Draft MOVES2009 Highway Vehicle
Population and Activity Data
United States
Environmental Protection
Agency
-------
Draft MOVES2009 Highway Vehicle
Population and Activity Data
Assessment and Standards Division
Office of Transportation and Air Quality
U.S. Environmental Protection Agency
v>EPA
NOTICE
This technical report does not necessarily represent final EPA decisions or
positions. It is intended to present technical analysis of issues using data
that are currently available. The purpose in the release of such reports is to
facilitate the exchange of technical information and to inform the public of
technical developments.
United States EPA-420-P-09-001
Environmental Protection . ^ „„_
Agency August 2009
-------
Table of Contents
1. Introduction 1
2. Data Sources 4
2.1. VIUS(andTIUS) 4
2.2. Polk NVPP® and TIP® 4
2.3. FHWA Highway Statistics 4
2.4. FTA National Transit Database 4
2.5. School Bus Fleet Fact Book 4
2.6. MOBILE6 5
2.7. Annual Energy Outlook & National Energy Modeling System 5
2.8. Transportation Energy Data Book 5
2.9. Oak Ridge National Laboratory Light-duty Vehicle Database 5
3. SourceTypeYear 6
3.1. 1999 SourceTypePopulation 6
3.2. 1990 SourceTypePopulation 10
3.3. SalesGrowthFactor 14
3.3. MigrationRate 16
4. SourceTypeModelYear 17
5. SourceTypeAge 18
5.1. SurvivalRate 18
5.2. Relative MAR 20
5.3. FunctioningACFraction 23
6. SourceTypeAgeDistribution 25
6.1. 1999 Motorcycles 25
6.2. 1999 Passenger Cars 25
6.3. 1999 Trucks 26
6.4. 1999 Intercity Buses 28
6.5. 1999 School Buses and Motor Homes 28
6.6. 1999 Transit Buses 28
6.7. 1990 Motorcycles 30
6.8. 1990 Passenger Cars 30
6.9. 1990 Trucks 30
6.10. 1990 Intercity Buses 31
6.11. 1990 School Buses and Motor Homes 31
6.12. 1990 Transit Buses 31
7. SourceBinDistribution 32
7.1. Motorcycles 34
7.2. Passenger Cars 35
7.3. Trucks 36
7.4. Buses 46
7.5. Refuse Trucks 50
7.6. Motor Homes 52
7.7. SourceBinDistributions for 2000-and-later 55
8. SourceUseType 62
8.1. SourceMass 62
-------
8.2. Road Load Coefficients 63
9. RoadTypeDistribution 66
10. Average Speed Distribution 67
ll.HPMSVTypeYear 70
H.l.HPMSBaseYearVMT 70
11.2. BaseYearOffNetVMT 70
11.3. VMTGrowthFactor 70
12. Temporal Distributions of VMT 74
12.1. MonthVMTFraction 74
12.2. DayVMTFraction 75
12.3.HourVMTFraction 75
13. Driving Schedule Tables 77
14. Drive Schedule Association 79
15. SourceTypeHour 83
IS.l.IdleSHOFactor 83
16. ZoneRoadType 85
17. Zone 86
17.1. StartAllocFactor 86
17.2. IdleAllocFactor 86
18. SCC Mappings 88
18.1. SCCVtypeDistribution 88
18.2. SCCRoadTypeDistribution 89
19. MonthGroupHour 91
20. Sample Trip Data 92
21. References 94
11
-------
List of Tables and Figures
Table 1-1. MOVES SourceTypes 1
Table 1-2. MOVES Database Elements Covered in This Report 2
Table 3-1. Vehicle Population Comparisons 1999 7
Table 3-2. Adjusted Vehicle Populations 7
Table 3-3a. VIUS 1997 Codes Used for Distinguishing Truck SourceTypes 8
Table 3-3b. VIUS 2002 Codes Used for Distinguishing Truck SourceTypes 8
Table 3-4. 1999 Truck Source Type Distribution and Populations 9
Table 3-5. 1999 Bus Population Comparisons 9
Table 3-6. 1999 SourceType Populations in Draft MOVES2009 10
Table 3-7. 1990 Vehicle Population Comparisons 11
Table 3-8. TIUS92 Codes Used for Distinguishing Truck SourceTypes 12
Table 3-9. 1990 Truck SourceType Distribution and Populations 12
Table 3-10. 1990 Bus Population Comparisons 13
Table 3-11. 1990 SourceType Populations in Draft MOVES2009 14
Table 3-12. SalesGrowthFactor by Calendar Year and Source Type 16
Table 4.1. AC Penetration Fractions in Draft MOVES2009 17
Table 5-1. SurvivalRate by Age and SourceType 20
Table 5.2. Equations for Calculating Annual Mileage Accumulation Rates used in MOVES 22
Figure 5.1. Relative Mileage Accumulation Rates in Draft MOVES2009 23
Table 5 -3. FunctioningACFraction by Age (All Use Types Except Motorcycles) 24
Figure 6.1 1999 Age Distributions for Passenger 26
Figure 6.2 1999 Age Distributions for Passenger and Light Commercial Trucks 27
Table 6-1. 1999 Age Fractions for MOVES Source Types 29
Table 7-1. Data Tables Used by SourceBinGenerator 33
Table 7-2. Motorcycle Engine Size and Average Weight Distributions for Selected Model Years 34
Table 7-3. Mapping Polk Fuel Codes to MOVES 35
Table 7-4. Mapping VIUS ENGTYP to MOVES FuelTypelD 36
Table 7-5. Diesel Fractions for Trucks 38
Table 7-6. Mapping VIUS Engine Size Categories to MOVES EngSizelD 39
Table 7-7. Mapping VIUS Average Weight to MOVES WeightClassID 42
Table 7.8. Light Truck Class 2 Weight Distribution 43
Table 7-9. Fraction of Light-Duty Trucks among Gasoline-Fueled Trucks 44
Table 7-10. Fraction of Light-Duty Trucks among Diesel-fueled Trucks 45
Table 7-11. Mapping National Transit Database Fuel Types to MOVES Fuel Types 46
Table 7-12. Fuel Fractions for Transit Buses 47
Table 7-13. Fuel Fractions for School Buses 48
Table 7-14. FTA Estimate of Bus Weights 48
Table 7-15. California School Buses 49
Table 7-16. Weight Distributions for Buses by Fuel Type 50
Table 7-17. Fuel Fractions for Refuse Trucks by Model Year 51
Table 7-18. Refuse Truck Size Weight Fractions by Fuel Type 52
Table 7-19. Diesel Fractions for Motor Homes 53
Table 7-20. Weight Fractions for Diesel Motor Homes by Model Year 54
Table 7-21. Weight Fractions for Gasoline Motor Homes by Model Year 55
Table 7-22. Supported Fuels and Technologies for 2000-and-later Model Years 56
Table 7.23. Fuel Fractions for 2002 and Newer Passenger Cars and Light Duty Trucks 58
Table 7.24. Fuel and Engine Technology Fractions for 2000-and-later Buses 58
111
-------
Table 7.25. Fuel and Engine Technology Fractions for 2002 and Newer Motor Homes and Single-Unit
Short-haul and Long-haul Trucks 59
Table 7.26. Fuel and Engine Technology Fractions for Refuse Trucks and Short-haul and Long-haul
Combination Trucks 61
Table 8-1. MOVES Weight Classes 63
Table 8-2. Road Load Coefficients for Heavy-Duty Trucks, Buses, and Motor Homes 64
Table 8-3. SourceUseType Characteristics 65
Table 9-1. Road Type Codes in MOVES 66
Table 9-2. Roadtype Distributions by Sourcetype 67
Table 10-2. MOVES Speed Bin Categories 68
Figure 10.1 Speed Distribution by Roadtype 69
Table 11-1. 1999 VMT by HPMS Vehicle Class 70
Table 11-2. VMTGrowthFactor Calculation for Passenger Cars and Light Trucks 72
Table 11-3. VMT Growth Factors in Draft MOVES2009 73
Table 12-1. MonthVMTFraction 74
Table 12-2. DayVMTFractions 75
Figure 12.1 Hourly VMT Fractions in Draft MOVES2009 76
Table 13-1. Default MOVES Drive Schedules 78
Table 14-1. Drive Schedule Mapping 79
Table 14.2 Proposed Drive Schedules for Passenger Cars, Passenger Trucks and Light Commercial
Trucks in Final MOVES2009 81
Table 14.2 Continued 82
Table 14.2 Continued 82
Table 18-1. SCC Mappings for Selected SourceTypes 89
Table 18-1. SCC RoadTypes 89
Table 19-1. Air Conditioning Activity Coefficients 91
Figure 19-1: Air Conditioning Activity Demand as a Function of Heat Index 91
Table 20.1. Source Data for Sample Vehicle Trip Information 92
Table 20.2. Synthesis of Sample Vehicles for Source Types Lacking Data 93
Table 20.2. Starts per Day by SourceType 93
IV
-------
1. Introduction
The Environmental Protection Agency's MOVES (Motor Vehicle Emission Simulator)
is a new set of modeling tools for estimating emissions produced by on-road (cars, trucks,
motorcycles, etc.) and nonroad (backhoes, lawnmowers, etc) mobile sources. This report
partially documents the Draft MOVES2009 version, released in April 2009. Draft MOVES2009
estimates greenhouse gases (GHG), criteria pollutants and selected air toxics from highway
vehicles. When finalized, MOVES2009 will serve as a replacment for MOBILE6.2
The primary vehicle classification in MOVES is "SourceType." (Also sometimes called
"SourceUseType").The MOVES SourceTypes are listed in Table 1-1, along with the associated
DOT Highway Performance Monitoring System (HPMS) vehicle classes.
To estimate emissions accurately, we must use accurate estimates of vehicle populations
and activity. This paper documents the sources and calculations used to produce the default
population and activity data in the DRAFT MOVES2009 database used to compute national
level emissions.a In particular, this paper will describe the data used to fill the tables and fields
listed 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,
Other Two-Axle/Four Tire,
Single Unit
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.
a For many uses, local inputs are required. EPA is currently developing draft technical guidance to describe these
requirements.
-------
Table 1-2. MOVES Database Elements Covered in This Report
Database Table Name*
SourceTypeYear
SourceTypeModelYear
SourceTypeAge
S ourceTy pe AgeDi stributi on
SourceBinDistribution*
SourceUseType
RoadTypeDistribution
AvgSpeedDistribution
HPMSVtypeYear
Month VMTFraction
DayVMTFraction
Hour VMTFraction
Drive Schedule
DriveScheduleSecond
Drive S chedul e As soci ati on
SourceTypeHour
ZoneRoadType
Zone
SCCVTypeDistribution
Fields
sourceTypePopulation
sal esGrowthF actor
migrationRate
ACPenetrati onFracti on
survivalRate
relativeMAR
functi oning ACFracti on
ageFraction
sourceBinActivityFraction
rollingTerm
rotatingTerm
dragTerm
sourceMass
roadTypeVMTFraction
avgSpeedFraction
HPMSBaseYearVMT
baseYearOffNetVMT
VMTGrowthF actor
month VMTFraction
day VMTFracti on
hour VMTFraction
average Speed
speed
sourceTypelD
roadTypelD
driveS chedul elD
isRamp
idleSHOFactor
SHOAllocFactor
idle All ocF actor
startAllocFactor
SHPAllocFactor
SCCVTypeFraction
-------
MonthGroupHour
SampleVehicleDay
SampleVehicleTrip
S ampl e Vehi cl ePopul ati on
AC Activity Terms (A, B &
C)
daylD
sourceTypelD
priorTripID
keyontime
keyOffTime
stmyFuelEngFraction
stmyFraction
*See also Table 7-1, listing tables and fields used by the SourceBinGenerator.
-------
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. VIUS(and TIUS)
Until 2002, the U.S. Census Bureau conducted the Vehicle Inventory and Use Survey
(VIUS)1 to collect data on the physical characteristics and activity of U.S. trucks every five
years. The survey is a sample of private and commercial trucks that were registered in the U.S.
as of July of the survey year. The survey excludes automobiles, motorcycles, government-owed
vehicles, ambulances, buses, motor homes and nonroad equipment. For MOVES, VIUS
provides information to characterize trucks by SourceType and to estimate age distributions.
Draft MOVES2009 uses data from both the 1997 and 20022 surveys. Before 1997, VIUS was
known as TIUS (Truck Inventory and Use Survey). To populate the 1990 base year, we used data
from the 1992 TIUS.3.
Note that Census Bureau has discontinued the VIUS survey. We request comments on
alternate data sources or approaches for determining truck populations in the future.
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®)4 and the Trucking Industry Profile (TIP®Net) Vehicles in Operation database.5 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 Draft MOVES2009, EPA used 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. 6 7 8 9 Hereafter,
references will be to FHWA MV-1, MV-10, VM-1, and VM-2. For the 1999 base year, we used
the 1999 statistics; for the 1990 base year, we used 1990 numbers.
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).10 For Draft MOVES2009,
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.u 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."12 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) 13'14 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. For Draft MOVES2009 we
used AEO2006 to forecast VMT growth and vehicle sales growth. For the final MOVES2009, we
propose updating these results with more recent forecasts.
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 200215 and Edition
23, published in October 2003.16 For Draft MOVES2009 we updated sales growth based on
Edition 27, published in 2008.17 and updated 1990 values using Edition 13, published in 1993.
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 - 2001.18 We used this database to determine weight
distributions for light trucks by model year.
-------
3. SourceTypeYear
The SourceTypeYear table stores three data fields—SourceTypePopulation,
SalesGrowthFactor, and Migration Rate. Each field is described below in terms of what
information it contains, the sources of the data used for the field, and, when applicable, certain
data points used in determining the field parameters.
3.1. 1999 SourceTypePopulation
The SourceTypePopulation field stores the total population of vehicles by SourceType
for a given base year and domain. For Draft MOVES2009, this is the entire United States in
1999. An additional base year is 1990. Some of the values are taken directly from the indicated
sources; other values needed to be derived from the available data.
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 (TEDB) 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.19 Finally, our 1999 Polk data set includes school buses and motor homes (which
can be counted separately), but does not include "non-school buses."
Motorcycle population estimates were available from both FHWA registration data and
from the Motorcycle Industry Council. The MIC estimate is based on 1998 sales estimates,
adjusted to subtract noped sales (nopeds are similar to mopeds, but lack pedals) and to account
for scrappage.
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 motorcycles, personal trucks and
automobiles.20 Data from the 2001 survey 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 five data sources are compared in Table 3-1.
Table 3-1. Vehicle Population Comparisons 1999
Data Source
FH W A MV-l(w Puerto
Rico and publically
ownedvehicles)
FHWAMV-10(w/o
Puerto Rico and
publically owned
vehicles)
Polk NVPP® & TIP®
NHTS (2001)
MIC (1998)21
Motorcycles
4,173,869
na
4,951,747
4,605,439
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 public vehicles and Puerto Pvican vehicles.
Adjusted Population = FHWA w public & PR * (Polk/FHWA w/0 public & PR)
This leads to the values in Table 3-2.b
Table 3-2. Adjusted Vehicle Populations
Automobiles
Trucks (total)
Population
(Draft MOVES2009)
130,163,354
83,348,540
Population
(proposed for final MOVES2009)
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 VIUS responses for Axle Arrangement, Primary
Area of Operation, Body Type and Major Use as detailed in Table 3-3a and Table 3.3b.
With these definitions and with vehicles that lack APxEAOP codes assigned
proportionally to the corresponding SourceTypes, we computed the distributions in Table 3-4.
b There was an error in the calculation of the value for total trucks used in Draft MOVES2009. We plan to correct
this error in the final version of MOVES2009 as indicated here.
7
-------
These distributions were multiplied by the total truck population from Table 3-2 to derive
population values for MOVES.
Table 3-3a. VIUS 1997 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
any
any
Major Use
personal
transportation
(MAJUSE=20)
any but personal
transportation
Any
Any
Any
Any
Any
Table 3-3b. VIUS 2002 Codes Used for Distinguishing Truck SourceT\
SourceType
Passenger
Trucks
Light
Commerical
Trucks
Refuse
Trucks
Single Unit
Short-Haul
Trucks
Single Unit
Long-Haul
Trucks
Combination
Short-Haul
Trucks
Combination
Long_Haul
Trucks
Axle Arrangement
axle_config in (1,6,7,8)
axle_config in (1,6,7,8)
axle config in
(2,3,4,5,9,10,11,12,13,14,15,16,17,18,19,20)
axle config in
(2,3,4,5,9,10,11,12,13,14,15,16,17,18,19,20)
axle config in
(2,3,4,5,9,10,11,12,13,14,15,16,17,18,19,20)
axle_config>=21
axle config>=21
Primary Area
of Operation
any
any
trip_primary in
(1,2,3,4)
trip_primary in
(1,2,3,4)
trip_primary in
(5,6) Long
Range
trip_primary in
(1,2,3,4)
trip_primary in
(5,6) Long
Range
Body Type
any
any
bodytype=21
Any except
bodytype=21
any
sample_strata=5
Combination
Trucks
sample_strata=5
Combination
Trucks
/pes.
Operator
Classification
opclass=5
opclass<>5
any
any
any
any
any
8
-------
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
(Draft MOVES2009)
57,424,819
19,184,642
88,970
4,489,140
265,520
1,088,815
806,633
83,348,540
Population
(final
MOVES2009)
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
2000".22 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. 1999 Bus Population Comparisons
Data Source
FHWAMV-1
FHWA MV- 10
(excludes PR)
FTA NTD
APTA23 ***
Polk TIP®
School Bus Fleet Fact
Book
Motorcoach
Census**
Total Buses
732,189
728,777
Intercity Buses
44,200
Transit Buses
55,706
75,087
School Buses
592,029*
460,178
429,086
* Includes some church & industrial buses.
* * Includes Canada.
*** Includes trolleybuses.
As Table 3-5 shows, estimates of bus populations 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. We request comment on ways to improve our national bus
population estimates.
For motorcycles we used the 1999 FHWA value from table MV-1. For comparison,
Table 3-1 also shows the 1998 population as estimated by the Motorcycle Industry Council based
on sales and estimated scrappage rates, and the 2001 population estimated by the 2001 NHTS.
The FHWA population estimates are noticeably lower than the other estimates. If time and
resources allow, EPA may investigate this further for future versions of the MOVES model.
For motor homes we used the population from the Polk TIP® database. In Table 3-1, this
value is compared to the estimate from the 2001 NHTS. As for motorcycles, the FHWA
registration count is noticeably lower than the household survey estimate. This could reflect
population growth in the years between the estimates, or it may reflect difference in the way
motor homes are defined in the two studies, or be an artifact of the method used to extrapolate
from the NHTS sample to the national population estimate. If time and resources allow, EPA
may investigate this further for future versions of the MOVES model.
Table 3-6 summarizes the 1999 vehicle populations used in Draft MOVES2009.
Table 3-6. 1999 SourceType Populations in Draft MOVES2009
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
57,424,800
19,184,600
84,454
55,706
592,029
88,607
4,489,140
265,520
902,949
1,088,820
806,633
3.2. 1990 SourceTypePopulation
Because SIPs require estimates of 1990 emissions, the MOVES database includes a 1990
base year. The SourceTypePopulation inputs for 1990 were developed using methods and data
similar to those used for 1999.
The primary sources for calendar year 1990 vehicle population data are the FHWA
Highway Statistics Tables MV-200, VM- 201 A, MV-10 and the Polk NVPP® databases. As
in 1999, the FHWA and Polk data differ in how vehicles are counted. (See previous section.)
Additionally, the 1990 Polk data does not include buses and motor homes. The National
Personal Transportation Survey includes data on personal trucks, automobiles and motorcycles.
10
-------
Data on motorcycles were also obtained from the Motorcycle Statistical Annual published by the
Motorcycle Industry Council. Values from all four sources are compared in Table 3-1.
Registration data on vehicles registered in Puerto Rico for year 1990 was obtained from
FHWA's Highway Statistics 1990.
Table 3-7. 1990 Vehicle Population Comparisons
Data Source
FHWA(w/ Puerto
Rico and Publicly
owned vehicles)1
FHWA (w/o Puerto
Rico and w/ Publicly
owned vehicles)2
PolkNVPP®
NPTS (1990)4
Motorcycle Industry
Council5
Motorcycles
4,278,286
4,259,461
na
2,089,523
4,310,000
Automobiles
135,022,124
133,700,497
123,276,600
120,712,000
na
Trucks (total)
54,673,458
54,470,430
56,023,0003
37,110,000
na
Buses (total)
629,943
626,987
na
na
na
Motor
Homes
na
na
na
821,000
na
1 Data on Puerto Rico was obtained from Highway Statistics 1990, published by the FHWA.
2 For these numbers, used data from FHWA Highway Statistics TableVM-201A, April 1997 and Table MV-200
(state motor vehicle registrations, by years 1990-1995).
3 As published in TEDB edition 23. Does not include Puerto Rico and publicly -owned vehicles.
41990 NPTS special report on travel modes- Chapters, the demography of the US Vehicle Fleet. The motorcycle
number is calculated using the appendix table and the proportion of MCs from Table 20 of the 2001 NHTS
Summary of Travel Trends.
5 The Motorcycle number was obtained as a sum of on-highway and dual motorcycles for year
1990 as published in the 1999 Motorcycle Statistical Annual.
For MOVES, total trucks are sub-classified into seven SourceTypes. The proportion of
total trucks in each subtype was estimated using TIUS92 responses for Axle Arrangement,
Primary Area of Operation, Body Type and Major Use as detailed in Table 3-8.
With these definitions and with vehicles that lack AREAOP codes assigned
proportionally to the corresponding SourceTypes, we computed the distributions in Table 3-9.
These distributions were multiplied by the total truck population from Table 3-7 to derive
population values for MOVES.
11
-------
Table 3-8. TIUS92 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
Any
Any
Major Use
personal
transportation
(MAJUSE=20)
any but personal
transportation
any
any
any
any
any
Table 3-9. 1990 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
67.32%
24.07%
0.11%
6.12%
0.23%
1.35%
0.81%
100.00%
Population
37,713,840
13,483,198
59,037
3,426,459
128,776
758,091
453,599
56,023,000
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 American Public Transit
Association (APIA) Fact Book, the School Bus Fleet Fact Book, and the Transportation Energy
Data Book.
12
-------
Table 3-10. 1990 Bus Population Comparisons
Data Source
FHWA
(w/o PR and with
Publicly-owned
Vehicles)*
FHWA (w/o PR
and w/o Publicly-
owned Vehicles)
APTA 1991
Transit Fact Book
TEDB**
School Bus Fleet
Fact Book***
Total Buses
626,9871
275,4931
Intercity Buses
20,6802
58,141
Transit Buses
60,585
59,753
School Buses
545,7223
508,261
391,714
FHWA Highway Statistics, Summary to 1995, Table MV-200
** Transportation Energy Data Book : Edition 13, March 1993, Table 3.29. 1990 buses. "Intercity Buses" is sum
of "Intercity Bus" and "Other;" "School Buses" includes other non-revenue buses.
*** Average of school years 1989-90 and 1990 -91, School Bus Fleet Fact Books 1990 and 1991.
Table 3-1 1 summarizes the 1990 vehicle populations used in Draft MOVES2009. For
motor homes we used the only available data from NPTS. We used the TEDB data for buses.
For trucks the TIUS data was used; the remaining values were based on FHWA data.
-------
Table 3-11. 1990 SourceType Populations in Draft MOVES2009
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
1990
Population
4,278,286
135,022,124
37,713,840
13,483,198
58,141
59,753
508,261
59,037
3,426,459
128,776
821,000
758,091
453,599
3.3. 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- in Draft MOVES2009, years 2000 through 2050.
Note that the sales growth factors are not used in the calculation of county-level or
project level emissions, where users must input local vehicle populations for each year that is
modeled. For MOVES2004, 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. For Draft MOVES2009, the sales data for passenger cars
and light trucks were updated to account for actual sales data and updated sales forecasts, but
rates for the remaining sourcetypes were not changed. Beyond 2030, the SalesGrowthFactor was
set to the 2030 value. For the final MOVES2009, we anticipate updating sales information, at
least for the dominant sourcetypes.
The data sources and methodologies by source type are described below:
-------
Passenger Cars and Passenger Trucks: SalesGrowthFactors for calendar year 2000
through 2005 were derived from total sales numbers reported in the TEDB26 Table
4.5. Factors for calendar years 2006 through 2030 were derived from new car sales
estimates presented in AEO2006 Supplemental Table 45, generated by NEMS.
Motorcycles: SalesGrowthFactors for calendar year 2000 and 2001 were computed
from sales values in the Motorcycle Industry Council Statistical Annual.24
SalesGrowthFactors for years 2006 through 2030 were set equal to passenger car
growth factors.
Commercial Trucks: SalesGrowthFactors for calendar year 2000 through 2005 were
derived from total light truck sales numbers reported in the TEDB26 Table 4.6.
Factors for Calendar year 2002 through 2030 differ from passenger trucks. It is
possible that they were mistakenly retained from an earlier version of the model. We
plan to investigate this further for the final MOVES2009. .
Buses, Single Unit Trucks & Motor Homes: Calendar years 2000-2001 were based on
sales as reported in TEDB23 Table 5.3 (gross weight range 10,000-33,000 Ibs).
Years 2004 through 2030 were calculated from medium-duty truck sales projections
from AEO2006Supp\Qmenta\ Table 55.
Combination Trucks, Refuse Trucks: Calendar years 2000-2001 were based on sales
as reported in TEDB23 Table 5.3 (gross weight range 33,001 and greater pounds).
Years 2004 through 2030 were calculated from heavy-duty truck sales projections
found in AEO2006 Supplemental Table 55.
The resulting SalesGrowthFactors by source type are shown in Table 3-12:
-------
Table 3-12. SalesGrowthFactor by Calendar Year and Source 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
2026
2027
2028
2029
2030+
Motorcycles
1.017
0.952
0.970
1.015
1.013
1.039
1.059
0.997
0.987
0.985
0.980
1.005
0.996
0.991
0.989
0.994
.001
.002
.005
.004
.007
.007
.009
.009
.009
.008
.010
.008
.007
.008
1.008
Passenger
Cars
1.017
0.952
0.962
0.939
0.986
1.021
1.059
0.997
0.987
0.985
0.980
1.005
0.996
0.991
0.989
0.994
.001
.002
.005
.004
.007
.007
.009
.009
.009
.008
.010
.008
.007
.008
.008
Passenger
Trucks
1.039
1.037
1.001
1.026
1.047
0.991
0.905
1.059
1.031
1.043
1.042
1.016
1.017
1.011
1.015
1.008
1.012
1.017
1.019
1.015
1.013
1.018
1.019
1.021
1.021
1.020
1.021
1.020
1.016
1.018
1.017
Light
Comm.
Trucks
1.039
1.037
1.001
1.026
1.047
0.991
0.998
1.047
1.007
1.039
0.994
1.015
0.983
0.996
1.011
1.019
1.022
1.016
1.015
1.009
1.011
1.012
1.015
1.015
1.016
1.015
1.016
1.015
1.012
1.014
1.013
Buses,
Single Unit
Trucks &
Motor
Homes
0.963
0.850
0.882
1.067
1.170
1.082
1.001
1.001
1.003
1.026
0.992
0.997
0.986
1.000
1.029
1.035
1.025
1.015
1.010
0.995
0.997
1.006
1.012
1.015
1.018
1.018
1.016
1.012
1.006
1.010
1.013
Combination
Trucks
0.809
0.660
0.923
1.042
1.310
1.130
1.010
0.940
0.990
.000
.000
0.990
.000
.010
.020
.020
.020
.020
.000
0.980
0.980
.000
.010
.010
.020
.020
.020
.010
.000
.010
.010
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 may be useful when modeling emissions on relatively small geographic scale.
For the default MOVES database, 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 ACPenetrationFraction, 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.
Default values in Draft MOVES2009 were taken from MOBILE6. 25 Market penetration
data by model year were gathered from Ward's Automotive Handbook for light-duty vehicles
and light-duty trucks for model years 1972 through the 1995 for cars and 1975-1995 for light
trucks. Rates in the first few years of available data are quite variable, so values for early model
years were estimated by applying the 1972 and 1975 rates for cars and trucks, respectively.
Projections beyond 1995 were developed by calculating the average yearly rate of increase in the
last five years of data and applying this rate 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, more likely on trucks than cars. For MOVES, the light-
duty vehicle rates were applied to passenger cars, and the light-duty truck rates were applied to
all other sourcetypes (except motorcycles, for which AC penetration is assumed to be zero).
Table 4.1. AC Penetration Fractions in Draft MOVES2009
1972-and-earlier
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+
Motorcycles
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
Passenger Cars
0.592
0.726
0.616
0.631
0.671
0.720
0.719
0.694
0.624
0.667
0.699
0.737
0.776
0.796
0.800
0.755
0.793
0.762
0.862
0.869
0.882
0.897
0.922
0.934
0.948
0.963
0.977
0.980
All Trucks and Buses
0.287
0.287
0.287
0.287
0.311
0.351
0.385
0.366
0.348
0.390
0.449
0.464
0.521
0.532
0.544
0.588
0.640
0.719
0.764
0.771
0.811
0.837
0.848
0.882
0.906
0.929
0.950
0.950
17
-------
5. SourceTypeAge
Three fields comprise SourceTypeAge in Draft MOVES2009: 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)
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. In MOVES, a separate SurvivalRate is applied to each age in each SourceType fleet.
These SurvivalRates in MOVES are used for all model years in a SourceType in all calendar
years.
SurvivalRates for Motorcycles were calculated based on regression of data provided by
the Motorcycle Industry Council (MIC).26
Survival rates for Passenger Cars, Passenger Trucks and Light Commercial Trucks came
from NHTSA's survivability Table 3 and Table 4.27 These survival rates are based on a detailed
analysis of Polk vehicle registration data from 1977 to 2002. We modified these rates to fit
them into the MOVES format:
• NHTSA rates for Light Trucks were used for both MOVES Passenger Trucks
and MOVES Light Commercial Trucks.
• MOVES calculates emissions to age 30 for both cars and trucks, but NHSTA car
rates were available only to age 25, so we extrapolated car rates to age 30 using
the estimated survival rate equation in section 3.1 of the NHTSA report.
• According to the NHTSA methodology, NHTSA "age= 1" corresponds to
MOVES "ageid =2," so the survival fractions were shifted accordingly.
• Because MOVES requires survival rates for MOVES ages < 2, the survival rates
for age 0 and age 1 were interpolated using a linear interpolation and assuming
that the survival rate prior to age 0 is 1.
• NHTSA defines survival rate as the ratio of the number of vehicles remaining in
the fleet at a given year as compared to a base-line year. MOVES calculations
require a value that is the ratio of a given year to the previous year, so we
transformed the NHTSA rates to MOVES rates using this ratio.
• Because MOVES ageid= 30 is intended to represent all ages 30-and-greater, the
survival rate for ageid=30 was set to 0.3.
• Quantitatively the formula used to derive the MOVES Survival rates was:
MOVES Survival Rate (ageid =0) = 1 - (1-NHTSA Survival Rate (age =2)/3)
MOVES Survival Rate (ageid =!) = !- (1- 2* NHTSA Survival Rate (age =2)/3)
18
-------
MOVES Survival Rate (age = 2 through 29) =
NHTSA Survival Rate (age = n-1)/ NHTSA Survival Rate (age = n-2)
MOVES Survival Rate (age = 30) = 0.3
The data for all other SourceTypes came from the Transportation Energy Data Book
(TEDB22, unchanged for version 23). We used the Heavy-Duty rates for the 1980 model year
(TEDB22, Table 6.11, same as TEDB26 Table 3.10). 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.
The TEDB survival rates were transformed into MOVES format in the same way as the
NHTSA rates. Survival rates for all "age 30" sourcetypes0 were set to 0.3. This is assumed to be
the fraction of all vehicles 30-and-older that survive an additional year.
SurvivalRates used in Draft MOVES2009 are shown in Table 5-1.
°Except motorcycles. See note below Table 5-1.
19
-------
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
0.990
0.990
0.980
0.970
0.960
0.960
0.950
0.940
0.930
0.920
0.920
0.910
0.900
0.890
0.890
0.880
0.870
0.860
0.850
0.850
0.840
0.830
0.820
0.820
0.810
0.800
0.790
0.780
0.780
0.770
0.760*
Passenger Cars
0.997
0.997
0.997
0.993
0.990
0.986
0.981
0.976
0.971
0.965
0.959
0.953
0.912
0.854
0.832
0.813
0.799
0.787
0.779
0.772
0.767
0.763
0.760
0.757
0.757
0.754
0.754
0.567
0.752
0.752
0.300
Passenger Trucks
Light Comm. Trucks
0.991
0.991
0.991
0.986
0.981
0.976
0.970
0.964
0.958
0.952
0.946
0.940
0.935
0.929
0.913
0.908
0.903
0.898
0.894
0.891
0.888
0.885
0.883
0.880
0.879
0.877
0.875
0.875
0.873
0.872
0.300
All Other
SourceTypes
1.000
1.000
1.000
1.000
0.990
0.980
0.980
0.970
0.970
0.970
0.960
0.960
0.950
0.950
0.950
0.940
0.940
0.930
0.930
0.920
0.920
0.920
0.910
0.910
0.910
0.900
0.900
0.900
0.890
0.890
0.300
* In draft MOVES2009, we neglected to set the age 30 motorcycle survival rate to 0.30. We
plan to fix this in the final MOVES2009.
We request comment on the survival rates used in MOVES and the possibility of
updating them.
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. This allows MOVES to maintain a constant MAR ratio
between ages and between the sourcetypes that make up each HPMS vehicle type even as
vehicle populations and the total VMT for an HPMS vehicle class changes over time. Table 1-2
(previous) lists the groupings of the MOVES SourceTypes within the six HPMS Vehicle Classes.
The following discussion refers to the Source Type ID numbers found in this table.
For many SourceTypes, the annual MARs were derived from the MARs developed for
MOBILE6. These were mapped from the MOBILE6 Vehicle Classes to the MOVES
20
-------
SourceTypes. We then used regression to smooth these initial MARs and to extend the MARs
from 25 to 30 ages.
5.2.1. Motorcycles
The MARs for motorcycles (category 11) were set to equal those in MOBILE6.
5.2.2. Passenger Cars, Passenger Trucks and Light Commerical Trucks
The MARs for passenger cars, passenger trucks and light commercial trucks (categories
21, 31 & 32) were taken from the NHTSA report on survivability and mileage schedules.28 In
the NHTSA analysis, annual mileage by age was determined for cars and for trucks using data
from the National Household Travel Survey. In this NHTSA analysis, vehicles that were less
than one year old at the time of the survey were classified as "age 1", etc. NHTSA used cubic
regression to smooth the VMT by age estimates.
We used NHTSA's regression coefficients to extrapolate mileage to ages not covered by
the report. We divided each age's mileage by the NHTSA "age 1" mileage to determine a
relative MAR. For consistency with MOVES age categories, we then shifted the relative MARs
such that the NHTSA "agel" ratio was used for MOVES age 0, etc. We used NHTSA's light
truck VMT to determine relative MARS for both passenger trucks and light commercial trucks.
5.2.3. Heavy Trucks
The initial MARs for truck categories 51, 52, 53, 61, and 62 in MOVES were calculated
based on weighting fractions assigned to the MOBILE6 truck classes. We used VIUS 1997
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.29 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 (HDDBT). This mileage data was obtained from the 1994 Federal
Transportation Administration survey of transit agencies. 30
For Intercity Buses (category 41), the initial MARs were taken from Motorcoach Census
2000.31 The data did not distinguish vehicle age, so the same MAR was used for each age. This
MAR is high compared to transit and school buses. We are not sure if this simply reflects the
very different type of driving done by these buses, or if it indicates a problem . We welcome
comments with ideas for validating or improving this estimate.
21
-------
5.2.4. Motor Homes
For motor homes (category 54), the initial MARs were taken from an independent
research study32 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 for motorcycles, trucks and buses.
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.
Table 5.2. Equations for Calculating Annual Mileage Accumulation Rates used in
MOVES
MOVES Source Type
Motorcycles
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
51
52
53
54
41
42
43
61
62
Regression Equation
na
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*
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. For example, all of the bus values are relative to each other. The relative MARs
for all sourcetypes are illustrated in Figure 5.1
22
-------
Figure 5.1. Relative Mileage Accumulation Rates in Draft MOVES2009
1.2
Relative Mileage Accumulation Rates
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
"TV
•11 • 21 31 32-*-41-»-42 i 43 51 52 53
54 61 62
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 MOBILE6.33 These
were applied to all source types except motorcycles, which were assigned a value of zero for all
years.
-------
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
24
-------
6. SourceTypeAgeDistribution
The age distribution of for each sourcetype is stored in the SoruceTypeAgeDistribution
table. Because sales are not constant, these distributions vary by calendar year. MOVES uses
age distributions for the base year combined with sales and scrappage information to compute
the age distribution in the calendar year selected for analysis.
This section first describes how the age distributions were determined for the primary
default base year of 1999, and then for the 1990 base year. Age distributions for the 1999 base
year are summarized in table 6-1. Age distributions for the 1990 base year are available in the
SourceTypeAgeDistribution table.
6.1. 1999 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. 1999 Passenger Cars
We considered three approaches to determine age fractions for passenger cars.
Our original approach (used for MOVES2004 and MOVES Demo) began with Polk
NVPP® 1999 data on car registration by model year. 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. However, because this method counts both
Model Year 1999 and Model Year 2000 as "Age 0", the Age 0 age fraction is inflated. When the
MOVES Total Activity Generator applies growth factors, the number of cars in future years is
inflated, and the fraction of passenger cars compared to other source types is skewed. Thus, we
rejected this approach.
A second approach was similar to the first, but with only Model Year 1999 vehicles
counted as "Age 0" in 1999.
Our third approach used passenger car sales data from Table 4.5 of the TEDB34 and
applied the NHTSA survival fractions, extrapolated to age 30 and shifted such that NHTSA age
n = MOVES age n+1. Survival fractions for MOVES age 0 and 1 were interpolated as described
in Section 5.1.
Not surprisingly, the age distributions resulting from the three approaches are very
similar, as illustrated in Figure 6.1. All show a fairly flat age distribution in the first eleven years
followed by a steep decline and a leveling off. The third approach provides a slightly more
generic age distribution than the second approach because the direct Polk data approach is for a
single year and the NHTSA survival fractions were derived by regression through many years of
data. For the Draft MOVES2009 default database, we selected the age distributions generated
with the third approach. For future versions of MOVES, we are considering updating these
values to better account for more recent data.
-------
Figure 6.1 1999 Age Distributions for Passenger
Passenger Car Age Distribution
-*— Original Polk
-B--Fblk2
-A- - - Cars-Sales & Scrappage
The passenger car age fractions used in MOVES are summarized in Table 6,1 at the
end of this section.
6.3. 1999 Trucks
To determine age fractions for refuse trucks, short-haul and long-haul single unit trucks
and short-haul and long-haul combination trucks, we used data from the VIUS database.
Vehicles in the VIUS database were assigned to MOVES source types as summarized in Table
3-3aandTable3.3b.
VIUS 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 VIUS 1997 questions "How did you obtain this vehicle?" (VIUS field
"OBTAIN" in VIUS 1997 or "ACQUIREHOW" in VIUS 2002) and "When did you obtain this
vehicle?" (VIUS field "ACQYR" in VIUS 1997 or "ACQUIREYEAR" in VIUS 2002) 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
26
-------
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.
Initially, we determined age fractions for passenger trucks and commercial trucks in the
same way as other trucks. However, when the new NHTSA survival rates for light duty trucks
became available, we reexamined this approach. We compared (1) our original approach with
VIUS data for 1997 and the TEDB scrappage rates, (2) a similar approach using VIUS data and
NHTSA survival rates, and (3) a "sales and scrappage" approach similar to that used for
passenger cars, combining passenger trucks and commercial light trucks and using TEDB sales
data. The results of the three approaches are illustrated in Figure 6.2.
Figure 6.2 1999 Age Distributions for Passenger and Light Commercial Trucks
Passenger and Commercial Light Truck Age Distributiions
0.12
0)
Q.
0)
^
o
J
'
o 0.04
0.02
0.08
0.06
-31 -Original VIUS
-32-OriginalVIUS
- 31 -VIUS & NHTSA
32-VIUS & NHTSA
-Trucks-Sales & Scrappage
10
15
Age
20
25
30
Use of the original VIUS data leads to a dip in 1996 and 1997 passenger trucks that is not
reflected by vehicle sales data. The other approaches all create similar trends of fairly steep
declines in age fractions until about age 7, a brief leveling off, another steep decline from about
age 12 to 17 and a final leveling off. For the MOVES default database, we selected the age
distribution generated with the "Sales and Scrappage" approach, which will be applied to both
passenger trucks and light commercial trucks. These rates are summarized in Table 6-1.
27
-------
6.4. 1999 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 above. 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. 1999 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. 1999 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.
28
-------
Table 6-1. 1999 Age Fractions for MOVES Source Types
source
type
age
0
1
2
3
4
5
6
1
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
11
0.0947
0.0935
0.0755
0.0681
0.0613
0.0570
0.0520
0.0433
0.0370
0.0355
0.0336
0.0388
0.0461
0.0422
0.0383
0.0345
0.0307
0.0270
0.0234
0.0198
0.0163
0.0129
0.0095
0.0062
0.0029
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
21
0.0646
0.0602
0.0610
0.0624
0.0626
0.0642
0.0597
0.0562
0.0543
0.0596
0.0608
0.0622
0.0549
0.0522
0.0419
0.0320
0.0226
0.0155
0.0129
0.0105
0.0080
0.0060
0.0045
0.0034
0.0026
0.0019
0.0014
0.0008
0.0006
0.0005
0.0000
31&32
0.1011
0.0906
0.0837
0.0791
0.0720
0.0700
0.0603
0.0502
0.0429
0.0450
0.0431
0.0422
0.0379
0.0351
0.0311
0.0244
0.0170
0.0127
0.0100
0.0100
0.0081
0.0066
0.0053
0.0041
0.0032
0.0031
0.0030
0.0029
0.0027
0.0026
0.0000
42
0.0624
0.0771
0.0742
0.0727
0.0627
0.0576
0.0504
0.0461
0.0492
0.0759
0.0609
0.0506
0.0489
0.0434
0.0394
0.0320
0.0321
0.0181
0.0082
0.0231
0.0071
0.0032
0.0007
0.0013
0.0009
0.0009
0.0002
0.0004
0.0003
0.0001
0.0002
43
0.0794
0.0660
0.0647
0.0594
0.0798
0.0406
0.0511
0.0435
0.0585
0.0696
0.0419
0.0526
0.0556
0.0512
0.0464
0.0374
0.0144
0.0111
0.0136
0.0138
0.0118
0.0104
0.0107
0.0073
0.0092
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
51
0.0498
0.0398
0.0340
0.0767
0.0926
0.0604
0.0544
0.0243
0.0696
0.0625
0.0514
0.0730
0.0610
0.0796
0.0442
0.0479
0.0145
0.0169
0.0156
0.0040
0.0043
0.0043
0.0000
0.0092
0.0027
0.0070
0.0001
0.0000
0.0000
0.0000
0.0000
52
0.0622
0.0520
0.0412
0.0466
0.0559
0.0572
0.0434
0.0344
0.0351
0.0435
0.0578
0.0531
0.0460
0.0580
0.0430
0.0251
0.0409
0.0220
0.0219
0.0239
0.0190
0.0225
0.0088
0.0112
0.0115
0.0125
0.0130
0.0265
0.0059
0.0032
0.0026
53
0.1697
0.1419
0.1124
0.0585
0.0609
0.1017
0.0783
0.0185
0.0138
0.0686
0.0748
0.0517
0.0129
0.0031
0.0064
0.0067
0.0000
0.0032
0.0024
0.0000
0.0002
0.0101
0.0006
0.0011
0.0005
0.0000
0.0021
0.0000
0.0000
0.0000
0.0000
54
0.0737
0.0456
0.0739
0.0487
0.0605
0.0608
0.0441
0.0408
0.0320
0.0442
0.0602
0.0563
0.0574
0.0447
0.0501
0.0531
0.0363
0.0221
0.0127
0.0017
0.0138
0.0191
0.0267
0.0169
0.0045
0.0000
0.0000
0.0000
0.0000
0.0000
0.0000
61 &41
0.0843
0.0672
0.0576
0.0506
0.0693
0.0562
0.0488
0.0379
0.0453
0.0535
0.0560
0.0550
0.0597
0.0528
0.0487
0.0400
0.0167
0.0147
0.0133
0.0180
0.0112
0.0090
0.0099
0.0038
0.0048
0.0048
0.0040
0.0036
0.0026
0.0006
0.0000
62
0.1668
0.1331
0.1140
0.1140
0.1186
0.0804
0.0643
0.0403
0.0304
0.0315
0.0320
0.0290
0.0080
0.0087
0.0115
0.0062
0.0013
0.0011
0.0035
0.0012
0.0010
0.0006
0.0010
0.0000
0.0009
0.0003
0.0003
0.0000
0.0002
0.0000
0.0000
29
-------
6.7. 1990 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 1990. However, data for
individual model years starting from 1978 and earlier were not available. A logarithmic
regression curve (R2 value = 0.82) was fitted to available data, which was then used to
extrapolate age fractions for earlier years beginning in 1978.
6.8. 1990 Passenger Cars
To determine age fractions for passenger cars, we began with Polk NVPP® 1990 data on
car registration by model year. However, this data presents a snapshot of registrations on July 1,
1990, and we needed age fractions as of December 31, 1990. 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 1990. Model Year 1989 cars were added to the previous
estimate of "Age 1" cars and Model Year 1990 and 1991 cars were added to the "Age 0" cars.
Also the data obtained was lumped together for ages 15+. Hence, regression estimates
were used to extrapolate the age fractions for individual ages 15+ based on an exponential curve
(R2 value =0.67) fitted to available data.
6.9. 1990 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 TIUS92 (1992 Truck Inventory and Use Survey) database. Vehicles in the
TIUS92 database were assigned to MOVES source types as summarized in Table 3-3.
TIUS92 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 TIUS92 questions "How was the vehicle obtained?" (TIUS field "OBTAIN")
and "When did you obtain this vehicle?" (TIUS field "ACQYR") to derive the model year of the
vehicles that were obtained new.
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 1992. This 1992 array probably overestimates the fraction of mid-
aged vehicles since the fraction of vehicles owned by their original owner clearly declines with
age; however, we believe the procedure is reasonable given the limited data available.
-------
6.10. 1990 Intercity Buses
As was true for the 1999 base year, 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.
6.11. 1990 School Buses and Motor Homes
Since we were unable to obtain the Polk TIP 1990 database, we used the 1999 age
fractions for School Buses and Motor Homes.
6.12. 1990 Transit Buses
For Transit Buses we used the MOBILE 6 age fractions since year 1990 data on transit
buses was not available from the Federal Transit Administration database.
-------
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 SourceBinDistributions could be input directly, MOVES usually generates the
values in this table using values in a collection of other tables. The SourceBinGenerator input
tables are described in Table 7-1.
This section describes how national default information was determined for MOVES.
Note that while previous versions of MOVES assigned fractions of vehicles to alternative fuels,
for Draft MOVES2009, we simplified the model by providing default fractions only for gasoline
and diesel vehicles. We expect to retain this simplified approach for the final MOVES2009.
Users wishing to model alternative fuels will still have the option of using the Alternative
Vehicle Fuels and Technology strategy to input their own fuel and engine technology fractions.
-------
Table 7-1. Data Tables Used by SourceBinGenerator
Generator Table Name
SourceTypePolProcess
FuelEngFraction
SizeWeightFraction
RegClassFraction
PollutantProcessModelYear
Sample VehiclePopulation
Key Fields
SourceTypelD
PolProcessID
SourceTypelD
ModelYearlD
FuelTypelD
EngTechID
SourceTypelD
ModelYearlD
FuelTypelD
EngTechID
WeightClassID
EngSizelD
SourceTypelD
ModelYearlD
FuelTypelD
EngTechID
RegClassID
PolProcessID
ModelYearlD
SourceType-
ModelYearlD
FuelTypelD
EngTechID
RegClassID
WeightClassID
EngSizelD
SCCVTypelD
Additional Fields
isSizeWeightReqd
isRegClassReqd
isMYGroupReqd
fuelEngFraction
SizeWeightFraction
regClassFraction
modelYearGroupID
stmyFuelEngFraction
stmyFraction
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.
Includes the fractions found in the
FuelEngFraction, RegClassFraction,
SizeWeightFraction and
SCCVTypeDistribution tables, but also
for combinations that do not exist in the
existing fleet. This table is only used
with the Alternative Vehicle Fuel &
Technology Strategy inputs to generate
alternate future vehicle fleet source
bins.
The MOVES 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.
33
-------
7.1. Motorcycles
For 1999-and-earlier motorcycle characteristics 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 1997 and for 1999.
Model year 2000 values were intended to be used for all 2000-and-later model years, however in
Draft MOVES2009, the 1999 value was used. For final MOVES2009, we intend to replace the
current 2000-and-later model year values with those based on the model year 2000 certification
data.
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(l)
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)*
0.029
0.043
0.928
Weight distribution
(EPA staff)
100%: <=5001bs
50%: <= 500 Ibs
50%: 5001bs -7001bs
30%: 500 lbs-700 Ibs
70%: > 7001bs
*Not entered in DraftMOVES2009, but planned for final.
7.1.3. RegClassFraction
All Motorcycles are assigned to the "Motorcycle" (MC) regulatory class.
34
-------
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.
The Polk fuel code was converted to the MOVES FuelTypelD using the mapping in
Table 6-3. .
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
5
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. While previous versions of MOVES included default values for alternative
fueled vehicles, DraftMOVES2009 includes only gasoline and diesel vehicles in the default
database. In model years where alternative vehicles were present,
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 sales35, 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
35
-------
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.
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.
The Vehicle Inventory and Use Survey (VIUS) conducted by the Census Bureau was the
primary source for information on truck distributions. Information from the 1997 and 2002
VIUS was supplemented with information from MOBILE6 and from the Oak Ridge National
Laboratory Light Duty Vehicle database.
VIUS 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 2002.
For years where no vehicles or only a few vehicles were surveyed by VIUS, we duplicated
fractions from the nearest available model year. The 2002 VIUS was used 1986 and later model
years and 1997 VIUS information was only used for the older model years not surveyed in the
2002 VIUS. In the Draft MOVES2009 release, the oldest model year observed diesel fractions
were applied to the older model years for combination trucks only. These older model years for
the other truck categories were assumed to have no diesel trucks.
7.3.1. FuelEngFraction
The VIUS 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 VIUS.
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. For the Draft MOVES2009
release, all non-gasoline trucks were set to be diesel fuel, so that the default fleet contains only
gasoline and diesel fuel trucks.
Table 7-4. Mapping VIUS ENGTYP to MOVES FuelTypelD
VIUS
1
2
3
4
5
Leaded gasoline
Unleaded gasoline
Diesel
Liquefied gas (petroleum
(LPG) or natural (LNG))
Other
MOVES
1
1
2
3or4
Gasoline
Gasoline
Diesel
CNG or
LPG
None
36
-------
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. The gasoline fractions can be estimated as one minus the diesel fractions listed
here.
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. The Census Bureau has discontinued the VIUS project, so
it will be necessary to use Polk data or other sources for this type of information for future
updates of these factors.
37
-------
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
0.00000
0.01633
0.03626
0.00000
0.00562
0.00833
0.00826
0.02875
0.01429
0.02557
0.01917
0.00792
0.02474
0.02167
0.00654
0.03755
Light
Commerical
Trucks
32
0.00000
0.00000
0.00000
0.00000
0.00906
0.08203
0.02876
0.00000
0.00000
0.00000
0.04185
0.05726
0.03149
0.29896
0.15086
0.21648
0.17784
0.07360
0.04131
0.11345
0.04988
0.05767
0.08897
0.13401
0.04579
0.06397
0.09397
0.06139
0.12999
0.04804
0.11866
Single-Unit
Short-haul
Trucks
52
0.00000
0.00000
0.06238
0.01695
0.04465
0.02377
0.02130
0.06518
0.32805
0.01731
0.11083
0.15791
0.16825
0.19327
0.67378
0.57100
0.52692
0.28809
0.50033
0.48870
0.51855
0.60288
0.66240
0.57597
0.62871
0.62889
0.65834
0.64296
0.68158
0.61441
0.73754
Single-Unit
Long-haul
Trucks
53
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.47356
1.00000
1.00000
0.06120
1.00000
1.00000
0.20453
0.87629
1.00000
1.00000
0.99148
0.31785
0.82097
0.89909
0.40003
0.82450
0.91614
1.00000
0.41192
0.89764
0.45123
0.88378
0.56891
0.61159
0.67638
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
0.96279
0.99402
0.98549
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
Combination
Long-haul
Trucks
62
1.00000
1.00000
1.00000
1.00000
1.00000
1.00000
1.00000
1.00000
1.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
.00000
0.99427
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
38
-------
7.3.2. SizeWeightFraction
Engine size distributions for trucks were determined using the VIUS 2002 database. The
VIUS database categorizes engine size by fuel type and the categories do not exactly match the
MOVES categories. We mapped from the VIUS engine size categories to the MOVES engine
size categories as described in Table 7-6. For comparison, the engine size ranges for both the
VIUS and MOVES categories are listed in cubic inches displacement.
EngSizelD
Table 7-6. Mapping VIUS Engine Size Categories to MOVES
Fuel Type
Gasoline
Gasoline
Gasoline
Gasoline
Gasoline
Gasoline
Gasoline
Diesel
Diesel
Diesel
Propane
Alcohol
Alcohol
Alcohol
Alcohol
Other
Other
Other
Other
Other
Other
Fuel Not
Reported
Vehicle Not
In Use
All
VIUS
Fuel_CID
code
1,2
3,4
5,6
7,8
9,10
11,12
13-18
20
21
22-36
38-41
43
44
45
46
48
49
50
51
52
53-56
58-61
63-66
19,37,42,47,5
7,62,67
VIUS CID
Range
1-129
130-149
150-179
180-209
210-239
240-299
300 & Up
1-249
250-299
300 & Up
All
1-229
230-269
270-339
340 & Up
1-99
100-149
150-199
200-249
250-299
300 & Up
All
All
Unknown
MOVES
EngSizelD
Code
20
2025
2530
3035
3540
4050
5099
3540
4050
5099
5099
3035
3540
4050
5099
20
2025
2530
3540
4050
5099
5099
5099
0
MOVES CID
Range
1-122
122-153
153-183
183-214
214-244
244-305
305 & Up
214-244
244-305
305 & Up
305 & Up
183-214
214-244
244-305
305 & Up
1-122
122-153
153-183
214-244
244-305
305 & Up
305 & Up
305 & Up
Unknown
Determining weight categories for light trucks was fairly complicated. The VIUS 1997
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
39
-------
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):
• VIUS 1997 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.
• VIUS 1997 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) VIUS 1997 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 VIUS 1997 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 VIUS 1997 average weight of 6,000 Ibs or less, we multiplied
the VIUS 1997 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 VIUS 1997 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 VIUS 1997 strata 1 and 2, thus a similar algorithm was applied.
• VIUS 1997 trucks of the Source Type in strata 3, 4, and 5 were assigned to the
appropriate MOVES weight class based on VIUS 1997 detailed average weight
information.
• VIUS 1997 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) VIUS 1997 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 VIUS 1997 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")
40
-------
• We did not believe the ORNL light duty vehicle database adequately represented
single unit trucks. Thus, the trucks with a VIUS 1997 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 VIUS 1997 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 VIUS 1997 strata 1 or 2. Thus we used the detailed VIUS 1997 average weight
information and engine size information to assign engine size and weight classes for all of these
trucks.
The VIUS 2002 contains an estimate of the average weight (vehicle weight plus cargo
weight) of 1998-2002 model year vehicle or vehicle/trailer combination as it was most often
operated when carrying a typical payload during 2002. These estimates were used to determine
the MOVES weightClassID categories for these trucks. Table 7.7 shows the weight ranges used
for each weightClassID. Any vehicles without a non-zero value for the average weight and
without a weight classification in the WeightAvgCK field were excluded from the analysis for
determining the average weight distributions.
Since there is a smaller number of gasoline trucks among the single unit and refuse
trucks, all model years (1998-2002) were combined to determine a single weight distribution to
use for these model years.
The average weight distributions for light duty trucks (sourceTypelD = 31, 32) and none
of the average weight distributions for any trucks for model years before 1998 were updated and
the VIUS 1997 estimates were retained.
In cases where distributions were missing (no survey information), distributions from a
nearby model year with the same source type was used. Weight distributions for all 2003 and
newer model years were set to be the same as for the 2002 model year for each source type.
-------
Table 7-7. Mapping VIUS Average Weight to MOVES WeightClassID
Where Weight Avg is not zero:
weightClassID
20
25
30
35
40
45
50
60
70
80
90
100
140
160
195
260
330
400
500
600
800
1000
1300
9999
WeightAvg Range
1-2000
2000-2499
2500-2999
3000-3499
3500-3999
4000-4499
4500-4999
5000-5999
6000-6999
7000-7999
8000-8999
9000-9999
10000-13999
14000-15999
16000-19499
19500-25999
26000-32999
33000-39999
40000-49999
50000-59999
60000-79999
80000-99999
100000-129999
130000 & Up
Where Weight Avg is zero:
weightClassID
140
160
195
WeightAvgCK
4 (10000-14000)
5 (14000-16000)
6 (16000-19500)
7.3.3. RegClassFraction
Trucks were split between the regulatory classes "Light-Duty Trucks" (LDT) and
"Heavy-Duty Trucks" (HOT) based on gross vehicle weight (GVW) (the maximum weight that a
truck is designed to carry.)
In particular, we used the VIUS response "PKGVW" in VIUS 1997 and ADM_GVW in
VIUS 2002 and the Davis & Truit report on Class 2b Trucks36 to determine GVW fractions by
fuel type. The VIUS fields are intended to identify the Polk weight class. Work for MOBILE6
using the VIUS precursor, TIUS 1992 indicated that the PKGVW measure in VIUS is
problematic. TIUS PKGVW 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 VIUS. However, "maximum
weight" was not available for smaller trucks, and the other measures of weight reported in VIUS
were not consistent with the need for an indicator of the relevant emission standards. When the
42
-------
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.
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
Ibs). Davis & Truitt37 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.
Table 7.8. Light Truck Class 2 Weight Distribution
Fuel Type
Gasoline
Diesel
Any
Class 2a
6001-8500 Ibs.
GVWR
74.7%
2.0%
76.7%
Class 2b
8501-10000 Ibs. GVWR
17.7%
5.6%
23.3%
Class 2b Fraction
19.2%
73.7%
The regulatory class fractions for trucks are listed below in Table 7-9 and Table 7-10.
Fractions of LDT for gasoline- and diesel-fueled vehicles are provided separately. The
remaining trucks are classified as HDT. 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. All 1986 and newer model year data was obtained from
VIUS 2002. The pre-1986 model year values are from VIUS 1997.
43
-------
Table 7-9. 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
1998
1999
2000
2001
2002
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
0.926321
0.951630
0.949331
0.951473
0.950769
0.958130
0.953552
0.953891
0.950555
0.945395
0.948863
0.950000
0.947357
0.930476
0.937397
0.935546
0.945155
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
0.818333
0.897109
0.890861
0.891322
0.911313
0.887311
0.905625
0.908697
0.872257
0.877733
0.861956
0.877692
0.891901
0.870745
0.884837
0.880982
0.897487
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
0.317167
0.458448
0.421998
0.525825
0.508253
0.405240
0.453636
0.672601
0.510745
0.453314
0.515149
0.447634
0.412569
0.366611
0.615046
0.537060
0.587987
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
0.429721
0
0
0
0
0
0.624370
0
0
0
0
0
0
0
0
0.429721
0
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
0
0
0
0
#N/A
0
#N/A
0
0
0
#N/A
0
0.082522
#N/A
0
#N/A
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
#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
44
-------
Table 7-10. 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
1998
1999
2000
2001
2002
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
0.456608
0.951630
0.254950
0.260932
0.260713
0.261741
0.262386
0.262899
0.298405
0.308964
0.289104
0.261310
0.263000
0.260865
0.358104
0.234050
0.282868
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
0.351492
0.088341
0.210368
0.144417
0.062091
0.176872
0.222906
0.149897
0.159601
0.200670
0.211153
0.356162
0.142366
0.214650
0.216855
0.342721
0.262352
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
0.021218
0.129185
0.054122
0.031919
0
0.111821
0.042603
0.156027
0.073051
0.117612
0.113798
0.120503
0.017443
0.155014
0.171699
0.120036
0.085967
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
0.028255
0
0.068212
0
0
0.184952
0.029801
0.538647
0.042628
0
0.084009
0
0
0
0.298503
0.188003
0
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
0
0
0
0
0
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
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
45
-------
7.4. Buses
Because buses are not included in VIUS 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 (AEO2006, 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
7-11. The resulting fractions by model year are summarized in Table 7-10.
Table 7-11.
Types
Mapping National Transit Database Fuel Types to MOVES Fuel
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
3
2
2
9
9
5
1
na
na
3
4
6
na
MOVES Fuel
Description
CNG
diesel
diesel
electric
electric
ethanol
gasoline
CNG
LPG
methanol
46
-------
Table 7-12. 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
estimates by model year.38) 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 7-13. In the future it would be desirable to obtain up-to-date,
detailed fuel information for school buses from Polk or some other source.
47
-------
Table 7-13. 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 FTA 2003 Report to
Congress39 that specified the number of buses in various weight categories. This information is
summarized in below in Table 7-14. 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 7-14. 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
48
-------
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 FTA Report to
Congress40and 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 data41 provided information on number of vehicles by gross
vehicle weight class and fuel as detailed in Table 7-15.
Table 7-15. 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 7-16.
7.4.3. RegClassFraction
All buses were assigned to the Heavy-Duty Truck regulatory class.
49
-------
Table 7-16. 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 VIUS.
7.5.1. FuelEngFraction
As for other trucks, we used the VIUS EngTyp field to estimate FuelType and Engine
Technology Fractions. The Refuse Trucks classified in VIUS as "CNG or LPG" are assigned to
diesel. All Refuse Trucks were assumed to have conventional internal combustion engines.
7.5.2. SizeWeightFraction
Because the sample of Refuse Trucks in VIUS was small, the same SizeWeight
distributions were used for model year groups. As for other trucks, the EngineSize group was
determined from the VIUS engine size categories and the WeightClass was determined from the
VIUS reported average weight.
-------
Table 7-17. Fuel Fractions for Refuse Trucks by Model Year
Model
Year
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
Gasoline
0.0155
0
0.2206
0.2132
0.1687
0
0.0231
0.1109
0
0.1120
0.0292
0.0415
0.0119
0
0.0201
0
0.0349
0.0184
0
0
Diesel
0.9845
1.0000
0.7794
0.7868
0.8313
1.0000
0.9769
0.8891
1.0000
0.8880
0.9708
0.9585
0.9881
1.0000
0.9799
1.0000
0.9651
0.9816
1.0000
1.0000
51
-------
Table 7-18. Refuse Truck SizeWeight Fractions by Fuel Type
Gasoline
Engine Size
3-3.5L
>5L
>5L
>5L
>5L
>5L
>5L
>5L
>5L
>5L
Sum
Diesel
Engine Size
3.5-4L
4-5L
4-5L
4-5L
>5L
>5L
>5L
>5L
>5L
>5L
>5L
>5L
>5L
>5L
>5L
>5L
Sum
Weight (Ibs.)
5000-6000
7000-8000
9000-10000
10000-14000
14000-16000
16000-19500
19500-26000
26000-33000
33000-40000
50000-60000
Weight (Ibs.)
10000-14000
10000-14000
14000-16000
16000-19500
9000-10000
10000-14000
14000-16000
16000-19500
19500-26000
26000-33000
33000-40000
40000-50000
50000-60000
60000-80000
80000-100000
100000-130000
Pre-1997
0.009074
0.148826
0.070720
0.135759
0.199961
0.055085
0.205341
0.022105
0.153129
0
1.000000
Pre-1998
0.007758
0
0
0
0.006867
0.011727
0.022960
0.063128
0.099782
0.102077
0.237485
0
0.336484
0.111730
0
0
1.000000
1997 and Newer
0
0
0
0.324438
0.593328
0
0
0
0
0.082234
1.000000
1998
0
0
0
0
0.009593
0
0
0
0.035378
0.019625
0.103922
0.283642
0.338511
0.196424
0
0.012904
1.000000
1999
0
0
0
0
0
0
0
0.011367
0.026212
0.067419
0.088975
0.275467
0.326902
0.193238
0.010420
0
1.000000
2000
0
0
0.015505
0
0
0
0
0.047200
0.052132
0.072106
0.085991
0.165624
0.384612
0.176831
0
0
1.000000
2001
0
0
0
0.011670
0
0.019438
0
0
0.018329
0.043877
0.042678
0.266357
0.315133
0.282517
0
0
1.000000
2002 and Newer
0
0.006614
0
0
0
0
0
0
0.026079
0
0.046966
0.194716
0.474469
0.224995
0.013081
0.013081
1.000000
7.5.3. RegClassFraction
Using the VIUS 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.42 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 7-19. 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.
Table 7-19. 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 7-20 and Table 7-21.
53
-------
Table 7-20. 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
54
-------
Table 7-21. 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
MOVES 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 7-22. Note that some fuel types that were supported in earlier versions
of MOVES (methanol and hydrogen) are not available in DraftMOVES2009.
The various hybrid types were split into "mild" 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 "mild" hybrids (like the
55
-------
early 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 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. 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).
,43
Table 7-22. 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
Electricity
Engine
Technology
Conventional
1C
Advanced 1C
CIC Hybrid
Mild
CIC Hybrid
Full
AIC Hybrid
Mild
AIC Hybrid
Full
Conventional
1C
Advanced 1C
CIC Hybrid
Mild
CIC Hybrid
Full
AIC Hybrid
Mild
Diesel AIC
Hybrid Full
Conventional
1C
Conventional
1C
Conventional
1C
Electric only
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
Combi-
nation
Short &
Long
Haul
Trucks
X
X
X
X
56
-------
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
Draft MOVES2009 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.
We analyzed passenger cars and light trucks separately. All vehicles were assigned to
either the gasoline or diesel fuel conventional engine technology category for all future years.
MOVES contains no projections for the use of hybrid or advanced engine technology or the use
of alternative fuels in future calendar years. The resulting fuelEngFractions for conventional
gasoline and diesel fueled vehicles are listed in Table 7.23.
We used the size and weight distributions from the 2002 model year for all 2003 and
newer model years. The size and weight distribution for 2002 model year gasoline conventional
internal combustion engines were used for all 2003 and newer model year technologies and fuel
types, other than diesel. The 2003 and newer model year diesel vehicles of all technologies use
the size and weight distribution for diesel conventional internal combustion engines of the 2002
model year.
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 (FtDT)
regulatory classes. We used the 2002 model year regulatory class distribution for gasoline
conventional internal combustion vehicles for all 2003 and newer model year technologies and
fuel types, other than diesel.. The 2003 and newer model year diesel vehicles of all technologies
use the regulatory class distribution for diesel conventional internal combustion vehicles of the
2002 model year.
57
-------
Table 7.23. Fuel Fractions for 2002 and Newer Passenger Cars and Light Duty
Trucks
Model Year
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
and newer
Passenger Cars
gasoline
0.9900
0.9900
0.9900
0.9900
0.9900
0.9900
0.9900
0.9900
0.9900
0.9900
diesel
0.0100
0.0100
0.0100
0.0100
0.0100
0.0100
0.0100
0.0100
0.0100
0.0100
Passenger Trucks
gasoline
0.9870
0.9870
0.9870
0.8597
0.5942
0.4264
0.2171
0.1994
0.1747
0.1658
diesel
0.0130
0.0130
0.0130
0.0123
0.0109
0.0100
0.0089
0.0088
0.0087
0.0087
Commerical Light Trucks
gasoline
0.9870
0.9870
0.9870
0.8597
0.5823
0.4234
0.2122
0.1935
0.1727
0.1500
diesel
0.0130
0.0130
0.0130
0.0121
0.0101
0.0089
0.0074
0.0072
0.0071
0.0070
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. Fuel and technology projections were not
available from AEO. The MOVES estimates for 1999 distributions of transit and school buses
are carried forward to 2050. These distributions are summarized in Table 7.24. Engine size
and vehicle weight distributions were also carried forward from 1999. All buses were assigned
to the Heavy-Duty Truck regulatory class.
Table 7.24. Fuel and Engine Technology Fractions for 2000-and-later Buses
Intercity Buses
Transit Buses
School Buses
Diesel CIC
1.00000
0.99399
0.95846
Gasoline CIC
0
0.00601
0.04154
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. Furthermore, AEO Table 146 lists technology penetrations for Class 4-6 freight
vehicles. All non-gasoline trucks, other than diesel, were assigned to the MOVES gasoline
-------
conventional combustion category. All diesel trucks with were assigned to the MOVES diesel
conventional internal combustion category. The resulting distributions are summarized in Table
7.25.
We used the engine size and vehicle weight distributions from 2002 for future years.
Where a future fuel was not part of the fleet in 2002, we used the 2002 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 2002, we used the 2002 size and
weight distribution for diesel conventional internal combustion vehicles.
Table 7.25. Fuel and Engine Technology Fractions for 2002 and Newer Motor
Homes and Single-Unit Short-haul and Long-haul Trucks
Model Year
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
2027
2028
2029
2030
203 land
Newer
Single Unit Short Haul
gasoline
0.2631
0.2924
0.2869
0.2809
0.2758
0.2710
0.2674
0.2642
0.2620
0.2602
0.2589
0.2579
0.2572
0.2566
0.2562
0.2560
0.2560
0.2561
0.2563
0.2565
0.2569
0.2573
0.2578
0.2586
0.2591
0.2594
0.2602
0.2608
0.2613
0.1532
diesel
0.7369
0.7076
0.7131
0.7191
0.7242
0.7290
0.7326
0.7358
0.7380
0.7399
0.7411
0.7421
0.7428
0.7434
0.7438
0.7440
0.7440
0.7439
0.7437
0.7435
0.7431
0.7427
0.7422
0.7414
0.7409
0.7406
0.7398
0.7392
0.7387
0.8468
Single Unit Long Haul
gasoline
0.0627
0.2924
0.2869
0.2809
0.2758
0.2710
0.2674
0.2642
0.2620
0.2602
0.2589
0.2579
0.2572
0.2566
0.2562
0.2560
0.2560
0.2561
0.2563
0.2565
0.2569
0.2573
0.2578
0.2586
0.2591
0.2594
0.2602
0.2608
0.2613
0.1532
diesel
0.9373
0.7076
0.7131
0.7191
0.7242
0.7290
0.7326
0.7358
0.7380
0.7399
0.7411
0.7421
0.7428
0.7434
0.7438
0.7440
0.7440
0.7439
0.7437
0.7435
0.7431
0.7427
0.7422
0.7414
0.7409
0.7406
0.7398
0.7392
0.7387
0.8468
Motor Home
gasoline
0.2237
0.2924
0.2869
0.2809
0.2758
0.2710
0.2674
0.2642
0.2620
0.2602
0.2589
0.2579
0.2572
0.2566
0.2562
0.2560
0.2560
0.2561
0.2563
0.2565
0.2569
0.2573
0.2578
0.2586
0.2591
0.2594
0.2602
0.2608
0.2613
0.1532
diesel
0.7763
0.7076
0.7131
0.7191
0.7242
0.7290
0.7326
0.7358
0.7380
0.7399
0.7411
0.7421
0.7428
0.7434
0.7438
0.7440
0.7440
0.7439
0.7437
0.7435
0.7431
0.7427
0.7422
0.7414
0.7409
0.7406
0.7398
0.7392
0.7387
0.8468
59
-------
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.
All non-gasoline trucks, other than diesel, were assigned to the MOVES gasoline conventional
combustion category. All diesel trucks with were assigned to the MOVES diesel conventional
internal combustion category.
Furthermore, AEO Table 146 lists technology penetrations for Class 7-8 freight trucks
with "higher cylinder pressure", "improved injection & combustion" and "waste heat/thermal
management". All trucks were assigned to the MOVES the conventional internal combustion
categories. The resulting distributions are summarized in Table 7.26.
We used the engine size and vehicle weight distributions from 2002 for future years.
Where a future fuel or engine technology was not part of the source type fleet in 2002, we used
the 2002 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 2002, we used the regulatory class distribution for diesel conventional internal
combustion vehicles.
60
-------
Table 7.26. Fuel and Engine Technology Fractions for Refuse Trucks and Short-
haul and Long-haul Combination Trucks
Model Year
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
2027
2028
2029
2030
203 land
Newer
Refuse Trucks
gasoline
0.0000
0.0332
0.0330
0.0328
0.0328
0.0330
0.0333
0.0336
0.0340
0.0344
0.0348
0.0353
0.0357
0.0362
0.0367
0.0371
0.0376
0.0380
0.0384
0.0388
0.0392
0.0396
0.0400
0.0404
0.0407
0.0411
0.0414
0.0417
0.0420
0.0164
diesel
1.0000
0.9668
0.9670
0.9672
0.9672
0.9670
0.9667
0.9664
0.9660
0.9656
0.9652
0.9647
0.9643
0.9638
0.9633
0.9629
0.9624
0.9620
0.9616
0.9612
0.9608
0.9604
0.9600
0.9596
0.9593
0.9589
0.9586
0.9583
0.9580
0.9836
Combination Short Haul
gasoline
0.0000
0.0330
0.0328
0.0327
0.0327
0.0329
0.0331
0.0335
0.0338
0.0342
0.0347
0.0351
0.0356
0.0361
0.0365
0.0370
0.0374
0.0379
0.0383
0.0387
0.0391
0.0395
0.0399
0.0403
0.0406
0.0409
0.0413
0.0416
0.0419
0.0164
diesel
1.0000
0.9670
0.9672
0.9673
0.9673
0.9671
0.9669
0.9665
0.9662
0.9658
0.9653
0.9649
0.9644
0.9639
0.9635
0.9630
0.9626
0.9621
0.9617
0.9613
0.9609
0.9605
0.9601
0.9597
0.9594
0.9591
0.9587
0.9584
0.9581
0.9836
Combination Long Haul
gasoline
0.0000
0.0330
0.0328
0.0327
0.0327
0.0329
0.0331
0.0335
0.0338
0.0342
0.0347
0.0351
0.0356
0.0361
0.0365
0.0370
0.0374
0.0379
0.0383
0.0387
0.0391
0.0395
0.0399
0.0403
0.0406
0.0409
0.0413
0.0416
0.0419
0.0164
diesel
1.0000
0.9670
0.9672
0.9673
0.9673
0.9671
0.9669
0.9665
0.9662
0.9658
0.9653
0.9649
0.9644
0.9639
0.9635
0.9630
0.9626
0.9621
0.9617
0.9613
0.9609
0.9605
0.9601
0.9597
0.9594
0.9591
0.9587
0.9584
0.9581
0.9836
61
-------
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:
)• v3 + (a + g • sin 9] • v.
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 sins^ 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:
s
z'=l, total # of ages
weightedvalue =
Qtj - unweightedvalue
7=1, 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: o,j
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.
62
-------
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 parameterized 44 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@50mph) recorded in the MSOD.45 The calculations
applied the following empirical equations:46
A = 0.7457*(0.35/50*0.447)
B = 0.7457*(0.10/(50*0.447)2)
* TRLHP@50mph
* TRLHP@5Qmph
63
-------
C = 0.7457*(0.55/(50*0.447)3) * TRLHP@50mph
The rolling resistance was multiplied by a factor of 5.
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,47 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)
C(kW*s-Vm3)
/M(tonne)
8500 to 14000 Ibs
(3.855 to 6.350
tonne)
0.0996
0
3.40 x 10"4
(mass is the average
mass of the weight
category)
1.47 5
masdkg)
14000 to 33000 Ibs
(6.350 to 14.968
tonne)
0.0875
0
1.97xlO"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 xlO"4
(mass is the
average mass of the
weight category)
2.89 _ 5
mas^kg)
Buses and
Motor Homes
0.0643
0
3.22 _5
mas^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.
64
-------
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
0.156461
0.22112
0.235008
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
For Final MOVES2009, we will add a new field to the SourceUseType table,
"fixedMassFactor," that will serve as the denominator in the Vehicle Specific Power (VSP)
equation, which generates a relationship between power and emissions that varies with the fixed
mass. (For more on VSP, see the Operating Mode Distribution Generator descriptions in the
Software Design and Reference Manual48) The fixed mass is fundamental to the calculation of
the emission rates in the MOVES emission rate tables. In Final MOVES2009, if a user wishes
to do "what if calculations varying the sourceMass, the fixedMassFactor should remain
constant. Such 'what if calculations are not possible in Draft MOVES2009, because increasing
the source mass would increase both the numerator and the denominator in the VSP equation,
leading to an incorrect decrease in emissions and energy consumption.
-------
9. RoadTypeDistribution
MOVES will calculate emissions separately for each road type and for "off-network"
activity. The road type codes used in MOVES are listed in Table 9-1. These road types are
aggregations of the HPMS functional facility types that are also used for SCC reporting.
Table 9-1. Road Type Codes in MOVES
RoadTypelD
1
2
o
5
4
5
Description
Off Network
Rural Restricted
Access
Rural Unrestricted
Access
Urban Restricted
Access
Urban Unrestricted
Access
HPMS functional Types
Off Network
Rural Interstate
Rural Principal Arterial, Minor
Arterial, Major Collector, Minor
Collector & Local
Urban Interstate & Urban
Freeway /Expressway
Urban Principal Arterial, Minor
Arterial, Collector & Local
SCCRoadTypelD
1
11
13, 15, 17, 19,21
23,25
27,29,31,33
For each SourceType, the RoadTypeVMTFraction field stores the fraction of total
VMT that is traveled on each of the 5 roadway types.
For MOVES2009, 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 HPMS
functional 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 facility type and allocated them to vehicle class in proportion to the values in
VM-1. We then calculated facility type VMT fractions for each HPMS Vehicle Type. We then
aggregated the values to the five MOVES road types.
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 Statistics. 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 and thus one of the four "on-network"
MOVES roadtypes.. No VMT is currently assigned to the "off-network" category in the national
defaults. See the discussion of BaseYearOffNetVMT in Section 11.2.
66
-------
Table 9-2. Roadtype Distributions by Sourcetype
RoadType ID
1
2
3
4
5
Total
Description
Off Network
Rural Restricted
Access
Rural Unrestricted
Access
Urban Restricted
Access
Urban
Unrestricted
Access
Motorcycles
0.0000
0.1040
0.3161
0.2177
0.3623
1.0000
Passenger
Cars
0.0000
0.0834
0.2891
0.2097
0.4178
1.0000
Other 2axle
- 4tire
vehicles
0.0000
0.0846
0.3055
0.2031
0.4068
1.0000
Buses
0.0000
0.1268
0.4821
0.1385
0.2526
1.0000
Single
unit
trucks
0.0000
0.1149
0.3972
0.1715
0.3165
1.0000
Combination
trucks
0.0000
0.3247
0.2941
0.2075
0.1737
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 restricted access roadways than the short-haul trucks. If such data becomes available we
would like to update the database.
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. The
values sum to one for each combination of SourceType, Road Type, Day, and Hour.
For Draft MOVES2009, the urban driving values were derived from the default speed
distributions (SVMT) in MOBILE6. The MOBILE6 speed fractions were adapted to MOVES
by converting the fraction of miles travelled to the fraction of time used, and by mapping from
the MOBILE6 road types to the MOVESroad types, with the MOBILE6 "freeway" values
mapped to the MOVES "urban restricted" roadtype and the MOBILE6 "arterial" values mapped
to the MOVES "urban unrestricted" roadtype. The time fractions were normalized to sum to one
for each hour of the day over all 14 speed bins. The values for the off-network roadway type
were set to null. The detailed distributions are available in the MOVES default database. Only
urban roadways obtain their values from the default MOBILE6 speed distributions. Average
speed used for rural driving relied on recent driving data collected in California under studies
performed for the California Department of Transportation (Caltrans). 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.
A contractor report describes the analysis done to develop speed distributions from these
datasets.49 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 ranged between 0.5 to 5 miles, a few of 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 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.
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
To import this information into MOVES, we started with the contractor-provided values
of "Time-weighted Distributions (% of time) of California Rural Chase Car Driving Data by
Average Link Speed for each HPMS Functional Class."50 These values were used directly for
the rural restricted access roadtype (2). For the MOVES rural unrestricted access roadtype, the
calculation required consolidating values on the five HPMS functional road classes to the single
MOVES roadtype. This was done separately for each HPMS Vehicle Class. For each vehicle
class, we used the roadtype distribution (see preceding section) to calculate the fraction of VMT
on each road class. We then changed to a time-basis by calculating the average speed on each
road class, dividing by the average speed and re-nomalizing. We then computed a sum-product
68
-------
of the speed bin fractions and the road class distributions to calculate the weighted-average speed
bin distribution for each vehicle class and assigned this distribution to each sourcetype in the
HPMS vehicle class.
Our use of the California rural data required a number of assumptions and extrapolations.
For Draft MOVES2009, the same rural speed distributions were used for all hours of the day.
And, while the California chase car data also only included light-duty vehicles, the resulting
speed distributions are also used for heavy-duty vehicles. Also 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.
National default speed distributions are available in the default database for each
roadtype, sourcetype and hourday, and are not provided here. However, for illustration, Figure
14.1, shows the speed distributions on different roadtypes for passenger cars for the time period
11 am. to noon on weekdays.
Figure 10.1 Speed Distribution by Roadtype
Speed Distributions by Roadtype
Passenger Cars, 11 am-Noon Weekdays
0.7
0.6 -
0.5
0.4
§0.3-
0.2
0.1
0
Avg. Speed of Speed Bin
I rural restricted n rural unrestricted Durban restricted Durban unrestricted
69
-------
ll.HPMSVTypeYear
Three fields comprise HPMSVTypeYear in Draft MOVES2009: 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 FHWAVM-1 and VM-2 tables as for
RoadTypeDistribution, but instead of calculating fractions, we calculated VMT sums by HPMS
Vehicle Class.
The resulting VMT for 1999 and 1990 by HPMS Vehicle Class is listed in Table 11-1.
Table 11-1. 1999 VMT by HPMS Vehicle Class
HPMS Vehicle Class
Motorcycles
Passenger Cars
Other 2 axle - 4 tire vehicles
Buses
Single unit trucks
Combination trucks
1990 VMT
9,557,000,000
1,408,270,000,000
574,571,000,000
5,726,000,000
51,901,000,000
94,341,000,000
1999 VMT
10,579,600,000
1,568,640,000,000
900,735,000,000
7,657,000,000
70,273,700,000
132,358,000,000
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. This
field is provided in case it is useful for modeling local areas. However, the reported HPMS
VMT values, used to calculate the national averages discussed here, are intended to include all
VMT. Thus, for Draft MOVES2009 national defaults, 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 Draft
MOVES2009. It is important to note that VMTGrowthFactor is a key component for estimates
of future activity in MOVES, because the level of total activity in future years for many emission
70
-------
processes is derived from projections of total VMT. For these processes, projections of future
populations based on sales growth, survival rates, etc. are only used to allocate total VMT.
Default estimates for VMTGrowthFactor were taken from FHWA Highway Statistics for
2000 through 2004, and from AEO2006 for years 2005-and-later. For passenger cars and light-
duty trucks, additional calculations were needed to allocate the more aggregate AEO estimates
for light-duty vehicles and trucks to the MOVES Source Types.
Calendar year 2000 through 2004 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
factors are simply total VMT for the calendar year divided by total VMT from the previous year.
Growth factors for calendar years 2005 through 2030 were calculated in the same manner
using NEMS projections of total VMT as reported in AEO2006. In the AEO analysis, VMT
projections are provided for total Light-Duty (AEO2006 Supplemental Table 48), total Medium-
Duty, and total Heavy-Duty (AEO2004 Supplemental Table 55). The growth factors derived
from the AEO2006 Medium-Duty VMT estimates were applied to the single-unit truck and bus
HPMS vehicle classes. The growth factors derived from the AEO2006 Heavy-Duty VMT
estimates were applied to the combination truck vehicle class.
Light-Duty VMT as reported in AEO2006 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 AEO2006.
Using theAEO2006 estimates of total light-duty VMT and vehicle population (i.e., stock) growth
rates listed in AEO Supplemental Table 46, we calculated the "per-vehicle" VMT implied from
these estimates (total VMT divided by population). Assuming that per-vehicle VMT growth is
the same for cars and light trucks, we multiplied the total light-duty per-vehicle VMT by the car
and light truck populations to project separate car and light truck VMT for future years and then
computed annual growth rates. Table 11-3 illustrates these calculation steps.
For final MOVES2009, we plan to update these growth factors using updated VMT
information and projections.
-------
Table 11-2. VMTGrowthFactor Calculation for Passenger Cars and Light Trucks
Calendar
Year
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
Vehicle Stock (million)
(From AEO2006)
LD
Total
211.553
216.805
221.645
226.700
231.613
236.476
241.264
246.004
250.608
254.971
259.118
263.028
266.797
270.473
274.100
277.634
281.091
284.535
287.982
291.469
295.016
298.626
302.333
306.124
309.952
313.854
317.825
LDV
130.782
131.992
133.602
135.178
136.561
137.746
138.705
139.664
140.540
141.283
141.888
142.395
142.866
143.308
143.740
144.156
144.585
145.036
145.515
146.029
146.582
147.172
147.816
148.503
149.223
149.988
150.802
LOT
80.771
84.813
88.043
91.522
95.052
98.730
102.559
106.340
110.068
113.688
117.231
120.633
123.932
127.165
130.360
133.478
136.506
139.499
142.467
145.439
148.435
151.454
154.517
157.621
160.729
163.866
167.023
VMT
(billion)
(From
AEO
2006)
LD
Total
2632.078
2619.176
2644.429
2693.347
2751.712
2818.227
2889.563
2946.387
3000.774
3055.248
3113.610
3171.164
3227.686
3288.719
3351.878
3414.157
3474.341
3535.598
3597.454
3660.468
3725.200
3791.240
3858.390
3927.069
3995.345
4064.186
4132.401
Per-
Vehicle
VMT
LD
Total
12.442
12.081
11.931
11.881
11.881
11.918
11.977
11.977
11.974
11.983
12.016
12.056
12.098
12.159
12.229
12.297
12.360
12.426
12.492
12.559
12.627
12.696
12.762
12.828
12.890
12.949
13.002
VMT by Type (pop * per-vehicle vmt)
LDV
Total
1627.147
1594.573
1593.996
1606.008
1622.438
1641.603
1661.235
1672.749
1682.825
1692.960
1704.944
1716.763
1728.375
1742.502
1757.751
1772.740
1787.100
1802.203
1817.763
1833.938
1850.899
1868.435
1886.440
1905.045
1923.515
1942.235
1960.744
Growth
0.980
.000
.008
.010
.012
.012
.007
.006
.006
.007
.007
.007
.008
.009
.009
.008
.008
.009
.009
.009
.009
.010
.010
.010
.010
.010
LOT
Total
1004.931
1024.603
1050.433
1087.339
1129.274
1176.623
1228.328
1273.638
1317.948
1362.288
1408.665
1454.400
1499.312
1546.217
1594.126
1641.418
1687.241
1733.395
1779.690
1826.530
1874.301
1922.804
1971.950
2022.024
2071.829
2121.952
2171.658
Growth
1.020
1.025
1.035
1.039
1.042
1.044
1.037
1.035
1.034
1.034
1.032
1.031
1.031
1.031
1.030
1.028
1.027
1.027
1.026
1.026
1.026
1.026
1.025
1.025
1.024
1.023
72
-------
Table 11-3. VMT Growth Factors in Draft MOVES2009
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
2027
2028
2029
2030
2031-
2050
Motorcycles
0.990
0.910
0.991
1.003
1.049
0.980
1.000
1.008
1.010
1.012
1.012
1.007
1.006
1.006
1.007
1.007
1.007
1.008
1.009
1.009
1.008
1.008
1.009
1.009
1.009
1.009
1.010
1.010
1.010
1.010
1.010
1.010
Passenger
Cars
1.021
1.012
1.019
1.008
1.020
0.980
1.000
1.008
1.010
1.012
1.012
1.007
1.006
1.006
1.007
1.007
1.007
1.008
1.009
1.009
1.008
1.008
1.009
1.009
1.009
1.009
1.010
1.010
1.010
1.010
1.010
1.010
Passenger
and Light
Comm.
Trucks
1.026
1.016
1.024
1.019
1.031
1.020
1.025
1.034
1.037
1.040
1.042
1.036
1.034
1.033
1.033
1.032
1.030
1.031
1.030
1.029
1.027
1.027
1.026
1.026
1.026
1.026
1.025
1.025
1.024
1.024
1.023
1.023
Buses
0.992
0.920
0.968
0.991
0.979
0.998
.007
.016
.013
.018
.021
.025
.023
.022
.023
.024
.025
.026
.026
.025
.025
.026
.027
.027
.027
.027
.028
.027
.027
.027
.026
.026
Single
Unit
Trucks
.004
.025
.048
.025
.043
0.998
.007
.016
.013
.018
.021
.025
.023
.022
.023
.024
.025
.026
.026
.025
.025
.026
.027
.027
.027
.027
.028
.027
.027
.027
.026
.026
Combination
Trucks
1.021
1.003
1.015
1.010
1.037
1.022
1.034
1.033
1.025
1.025
1.026
1.025
1.023
1.022
1.022
1.023
1.023
1.025
1.025
1.021
1.020
1.020
1.021
1.021
1.022
1.023
1.024
1.024
1.023
1.023
1.023
1.023
73
-------
12. Temporal Distributions of VMT
MOVES can estimate emissions for every hour of every day of the year. For this reason, for
national scale runs ("macroscale") annual VMT estimates need to be allocated to months, days,
and hours.
A 1996 report from the Office of Highway Information Management (OHIM)51 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
Draft MOVES2009.
The report does not specify VMT by SourceType or Vehicle Type. Thus, we currently
use the same value for all SourceTypes. For Final MOVES2009, we plan to update the MOVES
pre-processor tools to aid local areas entering VMT with accurate local temporal distributions.
12.1. MonthVMTFraction
For Month VMTFraction, we use the data from the OHIM 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 MOVES, 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. Month VMTFraction
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
74
-------
12.2. DayVMTFraction
The OHIM 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 OHIM report is not disaggregated by month or
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 OHIM report.
For the DayVMTFraction needed for MOVES2009, we first summed the reported
percentages for each day of the week and converted to fractions. 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. Because MOVES2009 classifies days into two types of days, "weekdays"
and "weekend," we then summed the daily fractions to compute fractions for each type of day.
Table 12-2. DayVMTFractions
Weekday
Weekend
Rural
0.2788
0.7212
Urban
0.2376
07624
We assigned the "Rural" fractions to the rural Roadtypes and the "Urban" fractions to the
urban Roadtypes. The correct distribution for "Off network" VMT is unknown. Since the
majority of U.S. travel is urban, the default DayVMTFraction for "Off network" will be assigned
the urban fractions. Note the MOVES2009 default VMT on "Off-network" roadtypes is zero.
12.3. HourVMTFraction
For HourVMTFraction we used the same data as for DayVMTFraction. We converted
the OHIM 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. The Off-Network roadtype was assigned the "Urban" fractions.
Figure 12.1 graphs the hourly VMT fractions.
75
-------
Figure 12.1 Hourly VMT Fractions in Draft MOVES2009
n OQ
n OR
i— n 07 -
> n ofi
-^ o o=;
a
O n 04 -
"5 o OT
o 0 09
"5 n 01
n U.U I -
ui o
c
Hourly VMT
• VT-^V^I \
f\ ' • "1-
/ \y ^
i r ^
7 %
. / 5
V f
r.*>k / <
**£#-
) 10 20 3
Hour of Day
0
— # — Off-Network and
Urban Roadtypes--
Weekend
— • — Off-Network and
Urban Roadtypes-
Weekday
Rural Roadtypes-
Weekend
— x— Rural Roadtypes-
Weekday
There is hourly VMT data available from Vehicle Travel Information System(VTRIS)
database maintained by FHWA that distinguishes hourly VMT by FHWA vehicle category.
Analysis of this data can provide HourVMTfractions differ by sourcetype. We request
comment on the relative priority of incorporating this data into MOVES defaults.
-------
13. Driving Schedule Tables
DriveSchedule refers to a second-by-second vehicle speed trajectory. Drive schedules
are used in MOVES to determine the operating mode distribution for most MOVES running
process emissions and energy consumption.
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, Draft MOVES2009 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 roadtype and sourcetype, based on the average speed
distribution. Briefly, for each speed bin in the speed distribution, the MOVES model selects the
two associated driving cycles with average speeds that bracket the speed bin's average speed.
The Vehicle Specific Power (VSP) distributions determined for each bracketing driving schedule
are averaged together, weighted by the proximity of the speed bin average speed to the driving
schedule average speeds. In this way, the VSP distribution of any roadtype's speed distribution
is determined from the available driving schedules. For more details, see the Operating Mode
Distribution Generator sections in the MOVES Software and Design Reference Manual.52
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 drive schedule name,
identification number, and the average speed of the drive schedule. DriveScheduleAssoc
defines the set of schedules which are available for each combination of source use type and road
type. This table also indicates which driving schedules describe freeway ramp type driving.
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. Note that the speed given in the drive schedule name is just a
nominal speed and not used in the MOVES calculations.
77
-------
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 1 (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)
15 mph (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 (3 51)
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
7.1
11.6
19.2
24.8
13.1
18.6
30.5
52.9
59.7
63.2
68.2
76
34.6
4.6
10.7
15.6
20.8
24.5
31.5
34.4
44.5
55.4
60.4
31
5.8
11.2
15.6
19.4
25.6
32.5
34.3
47.1
54.2
59.4
25.3
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 DriveScheduleAssociation, 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
Restricted Access Roadtypes
Light-Duty Freeway Schedules
Light-Duty Low Speed 1
New York City
Non-Freeway LOS EF
Medium Heavy -Duty Freeway
Medium Heavy -Duty Non-Freeway
Medium Heavy -Duty Freeway
Medium Heavy -Duty Non-Freeway
Heavy Heavy-Duty Freeway
Heavy Heavy -Duty Non-Freeway
Heavy Heavy-Duty Freeway
Heavy Heavy -Duty Non-Freeway
Unrestricted Access Roadtypes
Light-Duty Non-Freeway Schedules
Freeway LOS E
Freeway LOS D
Freeway LOS AC
Freeway High Speed 1
Freeway High Speed 2
Freeway High Speed 3
Freeway Ramp
Medium Heavy -Duty Freeway
Medium Heavy -Duty Non-Freeway
Bus Non-Freeway
Medium Heavy -Duty 50mph Freeway
Medium Heavy -Duty 60mph Freeway
MD Freeway Ramp
Refuse Truck Urban
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.001.53 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.54
High Speed 2 and 3 were developed specifically for MOVES. 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 improvements in vehicle
performance over the past decades dictate 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
79
-------
vehicle instrumented as part of EPA's On-Board Emission Measurement "Shootout" program,55
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 should be noted, however, that these schedules are only applied in Draft
MOVES2009 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 MOVES,
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."56 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, and 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 MOVES, 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 MOVES 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.
The freeway and non-freeway driving cycles are intended to cover most of the driving on
these respective roadtypes. However, some speed distributions for non-freeway roadtypes will
include average speeds faster than the fastest non-freeway cycles. The reverse will be true for
some freeway speed distributions. In these cases, the model will use appropriate average speed
drive schedules from a different roadtype. This mapping is summarized in table 14-1, which
illustrates, for example, that low-speed freeway driving is modeled using non-freeway driving
schedules. This mapping is appropriate since, when the average speed is very low or very high,
the roadtype has little impact on the driving pattern.
For Final MOVES2009 we plan to incorporate additional driving schedules and to
replace many of the older driving schedules, which we do not think adequately represent today's
vehicles and driving behavior. A contractor has developed 45 driving schedules for light-duty
vehicles.57 These are based on urban and rural data collected in California in 2000 and 2004.
The proposed mapping of driving cycles to roadtypes for Final MOVES2009 is summarized in
Table 14-2, below. This mapping would apply to passenger cars, passenger trucks and light
commercial trucks. We also hope to have additional driving schedules for motorcycles, but they
80
-------
are not available at this time. Other sourcetypes would use the driving schedules currently used
in Draft MOVES2009.
Table 14.2 Proposed Drive Schedules for Passenger Cars, Passenger Trucks
and Light Commercial Trucks in Final MOVES2009
Rural Restricted Access Roadtype (2)
ID
158
1009
1010
1011
1012
1013
1014
1015
1030
1031
1032
1033
102
101
Avgspeed
76.0
73.8
55.3
49.1
44.4
42.5
38.6
30.0
25.4
21.7
17.2
8.7
7.1
2.5
Road Classification
LD High Speed Freeway 3
Rural Interstate
Rural Principal Arterial
Rural Principal Arterial
Rural Major Arterial
Rural Minor Arterial
Rural Collector
Rural Local
Urban Principal Arterial
Urban Principal Arterial
Urban Principal Arterial
Urban Principal Arterial
LD New York City
LD Low Speed 1
DriveScheduleDesc
Final FC01LOSAF Cycle (C10R04-00854)
Final FC02LOSAC Cycle (C15R02-00646)
Final FC02LOSDF Cycle (C10R05-00513)
Final FC06LOSAF Cycle (C15R01-00276)
Final FC07LOSAF Cycle (C10R07-00913)
Final FC08LOSAF Cycle (C10R05-00330)
Final FC09LOSAF Cycle (C15R06-00563)
Final FC14LOSC Cycle (C10R04-00104)
Final FC14LOSD Cycle (C15R01-00836)
Final FC14LOSE Cycle (C15R03-00606)
Final FC14LOSF Cycle (C15R05-00424)
Urban Restricted Access Roadtype (4)
ID
158
1009
1023
1024
1025
1026
1014
1029
1035
1034
1028
1030
1036
1037
1040
1031
1041
1032
1043
1038
1044
1042
1039
1033
102
Avgspeed
76.0
73.8
66.4
63.7
52.8
43.3
38.6
31.0
29.5
26.6
25.5
25.4
23.3
21.9
21.8
21.7
18.6
17.2
15.7
14.6
12.0
11.2
10.5
8.7
7.1
Road Classification
LD High Speed Freeway 3
Rural Interstate
Urban Freeway
Urban Freeway
Urban Freeway
Urban Freeway
Rural Collector
Urban Principal Arterial
Urban Minor Arterial
Urban Minor Arterial
Urban Principal Arterial
Urban Principal Arterial
Urban Minor Arterial
Urban Minor Arterial
Urban Collector
Urban Principal Arterial
Urban Collector
Urban Principal Arterial
Urban Local
Urban Minor Arterial
Urban Local
Urban Collector
Urban Minor Arterial
Urban Principal Arterial
LD New York City
DriveScheduleDesc
Final FC01LOSAF Cycle (C10R04-00854)
Final FC12LOSB Cycle (C15R08-00003)
Final FC12LOSC Cycle (C15R04-00582)
Final FC12LOSD Cycle (C15R09-00037)
Final FC12LOSE Cycle (C15R10-00782)
Final FC08LOSAF Cycle (C10R05-00330)
Final FC14LOSB Cycle (C15R07-00177)
Final FC16LOSB Cycle (C15R03-00219)
Final FC16LOSA Cycle (C15R05-00755)
Final FC14LOSA Cycle (C15R03-00651)
Final FC14LOSC Cycle (C10R04-00104)
Final FC16LOSC Cycle (C15R05-00252)
Final FC16LOSD Cycle (C15R02-00561)
Final FC17LOSAC Cycle (C15R01-00333)
Final FC14LOSD Cycle (C15R01-00836)
Final FC17LOSD Cycle (C15R05-00480)
Final FC14LOSE Cycle (C15R03 -00606)
Final FC19LOSAC Cycle (C15R08-00267)
Final FC16LOSE Cycle (C15R05-00799)
Final FC19LOSDF Cycle (C15R03-00074)
Final FC17LOSEF Cycle (C15R02-00734)
Final FC16LOSF Cycle (C10R02-00249)
Final FC14LOSF Cycle (C15R05-00424)
81
-------
Table 14.2 Continued
Rural and Urban Unrestricted Access Roadtypes (3 & 5)
ID
158
1009
1017
1018
1024
1019
1022
1025
1020
1026
1014
1029
1030
1021
1027
1032
1033
102
101
Avgspeed
76.0
73.8
66.4
64.4
63.7
58.8
53.9
52.8
46.1
43.3
38.6
31.0
25.4
20.6
19.0
17.2
8.7
7.1
2.5
Road Classification
LD High Speed Freeway 3
Rural Interstate
Urban Interstate
Urban Interstate
Urban Freeway
Urban Interstate
Urban Freeway
Urban Freeway
Urban Interstate
Urban Freeway
Rural Collector
Urban Principal Arterial
Urban Principal Arterial
Urban Interstate
Urban Freeway
Urban Principal Arterial
Urban Principal Arterial
LD New York City
LD Low Speed 1
DriveScheduleDesc
Final FC01LOSAF Cycle (C10R04-00854)
Final FC11LOSB Cycle (C10R02-00546)*
Final FC11LOSC Cycle (C15R09-00849)
Final FC12LOSC Cycle (C15R04-00582)
Final FC11LOSD Cycle (C15R10-00068)
Final FC12LOSA Cycle (C15R02-00501)
Final FC12LOSD Cycle (C15R09-00037)
Final FC11LOSE Cycle (C15R1 1-00851)
Final FC12LOSE Cycle (C15R10-00782)
Final FC08LOSAF Cycle (C10R05-00330)
Final FC14LOSB Cycle (C15R07-00177)
Final FC14LOSC Cycle (C10R04-00104)
Final FC11LOSF Cycle (C15RO 1-00876)
Final FC12LOSF Cycle (C15R08-00294)
Final FC14LOSE Cycle (C15R03 -00606)
Final FC14LOSF Cycle (C15R05-00424)
*1009 was originally characterized as LOSE, but now considered AB.
82
-------
15. SourceTypeHour
The SourceTypeHour table provides one data field: IdleSHOFactor.
15.1. IdleSHOFactor
The IdleSHOFactor field is the number 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. In Draft MOVES2009, 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.
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 2004 paper by
Lutsey, et al., 58 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.59 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 of 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,"60
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. Weekday data was used for both weekday and
weekend fractions.
-------
Extended Idle Activity
Ratio of Extended Idle Time to Driving Time by Hour
n r\A _,
o ms -
n m
2 o OPS -
u! n n?
? n ni5
*2 n m
— n oo^
n
*-«, *-*"*""'
^S. X^
V S
N^ X
^^-^ _^
^^^— A A A^*r--'^
»
0 4 8 12 16 20 24
Hour
Note that the Draft MOVES2009 defaults assume no anti-idling measures or truck-stop-
electrification efforts. In future versions of MOVES, we intend to make it easier for users to
modify the inputs of extended idling behavior to account for new or locally available data on
such activity.
-------
i6. ZoneRoadType
The SHOAllocFactor field is used to determine the hours of vehicle operation in each
zone on each of the MOVES roadway types.
While geographic allocations clearly change over time, for national runs using Draft
MOVES2009 this table is used for all calendar years. Note that the allocation factors are not
used when a user selects the "County" scale. At the "National" scale, users may choose to do
multiple runs, with year-specific factors entered for each specific calendar year run.
The spatial allocation of source hours operating distributes the domain-wide estimates of
hours of operation to the zones. In draft MOVES2009, the default domain is the nation and the
zones are counties. The national source hours of operation (SHO) are calculated from estimates
of VMT and speed.
The estimate for the VMT by county comes from the 1999 National Emission Inventory
(NEI) analysis documented by Pechan & Associates.61 These estimates are based on the
Highway Performance Monitoring System (HPMS) data collected by the Federal Highway
Administration62 for use in transportation planning and vehicle type breakdowns from the EPA
MOBILE6 Emission Factor model.63 The NEI VMT estimates were incorporated into the
National Mobile Inventory Model (NMIM) county database.
To calculate default inputs for Draft MOVES2009, the 1999 NEI VMT estimates were
obtained from the NMEVI database for each county and HPMS facility type. The average speed
estimates were taken directly from Table 8 of the NEI documentation. VMT estimates for each
MOVES road type(i) were determined for each county(j) in the nation and the allocation was
calculated using the following formula, where k refers to the HPMS facility types within a
MOVES road type, and m refers to the VMT for each source type.
County Allocation(ij) = (Sum(j)(( County VMT (i,j,k,m)/Average Speed(k,m))) /
(Sum(ij)((CountyVMT(i,j,k,m)/AverageSpeed(k,m)))
The county allocation values for each roadway type sum to one for the nation. Although
the data is from 1999 calendar year estimates, the same allocations are used for all calendar
years.
-------
17. Zone
In Draft MOVES2009, activity data and meteorological data are assigned to zones rather
than counties. By creating and populating their own zones, users may customize geographical
boundaries to better represent non-attainment areas and climate differences that do not
necessarily follow county boundaries. However, for the national default database, zones and
counties are equivalent.
The Zone table provides values for four fields: CountylD, StartAllocFactor,
IdleAllocFactor, and SHPAllocFactor. CountylD is the identifier for the county in which the
zone is located. StartAllocFactor geographically allocates domain-wide start activity.
IdleAllocFacor allocates extended idle activity, and SHPAllocFactor allocates time parking
(important for evaporative emissions).
While geographic allocations clearly change over time, for national runs using Draft
MOVES2009 this table is used for all calendar years. Note that the allocation factors are not
used when a user selects the "County" scale. At the "National" scale, users may choose to do
multiple runs, with year-specific factors entered for each specific calendar year run.
17.1. StartAllocFactor
The StartAllocFactor distributes the domain-wide estimates of the number of trip starts to
the zones. In the default database for Draft MOVES2009, the domain is the nation and the zones
are counties. There is no national data on the number of trip starts by county, so for Draft
MOVES2009, we have used VMT will to determine this allocation.
The estimate for the VMT by county comes from the 1999 National Emission Inventory
(NEI) analysis.64 The NEI estimates are based on the Highway Performance Monitoring System
(HPMS) data collected by the Federal Highway Administration65 for use in transportation
planning and vehicle type breakdowns from the EPA MOBILE6 Emission Factor model.66 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, where "i"
represents each individual county.
CountyAllocation(i) = ( County VMT(i) / Sum(CountyVMT(i) )
The county allocation values 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
86
-------
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.67 These estimates were used to
determine the allocation to each State(i) using the following formula:
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.68 The NEI estimates are based on the
Highway Performance Monitoring System (HPMS) data collected by the Federal Highway
Administration69 for use in transportation planning and vehicle type breakdowns from the EPA
MOBILE6 Emission Factor model.70 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 where "j" refers to the counties in
each particular state.
IdleAllocFactor(i) = StateAllocation(i) * (CountyVMT(j) / Sum(CountyVMT(j))
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.
-------
18. SCC Mappings
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 mapping from
the MOVES roadtype on which the emissions occur to the HPMS Facility Type used in the SCC
codes. This mapping is captured in the SCCRoadTypeDistribution table described below.
18.1. SCCVtypeDistribution
Because the existing SCCs only identify gasoline and diesel-fueled vehicles, it was
necessary to map alternatively-fueled vehicles to one of these categories. All alternative-fuel
vehicles were mapped to the diesel SCC. In the future, SCCs may be revised to explicitly handle
alternative fuels.
For most SourceTypes, the mapping to SCCVtype 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 VIUS 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.71 The resulting truck mappings are too complex to summarize here, but are
available in the MOVES database.
88
-------
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
18.2. SCCRoadTypeDistribution
Each SCC includes a suffix that indicates the HPMS Facility Class on which the
emissions occur. Because MOVES calculations are done for MOVES roadtypes, the
SCCRoadTypeFraction provides an allocation of emissions on each MOVES roadtype to the
appropriate SCCRoadTypes.
Table 18-1. SCC RoadTypes
SCCRoadTypelD
11
13
15
17
19
21
23
25
27
29
31
33
1
SCCRoadTypeDesc
Rural Interstate
Rural Principal Arterial
Rural Minor Arterial
Rural Major Collector
Rural Minor Collector
Rural Local
Urban Interstate
Urban Freeway/Expressway
Urban Principal Arterial
Urban Minor Arterial
Urban Collector
Urban Local
Off-Network
Because roadtype distributions vary geographically, the mapping of MOVES roadTypes
to SCCRoadTypes varies by zone (in this case, county). For SCCRoadTypeDistribution we
determined the proportion of hours of operation on a given MOVES roadtype within a county
that occurred on each SCCRoadType. Hours of operation were estimated by dividing the 1999
National Emission Inventory (NEI) VMT by the 1999 NEI average speed. Both measures were
documented by Pechan & Associates.72 The NEI VMT estimates are based on the Highway
Performance Monitoring System (HPMS) data collected by the Federal Highway
-------
Administration73 for use in transportation planning and vehicle type breakdowns from the EPA
MOBILE6 Emission Factor model.74 The VMT estimates were obtained from the NMIM
database for each county and HPMS facility type. The average speed estimates are taken from
Table 8 of the NEI documentation.
The SCCRoadType fractions were calculated using the following formula, where i refers
to the county, j refers to the MOVES roadtype, k refers to the SCCRoadType within a MOVES
road type, and m refers to the VMT for each source type.
SCCRoadTypeFraction(i,j,k) = Sum(j,j,k)( VMT(k,m)/Average Speed(k,m)) /
Sum(i,j)((VMT(k,m)/AverageSpeed(k,m))
In cases where a county had no VMT for a given roadtype, the average values were used.
The SCCRoadTypeFraction for OffNetwork travel was set to 1 (mapping all "off-network"
emissions to this new roadtype. The SCCRoadType fractions for each roadway type will sum to
one for each county. Although the data is from 1999 calendar year estimates, the same
allocations will be used for all calendar years.
90
-------
19. MonthGroupHour
AC Activity Terms A, B and C are coefficients for a quadratic equation that calculates air
conditioning activity demand as a function of the heat index. These terms 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."75 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 Draft MOVES2009, the default data uses one set of coefficients for all
MonthGroups and Hours. These default coefficients represent an average A/C activity demand
function over the course of a full day. The coefficients are listed in Table 19.1.
Table 19-1. Air Conditioning Activity Coefficients
A
-3.63154
B
0.072465
C
-0.000276
The A/C activity demand function that results 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
70 75 80 85 90 95
Heatlndex(F)
100
105
110
-------
2O. Sample Trip Data
To estimate start and evaporative emissions, it is important to estimate the number of
starts by time of day, and the duration of time between vehicle trips. (This between-trip duration
is often called "soak time." To determine typical patterns of trip starts and ends, MOVES uses
information from instrumented vehicles. This data is stored in two tables: SampleVehicleDay
and SampleVehicleTrip.
The first table, SampleVehicleDay, lists a "sample population" of vehicles, each with an
identifier (vehID), an indication of vehicle type (sourceTypelD), and a "dayID" that indicates
whether the vehicle is part of the weekend or weekday vehicle population.
The second table, SampleVehicleTrip, lists the trips made by each of these vehicles. It
records the vehID, day ID, a trip number (tripID), the hour of the trip (hourlD), the trip number of
the prior trip (priorTripID), and the times at which the engine was turned on and off for the trip
(keyOnTime and keyOffTime, each recorded in minutes since midnight of the day of the trip).
To account for overnight soaks, many first trips reference a prior trip with a null value for
keyOnTime and a negative value for keyOffTime. And, to account for vehicles that sit for one
or more days without driving, the SampleVehicleDay table includes some vehicles that have no
trips in the SampleVehicleTrip table.
The data and processing algorithms used to populate these tables are detailed in two
contractor reports.76'77 The data comes from a variety of instrumented vehicle studies,
summarized in Table 20.1. This data was cleaned, adjusted, sampled and weighted to develop a
distribution intended to represent average urban activity across the U.S. For vehicle classes that
were not represented in the available data, the contractor synthesized trips using trip-per-
operating hour information from MOBILE6 and soak time and time-of-day information from
sourcetypes that did have data. The application of synthetic trips is summarized in Table 20.2.
The resulting trip per day estimates are summarized and compared to MOBILE6 in Table 20.3.
Table 20.1. Source Data for Sample Vehicle Trip Information
Study
3-City
Minneapolis
Knoxville
Las Vegas
Battelle
TxDOT
Study Area
Atlanta, GA;
Baltimore, MD;
Spokane, WA
Minneapolis/St.
Paul, MN
Knoxville, TN
Las Vegas, NV
California,
statewide
Houston, TX
Study Years
1992
2004-2005
2000-2001
2004-2005
1997-1998
2002
Vehicle Types
Passenger cars &
trucks
Passenger cars &
trucks
Passenger cars &
trucks
Passenger cars &
trucks
Heavy duty trucks
Heavy, heavy duty
diesel dump trucks
Number of
Vehicles
321
133
377
350
120
4
92
-------
Table 20.2. Synthesis of Sample Vehicles for Source Types Lacking Data
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
Based on Direct
Data?
No
Yes
Yes
No
No
No
No
No
Yes
No
No
Yes
Yes
Synthesized From
Passenger Cars
n/a
n/a
Passenger Trucks
Combination long-haul trucks
Single -unit short-haul trucks
Single -unit short-haul trucks
Combination short-haul trucks
n/a
Combination long-haul trucks
Passenger Cars
n/a
n/a
Table 20.2. Starts per Da^
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
/ by SourceType
Draft MOVES2009
Weekday
0.78
5.89
5.80
6.05
2.77
4.58
5.75
3.75
6.99
4.29
0.57
5.93
4.29
Draft MOVES2009
Weekend
0.79
5.30
5.06
5.47
0.88
3.46
1.26
0.92
1.28
1.29
0.57
1.16
1.29
MOBILE6*
1.35
6.75
7.38
7.38
6.88
6.88
6.88
6.88
6.88
6.88
6.88
6.88
6.88
* Note, MOBILE6 distinguished "starts" and "trips."
include some very short "trips."
MOVES does not, but MOVES does
93
-------
21. References
1 U.S. Census Bureau, 1997 Vehicle Inventory and Use Survey, CD-EC97-VIUS. January 2000.
Online at www.census.gov/prod/www/abs/vius-pdf.html
2 2 U.S. Census Bureau, 2002 Vehicle Inventory and Use Survey. Online at
www.census.gov/svsd/www/vius/2002.html
3 U.S. Census Bureau, 1992 Truck Inventory and Use Survey. Online at
www.census.gov/svsd/www/92vehinv.html
4 R.L. Polk & Co., National Vehicle Population Profile.® Southfield, MI. 1999. Information
online at http://usa.polk.com/Products/l_nvpp.htm.
5 R.L. Polk & Co, Trucking Industry Profile TIP® Vehicles in Operation. Southfield, MI. 1999.
Information online at http://usa.polk.com/Products/14_tipnet.htm.
6 U.S. Federal Highway Administration. Highway Statistics, 1999. Table MV-1, "State Motor
Vehicle Registrations," October 2000. Online at
www.fhwa.dot.gov/ohim/hs99/index.htm
7 U.S. Federal Highway Administration. Highway Statistics, 1999. Table MV-10, "Bus
Registrations,"October 2000. Online at www.fhwa.dot.gov/ohim/hs99/index.htm
8 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 www.fhwa.dot.gov/ohim/hs99/index.htm
9 U.S. Federal Highway Administration. Highway Statistics, 1999. Table VM-2, "Functional
System Travel," January 2002. Online at www.fhwa.dot.gov/ohim/hs99/index.htm
10 U.S. Federal Transit Administration. National Transit Database 1999, Table 29. "Age
Distribution of Active Revenue Vehicle Inventory: Details by Transit Agency." Online at
www. ntdprogram. com
11 Bobit Publications, School Bus Fleet Fact Book. Torrance. CA, 1999.
www. schoolbusfleet. com
12 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
www.epa.gov/otaq/models/mobile6/m6flt002.pdf
94
-------
13 Energy Information Adminstration. Annual Energy Outlook 2003 (AEO2003\ Report #:
DOE/EIA-0383 (2003), released January 9, 2003. Online at
www.eia.doe.gov/oiaf/archive/aeo03/index.html
14 Energy Information Administration, Supplemental Tables to the Annual Energy Outlook 2006,
Transportation Demand Sector, February 2006. Online at
www.eia.doe.gov/oiaf/archive/aeo06/supplement/index.html
15 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.
16 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.
17 Davis, Stacy C., Susan W. Diegel and Robert G. Boundy, Transportation Energy Data Book,
Edition 27. Center for Transportation Analysis, Oak Ridge National Laboratory. ORNL-
6991. 2008. Onlinewww-cta.ornl.gov/data/
18 Ward's Automotive Inc. www.wardsauto.com/
19 Hart, Larry. R.L. Polk & Company. Personal communication, June 16, 2003.
20 National Household Transportation Survey (NHTS). 2001 Online at
http: //nhts. ornl. gov/downl oad. shtml
21
22
Motorcycle Industry Council, 1998 Population. Available in EPA Docket A-2000-01, IIB-22
American Bus Association. "Motorcoach Census 2000," conducted by R. L. Banks and
Associates, Inc. July 2000. www.buses.org/industry/ABA-RLBanksReport.pdf
23 American Public Transportation Association, 2007 Public Transportation Fact Book as cited in
TEDB 27, Table 5.13 (page 5-20).
24 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.
www.mic.org/
25 John Koupal. M6.ACE.001 "Air Conditioning Activity Effects in MOBILE6," EPA420-R-01-
054, November 2001. www.epa.gov/otaq/models/mobile6/r01054.pdf
26 Motorcycle Industry Council, "On-Highway Motorcycles 1998 Population Estimate."
November 21, 1999. Available in the U.S. EPA docket: A-2000-01, II B-22.
27 NHTS A. "Vehicle Survivability and Travel Mileage Schedules," DOT HS 809 952, January
2006.www-nrd.nhtsa.dot.gov/Pubs/809952.PDF
95
-------
28 NHTSA, 2006.
29 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.
30 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.
31 American Bus Association, July 2000.
32 Good Sam Club, "Highways Member Study 2000." TL Enterprises, Inc., Ventura, California.
(805) 667-4100.
33 Koupal, November 2001.
34 Davis and Diegel, 2007
35 Electric Drive Association, http://www.electricdrive.org/
36 Davis and Truitt, March 2002.
37
Davis and Truitt, March 2002.
38 Union of Concerned Scientists, www.ucsusa.org/
39
U.S. Federal Transit Administration, December 2003.
40 U.S. Federal Transit Administration, December 2003.
41 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.
42 Brian, Mac. Recreational Vehicle Industry Association. Phone conversation, October 29,
2003.
43 Nam, Edward and Robert Giannelli, "Fuel Consumption Modeling of Conventional and
Advanced Technology Vehicles in the Physical Emission Rate Estimator (PERE),"
EPA420-P-05-001, February 2005. Available online at
www.epa.gov/otaq/models/ngm/420p05001.pdf
-------
44 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.
www.epa.gov/epahome/cfr40.htm
45 USEPA. "IM240 and Evap Technical Guidance," EPA420-R-00-007, April 2000. Online at
www.epa.gov/otaq/regs/im/r00007.pdf
46 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.
47 Petrushov, V.A., "Coast Down Method in Time-Distance Variables," SAE 970408, February
24, 1997. www.sae.org/
48 U.S. Environmental Protection Agency, "Draft Motor Vehicle Emission Simulator (MOVES)
2009 Software Design and Reference Manual," EPA420-b-09-007, March 2009. Online at
www.epa.gov/otaq/models/moves/420b09007.pdf
49 Sierra Research, Inc. Memo from Tom Carlson to John Koupal, "Analysis of Rural Average
Speed Distributions for MOVES," Purchase Order EP05B00129, December 1, 2004.
50 Sierra Research, Inc. Memo from Tom Carlson to John Koupal, "Analysis of Rural Average
Speed Distributions for MOVES," Purchase Order EP05B00129, December 1, 2004
51 Festin, Scott. "Summary of National and Regional Travel Trends: 1970-1995," Office of
Highway Information Management, Dept. of Transportation, May 1996. Online at
www.fhwa.dot.gov/ohim/bluebook.pdf
52 U.S. EPA, March 2009.
53 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
www.epa.gov/otaq/models/mobile6/m6spd001.pdf
54 USEPA Combined Federal Register. CFR 40, 86, Appendix I.
www.epa.gov/epahome/cfr40.htm
55 Hart, Constance. "EPA's Onboard Analysis Shootout: Overview and Results." EPA420-R-02-
026, October 2002. Online at
www.epa.gov/otaq/ngm.htm
56 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.
97
-------
57 Sierra Research, "Development of Generic Link-Level Driving Cycles." SR2009-05-02 EPA
Contract EP-C-05-037, Work Assignment 3-02, May 5, 2009.
58 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.
59 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
www.fmcsa.dot.gov/Pdfs/050200p.pdf
60 Eastern Research Group, August 2003.
61 Pechan, E.H. & Associates, Inc. "Documentation for the Onroad National Emissions
Inventory (NEI) For Base Years 1970-2002," prepared for EPA Office of Air Quality
Planning and Standards, January 2004. Online at
ftp://ftp.epa.gov/EmisInventory/2002fmalnei/documentation/mobile/onroad_nei_basel970
_2002.pdf.
62 U.S. Federal Highway Administration (FHA). Highway Performance Monitoring System
Field Manual. OMB No. 21250028, December, 2000.
63 Jackson, September 2001.
64 .Pechan & Associates, Inc. October 2002.
65
66
U.S. Federal Highway Administration, December 2000.
Jackson, September 2001.
67 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 www.tfhrc.gov/safety/pubs/011587
68 Pechan & Associates, Inc. October 2002.
69 U.S. Federal Highway Administration, December 2000.
70 Jackson, September 2001.
71 Davis and Truitt, March 2002.
72 Pechan, E.H. & Associates, Inc., January 2004.
-------
73 U.S. Federal Highway Administration (FHA). Highway Performance Monitoring System
Field Manual. OMB No. 21250028, December, 2000. Online at
www.fhwa.dot.gov/ohim/hpmsmanl/hpms.htm
74
75
Jackson, September 2001.
Koupal, November 2001.
76 Sierra Research, "Development of Trip and Soak Activity Defaults for Passenger Cars
andTrucks in MOVES2006," SR2006-03-04, EPA Contract EP-C-05-037, Work
Assignment No. 0-01, March 27, 2006.
77 Sierra Research, "Development of Trip and Soak Activity Defaults for Passenger Cars and
Trucks in MOVES," SR2007-06-01, EPA Contract EP-C-05-037, Work Assignment No. 1-
01, June 29, 2007.
99
------- |