NATIONAL
MOBILE
INVENTORY
MODEL
(
NMIM)
BASE
AND
FUTURE
YEAR
COUNTY
DATABASE
DOCUMENTATION
AND
QUALITY
ASSURANCE
PROCEDURES
Prepared
for:

U.
S.
Environmental
Protection
Agency
Office
of
Transportation
and
Air
Quality
(
OTAQ)
Air
Modeling
Division
2000
Traverwood
Drive
Ann
Arbor,
MI
48105
Prepared
by:

Eastern
Research
Group,
Inc.
14555
Avion
Parkway
Suite
200
Chantilly,
VA
20151­
1102
15
April
2003
EPA
Contract
No.
68­
C00­
112
Work
Assignment
3­
05
i
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
ii
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
CountyYear
Data
Sources
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
1
5­
2
County
FIPS
Codes
in
NEI
Stage
2
Refueling
Data
Not
Used
in
NMIM.
.
.
.
.
5­
3
5­
3
Examples
of
Differences
in
I/
M
Program
Data
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
7
5­
4
New
External
IM
Program
Files
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
5­
8
iii
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
1­
1
1.0
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
(
NMIM)
is
an
interim
product
supporting
creation
of
MOVES.
NMIM
combines
mobile
sources
emission
factor
modeling
with
areaspecific
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
NMIM,
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,
NMIM
implements
some
MOVES
architecture
features.
Specifically,
the
NMIM
database
is
based
on
the
MySQL
opensource
database
management
system,
and
the
Java
language
is
used
as
appropriate
for
software
components.
The
NMIM
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
2
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
3
Figure
1­
1.
Data
Relationship
Diagram
2­
1
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
DataSourceID
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
2
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
NMIM
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
NMIM
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
=
[

XXAA
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
3
TABLE
2­
1
HPMSRoadtype
Values
Roadway
Type
ID
Roadway
Type
Description
11
Interstate:
Rural
13
Other
Principal
Arterial:
Rural
15
Minor
Arterial:
Rural
17
Major
Collector:
Rural
19
Minor
Collector:
Rural
21
Local:
Rural
23
Interstate:
Urban
25
Other
Freeways
and
Expressways:
Urban
27
Other
Principal
Arterial:
Urban
29
Minor
Arterial:
Urban
31
Collector:
Urban
33
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
NMIM
database.
2­
4
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
Vehicle
Class
Abbreviation
Vehicle
Class
Description
1
LDGV
Light­
Duty
Gasoline
Vehicles
(
Passenger
Cars)

2
LDGT1
Light­
Duty
Gasoline
Trucks
1
(
0­
6,000
lbs.
GVWR,
0­
3750
lbs.
LVW)

3
LDGT2
Light­
Duty
Gasoline
Trucks
2
(
0­
6,000
lbs.
GVWR,
3751­
5750
lbs.
LVW)

4
LDGT3
Light­
Duty
Gasoline
Trucks
3
(
6,001­
8,500
lbs.
GVWR,
0­
5750
lbs.
ALVW)

5
LDGT4
Light­
Duty
Gasoline
Trucks
4
(
6,001­
8,500
lbs.
GVWR,
5751
lbs.
and
greater
ALVW)

6
HDGV2B
Class
2b
Heavy­
Duty
Gasoline
Vehicles
(
8501­
10,000
lbs.
GVWR)

7
HDGV3
Class
3
Heavy­
Duty
Gasoline
Vehicles
(
10,001­
14,000
lbs.
GVWR)

8
HDGV4
Class
4
Heavy­
Duty
Gasoline
Vehicles
(
14,001­
16,000
lbs.
GVWR)

9
HDGV5
Class
5
Heavy­
Duty
Gasoline
Vehicles
(
16,001­
19,500
lbs.
GVWR)

10
HDGV6
Class
6
Heavy­
Duty
Gasoline
Vehicles
(
19,501­
26,000
lbs.
GVWR)

11
HDGV7
Class
7
Heavy­
Duty
Gasoline
Vehicles
(
26,001­
33,000
lbs.
GVWR)

12
HDGV8A
Class
8a
Heavy­
Duty
Gasoline
Vehicles
(
33,001­
60,000
lbs.
GVWR)

13
HDGV8B
Class
8b
Heavy­
Duty
Gasoline
Vehicles
(>
60,000
lbs.
GVWR)

14
LDDV
Light­
Duty
Diesel
Vehicles
(
Passenger
Cars)

15
LDDT12
Light­
Duty
Diesel
Trucks
1
and
2
(
0­
6,000
lbs.
GVWR)

16
HDDV2B
Class
2b
Heavy­
Duty
Diesel
Vehicles
(
8501­
10,000
lbs.
GVWR)

17
HDDV3
Class
3
Heavy­
Duty
Diesel
Vehicles
(
10,001­
14,000
lbs.
GVWR)

18
HDDV4
Class
4
Heavy­
Duty
Diesel
Vehicles
(
14,001­
16,000
lbs.
GVWR)

19
HDDV5
Class
5
Heavy­
Duty
Diesel
Vehicles
(
16,001­
19,500
lbs.
GVWR)

20
HDDV6
Class
6
Heavy­
Duty
Diesel
Vehicles
(
19,501­
26,000
lbs.
GVWR)

21
HDDV7
Class
7
Heavy­
Duty
Diesel
Vehicles
(
26,001­
33,000
lbs.
GVWR)
TABLE
2­
2
Continued
Vehicle
Class
ID
Vehicle
Class
Abbreviation
Vehicle
Class
Description
2­
5
22
HDDV8A
Class
8a
Heavy­
Duty
Diesel
Vehicles
(
33,001­
60,000
lbs.
GVWR)

23
HDDV8B
Class
8b
Heavy­
Duty
Diesel
Vehicles
(>
60,000
lbs.
GVWR)

24
MC
Motorcycles
(
Gasoline)

25
HDGB
Gasoline
Buses
(
School,
Transit
and
Urban)

26
HDDBT
Diesel
Transit
and
Urban
Buses
27
HDDBS
Diesel
School
Buses
28
LDDT34
Light­
Duty
Diesel
Trucks
3
and
4
(
6,001­
8,500
lbs.
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
NMIM
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
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
6
Table
2­
3
presents
the
16
vehicle
types
contained
in
the
M6VType
table.

TABLE
2­
3
M6VType
Values
Vehicle
Type
ID
Vehicle
Type
Description
1
Light­
Duty
Vehicles
(
Passenger
Cars)

2
Light­
Duty
Trucks
1
(
0­
6,000
lbs.
GVWR,
0­
3,750
lbs.
LVW)

3
Light­
Duty
Trucks
2
(
0­
6,000
lbs.
GVWR,
3,751­
5,750
lbs.
LVW)

4
Light­
Duty
Trucks
3
(
6,001­
8,500
lbs.
GVWR,
0­
5,750
lbs.
ALVW)

5
Light­
Duty
Trucks
4
(
6,001­
8,500
lbs.
GVWR,
5,751
lbs.
and
greater
ALVW)

6
Class
2b
Heavy­
Duty
Vehicles
(
8,501­
10,000
lbs.
GVWR)

7
Class
3
Heavy­
Duty
Vehicles
(
10,001­
14,000
lbs.
GVWR)

8
Class
4
Heavy­
Duty
Vehicles
(
14,001­
16,000
lbs.
GVWR)

9
Class
5
Heavy­
Duty
Vehicles
(
16,001­
19,500
lbs.
GVWR)

10
Class
6
Heavy­
Duty
Vehicles
(
19,501­
26,000
lbs.
GVWR)

11
Class
7
Heavy­
Duty
Vehicles
(
26,001­
33,000
lbs.
GVWR)

12
Class
8a
Heavy­
Duty
Vehicles
(
33,001­
60,000
lbs.
GVWR)

13
Class
8b
Heavy­
Duty
Vehicles
(>
60,000
lbs.
GVWR)

14
School
Buses
15
Transit
and
Urban
Buses
16
Motorcycles
3­
1
3.0
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.
3­
2
TABLE
3­
1
Diesel
Sulfur
Values
Diesel
ID
Diesel
Sulfur
Value
(
parts
per
million
(
ppm))
Highway
Applicability
NonRoad
Applicability
1
500
Assigned
to
all
counties
for
calendar
years
1999
through
2003.
Not
applicable.

2
11
Assigned
to
all
counties
for
calendar
years
1999
through
2004.
Assigned
to
all
California
counties
for
calendar
years
2006
through
2008,
assigned
to
all
counties
for
calendar
years
2009
through
2050.

3
2700
Not
applicable.
Assigned
to
all
counties
except
those
in
California
for
calendar
years
1999
through
2005.

4
120
Not
applicable.
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
3­
3
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
4
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
Pinal)
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.
3­
5
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.
3­
6
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.
3­
7
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
8
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
9
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.
3­
10
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
FHWA's
estimate
of
0.65%
ethanol
for
the
state.

Massachusetts
Berkshire,
Franklin,
Hampden,
and
Hampshire
counties
were
represented
by
data
from
the
RFG
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
RFG
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
RFG
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
11
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
12
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
13
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
14
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
15
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
FHWA
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
16
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
17
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.

Oregon
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
18
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
19
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
20
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
ethanolbased
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
21
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
22
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
FHWA.

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
FHWA.

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
FHWA's
statewide
estimate
of
0.01%
ethanol
market
share.
3­
23
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
24
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
III
(
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
III
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,
III,
and
V
were
used
for
the
2004
Tier
2
sulfur
standards
(
MathPro,
1999a).
3­
25
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:

P
120
=
P
40
+
([
S
120
 
S
40]/[
S
ref
 
S
40])
x
(
P
ref
 
P
40)

Where
S
ref
=
sulfur
content
of
reference
fuel;
S
120
=
sulfur
content
of
120
ppm
sulfur
content
fuel;
S
40
=
sulfur
content
of
40
ppm
sulfur
content
fuel;
P
ref
=
value
of
other
parameter
for
reference
fuel;
P
120
=
value
of
other
parameter
for
120
ppm
sulfur
content
fuel;
and
P
40
=
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
(
MAF)
for
each
relevant
parameter
was
calculated
by
ratioing
the
120
ppm
sulfur
content
fuel
parameter
by
the
reference
fuel
parameter
(
i.
e.,
MAF
=
P
120/
P
ref).
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
=
P
120
 
P
ref).

Four
sets
of
adjustment
factors
were
developed
for
2004
fuel
in
PADDs
I,
II,
III,

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
26
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
MAF
and
AAF
values
for
PADD
IV
were
calculated
in
similar
manner
to
those
in
PADDs
I,
II,
III,
and
V
(
i.
e.,
MAF
=
P
150/
P
base
and
AAF
=
P
150
 
P
base);
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.

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).
3­
27
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
=
P
Phase3/
P
Phase2
and
AAF
=
P
Phase3
 
P
Phase2).
The
2004
MAF
and
AAF
values
were
also
used
in
two
Arizona
counties
(
Maricopa
and
Pinal)
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
=
P
90/
P
ref
and
AAF
=
P
90
 
P
ref).
This
resulted
in
five
sets
of
adjustment
factors
for
2005
fuel
in
PADDs
I,
II,
III,
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.

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.,
MAF
=
P
30/
P
ref
and
AAF
=
P
30
 
P
ref).
This
resulted
in
five
sets
of
adjustment
factors
for
2006
fuel
in
PADDs
I,
II,
III,
and
V
(
i.
e.,
summer
conventional,
summer
RFG,
winter
conventional,
winter
RFG,
and
summer
conventional
with
gasohol).
3­
28
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.,
MAF
=
P
30/
P
base
and
AAF
=
P
30
 
P
base).
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.

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
#
1a
and
2010
RFS
#
2
modeled
fuels
(
both
conventional
and
RFG)
were
obtained
separately
for
PADD
I,
II,
and
III.
Weighted
gasoline
parameters
were
derived
based
upon
volumes
of
MTBE­
and
ETOH­
blended
fuels
in
PADD
II
and
III.
3­
29
The
MAF
for
each
relevant
parameter
was
calculated
by
ratioing
the
RFS
#
2
fuel
parameter
by
the
Reference
#
1a
fuel
parameter
(
i.
e.,
MAF
=
P
RFS#
2/
P
Ref#
1).
The
AAF
for
each
relevant
parameter
was
calculated
by
subtracting
the
Reference
#
1a
fuel
parameter
from
the
RFS
#
2
fuel
parameter
(
i.
e.,
AAF
=
P
RFS#
2
 
P
Ref#
1).
The
oxygenate
contents
and
market
shares
were
then
adjusted
to
represent
expected
conditions
occurring
due
to
a
MTBE
ban.
The
PADD
II
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
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
CountyYearMonth
tables.
Each
of
these
components
is
described
in
detail
below.

Seasonal
Fuel
Data
3­
30
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
­
Olefins
­
MTBE
­
ETBE
­
TAME
­
ETOH
The
factors
for
the
parameters
listed
for
below
for
gasolines
A
through
W
are
additive:
3­
31
­
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
­
Olefins
­
Sulfur
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
:
3­
32
­
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
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,

Olefins,
Aromatics,
Benzene,
E200,
and
E300
by
the
percentage
of
oxygenate
fuel
sale
from
Federal
Highway
Survey,
as
shown
below:
3­
33
Composited
parameter
value

(
(
Fuel
1
parameter
×
Fuel
1
percentage
of
oxygenate
fuel
sale
federal
highway
survey)

(
Fuel
2
parameter
×
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
34
TABLE
3­
2
Sample
Calculation
for
Composited
Seasonal
Fuel
for
FIPS
39001:
Adams,
OH
Fuel
1
Fuel
1
Fuel
2
Fuel
2
Composited
Fuel
Composited
Fuel
Season
summer
winter
summer
winter
summer
winter
RVP
9.45
14.17
8.59
13.86
8.93
13.98
Sulfur
297.7
249.1
406.2
384.5
363.08
330.69
Olefins
7.26
7.65
10.378
9.634
9.14
8.84
Aromatics
25.75
19.76
31.057
26.71
28.95
23.95
Benzene
0.98
0.92
1.26
1.1773
1.15
1.07
E200
55.84
60.10
46.72
53.80
50.34
56.30
E300
81.39
84.18
78.90
82.62
79.89
83.24
MTBE
volume
0.02
0.01
4.41
1.33
4.41
1.33
MTBE
market
share
0
0
100
100
60.26
60.26
ETOH
volume
10.06
9.93
0
0
10.06
9.93
ETOH
market
share
100
100
0
0
39.74
39.74
ETBE
volume
0
0
0
0
0
0
ETBE
market
share
0
0
0
0
0
0
TAME
volume
0
0
0
0.009
0
0
TAME
market
share
0
0
0
0
0
0
Oxygenate_
Fuel_
Sale
_
Percentage
39.74
39.74
60.26
60.26
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.
NMIM
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
for
3­
35
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
ASTM
RVP
Class
ASTM
RVP
Schedule
January
E
15
February
E
15
March
D
13.5
April
D
13.5
May
C
11.5
June
C
11.5
July
C
11.5
August
C
11.5
September
C
11.5
October
C
11.5
November
D
13.5
December
E
15
Summer
(
June
value)
C
11.5
Winter
(
January
value)
E
15
3­
36
Interpolation
Factor
Calculation
Monthly
Interpolation
Factor

(
Monthly
RVP
Class

Summer
RVP
class)
(
Winter
RVP
Class

Summer
RVP
Class)

TABLE
3­
4
Monthly
Interpolation
Factor
Calculation
for
FIPS
39001:
Adams,
OH
Month
ASTM
RVP
Class
ASTM
RVP
Schedule
Interpolation
Factor
January
E
15
1
February
E
15
1
March
D
13.5
0.571
April
D
13.5
0.571
May
C
11.5
0
June
C
11.5
0
July
C
11.5
0
August
C
11.5
0
September
C
11.5
0
October
C
11.5
0
November
D
13.5
0.571
December
E
15
1
Summer
(
June
value)
C
11.5
Winter
(
January
value)
E
15
Monthly
Interpolation
Calculation
Interpolated
Monthly
Value

summer
value

monthly
interpolation
factor
×
(
winter
value

summer
value)
3­
37
TABLE
3­
5
Sample
Monthly
Interpolation
for
Olefins
Calendar
Year
1999
for
FIPS
39001:
Adams,
OH
Season/
Month
Seasonal
Volume
Percent
Olefins
Summer
Volume
Percent
Olefins
Interpolation
Factor
Winter
Volume
Percent
Olefins
­
Summer
Volume
Percent
Olefins
Interpolated
Monthly
Volume
Percent
Olefins
Summer
9.14
Winter
8.84
January
9.14
1
­
0.3
8.84
February
9.14
1
­
0.3
8.84
March
9.14
0.571
­
0.3
8.97
April
9.14
0.571
­
0.3
8.97
May
9.14
0
­
0.3
9.14
June
9.14
0
­
0.3
9.14
July
9.14
0
­
0.3
9.14
August
9.14
0
­
0.3
9.14
September
9.14
0
­
0.3
9.14
October
9.14
0
­
0.3
9.14
November
9.14
0.571
­
0.3
8.97
December
9.14
1
­
0.3
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
38
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
NMIM
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
39
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
40
3.4
Natural
Gas
The
NaturalGas
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
NaturalGas
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
Natural
Gas
Sulfur
Value
1
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
41
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
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
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
NaturalGas
data
sources.
3­
42
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
43
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
Gasoline2
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
44
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
45
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.
4­
1
4.0
VEHICLE
TABLES
Emissions
inventory
calculations
are
significantly
impacted
by
vehicle
population
and
travel
data.
The
sources
of
this
information
in
the
NMIM
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
2
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
NMIM
"
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
NMIM
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
3
TABLE
4­
1
National­
Average
VMT
Fraction
by
Road
Type
Used
for
California
VMT
Data
Vehicle
Class
HPMS
Roadway
Type
Total
11
13
15
17
19
21
23
25
27
29
31
33
1
0.097
0.091
0.064
0.078
0.022
0.049
0.140
0.053
0.149
0.117
0.051
0.091
1.000
2
0.101
0.094
0.066
0.081
0.022
0.049
0.138
0.052
0.144
0.114
0.049
0.089
1.000
3
0.101
0.094
0.066
0.081
0.022
0.049
0.137
0.052
0.144
0.114
0.049
0.089
1.000
4
0.101
0.093
0.066
0.080
0.022
0.049
0.139
0.052
0.145
0.114
0.049
0.089
1.000
5
0.101
0.093
0.066
0.080
0.022
0.049
0.139
0.052
0.145
0.115
0.049
0.089
1.000
6
0.106
0.094
0.066
0.080
0.022
0.049
0.138
0.052
0.143
0.113
0.049
0.087
1.000
7
0.136
0.120
0.082
0.101
0.027
0.059
0.114
0.042
0.117
0.092
0.039
0.070
1.000
8
0.136
0.119
0.081
0.100
0.027
0.058
0.115
0.043
0.118
0.093
0.039
0.071
1.000
9
0.135
0.120
0.082
0.102
0.027
0.060
0.114
0.042
0.117
0.092
0.039
0.070
1.000
10
0.134
0.120
0.084
0.103
0.027
0.061
0.113
0.042
0.116
0.091
0.039
0.070
1.000
11
0.135
0.120
0.083
0.102
0.027
0.060
0.114
0.042
0.117
0.092
0.039
0.070
1.000
12
0.159
0.144
0.101
0.125
0.033
0.074
0.085
0.032
0.090
0.071
0.030
0.054
1.000
14
0.101
0.097
0.069
0.083
0.021
0.055
0.131
0.049
0.145
0.113
0.049
0.087
1.000
15
0.103
0.092
0.064
0.078
0.021
0.048
0.140
0.053
0.147
0.116
0.049
0.090
1.000
16
0.111
0.097
0.067
0.080
0.021
0.047
0.138
0.051
0.141
0.112
0.048
0.086
1.000
17
0.138
0.122
0.084
0.103
0.027
0.060
0.112
0.041
0.115
0.090
0.038
0.069
1.000
18
0.139
0.122
0.084
0.103
0.027
0.059
0.112
0.041
0.115
0.090
0.038
0.069
1.000
19
0.140
0.122
0.083
0.101
0.026
0.058
0.114
0.042
0.115
0.091
0.039
0.070
1.000
20
0.138
0.122
0.085
0.103
0.027
0.060
0.111
0.041
0.114
0.090
0.038
0.069
1.000
21
0.138
0.122
0.085
0.103
0.028
0.061
0.111
0.041
0.114
0.090
0.039
0.069
1.000
22
0.161
0.144
0.100
0.122
0.033
0.072
0.089
0.032
0.090
0.071
0.030
0.054
1.000
23
0.161
0.144
0.100
0.122
0.033
0.072
0.089
0.032
0.090
0.071
0.031
0.055
1.000
24
0.107
0.101
0.071
0.087
0.024
0.055
0.130
0.049
0.137
0.109
0.047
0.084
1.000
25
0.159
0.140
0.095
0.119
0.031
0.069
0.094
0.034
0.095
0.075
0.032
0.057
1.000
26
0.162
0.142
0.097
0.119
0.031
0.068
0.093
0.033
0.093
0.073
0.031
0.056
1.000
27
0.161
0.142
0.098
0.120
0.031
0.069
0.092
0.033
0.093
0.073
0.031
0.056
1.000
28
0.103
0.094
0.066
0.081
0.021
0.049
0.138
0.052
0.145
0.114
0.049
0.089
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
4
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
5
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
6
FIGURE
4­
1.
VMT
Growth
Data
Sources
and
Methods
4­
7
Base
data
"
B00BaseYearVMT"
is
the
data
in
the
BaseYearVMT
table,

discussed
in
Section
4.1.
Base
data
"
F00HDRule"
is
the
data
from
the
HD2007
Rule
appendices.

In
step
"
B01BaseYearVMT"
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
NMIM
tables.

In
step
"
F01ConvertHD"
the
HD2007
data
were
converted
from
a
MOBILE5
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
MOBILE5
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
MOBILE5
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
MOBILE5
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
MOBILE5
classes.
4­
8
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
"
NoHDGV8a,"
the
HDGV8a
(
gasoline)
VMT
was
added
to
class
HDDV8a
(
diesel),
and
then
HDGV8a
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
9
TABLE
4­
2
Post­
fixes
to
Vehicle
Class
Conversions
Case
State
FIPS
County
FIPS
NoLDDT
6
3,
51,
91
NoLDDV
6
3,
91
NoMC
47
65
6
3,
49,
91
NoHDGV8a
16
25
20
33,
67,
71,
189,
199
28
55
30
19,
37,
59,
69
31
5,
7,
9,
75,
91,
103,
113,
115,
117,
171,
183
35
21
38
87
46
17
48
23,
33,
75,
79,
169,
247,
261,
263,
269,
301,
311,

345,
357,
383,
393,
443,
495
8
53
In
step
"
F02FixFIPS"
the
differences
between
area
FIPS
code
assignments
were
resolved
by
converting
the
HD2007
area
assignments
to
the
NMIM
basis.
The
basis
used
for
the
FIPS
reassignments
was
an
NEI
document
provided
by
EPA.
Most
of
the
differences
between
HD2007
and
NMIM
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
F02FixFIPS
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
FIPS
code
conversion,
A
­>
A.
4­
10
°
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
"
T01Totals"
the
VMT
totals
for
the
lower
49
were
calculated
for
the
anchor
years,
from
the
output
of
steps
B01CombineBaseYear
and
F02FixFIPS.
The
result
was
total
VMT
by
vehicle
class
for
all
four
anchor
years.

In
step
"
F03AddMissingST"
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
F03AddMissingST
was
a
set
of
complete
VMT
values
for
the
future
anchor
years,
for
the
states
not
covered
in
the
HD2007
data.

In
step
"
S01AnchorYearsComplete"
the
anchor
year
VMT
data
from
steps
B01BaseYearVMT,
F02FixFIPS,
and
F03AddMissingST
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
11
discussed
in
step
F01ConvertHD.
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
12
FIGURE
4­
2.
Anchor
Years
and
Interpolation
Spans
for
VMT
Growth
In
step
"
S03DeltaVMT"
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
13
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.

B01BaseYearVMT:
VMT
totals
by
vehicle
were
checked
at
the
state
level
and
the
county
level.

F01ConvertHD:
The
vehicle
class
conversion
is
not
reversible,
and
does
not
preserve
VMT
by
all
8
MOBILE5
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
14
then
processed
into
VMT
totals
by
county
for
the
five
MOBILE5
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.

T01Totals:
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.

F03AddMissingST:
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
T01Totals
data.

S01AnchorYearsComplete:
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
F03AddMissingST
steps.
4­
15
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
F01ConvertHD
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.

S03DeltaVMT:
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
16
TABLE
4­
3
Original
and
Adjusted
VMTMonthAllocation
Values
Vehicles:
Light
Duty
(
LDV,
LDT,
MC)
HDV
Roadway:
Rural
Urban
All
Month
Allocation
Values
Original
Adjusted
Original
Adjusted
Original
Adjusted
Jan
0.0744
0.0744
0.0806
0.0806
0.0861
0.0862
Feb
0.0672
0.0672
0.0728
0.0728
0.0778
0.0778
Mar
0.0805
0.0805
0.0859
0.0860
0.0842
0.0842
Apr
0.0779
0.0779
0.0832
0.0833
0.0815
0.0815
May
0.0805
0.0805
0.0859
0.0860
0.0842
0.0842
Jun
0.0942
0.0942
0.0864
0.0865
0.0815
0.0815
Jul
0.0974
0.0975
0.0893
0.0894
0.0842
0.0842
Aug
0.0974
0.0974
0.0893
0.0894
0.0842
0.0842
Sep
0.0844
0.0844
0.0808
0.0809
0.0824
0.0824
Oct
0.0872
0.0872
0.0835
0.0836
0.0852
0.0852
Nov
0.0844
0.0844
0.0808
0.0809
0.0824
0.0824
Dec
0.0744
0.0744
0.0806
0.0806
0.0861
0.0862
Sum
0.9999
1.0000
0.9991
1.0000
0.9998
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
17
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
18
TABLE
4­
4
Conversion
of
Roadway
and
Vehicle
Types
for
VMTMonthAllocation
Data.

Vehicle
Types
MOBILE6
VehicleTypes
LDV,
LDT,
MC
1:
LDV
2:
LDT1
3:
LDT2
4:
LDT3
5:
LDT4
16:
MC
HDV
6:
HDV2B
7:
HDV3
8:
HDV4
9:
HDV5
10:
HDV6
11:
HDV7
12:
HDV8A
13:
HDV8B
14:
HDBS
15:
HDBT
Roadways
HPMS
Codes
Rural
11:
Rural,
Interstate
13:
Rural,
Other
Principal
Arterial
15:
Rural,
Minor
Arterial
17:
Rural,
Major
Collector
19:
Rural,
Minor
Collector
21:
Rural,
Local
Urban
23:
Urban,
Interstate
25:
Urban,
Non­
Interstate
Freeway
27:
Urban,
Other
Principal
Arterial
29:
Urban,
Minor
Arterial
31:
Urban,
Collector
33:
Urban,
Local
5­
1
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
CountyYear
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
CountyYear
are
described
below.

Data
Source
The
Data
Sources
used
to
populate
the
CountyYear
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
00tables.
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
2
Data
Population
Methodology
The
procedures
used
to
populate
the
CountyYear
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
00Tables.
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
CountyYear
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
CountyYear
table
were
noted.
The
differences
in
FIPS
codes
were
expected
because
the
NEI
data
were
based
on
FIPS
numbering
with
several
differences
from
the
NMIM
numbering.
Table
5­
2
lists
the
fifteen
FIPS
codes
used
in
stg2dat.
xls
that
were
remapped
or
corrected
for
use
in
NMIM.
If
no
information
existed
for
a
particular
county,
it
was
assumed
no
Stage
2
refueling
program
was
in
effect
in
that
county.
5­
3
TABLE
5­
2
County
FIPS
Codes
in
NEI
Stage
2
Refueling
Data
Not
Used
in
NMIM.

County
in
stg2dat.
xls
Comments
FIPS
ST
COUNTYNM
02010
AK
Aleutian
Islands
Ed
Correct
for
new
counties
and
post­
1980
subdivision
in
AK.

02140
AK
Kobuk
Ed
02231
AK
Skagway­
Yakutat
Ed
02990
AK
Upper
Yukon
Ed
02991
AK
Seward
Ed
02992
AK
Kuskokwim
Ed
02993
AK
Bristol
Bay
Borough
02994
AK
Angoon
Ed
02996
AK
Cordova­
Mccarthy
Ed
02998
AK
Outer
Kethcikan
Ed
02999
AK
Barrow
Ed
12025
FL
Dade
Co
Dade
renamed
Miami­
Dade
County.

29193
MO
Ste.
Genevieve
Co
Corrected
numbering
to
29186,
per
1979
FIPS
correction.

30113
MN
Yellowstone
Natl
Par
Yellowstone
NP
assigned
to
neighboring
counties.

46131
SD
Washbaugh
Co
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
4
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
CountyYear
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
CountyYear
are
described
below.
The
contents
of
the
I/
M
files,
and
the
I/
M
program
implementation
schedule
that
is
reflected
in
CountyYear,
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
CountyYear
table,
for
the
1999
bas
year
only.
After
the
base
year,
I/
M
programs
are
added,
modified,
and
dropped,
and
the
CountyYear
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
CountyYear
for
the
first
implementation
year
and
all
following
years.
I/
M
programs
were
assumed
to
remain
in
place
indefinitely
once
started.
5­
5
I/
M
Programs
and
Implementation
Schedules
Several
initial
processing
steps
were
performed
on
the
I/
M
data.
For
the
NEI
I/
M
files,
program
ParseIM.
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
6
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
CountyYear
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
7
TABLE
5­
3
Examples
of
Differences
in
I/
M
Program
Data
State
Differences
and
Resolution
Colorado
El
Paso
County
(
Colorado
Springs)
uses
Denver
program
in
NEI.
Resolution:
Re­
assign
to
use
the
Colorado
Springs
file
CO95C.
IM.

Florida
No
information
on
Florida
in
Counties/
Model
Resolution:
I/
M
program
discontinued
after
1999.

Georgia
NEI
files
use
1992
program
start,
but
Model
says
10/
1998
Resolution:
Leave
NEI
start
in
place
for
better
NMIM/
NEI
consistency.

Idaho
Not
listed
in
Counties.
Resolution:
Use
NEI/
Model
as­
is.

Kentucky
Northern
Kentucky
counties:
Not
listed
in
Counties.
Resolution:
Use
NEI/
Model
as­
is.

Louisiana
Not
listed
in
NEI.
Resolution:
Add
new
program
2002
start.

Maryland
Counties
missing
Baltimore
City
(
likely
a
FIPS
code
issue)
Resolution:
Use
NEI/
Model
as­
is.

Massachusetts
Model
indicates
MA31
test,
NEI
uses
Idle.
Resolution:
Transition
to
MA31
test
in
MA95.
IM.

Minnesota
No
information
on
Minnesota
in
Counties/
Model
Resolution:
I/
M
program
discontinued
after
1999.

Missouri
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.

New
Hampshire
Not
listed
in
NEI.
Resolution:
Add
new
OBD
program
2002
start.

New
Jersey
Model
indicates
ASM5015
test,
NEI
uses
Idle.
Resolution:
Transition
to
ASM5015
test.

New
York
Counties
indicates
OTR
Low­
Enhanced
program
for
non­
NYC
counties.
Resolution:
Add
separate
program
for
upstate
counties.

North
Carolina
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.

Oregon
Counties
includes
Columbia,
Yamhill
counties.
Resolution:
Add
counties
to
enhanced
program.

Pennsylvania
Counties
includes
several
counties
added.
Resolution:
Add
counties
to
new
program.

Rhode
Island
Model
indicates
RI2000
test,
NEI
uses
Idle.
Resolution:
Transition
to
RI2000
test.

Utah
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
8
5­
9
TABLE
5­
4
New
External
IM
Program
Files
File
Comments
GA01.
IM
Add
counties
to
program
in
GA99.
IM.

IN01.
IM
Add
county
to
program
in
IN97.
IM.

LA00.
IM
New
OBD
program.

NH02.
IM
New
OBD
program.

NY01.
IM
New
program
for
upstate
counties.

NC01.
IM
Add
counties
to
program
in
NC87.
IM.

OR01P.
IM
Add
counties
to
program
in
OR98P.
IM.

PA01OLE.
IM
New
program
for
additional
counties.

To
modify
the
base
year
file
to
add
future
year
I/
M
programs,
the
"
rewriteIM.
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
5­
10
baseline
for
2007,
the
modified
files
showed
lower
emissions
in
general,
due
primarily
to
the
added
OBD
programs.
6­
1
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_
CHIEF_
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
2
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
NMIM
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
3
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:

CountyYear;
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:

CountyYear;
field:
ATPFileName).
ATP
programs
are
typically
associated
with
IM
programs.

The
representative
county
mapping
methodology
assumed
that
if
the
IM
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
NMIM
database
contains
nine
fuel
identification
fields
for
each
county
by
year
and
month
(
table:
CountyYearMonth;
fields
HwyDieselID,

HwyGasolineID,
NRGasolineID,
NRDieselID,
HwyGasolineIdA,
HwyGasolineIdB).
The
predominant
fuel
used
in
each
county
is
described
by
the
highway
gasoline
identification
number
(
table:
CountyYearMonth;
field:
HwyGasolineID);
therefore
this
field
only
was
used
in
the
representative
county
mapping
process.

Criteria
6:
Time
Frame
 
The
intent
of
the
NMIM
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
4
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
5
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
"
LDG
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:

LDG
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
6
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.
7­
1
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
2
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.
8­
1
8.0
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
PADD
IV.
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.
8­
2
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
and
Guidance
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.
A­
1
APPENDIX
A
INDEX
TO
DATA
FILES
AVAILABLE
ELECTRONICALLY
NMIM
Table
Name
File
Name
Description
Reference
Tables
Folder
HPMSRoadType
vehicles2.
xls
The
spreadsheet
was
taken
from
the
June
2002
National
Emissions
Inventory
(
NEI)
update
files.

M6VType
Appendix
B
of
MOBILE6
User's
Guide
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).

M6VClass
Appendix
B
of
MOBILE6
User's
Guide
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
CountyYearMonth
Gasoline
Gasoline2
CYMMTBEPhsOut
GasMTBEPhsOut
Gas2MTBEPhsOut
CYM_
Gas.
mdb
CYM_
GasMTBE.
mdb
030425_
gasoline
assignments
and
parameters.
xls
Final_
Fuel1999V3_
032
703.
xls
RFG_
Fuel00_
v1.
xls
Access
97
databases
which
produces
the
CountyYearMonth,
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.

Diesel
sulfur.
xls
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?]

NaturalGas
sulfur.
xls
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
NMIM
Table
Name
File
Name
Description
A­
2
CountyYear
stage2dat.
xls
00tables.
wpd
trends99_
pointer.
dbf
trends99_
im.
xls
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,
00tables.
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.

Vehicle
Tables
Folder
AverageSpeed
vmt99_
f_
m6_
with_
8Sta
tesUpdate.
DBF
Modeling
files
for
June
2002
update
to
the
1999
NEI,
extracted
from
the
VMT
table.

BaseYearVMT
vmt99_
f_
m6_
with_
8Sta
tesUpdate.
DBF
vmt99_
n_
m6.
dbf
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.

VMTGrowth
BaseYearVMT
table
V­
2.
xls
V­
3.
xls
V­
4.
xls
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).

VMTMonthAllocation
1999
NEI
document
"
Onroad
Source
Methodologies"
dated
10/
2001
Additional
Tables
Folder
County
EPA_
CHIEF_
county_
fips.
xls
The
spreadsheet
was
retrieved
from
http://
www.
epa.
gov/
ttn/
chief/
codes/
index.
html.

CountyMonth
mxmntp99.
dbf
pr_
vi_
temps.
xls
Fall
2001
update
to
the
1999
NEI.

State
trends99_
pointer.
dbf
June
2002
update
to
the
1999
NEI,
and
Fall
2001
update
to
the
1999
NEI.
See
1999
Updated
M6
Input
Files\*.*
under
CountyYear
for
source
for
NLEV
file
references.

DB
Documentation
Folder
Not
applicable
CountyDB.
pdf
CountyDB.
rtf
County
database
documentation
SQLScripts
Folder
NMIM
Table
Name
File
Name
Description
A­
3
All
NMIM
Tables
Load_
Data_
Tables.
sql
Load_
Data_
BaseYearV
MT.
sql
Load_
Data_
VMTGrowt
h.
sql
QA_
Script_
1.
sql
QA_
Script_
2.
sql
QA_
Script_
3.
sql
Data
Loading
Scripts
Quality
Assurance
Scripts
Data
Files
Folder
Vehicle
tables
GenerateVMTMoAlloc.
awk
Code
to
Produce
VMT
Data
County
Rep_
County_
Mapping.
s
ql
Representative
County
Mapping
Code
