Instructions and Best Practices for Development and Submittal of Onroad Inputs for the 2020 National Emissions Inventory (NEI) February 2021


Instructions and Best Practices for Development and Submittal of Onroad Inputs for the 2020

National Emissions Inventory (NEI)

February 2021

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

Contacts:

Janice Godfrey, Alison Eyth

-------
ITS

1	Introduction	4

2	Overview of Onroad Inventory Preparation	4

3	MOVES3 Onroad County Database Content	7

4	Data used to generate MOVES CDBs	15

4.1 County Database Naming Convention	17

5	Steps to Submit CDBs	18

5.1	Creating County Databases (CDBs)	18

5.2	Use the CDB QATool to Create a QA Report to include with your submittal	19

5.3	Create Your CDB Checklist	22

5.4	Documentation	24

5.5	Bundle CDB Submittal Components and Create the XML for EIS Submittal	24

6	Key Data Sources for Inputs	29

6.1	Representative Counties	30

6.2	Hotelling Hours Per Day	31

6.3	Age Distribution	31

6.4	Hourly Average Speed Distribution	32

6.5	VMT Distribution by Road Type	34

6.6	Average Speed Distribution	35

6.7	VMT/VPOP Ratio	36

6.8	Hour, Day, and Month VMT Fractions	38

6.9	QA Resolution	40

6.10	Considerations specific to 2020	40

7	Data Sources	41

7.1 Additional Resources	41

Table of Figures

Figure 1. Schematic of Onroad Inventory Preparation	5

Figure 2. 2017 State/Local Submitted Data for 2017 NEI	16

Figure 3. Example State Submissions	17

Figure 4. Example of QA Report Results	21

Figure 5. CDB Checklist	23

Figure 6. Complete Agency Submittal Checklist	24

Figure 7. OnRoad CDB Zipping Example	25

2

-------
Figure 8. Main Menu Bridge Tool	26

Figure 9. CERS XML Wrapper Example	28

Figure 10. EIS Gateway - Agency Organization Detail: Feedback	29

Figure 11. Representative County Map	30

Figure 12. 2017 Age Distribution Example	32

Figure 13. Speed Distribution Example	34

Figure 14. VMT Distribution Example	35

Figure 15. Hourly Average Speed Profile Example	36

Figure 16. VMT/VPOP Ratio Example Plot	38

Figure 17. VMT Fraction Vehicle Example	39

Figure 18. VMT Fraction Non-Existing Road Type Example	40

Table of Tables

Table 1. MOVES CDB Tables and Contents	9

Table 2. MOVES Speed Bins	33

Table 3. Enforced Maximum Allowable VMT/VPOP Ratios in 2020 NEI	37

3

-------
1 Introduction

The U.S. Environmental Protection Agency (EPA)'s Air Emissions Reporting Rule requires state and local
agencies who submit to the National Emissions Inventory (NEI) to submit model inputs for mobile
sources, rather than emissions, in the form of county-specific databases (CDBs). Emissions estimates may
be submitted in addition to model inputs, and emissions estimates will still be accepted from tribes without
inputs. If you do not submit onroad input data, EPA will generate emission estimates using EPA-developed
county-specific inputs.

The EPA's Motor Vehicle Emission Simulator—commonly referred to as MOVES—is a set of modeling
tools for estimating air pollution emissions produced by onroad (highway) and nonroad mobile sources.
For the 2020 NEI, EPA will use the most current version of the Motor Vehicle Emission Simulator
(MOVES) model, MOVES3, to estimate both onroad and nonroad emissions. For onroad mobile sources,
MOVES3 is used to generate emission factors, which are then used to calculate onroad emission
inventories by multiplying emission factors by the appropriate emission-related activity for each county
and SCC. Vehicle population and other types of activity data are ever-changing as new historical data
becomes available and new projections are generated. The EPA receives MOVES CDBs from many state
and local (S/L) air agencies to populate the various MOVES inputs. The EPA also develops a set of input
activity that is used for S/L areas that do not provide MOVES CDBs. This document is meant to serve both
as instructions and a guide to best practices when developing S/L CDBs for the NEI. Further information
on development of data for use by MOVES3 is provided in the EPA document, "Population and Activity
of Onroad Vehicles in MOVES 3". (EPA-420-R-20-023, November 2020). Some changes with MOVES3
over previous versions of MOVES that affect CDB submissions include:

• No need to estimate ramp fractions

• Source type 41 is now "Other Buses" (non-school, non-transit), rather than "Intercity Buses"

• A few additional source and fuel type combinations are allowed (e.g., "Other Buses" can be
gasoline, diesel, or CNG; instead of just diesel)

2 Overview of Onroad Invei paration

Onroad mobile source emissions result from motorized vehicles operating on public roadways. These
include passenger cars, motorcycles, minivans, sport-utility vehicles, light-duty trucks, heavy-duty trucks,
and buses. The sources are further divided by the fuel they use, including diesel, gasoline, E-85, and
compressed natural gas (CNG) vehicles. The sector characterizes emissions from parked vehicle processes
(e.g., starts, hot soak, and extended idle) as well as from on-network processes (i.e., from vehicles as they
move along the roads). Except for California, all onroad emissions are generated using the SMOKE-
MOVES framework that leverages MOVES-generated emission factors, county and SCC-specific activity
data, and hourly meteorological data. The onroad source classification codes (SCCs) in the modeling
platform are more finely resolved than those in the National Emissions Inventory (NEI). The NEI SCCs
distinguish vehicles and fuels while the SCCs used in the modeling platform also distinguish between
emissions process (e.g., running exhaust, start exhaust and evaporative emissions), and road types.

Figure 1 shows the data flow used to create an onroad emissions inventory. The emission rate (i.e.,
"lookup") tables input to SMOKE-MOVES are generated by MOVES. These tables differentiate emissions
by process, fuel type, vehicle type, road type, temperature, speed, hour of day, and day of week. To

-------
generate the MOVES emission rates that could be applied across the U.S., MOVES is ran to produce
emission factors for a series of temperatures and speeds for a set of "representative counties," to which
every other county in the country is mapped. Representative counties are used because it is impractical to
generate a full set of emission factors for the more than 3,000 counties in the U.S. The representative
counties for which emission factors are generated are selected according to their state, elevation, fuels used
in the region, vehicle age distribution, and inspection and maintenance programs. Every county in the
country is then mapped to a representative county based on its similarity to the representative county with
respect to those attributes. For vehicle age distributions and fuel types, rather than choose values specific to
each representative county, a weighted average was computed for all counties represented by each
representative county, and the mean of those averages was used.

Met. Data
from WRF

Meteorological
Preprocessor
(Met4Moves)

Temperature ranges,
rel. humidity

Use representative
EFs and county/
grid-specific
activity data and
meteorology to
create emissions
for all counties

SMOKE

MOVES

Run MOVES to get
emission factors
(EF) for

representative
counties for each
temperature and
speed needed

Activity Data
(VMT, VPOP,
+ Hoteling
hours for all
counties)

AQ model-ready files / NEI

Figure 1. Schematic of Onroad Inventory Preparation

This document addresses submitting onroad M0VES3 CDS inputs only. For information on submitting
MOVES3-Nonroad CDBs, see the companion document "Instructions for Submitting M O V E S 3 - N on road
Inputs to the 2020 National Emissions Inventory (NEI)"

MOVES3 includes significant updates to many input tables to include fuels and inspection and
maintenance (I/M) program data and these data should be used instead of the data from versions of
MOVES2014. EPA recommends that data submitters use local county-specific data for activity, fleet
information, speeds, and temporal profiles if local data are available. Any local data originally derived for
use with versions of MOVES2014 would likely need to be updated as well. EPA recommends creating

-------
new input databases using M0VES3 rather than attempting to convert and update existing input databases.
However, for existing input databases that still contain the latest available information, MOVES3 includes
scripts in the "Tools" menu that will convert input databases created with any version of MOVES2014 to
MOVES3 format. EPA has created a set of 2020 starting point CDBs for S/L agencies to download and
optionally replace certain tables where local data is available. The starting point CDBs are already
formatted for MOVES3, so agencies do not need to use any conversion tools if working from these. The
starting point CDBs contain entirely placeholder data from the MOVES3 database and in most cases, these
table do not represent final "EPA default" data for NEI. The MOVES3 GUI includes instructions and
additional help files for using these conversion tool. Note: onroad CDB submissions must be checked
using a quality assurance (QA) script (provided by EPA) that is specific to MOVES3. The MOVES3
onroad CDB QA script will be available by the spring of 2021.

Tip: the CDBs for the 2020 NEI are not due from states until January 2022. So, there
will be a year or so from the time of the model release to the NEI due date. EPA will
publish a set of CDBs that the states can start from for their submissions. These will be
provided in MOVES3 format by spring of2021. Additionally, EPA will provide a
converter for CDBs prepared using MOVES2014 b. Outreach on the convertor's use will
be provided through the MJO MOVES workgroup or some other forum.

MOVES3 inputs must be submitted to the EPA Emissions Inventory System (EIS) for each county as a
County Database (CDB), which consists of a set of MariaDB database tables specifically formatted to store
county-specific inputs for MOVES3. Prior to submitting MOVES3 inputs, agencies may download EPA's
draft 2020 CDBs for your submitting agency from the 2020 NEI FTP site

(ftp://newftp.epa.gov/air/nei/202Q/doc/supportine data/onroad/inil ') to use as a template. Note: if

your agency or browser does not support access to ftp:// sites, you may replace the ftp://newftp.epa.gov
portion of any FTP links with https://gaftp.epa.gov and you will be able to access the same data. The
emissions in the NEI are developed for all months of the year and all counties in every state, plus the
District of Columbia, Puerto Rico, and US Virgin Islands. This means that EPA will need to have county-
level information for every county. For the NEI, MOVES3 will be run at the county scale, which requires a
separate county database for each county that contains the data specific to that county.

Since the MOVES3 inputs are county-based, tribal agencies should run MOVES3 and submit emissions.
However, tribal agencies may use the input information from adjacent counties to prepare local inputs that
may be suitable for their tribal area MOVES3 runs.

On a county-by-county basis, agencies can change the CDBs as needed to reflect their own input
data, or they may choose to use the EPA-provided inputs. Agencies must check their CDBs prior
to submittal using the provided QA Tool that will generate a QA Report to include with
submittals. Agencies should submit their CDB submittal package to the EIS QA environment to
confirm there are no errors prior to submitting to the production environment. Agencies that want
to accept EPA's defaults and submit nothing may do so via a 'support request' message that states
that intent, through the EIS gateway.

-------
Tip: Indicate in your submittal checklists which tables contain any local data, using
footnotes to indicate any limitations or caveats (e.g. local data age distributions for only
certain source types).

The following sections describe how to create/revise onroad CDBs, QA check them, and submit
them to EIS.

Supporting tools and files referenced here can be downloaded from the 2020 onroad
supporting data FTP site:

ftp://newftp.epa.eov/air/nei/2020/doc/supportine data/onroad/.

3 III-" I, *1 'nroad Coum I 1 itabase Content

States are asked to supply the CDB tables listed in Table 1. MOVES CDB Tables and Contents
below, with a focus on those marked as Medium and High priority for the NEI. Table 1 indicates
the contents of the CDBs and the data EPA plans to use as defaults. Note that the values in the
2020 starting point CDBs are largely a placeholder developed from the MOVES3 database; the
EPA default data will be developed later during the 2020 NEI process. It is expected that states
will have more accurate information in many cases than MOVES3 placeholder values and EPA
default data, so we encourage submittals.

The CDB tables should contain the complete set of information needed to run MOVES3 for all
the vehicles in a county for all months of a single calendar year (2020) using county-specific
information.

As stated in the section above, starting point 2020 CDBs based on MOVES3 are available here:

ftp://newftp.epa.eov/air/nei/2020/doc/support.ine data/onroad/initial CDBs/. These CDBs may
be used by states as a starting point for generating CDBs to provide to EPA or states may provide
CDBs created independently using MOVES3. Any CDBs submitted to EIS for the 2020 NEI
should be complete and ready to run for the calendar year 2020.

Tip: The EIS Production Submission Window opens for S/L submittals 7/1/21. S/Ls last-
day for EIS submittal of Point, Onroad Mobile, Nonroad Mobile and I/vents data
category emissions is 1/15/2022. More information can be found here:

https://www.eva.sov/air-emissions-inventories/2020-national-emissions-inventory-nei-
documentation

Table 1 contains many tables that are marked as "Not needed" to be supplied by agencies. For example,
EPA uses refinery production compliance data to develop the fuel supply. S/Ls are therefore not expected
to supply fuel property information. The EPA uses meteorological data derived from the Weather Research
and Forecasting (WRF) model to develop MOVES3-ready meteorological inputs. S/Ls are therefore not

-------
expected to populate fuelformulation, fuelsupply, or zonemonthour tables in their submitted CDBs. For
more information on table information see the IV	Hub:

https:// eithub. com/LI SEP A/EPA. ]	iodel/blob/master/docs/MOVESDatabaseTables.md

8

-------
Table 1. MOVES CDB Tables and Contents

CDB Tables and
Contents

Description of
Content

EPA-Default CDB Table
Content

Priority Level
for NEI

auditlog**

Information
about the
creation of
the

database



Not needed

avft**

Diesel sales
fractions

The starting point CDBs
have MOVES3
placeholder values. EPA
defaults for NEI will
likely come from 2020
county-level registration
data.

Low

(EPA has a
good source)

avgspeeddistribution

Average speed
distributions. See
Section 6.6 for
more information.

Data specific to the
county or group of
similar counties.
EPA's starting point
CDBs have national
average placeholder
values for this table
that will change
when 2020-specific
data become
available.

Medium

county

Description of the
county

EPA provided data.

Not needed

countyyear*

Description of
the Stage 2
program

EPA's starting point
CDBs do not include this
table and it is not required
for a county-scale run.
S/L agencies may
optionally provide this
data if local data is
different than MOVES3.

Not needed

dayvmtfraction

VMT

distribution
across the type
of day.

Data specific to the
county or a group of
similar counties. This
may be default data until
2020-specific data
become available.

High

9

-------
CDB Tables and
Contents

Description of
Content

EPA-Default CDB Table
Content

Priority Level
for NEI

emissionratebyage*

Emission rates to
reflect adoption of
California
emission
standards.

The

Emi s si onRateBy Age
tables for some counties
have been populated
using the appropriate
data described in the
guidance for

states adopting California
emission standards. See
notes below ***

Not needed

(EPA already
accounts for
this)

fuelformulation

Fuel properties

EPA's starting point 2020
CDBs already have
placeholder values for this
table. Final NEI data will
come from 2020 refinery
data.

Not needed

fuel supply

Fuel differences
by month of the
year

EPA's starting point
2020 CDBs already have
placeholder values for
this table. Final NEI data
will come from 2020
refinery data.

Not needed

fuel supplyy ear*

Year for the fuel
properties

Set to 2020 in CDBs or can
leave it blank.

Not needed

fuelusagefraction

Fuel use
by flexi-
fuel

vehicles.

Based on EPA estimates
for 2020. EPA's starting
point CDBs already have
values for this table from
MOVES3 that will likely
remain the final 2020
values. S/L agencies may
override these with local
data.

Not needed

hotellingactivity distribution

**

Distribution of
hotelling hours
using extended
idle or

auxiliary power
units.

Based on EPA estimates
for 2020. EPA will
consider submitted
hotelling activity
distributions in place of
EPA estimates. This table
is empty in the starting
point CDBs and is not
required to be filled by S/L
agencies.

Low

10

-------
CDB Tables and
Contents

Description of
Content

EPA-Default CDB Table
Content

Priority Level
for NEI

hotellingagefraction

Allocates
hotelling activity
across vehicle
ages

Empty (using national
average EPA estimates).

Not needed

hotellinghourfraction

Allocates
hotelling activity
by day type across
hours of the day

Empty (using national
average EPA estimates).

Not needed

hotellinghoursperday

Hours of hotelling
by day type.

See Section 6.2
for further
information.

County-specific hours for
2020 EPA will consider
submitted hotelling hours
in place of EPA estimates.
This table is empty in the
starting point CDBs and is
not required to be filled by
S/L agencies.

Not needed

hotellingmonthadjust

Adjusts hotelling
activity by month

Empty (using national
average EPA estimates).

Not needed

hourvmtfraction

VMT distribution
across the
hours of the day.
See Section 6.8
for more
information.

Data specific to the county
or a group of similar
counties. EPA's starting
point CDBs have national
average placeholder values
for this table that will
change when 2020-specific
data become available.

High

hpmsvtypeday**

Daily VMT by
HPMS vehicle
type

Alternate method
of providing
VMT by HPMS
vehicle types,
day types, and
months.

This table is empty in
EPA's starting point
CDBs because the
S ourceT y pe Y ear VMT
table is used.

High

(note only 1
of the 4 VMT
tables should
be populated)

hpmsvtypeyear**

Total annual
VMT by HPMS
vehicle type

Alternate method
of providing
VMT by HPMS

This table is empty in
EPA's starting point
CDBs because the
S ourceT y pe Y ear VMT
table is used. EPA
default VMT in the NEI
will likely come from

High

(note only 1
of the 4 VMT
tables should
be populated)

11

-------
CDB Tables and
Contents

Description of
Content

EPA-Default CDB Table
Content

Priority Level
for NEI



vehicle type.

2020 FHWA data and
other sources. States may
provide VMT estimates
by MOVES3 source type
in SMOKE format
directly for use in the
2020 NEI.



idledayadjust**

Adjusts off-
network idling
activity by day
type

Empty (using national
average EPA estimates).

Not needed.

idlemodelyeargrouping* *

Total amount of
off-network idling
activity as a
fraction of source
hours operating by
source type and
model year range

Empty (using national
average EPA estimates).

Not needed.

idlemonthadjust**

Adjusts off-
network idling
activity by month

Empty (using national
average EPA estimates).

Not needed.

imcoverage

Description of
the Inspection
and Maintenance
program

I/M program description
in the starting point CDBs
is from MOVES3
database for the year
2020. Update only if this
data needs correction.

High

monthvmtfracti on

VMT

distribution
across the
months of the
year.

The starting point
CDBs have
MOVE S3
placeholder values.
EPA defaults for
NEI will likely come
from 2020-specific
data when it
becomes available.

High
(Capture
pandemic
effects on
2020)

onroadretrofit**

Emission
adjustments
to reflect

Empty (no retrofits
assumed).

Not needed

12

-------
CDB Tables and
Contents

Description of
Content

EPA-Default CDB Table
Content

Priority Level
for NEI



local retrofit
programs.





roadty pedi stributi on

VMT

distribution
across the road
types. See
Section 6.5 for
more

information

The starting point CDBs
have MOVES3
placeholder values that
are nationwide average.
EPA default

RoadTypeDistribution in
the NEI will by county-
specific and likely come
from 2020 FHWA data
and other sources.

High

sourcety peagedi stributi on

Distribution of
vehicle ages. See
Section 6.3 for
more information.

The starting point CDBs
have MOVES3
placeholder values. EPA
defaults for NEI will
likely come from 2020
county-level registration
data.

High

sourcetypeday vmt* *

Daily VMT by
source type.

Alternate method
to provide VMT
by source types,
day types, and
months.

This table is empty in
EPA's starting point CDBs
because the
S ourceT y pe Y ear VMT
table is used.

High

(note only 1
of the 4 VMT
tables should
be populated)

sourcetypeyear

Vehicle
populations

The starting point CDBs
have MOVES3
placeholder values. EPA
defaults for NEI will
likely come from 2020
county-level registration
data.

High

sourcetypeyearvmt

Total annual
VMT by source
type.

Preferred method
to provide VMT
by source types.

The starting point CDBs
have MOVES3
placeholder values. EPA
default VMT in the NEI
will likely come from 2020
FHWA data and other
sources.

High

(note only 1
of the 4 VMT
tables should
be populated)

13

-------
CDB Tables and
Contents

Description of
Content

EPA-Default CDB Table
Content

Priority Level
for NEI

starts**

Number of engine
starts

Empty (using national
average EPA estimates).

Not needed

startsageadjustment* *

Adjustment
factors to reflect
differences in
starts per day by
age.

Empty (using national
average EPA estimates).

Not needed

start shourfracti on* *

Distribution of
starts by hour of
the day

Empty (using national
average EPA estimates).

Not needed

startsmonthadjust* *

Variation of
the starts per
vehicles by
month.

Empty (using national
average EPA estimates).

Not needed

startsperday**

Engine starts per
day/hour per
vehicle.

Empty (using national
average EPA estimates).

Not needed

start sopmodedi stributi on * *

Engine soak
distributions.

Empty (using national
average EPA estimates).

Not needed

startsperdaypervehicle* *

Engine starts per
day per vehicle.

Empty (using national
average EPA estimates).

Not needed

state

Description of the
state

EPA-provided data.

Not needed

totalidlefraction* *

Total amount of
off-network idling
activity as a
fraction of source
hours operating by
source type,
model year range,
month, and day
type

Empty (using national
average EPA estimates).

Not needed

year

Year of the
database

Set to 2020.

Not needed

zone

Allocations of
starts,

extended idle
and vehicle
hours parked
to the county

Allocations must all be 1.0
(100%). This data should
not be changed.

Not needed

zonemonthhour

Temperature
and relative
humidity

The starting point
CDBs have
MOVES3.

Not needed

14

-------
CDB Tables and
Contents

Description of
Content

EPA-Default CDB Table
Content

Priority Level
for NEI



values

placeholder values
only to pass QA
checks. EPA will use
WRF meteorology
for the NEI, so states
do not need to
override this table.



zoneroadtype

Allocation of
road types to
the county

Allocations must all be 1.0
(100%). This data should
not be changed.

Not needed

*Tables that are not created by MOVES County Database Manager (CDM), but are
generated as empty tables by the QA tool that checks CDBs for EIS submittal

**Tables that can be empty but must be present in EIS submittal

*** The following states are given early NLEV programs in the EPA defaults:

•	Connecticut (9)

•	Delaware (10)

•	District of Columbia (11)

•	Maryland (24)

•	New Hampshire (33)

•	New Jersey (34)

•	Pennsylvania (42)

•	Rhode Island (44)

•	Vermont (50)

•	Virginia (51)

EPA accounts for states that have adopted LEV standards.

4 Data used to generate

MOVES CDBs may include the following input data: vehicle miles traveled (VMT), vehicle
population, vehicle starts, average speed distribution, fuels, hotelling, age distributions, hourly
average speed profiles, VMT fraction (by month, day, hour, and road type), inspection and
maintenance (I/M) program descriptions, and stage II refueling program effectiveness. States
often develop some of these inputs based on output data from their state departments of
transportation, which may include travel demand model. The EPA also develops a set of input
values to use in the event a state does not supply inputs. These input values will be developed
later in the NEI process, so are not present in the starting point CDBs that contain entirely
MOVES3 default information. For NEI defaults, the EPA uses data from various sources and
studies (e.g., FHWA, Coordinating Research Council studies such as ( Ki \ 100, t Ki. \ II ">)
to create a set of CDBs for inputs not provided by the S/Ls. Figure 2 shows state and local
regions in dark blue that submitted CDBs for the 2017 NEI. California is the exception; they use

15

-------
their own model and send emissions to EPA. There are also a few tribes that send emissions to
EPA instead of CDBs.

Figure 2. 2017 State/Local Submitted Data for 2017 NEI

Note that not all states submit data for all counties and/or MOVES inputs. Figure 3 provides a
snapshot of the number of counties within each submitting state that submitted data in the
identified MOVES tables for the 2017 NEI. For example, Georgia submitted sourcetypeyear for
all counties in the state, but submitted startsperday for 20 counties. The complete table for all
submitting agencies can be found in the 2017 NEI Technical Support Document (TSD).

16

-------
Arizona
(Maricopa)

1

1

1

1

1

1







1

1

1

1

1

1

1



Arizona
(Pima)

1

1



1











1

1

1

1

1

1

1



Connecticut

8 8

8





8





8

8

8

8

8

8 8

Delaware

3

3

3

3

3

3

3

3

3

3

3

3

3

3

3

3



District of
Columbia

1

1

1

1

1





1

1

1

1

1

1

1



Florida

67

67











67

67

67

67

67

67

67



Georgia



24

13

1



47







24

159

13

159

24 159

159

159

20

Idaho

44

44



44

44

44

44





44

44

44

44

44

44

44



Illinois



102

102

102

102

102

102





102

102

11

102

102

102

102



Kentucky
(Jefferson)

1















1





1

1

1



Maine

16

16

16

16

16





16

16

1

16

16

16

16



Maryland

24

24

24

24

24

24

24





24

24

24

24

24

24

24



Massachus.

14















14

14



14

14

14



Missouri

48



5

115





115





115

115

5

115









Figure 3. Example State Submissions

After the submission window closes, the EPA merges default data and state-submitted data to
come up with a complete set of CDB data.

4.1 County Database Naming Convention

To keep track of the thousands of CDBs used in the NEI calculations, EPA has established a
naming convention for CDBs that differentiate between databases and make automation of
running and processing the inputs and outputs from MOVES easier.

The naming convention for each CDB folder has 20 characters. The first 6 characters identify the
county, the next 5 indicate the calendar year of the county database, and the last characters
indicate the date on which the database was created.

The first 6 characters consist of the letter "c" followed by the 5-digit Federal Information
Processing Standard (FIPS) code for the county, including a leading zero when necessary. The
next 5 characters are the letter "y", followed by a 4-digit calendar year. This calendar year
indicates the calendar year of the data contained in the database. A CDB can only contain data
from a single calendar year. The last 8 digits, following and underscore character, are the date on
which the database was created in a YYYYMMDD format.

An example of a CDB name is "c26161y2020_20210601" where this CDB names indicates
"c26161" refers to the county FIPS code (in this case Washtenaw County, Michigan). "y2020"

17

-------
refers to the calendar year for the county database and "20210601" identifies the database
modification date of June 1, 2021, in YYYYMMDD format.

5 Steps to Subn

The steps for submitting CDBs are discussed in more detail in the subsections that follow:

1. Create CDBs

a. Edit existing CDBs to meet NEI requirements or

b. Create new CDBs from scratch or

c. Revise EPA inputs.

2. If necessary, run converter tool to convert CDBs prepared using MOVES2014b to
MOVES3. Note that the starting point CDBs are already in MOVES3 format.

3. Run EPA's MOVES3 QA Tool and revise the CDB as needed until the QA Tool
creates a QA report that confirms no errors exist.

4. Create a Checklist that indicates where changes have been made to the starting
point CDBs or where local data were provided in S/L-developed CDBs. Please
add footnotes as necessary to convey any nuances in how much data is local for a
particular table(s).

5. Provide documentation for the Agency- supplied inputs.

6. Submit the files to the EIS.

5.1 Creating County Databases (CDBs)

The NEI relies on MOVES runs at the County Scale which requires input CDBs as a way to
provide the model with data representative of the county.

There are many ways to create a CDB for submission. We prefer that the submitter starts from
EPA's 2020 starting point CDBs to perform county specific edits. However, if starting from
older CDBs prior to MOVES3, please see Tools to Develop or Convert MOVES Inputs for the
latest tools to develop or convert MOVES inputs.

Tip: Please ensure that the IMCoverage table covers flex-fuel vehicles in
addition to gasoline vehicles.

Tip: Please ensure that the IMCoverage table contains correct countylDs.

Tip: Please ensure you use the correct table structure (keys, column order, etc.) on all
CDB tables. This is guaranteed if you either create CDBs with the MOV'ESS County
Data Manager or use EPA's starting point CDBs.

-------
5.2 Use the CDB QA Tool to Create a QA Report to include with your submittal

This section is a placeholder and will be updated once the MOVES3 QA Tool becomes
available.

For Agencies submitting onroad CDBs based on MOVES3, from 2020 on road supporting data
FTP site: ftp://newftp.epa.eov/air/nei/2020/doc/supportine data/onroad/. we will update this
section once the actual package is available with specific file names and instructions. The QA
Report generated by the QA Tool script verifies all table contents meet range, naming
convention, format and other checks. The report confirms that each CDB contains the
appropriate number of tables and that the values within those tables are valid. It is very important
that the QA tool be properly run for each submitted CDB, because if it is not and the CDB
contains errors that are not resolved, the data may not work when we go to generate the NEI and
we will need to follow up with you at that time, or we may not be able to use your submitted
data.

EIS will check to see that each county listed in this report has an associated CDB in the
submission. Only include counties in the QA report that you will be submitting. If the counties in
the QA report do not match those in the CDB folder, or if any errors are indicated in the QA
report, EIS will indicate a critical error and will not accept the submission. The format is .txt, but
this file can be opened as a table using Microsoft Excel to make it more readable.

IMPORTANT NOTE: Some tables may not be populated. If you create the CDBs using the
MOVES County Data Manager (CDM) from scratch, the CDM will create most of these tables.
Your CDB will be missing two tables:

o EmissionRateByAge

o County Year

Running the QA Tool will create these tables in your CDB, but the tables will each contain zero
rows (i.e., they will be empty). Leaving these tables empty will not cause the CDBs to fail the
script if it is run again. Thus, if you create the CDBs using the CDM, your initial run of the QA
will indicate errors (missing tables). However, if you run the QA Tool again, the tables have
been created by the previous check and the "missing tables" error will no longer appear in the
report. You can then send the second QA report (with no errors) to the EIS.

Also, the QA tool uses tables from the default MOVES onroad database that was distributed with
MOVES3. If you do not have MOVES3 installed, you will need to obtain a copy of the database
folder specified in the QA script and place it in your MariaDB\data directory.

Name your QA report with your agency Program System Code (PSC), such as
"PSC_QA_Report.txt". For example, Delaware's state agency PSC is "DEDNR" and their QA
report would be named "DEDNR_QA_Report.txt".

If your state has many counties, you may wish to automate the checking process. Below is an
example of a batch file written to check the three counties for the state of Delaware. The batch
file deletes the old version of the report text file (named with the PSC), clears the MariaDB

-------
buffer of previous work and drops the database used to store the aggregated results from the
individual checks. Then MariaDB is directed to each county database (i.e.,
cl0001y2017_20180601) using the QA Tool script (CDB_QA_Checks_MOVES3.sql. Note that
the directory path must be included so that the file can be properly located by MariaDB. The path
you use can be different than the location (C:\MariaDB\) shown in this example. The final line in
the batch file exports the results into the report text file. A directory path can be added to this file
name as well to help locate the file once it has been populated.

del DEDNR_QA_Report.txt

MariaDB -P 3307 -uXXXXX -pYYYYY -e "flush tables;"

MariaDB -P 3307 -uXXXXX -pYYYYY -e "drop database if exists all_cdb_checks;"

MariaDB -P 3307 -uXXXXX -pYYYYY cl0001y2020_20210601 <

C: \Mari aDB\CDB_Q AChecksMO VES 3. sql

MariaDB -P 3307 -uXXXXX -pYYYYY cl0003y2020_20210601 <

C: \Mari aDB\CDB_Q AChecksMO VES 3. sql

MariaDB -P 3307 -uXXXXX -pYYYYY cl0005y2020_20210601 <

C: \Mari aDB\CDB_Q AChecksMO VES 3. sql

MariaDB -P 3307 -uXXXXX -pYYYYY -e "select * from

all cdb checks.all county database checks;" > DEDNR_QA_Report.txt"

The 3307 is the most likely* port number for MariaDB on a computer which already has MySQL
installed, XXXXX value is the MariaDB username, and the YYYYY value is the MariaDB
password. The script (CDB_QA_Checks_MOVES3.sql) is located in the C:\MariaDB\ directory
in this example. The file generated by this script (DEDNR _QA_Report.txt) is the quality
assurance (QA) report required by the EIS process.

*Note, your MariaDB port number is most likely 3307. The port number may be confirmed by
looking inside your MOVES3 installation directory file:

C:\Users\Public\EPA\MOVES\MOVES3.0\MySQL.txt. IfMySQL.txt is not present under
C:\Users\Public\EPA\MOVES\MOVES3.0\, then your port number is 3306.

-------
Figure 4. Example of QA Report Results

DEDNR QA Report results.xlsx - Excel

Brzezinslci, David S3

Insert Page Layout Formulas Data Review View 0 Tell me what you want to do

*

fa-
•!*

B I U

-	a a — =

|U	^ - s: s

nt	j|

./*	countylD

-	- §*' General

i' $ - % '

Alignment	<*	Number

m W &

Conditional Format as Cell
Formatting" Table" Styles"
Styles

gf3 Insert "
Delete *
gp Format"
Ceils

|: ir P

Sort & Find &
^ ' Filter" Select"



A

B

C

D

E

F

G

H

1

J

1

countylD jstatus

tableName

checkNumber

testDescription

testValue

count

dataBaseName

daylD

fuelFormulationID

2

10001 Completed

onRoadRetroFit

NULL

Table Check:

NULL

NULL

cl0001y2017_20181211

NULL

NULL

3

10001 Completed

importStartsOpmoDedistribution

NULL

Table Check:

NULL

NULL

cl0001y2017_20181211

NULL

NULL

4

10001 Completed

hotellingHours

NULL

Table Check:

NULL

NULL

cl0001y2017_20181211

NULL

NULL

5

10001 Completed

hotellingActivityDistribution

NULL

Table Check:

NULL

NULL

cl0001y2017_20181211

NULL

NULL

6

10001 Completed

fuelUsageFraction

NULL

Table Check:

NULL

NULL

cl0001y2017_20181211

NULL

NULL

7

10001 Error

avgSpeedDistribution

406

distribution error

0

NULL

Cl0001y2017_20181211

NULL

NULL

8

10001 Completed

VMT Tables

217

VMT tables with data: OK

1

NULL

cl0001y2017_20181211

NULL

NULL

9

10001 Completed

sourceTypeYearVmt

216

Number of Rows

13

NULL

cl0001y2017_20181211

NULL

NULL

10

10001 Completed

sourceTypeDayVmt

215

Number of Rows

0

NULL

cl0001y2017_20181211

NULL

NULL

11

10001 Completed

hpmsVTypeDay

214

Number of Rows

0

NULL

cl0001y2017_20181211

NULL

NULL

12

10001 Completed

hpmsVTypeYear

213

Number of Rows

0

NULL

cl0001y2017_20181211

NULL

NULL

13

10001 Warning

roadType

200

Number of Rows

0

NULL

cl0001y2017_20181211

NULL

NULL

14

10001 Completed

emissionRateByAge

NULL

Table Check:

NULL

NULL

cl0001y2017_20181211

NULL

NULL

15

10001 Completed

zoneRoadType

NULL

Table Check:

NULL

NULL

cl0001y2017_201S1211

NULL

NULL

16

10001 Completed

zoneMonthHour

NULL

Table Check:

NULL

NULL

cl0001y2017_20181211

NULL

NULL

17

10001 Completed

zone

NULL

Table Check:

NULL

NULL

cl0001y2017_20181211

NULL

NULL

18

10001 Completed

year

NULL

Table Check:

NULL

NULL

cl0001y2017_20181211

NULL

NULL

19

10001 Completed

sou rceTypeYear	

NULL

Table Check:

NULL

NULL

cl0001y2017_20181211

NULL

NULL

2G

10001 Completed

roadTypeDistribution

NULL

Table Check:

NULL

NULL

Cl0001y2017_20181211

NULL

NULL

21

10001 Completed

roadType

NULL

Table Check:

NULL

NULL

cl0001y2017_20181211

NULL

NULL

22

10001 Completed

monthVmtFraction

NULL

Table Check:

NULL

NULL

cl0001y2017_20181211

NULL

NULL

23

10001 Completed

im Cove rage

NULL

Table Check:

NULL

NULL

cl0001y2017_20181211

NULL

NULL

24

10001

Completed

hpmsVTypeYear

NULL

Table Check:

NULL

NULL

cl0001y2017_20181211

NULL

NULL

25

10001

Completed

hourVmtFraction

NULL

Table Check:

NULL

NULL

cl0001y2017_20181211

NULL

NULL

26

10001

Completed

fuelSupplyYear

NULL

Table Check:

NULL

NULL

cl0001y2017_20181211

NULL

NULL

27

10001

Completed

fuelSupply

NULL

Table Check:

NULL

NULL

cl0001y2017_20181211

NULL

NULL

28

10001

Completed

fuelFormulation

NULL

Table Check:

NULL

NULL

cl0001y2017_20181211

NULL

NULL

DEDNR QA Report

©

10-

In the example in Figure 4, there is an error with the AvgSpeedDistribution table. The
submission has a distribution that adds to zero (0.0). This entry (i.e., all zeros) can inadvertently
eliminate any VMT associated with that combination and cause an incorrect result. The sum of
the fractions in the AvgSpeedDistribution table must add to 1.0 for every combination of
SourceTypelD, RoadTypelD. and HourDaylD, even if the Source Type, Road Type or
HourDaylD do not have VMT associated with that combination. Never fill distribution tables
with zeros. You can use the EPA-developed distribution if no county-specific values are
available.

Tip: Speeds distributions should be different for different vehicles and road
types (e.g., vehicles likely spend a higher fraction of time in higher speed bins
on rural roads than on urban roads). See Section 6.4 below for further
instructions.

21

-------
The RoadType table has a warning (no rows). Since MOVES will use default values for this
table if no rows are supplied, the warning is provided to inform the user that no data has been
supplied, but MOVES will run successfully without user input.

Tip: Please ensure the RoadType table uses the correct structure.

The "status" category may not contain any "Error" entries to successfully pass QA. The report
will pass if only these entries appear:

1) Completed - this indicates a successful check.

2) Warning - This indicates the user-supplied values that may not have been intended (i.e.,
distributions that sum to zero), but will not cause MOVES to generate inappropriate results.

3) Comment - This will indicate that the table contains no user supplied rows to be checked.

5.3 Create Your CDB Checklist

From the ftp://newftp.epa.eov/air/nei/2020/doc/siipportine data/onroadA download the CDB
checklist, "MOVES Onro; mty Checklist.xlsx". This spreadsheet will contain rows for
every county in the nation. You can trim this list to only include the counties in your state.

This checklist is intended to indicate which tables your agency has revised from EPA's
placeholder values with local data for each county's CDB. Please feel free to add footnotes as
necessary to convey any nuances in the amount of local data in a particular table, for example -
local age distributions for only certain source types. The CDB checklist also has a column to
allow your agency to indicate counties for which you accept EPA default estimates for the
submittal. Note that the starting point CDBs do not contain EPA defaults for the 2020 NEI; these
will be developed later on. The checklist you submit should include all the counties in your state,
even if you are only submitting CDBs for some of the counties. See Figure 5.

-------
AutoSave (• Off) g] ^ v	*7 57 MOVES_Onroad_County_Checklistxlsx - Excel

Godfrey, Janice GJ

P Search

File Home Insert	Page Layout Formulas Data	Review View Help Acrobat i3 Share 5

J ^ CUt	iCalibri ^K> - A" A"	&¦« $ Wrap Text General - |g§ [S [07	§§ [§§ £-AutoSum Ay /0

D?|Copy *	~	0Fill"	'

Paste ~l-J '	«	— — — — — pcj., „ & n/»«-,ooo Conditional Format as Cell Insert Delete Format	Sort & Find &

- ^ Format Painter	BiU.m-0.-A,-	= „!=:!= S Merge & Center - $-%5 48 Fom,atting - Table - Styles - - - - *>Clear- Filter - Select -

Clipboard ft	Font ft	Alignment ft Number ft! Styles	Cells	Editing	Sensitivity '



AU16



























































































A

B

C

D

E

F

G

H

,

J

K

L

M

N

0

P

Q

R

S

T

U

V

w

X

Y

z

AA

AB

AC

AD

AE

AF

AG

AH

Al

AJ

AK

AL

AM

AN

AO

AP

AQ

AR

AS

1

County



Tables











2

stateid

statename

countyid countyname

Accept EPA Default Values

_c?

1

1

£

£
E

1

|

£

a

1

1

£

¦t

1

hotellingmonthadjust

c

0
u

1

E
|

f
|

f

1

t

idlemodelyeargrouping

T3

O

E

!

1

E

1

1

sourcetypeagedistribution

E

1

1

Q.
>
111

sourcetypeyearvmt

C

1
1

C

c
o
t;

E

1

1

E
t

1

1

c









E

|

3

1

ALABAMA

1001 Autauga County





















































4

1 ALABAMA

1003 Baldwin County

















































5

1 ALABAMA

1005 Barbour County

















































6

1 ALABAMA

1007 Bibb County











































7

1

ALABAMA

1009 Blount County













































8

1 ALABAMA

1011 Bullock County

















































9

1 ALABAMA

1013 Butler County





















































10

1 ALABAMA

1015 Calhoun County





















































11

1

ALABAMA

1017 Chambers County















































12

1 ALABAMA

1019 Cherokee County



















































13

1 ALABAMA

1021 Chilton County

















































14

1 ALABAMA

1023 Choctaw County

















































15

1 ALABAMA

1025 Clarke County















































16

1 ALABAMA

1027 Clay County















































17

1 ALABAMA

1029 Cleburne County

















































18

1 ALABAMA

1031 Coffee County







































19

1 ALABAMA

1033 Colbert County















































20

1 ALABAMA

1035 Conecuh County

























































21

1 ALABAMA

1037 Coosa County























































Figure 5. CDB Checklist.

Name your checklist with your agency Program System Code (PSC), such as
"PSC_Checklist.xls (or xlsx)". For example, Delaware's state agency PSC is "DEDNR" and their
checklist would be named "DEDNR_Checklist.xls".

The purpose of the checklist is to provide a method to determine which parts of the state
submission contain new information. This will greatly assist EPA in using this information for
making projections to other calendar years.

Figure 6 shows an example of a complete agency submittal checklist. This S/L went above and
beyond the detail we ask for and it was very useful to our efforts to streamline the submission
and QA processes. We encourage other agencies to use footnotes similar to Figure 6, especially
if there are any tables with partial local data.

23

-------
County

Tables

« £ 9- S

stateid statename

9 CONNECTICUT
9 CONNECTICUT
9 CONNECTICUT
9 CONNECTICUT
9 CONNECTICUT
9 CONNECTICUT
9 CONNECTICUT
9 CONNECTICUT

countyid courrtyname
9001 Fairfield County
9003 Hartford County
9005 Litchfield County
9007 Middlesex County
9009 New Haven County
9011 New London County
9013 Tolland County
9015 Windham County

m m m

m m m
m m m

**Tablesthat can be empty but must be present in EIS submittal

x Tables were created from scratch using local data instead of using data provided in EPA provided default CDBs
m Tables were created from scratch using MOVES2014a default data instead of using data provided in EPA provided default CDBs

Notes

Connecticut left these tables blank but CT believes that MOVES2014b defaults are not appropriate and EPA should use their own calculated

1 defaults. We also did not include the 2014 EPA CDB defaultsthat were provided because, ideally, EPA would calculate an updated 2017
hotellinghours input usingthe latest available information.

This additional table will be ignored forthe 2017 NEI runs, however EPA should use this data for regulated analyses (conformity, etc.) and future

2 year projection analyses. To use this MOVES input EPA would need to rename the table files to "imcoverage" and adjust dates if the MOVES run
is for a calendaryearotherthan 2017.

Connecticut provided a sourcetype age distribution that would be appropriate for a conformity analysis usingthe best available information to

3 us at the time we prepared these inputs for 2017. EPA may update this input with more recent information (e.g. EPA 2017 VIN decode) with the
exception of Motorcycles (11) and School Buses (43) which is based on actual local data decoded from a 2017 CT DMV registration pull.

Figure 6. Complete Agency Submittal Checklist

5.4 Documentation

All submissions must include documentation. At a minimum, the documentation should address
all changes made by the state to the EPA provided CDBs. For each change, the document should
briefly state the source of the state supplied information used to populate the CDB. References to
other documents with more detail are encouraged.

If you created the CDBs from scratch (not editing the EPA provided CDBs), please document
where local data were used and where default data from MOVES were used.

You may include additional documentation files which are referenced by the main
documentation. These additional files can be in any format (e.g., .pdf or .xls).

Name your documentation with your agency Program System Code (PSC), such as
"PSC_ Documentation doc (or docx)". For example, Delaware's state agency PSC is "DEDNR"
and their documentation would be named "DEDNR_Documentation.doc". Any additional files
provided can keep their original names and do not need to conform to any standard.

5.5 Bundle CDB Submittal Components and Create the XML for EIS Submittal

Once you have prepared the parts of your submittal, you'll need to zip them together in a specific
way and reference them with the EIS/CDX required XML file. The CDBs are folders located in
the MariaDB/data directory on your system. Once you have completed creating/editing the
CDBs for your state, these folders and their contents are to be included in the zip file for
submission.

-------
Figure 7 below shows that the individual CDBs are zipped into one zip file. That zip file and the
remaining parts (QA report, checklist, and documentation) are then zipped into a zip folder.

OnRoad CDB Example

I

j, Sample_CDB_Submission.Hip
= Sample_CDB_Submission.xml

—>

V

4.

5arnple_CDB_File.zip

I

I PSC_County_Databases.zip

c CCCCCy202 0_ YYYYMMDD
cCCCCCy20 20_ YYYYMMDD
cCCCCCy2020_ YYYYMMDD

= P5C_QA_Report.txt
— PSC_Checklist.xls (or xlsx)

= PSC_Documentation.doc (ordocx)
*PSC = Program System Code	

Figure 7. OnRoad CDB Zipping Example

To create the XML file Use the EIS Access Bridge Tool

1.	Download the "Nonpoint/Onroad/Nonroad Bridge Tool".

2.	Open the file in Microsoft Access.

3.	Choose the "Export Onroad/NonRoad XML Wrapper" from the Main Menu (Figure 8).

25

-------
=u Main Menu\

Main Menu

Area Emissions Inventory System Bridge Tool

(

Import from CERS XML

y*—

Export To CERS XML

Edit Data Tables

View QA Report

View XML Component Scan Report

Export Onroad/Nonroad XML Wrapper

Figure 8. Main Menu Bridge Tool

4. Fill out the form with the appropriate information Figure 9.

26

-------
25 Main Menu

Main Menu

Area Emissions Inventory

(

Import from CERS XML



Export To CERS XML

Edit Data Tables

View QA Report

View XML Component Scan Report

Export Onroad/Nonroad XML Wrapper

\

-------
Export CERS XML Wrapper for Activity Data

Export CERS XML Wrapper for Activity Data

File Location: lc:\Your_locatiorAFilenaiTie,xml

Browse..,



Complete / Verify Header Information

Data Category:

Onroad |v



Emissions Year:

2020









Submission Type:

Production v



Activity Database Type:

coa p



Activity Database File:

YourZippedFi1eName.zip



User identifier:

[VourEIStD	



Author Name:

Your Name



Organization Name:

Your Organization



Program system code:

lYOUf PSC



EPA Dataset:





Keywords:

Optionally fill



Submission comment

Optionally fill



Begin Export	Cancel

Figure 9. CERS XML Wrapper Example

a.	File Location = The name and location where the resulting XML file will be
generated.

b.	Data Category: Choose "Onroad"

c.	Emissions Year: Set to 2020

d.	Submission Type: Choose either "QA" or "Production"

e.	Activity Database Type: Set to "CDB"

f.	Activity Database File: The name of the packaged set of zip files. Using the
earlier example: "Sample_CDB_Submission.zip"

g.	User Identifier: Your EIS User ID

h.	Author Name: Your name (optional)

i.	Organization Name: The name of your organization (optional)

j. Program System Code: The program system code of your organization
k. EPA Dataset: Leave blank

28

-------
1. Keywords: Any keywords you would like to submit about the document
(optional)

m. Comment: Any comments about the document (optional)

5. Press the "Begin Export" button. Your file will be generated to the location specified.

NOTE: Each CDB submittal is a total replacement to any CDBs previously submitted by that
agency. Thus, if a state submits every county in one submittal, then submits only one county in a
subsequent submittal, only the one county will be present in EIS. You can check the feedback on
your submittal in EIS by choosing your agency and the "Feedback Reports" tab as shown in
Figure 10 below:

Mia 1 _

EIS Gateway

1 'Jonathan Miller; EIS: Content Manager Role. Inventory Developer Role. Authenticated Role, Inventory Selector Role. Account Manager Role

L$.~l feu?

View/Add/Edit

» Facility Inventory and Point
Emissions

» Potential Duplicate Facilities
» Merge Processes
» Nonpoint/ Onroad/ Nonroad
Emissions
» Event Emissions
» NCD Activity Data
» CDB Activity Data
» Inventory Selection
» Schedule Augmentation
» Data Tagging

| Reports |

» Request Reports
» Report Downloads
» Large File Download
» Feedback Reports
» Agency Submission History
Report

RpFFBFNrp Rata

Agency Organization Detail
Current Agency

Agency Description: Alabama Department of Environmental Management

Agency Type: State
ETL Process Group: 2

Agency Responsibilities

Agency Members

Program System Codes

Allow Access

Feedback Reports

Nonpoint Survey

Submission History

Submitter Data Category Type

d438e5d0-r75a-4ca9-8f98-728902dcc17d

COMPLETED

Anna Wood

Point

PRODUCTION

2017/12/27 11:33:34 AM

Download Report

ca820725-6990-477f-ada0-2b246df75fa5

COMPLETED

Anna Wood

Facility Inventory

PRODUCTION

2017/12/27 10:35:20 AM

Download Report

0d6c79ba-37e5-46d7-bb2e-038b2b6cb919

COMPLETED

Anna Wood

Facility Inventory

PRODUCTION

2017/12/22 02:09:42 PM

Download Report

7261d47d-debc-47fa-aade-cb18a651 ece€

COMPLETED

Anna Wood

Facility Inventory

PRODUCTION

2017/12/2211:54:57 AM

Download Report

ed127e6a-d893-41 bc-394d-6327d2bb7d4e

COMPLETED

Anna Wood

Point

PRODUCTION

2016/12/28 02:42:37 PM

Download Report

_29d883S3- 7eed-4ddd-8e2a-326de07bfa95

COMPLETED

Anna Wood

Facility Inventory

PRODUCTION

2016/12/28 11:45:08 AM

Download Report

a t c61 Sfb-5184-4123-a01 c-0f318e64f003

COMPLETED

Anna Wood

Facility Inventory

PRODUCTION

2016/12/27 03:27:55 PM

Download Report

a6c50cb0-446b-4207-a
-------
6.1 Representative Counties

Representative counties are developed based on:

• Their state (counties represent others in the same state),

• Light-duty age distributions,

o Note: representative county CDBs use a population-weighted average of the
member county age distributions.

• Altitude,

• Fuel region, and

• Inspection and maintenance programs.

Figure 11 shows the roughly 300 representative counties used for a recent project. These will be
reassessed for the 2020 NEI.

Reference County Groups MYR

Figure 11. Representative County Map

-------
6.2 Hotelling Hours Per Day

Hotelling Hours refers to the time spent idling by diesel long-haul combination trucks during
federally mandated rest periods of long-haul trips. Submitting agencies have the option to
directly provide MOVES with the number of hotelling hours.

Tip: 'Please ensure that 'HotellingHoursPerDay' and

"HotellingActtvityDistributton" represent the submission year. They may also
be left empty.

6.3 Age Distribution

Age distribution represents the fraction of vehicles by age and source type for calendar year. For
S/L submittals it is important that this distribution be representative of the NEI year. Most states
rely on state vehicle registrations to determine age distribution. The EPA has found that if the
data pull occurs too early in the calendar year, newer model cars will not have had the chance to
infiltrate the vehicle population and the age distribution may be weighted toward older vehicles,
which may result in unrealistically higher mobile emissions. The EPA suggests pulling the
vehicle registrations mid-year (around July 1). EPA expects that agency-submitted age
distributions developed from DMV data may differ from EPA defaults developed from
nationally compiled county-level registration data. Figure 12 shows a comparison between an
S/L submitted age distribution showing relatively fewer newer vehicles (and a larger fraction of
older vehicles) compared to the EPA developed age distributions. EPA will consider all agency
submitted age distributions and accept them as long as they contain realistic features. The EPA
develops age distributions for all counties and for representative counties, where the latter is a
population-weighted average of member counties' age distributions. Long-haul trucks are an
exception to the county group averaging; they reflect national averages because long-haul trucks
frequently travel across many states. Please see for further details on the age

distribution development for the 2017 NEI.

Tip: MOVES3 only contains registration-based age distributions for two
analysis years: 1990 and, 2014. The age distributions for all other analysis
years in MOVES3 were projected forwards or backwards from the 2014 base
age distribution. S/L agencies are encouraged to replace the MOVES3 default
age distributions in the starting point CDBs with local data. Any MOVES3
defaults remaining in the submittals will be replaced with EPA defaults for the
2020 NEI when the data become available.

-------
Source Type 31: Passenger Truck

Age (Years Old)

Figure 12. 2017 Age Distribution Example

Age distribution was found to be a key parameter that impacted mobile emissions in the 2017
NEI and 2016 platform. Many states provided age distributions with issues such as outdated year
for age distribution draw (e.g., 2013), light-duty vehicle recession "dip" shifted by several years
away from the actual recession year of 2009, or unrealistically high fraction of older (or newer)
vehicles. States should also consider using 2020 nationwide EPA defaults for long-haul trucks
because they drive long distances from the counties and even states in which they are registered.
The EPA default long haul truck age distributions will be shared with agencies through the MJO
MOVES workgroup later on in the NEI development process.

The EPA ran a sensitivity analysis on age distribution for the 2017 NEI by substituting the EPA
default age distribution (derived from the CRC A-115 project with adjustments based on
submitted 2017 LD age distributions) in place of S/L submitted age distributions. The EPA
default age distributions were consistent with the 2016vl modeling platform age distributions,
but specific to the year 2017 and included adjustments to remove antiques and older vehicles.
The sensitivity results illuminated the importance of a mid-year determination of the distribution
and having the recession dip in the correct year.

6.4 Hourly Average Speed Distribution

Average speed is the distance traveled (in miles) divided by the time (in hours). The EPA urges
users to develop the most detailed local speed information that is reasonable to obtain. The speed

-------
distributions must vary by road type, hour, and weekday/weekend, and it is desirable that they
also vary by source type (or groups of source types) though vehicle specificity is not always
possible. EPA default speed distribution profiles in the 2017 NEI were derived from the
100 study, which included the use of GPS-based telematics data at the county level with
differences for light-duty personal vehicles and medium/heavy commercial trucks. Compared to
the EPA default speed distributions, some submitted speed distributions look very different than
. Some of the less desirable features in the submitted speeds included:

• Zero time in bin 1 (speeds 0 to 2.5 mph), even on unrestricted roads where vehicles have
to come to a stop at lights or stop signs (surface streets with intersections)

• Speed distributions that were the same for all source types

• No variation in speeds by hour of day or weekday/weekend

• No variation in speeds by road type

Speeds distributions should be different for different vehicles and road types (e.g., vehicles likely
spend a higher fraction of time in higher speed bins on rural roads than on urban roads). If speed
distributions are not realistic, EPA will not accept the submitted data because the EPA default
data likely better represents average speeds in the area. Table 2 displays the speed bins in
MOVES. Speed fractions need to sum to 1 over the 16 bins for each combination of county, road
type, and vehicle type.

Table 2. MOVES Speed Bins

Bin

Midpoint (mph)

Ranae

2.5

speed < 2.5mph

2.5rnph <= speed < 7.5mph

7.5tnph <= speed < 1 2.5mph

1 5

1 2.5rnph <= speed < 1 7.5rnph

1 7.5mph <= speed <22.5mph

22.5mph <= speed < 27.5mph

27.5mph <= speed < 32.5mph

32.5mph <= speed < 37.5mph

37.5mph <= speed < 42.5mph

42.5mph <= speed < 47.5tnph

1 1

4 7.5mph <= speed < 52.5mph

52.5mph <= speed < 57.5mph

57.5mph <= speed < 62.5mph

62.5mph <= speed < 67.5mph

1 5

67.5mph <= speed < 72.5mph

72.5mph <= speed

Figure 13 shows an 8 am speed distribution for an urban non-freeway. The fraction of time spent
in the 16 speed bins sums to 1 for this hour example (and for each hour). Notice that the state
submitted data (purple) has the same profile for all source types, whereas the CRC A-100 data

-------
distinguishes between light duty, medium, and heavy duty. This particular submittal in Figure 13
did a good job in capturing time spent at bin 1 speeds (0 to 2.5 mph). Unfortunately, sometimes
travel demand model (TDM) based approaches will average over links using already-averaged
hourly speeds, and this can miss the lowest speed bins. Emissions are sensitive to speed, and
particularly lower speed

Regarding speed differences by vehicle type, different classes of vehi cles drive the same speeds
on the same road, with perhaps the exception of reduced speed limits for trucks in the rightmost
lane of some freeways. The MOVES road type "Unrestricted Access" includes a broad group of
road types such as arterials, collectors, and small local roads. Analysis of telematics data (CRC
A-100) shows that HDVs tend to have faster overall speeds than passenger cars and trucks on
Unrestricted Access road types, likely because they are traveling more on the faster arterial as
opposed to the residential roads.

-o-d Submittal • Source Type 21
Submittal - Source Type 31
O-a-O Submittal • Source Type 32
A-A-A Submittal • Source Type 41

i Submittal - Source Type 42
Submittal • Source Type 43
Submittal - Source Type 51
• • 1 Submittal ¦ Source Type 52
Submittal ¦ Source Type 53
Submittal - Source Type 54
Submittal ¦ Source Type 61
EH3-Q Submittal • Source Type 62

/proj1/EPA_2017_NEI_CDBs/2017_CDB_Review/Plotting/compare_speeds.sas 18NOV19 22:21

Figure 13. Speed Distribution Example

6.5 VMT Distribution by Road Type

The fraction of VMT by road type varies from area to area and can have a significant effect on
overall emissions from onroad mobile sources. For each source type, the RoadTypeVMTFraction
field in the RoadTypeDi stributi on table stores the fraction of total VMT for each source type that

-------
is traveled on each of the MOVES four road types. Figure 14 illustrates where a few states
submitted non-zero values for VMT in counties that did not have rural restricted road type.
Although this figure is from the 2016 platform, submittals were similar for the 2017 NEI.
Erroneous gap filling should be avoided. Road type VMT should be county-specific and based
on the actual road network within that county. In order to avoid unrealistic vehicle activity and
emissions appearing to be distributed across an entire state, county-specific road type
distributions should be supplied to EPA.

Tip: Please do not provide empty tables for the roadTypeDistribution' tables.

Please ensure that the values in the RoadTypeDistributionx table sum to one
for all source types.

ir-

2016fh (vl) VMT: R2 | Rural Restricted

/ \

r j u

:( V

k' ' • v • :<^T\ . y

_v',

—*

• H

v..

i- f. .

v iSzM

J M<4oj \

> 7.28e+0?
6 7?e+07
6lfce*07
U«Qe+07
Utl4e+07
48€+07
3.92e+07 j
13.36e+07 1
?80e*07
¦I 2 24e+07
TJ68e+07
1.12e 1-07
< S.60e+06

Figure 14. VMT Distribution Example

6.6 Average Speed Distribution

Average speed is used in MOVES to convert VMT inputs into the source hours operating units
that MOVES uses for internal calculations. It is also used to select appropriate driving cycles,
which are then used to calculate exhaust running operating mode distributions. Hourly speed
profiles should vary by weekday versus weekend and by road type. Figure 15 is an example
where the submitted state data held three road types at a constant speed, while variation was
shown with the CRC A-100 data. Distributions should reflect AM and PM congestion/rush hour
slowdowns in counties where such slowdowns exist. Please note that the EPA will not accept
average speed profiles in 2020 that are identical for weekdays and weekends.

-------
County ID = 24039, Maryland: Somerset
URBAN (Salisbury; MD-DE)

Weekday

£" 5Q,

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hour

/proj1/EPA_2017_NEI_CDBs/2017_CDB_Review/Plotting/compare_speeds.sas 12NOV19 11:18

Figure 15. Hourly Average Speed Profile Example
6.7 VMT/VPOP Ratio

We want to avoid too high ratios of VMT to vehicle population (VPOP) (e.g., a rural county with
an interstate highway running through while few vehicles registered there), because in the past
this has triggered MOVES to produce negative emission factors for some emission types that
occur while vehicles are parked. For the 2017 NEI the EPA ranked VMT/VPOP ratios by state
and source type to look at where this was happening and why. The EPA determined a threshold
ratio to calculate more appropriate VPOP to reflect number of vehicles on the roadways, as
opposed to registered in a county, and developed more moderate VMT/VPOP ratios. The
maximum allowable ratios are shown in Table 3. These maximum allowable ratios were derived
by examining the ratios nationwide for all vehicle types and were set to around the 95th

CRC A10O RU Freeway (RT 2) - South Region Average of non-Core Counties inside MSAs.
CRC A10O RU Non-Fwy (RT 3) - South Region Average of non-Core Counties *istde MSAs.
CRC A100 UR Freeway (RT 4) - South Regon Average of non-Core Counties inside MSAs.
CRC A100 UR Nort-Fwy (RT 5) - South Region Average of non-Core Counties inside MSAs.

Submittal - RU Freeway (RT 2)

B-S43 Submittal - RU Non-Fwy (RT 3)
—— Submittal - UR Fieeway (RT 4)
4-*-* Submittal - UR Non-Fwy (RT 5)

-------
percentile of vehicles in that source type. The exception was combination long-haul truck, which
was set to 150,000 miles per vehicle per year.

Table 3. Enforced Maximum Allowable VMT/VPOP Ratios in 2020 NEI

MOVES source type

Source type description

Maximum V MT/V POP ratio
(miles per year)

Motorcycle

7,500

Passenger Car

31,000

Passenger Truck

31,000

Light Commercial Truck

31,000

Other Buses (non-school, non-transit)

130,000

Transit Bus

90,000

School Bus

30,000

Refuse Truck

60,000

Single Unit Short-haul Truck

45,000

Single Unit Long-haul Truck

60,000

Motor Home

7,000

Combination Short-haul Truck

150,000

Combination Long-haul Truck

150,000

States can do this type of analysis on their own. Figure 16 shows the miles/vehicle for
combination long-haul trucks for 2017. This shows where some of the higher ratios occurred and
where vehicle population might need to be further analyzed and possibly increased to be more
representative of the vehicles on the roadways as opposed to just those registered in the county.

Tip: A useful quality check on population and VMT inputs is to divide VMT by
source type by source type population to estimate VMT per vehicle, and then
determine whether these estimates are reasonable.

-------
2017 draft (with state data) onroad VMT/VPOP ratio: Combination Long-Haul Trucks

h 1.80e+05

V 1.60e+05

V 1.40e+05
I 1.20e+05

jy
y

- 1.00e+05 4

in
V

- 8.00e+04

- 6.00e+04

- 4.00e+04

- 2.00e+04

Figure 16. VMT/VPOP Ratio Example Plot

6.8 Hour, Day, and Month VMT Fractions

WIT fraction is the fraction of VMT occurring in each hour of the day, weekday/weekend, and
month on each road type. This fraction will vary from location to location and the EPA
encourages S/L agencies to submit local VMT fractions accordingly. VMT fraction also varies
by vehicle type. Certain vehicles will drive more miles during certain times of the day. For
example, school buses are not likely to be out on the road at night. Figure 17 shows a weekday
example where a state submitted the same hourly VMT fractions for all vehicle types on rural
unrestricted road type, whereas the EPA default data from CRC A-100 showed some variability
by vehicle type.

-------
Rural Non-Freeway (Road Type 3)
Weekday

Hour of Day

.1 0.09

I 0.08

l£ 0.07
0.06

> 0.05
0.04
0.03
0.02
0.01
0.00

CRC A10O - Source Types 1101/11.
CRC A10O - Source Types 32'52.
CRC A100 - Source T,pe» 53.«1.«2
Submittal - Source Type 11
Sjbmitt* - Source Type 21
Submittal Source Type 31
Submittal - Source Type 32
Submittal - Source Type 41

Grouping is Individual.
Grouping is lndvxiuaJ.
Grauprg is lndimjg«l.

Sufcrnfflal
SubmOal
SubmeJal
Subm«jl
SufcmeJil
SiAmCTal
Siitometal

- Source

- Source
¦ Source

- Scurce

Figure 17. VMT Fraction Vehicle Example

In some cases, a county may not have a certain road type and will elect to put 100% of VMT in
hour 1 for road types that do not exist (see Figure 18). There are two better options in this
situation. EPA would prefer that states either substitute a real distribution from a similar road
type in the area or do not update the starting point CDB data for those road types.

-------
Rural Freeway (Road Type 2)
Weekday

8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

Hour of Day

CRC At 00 Source Types 11/21/31. Grouping is South Region Average of Core Counties inside MS As.
CRC A100 - Source Types 3252. Grouping is South Region Average of Cor* Counbes inside MSAs
CRC A100 - Source type* 53AM/62. Grouping is South Region Average of Core Counties iwude MSA*.
1 Submittal ¦ Source Type 11
1 Submittal - Source Type 21
SubmrBaf - 5«wrc« Type 31
1 Submittal Source Type 32
Submittal - Source Type 41
i Submittal - So wee Type 42
Submittal - Source Type 43
i Submittal - Source Type 51

¦ Submittal - Source Type 52

¦ Submittal - Sowee Type 53

¦ Submittal • Source Type 54

¦ Submittal - Source Type $1
! Submittal - Source Type 62

Figure 18. VMT Fraction Non-Existing Road Type Example

Hour VMT fraction should also vary by day type (weekday or weekend day) and vehicle type.
For the 2017 NEI the EPA did not accept weekday=weekend profiles.

6.9 QA Resolution

The EPA resolves data problems by coordinating with S/L agencies individually and/or
presenting intentions during monthly meetings with the multi-jurisdictional organization (MJO)
MOVES workgroup. In some cases, during the 2017 NEI development, the S/L agency preferred
to submit a corrected CDB. In other cases, the agency provided the EPA with instructions for a
spot correction to a table or accepted the EPA's proposed update.

6.10 Considerations specific to 2020

• Due to the effects of COVID-19 on travel, S/Ls should consider submitting monthly
VMT fractions (via the MonthVMTFraction table, see Table 1) in order to represent
the decreased VMT during the spring months when states were shut down. EPA
would also gladly accept month total VMT by source type, HPMS vehicle type, or

see.

-------
7 Data Sources

FHWA Home / Policy & Government Ulairs / Highway Policy Information / Highway
Statistics 2017

CRC A-100

CRC A -115

7.1 Additional Resources

For additional assistance contact:

Submittal Issues -Janice Godfrey; 919-541-3391
Inventory Issues -Janice Godfrey; 919-541-3391
CDB Content Issues -Jaehoon Han; 734-214-4299
MOVES Issues -mobile@epa.gov

-------