vvEPA
United States
Environmental Protection
Agency
Air and Radiation EPA420-R-03-017
April 2003
National Mobile Inventory
Model (NMIM) Base and
Future Year County Database
Documentation and Quality
Assurance Procedures
-------
EPA420-R-03-017
April 2003
National Mobile Inventory Model (NMIM) Base and
Future Year County Database Documentation and
Quality Assurance Procedures
Assessment and Standards Division
Office of Transportation and Air Quality
U.S. Environmental Protection Agency
Prepared for EPA by
Eastern Research Group, Inc.
EPA Contract No. 68-COO-l 12
Work Assignment No. 3-05
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 which
may form the basis for a final EPA decision, position, or regulatory action.
-------
TABLE OF CONTENTS
Page
1.0 INTRODUCTION 1-1
2.0 REFERENCE TABLES 2-1
2.1 DataSource 2-1
2.2 HPMSRoadType 2-2
2.3 M6VClass 2-3
2.4 M6VType 2-5
3.0 FUEL TABLES 3-1
3.1 Diesel 3-1
3.2 GasMTBEPhsOut 3-2
3.3 Gas2MTBEPhsOut 3-39
3.4 Natural Gas 3-40
3.5 CountyYearMonth 3-41
3.6 Fuel Tables Required to Model No MTBE Phase Out Scenario 3-42
3.6.1 Gasoline 3-42
3.6.2 Gasoline2 3-43
3.6.3 CountyYearMonth 3-44
4.0 VEHICLE TABLES 4-1
4.1 AverageSpeed 4-1
4.2 BaseYearVMT 4-2
4.3 VMTGrowth 4-4
4.4 VMTMonthAllocation 4-15
5.0 INSPECTION AND MAINTENANCE (I/M) PROGRAM TABLES 5-1
5.1 County Year 5-1
6.0 ADDITIONAL TABLES 6-1
6.1 County 6-1
6.2 CountyMonth 6-4
6.3 State 6-5
7.0 INTERNAL QUALITY ASSURANCE TABLES 7-1
7.1 Minimum and Maximum Field Values 7-1
7.2 Null Values 7-1
7.3 Zero Values 7-1
7.4 Table Relationships 7-2
8.0 REFERENCES 8-1
APPENDIX A Index to Data Files Available Electronically
-------
LIST OF TABLES
Page
2-1 HPMSRoadtype Values 2-3
2-2 M6VClass Values 2-4
2-3 M6VType Values 2-6
3-1 Diesel Sulfur Values 3-2
3-2 Sample Calculation for Composited Seasonal Fuel for FIPS 39001:
Adams, OH 3-34
3-3 ATSM RVP Class Assignment for FIPS 39001: Adams, OH 3-35
3-4 Monthly Interpolation Factor Calculation for FIPS 39001: Adams, OH ... 3-36
3-5 Sample Monthly Interpolation for Olefins Calendar Year 1999 for
FIPS 39001: Adams, OH 3-37
3-6 Natural Gas Sulfur Values 3-40
4-1 National-Average VMT Fraction by Road Type Used for California
VMT Data 4-3
4-2 Post-fixes to Vehicle Class Conversions 4-9
4-3 Original and Adjusted VMTMonthAllocation Values 4-16
4-4 Conversion of Roadway and Vehicle Types for VMTMonthAllocation
Data 4-18
5-1 County Year Data Sources 5-1
5-2 County FIPS Codes in NEI Stage 2 Refueling Data Not Used in NMEVI 5-3
5-3 Examples of Differences in I/M Program Data 5-7
5-4 New External EVI Program Files 5-8
11
-------
LIST OF FIGURES
Page
1-1. Data Relationship Diagram 1-3
4-1. VMT Growth Data Sources and Methods 4-6
4-2. Anchor Years and Interpolation Spans for VMT Growth 4-12
4-3. Percentage Growth Rate for the VMT Growth Table 4-13
in
-------
i.o INTRODUCTION
To keep pace with new analysis needs, new modeling approaches, and new data,
the EPA's Office of Transportation and Air Quality (OTAQ) is currently working the Multi-scale
MOtor Vehicles and equipment Emission System (MOVES). MOVES will estimate emissions
for on-road and off-road sources, cover a broad range of pollutants, and allow multiple scale
analysis, from fine-scale analysis to national inventory estimation. When fully implemented,
MOVES will replace both MOBILE6 and NONROAD. MOVES will not necessarily be a single
piece of software, but instead will encompass the tools, algorithms, underlying data and guidance
necessary for use in all official analyses associated with regulatory development, compliance
with statutory requirements, and national/regional inventory projections. Additional detail on
EPA's MOVES program can be found at http://www.epa.gov/otaq/ngm.htm.
EPA's National Mobile Inventory Model (NMEVI) is an interim product
supporting creation of MOVES. NMEVI combines mobile sources emission factor modeling with
area-specific data to produce national emission inventories at county level using MOBILE6.3 and
NONROAD. NMIM inventories will support EPA regulatory analysis and policy setting
activities.
To support development of NMEVI, ERG created and populated a data set that
contains the area-specific county-level data required for emissions inventory modeling. There are
two distinct components of this data: complete "baseline" data for 1999, and the future-year
(post-1999) data to project beyond the baseline. As an interim product, NMEVI implements some
MOVES architecture features. Specifically, the NMEVI database is based on the MySQL open-
source database management system, and the Java language is used as appropriate for software
components. The NMEVI data set ERG produced includes a MySQL-based database and also a
set of non-database data files (primarily MOBILE6 input files). This report documents the
development of a data set that contains the area-specific county-level data required for emissions
inventory modeling, including "baseline" data for 1999, and the future-year (post-1999) data.
Figure 1-1 presents a data relationship diagram illustrating the data set.
1-1
-------
This report is organized as follows:
Section 2.0: Reference Tables;
Section 3.0: Fuel Tables;
Section 4.0: Vehicle Tables;
Section 5.0: Inspection and Maintenance (I/M) Tables;
Section 6.0: Additional Tables; and
Section 7.0: Internal QA/QC Tables.
1-2
-------
COUNTYDB20020904 - Display! / •--Ktoin Subject Area:-
NMIM County Database
Physical Design -
9/4/2002
Slate
Not9 that all SMALLINTattribut9S (except Year. BaseYear. NRRepFIPScnlyld. HivyRepFIPSCntyld.
FlPSCountyld, and all Data Source related attributes) are actually TINYINT and will be implemented assuch in
MySQL (This is because ERwin does not support TINYINT, and SMALLINT is used herein instead)
FlPSSIatett SMALLINT
Rtn
StndsDalaSource: SMALLINT
NLEVFileName: CHARM 2i
StateAbbr: CHAR(2)
SialeName:CHAR(25:i
'
County
FlPSCountyld: SMALLINT
RPSStateld: SMALLINT
Altitude: CHAW 11
CountyName: CHARiK)
RPSSlateld: SMALLINT
NRRepFIPSCntyld: SMALLINT
HwyRepRPSCntyld: SMALLINT
:
AverageSpeed
1 FtoadType: SMALLINT |
VType: SMALLINT
DaUScurce: SMALLINT I
AvgSpeed: NUMBER[21 !
BaseYearVMT
BaseYear: SMALLINT
^Class: SMALLINT
RoadType: SMALLINT
FlPSCcurrtyld: SMALLINT
FlPSStateld: SMALLINT
DataScurce: SMALLINT
I'MT: NUMBER.; 10.4)
MSVType
VTypa: SMALLINT
VTypeCesc: CHHR(80)
HPMSRoadType
RoadType: SMALLINT
RoadDesc: CHAR(SO)
i——
.INT
M |
VMTMonthAllocation
'"tfType: SMALLINT
RoadType: SMALLINT
Montti: SMALLINT
DalaSouree: SMALLINT
AllocFaclcr: NUMBER(21
Co LI nty Month
FlPSCountyld: SMALLINT
FlPSSlateld: SIvWLLINT
Month: SMALLINT
TempDataSourcs: SMALL) NT
MaxTemp: NUMBER(4,1)
MinTfmp:NUMBER(4,i;i'
AvgTerap:NUMBER(4.1)
-
CountyYear
Year: SMALLINT
FlPSCountyld: SMALLINT
FlPSStaled: SMALLINT
DataScurce: SMALLINT
Stage2Pct NUMBER(4.1)
ATPFileNaire:CHARi;i2}
IMFileName: CHAR(12)
VMTGrowth
Year: SMALLINT
VCIaas: SMALLINT
FlPSCounlyld: SMALLINT
FlPSStateld: SMALLINT
DataScurce: SMALLINT
VMTGiowlhRale: NUMBERi:4,1j
I I
M6VCIass
vaase: SMALL INT
VCIassAbtr: CHAR(S)
VCIassDeso: CHAR(SO)
CountyYearMonth
12
FlPSCountyld: SMALLINT
FlPSStaleld: SMALLINT
Year: SMALLINT
Monlh: SMALLINT
NGId: SMALLINT
HwyDieselld: SMALLINT
H»yGa5dineld: SMALLINT
NRGasolineld: SMALLINT
NRDieselld: SMALLINT
HwyFuelDataScurce: SMALLINT
NRFuelDataSouroe: SMALLINT
HwyGaEdinedA: SMALLINT
HwvGaidinekJB: SMALLINT
'
' C
NaturalGas
NGId: SMALLINT
NGSuHur:NUMBER(t
Gasoline2
Gasdireld: SMALLINT
MktShars: NUMBER(3,2)
RVP: NUMBER(3,1)
GaiSulfur: NUMBER(6,2)
GasMaxSullur: NUMBER-6.2.
RVPOxyWaiverNUMBERn.i
ETOHVolume: NUMBERfSil)
ToElhsrVdu me: NUMBERi'3.1)
MTBEVolutTB:NUr.1BER(3.11
ETBEVdume:NUMBERf3.1i
TAMEVolume: NUMBER(3.1)
AiomatioContenl: NUMBER(3.1)
Olefi[Content NUMBER(3.1)
BenzeneContent NUMBERS,1)
E200: NUMBER(3.1i
E300:NUMBER(4,1'i
RFG:CHAR(1) '
1(6,2) [
Diesel
Dieselld: SMALLINT
DieselSulfur: NUMBER(6,21
Gasoline
Gasolineld: SMALLINT
I
DataSource
DataSoureeld: SMALLINT
Author: CHAR(2S)
Dale: DATE
Sponsor: CHAR(301
DocumenlU: CHAR(30)
QualityLevehCHARfl)
RVP: NUMBERl'3,1)
GasSu»ur:NUMBER(6,2)
GasMaxSulfur:NUMBER.:i.2.
RVPOxyWaiver NUMBERMj
ETOHVdume: NUMBER(3',1)
ETOHMklShare: NUMBER! 3.2.
ToŁtheiVolurae: NU MBERft. 11
ToEtheiMklShare: NU MBERi'3,2)
MTBEVdume: NUMBER(3.11
MTBEMWShare: N UM BBV3.3)
ETBEVolume: NUMBER(J.I)
ETBEMklShare: NUMBER(3,2)
TAMEVdume:NUMBER(3,1)
TAMEMktShare: N UM BERl 3.2'.
AromalioContent: NUMBER(H)
OlefinContenl: NUMBERf3,1)
BemeneContent:NUMBERi2.1i
E200:NUMBER(3,11
E3]0:NUMBER(4,1J
RFC: CHAR(1)
1,1 /1.1 -11:02:49 AM. &E..-02
Figure 1-1. Data Relationship Diagram
1-3
-------
2.0 REFERENCE TABLES
NMIM includes a series of reference, or look-up, tables that contain definitions of
codes used in certain fields in other tables. Each of these reference tables are described below.
2.1 DataSource
The DataSource table contains reference information for the documents, data
bases, and other sources of information used to populate NMIM data tables.
Data Source
Multiple references were reviewed to gather the information used to populate
NMIM tables. DataSource provides additional detail associated with each reference.
Data Population Methodology
The DataSource table was populated manually as each NMIM table was added to
the database.
Quality Assurance Procedures
The contents of the DataSource field in each table were visually compared with
the contents of the DataSourcelD field in the DataSource table verify that all sources were
documented. In addition, the null value, zero value, maximum and minimum value, parent-child,
and child-parent QA/QC checks described in Section 7.0 were also completed for this table.
2-1
-------
2.2 HPMSRoadType
The HPMSRoadType table contains the 12 Highway Performance Monitoring
System (HPMS) roadway types and the unique identifier assigned to each type.
Data Source
The HPMS roadway types were extracted from the vehicles2.xls file from the
June 2002 National Emissions Inventory (NEI) update files.
Data Population Methodology
The data were exported from Microsoft Excel to a comma-separated value (csv)
file. The csv file was then imported into the NMEVI database. The original spreadsheet included
3-digit roadway types rather than 2-digit roadway types. This 3-digit roadway type was
converted to the corresponding 2-digit roadway type before being imported into the NMEVI
database.
Quality Assurance Procedures
ERG compared the 12 HPMS roadway types against the corresponding portion of
SCC codes retrieved from http://www.epa.gov/ttn/chief/codes/index.html where SCC = [XX-
AA-BBB-CC-D] and CC = HPMS roadway type. In addition, the null value, zero value,
maximum and minimum value, parent-child, and child-parent QA/QC checks described in
Section 7.0 were also completed for this table.
Table 2-1 presents the 12 roadway types contained in the HPMSRoadtype table.
2-2
-------
TABLE 2-1
HPMSRoadtype Values
Roadway Type ID
11
13
15
17
19
21
23
25
27
29
31
33
Roadway Type Description
Interstate: Rural
Other Principal Arterial: Rural
Minor Arterial: Rural
Major Collector: Rural
Minor Collector: Rural
Local: Rural
Interstate: Urban
Other Freeways and Expressways: Urban
Other Principal Arterial: Urban
Minor Arterial: Urban
Collector: Urban
Local: Urban
2.3
M6VClass
The M6VClass table contains the 28 vehicle classes used in MOBILE6 and the
unique identifier assigned to each.
Data Source
Vehicle classes were obtained from Section 1.2.3 of the MOBILE6 User Guide
(EPA 420-R-02-028, October 2002), also available from http://www.epa.gov/otaq/m6.htm.
Data Population Methodology
Data were entered in a Microsoft Excel spreadsheet and then exported to a csv
file. The csv file was then imported into the NMEVI database.
2-3
-------
Quality Assurance Procedures
The contents of M6VClass were printed and visually compared to the list of
MOBILE6 vehicle classes from the MOBILE6 User's Guide. In addition, the null value, zero
value, maximum and minimum value, parent-child, and child-parent QA/QC checks described in
Section 7.0 were also completed for this table.
Table 2-2 presents the 28 vehicle classes contained in the M6VClass table.
TABLE 2-2
M6VClass Values
Vehicle
Class ID
1
2
o
6
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Vehicle Class
Abbreviation
LDGV
LDGT1
LDGT2
LDGT3
LDGT4
HDGV2B
HDGV3
HDGV4
HDGV5
HDGV6
HDGV7
HDGV8A
HDGV8B
LDDV
LDDT12
HDDV2B
HDDV3
HDDV4
HDDV5
HDDV6
HDDV7
Vehicle Class Description
Light-Duty Gasoline Vehicles (Passenger Cars)
Light-Duty Gasoline Trucks 1 (0-6,000 Ibs. GVWR, 0-3750 Ibs. LVW)
Light-Duty Gasoline Trucks 2 (0-6,000 Ibs. GVWR, 3751-5750 Ibs. LVW)
Light-Duty Gasoline Trucks 3 (6,001-8,500 Ibs. GVWR, 0-5750 Ibs. ALVW)
Light-Duty Gasoline Trucks 4 (6,001-8,500 Ibs. GVWR, 5751 Ibs. and greater
ALVW)
Class 2b Heavy-Duty Gasoline Vehicles (8501-10,000 Ibs. GVWR)
Class 3 Heavy-Duty Gasoline Vehicles (10,001-14,000 Ibs. GVWR)
Class 4 Heavy-Duty Gasoline Vehicles (14,001-16,000 Ibs. GVWR)
Class 5 Heavy-Duty Gasoline Vehicles (16,001-19,500 Ibs. GVWR)
Class 6 Heavy-Duty Gasoline Vehicles (19,501-26,000 Ibs. GVWR)
Class 7 Heavy-Duty Gasoline Vehicles (26,001-33,000 Ibs. GVWR)
Class 8a Heavy-Duty Gasoline Vehicles (33,001-60,000 Ibs. GVWR)
Class 8b Heavy-Duty Gasoline Vehicles (>60,000 Ibs. GVWR)
Light-Duty Diesel Vehicles (Passenger Cars)
Light-Duty Diesel Trucks 1 and 2 (0-6,000 Ibs. GVWR)
Class 2b Heavy-Duty Diesel Vehicles (8501-10,000 Ibs. GVWR)
Class 3 Heavy-Duty Diesel Vehicles (10,001-14,000 Ibs. GVWR)
Class 4 Heavy-Duty Diesel Vehicles (14,001-16,000 Ibs. GVWR)
Class 5 Heavy-Duty Diesel Vehicles (16,001-19,500 Ibs. GVWR)
Class 6 Heavy-Duty Diesel Vehicles (19,501-26,000 Ibs. GVWR)
Class 7 Heavy-Duty Diesel Vehicles (26,001-33,000 Ibs. GVWR)
2-4
-------
TABLE 2-2
Continued
Vehicle
Class ID
22
23
24
25
26
27
28
Vehicle Class
Abbreviation
HDDV8A
HDDV8B
MC
HDGB
HDDBT
HDDBS
LDDT34
Vehicle Class Description
Class 8a Heavy-Duty Diesel Vehicles (33,001-60,000 Ibs. GVWR)
Class 8b Heavy-Duty Diesel Vehicles (>60,000 Ibs. GVWR)
Motorcycles (Gasoline)
Gasoline Buses (School, Transit and Urban)
Diesel Transit and Urban Buses
Diesel School Buses
Light-Duty Diesel Trucks 3 and 4 (6,001-8,500 Ibs. GVWR)
2.4
M6VType
The M6VType table contains a list of the 16 consolidated vehicle types used in
MOBILE 6 and the unique identifier assigned to each.
Data Source
Vehicle types were obtained from Table 1 of Appendix B of the MOBILE 6
User's Guide (EPA 420-R-02-028, October 2002), also available from
http://www.epa.gov/otaq/m6.htm.
Data Population Methodology
Data were entered in a Microsoft Excel spreadsheet and then exported to a
comma-delimited (csv) file. The csv file was then imported into the NMEVI database.
Quality Assurance Procedures
The contents of M6VType were printed and visually compared to the list of
MOBILE6 vehicle types from the MOBILE6 User's Guide. In addition, the null value, zero
2-5
-------
value, maximum and minimum value, parent-child, and child-parent QA/QC checks described in
Section 7.0 were also completed for this table.
Table 2-3 presents the 16 vehicle types contained in the M6VType table.
TABLE 2-3
M6VType Values
Vehicle Type
ID
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Vehicle Type Description
Light-Duty Vehicles (Passenger Cars)
Light-Duty Trucks 1 (0-6,000 Ibs. GVWR, 0-3,750 Ibs. LVW)
Light-Duty Trucks 2 (0-6,000 Ibs. GVWR, 3,751-5,750 Ibs. LVW)
Light-Duty Trucks 3 (6,001-8,500 Ibs. GVWR, 0-5,750 Ibs. ALVW)
Light-Duty Trucks 4 (6,001-8,500 Ibs. GVWR, 5,751 Ibs. and greater ALVW)
Class 2b Heavy-Duty Vehicles (8,501-10,000 Ibs. GVWR)
Class 3 Heavy-Duty Vehicles (10,001-14,000 Ibs. GVWR)
Class 4 Heavy-Duty Vehicles (14,001-16,000 Ibs. GVWR)
Class 5 Heavy-Duty Vehicles (16,001-19,500 Ibs. GVWR)
Class 6 Heavy-Duty Vehicles (19,501-26,000 Ibs. GVWR)
Class 7 Heavy-Duty Vehicles (26,001-33,000 Ibs. GVWR)
Class 8a Heavy-Duty Vehicles (33,001-60,000 Ibs. GVWR)
Class 8b Heavy-Duty Vehicles (>60,000 Ibs. GVWR)
School Buses
Transit and Urban Buses
Motorcycles
2-6
-------
s.o FUEL TABLES
NMIM contains several tables used to describe base and future year fuel
parameters, including fuel formulation information, market share information, fuel diesel
content, and natural gas content. Each fuel table is described in detail below.
3.1 Diesel
The Diesel table specifies the sulfur content of various diesel fuels used in the
base year, or anticipated to be used in future years.
Data Source
The Diesel sulfur values were extracted from the Future tab of a spreadsheet titled
sulfur.xls, forwarded by Dave Brzezinski, USEPA, on September 19, 2002.
Data Population Methodology
Because of the limited number of diesel fuels in the baseline and future years, the
Diesel table was populated manually. Two diesel records were added to the database as shown in
Table 3-1.
5-1
-------
TABLE 3-1
Diesel Sulfur Values
Diesel
ID
1
2
3
4
Diesel Sulfur Value
(parts per million
(ppm))
500
11
2700
120
Highway Applicability
Assigned to all counties
for calendar years 1999
through 2003.
Assigned to all counties
for calendar years 1999
through 2004.
Not applicable.
Not applicable.
NonRoad
Applicability
Not applicable.
Assigned to all California counties for
calendar years 2006 through 2008, assigned to
all counties for calendar years 2009 through
2050.
Assigned to all counties except those in
California for calendar years 1999 through
2005.
Assigned to all California counties for
calendar years 1999 through 2006.
Quality Assurance Procedures
The contents of Diesel were printed and visually compared to the diesel fuel
specification information provided in the data sources listed above. In addition, the null value,
zero value, maximum and minimum value, parent-child, and child-parent QA/QC checks
described in Section 7.0 were also completed for this table.
3.2
GasMTBEPhsOut
The GasMTBEPhsOut table contains fuel formulation and market share
information for base and future years.
Data Source
The baseline gasoline parameters used in this analysis were collected for calendar
year 1999. The gasoline parameters and county fuel mappings were obtained from a U.S. EPA
5-2
-------
guidance document that described the toxics module of MOBILE6.2 (U.S. EPA, 2002a). These
gasoline parameters were derived from several surveys: U.S. EPA's reformulated gasoline (RFG)
survey (U.S. EPA, 2000), the U.S. EPA Oxygenated Fuel Program Summary (U.S. EPA, 2001),
the TRW (previously NIPER) fuel survey (TRW, 1999), and the Alliance of Automobile
Manufacturers' (AAMA) North American Gasoline and Diesel Fuel Survey (AAMA, 1999). The
TRW fuel survey reports the data in several tables, including Table 9 (Motor Gasoline Survey,
Season [Summer/Winter], Year [1999/2000], and Average Data for Different Brands) and Table
10 (Motor Gasoline Survey, Season [Summer/Winter], Year [1999/2000], and Average Data for
Different Brands Containing Alcohols).
Data for the percent market share of oxygenated fuel sales were obtained from
Oxygenate Type Analysis Tables (1995-2000) (U.S. EPA, 2001) and the Federal Highway
Administration website (FHWA 1999).
The following section presents the methodologies and assumptions for selecting
parameters by state.
Calendar Year 1999 - NMIM Base Year
The data sources used to develop data for the 1999 base year by state are
described below.
All States
If methyl-tertiary butyl ether (MTBE) percent volume content was less than 0.1
percent, MTBE content was assumed to be zero, thus resulting in zero percent MTBE market
share. If ethanol percent volume content was less than 0.1 percent, ethanol content was assumed
to be zero and resulted in zero percent ethanol market share.
For any area that TRW reported MTBE, tert-amyl methyl ether (TAME), or ethyl
tert-butyl ether (ETBE) content as non-zero, the model assumed the entire market is attributed to
3-3
-------
MTBE because it was not possible to distinguish the market share between these specific
oxygenates.
For any area that reported a FHWA gasohol sale fraction, in addition to TRW data
for both regular gasoline and alcohol-containing gasoline, the fuel parameters for both sets of
TRW were reported and assigned 100% market share MTBE or ethanol, respectively. The
corresponding FHWA gasohol sale fractions were reported in a separate column.
Maximum sulfur values for 1999 through 2003 were assigned a value of 1,000
ppm based on Summer and Winter Reformulated Gasoline Parameters tables in Section 2.8.10.1
of the MOBILE6.2 User's Guide (U.S. EPA, 2002b)).
Alabama
The FHWA reported the ethanol market share as 0.16%. However, data from
TRW Table 9 (District 3) with 100% MTBE market share were used to represent the entire state
because TRW Table 10 did not report any alcohol-containing gasoline samples for District 3.
Alaska
All counties in Alaska were represented by fuel parameter data from the AAMA
survey for Fairbanks, Alaska. MTBE market share was 90.8% and ethanol market share was
9.2% based on FHWA data.
Arizona
Two counties in the Phoenix area (Maricopa and Final) were represented by fuel
parameter data from the AAMA survey for Phoenix, Arizona. Oxygenate fuel market share for
these counties was 100% MTBE for the summer and 100% ethanol for the winter, as provided by
the U.S. EPA Oxygenated Fuel Program Summary.
5-4
-------
The remaining counties in Arizona were represented by fuel parameter data from
TRW Table 9 (District 12). For this region, FHWA reported 100% ethanol market share for both
summer and winter. This oxygenate market share data were consistent with the statewide annual
average of 7.6% reported by FHWA.
Arkansas
All counties in Arkansas were represented by fuel parameter data from TRW
Table 9 (District 3) with 100% MTBE market share.
California
Six California counties (Los Angeles, Orange, Riverside, San Bernardino, San
Diego, and Ventura) were included as federal RFG program areas. In addition, California
administers its own RFG program, but does not sample during the winter. Therefore, California
fuel parameter data were obtained from the TRW survey and the AAMA survey.
Counties in the San Francisco Bay area (Alameda, Contra Costa, Marin, Napa,
San Francisco, San Mateo, Santa Clara, Solano, and Sonoma) were represented by data from the
AAMA survey for San Francisco, California, including a 50/50 split market share between
MTBE and ethanol during summer and 100% MTBE market share during winter.
All other counties in California were represented by data from the AAMA survey
for Los Angeles, including 100% MTBE market share for both summer and winter.
Colorado
Counties in the Denver area (Adams, Arapahoe, Denver, Douglas, and Jefferson)
were represented by data from the AAMA survey for Denver, Colorado, including a year round
100% ethanol market share.
5-5
-------
All other counties in Colorado were represented by data from TRW Table 9
(District 10) for the summer and TRW Table 10 (District 10) for the winter. These data include a
100% MTBE market share during summer and a 100% ethanol market share during winter,
which were consistent with the annual 27.27% ethanol statewide average, as reported by FHWA.
Connecticut
All counties, except Fairfield County, were represented by data from the RFG
survey for Hartford, Connecticut. These data include a 99.15% MTBE and 0.85% TAME market
share for the summer and a 95% MTBE, 4% ethanol, and 1% TAME market share for the winter.
Fairfield County was represented by data from the RFG survey for New York-
New Jersey-Long Island, including 100% MTBE market share for the summer and 98.14%
MTBE and 1.86% ethanol market share for the winter. These data exhibit a small discrepancy
with 2.27% ethanol for the state reported by FHWA.
Delaware
All counties, except Sussex County, were represented by data from the RFG
survey for Philadelphia-Wilmington-Trenton, including 100% MTBE market share data for the
summer and 98.55% MTBE and 1.45% ethanol market share for the winter.
Sussex County was represented by data from the RFG survey for Sussex County,
Delaware, including 100% MTBE market share for both summer and winter.
District of Columbia
Washington D.C. was represented by data from the RFG survey for Washington
D.C. including 100% MTBE market share. This was consistent with the 0% ethanol market
share reported by FHWA.
-------
Florida
Dade County was represented by data from the AAMA survey for Miami, FL,
including 100% MTBE market share.
All other counties in Florida were represented by data from TRW Table 9 (District
4), including 100% MTBE market share. These data were consistent with 0% ethanol statewide
as reported by FHWA.
Georgia
Counties in the Atlanta area (Barrow, Bartow, Carroll, Cherokee, Clayton, Cobb,
Coweta, DeKalb, Douglas, Fayette, Forsyth, Fulton, Gwinnett, Henry, Newton, Paulding,
Pickens, Rockdale, Spalding, and Walton) were represented by data from the AAMA survey for
Atlanta, Georgia, including 100% MTBE market share.
The remaining counties in Georgia were represented by data from TRW Table 9
(District 3), including 100% MTBE market share. These data were consistent with 0% ethanol
statewide as reported by FHWA.
Hawaii
All counties in Hawaii were represented by fuel parameter data from TRW Table
9 (District 14-Northern California), including 100% MTBE market share.
Iowa
All counties in Iowa were represented by fuel parameter data from both TRW
Table 9 and TRW Table 10 (District 7) for both summer and winter because the oxygenate
market share was unknown. If the data originated from TRW Table 9, then MTBE market share
was 100%. If the data originated from TRW Table 10, then ethanol market share was 100%.
3-7
-------
FHWA survey data were used to assign market shares of 55.18% MTBE and 44.82% ethanol for
all counties.
Idaho
All counties in Idaho were represented by fuel parameter data from TRW Table 9
(District 9), including 100% MTBE market share. This was consistent with 0% ethanol market
share statewide, as reported by FHWA.
Illinois
Counties in the Chicago area (Cook, DuPage, Grundy, Kane, Kendall, Lake,
McHenry, and Will) were represented by data from the AAMA survey for Chicago-Lake County,
Illinois, including 100% ethanol market share.
Counties in the St. Louis area (Clinton, Jersey, Madison, Monroe, and St. Clair)
were represented by fuel parameter data from the AAMA survey for St. Louis, Missouri. Market
share data from the RFG survey for St. Louis, Missouri were 80.34% MTBE and 19.66% ethanol
for the summer and 54.95% MTBE and 45.05% ethanol for the winter.
The remaining counties were represented by fuel parameter data from both TRW
Table 9 and TRW Table 10 (District 7, except Adams County uses data from District 5) for both
winter and summer because the oxygenate market share was unknown. If the data originated
from TRW Table 9, then MTBE market share was 100%. If the data originated from TRW Table
10, then ethanol market share was 100%. FHWA survey data were used to assign market shares
of 50.78% MTBE and 49.22% ethanol for all counties.
Indiana
Lake and Porter counties were represented by data from the AAMA survey for
Chicago-Lake County, Illinois, including 100% ethanol market share.
3-8
-------
The remaining counties were represented by fuel parameter data from both TRW
Table 9 and TRW Table 10 (District 6, except Adams County uses data from District 5) for both
winter and summer because the oxygenate market share was unknown. If the data originated
from TRW Table 9, then MTBE market share was 100%. If the data originated from TRW Table
10, then ethanol market share was 100%. FHWA survey data were used to assign market shares
of 68.92% MTBE and 31.08% ethanol for all counties.
Kansas
All counties were represented by fuel parameter data from both TRW Table 9 and
TRW Table 10 (District 7) for both winter and summer because the oxygenate market share was
unknown. If the data originated from TRW Table 9, then MTBE market share was 100%. If the
data originated from TRW Table 10, then ethanol market share was 100%. FHWA survey data
were used to assign market shares of 96.39% MTBE and 3.61% ethanol for all counties.
Kentucky
Boone, Campbell, and Kenton counties were represented by data from the RFG
survey for Covington, Kentucky. These data include 22.53% MTBE and 77.47% ethanol market
share for the summer and 25.51% MTBE and 74.49% ethanol market share for the winter.
Bullitt, Jefferson, and Oldham counties were represented by data from the RFG
survey for Louisville, Kentucky. These data include 76.25% MTBE and 23.75% ethanol market
share for the summer and 72.61% MTBE and 27.39% ethanol market share for the winter.
All other counties were represented by data from TRW Table 9 (District 6),
including 100% MTBE market share.
These data may slightly overestimate the ethanol sales when compared to
FHWA's statewide market share estimate of 1.52% ethanol.
5-9
-------
Louisiana
Parishes in the New Orleans area (Jefferson, Orleans, Plaquemines, St. Bernard,
St. Charles, St. James, St. John the Baptist, and St. Tammany) were represented by data from the
AAMA survey for New Orleans, Louisiana, including a 100% MTBE market share.
All other counties were represented by TRW Table 9 (District 3) with 100%
MTBE market share.
These data were slightly inconsistent with FFIWA's estimate of 0.65% ethanol for
the state.
Massachusetts
Berkshire, Franklin, Hampden, and Hampshire counties were represented by data
from the JAFG survey for Springfield, Massachusetts. This data includes 98.74% MTBE and
1.26% TAME market share for the summer and 95.83% MTBE and 4.17% ethanol market share
for the winter.
All other counties were represented by fuel parameter data from the JAFG and
AAMA surveys for Boston-Worchester, Massachusetts. Market share data obtained from the
RFG survey include 96.51% MTBE and 3.49% TAME market share for the summer and 91.67%
MTBE, 4.17% ethanol, and 3.92% TAME market share for the winter.
Maryland
Cecil, Kent, and Queen Anne's counties were represented by fuel parameter data
from the JAFG survey for Philadelphia-Wilmington-Trenton, including 100% MTBE market
share in the summer and 98.55% MTBE and 1.45% ethanol market share in the winter.
3-10
-------
Calvert, Charles, Frederick, Montgomery, and Prince George's counties were
represented by fuel parameter data from the RFG survey for Washington D.C., including 100%
MTBE market share for both summer and winter.
Counties in the Baltimore area (Anne Arundel, Baltimore, Baltimore City, Carroll,
Harford, and Howard) were represented by fuel parameter data from the RFG survey for
Baltimore, Maryland. These data include 99.45% MTBE and 0.46% TAME market share for the
summer and 99.44% MTBE and 0.56% ethanol market share for the winter.
All other counties were represented by data from TRW Table 9 (District 1) and
100% MTBE market share.
Maine
Seven counties in Maine (i.e., Androscoggin, Cumberland, Kennebec, Knox,
Lincoln, Sagadahoc, and York counties) "opted-out" of the federal RFG program effective March
10, 1999; the RFG survey data were not used for these seven counties for 1999. These seven
counties were represented by fuel parameter data from TRW Table 11 and 100% MTBE market
share.
All other counties were represented by fuel parameter data from TRW Table 9
(District 1) and 100% MTBE market share. These assumptions were consistent with the 0%
statewide ethanol consumption reported by FHWA.
Michigan
Counties in the Detroit area (Lapeer, Macomb, Monroe, Oakland, St. Clair, and
Wayne) were represented by fuel parameter data from the AAMA survey for Detroit, Michigan
and 100% ethanol market share.
3-11
-------
All other counties were represented by data from both TRW Table 9 and TRW
Table 10 (District 5) because the oxygenate market share was unknown. If the data originated
from TRW Table 9, then MTBE market share was 100%. If the data originated from TRW Table
10, then ethanol market share was 100%. FHWA survey data were used to assign market shares
of 93.07% MTBE and 6.93% ethanol for all counties.
Minnesota
Counties in the Minneapolis/St. Paul area (Anoka, Carver, Chisago, Dakota,
Hennepin, Isanti, Ramsey, Scott, Sherburne, Washington, and Wright) were represented by data
from the AAMA survey for Minneapolis/St. Paul, Minnesota and 100% ethanol market share for
both summer and winter, based on low measured MTBE concentrations (0.1%).
All other counties were represented by data from both TRW Table 9 and TRW
Table 10 (District 5) because the oxygenate market share was unknown. If the data originated
from TRW Table 9, then MTBE market share was 100%. If the data originated from TRW Table
10, then ethanol market share was 100%. FHWA survey data were used to assign market shares
of 8.26% MTBE and 91.74% ethanol for all counties for both summer and winter.
Missouri
Five counties in Missouri (Franklin, Jefferson, St. Charles, St. Louis, and the city
of St. Louis) "opted-in" to the federal RFG program effective June 1, 1999. These five counties
were represented by data from the RFG and AAMA surveys for St. Louis, including a market
share of 80.34% MTBE and 19.66% ethanol in the summer and 54.95% MTBE and 45.05% in
the winter.
Counties in the Kansas City area (Cass, Clay, Clinton, Jackson, Lafayette, Platte
and Ray) were represented by fuel parameter data from the AAMA survey for Kansas City,
Missouri and 100% MTBE market share.
3-12
-------
The remaining counties were represented by data from TRW Table 9 (District 7)
with 100% MTBE market share.
FHWA reported a 5.34% ethanol sale fraction for the entire state of Missouri.
Mississippi
Fuel parameter data from TRW Table 9 (District 3) and 100% MTBE market
share were used to represent all counties in Mississippi. These data were consistent with
FHWA's estimate of 0% ethanol sales market share.
Montana
Yellowstone County was represented by data from the AAMA survey for Billings,
Montana and 100% MTBE market share for both summer and winter.
Missoula County was represented by data from TRW Table 10 with 100% ethanol
market share in winter, per U.S. EPA's Oxygenated Fuel Program Summary.
All other counties were represented by fuel parameter data from TRW Table 9
(District 9) and 100% MTBE market share.
Nebraska
All counties in Nebraska were represented by data from both TRW Table 9 and
TRW Table 10 (District 7) because the oxygenate market share was unknown. If the data
originated from TRW Table 9, then MTBE market share was 100%. If the data originated from
TRW Table 10, then ethanol market share was 100%. FHWA survey data were used to assign
market shares of 74.78% MTBE and 25.22% ethanol for all counties.
3-13
-------
Nevada
Clark and Nye counties were represented by data from the RFG survey for Las
Vegas, Nevada. Based on the U.S. EPA Oxygenated Fuel Program Summary, these counties
were assigned 100% ethanol market share for the winter. The summer market share was
assigned a value of 100% MTBE to be more consistent with the FHWA estimate of 0% ethanol
market share.
Carson City, Esmeralda, Lincoln, and Mineral counties were represented by data
from TRW Table 9Se (District 12) for the summer TRW Table 10 (District 12) for the winter.
Based on the U.S. EPA Oxygenated Fuel Program Summary, these four counties were assigned
100% ethanol market share for the summer and 100% MTBE market share for the winter. Note
that the gas sulfur values were reported as 0 for these four counties.
All other counties were assigned data from TRW Table 9 (District 14) with 100%
MTBE market share. This assumption allows the data to be more consistent with the FHWA
estimate of 0% ethanol market share.
New Hampshire
Hillsboro and Merrimack counties were represented by fuel parameter data from
the RFG survey for the Manchester, New Hampshire area. These data include 100% MTBE
market share for the summer and 99.16% MTBE and 0.84% TAME market share in the summer.
Rockingham and Strafford counties were represented by data from the RFG
survey for the Portsmouth-Dover, New Hampshire area. These data include 100% MTBE market
share for both summer and winter.
All other counties were represented by data from TRW Table 9 (District 1) with
100% MTBE.
3-14
-------
New Jersey
Atlantic and Cape May counties were represented by data from the RFG survey
for Atlantic City, New Jersey, including 100% MTBE market share for summer and 96.84%
MTBE and 2.11% ethanol market share for winter..
Warren County was represented by data from the RFG survey for Warren County,
including 100% MTBE market share for both summer and winter.
Burlington, Camden, Cumberland, Gloucester, Mercer, and Salem counties were
represented by data from the RFG survey for Philadelphia-Wilmington-Trenton. This includes
100% MTBE market share for the summer and 98.55% MTBE and 1.45% ethanol market share
for the winter.
All other counties were represented by data from the RFG survey for the New
York-New Jersey-Long Island-Connecticut region. These data include 100% MTBE market
share for the summer and 98.14% MTBE and 1.86% ethanol market share for the winter.
These assumptions slightly underestimate the FFIWA statewide estimate of 2.10%
ethanol sales market share.
New Mexico
Bernalillo, Sandoval, and Valencia counties were represented by data from the
RFG survey for the Albuquerque area. Based on the U.S. EPA Oxygenated Fuel Program
description, these counties were assigned 100% ethanol market share for the winter. For the
summer, 100% ethanol market share was assumed based on low measure concentrations of
MTBE (0.1%) versus ethanol (0.8%).
All other counties were represented by data from TRW Table 9 (District 11) with
100% MTBE market share for the summer and TRW Table 10 (District 11) with 100% ethanol
3-15
-------
for the winter. There were no data for summer alcohol fuels in NIPER District and winter
MTBE levels were measured as 0 in Table 9.
New York
Dutchess and Putnam counties were represented by data from the RFG survey for
Poughkeepsie, New York with RFG survey market share. These data include 100% MTBE
market share for the summer and 95.14% MTBE and 4.86% ethanol market share for the winter.
Counties in the New York City area (Bronx, Kings, Nassau, New York, Orange,
Queens, Richmond, Rockland, Suffolk, and Westchester) were represented by data from the RFG
survey data for the New York-New Jersey-Long Island-Connecticut region. These data include
100% MTBE market share for the summer and 98.14% MTBE and 1.86% ethanol market share
for the winter.
The remaining counties were represented by data from TRW Table 9 (District 1)
and 100% MTBE market share. TRW Table 10 was not provided in this data set.
North Carolina
All counties were represented by fuel parameter data from TRW Table 9 (District
2) and 100% MTBE market share. FHWA reported 7.47% gasohol sales in North Carolina, but
the TRW survey did not collect any gasoline containing alcohol in this area.
North Dakota
All counties in North Dakota were represented by data from both TRW Table 9
and TRW Table 10 (District7) because the oxygenate market share was unknown. If the data
originated from TRW Table 9, then MTBE market share was 100%. If the data originated from
TRW Table 10, then ethanol market share was 100%. FHWA survey data were used to assign
market shares of 87.64% MTBE and 12.36% ethanol for all counties.
3-16
-------
Ohio
Counties in the Cleveland area (Ashtabula, Cuyahoga, Geauga, Lake, Lorain, and
Medina) were represented by fuel parameter data from the RFG survey for Cleveland. The
ethanol market share was assumed to be 100%, based on low measured concentrations of MTBE
(-0.1%)
The remaining counties in Ohio were represented by data from both TRW Table 9
and TRW Table 10 (District 6) because the oxygenate market share was unknown. If the data
originated from TRW Table 9, then MTBE market share was 100%. If the data originated from
TRW Table 10, then ethanol market share was 100%. FHWA survey data were used to assign
market shares of 60.26% MTBE and 39.74% ethanol for all counties.
Oklahoma
All counties in Oklahoma were represented by fuel parameter data from TRW
Table 9 (District 8) and 100% MTBE market share. These data were consistent with FHWA's
statewide estimate of 0% ethanol market share.
Baker County was represented by from TRW Table 9 (District 9) and 100%
MTBE market share.
All other counties in Oregon, with the exception of Clackamas, Columbia,
Jackson, Josephine, Klamath, Multnomah, Washington, and Yamhill counties for the winter
season, were represented by fuel parameter data from TRW Table 9 (District 13), including
100% MTBE market share.
For the summer season, these eight counties were represented by TRW Table 10
with 100% ethanol market share, based on U.S. EPA Oxygenated Fuel Program descriptions.
3-17
-------
These assumptions were consistent with FHWA's statewide estimate of 7.3% ethanol market
share.
Pennsylvania
Bucks, Chester, Delaware, Montgomery, and Philadelphia counties were
represented by data from the RFG survey for Philadelphia. These data include 100% MTBE
market share for the summer and 98.55% MTBE and 1.45% ethanol market share for the winter.
For the summer season only, Alleghany, Armstrong, Butler, Fayette, Washington,
and Westmoreland counties were represented by data from the AAMA survey for the Pittsburgh
region with 100% MTBE market share. For the winter season, these counties were represented
by data from TRW Table 9 with 100% MTBE market share.
The remaining counties in Pennsylvania were represented by data from TRW
Table 9 (District 1) with 100% MTBE market share. There were no alcohol-containing samples
in the NIPER District 1 surveys, which contradicts FHWA's statewide estimate of 2.11% ethanol
market share.
Rhode Island
All counties were represented by data from the RFG survey for the state of Rhode
Island, including 100% MTBE market share for the summer and 97.52% MTBE and 2.48 ethanol
market share for the winter.
South Carolina
Fuel parameter data from TRW Table 9 (District 3) and 100% MTBE market
share were used to represent all counties in South Carolina. These data were consistent with
FHWA's statewide estimate of 0% ethanol market share.
3-18
-------
South Dakota
All counties in South Dakota were represented by data from both TRW Table 9
and TRW Table 10 (District 7) because the oxygenate market share was unknown. If the data
originated from TRW Table 9, then MTBE market share was 100%. If the data originated from
TRW Table 10, then ethanol market share was 100%. FHWA survey data were used to assign
market shares of 57.32% MTBE and 42.68% ethanol for all counties.
Tennessee
Fuel parameter data from TRW Table 9 (District 3) and 100% MTBE market
share were used to represent all counties in Tennessee. These data were consistent with FHWA's
statewide estimate of 0% ethanol market share.
Texas
Bexar, Comal, Guadalupe, and Wilson counties were represented by data from the
AAMA survey data for San Antonio, Texas. These counties were assigned 100% MTBE market
share, based on low measured ethanol concentrations (-0.1%).
Collin, Dallas, Denton, and Tarrant counties were represented by data from the
RFG survey data for the Dallas-Fort Worth region, including 100% MTBE market share in the
summer and 94.15 MTBE% and 5.85% TAME market share in the winter.
Brazoria, Chambers, Fort Bend, Galveston, Harris, Liberty, Montgomery, and
Waller counties were represented by data from the RFG survey data for the Houston-Galveston
area. These data include 97.69% MTBE and 1.82 ethanol market share for the summer and
99.53% MTBE and 0.47% ethanol market share for the winter.
3-19
-------
Counties in the eastern part of the state were represented by data from TRW Table
9 for District 8. These counties were assigned 100% MTBE market share because District 8 does
not have survey information for fuels with alcohol.
Counties in the western part of the state were represented by data from TRW
Table 10 for District 11. These counties were assigned 100% ethanol market share for the winter
season because measured MTBE levels were zero. These counties were assigned 100% MTBE
market share during the summer season because District 11 does not have survey information for
fuels with alcohol in the summer.
These assumptions, primarily the assumption that western counties use ethanol-
based fuel in the winter, were relatively consistent with FHWA's statewide estimate of 4.95%
ethanol market share.
Utah
Data from TRW Table 9 (District 10) with 100% MTBE market share were used
to represent all counties in Utah, except Utah and Weber counties during the winter season. For
this season, data from TRW Table 10 (District 10) were used to represent Utah and Weber
counties. Utah and Weber counties were assigned with 100% ethanol market share, based on the
U.S. EPA Oxygenated Fuel Program description. These assumptions may not fully account for
FHWA's statewide estimate of 10.67% ethanol market share.
Virginia
Counties in the Washington D.C. area (Alexandria City, Fairfax City, Falls
Church City, Manassas City, Manassas Park City, Arlington, Fairfax, Loudoun, Prince William,
and Stafford) were represented by data from the RFG survey for Washington D.C. for both fuel
parameters, including 100% MTBE market share.
3-20
-------
Counties in the Richmond area (Colonial Heights City, Hopewell City, Richmond
City, Hanover, and Henrico counties) were represented by data from the RFG survey for
Richmond for both fuel parameters, including 100% MTBE market share.
Counties in the Norfolk area (Chesapeake City, Hampton City, Newport News
City, Norfolk City, Poquoson City, Portsmouth City, Suffolk City, Virginia Beach,
Williamsburg, Charles City, Chesterfield, James City, and York) were represented by data from
the RFG survey for Norfolk-Virginia Beach for both fuel parameters, including 100% MTBE
market share.
All other counties were represented by data from TRW Table 10 (District 6),
including 100% ethanol market share.
The FHWA reported a statewide 8.61% ethanol market share for Virginia.
Vermont
Fuel parameter data from TRW Table 9 (District 1) and 100% MTBE market
share were used to represent all counties in Vermont. These data were consistent with FHWA's
statewide estimate of 0% ethanol market share.
Washington
Island, King, and Snohomish counties were represented by data from the AAMA
survey for the Seattle, Washington area. These data include 100% ethanol market share during
winter, based on low measured MTBE concentrations (0.1%), and 100% MTBE market share
during summer.
Adams County was represented with data from TRW Table 9 (District 9) and
100% MTBE market share for both summer and winter.
3-21
-------
Clark and Spokane counties were represented with data from TRW Table 10
(District 13) and 100% ethanol market share during winter per the Oxygen Fuel Program
description. For the summer season, these counties were represented by data from TRW Table 9
(District 13) and 100% MTBE market share.
All other counties were represented by data from TRW Table 9 (District 13) and a
100% MTBE market share summer, but no defined market share for winter.
These assumptions may over predict the statewide ethanol market fraction when
compared to the 9.93% as reported by FFEWA.
Wisconsin
Kenosha, Milwaukee, Ozaukee, Racine, Washington, and Waukesha counties
were represented by data from the RFG survey for the Milwaukee-Racine region for both fuel
parameters and market share. These data include 100% ethanol market share, which accounts for
the statewide 10.98% ethanol market share reported by FFEWA.
All other counties were represented by data from TRW Table 9 (District 5),
including 100% MTBE market share.
West Virginia
Fuel parameter data from TRW Table 9 (District 6) and 100% MTBE market
share were used to represent all counties in West Virginia. These data were consistent with
FFEWA's statewide estimate of 0.01% ethanol market share.
3-22
-------
Wyoming
Fuel parameter data from TRW Table 9 (District 9) and 100% MTBE market
share were used to represent all counties in Wyoming. These data were consistent with FHWA's
statewide estimate of 0% ethanol market share.
Puerto Rico and the U.S. Virgin Islands
Gasoline parameters and county fuel mappings for Puerto Rico and the U.S.
Virgin Islands were not included in the U.S. EPA guidance document referenced above. It was
assumed that gasoline in Puerto Rico and the U.S. Virgin Islands was similar to Hawaii.
Future Years
The future year gasoline parameters were calculated using adjustment factors that
were applied to the base year gasoline parameters. In general, multiplicative adjustment factors
were used to calculate future year gasoline parameters (i.e., future year parameter = base year
parameter x adjustment factor). However, additive adjustment factors were used to calculate
future year parameters for E200, E300, and oxygenate market shares (i.e., future year parameter =
base year parameter + adjustment factor). The estimation of the future year gasoline parameters
is described below:
Calendar Year 2000
For most counties, the 2000 gasoline parameters were identical to the 1999
gasoline parameters. The only exception was that updated U.S. EPA RFG survey data for 2000
replaced the 1999 gasoline parameters for the 154 non-California RFG area counties (U.S. EPA,
2000).
3-23
-------
Calendar Years 2001 through 2003
The 1999 gasoline parameters (and 2000 gasoline parameters for the 154 non-
California RFG area counties) were used to represent the 2001, 2002, and 2003 calendar years
(i.e., multiplicative adjustment factors for these years were set to 1.0 and additive adjustment
factors set to 0.0). The phase-in of Phase 3 RFG in California had initially been set to begin in
2003. However, this phase-in has since been pushed back by one year and is scheduled to begin
in 2004. Therefore, multiplicative adjustment factors of 1.0 and additive adjustment factors of
0.0 were also applied to California for the 2001 through 2003 calendar years.
Calendar Year 2004
Beginning in 2004, Tier 2 motor vehicle emissions standards and gasoline sulfur
control requirements will be phased in throughout the country (Federal Register, 2000; Federal
Register, 2001). Fuel parameters were obtained from cost analyses conducted for the National
Petrochemical and Refiners Association (NPRA) (MathPro, 1998) and the American Petroleum
Institute (API) (MathPro, 1999a). The NPRA analysis focused only on Petroleum
Administration Defense District (PADD) IV (i.e., Montana, Idaho, Utah, Wyoming, and
Colorado). The API analysis included PADDs I, II, and HI (i.e., 38 Eastern and Plains states,
Puerto Rico, and the U.S. Virgin Islands). US EPA staff indicated that data derived from the API
analysis for PADDs I, II, and HI should also be used for PADD V (i.e., Arizona, Nevada, Oregon,
Washington, Alaska, and Hawaii; excluding California).
The Tier 2 sulfur standards include refinery average limits, corporate pool average
limits, and per-gallon cap limits (Federal Register, 2000), and are applicable for most of the
country, excluding the Geographic Phase-In Area (GPA) described below. The years of 2004
and 2005 are phase-in years with the final limits being implemented in 2006. Additional
discussion with U.S. EPA staff indicated that appropriate "at the pump" sulfur contents were 120
parts per million (ppm), 90 ppm, and 30 ppm (for 2004, 2005, and 2006 and beyond,
respectively) (Manners, 2002). With the exception of the GPA, the API gasoline parameter data
for PADDs I, II, IE, and V were used for the 2004 Tier 2 sulfur standards (MathPro, 1999a).
3-24
-------
The API analysis contained modeled gasoline parameters for conventional
gasoline (summer and winter) and RFG (summer and winter). The modeled gasoline parameters
were based on a 2004 reference fuel and two 40 ppm sulfur content fuels (one modeled with the
OCTGAIN process and the other with the CD TECH process). In addition to the expected 2004
sulfur content of 120 ppm, the other gasoline parameters were calculated by interpolation using
the following equation:
([$120 ~~
Where Sref = sulfur content of reference fuel;
S120 = sulfur content of 120 ppm sulfur content fuel;
S40 = sulfur content of 40 ppm sulfur content fuel;
Pref = value of other parameter for reference fuel;
P120 = value of other parameter for 120 ppm sulfur content fuel; and
P40 = value of other parameter for 40 ppm sulfur content fuel.
The sulfur content and other parameter values for the 40 ppm sulfur content fuel
were averages of the OCTGAIN and CD TECH modeled fuels. This interpolation method was
used to determine fuel parameter values for the 120 ppm sulfur content fuel. The multiplicative
adjustment factor (MAP) for each relevant parameter was calculated by ratioing the 120 ppm
sulfur content fuel parameter by the reference fuel parameter (i.e., MAP = Pi2o/Pref). The additive
adjustment factor (AAF) for each relevant parameter was calculated by subtracting the reference
fuel parameter from the 120 ppm sulfur content fuel parameter (i.e., AAF = P120 - Pref).
Four sets of adjustment factors were developed for 2004 fuel in PADDs I, n, HI,
and V (i.e., summer conventional, summer RFG, winter conventional, and winter RFG). A fifth
set of adjustment factors were also developed for those conventional gasoline areas that use
gasohol during the summer. These adjustment factors are identical to the summer conventional
except that the oxygenate adjustment factors were set to 1.0.
3-25
-------
Tier 2 - Geographic Phase-In Area (GPA)—Amendments to the Tier 2 sulfur
standards provided for an additional phase-in year (i.e., 2006) in a defined Geographic Phase-In
Area (GPA) (Federal Register, 2001). The GPA is established ensure a smooth transition to low
sulfur gasoline nationally and to mitigate the potential of gasoline supply shortages in certain
parts of the country. The GPA is defined as eight states (i.e., Montana, Idaho, Utah, Wyoming,
Colorado, New Mexico, North Dakota, and Alaska) plus 74 adjacent counties in six other states
(i.e., Washington, Oregon, Nevada, Arizona, South Dakota, and Nebraska). Additional
discussion with U.S. EPA staff indicated that appropriate "at the pump" sulfur contents for the
GPA were 150 ppm, and 30 ppm (for 2004-2006 and 2007 and beyond, respectively) (Manners,
2002). Because PADD IV roughly corresponds with the GPA, the NPRA gasoline parameter
data for PADD IV were used for the 2004 Tier 2 sulfur standards in the GPA (MathPro, 1998).
The NPRA analysis contained modeled gasoline parameters for high and low
sulfur gasolines (summer and winter). The modeled gasoline parameters were based on a 1996
baseline fuel and a 150 ppm sulfur content fuel. Pooled fuel parameters were estimated for both
the baseline fuel and the 150 ppm sulfur content fuel assuming a pool fuel mix of 75 percent high
sulfur gasoline and 25 percent low sulfur gasoline. The 2004 MAP and AAF values for PADD
IV were calculated in similar manner to those in PADDs I, n, IE, and V (i.e., MAP = Pi50/Pbase
and AAF = P150 - Pbase); the only significant difference is that fuel parameter interpolation was
not needed because the NPRA analysis included the appropriate sulfur content fuel (i.e., 150
ppm).
Two sets of adjustment factors were developed for 2004 fuel in PADD IV (i.e.,
summer and winter). A third set of adjustment factors were also developed for those areas that
use gasohol during the summer. These adjustment factors are identical to the summer
conventional except that the oxygenate adjustment factors were set to 1.0.
3-26
-------
California Phase 3 RFG—In addition to the phase-in of Tier 2 sulfur standards
throughout the country, the phase-in of California Phase 3 RFG also begins in 2004. As
previously mentioned, this phase-in was initially scheduled to begin in 2003, but was pushed
back 1 year. In support of California Phase 3 RFG, a standard analysis was conducted for the
California Energy Commission (CEC) that modeled 18 different fuel scenarios (MathPro,
1999b). The two fuel scenarios that were used were a MTBE-containing Phase 2 RFG fuel and a
Phase 3 RFG fuel containing no oxygenates (i.e., representing the effects of an MTBE ban).
The 2004 MAF and AAF values for California were calculated in similar manner
to those in PADDs I through V (i.e., MAF = Pphase3/PPhase2 and AAF = PPhase3 - PPhase2). The 2004
MAF and AAF values were also used in two Arizona counties (Maricopa and Final) where a
similar, but not identical, fuel will be implemented.
Calendar Year 2005
In 2005, the non-GPA sulfur content was reduced from 120 ppm to 90 ppm based
upon discussions with U.S. EPA staff (Manners, 2002). The interpolation method described for
2004 non-GPA fuels was used to determine appropriate adjustment factors for the 2005 non-
GPA fuels as well. The only change was basing the interpolation on a 90 ppm fuel instead of a
120 ppm fuel (i.e., MAF = P90/Pref and AAF = P90 - Pref). This resulted in five sets of adjustment
factors for 2005 fuel in PADDs I, II, in, and V (i.e., summer conventional, summer RFG, winter
conventional, winter RFG, and summer conventional with gasohol).
The 2005 GPA and California Phase 3 RFG fuels were unchanged relative to the
2004 fuels. As a result, the 2005 GPA and California adjustment factors are identical to 2004.
3-27
-------
Calendar Year 2006
In 2006, the non-GPA sulfur content was reduced from 90 ppm to 30 ppm based
upon discussions with U.S. EPA staff (Manners, 2002). The interpolation method described for
2004 non-GPA fuels was used to determine appropriate adjustment factors for the 2006 non-
GPA fuels as well. The only change was basing the interpolation on a 30 ppm fuel instead of a
120 ppm fuel (i.e., MAP = P30/Pref and AAF = P30 - Pref). This resulted in five sets of adjustment
factors for 2006 fuel in PADDs I, II, IE, and V (i.e., summer conventional, summer RFG, winter
conventional, winter RFG, and summer conventional with gasohol).
The 2006 GPA and California Phase 3 RFG fuels were unchanged relative to the
2004 fuels. As a result, the 2006 GPA and California adjustment factors are identical to 2004.
Calendar Year 2007
In 2007, the GPA sulfur content was reduced from 150 ppm to 30 ppm based
upon discussions with U.S. EPA staff (Manners, 2002). The interpolation method described for
2004 non-GPA fuels was used to determine appropriate adjustment factors for the 2007 GPA
fuels as well. The only change was basing the interpolation on a 30 ppm fuel instead of a 120
ppm fuel (i.e., MAP = P30/Pbase and AAF = P30 - Pbase). This resulted in three sets of adjustment
factors for 2007 in the GPA (i.e., summer, winter, and summer with gasohol).
The 2007 non-GPA and California Phase 3 RFG fuels were unchanged relative to
the 2006 fuels. As a result, the 2007 GPA and California adjustment factors were identical to
2006.
Calendar Years 2008 and 2009
In 2008 and 2009, it was assumed that there were no fuel changes for any fuels
(i.e., non-GPA, GPA, and California). As a result, all gasoline parameters for 2008 and 2009
were identical to 2007.
3-28
-------
Calendar Years 2010 through 2050
Beginning in 2010, a potential ban of MTBE-containing fuels was modeled. Fuel
parameters were derived from detailed refinery modeling runs conducted for U.S. EPA (Abt,
2003). Gasoline parameters for the 2010 Reference #la and 2010 RFS #2 modeled fuels (both
conventional and RFG) were obtained separately for PADD I, II, and HI. Weighted gasoline
parameters were derived based upon volumes of MTBE- and ETOH-blended fuels in PADD II
and HI.
The MAP for each relevant parameter was calculated by ratioing the RFS #2 fuel
parameter by the Reference #la fuel parameter (i.e., MAP = PRFS#2/PRef#i)- The AAF for each
relevant parameter was calculated by subtracting the Reference #la fuel parameter from the RFS
#2 fuel parameter (i.e., AAF = PRFS#2 - PR^I)- The oxygenate contents and market shares were
then adjusted to represent expected conditions occurring due to a MTBE ban. The PADD n
adjustment factors were applied to PADD V. No changes related to a MTBE ban were made to
California (where MTBE was already phased-out as of 2004) or to the GPA.
Oxygenate Volume and Market Share Analysis for 2000 through 2050
Because oxygenate volume and market share data were not available for calendar
years past 1999, an analysis of the average market share for each oxygenate at the PADD level
was performed. The total weight percent oxygenate data available from the Future Year Fuel
Data spreadsheet were used in combination with the MOBILE6 oxygenate conversion factors to
determine individual oxygenate volumes. These PADD oxygenate volumes and market shares
were then transferred to the future year spreadsheet prior to developing the gasoline table.
Data Population Methodology
The GasMTBEPhsOut data were populated using information from spreadsheets
containing seasonal fuel data for various years as described in Section 3.2.1 and programming
utilities written using Microsoft Access. These programming utilities prepared composite
3-29
-------
seasonal gasolines for counties that reported multiple winter and summer fuels, applied
multiplicative or additive parameters for appropriate years, interpolated seasonal fuel parameters
to monthly fuel parameters, determined the set of unique gasolines resulting from the
interpolation program, and populated the Gasoline, Gasoline2, GASMTBEPhsOut, and
County YearMonth tables. Each of these components is described in detail below.
Seasonal Fuel Data
The seasonal fuel data spreadsheets were populated using the data sources
described in the Seasonal Fuel Data portion of this section. The format of each is as follows:
• Future Year Fuel Data: This spreadsheet includes three worksheets -
Factors, Gasoline Assignment, and Notes.
Factors worksheet. This worksheet is divided into two sections. The
upper section provides the gasoline parameters used to develop the
multiplicative and additive factors for each PADD or area for future years.
This portion of the spreadsheet includes the following columns: area, fuel
type, fuel description, RVP, oxygen (weight %), aromatics (volume %),
benzene (volume %), olefins (volume %), sulfur (parts per million), E200
(volume % off), and E300 (volume % off). The lower portion of the
spreadsheet provides the additive and multiplicative factors to be applied
to base year, year 2000, or year 2009 gasolines to determine future year
gasoline parameters. It contains columns specifying the area, fuel type,
factor or gasoline identifier (i.e., letters A through CC) and the additive or
multiplicative factors for each gasoline parameter. The factors for the
parameters listed for below for gasolines A through W are multiplicative:
-RVP
-Aromatics
-Benzene
3-30
-------
-Olefms
-MTBE
-ETBE
-TAME
-ETOH
The factors for the parameters listed for below for gasolines A through W
are additive:
-MTBE_M
-ETBE_M
-TAME_M
-ETOH_M
-E200
-E300
Actual values rather than multiplicative or additive factors were provided
for the parameters listed for below for gasolines A through W :
-Sulfur
The factors for the parameters listed for below for gasolines X through CC
are multiplicative:
-RVP
-Aromatics
-Benzene
-Olefms
- Sulfur
3-31
-------
The factors for the parameters listed for below for gasolines X through CC
are additive:
-E200
-E300
Actual values rather than multiplicative or additive factors were provided
for the parameters listed for below for gasolines A through W :
-MTBE
-MTBE_M
-ETBE
-ETBE_M
-TAME
-TAME_M
-ETOH
-ETOH_M
Gasoline Assignment Worksheet. This worksheet indicates for each county
the source of gasoline data for winter and summer for each year from 1999
through 2050. These references include the 1999 Fuel Data spreadsheet,
2000 Fuel Data Spreadsheet, or Gasoline Identifier A through CC.
Notes Worksheet. This worksheet provides any special instructions for
application of the factors and values provided in the Factors worksheet.
Seasonal Fuel Compositing
The fuel data available for several counties indicated that multiple formulations
may be used in a given season. Information at this level of detail were available only from the
sources consulted to prepare the 1999 Fuels Data spreadsheet, and are indicated by an entry in the
3-32
-------
Percentage of Oxygenate Fuel Sale from Federal Highway Survey column. Because NMIM can
only use a single fuel for each month in a particular county, a programming routine was
developed that identified the counties with multiple fuels and composited the two fuels. The
methodology used to calculate the composited fuel weighted the value of the fuel parameters
RVP, Sulfur, Olefms, Aromatics, Benzene, E200, and E300 by the percentage of oxygenate fuel
sale from Federal Highway Survey, as shown below:
Composited parameter value =
((Fuel 1 parameter x Fuel 1 percentage of oxygenate fuel sale federal highway survey) +
(Fuel 2 parameter x Fuel 2 percentage of oxygenate fuel sale federal highway survey))
100
For the oxygenate volume and market share parameters, the composited values
were set equal to the higher of the two possible values. The results of a sample calculation are
provided in Table 3-2.
3-33
-------
TABLE 3-2
Sample Calculation for Composited Seasonal Fuel for FIPS 39001: Adams, OH
Season
RVP
Sulfur
Olefins
Aromatics
Benzene
E200
E300
MTBE volume
MTBE market share
ETOH volume
ETOH market share
ETBE volume
ETBE market share
TAME volume
TAME market share
Oxygenate_Fuel_Sale_
Percentage
Fuell
summer
9.45
297.7
7.26
25.75
0.98
55.84
81.39
0.02
0
10.06
100
0
0
0
0
39.74
Fuell
winter
14.17
249.1
7.65
19.76
0.92
60.10
84.18
0.01
0
9.93
100
0
0
0
0
39.74
Fuel 2
summer
8.59
406.2
10.378
31.057
1.26
46.72
78.90
4.41
100
0
0
0
0
0
0
60.26
Fuel 2
winter
13.86
384.5
9.634
26.71
1.1773
53.80
82.62
1.33
100
0
0
0
0
0.009
0
60.26
Composited
Fuel
summer
8.93
363.08
9.14
28.95
1.15
50.34
79.89
4.41
60.26
10.06
39.74
0
0
0
0
Composited
Fuel
winter
13.98
330.69
8.84
23.95
1.07
56.30
83.24
1.33
60.26
9.93
39.74
0
0
0
0
Interpolation
The fuels data provided by the sources described in Seasonal Fuel Data portion of
this section were only available on a seasonal (i.e., summer or winter) basis. NMEVI requires
fuels data on a monthly basis. To distribute the seasonal fuels over the 12 months in a year, a
programming utility was developed that interpolated the values in a manner similar to that used
by Pechan Associates for RVP values in the 1999 NEI analysis. This methodology uses the
Pechan ASTM RVP classifications by state from the NEI documentation and the RVP schedule
3-34
-------
for ASTM classes A through E. Although this methodology was applied to RVP values only in
the Pechan analysis, it was applied to all fuel parameters for the NMIM effort.
This method was used for the RVP interpolation, because it minimized
differences between NMIM and the NEI results. The RVP schedule presents a stepwise change in
gasoline composition from summer to winter RVP conditions and back. Since the method was to
be used for this key gasoline composition parameter, it was chosen for the other gasoline
parameters in order to keep all the conversions on the most consistent basis possible. Applying
the method in this manner provides stepwise changes in every gasoline parameter on the same
schedule as RVP, over each parameter's winter through summer range of values. The results of a
sample calculation are provided in Tables 3-3 through 3-5.
TABLE 3-3
ATSM RVP Class Assignment for FIPS 39001: Adams, OH
Month
January
February
March
April
May
June
July
August
September
October
November
December
Summer (June value)
Winter (January value)
ASTM RVP Class
E
E
D
D
C
C
C
C
C
C
D
E
C
E
ASTM RVP
Schedule
15
15
13.5
13.5
11.5
11.5
11.5
11.5
11.5
11.5
13.5
15
11.5
15
3-35
-------
Interpolation Factor Calculation
,, .,, T . , .. _, . (Monthly RVP Class - Summer RVP class)
Monthly Interpolation Factor =
(Whiter RVP Class - Summer RVP Class)
TABLE 3-4
Monthly Interpolation Factor Calculation for FIPS 39001: Adams, OH
Month
January
February
March
April
May
June
July
August
September
October
November
December
Summer (June value)
Winter (January value)
ASTM RVP Class
E
E
D
D
C
C
C
C
C
C
D
E
C
E
ASTM RVP
Schedule
15
15
13.5
13.5
11.5
11.5
11.5
11.5
11.5
11.5
13.5
15
11.5
15
Interpolation
Factor
1
1
0.571
0.571
0
0
0
0
0
0
0.571
1
Monthly Interpolation Calculation
Interpolated Monthly Value = summer value +
monthly interpolation factor x (winter value - summer value)
3-36
-------
TABLE 3-5
Sample Monthly Interpolation for Olefins Calendar Year 1999 for
FIPS 39001: Adams, OH
Season/Month
Summer
Winter
January
February
March
April
May
June
July
August
September
October
November
December
Seasonal
Volume
Percent
Olefins
9.14
8.84
Summer
Volume
Percent
Olefins
9.14
9.14
9.14
9.14
9.14
9.14
9.14
9.14
9.14
9.14
9.14
9.14
Interpolation
Factor
1
1
0.571
0.571
0
0
0
0
0
0
0.571
1
Winter Volume
Percent Olefins -
Summer Volume
Percent Olefins
-0.3
-0.3
-0.3
-0.3
-0.3
-0.3
-0.3
-0.3
-0.3
-0.3
-0.3
-0.3
Interpolated
Monthly
Volume
Percent
Olefins
8.84
8.84
8.97
8.97
9.14
9.14
9.14
9.14
9.14
9.14
8.97
8.84
Identification of Unique Gasolines and Population of Gasoline
Following the generation of the full set of monthly gasoline parameters for all
counties, the Microsoft Access programming utility identified the unique set of gasoline
formulations, assigned each a gasoline identification number, and populated the Gasoline,
Gasoline2, GasMTBEPhsOut, Gas2MTBEPhsOut, CountyYearMonth, and CYMMTBEPhsOut
tables.
3-37
-------
Quality Assurance Procedures
The results of the gasoline program were confirmed with "hand" calculations
completed using a spreadsheet. Several base year and at least one future year gasoline
calculation were verified. Oxygenate market share totals were verified by querying the database
to determine if they added to 100 percent for gasolines based on gasoline assignments A through
W in the Gasoline Assignment worksheet. In several cases, the sum of market share data were
either slightly less than or slightly greater than 100%. Upon further investigation, it was also
noted that there were cases where oxygenate volume data were greater than zero but the
corresponding oxygenate market share data were equal to zero, as well as cases where the where
oxygenate market share data were greater than zero but the corresponding oxygenate volume
were equal to zero. Through a review of the raw data and interpolation methodology, it was
determined that these issues were the result of the raw data that were available and the precision
of NMEVI database. For all gasolines where this was noted, the market share or volume data
were reset to zero, and the sum of the market shares for the remaining oxygenates were
renormalized to 100 percent.
In addition, the null value, zero value, maximum and minimum value, parent-
child, and child-parent QA/QC checks described in Section 7.0 were also completed for this
table.
3-38
-------
3.3 Gas2MTBEPhsOut
Multiple fuel formulation data for the same season were available for several
counties in several states. These counties reported using multiple fuel formulations in the same
season. Gasoline2 contains the fuel formulation and market share information for each
individual gasoline. The design of this Gasoline2 table limits the number of fuels that can be
associated with a county to a maximum of two fuels.
Data Source
The data sources used to prepare the fuel formulation information used to
populate Gasoline2 are described in Section 3.2.
Data Population Methodology
To prepare the Gasoline table, counties for which multiple fuels were available for
a specific season were combined into one fuel using a weighted average based on each fuel's
market share. To prepare the Gasoline2 table, each of these fuels were interpolated separately
and added to the table using the interpolation and unique gasoline identification methodology
described in Section 3.2.
Quality Assurance Procedures
The results of the Gasoline2 program were confirmed with "hand" calculations
completed using a spreadsheet. Market share totals were verified by querying the database to
determine if they added to 100 percent. In addition, the null value, zero value, maximum and
minimum value, parent-child, and child-parent QA/QC checks described in Section 7.0 were also
completed for this table.
3-39
-------
3.4 Natural Gas
The Natural Gas table specifies the sulfur content of various natural gas fuels used
in the base year, or anticipated to be used in future years.
Data Source
The Natural Gas sulfur values were extracted from the Pechan tab of a spreadsheet
titled sulfur.xls, forward from Dave Brzezinski, USEPA, on September 19, 2002.
Data Population Methodology
Because of the limited number of natural gas fuels in the baseline and future
years, the NaturalGas table was populated manually. One record was added to the database as
shown in Table 3-6.
TABLE 3-6
Natural Gas Sulfur Values
Natural Gas ID
1
Natural Gas Sulfur Value
30
Based on the information in sulfur.xls, Natural Gas ID 1 was inserted into
CountyYearMonth for all counties for all years.
Quality Assurance Procedures
The contents of NaturalGas were printed and visually compared to the natural gas
fuel specification information provided in the data source listed above. In addition, the null
value, zero value, maximum and minimum value, parent-child, and child-parent QA/QC checks
described in Section 7.0 were also completed for this table.
3-40
-------
3.5 CountyYearMonth
The CountyYearMonth table includes for each month in the base and future years
for each county, an indication of the fuel formulation, diesel formulation, natural gas
formulation, and data source information for both on-road and non-road fuels. It also contains a
reference to alternate gasoline formulation information for the base year where appropriate.
Data Source
The data sources and the Microsoft Access programming utility described in
Section 3.2 were used to assign highway gasoline identification information in the
CountyYearMonth table. The diesel identification information was populated using the Diesel
table data sources. The natural gas identification information was populated using the
Natural Gas table data sources.
Data Population Methodology
The Microsoft Access programming utility described in Section 3.2 was used to
populate CountyYearMonth.
Quality Assurance Procedures
The gasoline assignments and corresponding formulation information for several
counties over several years generated by the Microsoft Access programming utility were
compared with results obtained using a spreadsheet calculation and verified for accuracy.
Queries of base and future year diesel and natural gas assignments were completed and the
results compared to the information available in the Diesel and Natural Gas data sources.
3-41
-------
3.6 Fuel Tables Required to Model No MTBE Phase Out Scenario
To allow NMIM to model a nationwide scenario of no phase out of MTBE, the
following additional fuels tables were developed: Gasoline, Gasoline2, and CountyYearMonth.
These tables were generated using a Microsoft Access utility based on the one
used to generate GasMTBEPhsOut, Gas2MTBEPhsOut, an CYMMTBEPhsOut. The following
modifications were made to the access utility:
• 2003 gasolines were copied forward as is through 2050 for all California counties.
• 2009 gasolines were copied forward as is through 2050 for all remaining counties.
3.6.1 Gasoline
The Gasoline table contains fuel formulation and market share information for
base and future years.
Data Source
The data sources used to prepare the fuel formulation information used to
populate Gasoline are described in Section 3.2 with the following exceptions:
2003 gasolines were copied forward as is through 2050 for all California counties.
• 2009 gasolines were copied forward as is through 2050 for all remaining counties.
3-42
-------
Data Population Methodology
To prepare the Gasoline table, the interpolation and unique gasoline identification
methodology described in Section 3.2 were followed up through calendar year 2003 for
California and up through calendar year 2009 for the remaining counties in the United States.
Quality Assurance Procedures
The results of the gasoline program were confirmed with "hand" calculations
completed using a spreadsheet. Several base year and at least one future year gasoline
calculation were verified. At least one calendar year 2004 gasoline and one calendar year 2050
gasoline for a California county were verified to be the same as the 2003 gasoline for the same
county. In addition, and at least one calendar year 2010 gasoline and calendar year 2050 gasoline
for a non-California county were verified to be the same as the 2009 gasoline for the same
county. Lastly, the null value, zero value, maximum and minimum value, parent-child, and
child-parent QA/QC checks described in Section 7.0 were also completed for this table.
3.6.2 Gasoline!
The Gasoline2 table is identical to Gas2MTBEPhsOut table. Because
Gas2MTBEPhsOut only includes gasolines for calendar year 1999, there are no differences
between Gas2MTBEPhsOut and Gasoline2
Data Source
The data sources used to prepare the fuel formulation information used to
populate Gasoline2 are described in Section 3.2.
3-43
-------
Data Population Methodology
To prepare the Gasoline2 table, the interpolation and unique gasoline
identification methodology described in Section 3.2 were followed.
Quality Assurance Procedures
The data were quality assured by comparing the parameters associated with
several specific Gasoline Identification numbers between Gasoline2 and Gas2MTBEPhsOut to
verify that no changes were made.
3.6.3 CountyYearMonth
The CountyYearMonth table includes for each month in the base and future years
for each county, an indication of the fuel formulation, diesel formulation, natural gas
formulation, and data source information for both on-road and non-road fuels for the MTBE
phase out with no oxygenate replacement scenario. It also contains a reference to alternate
gasoline formulation information for the base year where appropriate.
Data Source
The data sources and the Microsoft Access programming utility described in
Section 3.2 were used to assign highway gasoline identification information in the
CountyYearMonth table, with the exception of the following:
• 2003 gasolines were copied forward as is through 2050 for all California counties.
• 2009 gasolines were copied forward as is through 2050 for all remaining counties
3-44
-------
The diesel identification information was populated using the Diesel table data
sources. The natural gas identification information was populated using the NaturalGas table
data sources.
Data Population Methodology
The Microsoft Access programming utility described in Section 3.2 was used to
populate CountyYearMonth.
Quality Assurance Procedures
At least one calendar year 2004 gasoline assignment and one calendar year 2050
gasoline assignment for a California county were verified to be the same as the 2003 gasoline
assignment for the same county. In addition, and at least one calendar year 2010 gasoline
assignment and one calendar year 2050 gasoline assignment for a non-California county were
verified to be the same as the 2009 gasoline assignment for the same county.
3-45
-------
4.0 VEHICLE TABLES
Emissions inventory calculations are significantly impacted by vehicle population
and travel data. The sources of this information in the NMEVI database are described in the
sections below.
4.1 AverageSpeed
The AverageSpeed table presents the average speed for each vehicle traveling on
each HPMS roadway type.
Data Source
Modeling files for June 2002 update to the 1999 NEI, extracted from the VMT
table, vmt99_f_m6_with8statesupdate.dbf
Data Population Methodology
The data in the VMT table was extracted and run through two processing steps.
The source data included speeds assigned to the 28 vehicle classes. The first processing step
determined every unique combination of road type, vehicle class, and speed. The second step
populated a table with speed based on the 16 vehicle classes and road type.
Quality Assurance Procedures
In the first processing step, checks verified that one and only one speed existed for
each combination of road type and vehicle class. The overall effect of the second step was to
"merge" gasoline and diesel vehicle classes into 16 vehicle types. Each type was then verified to
confirm that the corresponding gasoline and diesel classes had the same speed, that there was
exactly one combination of each, and that each required index combination in the output was
populated.
4-1
-------
4.2 BaseYearVMT
The BaseYearVMT table contains annual vehicle miles traveled (VMT) data for
every county-level area, for every combination of vehicle class and road type. This is VMT for
the NMEVI "base year," calendar year 1999.
Data Source
Base year VMT data were collected from two related sources. The first was the
file vmt99_f_m6_with8statesupdate.dbf in the modeling files for June 2002 Update to the 1999
NEI. These data were used for all 50 states and Washington, DC. The second source was the file
vmt99_n_m6.dbf in the modeling files for the Fall 2001 1999 NEI update. This second source
was used for VMT for Puerto Rico and Virgin Islands.
Data Population Methodology
Three processes were used to populate the base year VMT table. For Washington,
DC and all 50 states except California, the VMT information from the first data source was used
as directly as possible. For these 50 state-level areas, a record was written for every record in the
original table that included only the fields required for import to the NMEVI table. The 3-digit
NEI "SCCRT" field was converted to a two-digit road type code by extracting the first two digits.
For California, the source data were not allocated by road type. The road type
allocation was made using a national average prepared from the data for the other 49 states and
DC. Total VMT excluding California was calculated from the source data, and VMT fraction by
road type and vehicle class was then calculated. Using the VMT fractions, each California record
in the source data were processed. The total VMT was allocated to the 12 road types, and 12
base year VMT records were written. Table 4-1 lists the VMT road type fractions used.
4-2
-------
TABLE 4-1
National-Average VMT Fraction by Road Type Used for California VMT Data
Vehicle
Class
1
2
3
4
5
6
7
8
9
10
11
12
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
HPMS Roadway Type
11
0.097
0.101
0.101
0.101
0.101
0.106
0.136
0.136
0.135
0.134
0.135
0.159
0.101
0.103
0.111
0.138
0.139
0.140
0.138
0.138
0.161
0.161
0.107
0.159
0.162
0.161
0.103
13
0.091
0.094
0.094
0.093
0.093
0.094
0.120
0.119
0.120
0.120
0.120
0.144
0.097
0.092
0.097
0.122
0.122
0.122
0.122
0.122
0.144
0.144
0.101
0.140
0.142
0.142
0.094
15
0.064
0.066
0.066
0.066
0.066
0.066
0.082
0.081
0.082
0.084
0.083
0.101
0.069
0.064
0.067
0.084
0.084
0.083
0.085
0.085
0.100
0.100
0.071
0.095
0.097
0.098
0.066
17
0.078
0.081
0.081
0.080
0.080
0.080
0.101
0.100
0.102
0.103
0.102
0.125
0.083
0.078
0.080
0.103
0.103
0.101
0.103
0.103
0.122
0.122
0.087
0.119
0.119
0.120
0.081
19
0.022
0.022
0.022
0.022
0.022
0.022
0.027
0.027
0.027
0.027
0.027
0.033
0.021
0.021
0.021
0.027
0.027
0.026
0.027
0.028
0.033
0.033
0.024
0.031
0.031
0.031
0.021
21
0.049
0.049
0.049
0.049
0.049
0.049
0.059
0.058
0.060
0.061
0.060
0.074
0.055
0.048
0.047
0.060
0.059
0.058
0.060
0.061
0.072
0.072
0.055
0.069
0.068
0.069
0.049
23
0.140
0.138
0.137
0.139
0.139
0.138
0.114
0.115
0.114
0.113
0.114
0.085
0.131
0.140
0.138
0.112
0.112
0.114
0.111
0.111
0.089
0.089
0.130
0.094
0.093
0.092
0.138
25
0.053
0.052
0.052
0.052
0.052
0.052
0.042
0.043
0.042
0.042
0.042
0.032
0.049
0.053
0.051
0.041
0.041
0.042
0.041
0.041
0.032
0.032
0.049
0.034
0.033
0.033
0.052
27
0.149
0.144
0.144
0.145
0.145
0.143
0.117
0.118
0.117
0.116
0.117
0.090
0.145
0.147
0.141
0.115
0.115
0.115
0.114
0.114
0.090
0.090
0.137
0.095
0.093
0.093
0.145
29
0.117
0.114
0.114
0.114
0.115
0.113
0.092
0.093
0.092
0.091
0.092
0.071
0.113
0.116
0.112
0.090
0.090
0.091
0.090
0.090
0.071
0.071
0.109
0.075
0.073
0.073
0.114
31
0.051
0.049
0.049
0.049
0.049
0.049
0.039
0.039
0.039
0.039
0.039
0.030
0.049
0.049
0.048
0.038
0.038
0.039
0.038
0.039
0.030
0.031
0.047
0.032
0.031
0.031
0.049
33
0.091
0.089
0.089
0.089
0.089
0.087
0.070
0.071
0.070
0.070
0.070
0.054
0.087
0.090
0.086
0.069
0.069
0.070
0.069
0.069
0.054
0.055
0.084
0.057
0.056
0.056
0.089
Total
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
For Puerto Rico and Virgin Islands, data from the second source and the
methodology described for the non-California data above was used. The source data were read
4-3
-------
and records were written for every Puerto Rico or Virgin Islands record, with the required index
fields and with the SCCRT field corrected to a two-digit road type.
After the source data were imported to the final table for all areas, records with
zero VMT were added for every combination of VMT and road type that was not represented in
the source data. For the data imported directly from the first source, this was approximately 30%
of all possible area-vehicle-road combinations.
Quality Assurance Procedures
The base year VMT data were checked for duplicate records, and VMT values
were checked by testing several VMT sums against the source data. For state-level areas, the
total VMT by vehicle class was compared. For county-level areas, total VMT, VMT by road
type, and VMT by vehicle class were checked.
4.3 VMTGrowth
The VMTGrowth table contains percentage growth factors for scaling VMT from
one calendar year to the following year. It holds growth factors for every vehicle class in every
county, for every calendar year from 1999 to 2050. To calculate VMT for calendar year 2010
from calendar year 2009, the VMTGrowthRate data for calendar year 2010 is used. The growth
factors for calendar year 1999 are zero. The factor is a positive or negative value representing the
percentage change that a specific vehicle class in a specific county will change from the previous
year to the year selected. To derive VMT for a calendar year 2025 case, the 1999 base VMT is
obtained and then multiplied by 1 + VMTGrowthRate(year) for every year from calendar year
2000 to calendar year 2025.
4-4
-------
Data Source
VMT growth data were collected from two primary sources:
The BaseYearVMT table, based on data from the calendar year 1999 NEI
modeling files; and
• VMT estimates for calendar year 2007, calendar year 2020, and calendar
year 2030 provided by the EPA from appendix tables in the support
documentation for the 2007 Heavy Duty Diesel Rule (HD2007) (files V-
2.xls, V-3.xls, and V-4.xls).
Four state-level areas (AK, HI, PR, and VI) were not included in the HD2007
data. Future-year VMT for these areas was estimated based on average VMT growth in the other
state-level areas.
Data Population Methodology
Several steps were required to prepare VMT Growth data. The overall process
included: A) preparing complete sets of VMT data for "anchor" calendar years 1999, 2007, 2020,
and 2030; B) interpolating between anchor years to derive complete sets of VMT for all years
from 1999 through 2030; C) and extrapolating the 2030 VMT and preparing complete sets of
data for 2031 through 2050; and D) computing a percentage VMT growth for each year from
2000 to 2050 using the VMT data.
Each of these overall processes involved several separate processing steps. Figure
4-1 shows an overall view of the processing.
4-5
-------
Future Years
Base Year 1999
BOOBaseYearVMT
CY 1999
All States, All FIPS
NMIM Roadway Types
MOBILES Vehicle Classes
FOOHDRule
CY 2007, 2020, 2030
Lower 48 States
FIPS Errors
MOBILE5 Vehicle Classes
BOlCombineBaseYear
CY 1999
All States, All FIPS
MOBILES Vehicle Classes
FOlConvertHD
CY 2007, 2020, 2030
Lower 48 States
FIPS Errors
MOBILES Vehicle Classes
F02FixFIPS
CY 2007, 2020, 2030
Lower 48 States
All FIPS
TOlTotals
CY 1999, 2007, 2020, 2030
Lower 48 States, All FIPS
VMT Totals
FOSAddMissingST
CY 2007, 2020, 2030
AK. HI. PR. VI
All Years
SOlAnchorYearsComplete
CY 1999. 2007. 2020. 2030
All States, All FIPS
S02InterpolateVMT
All CY 1999-2050
All States, All FIPS
SOSDeltaVMT
All CY 1999-2050
All States, All FIPS
VMT Growth Factors
FIGURE 4-1. VMT Growth Data Sources and Methods
4-6
-------
Base data "BOOBaseYearVMT" is the data in the BaseYearVMT table,
discussed in Section 4.1. Base data "FOOHDRule" is the data from the HD2007 Rule
appendices.
In step "BOlBaseYearVMT" the BaseYearVMT table was queried for the sum
of VMT by area and vehicle class, year, and area. This result was saved as the "complete" set of
VMT for anchor year 1999. It included every required combination of vehicle class and area, on
the same basis as other NMEVI tables.
In step "FOlConvertHD" the HD2007 data were converted from a MOBILES
vehicle class basis to a MOBILE6 vehicle class basis. The methodology presented in Section 5
of the MOBILE6 User's Guide was used to convert to the 16 MOBILE6 vehicle types, and
relative VMT was calculated from MOBILE6 defaults to assign VMT for the 16 types into the 28
vehicle classes. There are some important aspects of the MOBILE6 User's Guide Chapter 5
method that impacted QA tests:
• The conversion method preserves total VMT for all vehicles;
• The method preserves VMT by MOBILES vehicle class for the five
vehicle class "groupings" listed in section 5.3.2 of the MOBILE6 User's
Guide;
• The method does not preserve VMT by the eight MOBILES vehicle
classes because it involves a fuel independent sum that is distributed into
classes based on MOBILE6 defaults; and
• The method allocates VMT to every MOBILE6 vehicle class for calendar
year 2007, and all classes except LDDT12 for calendar year 2020 through
2030.
The conversion to fuel-independent vehicle groupings and then back to vehicle
classes means that, in general, the MOBILES diesel-gasoline ratios are not preserved, and the QA
checks had to compare with source data after at least one processing step. Also, this step cannot
be reversed, there is no path to convert the VMT by 28 classes back to the source VMT by eight
MOBILES classes.
4-7
-------
In the MOBILE6 default population, vehicle class LDDT12 is not sold after
model year 1986, and the last age-25 vehicles in this class are retired after calendar year 2010.
LDDT12 VMT will be zero for calendar year 2011 and later. After converting the source data,
calendar year 2007 included VMT for the class, and calendar years 2020 and 2030 did not.
The conversion method assigned VMT without any reference to the calendar year
1999 data. This created some conflicts where a class was not present in a specific area in
calendar year 1999 but had VMT in calendar year 2007. These cases were located using QA
checks, and a set of post-fixes was applied in step F01 after the basic vehicle class conversion.
Table 4-2 lists the areas and vehicle classes that were included in the post-fixes. In the table, the
"case" labels are in the form "NoVVVV" to indicate that VMT for class VVVV should be zeroed
in future years, because it is zero in calendar year 1999. To zero the VMT for a class, its VMT
was first added to the class with the same type but different fuel: for case "NoHDGVSa," the
HDGVSa (gasoline) VMT was added to class HDDVSa (diesel), and then HDGVSa was set to
zero. For the "NoLDDT" case, the LDDT12 VMT was allocated to LDGT1 and LDGT2 based
on the existing relative VMT in the two classes. The same method was used for LDDT34,
LDGT3, and LDGT4. For the "NoMC" case, the motorcycle VMT was added to the LDGV
VMT.
4-8
-------
TABLE 4-2
Post-fixes to Vehicle Class Conversions
Case
NoLDDT
NoLDDV
NoMC
NoHDGVSa
State FIPS
6
6
47
6
16
20
28
30
31
35
38
46
48
8
County FIPS
3,51,91
3,91
65
3,49,91
25
33,67,71, 189, 199
55
19, 37, 59, 69
5, 7, 9, 75, 91, 103, 113, 115, 117, 171, 183
21
87
17
23, 33, 75, 79, 169, 247, 261, 263, 269, 301, 311,
345,357,383,393,443,495
53
In step "F02FixFIPS" the differences between area FIPS code assignments were
resolved by converting the HD2007 area assignments to the NMTM basis. The basis used for the
FIPS reassignments was an NEI document provided by EPA. Most of the differences between
HD2007 and NMEVI FIPS codes were addressed in this document.
The VMT reassignments were handled as a set of special cases. There were a set
of six cases for various types of area reassignments, including a base case with no conversion.
Each county-level area in the F01 output was converted by one of the six cases and added to the
step F02FixFEPS output. The cases were run independently and verified using QA checks. Some
of the checks were specific to the conversion case, and others compared all of the input and
output data. The cases include:
• Copy unchanged (base case): No FEPS code conversion, A -> A.
4-9
-------
Replace: The previous code is replaced with no other changes, A->B.
Split with one new: Area split in two, one part keeping the same code, A -
>A+B.
• Split with both new: Area split in two, with previous code dropped, A ->
B+C.
Split two to three: New area split from two existing, B+C-> A+B+C.
Split to two existing: Area reassigned to two other areas, A+B+C -> B+C.
The output for step F02FixFIPS was a set of complete VMT records for the
anchor calendar years 2007, 2020, and 2030, for the "lower 49" state-level areas (the lower 48
states and Washington, DC).
In step "TOlTotals" the VMT totals for the lower 49 were calculated for the
anchor years, from the output of steps BOlCombineBaseYear and F02FixFIPS. The result was
total VMT by vehicle class for all four anchor years.
In step "FOSAddMissingST" the future anchor year VMT for AK, HI, PR, and
VI was estimated based on "national average" values from the lower 49 totals. The average
VMT growth by vehicle class for 1999 through 2007 was calculated from the totals, and the
calendar year 2007 VMT for each county-level area in AK, HI, PR, and VI was calculated from
these growth factors and the base year VMT data. The same procedure was used to extrapolate
the calendar year 2007 VMT to calendar year 2020, and for calendar year 2020 to calendar year
2030. The final output from step FOSAddMissingST was a set of complete VMT values for the
future anchor years, for the states not covered in the HD2007 data.
In step "SOlAnchorYearsComplete" the anchor year VMT data from steps
BOlBaseYearVMT, F02FixFIPS, and FOSAddMissingST were compiled to prepare for the
following steps. QA checks that compared VMT changes from one anchor year to the next were
run at this time. The VMT changes identified cases where the NEI data and the converted
HD2007 data had conflicts for specific areas and vehicles, such as the list of special cases
4-10
-------
discussed in step FOlConvertHD. Completeness checks were run to verify that records for every
county and every vehicle class in every anchor year existed.
In step "S02InterpolateVMT" the anchor year VMT was copied, and VMT data
were interpolated or extrapolated for every additional calendar year from 2000 to 2050. The
range of calendar years was handled as four "spans," calendar year 1999 through calendar year
2007, calendar year 2007 through calendar year 2020, calendar year 2020 through calendar year
2030, and calendar year 2030 through calendar year 2050. For the first three spans, VMT for
intermediate years was interpolated, and for the final span the calendar year 2030 VMT was
extrapolated.
The interpolation/extrapolation method assumes constant growth in VMT miles,
rather than a constant growth ratio. In extrapolating, the annual VMT growth in miles from the
last interpolated year was used for all following years. Figure 4-2 illustrates the handling of the
calendar year spans. The VMT data plotted in the figure is for Cochran County, TX (FIPS
48017).
4-11
-------
Figure 4-2. Anchor Years and Interpolation Spans for VMT Growth
30
25
20
H
10
0
1995 2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Calendar year
-------
FIGURE 4-2. Anchor Years and Interpolation Spans for VMT Growth
In step "SOSDeltaVMT" the VMT by calendar year results from
S02InterpolateVMT were compared and growth rates were generated. The growth rates were set
to zero for calendar year 1999, and calculated as a percentage change in VMT for each calendar
year from 2000 through 2050. With a constant growth in VMT between spans, the percentage
growth changed for each year. Figure 4-3 illustrates the growth factor change characteristics for
the same data plotted in Figure 4-2.
4-12
-------
Figure 4-3. Percentage Growth Rate for the VMT Growth Table
1 A.
1 .*4
1 9
1 .Z
i
1
^ n 8
ox U.O
o
r^ 0 &
O u.o
n A
U.4
09
.z
n
i
1 n
UD",
'"*
D D 1
1 D D D D 1
1 D D D D
D D D D 1
1 D D D D 4
t a a a a i
1
u
1995 2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050
Calendar Year
-------
FIGURE 4-3. Percentage Growth Rate for the VMT Growth Table
Quality Assurance Procedures
A variety of QA checks were made in the overall process, because the individual
processing steps were quite different and handled a number of special cases. Some of the QA
checks can be summarized by processing step.
BOlBaseYearVMT: VMT totals by vehicle were checked at the state level and
the county level.
FOlConvertHD: The vehicle class conversion is not reversible, and does not
preserve VMT by all 8 MOBILES classes. The output from this step was checked against the
source data to confirm overall VMT totals at the state and county level. The source data were
4-13
-------
then processed into VMT totals by county for the five MOBILES vehicle "groupings" listed in
the MOBILE 6 User's Guide, and the corresponding totals were calculated and compared for the
output data after the vehicle class post-fixes. The post-fixes preserved VMT by the five
"groupings" for all cases except the "NoMC" case. For the four "NoMC" cases, the input
motorcycle VMT was added to the input LDV VMT and checked against output LDV VMT.
F02FixFIPS: This step preserves total VMT by state, and it preserves total VMT
by county for the "copy-unchanged" base case counties. Each individual FIPS correction
preserves VMT for the counties involved. QA tests were made for total VMT by class at the
state level. For the copy-unchanged counties and the renamed counties, the input and output
VMT was compared directly by vehicle class at the county level. The other FIPS corrections
involved one to three input counties and two to three output counties. For these cases, VMT was
totaled over the input and output counties and then compared.
TOlTotals: There were relatively few checks that could be made to this data. The
distribution of VMT by vehicle class was calculated and compared to the distribution from the
larger states. This was not an exact comparison, but was used to qualitatively assess the
difference between the final VMT distribution and the distributions for individual states.
FOSAddMissingST: This step was designed to preserve overall VMT growth
exactly, and relative VMT growth by vehicle class only approximately. The total percentage
VMT growth by county for the four states calculated was compared to the percentage VMT
growth calculated from the TOlTotals data.
SOlAnchorYearsComplete: This step consisted of consolidating data from
several preceding steps. There were no data manipulations in this step to be validated, but it was
a convenient point to perform checks across all of the anchor year data. Checks were run for
completeness to verify that every required combination of county and vehicle class was created in
the F02FixFIPS and FOSAddMissingST steps.
4-14
-------
S02InterpolateVMT: The processing routines for this step compared adjacent
anchor years for VMT by vehicle class and county. This comparison determined the cases where
a vehicle class in a county was added or dropped from one anchor year to the next. Resolving
these conflicts led to the set of post-fixes applied in the FOlConvertHD step. A separate test was
run for the input and output data for this step, checking for a complete set of unique records for
every state, county, and vehicle class. In this test, the state and county FIPS IDs were also
compared to data read from the EPA CHIEF FIPS list.
SOSDeltaVMT: Because the final calculated growth factors were stored with
fixed numeric precision, it was not possible to perform exact comparisons between the input
VMT and the cumulative growth factors. Qualitative checks verified that the input VMT,
multiplied by appropriate growth factors, matched the output VMT within a reasonable error
tolerance. The calculated growth factors were also checked against the range limits allowed for
the VMTGrowthRate field in the database design. Also, growth factors of-100% were checked
against the expected cases for dropped vehicle classes.
4.4 VMTMonthAllocation
This table contains, for a combination of vehicle class and road type, the fraction
of annual VMT that should be allocated to each month of the year.
Data Source
The data were copied from a table in the October 2001 Draft 99 NEI
documentation. In Table 4-3, the columns marked "Original" list the actual values copied from
the source document.
4-15
-------
TABLE 4-3
Original and Adjusted VMTMonthAllocation Values
Vehicles:
Roadway:
Month
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Sum
Light Duty (LDV, LDT, MC)
Rural
Urban
HDV
All
Allocation Values
Original
0.0744
0.0672
0.0805
0.0779
0.0805
0.0942
0.0974
0.0974
0.0844
0.0872
0.0844
0.0744
0.9999
Adjusted
0.0744
0.0672
0.0805
0.0779
0.0805
0.0942
0.0975
0.0974
0.0844
0.0872
0.0844
0.0744
1.0000
Original
0.0806
0.0728
0.0859
0.0832
0.0859
0.0864
0.0893
0.0893
0.0808
0.0835
0.0808
0.0806
0.9991
Adjusted
0.0806
0.0728
0.0860
0.0833
0.0860
0.0865
0.0894
0.0894
0.0809
0.0836
0.0809
0.0806
1.0000
Original
0.0861
0.0778
0.0842
0.0815
0.0842
0.0815
0.0842
0.0842
0.0824
0.0852
0.0824
0.0861
0.9998
Adjusted
0.0862
0.0778
0.0842
0.0815
0.0842
0.0815
0.0842
0.0842
0.0824
0.0852
0.0824
0.0862
1.0000
Data Population Methodology
The source data had been published as fractions showing four decimal places.
Because of rounding errors, the total annual allocation did not sum to one. In order to force the
annual sums to be one, the original values were adjusted by a correction factor and then rounded
again to four decimal places. In Table 4-3, the columns marked "Adjusted" list the adjusted
values used in the database.
4-16
-------
The source data included three 12-month allocation profiles that were each used
for particular combinations of road type and vehicle class. In order to generate all of the input
records required for the NMIM table, each of the allocation profiles was written out to all of the
combinations of vehicle class and road type for which it applied. The light-duty/rural profile, for
example, was written out for every combination of six light-duty vehicle types and six rural road
types. Table 4-4 shows how the original combinations of vehicle type and road class were
applied to the 16 vehicle types and 12 roadway types used in the database.
Quality Assurance Procedures
The QA check for completeness required that every combination of month,
vehicle type, and road type identified a unique record with an allocation factor within the table's
valid data range. The check for annual totals requires that, for every combination of vehicle and
road, the twelve monthly allocation factors should sum to one. The source data were adjusted, as
shown in Table 4-3, to meet this requirement.
4-17
-------
TABLE 4-4
Conversion of Roadway and Vehicle Types for VMTMonthAllocation Data.
Vehicle Types
LDV, LOT, MC
HDV
Roadways
Rural
Urban
MOBILE6 VehicleTypes
1:
2:
3:
4:
5:
16
6:
7:
LDV
LDT1
LDT2
LDT3
LDT4
:MC
HDV2B
HDV3
8: HDV4
9:
10
11
12
13
14
15
HDV5
:HDV6
:HDV7
: HDV8A
: HDV8B
:HDBS
:HDBT
HPMS Codes
11
13
15
17
19
21
23
25
27
: Rural, Interstate
: Rural, Other Principal Arterial
: Rural, Minor Arterial
: Rural, Major Collector
: Rural, Minor Collector
: Rural, Local
: Urban, Interstate
: Urban, Non-Interstate Freeway
: Urban, Other Principal Arterial
29: Urban, Minor Arterial
31
: Urban, Collector
33: Urban, Local
4-18
-------
5.0
INSPECTION AND MAINTENANCE (I/M) PROGRAM TABLES
Several types of inspection and maintenance (I/M) program data are used by
NMIM: Stage 2 refueling program efficiency, anti-tampering program information, an I/M
program information for multiple vehicle classes over multiple years. The sections below
describe this information.
5.1
County Year
The County Year table contains stage 2 refueling program efficiency data,
references to external anti-tampering program files, and references to external I/M program files
for all counties from calendar year 1999 through calendar year 2050. The data sources, data
population methodologies, and QA procedures used for each type of data in County Year are
described below.
Data Source
The Data Sources used to populate the County Year table are listed in Table 5-1.
TABLE 5-1
CountyYear Data Sources
Type of Data
Data Source
Stage 2 refueling program efficiency
NEI Fall 2001 Update files, stage2dat.xls and
OOtables.wpd
Anti-tampering program file name and files
NEI Fall 2001 Update, MOBILE6 input files and
Trends99 Pointer.dbf
I/M program file name and files
Base Year: NEI Fall 2001 Update MOBILE6 input
files and Trends99_im.xls.
OBD Schedules: File Model.wpd, "Major Elements of
Operating I/M Programs (as of 3/02)". A 12/1999
version of this document is available on the EPA
OTAQ Web site:
http://www.epa.gov/otaq/epg/b99008.pdf.
Other future programs: File Counties.wpd, "States and
Counties with I/M programs".
5-1
-------
Data Population Methodology
The procedures used to populate the County Year table are described in the
sections below.
Stage 2 Refueling Efficiency
The Stage 2 refueling efficiency programs were designated with either a "1" or a
"0" in the spreadsheet stg2dat.xls for each county. The contents of OOTables.wpd indicate that a
value of "0" in stg2dat.xls means that Stage 2 refueling programs are not in effect and a 0%
applies, while a values of "1" indicates that Stage 2 refueling programs are in effect and assumed
to be 95% effective.
Records from the NEI Fall 2001 Update stg2dat.xls file were joined by FIPS code
with records in the County Year table. For counties which had a Stage 2 refueling program in
effect, the Stage2Pct field was populated with a 95. The base year Stage 2 refueling efficiency
values were assumed to be in effect for all future years.
Fifteen mismatches between the NEI Fall 2001 Update stg2dat.xls spreadsheet
and the FIPS codes in the County Year table were noted. The differences in FIPS codes were
expected because the NEI data were based on FIPS numbering with several differences from the
NMEVI numbering. Table 5-2 lists the fifteen FIPS codes used in stg2dat.xls that were remapped
or corrected for use in NMEVI. If no information existed for a particular county, it was assumed
no Stage 2 refueling program was in effect in that county.
5-2
-------
TABLE 5-2
County FIPS Codes in NEI Stage 2 Refueling Data Not Used in NMIM.
County in stg2dat.xls
FIPS
02010
02140
02231
02990
02991
02992
02993
02994
02996
02998
02999
12025
29193
30113
46131
ST
AK
AK
AK
AK
AK
AK
AK
AK
AK
AK
AK
FL
MO
MN
SD
COUNTYNM
Aleutian Islands Ed
KobukEd
Skagway-Yakutat Ed
Upper Yukon Ed
Seward Ed
Kuskokwim Ed
Bristol Bay Borough
Angoon Ed
Cordova-Mccarthy Ed
Outer Kethcikan Ed
Barrow Ed
Dade Co
Ste. Genevieve Co
Yellowstone Natl Par
Washbaugh Co
Comments
Correct for new counties and post- 1980 subdivision in AK.
Dade renamed Miami-Bade County.
Corrected numbering to 29186, per 1979 FIPS correction.
Yellowstone NP assigned to neighboring counties.
Washbaugh absorbed into neighboring counties.
Anti-tampering Program File Names and Files
The base year anti-tampering program data were retrieved by searching the NEI
Fall 2001 Update MOBILE6 input files for the command "anti-tamp." For each MOBILE6 file
that included this command, a corresponding anti-tampering program file was created by copying
the series of parameters that followed the command into a new file. For example, the MOBILE6
input file N0202010.IN includes the following:
ANTI-TAMP
86 68 50 22222 11111111 1 22 095. 22112222
These data were copied to a text file and saved as atp02020.txt. The counties to
which this anti-tampering program applied were determined using the Trends99_pointer.dbf file,
5-3
-------
which notes all of the counties that used the MOBILE6 input file from which the information
was extracted.
Records from the Trends99_pointer.dbf file were joined by FIPS code with
records in the County Year table. For counties that participated in the anti-tampering program, the
ATPFileName field was populated with the appropriate file name. The base year anti-tampering
program data were assumed to be in effect for all future years.
I/M Program File Names
The methods used to develop the I/M program file names and load the
combination of file names and program schedules into table County Year are described below.
The contents of the I/M files, and the I/M program implementation schedule that is reflected in
County Year, is described in more detail in the following section.
The files that describe I/M programs in use in the base year were all derived from
the NEI modeling files. The names of the files were preserved, although the file contents were
updated as needed to reflect future year programs. For these files and I/M programs, the mapping
of counties to I/M files in Trends99_im.xls is identical to the mapping defined in the County Year
table, for the 1999 bas year only. After the base year, I/M programs are added, modified, and
dropped, and the County Year data reflects this.
Additional files were required to describe programs implemented after 1999. The
file names were developed on a case by case basis, but the naming conventions matched the NEI
files as closely as possible.
The data used to load the IMFileName field was written using the information in
the final schedule described below. The schedule table included every county that would
implement a program in any year from 1999 forward. For every such county, records were
generated for import to County Year for the first implementation year and all following years. I/M
programs were assumed to remain in place indefinitely once started.
5-4
-------
I/M Programs and Implementation Schedules
Several initial processing steps were performed on the I/M data. For the NEI I/M
files, program ParselM.awk was used to read all of the files and extract the I/M program details.
Some minor changes were made to the files at this point to improve consistency between the
programs. For example, in the ctim98.im file, the upper model year limit was increased from
2020 to 2050.
The program list in the "States and Counties with I/M programs" document
(Counties.wpd) was reformatted and loaded into a spreadsheet-based table. A FIPS code was
identified for all of the counties listed, and a simple program name was generated from the brief
program description in the document.
A number of changes were made to the program list in the "Major Elements of
Operating I/M Programs" document (Model.wpd) to help extract program descriptions. The table
was exported to a spreadsheet and reformatted. Each "program" row in the original table
represents a set of two to five I/M "programs" in MOBILE6 input data. A table of simplified
program names was generated for use in merging the program descriptions.
As a start for resolving differences in the three sources, a table was prepared
listing the states which had I/M programs defined in each source. There were differences for five
states, and each one was examined to determine the source of the differences. There were cases
in which the NEI data were missing post-1999 programs as expected (LA, NH), cases in which
the NEI included discontinued programs not listed in the other sources (FL, MN), and one case in
which the Counties list was missing a program that was included the other two sources (ID). At
this point, a merged list of state programs was prepared that included all programs listed in the
three sources.
The sources were then compared on a county-by-county basis, and the merged list
of state programs was expanded to include all counties listed in the sources. All of the state-level
and county-level differences in the merged list were resolved, so that each county was assigned to
5-5
-------
a specific state-level I/M program, and the state-level programs and county assignments were as
consistent as possible with the data sources. The majority of the differences noted at this level
could be tracked to the state-level differences, particularly the cases in which programs were
added after 1999, and the cases in which the NEI data described programs not included in the
other sources. In addition, there were a number of cases in which the Counties.wpd list was
missing a county or had an error in county names. Table 5-3 lists the major differences between
the data sources and the way that each was resolved.
This merged list of counties was next used as the starting point for generating a
master schedule table. Each county was mapped to the first year in which it had a program, the
year in which the program added OBD testing, and the year in which any other program changes
were made. The specific program files to be used were also identified.
After cross-checking with the actual I/M files, the master schedule table was used
to generate the data required for the County Year table.
To develop the contents of the I/M files, there were two general cases. For
programs included in the 1999 base year, the base year NEI file was modified to add future-year
changes (primarily OBD program and exhaust test changes). For new programs, the base year
I/M files were used as examples in developing new files. "Generic" program files were first
developed for Enhanced, LowEnhanced, and OTRLowEnhanced programs. These files were used
to develop state- and program-specific I/M files. Table 5-4 lists the eight new program files that
were developed.
5-6
-------
TABLE 5-3
Examples of Differences in I/M Program Data
State
Colorado
Florida
Georgia
Idaho
Kentucky
Louisiana
Maryland
Massachusetts
Minnesota
Missouri
New Hampshire
New Jersey
New York
North Carolina
Oregon
Pennsylvania
Rhode Island
Utah
Differences and Resolution
El Paso County (Colorado Springs) uses Denver program in NEI.
Resolution: Re-assign to use the Colorado Springs file CO95C.IM.
No information on Florida in Counties/Model
Resolution: I/M program discontinued after 1999.
NEI files use 1992 program start, but Model says 10/1998
Resolution: Leave NEI start in place for better NMIM/NEI consistency.
Not listed in Counties.
Resolution: Use NEI/Model as-is.
Northern Kentucky counties: Not listed in Counties.
Resolution: Use NEI/Model as-is.
Not listed in NEI.
Resolution: Add new program 2002 start.
Counties missing Baltimore City (likely a FIPS code issue)
Resolution: Use NEI/Model as-is.
Model indicates MAS 1 test, NEI uses Idle.
Resolution: Transition to MAS 1 test in MA95.IM.
No information on Minnesota in Counties/Model
Resolution: I/M program discontinued after 1999.
Counties list does not include Franklin county.
Resolution: Use NEI/Model as-is.
Model indicates IM240 test, NEI uses Idle.
Resolution: Transition to IM240 test.
Not listed in NEI.
Resolution: Add new OBD program 2002 start.
Model indicates ASM5015 test, NEI uses Idle.
Resolution: Transition to ASM5015 test.
Counties indicates OTR Low-Enhanced program for non-NYC counties.
Resolution: Add separate program for upstate counties.
NEI and Counties show several differences in NC county list.
Resolution: Add Cabarrus, Orange, Union counties to Basic.
Retain NEI counties not in Counties list: Davidson, Davie, Granville.
Counties includes Columbia, Yamhill counties.
Resolution: Add counties to enhanced program.
Counties includes several counties added.
Resolution: Add counties to new program.
Model indicates RI2000 test, NEI uses Idle.
Resolution: Transition to RI2000 test.
Counties indicates that Weber and Utah counties are in different programs,
NEI and Model have Utah grouped with Weber.
Resolution: Use NEI/Model data.
5-7
-------
TABLE 5-4
New External IM Program Files
File
GA01.IM
IN01.IM
LAOO.IM
NH02.IM
NY01.IM
NC01.IM
OR01P.IM
PA01OLE.IM
Comments
Add counties to program in GA99.IM.
Add county to program in IN97.IM.
New OBD program.
New OBD program.
New program for upstate counties.
Add counties to program in NC87.IM.
Add counties to program in OR98P.IM.
New program for additional counties.
To modify the base year file to add future year I/M programs, the
"rewritelM.awk" script was used. The base year details were read in, and a set of modified
programs was written to the output file.
Quality Assurance Procedures
MySQL queries were run to ensure that there were 52 records for each of the
3,222 counties (one of each year), and 3,222 records for each of the 52 years.
The state-level and county-level program comparisons were checked manually
against the source data. The master program schedule table was checked against the county-level
comparison, and the file names in the schedule were checked against the trends99_im.xls data.
When the I/M files were prepared, a set of MOBILE6 runs were made that
exercised every program file for calendar years 1999 and 2007. For the programs present in the
NEI data, a baseline run was completed using the original I/M files for the same years. The ratio
of emissions for the baseline results and the results with the new files was calculated, and a table
of ratios by I/M file and emissions type was reviewed. For the base year, the ratios were 1.0 or
were different for some special cases that were expected. When the NEI files were used as
-------
baseline for 2007, the modified files showed lower emissions in general, due primarily to the
added OBD programs.
5-9
-------
6.0 ADDITIONAL TABLES
NMIM includes several additional data tables that store county Federal
Information Processing Standard (FIPS) codes, representative county mapping information,
climate data, altitude data, and information about states using a non-standard phase-in for low
emissions vehicles (LEV). Each of these tables is describe below.
6.1 County
The County table includes FIPS codes for each county or equivalent political
subdivision of one of the states or territories of the USA, county altitude data, and representative
county mapping information.
Data Source
The FIPS codes for each county or equivalent political subdivision of one of the
states or territories of the USA were extracted from FIPSCNTY field in the
EPA_CHTEF_county_fips.xls file retrieved from http://www.epa.gov/ttn/chief/codes/index.html.
Per the guidance specified in Documentation for the Draft 1999 National
Emissions Inventory for Criteria Air Pollutants Onroad Source Methodologies (page 10)
Colorado, Nevada, New Mexico, and Utah were designated as high altitude areas while all other
states were designated as low altitude areas.
Representative county identification numbers were generated using a series of
MySQL queries that determined unique counties based on a number of parameters.
Data Population Methodology
The procedures used to populate the County table are described in the sections
below.
6-1
-------
Altitude Data
Per the guidance specified in Documentation for the Draft 1999 National
Emissions Inventory for Criteria Air Pollutants Onroad Source Methodologies (page 10)
Colorado, Nevada, New Mexico, and Utah were designated as high altitude areas while all other
states were designated as low altitude areas.
Representative County Mapping
In order to shorten the time required to complete a National run of NMIM, a
smaller number of counties that "represent" the full complement of counties can be used. After
MOBILE6 has been run for each of the representative counties, the results can then be mapped to
each individual county during post processing, and the actual vehicle miles traveled (VMT) in
each county could be used to generate the final emissions inventory.
The use of representative counties is a compromise between accuracy and
computational time and effort. The degree to which counties much match in order for one to
represent another can be altered in order to optimize the balance between accuracy and time. The
remainder of this section describes the available criteria for matching counties and those that
were used in the county mapping process.
Criteria 1: Same State—Due to the structure of the NMEVI database, a county
may only represent counties in the same state.
Criteria 2: Meteorological Data—The NMIM database currently stores the
maximum, minimum, and average temperature for each county (table: CountyMonth).
However, this information was only available at the state level, therefore each county in the
same state has the same meteorological data. This information was not considered in the
representative county mapping process.
6-2
-------
Criteria 3: Inspection and Maintenance (IM) Program—The database stores
the name of a file that describes the IM program to use for each county for each year (table:
County Year; field: IMFileName). A representative county must use the same IM file name in
database.
Criteria 4: Anti-Tampering Program (ATP)—The database stores the name of
a file that describes the anti-tampering program in place for each county for each year (table:
County Year; field: ATPFileName). ATP programs are typically associated with IM programs.
The representative county mapping methodology assumed that if the EVI program is similar, then
the ATP program is also similar. This information was not considered in the representative
county mapping process.
Criteria 5: Type of Fuel—The NMEVI database contains nine fuel identification
fields for each county by year and month (table: CountyYearMonth; fields HwyDiesellD,
HwyGasolinelD, NRGasolinelD, NRDiesellD, HwyGasolineldA, HwyGasolineldB). The
predominant fuel used in each county is described by the highway gasoline identification number
(table: County YearMonth; field: HwyGasolinelD); therefore this field only was used in the
representative county mapping process.
Criteria 6: Time Frame—The intent of the NMEVI database is to encompass all
the data needed to run the model from 1999 through 2050. The current set of available data
indicates that there are currently no meaningful changes after 2010. Therefore, the representative
county mapping was performed using 1999 through 2010 data.
The number of unique combinations of the criteria described above were queried
from the database. The counties were then grouped by each unique combination. All counties in
a group were assigned a representative county ID based on the lowest FIPS county code
associated with the counties in the group. Note that to support all of the functionality required to
complete representative county mapping, the MySQL code was written in version 4.0.5 beta.
6-3
-------
Quality Assurance Procedures
Several representative counties were verified visually to determine that the I/M
program files and gasoline assignments were the same. In addition, a MySQL query was created
to confirm that counties in only the four appropriate states were designated as high altitude.
6.2 CountyMonth
The CountyMonth table includes monthly climate information for each county.
This table is not dependent on year.
Data Source
Monthly temperature data for each state were collected from the 1999 NEI Fall
2001 Update files using the Max# and Min# fields in mxmntp99.dbf and pr_vi_temps.xls. Each
county within each state was assumed to experience the same monthly average temperatures.
Data Population Methodology
Records from the mxmntp99.dbf file were joined by FIPS code with records in the
CountyMonth table. With the exception of California and Texas, each state had one record that
was used to populate all months for all counties in each state. For California and Texas, which
had two monthly temperature data records, the record containing higher temperatures was used.
Monthly average temperatures for Puerto Rico and the Virgin Islands were populated in the same
manner from the pr_vi_temps.xls file.
Quality Assurance Procedures
MySQL queries were run to ensure that there were 12 monthly records for each of
the 3,222 counties, and 3,222 records for each month of the year. The monthly average
temperature data were printed and visually compared to the data sources listed above.
6-4
-------
6.3 State
The State table provides the Federal Information Processing Standard (FIPS)
codes associated with each of the 50 states, as well as Puerto Rico and the Virgin Islands. The
name of external data files required for states using a non-standard phase-in for low emissions
vehicles (LEV) is also included.
Data Source
The state identification information was populated using the June 2002 NEI
Update files.
The LEV indications were populated using the NEI Fall 2001 Update, MOBILE6
input files, and the Trends99_Pointer.dbf file.
Data Population Methodology
The LEV external data file name was retrieved by searching the NEI Fall 2001
Update MOBILE6 input files for the command "LOG IMP." For each MOBILE6 file that
included this command, the corresponding LEV file name followed the command. For example,
the MOBILE6 input file N5000110.IN includes the following:
LOG IMP
vtimp.d
These LEV files were copied to a central location. The counties to which this
LEV program apply were determined using the Trends99_pointer.dbf file, which notes all of the
counties that used the MOBILE6 input file from which the information was extracted.
6-5
-------
Records from the trends99_im.xls file were joined by FIPS code with records in
the State table. For states using an LEV program, the appropriate LEV program file names were
copied into the NLEVFileName field.
Quality Assurance Procedures
A MySQL query was run to confirm the appropriate number of states contained an
LEV filename. The State table was also printed and visually compared to the data sources listed
above.
6-6
-------
7.0 INTERNAL QUALITY ASSURANCE TABLES (NOT DELIVERED)
ERG created a series of internal quality assurance tables that presented basic
statistics on the data contained in the NMIM database. These tables, which were reviewed for
anomalies after each database repopulation, are described briefly below.
7.1 Minimum and Maximum Field Values
For each numeric field in each table, a record was inserted into a new table which
contained the table name, field name, maximum value of the field, and minimum value of the
field. The maximum and minimum values were compared with known maximum and minimum
values for state FIPS; county FIPS; gasoline parameters such as E200, E300, and others listed in
the MOBILE6 User's Guide; and from other sources.
7.2 Null Values
For each field in each table, a count was made of the number of records
containing a null value. A record was inserted for each field which contained the table name,
field name, and count of null values. These records were reviewed to ensure that fields for which
null values were not expected were not included in the table.
7.3 Zero Values
For each numeric field in each table, a count was made of the number of records
containing a zero value. A record was inserted in a new table for each field which contained the
table name, field name, and count of zero values. These records were reviewed to ensure that
fields for which zero values were not expected were not included in the table.
7-1
-------
7.4 Table Relationships
For each field in each table with a Child to Parent relationship, a count was made
of the number of records not having a matching parent record. A record was inserted in a new
table for each field which contained table name, field name and count of records missing a parent
record. These records were reviewed to ensure that tables for which missing parent records were
not expected were not included in the table.
Likewise, for each field in each table with a Parent to Child relationship, a count
was made of the number of records not having a matching child record. If this count was greater
than zero, a record was inserted in a new table which contained the parent table name, child table
name, field name, and the value of the field without a matching child record. These records were
reviewed to ensure that tables for which missing child records were not expected were not
included in the table.
7-2
-------
s.o REFERENCES
AAMA, 1999. North American Gasoline and Diesel Fuel Survey. Alliance of Automobile
Manufacturers.
Abt, 2003. Refinery Modeling: Legislative and Regulatory Developments - Effects on Gasoline
Supply, Federal Oxygenate Case and Fuel Control Cases (Benzene, Toxics and Sulfur Control).
EPA Contract 68-C-01-164, Work Assignment 1-7 & 1-7.1, Sub-Tasks 3.2, 4.1, and 4.2 for
2010. Prepared for U.S. Environmental Protection Agency by Abt Associates Inc., Bethesda,
Maryland. January 14.
Federal Register, 2000. Control of Air Pollution From New Motor Vehicles: Tier 2 Motor
Vehicle Emissions Standards and Gasoline Sulfur Control Requirements. Final Rule. 65 CFR
6698. February 10.
Federal Register, 2001. Control of Air Pollution From New Motor Vehicles; Amendment to the
Tier 2/Gasoline Sulfur Regulations. Direct Final Rule. 66 CFR 19296. April 13.
FHWA, 1999. Federal Highway Administration (FHWA) website for oxygenated fuel sale
percentage. Table MF-33E - Estimated Use of Gasohol and Table MF-21 - Motor-Fuel Use.
Internet address: http://www.fhwa.dot.gov/ohim/hs99/mfpage.htm
Manners, 2002. Personal communication between Mary Manners (U.S. EPA, Ann Arbor,
Michigan; 734-214-4873) and Marty Wolf (ERG). August 23.
MathPro, 1998. Costs of Alternative Sulfur Content Standards for Gasoline in PADDIV. Final
Report. Prepared for the National Petrochemical and Refiners Association by MathPro Inc.,
West Bethesda, Maryland. December 30.
MathPro, 1999a. Costs of Meeting 40 PPM Sulfur Content Standard for Gasoline in PADDS 1-
3, Via Mobil and CD Tech Desulfurization Processes. Final Report. Prepared for the American
Petroleum Institute by MathPro Inc., West Bethesda, Maryland. February 26.
MathPro, 1999b. Analysis of California Phase 3 RFG Standards. Prepared for the California
Energy Commission by MathPro Inc., West Bethesda, Maryland. December 7.
TRW, 1999. National Institute of Petroleum and Energy Research (NIPER or TRW) Fuel
Survey.
U.S. EPA, 2000. Reformulated Gasoline Survey Data for 2000. U.S. Environmental Protection
Agency, Office of Transportation and Air Quality, Ann Arbor, Michigan. Internet address:
http://www.epa.gov/otaq/consumer/fuels/mtbe/oxy-95-00.pdf.
3-1
-------
U.S. EPA, 2001. U.S. EPA Oxygenated Fuel Program Summary, State Winter Oxygenated Fuel
Program Requirements for Attainment or Maintenance of CO NAAQS, U.S. Environmental
Protection Agency, Office of Transportation and Air Quality, Ann Arbor, Michigan. October.
Internet address: http://www.epa.gov/otaq/regs/fuels/oxy-area.pdf.
U.S. EPA, 2002a. Technical Description of the Toxics Module for MOBILE6.2 andGuidance on
Its Use for Emission Inventory Preparation. EPA420-R-02-029. U.S. Environmental Protection
Agency, Office of Transportation and Air Quality, Ann Arbor, Michigan. November.
U.S. EPA, 2002b. MOBILE6.2 User's Guide. EPA 420-R-02-028. U.S. Environmental
Protection Agency, Office of Transportation and Air Quality, Ann Arbor, Michigan. October.
8-2
-------
APPENDIX A
INDEX TO DATA FILES AVAILABLE ELECTRONICALLY
NMIM Table Name
File Name
Description
Reference Tables Folder
HPMSRoadType
M6VType
M6VClass
vehicles2.xls
Appendix B of
MOBILE6 User's Guide
Appendix B of
MOBILE6 User's Guide
The spreadsheet was taken from the June 2002 National
Emissions Inventory (NEI) update files.
The 16 vehicle classes were obtained from Table 2 of
Appendix B (pg 245) of the MOBILE6 User's Guide
(EPA420-R-02-028, October 2002).
The 16 vehicle classes were obtained from Section 1.2.3
(pg 14) of the MOBILE6 User's Guide (EPA420-R-02-
028, October 2002).
Fuel Tables Folder
County YearMonth
Gasoline
Gasoline2
CYMMTBEPhsOut
GasMTBEPhsOut
Gas2MTBEPhsOut
Diesel
NaturalGas
CYM Gas.mdb
CYM_GasMTBE.mdb
030425_gasoline
assignments and
parameters.xls
Final Fuell999V3 032
703.xls
RFG_FuelOO_vl.xls
sulfur.xls
sulfur.xls
Access 97 databases which produces the
County YearMonth, Gasoline, Gasoline2,
CYMMTBEPhsOut, GasMTBEPhsOut, and
Gas2MTBEPhsOut tables. The user can open the form
frmExecFunctions in each database and click the buttons
in order to re-create the data. The queries with names
beginning "Export_" should be imported into the County
database.
Gasoline assignments and factors for 1999 through 2050.
1999 seasonal gasoline parameters by county
Updated seasonal gasoline parameters by county for year
2000.
The spreadsheet contains the assumptions Pechan used
for non-road sulfur content of diesel and CNG fuels,
forwarded by Dave Brzezinski, EPA. [two additional
records were added, where did they come from?]
The spreadsheet contains the assumptions Pechan used
for non-road sulfur content of diesel and CNG fuels,
forwarded by Dave Brzezinski, EPA.
I/M Program Tables Folder
CountyYear
stage2dat.xls
OOtables.wpd
trends99_pointer.dbf
trends99 iraxls
model, wpd
counties.wpd
IMFiles\*.*
1999 Updated M6 Input
Files\*.*
Fall 2001 update to the 1999 NEI files: Stage 2 refueling
program efficiency (stage2dat.xls, OOtables.wpd),
Anti-tampering program file name and files
(trends99_pointer.dbf), I/M program file name and files
(Trends99_im.xls), OBD Schedules (Model. wpd), other
future programs (Counties.wpd).
Source of "cutpoint" file references in the I/M files.
Source for ATP Info.
A-l
-------
NMIM Table Name
File Name
Description
Vehicle Tables Folder
Average Speed
BaseYearVMT
VMTGrowth
VMTMonthAllocation
vmt99_f_m6_with_8Stat
esUpdate.DBF
vmt99_f_m6_with_8Stat
esUpdate.DBF
vmt99_n_m6.dbf
BaseYearVMT table
V-2.xls
V-3.xls
V-4.xls
Modeling files for June 2002 update to the 1999 NEI,
extracted from the VMT table.
Modeling files for June 2002 update to the 1999 NEI for
the 50 state and Washington, DC, and Fall 2001 update
to the 1999 NEI for Puerto Rico and the Virgin Islands.
VMT estimates for calendar year 2007, calendar year
2020, and calendar year 2030 provided by the EPA from
appendix tables in the support documentation for the
2007 Heavy Duty Diesel Rule (HD2007).
1999 NEI document "Onroad Source Methodologies"
dated 10/2001
Additional Tables Folder
County
CountyMonth
State
EPA_CHIEF_county_
fips.xls
mxmntp99.dbf
pr_vi_temps.xls
trends99_pointer.dbf
The spreadsheet was retrieved from
http://www.epa.gov/ttn/chief/codes/index.html.
Fall 2001 update to the 1999 NEI.
June 2002 update to the 1999 NEI, and Fall 2001 update
to the 1999 NEI. See 1999 Updated M6 Input Files\*.*
under County Year for source for NLEV file references.
DB Documentation Folder
Not applicable
CountyDB.pdf
CountyDB.rtf
County database documentation
SQLScripts Folder
All NMIM Tables
Load_Data_Tables. sql
Load Data BaseYearV
MT.sql
Load_Data_VMTGrowt
h.sql
QA_Script_l.sql
QA_Script_2.sql
QA_Script_3.sql
Data Loading Scripts
Quality Assurance Scripts
Data Files Folder
Vehicle tables
County
Generate VMTMo Alloc .
awk
Rep County Mapping, s
ql
Code to Produce VMT Data
Representative County Mapping Code
A-2
------- |