestimate the fraction of mid-aged
vehicles since the fraction of vehicles purchased new likely declines with time; however, we
believe the procedure is reasonable given the limited data available.
We then used the sales growth for 1997 and 1998 from TEDB22 Tables 7.6 and 8.3 and
the scrappage rates from TEDB22 Tables 6.10 and 6.11 to grow the population to the 1999 base
year and then we calculated age fractions. These fractions are summarized in Table 6-1.
-------
6.4. Intercity Buses
We were not able to identify a data source for estimating age distributions of intercity
buses. Because the purchase and retirement of these buses is likely to be driven by general
economic forces rather than trends in government spending, we will use the age distribution that
was derived for short-haul combination trucks, described previously. While we believe this
choice is reasonable given the lack of data, we welcome suggestions of improved data sources or
algorithms to improve the intercity bus age fractions used in future versions of the MOVES
database.
6.5. School Buses and Motor Homes
To determine the age fractions of School Buses and Motor Homes, we used information
from the Polk TIP® 1999 database. School Bus and Motor Home counts were available by
model year. Unlike the Polk data for passenger cars, these counts reflect registration at the end
of the calendar year and, thus, did not require adjustment. We converted model year to age and
calculated age fractions. These are summarized in Table 6-1.
6.6. Transit Buses
To determine the age fractions for Transit Buses, we used data from the Federal Transit
Administration database. In particular, we used responses to 1999 Form 408, which included
counts of in-use vehicles by year of manufacture.
To properly account for the fraction of Age 0 vehicles at the end of 1999, it was
necessary to adjust the counts for model-year-1999 vehicles to account for the different reporting
periods of the various transit organizations. The counts were adjusted proportionally depending
on the month in which the fiscal year ended. The adjusted counts were used to calculate the age
fractions.
24
-------
Table 6-1. 1999 Age Fractions for MOVES Source Types
Age
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
SourceType
11
0.094714
0.093476
0.075459
0.068103
0.061255
0.056974
0.051956
0.043254
0.037016
0.035495
0.033575
0.038755
0.046127
0.042189
0.038313
0.034497
0.030742
0.027049
0.023416
0.019845
0.016334
0.012885
0.009497
0.00617
0.002904
0
0
0
0
0
0
21
0.081475
0.06084
0.062471
0.058685
0.069053
0.060901
0.061327
0.056453
0.057629
0.057889
0.061101
0.058577
0.053134
0.047721
0.039086
0.030594
0.018665
0.012461
0.010356
0.008542
0.01012
0.007927
0.006039
0.003858
0.002324
0.002771
0
0
0
0
0
31
0.08535
0.077167
0.07186
0.09132
0.096596
0.094721
0.080811
0.063061
0.059255
0.050885
0.051119
0.045676
0.042863
0.037329
0.019511
0.013223
0.007979
0.003568
0.002169
0.001659
0.001279
0.001328
0.00051
0.000413
8.11E-05
3.66E-05
0.00012
7.54E-05
6.78E-06
2.53E-05
2.89E-06
32
0.117527
0.106259
0.098951
0.09387
0.098437
0.082957
0.07331
0.047932
0.04581
0.041843
0.044191
0.038966
0.033072
0.028084
0.018968
0.013196
0.004868
0.001972
0.003003
0.001351
0.001305
0.002379
0.000791
0.000641
8.66E-05
6.19E-05
5.64E-05
3.72E-05
6.16E-05
9.05E-06
5.35E-06
41
0.09124
0.072783
0.062335
0.054827
0.075047
0.061137
0.053356
0.041102
0.048557
0.05662
0.057585
0.05531
0.058511
0.050309
0.0449
0.035235
0.014107
0.011903
0.010321
0.013225
0.00779
0.005953
0.006081
0.002208
0.002572
0.00236
0.00184
0.001547
0.001016
0.000223
0
42
0.062424
0.077118
0.074172
0.072682
0.062732
0.057646
0.050384
0.04612
0.049186
0.075902
0.060917
0.05059
0.04886
0.043448
0.039355
0.032043
0.032094
0.018085
0.008203
0.023051
0.007056
0.003185
0.000651
0.001267
0.000873
0.000856
0.000223
0.000377
0.000274
5.14E-05
0.000171
43
0.079424
0.065977
0.064695
0.059353
0.07985
0.040606
0.051098
0.043455
0.058484
0.069647
0.041921
0.052586
0.055609
0.051232
0.046432
0.037449
0.014449
0.011057
0.013562
0.013786
0.011776
0.010385
0.010678
0.007299
0.009192
0
0
0
0
0
0
51
0.053361
0.042567
0.036457
0.082125
0.09917
0.064919
0.058788
0.026059
0.073815
0.065308
0.052269
0.072611
0.059178
0.074901
0.040258
0.041688
0.012119
0.01351
0.011952
0.002947
0.002976
0.002823
0
0.005242
0.001453
0.00345
5.3E-05
0
0
0
0
52
0.071217
0.059567
0.047165
0.053342
0.064028
0.065817
0.050153
0.039507
0.039877
0.048617
0.062773
0.056501
0.047723
0.058442
0.041881
0.023386
0.036565
0.018857
0.017918
0.018627
0.014052
0.01564
0.005687
0.006819
0.006557
0.00654
0.006311
0.01192
0.00241
0.0012
0.000902
53
0.173083
0.144768
0.114627
0.059639
0.062123
0.104159
0.08065
0.018942
0.013911
0.068272
0.072438
0.048926
0.011957
0.002737
0.005549
0.005545
0
0.002432
0.001752
0
0.000157
0.006238
0.000338
0.000596
0.000259
0
0.000901
0
0
0
0
54
0.073713
0.045616
0.07393
0.048698
0.060515
0.060804
0.044092
0.040781
0.031984
0.04417
0.060195
0.056322
0.05743
0.044695
0.050087
0.053065
0.036308
0.02215
0.012724
0.001733
0.013844
0.019059
0.0267
0.016931
0.004455
0
0
0
0
0
0
61
0.09124
0.072783
0.062335
0.054827
0.075047
0.061137
0.053356
0.041102
0.048557
0.05662
0.057585
0.05531
0.058511
0.050309
0.0449
0.035235
0.014107
0.011903
0.010321
0.013225
0.00779
0.005953
0.006081
0.002208
0.002572
0.00236
0.00184
0.001547
0.001016
0.000223
0
62
0.168908
0.134739
0.115399
0.115399
0.12002
0.081687
0.065744
0.040851
0.03045
0.031176
0.030707
0.027311
0.007322
0.007708
0.009884
0.005095
0.001042
0.0008
0.002546
0.000825
0.000648
0.000383
0.000584
0
0.000427
0.000131
0.000146
0
6.83E-05
0
0
25
-------
7. SourceBinDistribution
The SourceBinDistribution describes the characteristics of a SourceType population as a
distribution among SourceBins. These SourceBins classify a vehicle by discriminators relevant
for emissions and energy calculations: fuel and engine technology, average vehicle weight and
engine displacement, model year group, and regulatory class.
While users can enter SourceBinDistributions directly, MOVES-GHG will usually
generate the values in this table using values in a collection of other tables, which, in turn, need
to be filled with accurate data. The SourceBinGenerator input tables are described in Table 7-1.
Table 7-1. Data Tables Used by SourceBinGenerator
Generator Table Name
SourceTypePolProcess
FuelEngFraction
SizeWeightFraction
RegClassFraction
PollutantProcessModelYear
Key Fields
SourceTypelD
PolProcessID
SourceTypelD
ModelYearlD
FuelTypelD
EngTechID
SourceTypelD
ModelYearlD
FuelTypelD
EngTechID
WeightClassID
EngSizelD
SourceTypelD
ModelYearlD
FuelTypelD
EngTechID
RegClassID
PolProcessID
ModelYearlD
Additional Fields
isSizeWeightReqd
isRegClassReqd
isMYGroupReqd
fuelEngFraction
SizeWeightFraction
regClassFraction
modelYearGroupID
Notes
Indicates which pollutant-processes the
source bin distributions may be applied
to and indicates which discriminators are
relevant for each sourceType and
polProcess (pollutant/process
combination)
Joint distribution of vehicles with a
given fuel type and engine technology.
Sums to one for each sourceType &
modelYear
Joint distribution of engine size and
weight. Sums to one for each
sourceType, modelYear and fuel/engtech
combination.
Fraction of vehicles in a given
"Regulatory Class." Sums to one for
each sourceType, modelYear and
fuel/engtech combination.
Assigns model years to appropriate
model year groups.
The Source Bin generator code determines which discriminators are relevant for a given
pollutant/process combination and multiplies the relevant fractions from the tables listed above
to determine the detailed SourceBinDistribution for each combination of Pollutant, Process,
SourceType, and Model Year.
More detailed descriptions of the SourceBin Distribution inputs for each SourceType
follow. The Inputs for 2000-and-later vehicles of all SourceTypes are described in Section 7.7.
26
-------
7.1. Motorcycles
For base year 1999, motorcycle distributions were assigned based on information from
EPA motorcycle experts and from the Motorcycle Industry Council.
7.1.1. FuelEngFraction
We assume all motorcycles are powered by conventional gasoline engines.
7.1.2. SizeWeightFraction
The Motorcycle Industry Council "Statistical Annual" provides information on
displacement distributions for highway motorcycles for model years 1990 and 1998. These were
mapped to MOVES engine displacement categories. Additional EPA certification data was used
to establish displacement distributions for model year 2000. We assumed that displacement
distributions were the same in 1969 as in 1990, and interpolated between the established values
to determine displacement distributions for all model years from 1990 to 1999.
We then applied weight distributions for each displacement category as suggested by
EPA motorcycle experts. The average weight estimate includes fuel and rider. The weight
distributions depended on engine displacement but were otherwise independent of model year.
This information is summarized in Table 7-2.
Table 7-2. Motorcycle Engine Size and Average Weight Distributions for
Selected Model Years
Displacement
Category
0-169 cc (1)
170-279 cc (2)
280+ cc (9)
1969 MY
distribution
(assumed)
0.118
0.09
0.792
1990 MY
distribution
(MIC)
0.118
0.09
0.792
1998 MY
distribution
(MIC)
0.042
0.05
0.908
2000 MY
distribution
(certification
data)
0.015
0.035
0.95
Weight
distribution
(EPA staff)
100%:
<= 500 Ibs
50%:
<= 500 Ibs
50%:
> 500 Ibs
<= 700 Ibs
30%:
> 500 Ibs,
<= 700 Ibs
70%:
> 7001bs
7.1.3. RegClassFraction
All Motorcycles are assigned to the "Motorcycle" (MC) regulatory class.
7.2. Passenger Cars
For base year 1999, passenger car distributions were derived from the 1999 Polk
NVPP®. The national files for domestic and imported cars were consolidated into a single file.
7.2.1. FuelEngFraction
The FuelEngFraction table assigns a fraction of each source type and model year to all
relevant combinations of fuel type bin and engine technology bin.
27
-------
The Polk fuel code was converted to the MOVES FuelTypelD using the mapping in
Table 6-3. Note that convertible and flexible fueled vehicles are counted as gasoline-powered
vehicles in MOVES2004 because most of these pre-1999 vehicles operate on gasoline most of
the time. This approach differs from that of the Department of Energy and, likely,
underestimates the number of vehicles operating on alternative fuels. On the national scale, the
resulting difference in emissions is expected to be negligible.
Table 7-3. Mapping Polk Fuel Codes to MOVES.
Polk
FUEL CD
C
D
E
F
G
N
P
R
V
X
FUEL_NAME
DSL TURBO
DIESEL
ELECTRIC
GAS TURBO
GAS
NATURAL GAS
PROPANE
METHANOL
CONVERTIBLE
FLEXIBLE
MOVES
FuelTypelD
2
2
9
1
1
o
6
4
6
1
1
Fuel Description
Diesel
Diesel
Electric
Gasoline
Gasoline
CNG
LPG
Methanol
Gasoline
Gasoline
For each model year, the car counts for the MOVES fuels were summed and fractions
were computed. Entries for which no fuel was reported were omitted from the calculations. For
the 1999-and-earlier cars, electric vehicles were assigned to EngTechID "30" (electric only). All
other 1999-and-earlier vehicles were assigned to EngTechID "1" (conventional). Additional
analysis indicated a likely error in the Polk data (an entry for 1983 Ford Escorts powered by
methanol). This fuel/engine fraction was removed and the 1983 values were renormalized.
7.2.2. SizeWeightFraction
The Polk cubic displacement values were converted to liters and assigned to the MOVES
engine size bins. The weight ID was assigned by adding 300 Ibs to the Polk curb weight and
grouping into MOVES weight bins. For each fuel type, model year, engine size, and weight bin,
the number of cars was summed and fractions were computed. In general, entries for which data
was missing were omitted from the calculations. However, because no curb weight data was
available from Polk for electric cars, additional analysis was performed. Based on data from the
Electric Drive Association on electric vehicle sales26, two-thirds of electric vehicles were
assigned to weight class 35 and one third was assigned to weight class 40. Also, further analysis
indicated a likely error in the Polk data (an entry for 1997 gasoline-powered Bentleys with
engine size 5099 and weight class 20). This fraction was removed and the 1997 values were
renormalized.
7.2.3. RegClassFraction
All Passenger Cars were assigned to the "Light-Duty Vehicle" (LDV) regulatory class.
28
-------
7.3. Trucks
This section describes how default Source Bin information was compiled for Passenger
Trucks, Light Commercial Trucks, Single-Unit Short-haul and Long-haul Trucks, and
Combination Short-haul and Long-haul Trucks. Source Bin information for Buses, Refuse
Trucks, and Motor Homes are described in separate sections following.
VIUS97 was the primary source for information on truck distributions. Information
from VIUS97 was supplemented with information from MOBILE6 and from the Oak Ridge
National Laboratory Light Duty Vehicle database.
VIUS97 records were assigned to SourceTypes as described above in Table 3-3. Not all
SourceTypes had data for all model years, and no data was available beyond model year 1997.
For years where no vehicles or only a few vehicles were surveyed by VIUS, we duplicated
fractions from the nearest available model year.
7.3.1. FuelEngFraction
The VIUS97 ENGTYP field was converted to the MOVES FuelTypelD using the
mapping in Table 7-4. Note, it was not possible to distinguish LPG and CNG vehicles using
VIUS97. Based on historical data, we assigned the pre-1990 LPG/LNG vehicles to LPG and the
1990-and-later vehicles to CNG. While these vehicles form a very small portion of the national
fleet, we would like to update this assignment if better information becomes available. Also, it
was not possible to identify the fuel used for the VIUS category "Other." Vehicles in this
category were omitted from the analysis and model year results were renormalized.
Table 7-4. Mapping VIUS97 ENGTYP to MOVES FuelTypelD
VIUS97
1
2
3
4
5
Leaded gasoline
Unleaded gasoline
Diesel
Liquefied gas (petroleum
(LPG) or natural (LNG))
Other
MOVES
1
1
2
3 or 4
Gasoline
Gasoline
Diesel
CNG or
LPG
None
There were no electric-fueled trucks, so all 1999-and-earlier trucks were assigned to
EngTechID "1" (conventional).
Table 7-5 summarizes the pre-1999 diesel fractions for MOVES general truck categories
by model year. Because alternative fuels form a very small portion of the 1999-and-earlier fleet,
the gasoline fractions can be estimated as one minus the diesel fractions listed here. For the
exact gasoline fractions and alternative fuel fractions, see the MOVES database.
For light trucks, fuel distribution information is also available from Polk. While the Polk
data cannot easily be mapped to the truck SourceTypes used in MOVES, if future resources
allow, it would be instructive to compare the Polk distributions to the combined passenger truck
and light commercial truck distributions. This could help estimate the uncertainty in the fuel
fraction estimates for these vehicles.
29
-------
Table 7-5. Diesel Fractions for Trucks
Source
Type
Model
Year
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
Passenger
Trucks
31
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.01392
0.00000
0.03557
0.00000
0.04182
0.00132
0.01633
0.01426
0.00188
0.00801
0.00706
0.00661
0.01442
0.01185
0.00995
0.01488
0.02141
0.00923
0.01002
0.01002
0.01002
Light
Commerical
Trucks
32
0.00906
0.00906
0.00906
0.00906
0.00906
0.08203
0.02876
0.00000
0.00399
0.00083
0.04185
0.05703
0.03142
0.29896
0.15086
0.21648
0.17724
0.04786
0.02941
0.05089
0.05186
0.05723
0.07682
0.05506
0.07803
0.07562
0.07338
0.05300
0.04391
0.04391
0.04391
Single-Unit
Short-haul
Trucks
52
0.06238
0.06238
0.06238
0.01695
0.04465
0.02377
0.02130
0.06518
0.32805
0.01731
0.11083
0.15791
0.16825
0.19327
0.67377
0.57100
0.52692
0.42720
0.60714
0.43232
0.33462
0.47071
0.62245
0.50514
0.58491
0.60593
0.58120
0.60411
0.55462
0.55462
0.55462
Single-Unit
Long-haul
Trucks
53
0.47356
0.47356
0.47356
0.47356
0.47356
0.47356
0.47356
1.00000
1.00000
0.06120
1.00000
1.00000
0.20453
0.87629
1.00000
1.00000
0.99148
0.97560
0.94441
0.30001
0.71929
0.81014
0.30680
0.80948
0.35251
0.53482
0.82016
0.40650
0.51978
0.51978
0.51978
Combination
Short-haul
Trucks
61
0.73282
0.73282
0.73282
0.73282
0.73282
0.73282
0.73282
0.73282
0.73282
0.73282
0.73282
0.73282
0.96590
0.94257
0.92500
0.91464
0.89852
0.95934
0.98601
0.96482
0.96268
0.98571
0.97919
0.97942
0.96928
0.99546
0.97506
0.98205
0.97667
0.97667
0.97667
Combination
Long-haul
Trucks
62
1.00000
1.00000
1.00000
1.00000
1.00000
1.00000
1.00000
1.00000
1.00000
1.00000
1.00000
1.00000
1.00000
1.00000
1.00000
1.00000
1.00000
0.98152
1.00000
0.83667
0.99984
0.99701
0.99732
1.00000
1.00000
0.99873
0.99694
0.99864
0.99973
0.99973
0.99973
7.3.2. SizeWeightFraction
Engine size distributions for trucks were determined using the VIUS97 database. The
VIUS97 database categorizes engine size by fuel type and the categories do not exactly match
the MOVES categories. We mapped from the VIUS97 engine size categories to the MOVES
engine size categories as described in Table 7-6. For comparison, the engine size ranges for both
the VIUS97 and MOVES categories are listed in cubic inches displacement.
-------
Table 7-6. Mapping VIUS97 Engine Size Categories to MOVES EngSizelD
VIUS PKCID |VIUS Range (CID)
EngSizelD
MOVES Range (CID)
Gasoline Engines
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
1-99
100-149
150-169
170-199
200-249
250-269
270-299
300-309
310-349
350-359
360-369
370-399
400-449
450-9999
Not reported
20
2025
2530
3035
3540
4050
4050
4050
5099
5099
5099
5099
5099
5099
0-122
122-153
153-183
183-214
214-244
244-305
244-305
244-305
305-6041
305-6041
305-6041
305-6041
305-6041
305-6041
Diesel Engines
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
1-249
250-299
300-349
350-369
370-399
400-429
430-449
450-469
470-499
500-549
550-599
600-649
650-699
700-749
750-799
800-849
850-9999
Not reported
3540
4050
5099
5099
5099
5099
5099
5099
5099
5099
5099
5099
5099
5099
5099
5099
5099
183-214
244-305
305-6041
305-6041
305-6041
305-6041
305-6041
305-6041
305-6041
305-6041
305-6041
305-6041
305-6041
305-6041
305-6041
305-6041
305-6041
Engines for other Fuels
34
35
36
37
38
39
40
41
42
43
1-99
100-149
150-199
200-249
250-299
300-349
350-399
400-449
450-9999
Not reported
20
2025
2530
3540
4050
5099
5099
5099
5099
0-122
122-153
153-183
214-244
305-6041
244-305
305-6041
305-6041
305-6041
31
-------
Determining weight categories for light trucks was fairly complicated. The VIUS97 data
combines information from two different survey forms. The first form was administered for
VIUS "strata" 1 and 2 trucks: pickup trucks, panel trucks, vans (including mini-vans), utility type
vehicles (including jeeps) and station wagons on truck chassis. The second form was
administered for all other trucks. While both surveys requested information on engine size, only
the second form requested detailed information on vehicle weight. Thus for strata 1 and 2
trucks, VIUS classifies the trucks only by broad average weight category (AVGCK): 6,000 Ibs or
less, 6,001-10,000 Ibs, 10,001-14,0001bs, etc. To determine a more detailed average engine size
and weight distribution for these vehicles, we used the Oak Ridge National Laboratory (ORNL)
light-duty vehicle database to correlate engine size with vehicle weight distributions by model
year.
In particular, for Source Types 31 and 32 (Passenger Trucks and Light Commercial
Trucks):
• VIUS97 trucks of the SourceType in strata 3, 4, and 5 were assigned to the
appropriate MOVES weight class based on VIUS detailed average weight
information.
• VIUS97 trucks of the SourceType in strata 1 and 2 were identified by enginesizelD
and broad average weight category.
• Strata 1 and 2 trucks in the heavier (10,001-14,000 Ibs, etc) VIUS97 broad categories
were matched one-to-one with the MOVES weight classes.
• For trucks in the lower broad categories (6,000 Ibs or less and 6001-10,000 Ibs), we
used VIUS97 to determine the fraction of trucks by model year and fuel type that fell
into each engine size/broad weight class combination (the "VIUS fraction")
• We assigned trucks in the ORNL light duty vehicle database to a weightclassID by
adding 3001bs to the recorded curb weight and determining the appropriate MOVES
weight class.
• For the trucks with a VIUS97 average weight of 6,000 Ibs or less, we multiplied the
VIUS97 fraction by the fraction of trucks with a given weightclassID among the
trucks in the ORNL database that had the given engine size and an average weight of
6,000 Ibs or less. Note, the ORNL database did not provide information on fuel type,
so the same distributions were used for all fuels.
• Because the ORNL database included only vehicles with a GVW up to 8500 Ibs, we
did not use it to distribute the trucks with a VIUS97 average weight of 6,001-10,000
Ibs. Instead these were distributed equally among the MOVES WeightClassIDs 70,
80, 90 and 100.
Source Types 52 and 53 (Long- and Short-haul Single Unit Trucks) also included some
trucks in VIUS97 strata 1 and 2, thus a similar algorithm was applied.
• VIUS97 trucks of the Source Type in strata 3, 4, and 5 were assigned to the
appropriate MOVES weight class based on VIUS97 detailed average weight
information.
-------
• VIUS97 trucks of the Source Type in strata 1 and 2 were identified by enginesizelD
and broad average weight category.
• Strata 1 and 2 trucks in the heavier (10,001-14,000 Ibs, etc) VIUS97 broad categories
were matched one-to-one with the MOVES weight classes.
• For trucks in the lower broad categories (6,000 Ibs-or-less and 6001-10,000 Ibs), we
used VIUS97 to determine the fraction of trucks by model year and fuel type that fell
into each engine size/broad weight class combination (the " VIUS fraction")
• We did not believe the ORNL light duty vehicle database adequately represented
single unit trucks. Thus, the trucks with a VIUS97 average weight of 6,000 Ibs or
less and an engine size less than 5 liters were distributed equally among the MOVES
weight classes 20, 25, 30, 35, 40, 45, 50, and 60. Because no evidence existed of
very light trucks among the vehicles with larger engines (5 liter or larger), these were
equally distributed among MOVES weight classes 40, 45, 50 and 60.
• The trucks with a VIUS97 average weight of 6,001-10,000 Ibs were distributed
equally among the MOVES weight classes 70, 80, 90 and 100.
SourceTypes 61 and 62 (Long- and Short-haul combination trucks) did not include any
vehicles of VIUS97 strata 1 or 2. Thus we used the detailed VIUS97 average weight information
and engine size information to assign engine size and weight classes for all of these trucks.
7.3.3. RegClassFraction
Trucks were split between the regulatory classes "Light-Duty Trucks" (LDT) and
"Heavy-Duty Trucks" (HDT) based on gross vehicle weight (GVW) (the maximum weight that a
truck is designed to carry.)
In particular, we used the VIUS97 response "PKGVW" and the Davis & Truit report on
Class 2b Trucks27 to determine GVW fractions by fuel type. The VIUS97 PKGVW field is
intended to identify the Polk weight class. Work for MOBILE6 using the VIUS97 precursor,
TIUS 1992 indicated that the PKGVW measure in VIUS97 is problematic. It is taken from the
truck VEST, but is not always consistent with the indicated average and maximum weight. (For
example, the reported "maximum weight" often exceeded the PKGVW.) These problems were
also seen in VIUS97. However, "maximum weight" was not available for smaller trucks, and the
other measures of weight reported in VIUS97 were not consistent with the need for an indicator
of the relevant emission standards. When the PKGVW led to unusual results, for example,
particularly high fraction of LDT among combination trucks, we checked additional VIUS fields
to determine if the PKGVW was mistaken. In some cases, the PKGVW was manually revised to
a higher value and fractions were recomputed. In other cases, the PKGVW was consistent with
the other fields, and the difference reflected the fact that our SourceType categories are based on
axle counts and trailer configurations rather than weight. For example, a 6-tire ("dually") pickup
that regularly pulls a trailer is classified as a "Combination Truck," although it is in the LDT
regulatory class. Some model years had relatively high fractions of such trucks. It is likely these
high values indicate a problem with small sample size for the model year, but they were left
unchanged for now.
33
-------
Also, because the split between the LDT and HDT regulatory class is at 8500 Ibs, it was
necessary to split the Polk GVW Class 2 into class 2a (6001-8500 Ibs) and class 2b (8501-10,000
98
Ibs). Davis & Truitt report that, on average, 23.3 percent of Class 2 trucks are in Class 2b;
97.4 percent of Class 2a trucks are powered by gasoline, and 76 percent of Class 2b trucks are
powered by gasoline. From this information, we estimate that 19.2 percent of gasoline-powered
Class 2 trucks are Class 2b and that 73.7 percent of diesel-powered class 2 trucks are Class 2b.
The regulatory class fractions for trucks are listed below in Table 7-7 and Table 7-8.
Fractions of LDT for gasoline- and diesel-fueled vehicles are provided separately. The
remaining trucks are classified as FtDT. Entries of "#N/A" indicate that no vehicles of that
SourceType and FuelType were surveyed in that model year. Values for alternative-fuel vehicles
are available in the MOVES database.
34
-------
Table 7-7 Fraction of Light-Duty Trucks among Gasoline-Fueled Trucks
Model
Year
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
SourceType
Passenger
Trucks
31
0.902303
0.879238
1
0.983681
0.956315
0.957791
0.953535
0.946371
0.966522
0.951185
0.887739
0.847443
0.863942
0.897151
0.959489
0.939455
0.95116
0.937822
0.933322
0.945257
0.956912
0.958257
0.956279
0.955606
0.967955
0.95438
0.957004
0.949354
0.943815
0.953521
0.954686
Light
Commercial
Trucks
32
#N/A
#N/A
#N/A
#N/A
#N/A
0.74768
0.59472
0.65248
0.724827
0.883189
0.793622
0.809907
0.776929
0.74161
0.893686
0.719863
0.903414
0.86782
0.869615
0.912692
0.896428
0.899821
0.895456
0.897817
0.903363
0.913522
0.923815
0.891218
0.883694
0.902039
0.924425
Single-Unit
Short-haul
Trucks
52
#N/A
#N/A
0.109337
0.046808
0.38324
0.683527
0.300171
0.132987
0.134558
0.125404
0.061817
0.45065
0.255077
0.171485
0.304625
0.544875
0.494159
0.332359
0.253229
0.30116
0.376371
0.580867
0.594791
0.530591
0.448187
0.624044
0.59655
0.593485
0.507255
0.655492
0.628248
Single-Unit
Long-haul
Trucks
53
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
0
0
0
#N/A
#N/A
0.62437
#N/A
#N/A
0.643456
0
#N/A
#N/A
0.808384
0
0
0.451689
0.334361
0
0.950832
0
0.596423
0.533585
0.015518
0.567117
0.501883
Combination
Short-haul
Trucks
61
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
0
0
0
0
0
0
0.028698
0
0
0
0
0
0
0
0
0
0
0
Combination
Long-haul
Trucks
62
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
0
#N/A
0.775934
0.808384
0
0
#N/A
#N/A
0
0
0
0
35
-------
Table 7-8. Fraction of Light-Duty Trucks among Diesel-fueled Trucks
Model
Year
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
SourceType
Passenger
Trucks
31
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
0
#N/A
0.892664
#N/A
0.54614
0.262872
0.259661
0.959781
0.262872
0.254531
0.25772
0.374882
0.262244
0.260222
0.262872
0.314785
0.262219
0.307815
0.26213
Light
Commercial
Trucks
32
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
0
0
0
#N/A
0
0
0.072135
0.397873
0.118825
0.271488
0.232866
0.243221
0.231416
0.133885
0.091469
0.158459
0.200201
0.176388
0.203415
0.157783
0.242925
0.243391
0.178734
0.193289
0.186986
Single-Unit
Short-haul
Trucks
52
#N/A
#N/A
#N/A
#N/A
0
0
0
0
0
0
0
0
0
0
0
0.047107
0.219283
0.019513
0.041111
0.034369
0.013541
0.078315
0.040286
0.012528
0.145307
0.095434
0.171617
0.182654
0.092071
0.12573
0.082764
Single-Unit
Long-haul
Trucks
53
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
0
0
0
0
0
0
0
0
#N/A
0
0
0
0
0.008054
0.069597
0.121637
0
0.078572
0.433434
0.132245
0.160888
0.119779
0.303336
Combination
Short-haul
Trucks
61
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
0.009394
0
0
0
0
0.006796
0.001254
0
0
0
0.000596
0.002601
0
0.001354
0
0.001135
0.000667
0.002372
Combination
Long-haul
Trucks
62
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
#N/A
0
0
0
0
0
0.000507
0
0
0
0
0
0
0
0.003932
7.4. Buses
Because buses are not included in VIUS97 and because the Polk data we had for school
buses was incomplete, the source bin fractions for buses is based on a variety of data sources and
assumptions. Values for transit buses, school buses, and intercity buses were calculated
separately.
7.4.1. FuelEngFraction
We followed the Energy Information Administration (EIA) in assigning all intercity
buses to conventional diesel engines (AEO2004, Supplemental Table 34}.
-------
The National Transit Database (NTD) responses to form 408 (Revenue Vehicle
Information Form) included information classifying transit buses to a variety of fuel types by
model year. The mapping from NTD fuel types to MOVES fuel types is summarized in Table
7-9. The resulting fractions by model year are summarized in Table 7-10.
Table 7-9. Mapping National Transit Database Fuel Types to MOVES Fuel Types
NTD code
BF
CN
DF
DU
EB
EP
ET
GA
GR
KE
LN
LP
MT
OR
NTD description
Bunker fuel
Compressed natural gas
Diesel fuel
Dual fuel
Electric battery
Electric propulsion
Ethanol
Gasoline
Grain additive
Kerosene
Liquefied natural gas
Liquefied petroleum gas
Methanol
Other
MOVES
Fuel ID
na
o
3
2
2
9
9
5
1
na
na
o
3
4
6
na
MOVES Fuel
Description
CNG
diesel
diesel
electric
electric
ethanol
gasoline
CNG
LPG
methanol
37
-------
Table 7-10. Fuel Fractions for Transit Buses
Model
Year
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
Gasoline
0
0
0
0
0
0
0
0
0
0
0.033981
0
0.002088
0.001894
0
0.001603
0
0.00079
0.001402
0.002377
0.00113
0.002941
0.003134
0.010769
0.003061
0.010711
0.009555
0.017963
0.012702
0.012003
0.005998
Diesel
1
1
1
1
1
1
1
1
1
1
0.966019
1
0.997912
0.992424
1
0.998397
0.999565
0.996447
0.998598
0.997623
0.998306
0.990271
0.978064
0.933903
0.918707
0.900625
0.835108
0.881825
0.810162
0.838409
0.878041
CNG
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.000435
0.002764
0
0
0
0.006787
0.018106
0.046417
0.07551
0.084796
0.153153
0.097613
0.174365
0.1487
0.113296
LPG
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.000743
0.00068
0.000893
0
0.000709
0.000462
0
0
Ethanol
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.000565
0
0
0
0.001361
0
0
0
0
0
0
Methanol
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.005941
0
0
0
0
0
0
0
Electric
0
0
0
0
0
0
0
0
0
0
0
0
0
0.005682
0
0
0
0
0
0
0
0
0.000696
0.002228
0.00068
0.002975
0.002184
0.001891
0.002309
0.000889
0.002666
All 1999-and-earlier electric buses were assigned to EngTechID "30" (electric only). All
other 1999-and-earlier buses were assigned to EngTechID "1" (conventional).
The available Polk data excluded fuel information on school buses and we were unable to
locate any other source for bus fuel fractions. (The Union of Concerned Scientists estimates that
about one percent of school buses are fueled by either CNG or propane, but does not provide
9Q
estimates by model year. ) Thus we used the diesel fractions from MOBILE6, which were
derived from Polk 1996 and 1997 data. We assigned non-diesel buses to gasoline. These
fractions are summarized in Table 7-11. In the future it would be desirable to obtain up-to-date,
detailed fuel information for school buses from Polk or some other source.
-------
Table 7-11. Fuel Fractions for School Buses
Model Year
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
Gasoline
1
1
1
0.991272
0.99145
0.976028
0.970936
0.95401
0.94061
0.736056
0.674035
0.676196
0.615484
0.484507
0.326706
0.265547
0.249771
0.229041
0.124036
0.089541
0.010041
0.120539
0.147479
0.114279
0.041539
Diesel
0
0
0
0.008728
0.00855
0.023972
0.029064
0.04599
0.05939
0.263944
0.325965
0.323804
0.384516
0.515493
0.673294
0.734453
0.750229
0.770959
0.875964
0.910459
0.989959
0.879461
0.852521
0.885721
0.958461
7.4.2. SizeWeightFraction
While the vast majority of buses of all types have engine displacement larger than five
liters (EngSizeID=5099), it was difficult to find detailed information on average bus weight.
For intercity buses, we used information from Table II-7 of the FT A 2003 Report to
Congress30 that specified the number of buses in various weight categories. This information is
summarized in below in Table 7-12. Note the FTA uses the term "over-the-road bus" to refer to
the class of buses roughly equivalent to the MOVES "intercity bus" category. The FTA weight
categories were mapped to the equivalent MOVES weight classes.
Table 7-12. FTA Estimate of Bus Weights
Weight (Ibs)
0-20,000
20,000-30,000
30,000-40,000
40,000-50,000
total
MOVES
Weight
ClassID
400
500
MOVES
Weight Range
(Ibs)
33,000-40,000
40,000-50,000
Number buses
(2000)
173,536
392,345
120,721
67,905
754,509
Bus type
school & transit
school & transit
school & transit & intercity
intercity
39
-------
Using our 1999 bus population estimates (in Table3-l), we were able to estimate the
fraction of all buses that were intercity buses and then to estimate the fraction of intercity buses
in each weight bin. In particular:
Estimated number of intercity buses in 2000:
754,509 * (84,4547(84,454+55,706+592,029)) = 87,028
Estimated number of intercity buses 30,000-40,000 Ibs:
Estimated intercity bus weight distribution:
This distribution was used for all model years.
87,028-67,905 = 19,123
Class 400 = 19,123/87,028 = 22%
Class 500 = 67,905/87,028 = 78%
For transit buses, we took average curb weights from Figure II-6 of the FT A Report to
Congress31and added additional weight to account for passengers and alternative fuels. The
resulting in-use weights were all in the range from 33,850 to 40,850. Thus all transit buses were
assigned to the weight class "400" (33,000 - 40,000 Ibs) for all model years. This estimate could
be improved if more detailed weight information for transit buses becomes available.
For school buses, we used information from a survey of California school buses. While
this data may not be representative of the national average distribution, it was the best data
source available. The California data32 provided information on number of vehicles by gross
vehicle weight class and fuel as detailed in Table 7-13.
Table 7-13. California School Buses
LHDV
MHDV
HHDV
Total
Gas
2740
467
892
4099
Diesel
4567
2065
11639
18271
Other
8
2
147
157
Total
7315
2534
12678
To estimate the distribution of average weights among the MOVES weight classes, we
assumed that the Light Heavy-Duty (LHDV) school buses were evenly distributed among
weightClassIDs 70, 80, 90, 100, and 140. Similarly, we assumed the Medium Heavy-Duty
(MHDV) school buses were evenly distributed among weightClassIDs 140, 160, 195, 260, and
330 and the Heavy Heavy-Duty (HHDV) school buses were evenly distributed among
weightClassIDs 195, 260, 330, and 440.
The final default weight distributions for buses are summarized in Table 7-14.
7.4.3. RegClassFraction
All buses were assigned to the Heavy-Duty Truck regulatory class.
-------
Table 7-14. Weight Distributions for Buses by Fuel Type
Weight Class
70
80
90
100
140
160
195
260
330
400
500
Intercity
Buses
(41)
Diesel
0.2197
0.7800
Transit Buses
(42)
Diesel & Gas
1.0000
School Buses (43)
Diesel
0.0500
0.0500
0.0500
0.0500
0.0726
0.0226
0.1819
0.1819
0.1819
0.1593
Gas
0.1337
0.1337
0.1337
0.1337
0.1565
0.0228
0.0772
0.0772
0.0772
0.0544
7.5. Refuse Trucks
Values for Refuse Trucks (Source Type 51) were computed from information in VIUS97.
7.5.1. FuelEngFraction
As for other trucks, we used the VIUS97 EngTyp field to estimate FuelType and Engine
Technology Fractions. The Refuse Trucks classified in VIUS97 as "CNG or LPG" are assigned
to CNG. All Refuse Trucks were assumed to have conventional internal combustion engines.
Table 7-15. Fuel Fractions for Refuse Trucks by Model Year
Model
Year
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
Gasoline
0
0.220605
0
0
0
0.076055
0.0175
0
0.014675
0.005532
0.071123
0
0
0
Diesel
1
0.779395
1
1
1
0.922351
0.9825
1
0.985325
0.994468
0.928877
1
1
1
CNG
0
0
0
0
0
0.001594
0
0
0
0
0
0
0
0
7.5.2. SizeWeightFraction
Because the sample of Refuse Trucks in VIUS97 was small, the same SizeWeight
distributions were used for all model years. As for other trucks, the EngineSize group was
-------
determined from the VIUS97 engine size categories and the WeightClass was determined from
the VIUS97 reported average weight.
Table 7-16. Refuse Truck SizeWeight Fractions by Fuel Type
Engine Size
3-3. 5L
3.5-4L
3.5-4L
3.5-4L
3.5-4L
3.5-4L
3.5-4L
4-5L
>5L
>5L
>5L
>5L
>5L
>5L
>5L
>5L
>5L
>5L
>5L
>5L
Weight
Class
60
100
140
160
195
260
400
140
80
100
140
160
195
260
330
400
500
600
800
1000
Gasoline
0.009008
0.001919
0.000000
0.147737
0.070203
0.134765
0.198498
0.054682
0.203838
0.021943
0.152009
0.003834
0.001563
Diesel
0.000317
0.007691
0.000822
0.000258
0.000741
0.001081
0.000074
0.000000
0.006808
0.011626
0.003302
0.022762
0.062584
0.098921
0.101197
0.235437
0.333582
0.110766
0.002031
CNG
1.000000
7.5.3. RegClassFraction
Using the VIUS97 data on gross vehicle weight, all Refuse Trucks were classified as
Heavy-Duty Trucks.
7.6. Motor Homes
Determining source bin distribution for Motor Homes required a number of assumptions
and interpolation due to the lack of detailed information. For each field, the following describes
the information available, assumptions made, and how data points were determined.
7.6.1. FuelEngFraction
Detailed information on motor home fuel distribution was not available. Staff of the
Recreational Vehicle Industry Association (RVIA) told us that the fraction of diesel motor
homes had been relatively constant at 10 to 20 percent for many years.33 This fraction began to
increase steadily in the mid-1990s and is now 40%. Based on this information, we used linear
interpolation to estimate the diesel fractions in Table 7-17. The remaining 1999-and-earlier
motor homes are assumed to be gasoline-fueled. We assumed all 1999-and-earlier motor homes
have conventional internal combustion engines.
42
-------
Table 7-17. Diesel Fractions for Motor Homes.
Model Year
1993-and-earlier
1994
1995
1996
1997
1998
1999
Fraction Diesel
0.150000
0.177778
0.205556
0.233333
0.261111
0.288889
0.316667
7.6.2. SizeWeightFraction
No detailed information was available on average engine size and weight distributions for
motor homes. We assumed all motor home engines were 5 L or larger. As a surrogate for
average weight, we used information on gross vehicle weight provided in the Polk TIP® 1999
database by model year and mapped the Polk GVW Class to the MOVES weight bins. These
values are likely to overestimate average weight and should be updated if better information
becomes available. The Polk TIP® information did not specify fuel type, so we assumed that the
heaviest vehicles in the Polk database were diesel-powered and the remainder are powered by
gasoline. This led to the weight distributions in Table 7-18 and Table 7-19.
43
-------
Table 7-18. Weight Fractions for Diesel Motor Homes by Model Year
Polk GVW
bin
MOVES
weight class
Model Year
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
3
140
4
160
5
195
6
260
7
330
8
400
Diesel
0.171431
0.637989
0.68944
0.423524
0.096922
0.462916
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.792112
0.340639
0.292308
0.574539
0.899344
0.537084
0.941973
0.868333
0.912762
0.932659
0.881042
0.855457
0.791731
0.72799
0.73298
0.173248
0
0
0
0
0
0
0
0
0
0.029828
0.018755
0.012168
0
0
0
0
0
0.000203
0.000835
0.001474
0.013381
0.085493
0.148917
0.128665
0.614798
0.619344
0.551548
0.345775
0.45546
0.635861
0.553807
0.666905
0.267
0
0
0.000436
0.000277
0.000387
0.001067
0
0.030174
0.049
0.014845
0.009183
0.010761
0.022962
0.022498
0.015469
0.043052
0.043628
0.063712
0.01901
0.471873
0.354386
0.163195
0.229529
0.193167
0.335069
0.736656
0.006629
0.002181
0.005531
0.00155
0.002667
0
0
0.03
0.030096
0.036732
0.083285
0.089534
0.087164
0.093335
0.082792
0.149939
0.296399
0.385085
0.144844
0.159622
0.17468
0.184208
0.111299
0.357508
0.233886
0
0
0.000277
0
0
0
0.027853
0.052667
0.042094
0.020592
0.023438
0.018667
0.013113
0.014289
0.012511
0.018387
0.020545
0.044356
0.037509
0.030531
0.026264
0.032456
0.028628
0.040423
0.029458
44
-------
Table 7-19. Weight Fractions for Gasoline Motor Homes by Model Year
Polk GVW
bin
MOVES
weight class
Model Year
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
3
140
4
160
5
195
6
260
7
330
8
400
Gasoline
1
1
1
1
1
1
0.747723
0.732235
0.714552
0.641577
0.692314
0.720248
0.606635
0.459429
0.551601
0.543354
0.612025
0.54464
0.583788
0.481099
0.52997
0.435959
0.221675
0.288222
0.170133
0
0
0
0
0
0
0.252277
0.267765
0.285448
0.358423
0.307686
0.279752
0.393365
0.540571
0.448399
0.456646
0.322022
0.373999
0.361277
0.361146
0.198479
0.289453
0.433334
0.581599
0.392451
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.065952
0.081361
0.054935
0.157755
0.271551
0.274588
0.344991
0.13018
0.288411
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.149004
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
7.6.3. RegClassFraction
We assigned all motor homes to the Heavy-Duty Truck regulatory class.
7.7. SourceBinDistributions for 2000-and-later
MOVES2004 was designed to support a wide variety of future fuels and engine
technologies, including compressed natural gas (CNG), liquified petroleum gas (LPG), and
conventional internal combustion (CIC) and advanced internal combustion (AIC) engines. In
particular, emission rates were developed to support the combinations of fuel and engine
technology listed by SourceType in Table 7-20.
The various hybrid types were split into "moderate" and "full" categories because there
are types of hybrids which get less of an efficiency increase from hybrid design due to larger
engines and smaller electrical components. The less efficient designs we called "moderate"
hybrids (like the Honda hybrids) to distinguish them from the more efficient, full hybrid designs
(like the Toyota Prius). Both of these categories have significantly different energy rates and
45
-------
potentially different market shares. Conventional categories are split from advanced categories
for a different reason. There have been significant improvements in internal combustion engines
over time. The conventional versus advanced split is a crude accounting of these improvements.
Hydrogen-fuel vehicles (including internal combustion, fuel cell and fuel cell-hybrid) are shown
to have a small market share in future years, based on NEMS projections; however, these
vehicles are not currently supported as the energy and emission rates for these vehicles are not
included in the initial release of draft MOVES2004. When these rates are added to the model, the
range of hydrogen technologies shown here will be supported. All of these technologies are
further defined in the report, "Fuel Consumption Modeling of Conventional and Advanced
Technology Vehicles in the Physical Emission Rate Estimator (PERE)," which is being written
for the MOVES model.
Table 7-20. Supported Fuels and Technologies for 2000-and-later Model Years.
Fuel
Gasoline
Gasoline
Gasoline
Gasoline
Gasoline
Gasoline
Diesel
Diesel
Diesel
Diesel
Diesel
Diesel
CNG
LPG
Ethanol
Methanol
Engine
Technology
Conventional
1C
Advanced 1C
CIC Hybrid
Moderate
CIC Hybrid
Full
AIC Hybrid
Moderate
AIC Hybrid
Full
Conventional
1C
Advanced 1C
CIC Hybrid
Moderate
CIC Hybrid
Full
AIC Hybrid
Moderate
Diesel AIC
Hybrid Full
Conventional
1C
Conventional
1C
Conventional
1C
Conventional
Motor-
cycles
X
Passenger
Cars,
Light
Passenger
&
Commerci
al Trucks
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Transit &
School
Buses;
Single-
Unit Short
Haul
Trucks &
Motor
Homes
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Intercity
Buses
X
X
Refuse
Trucks
X
X
X
X
X
X
Single
Unit Long
Haul
Trucks
X
X
X
X
X
X
X
X
Combi-
nation
Short &
Long
Haul
Trucks
X
X
X
X
46
-------
Fuel
Gaseous
Hydrogen
Gaseous
Hydrogen
Gaseous
Hydrogen
Gaseous
Hydrogen
Liquid
Hydrogen
Liquid
Hydrogen
Electricity
Engine
Technology
1C
Advanced 1C
AIC Hybrid
Fuel Cell
Hybrid
Fuel Cell
Fuel Cell
Hybrid
Fuel Cell
Electric only
Motor-
cycles
Passenger
Cars,
Light
Passenger
&
Commerci
al Trucks
To Be
Added
To Be
Added
To Be
Added
To Be
Added
To Be
Added
To Be
Added
X
Transit &
School
Buses;
Single-
Unit Short
Haul
Trucks &
Motor
Homes
To Be
Added
To Be
Added
To Be
Added
To Be
Added
To Be
Added
To Be
Added
X
Intercity
Buses
Refuse
Trucks
To Be
Added
To Be
Added
To Be
Added
To Be
Added
To Be
Added
To Be
Added
X
Single
Unit Long
Haul
Trucks
Combi-
nation
Short &
Long
Haul
Trucks
The inputs for determining default SourceBinDistributions for model years 2000-and-
later were generally based on fuel and engine technology projections from AEO2004 and on the
1999 calendar year regulatory class, size and weight distributions used in MOVES.
7.7.1. Motorcycles
We assumed that all 2000-and-later motorcycles were fueled by conventional gasoline
engines, with the same size and weight distributions as in 1999. All motorcycles are in the
"Motorcycle" regulatory class.
7.7.2. Passenger Cars, Light Passenger Trucks and Light Commercial Trucks
MOVES2004 supports a wide range of fuels and future engine technologies for passenger
cars and light trucks.
The FuelEngFractions for these vehicles were determined from AEO2004. Supplemental
Table 45 of the AEO2004 lists projected sales by technology type for light duty vehicles.
Supplemental Table 56 lists projected technology penetrations for light duty vehicles. These
values were mapped to the MOVES fuels and technologies to project fractions for model years
2001 through 2025. Fractions from 2001 were applied to model year 2000. Fractions from 2025
were applied to model years 2026 through 2050.
In particular, we analyzed passenger cars and light trucks separately. We assumed that
vehicles listed as "flexible fueled" in AEO Table 45 would primarily operate on gasoline and
47
-------
mapped them to the MOVES category for gasoline-powered conventional internal combustion
engines. Also, we assigned half of the AEO Table 45 gasoline hybrid vehicles to the MOVES
gasoline-powered conventional internal combustion hybrid—"moderate" category and half to the
gasoline-powered conventional internal combustion hybrid—"full" category. None were
assigned to the MOVES advanced internal combustion hybrid categories. We made a similar
split for diesel hybrids. Finally, to split the AEO "conventional" gasoline and diesel vehicles
into MOVES conventional and advanced internal combustion categories, we used information
from AEO Table 56, assigning all vehicles with gasoline direct injection or cylinder deactivation
to the advanced internal combustion category. The resulting fuelEngFractions are listed in
Table 7.21 and 7.22.
We used the size and weight distributions from 1999 for future years. Where a future
fuel was not part of the fleet in 1999, we used the 1999 size and weight distribution for gasoline
conventional internal combustion engines. Where a future diesel engine technology was not part
of the fleet in 1999, we used the 1999 size and weight distribution for diesel conventional
internal combustion engines.
All Passenger Cars were assigned to the Light Duty Vehicle (LDV) regulatory class.
Light Trucks were distributed among the Light Duty Truck (LDT) and Heavy Duty Truck (HDT)
regulatory classes. Where a future fuel was not part of the fleet in 1999, we used the regulatory
class distribution for gasoline conventional internal combustion vehicles. Where a future diesel
engine technology was not part of the fleet in 1999, we used the 1999 regulatory class
distribution for diesel conventional internal combustion vehicles.
-------
Table 7.21. Fuel and Engine Technology Fractions for 2000-and-later Passenger Cars
Model
Year
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026+
Gasoline
CIC
0.99701
0.99701
0.99512
0.99096
0.96369
0.95946
0.95173
0.94111
0.92946
0.90863
0.88936
0.86848
0.84843
0.82957
0.81089
0.79319
0.77795
0.76278
0.75130
0.73958
0.73156
0.71839
0.71063
0.70685
0.70357
0.70063
0.70063
Gasoline
AIC
0.00026
0.00026
0.00136
0.00272
0.00545
0.00837
0.01481
0.02347
0.03359
0.04520
0.06093
0.07365
0.09082
0.10790
0.12427
0.14018
0.15510
0.17002
0.18112
0.19264
0.20051
0.21362
0.22125
0.22496
0.22794
0.23068
0.23068
Gasoline
CIC
Hybrid
Mod
0.00078
0.00078
0.00120
0.00250
0.01433
0.01495
0.01562
0.01635
0.01709
0.02167
0.02146
0.02550
0.02547
0.02613
0.02645
0.02711
0.02718
0.02726
0.02731
0.02733
0.02733
0.02727
0.02725
0.02723
0.02720
0.02716
0.02716
Gasoline
CIC
Hybrid
Full
0.00078
0.00078
0.00120
0.00250
0.01433
0.01495
0.01562
0.01635
0.01709
0.02167
0.02146
0.02550
0.02547
0.02613
0.02645
0.02711
0.02718
0.02726
0.02731
0.02733
0.02733
0.02727
0.02725
0.02723
0.02720
0.02716
0.02716
Diesel
CIC
0.00081
0.00081
0.00074
0.00096
0.00184
0.00191
0.00186
0.00235
0.00239
0.00242
0.00285
0.00282
0.00284
0.00290
0.00292
0.00298
0.00306
0.00314
0.00320
0.00332
0.00344
0.00356
0.00371
0.00384
0.00407
0.00430
0.00430
Diesel
AIC
-
~
-
-
0.00000
0.00000
0.00002
0.00003
0.00004
0.00004
0.00006
0.00006
0.00008
0.00009
0.00010
0.00011
0.00012
0.00013
0.00013
0.00014
0.00014
0.00014
0.00014
0.00014
0.00014
0.00014
0.00014
Diesel
CIC
Hybrid
Mod
-
~
-
-
-
~
-
-
-
-
0.00177
0.00181
0.00326
0.00340
0.00423
0.00436
0.00439
0.00439
0.00438
0.00440
0.00441
0.00442
0.00443
0.00441
0.00444
0.00445
0.00445
Diesel
CIC
Hybrid
Full
-
~
-
-
-
~
-
-
-
-
0.00177
0.00181
0.00326
0.00340
0.00423
0.00436
0.00439
0.00439
0.00438
0.00440
0.00441
0.00442
0.00443
0.00441
0.00444
0.00445
0.00445
CNG
CIC
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
0.00002
Ethanol
(E85)
CIC
0.00007
0.00007
0.00007
0.00007
0.00007
0.00007
0.00007
0.00008
0.00008
0.00009
0.00010
0.00010
0.00011
0.00011
0.00012
0.00013
0.00013
0.00014
0.00014
0.00014
0.00015
0.00016
0.00016
0.00016
0.00017
0.00017
0.00017
Gaseous
Hydrogen
Fuel Cell
Hybrid
-
~
-
-
-
~
-
-
-
-
~
-
-
0.00012
0.00009
0.00020
0.00023
0.00024
0.00046
0.00045
0.00044
0.00047
0.00048
0.00049
0.00056
0.00058
0.00058
Electric
0.00028
0.00028
0.00028
0.00027
0.00027
0.00026
0.00025
0.00025
0.00024
0.00024
0.00023
0.00023
0.00024
0.00024
0.00024
0.00024
0.00025
0.00025
0.00025
0.00025
0.00025
0.00025
0.00025
0.00025
0.00025
0.00025
0.00025
49
-------
Table 7.22. Fuel and Engine Technology Fractions for 2000-and-later Light Trucks
Model
Year
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026+
Gasoline
CIC
0.95723
0.95723
0.95973
0.95233
0.90805
0.88864
0.87335
0.83671
0.83460
0.83143
0.81513
0.80855
0.79112
0.77390
0.75524
0.73311
0.71683
0.69893
0.68217
0.66799
0.65718
0.65068
0.64469
0.63849
0.63780
0.63376
0.63376
Gasoline
AIC
0.00019
0.00019
0.00080
0.00157
0.00322
0.00534
0.02126
0.05941
0.05951
0.05994
0.06674
0.07359
0.09034
0.10570
0.12390
0.14355
0.15904
0.17593
0.19204
0.20529
0.21429
0.22022
0.22561
0.23209
0.23140
0.23473
0.23473
Gasoline
CIC
Hybrid
Mod
-
-
-
~
0.01026
0.01887
0.01953
0.01916
0.01964
0.02013
0.02046
0.02097
0.02146
0.02196
0.02245
0.02221
0.02222
0.02196
0.02196
0.02195
0.02128
0.02127
0.02127
0.02127
0.02124
0.02118
0.02118
Gasoline
CIC
Hybrid
Full
-
-
-
~
0.01026
0.01887
0.01953
0.01916
0.01964
0.02013
0.02046
0.02097
0.02146
0.02196
0.02245
0.02221
0.02222
0.02196
0.02196
0.02195
0.02128
0.02127
0.02127
0.02127
0.02124
0.02118
0.02118
Diesel
CIC
0.04214
0.04214
0.03904
0.04567
0.06775
0.06768
0.06494
0.06134
0.06240
0.06398
0.07237
0.07081
0.06935
0.06969
0.06848
0.06801
0.06839
0.06860
0.06843
0.06941
0.07011
0.07066
0.07128
0.07108
0.07239
0.07324
0.07324
Diesel
AIC
-
-
-
~
0.00003
0.00010
0.00086
0.00369
0.00368
0.00368
0.00414
0.00443
0.00534
0.00606
0.00680
0.00749
0.00781
0.00807
0.00828
0.00827
0.00827
0.00827
0.00827
0.00827
0.00827
0.00827
0.00827
Diesel
CIC
Hybrid
Mod
-
-
-
~
-
-
-
~
-
-
-
-
~
-
-
0.00121
0.00123
0.00177
0.00177
0.00179
0.00305
0.00306
0.00308
0.00308
0.00311
0.00312
0.00312
Diesel
CIC
Hybrid
Full
-
-
-
~
-
-
-
~
-
-
-
-
~
-
-
0.00121
0.00123
0.00177
0.00177
0.00179
0.00305
0.00306
0.00308
0.00308
0.00311
0.00312
0.00312
CNG
CIC
0.000005
0.000005
0.000005
0.000004
0.000005
0.000004
0.000004
0.000004
0.000004
0.000004
0.000004
0.000004
0.000004
0.000004
0.000004
0.000004
0.000004
0.000004
0.000004
0.000004
0.000004
0.000004
0.000004
0.000004
0.000004
0.000004
0.000004
LPG
CIC
0.00000
0.00000
0.00001
0.00001
0.00000
0.00001
0.00001
0.00001
0.00001
0.00001
0.00001
0.00001
0.00001
0.00001
0.00001
0.00001
0.00001
0.00001
0.00001
0.00001
0.00001
0.00001
0.00001
0.00001
0.00001
0.00001
0.00001
Gaseous
Hydrogen
Fuel Cell
Hybrid
-
-
-
~
-
0.00000
0.00000
0.00000
0.00000
0.00012
0.00011
0.00011
0.00025
0.00010
0.00009
0.00020
0.00021
0.00021
0.00042
0.00039
0.00037
0.00036
0.00034
0.00030
0.00031
0.00028
0.00028
Electric
0.00043
0.00043
0.00043
0.00043
0.00043
0.00048
0.00052
0.00052
0.00052
0.00058
0.00058
0.00057
0.00067
0.00062
0.00059
0.00078
0.00081
0.00081
0.00119
0.00115
0.00111
0.00112
0.00110
0.00107
0.00112
0.00110
0.00110
50
-------
7.7.3 Buses
Historically, school buses and transit buses have used a wide range of alternative fuels,
while intercity buses have been powered almost exclusively by conventional diesel engines. For
MOVES we anticipate this trend will continue. Because fuel and technology projections were not
available from AEO, for MOVES defaults we carried 1999 distributions forward to 2050. These
distributions are summarized in Table 7.23. Engine size and vehicle weight distributions were
also carried forward from 1999. All buses were assigned to the Heavy-Duty Truck regulatory
class.
Table 7.23. Fuel and Engine Technology Fractions for 2000-and-later Buses
Intercity Buses
Transit Buses
School Buses
Diesel CIC
1
0.878041
0.958461
Gasoline CIC
0
0.005998
0.041539
CNG CIC
0
0.113296
0
Electric
0
0.002666
0
7.7.4. Motor Homes and Single Unit Short-haul and Long-haul Trucks
For Motor Homes and Single Unit Short-haul and Long-haul Trucks, MOVES uses the
AEO2004 projections for medium duty vehicles. AEO Table 55 lists sales projections for
medium-duty freight trucks powered by diesel, gasoline, liquified petroleum gas and compressed
natural gas. These projections were used to compute future distributions for these fuels.
Furthermore, AEO Table 146 lists technology penetrations for Class 4-6 freight vehicles.
Gasoline direct injection trucks were assigned to the MOVES gasoline advanced internal
combustion category and diesel trucks with "turbo, direct injection, thermal" engine
improvements were assigned to the MOVES diesel advanced internal combustion category. The
resulting distributions are summarized in Table 7.24.
We used the engine size and vehicle weight distributions from 1999 for future years.
Where a future fuel was not part of the fleet in 1999, we used the 1999 size and weight
distribution for gasoline conventional internal combustion vehicles. Where a future diesel
engine technology was not part of the source type fleet in 1999, we used the 1999 size and
weight distribution for diesel conventional internal combustion vehicles.
-------
Table 7.24. Fuel and Engine Technology Fractions for 2000-and-later Motor
Homes and Single-Unit Short-haul and Long-haul Trucks
Model
Year
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026+
Gasoline
CIC
0.23050
0.23050
0.22307
0.21608
0.20963
0.20377
0.19817
0.19271
0.18801
0.18386
0.18015
0.17683
0.17394
0.17144
0.16918
0.16712
0.16517
0.16350
0.16177
0.12344
0.12249
0.11863
0.11721
0.11635
0.11558
0.11489
0.11489
Gasoline
AIC
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.036729
0.036198
0.038798
0.039069
0.038783
0.038527
0.038296
0.038296
Diesel CIC
0.762812
0.762812
0.767655
0.772364
0.775387
0.777286
0.777746
0.777069
0.77743
0.778309
0.779640
0.781240
0.783217
0.785759
0.788316
0.791025
0.793481
0.606634
0.604079
0.601740
0.600768
0.602232
0.603619
0.604494
0.605453
0.606397
0.606397
Diesel AIC
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.189632
0.193720
0.197787
0.200256
0.200744
0.201206
0.201498
0.201818
0.202132
0.202132
CNG CIC
0.004410
0.004410
0.006332
0.008224
0.011180
0.014343
0.018510
0.022619
0.026181
0.029125
0.031436
0.033166
0.034098
0.034260
0.034146
0.033797
0.033558
0.033040
0.032875
0.032762
0.032566
0.032155
0.031739
0.031478
0.031337
0.031155
0.031155
LPG CIC
0.002283
0.002283
0.002944
0.003337
0.003804
0.004598
0.005574
0.007605
0.008376
0.008708
0.008770
0.008761
0.008749
0.008538
0.008356
0.008055
0.007786
0.007191
0.007558
0.007539
0.007722
0.007446
0.007157
0.007396
0.007283
0.007130
0.007130
52
-------
7.7.5. Refuse and Combination Trucks
For Refuse, Short-haul and Long-haul Combination Trucks, MOVES uses the AEO2004
projections for heavy-duty freight trucks. AEO Table 55 lists sales projections for heavy-duty
freight trucks powered by diesel, gasoline, liquified petroleum gas and compressed natural gas.
For refuse trucks, these projections were used directly to compute future distributions for these
fuels. However, because MOVES does not support LPG- and CNG-fueled combination trucks,
for combination trucks, these vehicles were assigned to the diesel category.
Furthermore, AEO Table 146 lists technology penetrations for Class 7-8 freight trucks.
Trucks with "higher cylinder pressure", "improved injection & combustion" and "waste
heat/thermal management" were assigned to the MOVES diesel advanced internal combustion
category. The resulting distributions are summarized in Table 7.25.
We used the engine size and vehicle weight distributions from 1999 for future years.
Where a future fuel or engine technology was not part of the source type fleet in 1999, we used
the 1999 size and weight distribution for diesel conventional internal combustion vehicles.
All Refuse Trucks were assigned to the Heavy-Duty Truck regulatory class.
Combination Trucks were distributed among the Light Duty Truck (LOT) and Heavy Duty
Truck (HDT) regulatory classes. Where a future fuel or technology was not part of the source
type fleet in 1999, we used the regulatory class distribution for diesel conventional internal
combustion vehicles.
53
-------
Table 7.25. Fuel and Engine Technology Fractions for Refuse Trucks and Short-
haul and Long-haul Combination Trucks
Model
Year
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026+
Refuse Trucks
Gasoline
CIC
0.012972
0.012972
0.012949
0.012969
0.013023
0.013110
0.013222
0.013355
0.013509
0.013679
0.013861
0.014050
0.014244
0.014440
0.014635
0.014825
0.015008
0.015186
0.015359
0.015523
0.015683
0.015840
0.015991
0.016137
0.016278
0.016415
0.016415
Diesel
CIC
0.984147
0.984147
0.983089
0.981930
0.980407
0.976878
0.971446
0.959230
0.929178
0.855626
0.702373
0.475144
0.329839
0.329806
0.329834
0.329892
0.329911
0.329942
0.329963
0.329937
0.329923
0.329937
0.329953
0.329951
0.329938
0.329930
0.329930
Diesel
AIC
0
0
0
0
0
0.001869
0.005396
0.015662
0.044167
0.116478
0.268796
0.495357
0.640275
0.640212
0.640266
0.640378
0.640415
0.640476
0.640517
0.640466
0.640438
0.640466
0.640497
0.640493
0.640468
0.640453
0.640453
CNG
CIC
0.001766
0.001766
0.002539
0.003394
0.004594
0.005902
0.007455
0.008941
0.010209
0.011237
0.012005
0.012531
0.012786
0.012787
0.012630
0.012403
0.012252
0.012083
0.011925
0.011853
0.011762
0.011617
0.011472
0.011366
0.011290
0.011208
0.011208
LPG
CIC
0.001114
0.001114
0.001423
0.001708
0.001976
0.002241
0.002482
0.002812
0.002937
0.002981
0.002965
0.002917
0.002855
0.002754
0.002636
0.002502
0.002413
0.002313
0.002237
0.002221
0.002194
0.002140
0.002086
0.002052
0.002026
0.001995
0.001995
Combination Trucks
Gasoline
CIC
0.012972
0.012972
0.012949
0.012969
0.013023
0.013110
0.013222
0.013355
0.013509
0.013679
0.013861
0.014050
0.014244
0.014440
0.014635
0.014825
0.015008
0.015186
0.015359
0.015523
0.015683
0.015840
0.015991
0.016137
0.016278
0.016415
0.016415
Diesel
CIC
0.987028
0.987028
0.987051
0.987031
0.986977
0.985021
0.981382
0.970983
0.942324
0.869844
0.717343
0.490592
0.345480
0.345347
0.345099
0.344797
0.344576
0.344338
0.344125
0.344010
0.343879
0.343695
0.343511
0.343369
0.343254
0.343133
0.343133
Diesel
AIC
0
0
0
0
0
0.001869
0.005396
0.015662
0.044167
0.116478
0.268796
0.495357
0.640275
0.640212
0.640266
0.640378
0.640415
0.640476
0.640517
0.640466
0.640438
0.640466
0.640497
0.640493
0.640468
0.640453
0.640453
54
-------
8. SourceUseType
The SourceUseType table lists average vehicle mass and three average road load
coefficients for each SourceType. The mass is listed in metric tons. The road load coefficients
are a rolling term "A," a rotatating term "B," and a drag term "C."
MOVES uses these coefficients to calculate vehicle specific power for each source type
according to the equation:
where A, B, and C are the road load coefficients in units of (kiloWatt second)/(meter tonne),
(kiloWatt second2)/(meter2 tonne), and (kiloWatt second3)/(meter3 tonne), respectively. Mis the
mass of the vehicle in kilograms, g is the acceleration due to gravity (9.8 meter/ second2), v is the
vehicle speed in meter/second, a is the vehicle acceleration in meter/second2, and sin(9 is the
(fractional) road grade.
The values in the SourceUseType table were averaged from values in the Mobile Source
Observation Database (MSOD). The values were weighted using the age and sourcebin
distributions described elsewhere in this report. In particular, the average values were computed
using the equation:
I
z=l, total # of ages
weightedvalue = •
A
aj • unweightedvalue
j=l , total # of sourcebins
j =1 , total # of sourcebins
IA
i =1, total # of ages
where the "unweighted value" was either the vehicle mid-point mass or one of the three different
road load coefficients determined from the road load-vehicle mass relations described below: Oj
were the sourceBinActivityFractions in the MOVES database and (3; were the ageFractions in the
MOVES database. Age fractions were matched to model years for calendar year 1999 (i.e.,
Model Year 1999 corresponds to vehicle agelD of 0; Model Yearl969 corresponds to agelD of
30.) Only sourcebins and ages with vehicles in the MSOD were used in these weightings. Thus,
the "total number of sourcebins" in the MSOD and "total number of ages" in the MSOD were
used to normalize the results.
8.1. SourceMass
The SourceMass was computed as the weighted average of the "mid-point" mass for the
Weight Class associated with each sourcebin. Sourcebins not represented in the MSOD were
excluded.
55
-------
Table 8-1. MOVES Weight Classes
Weight
ClassID
0
20
25
30
35
40
45
50
60
70
80
90
100
140
160
195
260
330
400
500
600
800
1000
1300
9999
5
7
9
Weight Class Name
Doesn't Matter
weight < 2000 pounds
2000 pounds <= weight < 2500 pounds
2500 pounds <= weight < 3000 pounds
3000 pounds <= weight < 3500 pounds
3500 pounds <= weight < 4000 pounds
4000 pounds <= weight < 4500 pounds
4500 pounds <= weight < 5000 pounds
5000 pounds <= weight < 6000 pounds
6000 pounds <= weight < 7000 pounds
7000 pounds <= weight < 8000 pounds
8000 pounds <= weight < 9000 pounds
9000 pounds <= weight < 10000 pounds
10000 pounds <= weight < 14000 pounds
14000 pounds <= weight < 16000 pounds
16000 pounds <= weight < 19500 pounds
19500 pounds <= weight < 26000 pounds
26000 pounds <= weight < 33000 pounds
33000 pounds <= weight < 40000 pounds
40000 pounds <= weight < 50000 pounds
50000 pounds <= weight < 60000 pounds
60000 pounds <= weight < 80000 pounds
80000 pounds <= weight < 100000 pounds
100000 pounds <= weight < 130000 pounds
130000 pounds <= weight
weight < 500 pounds (for MCs)
500 pounds <= weight < 700 pounds (for MCs)
700 pounds <= weight (for MCs)
Midpoint
Weight
[NULL]
1000
2250
2750
3250
3750
4250
4750
5500
6500
7500
8500
9500
12000
15000
17750
22750
29500
36500
45000
55000
70000
90000
115000
130000
350
600
700
8.2. Road Load Coefficients
The information available on road load coefficients varied by regulatory class.
Motorcycle road load coefficients are typically parameterized34 with mass dependent A
and C terms which take into account rolling resistance and aerodynamic drag. Parameters
adopted here are from the UN report:
A = 0.088M and C= 0.26 + 1.94xlO-4M
where M is the inertial mass of the motorcycle and driver and has units of metric tonnes.
For vehicles with a weight of 8500 Ibs or less, the road load coefficients were derived
from the track road load horspower (TRLHP@5omph) recorded in the MSOD.35 The calculations
applied the following empirical equations:36
-------
* TRLHP@50mph
* TRLHP@50mph
A = 0.7457*(0.35/50*0.447)
B = 0.7457*(0.10/(50*0.447)2)
C = 0.7457*(0.55/(50*0.447)3) * TRLHP@50mph
For the heavier vehicles, no road load parameters were available in the MSOD. Instead
EPA used the relationships of road load coefficent to vehicle mass from a study done by V. A.
Petrushov,37 as shown in Table 8-2. The mid-point mass for the sourcebin was used as the
vehicle mass.
Table 8-2. Road Load Coefficients for Heavy-Duty Trucks, Buses, and Motor
Homes
A(kW*s/m)/
M(tonne)
B(kW*s2/m2)/
M(tonne)
30
v^vv -o /mj)
/M(tonne)
8500 to 14000 Ibs
(3.855 to 6.350
tonne)
0.0996
0
3.40 xlO'4
(mass is the average
mass of the weight
category)
1.47 5
+ 5.22x10
masdkg)
14000 to 33000 Ibs
(6.350 to 14.968
tonne)
0.0875
0
1.97X10'4
(mass is the average
mass of the weight
category)
1.93 _5
+ 5.90x10
masfkg)
>33000 Ibs
(>14.968 tonne)
0.0661
0
1.79 x 10'4
(mass is the
average mass of the
weight category)
2.89 _ 5
masdkg)
Buses and
Motor Homes
0.0643
0
3.22 _5
masf(kg)
In both cases, values of A, B, and C were computed for each SourceBin-associated
vehicle in the MSOD and a weighted average was computed as described above. The final
SourceMass and road load coefficients for all SourceTypes are listed in Table 8-3.
57
-------
Table 8-3. SourcellseType Characteristics
Source
TypelD
11
21
31
32
41
42
43
51
52
53
54
61
62
HPMS
Vtype ID
10
20
30
30
40
40
40
50
50
50
50
60
60
SourceType
Name
Motorcycle
Passenger Car
Passenger Truck
Light Commercial
Truck
Interstate Bus
Urban Bus
School Bus
Refuse Truck
Single-Unit
Commercial Truck
Single-Unit Delivery
Truck
Motor Home
Combination
Commercial Truck
Combination
Delivery Truck
Rolling
TermA
(kW-s/m)
0.0251
0.031292
0.044224
0.047002
1.295151
1.0944
0.746718
1.417049
0.561933
0.498699
0.617371
1.963537
2.081264
Rotating
TermB
(kW-s2/m2)
0
0.002002
0.002838
0.003039
0
0
0
0
0
0
0
0
0
Drag
TermC
(kW-s3/m3)
0.000315
0.000493
0.000698
0.000748
0.003715
0.003587
0.002176
0.003572
0.001603
0.001474
0.002105
0.004031
0.004188
Source
Mass (metric
tons)
0.285
1.478803
1.866865
2.059793
19.59371
16.55604
9.069885
20.68453
7.641593
6.250466
6.734834
29.32749
31.40378
58
-------
9. RoadTypeDistribution
MOVES will calculate emissions separately for each HPMS facility type and for "off-
network" activity. The road type codes used in MOVES are listed in Table 9-1.
Table 9-1. Road Type Codes in MOVES
RoadTypelD
1
11
13
15
17
19
21
23
25
27
29
31
33
Description
Off Network
Rural Interstate
Rural Other Principal Arterial
Rural Minor Arterial
Rural Major Collector
Rural Minor Collector
Rural Local
Urban Interstate
Urban Other Freeways and Expressways
Urban Other Principal Arterial
Urban Minor Arterial
Urban Collector
Urban Local
For each SourceType, the RoadTypeVMTFraction field stores the fraction of total
VMT that is traveled on each of the 13 roadway types.
For MOVES2004, we used data from 1999 FHWA Highway Statistics, Tables VM-1 and
VM-2. VM-1 provides detail on VMT by vehicle type; VM-2 provides detail by roadway type.
At the time of this analysis, VM-1 (October 2000) had not been updated, but VM-2 was updated
in January 2002. We used the total values from the more recent VM-2 to distribute VMT by
roadway type and allocated them to vehicle class in proportion to the values in VM-1. We then
calculated road type VMT fractions for each HPMS Vehicle Type.
The FHWA Highway Statistics is currently considered the best available source for
national information regarding vehicle miles traveled. However, there are problems and
constraints associated with using the (mostly) self-reported data in Highway Statistics65. In
many cases, locally derived VMT data may be more accurate when modeling local areas.
The VMT distributions in Table 9-2 assume that all VMT reported by HPMS is
accumulated on one of the 12 HPMS roadway types. No VMT is currently assigned to the "off-
network" category in the national defaults. See the discussion of BaseYearOffNetVMT in
Section 11.2.
59
-------
Table 9-2. Road Type Fractions by HPMS Vehicle Type
RoadTypelD
1
11
13
15
17
19
21
23
25
27
29
31
33
Total
Motorcycles
0.0000
0.1040
0.0928
0.0643
0.0845
0.0235
0.0509
0.1598
0.0579
0.1326
0.1060
0.0444
0.0792
1.0000
Passenger
Cars
0.0000
0.0834
0.0870
0.0603
0.0753
0.0210
0.0454
0.1429
0.0668
0.1529
0.1223
0.0513
0.0914
1.0000
Other 2 axle -
4 tire vehicles
0.0000
0.0846
0.0908
0.0630
0.0807
0.0225
0.0486
0.1381
0.0650
0.1489
0.1190
0.0499
0.0890
1.0000
Buses
0.0000
0.1268
0.1060
0.0735
0.1608
0.0448
0.0969
0.0982
0.0404
0.0925
0.0739
0.0310
0.0552
1.0000
Single unit
trucks
0.0000
0.1149
0.1174
0.0815
0.1054
0.0294
0.0635
0.1209
0.0506
0.1158
0.0926
0.0388
0.0692
1.0000
Combination
trucks
0.0000
0.3247
0.1192
0.0827
0.0490
0.0137
0.0296
0.1798
0.0278
0.0636
0.0508
0.0213
0.0380
1.0000
We are currently assuming identical VMT distributions for all SourceTypes within an
HPMS Vehicle Type. However the MOVES model is designed to allow roadway type allocation
by SourceType and one would expect the different SourceTypes to have different roadway type
allocations. For example, the long-haul trucks generally would have a greater fraction of travel
on rural interstates than the short-haul trucks. If such data becomes available we would like to
update the database.
60
-------
10. Average Speed Distribution
The AvgSpeedDistribution table provides the fraction of driving time for each
SourceType, Road Type, Day, Hour, and Speed Bin in a field called AvgSpeedFraction.
values sum to one for each combination of SourceType, Road Type, Day, and Hour.
The
For MOVES2004, the urban driving values were derived from the default speed
distributions (SVMT) in MOBILE6. The MOBILE6 speed fractions were adapted to MOVES
converting the fraction of miles travelled to the fraction of time used, and by mapping from the
MOBILE6 road types to the MOVES road types. This road type mapping is detailed in Table 10-
1. The time fractions were normalized to add to one for each hour of the day over all 14 speed
bins. The values for the off-network roadway type in MOVES2004 were set to null. The detailed
distributions are available in the MOVES default database. See Table 9-1 for a description of the
MOVES road type numbers used in Table 10-1. Only urban roadways obtain their values from
the default MOBILE6 speed distributions.
Table
10-1. Map
MOVES
Road Type
ping of MOVES Road Types to MOBILES Road Types
MOBILE6 Road Type
Arterial/Collector
27,29,31
Freeway
23,25
Local
33
Ramp
—
Average speed used for rural driving relied on an analysis of recent driving data collected
entirely in California under studies performed for the California Department of Transportation
(Caltrans) performed by Sierra Research, Inc, 38 which is summarized here. Under these Caltrans
driving studies, instrumented "chase cars" were equipped with laser rangefinders mounted
behind the front grill of each chase car. The studies were performed in the Sacramento area, the
San Francisco Bay area and the San Joaquin Valley. Another driving study was also conducted
in the South Coast (i.e., Los Angeles Basin), but was conducted entirely in urbanized areas.
Thus, this data was not used for the rural area analysis.
The datasets contained driving in both urban and rural areas. In the post-processing that
was performed under each of these studies, the type of roadway the vehicle was traveling on
during each second was also recorded in the output dataset. Since the datasets contained the
Highway Performance Monitoring System (HPMS) Functional Class designation, it was easy to
divide the driving data from these studies into rural functional class groups for creating average
speed distributions. (The urban area travel in these datasets was discarded for this analysis.)
The average speed was calculated over each link traverse for the individual links in each
data set. A link traverse is defined as a one-way driving traverse of the entire extent of a
roadway link. A review of the links identified in the data showed that although distances of most
links identified as above ranged between 0.5 to 5 miles, a few of the them were ten miles or
longer. These longer links were generally restricted to limited access freeways and highways or
remote county roads. In rural areas, the difference in average speeds calculated over
conventionally defined links versus longer link sections as identified in the route-based driving
studies is not likely to be significant because of the general lack of traffic congestion on these
61
-------
rural roads.
Once the average speed was calculated for each link traverse, it was allocated into one of
sixteen speed bins defined by EPA for the purpose of calculating speed distributions for use in
MOVES. The MOVES speed bins are shown in Table 10-2. An important point is that although
the technical memo prepared by Sierra Research presents distributions based on the number of
observations in each bin (i.e. unweighted), the distributions contained in the
AvgSpeedDistribution table are weighted by the travel time on each link traverse, since
AvgSpeedFraction is meant to capture the fraction of time spent in each bin.
Table 10-2. MOVES Speed Bin Categories.
Bin
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Average Speed (mph)
2.5
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
Average Speed Range (mph)
speed < 2.5 mph
2.5 mph <= speed < 7.5 mph
7.5 mph <= speed < 12.5 mph
12.5 mph <= speed < 17.5 mph
17.5 mph <= speed < 22.5 mph
22.5 mph <= speed < 27.5 mph
27.5 mph <= speed < 32.5 mph
32.5 mph <= speed < 37.5 mph
37.5 mph <= speed < 42.5 mph
42.5 mph <= speed < 47.5 mph
47.5 mph <= speed < 52.5 mph
52.5 mph <= speed < 57.5 mph
57.5 mph <= speed < 62.5 mph
62.5 mph <= speed < 67.5 mph
67.5 mph <= speed < 72.5 mph
72.5 mph <= speed
The existing data from the studies used in this analysis were collected entirely in
California. Thus, use of these California results to represent national rural speed distributions
must include the critical assumption that average speeds within each HPMS functional class do
not significantly vary across the U.S on rural roadways. The California chase car data also only
included light-duty vehicles. Heavy-duty vehicles represent a significant fraction of rural area
travel and were not targeted during these chase car studies.
62
-------
11. HPMSVTypeYear
Three fields comprise HPMSVTypeYear in MOVES2004: HPMSBaseYearVMT,
BaseYearOffNetVMT, and VMTGrowthFactor.
11.1. HPMSBaseYearVMT
The HPMSBaseYearVMT field stores the base year VMT for each HPMS Vehicle Type.
This VMT was calculated from the FHWA VM-1 and VM-2 tables as for
RoadTypeDistribution, but instead of calculating fractions, we calculated VMT sums by HPMS
Vehicle Class. The resulting 1999 VMT by HPMS Vehicle Type is shown in Table 11-1.
11.2. BaseYearOffNetVMT
Off Network VMT refers to the portion of activity that is not included in travel demand
model networks or any VMT that is not otherwise reflected in the other twelve categories.
However, the reported HPMS VMT values, used to calculate the national averages discussed
here, are intended to include all VMT. Thus, for MOVES2004, the BaseYearOffNetVMT will
be zero for all Vehicle Types.
11.3. VMTGrowthFactor
The VMTGrowthFactor field stores a multiplicative factor indicating changes in total
vehicle miles for calendar years after the base year. Total VMT data are reported according the
HPMS vehicle classes discussed previously, i.e. passenger car, other 2-axle / 4-tire vehicle,
single-unit truck, combination truck, bus and motorcycle. VMTGrowthFactor is expressed
relative to the previous year's VMT; for example, 1 means no change from previous year VMT,
1.02 means a two percent increase in VMT, and 0.98 means a two percent decrease in VMT.
VMTGrowthFactor is used in the Total Activity Generator calculation of VMT for
calendar years after the base year, meaning calendar years 2000 through 2050 in MOVES2004.
It is important to note that VMTGrowthFactor is the key component for estimates of future
activity in MOVES, because the level of total activity in future years for most emission processes
— running, start and extended idle in MOVES2004 — is derived from projections of total VMT.
Projections of future populations based on sales growth, survival rates, etc. only are used to
allocate total VMT.
The sources for default estimates for VMTGrowthFactor are FHWA Highway Statistics
for 2000 through 2002 and AEO2004 for 2003 onward. Some additional analyses were required
to allocate the more aggregate AEO estimates for light-duty vehicles and trucks to the MOVES
Source Types.
Calendar years 2000 through 2002 growth factors were derived from estimates of total
VMT data as reported by FHWA's Highway Statistics, Table VM-1. Total VMT data are
reported according the HPMS vehicle classes discussed previously, i.e. passenger car, other 2-
axle / 4-tire vehicle, single-unit truck, combination truck, and bus. For these years the growth
63
-------
factors are simply total VMT for the calendar year divided by total VMT from the previous year.
The VMTGrowthFactors for 2000 through 2002 are shown in Table 11-1.
Table 11-1. BaseYearVMT and VMTGrowthFactor by HPMS Vehicle Class
HPMS Vehicle Class
Motorcycles
Passenger Cars
Other 2 axle - 4 tire vehicles
Buses
Single unit trucks
Combination trucks
1999 VMT
10,579,571,538
1,568,637,135,533
900,735,282,077
7,656,997,688
70,273,725,843
132,358,287,321
2000 Growth
0.99
1.021
1.026
0.992
1.004
1.021
2001 Growth
0.91
1.012
1.016
0.92
1.025
1.003
2002 Growth
0.991
1.019
1.024
0.968
1.048
1.015
Growth factors for calendar years 2003 through 2025 were calculated in the same manner
as 2000-2002 using NEMS projections of total VMT as reported inAEO2004. These estimates
are broken down by total Light-Duty (AEO2004 Supplemental Table 48), total Medium-Duty,
and total Heavy-Duty (AEO2004 Supplemental Table 55). The growth factors derived from the
AEO2004 Medium-Duty VMT estimates were applied to the single-unit truck and bus HPMS
vehicle classes. The growth factors derived from the AEO2043 Heavy-Duty VMT estimates
were applied to the combination truck vehicle class. VMTGrowthFactors derived from medium
and heavy-duty vehicle AEO2004 VMT projections are shown in Table 11-2.
64
-------
Table 11-2. 2003 and later VMTGrowthFactors for
Medium-Duty & Heavy-Duty Trucks
Calendar Year
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
Medium-Duty
(Single Unit Trucks, Buses)
.003
.036
.030
.018
.016
.015
.018
.023
.025
.026
.028
.026
.025
.028
.030
.029
.026
.029
.027
.030
.030
.030
.031
Heavy-Duty
(Combination Trucks)
1.007
1.041
1.041
1.033
1.029
1.025
1.026
1.029
1.028
1.027
1.028
1.026
1.024
1.025
1.027
1.026
1.022
1.023
1.021
1.024
1.025
1.025
1.028
Light-Duty VMT as reported in AEO2004 Supplemental Table 48 applies to total light-
duty growth from both cars and trucks; as such they do not reflect the higher growth rate of light
trucks relative to passenger cars brought on by steadily increasing sales of light duty trucks.
Separate VMTGrowthFactors for the Passenger Car and Other 2-axle/4-wheel Vehicle classes
were therefore developed based on estimates of car and light truck populations from AEO2004.
Using theAEO2004 estimates of total light-duty VMT and vehicle population (i.e., stock) growth
rates shown in Table 11-3, we calculated the "per-vehicle" VMT growth implied from these
estimates (total VMT growth divided by population growth). Assuming that per-vehicle VMT
growth is the same for cars and light trucks, we multiplied the total light-duty per-vehicle VMT
growth factors by car and light truck population growth factors presented in AEO Supplemental
Table 46. This produced the separate car and light truck VMT growth factors shown in Table
11-3, as the product of vehicle population growth and per-vehicle travel growth. The "car" rates
derived from NEMS were applied to the MOVES source types Passenger Car and Motorcycle,
and the "light truck" rates were applies to the MOVES source types Passenger Truck and Light
Commercial Truck.
-------
Table 11-3. 2003 and later VMTGrowthFactor Calculation for
Passenger Cars and Trucks
Calendar
Year
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
AEO Total Light-Duty Growth Factors
VMT
.019
.034
.027
.025
.023
.023
.023
.023
.023
.024
.025
.022
.021
.020
.020
.020
.020
.020
.020
.020
.020
.021
.022
Population
1.024
1.025
1.024
1.024
1.023
1.021
1.020
1.019
1.019
1.017
1.017
1.016
1.015
1.015
1.015
1.015
1.015
1.014
1.013
1.013
1.013
1.013
1.013
Per- Vehicle
VMT
0.995
.009
.002
.001
.001
.001
.002
.004
.005
.007
.008
.006
.006
.006
.005
.005
.005
.006
.006
.007
.007
.008
.009
AEO Population Growth
Factors
Cars
1.003
1.001
1.001
1.001
1.001
1.000
1.000
1.000
1.000
1.000
1.000
1.000
0.999
1.000
1.001
1.001
1.002
1.002
1.001
1.001
1.001
1.001
1.002
Light
Trucks
1.063
1.064
1.061
1.057
1.053
1.049
1.046
1.043
1.040
1.037
1.035
1.033
1.031
1.029
1.028
1.028
1.027
1.026
1.024
1.023
1.022
1.022
1.022
Calculated
VMTGrowthFactor
Cars
0.997
.010
.003
.002
.002
.002
.002
.004
.005
.007
.008
.006
.005
.005
.006
.006
.007
.007
.007
.008
.009
.010
.011
Light
Trucks
1.057
1.073
1.063
1.058
1.054
1.051
1.048
1.047
1.045
1.044
1.044
1.039
1.037
1.035
1.034
1.033
1.032
1.031
1.030
1.030
1.030
1.030
1.031
66
-------
12. Temporal Distributions of VMT
MOVES can estimate emissions for every hour of every day of the year. For this reason,
annual VMT estimates need to be allocated to months, days, and hours.
A 1996 report from the Office of Highway Information Management (OHEVI)39 describes
analysis of a sample of 5,000 continuous traffic counters distributed through the United States.
EPA obtained the data used in the report and used it to generate inputs in the form needed for
MOVES2004.
The report does not specify VMT by SourceType or Vehicle Type. Thus, we currently
use the same value for all SourceTypes.
12.1. MonthVMTFraction
For Month VMTFraction, we will use the data from the OHEVI report's Figure 2.2.1
"Travel by Month, 1970-1995," but modified to fit MOVES specifications.
The figure shows VMT/day, normalized to January=l. For MOVES2004, we need the
fraction of total VMT per month, with different values for leap year and non-leap year. We
computed the fractions using the report values and the number of days in each month.
Table 12-1. MonthVMTFraction
Month
January
February
March
April
May
June
July
August
September
October
November
December
Normalized
VMT/day
1.0000
1.0560
1.1183
1.1636
1.1973
1.2480
1.2632
1.2784
1.1973
1.1838
1.1343
1.0975
MOVES
not Leap
Year
0.0731
0.0697
0.0817
0.0823
0.0875
0.0883
0.0923
0.0934
0.0847
0.0865
0.0802
0.0802
MOVES
Leap Year
0.0729
0.0720
0.0815
0.0821
0.0873
0.0881
0.0921
0.0932
0.0845
0.0863
0.0800
0.0800
12.2. DayVMTFraction
The OHEVI report provides VMT percentage values for each day and hour of a typical
week for urban and rural roadway types for various regions of the United States for both 1992
and 1995. The data obtained from the OHEVI report is not disaggregated by month or
67
-------
SourceType. The same values will be used for every month and SourceType. We used 1995
data (which is very similar to 1992) as it is displayed in Figure 2.3.2 of the OHTM report.
For the DayVMTFraction needed for MOVES2004, we summed the reported percentages
for each day. Note, the report explains that data for "Sam" refers to data collected from Sam to
4am. Thus data labeled "midnight" belongs to the upcoming day.
Table 12-2. DayVMTFractions
Mon
Tues
Wed
Thurs
Fri
Sat
Sun
Total
Rural
0.1363
0.1352
0.1387
0.1442
0.1668
0.1447
0.1342
1.0000
Urban
0.1442
0.1489
0.1516
0.1536
0.1641
0.1304
0.1073
1.0000
We assigned the "Rural" fractions to the rural Roadtypes (11-21) and the "Urban"
fractions to the urban Roadtypes (23-33). The correct distribution for "Off network" VMT is
unknown. Since the majority of U.S. travel is urban, any VMT assigned to "Off network" will
be assigned the urban distribution of DayVMTFractions. The MOVES2004 default VMT
fraction on "Off-network" is zero.
12.3. HourVMTFraction
For HourVMTFraction we used the same data as for DayVMTFraction. We converted
the OHEVI report data to percent of day by dividing by the DayVMTFraction.
There are separate sets of HourVMTFractions for "Urban" and "Rural" roadway types.
Roadway types were assigned as for DayVMTFraction. All SourceTypes use the same
HourVMTFraction distributions. Table 12-3 shows only the "Urban" HourVMTFractions.
68
-------
Table 12-3. HourVMTFractions
Hour
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Sunday
0.0235
0.0161
0.0121
0.0079
0.0064
0.0085
0.0147
0.0208
0.0292
0.0424
0.0542
0.0628
0.0731
0.0750
0.0757
0.0760
0.0756
0.0722
0.0646
0.0544
0.0458
0.0384
0.0299
0.0207
Monday
0.0093
0.0059
0.0047
0.0044
0.0070
0.0188
0.0463
0.0700
0.0611
0.0509
0.0513
0.0557
0.0593
0.0594
0.0634
0.0721
0.0781
0.0786
0.0590
0.0428
0.0340
0.0298
0.0224
0.0156
Tuesday
0.0095
0.0059
0.0048
0.0046
0.0071
0.0191
0.0478
0.0723
0.0627
0.0512
0.0499
0.0538
0.0570
0.0573
0.0616
0.0711
0.0779
0.0788
0.0595
0.0431
0.0344
0.0310
0.0235
0.0162
Wednesday
0.0098
0.0061
0.0049
0.0046
0.0070
0.0189
0.0475
0.0719
0.0625
0.0507
0.0494
0.0537
0.0567
0.0570
0.0614
0.0707
0.0774
0.0785
0.0600
0.0437
0.0353
0.0317
0.0240
0.0167
Thursday
0.0103
0.0066
0.0053
0.0048
0.0071
0.0186
0.0464
0.0701
0.0611
0.0502
0.0494
0.0538
0.0568
0.0572
0.0614
0.0704
0.0768
0.0777
0.0601
0.0447
0.0360
0.0325
0.0252
0.0176
Friday
0.0103
0.0068
0.0055
0.0049
0.0068
0.0172
0.0421
0.0644
0.0570
0.0487
0.0498
0.0549
0.0584
0.0592
0.0634
0.0709
0.0750
0.0739
0.0602
0.0474
0.0373
0.0339
0.0291
0.0228
Saturday
0.0198
0.0131
0.0100
0.0071
0.0072
0.0119
0.0215
0.0318
0.0423
0.0517
0.0602
0.0669
0.0699
0.0686
0.0685
0.0687
0.0675
0.0643
0.0595
0.0495
0.0404
0.0378
0.0341
0.0277
69
-------
13. DriveSchedule
DriveSchedule refers to a second-by-second vehicle speed trajectory which is used in the
determination of operating mode distribution, defined (for the running energy consumption
pollutant/process) by Vehicle Specific Power (VSP) and vehicle speed. A key feature of
MOVES is the capability to accommodate any number of drive schedules to represent driving
patterns across source type, roadway type and average speed. For the national default case,
MOVES2004 employs 40 drive schedules, mapped to specific source types and roadway types.
The average speed of a driving schedule is used to determine the weighting of that schedule for a
given roadway, based on the average speed of the roadway. Briefly, the calculated VSP
distribution determined for a given driving schedule and the next nearest driving schedule which
brackets the roadway average speed, are averaged together, weighted by the proximity of the
roadway average speed to the driving schedule average speeds. In this way, the VSP distribution
of any roadway average speed can be determined from two driving schedules, whose average
speeds bracket the roadway average speed. This is presented in detail in the discussion of the
Operating Mode Distribution Generator in the MOVES Design Documentation.
For brevity, the entire body of drive schedule information is not presented in this
document. The reader is referred to the MOVES database, where three MOVES database tables
encompass drive schedule information. DriveSchedule provides the average speed of traffic on
the road type and the drive schedule name. DriveScheduleAssoc defines the set of schedules
which represent combinations of source use type and road type. The schedules within a set are
differentiated by the average speed of traffic on the road type. Although not defined as unique
road types, freeway ramp cycles are accounted for as separate schedules; they will be associated
with interstates and freeways. DriveScheduleSecond contains the second-by-second vehicle
trajectories for each schedule. In some cases the vehicle trajectories are not contiguous; that is,
they represent several unconnected microtrips.
Table 13-1 shows a complete list of the driving schedules used in the default case and
their associated average speed.
70
-------
Table 13-1. Default MOVES Drive Schedules
Drive Schedule Set
Light-Duty Non-Freeway
Light-Duty Freeway
Medium Heavy-Duty Non-Freeway
Medium Heavy-Duty Freeway
Heavy Heavy -Duty Non-Freeway
Heavy Heavy -Duty Freeway
Bus Non-Freeway
Refuse Truck
DriveScheduleName(ID)
Low Speed (101)
New York City (102)
Non-Freeway LOS EF (103)
Non-Freeway LOS CD (104)
Non-Freeway LOS AB (105)
Freeway LOS G (151)
Freeway LOS F (152)
Freeway LOS E (153)
Freeway LOS D (154)
Freeway LOS AC (155)
Freeway High Speed 1 (156)
Freeway High Speed 2 (157)
Freeway High Speed 3 (158)
Freeway Ramp (199)
5mph(201)
10 mph (202)
15mph(203)
20 mph (204)
25 mph (205)
30 mph (206)
30 mph (251)
40 mph (252)
50 mph (253)
60 mph (254)
Ramp (299)
5 mph (301)
10 mph (302)
15 mph (303)
20 mph (304)
25 mph (305)
30 mph (306)
30 mph (351)
40 mph (3 52)
50 mph (353)
60 mph (3 54)
Ramp (399)
Low Speed Urban (401)
30 mph flow (402)
45 mph flow (403)
Refuse Truck Urban (501)
AverageSpeed (mph)
2.5
7.05
11.55
19.23
24.75
13.13
18.61
30.49
52.87
59.66
63.23
68.21
75.99
34.6
1.81
10.53
15.55
20.37
24.36
30.83
37.37
45.3
55.5
60.06
29.2
1.19
10.75
15.22
19.81
24.87
30.81
34.9
46.89
54.33
59.5
26.7
15*
30*
45*
2.2
* Speed represents average of traffic the bus is traveling in, not the average speed of the bus, which is lower
due to stops.
-------
14. Drive Schedule Association
The DriveSchedules listed in Table 13-1 are associated with specific SourceTypes and
RoadTypes as summarized in Table 14-1. This table is an aggregated representation of the
information in Drive Schedule Association, which contains a mapping of every schedule to each
SourceType across each of the 12 HPMS roadway types.
Table 14-1. Drive Schedule Mapping
Source Use Type
Motorcycle
Passenger Car
Passenger Truck
Commercial Truck
Intercity bus
Single Unit Short Haul
Single Unit Long Haul
Motor Home
Transit bus
School Bus
Refuse Truck
Combination Short Haul
Combination Long Haul
Interstate,
Freeway/Expressway
Arterial, Local
Collector
Light-Duty Non-Freeway Schedules
Light-Duty Freeway Schedules
Light-Duty Ramp Schedule
Medium Heavy-Duty Non-Freeway
Medium Heavy-Duty Freeway
Medium Heavy-Duty Freeway
Bus Non-Freeway
Medium Heavy-Duty Freeway
(50 mph & 60 mph)
Bus Non- Refuse Truck Local
Freeway
Heavy Heavy-Duty Freeway
Heavy Heavy-Duty Non-Freeway
The default drive schedules listed in Tables 13-1 and 14-1 were developed from several
sources. The majority of the light-duty cycles are identical to those developed for MOBILE6
and documented in report M6.SPD.001.40 What we now refer to as "non-freeway" schedules are
the same as the "arterial" cycles used in MOBILE6; the name change was made to reflect the
application of these schedules to all non-freeway operation, including local roadways. The light-
duty schedules not included in the MOBILE6 work are Low Speed, New York City, High Speed
2 and High Speed 3. Low Speed is a historic cycle used in the development of speed corrections
for MOBILES and is meant to represent extreme stop-and-go "creep" driving. The New York
City Cycle is a historic test schedule representing congested urban travel with lots of stop-and-
go. It is used in EPA's running loss certification test procedure.41
High Speed 2 and 3 were developed specifically for MOVES2004. High Speed 1 was
the highest speed schedule in MOBILE6, with an average speed of 63 mph. EPA received many
comments with respect to MOBILE6 that this was not sufficient to capture the range of high
speed freeway driving in-use. The increase in speed limits as well as vehicle performance within
the past decade dictates the need to represent more extreme driving; High Speed 2 and 3 were
developed to represent these conditions. High Speed 2 is a 240-second segment of the US06
certification compliance cycle, with an average speed of 68 mph and a maximum of 80 mph.
High Speed 3 is 580-second segment of freeway driving from an in-use vehicle instrumented as
part of EPA's On-Board Emission Measurement "Shootout" program,42 with an average speed of
76 mph and a maximum of 90 mph. The addition of these schedules will serve to increase the
capacity of MOVES to reflect the higher speed freeway operation seen on the road today. It
72
-------
should be noted, however, that these schedules are only applied in MOVES2004 if
AverageSpeedDistribution contains operation in the highest speed bins; i.e. 70 mph and greater.
Medium-Duty and Heavy-Duty schedules were developed specifically for MOVES2004,
based on work performed for EPA by Eastern Research Group (ERG), Inc. and documented in
the report "Roadway-Specific Driving Schedules for Heavy-Duty Vehicles."43 ERG analyzed
data from 150 medium and heavy-duty vehicles instrumented to gather instantaneous speed and
GPS measurements. ERG segregated the driving into freeway and non-freeway driving for
medium and heavy-duty vehicles, then further stratified vehicles trips according the pre-defined
ranges of average speed covering the range of vehicle operation. ERG characterized
representative driving within each speed range, using distributions of vehicle specific power
(VSP), speed and acceleration. Driving schedules were then developed for each speed bin by
creating combinations of idle-to-idle "microtrips" until the representative target metrics were
achieved. The schedules developed by ERG are, thus, not contiguous schedules which would be
run on a chassis dynamometer, but are made up of non-continguous "snippets" of driving meant
to represent target distributions. For use in MOVES2004, the highway heavy-duty schedules
developed by ERG were modified to isolate operation on freeway ramps. The segments of
freeway microtrips identified by ERG as taking place on on-and off-ramps were extracted and
used to create medium-duty and heavy-duty ramp schedules (299 and 399). Thus, the schedules
which represent on-freeway driving do not contain ramp operation. Another minor modification
to the schedules for use in MOVES2004 was made to the time field in order to signify, within a
drive schedule, when one microtrip ended and one began. The time field increments two
seconds instead of one when each new microtrip begins. This two second increment signifies
that these should not be regarded by the model as contiguous operation.
It is possible (but unlikely) that users will specify average speeds which exceed the range
of schedules that apply to arterial and local roadways. In these cases, freeway schedules will be
sometimes used to model these unusually high average speed cases. Logically, any roadway
whose average speed approaches those of freeways is functionally approaching the behavior of a
freeway schedule. Similarly, in cases where average freeway speeds are unusually low, non-
freeway driving schedules may be used. The cases in which freeway schedules are available for
non-freeway driving, and vice versa, are indicated in the mapping shown in Table 14-1.
73
-------
15. SourceTypeHour
SourceTypeHours consists of two fields: StartsPerSHO and IdleSHOFactor.
15.1. StartsPerSHO
The StartsPerSHO field stores the factor used to determine the number of engine starts
(trips) per hour of vehicle operation for each Source Type by day of the week and hour of the
day. Each trip is assumed to begin with an engine start. After MOVES calculates the hours of
operation, MOVES calculates the number of trips from the amount of hours of source operation.
MOVES allows for unique values for trip starts per source hour of operation (SHO) for each
source use type, each day of the week and each hour of the day.
Three basic sources for information regarding the number of engine starts per hour of
vehicle operation were used for MOVES. The report, "Roadway-Specific Driving Schedules for
Heavy-Duty Vehicles,"44 combines data from several instrumented truck studies. The data was
used to directly determine the trip starts and hours of vehicle operation by hour of the day for
heavy-duty vehicles. Only non-parcel truck data was used. The data was grouped into medium
heavy-duty trucks (19,501 Ibs GVWR to 33,000 Ibs. GVWR) and heavy heavy-duty trucks
(greater than 33,000 Ibs. GVWR). Data from weekdays and weekend days were grouped. All
weekdays use the same hourly values and both weekend days use the same hourly values.
The estimate for light-duty passenger vehicles and light-duty trucks are derived from the
instrumented vehicle data collected for the FTP Study in Spokane, Baltimore and Atlanta.45
From this data the number of engine starts and hours of vehicle operation can be directly
determined for each hour of the day. Data from weekdays and weekend days were grouped. All
weekdays use the same hourly values and both weekend days use the same hourly values.
Engine start estimates for motorcycles and buses were derived from MOBILE6 estimates
for the number of engine starts, the number of miles traveled and the average speeds. The
number of engine starts per day are taken from the MOBILE6 default values. These values were
carried over from previous versions of the MOBILE model and, to our knowledge, are not
documented. The derivation of the number of miles traveled is described in the technical report,
"Fleet Characterization Data for MOBILE6: Development and Use of Age Distributions,
Average Mileage Accumulation Rates and Projected Vehicle Counts for Use in MOBILE6."46
The derivation of average speeds is described in the technical report, "Development of
Methodology for Estimating VMT Weighting by Facility Type."47
Engine start estimates for refuse trucks and motor homes are assumed to be the same as
for transit buses. This rough estimate is based on the assumption that, as in the case for transit
buses, refuse trucks and motor homes are started infrequently as compared to the hours of
operation.
Table 15-1 summarizes the average trip starts per source hour operating (SHO) from the
various data sources. Though not the values used in the MOVES database (MOVES allows the
number of trip starts per SHO to vary by hour of the day), the table shows the relative differences
between the various vehicle classes and summarizes data sources.
74
-------
Table 15-1. Data Sources for Trip Starts Per Source Hour of Operation (SHO)
Data Sources for Trip Starts Per Source Hour of Operation (SHO)
SourceTypelD
11
21
31
32
41
42
43
51
52
53
54
61
62
SourceTypeName
Motorcycle
Passenger Car
Passenger Truck
Light Commercial Truck
Intercity bus
Transit bus
School Bus
Refuse Truck
Single-Unit Commercial Truck
Single-Unit Delivery Truck
Motor Home
Combination Commercial Truck
Combination Delivery Truck
Source of Data
MOBILE6
LDData
LDData
LD Data
Transit bus
MOBILE6
MOBILE6
Transit bus
MDData
MDData
Transit bus
HD Data
HD Data
Trip Starts per SHO Average
3.718
5.631
5.631
5.631
1.879
1.879
6.740
1.879
3.404
3.404
1.879
1.231
1.231
Using MOBILE6 Trip Information
The start estimates for motorcycles and buses were derived from MOBILE6 estimates for
the number of engine starts, the number of miles traveled, and the average speeds. MOBILE6
divides the on-highway vehicle fleet into 28 separate vehicle classes. These classes are briefly
described in Table 15-2.
Each vehicle class has estimates for the number of trips per day (engine starts), the
number of miles traveled each day and the average speed traveled on all roadway types (defined
as trip distance divided by full trip time, including delay). From these three parameters, it is
possible to calculate the number of trips per hour of engine operation. Table 15-3 shows how
this is calculated for each of the vehicle classes in MOBILE6.
Of the values in Table 15-3, only the estimates for Transit Buses (HDDBT), School
Buses (HDDBS) and Motorcycles (MC) were needed for MOVES. These same estimates were
used for every hour of all days.
75
-------
Table 15-2. MOBILES Vehicle Classifications
Index
I
1
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Class
LDGV
LDGT1
LDGT2
LDGT3
LDGT4
HDGV2B
HDGV3
HDGV4
HDGV5
HDGV6
HDGV7
HDGV8A
HDGV8B
LDDV
LDDT12
HDDV2B
HDDV3
HDDV4
HDDV5
HDDV6
HDDV7
HDDV8A
HDDV8B
MC
HDGB
HDDBT
HDDBS
LDDT34
Description
Light-Duty Gasoline Vehicles (Passenger Cars)
Light-Duty Gasoline Trucks 1 (0-6,000 Ibs. GVWR, 0-3750 Ibs. LVW)
Light-Duty Gasoline Trucks 2 (0-6,001 Ibs. GVWR, 3751-5750 Ibs. LVW)
Light-Duty Gasoline Trucks 3 (6,001-8500 Ibs. GVWR, 0-3750 Ibs. LVW)
Light-Duty Gasoline Trucks 4 (6,001-8500 Ibs. GVWR, 3751-5750 Ibs. LVW)
Class 2b Heavy-Duty Gasoline Vehicles (8501-10,000 Ibs. GVWR)
Class 3 Heavy-Duty Gasoline Vehicles (10,001-14,000 Ibs. GVWR)
Class 4 Heavy-Duty Gasoline Vehicles (14,001-16,000 Ibs. GVWR)
Class 5 Heavy-Duty Gasoline Vehicles (16,001-19,500 Ibs. GVWR)
Class 6 Heavy-Duty Gasoline Vehicles (19,501-26,000 Ibs. GVWR)
Class 7 Heavy-Duty Gasoline Vehicles (26,001-33,000 Ibs. GVWR)
Class 8a Heavy-Duty Gasoline Vehicles (33,001-60,000 Ibs. GVWR)
Class 8b Heavy-Duty Gasoline Vehicles (>60,000 Ibs. GVWR)
Light-Duty Diesel Vehicles (Passenger Cars)
Light-Duty Diesel Trucks 1 (0-6,000 Ibs. GVWR)
Class 2b Heavy-Duty Diesel Vehicles (8501-10,000 Ibs. GVWR)
Class 3 Heavy-Duty Diesel Vehicles (10,001-14,000 Ibs. GVWR)
Class 4 Heavy-Duty Diesel Vehicles (14,001-16,000 Ibs. GVWR)
Class 5 Heavy-Duty Diesel Vehicles (16,001-19,500 Ibs. GVWR)
Class 6 Heavy-Duty Diesel Vehicles (19,501-26,000 Ibs. GVWR)
Class 7 Heavy-Duty Diesel Vehicles (26,001-33,000 Ibs. GVWR)
Class 8a Heavy-Duty Diesel Vehicles (33,001-60,000 Ibs. GVWR)
Class 8b Heavy-Duty Diesel Vehicles (>60,000 Ibs. GVWR)
Motorcycles (Gasoline)
Gasoline Busses (School, Transit and Urban)
Diesel Transit and Transit busses
Diesel School Busses
Light-Duty Diesel Trucks 1 (6,001-8500 Ibs. GVWR)
76
-------
Table 15-3. MOBILES Starts Per Day, Miles Driven Per Day and Average Speed
And Calculated Starts Per Source Hour Operating
Index
1
2
o
5
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Class
LDGV
LDGT1
LDGT2
LDGT3
LDGT4
HDGV2B
HDGV3
HDGV4
HDGV5
HDGV6
HDGV7
HDGV8A
HDGV8B
LDDV
LDDT12
HDDV2B
HDDV3
HDDV4
HDDV5
HDDV6
HDDV7
HDDV8A
HDDV8B
MC
HDGB
HDDBT
HDDBS
LDDT34
Starts/day
Weekday
7.28
8.06
8.06
8.06
8.06
6.88
6.88
6.88
6.88
6.88
6.88
6.88
6.88
7.28
8.06
6.65
6.65
6.65
6.65
6.65
6.65
6.65
6.65
1.35
6.88
6.65
6.65
8.06
Starts/day
Weekend
5.41
5.68
5.68
5.68
5.68
6.88
6.88
6.88
6.88
6.88
6.88
6.88
6.88
5.41
5.68
6.65
6.65
6.65
6.65
6.65
6.65
6.65
6.65
1.35
6.88
6.65
6.65
5.68
CY2000
Miles/day
29.4755
35.2916
35.2916
34.0771
34.0771
35.6267
30.9094
20.3003
27.6105
26.9164
22.8339
21.3321
21.3321
19.4586
10.7539
45.4056
49.4674
62.2014
65.185
65.0443
61.6706
108.9881
168.0957
10.0204
27.2301
97.6678
27.2301
43.8645
Average
Speed
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
27.6
Calculated
SHO
.067953
.278681
.278681
.234678
.234678
.290822
.119906
0.735518
1.00038
0.975232
0.827315
0.772902
0.772902
0.705022
0.389634
1.64513
1.792297
2.253674
2.361775
2.356678
2.234442
3.948844
6.090424
0.363058
0.986598
3.538688
0.986598
1.589293
Weekday
Starts/SHO
6.817
6.303
6.303
6.528
6.528
5.330
6.143
9.354
6.877
7.055
8.316
8.902
8.902
10.326
20.686
4.042
3.710
2.951
2.816
2.822
2.976
1.684
1.092
3.718
6.973
1.879
6.740
5.071
Source Hours Operating = (Miles per Day) / (Miles per Hour) = SHO
77
-------
15.2. IdleSHOFactor
The IdleSHOFactor field stores the factor used to determine the number of hours of
extended idling for each Source Type by day of the week and hour of the day. Extended idling,
also referred to as "hoteling," is defined as any long period of discretionary idling that occurs
during long distance deliveries by heavy-duty trucks.
No sources exist that directly measure extended idling in order to determine the total
hours of extended idling estimated for heavy-duty trucks. However, hoteling mainly occurs
among the largest (Class 8) trucks, which are now almost exclusively diesel. A paper by Lutsey,
et al., 48 recently submitted to the Transportation Research Board, provides some insights on how
truck hoteling relates to overall truck activity.
Federal law limits the number of hours which long haul truck drivers can operate each
day. These regulations are described in the Federal Register.49 Using the distribution of truck
hoteling duration times (shown in Figure 1 of the Lutsey, et al. paper) and assuming that long
haul truck drivers travel an average 10 hours a day when engaged in hoteling behavior, we can
estimate the average duration of hoteling as 5.9 hours for every 10 hours of long-haul truck
driving. However, for MOVES we need to know the fraction of hours spent hoteling versus
hours of vehicle operation by time of day. This value can be derived from the known truck
activity.
In particular, the report, "Roadway-Specific Driving Schedules for Heavy-Duty
Vehicles,"50 combines data from several instrumented truck studies. The data contains detailed
information about truck driver behavior; however, none of the trucks in any of the studies was
involved in long haul, interstate activity. We assumed that all long haul truck trips have the
same hourly truck trip distribution as the heavy heavy-duty trucks in the instrumented studies
and that all long haul trips are 10 hours long, and thus deduced an hourly distribution of long
haul trip ends. The distribution of hoteling durations from the Lutsey report was applied to these
trip-end distributions. From these calculations, we estimated the number of hours of truck
operation and hours of truck hoteling. For MOVES, we then calculated the ratio of hoteling
hours to truck operation hours for each hour of the day. Only weekday data was used.
In MOVES, only the long haul combination truck sourcetype is assumed to have hoteling
activity. All other source use types have hoteling activity fractions set to zero.
78
-------
i6. ZoneYearRoadType
The SHOAllocFactor field stores the factor used to determine the hours of vehicle
operation in each zone in each calendar year on each of the roadway types.
The spatial allocation of source hours operating distributes the domain-wide estimates of
hours of operation to the zones. In the macro-scale implementation of the model, the domain is
the nation and the zones are counties. The nationwide hours of operation are not known
(measured). However, roadway vehicle miles traveled (VMT) information is available in detail.
Since the allocation is by roadway type, it is reasonable to assume that the average speeds by
roadway type are the same in every county, which would make the hours of operation directly
proportional to the VMT on each roadway type. So, VMT will be used to determine the
allocation of source hours operating to counties.
The estimate for the VMT by county comes from the 1999 National Emission Inventory
(NEI) analysis documented by Pechan & Associates.51 The NEI estimates are based on the
Highway Performance Monitoring System (HPMS) data collected by the Federal Highway
Administration52 for use in transportation planning and vehicle type breakdowns from the EPA
MOBILE6 Emission Factor model.53 The NEI VMT estimates have been incorporated into the
National Mobile Inventory Model (NMIM) county database.
The VMT estimates were obtained from the NMIM database. VMT estimates for each
roadway type were determined for each county in each state and the allocation calculated using
the following formula.
CountyAllocation(i) = ( CountyVMT(i) / Sum(CountyVMT(i) )
The roadway types in the NMEVI database match the roadway types used in
MOVES2004. The county allocation values for each roadway type will sum to one for the
nation. Although the data is from 1999 calendar year estimates, the same allocations will be
used for all calendar years.
79
-------
17. ZoneYear
Zone Year consists of two fields: StartAllocFactor and IdleAllocFactor.
17.1. StartAllocFactor
The StartAllocFactor field stores the factor used to determine the number of starts in each
zone in each calendar year.
The trip start allocation distributes the domain-wide estimates of the number of trip starts
to the zones. In the macro-scale implementation of the model, the domain is the nation and the
zones are counties. Nationally, the number of trip starts are not known (measured), but roadway
vehicle miles traveled (VMT) is documented. Since the number of trips is roughly proportional
to the VMT, VMT will be used to determine the allocation of trip starts to counties.
The estimate for the VMT by county comes from the 1999 National Emission Inventory
(NEI) analysis.54 The NEI estimates are based on the Highway Performance Monitoring System
(HPMS) data collected by the Federal Highway Administration55 for use in transportation
planning and vehicle type breakdowns from the EPA MOBILE6 Emission Factor model.56 The
NEI VMT estimates have been incorporated into the National Mobile Inventory Model county
database.
The VMT estimates were obtained from the NMIM database. VMT estimates for each
county in each state and the allocation calculated using the following formula.
CountyAllocation(i) = ( CountyVMT(i) / Sum(CountyVMT(i) )
The county allocation values will sum to one for the nation. Although the data is from
1999 estimates, the same allocations will be used for all calendar years.
17.2. IdleAllocFactor
The IdleAllocFactor field stores the factor used to determine the hours of extended idling
in each zone in each calendar year.
No sources exist that directly measure extended idling in order to allocate the hours of
extended idling estimated for heavy-duty trucks. However, extended idling (or hoteling) occurs
primarily on long-haul trips across multiple states, which suggests that travel on rural and urban
interstates would best represent long-haul trips. Extended idling mainly occurs among the
largest (Class 8) trucks, which are now almost exclusively diesel. Since we have estimates for
the amount of rural and urban interstate VMT by Class 8 heavy-duty diesel trucks in each county
of the nation, we can use this estimate to create a national allocation factor for extended idling
hours.
The actual total demand for overnight parking by trucks has been estimated by the
Federal Highway Administration on a state by state basis.57 These estimates were used to
determine the allocation to each State(i) using the following formula:
80
-------
StateAllocation(i) = StateParkingDemand(i) / Sum( StateParkingDemand(i))
The State allocation values will sum to one for the entire country. This method results in
no idling in Washington, D.C., Hawaii, Virgin Islands, or Puerto Rico, which make sense, since
none of these areas have VMT associated with rural or urban interstates.
The estimate for the VMT from Class 8 heavy-duty diesel trucks by county comes from
the 1999 National Emission Inventory (NEI) analysis.58 The NEI estimates are based on the
Highway Performance Monitoring System (HPMS) data collected by the Federal Highway
Administration59 for use in transportation planning and vehicle type breakdowns from the EPA
MOBILE6 Emission Factor model.60 The NEI VMT estimates have been incorporated into the
National Mobile Inventory Model (NMIM) county database.
The VMT estimates were obtained from the NMIM database. VMT estimates for Class
8 heavy-duty diesel trucks on rural and urban interstates were determined for each county in each
state and the allocation calculated using the following formula.
CountyAllocation(i) = StateAllocation * (CountyVMT(i) / Sum(CountyVMT(i))
The county allocation values will sum to one for the entire country. The sum of the
county allocations for a given State will equal the State allocation for that State, as determined
earlier.
81
-------
18. SCCVTypeDistribution
For some uses, particularly the preparation of national inventories, modelers will need to
produce output aggregated by EPA's Source Category Codes (SCC). The EPA's highway
vehicle SCC were derived from MOBILES and MOBILE6 and do not directly correspond to the
MOVES SourceTypes. For example, depending on its fuel and Gross Vehicle Weight (GVW)
limits, a vehicle in the MOVES Passenger Truck category may be coded with one of eight
SCCs—including the SCC for a Light-Duty Gasoline Truck 1, a Light-Duty Gasoline Truck 2, a
Heavy-Duty Gasoline Truck, a Light-Duty Diesel Truck, or one of the four codes for Heavy-
Duty Diesel Vehicle.
The MOVES model is designed to aggregate emissions to the user's choice of
SourceType or SCC using the SCCVTypeDistribution table. For each combination of
SourceType, Model Year and FuelType, the SCCVTypeDistribution table lists IDs for the
possible SCC and the fraction of vehicles assigned to each SCC. The full SCC also includes a
suffix that indicates roadway type. This is a simple mapping from the MOVES roadtype on
which the emissions occur.
While the existing SCCs only identify gasoline and diesel-fueled vehicles, it was
necessary to map alternatively-fueled vehicles to SCCs. All alternative-fuel vehicles were
mapped to the diesel SCC, with the same distribution between light and heavy-duty categories as
diesels in that model year. In the future, SCCs may be revised to explicitly handle alternative
fuels.
For most SourceTypes, the mapping to SCC was straightforward. These mappings are
summarized in Table 18-1. However, the trucks span a wide range of GVWs and, thus, a wide
range of SCCs. We used VIUS97 values for GVW to determine the truck SCC fractions by
model year. To separate Light-Duty Trucks 1 and Light-Duty Trucks 2, which are distinguished
by Loaded Vehicle Weights, we used information from the Oak Ridge National Laboratory
Light-Duty Vehicle database. And to separate Class 2a and 2b trucks, we used information from
Davis and Truitt.61 The resulting truck mappings are too complex to summarize here, but are
available in the MOVES database.
Table 18-1. SCC Mappings for Selected SourceTypes
Source
Type ID
11
21
21
41
41
42
42
43
43
54
54
SourceType
Motorcycle
Passenger Car
Passenger Car
Intercity Bus
Intercity Bus
Transit Bus
Transit Bus
School Bus
School Bus
Motor Home
Motor Home
Fuel Type
gasoline
gasoline
other
gasoline
other
gasoline
other
gasoline
other
gasoline
other
SCC-ID
5
1
6
4
12
4
12
4
12
4
10
SCC
prefix
2201080
2201001
2230001
2201070
2230075
2201070
2230075
2201070
2230075
2201070
2230073
Abbreviated
Description
Motorcycles
LDGV
LDDV
HDGV&B
HDDB
HDGV&B
HDDB
HDGV&B
HDDB
HDGV&B
M-HDDV
82
-------
19. MonthGroupHour
ACActivityTerms A, B and C are coefficients for a quadratic equation that calculates air
conditioning activity demand as a function of the heat index. They are applied in the calculation
of the A/C adjustment in the energy consumption calculator. The methodology and the terms
themselves were originally derived for MOBILE6 and are documented in the report "Air
Conditioning Activity Effects in MOBILE6."62 They are based on analysis of air conditioning
usage data collected in Phoenix, Arizona, in 1994. In MOVES, ACActivityTerms are allowed
to vary by monthGroup and Hour, in order to provide the possibility of different A/C activity
demand functions at a given heat index by season and time of day (this accounts for differences
in solar loading observed in the original data). However, for MOVES2004, the default data uses
one set of coefficients, to be applied across all MonthGroups and Hours. These default
coefficients represent an average A/C activity demand function over the course of a full day.
These coefficients are: -3.63 for A, 0.0725 for B, and -0.00028 for C. The A/C activity demand
function that would result from these coefficients is shown in Figure 19-1. A value of 1 means
the A/C compressor is engaged 100 percent of the time; a value of 0 means no A/C compressor
engagement.
Figure 19-1: Air Conditioning Activity Demand as a Function of Heat Index
0.9
0.8
0.7
•a 0.6
O
0.5
o
to.4
0.3
0.2
0.1
70 75 80 85 90 95
Heat Index (F)
100
105
110
-------
20. ZoneMonthHour
The ZoneMonthHour table contains environmental parameters that may affect energy
consumption, such as temperature, relative humidity. This table also contains the heat index
value, which is derived from the temperature and humidity. The heat index is used in the
calculation of air conditioning usage.
Temperature and relative humidity are linked, since the value of relative humidity is in
units of percent, which will vary, depending on the temperature. Values of temperature should
not be changed, unless the corresponding relative humidity value can also be determined.
The MOVES model allows temperature and relative humidity to vary by month, hour and
zone. In the macroscopic implementation of MOVES2004, Zone is defined as County. There is
an average temperature value (in degrees Fahrenheit) and relative humidity value (in percent) for
each hour of an average day for each month of the year for each county. The same temperatures
and humidity values are used for all calendar years.
The temperature and humidity values in the ZoneMonthHour table were derived from
data from the National Climatic Data Center (NCDC)63. The NCDC is the national and
international depository for weather observations. As part of its many duties, the NCDC
publishes and maintains many climatic data sets. Among these databases are historical and
current daily and monthly average maximum and minimum temperatures and dew point
measurements. However, it was necessary to obtain the daily maximum and minimum
observations for all stations for all years of interest, and compute the long and short term
averages from scratch in order to resolve missing monthly averages.
The daily maximum and minimum temperature data for all available stations were
processed into monthly averages. These stations covered all classifications, including First-
Order, Second-Order, (both Automated Surface Observing System (ASOS) and Automated
Weather Observing System (AWOS)) and cooperative. Meteorologists and Climatologists
routinely refer to the major National Weather Service (NWS) observation stations as "First
Order" stations. These usually include large cities and metropolitan airports. Second Order
generally means hourly airways observations are taken, but not in accordance with first order
requirements. Most are FAA-operated stations. Smaller stations are referred to as "Co-Operative
(Co-Op)" or "Third Order" stations. These stations are usually found in small towns, or rural
areas, and number in the thousands in the U.S. Following NCDC guidelines, a month's averages
were considered valid when no more than 5 days had missing data during that month. The data
were then organized to determine if the station has enough valid data to be included in
subsequent analyses. Using NCDC guidelines, a year of data is valid only if all of the months
have data.
After these filters were applied, the average monthly maximum and minimum
temperature data were adjusted to the common midnight-to-midnight observational period. This
adjustment is necessary since many of the cooperative stations take their observations either
early in the morning or late in the afternoon rather than at midnight. These observation times
-------
induce a bias into the monthly temperature averages. The contractor obtained the appropriate
correction values from the NCDC and applied them to the monthly averages.
Population centroids (latitude and longitude) for each county were obtained from the
2000 United States Census. Population, rather than geographic, centroids were used to provide
the best estimate of where the county's VMT would occur. From each county's centroid, the
distance and direction to each weather station was calculated. The nearest site in each of the
eight compass directions (an octal search) was used to identify the nearby measurements. The
distance was computed using the standard great circle navigation method and the constant course
direction was computed using the standard rhumb line method. A rhumb line is a line on a
sphere that cuts all meridians at the same angle; for example, the path taken by a ship or plane
that maintains a constant compass direction. For each octant, the stations were sorted by
distance. The station closest to the centroid for each octant was chosen for further processing. If
the closest station was more than 200 miles away, that octant was ignored. (Such situations
occurred near the oceans and the along the Canadian and Mexican borders. The temperatures
from these eight (or less) stations were then weighted together using inverse-distance weighting.
Relative humidity is a calculated value that depends on both temperature and dew point.
Average hourly dew points were computed employing the same octal search, inverse-distance
weighting scheme as used for temperature. The relative humidity was then computed from the
resulting hourly temperature and dew point pairs.
-------
21. Fuel Types
Energy consumption, expressed as fuel consumption, will vary depending on the fuel
used. MOVES2004 expresses fuel as one of nine categories. These categories are shown in
Table 21.1 below.
Table 21-1. Fuel Types
fuelTvoeld
1
2
3
4
5
6
7
8
9
fuelTypeDesc
Gasoline
Diesel Fuel
Compressed Natural Gas (CNG)
Liquid Propane Gas (LPG)
Ethanol (E85 or E95)
Methanol (M85 or M95)
Gaseous Hydrogen
Liquid Hydrogen
Electricity
The fuel types are represented by fractions of each source use type and model year
combination (SourceTypeModelYearlD) in the FuelEngFraction table by the FuelEngFraction
field. The sum of the FuelEngFraction values will be one for each SourceTypeModelYearlD.
Although the fractions of engines by fuel type is constant, the overall total number of engines of
each fuel type will vary by calendar year depending on the number of engines of each source use
type and model year in that calendar year.
21.1. FuelSubType
The properties of specific fuels in the broad FuelType categories can vary widely. These
differences are captured as fuel subtypes. The FuelSubtypes used by MOVES2004 are shown in
Table 21-2 below.
The fractions of vehicles which are use the various fuel subtypes are stored in the
MarketShare field of the FuelSupply table. The MarketShare values for a FuelType will sum to
one for each combination of CountylD, YearlD and MonthGroupID. Market share values may
vary by calendar year and month grouping for each county.
86
-------
Table 21-2. Fuel SubTypes
fuelSubtvoelD
10
11
12
20
21
22
30
40
50
60
70
80
90
fuelTvoelD
1
1
1
2
2
2
O
4
5
6
7
8
9
fuelSubtvpeDesc
Conventional Gasoline
Reformulated Gasoline (RFG)
Gasohol (E10)
Conventional Diesel Fuel
Biodiesel
Fischer- Tropsch Diesel
Compressed Natural Gas (CNG)
Liquid Propane Gas (LPG)
Ethanol (E85 or E95)
Methanol (M85 or M95)
Gaseous Hydrogen
Liquid Hydrogen
Electricity
21.2. FuelSupply
Each individual engine within a SourceType is assumed to be built to be powered by only
one of the FuelTypes shown in Section 21.1. However, within a Fuel Type, an engine can be run
on any of the FuelSubtypes within their FuelType shown in Section 21.2, depending on the
availability of the alternatives and other motivational factors. As a result, the fuel consumption
of each FuelSubtype may depend on the time and location, as well as the count and activity of
the SourceTypes.
MOVES2004 allows the FuelSubtype to vary by time (calendar year and season) and the
location (county). The distribution of fuel consumption between the various FuelSubtypes is
stored in the FuelSupply table. Table 21-3 describes the fields in the FuelSupply table.
Table 21-3. FuelSupply Table Description
Field Name
county ID
yearlD
monthGroupID
fuelSubtypelD
marketShare
Description
A political and territorial subdivision of a State (see definition of State) as
defined by FIPS standard codes.
Calendar year (4 digit integer). The valid range is 1990-2050.
Integer value which indicates a particular grouping of months. l=Summer,
2=Fall, 3=Winter, 4=Spring.
Identifies a particular kind of fuel within a FuelType. e.g. Gasoline may be
conventional, or Reformulated Gasoline (RFG), diesel may be conventional,
biodiesel, Fischer- Troppes, etc.
Decimal Fraction of the supply of this FuelType which this FuelSubtype
constitutes. Defaults to 1.0 for the lowest numbered fuelSubtypelD, 0.0 for
all others, if no record is present.
87
-------
The MOVES2004 database contains default values for Fuel Subtype market shares for
each season (MonthGroupID) in each year (YearlD) for each county (CountylD). These values
were derived from a more detailed set of fuel descriptions developed for the National Mobile
Inventory Model (NMIM) County database for the National Emission Inventory (NEI)64.
The NMIM fuel parameters were derived from several surveys: U.S. EPA's reformulated
gasoline (RFG) survey (U.S. EPA, 2000), the U.S. EPA Oxygenated Fuel Program Summary
(U.S. EPA, 2001), the TRW (previously NIPER) fuel survey (TRW, 1999), and the Alliance of
Automobile Manufacturers' (AAMA) North American Gasoline and Diesel Fuel Survey
(AAMA, 1999). The TRW fuel survey reports the data in several tables, including Table 9
(Motor Gasoline Survey, Season [Summer/Winter], Year [1999/2000], and Average Data for
Different Brands) and Table 10 (Motor Gasoline Survey, Season [Summer/Winter], Year
[1999/2000], and Average Data for Different Brands Containing Alcohols). Data for the percent
market share of oxygenated fuel sales were obtained from Oxygenate Type Analysis Tables
(1995-2000) (U.S. EPA, 2001) and the Federal Highway Administration website (FHWA 1999).
The survey fuels were assigned to individual counties by region and many fuel
parameters were combined to generate a single set of fuel parameters for each county. Separate
fuels were derived for Summer, Winter and Spring/Fall. Future calendar years fuel properties
were derived accounting for the phase-in of Phase 3 RFG in California, the Tier 2 motor vehicle
emissions standards and gasoline sulfur control requirements and other expected changes in fuel
properties due to regulations. For MOVES2004, each fuel was assigned to one of the
Fuel Subtypes and market shares were derived from the market share field used in NMIM.
88
-------
22. Peer Review
This section includes the complete comments received in November 2004 from the
formal EPA peer review of the initial draft "MOVES2004 Highway Vehicle Population and
Activity Data" report. EPA responses are in italics. The review was done by:
Debbie A. Niemeier, Professor and Chair
Department of Civil and Environmental Engineering
University of California, Davis 95616
530-752-0586 (phone)
530-752-7872 (fax)
dni emei er@ucdavi s. edu
General Comments
In general, EPA has developed the underlying vehicle population and activity databases
by integrating a number of different data. The methods and assumptions used to combine
external data sources and to populate the EPA databases reflect many of the same assumptions
applied in MOBILE6. There are likely many reasonable approaches to assembling a complete
census of vehicle population/activity, and thus, a variety of opinions about the assumptions
applied or the ways in which the various source data are combined. That is, some assumptions
are inevitable regardless of how the underlying databases are developed. Within the scope and
time provided for reviewing the technical documentation, my overall assessment is that EPA has
taken a reasonable approach to assembling the vehicle population/activity data required for the
operation of MOVES. Within this assessment, however, I did have a number of questions which
I elaborate on in this document.
The review begins with a brief comment on the technical documentation. This is the
starting point of the discussion because I believe many of the questions in the subsequent section
could be cleared up with additional detail in the technical documentation. The main questions in
the second section are related to specific variables/databases and the way in which some
assumptions about various databases are applied. The report concludes with a brief summary of
longer term suggestions that might be useful for EPA to consider during the development period
of the complete MOVES model.
Suggestions Related to the Documentation
As it stands the documentation could use significantly more references and/or details or
appendices. This may in fact be EPA's longer term intent. The current documentation does not
seem to provide enough details on how databases are combined or manipulated into their final
form. For example, for the MOBILE6 emissions model, the EPA calculated future year vehicle
populations by vehicle class and age by setting vehicle counts for the xth year equal to the sum
of vehicle counts for (x-l)th year multiplied by (1-scrappage rate for the xth year) plus the new
sales for xth year. That is, each year's vehicle population forecast is based on the vehicle
population estimated from the previous year.
89
-------
For MOBILE6,1 believe the 1996 vehicle population was used as the baseline for 1997
and forward estimates and survival rates were based on the 1996 World Vehicle Forecasts and
Strategies' Report (Pemberton, 1996). As I recall, EPA generally estimated scrappage rates as an
increasing trend over time. For example, for the periods 1995-1999, 2000-2004, 2005-2009,
2010-2014, and 2015-2020, the scrappage rates (as of percentage of the total in-use fleet) are
estimated as 5.77%, 5.7%, 6.01%, 6.34%, and 6.56%. Given that MOVES, in its current version,
is being estimated for the 1999 baseline only, some of this is not applicable, however, some of it
is, even to set the 1999 baseline. A simple reference would help to document whether MOBILE6
methods are being used (e.g., consider SurvivalRatea or SalesGrowthFactor, which seems to be
computed in a slightly different manner from that applied in MOBILE6). Additional details,
some of which I've tried to identify below, about the assumptions used to distill the main
ingredients of most of the tables would be helpful, including identifying when the basic methods
are similar to or diverge from MOBILE6.
EPA Response: Additional text was added to Section 3.2 and 5.1 indicating how MOVES
differs from MOBILE6.
There is also somewhat of an incongruence that arises in the documentation. It is clear
both from EPA's letter of request to review the documentation and from various statements in
the report that the main emphasis of this particular version is on producing national estimates.
However, there is also text (and some modeling capabilities) that suggests that MOVES is
"ready" for more localized estimation (e.g., at the roadway level or for more resolved time
periods). I personally would prefer the documentation to be consistent - either the model is
acceptable in EPA's view for use at the local level or references to localized model capabilities
should be taken out and perhaps summarized in a concluding chapter that identifies next steps.
EPA Response: The design of the MOVES model was intended to accommodate both
national (macroscale) and local (mesoscale) modeling. Modeling of local areas will require
areas to provide detailed roadway specific information that will not be provided by EPA. This
document only describes the information provided by EPA for the MOVES model for national
modeling. Because of the design of the MOVES model, it is difficult to avoid discussing the
model inputs in terms that exclude the local modeling input options. We will try to make clear
the distinctions between national (EPA supplied) inputs and local (user supplied) inputs in this
documentation.
In general, the technical documentation feels hurried. In many places (again, I've tried to
identify some of them below), the lack of detail on the methods, assumptions, and rationale for
these assumptions in the documentation makes truly understanding the development of the
databases a bit hard to discern. I would also suggest staying away from language used in the
introduction referring to "accuracy." The number of assumptions and lack of independent
a In SurvivalRate, the text is confusing and suggests that rates are based on 1990 baseline. Enough detail should be
added to the technical documentation to make clear how each of the datasets are phased "up" to the 1999 baseline
year. These kinds of details are directly related to the assembly of the databases themselves and might influence
results in ways that users should be made aware of.
90
-------
verification of the vehicle population and activity census makes it difficult to assess (or claim)
accuracy.
Questions Related to Database Details
Pg. 10 Please clarify the statement: "Some of the values are available directly from other
sources; other values were derived from the available data." EPA Response: The statement was
rewritten, "Some of the values are taken directly from the indicated sources; other values needed
to be derived from available data and are not found explicitly in any of the data sources."
Pg. 15 The discussion of the migration variable, which is set at 1 for this release, is an
example where EPA seems to imply more localized modeling is acceptable. I would suggest
gathering these kinds of statements into a final chapter on next releases. EPA Response: The
MOVES design includes migration rates. Discussion of migration is appropriate, even if the
value for national modeling is set to one. No changes were made.
Pg. 18 Provide the mapping from MOBILE6 to MOVES for the Relative MARs. Many
of the regression equations are of squared and exponential forms, are these functional forms
reasonable from an applied perspective? EPA Response: The functional forms were chosen to
best represent the form of the observed data. No changes were made.
Pg. 27 Here is an example of where the report implicitly emphasizes use of MOVES for
national estimates (or perhaps cautions against localized): "On a national scale..." EPA
Response: This is certainly an issue. Local areas will (hopefully) have a better idea of how
flexible fueled vehicles are operated in their areas. However, this document is clearly not
intended to be guidance on how local areas might change the assumptions used for the national
averages. No changes were made.
Pg. 45 Some underlying rationale for the decisions made with respect to splits derived
for the data in AEO Table 45 should be provided. Why is splitting gas hybrids between moderate
and full a reasonable assumption? EPA Response: Text was added to Section 7.7 to better
explain the need for the various engine technology categories shown in Table 7-20.
Pg. 45 Why not use 2004 or 2005 size and weight distributions for future years instead of
1999? EPA Response: The statement was rewritten, "The inputs for determining default
SourceBinDistributions for model years 2000-and-later were generally based on fuel and engine
technology projections from AEO2004 and on the 1999 calendar year regulatory class, size and
weight distributions used in MOVES." Regulatory class, size and weight distributions for other
calendar years are not yet available.
Pg. 56 I would suggest adding at least a mention of the problems and constraints
associated with using the (mostly) self-reported data in Highway Statistics. See Hendren and
Niemeier65 (2001) for some background. EPA Response: Text was added to Section 9 to
caution the reader.
-------
Pg. 57 Here is also where WIM data (discussed later) might be useful. EPA Response:
The statistical tools necessary to use the use weigh-in-motion data to supplement or replace
vehicle census data have not yet been developed. No changes were made.
Pg. 58 There are serious limitations to the data used to develop the speed distributions
for MOBILE6. Suggest instead of just translating these data, EPA utilize the new California
chase data that was collected as part of the CAMP effort. These data provide a much more robust
sample in terms of sample size and representativeness. In the mapping on Table 10-1, where are
collectors? EPA Response: The California chase data was not yet available at the time the
national average estimates for MOVES were developed. However, data from these studies is
now becoming available and the average speed distributions for rural roadways from the
California studies will now be used instead of the MOBILE6 estimates. All twelve of the HPMS
roadway types are represented in Table 10-1, including collectors.
Pg. 59 The BaseYearOffNetVMT seems to conflict with what is implied in Table 9-1,
where all of the functional classes are included. Yet, most travel models don't include local
roads, which (I think) would actually be captured in this parameter. When you look at Table 9-2,
source type fractions appear for local roads (and there can be collectors not included in the travel
networks as well). Need to clarify whether these fractions and types are used or not in the current
version. EPA Response: Text was added to Section 9 and 11.2 to clarify the meaning of "off
network" VMT.
Pg. 64 Please clarify the statement "The data does not vary by month or SourceType."
Do the automatic counters give a breakdown by source type? "Do not vary" seems to imply there
very little month to month variation. There have been studies through the years suggesting
monthly variation between summer and winter for example. Perhaps provide a standard error to
justify this statement? Also on pg. 64, there is a statement "The correct distribution for "off
network" VMT..." that seems to conflict with the BaseYearOffNetVMT discussion? EPA
Response: The statement in Section 12.2 was rewritten, "The data obtained from the OH1M
report is not disaggregated by month or SourceType. The same values will be used for every
month and SourceType." The discussion in Section 12.2 was also rewritten to clarify that the
urban day of the week distribution is applied to the off network VMT (if any).
Pg. 64 Are the hourly VMT fractions computed in Table 12-3 to be applied across all
roadway types (e.g., rural versus urban)? EPA Response: No. There are separate hourly VMT
fractions for urban and rural driving. Text was added to Section 12.3 to clarify the content of
Table 12-3 (urban only).
Pg. 67 VSP is typically applied on a second by second basis. How do the driving
schedules combine with VSP? And if the drive cycle acceleration and speed are inputs to the
VSP calculation, is this averaged over the drive cycle or calculated sec by sec? EPA Response:
Text has been added to Section 13 to discuss briefly how driving schedules are combined.
92
-------
Pg. 68 Table 14-1 seems to imply that freeway schedules are used for arterial and local
roads? Is this correct? EPA Response: That is correct. Text has been added to Section 14 to
discuss this fact.
Longer Term Considerations
There are some interesting longer-term fundamental issues related to the vehicle
population/activity data required for MOVES that EPA could begin to assess. One main issue
worth considering is the value of continuing to construct what is essentially a vehicle and activity
census, which requires a great many assumptions and sometimes less than optimal use of less
than optimal databases. The alternative would be to concentrate on the development of statistical
sampling and modeling methods that would provide the ability to statistically produce a robust
vehicle profile.
For example, in the case of mileage accrual, Miller et al (2001)66 noted that, in contrast to
that represented in MOBILE6, mileage accrual is nonlinearly related to vehicle age, and the
distribution of mileage accruals for vehicles of the same age is likely to be normal. Miller et al.
also argued that the reason such a discrepancy exists between MOBILE6 estimates and observed
data is because vehicles with different odometer readings will likely have different scrappage
rates. For example, a vehicle with a higher odometer reading is likely to have a higher scrappage
rate than a vehicle with a lower odometer reading, even if they are of comparable age.
The way in which a vehicle population and activity census is developed necessarily
involves many assumptions that might better be captured in a statistical model. Each time an
update is required, a sampling protocol could be implemented and model parameters updated.
This at least would provide the opportunity to assess issues related to variability and precision.
The use of weigh-in-motion data would also provide a better linkage between vehicle types and
activity for freeway related travel. WEVI stations are usually located to provide reasonable
representation of freeway activity, particularly for heavy duty vehicles. It might be useful to
examine these data with respect to MOVES and the ability to define statistical relationships
instead of relying on full development of a census.
EPA Response: EPA is looking at weigh-in-motion data as a source of information
about the distribution of vehicles on roadways and will incorporate the information into MOVES
as it becomes available.
EPA should consider developing a mapping scheme between travel models and the
MOVES vehicle categories. Right now, not only is mapping done for the MOVES model (such
as that shown in the technical documentation), but then two other steps of mapping are also
performed. The first with the travel model, in which at best, there are only general categories of
LDV, LDT and "goods movement," and the second, when the emissions inventory is prepared
for photochemistry. The result of all this series of mapping between vehicle sources is almost
certainly a cause for error propagation.
93
-------
23. References
1 U.S. Census Bureau, 1997 Vehicle Inventory and Use Survey, CD-EC97-VIUS. January 2000.
Online at http://www.census.gov/prod/www/abs/vius-pdf.html
2 R.L. Polk & Co., National Vehicle Population Profile.® Southfield, MI. 1999. Online at
http://www.polk.com/products/nvpp.asp
3 R.L. Polk & Co, TIP® Vehicles in Operation. Southfield, MI. 1999. Online at
http://www.polk.com/products/tip_jiet.asp
4 U.S. Federal Highway Administration. Highway Statistics, 1999. Table MV-1, "State Motor
Vehicle Registrations," October 2000. Online at
http://www.fhwa.dot.gov/ohim/hs99/index.htm
5 U.S. Federal Highway Administration. Highway Statistics, 1999. Table MV-10, "Bus
Registrations,"October 2000. Online at http://www.fhwa.dot.gov/ohim/hs99/index.htm
6 U.S. Federal Highway Administration. Highway Statistics, 1999. Table VM-1, "Annual
Vehicle Distance Travelled in Miles and Related Data by Highway Vehicle Category and
Vehicle Type," October 2000. Online at http://www.fhwa.dot.gov/ohim/hs99/index.htm
7 U.S. Federal Highway Administration. Highway Statistics, 1999. Table VM-2, "Functional
System Travel," January 2002. Online at http://www.fhwa.dot.gov/ohim/hs99/index.htm
8 U.S. Federal Transit Administration. National Transit Database 1999, Table 29. "Age
Distribution of Active Revenue Vehicle Inventory: Details by Transit Agency." Online at
http: //www. ntdprogram. com
9 Bobit Publications, School Bus Fleet Fact Book. Torrance. CA, 1999.
http://www.schoolbusfleet.com
10 Browning, Louis, Michael Chan, Doug Coleman, and Charlotte Pera. ARCADIS Geraghty &
Miller Inc. "Update of Fleet Characterization Data for Use in MOBILE6 - Final Report."
M6.FLT.002, EPA420-P-98-016, June 1998. Online at
http://www.epa.gov/otaq/models/mobile6/m6flt002.pdf
11 Energy Information Adminstration. Annual Energy Outlook 2003 (AEO2003\ Report #:
DOE/EIA-0383 (2003), released January 9, 2003. Online at
http ://www. eia. doe. gov/oiaf/archive/aeo03/index. html
19
Davis, Stacy C. and Susan W. Diegel, Transportation Energy Data Book, Edition 22. Center
for Transportation Analysis, Oak Ridge National Laboratory. ORNL-6967. September
2002. Online at http://www-cta.ornl.gov/data/ or
http ://www-cta. ornl. gov/cta/Publi cations/pdf/ORNL-6967. pdf
94
-------
13 Davis, Stacy C. and Susan W. Diegel, Transportation Energy Data Book, Edition 23. Center
for Transportation Analysis, Oak Ridge National Laboratory. ORNL-6967. October 2003.
Online http://www-cta.ornl.gov/data/
14 Ward's Automotive Inc. http://www.wardsauto.com/
15
Hart, Larry. R.L. Polk & Company. Personal communication, June 16, 2003.
16 National Household Transportation Survey (NHTS). Online at
http://nhts.ornl.gov/2001/index.shtml
17 American Bus Association. "Motorcoach Census 2000," conducted by R. L. Banks and
Associates, Inc. July 2000. http://www.buses.org/industry/ABA-RLBanksReport.pdf
18 Motorcycle Industry Council. Motorcycle Statistical Annual Magazine. 2000 and 2002. 2
Jenner St., Suite 150, Irvin, CA 92718. Phone: 714 727-4211, Fax: 714 727-4217.
http ://www.mic.org/
19 John Koupal. M6.ACE.001 "Air Conditioning Activity Effects in MOBILE6," EPA420-R-01-
054, November 2001. http://www.epa.gov/otaq/models/mobile6/r01054.pdf
20 Motorcycle Industry Council, "On-Highway Motorcycles 1998 Population Estimate."
November 21, 1999. Available in the U.S. EPA docket: A-2000-01, IIB-22.
21 Davis, Stacy C. and Lorena F. Truitt. "Investigation of Class 2b Trucks (Vehicles of 8,500 to
10,000 Ibs GVWR)," Oak Ridge National Laboratory. ORNL/TM-2002.49, March 2002.
99
U.S. Federal Transit Administration (FTA). "Study & Report to Congress: Applicability of
Maximum Axle Weight Limitations to Over-the-Road and Public Transit Buses,"
December 2003.
23
American Bus Association, July 2000.
9/1
Good Sam Club, "Highways Member Study 2000." TL Enterprises, Inc., Ventura, California.
(805) 667-4100.
25 Koupal, November 2001.
r\r
Electric Drive Association. See http://www.electricdrive.org/edtech/ele_ev_market.htm
27 Davis and Truitt, March 2002.
28 Davis and Truitt, March 2002.
95
-------
9Q
Union of Concerned Scientists, http://www.ucsusa.org/
30 U.S. Federal Transit Administration, December 2003.
31 U.S. Federal Transit Administration, December 2003.
32 Yuji Horie, Craig Tranby and Steven Sidawi, Valley Research Corporation. "On-Road Motor
Vehicle Activity Data: Volume I - Bus Population and Activity Pattern, Final Report."
Tables 3-9 & 2-2. Contract A132-182. Prepared for California Air Resources Board,
September 1994.
33 Brian, Mac. Recreational Vehicle Industry Association. Phone conversation, October 29,
2003.
34 USEPA Code of Federal Regualtions. (CFR) 40 section 86.529-78 and United Nations (UN)
"Worldwide Harmonised Motorcycle Emissions Certification Procedure", Informal
document No. 15, 46th GRPE, 19-23 May 2003, agenda item number 3.
http ://www. epa.gov/epahome/cfr40. htm
35 USEPA. "IM240 and Evap Technical Guidance," EPA420-R-00-007, April 2000. Online at
http://www.epa.gov/otaq/regs/im/r00007.pdf
36 Warila,J. "Derivation of Mean Energy Consumption Rates within the MOVES Modal
Framework", 14th Coordinating Research Council On-Road Vehicle Emissions Workshop
Poster Session, San Diego, California, March 29-31, 2004.
37 Petrushov, V.A., "Coast Down Method in Time-Distance Variables," SAE 970408, February
24, 1997. http://www.sae.org/
TO
Sierra Research, Inc. Memo from Tom Carlson to John Koupal, "Analysis of Rural Average
Speed Distributions for MOVES," Purchase Order EP05B00129, December 1, 2004.
39 Festin, Scott. "Summary of National and Regional Travel Trends: 1970-1995," Office of
Highway Information Management, Dept. of Transportation, May 1996. Online at
http://www.fhwa.dot.gov/ohim/bluebook.pdf
40 Sierra Research, Inc. M6.SPD.001 "Development of Speed Correction Cycles." EPA Contract
No. 68-C4-0056, Work Assignment 2-01, June 26, 1997. Online at
http://www.epa.gov/otaq/models/mobile6/m6spd001.pdf
41 USEPA Combined Federal Register. CFR 40, 86, Appendix I.
http ://www. epa.gov/epahome/cfr40. htm
42 Hart, Constance. "EPA's Onboard Analysis Shootout: Overview and Results." EPA420-R-02-
026, October 2002. Online at
http ://www. epa. gov/otaq/ngm. htm
96
-------
43 Eastern Research Group, Inc. (ERG), "Roadway-Specific Driving Schedules for Heavy-Duty
Vehicles." EPA Contract 68-C-OO-l 12, Work Assignment 3-07, August 15, 2003.
44 Eastern Research Group, August 2003.
45 Sierra Research, Inc. "Travel Trip Characteristics Analysis." EPA Contract 68-C1-0079, Work
Assignment 2-05, September 30, 1994.
46 Jackson, Tracie R. "Fleet Characterization Data for MOBILE6: Development and Use of Age
Distributions, Average Mileage Accumulation Rates and Projected Vehicle Counts for Use
in MOBILE6." M6.FLT.007, EPA420-R-01-047, September 2001. Online at
http://www.epa.gov/otaq/models/mobile6/r01047.pdf
47 Systems Application International, Inc. M6.SPD.003 "Development of Methodology for
Estimating VMT Weighting by Facility Type" EPA420-R-01-009, April 2001. Online at
http://www.epa.gov/otaq/models/mobile6/r01009.pdf
4R
Lutsey, Nicholas, Christie-Joy Brodrick, Daniel Sperling, and Carollyn Oglesby. "Heavy-Duty
Truck Idling Characteristics - Results from a Nationwide Truck Survey." Annual Meeting
of the Transportation Research Board, January 2004.
49 USEPA Combined Federal Register Vol. 65, No. 85, Tuesday, May 2, 2000. Proposed Rules,
49 CFR Parts 350, 390, 394, 395 and 398. Online at
http://www.fmcsa.dot.gov/Pdfs/050200p.pdf
50 Eastern Research Group, August 2003.
51 Pechan, E.H. & Associates, Inc. "Documentation for the Onroad National Emissions
Inventory (NET) For Base Years 1970-2002," prepared for EPA Office of Air Quality
Planning and Standards, January 2004.
(ftp://ftp. epa.gov/EmisInventory/fmalnei99ver3/haps/documentati on/onroad/nei_onroadja
n04.pdf).
52 U.S. Federal Highway Administration (FHA). Highway Performance Monitoring System
Field Manual. OMB No. 21250028, December, 2000. Online at
http://www.fhwa.dot.gov/ohim/hpmsmanl/hpms.htm
53 Jackson, September 2001.
54 .Pechan & Associates, Inc. October 2002.
55 U.S. Federal Highway Administration, December 2000.
56 Jackson, September 2001.
97
-------
57 Fleger, Stephen A., Robert P. Haas, Jeffrey W. Trombly, Rice H. Cross III, Juan E. Noltenius,
Kelley K. Pecheux, and Kathryn J. Chen. "Study of Adequacy of Commercial Truck
Parking Facilities." Table 7. U.S. Federal Highway Administration, FHWA-RD-01-158,
March 2002. Online at http://www.tfhrc. gov/safetv/pub s/0115 8/index.htm
58
Pechan & Associates, Inc. October 2002.
59 U.S. Federal Highway Administration, December 2000.
60 Jackson, September 2001.
61 Davis and Truitt, March 2002.
62 Koupal, November 2001.
63 Air Improvement Resources, Inc. (AIR), "Derivation of By-Month, By-County, By-Hour
Temperature and Relative Humidity with Monthly Data," EPA Order 4C-S082-NTSA,
December 8, 2004.
64 Eastern Research Group (ERG), "National Mobile Inventory Model (NMIM) Base and Future
Year County Database Documentation and Quality Assurance Procedures," EPA Contract
No. 68-COO-112, Work Assignment 3-05. April 15, 2003.
65 Hendren, T., D. Niemeier (2001). "State transportation expenditure reporting: Questions for
the 21st century, Public Works Management and Policy," 5(3): 179-198. Online at
http ://pwm. sagepub. com/cgi/reprint/5/3/179
66 Miller, T. L., W.T. Davis, G.D. Reed, P. Doraiswamy, A. Tang, and P. Sanhueza, "Corrections
to Mileage Accumulation Rates for Older Vehicles and The Effect on Air Pollution
emissions," Transportation Research Record, No. 1750, 80th Annual Meeting of
Transportation Research Board, National Research Council, National Academy Press,
Washington D.C., January 2001.
------- |