Light-Duty Vehicle Greenhouse Gas
Emissions Inventory for Air Quality
Modeling Technical Support Document
United States
Environmental Protection
Agency
-------
Light-Duty Vehicle Greenhouse Gas
Emissions Inventory for Air Quality
Modeling Technical Support Document
Assessment and Standards Division
Office of Transportation and Air Quality
U.S. Environmental Protection Agency
SER&
United States
Environmental Protection
Agency
EPA-420-R-10-011
April 2010
-------
TABLE OF CONTENTS
ACRONYMS Hi
LIST OF TABLES iv
LIST OF APPENDICES iv
1 Introduction 1
2 2005 Emission inventories and their preparation 2
2.1 Custom configuration for emissions modeling for LDGHG 3
2.2 Point sources 6
2.3 2005 Nonpoint sources 10
2.4 2005 Mobile sources 10
3 VOC speciation changes that represent fuel changes 11
4 2030 Reference Case 13
4.1 2030 Reference Case Point sources 13
4.1.1 Part 1: Projecting 2005 to 2030 Reference for ptnonipm 14
4.1.2 Part 2: Additional OTAQ-supplied emissions data for ptnonipm 17
4.2 2030 Reference Case Nonpoint sources 18
4.2.1 Part 1: Projecting 2005 to 2030 Reference case for nonpt 19
4.2.2 Part 2: Additional OTAQ-supplied emissions data for nonpt 19
4.3 Mobile sources 20
4.3.1 US Aircraft, locomotive, and non-c3 commercial marine (alm_no_c3) 20
4.3.2 Canada and Mexico onroad mobile sources (othon) 20
4.3.3 C3 commercial marine sources from all waters (seca_c3) 20
4.3.4 US nonroad mobile sources (nonroad) 22
4.3.5 Onroad mobile sources (on_moves_runpm, on_moves_startpm, and on_noadj) 23
5 2030 Control Case 28
5.1 2030 Control Case Point sources 28
5.2 2030 Control Case Nonpoint sources 29
5.3 2030 Control Case Mobile sources 29
11
-------
ACRONYMS
AEO Annual Energy Outlook
BEIS Biogenic Emission Inventory System
btp Bulk plant terminal-to-pump
C3 Category 3 (commercial marine vessels)
CAMD EPA's Clean Air Markets Division
CAP Criteria Air Pollutant
CARB California Air Resources Board
CEM Continuous Emissions Monitoring
CMAQ Community Multiscale Air Quality
DOE Department of Energy
EO 0% Ethanol gasoline
E10 10% Ethanol gasoline
E85 85% Ethanol gasoline
EISA Energy Independence and Security Act of 2007
EGU Electric Generating Utility
FAA Federal Aviation Administration
FIPS Federal Information Processing Standard
HAP Hazardous Air Pollutant
HDGV Heavy-duty Gasoline Vehicles
IPM Integrated Planning Model
LDGHG Light Duty Greenhouse Gas
LDGT1 Light-duty Gasoline Trucks, 0-6000 pounds gross vehicle weight
LDGT2 Light-duty Gasoline Trucks, 6000-8500 pounds gross vehicle weight
LDGV Light-duty Gasoline Vehicles
MOBILE6 Mobile Source Emission Factor Model, version 6
MOVES Motor Vehicle Emissions Simulator
NEEDS National Electric Energy Database System
NEI National Emission Inventory
NMIM National Mobile Inventory Model
OAQPS EPA's Office of Air Quality Planning and Standards
ORL One Record per Line (a SMOKE input format)
ORNL Oak Ridge National Laboratory
MP Multipollutant
PFC Portable Fuel Container
rtb Refinery-to-bulk terminal
RFS1 Renewable Fuel Standard program
RFS2 Revised annual renewable fuel standard
SMOKE Sparse Matrix Operator Kernel Emissions
SCC Source Category Code
TAF Terminal Area Forecast
VOC Volatile Organic Compound
WRAP Western Regional Air Partnership
in
-------
LIST OF TABLES
Table 1. List of cases run in support of the LDGHG air quality modeling
Table 2. Sectors Used in Emissions Modeling for the LDGHG Platform
Table 3. Comparison of model species in the 2005 v4 platform and the LDGHG platform
Table 4. Description of differences in ancillary data between the LDGHG 2005 case and the 2005 v4
platform
Table 5. Emissions from ethanol plants added to 2005 vl point inventory (tons/year)
Table 6. Quantifying MOVES Diesel Exhaust PM Error for 2005 Base Case
Table 7. Summary of VOC speciation profile approaches by sector across cases
Table 8. Explanation of VOC profile codes listed in Table 7.
Table 9a. Impact on the ptnonipm Sector of Not Applying Base Case Controls to the LDGHG 2030
Reference Case.
Table 9b. Impact on Total Anthropogenic Emissions of Not Applying Base Case Controls to the LDGHG
2030 Reference Case.
Table 10. HAP emission ratios for generation of HAP emissions from criteria emissions for C3 commercial
marine vessels.
Table 11. Components of 2030 Nonroad Sector for Reference and Control Cases.
Table 12. MOVES and NMIM Inventory Components of 2030 Onroad Sectors for Reference and Control
Cases.
LIST OF APPENDICES
APPENDIX A: Equations to adapt pre-speciated diesel emissions from MOVES to air quality modeling
species needed for CMAQ.
APPENDIX B: Inventory Data Files Used for Each LDGHG Modeling Case - SMOKE Input Inventory
Datasets
APPENDIX C: Ancillary Data Files Used for LDGHG 2005 Case Compared to 2005 v4 Platform Data Files
APPENDIX D: Growth and Control Assumptions and Affected Pollutants for the 2030 Reference Case
IV
-------
1 Introduction
This document provides the details of emissions data processing done in support of the Environmental
Protection Agency (EPA) and National Highway Traffic Safety Administration (NHTSA) joint rulemaking
effort to establish Light Duty Vehicle Greenhouse Gas (LDGHG) Emissions Standards and Corporate
Average Fuel Economy (CAFE) Standards. The emissions and modeling effort is hereafter referred to as
LDGHG and consists of three emissions cases. Table 1 provides of list of the emissions cases created for
this modeling effort.
Table 1. List of cases run in support of the LDGHG air quality modeling
Case Name
Internal EPA
Abbreviation
Description
2005 basecase
2005cp_tox
2005 case done with average year fires data and an average
year temporal allocation approach for Electrical Generating
Units (EGUs), to use for computing relative response factors
with 2030 scenarios
2030 Reference case
2030cp_tox
2030 "baseline" scenario representing the best estimate for the
future year without implementation of national CC>2 emissions
standards.
2030 Control case
2030cp_tox_ldghg
2030 "control" case scenario representing implementation of
national CC>2 emissions standards for light-duty vehicles.
These standards will require these vehicles to meet an
estimated combined average emissions level of 250
grams/mile of CO2 in model year 2016.
The data used in the 2005 emissions cases are often the same as those described in the 2005-based, v4
platform document (http://www.epa.gov/ttn/chief/emch/index.htmltf2005). The LDGHG cases use some
different emissions data than the official v4 platform for two reasons. First, the LDGHG Standard was
evaluated in comparison to the modeling performed for the Revised annual Renewable Fuel Standard
(RFS2); therefore, RFS2-specific inputs were retained for LDGHG. The 2005 RFS2 modeling was
performed from December 2008 through March of 2009, prior to the completion of the 2005 v4 platform.
Second, the LDGHG modeling used data intended only for the rule development and not for general use. All
of the documentation provided here describes what was done differently and specifically for the LDGHG
effort in contrast to what is used in the v4 platform.
In LDGHG, we used a 2005 base case approach for the year 2005 emissions scenario. While there is not
documentation on the approach specific to 2005, this approach is similar to that in OAQPS's 2002-based v3
platform (http://www.epa.gov/ttn/chief/emch/index.htmltf2002). A base case approach uses average year
fires and EGU temporal profiles from 3 years of EGU data. We use a base case approach because we want
to reduce year-specific variability in these components of the inventory: for example large fires vary in
location and day of the year each year, and EGU shutdowns and high use on high energy demand days also
vary by year. By using a base case approach, these two aspects of the inventory are maintained in future
year modeling and therefore do not introduce potentially spurious year-specific artifacts in air quality
modeling estimates. In addition, the same biogenic emissions data as the v4 platform was used not only in
the 2005 case for LDGHG, but also in both future-year cases run for LDGHG. For LDGHG, the only
significant data changes between the 2005 and future-year cases are emission inventories and speciation
approaches.
-------
For this effort, we have created and provided state, county, and source category code (SCC) emission
summaries for the nonpoint and mobile sectors. For point sources, we have posted the actual inventory
datasets that we used for emissions modeling. These data have been provided to the EPA docket for this
rule. In addition, the data can be found associated with the "LDGHG 2005 and 2030 emissions data" link on
the Clearinghouse for Inventories and Emissions Factors (CHIEF) website at
http://www.epa.gov/ttn/chief/emch/index.htmltf2005.
In the remainder of this document, we provide a description of the approaches taken for the emissions in
support of air quality modeling for LDGHG. In Section 2, we describe the ancillary data and 2005 inventory
differences from the v4 platform. In Section 3, we describe the speciation differences among each of the
cases run for LDGHG. In Section 4, we describe the 2030 Reference case as compared to the 2005 base
case, and in Section 5, we describe the 2030 Control Case in comparison to the 2030 Reference case.
2 2005 Emission inventories and their preparation
As mentioned previously, the 2005 emissions modeling approach for LDGHG used much of the same data
and approaches as the 2005 v4 platform. In this section, we identify the differences between the data used
for LDGHG and that used for the 2005 v4 platform. Section 2.1 provides ancillary data differences that
impact multiple sectors and Sections 2.2 through 2.4 provides differences for the point, area, and mobile
sectors.
Table 2 below lists the platform sectors used for the LDGHG modeling platform. It also indicates which
platform sectors include HAP emissions and the associated sectors from the National Emission Inventory
(NEI). Subsequent sections refer to these platform sectors for identifying the emissions differences between
the v4 platform and the LDGHG platform.
Table 2. Sectors Used in Emissions Modeling for the LDGHG Platform
Platform Sector
IPM sector: ptipm
Non-IPM sector:
ptnonipm
Average-fire sector:
avefire
Agricultural sector:
ag
Area fugitive dust
sector: afdust
Remaining nonpoint
sector: nonpt
Nonroad sector:
nonroad
Aircraft,
locomotive, marine:
aim no c3
C3 commercial
marine: seca c3
2005 NEI
Sector
Point
Point+
N/A
Nonpoint
Nonpoint
Nonpoint+
Mobile:
Nonroad
Mobile:
Nonroad
Mobile:
nonroad
Description
NEI EGU units at facilities mapped to the IPM model using the
National Electric Energy Database System (NEEDS) database.
All NEI point source units not matched to the ptipm sector.
Average-year wildfire and prescribed fire emissions derived from
the 2001 Platform avefire sector, county and annual resolution.
NH3 emissions from NEI nonpoint livestock and fertilizer
application.
PM10 and PM2 5 emissions from fugitive dust sources in the NEI
nonpoint inventory.
All nonpoint sources not otherwise included in other emissions
modeling sectors.
Nonroad emissions from National Mobile Inventory Model
(NMIM) using NONROAD2005, other than for California, for
which emissions submitted by the California Air Resources Board
(CARB) were used. CARB data used for HAPs are annual,
allocated to monthly using NMIM, while other data are monthly.
Aircraft, locomotive, commercial marine except for category 3
(C3) commercial marine vessels
C3 commercial marine vessels
Contains HAP
emissions?
Yes
Yes
Yes
No
No
Yes
Yes
Yes
Yes
-------
Platform Sector
Onroad, except
gasoline PM:
on_noadj
Onroad starting
exhaust PM:
on moves startpm
Onroad running
exhaust PM
on moves runpm
Biogenic: biog
Other point sources
not from the NEI:
othpt
Other nonpoint and
nonroad not from
the NEI: othar
Other onroad
sources not from the
NEI: othon
2005 NEI
Sector
Mobile:
onroad+
Mobile:
onroad+
Mobile:
onroad+
N/A
N/A
N/A
N/A
Description
A combination of onroad mobile sources from MOBILE6,
MOVES, and 2005 NEI v2 data from California. MOVES-based
data used for:
1) onroad diesel exhaust PM, CO, NOx, VOC, some VOC HAPs
2) onroad gasoline exhaust CO and NOx
3) onroad gasoline evaporative and exhaust VOC, and some VOC
HAPs.
More details are provided in the 2005 v4 platform documentation.
MOVES-based onroad mobile start gasoline exhaust PM data.
More details provided in the 2005 v4 platform documentation.
MOVES-based onroad mobile running gasoline exhaust PM data.
More details provided in the 2005 v4 platform documentation.
Hour-specific emissions generated from the Biogenic Emission
Inventory System (BEIS), version 3.14 model (includes emissions
in Canada and Mexico) run with the Sparse Matrix Operator
Kernel Emissions (SMOKE) modeling system.
Point sources from Canada's 2000 inventory, Mexico's 1999
inventory, and off-shore point sources from the 2001 platform
Canada 2000 and Mexico 1999 nonpoint and nonroad mobile
inventories
Canada 2000 and Mexico 1999 onroad mobile inventories
Contains HAP
emissions?
Yes
No
No
No
No
No
No
Some data included in modeling sector has been revised beyond what is included in the 2005 NEI vl or v2.
As with the 2005 v4 platform, we processed all emissions data with a custom version of the Sparse Matrix
Operator Kernel Emissions (SMOKE) modeling system, version 2.5. Users seeking to replicate modeling
done for this effort can use version 2.6 of SMOKE. More details about SMOKE including user
documentation are available at its website (http://www.smoke-model.org).
2.1 Custom configuration for emissions modeling for LDGHG
Unlike the 2005 v4 platform, the configuration for LDGHG modeling included additional hazardous air
pollutants (HAPs) and used slightly older ancillary data. Both of these differences are described in this
section.
Table 3 lists the additional HAP pollutants processed for the LDGHG platform, which were not included in
the 2005 v4 platform. However, since using the full multipollutant HAP version of the Community
Multiscale Air Quality (CMAQ) model would have taken longer than the time available for our project, we
used a "lite" version of the multipollutant CMAQ that required emissions only for the species flagged in the
third column of Table 3. Additional model species that appear in model-ready data files are listed in the
right two columns of the table, but we did not run these additional HAPs through CMAQ for this effort.
-------
Table 3. Comparison of model species in the 2005 v4 platform and the LDGHG platform
Description
Carbon Monoxide
Nitrogen Oxide
Nitrogen Dioxide
Nitrous acid
Ammonia
Sulfur dioxide
Sulfuric acid vapor
PM2 5 Elemental carbon
PM25 Organic carbon
PM2 5 primary nitrate
PM2 5 primary sulfate
PM25 other
PM Coarse (PM10-PM2 5)
Acetaldehyde
Higher aldehydes
Ethene
Ethane
Ethanol
Formaldehyde
Internal olefin carbon bond
Isoprene
Methanol
Nonreactive VOC
Nonvolatile (from VOC mass)
Terminal olefin carbon bond
Parrafin carbon bond
Toluene and other monoalkyl
aromatics
Unknown VOC
Unreactive VOC
Xylene and other polyalkyl
aromatics
Sequiterpenes
Terpene
Benzene
Chlorine
Hydrochloric acid
Divalent gaseous mercury
Elemental mercury
Paniculate mercury
Naphthalene from the HAP
inventory
Acrolein from the HAP
inventory
Acetaldehyde from the HAP
inventory
1,3 Butadiene from the HAP
inventory
2005 v4
platform
species
CO
NO
NO2
HONO
NH3
S02
SULF
PEC
POC
PNO3
PSO4
PMFINE
PMC
ALD2
ALDX
ETH
ETHA
ETOH
FORM
IOLE
ISOP
MEOH
NR
NVOL
OLE
PAR
TOL
UNK
UNR
XYL
SESQ
TERP
BENZENE
CL2
HCL
HGIIGAS
HGNRVA
PHGI
LDGHG platform
Species in CMAQ
MPlite
CO
NO
NO2
HONO
NH3
S02
SULF
PEC
POC
PNO3
PSO4
PMFINE
PMC
ALD2
ALDX
ETH
ETHA
ETOH
FORM
IOLE
ISOP
MEOH
NR
NVOL
OLE
PAR
TOL
UNK
UNR
XYL
SESQ
TERP
BENZENE
CL2
HCL
HGIIGAS
HGNRVA
PHGI
NAPHTHALENE
ACROLEIN
ALD2_PRIMARY
BUTADIENE13
Additional LDGHG
platform HAP
Species*
ACRYLONITRILE
BR2 C2 12
CARBONTET
CHCL3
CL_ETHE
CL2 C2 12
CL2 ME
CL3_ETHE
CL4_ETHANE1122
CL4 ETHE
DICHLOROBENZE
NE
DICHLOROPROPE
NE
ETOX
HEXAMETHY_DIIS
HYDRAZINE
MAL_ANHYDRIDE
PROPDICHLORIDE
QUINOLINE
TOL_DIIS
TRIETHYLAMINE
DIESEL PEC
DIESEL POC
DIESEL PMFINE
DIESEL_PNO3
DIESEL_PMC
DIESEL PSO4
BERYLLIUM_C
BERYLLIUM F
CADMIUM_C
CADMIUM_F
CHROMHEX_C
CHROMHEX F
CHROMTRI C
CHROMTRI_F
LEAD_C
LEAD F
MANGANESE C
MANGANESE F
NICKEL_C
NICKEL_F
Description
Acrylonitrile
1,2 Dibromoethane
Carbontet
Chloroform
Vinyl Chloride
1,2 Dichloroethane
Methylene Chloride
Trichloroethylene
1,1,2,2 Tetrachloroethane
Perchloroethylene
Dichlorobenzene
Dichloropropene
Ethylene Oxide
Hexamethylene 1,6-
Diisocyanate
Hydrazine
Maleic Anhydride
Propdichloride
Quinoline
2,4-Toluene Diisocyanate
Triethylamine
Diesel PM2 5 elemental carbon
Diesel PM2 5 organic carbon
Diesel PM2 5 primary nitrate
Diesel PM2 5 primary sulfate
Diesel PM2 5 other
Diesel coarse PM
Coarse Beryllium
Fine Beryllium
Coarse Cadmium
Fine Cadmium
Coarse Hexavalent Chromium
Fine Hexavalent Chromium
Coarse Trivalent Chromium
Fine Trivalent Chromium
Coarse Lead
Fine Lead
Coarse Manganese
Fine Manganese
Coarse Nickel
Fine Nickel
-------
Description
Methane from the HAP
inventory
Formaldehyde from the HAP
inventory
M-isomer of Xylene from the
HAP inventory
O-isomer of Xylene from the
HAP inventory
P-isomer of Xylene from the
HAP inventory
Toluene from the HAP
inventory
2005 v4 LDGHG platform
platform Species in CMAQ
species MP lite
CH4
FORM_PRIMARY
MXYL
OXYL
PXYL
TOLU
Additional LDGHG
platform HAP
Species* Description
* These species are created by the emissions configuration, but were not modeled.
In addition to the model species differences, the LDGHG platform had a few additional custom aspects in the
2005 cases. Table 4 lists the datasets used by the LDGHG platform that are different from the v4 platform,
including a description of the impact of the differences. These differences stem from the 2005 v4 platform
having been done after the RFS2 platform, which was used as the basis for the LDGHG platform, resulting
in newer inventory data used for the v4 platform. These inventory differences are described more in later
sections of this document. In addition, Appendix B provides a more detailed comparison of the ancillary
datasets for the 2005 v4 platform versus the LDGHG platform.
Another consideration is the speciation across the LDGHG future-year cases as compared to 2005. Section 3
provides a detailed account of these differences. Otherwise, the future-year ancillary data were largely the
same as those in 2005, with no substantial differences. All ancillary data files can be found at the 2005-
based platform website (http://www.epa.gov/ttn/chief/emch/index.htmltf2005).
-------
Table 4. Description of differences in ancillary data between the LDGHG 2005 case
and the 2005 v4 platform
Ancillary Data Type
Spatial cross references
Temporal cross-
references
Temporal profiles
Speciation cross-
references and
Speciation profiles
Inventory tables
NonHAP exclusions
data
Difference between 2005 v4 platform and LDGHG platform
The 2005 v4 platform data are updated to support the newer (2006) Canadian
inventories. The LDGHG files contain references to diesel exhaust start
emissions from MOVES, that were not available for the v4 platform.
The 2005 v4 platform data are updated to support the 2006 Canadian data,
additional source category codes (SCCs) found in the 2005 v2 NEI point
inventory, and a revised oil and gas inventory. The LDGHG files contain
references to diesel exhaust start emissions from MOVES.
The 2005 v4 platform dataset adds additional profiles from Environment
Canada to support processing of the 2006 Canadian inventory.
The LDGHG data files are configured to support the multi-pollutant (MP)
version of CMAQ, whereas the 2005 v4 platform data file is configured to
support only the non-MP version. Therefore, the LDGHG data files include
profiles for additional HAP species, including HAP VOCs, HAP metals,
chromium, and diesel PM. The 2005 v4 platform data files include profiles for
passing through the pre-speciated VOCs for the 2006 Canadian inventory.
Furthermore, new headspace vapor VOC speciation profiles were used based on
new test data for EO and E10 fuels for LD GHG.
The LDGHG data file is configured to support the MP version of CMAQ,
whereas the 2005 v4 platform data file is configured to support only the non-
MP version.
The 2005 v4 platform data has been updated with new oil and gas SCCs not
used for the LDGHG platform.
2.2 Point sources
The 2005 emissions from the U.S. point source sectors (ptipm and ptnonipm) used for LDGHG differ from
the v4 platform primarily because the emissions are based on the 2005 NEI version 1, rather than the 2005
NEI version 2. While the NEI version 2 was available in time for this work, we intentionally remained
consistent with the RFS2 modeling approach that had just been completed. The original emissions were
created from the NEI database on June 10, 2008. These emissions were further modified, further changing
the 2005 NEI version 1, as follows:
1) Inventory split into ptipm and ptnonipm sectors using the IPM column of the SMOKE-ready datasets
2) Applied fugitive dust transport fractions to the appropriate SCCs (SCC list available at
http://www.epa.gov/ttn/chief/emch/dustfractions)
3) Further analyzed HC1 emissions for missing ptipm sector HC1 sources and moved emissions originally
labeled as ptnonipm to correctly be placed with their associated criteria air pollutants (CAP) emissions
in the ptipm sector
4) Removed emissions of benzene, acetaldehyde, formaldehyde, methanol, to support the "no hap use"
approach for the ptipm and ptnonipm sectors. This approach creates emissions of these HAPs from
VOC speciation profiles rather than the HAP emissions inventory, because the CAP and HAP emissions
were not consistent enough to be able to use the HAP inventory for VOC speciation.
5) Added an additional 47 ethanol plants. Unless otherwise noted in Table 5 (a list of ethanol facilities and
their emissions), we adjusted the original VOC emissions provided by a factor of 0.65 (a reduction) to
offset the ethanol-heavy speciation profile (99.6% ethanol) that we had available for speciation of
ethanol plants. This prevented the overstating of ethanol emissions. In addition, we multiplied all
-------
emissions by 17(453.6 x 2000) x 1,000,000= 1.1023113109 (converting from grams/yr per million
gallons to tons/yr) to correct the units of the emissions (from grams per million gallon per year to
tons/year). These changes resulted in roughly a 10.23% increase in the emissions values initially
provided by OTAQ. The list of these facilities and their emissions is available in Table 5.
6) Removed facilities that closed between 2002 and 2005, since much of the ptnonipm portion of the 2005
vl NEI was carried forward from the 2002 v3 NEI. This change was also made in creating the v4
platform. The list of removed sources is available in the Excel® file
"Closures_applied_to_2002v3_point_for_2005af.xls" provided with the 2005 v4 platform
documentation.
7) Removed Minnesota airport ground support equipment (SCCs 2265008005 and 2270008005), to prevent
double counting with the nonroad sector.
8) Other minor adjustments:
o Removed exclamation marks, asterisks, and embedded double quotes in facility names, and
other key text fields, changed "PM25" and "PM25-PRI" to "PM2_5", and changed the
state/county Federal Information Processing Standard (FIPS) code field for tribal records from
00000 to 88TTT, where TTT is the tribal code.
o Removed two SCC=201002 records because this is an invalid SCC; both records had zero
emissions
To implement the inventory processing, we split the 2005 ptnonipm CAP and HAP inventories into five
separate datasets to facilitate replacement and projection for the 2030 reference and control scenarios. These
datasets are:
1) A dataset with one ethanol plant (Chippewa) CAP and HAP emissions for which the emissions are
held at 2005 values for the 2030 LDGHG reference and control scenarios;
2) A dataset with three ethanol plants' CAP and HAP emissions that are replaced in both the 2030
LDGHG reference and control scenarios;
3) A dataset with CAP and HAP emissions for 43 additional ethanol plants not available in the 2005
vl NEI, which also have different emissions in the 2030 reference and control scenarios;
4) A dataset with CAP emissions from the all other nonEGU sources from the 2005 NEI vl
5) A dataset with HAP emissions from the all other nonEGU sources from the 2005 NEI vl
In addition to differences in the U.S. point sources, further differences exist from the v4 platform for the
Canadian point emissions. The Mexico point emissions are identical to those documented for the 2002 v3
platform and the 2005 v4 platform. The LD GHG modeling inventories included year 2000 Canadian
emissions from the 2002 v3 platform and did not include the updated 2006 Canadian emissions, to be
consistent with the modeling done for RFS2. We did not model mercury and therefore did not use Canadian
mercury emissions data. The offshore sources were different from what was used in the 2005 v4 platform
because they were not updated with the new offshore inventory available in the 2005 v2 NEI. For more
information on the Canadian and Mexican emissions used for this effort, please refer to the 2002 v3 platform
documentation at http://www.epa.gov/ttn/chief/emch/index.htmltf2002.
In addition, we processed emissions for LDGHG using the 3-d emissions option for all point source sectors
rather than the "inline" point source option that we used for the 2005 v4 platform. This approach has
essentially no effect on the modeling results. Using the inline approach makes the CMAQ emissions data
files smaller, but that option was not available for multi-pollutant CMAQ in time for use on this effort.
-------
Table 5. Emissions from ethanol plants added to 2005 vl point inventory (tons/year)
County Name
Swift Co
Chippewa Co
Dodge Co
Sibley Co
Dawson Co
Stevens Co
Wichita Co
Sedgwick Co
Roosevelt Co
Linn Co
Clinton Co
Platte Co
Lyon Co
Peoria Co
Pembina Co
Tazewell Co
Monroe Co
Des Moines Co
Washington Co
Christian Co
Codington Co
Kossuth Co
Riverside Co
Cerro Gordo
Co
Hardin Co
Brown Co
Beadle Co
Pierce Co
Kearney Co
Stearns Co
Crawford Co
State Name
Minnesota
Minnesota
Minnesota
Minnesota
Nebraska
Minnesota
Kansas
Kansas
New Mexico
Iowa
Iowa
Nebraska
Minnesota
Illinois
North
Dakota
Illinois
Wisconsin
Iowa
Nebraska
Kentucky
South
Dakota
Iowa
California
Iowa
Iowa
South
Dakota
South
Dakota
Nebraska
Nebraska
Minnesota
Illinois
St/Co
FIPS
27151
27023
27039
27143
31047
27149
20203
20173
35041
19113
19045
31141
27083
17143
38067
17179
55081
19057
31177
21047
46029
19109
06065
19033
19083
46013
46005
31139
31099
27145
17033
PLANT
Chippewa Valley Ethanol Co LLLP (*)
Granite Falls WWTP (+)
Al-Corn Clean Fuel (+)
Heartland Corn Products (+)
Cornhusker Energy Lexington (CEL)
Diversified Energy Company (DENCO),
LLC
ESE Alcohol Inc.
Abengoa Bioenergy Corporation
Abengoa Bioenergy Corporation
Archer Daniels Midland (ADM)
Archer Daniels Midland (ADM)
Archer Daniels Midland (ADM)
Archer Daniels Midland (ADM)
Archer Daniels Midland (ADM)
Archer Daniels Midland (ADM)
Aventine Renewable Energy, Inc. (formerl
Badger State Ethanol, LLC
Big River Resources, LLC
Cargill (was PGLA-1CO)
Commonwealth Agri-Energy, LLC
Glacial Lakes Energy, LLC (OLE)
Global Ethanol
Golden Cheese Company of CA
Golden Grain Energy LLC
Hawkeye Renewables, LLC
Heartland Grain Fuels, LP
Heartland Grain Fuels, LP
HuskerAg,LLC
KAAPA Ethanol, LLC
Land O' Lakes / Melrose Dairy Proteins
Lincolnland Agri-Energy
CO
29.7
0.0
28.5
51.4
10.5
52.4
3.1
52.4
62.8
964.7
733.2
366.6
154.4
439.8
52.4
385.9
108.9
108.9
328.0
41.9
104.7
209.4
10.5
146.6
115.2
20.9
29.3
54.5
125.7
5.4
58.6
NOX
40.4
0.0
45.0
62.9
30.3
151.6
9.1
151.6
181.9
1,338.4
1,017.2
508.6
214.1
1,273.1
151.6
535.4
315.3
315.3
455.1
121.3
303.1
606.3
30.3
424.4
333.4
60.6
84.9
157.6
363.8
15.8
169.8
voc
155.1
0.7
207.6
378.3
14.3
71.6
4.3
71.6
86.0
418.1
317.7
158.9
66.9
601.9
71.6
167.2
149.0
149.0
142.1
57.3
143.3
286.6
14.3
200.6
157.6
28.7
40.1
74.5
172.0
7.5
80.2
SO2
0.6
0.0
0.5
0.4
38.6
192.9
11.6
192.9
231.5
1,266.3
962.4
481.2
202.6
1,620.4
192.9
506.5
401.2
401.2
430.5
154.3
385.8
771.6
38.6
540.1
424.4
77.2
108.0
200.6
463.0
20.1
216.0
PM25
19.7
0.0
0.0
36.7
1.1
6.5
0.4
5.5
6.6
546.7
415.5
207.8
87.5
365.5
43.5
218.7
13.4
13.4
185.9
5.2
12.9
25.8
1.3
15.3
14.2
2.2
3.1
5.7
15.5
0.7
7.2
PM10
28.4
0.0
24.0
61.6
12.1
60.6
3.6
60.6
72.8
1,248.9
949.2
474.6
199.8
509.3
60.6
499.6
126.1
126.1
424.6
48.5
121.3
242.5
12.1
169.8
133.4
24.3
34.0
63.1
145.5
6.3
67.9
ACROLEIN
0.0001
0.0000
0.0618
0.0000
0.0557
0.2783
0.0166
0.2783
0.3339
2.7827
2.1148
1.0574
0.4452
2.3375
0.2783
1.1131
0.5788
0.5788
0.9458
0.2226
0.5567
1.1131
0.0557
0.7793
0.6122
0.1113
0.1558
0.2894
0.6678
0.0290
0.3119
-------
County Name
Cherokee Co
Jefferson Co
Tazewell Co
Lincoln Co
Hamilton Co
Roberts Co
Delaware Co
Ida Co
Finney Co
Sioux Co
Red Willow Co
Loudon Co
Hitchcock Co
Winnebago Co
Gove Co
Mitchell Co
State Name
Iowa
Colorado
Illinois
Nebraska
Nebraska
South
Dakota
Iowa
Iowa
Kansas
Iowa
Nebraska
Tennessee
Nebraska
Wisconsin
Kansas
Georgia
St/Co
FIPS
19035
08059
17179
31111
31081
46109
19055
19093
20055
19167
31145
47105
31087
55139
20063
13205
PLANT
Little Sioux Corn Processors
Merrick & Company (Coors Brewery)
MGP Ingredients, Inc.
Midwest Renewable Energy, LLC (MRE)
Nebraska Energy
North Country Ethanol (NCE)
Permeate Refining
Quad-County Corn Processors
Reeve Agri-Energy
Siouxland Energy & Livestock Coop
(SELC)
SW Energy, LLC.
Tate & Lyle
Trenton Agri-Products, LLC. (TAP)
Utica Energy, LLC
Western Plains Energy, LLC
Wind Gap Farms (Anheuser/Miller
Brewery)
Total
CO
184.3
4.2
188.5
41.9
104.7
52.4
3.1
56.5
25.1
46.1
0.2
254.7
83.8
108.9
62.8
0.8
6,074.3
NOX
533.5
12.1
545.6
121.3
303.1
151.6
9.1
163.7
72.8
133.4
0.6
353.3
242.5
315.3
181.9
2.4
12,610.4
voc
252.2
5.7
257.9
57.3
143.3
71.6
4.3
77.4
34.4
63.1
0.3
110.4
114.6
149.0
86.0
1.1
5,923.2
S02
679.0
15.4
694.4
154.3
385.8
192.9
11.6
208.3
92.6
169.8
0.8
334.3
308.6
401.2
231.5
3.1
14,417.5
PM25
19.3
0.5
23.2
5.2
10.9
6.5
0.4
7.0
3.1
4.8
0.0
144.3
10.3
11.4
6.6
0.1
2,537.1
PM10
213.4
4.9
218.3
48.5
121.3
60.6
3.6
65.5
29.1
53.4
0.2
329.7
97.0
126.1
72.8
1.0
7,456.6
ACROLEIN
0.9795
0.0223
1.0020
0.2226
0.5565
0.2783
0.0166
0.3005
0.1336
0.2449
0.0011
0.7346
0.4452
0.5788
0.3339
0.0045
24.0
(*) Data taken from the 2002 NEI
(+) Data taken from the 2005 NEI
data by 0.65.
rather than provided by OTAQ for 2005. Units conversion not performed on this facility and no adjustment of the emissions by 0.65.
vl, rather than provided by OTAQ for 2005. Units conversion not performance on these facilities and no adjustments of the emissions
-------
2.3 2005 Nonpoint sources
The emissions from the agricultural ammonia (ag) and nonpoint fugitive dust (afdust) sectors are
the same as the v4 platform. For the "other" nonpoint (nonpt) sector, the only difference from
the v4 platform is that these emissions do not include the oil and gas extraction emissions (SCCs
matching 23100XXXXX) provided by the Western Regional Air Partnership (WRAP) for the
western states. These updated oil and gas extraction emissions were provided after the modeling
platform development for RFS2 and LDGHG.
For the Canadian and Mexican nonpoint sector (othar), we used the same inventories as the 2002
v3 platform, and did not yet have the updated 2006 Canadian inventory used in the v4 platform.
The avefire emissions are the same as those in the 2005ci_tox_05b case mentioned above, with
the exception that we added 1-3-butadiene, acrolein, and xylenes, and toluene using ratios to
PM2 5 available in the 2005 platform documentation. These factors were also used to create
emissions for benzene, acetaldehyde, and formaldehyde in the 2005 v4 platform and were
unchanged for LDGHG; however, this sector was processed as "no-integrate"; therefore,
emissions for benzene, acetaldehyde, and formaldehyde were obtained from speciated VOC.
2.4 2005 Mobile sources
Mobile sources include three US onroad sectors (on_noadj, on_moves_startpm,
on_moves_runpm) and three US nonroad sectors (nonroad, alm_no_c3, and seca_c3). In
addition, it includes Canadian and Mexican emissions in a separate onroad sector (othon) and
nonroad/nonpoint sector (othar).
For onroad mobile, the on_moves_startpm and on_moves_runpm emissions inventory data are
from an updated draft version of the Motor Vehicle Emission Simulator (MOVES) compared to
the 2005 v4 platform. The LDGHG emissions also keep additional pollutants as described in
Section 2.1. In addition for these MOVES sectors, the temperature adjustment calculations
applied to PM2.5 species were the same as the v4 platform, since these adjustments had not
changed between draft MOVES versions.
The on_noadj sector for LDGHG also uses an updated version of MOVES, and this updated
version also provides onroad diesel exhaust emissions for CO, NOx, PM, and VOC HAPs. Note
that PM emissions from these diesel sources are not subject to temperature adjustments like the
on_moves_startpm and on_moves_runpm sectors. These MOVES-based onroad diesel
emissions replace the MOBILE6-based NMIM emissions in the 2005v4 platform.
For onroad gasoline exhaust PM emissions, the allocation of MOVES PM2.5 emissions to
SMOKE-ready format PM species is documented in Appendix B of the 2005v4 Platform
Documentation. However, LDGHG MOVES data also provides diesel exhaust PM2.5
emissions, and these equations are provided in this document in Appendix A of this document.
A small processing error impacted the MOVES diesel PM exhaust emissions used for this
modeling. The fraction of metals in the speciation profile (Fmetaj) in equation 4 in Appendix A
10
-------
was inadvertently squared, resulting in an underestimate in metals, which caused a very small
underestimate in inventory organic carbon (POC) and a corresponding equal overestimate in
inventory PMFINE (other PM species not already classified). Total PM2.5 was not impacted.
This error impacts all LDGHG scenarios, and the small impact of this error is shown in the
national totals provided in Table 6.
Table 6. National summary of MOVES Diesel Exhaust PM Error for 2005 Base Case
Pollutant
PEC (Elemental Carbon)
PSO4 (PM Sulfate)
PNO3 (PM Nitrate)
metal_bad (Eq 4 in Appendix A)
metal (Eq 4 in Appendix A)
POC_bad (Organic Carbon)
POC (Organic Carbon)
PMFINE bad (Other PM)
PMFINE (Other PM)
PMC (PM-Coarse, orPMlO minus PM2.5)
PM2 5 total bad
PM2_5_total
HDDV
90,734
6,812
134
1
313
45,315
45,054
11,661
11,921
13,300
154,656
154,656
LDDV/LDDT
3,107
154
6
0
14
1,300
1,288
319
331
420
4,886
4,886
HDDV
Error
260
-260
LDDV/LDDT
Error
12
-12
The nonroad emissions inventory data are the same as the v4 platform, with additional HAPs
being kept for LDGHG as well. The alm_no_c3 emission sector does use different data from
that of the v4 platform. Specifically, the aircraft emissions remain in this sector and are older
data from that of the v4 platform. In the v4 platform, the aircraft emissions had been revised to
be consistent with the 2005 NEI v2 and included in the ptnonipm sector. The airport emissions
used in LDGHG were from the 2002 NEI, version 3, acquired March 27, 2007 and used in the
2002 v3.1 platform.
Additionally, the onroad emissions for Canada and Mexico for LDGHG (othon sector) differ
from the v4 platform because we used the older Canadian data. The data we used reflect 2000
emissions and are the same data used in the v3 and v3.1 2002-based platforms. The 2005 v4
platform uses 2006 Canadian inventory data.
3 VOC speciation changes that represent fuel changes
A significant detail that changes for each of the LDGHG modeling cases is the VOC speciation
profiles used to split total VOC emissions into the various VOC model species needed for
CMAQ. In this section, we summarize the various speciation profile information used in
configuring the various cases, and we include Table 7 to provide a summary of the VOC
speciation approach for each of the future-year cases.
The approaches taken in the LD GHG 2005 case below are not the same as the 2005 v4 platform,
and so the first 2005 column shown represents the 2005 v4 platform and the second 2005 column
shown is for LD GHG. Two new headspace vapor profiles were created for LD GHG modeling
that were not available for the 2005 v4 platform development. The approaches used for each of
11
-------
the future-year cases are customized for those cases, and they include the impact of fuel changes
for each of the future-year cases on emissions from the on_noadj sector, the nonroad sector, and
parts of the nonpt and ptnonipm sectors. The speciation changes from fuels in the nonpt sector
include changes for portable fuel containers (PFCs) and some parts of the bulk-plant-to-pump
(btp) and refinery-to-bulk terminal (rbt) emissions. The speciation changes from fuels in the
ptnonipm sector include the remainder of the emissions for the btp and rbt emissions. Mapping
of fuel distribution SCCs to btp and rbt emissions categories can be found in Appendix A of the
RFS2 Emissions Inventory for Air Quality Modeling Technical Support Document (EPA Report
No. 420-R-10-005, January 2010, http://www.epa.gov/otaq/renewablefuels/420rl0005.pdf).
A general indication of the VOC speciation approach is provided in Table 7.
Table 7. Summary of VOC speciation profile approaches by sector across cases
Inventory
type and
mode
Mobile
Exhaust
Mobile
Evaporative
Other
sources:
nonroad
refueling,
PFCs,
btp,
rbt
VOC speciation approach
for fuels
Tier 1 EO and E10 combinations
Tier 1 EO or E10 by county
Tier 2 EO or E10 by county
Tier 1 E10
Tier2E10
EO and E10 combinations
EO or E10 by county
E10
EO headspace (old)
EO headspace (new)
EO headspace (new) or E10
headspace (new) by county
E10 headspace (new)
VOC
Profile
Codes
8750
8751
8750
8751
8756
8767
8751
8757
8753
8754
8753
8754
8754
8737
8762
8762
8763
8763
2005 v4
platform
on noadj
nonroad
on_noadj
nonroad
All listed
2005
LDGHG
case
on noadj
nonroad
on_noadj
nonroad
All listed
2030
Reference
Case
nonroad
on_noadj
on noadj
nonroad
All listed
2030
Control
Case
nonroad
on noadj
on_no_adj
nonroad
All listed
Table 8 provides the purpose of the VOC speciation profile codes used in the table:
12
-------
Table 8. Explanation of VOC profile codes listed in Table 7
Exhaust
8750 Tierl EO
8751 Tierl E10
8756 Tier2 EO
8757 Tier2E10
Evaporative
8753 EO
8754 E10
Refueling
8737 EO headspace vapor (old)
8762 EO headspace vapor
8763 E10 headspace vapor
Appendix C summarizes the data file names used for all of the data files that are updated from
the v4 platform. All ancillary data files are available on the 2005-based platform website
previously referenced.
4 2030 Reference Case
The 2030 Reference case is intended to represent the emissions associated with use of the most
likely volume of ethanol in the absence of the LDGHG CC>2 reductions and RFS2 rule and
Energy Independence and Security Act of 2007 (EISA) renewable fuel requirements. For this
case, the ethanol volume was projected for 2030 using the Department of Energy, Energy
Information Administration in the 2007 Annual Energy Outlook (AEO) report. That year's AEO
projections were used because Department of Energy (DOE) started accounting for EISA in their
2008 AEO projections. Therefore, the 2030 LDGHG Reference scenario shares several of the
same emissions (especially for stationary sources) as the 2022 RFS-2 AEO case (please see the
RFS-2 TSD, available at http://www.epa.gov/otaq/renewablefuels/420rl0005.pdf). A list of
inventory datasets used for this and all cases is provided in Appendix C. A list of all growth and
control assumptions for this Case is provided in Appendix D. Section 5 describes the projection
differences between the 2030 Reference and 2030 Control cases.
4.1 2030 Reference Case Point sources
The point sources for the 2030 Reference case included US EGU point sources (ptipm), US
nonEGU point sources (ptnonipm) and sources from Mexico, Canada, and the Gulf of Mexico
(othpt). The US EGU point sources for both the reference and control 2030 cases use an
Integrated Planning Model (IPM) run for criteria pollutants, HCl, and mercury in 2020 (though
Hg was not modeled). We used 2020 because it was the closest readily available year to the
2030 year used for LD GHG modeling. With very few exceptions, we have used year 2020 as
the furthest "out year" for stationary source projections. Also, by using many of the same
emissions inputs as RFS-2 that are not impacted by the LDGHG control strategy, we could better
isolate the impacts of LDGHG-related controls versus RFS-2 controls and obtain a more 'apples
to apples' comparisons of air quality and emissions-related impacts between RFS-2 and
LDGHG.
The code number used by EPA's Clean Air Markets Division (CAMD) to denote the run is
EPA30Draft_BC_42l. While these IPM emissions are different from those used in the v4
platform, they are consistent with the 2020 emissions used in the v3 platform. OAQPS post-
processed these data in the same way as described in the 2005 v4 platform documentation for the
"base case" to create daily emissions that include temporal allocation information from three
years of Continuous Emissions Monitoring (CEM) data. The temporal allocation approach is the
13
-------
same as for the LDGHG 2005 base case (2005cp_tox_05b), to eliminate artificial differences in
temporal allocation between the base and future years.
For the Mexican and offshore point source emissions, we held the data constant with the 2005
base case. This means that the Mexican emissions were based on year 1999, since Mexico has
not provided emissions projections to date. We used 2020 emissions projections for Canadian
emissions, provided by Environment Canada, which are consistent with the base year Canadian
data. The Canadian data are the same as those used for the 2002-based v3 platform. The data
used for LDGHG future year reference and control cases are different from our v4 platform.
For the US nonEGU emissions (sector ptnonipm), there were two main pieces for the 2030
LDGHG ptnonipm inventories: data projected from 2005 values and data that were replaced by
OTAQ-generated data that are intended to represent emissions estimates for any year beyond
2016. Referring to the five parts listed in Section 2.2, datasets (4) and (5) were projected from
2005 to 2030 values, datasets (2) and (3) were replaced, and dataset (1) was unchanged.
4.1.1 Part 1: Projecting 2005 to 2030 Reference for ptnonipm
We applied both control and growth factors to a subset of the 2005 ptnonipm to create the 2030
Reference data. We started with 2020 projection factors from the 2002 v3.1 platform for most of
the LDGHG year 2030 projections. Given the uncertainty of emissions projections and the lack
of any additional controls coming into effect between 2020 and 2030, we decided that using
2020 as the year for most of our non-EGU point and nonpoint projections would be sufficient;
exceptions to this are described below. This approach matched the year used for the ptipm
sector. Furthermore, we did not have to adjust the factors for a 2005-base year in most cases,
because most of the 2005 vl nonEGU point emissions data were populated with 2002 emissions
values.
The 2002 v3.1 platform growth and control factors had been revised from the 2002v3.0 platform
projection factors that are documented in the 2002 v3 emissions modeling platform
documentation (see http://www.epa.gov/ttn/chief/emch/index.html#2002). These updates
included Hazardous Waste Incineration adjustments and Small and Large Municipal Waste
Combustor closures and adjustments. The following describes how we further modified the
2002v3.1 projection factors for the 2030 Reference case.
We intended to use a SMOKE "control" packet (data file) to apply control factors that implement
known emissions reductions from point sources for national rules. For the Reference case, the
control packet was intended to be the same as the revised control packet from the 2002 v3.1
factors to give key VOC HAPs the same control factors as those for VOC. The VOC HAPs (and
CAS numbers) of interest for this effort were: 1,3-butadiene (CAS=106990), acrolein
(CAS=107028), formaldehyde (CAS=50000), methanol (CAS=67561), benzene (CAS=71432),
acetaldehyde (CAS=75070), and naphthalene (CAS=91203). However, the "control" packet
inadvertently was not applied for either the 2030 Reference or 2030 Control cases. This
emissions processing error results in the same amount of missed reductions from both the 2030
Reference and Control Cases. These "missed" reductions, and the percent of the ptnonipm
inventory that is under-controlled are provided in Table 9a. Appendix D documents which
14
-------
control and projection programs were erroneously not included in the 2030 Reference and
Control cases. No other sectors were impacted by this error.
15
-------
Table 9a. Impact on the ptnonipm Sector of Not Applying Base Case Controls to the LDGHG
2030 Reference Case
Pollutant
CO
NH3
NOX
PM10
PM2 5
SO2
VOC
2030
Reference
Case:
Corrected
3,194,382
153,039
2,205,555
586,770
356,113
2,102,054
1,165,031
2030
Reference
Case:
Modeled
3,206,284
157,947
2,408,024
613,850
372,081
2,290,193
1,414,238
Missing
Reductions
11,902
4,907
202,469
27,080
15,969
188,139
249,208
Percent of
ptnonipm Inventory
Erroneously Not
Reduced
0.4%
3.1%
8.4%
4.4%
4.3%
8.2%
17.6%
While these fractions are somewhat large for only the ptnonipm sector, the fractions when
compared to the entire anthropogenic inventory are much smaller. Table 9b shows the impact
across all anthropogenic sectors, which is closer to what the air quality model uses. The VOC
impact is further lessened in the air quality model because of the biogenic VOC emissions, which
are much larger than the anthropogenic in most cases. In addition, since the discrepancy
occurred consistently in both the 2030 Reference and Control cases, the impact on the affect of
the control strategy is further limited.
Table 9b. Impact on Total Anthropogenic Emissions of Not Applying Base Case Controls to the
LDGHG 2030 Reference Case
Pollutant
CO
NH3
NOX
PM10
PM2 5
SO2
VOC
2030
Reference
Case:
Corrected
53,960,990
4,258,901
11,516,138
12,720,907
4,031,794
9,170,404
12,093,860
2030
Reference
Case:
Modeled
53,972,892
4,263,808
11,718,607
12,747,988
4,047,763
9,358,543
12,343,067
Missing
Reductions
11,902
4,907
202,469
27,080
15,969
188,139
249,208
Percent of
Total Anthropogenic
Inventory
Erroneously Not
Reduced
0.0%
0.1%
1.7%
0.2%
0.4%
2.0%
2.0%
We also used SMOKE with a "projection" packet to apply growth adjustments, some control
adjustments, and plant closures. For LDGHG, the projection packet for ptnonipm consisted of
several components. First, we modified the data from the 2002v3.1 platform packet to remove
entries for which we had new factors for this effort, such as onroad refueling and aircraft. In
addition, we added the same list of VOC HAPs to the projection factors as we had added to the
control packet for the VOC-specific entries already present (e.g., landfills in nonpt sector). Next,
16
-------
we added the LDGHG Reference case-specific projection factors from several OTAQ sources, as
follows:
1. Onroad refueling: We retained the same list of refueling ptnonipm and nonpt
state/county FIPS codes and SCCs from the 2002 v3 emissions modeling platform. For
California, we used the 2005v2 California-submitted data and 2005 NMIM and 2030
NMIM reference case VOC and VOC-HAP refueling emissions to obtain annual
adjustment ratios by pollutant and county. The formula for the projection factors was
Factor 2030 = NMIM Emis203o /NMIM Emis2oos (note: we applied different refueling
adjustment factors to the Reference and Control cases because of different onroad
gasoline refueling inventories in the 2 cases. These factors were applied to the refueling
SCCs from the 2005 base case.
2. Year 2025 aircraft Federal Aviation Administration (FAA) takeoff and landing data from
the Terminal Area Forecast (TAP) data: This is the furthest out-year of FAA TAP data
activity and replaces the 2020 factors for both LDGHG 2030 scenarios.
3. We applied adjustments to refinery emissions by state and SCC, using the same factor for
all pollutants to represent activity adjustments (note: different adjustments were made to
the control case). The state-level adjustments and the list of SCCs affected were
provided by Rich Cook on 11/25/2008 in the Excel® workbook
"RFS2_Refmery_Adjust.xls"; this is the same set of factors used in the RFS-2 AEO
(baseline) scenario. The LDGHG control scenario applies an adjustment on top of these
LDGHG reference case emissions.
4. For gasoline distribution SCCs (both ptnonipm and nonpt SCCs), we additionally applied
VOC and VOC HAP adjustments to SCCs representing emissions from bulk-plant-to-
pump (btp) and refinery-to-bulk terminal (rtb) processes. These SCC-level adjustments
impact VOC and VOC HAPs in both the ptnonipm and nonpt sectors. The adjustments
were provided by OTAQ on 12/16/2008 in the Excel® workbook
"2005ai_tox_SCC_50state_CAPHAP-20081216.xls". Different adjustments were
applied to the LDGHG control scenario, based on assumed changes in the ratios of EO to
E10. The LDGHG control scenario applies an adjustment on top of these LDGHG
reference case emissions.
We applied the control and projection factors to the ptnonipm 2005 inventory described in
Section 2.2, resulting in a "2030" ptnonipm inventory. To this inventory, we added additional
OTAQ-supplied inventory data as described in Step 2 below. The configuration of the
processing was designed specifically to prevent inadvertent double counting of emissions.
4.1.2 Part 2: Additional OTAQ-supplied emissions data for ptnonipm
In addition to the data preparation described above, we added the following data supplied by
OTAQ:
17
-------
1. Ethanol plant additions: These emissions data completely replace all of the ethanol
plants listed in Table 5 except the first one. OTAQ provided the emissions, stack
parameters, locations and the same SCC assignments were made as described for 2005:
PM2.5 (30299999) and VOC and all other pollutants (30125010). The original VOC
emissions were supplied by OTAQ in Excel® and OAQPS further adjusted the data to
correct for units conversions and to adjust VOC downward to account for too much
ethanol in the speciation profile. These were the same adjustments made to the 2005 data
and are described in Section 2.2. Further, we changed the original data to use the same
Stack IDs for both PM2.5 and PMio, to ensure SMOKE could properly calculate PMC as
PMi0-PM2 5 after matching emissions sources. Lastly, we removed the benzene,
formaldehyde, acetaldehyde, and methanol inventory emissions from these sources
because they are "no integrate" sources for VOC HAPs (we do not use the VOC HAPs to
augment speciation information for this sector), and so these model species were to be
calculated from VOC speciation.
2. In addition, we retained the Chippewa plant from 2005 (the emissions at this plant were
only changed for an emissions scenario not related to LDGHG modeling).
4.2 2030 Reference Case Nonpoint sources
The nonpoint sources are sources aggregated to the county (or Canadian Province) and process
level and included US fugitive dust sources (afdust sector), US agricultural NH3 (ag sector), US
average fires (avefire sector), other US nonpoint sources (nonpt sector), and emissions for
Canada and Mexico (othar and othar_hg). Of these, the nonpt sector contained the most detailed
changes for the RFS2 modeling, as explained further below.
The emissions used for this case for the afdust and ag sectors are the same as those used in 2020
in the 2002-based v3 platform, previously referenced. Additionally, the avefire emissions are the
same as those in the 2005cp_tox_05b case mentioned above, and were intentionally held
constant between the 2005 base case and the 2030 Reference and Control cases.
The emissions for Canada and Mexico were the same as the 2002 v3 platform referenced earlier,
with one minor exception: we modified the Canada nonroad sources to remove C3 commercial
marine SCC (2280003010), which was just one record in British Columbia and which prevented
double-counting with the seca_c3 sector. This means that the LDGHG platform differs from the
v4 platform in that older estimates of 2020 emissions were used for nonpoint sources in Canada,
rather than 2006 estimates retained between the base and future years in the v4 platform. This
sector includes both nonroad mobile and nonpoint sources.
The nonpt sector is comprised of several different pieces. Like the ptnonipm sector, the nonpt
sector required two steps to compile the 2030 Reference case inventories: data projected from
2005 values and data that are replaced by OTAQ-generated data in 2030.
We discovered a very minor error in our creation of 2030 Reference and Control case nonpt
emissions. We inadvertently dropped NEI nonpoint inventory oil and gasoline production
emissions in WRAP states (excluding California) from both the 2030 Reference and Control
18
-------
cases. This error resulted in 1 ton of CO and 31 tons of NOx dropped in New Mexico, and a
cumulative 2,184 tons of VOC dropped in Arizona, Montana, Nevada, New Mexico, North
Dakota, South Dakota, and Wyoming.
4.2.1 Part 1: Projecting 2005 to 2030 Reference case for nonpt
The same steps taken and described for the ptnonipm sector (Section 4.1.1) were also taken for
the nonpt sector. In fact, all of the information that we added to the SMOKE projection and
control packets has already been described in that section and the parts that also applied to the
nonpt sector have already been identified in that section. Unlike the ptnonipm sector, we did not
miss applying controls to this sector because all of the nonpoint sector reductions are done using
a different packet than the missed control packet described for ptnonipm.
In addition to the documentation provided previously for ptnonipm, the 2020 projection factors
for Residential Wood Combustion SCCs in this nonpt sector were not updated for 2030 because
the difference would have been very small and these emissions are highly uncertain.
The refueling and gasoline distribution projection approaches described in the point source
section also apply to the nonpoint sector, since those approaches affect SCCs in both sectors.
4.2.2 Part 2: Additional OTAQ-supplied emissions data for nonpt
In addition to the nonpt emissions projected from 2005 to 2030, several additional OTAQ-
provided emission inventories were created to complete the emissions needed for the 2030
Reference case nonpt sector:
1. Ethanol plant additions: These are conceptually the same as the ptnonipm plant additions
but are for plants with unknown coordinates, and are therefore aggregated to the county -
SCC level. SCC assignments determine spatial allocation using spatial surrogate
ancillary data. The nonpoint ethanol plants were created from an Excel® spreadsheet
provided by OTAQ on 11/17/2008: Corn_EtOH_Plant_Inv_2022-aeo.xls. The same
VOC adjustments and SCC assignments detailed for ptnonipm were made here as well.
These additional emissions are expected to be implemented by year 2022 and are held at
the same level for year 2030 LDGHG modeling.
2. Ethanol transfer additions: These new VOC emissions account for vapor loss during
transport and loading by truck and rail are assigned three SCCs (30205031, 30205052,
and 30205053). Ethanol transfer emissions were provided for an RFS-2 AEO scenario on
12/11/2008 by OTAQ in an Excel® workbook called "EtOH_transport_vapor_AEO.xls".
These additional emissions are expected to be implemented by year 2022 and are held at
the same level for year 2030 LDGHG modeling.
3. Biodiesel plant additions: These new emissions are assigned SCC 2102006001 and also
contain an units correction (grams/year to tons/year) of 1.102311309E-06, applied by
OAQPS after receiving the data from OTAQ. These data were created from an Excel®
spreadsheet provided by OTAQ on 11/21/2008: Biodsl_Plant_Inv_2022-aeo-prelim.xls.
We selected the SCC 2102006001 (Stationary Source Fuel Combustion;Industrial;Natural
19
-------
Gas;All Boiler Types) to represent these sources in consultation with Ron Ryan. The
VOC was not adjusted by 65% like it was for ethanol and cellulosic plants. OAQPS
converted emissions from grams to tons. These additional emissions are expected to be
implemented by year 2022 and are held at the same level for year 2030 LDGHG
modeling.
4. Portable Fuel Containers: These MSAT-based emissions were provided by OTAQ for
the year 2030 for the 2002v3 platform and are unchanged for the 2030 LDGHG
Reference case. We applied adjustment factors (provided by OTAQ) to the LDGHG
control case.
4.3 Mobile sources
The mobile sources included many different approaches, depending on the modeling sector.
Each of these is described in a separate subsection below.
4.3.1 US Aircraft, locomotive, and non-c3 commercial marine
(alm_no_c3)
This sector was projected from the base year (2005) to the LDGHG future year (2030); the base
year emissions are actually year 2002 and the same as those used in the 2002v3 platform. There
are two components for projecting these emissions to year 2030:
1. FAA TAP data for year 2025. This data was updated in December 2007 and, as
discussed in Section 4.1.1, the furthest out year of 2025 is used to approximate year 2030
emissions.
2. Updated locomotives and C1/C2 CMV factors. OTAQ provided updated loco-marine
Final Rulemaking (FRM) national-based adjustment factors for locomotive-marine
emissions on 12/17/2009: "alm_no_c3_2030LDGHG_vs_2005_pcQA_locofix.xls".
We applied HAP factors to VOC to obtain 1,3-butadiene, acetaldehyde, acrolein, benzene, and
formaldehyde. The remaining HAPs -metals and other non-VOC HAPs not already provided-
are held at base year levels (the 2002 emissions estimates used in the 2005 basecase).
The aim no c3 emissions are the same for the 2030 Reference and Control cases.
4.3.2 Canada and Mexico onroad mobile sources (othon)
The data in this sector are the same as 2020 data released in 2002-based v3 platform, previously
referenced. These are different from 2005 v4 platform because the 2006 Canadian data were not
available for use in this project.
4.3.3 C3 commercial marine sources from all waters (seca_c3)
The seca_c3 sector emissions data were provided by OTAQ in an ASCII raster format used since
the ECA-IMO project began in 2005. The (S)ECA C3 year 2030 base case from the ECA/EVIO
project was unchanged and used for both the 2030 LDGHG Reference and 2030 LDGHG
20
-------
Control cases. The 2030 base seca_c3 inventory was provided by OTAQ 03/27/2008:
taskl l_2030base.zip. OTAQ also provided factors to compute HAP emission (based on
emissions ratios) on 2/28/2008, which OAQPS applied to either VOC or PM2.5 to obtain HAP
emissions values. Table 10 below shows these factors and whether they were applied to VOC or
PM2.5. As with the 2005 case, this sector uses CAP-HAP VOC integration.
Table 10. HAP emission ratios for generation of HAP emissions from
criteria emissions for C3 commercial marine vessels
Pollutant
Acetaldehyde
Benzene
Formaldehyde
Benz[a]Anthracene
Benzo[a]Pyrene
Benzo[b]Fluoranthene
Benzo[k]Fluoranthene
Chrysene
lndeno[1 ,2,3-c,d]Pyrene
Acenaphthene
Acenaphthylene
Anthracene
Benzo[g,h,i,]Perylene
Fluoranthene
Fluorene
Naphthalene
Phenanthrene
Pyrene
Beryllium
Cadmium
Chromium VI
Chromium III
Lead
Manganese
Nickel
Selenium
Apply to
VOC
VOC
VOC
PM2_5
PM2_5
PM2_5
PM2_5
PM2_5
PM2_5
PM2_5
PM2_5
PM2_5
PM2_5
PM2_5
PM2_5
PM2_5
PM2_5
PM2_5
PM10
PM10
PM10
PM10
PM10
PM10
PM10
PM10
Pollutant Code
75070
71432
50000
56553
50328
205992
207089
218019
193395
83329
208968
120127
191242
206440
86737
91203
85018
129000
7440417
7440439
18540299
16065831
7439921
7439965
7440020
7782492
Factor
0.0002286
9.795E-06
0.0015672
5.674E-07
1 .844E-07
1 .56E-07
1 .56E-07
9.929E-08
1.418E-08
3.404E-07
5.248E-07
5.248E-07
1.277E-07
3.12E-07
6.95E-07
1.987E-05
7.943E-07
5.532E-07
5.459E-07
7.642E-06
2.948E-06
1.343E-05
3.002E-05
5.732E-05
0.0016377
1.337E-05
OAQPS converted emissions to SMOKE point source ORL format allowing for emissions to be
allocated to modeling layers above the surface layer. OAQPS corrected emissions for one
state/county FIPS code fix in Rhode Island. All non-US emissions (i.e., in waters considered
outside of US territory) are simply assigned a dummy state/county FIPS code=98001. Due the
huge size of these data, the CAP emissions are in one ORL file and the HAP emissions for
benzene, acetaldehyde, and formaldehyde are in a separate ORL file; all other HAPs are not
provided for LDGHG as they are uninvolved in air quality model chemistry and are not of
interest for this rulemaking. The emissions spatial extent includes waters off of the coasts of the
US, Canada, and Mexico, as well as emissions in major waterways and the Great Lakes. The
21
-------
SMOKE-ready data have also been cropped from the original data provided by OTAQ to cover
only the 36km CMAQ domain, which is the largest domain used for this effort.
4.3.4 US nonroad mobile sources (nonroad)
All states except California:
OTAQ provided several runs of NMIM emissions that were blended together to create the 2030
Reference and Control case nonroad sector emissions. Table 11 shows how the various NMIM
runs were combined to create the reference and control case non-California nonroad mobile
inventories. The first component "2002v3-based 2030 Base Case" is from the 2030 Base case in
our 2002v3 platform for the SCCs listed in Table 11, and these emissions are the same in the
reference and control cases. OTAQ also provided diesel recreational marine (pleasure craft)
emissions in November 2009 that are used in both the reference and control cases. An NMIM
run specific to the Reference case "LdGhgN2030Aeo.txt" was provided in November 2009 for
the remainder of the nonroad mobile sector: gasoline engine and recreational marine sources.
The only difference in the control case is a substitution of the gasoline engine and recreational
marine emissions with the "E10" NMIM run: "LdGhgN2030el0.txt".
Table 11. Components of 2030 Nonroad Sector for Reference and Control Cases
NMIM file
2002v3-based 2030 Base Case
LdGhgN2030eO_nponzsegl ldies.txt
LdGhgN2030Aeo.txt
LdGhgN2030el0.txt
SCCs
2267x
2268x
2270x
2285002015,
2285006015
22820200x
2260x
2265x
228200x,
22820 Ix
2260x
2265x
228200x,
22820 Ix
Description of Nonroad
SCCs
LPG equipment
CNG equipment
Diesel engines
Railway maintenance
Diesel recreational-
marine
2-stroke gasoline engines
4-stroke gasoline engines
Gasoline recreational
marine
2-stroke gasoline engines
4-stroke gasoline engines
Gasoline recreational
marine
Use
Both Reference and Control
Cases
Both Reference and Control
Cases
Reference Case only
Control Case only
All NMIM data are based on AEO2007 fuels and NMIM county database NCD20080727. We
also reassigned NMIM evaporative and refueling xylene (compound XYL or
CAS=EVP_1330207, RFL_1330207) into MXYL (CAS=EVP_108383, RFL_108383) and
OXYL (CAS=EVP_95476, RFL_95476) using a 68% and 32% ratio to both evaporative and
refueling XYL, respectively. We also split NMEVI exhaust xylene (CAS=EXH 1330207) into
MXYL (CAS=EXH_108383) and OXYL (CAS=EXH_95476) using a 74% and 26% ratio to
XYL, respectively. We converted emissions from monthly totals to monthly average-day values
based the on number of days in each month. CO2 and all of California emissions were removed
prior to creating SMOKE ORL files.
22
-------
California nonroad:
California monthly nonroad emissions are year 2030 and are based on March 2007 California Air
Resources Board (CARB) data (Martin Johnson: mjohnson@arb.ca.gov). NH3 emissions are
from NMIM runs for California (same data as were used in 2030 from the 2002 v3 platform).
We allocated refueling emissions to the gasoline equipment types based on evaporative mode
VOC emissions from the 2002 v3 platform 2030 NMEVI data, and the refueling emissions were
computed by multiplying SCC 2505000120 emissions by 0.61, to adjust to remove double
counting with Portable Fuel Container inventory for California. We estimated HAP emissions
by applying HAP-to-CAP ratios computed from the California data provided for the 2005 NEI
v2, collected by EPA on 12/2007. This was done because the CARB submittal from March 2007
did not include estimates for HAPs. We retained only those HAPs that are also estimated by
NMEVI for nonroad mobile sources; all other HAPs were dropped.
4.3.5 Onroad mobile sources (on_moves_runpm, on_moves_startpm,
and on_noadj)
As in 2005, the on_moves_runpm and on_moves_startpm sectors include emissions from onroad
gasoline sources for PM, which need temperature adjustment factors. The temperature
adjustment factors were specific to 2030 (different from those used in 2005) and we used the
same adjustment factors in both of the 2030 cases. The temperature adjustments have the
limitation that they were based on the use of MOVES default inputs rather than county-specific
inputs, because a county-specific database for input to MOVES was not available at the time this
approach was needed. Further, the version of MOVES used for all runs was a draft version and
has since changed.
Also like 2005, the on_noadj sector includes non-PM MOVES data for gasoline and diesel
vehicles for some pollutants, NMIM-based data for motorcycles and the remaining pollutants for
onroad gasoline and diesel, as well as all California onroad mobile emissions. The detailed
approaches described here are the same as those for 2005 (except for the NMIM and MOVES
data used), but are included here for convenience.
Table 12 shows the various pieces of onroad mobile NMIM and MOVES-based emissions
inventories that were used to create the 2030 Reference and Control case onroad mobile sectors:
on_moves_startpm, on_moves_runpm, and on_noadj. Unless otherwise noted, the emissions are
part of the "no-adjust" on_noadj sector. As Table 12 shows, several pieces of NMIM and the
state-level MOVES data used in this project were different between the 2030 Reference and
Control cases; the only exceptions are the NMEVI onroad diesel vehicles and the adjustment
factors (used in the on_moves_startpm and on_moves_runpm) that were the same for each case.
23
-------
Table 12. MOVES and NMIM Inventory Components of 2030 Onroad Sectors for Reference
and Control Cases
MOVES and NMIM files
LdGhgO2030el0.txt
(NMIM)
LdGhgO2030Aeo.txt
(NMIM)
LgrDiesExh2030_8 12_Ref.txt
(MOVES exhaust)
LgrDiesExh2030_67_Ref.txt
(MOVES exhaust)
LgrDiesExh2030_67_Ctltxt
(MOVES exhaust)
LgrGasExh2030aeoR2_Ref.txt
(MOVES exhaust)
LgrGasExh2030elOR2 Ctl.txt
(MOVES exhaust)
LgrGasPm2030aeo Ref.txt
(MOVES exhaust)
LgrGasPm2030elO Ctl.txt
(MOVES exhaust)
LgrEvapAeoR2_Ref.txt
(MOVES evaporative)
LgrEvapE10R3_Ctl.txt
(MOVES evaporative)
SCCs
223x
ALL
220x
223007x
2230001x
223006x
2230001x
223006x
220 lOx
except
220108x
220 lOx
except
220108x
220 lOx
except
220108x
220 lOx
except
220108x
220 lOx
except
220108x
220 lOx
except
220108x
Description of
Onroad SCCs
All diesel vehicles
All Onroad Emissions
All gasoline vehicles
and trucks
Diesel vehicles classes
8-12 (HDDV)
Diesel vehicles class 6
(LDDV)
Diesel vehicles class 7
(LDDT)
Diesel vehicles class 6
(LDDV)
Diesel vehicles class 7
(LDDT)
Non-motorcycle
gasoline vehicles and
trucks
Non-motorcycle
gasoline vehicles and
trucks
Non-motorcycle
gasoline vehicles and
trucks
Non-motorcycle
gasoline vehicles and
trucks
Non-motorcycle
gasoline vehicles and
trucks
Non-motorcycle
gasoline vehicles and
trucks
Use
Reference Case: Used only to
1) apportion diesel vehicle emissions
to county -level for MOVES
pollutants,
2) as-is for all non-MOVES pollutants
Control Case: Used to
1) apportion emissions to county-level
for all MOVES pollutants,
2) as-is for all non-MOVES pollutants
Reference Case only: Used for
1) all motorcycle emissions,
2) as-is for all non-MOVES pollutants,
3) apportioning MOVES gasoline
vehicle emissions to county -level
Both Reference and Control Cases: PM,
CO, NOx, VOC, acetaldehyde, benzene,
formaldehyde, acrolein, 1,3 -butadiene
Reference Case only: PM, CO, NOx,
VOC, acetaldehyde, benzene,
formaldehyde, acrolein, 1,3 -butadiene
Control Case only: PM, CO, NOx,
VOC, acetaldehyde, benzene,
formaldehyde, acrolein, 1,3 -butadiene
Reference Case only: CO, NOx, VOC,
acetaldehyde, benzene, formaldehyde,
acrolein, 1,3 -butadiene
Control Case only: CO, NOx, VOC,
acetaldehyde, benzene, formaldehyde,
acrolein, 1,3 -butadiene
Reference Case only: PM-only, subject
to temperature adjustments
(on_moves_startpm and
on moves runpm sectors)
Control Case only: PM-only, subject to
temperature adjustments
(on_moves_startpm and
on moves runpm sectors)
Reference Case only: VOC and benzene
Control Case only: VOC and benzene
on_moves_startpm and on_moves_runpm
For the on_moves_runpm and on_moves_startpm sectors, the same preprocessing as was done in
2005 was done here, but using a 2030 Reference case NMIM runs to create the monthly county -
to-state ratios by state and SCC and using the 2030 PM adjustment factors.
24
-------
OTAQ supplied four input files for the reference and control cases that contain state-level
MOVES-based onroad gasoline and diesel emissions by month for the following pollutants (NEI
pollutant IDs are provided in parentheses):
1. Onroad Gasoline Exhaust: VOC, NOX, CO, 1,3-butadiene (106990), acetaldehyde
(75070), acrolein (107028), benzene (71432), and formaldehyde (50000);
2. Onroad Diesel Exhaust: VOC, NOX, CO, 1,3-butadiene (106990), acetaldehyde (75070),
acrolein (107028), benzene (71432), formaldehyde (50000), and MOVES-defined PM
species OC, EC, and SO4, components of PM2.5 -PM25OC, PM25EC, and PM25SO4
respectively;
3. Evaporative: Non-refueling VOC and benzene (71432) -gasoline vehicles only;
4. Onroad Gasoline MOVES-speciated PM at 72°F: MOVES-defined PM species OC, EC,
and SO4, components of PM2.5 -PM25OC, PM25EC, and PM25SO4 respectively.
Emissions are computed at 72°F and we used SAS® to convert the MOVES-based PM2.5
species into the following SMOKE-ready pollutants:
• PEC_72: unchanged from MOVES-based PM25EC, subject to temperature
adjustment below 72°F.
• POC_72: modified MOVES-based PM25OC to remove metals, PNO3 (computed
from MOVES-based PM25EC), NH4 (computed from MOVES-based PM25SO4
and PNO3), and MOVES-based PM25SO4. Subject to temperature adjustment
below 72°F.
• PSO4: unchanged from MOVES-basedPM25SO4, not subject to temperature
adjustment.
• PNO3: computed from MOVES-based PM25EC, not subject to temperature
adjustment.
• OTHER: sum of computed metals (fraction of MOVES-based PM25EC) and
NH4 (computed from PNO3 and PSO4), not subject to temperature adjustment.
• PMFINE_72: Computed from OTHER and fraction of POC_72. Subject to
temperature adjustment below 72°F.
• PMC_72: Computed as fraction of sum of PMFINE_72, PEC_72, POC_72,
PSO4, and PNO3. Subject to temperature adjustment below 72°F.
The first three inputs listed above are processed in the on_noadj sector because temperature
adjustments are not needed for these inventories. The fourth input listed above is further broken
into the on_moves_runpm and on_moves_startpm sectors; these emissions require separate
temperature adjustments after some intermediate emissions processing.
MOVES gasoline emissions were used for light-duty gasoline vehicles (LGDV), light-duty
gasoline trucks 0-6000 pounds gross vehicle weight (LDGT1), light-duty gasoline trucks 6000-
8500 pounds gross vehicle weight (LDGT2), and heavy-duty gasoline trucks (HDGV).
Motorcycle emissions were not available from MOVES at the time of this project and so
emissions from that vehicle class came from the case-specific NMIM runs.
25
-------
MOVES-based, monthly state-level emissions were first allocated to counties using county-SCC
specific state-to-county ratios that we created from the 2030 Reference case NMIM run.
California MOVES-based emissions are discarded; they do not replace the existing California
inventories (discussed in on_noadj sector). MOVES data were provided by OTAQ in October
2009.
In each MOVES file, "start" emissions are represented by SCCs with the SCC characters 8-9
equal to "00": 2201001000 (LDGV), 2201020000 (LDGT1), 2201040000 (LDGT2),
2201070000 (HDGV), 2230001000 (LDDV), 22300601000 (LDDT), and 223007X000
(HDDV). These start emissions are assigned to urban and rural SCCs based on the county-level
ratio of NMIM emissions from urban versus local roads. For example, LDGV start emissions
(2201001000) were split into urban (2201001370) and rural (2201001350) based on the ratio of
LDGV emissions from urban (2201001330) and rural (2201001210) local roads.
Finally, the set of emissions are broken into 3 sets that are processed in 3 separate sectors for
LDGHG:
1. on_moves_startpm: monthly MOVES-based onroad gasoline "start" PM emissions
subject to temperature adjustments, and onroad gasoline PM species not subject to
temperature adjustments (e.g., PNO3 and PSO4). These are limited to 8 SCCs
(urban/rural and 4 vehicle types) representing parking areas for the following pollutants:
PEC_72, POC_72, PNO3, PSO4, OTHER, PMFINE_72, and PMC_72.
2. on moves runpm: monthly MOVES-based onroad gasoline "running" PM emissions
subject to a different set of temperature adjustments compared to "start" emissions;
similar to the on_moves_startpm sector, this sector includes all onroad gasoline PM
species, not just those subject to temperature adjustments. The same pollutants are
provided as on_moves_startpm.
3. on_noadj MOVES-based emissions: The remaining monthly diesel exhaust and non-PM
gasoline MOVES-based emissions that are also not subject to temperature adjustments -
see inputs (1) and (2) above. These emissions are modeled in the on_noadj sector
discussed below and include both start and running emissions for all diesel
(MOVES_based) pollutants and gasoline non-PM pollutants.
on_noadj
The on_noadj sector contains all US onroad mobile emissions that do not get a PM temperature
adjustment. There are three sources of data that are pre-processed to create three sets of monthly
inventories for this sector.
1. MOVES-based diesel exhaust and non-PM gasoline: These are the monthly non-PM
MOVES-based emissions discussed in item #3 in "Outputs" in the on_moves_runpm and
on_moves_startpm sector discussion. In short, these are non-California, select pollutants
exhaust and evaporative (non-refueling) onroad gasoline LDGV, LDGT1, LDGT2,
HDGV, and onroad diesel (HDDV -all classes) emissions.
2. California onroad inventory: California year 2030 complete CAP/HAP onroad inventory.
California monthly onroad emissions are year 2030 and are based on March 2007
California Air Resources Board (CARB) data (Martin Johnson: mjohnson@arb.ca.gov).
26
-------
NH3 emissions are from NMIM runs for California (same data as were used in 2030v3).
We estimated HAP emissions by applying HAP-to-CAP ratios computed from California
2005 NEI submittal provided by EPA in 12/2007. This was done because the CARB
submittal from March 2007 did not include estimates for HAPs. We retained only those
HAPs that were also estimated by NMIM for nonroad mobile sources; all other HAPs
were dropped.
3. Remaining onroad NMIM-based onroad inventory: The remainder of the non-California
onroad inventory not replaced by MOVES. This includes monthly emissions for all
onroad diesel, all motorcycles, all refueling, and onroad LDGV, LDGT1, LDGT2,
HDGV, and HDDV emissions for pollutants not covered by MOVES (e.g., SO2, NH3).
All NMIM onroad data are based on a composite Reference case inventory from Table
10, which uses AEO2007 fuels and NMEVI county database NCD20080727.
The remainder of this section discusses the pre-processing required to create monthly ORL files
for the remainder of the on_noadj sector (#3 above).
Table 10 shows how two NMIM inventories were combined to create the Reference case NMIM
emissions. OTAQ provided a "reference and control" set (LdGhgO2030el0.txt) in October 2009
of NMEVI ElO-penetration emissions to be used for both the Reference and Control case. This is
the complete NMIM scenario for the control case, but is used only for diesel vehicle emissions in
the Reference case scenario. This reference and control set of monthly emissions includes all 50
states plus DC and all CAPs and HAPs of interest. A Reference case-specific
(LdGhgO2030Aeo.txt) set of NMIM monthly emissions was also provided in October 2009 by
OTAQ for onroad gasoline emissions. Onroad gasoline emissions in the control case use the
"reference and control" case ElO-based inventory.
Similar to nonroad pre-processing, we also reassigned NMEVI evaporative and refueling xylene
(compound XYL or CAS=EVP_1330207, RFL_1330207) into MXYL (CAS=EVP_108383,
RFL_108383) and OXYL (CAS=EVP_95476, RFL_95476) using a 68% and32% ratio to
both evaporative and refueling XYL, respectively. We also split NMEVI exhaust xylene
(CAS=EXH_1330207) into MXYL (CAS=EXH_108383) and OXYL (CAS=EXH_95476)
using a 74% and 26% ratio to XYL, respectively.
Emissions were converted from monthly to average-day based the on number of days in each
month. CO2 and all of California emissions were removed prior to creating the NMIM-only
parts of the final inventory files.
We also removed refueling emissions from the raw NMIM data prior to creating the SMOKE
inputs, since these emissions were included in the nonpt and ptnonipm sectors. The NMIM
refueling data were used with existing 2005 refueling emissions to create projection ratios for
ptnonipm and nonpt refueling, as previously described in Section 4.1.1.
27
-------
5 2030 Control Case
The 2030 LDGHG Control Case was intended to represent the implementation of national CO2
emissions standards for light-duty vehicles. A list of inventory datasets used for this and the
2030 reference and 2005 cases is provided in Appendix A.
5.1 2030 Control Case Point sources
The point sources for the 2030 Control Case include the same emissions as the 2030 Reference
Case for the following point source sectors: US EGU point source (ptipm), sources from Mexico,
Canada, and the Gulf of Mexico (othpt).
As described in Section 4.1.1, the base case controls were erroneously not applied to both the
2030 Reference, and by extension, the 2030 Control case ptnonipm inventories. No other sectors
are impacted by this error.
For the nonEGU point sources (ptnonipm), some of the 2030 Reference case emissions were
adjusted to reflect the 2030 Control case. These differences from the 2030 Reference case are:
• Gasoline Distribution: These SCC- adjustments were supplied by OTAQ on 11/09/2009
in the Excel® workbook "LDGHG_SCC_GasDistrib_Adjust2.xls". These adjustments
reduce all pollutant emissions from the Reference to the Control case for Refinery-To-
Bulk-Terminal (rbt) and Bulk-Terminal-To-Pump (btp) by approximately 18.3% to
18.8%, respectively. On 12/28/2009, OTAQ also provided an additional 25 rbt-related
SCCs that were subject to the same projection factors. These adjustments impact both
the ptnonipm and nonpt sectors.
• Oil Refining, Crude Oil Transport, and Crude Oil Production: These SCC-level
adjustments were supplied by OTAQ on 11/09/2009 in the Excel® workbook
"Oil_Prod_Transp_Refine_2030_adjust.xls". These adjustments reduce all pollutant
emissions from the Reference to the Control case for refinery-related SCCs by
approximately 6.1% and for transport and production-related SCCs by approximately
0.6%. These adjustments impact both the ptnonipm and nonpt sectors.
• Onroad refueling: We used the same overall projection approach for onroad refueling as
was used in the 2030 Reference case for the ptnonipm and nonpt sectors. Instead of
using the 2030 Reference case NMIM run, we used the Control case-specific NMIM run
for the 2030 refueling ratio calculation. We applied these factors to the refueling SCCs
from the 2030 Reference case using this formula: Factor2030controi = NMIM Emis203ocontroi
/NMEVI Emis2030Refemce. These adjustments impact both the ptnonipm and nonpt sectors.
• The headspace vapor VOC speciation profile changed to reflect the assumption of 100%
E10 in the 2030 control case fpr rbt and btp emissions, the specifics of which are
provided in Section 3.
28
-------
5.2 2030 Control Case Nonpoint sources
The nonpoint sources for the 2030 Control Case include the same emissions as the 2030
Reference Case for the following sectors: US fugitive dust sources (afdust sector), US
agricultural NH3 (ag sector), the US average fires (avefire sector), and the nonpoint emissions
for Canada and Mexico (othar).
The 2030 Control Case emissions for the "other nonpoint" sector (nonpt) emissions differed
from the 2030 Reference case in the following ways:
• Created revised PFC emissions by applying adjustment factors by county and pollutant,
based on an Excel® spreadsheet provided by OTAQ on 1/14/2009: "aeo_to_eisa.xls".
These adjustments reduced PFC emissions as compared to the 2030 Reference case.
• As previously mentioned for the point sources, we changed refueling, rbt/btp, and oil
refining, transport, and production emissions projections, which also affected the nonpt
sector.
• The headspace vapor VOC speciation profile changed to reflect the assumption of 100%
E10 in the 2030 control case fpr rbt and btp emissions, the specifics of which are
provided in Section 3.
5.3 2030 Control Case Mobile sources
Compared to the 2030 Reference case, the mobile source emissions included changes for all
onroad mobile (on_noadj, on_moves_startpm, and on_moves_runpm), and non-California
nonroad mobile sources (nonroad). The aircraft, locomotive, and non-C3 commercial marine
(alm_no_c3), C3 commercial marine (seca_c3) and the Canada and Mexico emissions (othon)
were unchanged.
For the nonroad sector, as discussed in Section 4.3.4, OAQPS substituted 2030 Control case
emissions for nonroad gasoline 2-stroke and 4-stroke equipment and gasoline recreational marine
sources. California emissions were not changed, but emissions in all other states were updated.
Otherwise the steps taken were the same as described for the 2030 Reference case.
For the US onroad mobile sectors, as discussed in Section 4.3.5, OAQPS substituted NMIM
2030 Control case emissions for onroad gasoline vehicles and motorcycles. In addition, the
MOVES-based emissions for the control case completely replaced the reference case MOVES
data with the exception of MOVES-based vehicle class 8-12 diesel (HDDV) that are shared by
both the reference and control cases.
We allocated the state-SCC MOVES data to county-SCC using ratios developed from the
Control case-specific NMIM county-SCC data. Other than the different data used for creating
the monthly county-SCC SMOKE-ready inventories, we used the same processing steps as
described for the 2005 base and 2030 Control cases.
Lastly, VOC speciation profile changes affected this sector, as described in Section 3.
29
-------
APPENDIX A
Equations to adapt pre-speciated diesel emissions from MOVES to air quality modeling species needed
for CMAQ
As shown in equation (1) below, MOVES provides total PM2 5, PEC and PSO4. A remainder term, R, makes
up the difference between the two species and the total PM2.5.
MOVES total PM2.5 = PEC + PSO4 + R (1)
The R term includes POM, which consists of POC and the hydrogen and oxygen atoms attached to the
carbon as part of the organic matter, PNO3, soil oxides and metals (also known as "crustal" and called
METAL here), ammonium, and water, and thus can be also written as:
R = POM + PNO3 + METAL + NH4 + H20 (2)
To correctly calculate the five PM2.5 species needed for CMAQ, we first needed to break out the POC,
PNO3, and PMFINE from R. Different calculations are used for light-duty diesel vehicles and heavy-duty
diesel vehicles, since the speciation profiles for these are different. The speciation profiles used for these
calculations are:
For both light duty diesel vehicles and heavy duty diesel vehicles, the SPECIATE 4.0 PM2.5 speciation
profiles "3914" (HHDV) and "92042" (LDDV) will be used to help calculate the other species. At the time,
OTAQ did not provide a justification for choosing this profile, but the fractions of metals and PNO3 are
small and so presumably the choice does not matter too much as long as the smallest of those fractions is
representative.
We computed the primary nitrate based on speciation profile 92011 from the SPECIATE4.1 database (Hsu et
al., 2006) using equation (3) shown below.
PNO3 = PEC x FN03 /FEC (3)
where,
FEC = Fraction of elemental carbon in speciation profile:
- LDDV: 57.4805% (based on profile 92042)l
- HDDV: 77.1241% (based on profile 3914)
FNOS = Fraction of nitrate in speciation profile
LDDV: 0.1141% (based on profile 3914, intentionally inconsistent)
- HDDV: 0.1141% (based on profile 3914)
To identify which sources should get the LDDV and which should get the HDDV approach, see Table 1,
below.
Since CMAQ's PMFINE species is the sum of soil oxides, metals, ammonium, and water, we needed to
calculate all of its components. First, the metals and ammonium are computed using equations (4) and (5).
Equation (5) is based on stoichiometric calculations.
METAL = PEC x Fmetal / FEC (4)
1 All profile fractions provided in email from Catherine Yanca on 11/6/2009, 1:49pm in attachment "Equations for diesel MOVES
speciation use in CMAQ 110609.doc"
30
-------
NH4 = (PNO3/MWN03 +2 x PSO4/MWSO4) x MWNH4 (5)
where,
Fmetai = Fraction of metals in speciation profile (0.0026632)
MWS04 = Molecular weight of sulfate (96.0576)
MWNOs = Molecular weight of nitrate (62.0049)
MWNH4 = Molecular weight of ammonium (18.0383)
The final component of PMFINE is the non-carbon mass of organic carbon. To calculate the non-carbon
mass, we first needed to compute organic carbon from the remainder term, R.
A key assumption is that POM is a factor of 1.2 greater than the mass of primary organic carbon, which is
also used in the CMAQ postprocessing software at EPA.
POM = 1.2 x POC (6)
Using this assumption and assuming that the H20 is negligible, the equation needed for the calculation of
POC is shown in equation (7) below.
POC = 5/6 x (R - METAL - NH4 - PNO3) (7)
From equation (6), the non-carbon portion of the organic carbon matter is 20%, of the POC. By definition,
PMFINE is the sum of the non-carbon portion of the mass, METAL and NH4. Thus, we computed
PMFINE_72 using equation (8) shown below.
PMFINE_72 = METAL + NH4 + 0.2 x POC_72 (8)
For mobile sources, we assumed that PMC is 8.6% of the PM2.5 mass. Equation (9) shows how we
calculated it.
PMC = 0.086 x (PMFINE + PEC + POC + PSO4 + PNO3)
Table A-l. List of SCC groups for application of LDDV or UDDV approach
Approach
LDDV
HDDV
SCC list
2230001000 through 2230060334
2230071 1 10 through 2230075330
2 Value provided by Catherine Yanca and Joe Somers to OAQPS in email provided 11/5/2009
31
-------
APPENDIX B
Inventory Data Files Used for Each LDGHG Modeling Case - SMOKE Input Inventory Datasets
In any of the following dataset names where the placeholder has been provided, this is intended to
mean 12 separate files with the placeholder replaced with either jan, feb, mar, apr, may, jun, jul, aug,
sep, oct, nov, or dec, each associated with a particular month of the year.
32
-------
Table B-l. List of inventory data associated with LDGHG modeling cases.
Case
2005 Base
(2005cp_tox_05b)
o
o
All Cases
2030 Reference
and Control
Sector
ptipm
ptnonipm
afdust
ag
aim no c3
nonpt
nonroad
on noadj
on moves runpm
on moves startpm
seca c3
avefire
othar
othon
othpt
ptipm
SMOKE Input Files
Annual: ptinv_ptipm cap2005nei 26aug2008 v2 orl.txt
Annual: ptinv_ptipm_hap2005agtox_vl_noBAFM_28aug2008_vO_orl.txt
Daily: ptday_ptipm caphap noncem 2005ag ida.txt
Daily: ptday_ptipm caphap cem 2005ci ida.txt
ptinv_ptnonipm xportfrac 2005cap remainder 24nov2008 vO orl.txt
ptinv_ptnonipm 2005hap noBAFM remainder 24nov2008 vO orl.txt
ptinvjrtnonipm 2005caphap chippewa delete for 2022EISA 19nov2008 vO orl.txt
ptinv_ptnonipm caphap ethanol_plant additions 2005 17dec2008 vl orl.txt
ptinv_ptnonipm_2005caphap_3ethanol_plants_delete_for_all2022_19nov2008_vO_orl.txt
arinv afdust 2002ad xportfrac 26sep2007 vO orl.txt
arinv_ag_cap2002nei_06nov2006_vO_orl.txt
arinv aim no c3 cap2002v3 30jul2008 vl orl.txt
arinv_alm_no_c3_hap2002v4_12sep2008_vO_orl.txt
arinv nonpt caphap ethanol_plant additions 2005 17dec2008 vO orl.txt
arinv nonpt_pf4 cap nopfc 08sep2008 vO orl.txt
arinv nonpt_pf4 hap nopfc Ilsep2008 vO orl.txt
arinv_pfc_2002_caphap_27dec2007_vO_orl.txt
arinv nonroad caps 2005v2 revised 08sep2008 vO orl.txt
arinv nonroad calif caphap 2005v2 02apr2008 vO orl.txt
arinv nonroad haps 2005v2 revised 05sep2008 vO orl.txt
mbinv_onroad_calif_caphap_2005v2__02apr2008_vO_orl.txt
mbinv noadj capshaps 2005v2 nmim not2moves4LDGHG 12NOV09 12nov2009 vO orl.txt
mbinv_on_noadj_MOVES_LDGHG_2005__17NOV09_l 7nov2009_vO_orl.txt
mbinv on moves runpm LDGHG 2005 12NOV09 12nov2009 vO orl.txt
mbinv_on_moves_startpm_LDGHG_2005__12NO V09_12nov2009_vO_orl.txt
ptinv seca c3 caps2005pf4 31jul2008 vO orl.txt
ptinv_seca_c3_haps_east2005pf4_31jul2008_vO_orl.txt
ptinv seca c3 haps centra!2005pf4 31jul2008 vO orl.txt
ptinv_seca_c3_haps_west2005pf4_31jul2008_vO_orl.txt
ptinv seca c3 haps NonUS east2005pf4 09sep2008 vO orl.txt
ptinv_seca_c3_haps_NonUS_central2005pf4_09sep2008_vO_orl.txt
ptinv seca c3 haps NonUS west2005pf4 09sep2008 vO orl.txt
arinv_avefire_2002ce_21dec2007_vO_ida.txt
arinv avefire 2002 hap 18nov2008 vO orl.txt
arinv nonroad mexico interior!999 21dec2006 vO ida.txt
arinv nonroad mexico border!999 21dec2006 vO ida.txt
arinv nonroad Canada 2000 inventory Ilfeb2008 v2 ida.txt
arinv_nonpoint_Canada2000_2 1 dec2006_vO_ida.txt
arinv nonpt mexico interior!999 21dec2006 vO ida.txt
arinv_nonpt_mexico_borderl999_21dec2006_vO_ida.txt
mbinv onroad Canada2000 07nov2006 vO ida.txt
mbinv_onroad_mexico_borderl999_21dec2006_vO_ida.txt
mbinv onroad mexico interior!999 21dec2006 vO ida.txt
ptinv_2000_canada_egu_07nov2006_vO_ida.txt
ptinv 2000 Canada nonegu 05dec2006 vO ida.txt
ptinv mexico border99 03mar2008 vl ida.txt
ptinv mexico interior99 05feb2007 vO ida.txt
ptinv offshore_point from 200 l_platform 07nov2006 vO ida.txt
Annual: ptinv_egu_summer_2020_pf31capsHgHCl_19feb2008_vO_orl.txt
Annual: ptinv_ptipm hap2005agtox vl noBAFM noHgHCL 07jan2009 v2 orl.txt
Daily: ptday_ptipm caphap cem 2022ci ida.txt
Daily: ptday_ptipm caphap noncem 2022ci ida.txt
Daily: ptday_ptipm hap cem 2005ci ida.txt
-------
Case
2030 Reference Only
(2030cp_tox_05b)
o
^.
Sector
afdust
ag
aim no c3
seca c3
ptnonipm
nonpt
nonroad
on noadj
on moves runpm
on moves startpm
ptnonipm
nonroad
on noadj
on moves runpm
on moves startpm
SMOKE Input Files
Daily: ptday_ptipm hap noncem 2005ag ida.txt
arinv afdust 2020ce 02b BASE 07apr2008 vO orl.txt
arinv_ag_2020ce_02b_BASE_08apr2008_vO_orl.txt
arinv aim no c3 cap2030cp 21dec2009 vO orl.txt
arinv_alm_no_c3_hap2030cp_21dec2009_vO_orl.txt
ptinv seca c3 BAF HAPs2030pf31 17dec2009 vO orl.txt
ptinv_seca_c3_caps2030pf31_10jun2008_vO_orl.txt
ptinv_ptnonipm xportfrac 2030cp tox cap remainder 17dec2009 vO orl.txt
ptinv_ptnonipm 2005caphap chippewa delete for 2022EISA 19nov2008 vO orl.txt
ptinv_ptnonipm 2030cp tox hap noBAFM remainder 17dec2009 vO orl.txt
ptinv_ptnonipm caphap ethanol_plant additions 2022AEO 05jan2009 vl orl.txt
arinv_nonpt_caphap_biodiesel_plant_additions_2022AEO_18dec2008_vO_orl.txt
arinv nonpt caphap ethanol_plant additions 2022AEO 15dec2008 vO orl.txt
arinv nonpt_pf4 2030cp tox cap nopfc 17dec2009 vO orl.txt
arinv nonpt_pf4 2030cp tox hap nopfc fixed with wrap oilgas 17dec2009 vO orl.txt
arinv_nonpt_voc_ethanol_transfer_additions_2022AEO_15dec2008_vO_orl.txt
arinv_pfc reference caphap2030LDGHG 14dec2009 vO orl.txt
arinv nonroad capshaps LDGHG 2030 REFERENCE 15DEC09 15dec2009 vO orl.txt
arinv nonroad calif caphap 2030v31 17apr2008 vO orl.txt
mbinv onroad calif caphap 2030v31 30apr2008 vO orl.txt
mbinv_on_noadj_nmim_not2moves_2030LDGHG_REFERENCE__10DEC09_10dec2009_vO_orl.txt
mbinv_on_noadj_MOVES_LDGHG_2030_REFERENCE__l lDEC09_lldec2009_vO_orl.txt
mbinv_on_moves_runpm_LDGHG_2030_REFERENCE__llDEC09_lldec2009_vO_orl.txt
mbinv on moves startpm LDGHG 2030 REFERENCE 11DEC09 Ildec2009 vO orl.txt
ptinv_ptnonipm 2005caphap chippewa delete for 2022EISA 19nov2008 vO orl.txt
ptinv_ptnonipm 2030cp tox Idghg hap noBAFM remainder 28dec2009 vO orl.txt
ptinv_ptnonipm caphap ethanol_plant additions 2022AEO 05jan2009 vl orl.txt
ptinv_ptnonipm xportfrac 2030cp tox Idghg cap remainder 28dec2009 vO orl.txt
arinv nonpt caphap biodiesel_plant additions 2022AEO 18dec2008 vO orl.txt
arinv nonpt caphap ethanol_plant additions 2022AEO 15dec2008 vO orl.txt
arinv nonpt_pf4 2030cp tox Idghg cap nopfc 28dec2009 vO orl.txt
arinv nonpt_pf4 2030cp tox Idghg hap nopfc fixed with wrap oilgas 28dec2009 vO orl.txt
arinv nonpt voc ethanol transfer additions 2022AEO 15dec2008 vO orl.txt
arinv_pfc_control_caphap2030LDGHG_14dec2009_vO_orl.txt
arinv nonroad calif caphap 2030v31 17apr2008 vO orl.txt
arinv_nonroad_capshaps_LDGHG_2030_CONTROL__15DEC09_15dec2009_vO_orl.txt
mbinv onroad calif caphap 2030v31 30apr2008 vO orl.txt
mbinv on noadj nmim not2moves 2030LDGHG CONTROL 10DEC09 10dec2009 vO orl.txt
mbinv_on_noadj_MOVES_LDGHG_2030_CONTROL__llDEC09_lldec2009_vO_orl.txt
mbinv on moves runpm LDGHG 2030 CONTROL feb 11DEC09 Ildec2009 vO orl.txt
mbinv_on_moves_startpm_LDGHG_2030_CONTROL__l lDEC09_lldec2009_vO_orl.txt
-------
APPENDIX C
Ancillary Data Files Used for LDGHG 2005 Case Compared to 2005 v4 Platform Data Files
To match the Datasets and Versions listed in this table to actual data files, combine the Dataset name and the
version number in the following pattern: __.txt, where is
the last date of change for that version and will have a unique value for the combination of Dataset Name
and Version number.
35
-------
Table C-l. Detailed list of ancillary data differences between the LDGHG 2005 and the 2005 v4 platform
Description
Area-source spatial
cross-reference
Onroad spatial
cross-reference
Area and onroad
temporal profiles
Area temporal
cross-reference
Onroad temporal
cross-reference
Grid descriptions
Inventory table
Inventory table
Non-HAP
Exclusions for nonpt
sector
Elevated
configuration file for
seca c3 sector
Speciation profiles
for TOG
Speciation profiles
speciated VOC
Environment
Variable
AGREF
MGREF
ATPRO,
MTPRO,
PTPRO
ATREF,
PTREF
MTREF
GRIDDESC
INVTABLE
INVTABLE
NHAP
EXCLUDE
PELV
CONFIG
GSPRO
GSPRO
Sectors
All sectors
All sectors
All sectors
All sectors
All sectors
All sectors
All sectors
avefire,
ptnonipm,
ptipm
nonpt
seca_c3
All sectors
othpt
2005 v4 platform
Dataset
amgref_us_can_mex_revised
amgref_us_can_mex_revised
amptpro_2005_us_can_revised
amptref_v3_3_revised
amptref_v3_3_revised
griddescjambertonly
invtable_hapcapintegate_cb05s
oa_nomp
invtable_hapcapnohapuse_cb05
soa_nomp
nhapexclude_nonpt_pf4
pelvconfig_seca_c3
gspro tog cb05 soa pf4 pretie
r2
gspro_speciated_voc
Vsn
5
7
0
1
0
24
4
4
3
0
1
0
2005 LDGHG platform
Dataset
amgref_us_can_allmex3
amgref_us_can_allmex3
amptpro_2005_us_can
amptref_v3_3
amptref_v3_3
griddescjambertonly
invtable_hapcap_cb05soa
n/a
nhapexclude_nonpt_pf4
n/a
gspro_tog_nohapuse_cb05_tx_
pf4_pretier2
n/a
Vsn
12
12
3
9
9
23
9
2
1
Comment and Impact
For v4 platform, revised for new Canadian inventory and
spatial surrogates. LDGHG includes MOVES SCCs for
diesel parking areas
For v4 platform, revised for new Canadian inventory and
spatial surrogates. LDGHG includes MOVES SCCs for
diesel parking areas
For v4 platform, revised for new Canadian inventory and
temporal profile. LDGHG includes HDD idling profiles
For v4 platform, added SCCs needed for 2005 v2 point
inventory and for WRAP oil and gas inventory. LDGHG
includes MOVES-based diesel parking SCCs
For v4 platform, added SCCs needed for 2005 v2 point
inventory (but not for WRAP oil and gas inventory).
Superceeded by v1 of the same dataset.
An older file used for LDGHG with an unrelated grid not
defined.
LDGHG used a full toxics approach for processing the
emissions and 2005 v4 platform used an approach without
most toxics. Impacts only the species included in the air
quality modeling. LDGHG also includes speciated PM from
MOVES diesel sources.
Approach for implementing "no HAP use" approach for
these sectors was different in LDGHG, but the result was
the same.
v3 includes WRAP oil and gas SCCs since these do not
have HAP VOCs either. v3 needed for the v4 platform but
not LDGHG as LDGHG does not include WRAP oil and gas
SCCs.
v4 platform used inline point sources, and so PELVCONFIG
file needed only in v4 platform.
Excluded ald2_primary and form_primary from v4 platform
since not needed
For the v4 platform, this speciation profile dataset passes
through the pre-speciated VOC point source species
provided by Environment Canada 2006 data.
Impact?
Yes
Yes
Yes
Yes
Yes
No
No
No
Yes
Yes
No
Yes
-------
Description
Speciation profiles
Canada PM
Speciation profiles
for biogenic
emissions
Speciation profiles
Other VOC HAP
Speciation profiles
CHROMIUM
Speciation profiles
METALS
Speciation profiles
MOVES PM
Speciation profile
8762/8763 for TOG
toBAF
Speciation profile
8762/8763 for
nontoxics
Speciation profile
8762/8763 for toxics
Speciation xref for
PM2.5 diesel SCCs
but do not produce
diesel
Speciation xref NO
DIESEL PM, othar
Speciation xref for
non-diesel PM2.5
Environment
Variable
GSPRO
GSPRO
GSPRO
GSPRO
GSPRO
GSPRO
GSPRO
GSPRO
GSPRO
GSREF
GSREF
GSREF
Sectors
othpt
biog
All sectors
All sectors
All sectors
All sectors
All sectors
All sectors
All sectors
All sectors
othar
All sectors
2005 v4 platform
Dataset
gspro_pm25_canada_2006_poi
nt
gspro_biogenics
n/a
n/a
n/a
n/a
n/a
n/a
n/a
gsref_no_dieselpm
gsref_no_dieselpm
gsref_pm25_pf4_nondiesel
Vsn
0
0
1
1
8
2005 LDGHG platform
Dataset
n/a
n/a
gspro_other_hapvoc_nobenz-
benz
gspro_chromium
gspro_hapmetals
gspro_speciated _pm
gspro cmaq cb05 hspace BA
F
gspro_cmaq_cb05_hspace_non
toxic
gspro_cmaq_cb05_hspace_toxi
c
n/a
gsref_no_dieselpm
gsref_pm25_pf4_nondiesel
Vsn
0
0
0
3
1
0
0
2
1
Comment and Impact
For the v4 platform, this speciation profile dataset provides
the Canadia-specific PM2.5 speciation profile recommended
by Environment Canada for point sources (1 % POC, 2%
PEC, 12% PS04, 85% PMFINE)
For the v4 platform, this dataset contains the biogenic VOC
speciation profiles previously included in gspro_static_cmaq
v7 for the RFS2 effort
For LDGHG, this dataset has the HAP VOC species that
get passed through from inventory to the multipollutant
inputs created for LDGHG.
Chromium speciation profiles needed only for multi-
polltutant approach used in LDGHG
HAP metal pass-through profiles needed only for multi-
pollutant approach used in LDGHG.
Speicated PM (non-72 degrees) added for MOVES diesel
exhaust in LDGHG.
For LDGHG, benzene from TOG for headspace profiles
8762 and 8763
For LDGHG, toxics from TOG for headspace profiles 8762
and 8763
For LDGHG, toxics from NONHAPTOG for headspace
profiles 8762 and 8763
Both the LDGHG and v4 platform did not create the
DIESEL_PM species needed for the multipollutant version
of CMAQ, the data set used for the v4 platform contains the
SCC cross-references for diesel SCCs without the
DIESEL_PM species.
Prevents creation of diesel PM species in Canada for
LDGHG
For the v4 platform, this dataset includes updates for 2005
v2 point sources as well as WRAP oil and gas inventory.
Note that future years of LDGHG used more current version
than "v1 " of this dataset, and the v8 dataset could be used
for replicating LDGHG runs.
Impact?
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
No
No
-------
oo
Description
Speciation xref for
VOC, not year-
specific
Speciation xref for
NONHAPVOC, not
year-specific
Speciation xref for
VOC, year-specific
Speciation xref for
NONHAPVOC,
year-specific
Speciation xref
static NOX -
MONO for mobile
sources
Speciation xref for
Integrate-HAPs
static
Speciation xref for
Canada PM
Speciation xref HAP
chromium nonroad
sectors
Speciation xref HAP
chromium on_noadj
sector
Speciation xref HAP
chromium stationary
sectors
Speciation xref HAP
metals nonroad
sectors
Environment
Variable
GSREF
GSREF
GSREF
GSREF
GSREF
GSREF
GSREF
GSREF
GSREF
GSREF
GSREF
Sectors
All sectors
All sectors
All sectors
All sectors
All sectors
All sectors
othpt
alm_no_c3
nonroad
seca_c3
on_noadj
nonpt
ptipm
ptnonipm
alm_no_c3
nonroad
seca_c3
2005 v4 platform
Dataset
gsref_voc_general
gsref_nonhapvoc_general_upda
te
gsref_voc_2005
gsref_nonhapvoc_2005
gsref_static_nox_hono_pf4
gsref_static_integratehap_emv4
gsref_pm25_canada_2006_poi n
t
n/a
n/a
n/a
n/a
Vsn
19
3
2
1
3
2
2
2005 LDGHG platform
Dataset
gsref_voc_general_ldghg
gsref_nonhapvoc_general
gsref_voc_2005_ldghg
gsref_nonhapvoc_2005_ldghg
gsref_static_nox_hono_pf4
n/a
n/a
gsref_chromium_nonroad
gsref_chromium_onroad
gsref_chromium_stationary
gsref_metals_nonroad
Vsn
0
1
0
0
2
1
1
0
0
Comment and Impact
LDGHG copied V20 file from v4 platform and replaced
headspace profile 8737 with 8762. For v4 platform,
includes updates for 2005 v2 point inventory, WRAP oil and
gas SCCs, changes to gasoline distribution SCCs and all
SCCs from HAP inventory.
For the v4 platform, includes WRAP oil and gas SCCs and
removed duplicates. LDGHG replaces profile 8737 with
headspace profile 8762.
The LDGHG file is the v4 file updated to include new
headspace profiles 8762.
LDGHG replaces headspace profile 8737 with 8762
The v4 platform dataset adds four Canadian SCCs that are
needed for processing the 2006 Canadian inventory.
For the v4 platform, this speciation cross-reference dataset
passes through the pre-speciated VOC point source
species provided by Environment Canada 2006 data.
For the v4 platform, this speciation cross-reference dataset
assigns the single Canadia-specific PM2.5 speciation profile
recommended by Environment Canada for all point
sources.
Enables creation of chromium species
Enables creation of chromium species
Enables creation of chromium species
Enables creation of HAP metal species
Impact?
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
-------
Description
Speciation xref HAP
metals stationary
sector
Speciation xref HAP
metals nonroad
sectors
Speciation xref HAP
metals on_noadj
sector
SCC Descriptions
Environment
Variable
GSREF
GSREF
GSREF
SCCDESC
Sectors
nonpt
ptipm
ptnonipm
nonroad
on_noadj
All sectors
2005 v4 platform
Dataset
n/a
n/a
n/a
sccdesc_pf31
Vsn
5
2005 LDGHG platform
Dataset
gsref_metals_stationary
gsref_metals_nonroad
gsref_metals_onroad
sccdesc_pf31
Vsn
1
0
0
8
Comment and Impact
Enables creation of HAP metal species
Enables creation of HAP metal species
Enables creation of HAP metal species
LDGHG updated to include MOVES diesel start SCCs and
Canada 2006 added (late for v4)
Impact?
Yes
Yes
Yes
Yes
VO
-------
APPENDIX D
Growth and Control Assumptions and Affected Pollutants for the 2030 LDGHG Reference Case
For nonEGU point and stationary area sources, the "2005" inventory data used 2002 emissions. As a result,
we used our 2002-based approaches for these sectors to project to 2030. Many of these controls have
effective dates between 2002 and 2005, and therefore would not be applied to a true 2005 inventory value.
As mentioned in Section 4.1.1, these controls were inadvertently not applied to the 2030 reference and
control cases.
Table D-l. Control Strategies and Projection Assumptions in the 2030 LDGHG Reference Case
Emissions Inventories
Control Strategies
(Grouped by Affected Pollutants or Standard and Approach Used to
Apply to the Inventory)
Non-EGU Point Controls
NOX SIP Call (Phase II):
Cement Manufacturing
Large Boiler/Turbine Units
Large 1C Engines
DOJ Settlements: plant SCC controls
Alcoa, TX
MOTIVA, DE
Refinery Consent Decrees: plant/SCC controls
Closures, pre-2007: plant control of 100%
Auto plants
Pulp and Paper
Large Municipal Waste Combustors
Small Municipal Waste Combustors
Plants closed in preparation for 2005 inventory
Industrial Boiler/Process Heater plant/SCC controls for PM
Large Municipal Waste Combustors (LMWC)
Small Municipal Waste Combustors (SMWC)
MACT rules, national, VOC: national applied by SCC, MACT
Boat Manufacturing
Polymers and Resins III (Phenolic Resins)
Polymers and Resins IV (Phenolic Resins)
Wood Building Products Surface Coating
Generic MACT II: Spandex Production, Ethylene manufacture
Large Appliances
Miscellaneous Organic NESHAP (MON): Alkyd Resins, Chelating Agents,
Explosives, Phthalate Plasicizers, Polyester Resins, Polymerized Vinylidene
Chloride
Manufacturing Nutritional Yeast
Oil and Natural Gas
Petroleum Refineries -Catalytic Cracking, Catalytic Reforming, & Sulfur
Plant Units
Pesticide Active Ingredient Production
Publicly Owned Treatment Works
Reinforced Plastics
Rubber Tire Manufacturing
Asphalt Processing & Roofing
Combustion Sources at Kraft, Soda, and Sulfite Paper Mills
Fabric Printing, Coating and Dyeing
Pollutants
Affected
NOx
NOx, SO2
NOx, PM, SO2
all
PM
PM, Hg, and
metals
PM, Hg, metals,
NOx, SO2
VOC
Approach or
Reference:
1
2
o
J
4
5
6
6
EPA, 2007
Applied?
No
No
No
No
No
Partially
Partially
Only
Municipal
Solid Waste
Landfills
were
applied
40
-------
Control Strategies
(Grouped by Affected Pollutants or Standard and Approach Used to
Apply to the Inventory)
Iron & Steel Foundries
Metal: Can, Coil
Metal Furniture
Miscellaneous Metal Parts & Products
Municipal Solid Waste Landfills
Paper and Other Web
Plastic Parts
Plywood and Composite Wood Products
Wet Formed Fiberglass Production
Wood Building Products Surface Coating
Carbon Black Production
Cellulose Products Manufacturing
Cyanide Chemical Manufacturing
Friction Products Manufacturing
Leather Finishing Operations
Miscellaneous Coating Manufacturing
Organic Liquids Distribution (Non-Gasoline)
Refractory Products Manufacturing
Sites Remediation
Solid Waste Rules (Section 129d/llld)
Hospital/Medical/Infectious Waste Incinerator Regulations
MACT rules, national, PM:
Portland Cement Manufacturing
Secondary Aluminum
MACT rules, plant-level, VOC:
Auto Plants
MACT rules, plant-level, PM & SO2:
Lime Manufacturing
MACT rules, plant-level, PM:
Taconite Ore
Stationary Area Assumptions
Municipal Waste Landfills: project factor of 0.25 applied,
Livestock Emissions Growth to year 2020
Residential Wood Combustion Growth and Changeouts to year 2020
Gasoline Stage II growth and control to 2030 Reference
Portable Fuel Container growth and control to year 2030
ECU Point Controls
CAIR/CAMR/CAVR
IPM Model 3.0
Onroad Mobile and Nonroad Mobile Controls (list includes all key
mobile control strategies but is not exhaustive)
Pollutants
Affected
NOx, PM, SO2
PM
VOC
PM, SO2
PM
VOC
NH3
all
VOC
VOC
NOx, SO2, PM
Approach or
Reference:
EPA, 2005
7
8
9
10
EPA, 2007
11
12
13
14
15
National Onroad Rules:
Tier 2 Rule
2007 Onroad Heavy -Duty Rule all
Final Mobile Source Air Toxics Rule (MSAT2)
Renewable Fuel Standard
Applied?
Partially
No
No
No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
41
-------
Control Strategies
(Grouped by Affected Pollutants or Standard and Approach Used to Pollutants Approach or
Apply to the Inventory) Affected Reference:
Local Onroad Programs:
National Low Emission Vehicle Program (NLEV) VOC 16
Ozone Transport Commission (OTC) LEV Program
National Nonroad Controls:
Clean Air Nonroad Diesel Final Rule - Tier 4
Control of Emissions from Nonroad Large-Spark Ignition Engines and
Recreational Engines (Marine and Land Based): "Pentathalon Rule"
Clean Bus USA Program all 17,18,19
Control of Emissions of Air Pollution from Locomotives and Marine
Compression-Ignition Engines Less than 30 Liters per Cylinder
Exhaust emission standards for marine spark-ignition engines and small
land-based nonroad engines
Aircraft, Locomotives, and Commercial Marine Assumptions
Aircraft:
Itinerant (ITN) operations at airports to year 2022 a11 20
Locomotives:
Energy Information Administration (El A) fuel consumption projections for
freight rail ,, 18, 21, (EPA,
Clean Air Nonroad Diesel Final Rule - Tier 4 2009)
Locomotive Emissions Final Rulemaking, December 17, 1997
Control of Emissions of Air Pollution from Locomotives and Marine
Commercial Marine:
EIA fuel consumption projections for diesel-fueled vessels
OTAQ ECA C3 Base 2022 inventory for residual-fueled vessels
Clean Air Nonroad Diesel Final Rule - Tier 4 all 21, (EPA, 2009)
Emissions Standards for Commercial Marine Diesel Engines, December 29,
1999
Tier 1 Marine Diesel Engines, February 28, 2003
Reference Case-specific inventory adjustments
Nonpt and ptnonipm:
Gasoline distribution by SCC: bulk-plant-to-pump (btp) and refinery -to- vnr FPA ^c\^c\
i 11 j. • i / j.1 \ V V_/l^- \—ii£\. Z\) \.\)
bulk terminal (rtb) processes ,,
Ethanol plant additions
ptnonipm:
Refinery adjustments by state and SCC all EPA, 2010
Ethanol plant additions all
nonpt:
Ethanol transfer additions VOC EPA, 2010
Biodiesel Plant additions all
Nonroad (not including California): EPA, 2010
AEO-specific gasoline equipment and gasoline rec-marine all
EO-based emissions for diesel rec-marine
Onroad (not including California): EPA, 2010
NMIM and MOVES AEO-specific emissions for onroad gasoline and diesel
Applied?
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
APPROACHES:
1.
Used Emission Budget Inventories report (EPA, 1999) for list of SCCs for application of controls, and for percent reductions
(except 1C Engines). Used Federal Register on Response to Court decisions (Federal Register, 2004) for 1C Engine percent
reductions and geographic applicability
For ALCOA consent decree, used http:// cfpub.epa.gov/compliance/cases/index.cfm; for MOTIVA: used information sent
by State of Delaware
Used data provided by Brenda Shine, EPA, OAQPS
Closures obtained from EPA sector leads; most verified using the world wide web.
Used data list of plants provided by project lead from 2001-based platform; required mapping the 2001 plants to 2002 NEI
42
-------
plants due to plant id changes across inventory years
6. Used data provided by Walt Stevenson, EPA, OAQPS
7. Same as used in CAIR, except added SCCs appeared to be covered by the rule: both reductions based on preamble to final
rule. (Portland Cement used a weighted average across two processes )
8. Percent reductions recommended and plants to apply to reduction to were based on recommendations by rule lead engineer,
and are consistent with the reference: EPA, 2007e
9. Percent reductions recommended are determined from the existing plant estimated baselines and estimated reductions as
shown in the Federal Register Notice for the rule. SO2 % reduction will therefore be 6147/30,783 = 20% and PM10 and
PM2.5 reductions will both be 3786/13588 = 28%
10. Same approach used in CAIR: FR notice estimates reductions of "PM emissions by 10,538 tpy, a reduction of about 62%."
Used same list of plants as were identified based on tonnage and SCC from CAIR.
11. Except for dairy cows and turkeys (no growth), based in animal population growth estimates from USD A and Food and
Agriculture Policy and Research Institute.
12. Expected benefits of woodstoves change-out program: http://www.epa.gov/woodstoves/index.html
13. VOC emission ratios of year 2022 AEO-specific from year 2005 from the National Mobile Inventory Model (NMIM) results
for onroad refueling including activity growth from VMT, Stage II control programs at gasoline stations, and phase in of
newer vehicles with onboard Stage II vehicle controls.
14. VOC and benzene emissions for year 2020 from year 2002 from MSAT rule (EPA, 2007c, EPA, 2007d)
15. http://www.epa.gov/airmarkets/progsregs/epa-ipm/docs/summarv2006.pdf
16. Only for states submitting these inputs: http://www.epa.gov/otaq/lev-nlev.htm
17. http://www.epa. gov/nonroad-diesel/2004fr.htm
18. http://www.epa.gov/cleanschoolbus/
19. http://www.epa.gov/otaq/marinesi.htm
20. Federal Aviation Administration (FAA) Terminal Area Forecast (TAP) System, December 2007:
http://www.apo.data.faa.gov/main/taf.asp
21. http://www.epa.gov/nonroad-diesel/2004fr.htm
Appendix D References
EPA, 2005. Clean Air Interstate Rule Emissions Inventory Technical Support Document, U.S.
Environmental Protection Agency, Office of Air Quality Planning and Standards, March 2005. Available at
http://www.epa.gov/cair/pdfs/fmaltech01.pdf
EPA, 2007. Guidance for Estimating VOC and NOx Emission Changes from MACT Rules, U.S.
Environmental Protection Agency Office of Air Quality Planning and Standards, Air Quality Policy
Division, Research Triangle Park, NC 27711, EPA-457/B-07-001, May 2007. Available at
http://www.epa.gov/ttn/naaqs/ozone/o3imp8hr/documents/guidance/200705_epa457_b-07-
001 emission changes mact rules.pdf
EPA, 2009. Regulatory Impact Analysis: Control of Emissions of Air Pollution from Locomotive Engines
and Marine Compression Ignition Engines Less than 30 Liters Per Cylinder. U.S. Environmental Protection
Agency Office of Transportation and Air Quality, Assessment and Standards Division, Ann Arbor, MI 48105,
EPA420-R-08-001a, May 2009. Available at: http://www.epa.gov/otaq/regs/nonroad/420r08001a.pdf
EPA, 2010 : RFS2 Emissions Inventory for Air Quality Modeling Technical Support Document, U.S.
Environmental Protection Agency, Assessment and Standards Division, Office of Transportation and Air
Quality, and Emissions Inventory Analysis Group, Office of Air Quality Planning and Standards, January
2010. Available at http://www.epa.gov/otaq/renewablefuels/420rl0005.pdf
43
------- |