Methodology
Report for Inventory
of U.S. Greenhouse
Gas Emissions
and Sinks by State:
1990-2022
~ j nr. i
~
EPA-430-R-24-006
U.S. Environmental Protection Agency
Office of Atmospheric Protection
Climate Change Division
August 2024
v=,EPA
-------
Contents
Section Page
Contents ii
Figures v
Tables vi
Acknowledgements viii
1 Introduction 1-1
1.1 Areas Where Differences Between State GHG Inventories and the EPA State-Level
Estimates May Occur 1-2
1.2 Institutional Arrangements for Compiling State-Level Inventory Estimates 1-4
1.3 Methods Overview 1-4
1.4 Summary of Updates Since Previous Report 1-5
1.5 QA/QC Procedures 1-7
1.5.1 State Expert Review 1-8
1.5.2 Peer Review 1-8
1.6 Uncertainty 1-8
1.7 Planned Improvements 1-9
1.8 References 1-9
2 Energy (NIR Chapter 3) 2-1
2.1 Emissions Related to Fuel Use 2-2
2.1.1 Fossil Fuel Combustion (NIR Section 3.1) 2-2
2.1.2 Carbon Emitted from NEUs of Fossil Fuel (NIR Section 3.2) 2-28
2.1.3 Geothermal Emissions 2-32
2.1.4 Incineration ofWaste (NIRSection 3.3) 2-33
2.1.5 International Bunker Fuels (NIRSection 3.10) 2-35
2.1.6 Wood Biomass and Biofuels Consumption (NIR Section 3.11) 2-36
2.2 Fugitive Emissions 2-38
2.2.1 Coal Mining (NIR Section 3.4) 2-38
2.2.2 Abandoned Underground Coal Mines (NIR Section 3.5) 2-41
2.2.3 Petroleum Systems (NIR Section 3.6) 2-45
2.2.4 Natural Gas Systems (NIR Section 3.7) 2-48
2.2.5 Abandoned Oiland Gas Wells (NIRSection 3.8) 2-54
3 Industrial Processes and Product Use (NIR Chapter 4) 3-1
3.1 Minerals 3-3
3.1.1 Cement Production (NIRSection 4.1) 3-3
3.1.2 Lime Production (NIR Section 4.2) 3-6
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State ii
-------
3.1.3 Glass Production (NIR Section 4.3) 3-11
3.1.4 Other Process Uses of Carbonates (NIR Section 4.4) 3-14
3.1.5 Carbon Dioxide Consumption (NIR Section 4.16) 3-18
3.2 Chemicals 3-20
3.2.1 Ammonia Production (NIR Section 4.5) 3-20
3.2.2 Urea Consumption for Nonagricultural Purposes (NIR Section 4.6) 3-23
3.2.3 Nitric Acid Production (NIR Section 4.7) 3-25
3.2.4 Adipic Acid Production (NIR Section 4.8) 3-27
3.2.5 Caprolactam, Glyoxal, and Glyoxylic Acid Production (NIR Section 4.9) 3-28
3.2.6 Carbide Production and Consumption (NIR Section 4.10) 3-31
3.2.7 Titanium Dioxide Production (NIR Section 4.11) 3-34
3.2.8 Soda Ash Production (NIR Section 4.12) 3-36
3.2.9 Petrochemical Production (NIR Section 4.13) 3-37
3.2.10 HCFC-22 Production (NIR Section 4.14) 3-47
3.2.11 Production of Fluorochemicals Other than HCFC-22 (NIR Section 4.15) 3-51
3.2.12 Phosphoric Acid Production (NIR Section 4.17) 3-53
3.3 Metals 3-55
3.3.1 Iron & Steel Production and Metallurgical Coke Production (NIR Section 4.18) 3-55
3.3.2 Ferroalloys Production (NIR Section 4.19) 3-61
3.3.3 Aluminum Production (NIR Section 4.20) 3-64
3.3.4 Magnesium Production and Processing (NIR Section 4.21) 3-67
3.3.5 Lead Production (NIR Section 4.22) 3-70
3.3.6 Zinc Production (NIR Section 4.23) 3-73
3.4 Product Use (Fluorinated Sources, N20) 3-76
3.4.1 Electronics Industry (NIR Section 4.24) 3-77
3.4.2 Substitution of Ozone-Depleting Substances (NIR Section 4.25) 3-83
3.4.3 Electrical Equipment (NIR Section 4.26) 3-87
3.4.4 SFs and PFCs from Other Product Use (NIR Section 4.27) 3-92
3.4.5 Nitrous Oxide from Product Uses (NIR Section 4.28) 3-94
4 Agriculture (NIR Chapter 5) 4-1
4.1 Livestock Management 4-1
4.1.1 Enteric Fermentation (NIR Section 5.1) 4-2
4.1.2 Manure Management (NIR Section 5.2) 4-3
4.2 Other (Agriculture) 4-5
4.2.1 Rice Cultivation (NIR Section 5.3) 4-5
4.2.2 Agricultural Soil Management (NIR Section 5.4) 4-7
4.2.3 Liming (NIR Section 5.5) 4-9
4.2.4 Urea Fertilization (NIRSection 5.6) 4-10
4.2.5 Field Burning of Agricultural Residues (NIR Section 5.7) 4-11
iii
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
5 Land Use, Land-Use Change, and Forestry (NIR Chapter 6) 5-1
5.1.1 Forest Land Remaining Forest Land (NIR Section 6.2) 5-2
5.1.2 Land Converted to Forest Land (NIR Section 6.3) 5-5
5.1.3 Cropland Remaining Cropland (NIR Section 6.4) 5-6
5.1.4 Land Converted to Cropland (NIR Section 6.5) 5-7
5.1.5 Grassland Remaining Grassland (NIR Section 6.6) 5-9
5.1.6 Land Converted to Grassland (NIR Section 6.7) 5-10
5.1.7 Wetlands Remaining Wetlands (NIR Section 6.8) 5-12
5.1.8 Flooded Land Remaining Flooded Land (NIR Section 6.8) 5-19
5.1.9 Land Converted to Wetlands (NIR Section 6.9) 5-20
5.1.10 Settlements Remaining Settlements (NIRSection 6.10) 5-23
5.1.11 Land Converted to Settlements (NIRSection 6.11) 5-29
5.1.12 Other Land Remaining Other Land (NIR Section 6.12) and Land Converted to
Other Land (NIR Section 6.13) 5-31
6 Waste (NIR Chapter 7) 6-1
6.1 Solid Waste Disposal 6-1
6.1.1 Landfills (NIR Section 7.1) 6-2
6.1.2 Composting (NIR Section 7.3) 6-6
6.1.3 Anaerobic Digestion at Biogas Facilities (Stand-Alone) (NIR Section 7.4) 6-10
6.2 Wastewater Management 6-12
6.2.1 Wastewater Treatment and Discharge (NIR Section 7.2) 6-12
List of Appendices
Appendix A Data Appendices
Appendix B State-Level GHG Data Caveats
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State iv
-------
Figures
Number Page
Figure 2-1. Adjustments to Energy Consumption for Emissions Estimates 2-3
Figure 2-2. Adjustments Made to Industrial Sector Energy Use to Account for Emissions Reported in
IPPU 2-8
Figure 2-3. Comparison of Gasoline Sector Allocation 2-12
Figure 2-4. Comparison of Diesel Fuel Sector Allocation 2-13
Figure 2-5. Comparison of Transportation Sector Fuel Use 2-14
Figure 2-6. Transportation Sector 2020 State-Level Allocation Examples 2-16
Figure 2-7. Adjustments Made to Industrial Sector Energy Use to Account for Emissions Reported as
NEUs 2-18
Figure 2-8. Adjustments Made to Transportation Sector Energy Use to Account for IBFs 2-20
Figure 2-9. Differences in State-Level Total and National Total FFC C02 Emissions 2-22
Figure 2-10. Mobile Source Non-C02 Calculation Methodology 2-24
Figure 2-11. Adjustments to Energy Consumption for Emissions Estimates 2-29
Figure 2-12. Differences in State-Level and NationalTotal NEU C02 Emissions 2-31
Figure 3-1. U.S. Transmission Lines Separated by State UsingGIS Processing Tool 3-89
V
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Tables
Number Page
Table 1-1. Category Estimates Updated Since Release of Previous Inventory by U.S. State 1-6
Table 2-1. Overview of Approaches for Estimating State-Level Energy Sector GHG Emissions 2-1
Table 2-2. Comparison of Approaches/Data Sources Used to Determine FFC Emissions 2-5
Table 2-3. Default Data Sources for Mobile Source Non-C02 Emissions 2-24
Table 2-4. Summary of Approaches to Disaggregate Waste Incineration Emissions Across Time
Series 2-33
Table 2-5. Example State Allocation Factors for the Illinois Coal Basin (Sealed Mines) 2-43
Table 3-1. Overview of Approaches for Estimating State-Level IPPU Sector GHG Emissions 3-1
Table 3-2. Summary of Approaches to Disaggregate the National Inventory for Cement Production
Across Time Series 3-4
Table 3-3. Summary of Approaches to Disaggregate the National Inventory for Lime Production
Across Time Series 3-6
Table 3-4. Summary of Approaches to Disaggregate the National Inventory for Glass Production
Across Time Series 3-11
Table 3-5. Summary of Approaches to Disaggregate the National Inventory for Ammonia Production
Across Time Series 3-21
Table 3-6. Summary of Approaches to Disaggregate the National Inventory for Nitric Acid
Production Across Time Series 3-25
Table 3-7. Summary of Approaches to Disaggregate the National Inventory for Ti02 Production
Across Time Series 3-34
Table 3-8. Summary of Approaches to Disaggregate the National Inventory for HCFC-22 Production
Across Time Series 3-47
Table 3-9. Facilities Producing HCFC-22 or Destroying HFC-23 Generated During HCFC-22
Production from 1990 to 2022 3-49
Table 3-10. Summary of Approaches to Disaggregate the National Inventory for Fluorochemical
Production Across Time Series 3-51
Table 3-11. Summary of Approaches to Disaggregate the National Inventory for Phosphoric Acid
Production Across Time Series 3-53
Table 3-12. Summary of Approaches to Disaggregate the National Inventory for Ferroalloys
Production Across Time Series 3-61
Table 3-13. Summary of Approaches to Disaggregate the National Inventory for Aluminum
Production Across Time Series 3-65
Table 3-14. Summary of Approaches to Disaggregate the National Inventory for Magnesium
Production Across Time Series 3-68
Table 3-15. Summary of Approaches to Disaggregate the National Inventory for Lead Production
Across Time Series 3-71
Table 3-16. Summary of Approaches to Disaggregate the National Inventory for Zinc Production
Across Time Series 3-74
Table 3-17. Summary of Approaches to Disaggregate the National Inventory for Semiconductor and
MEMS Manufacturing Across Time Series 3-78
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State vi
-------
Table 3-18. Summary of Approaches to Disaggregate the National Inventory for Fluorinated Heat
Transfer Fluids Across Time Series 3-80
Table 3-19. Summary of Approaches to Disaggregate the National Inventory for Photovoltaics Across
Time Series 3-81
Table 3-20. Summary of Approaches to Disaggregate the National Inventory for Electrical Equipment
Across Time Series 3-87
Table 3-21. Summary of Approaches to Disaggregate the National Inventory for Manufacture of
Electrical Equipment Across Time Series 3-91
Table 4-1. Overview of Approaches for Estimating State-Level Agriculture Sector GHG Emissions 4-1
Table 4-2. Approaches to Estimate Enteric Fermentation Methane Across Time Series 4-2
Table 4-3. Approaches to Estimate Manure Management Methane and N20 Across Time Series 4-4
Table 5-1. Overview of Approaches for Estimating State-Level LULUCF Sector GHG Emissions and
Sinks 5-1
Table 6-1. Overview of Approaches for Estimating State-Level Waste Sector GHG Emissions and Sinks 6-1
Table 6-2. Summary of Approaches to Disaggregate the National Inventory for MSW Landfills Across
Time Series 6-2
Table 6-3. Summary of Availability and Sources for Composting Data 6-7
Table 6-4. Pulp and Paper Permits Manually Removed From or Added to Analysis 6-16
Table B-1. State Level-GHG Data Differences with National GHG Data B-l
vii
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Acknowledgements
The Environmental Protection Agency would like to acknowledge the many individual and organizational
contributors to this report, without whose efforts this report would not be complete. Although the complete
list of researchers, government employees, and consultants who have provided technical and editorial
support is too long to list here, EPA would like to thank some key contributors and reviewers whose work has
significantly improved this year's report.
Within EPA's Office of Atmospheric Protection (OAP), development and compilation of emissions from fuel
combustion was led by Vincent Camobreco. Sarah Roberts (EPA Office of Transportation and Air Quality
(OTAQ)) directed the work to compile estimates of emissions from mobile sources. Work on fugitive methane
emissions from the Energy sector was directed by Julie Powers, Sarah Busch, Vasco Roma and Chris Sherry.
Development and compilation of emissions estimates for the Waste sector were led by Lauren Aepli and
Mausami Desai. John Steller and Kenna Rewcastle directed work to compile estimates for the Agriculture and
the Land Use, Land-Use Change, and Forestry chapters with support from Jake Beaulieu and Alex Hall (EPA
Office of Research and Development (ORD)) on compiling the inventories for C02 and CH4 associated with
flooded lands. Development and compilation of Industrial Processes and Product Use (IPPU) C02, CH4, and
N20 emissions was directed by Amanda Chiu, Vincent Camobreco, Karen Place, and Ga-Young Park.
Development and compilation of emissions of HFCs, PFCs, SFs, and NF3 from the IPPU sector was directed
by Deborah Ottinger, Dave Godwin, and Stephanie Bogle. Cross-cuttingworkwas directed by Mausami
Desai. We thank Bill Irving for general advice, guidance, and cross-cutting review.
We thank the U.S. Forest Service (USFS) (Grant Domke, Brian Walters, James Smith, and Courtney Giebink)
for compiling the state-level inventories for C02, CH4, and N20 fluxes associated with forest land.
We thank the Department of Agriculture's Agricultural Research Service (USDA-ARS) (Stephen Del Grosso)
and the Natural Resource Ecology Laboratory and Department of Statistics at Colorado State University
(CSU) (Stephen Ogle, Bill Parton, Shannon Spencer, Alisa Keyser, Lauren Hoskovec, Ram Gurung, Ryan
Scheiderer, Veronica Thompson, Stephen Williams, and Guhan Dheenadayalan Sivakami) for compilingthe
inventories for CH4 emissions, N20 emissions, and C02 fluxes associated with soils in croplands, grasslands
and settlements.
We thank the National Oceanic and Atmospheric Administration (NOAA) (Nate Herold, Ben DeAngelo and
Meredith Muth), Silvestrum Climate Associates (Stephen Crooks, Lisa Schile Beers, Rebeca Brenes), the
Smithsonian Environmental Research Center (J. Patrick Megonigal, James Holmquist, Jaxine Wolfe and Meng
Lu), and Florida International University (Tiffany Troxler) as well as members of the U.S. Coastal Wetland
Carbon Working Group for compiling inventories of land use change, soil carbon stocks and stock change,
CH4 emissions, and N20 emissions from aquaculture in coastal wetlands. We also thank NOAA's Global
Monitoring Lab (Stephen Montzka and Lei Hu) for information on atmospheric measurements and derived
emissions of HFCs and SFs.
We thank Eastern Research Group for their analytical support and cross-cutting support. Deborah Bartram,
Kara Edquist, and Madison Eaton supported development of emissions estimates for wastewater. Kara
Edquist, Cortney Itle, Amber Allen, Spencer Sauter, Sarah Wagner, and Madison Eaton supported the
development of emission estimates for manure management, enteric fermentation, peatlands (included in
wetlands remainingwetlandsj, and landfilled yard trimmings and food scraps (included in settlements
remaining settlements). Brandon Long, Gopi Manne, and Marty Wolf supported the development of
estimates for natural gas and petroleum systems. Gopi Manne and Tara Stout supported the development of
emission estimates for coal mine methane. Charlie Goff, Hannah Stroud and Chris Salois supported annual
updates to EPA's GHG Data Explorer. Charlie Goff supported production of this methodology report.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
viii
-------
We thank Mollie Averyt, Deborah Harris, Rebecca Ferenchiak, Fiona Wissell, Bikash Acharya, Johanna
Garfinkel, David Landolfi, Emily Carr, Georgia Kerkezis, Isabella Scornaienchi, Katie O'Malley, Maris Welch,
Emily Adkins, Zeyu Hu, Alex Da Silva, Sneha Balakrishnan, Kenny Yerardi, Seth Hartley, Ajo Rabemiarisoa,
Julia Garbow, Angus Dillon, and Becky Petrou O'Rourke for technical support in compiling individual
analyses for specific report chapters including many fluorinated emission sources, including aluminum and
magnesium production. We thank Katie O'Malley for her support in updating the companion crosswalk
document summarizing how methods and default data in the State Inventory Tool (SIT) align with methods
and data described in this report.
We thank the RTI International team: Kate Bronstein, Emily Thompson, Jeff Coburn, Debra Kantner, and Keith
Weitz for their analytical support in development of the estimates of emissions from landfills; Melissa
Icenhour, David Randall, Karen Schaffner, Riley Vanek, Libby Robinson, Matt Hakos, Jeremy Kaelin, Betty
Gatano, and Keegan Waggener for their analytical support in development of IPPU C02, CH4, and N20
emissions; and Karen Schaffner and Haley Key for their analytical support in the development of the
estimates of emissions from fluorochemical production.
We thank the SAIC team: Martin Huppert, Ethan Horvath, and Archana Podduturi for support developing and
maintaining the state-level GHG inventory database.
Data Contributors
Other EPA offices and programs also contributed data, analysis, and technical review for this report. OAP's
Greenhouse Gas Reporting Program staff facilitated aggregation and review of facility-level data for use in the
Inventory, in particular aggregation of confidential business information data. Office of Air Quality Planning
and Standards (OAQPS) with contributions from OTAQ provided analysis for precursor estimates and review
for several of the source categories (i.e., natural gas and petroleum systems) included in this report. ORD
conducted field research and developed estimates associated with flooded lands. The Office of Land and
Emergency Management (OLEM) also contributed analysis and research.
The Energy Information Administration (EIA) and the Department of Energy (DOE) contributed invaluable data
and analysis on numerous energy-related topics. Kevin Nakolan and Michael Francis at EIA provided valuable
information on state-level energy data that are used in fossil fuel combustion estimates. We also thank Chris
Tremper, Soudeh Motamedi, and Ashley Ruocco at the Department of Energy for providing data and
information on emissions of SFs and PFCs from Other Product Use.
We thank the Department of Defense (DOD) (David Asiello, DoD and Matthew Cleaver of Leidos) for
compilingthe data on military bunker fuel use.
Other government agencies have contributed data as well, including the U.S. Geological Survey (USGS), the
Federal Highway Administration (FHWA), the Department of Transportation (DOT), the Bureau of
Transportation Statistics (BTS), the Department of Commerce, the Mine Safety and Health Administration
(MSHA), and the National Agricultural Statistics Service (NASS). We thank the United States Department of
Agriculture's Economic Research Service (USDA-ERS) (Jeffrey Hopkins) for providing data on agricultural
energy use.
Suggested Citation
EPA (U.S. Environmental Protection Agency) (2024) Methodology Report: Inventory of U.S. Greenhouse Gas
Emissions and Sinks by State: 1990-2022. EPA-430-R-24-006. Available online at:
https://www.epa.gov/ghgemissions/methodology-report-inventory-us-greenhouse-gas-emissions-and-
sinks-state-1990-2022.
ix
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
1 Introduction
This report describes methods used to compile the annual publication of U.S. anthropogenic
greenhouse gas (GHG) emissions and sinks disaggregated by U.S. state and consistent with the Inventory of
U.S. Greenhouse Gas Emissions and Sinks (national Inventory hereafter). By April of each year, the U.S.
Environmental Protection Agency (EPA) prepares the official national Inventory, presenting time series
estimates by gas, source/sink, and sector. The latest annual report includes estimates from 1990-2022 and
is available here: https://vwwv.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
This state-level report is complementary publication released annually after the national Inventory report.
EPA recognizes that a number of states have compiled or are developing their own state-level GHG
inventories on a regular or periodic basis. The state-level inventory data presented here should not be viewed as
official data of any state government, and EPA provides users information on where they can find official
state-level data from EPA's website here: https://www.epa.gov/ghgemissions/learn-more-about-official-
state-greenhouse-gas-inventories. In addition, for states where an official inventory is available, EPA's GHG
Data Explorer provides links along with the published state-level data so that when users query information
for a particular state, the link to view the official state inventory will be shown. States themselves may find
this information useful to facilitate comparisons, for quality assurance and quality control (QA/QC), to
supplement and complement existing state efforts, or to serve as official estimates, depending on their own
circumstances and policy needs.
The state-level estimates described in this document are consistent with the national Inventory,
meaning they:
Adhere to international standards, includingthe Intergovernmental Panel on Climate Change (IPCC)
Guidelines and United Nations Framework Convention on Climate Change (UNFCCC) transparency
reporting system. The emissions and removals presented in this report are organized by source and
sink categories within IPCC sectors (energy; industrial processes and product use [IPPU];
agriculture, land use, land-use change, and forestry [LULUCF]; and waste) and their respective
source and sink categories.
Are based on the same methodologies as the national Inventory and reflect the latest
methodological improvements in the national Inventory, including the use of Greenhouse Gas
Reporting Program (GHGRP) data.
Cover the complete time series consistent with the national Inventory, starting with 1990 through the
latest national Inventory year (i.e., 2022).
Cover all anthropogenic sources and sinks, and all seven gases (carbon dioxide [C02], methane
[CH4], nitrous oxide [N20], hydrofluorocarbons [HFCs], perfluorocarbons [PFCs], sulfur hexafluoride
[SFs], and nitrogen trifluoride [NF3]). The completeness and geographic disaggregation of the report
are consistent with the national Inventory, meaning in addition to estimates for states, the methods
also address emissions and removals occurring in the District of Columbia, U.S. territories, and
tribal lands.
Use estimates that were compiled to avoid double counting or gaps in emissions coverage between
states, ensuring that state totals, when summed, will equal totals in the national Inventory. This is
important for those looking for consistent, comparable, and complete state data for analyses and
other purposes where double counting or omissions would be problematic.
1-1
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
This report's chapters are organized by UNFCCC and Paris Agreement reporting sectors1 and their
respective source and sink categories. Domestic and international users alike will recognize this format given
its long-established use by countries for UNFCCC and Paris Agreement reporting. The chapter and category
section titles all include a reference to the corresponding section in the national Inventory report (NIR), such
as NIR Section 3.1, to facilitate understanding national Inventory methods in relation to approaches applied
to allocate national emissions to the state level. For each category, we recommend reading this report in
conjunction with the referenced national Inventory sections. Each category section within a chapter includes
a background discussion, a description of methods/approaches, and a discussion of planned improvements.
The background includes a brief overview of the source or sink category consistent with the national
Inventory. The methods section includes the approach to develop state-level estimates and the gases
covered. The planned improvements indicate areas for improvement identified during this first effort to
disaggregate state-level emissions and sinks.
1.1 Areas Where Differences Between State GHG Inventories and the
EPA State-Level Estimates May Occur
EPA recognizes that there will be differences between EPA's state-level estimates and some inventory
estimates developed independently by individual state governments. Inventories compiled by states may
differ for several reasons and differences do not necessarily mean that one set of estimates is more
accurate, or "correct." EPA has strived to ensure the coverage, methodological, and accounting approaches
are clearly described so users can understand differences with how states may compile their inventories.
The results should be viewed as complementary and supplement existing state data. Differences between
EPA and official state estimates include:
Organization of sectors. EPA has organized estimates by sector and their respective source and
sink categories consistent with the national Inventory and international reporting guidelines.
Standardization of sectors in international reporting allows countries to compare data and supports
cooperation on climate action. States may use alternate organization of data for presenting
emissions and sinks, such as economic sectors, rather than IPCC sectors. Some states may use
IPCC sectors as the basis of their inventory, but allocate some categories differently across sectors,
such as reporting some IPPU categories in the energy sector (e.g., SFsfrom electrical transmission
and distribution). Comparability also depends on similar coverage. The completeness and
geographic disaggregation of the estimates are consistent with the national Inventory, meaning in
addition to estimates for states, the methods also address emissions and removals occurring in the
District of Columbia, U.S. territories, and tribal lands.
Methods and data. In some cases, EPA may be using different methodologies, activity data, and
emissions factors, or may have access to the latest facility-level information through EPA's
Greenhouse Gas Reporting Program (GHGRP). EPA used as a basis, or starting point, either the same
methods or methods based on those used to compile the national-level estimates. States may use
the same methods but use different sources of activity data.
1 The international reporting guidelines under the UNFCCC and the Paris Agreement require reporting of GHG emissions and
removals across five sectors: energy, IPPU, agriculture, LULUCF, and waste. Note that while the UNFCCC reporting
guidelines require using methods from the 2006 IPCC Guidelines for estimating GHG emissions and removals, they require
separate, rather than combined, reporting of emissions and sinks from the agriculture, forestry, and other land use sector as
presented in the IPCC guidelines.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Section 1 Introduction
Accounting approaches. In other cases, states may have adopted different accounting decisions
that differ from those adopted by the IPCC, UNFCCC, and Paris Agreement (e.g., use of different
category definitions and emission scopes consistent with state Laws and regulations). For example,
EPA's approach is to focus on emissions that occur within geographic state boundaries ("Scope 1"),
whereas some states include emissions that are caused by activity within their borders but which
actually occur in other states ("Scope 2 or 3"), or they use consumption-based accounting
approaches. For example, some states include emissions from imported electricity, or electricity
production that occurs outside state boundaries. EPA's use of geographic state boundaries to
allocate emissions is consistent with the methodological framework in the IPCC guidelines.2
Differences in accounting approaches also include differences in the approach to estimating
transportation, cross-border aviation and marine emissions, or treatment of biogenic C02. For
example, EPA does not include biogenic C02 emissions in state energy sector totals because, in
accordance with IPCC methodological guidelines, C02 emissions and removals due to the
harvesting, combustion, and growth of biomass are included in the carbon stock (C stock) changes
of the relevant land use category of the agriculture and LULUCF sectors, where the biomass
originates, and including these emissions in energy sector totals would result in double counting.3
Users of state GHG data should take care to review and understand differences in accounting approaches
to ensure that any comparisons of estimates are based on an equivalent or an apples to apples
comparison of estimates.
Time series. EPA has developed state-level estimates for 1990-2022 consistent with the national
Inventory published in April 2024 and current UNFCCC reporting requirements. States may estimate
emissions and sinks over a different time period based on state goals, designation of different base
years, legislation, and available state data. Some states may not estimate back to 1990 and include
only more recent years. Other states may have previously published estimates for earlier years, but
not recalculated or otherwise updated these estimates in more recent publications despite changes
in methods, activity data, or emissions factors. Similarly, new emissions sources may be added in
recent years but not estimated for more distant years.
Global warming potentials (GWPs). States may use different metrics for C02 equivalency of non-
C02 gases, such as different values for GWPs. Consistent with the national Inventory, in this report
EPA is using 100-year GWPs from IPCC's Fifth Assessment Report (AR5) to calculate C02
equivalency of non-C02 emissions, as required in reporting annual inventories to the UNFCCC and
the Paris Agreement. EPA shifted to using 100-year GWPs from AR5 in 2023. Recent decisions4 under
the UNFCCC and the Paris Agreement require members of the Conference of Parties to use 100-year
GWP values from AR5 for calculating C02 equivalents in their national reporting (IPCC 2013) by the
2 Per the 2006 IPCC Guidelines, national inventories include GHG emissions and removals taking place within national
territory and offshore areas over which the country has jurisdiction with some minor exceptions. For example, one exception
is "CO2 emissions from road vehicles should be attributed to the country where the fuel is sold to the end user." See Volume
1, Chapter 8, Section 8.2.1, on Coverage, available online at https://www.ipcc-
nggip.iges.or.jp/pub[ic/2006g[/pdf/1 Vo[ume1/V1 8 Ch8 Reporting Guidance.pdf.
3 See Q2-10 of Frequently Asked Questions on general guidance and other inventory issues: https://www.ipcc-
nggip.iges.or.jp/faq/faq.html.
4 See paragraphs 1 and 2 of the decision on common metrics adopted at the 27th UNFCCC Conference of Parties (COP27),
available online at https://unfccc.int/sites/default/files/resource/cp2022 10a01 adv.pdf. and paragraph 37 of the Annex
to decision 18/CMA.1, available online at https://unfccc.int/sites/default/files/resource/CMA2018 03a02E.pdf.
1-3
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
end of 2024. This requirement reflects updated science and ensures that national GHG inventories
reported by all nations are comparable.
1.2 Institutional Arrangements for Compiling State-Level Inventory
Estimates
In preparing the state-level inventory, EPA took advantage of existing data arrangements used to
compile the national Inventory (see Chapter 1.2 of the national Inventory). EPA acknowledges the additional
contributions from the U.S. Department of Agriculture's U.S. Forest Service (USDA-USFS) and National
Oceanic and Atmospheric Administration (NOAA). USDA-USFS has ongoing efforts to prepare state-level
data5 to track emissions and sinks from land use and land use change in forested lands and settlement
lands. NOAA has compiled the state-level emissions and removals from coastal wetlands. EPA also
acknowledges additional effort from USDA's National Agricultural Statistics Service (NASS) and Office of
Chief Economist (OCE) for providing state-level data on energy use in agriculture and from the Department of
Energy's Energy Information Administration (EIA) for providing state-level energy use data. Finally, EPA
acknowledges contributions and investments from USDA-OCE that will facilitate addressing some of the
planned improvements outlined in Chapters 4 and 5 of this report.
EPA also collects GHG emissions data from individual facilities and suppliers of certain fossil fuels and
industrial gases through its GHGRP.6 The GHGRP does not provide full economywide coverage of total
annual U.S. GHG emissions and sinks (e.g., the GHGRP does not collect data on emissions from the
agricultural, land use, and forestry sectors), but it is an important input to the calculations of state-level
estimates in the national Inventory. In general, the threshold for reporting is 25,000 metric tons or more of
C02 equivalent per year. Facilities in most source categories subject to GHGRP began reporting for the
reporting year (RY) 2010, while additional types of industrial operations began reporting for RY 2011. When
incorporating these data from GHGRP, consistent with the national Inventory, EPA considers good practice
guidance from the 2019 Refinement to the 2006IPCC Guidelines (Volume 1, Chapter 2)7 and IPCC's Use of
Facility-Specific Data in National GHG Inventories technical bulletin8 to ensure completeness, time series
consistency, and transparency in state-level methods and associated estimates.
Data presented in this state-level inventory report and EPA's GHGRP are complementary. As discussed
across this report, in addition to annual emissions information, the GHGRP also provides other annual
information such as activity data and emissions factors that can improve and refine state-level trends over
time. More information on the relationship between GHGRP and the national Inventory is available online at
https://www.epa.gov/ghgreporting/greenhouse-gas-reporting-program-and-us-inventorv-greenhouse-gas-
emissions-and-sinks.
1.3 Methods Overview
In developing the state-level estimates consistent with the national Inventory, EPA used as a basis, or
starting point, the same methods or methods based on those used to compile the national-level estimates.
From this starting point, there were three different approaches taken to arrive at state-level estimates:
5 https://research.fs.usda.gov/treesearch/66Q35
6 https://www.epa.gov/ghgreporting
7 https://www.ipcc-nggip.iges.or.jp/public/9019rf/pdf/1 Volume1/19R V1 ChO? DataCollection.pdf
8 https://www.ipcc-nggip.iges.or.jp/public/tb/TFI Technical Bulletin 1.pdf
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
1-4
-------
Section 1 Introduction
Approach 1. Estimates were built by applying national methods directly to geographically
disaggregated data (at state or finer level). For example, estimates of forest land remaining forest
land and of lands converted to forest land are built from existing data sets that already disaggregate
to the state level (see Section 5.1.1). Also, portions of fossil fuel combustion emissions were based
on the same approach as the national estimates using state disaggregated energy consumption data
(see Section 2.1.1).
Approach 2. Estimates were disaggregated from national-level estimates using geographic proxies
or other indicators (e.g., population, production capacity, GHGRP). This approach was used for
categories where the type of state data used in Approach 1 were not available or were incomplete.
For example, Approach 2 is used to estimate state-level emissions from other process uses of
carbonates (see Section 3.1.4) where state-level population is used as a proxy to allocate national
emissions. A key factor in Approach 2 is how well emissions correlate with proxies, and where
multiple options exist, how to choose among them.
Hybrid approach. Under this approach, estimates used a combination of Approach 1 and Approach
2 methods over the time series because data availability limited the use of Approach 1 for all years of
the time series. For example, some estimates may use EPA's GHGRP, which began collecting data in
2010, as a basis for national- and state-level estimates. For these categories, EPA uses Approach 1
for 2010-2022 and uses Approach 2 for earlier years of the time series to arrive at state-level
estimates, using IPCC guidance to ensure consistency over the time series to the extent possible.
For example, the Hybrid approach is used to estimate state-level C02 and PFC emissions from
aluminum production (see Section 3.3.3).
Across this report, in addition to a sector-level summary, under each category, EPA has indicated the
approach used to disaggregate national estimates to the state level. Where appropriate for explaining
methods used under Approach 2 or the Hybrid approach, EPA has included equations to enhance
understanding of the implementation of disaggregation methods. EPA has also included data appendices to
provide underlying data to estimate emissions and sinks.
1.4 Summary of Updates Since Previous Report
Each year, many emission and sink estimates in the national Inventory are recalculated and revised, as
efforts are made to improve the estimates through the use of better methods and/or data with the goal of
improving inventory quality and reducing uncertainties, including the transparency, completeness,
consistency, and overall usefulness of the report. The same is the case with state-level estimates where
updates were made to improve inventory quality. In general, when methodological changes have been
implemented, the previous national Inventory's time series (i.e., 1990-2021) was recalculated to reflect the
change. Note that the most common reason for recalculating national GHG emission estimates is to update
recent historical activity data. Changes in historical data are generally the result of changes in statistical data
supplied by other U.S. government agencies, and do not necessarily impact the entire time series. The table
below reflects more significant changes beyond routine data updates (e.g., use of a new data source, change
in national Inventory method, changes in disaggregation method).
A summary of methodological changes and historical data updates made to the state-level data is
presented below by category. Table 1-1 notes whether changes are due to refinements in the national
Inventory methods and data, including new categories, and/or due to an update that refined the approach
and data used to disaggregate national estimates to the state level. Note that when category-level changes in
absolute state-level emissions or removals for a state between this version and the previous state report are
due to recalculations and improvements implemented in the national Inventory, changes are indicated only
1-5
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
in the national-Level column in Table 1-1 below, as the approach to disaggregation of the updated national
estimates to the state level remains unchanged. Categories not listed had no changes for either the national
or state-level estimates. See the recalculations sections of each category for more detail on the updates
within this report.
Table 1-1. Category Estimates Updated Since Release of Previous Inventory by U.S. State
IPCC Sector
Category
Changes to Inventory (i.e., Refined
Method/Data or New Category)
National-Level
State-Level
E
Fuel Combustion
E
Non-Energy Use of Fuels
E
Coal Mining
E
Abandoned Mines
E
Petroleum Systems (updated data sources,
revision of methodology to use basin-level data
for certain segments)
E
Natural Gas Systems (updated data sources,
revision of methodology to use basin-level data
for certain segments)
E
Abandoned Oil and Gas Wells (updated data
sources)
I
Glass Production
I
Other Process Uses of Carbonates (includes
new subcategories ceramics and non-
metallurgical magnesia productions)
I
C02 Emissions from C02 Consumption
I
Ammonia Production
I
C02 from Urea Use
I
Adipic Acid
I
C02 from Carbide Production
I
Titanium Dioxide Production
I
Petrochemicals
I
Phosphoric Acid Production
I
Nitric Acid Production
I
Caprolactam, Glyoxal and Glyoxylic Acid
Production
I
Iron and Steel Production
I
Magnesium Production
I
Lead Production
I
Zinc Production
I
Electronics Industry
I
ODS Substitutes
I
Electrical Transmission and Distribution
I
HCFC-22 Production
I
Production of Fluorochemicals Other Than
HCFC-22 (new category)
I
SFs and PFCs from Other Product Use (new
category)
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
1-6
-------
Section 1 Introduction
Changes to Inventory (i.e., Refined
IPCC Sector
Category
Method/Data or New Category)
National-Level
State-Level
I
N20 from Product Use
A
Enteric Fermentation
A
Manure Management
A
Rice Cultivation
A
Agricultural Soil Management
A
Liming
A
Urea Fertilization
A
Field Burning of Agricultural Residues
L
Forest Land Remaining Forest Land
L
Land Converted to Forest Land
L
Cropland Remaining Cropland
L
Land Converted to Cropland
L
Grassland Remaining Grassland
L
Land Converted to Grassland
L
Wetlands Remaining Wetlands
L
Land Converted to Wetlands
L
Settlements Remaining Settlements
(subcategory N20 from soils, subcategory
landfilled yard trimmings and food scraps)
L
Land Converted to Settlements
W
Landfills
W
Composting
W
Anerobic Digestion at Biogas Facilities
W
Wastewater Treatment and Discharge
E = Energy Sector; I = Industrial Processes and Product Use; A = Agriculture; L = Land Use Change, Land Use Change and
Forestry; W = Waste
1.5 QA/QC Procedures
In disaggregating emissions and sinks from the national Inventory, EPA implemented QC procedures
during the compilation process to ensure quality, transparency, and credibility of the state GHG data. EPA
implemented general QC procedures adapted from the existing QA/QC plan9 for the national Inventory to
ensure that data processing and application of methods could easily identify and correct errors (i.e.,
data/unit transcription, computation, and trend checks). EPA also implemented additional category-specific
QC procedures to assess disaggregation approaches (e.g., comparisons with other data such as available
state GHG inventories) to further review methods and resulting estimates, including comparing category
estimates to available state GHG inventories and comparing the sum of state estimates to national
estimates. When additional category-specific QC procedures were implemented, the procedure and findings
are discussed in the respective category section.
EPA also implemented QA procedures outlined by EPA and IPCC as QAgood practices (i.e., external
review by experts not directly involved in compiling the data). EPA conducted a peer review in fall 2021 and is
sharing this methodology report for an annual 30-day state expert review in summer 2024. Both reviews are
9 See the introduction (Section 1.6) and Annex 8 of the national Inventory for more information on the QA/QC plan, available
online at https://www.epa.gov/ghgemissions/inventory-us-greenhouse-gas-emissions-and-sinks-1990-90??.
1-7
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
described further below. The QA/QC findings also informed the overall improvement planning, and specific
improvements are noted in the planned improvements sections of respective categories.
1.5.1 State Expert Review
Technical staff from each state (e.g., environmental agencies, other state agencies, institutions) were
provided with an opportunity to review the draft data and a draft of this methodology report from July 18
August 19, 2024. The methodology report and state-level estimates were shared with state experts from all
50 U.S. states and the District of Columbia for review.
EPA gratefully acknowledges all the state experts for their review. EPA asked state experts for feedback
on this methodology report, its data appendices, and the resulting estimates.
EPA received 2 sets of technical comments including 12 unique comments from state experts.
Responses to comments from the latest reviewwill be published at
https://www.epa.gov/ghgemissions/state-ghg-emissions-and-removals. and indicate how EPA addressed
comments. See category-specific planned improvement discussions throughout this report reflecting
updates planned for future publications of these data.
1.5.2 Peer Review
The methodology report and the resulting state-level estimates for the 1990-2019 data were
independently peer reviewed from September 17 to November 1, 2021. Seventeen external experts
participated in a process independently coordinated by RTI International and an EPA peer-review
coordinator.
EPA gratefully acknowledges all the peer reviewers for their useful comments. The peer review report
and responses from EPA are available online here: https://www.epa.gov/ghgemissions/state-ghg-emissions-
and-removals. The information and views expressed in this report do not necessarily represent those of the
peer reviewers, who also bear no responsibility for any remaining errors or omissions. Details describing this
review can be found below. Peer review of the report followed the procedures in EPA's Peer Review
Handbook, 4th edition (EPA/100/B-15/001) for reports that do not provide influential scientific information.
The review was managed by a contractor under the direction of a designated EPA peer review leader,
who coordinated the preparation of a peer review plan, the scope of work for the review contract, and the
charge for the reviewers. The peer review leader played no role in producing the draft report. Each sectoral
reviewer was charged with reviewing the Introduction, the sector or subsector of the report relevant to their
expertise, resulting estimates, and data appendices. Peer reviewers were charged with ma king specific
comments and edits as well as providing a written response to a set of general and category-specific charge
questions. The EPA author team then responded to and addressed all comments from the peer reviewers in a
written summary and revised the report accordingly.
1.6 Uncertainty
EPA has not assessed state-specific or category-level quantitative uncertainties for the activity data and
other parameters used to estimate state-level emissions and removals for this current publication but has
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
1-8
-------
Section 1 Introduction
included qualitative information on how uncertainties compare to those assessed quantitively for each
category in the national Inventory,10
The uncertainties of state-level emissions estimates are generally expected to be comparable to or
higher than the uncertainties of national-level emissions estimates for two reasons. First, where emissions
are estimated at the national level and then allocated to states based on proxy or surrogate data and
indicators other than those used to estimate emissions (i.e., where Approach 2 is used), uncertainties in the
relationship between the allocation indicator and the emissions increase the uncertainty of the allocation.
For example, where total U.S. production is multiplied by an emissions factor to obtain total national-level
emissions, but production capacity rather than production is used to allocate the U.S. emissions to facilities
and states, variation in each facility's capacity utilization will not be reflected in the estimates, increasing
their uncertainty. Second, for some categories where state-level emissions are estimated using the same
facility-based methods as are used for national-level emissions (i.e., where Approach 1 is used), state-level
uncertainties will generally be higher than national-level uncertainties (in percentage terms), assuming the
uncertainties in the estimates for each facility and state are independent of each other. For example, EPA
estimates the uncertainties in emissions from aluminum production at individual smelters to be +6/-6%,
+16/-16%, and +20/-20% for C02, perfluoromethane, and perfluoroethane emissions, respectively. When
propagated to the national level across the seven smelters that operated in 2021, these uncertainties decline
to -2%/+3% for C02 and +8/-8% for PFCs. Since the states with aluminum production each have just one to
two smelters, the uncertainties in the state-level emissions will be closer to the uncertainties in the
emissions for individual smelters than to the uncertainties in the national-level emissions.
For more information on uncertainties with national-level GHG estimates, see Section 1.7 of the
Introduction chapter to the national Inventory. Category-specific uncertainties for national estimates are
included in the category-specific methodological discussions across the national Inventory report.
1.7 Planned Improvements
Across this report, per EPA's QC and feedback from the previous peer and state reviews, EPA has
outlined areas for improving future annual publications of these data at the category level across the report.
Based on feedback, EPA continues to prioritize the following cross-cutting improvements for future annual
publications of these data:
Finalize state-level key category analyses consistent with IPCC guidance and international reporting
guidelines to help identify categories that are more significant at the state level and publish in fall of
2024.
1.8 References
IPCC (Intergovernmental Panel on Climate Change) (2013) Climate Change 2013: The Physical Science
Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on
Climate Change. T.F. Stocker, D. Qin, G.-K. Plattner, M.B. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y.
Xia, V. Bex, and P.M. Midgley (eds.). Cambridge University Press.
10 Within the forest [and remaining forest [and and [ands converted to forest [and categories, USFS has quantified
uncertainties for state-[eve[ estimates for net CO2 flux from forest ecosystem carbon poo[s and non-CCb emissions from
forest fires that are the basis for the estimates a[so in the nationa[ Inventory. The quantified uncertainties are avai[ab[e in
USDA-USFS Resource Bulletin WO-101, avai[ab[e at https://www.fs.usda.gov/research/treesearch/66035.
1-9
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
2 Energy (NIR Chapter3)
For this methodology report, energy emissions are broken into two main categories: emissions
associated with fuel useincluding fossil fuel combustion (FFC) and nonenergy use (NEU)and fugitive
emissions mainly from fuel production. The energy emissions presented here include some categories that
are not added to energy sector totals in the national Inventory but are instead presented as memo items,
including international bunker fuels (IBFs)11 and biomass emissions,12 consistent with UNFCCC reporting
guidelines. This approach directly affects state-level energy sector estimates and, in some cases, may
account for differences with official estimates published by individual state governments. For more
information on energy sector emissions, see Chapter 3 of the national Inventory. Table 2-1 summarizes the
different approaches used to estimate state-level energy emissions and completeness across states.
Geographic completeness is consistent with the national Inventory. The sections below provide more detail
on each category.
Table 2-1. Overview of Approaches for Estimating State-Level Energy Sector GHG Emissions
Category
Gas
Approach
Geographic Completeness8
FFC
co2,
ch4,
n2o
Hybrid approach
Approach 1 used for most
fuels and sectors
Approach 2 proxy data used
to allocate national totals for
some fuels and sectors
Includes emissions from all
states, the District of Columbia,
tribal lands, and territories (i.e.,
American Samoa, Guam, Puerto
Rico, Northern Mariana Islands,
U.S. Virgin Islands, and other
outlying minor islands) as
applicable.
NEUs of Fossil Fuels
co2
Approach 2
Includes emissions from all
states, the District of Columbia,
tribal lands, and territories (i.e.,
American Samoa, Guam, Puerto
Rico, Northern Mariana Islands,
U.S. Virgin Islands, and other
outlying minor islands) as
applicable.
Geothermal Emissions
co2
Approach 2
Includes emissions from all
states, the District of Columbia,
and tribal lands as applicable.8
Incineration of Waste
co2,
ch4,
n2o
Hybrid approach
2011-2022: Approach 1
1990-2010: Approach 2
Includes emissions from all
states, the District of Columbia,
and tribal lands as applicable.8
IBFs (memo item)
co2,
ch4,
n2o
Approach 2
Includes emissions from all
states, the District of Columbia,
and tribal lands as applicable.
11 Emissions from IBFs are not included specifically in summing energy sector totals. The values are presented for
informational purposes only, in line with the 2006IPCC Guidelines and UNFCCC reporting obligations.
12 Emissions from wood biomass, ethanol, and biodiesel consumption are not included specifically in summing energy
sector totals. The values are presented for informational purposes only, in line with the 2006 IPCC Guidelines and UNFCCC
reporting obligations. Net carbon fluxes from changes in biogenic carbon reservoirs are accounted for in the estimates for
LULUCF.
2-1
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Category
Gas
Approach
Geographic Completeness8
Wood Biomass and
C02
Approach 2
Includes emissions from all
Biofuels Consumption
states, the District of Columbia,
(memo item)
and tribal lands as applicable.8
Coal Mining
ch4
Approach 1: Active Underground
Mines
Approach 1: Surface Mining and
Post-mining Activities
Includes emissions from all
states and the District of
Columbia as applicable.8
Abandoned
ch4
Hybrid approach
Includes emissions from all
Underground Coal
1990-2019: Approach 2
states and the District of
Mines
2020-2022: Approach 1
Columbia as applicable.8
Petroleum and Natural
co2,
Hybrid
Includes emissions from all
Gas Systems
ch4,
Both approaches are used.
states, the District of Columbia,
n2o
Approach 1 is used for
basin-level sources.
Approach 2 is used for non-
basin-level sources.
and territories (i.e., American
Samoa, Guam, Puerto Rico, U.S.
Virgin Islands, Northern Mariana
Islands, and other outlying minor
islands) as applicable.8
Abandoned Oil and
co2,
Approach 1
Includes emissions from all
Gas Wells
ch4
states, the District of Columbia,
and territories (i.e., American
Samoa, Guam, Puerto Rico, U.S.
Virgin Islands, Northern Mariana
Islands, and other outlying minor
islands) as applicable.8
a Emissions are not likely occurring in U.S. territories; due to a lack of available data and the nature of this category,
territories not listed are not estimated.
2.1 Emissions Related to Fuel Use
This section presents the methodology used to estimate the fuel use portion of emissions, which
consists of the following sources:
FFC (C02, CH4, N20)
Carbon emitted from NEUs of fossil fuels (C02)
Geothermal emissions (C02)
Incineration of waste (C02, CH4, N20)
IBFs(C02,CH4, N20)
Wood biomass and biofuels consumption (C02)
2.1.1 Fossil Fuel Combustion (NIR Section 3.1)
2.1.1.1 Background
Emissions from FFC include the GHGs C02, CH4, and N20. C02is the primary gas emitted from FFC and
represents the largest share of U.S. total GHG emissions. The methods to estimate C02 emissions from FFC
and the methods to estimate CH4 and N20 emissions from stationary and mobile combustion rely in large
part on the same underlying data. However, there are some differences; therefore, the methods used to
estimate C02 and non-C02 emissions are presented separately.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-2
-------
Section 2 Energy (NIR Chapter 3)
2.1.1.2 Methods/Approach
The approach for determining national-Level FFC emissions is based on multiplying emissions factors
times activity data on fuel consumption. The activity data on fuel consumption were taken from national-
level energy balances prepared for ElA's Monthly Energy Review (MER) estimates (EIA 2024a). EIA prepares
national-level energy statistics that consider energy production imports/exports and stock changes to
determine energy supply/consumption. The fuel consumption information is used as a starting point for
determining emissions.13 The approach starts with determining fuel use by fuel type because different types
of fuels have different carbon content (C content) and therefore different emissions factors. The information
is also broken out by energy-consuming sectors of U.S. society to provide more detail and information on
trends; the sectors included are residential, commercial, industrial, transportation, and electric power. Data
from U.S. territories were also included in the analysis per international reporting requirements. Several
adjustments were made to the data to account for fuel use and emissions that are either excluded or
reported in other parts of the national Inventory, as shown in Figure 2-1.
Figure 2-1. Adjustments to Energy Consumption for Emissions Estimates
Determine Fuel
Use by Type and
Sector
Subtract Fuel Use
Accounted for in
IPPU
Subtract Biofuels
and Exported C02
Adjust Sectoral Allocation
of Distillate Fuel Oil and
Motor Gasoline
Subtract Consumption
for Non-Energy Use
(NEU)
Subtract
International
Bunker Fuel (IBFs)
Calculate C02
Emissions
Residential Sector
Wood
Adjust dlesel fuel 1
Energy Use
Commercial Sector
Wood&
Adjust diesel fuel
and gasoline
Energy Use
ethanol
_»
Coking & other coal, natural gas,
asphalt, diesel fuel, HGL,
lubricants, misc. prod, naphtha,
otheroil, pentanes plus, pet coke,
still gas, special naphtha and waxes
Industrial Sector
U
Coking & other coal,
natural gas, dlesel &
Wood&
H
Coal linked to
-
Adjust dlesel fuel
and gasoline
Energy Use
residual fuel
n
ethanol
exported C02
Transportation
Sector Energy Use
Electric Power
Sector Energy Use
CH< emissions
accounted for In
the IPPU sector
Adjust dlesel fuel
and gasoline
Lubricants and misc.
rCOj emissions
excluded from the
Inventory
COj emissions
accounted for as
NEU
COj, CH4 and W
emissions reported
as a memo item
d NjO
ported
item
Calculate non-C02
Emissions
I Counted as part of FFC
| Counted elsewhere
f Reported as memo items
I Not part of Inv. totals
This section describes how national-level estimates for FFC were disaggregated to the state level for the
following separate sources:
FFC C02
Stationary non-C02 emissions
Mobile non-C02 emissions
13 The energy balance data include information on all energy sources. Emissions estimates exclude data on non-emitting
sources (e.g., nuclear, wind, solar); however, those data are considered when looking at overall energy use and efficiency.
2-3
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
This section also discusses how energy use data were broken out at the state Level as part of the
adjustments noted in Figure 2-1 and then used to report emissions elsewhere in the national Inventory.
Emissions from energy use that were excluded from FFC are discussed in other sections of the report as
follows:
For energy used in the IPPU sector, see Section 3.
For biofuel use, see Section 2.1.6.
For NEUs of fuels, see Section 2.1.2.
For IBFs, see Section 2.1.5.
Disaggregating FFC emissions to the state level largely followed the same process and energy
consumption data that are used at the national level. However, in several instances, the data used to
develop national estimates are not available at the state level, and additional steps were needed to distribute
national-level emissions across the states while maintaining consistency with national-level totals.
Therefore, Approach 3, the Hybrid approach as described in Section 1.3 of the Introduction chapter, was
used to determine state-level emissions for FFC, including some data that were directly used in the national
Inventory and some surrogate data as discussed in the following sections.
2.1.1.2.1. FFC C02 State-Level Breakout
C02 emissions from FFC at the national level are estimated with a Tier 2 method described by the IPCC
in the 2006 IPCC Guidelines for National Greenhouse Gas Inventories (IPCC 2006). As discussed above, this
method is based on multiplying activity data on fuel use (that have been adjusted to allocate and report data
consistent with UNFCCC reporting guidelines and avoid double counting) by emissions factors to determine
emissions.
Determining adjusted fuel use activity data is based on the seven steps discussed in Table 2-2 below.
The result of these seven steps is an adjusted amount of fuel use activity data that are then used to
determine FFC C02 emissions. In Appendix A to this document (included as separate Excel files), the
"National 2022 FFC C02" Tab provides more details on an example of the adjustments made to the national-
level energy use data to determine adjusted fuel use activity data for 2022. Three additional steps (Steps 8-
10 in Table 2-2) are required to determine C02 emissions in the national Inventory, also discussed below.
Ideally, to determine state-level FFC C02 emissions estimates, the same approach could be used, and
adjusted energy use, as shown in the "National 2022 FFC C02" Tab of Appendix A, could be developed for
each state. However, the national-level emissions were developed based on multiple factors and inputs,
some of which were not available or readily published at the state level. Therefore, a Hybrid approach was
taken where state-level data were used when available. In cases where state-level data were not available,
national-level estimates were used with available surrogate data to determine state-level percentages of
each fuel use. Table 2-2 shows a high-level comparison of the different data sources used for the different
steps to determine national-level and state-level estimates.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-4
-------
Section 2 Energy (NIR Chapter 3)
Table 2-2. Comparison of Approaches/Data Sources Used to Determine FFC Emissions
Calculation Step
National-Level Estimates
State-Level Estimates
Determine Activity Data
Step 1: Determine Total
Fuel Consumption by
Fuel Type and Sector
Based on EIA MER
Based on EIA SEDS (adjusted to match
national totals as applicable)
Step 2: Subtract Uses
That Are Accounted For
in the IPPU Sector
Taken from industry data or based
on national-level emissions
National-level data allocated to states
based on state-level emissions estimates
for each IPPU category in question as
calculated in Chapter 1
Step 3: Adjust for
Biofuels and Petroleum
Denaturant
Based on national-level data from
EIA MER
Not needed (see Step 5)
Step 4: Adjust for C02
Exports
Based on industry data and
Canadian import data
Based on industry data and Canadian
import data
Step 5: Adjust Sectoral
Allocation of Diesel Fuel
and Gasoline
Based on bottom-up
transportation sector data on fuel
use by vehicle type
National-level data (already excluding
biofuels) allocated to states based on
state-level fuel use data (not vehicle type
specific)
Step 6: Subtract
Consumption for NEUs
Based on data from EIA MER
National-level data allocated to states
based on SEDS
Step 7: Subtract
Consumption of IBFs
Based on data from Federal
Aviation Administration (FAA) and
other national-level sources
National-level data allocated to states
based on SEDS and other sources
Calculate C02 Emissions
Step 8: Determine the
Carbon Content of Each
Fuel Consumed
National-level average C content
values
National-level average C content values
Step 9: Estimate C02
Emissions
Multiply C content by activity data
and oxidation percentage
Multiply C content by activity data and
oxidation percentage
Step 10: Allocate
Transportation
Emissions by Vehicle
Type
Allocated at the national level
based on data from Step 5
Not currently done
The following discussion details what data were used for each step in Table 2-2 to determine national-
and state-level FFC emissions. Appendix A, Table A-1 in the "State FCC C02"Tab, provides more details on
where state-level data were used directly and where other data were used to make adjustments to
disaggregate national numbers across fuel types and sectors for each of the steps identified.
2.1.1.2.2. Step 1: Determine Total Fuel Consumption by Fuel Type and Sector
As discussed above, national-level data on fuel supply/consumption comes from ElA's MER. Because
not all fuel supplied/consumed directly results in GHG emissions, or it could be included as part of other
emissions reporting in the national Inventory, adjustments have to be made as shown above in Table 2-2 and
described in the following steps. State-level energy data are available from ElA's State Energy Data System
(SEDS). Those data are broken out by fuel type and sector (residential, commercial, industrial,
transportation, and electric power) and are available for the years 1960-2022 (EIA 2024b). SEDS estimates
energy consumption using data from surveys of energy suppliers that report consumption, sales, or
distribution of energy at the state level. Most SEDS estimates rely directly on collected state-level
consumption data. For example, SEDS uses state-level sales survey data and other proxies of consumption
2-5
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
to allocate the national petroleum product-supplied totals to the states. The sums of the state estimates
equal the national totals as closely as possible for each energy type and end-use sector, and energy
consumption estimates are generally comparable to the national statistics in ElA's MER because both data
sets rely largely on the same survey returns for producers and consumers.
However, the totals across all states (and the District of Columbia) from SEDS do not always match the
U.S. total energy data used in the national Inventory, which is based on the EIA February 2024 MER estimates
(EIA 2024a). The main differences are for coal and natural gas and primarily in the industrial sector. For coal,
there are differences in both energy content and short tons, but the differences are not consistent across
time or sectors. For natural gas, the difference is mainly in the energy content. The reason for the differences
is that SEDS uses state-level energy content conversion factors for coal and natural gas, while the MER uses
national-level conversion factors. These different calculations sometimes cause the sums of the SEDS states
to be different than the MER values. Although the percentage differences are not large (max 5.2% for coal and
1.4% for natural gas in the industrial sector), they cause noticeable differences when comparing emissions
totals across all states to national totals, especially by sector.
The petroleum categories generally line up well across state-level and national totals. There are only
minor differences in petroleum coke, mainly in the industrial sector. For petroleum coke, there are
differences in energy content and barrels, but the difference in energy content appears in 2004, which is
when petroleum coke heating values were changed from a constant value to values based on marketable
and catalyst coke. Again, this difference is because of different national-level and state-level conversion
factors. Since 2004, the MER has used an annual national-level "quantity-weighted" average petroleum coke
conversion factor (instead of a fixed factor). SEDS applies the marketable and catalyst coke conversion
factors to the state-level consumption of each petroleum coke category within each state.
For diesel fuel and gasoline, the totals generally line up, but there are differences across sectors. These
differences are discussed in Step 5 below.
In addition to the differences in gasoline and diesel fuel across sectors over the time series, there are
also differences in some petroleum fuels across sectors, specifically in 2022. This is because the SEDS
represents the latest data from EIA in terms of sector breakouts that were not reflected in the national
Inventory 2022 values that relied on older EIA data. Again, the totals for the fuels line up, but there are
differences across sectors. The updated SEDS data were used in the state-level breakout because they
represent the latest data available. This results in differences in 2022 results across sectors for the state
totals versus the national Inventory. However, the national Inventory numbers will be updated to match the
2022 SEDS data during the next national Inventory cycle.
Furthermore, some of the fuel use reported in SEDS is different from the reporting in the national
Inventory. For example, natural gas reported in SEDS includes supplemental gas, which is included in the
national Inventory under the primary fuel used to make the supplemental gas, so including supplemental gas
in state level results would result in double counting. Gasoline and distillate fuels in SEDS include biofuels,
which are reported separately in the national Inventory. These differences make it difficult to use the SEDS
data directly to determine state-level fuel use data, in a manner consistent with the national Inventory.
Therefore, the following approach was used in determining fuel use by type by sector at the state level:
If SEDS data totals matched the national totals and there were no further adjustments needed (as
per Steps 2-7), the SEDS data were used directly to represent state-level energy use.
For fuels where the SEDS totals did not match the national totals (i.e., coal, natural gas, and
petroleum coke), fuel use in each sector was adjusted to match the national totals used in the
national Inventory. This calculation was based on the percentage of each fuel used in each state
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-6
-------
Section 2 Energy (NIR Chapter 3)
from the SEDS data. For the industrial sector, this adjustment was made after subtracting for uses in
the IPPU sector (see Step 2 below).
For other fuels where sector totals did not match up (e.g., gasoline and diesel fuel), totals for each
fuel type were generally taken from the national Inventory (see Step 5), and the SEDS data or other
proxy data sources were used to determine state-level percentages of each fuel use.
This approach generally results in state-level energy use data that are consistent with national totals
used in the national Inventory. More details on further adjustments made during the different steps are
discussed below.
Appendix A has details on how the SEDS data were adjusted to determine state-level energy use by fuel
type and sector. Tables A-2 through A-6 in the "FFC C02 Residential" Tab describe the residential sector
adjustments. Tables A-9 through A-13 in the "FFC C02 Commercial" Tab describe the commercial sector
adjustments. Tables A-44 through A-47 in the "FFC C02 Industrial" Tab describe the industrial sector
adjustments for petroleum coke and HGL; the remaining industrial sector adjustments are described further
in Steps 2 and 3 below. Tables A-50 and A-51 in the "FFC C02 Transportation" Tab describe the
transportation sector adjustments. Tables A-52 through A-56 in the "FFC C02 Electricity" Tab describe the
electricity production sector adjustments.
2.1.1.2.3. Step 2: Subtract Uses That Are Accounted for in the IPPU Sector
In the national Inventory, portions of fuel consumption data for several fuel categories (coking coal,
other coal, natural gas, residual fuel, and distillate fuel) are reallocated from the energy sector to the IPPU
sector because these portions were consumed as raw materials during nonenergy-related industrial
processes. As per IPCC Guidelines that distinguish between the energy and IPPU sector reporting, emissions
from fuels used as raw materials are presented as part of IPPU and are removed from the energy use
estimates (IPCC 2006, Volume 3, Chapter 1). Portions of fuel use were therefore subtracted from the
industrial sector fuel consumption data before determining combustion emissions. Note that other
adjustments were also made to the NEU calculations to reflect energy use accounted for under IPPU; see
Step 6 and the NEU emissions discussion below.
The adjustments vary over time and represent from about 4% to 8% of total unadjusted industrial sector
energy use, as shown in Figure 2-2.
2-7
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Figure 2-2. Adjustments Made to Industrial Sector Energy Use to Account for Emissions Reported in IPPU
1,800
M 1,600
£ 1.400
9%
E
o
1,200
¦a 1,000
a>
u
2
+-»
-Q
3
to
V)
3
¦M
CQ
800
600
400
200
OHfNrO^LDONCOffl
CT>Cr>Cr>Cr)(T>Cr>Cr>
(jicDo^fDcrio^o^cricTicr)
OH(NrO^LniDNCOCnOHfNr0^miDNCO(nOH(N
0000000000«-HiItItIiIiI H H (N (N tN
ooooooooooooooooooooooo
CN(NrM(N(N(NCNrMCN(N
(/>
Amount Removed
% of Unadjusted Industrial Sector Total
Adjustments for each fuel type were made based on industry data or assumptions about fuel use based
on emissions reported under IPPU. The following bullets discuss the assumptions made regarding the
different industrial sector fuel types at the national and state levels to reflect their use in IPPU:
Coking coal. Coking coal is used to make coke that, in turn, is used in industrial processes. The
national total amount of coking coal used in IPPU was back-calculated based on the amount of
coking coal needed to make the coke used as input to iron and steel (l&S) and lead and zinc
production (approximately 94% is used in l&S). National-level coke use in l&S production was based
on industry data that are not available at the state level. Coke used in lead and zinc production was
based on the amount of carbon emitted from the processes and is also not available specifically at
the state level. Therefore, the national total amount of coking coal used in IPPU was allocated per
state based on the percentage of total coking coal used per state from the SEDS data. This approach
assumes that coke use in l&S and lead and zinc production is proportional to the amount of coking
coal used in a state. This assumption may not be the case because state-level coking coal use is
based on coke production in a given state, not necessarily coke use. The coke could be produced in
one state and shipped for use in another state. However, given the lack of specific data, coking coal
use was determined to be a good surrogate for coke use within a given state because coke
production is often integrated with l&S production where the coke is used. As one further
adjustment, if the amount of coking coal used in IPPU was greater than the total coking coal reported
in the national energy statistics, the amount of coking coal used in the energy sector results were
zeroed out to avoid negative values (this only occurs in 1990,1991,1992, and 1997), and additional
other coal use was subtracted to make up the difference (see "Other coal" below). Appendix A,
Tables A-19 and A-20 in the "FFC C02 Industrial" Tab, describe the coking coal used in IPPU.
Other coal. Two adjustments were made to account for other coal used in the industrial sector. The
first adjustment was to subtract the extra amount of coking coal required for years where the coking
coal adjustment was more than the coking coal total (see above). Similar to coking coal, this
adjustment was based on the percentage of coking coal consumption per state from SEDS.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-8
-------
Section 2 Energy (NIR Chapter 3)
Appendix A, Tables A-21 and A-22 in the "FFC C02 Industrial" Tab, describe this adjustment. The
second adjustment was to subtract coal directly used in the l&S sector. In addition to being used
indirectly to produce coke, coal can be used directly as a process input to l&S production; note that
this does not include coal combusted at l&S facilities to produce power. Other national-level coal
used in l&S production was based on industry data that are not available at the state level. Therefore,
this adjustment was based on the percentage of l&S emissions per state. l&S emissions per state
were taken from the IPPU breakout for l&S, as described in Section 3.3.1. Appendix A, Table A-24 in
the "FFC C02 Industrial" Tab, describes this adjustment. An IPPU-adjusted other coal total was then
calculated by subtracting the adjustments described above (note: this also included the
adjustments for conversion of fuels and C02 exports as described in Step 4 below). Appendix A,
Table A-25 in the "FFC C02 Industrial" Tab, shows this total. The total other coal use was then
adjusted to match the total other coal from the national Inventory (as per Step 1); this adjustment
was based on the percentage of other coal used after the IPPU adjustment. Appendix A, Table A-26 in
the "FFC C02 Industrial" Tab, describes this adjustment.
Natural gas. Two adjustments were made to account for natural gas used in the industrial sector.
The first adjustment was to subtract the amount of natural gas consumption that was used in
ammonia production from energy sector natural gas use. The national-level natural gas used in
ammonia production was back- calculated based on the amount and C content of natural gas
needed to produce ammonia C02 emissions. Therefore, the state-level natural gas used for
ammonia was based on the percentage of ammonia emissions per state. Ammonia emissions per
state were taken from the IPPU breakout for ammonia, as described in Section 3.2.1. Appendix A,
Tables A-27 through A-29 in the "FFC C02 Industrial" Tab, describe this adjustment. The second
adjustment was to subtract natural gas directly used in l&S. National-level natural gas used in l&S
production was based on industry data that are not available at the state level. Therefore, similar to
other coal, the adjustment was based on the percentage of l&S emissions per state from the IPPU
breakout for l&S, as described in Section 3.3.1. Appendix A, Table A-30 in the "FFC C02 Industrial"
Tab, describes this adjustment. An IPPU-adjusted natural gas total was then calculated by
subtracting the adjustments described above. Appendix A, Table A-31 in the "FFC C02 Industrial"
Tab, shows this total. The total natural gas use was then adjusted to match the total natural gas use
from the national Inventory (as per Step 1); this adjustment was based on the percentage of natural
gas used after the IPPU adjustment. Appendix A, Table A-32 in the "FFC C02 Industrial" Tab,
describes this adjustment.
Residualfuel. The residual fuel use was adjusted to subtract the amount of residual fuel used in
carbon black production. Carbon black was the only IPPU use of residual oil. The national-level
residual oil used in IPPU was based on NEUs of residual oil from EIA data, which are not available at
the state level. Therefore, the residual oil IPPU state-level adjustment was based on the percentage
of carbon black emissions per state. Carbon black emissions per state were taken from the IPPU
breakout for petrochemicals, as described in Section 3.2.9, and the percentage for carbon black
specifically was used. Carbon black emissions were determined to be a good surrogate for residual
oil use because the emissions from carbon black production would be directly proportional to
residual oil use. Appendix A, Tables A-33 and A-34 in the "FFC C02 Industrial" Tab, describe this
adjustment. An IPPU-adjusted residual fuel total was then calculated. Appendix A, TableA-35 in the
"FFC C02 Industrial" Tab, shows this total. The total residual fuel use was then adjusted to match
the total residual fuel from the national Inventory (similar to what was done for coal and natural gas
in Step 1); this adjustment was based on the percentage of residualfuel used after the IPPU
adjustment. After the adjustment, the residual fuel use summed across states did not match the
national totals anymore (likely due to the distribution of adjustment based on petrochemical
2-9
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
production, which resulted in negative emissions in some states that were then zeroed out).
Appendix A, Table A-36 in the "FFC C02 Industrial" Tab, describes this adjustment.
Distillate fuel. Distillate fuel use was adjusted to subtract the amount of distillate fuel directly used
in l&S production. National-level diesel fuel used in l&S production was based on industry data that
are not available at the state level. Therefore, similar to other coal and natural gas direct use in l&S,
the adjustment was based on the percentage of l&S emissions per state from the IPPU breakout for
l&S, as described in Section 3.3.1. Appendix A, Tables A-37 and A-38 in the "FFC C02 Industrial" Tab,
describe this adjustment. An IPPU-adjusted distillate fuel total was then calculated. Appendix A,
Table A-39 in the "FFC C02 Industrial" Tab, shows this total. This total was adjusted further based on
reallocation of diesel fuel use across sectors, as shown in Step 5 below.
2.1.1.2.4. Step 3: Adjust for Biofuels and Petroleum Denaturant
Fuel consumption estimates used for C02 calculations were adjusted downward to exclude fuels with
biogenic origins consistent with the IPCC Guidelines. C02 emissions from ethanol and biodiesel
consumption are not included in fuel combustion totals in line with the 2006 IPCC Guidelines and UNFCCC
reporting obligations to avoid double counting with net carbon fluxes from changes in biogenic carbon
reservoirs accounted for in the estimates for LULUCF. C02 emissions from biogenic fuels under fuel
combustion are estimated separately and reported as memo items for informational purposes under the
energy sector. Furthermore, for several years of the time series, denaturant used in ethanol production was
double counted in both transportation and industrial sector energy use statistics. It was therefore subtracted
from transportation sector energy use to avoid double counting. Fuels with biogenic origins (ethanol and
biodiesel) and ethanol denaturant adjustments at the state level are handled by adjusting gasoline and diesel
fuel use based on the total non-biogenic components of those fuels only (which also include any
adjustments for denaturant), as described in Step 5 below. So, in effect, the state-level energy use
calculations used to determine FFC emissions for gasoline and diesel fuel combine this Step 3 with Step 5
below. See Section 2.1.6 for more detail on biofuel use at the state level used to calculate biomass C02 as a
memo item.
2.1.1.2.5. Step 4: Adjust for C02 Exports
Since October 2000, the Dakota Gasification Plant has been exporting C02 produced in a coal
gasification process to Canada by pipeline. Because this C02 is not emitted to the atmosphere in the United
States, the coal that is gasified to create the exported C02 is subtracted from fuel consumption statistics
used to calculate combustion emissions in the national Inventory. Consistent with the approach currently
used in the national Inventory, the coal used to produce exported C02 from the Dakota gas plant to Canada
was subtracted from other coal use to determine state-level emissions. This was all assumed to be
subtracted from North Dakota, the location of the Dakota gas plant. Appendix A, Table A-23 in the "FFC C02
Industrial" Tab, describes this adjustment.
2.1.1.2.6. Step 5: Adjust Sectoral Allocation of Distillate Fuel Oil and Motor Gasoline
Motor gasoline and diesel fuel are used across all sectors. The total amount of motor gasoline and
diesel fuel consumed as reported in the MER is based on petroleum supply data from refineries. Gasoline
use is allocated across the sectors in proportion to aggregations of categories reported in the U.S.
Department of Transportation's Federal Highway Administration (FHWA) highway statistics data (FHWA
1996-2022).14 Diesel fuel use is allocated to the electric power sector based on industry surveys. The
14 FHWA forms MF-21 and MF-24 are used in the calculations.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-10
-------
Section 2 Energy (NIR Chapter 3)
remaining diesel fuel use is allocated across the remaining sectors in a similar way to gasoline use based on
sales data to different categories. Through 2020, the allocation was based on data from ElA's fuel oil and
kerosene sales (FOKS) data (EIA 2022). EIA suspended the FOKS report after data year 2020. Starting in 2021,
diesel fuel use is allocated to sectors based on data from SEDS. For 2021 forward, SEDS uses several
external sources, regressions, and historical sector and state shares to estimate the data that were in the
FOKS report. For the national Inventory, data are needed on fuel use by vehicle type to determine emissions,
so a bottom-up method is used to estimate transportation sector gasoline and diesel fuel use. The national
Inventory determines gasoline and diesel fuel use by vehicle type based on FHWA data and outputs from
EPA's MOtor Vehicle Emissions Simulator (MOVES) model (EPA 2022). The national Inventory then allocates
the remaining fuel use to the remaining sectors based on the proportions in the EIA data. The differences in
the EIA and national Inventory gasoline and diesel fuel allocation approach across sectors are shown below
in Figure 2-3 and Figure 2-4, including information on the categories of use included in each sector and data
for 2022 as an example.
2-11
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Figure 2-3. Comparison of Gasoline Sector Allocation
From the MER, Based on Refinery
Supply Data
Total Supply I 2022 ('000 Gal)
Motor Gasoline
EIA National Energy Balance
From FHWA Data
I Industrial Sector
I 2022 ('000 Gal)
I % of Total 1
Ag
280,522
Construction
307,460
Ind. & comm.
2,045,626
Sector Total:
2,633,608
11.7
| Transportation Sector | 2022 ('000 Gal) |
Highway 127,182,516
Boating 2,306,652
Rec. vehicles* 1,642,447
Sector Total: 131,131,615
1S5-6 i
1 Total
1 2022 ('000 Gal) 1
Motor Gasoline
137,450,452
* Note FHWA added the lawn and garden and recreational vehicle use categories in 2015 which causes a time series discontinuity in the split
between the different sectors. EPA for the national Inventory back calculated fuel use for those categories and adjusts the FHWA data accordingly.
National Sector Split
Non-Transportation
Percent
| Commercial Sector | 2022 ('000 Gal) |
Public non-highway 111,820
Lawn and garden* 3,120,695
Misc. 452,714
Sector Total: 3,685,229
Commercial Sector
Industrial Sector I 2022 ('000 Gal)
Transportation Sector I 2022 ('000 Gal)
EPA National Inventory
From the MER, Based on
Refinery Supply Data
From FHWA &
DOT Data
Non-Transportation
Percent from EIA
National Sector
Split
| Total Supply |
| 2022 ('000 Gal)
Motor Gasoline 135,055,736
1
Transportation Sector 1
1 2022 ('000 Gal) 1
Automobiles
45,107,345
Motorcycles
943,947
Buses
367,988
Light Trucks
77,230,862
Other Trucks
3,532,373
Transportation Sector
2022 ('000 Gal) |
1
Motor Gasoline 128,548,462
Boats (Recreational) 1,365,946
Sector Total: | 128,548,462
t
Remaining
6,507,274 -
Commercial Sector
1 2022 ('000 Gal) 1
Motor Gasoline
3,996,693
Industrial Sector
Motor Gasoline
2,510,581
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-12
-------
Section 2 Energy (NIR Chapter 3)
Figure 2-4. Comparison of Diesel Fuel Sector Allocation
EIA National Energy Balance
From the MER, Based on Refinery
Supply Data
From Electric
Power Oata
From SEDS Data
National Sector Split
Non-Transportation
Non- Electric Percent
1 Total Supply
1 2022 ('000 Gal) 1
Diesel Fuel
61,712,034
Remaining
61,104,614
| Residential Sector
| 2022 ('000 Gal)
% of Total
|
Residential 3,486,798 | 5.7
h-J
Diesel Fuel
2022 ('000 Gal)
| Commercial Sector
1 2022 ('000 Gal)
iisb
Commercial
2,415,840
|4.o y
| Industrial Sector
| 2022 ('000 Gal)
| % of Total |
Industrial
8,727,390
1H
Diesel Fuel
8.727.400
Transportation Sector
46,474,554 | 76.1
_r
Transportation Sector I 2022 ('000 Gal)
1 Total
1 2022 ('000 Gal) 1
Diesel Fuel
61,104,582
Electric Power Sector I 2022 ('000 Gal)
Electric Power Sector I 2022 ('000 Gal)
* Note FHWA added the lawn and garden and recreational vehicle use categories In 2015 which causes a time series discontinuity In the split
between the different sectors, EPA for the national Inventory back calculated fuel use for those categories and adjusts the FHWA data accordingly.
EPA National Inventory
From the MER, Based on
Refinery Supply Data
From Electric
Power Data
From FHWA/
DOT Data
Non-Transportation
Non- Electric Percent
National Sector
Split
| Electric Power Sector |
| 2022 ('000 Gal)
1 J
Electric Power Sector
Diesel Fuel 607,421
Diesel Fuel 607,421
Transportation Sector I 2022 ('000 Gal)
Remaining
61,104,614
Passenger Cars
Buses
Light-Duty Trucks
Medium- and Heavy-Duty Trucks
Recreational Boats
Ships and Non-Recreational Boats
Rail
Sector Total: [~
278,375
2,237,279
3,529,399
39,188,894
308,809
810,026
3,393,700
49,746,481
Transportation Sector I 2022 ('000 Gal)
24%
Residential Sector
2022 ('000 Gal) 1
Diesel Fuel 2,707,006
Remaining
Commercial Sector
2022 ('000 Gal)
11,358,132
Diesel Fuel 1,875,553
\ J
Industrial Sector
2022 ('000 Gal) |
Diesel Fuel 6,775,573
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
The bottom-up approach used by the national Inventory to determine transportation sector fuel use
generally results in less allocation of gasoline to the transportation sector (and more to other sectors) and
more diesel fuel allocated to the transportation sector (and less to other sectors) compared with the original
MER energy balance data, as shown below in Figure 2-5.
Figure 2-5. Comparison of Transportation Sector Fuel Use
Differences in Transportation Sector Gasoline Use
m
200
100
-100
-200
-300
-400
-500
-600
-700
-800
2%
1%
0%
a>
-1% o
-2%
-3%
TO
a
-4% to
-5%
ic
oaiaiCTiCTiaiaicncTiaiooooooooooooooooooooooo
HHHHHHHHHHIN(N(NMN(N«N(NCN{MCN(NO((NtNNN(N(M(SIN(N
-6% o
-7% °
-8%
Nat Inv-EIA MER
> of Total
Differences in Transportation Sector Diesel Use
3
+¦*
CO
OHtNro^i^iDh co oi
(7>cncy>o,,'>cT>o">o
t m id n co cn o
O O O O O O rH
ooooooooooo
lOiDi^-cocnoHM
'IfI'I iI <I (N (N CN
oooooooo
HHHHHHHHHHfNfM(N|fNfSfS(N(N^(NfM(NfMtNlrM(N(NfNfM(NfN(NfN
Nat Inv-EIA MER
> of Total
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Section 2 Energy (NIR Chapter 3)
The national-Level data on gasoline and diesel fuel use by vehicle type used in the bottom-up analysis
was not readily available at the state level. Therefore, the following assumptions and adjustments were made
to distillate fuel and motor gasoline consumption at the state level across the different sectors to reflect the
national Inventory bottom-up transportation fuel use approach:
Transportation sector. The total amount of distillate fuel and motor gasoline used in the
transportation sector was taken from the national Inventory totals (these totals already subtract
biofuel use, subtract denaturants if needed, and are based on multiple factors to determine
transportation sector fuel use). This total amount of distillate fuel and motor gasoline use and
emissions was allocated across states based on the percentage of fuel use by state in gallons from
FHWAdata (FHWA 2022a, 2022b). For distillate fuel, the total was based on FHWAform MF-225, and
the motor gasoline total was based on FHWAform MF-226, both of which have time series of fuel use
by state. Appendix A, Tables A-48 and A-49 in the "FFC C02 Transportation" Tab, describe this
adjustment. The FHWA data reflect on-highway fuel use, but, as seen in Figure 2-3 and Figure 2-4
above, the transportation sector fuel use includes some mobile sources that are considered off-
highway (e.g., recreational boating, railroads). However, because the majority of the motor gasoline
and diesel fuel use is for on-highway purposes, using FHWA data to allocate transportation sector
fuel use to the state level is reasonable. Note that FHWA state-level fuel consumption data are
representative of the point-of-sale and not the point-of-use, so fuel sold in one state that may be
combusted in other states is assigned to the state where the fuel was purchased. This approach is
consistent with IPCC Guidelines (IPCC 2006) for country-level reporting that indicate that "where
cross-border transfers take place in vehicle tanks, emissions from road vehicles should be
attributed to the country where the fuel is loaded into the vehicle." Therefore, when applying the
IPCC approach to the state-level inventory, vehicle emissions are attributed to the state where the
vehicle fuel is sold. This approach could introduce some differences in state-level transportation
sector fuel use and emissions allocations reported here and those reported by individual states. For
example, in addition to fuel sales data, state-level vehicle miles traveled (VMT) data are another
potential surrogate for allocating fuel use to the state level, but that approach does not account for
vehicle and fleet fuel economy variability between states. EPA will consider alternative or
complementary approaches to allocate transportation fuel across states, including VMT data and
other sources. For example, the National Emissions Inventory (NEI) uses county-level fleet and
activity data to generate a bottom-up inventory (EPA 2020).15 Figure 2-6 shows the transportation
sector emissions in 202016 from the top 10 emitting states using different allocation approaches. As
seen in the figure, the approach used will lead to different allocations across states.
15 Note the NEI uses a bottom-up method for determining transportation sector fuel use and emissions based on VMT and
assumed vehicle fleet fuel efficiency at the county level through the MOVES model. However, applying that approach across
all states could lead to differences with national totals. The approach used here is to allocate national totals to states and
not perform a bottom-up analysis for each state.
16 2020 is shown because that is the latest year of NEI data that are produced every three years.
2-15
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Figure 2-6. Transportation Sector 2020 State-Level Allocation Examples
O
u
200
150
100
50
California Florida Georgia Illinois Michigan New North Ohio Pennsylvania Texas
York Carolina
1 Inv
SIT Fuel
'SEDS
SITVMT
NEI
Residential sector. The total amount of distillate fuel used in the residential sector was taken from
the national Inventory totals. It was allocated across states based on the percentage of existing fuel
use in the residential sector per state from SEDS. Appendix A, Tables A-7 and A-8 in the "FFC C02
Residential" Tab, describe this adjustment. Based on the reallocation of sector fuel use, the
residential sector fuel use from the national Inventory is different from the value in SEDS; therefore,
the state-level allocation from SEDS may not represent exactly the fuel values from the national
Inventory. However, residential sector fuel use represented by the national Inventory should be
consistent with what is included in SEDS (e.g., home heating); therefore, the SEDS state-level
breakout is assumed to be representative.
Commercial sector. The total amount of distillate fuel and motor gasoline used in the commercial
sector was taken from the national Inventory totals. It was allocated across states based on the
percentage of existing fuel use in the commercial sector per state from SEDS. Appendix A, Tables A-
14 to A-18 in the "FFC C02 Commercial" Tab, describe this adjustment. Based on the reallocation of
sector fuel use, the commercial sector fuel use from the national Inventory is different from the
value in SEDS; therefore, the state-level allocation from SEDS may not represent the exact fuel
values from the national Inventory. However, commercial sector fuel use represented by the national
Inventory should be consistent with what is included in SEDS (e.g., construction equipment);
therefore, the SEDS state-level breakout is assumed to be representative.
Industrial sector. The total amount of distillate fuel and motor gasoline used in the industrial sector
was taken from the national Inventory totals. Distillate fuel was allocated across states based on the
percentage of existing fuel use in the industrial sector per state after the IPPU adjustments
described in Step 2. Motor gasoline was allocated across states based on the percentage of existing
fuel use in the industrial sector per state from SEDS. Appendix A, Tables A-40 and A-43 in the "FFC
C02 Industrial" Tab, describe this adjustment. Based on the reallocation of sector fuel use, the
industrial sector fuel use from the national Inventory is different from the value in SEDS; therefore,
the state-level allocation from SEDS may not represent the exact fuel values from the national
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-16
-------
Section 2 Energy (NIR Chapter 3)
Inventory. However, industrial sector fuel use represented by the national Inventory should be
consistent with what is included in SEDS (e.g., process energy use); therefore, the SEDS state-level
breakout is assumed to be representative.
Electric power sector. The total amount of distillate fuel used in the electric power sector was taken
from the national Inventory totals. It was allocated across states based on the percentage of existing
fuel use in the electric power sector per state from SEDS. Appendix A, Tables A-57 and A-58 in the
"FFC C02 Electricity" Tab, describe this adjustment. The electric power sector fuel use was not
adjusted in the national Inventory compared with what is represented in SEDS; therefore, the SEDS
state-level breakout is considered representative.
2.1.1.2.7. Step 6: Subtract Consumption for NEU
The energy statistics include consumption of fossil fuels for nonenergy purposes. Most fossil fuels
consumed are combusted to produce heat and power. However, some are used directly for NEU as
construction materials, chemical feedstocks, lubricants, solvents, and waxes.17 For example, asphalt and
road oil are used for roofing and paving, and hydrocarbon gas liquids are used to create intermediate
products. In the national Inventory, emissions from these NEUs are estimated separately under the Carbon
Emitted and Stored in Products from NEUs source category. Therefore, the amount of fuels used for
nonenergy purposes needs to be subtracted from fuel consumption data for determining combustion
emissions.
The adjustments vary over time and represent about 25% to 30% of total unadjusted industrial sector
energy use, as shown in Figure 2-7.
17 Under IPCC Inventory guidance, emissions from these nonenergy sources should be reported as part of IPPU. However,
because of national circumstances and the inability to separate these uses from the national energy balance, the United
States reports these emissions as part of energy. This is an area for future planned improvement as part of the national
Inventory, and any updates will be carried over to the state-level reporting.
2-17
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Figure 2-7. Adjustments Made to Industrial Sector Energy Use to Account for Emissions Reported as NEUs
bJO
i
a>
c
T3
Q)
+¦>
(J
C5
i
-Q
3
CO
CO
3
¦41
CQ
7,000
6,000
5,000
4,000
3,000
2,000
1,000
0
35%
30% S
OHtNrO'tLntDNcoo^
o»irsjro^LOh--ooo>)0'-HfNrO'^-i-ntr)h-coo^O'-HfN
OOOOOOOOOOfIt-Hi1»liIfli1»1«liICNCNCN
ooooooooooooooooooooooo
fN
¦a 3
£ >.
¦a £?
c
(O
c
3
1 Amount Removed
> of Unadjusted Industrial Sector Total
Adjustments for each fuel type were made at the national level based on data and assumptions from EIA
as used in the national energy balance. More detail on the amount and types of fuels used for NEU at the
national level are shown in Appendix A in the "National 2022 NEU C02" Tab.
The following approaches were taken to determine the amounts of different fuels used for NEUs that
needed to be subtracted from energy combustion estimates at the state level. The subtractions were all
made in the industrial sector except for lubricants; those subtractions were used in both the industrial and
transportation sectors and for NEU from territories. The fuels requiring subtraction are:
Coking coal. As per the national Inventory, the amount of coking coal used for NEUs was
determined to be the total of the adjusted coking coal (after subtracting for IPPU use, per Step 2).
Therefore, the state-level totals from Step 2 for coking coal were used to represent NEUs. Appendix
A, Table A-59 in the "NEU" Tab, shows this state-level breakout.
Other coal. The coal used to produce synthetic natural gas at the Eastman gas plant (based on data
from the national Inventory) was assumed to be used for chemical feedstock and therefore was
accounted for under NEU. This other coal NEU was allocated across states by assuming it all
occurred in Tennessee, the location of the Eastman facility. Appendix A, Table A-60 in the "NEU"
Tab, shows this state-level breakout.
Natural gas. The total national-level amount of natural gas used for NEUs was taken from the
national Inventory (based on data from EIA) and represents natural gas used for chemical plants and
other uses. Natural gas used for NEUs was allocated across states based on the percentage of
petrochemical emissions per state. This is an area where there was not any specific data on natural
gas used for NEU in chemical plants and other uses by state. Using petrochemical emissions to
allocate natural gas NEU use by state was considered a reasonable approach as emissions are a
good indication of petrochemical production in a state, and therefore a good indication of how much
NEU fuel was used in that state. Petrochemical emissions per state were taken from the IPPU
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-18
-------
Section 2 Energy (NIR Chapter 3)
breakout for petrochemicals, as described in Section 3.2.9, and the total percentage for all
petrochemicals was used. Appendix A, Table A-61 in the "NEU" Tab, shows this state-level breakout.
LPG, still gas, and petroleum coke. The national-level amount of each of these fuels used for NEUs
was taken from the national Inventory (from EIA data) and assumed to be used primarily as chemical
feedstocks. The amount of NEUs for each fuel was allocated across states based on the percentage
of each total fuel use in the industrial sector per the original state-level data from SEDS. The SEDS
data include NEU and fuel combustion uses of fuel so this approach assumes that the percentage of
these fuel products used in NEU applications per state are proportional to the fuel combustion uses
of these fuel products in a given state. This assumption was considered reasonable as the fuel
combustion and NEU applications of these fuel products are likely to be in the same types of
chemical facilities. Appendix A, Tables A-63 through A-65 and Tables A-69 through A-72 in the "NEU"
Tab, show these state-level breakouts.
Distillate fuel. The total national-level amount of distillate fuel used for NEUs was taken from the
national Inventory (based on data from EIA). Distillate fuel used for NEUs was allocated across
states based on the percentage of distillate fuel use in the industrial sector per state after IPPU
adjustments described in Step 2. As per the previous group of fuel products, this approach assumed
that the percentage of distillate fuel used in NEU applications per state is proportional to fuel
combustion uses of distillate fuel in a given state. The national-level data on distillate fuel used in
NEU applications are based on industry surveys for nonfuel uses in the chemical industry. Therefore,
the assumption that NEUs of distillate fuel are proportional to the total industrial sector amount of
distillate fuel use in a given state may not be completely representative because fuel or other uses of
distillate fuel in the industrial sector could be very broad. However, it was felt to be a reasonable
approach because specific state-level distillate fuel used in NEU applications was not readily
available and the percentage of NEUs of distillate fuel was a small fraction of overall industrial
sector distillate fuel use (less than 1%). EPA will continue to examine other possible sources for
distillate fuel NEU state-level data for future reports. Appendix A, Table A-74 in the "NEU" Tab,
shows this state-level breakout.
Asphalt and road oil, lubricants (in both the industrial and transportation sectors), naphtha
(<401 °F), other oil (>401 °F), special naphtha, waxes and miscellaneous products. As per the
national Inventory, the total amounts of these fuel products were all assumed to be used in NEUs.
Therefore, the total state-level data from SEDS were used to represent NEUs for these fuel products.
Appendix A, Tables A-62, A-66 through A-68, A-73, and A-75 through A-77 in the "NEU" Tab, show
these state-level breakouts.
Emissions associated with NEUs were calculated and reported separately from FFC emissions. Some
further adjustments were made to NEU, and carbon factors were applied; see further discussion in Section
2.1.2 below.
2.1.1.2.8. Step 7: Subtract Consumption oflBFs
The energy statistics include consumption of fossil fuels that are ultimately used for international
bunkers. In the national Inventory, emissions from IBF consumption are not included in national totals and
are instead reported separately as a memo item, as required by the IPCC and UNFCCC inventory reporting
guidelines. There are other international organizations, including the International Civil Aviation Organization
and the International Maritime Organization, that consider global action from these sectors. Therefore, the
amount of each fuel type used for international bunkers was subtracted from fuel consumption data when
determining fuel combustion emissions. The adjustments vary over time and represent about 4% to 7% of
total unadjusted transportation sector energy use, as shown in Figure 2-8.
2-19
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Figure 2-8. Adjustments Made to Transportation Sector Energy Use to Account for IBFs
1,800
w> 1,600
= 1,400
TJ
2 M
I- a
"O E
CD LU
"a u
c
oHMPi^L/noMOoiOHMn^inmsoscioHN
0000000000»H^Hi-HtI H H H H r-lp-IfNCNlfN
ooooooooooooooooooooooo
rMCNCN of Unadjusted Transportation Sector Total
Adjustments for each fuel type were made at the national Level based on data and assumptions from
different data sources, including FAA flight data and information on international shipping; see the national
Inventory report for more details. More details on the amount and types of fuels used for IBFs at the national
level are shown in Appendix A in the "National 2022 FFC C02" Tab.
The following approaches were taken to determine the state-level amounts of different fuels used for
IBFs that needed to be subtracted from energy combustion estimates. The subtractions were all made in the
transportation sector:
Residual fuel and distillate fuel. The total national-level amount of residual and distillate fuel used
for IBF was taken directly from the national Inventory (I BF subtractions). The fuels used for IBF were
allocated across states based on the percentage of fuel use for bunkers from the EIA FOKS data (EIA
2022). This approach was considered reasonable because the FOKS data have information directly
on bunker fuel used at the state level.18 Appendix A, Table A-78 and Table A-79 in the "IBF" Tab, show
these state-level breakouts.
Jet fuel. The total national-level amount of jet fuel used for IBF was taken directly from the national
Inventory (IBF subtractions). Jet fuel used for IBF was allocated across states based on the
percentage of total jet fuel use in the transportation sector by state per the original state-level data
from SEDS. Appendix A, Table A-80 and Table A-81 in the "IBF" Tab, show that state-level breakout
data on jet fuel specifically used for international flights were difficult to find at the state level. The
approach used here to allocate IBFs by state based on the total amount of jet fuel used by state
could potentially lead to an overestimation of I BF emissions for some states with below-average
international flight activity or underestimation for other states with significantly greater than average
18 Note that the FOKS data publication was suspended with the 2020 data release; for this cycle, the same percentage by
state for 2020 was applied to 2022.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-20
-------
Section 2 Energy (NIR Chapter 3)
international flight activity. This is an area of future planned improvements. Also note that this
adjustment is for IBFs. Fuel use and emissions from interstate flights are still included in the
national- and state-level FFC emissions. They were allocated to the state where the jet fuel is
purchased/sold as per the SEDS data.
The result of these previous seven steps is an adjusted amount of fuel use activity data that is then used
to determine FFC C02 emissions. Three additional steps are then required to determine C02 emissions, as
discussed further below.
2.1.1.2.9. Step 8: Determine the C Content of All Fuels
To determine emissions, the amount of carbon per unit of energy in each fuel was needed. Because
different fuels have different C contents, a different factor was determined for each fuel type. The total
carbon estimate defines the maximum amount of C that could potentially be released to the atmosphere if
all of the carbon in each fuel was converted to C02. Fuel-specific C content coefficients for each fuel type
were taken from the national Inventory; see Annex 2 of the national Inventory for more details on carbon
factors used. The national total factors for each fuel used in the national Inventory were applied for fuel use
at the state level. This was considered a reasonable assumption since fossil fuels are widely traded and
regulated, and C contents within the United States do not vary appreciably. Two possible exceptions to this
are coal and gasoline where state-specific C contents could vary based on the type of coal used and the
gasoline blend and grade used. Those fuel emissions factors in the national Inventory were based on
weighted averages of state-level factors. For these factors, EPA will look into using specific state-level
factors in the state-level estimates in future reports.
2.1.1.2.10. Step 9: Estimate C02 Emissions
Total C02 emissions for each fuel are the product of the adjusted energy consumption (from the
previous methodology Steps 1-7), the C content of the fuels consumed (from Step 8), and the fraction of
carbon that is oxidized. Carbon emissions were multiplied by the molecular-to-atomic weight ratio of C02 to
carbon (44/12) and the fraction of carbon that was oxidized to obtain total C02 emitted from FFC. The
fraction oxidized was assumed to be 100% for petroleum, coal, and natural gas.
State-level fuel use byfueltype per sector from Steps 1-7was multiplied by national-level carbon
factors from Step 8 (and also multiplied by molecular weight ratios and oxidation fractions) to determine
state-level emissions by fuel type and by sector.
2.1.1.2.11. Step 10: Allocate Transportation Emissions by Vehicle Type
As discussed in Step 5 above, fuel use at the national level was determined by vehicle type (e.g.,
passenger cars, passenger trucks, buses, medium- and heavy-duty trucks) in the transportation sector
because non-C02 emissions differ by vehicle type and emission control system design. Activity data were
needed by vehicle type to use higher tier methods for non-C02 emissions. The national Inventory is,
therefore, also able to provide the same level of detail for C02 emissions by vehicle type from transportation.
For fuel types other than jet fuel, fuel consumption data by vehicle type and transportation mode were used
to allocate emissions by fuel type calculated for the transportation end-use sector in the national Inventory.
However, as also discussed in Step 5 above, state-level information on fuel use by vehicle type was not
readily available. For C02 emissions, vehicle type is not critical for determining transportation sector
emissions because the calculations are based primarily on fuel use, independent of vehicle type. A state-
level C02 emissions breakout by vehicle type was not done at this time, but this is an area of future planned
improvements.
2-21
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
The above calculations resulted in state-level GHG estimates that generally add up to the total
estimates in the national Inventory, with small differences occurring at the more disaggregated sector level,
as shown below in Figure 2-9 for FFC C02 emissions. The differences are due to the vintage of the different
data sources used. As discussed above in Step 1, the national Inventory was based on the February 2024
MER, while the state-level values were based on the May 2024 SEDS. The SEDS used updated information on
the sector allocation of some fuels, which will be reflected in the next national Inventory report. There is also
a slight difference in total emissions. The percentage differences in the 2022 sector totals are small: a 0.3%
difference in the residential sector and 0.4% in the commercial sector. The percentage difference in total
emissions is also very minor, a 0.0001% difference.
Figure 2-9. Differences in State-Level Total and National Total FFC C02 Emissions
Sector Difference Inv Totals - State Totals
o
u
1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
O H (N CO If) li)
cncncncr>cncncr>cncr>oo
ClCT)CT^Cri(T)CT)01CTlCT)0^00
NCO(J>OHtNf0^mi£lNCOOlOH(NrO LT) (£>
O O O O O O O O tI »I 'I tI rH r-H tI
oooooooooooo
(N(N(N
Residential
Commercial
Industrial
Transportation
Electric Power
Total Difference Inv Totals - State Totals
0.006
0.005
0.004
"0.003
0.002
0.001
0.000
-0.001
-0.002
1
1 ... 1.1
l__l- 1 ¦ I
¦ . "1
1 ¦
cn (T> (T> (T) CT) O") O") CX>
OHtNrO-t^lCNCOQOHfNrO^lO^NCOaiOHfN
OOOOOOOOOOiItItIfItItliIt|iHt|(NCN
-------
Section 2 Energy (NIR Chapter 3)
2.1.1.2.12. Stationary Non-C02 State-Level Breakout
Stationary non-C02 emissions include CH4 and N20 emissions from four energy consumption sectors
(residential, commercial, industrial, and electric power) and four fuel types (coal, fuel oil, natural gas, and
wood).
Non-C02 emissions from FFC at the national level were estimated in line with Tier 1 and 2 methods
described by the IPCC in the 2006IPCC Guidelines for National Greenhouse Gas Inventories (IPCC 2006). For
most categories, a Tier 1 approach was used, which multiplies the adjusted activity data on fuel use by
default emissions factors to determine emissions. The electric power sector used a Tier 2 approach that
relied on the adjusted fuel use activity data and country-specific emissions factors by combustion
technology type.
National-level emissions for all sectors were allocated across states based on the same percentage as
C02 emissions from those sectors and fuel types, as described in the previous section. Appendix A, Tables A-
89 through A-104 in the "Stationary non-C02" Tab, show the percentage breakout of each fuel across sectors
that were used in the analysis. For the residential, commercial, and industrial sectors, it is reasonable to
assume non-C02 emissions by fuel type would be proportional to C02 emissions across states because the
fuel use activity data are the same and only one non-C02 emissions factor was applied per fuel type per
category for each gas.
Electric power sector non-C02 emissions could differ across states based on the type of combustion
technology used, but the analysis was unable to assess these potential differences. The overall impact of
these simplifying assumptions on total state combustion emissions is expected to be small.
2.1.1.2.13. Mobile Non-C02 State-Level Breakout
Mobile non-C02 emissions include CH4 and N20 emissions. National-level estimates of CH4 and N20
emissions from mobile combustion are calculated by multiplying emissions factors by measures of activity
for each fuel and vehicle type (e.g., light-duty gasoline trucks). Activity data include VMT for onroad vehicles
and fuel consumption for nonroad mobile sources. State-level mobile non-C02 emissions were calculated
for four main categories of mobile source emissions: gasoline highway, diesel highway, alternative fuel
highway, and nonhighway. More detail on the approach and what is included under each of the categories is
shown in Figure 2-10 below (EPA 2024a).
2-23
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Figure 2-10. Mobile Source Non-C02 Calculation Methodology
Highway Vehicles
Gasoline
Light Duty (LD)
Heavy Duty (HD)
Truck
Motorcycle
Diesel
Light Duty
Heavy Duty
Truck
Highway 2
Highway 3
Alternative
I
fuels (ethonol,
Alternative Fuel 1
compressed
Vehicles (AFV) 1
natural gas, etc.)
I
1 «V 1
Mobile Combustion Module
Highway 1
Vehicle Miles
Traveled (VMT) by
Vehicle Type
I
Highway 2
Vehicle Age by
Vehicle Type
i
VMT by Vehicle Age
i
Emission Control
System by VMT and
Vehicle Age
Emission Factor by:
Emission Control
System
Fuel Type, and
Vehicle Type
Aviation
Locomotives
Other
Emission Factor by:
Fuel Type
Non-Highway Vehicles
Naphtha
Aviation
Kerosene
Gasoline
Residual Fuel Oil
Marine
Distillate Fuel Oil (Diesel)
Gasoline
Residual Fuel Oil
Railway
Distillate Fuel Oil (Diesel)
Coal
Diesel (for farm, construction
or other activity)
Gasoline (for farm,
construction or other activity)
Key
Fuel Type
Vehicle Type
Tab Name
Emissions Factors and Inputs
Highway VMT
i
Highway Emissions
Non-Highway Fuel
Consumption
+
1
Non-Highway
Emissions
Mobile Combustion Emissions
The approach to estimate mobile non-C02 emissions was to develop state-level estimates by fuel
type/category and use those estimates to develop the percentage of emissions by state. The percentage of
emissions by state were then applied to the national totals from the national Inventory to disaggregate
national totals at the state level. Table 2-3 shows the default data type and source used in developing the
state-level estimates. Appendix A, Tables A-105 through A-116 in the "Mobile Non-C02"Tab, show the
percentages of emissions by vehicle type by state that were used in the analysis.
Table 2-3. Default Data Sources for Mobile Source Non-C02 Emissions
Source/Category
Type of Input
Default Source
Highway Vehicles
Emissions Factors
and VMT
CH4and N20 emissions factors (g/km
traveled) for each type of control
technology
Not state specific, using national
factors; see Annex 3.2 of the national
Inventory
State total VMT, 1990-present, for all
vehicle types
VMT by state for each year from FHWA
Table VM-2. Apportioned to vehicle
type based on national vehicle type
distributions from FHWATable VM-1.
The fuel type distribution within each
vehicle type (i.e., the distribution
between gasoline and diesel) was
taken from the national Inventory
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-24
-------
Section 2 Energy (NIR Chapter 3)
Source/Category
Type of Input
Default Source
Highway Vehicles
Allocating VMT by
Model Year
Annual vehicle mileage accumulation
(miles) for each modelyear in use and
age distribution of vehicles (%) in the
current year
Not state specific, using national
factors; see Annex 3.2 of the national
Inventory
Highway Vehicles
Allocating Control
Technology by Model
Year
Percentage of vehicles with each
control type, 1960-present
Not state specific, using national
factors; see Annex 3.2 of the national
Inventory
Aviation
N20 and CH4 emissions factors (g/kg
fuel) for each type of fuel
Not state specific, using national
factors; see Annex 3.2 of the national
Inventory
Aviation fuel consumption (million
BTU), 1990-present by fuel type
EIASEDS(EIA 2024b)
Marine
N20 and CH4 emissions factors (g/kg
fuel) for each type of fuel
Not state specific, using national
factors; see Annex 3.2 of the national
Inventory
Marine fuel consumption (gallons),
1990-present
Gasoline from FHWA Highway
Statistics, Table MF-24, boating
column; other fuels from EIA SEDS
Locomotive
N20 and CH4 emissions factors (g/kg
fuel) for each type of fuel
Not state specific, using national
factors; see Annex 3.2 of the national
Inventory
Locomotive fuel consumption (gal or
tons), 1990-present
EIA FOKS
Other Nonhighway
N20 and CH4 emissions factors (g/kg
fuel) for diesel and gasoline tractors,
construction equipment, and other
equipment
Not state specific, using national
factors; see Annex 3.2 of the national
Inventory
Fuel consumption (gal), 1990-present,
for agriculture equipment
Gasoline from FHWA Table MF-24,
agriculture column, diesel fuel from EIA
FOKS
Fuel consumption (gal), 1990-present,
for construction equipment
Gasoline from FHWA Table MF-24,
construction column, diesel fuel total
from the national Inventory
apportioned based on gasoline
percentage
Fuel consumption (gal), 1990-present,
for other equipment
Gasoline from FHWA Table MF-24,
industrial and commercial column plus
totals from other small sources from
the national Inventory, diesel fuel from
EIA FOKS
Alternative Fuel
Vehicles
CH4 and N20 emissions factors (g/km
traveled) for each type of alternative
fuel (methanol, ethanol, LPG, liquefied
natural gas, compressed natural gas)
Not state specific, using national
factors; see Annex 3.2 of the national
Inventory
State total VMT, 1990-present, for
alternative fuel vehicles
Based on national totals and
assumptions on alternative fuel vehicle
use by state from EIA alternative fuel
vehicle data
2-25
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
The bottom-up approach to develop mobile source non-C02 state-level estimates by fuel type/category
described above results in a different overall emissions total compared with the national Inventory values.
That is why the estimates are used to develop the percentage of emissions by state that are applied to the
national totals from the national Inventory to disaggregate national totals at the state level. The approach
above could also overestimate or underestimate state emissions by assuming a national average of vehicle
age distribution across states when each state could have a different mix of vehicle fleet age distribution.
However, the approach is considered reasonable, and the overall impact of these simplifying assumptions
on state emissions is expected to be small.
2.1.1.2.14. Breaking Out Data by Economic Sector
The EIA data used for this analysis report fuel use for five sectors (residential, commercial, industrial,
transportation, and electric power). The reporting of emissions at the state level in this analysis also included
emissions from FFC in the agriculture economic sector (which is not the case with the agriculture sector as
defined by the IPCC). Agriculture sector fuel use at the national level was based on supplementary sources
of data because EIA includes agriculture equipment in the industrial fuel-consuming sector. State-level
agriculture fuel use estimates were obtained from USDA survey data. Agricultural operations are based on
annual energy expense data from the Agricultural Resource Management Survey (ARMS) conducted by the
USDA National Agricultural Statistics Service (NASS). NASS uses the annual ARMS to collect information on
farm production expenditures, including expenditures on diesel fuel, gasoline, LPG, natural gas, and
electricity use. A USDA publication (USDA 2023) shows national totals, as well as select states and ARMS
production regions. State estimates were survey-derived for 15 states (Alaska, California, Florida, Georgia,
Iowa, Illinois, Indiana, Kansas, Minnesota, Missouri, North Carolina, Nebraska, Texas, Washington, and
Wisconsin) and model-derived for the remaining states using data and methods developed by the Economic
Research Service of USDA.
These supplementary data were subtracted from the industrial fuel use reported by EIA to obtain
agriculture fuel use. C02 emissions from FFC as well as CH4 and N20 emissions from stationary and mobile
combustion were then apportioned to the agriculture economic sector based on agricultural fuel use.
Calculations for the agricultural sector emissions breakout are shown in Tables 130 through 135 in Appendix
A.
2.1.1.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02 and non-C02 emissions from
FFC was calculated using the 2006 IPCC Guidelines Approach 2 methodology (IPCC 2006). As described
further in Chapter 3 and Annex 7 of the national Inventory (EPA 2024b), levels of uncertainty in the national
estimates in 2022 for FFC were -2%/+4% for C02, 31 %/+122% for stationary source CH4, -33%/+35% for
stationary source N20, -4%/+30% for mobile source CH4, and -8%/+20% for mobile source N20.
The uncertainty estimates for the national Inventory largely account for uncertainty in the magnitude of
emissions and consider uncertainty in activity data and emissions factors used to develop the national
estimates. State-level estimates of annual emissions will likely have a higher relative uncertainty compared
with these national estimates as a result of the additional requirement in some cases of apportioning
national emissions to each state using spatial proxy and supplemental surrogate data sets. As discussed
above, the steps involved in determining state-level FFC emissions could result in some overestimation or
underestimation of state-level emissions. The sources of uncertainty for this category are consistent over
time because the same approaches are applied across the entire time series. As with the national Inventory,
the state-level uncertainty estimates for this category may change as the understanding of the uncertainty of
estimates and the underlying data sets and methodologies improves.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-26
-------
Section 2 Energy (NIR Chapter 3)
2.1.1.4 Recalculations
Consistent with recalculations at the national level, EIA updated distillate fuel oil consumed by the
transportation sector and propane consumed by the industrial sector for 2010 and subsequent years relative
to the previous Inventory. In addition, consistent with the national Inventory, EIA shifted all 2022 product
supplied totals for natural gasoline and unfinished oils to crude oil transfers. This change was made to reflect
the fact that natural gasoline and unfinished oils are used as feedstocks in crude oil production instead of
directly consumed as an end-use fuel. EPA made the same adjustment across the time series. This change
impacted industrial energy consumption across the time series as well as non-energy-use consumption,
which impacts industrial energy consumption values. This change also impacted the HGL carbon content
coefficient used to calculate emissions.
2.1.1.5 Planned Improvements
For coking coal, the percentage subtracted by state could be based on other factors like BOF l&S
production in each state, as opposed to the percentage of total coking coal use. In some cases, a state could
have negative emissions for all fuels if the amount subtracted, as determined from assumed distribution,
was greater than consumption data from SEDS for that state. These negative values were corrected to zero,
but alternative ways to readjust them across other states will be considered.
For petrochemical feedstocks, natural gas NEU was allocated across states based on GHGRP
petrochemicals emissions data per state, while other fuels' NEUs were allocated based on the underlying
SEDS data. Allocating across states based on the underlying SEDS data ensures that in no states is NEU
larger than in the original SEDS data, which would result in negative numbers associated with subtracting
NEU (it is not an issue for natural gas because use is so high overall compared with NEU). However, EPA will
explore different percentages or a way to use GHGRP petrochemical data without resulting in negative use in
any given state.
EPA will look for better ways to allocate jet fuel bunker data across states as opposed to basing it on
percentage of total use (e.g., FAA data, assumptions based on states with international airports and flights).
EPA will look into more state-level activity data for different mobile combustion sources to better
allocate mobile non-C02emissions.
The coal carbon factors in the national Inventory are based in part on state-level data. It might be
possible to build out weighted state-level coal carbon factors that would still amount to the national totals.
For natural gas, state-level heat content data could be used to develop state-level carbon factors for natural
gas, but they would have to be compared with national totals. It might be possible to develop gasoline and
distillate fuel factors per state for the transportation sector, but EPA would have to ensure they are
consistent with the national-level factors.
EPA will look into allocating power sector non-C02 emissions based on other sources like eGRID and
EPA Air Markets Program Data, for instance.
The national Inventory distributes electricity emissions across end-use sectors to present results with
electricity distributed by sector. That calculation was not done at the state level. The national Inventory also
breaks out transportation sector emissions by vehicle type; that calculation was also not done at the state
level. EPA will look into reporting these disaggregated data in future state-level reports.
2.1.1.6 References
EIA (U.S. Energy Information Administration) (2022) Fuel Oil and Kerosene Sales. U.S. Department of Energy.
Available online at: http://www.eia.gov/petroleum/fueloilkerosene.
2-27
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
EIA (2024a) February2024: Monthly Energy Review. DOE/EIA-0035(2024/02). U.S. Department of Energy.
Available online at: https://www.eia.gov/totalenergy/data/monthlv/previous.php.
EIA (2024b) State Energy Data System (SEDS): 1960-2022 (Complete). June 28, 2024. U.S. Department of
Energy. Available online at: https://vwwv.eia.gov/state/seds/seds-data-complete.php?sid=US.
EPA (U.S. Environmental Protection Agency) (2020) 2020 National Emissions Inventory (NEI) Data. Available online
at: https://www.epa.gov/air-emissions-inventories/2020-national-emissions-inventorv-nei-data.
EPA (2022) MOtor Vehicle Emissions Simulator (MOVES3). Available online at https://www.epa.gov/moves.
EPA (2024a) User's Guide for Estimating Methane and Nitrous Oxide Emissions from Mobile Combustion
Using the State Inventory Tool. Available online at: https://www.epa.gov/svstem/files/documents/2024-
02/mobile-combustion-users-guide 508.pdf.
EPA (2024b) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2022. EPA430-R-24-004.
Available online at: https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
FHWA (Federal Highway Administration) (1996-2022) Highway Statistics. U.S. Department of Transportation.
Available online at: http://vwwv.fhwa.dot.gov/policv/ohpi/hss/hsspubs.htm.
FHWA (2022a) Table MF-225. In: Private and Commercial Highway Use of Special Fuel, by State, 1949-2022.
U.S. Department of Transportation. Available online at:
https://www.fhwa.dot.gov/policvinformation/statistics/2022/mf225.cfm.
FHWA (2022b) Table MF-226. In: Highway Use of Gasoline by State, 1949-2022. U.S. Department of
Transportation. Available online at:
https://www.fhwa.dot.gov/policvinformation/statistics/2022/mf226.cfm.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
USDA (U.S. Department of Agriculture) (2023) U.S. Farm Production Expenditures, 2022. Available online at:
https://www.nass.usda.gov/Publications/Highlights/2023/2022 Farm Expenditures FINAL version%202.pdf.
2.1.2 Carbon Emitted from NEUs of Fossil Fuel (NIR Section 3.2)
2.1.2.1 Background
In addition to being combusted for energy, fossil fuels are consumed for NEUs. The fuels used for these
purposes and the nonenergy applications of these fuels are diverse, including feedstocks for manufacturing
plastics, rubber, synthetic fibers, and other materials; reducing agents for producing various metals and
inorganic products; and products such as lubricants, waxes, and asphalt. C02 emissions arise via several
pathways. Emissions may occur when manufacturing a product, as is the case in producing plastics or
rubber from fuel-derived feedstocks. Additionally, emissions may occur during a product's lifetime, such as
during solvent use. As discussed above in the FFC section, emissions from these NEUs are estimated
separately and, therefore, the amount of fuels used for nonenergy purposes are subtracted from fuel
consumption data. Given the linkages between NEUs and combustion emissions, the NEU adjustments and
calculations are presented here.
2.1.2.2 Methods/Approach
FFC C02 emissions calculations discussed above (as per Step 6) were adjusted for fuels used for NEUs.
C02 emissions arise from NEUs via several pathways, including emissions from the manufacture of a
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-28
-------
Section 2 Energy (NIR Chapter 3)
product and during the product's useful Lifetime and ultimate disposal. The approach for determining
national-level NEU emissions is based for the most part on NEU activity data, C contents and assumed C
storage factors. The activity data on NEU by fuel were taken from the FFC adjustments. Then, several
adjustments were made to the data to account for fuel exports and IPPU emissions that are either excluded
or reported in other parts of the national Inventory, as shown in Figure 2-11. C storage factors are based on
the end use of the fuel and assumed fate of the carbon in the products. Appendix A in the "National 2022
NEU C02" Tab provides more details on an example of the adjustments made to the national-level NEU data
to determine adjusted NEU activity data for 2022.
Figure 2-11. Adjustments to Energy Consumption for Emissions Estimates
Determine NEU by Fuel Type and Sector
(From Step 6 of FFC)
Subtract Fuels
Exported
Subtract Fuel Use
Accounted for in IPPU
Determine C
Storage Factor
Calculate C02
Emissions
COj emissions C02 emissions
excluded from accounted for in
the Inventory the IPPU sector
i Counted as part of FFC
I Counted elsewhere
i Not part of Inv. totals
NEU emissions at the state level were calculated based on the same approach as used to determine
national-level NEU emissions. The following steps describe the approach used to determine state-level NEU
emissions.
2.1.2.2.1. Step 1: Determine Total NEU by Fuel Type and Sector
State-level NEU energy data by sector and fuel type were calculated from Step 6 of the FFC
calculations, as discussed above. The NEU adjustments to the FFC data were used as the input to the NEU
calculations. The same state-level breakout of the NEU data used in the FFC calculations was used here.
2.1.2.2.2. Step 2: Adjust for Portions of NEU in Exported Products
State-level NEU energy data calculated from Step 6 above were adjusted to account for exports. Natural
gas, HGL, naphtha (<401 °F), other oil (>401 °F), and special naphtha were adjusted down to subtract out net
exports of these products that are not reflected in the raw NEU data from EIA. Consumption values were also
adjusted to subtract net exports of HGL components (e.g., propylene, ethane). Similar to exported C02
discussed in the FFC calculations, because any potential C02 emissions from exported products are not
emitted to the atmosphere in the United States, the fuel used to create the exported products is subtracted
from statistics used to calculate NEU emissions. The national-level total export energy adjustment data were
taken from the national Inventory. The export adjustments were allocated to states based on the total
amount of NEU fuel use by state from Step 1 under the simplifying assumption that the share of nonenergy
fuels exported matched the amount of nonenergy fuels used by a given state. This assumption could lead to
an overestimation or underestimation of NEU emissions in a given state based on the actual amount of
product exported. However, it was felt to be reasonable given the lack of export data by state and the small
overall adjustment made (2022 export adjustments represent 6.5% of unadjusted nonenergy fuel use).
Appendix A, Tables A-82 through A-86 and Table A-88 in the "NEU Adj" Tab, show these adjusted totals.
2-29
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
2.1.2.2.3. Step 3: Adjust for Portions of NEU Accounted for in IPPU
State-Level NEU energy data were also adjusted down to account for other oil (>401 °F) and petroleum
coke use in IPPU. As per Step 2 in the FFC calculations, emissions from fuels used as raw materials
presented as part of IPPU were removed from the NEU estimates. Portions of nonenergy fuel use were,
therefore, subtracted from the industrial sector nonenergy fuel consumption data before determining NEU
emissions. The national-level total IPPU energy adjustment data for NEU were taken from the national
Inventory. The IPPU adjustments were allocated to states based on the total amount of nonenergy fuel use by
state from Step 1 under the simplifying assumption that the share of nonenergy fuels used in IPPU matched
the amount of nonenergy fuels used by a given state. This assumption could lead to an overestimation or
underestimation of NEU emissions in a given state based on the actual amount of fuel used in IPPU.
However, it was felt to be reasonable given the lack of data by state on NEU fuels used in IPPU and the small
overall adjustment made (2022 IPPU adjustments represent 1.0% of unadjusted NEU fuel use). Appendix A,
Tables A-86 and A-87 in the "NEU Adj" Tab, show these adjusted totals.
2.1.2.2.4. Step 4: Determine C Storage Factor by Fuel Type
C02 emissions can arise from NEUsvia several pathways. Emissions mayoccurwhen manufacturing a
product, as is the case when producing plastics or rubber from fuel-derived feedstocks, or emissions may
occur duringthe product's lifetime, such as during solvent use. Carbon can also be stored from NEUs such
as in a final product like plastics or asphalt. Overall, at a national level in 2022, about 70% of the total carbon
consumed for NEUs is stored in products (e.g., plastics) and not released to the atmosphere. For state-level
calculations, the storage factors per fuel type were taken from the national Inventory values and vary across
fuel types and, for some fuels, over time. See Annex 2.3 of the national Inventory for more details on storage
factors used.
2.1.2.2.5. Step 5: Calculate NEU C02 Emissions
Emissions from NEUs were calculated based on multiplying the adjusted NEU fuel use by state (from
Steps 1-3) by the national-level carbon factors by fuel type (same as used in the FFC calculations, including
oxidation and molecular weight ratio with the exception that HGLs and still gas have separate carbon factors
for combustion and NEUs) and by the fraction of carbon emitted, which is equal to 1 minus the storage factor
of each fuel type (from Step 4). See Annex 2.2 of the national Inventory for more details on carbon factors
used.
There are some small differences in the NEU-calculated state-level emissions totals compared with
what is reported in the national Inventory, as shown in Figure 2-12 below. As with FFC, these differences
represent a very small percentage of total NEU emissions (the maximum percentage difference over time is
around 0.015% of total NEU emissions).
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-30
-------
Section 2 Energy (NIR Chapter 3)
Figure 2-12. Differences in State-Level and NationalTotal NEU C02 Emissions
0.0100
0.0100%
0.0050
0.0000 I I |
3 -0.0050
-0.0100
-0.0150
-0.0200
"I
¦¦I
ill III
0.0050%
0.0000%
-0.0050% J
-0.0100% '
-0.0150%
nj
o
-0.0250 -0.0200%
OHtNlll^lfllONlOOlOHNm^iniOhCOOlOHNm^lfllOMOOlOHM
oiaiCT>(T>ooooooooooooooooooooooo
HHHHHHHHHHOINOI(SOJN(StMrMNOlrMCV(NN(MOI(N(MfV(MlNrM
NEU
2.1.2.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02 from NEUs was calculated
using the 2006 IPCC Guidelines Approach 2 methodology (IPCC 2006). As described further in Chapter 3 and
Annex 7 of the national Inventory (EPA 2024), levels of uncertainty in the national estimates in 2022 were
-31%/+62% for C02. State-level estimates are expected to have a higher uncertainty because some of the
national-level data were apportioned to each state. For example, the allocations of export and IPPU
adjustments are likely to add to the uncertainty at a state level compared with the national totals.
2.1.2.4 Recalculations
Consistent with national estimates, EIA updated energy consumption statistics across the time series
relative to the previous national Inventory. EIA shifted all 2022 product supplied totals for natural gasoline
and unfinished oils to crude oil transfers. This change was made to reflect the fact that, in actuality, nearly
the full volume of these fuels is used as a feedstock in crude oil production, instead of directly consumed as
an end-use fuel. Under ElA's guidance, EPA shifted all product supplied totals for natural gasoline to crude
oil transfers for the time series. Natural gasoline was entirely recategorized, which resulted in zero emissions
for the time series from 1990 to 2022. Natural gasoline previously made up 1.7% of total emissions on
average across the time series for non-energy uses of fossil fuels. Also, to better align with EIA methodology,
the non-energy-use consumption of HGLs is now calculated for the entire time series by assuming that 100%
of ethane, ethylene, and propylene consumption is for non-combustion use and 85% of normal butane,
butylene, isobutane, and isobutylene consumption is for non-combustion use. Non-energy-use consumption
of propane is calculated by subtracting the non-energy consumption of all other HGLs from the total non-
combustion consumption of HGLs as published by the EIA.
2.1.2.5 Planned Improvements
Planned improvements for state-level NEU estimates are consistent with EPA's planned improvements
for national NEU estimates, which are discussed in Section 3.2 of the national Inventory report (EPA 2024).
2-31
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
EPA will also Look into the export and IPPU adjustments to see if they could be done based on state-Level
data, if these data are available, as opposed to assuming the percentage based on SEDS state-level totals.
2.1.2.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
2.1.3 Geothermal Emissions
2.1.3.1 Background
Although not a fossil fuel, geothermal energy does cause C02 emissions, which are included in the
national Inventory. The source of C02 is non-condensable gases in subterranean heated water that is
released during the process.
2.1.3.2 Methods/Approach
National-level geothermal electricity production emissions were estimated by multiplying technology-
specific net generation by technology-specific C contents based on geotype (i.e., flash steam and dry steam).
For state-level geothermal emissions, the total national-level geothermal emissions were taken from
the national Inventory (EPA 2024) and allocated across states based on the amount of geothermal energy
consumed by each state from the SEDS data (EIA 2024). All geothermal emissions were assumed to be in the
electricity sector. Almost every state reported some level of geothermal energy consumption across the time
series.
2.1.3.3 Uncertainty
Given its small contribution to the overall FFC portion of the national Inventory (0.008% in 2022), an
uncertainty analysis was not performed for C02 emissions from geothermal production.
2.1.3.4 Recalculations
No recalculations were applied for this current report.
2.1.3.5 Planned Improvements
EPA will consider if geothermal emissions could be allocated by the type of geothermal production per
state (because different types have different emissions factors) if that data are available.
2.1.3.6 References
EIA (U.S. Energy Information Administration) (2024) State Energy Data System (SEDS): 1960-2022
(Complete). June 28, 2024. U.S. Department of Energy. Available online at:
https://www.eia.gov/state/seds/seds-data-complete.php?sid=US.
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-32
-------
Section 2 Energy (NIR Chapter 3)
2.1.4 Incineration of Waste (NIR Section 3.3)
2.1.4.1 Background
In the context of this section, waste includes all municipal solid waste (MSW) and scrap tires. In the
United States, incineration of MSW tends to occur at waste-to-energy facilities or industrial facilities where
useful energy is recovered; thus, emissions from waste incineration are accounted for as part of the energy
sector. Similarly, scrap tires are combusted for energy recovery in industrial and utility boilers, pulp and
paper mills, and cement kilns. Incinerating waste results in conversion of the organic inputs to C02. Thus, the
C02 emissions from waste incineration are calculated by estimating the quantity of waste combusted and an
emission factor based on the fraction of the waste that is carbon-derived from fossil sources.
2.1.4.2 Methods/Approach
The different categories of national-level waste incinerations emissions include C02 emissions from
MSW fossil components (plastics, synthetic rubber, and synthetic fibers), tire fossil components (synthetic
rubber and carbon black), and non-C02 emissions of CH4 and N20 from total waste combustion. Any net C02
that ultimately results from incinerated biogenic waste is counted through C stock change methodologies in
the agriculture and LULUCF sectors discussed in Chapters 4 and 5 of this report.
National emissions from all the categories were allocated to states based on the percentage of total
MSW combusted. The amount of waste combusted by state was estimated based on several different
sources depending on the year of data, as shown in Table 2-4. This is the same approach as currently used in
the national Inventory (EPA 2024a). The national Inventory has more information on the data sources used.
Table 2-4. Summary of Approaches to Disaggregate Waste Incineration Emissions Across Time
Series
Time Series
Range
Summary of Data Used
1990-2005
Waste combusted by state was based on BioCycle report data.
2006-2010
Waste combusted was based on data from BioCycle, EPA, EIA and the Energy
Recovery Council (ERC) on waste combustion.
2011-2021
Waste combustion data were based on the U.S. EPA GHGRP.
The methodology used for 1990-2005 was to estimate waste combusted by state based on data from
multiple years of BioCycle reports.
The methodology used for 2006-2010 was to estimate waste combusted by state based on data from
the BioCycle reports, EPA Facts and Figures, EIA (EIA 2006-2010), and Energy Recovery Council data.
The methodology used for 2011-2022 was to estimate waste combustion based on EPA's GHGRP (EPA
2024b). The GHGRP reports facility-level emissions of GHG by fuel type from Subpart C data. The CH4 and
N20 data from MSW combustion by facility/unit can be divided by default CH4 and N20 emissions factors to
back-calculate tons of MSW combusted.
See Appendix A, Table A-117 in the "Waste Incineration" Tab, for the percent of MSW combusted
assumed by state byyear from the different sources, as well as the national Inventory report, for more
information on the data sources and methodology used.
The approach used assumed that individual states' waste combustion emissions are proportional to
their share of waste combusted. This assumption is considered reasonable because currently there is no
distinction in the national Inventory on different MSW compositions and fossil component (e.g., plastics)
2-33
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
percentages across states. There could potentially be differences in waste compositions and, therefore,
emissions across states (e.g., because of state waste management policies). The EPA update to the national-
level waste incineration emissions estimates could provide more information on state-level C02 emissions
factors per ton of MSW. This is an area for future work. Assuming scrap tire emissions are produced in
proportion to MSW combustion per state could lead to overestimating or underestimating tire combustion
emissions at the state level. However, given the lack of readily available data, the assumption that tire
combustion emissions occur in proportion to MSW tons combusted in a given state is considered
reasonable.
2.1.4.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02 and N20 from waste
incineration was calculated using the 2006 IPCC Guidelines Approach 2 methodology (IPCC 2006). As
described further in Chapter 3 and Annex 7 of the national Inventory (EPA 2024a), levels of uncertainty in the
national estimates in 2022 were -17%/+16% for C02 and -54%/+164% for N20. State-level estimates are
expected to have a higher uncertainty because the national-level data were apportioned to each state based
on MSW tonnage. In particular, assuming emissions are proportional to total MSW combusted adds
uncertainty associated with different waste compositions across different states. Furthermore, assuming tire
combustion emissions are proportional to MSW tonnage also adds uncertainty associated with the
differences in tire and MSW combustion across states.
2.1.4.4 Recalculations
No recalculations were applied for this current report.
2.1.4.5 Planned Improvements
EPA will look into separating emissions by state based on the category of emissions (e.g., MSW
combustion versus tire combustion). EPA will also consider developing state-level MSW carbon factors
based on the GHGRP state-level data.
2.1.4.6 References
EIA (U.S. Energy Information Administration) (2006-2010) Form EIA-923 Detailed Data with Previous Form
Data (EIA-906/920). U.S. Department of Energy. Available online at:
https://www.eia.gov/electricitv/data/eia923/.
EPA (U.S. Environmental Protection Agency) (2024a) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
EPA (2024b) Data Sets. Available online at: https://www.epa.gov/ghgreporting/ghg-reporting-program-data-
sets.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-34
-------
Section 2 Energy (NIR Chapter 3)
2.1.5 International Bunker Fuels (NIR Section 3.10)
2.1.5.1 Background
Emissions resultingfrom the combustion of fuels used for international transport activities, termed IBFs
under the UNFCCC, are not included in national emissions totals but are reported separately based on the
location of the fuel sales. Two transport modes are addressed under the IPCC definition of IBFs: aviation and
marine. GHGs emitted from the combustion of IBFs, like other fossil fuels, include C02, CH4, and N20 for
marine transport modes and C02 and N20 for aviation transport modes. Emissions from ground transport
activitiesby road vehicles and trainseven when crossing international borders are allocated to the
country where the fuel was loaded into the vehicle and, therefore, are not counted as IBF emissions.
Although reporting on IBFs is a memo item in national-level reports, it does affect the total jet fuel
emissions that are reported because it is a subtraction from total jet fuel use. The same is true at the state
level, where subtracting IBFs affects jet fuel emissions that are reported in a given state (see Step 7 of the
FFC emissions calculations).
2.1.5.2 Methods/Approach
As noted, emissions resultingfrom the combustion of IBFs are not included in national emissions totals
but are reported separately as a memo item based on the location of fuel sales. The same approach was
used at the state level, where estimates of bunker fuels were determined by state and reported as memo
items. Although bunker fuels are memo items and do not affect state-level total GHG emissions, the
allocation of bunker fuels across states could affect the total amount of jet fuel used per state, including
domestic jet fuel use and emissions. Bunker fuel emissions include C02, CH4, and N20 emissions from jet
fuel, diesel fuel, and residual fuel. The jet fuel emissions are broken into commercial and military use. See
Appendix A, Tables A-78 through A-81 in the "IBF" Tab, for details on IBF energy use breakout by state.
The approach used here at the state level to allocate and report IBF and other cross state transportation
sector emissions to the state where the fuel is sold is considered reasonable. However, it is an accounting
decision and may differ from how individual states account for those cross state and international fuel use
emissions in their own inventories.
2.1.5.2.1. Jet Fuel
National-level jet fuel C02 emissions from commercial aircraft came directly from FAA emissions data.
C02 emissions from military use were based on fuel use data multiplied by the national Inventory C02
emissions factor. National-level CH4 and N20 emissions were based on fuel use data multiplied by an
emissions factor, and CH4 emissions from jet fuel use were assumed to be zero. N20 emissions were split
between commercial and military based on the percentage of total C02 emissions.
Jet fuel emissions from bunker fuels were allocated to states based on jet fuel use sales data from SEDS
(EIA2024).
2.1.5.2.2. Residual and Diesel Fuel
National-level residual and diesel fuel emissions were based on fuel use data multiplied by emissions
factors for the different emissions. The emissions were allocated to states based on EIA FOKS data for
bunker fuel use for diesel and residual fuels (EIA 2022).19
19 Note that the FOKS data were suspended with the 2020 data. For this cycle, the same percentage by state for 2020 was
applied to 2022.
2-35
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
2.1.5.3 Uncertainty
A quantitative uncertainty analysis associated with the national estimates of C02, CH4, and N20 from
IBFs was not calculated because the estimates are only considered memo items. However, there is a
qualitative discussion of uncertainty associated with national-level IBF emissions in the national Inventory.
State-level estimates are expected to have a higher uncertainty because of the assumptions related to
allocating IBF fuels to the state level. For example, a high degree of uncertainty is associated with allocating
jet fuel bunkers to states based on the total amount of jet fuel used per state.
2.1.5.4 Recalculations
No recalculations were applied for this current report.
2.1.5.5 Planned Improvements
As discussed previously, the approach used here to allocate bunker fuels by state based on the total
amount of jet fuel used by state could potentially lead to an overestimation or underestimation of bunker fuel
emissions for some states. Therefore, EPA will look into data specific to jet fuel bunkers by state, such as
flight-level data on departures and destinations.
Currently, the approach used here allocates total IBF use to the 50 states and the District of Columbia.
EPA will examine if it is possible to allocate some jet fuel and marine bunkers to territories as they are also
covered as part of the National Inventory.
2.1.5.6 References
EIA (U.S. Energy Information Administration) (2022) Fuel Oil and Kerosene Sales. U.S. Department of Energy.
Available online at: http://www.eia.gov/petroleum/fueloilkerosene.
EIA (2024) State Energy Data System (SEDS): 1960-2022 (Complete). June 28, 2024. U.S. Department of
Energy. Available online at: https://www.eia.gov/state/seds/seds-data-complete.php?sid=US.
2.1.6 Wood Biomass and Biofuels Consumption (NIR Section 3.11)
2.1.6.1 Background
In line with the reporting requirements for national-level inventories submitted under the UNFCCC, C02
emissions from biomass combustion are estimated separately from fossil fuel C02 emissions and are not
directly included in the energy sector contributions to U.S. totals. In accordance with IPCC methodological
guidelines, any such emissions are calculated by accounting for net carbon fluxes from changes in biogenic
carbon reservoirs in the agriculture, land use, land-use change and forestry sector. Biomass non-C02
emissions are reported as part of emissions totals and are included under fossil fuel non-C02 emissions for
both stationary and mobile sources.
2.1.6.2 Methods/Approach
The combustion of biomass fuelssuch as wood, charcoal, and biomass- and wood waste-based fuels
such as ethanol, biogas, and biodieselgenerates C02 in addition to the CH4 and N20 covered earlier. In line
with the reporting requirements for inventories submitted under the UNFCCC, C02 emissions from biomass
combustion have been estimated separately from fossil fuel C02 emissions and are not directly included in
the energy sector contributions to U.S. totals. In accordance with IPCC methodological guidelines, any such
emissions were calculated by accounting for net carbon fluxes from changes in biogenic carbon reservoirs in
the agriculture, land use, land use change, and forestry sector.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-36
-------
Section 2 Energy (NIR Chapter 3)
Therefore, C02 emissions from wood biomass and biofuel consumption were not included specifically
in summing energy sector totals. However, they are presented here for informational purposes and to provide
detail on wood biomass and biofuels consumption. See Appendix A, Tables A-118 through A-129 in the
"Biomass C02" Tab, for the breakout of biomass C02 emissions by fuel type and sector.
2.1.6.2.1. BiomassEthanol, Transportation
National-level ethanol C02 emissions from the transportation sector were taken from the national
Inventory. Emissions were allocated to states based on the percentage of gasoline used in the transportation
sector by state, which is based on FHWA data (FHWA 2022a).
2.1.6.2.2. BiomassEthanol, Industrial
National-level ethanol C02 emissions from the industrial sector were taken from the national Inventory.
Emissions were allocated to states based on the percentage of gasoline used in the industrial sector by
state, which is based on SEDS data (EIA 2024).
2.1.6.2.3. BiomassEthanol, Commercial
National-level ethanol C02 emissions from the commercial sector were taken from the national
Inventory. Emissions were allocated to states based on the percentage of gasoline used in the commercial
sector by state, which is based on SEDS data (EIA 2024).
2.1.6.2.4. BiomassBiodiesel, Transportation
National-level biodiesel C02 emissions from the transportation sector were taken from the national
Inventory. Emissions were allocated to states based on the percentage of diesel fuel used in the
transportation sector by state, which is based on FHWA data (FHWA 2022b).
2.1.6.2.5. BiomassWood, Industrial/Residential/Commercial/Electric Power
National-level wood C02 emissions from all sectors were taken from the national Inventory. Emissions
were allocated to states based on the percentage of wood used in each sector by state, which is based on
SEDS data (EIA 2024).
2.1.6.3 Uncertainty
A quantitative uncertainty analysis associated with the national estimates of C02, CH4, and N20 from
wood biomass and biofuels combustion has not been considered a priority and has not been estimated. The
priority is to estimate uncertainty for estimates that get rolled into national totals as opposed to estimates
that are considered memo items. However, a qualitative discussion of uncertainty is associated with
national-level wood biomass and biofuels combustion emissions in the national Inventory. State-level
estimates are expected to have a higher uncertainty because of the assumptions related to allocating
emissions to the state level based on fuel use data.
2.1.6.4 Recalculations
No recalculations were applied for this current report.
2.1.6.5 Planned Improvements
For C02 emissions from wood fuels, there is likely considerable variation among states. EPA will look
into other data sources, including from the USFS, on wood used as a fuel.
EPA will look into variability in ethanol consumption across states. It is not likely that ethanol is blended
in the same percentage annually across all states.
2-37
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
2.1.6.6 References
EIA (U.S. Energy Information Administration) (2024) State Energy Data System (SEDS): 1960-2022
(Complete). June 28, 2024. U.S. Department of Energy. Available online at:
https://www.eia.gov/state/seds/seds-data-complete.php7sicNUS.
FHWA(Federal Highway Administration) (2022a)Table MF-226. In: Highway Use of Gasoline by State, 1949-
2022. U.S. Department of Transportation. Available online at:
https://www.fhwa.dot.gov/policvinformation/statistics/2022/mf226.cfm.
FHWA (2022b) Table MF-225. In: Private and Commercial Highway Use of Special Fuel, by State, 1949-2022.
U.S. Department of Transportation. Available online at:
https://www.fhwa.dot.gov/policvinformation/statistics/2022/mf225.cfm.
2.2 Fugitive Emissions
This section presents the methodology used to estimate the fugitive portion of energy emissions and
consists of the following sources:
Coal mining (CH4, C02)
Abandoned underground coal mines (CH4)
Petroleum and natural gas systems (C02, CH4, N20)
Abandoned oil and gas wells (C02, CH4)
2.2.1 Coal Mining (NIR Section 3.4)
2.2.1.1 Background
Three types of coal mining-related activities release CH4to the atmosphere: underground mining,
surface mining, and post-mining (i.e., coal-handling) activities. For the national Inventory, EPA compiles
emissions estimates for each mine into a national total for active underground mines and compiles coal
production data to estimate emissions from surface coal mining and post-mining activity.
2.2.1.2 Methods/Approach
The methods used to determine state-level estimates for coal mining fugitive emissions consists of two
separate sources consistent with the national Inventory:
Active underground mines
Surface mining and post-mining activities
2.2.1.2.1. Active Underground Mines
To compile national estimates of CH4 emissions from active underground coal mines for the national
Inventory, EPA develops emissions estimates for each mine and sums them to a national total. The approach
to arrive at state-by-state estimates of CH4 emissions from active underground mines is consistent with the
national methods (i.e., using Approach 1 as defined in the Introduction of this report). Rather than summing
estimates to a national total, EPA instead totals these mine-specific estimates into a state-level total for
each state, based on the estimates for each of the mines located in a state. These state-level estimates are
also published in Annex 3.4 to the national Inventory (EPA 2024).
As described in Section 3.4 of the national Inventory, EPA uses an IPCCTier 3 method for estimating CH4
emissions from underground coal mining. These emissions have two sources: ventilation systems and
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-38
-------
Section 2 Energy (NIR Chapter 3)
degasification systems. Emissions are estimated using mine-specific data, then summed to determine total
CH4 Liberated. The CH4 recovered and used is then subtracted from this total, resulting in an estimate of net
emissions to the atmosphere. See Section 3.4 of the national Inventory (EPA 2024) for more detail.
To estimate CH4 liberated from ventilation systems, EPA uses data collected through its GHGRP20
(Subpart FF, "Underground Coal Mines") (EPA 2023), data provided by the U.S. Mine Safety and Health
Administration (MSHA) (MSHA 2023), and occasionally data collected from other sources on a site-specific
level (e.g., state gas production databases). Since 2011, the nation's "gassiest" underground coal mines
those that liberate more than 36,500,000 actual cubic feet of CH4 per year (about 17,525 metric tons C02
equivalent)have been required to report to EPA's GHGRP (EPA 2023).21 Mines that report to EPA's GHGRP
must report quarterly measurements of CH4 emissions from ventilation systems; they have the option of
recording and reporting their own measurements or using the measurements taken by MSHA as part of that
agency's quarterly safety inspections of all mines in the United States with detectable CH4 concentrations.22
More information can be found in the national Inventory (Chapter 3, Section 3.4 and Annex 3.4) at
https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-inventorv-2024-chapter-3-energy.pdf.
EPA estimates fugitive C02 emissions from underground mining using an IPCCTier 1 method. Emission
estimates are based on the IPCCTier 1 emission factor (5.9 m3/metric ton) and annual coal production from
underground mines from EIA (IPCC 2019; EIA 2023, Table 1). The underground mining default emission factor
accounts for all the fugitive C02 likely to be emitted from underground coal mining.
2.2.1.2.2. Surface Mining and Post-mining Activities
Mine-specific data are not available for estimating CH4 emissions from surface coal mines or for post-
mining activities. For surface mines, basin-specific coal production obtained from ElA's Annual Coal Report
(EIA 2023) are multiplied by basin-specific CH4 contents (EPA 1996, 2005) and a 150% emissions factor (to
account for CH4from overburden and underburden) to estimate CH4 emissions (King 1994, Saghafi 2013). For
post-mining activities, basin-specific coal production is multiplied by basin-specific gas contents and a mid-
range 32.5% emissions factor for CH4 desorption during coal transportation and storage (Creedy 1993).
Basin-specific in situ gas content data were compiled from the American Association of Petroleum
Geologists (AAPG 1984) and U.S. Bureau of Mines (1986).
To determine state-level CH4 emissions estimates for surface coal mining and post-mining activities,
emissions estimates are apportioned based on the coal production in each state, as reported in the EIA
Annual Coal Report (i.e., using Approach 1 as defined in the Introduction of this report). The appropriate
basin-specific CH4 content for the coal produced in a state was assigned based on the coal basin within
which the state is located. For post-mining activities, these emissions are assigned to the state where the
coal was produced, even if a portion of such emissions may occur outside the state, such as during
interstate transport and storage before use. More information can be found in the national Inventory (Chapter
3, Section 3.4 and Annex 3.4). EPA estimates fugitive C02 emissions from surface mining using an IPCCTier 1
method. Emission estimates are based on the IPCCTier 1 emission factor (0.44 m3/metric ton) and annual
coal production from surface mines (EIA 2023, Table 1). IPCC methods and data to estimate fugitive C02
20 In implementing improvements and integrating data from its GHGRP, EPA follows the latest guidance from the IPCC in its
Use of Facility-Specific Data in National Greenhouse Gas Inventories technical bulletin (IPCC 2011).
21 Underground coal mines report to EPA under Subpart FF of the GHGRP (40 CFR Part 98). In 2022, 61 underground coal
mines reported to the program.
22 MSHA records coal mine CH4 readings with concentrations of greater than 50 ppm (parts per million) of CH4. Readings
below this threshold are considered nondetectable.
2-39
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
emissions from post-mining activities (for both underground and surface coal mining) are currently not
available.
2.2.1.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of CH4 and C02 emissions from coal
mining was calculated using the 2006 IPCC Guidelines Approach 2 methodology (IPCC 2006), which is
described further in Chapter 3 of the national Inventory (EPA 2024). The level of uncertainty in the 2022
national CH4 estimate is -20%/+9%; for the national fugitive C02 estimate, the level of uncertainty is
-69%/+75%. Because CH4 emissions estimates from underground mine ventilation and degasification
systems were based on actual measurement data from EPA's GHGRP and from MSHA, uncertainty is
relatively low. Surface mining and post-mining CH4 emissions, which are based on coal production and the
application of emissions factors, are associated with considerably more uncertainty than underground
mines because of the difficulty in developing accurate basin-level emissions factors from field
measurements. However, because underground mine emissions constitute the majority of total coal mining
emissions, the uncertainty associated with underground emissions is the primary factor that determines the
overall uncertainty of the CH4 emissions estimates. The major sources of uncertainty for estimates of fugitive
C02 emissions are the Tier 1 IPCC default emission factors used for underground mining (-50%/+100%) and
surface mining (-67%/+200%) (IPCC 2019).
National-level emissions estimates for underground mines were developed by aggregating mine-level
estimates. Similarly, state-level emissions estimates for underground mines were developed by aggregating
mine-level estimates for all the coal mines located within each state. The relatively low uncertainty
associated with underground mine emissions at the national level is assumed to be the same for state-level
underground mine emissions estimates. State-level emissions estimates for surface mining and post-mining
emissions are associated with higher uncertainty than underground estimates because they are based on
coal production within a state and the application of emissions factors. Because state-level estimates are
based on the coal production within a state, the uncertainty associated with surface mining and post-mining
emissions at the national level is assumed to be the same for state-level estimates. However, as with the
national estimates, underground emissions account for the majority of state-level coal mining emissions,
and the uncertainty associated with underground emissions is the primary factor that determines overall
uncertainty for state-level emissions estimates.
2.2.1.4 Recalculations
No recalculations were applied for this current report.
2.2.1.5 Planned Improvements
Any planned improvements for state-level coal mining estimates will align with any improvements EPA
implements for improving national estimates for coal mining, which are discussed in Section 3.4 of the
national Inventory report (EPA 2024). EPA is assessing possible improvements for future national reports, but
at this time has no specific planned improvements for the national Inventory report for estimating CH4 and
C02 emissions from underground and surface mining and CH4 emissions from post-mining.
2.2.1.6 References
AAPG (American Association of Petroleum Geologists) (1984) Coalbed Methane Resources of the United
States. AAPG Studies in Geology Series #17.
Creedy, D.P. (1993) Methane Emissions from Coal Related Sources in Britain: Development of a
Methodology. Chemosphere, 26:419-439.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-40
-------
Section 2 Energy (NIR Chapter 3)
EIA (U.S. Energy Information Administration) (2023) Annual Coal Report2022. DOE/EIA-0584. U.S.
Department of Energy.
EPA (U.S. Environmental Protection Agency) (1996) Evaluation and Analysis of Gas Content and Coal
Properties of Major Coal Bearing Regions of the United States. EPA/600/R-96-065.
EPA (2005) Surface Mines Emissions Assessment. Draft.
EPA (2023) 2022 Envirofacts. Subpart FF: Underground Coal Mines. Available online
at:https://enviro.epa.gov/querv-builder/ghg/SUBPART%20FF%20-
%20UNDERGROUND%20COAL%20MINES.
EPA (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2022. EPA430-R-24-004. Available
online at: https://vwwv.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.jp/public/2006gl/.
IPCC (2011) Use of Facility-Specific Data in National Greenhouse Gas Inventories. Available online at:
https://www.ipcc-nggip.iges.or.ip/public/tb/TFI Technical Bulletin 1.pdf.
IPCC (2019) 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories. E.C.
Buendia, K. Tanabe, A. Kranjc, J. Baasansuren, M. Fukuda, S. Ngarize A. Osako, Y. Pyrozhenko, P.
Shermanau, and S. Federici (eds.). Available online at: https://www.ipcc.ch/report/2019-refinement-to-
the-2006-ipcc-guidelines-for-national-greenhouse-gas-inventories/.
King, B. (1994) Management of Methane Emissions from Coal Mines: Environmental, Engineering, Economic
and Institutional Implication of Options. Neil and Gunter Ltd.
MSHA (Mine Safety and Health Administration) (2023) Data Transparency at MSHA. Available online at:
http://www.msha.gov/.
Saghafi, A. (2013) Estimation of Fugitive Emissions from Open Cut Coal Mining and Measurable Gas Content.
13th Coal Operators' Conference, University of Wollongong, The Australian Institute of Mining and
Metallurgy & Mine Managers Association of Australia. 306-313.
U.S. Bureau of Mines (1986) Results of the Direct Method Determination of the Gas Contents of U.S. Coal
Basins. Circular 9067.
2.2.2 Abandoned Underground Coal Mines (NIR Section 3.5)
2.2.2.1 Background
Underground coal mines continue to release CH4 after closure. As mines mature and coal seams are
mined through, mines are closed and abandoned. Many are sealed, and some flood when groundwater or
surface water intrudes into the mine void. Shafts or portals are generally filled with gravel and capped with a
concrete seal, while vent pipes and boreholes are plugged in a manner similar to oil and gas wells. Some
abandoned mines are vented to the atmosphere to prevent the buildup of CH4 that may find its way to
surface structures through overburden fractures. As work stops within the mines, CH4 liberation decreases,
but it does not stop completely. Following an initial decline, abandoned mines can liberate CH4 at a near-
steady rate over an extended period of time, or, if flooded, produce gas for only a few years. The gas can
migrate to the surface through the conduits described above, particularly if they have not been sealed
adequately. In addition, diffuse emissions can occur when CH4 migrates to the surface through cracks and
fissures in the strata overlying the coal mine.
2-41
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
2.2.2.2 Methods/Approach
For the national Inventory, EPA estimates national-Level CH4 emissions from abandoned underground
coal mines using the Abandoned Mine Methane (AMM) model.23 The AMM model predicts mine-level CH4
estimates from the time of abandonment through the inventory year of interest. The flow of CH4 from the coal
to the mine void is primarily dependent on the mine's emissions when active and the extent to which the
mine is flooded, sealed, or vented. For each abandoned mine, the AMM model accounts for mine status,
date of abandonment, and the reported average daily emission rate at the time of abandonment to estimate
emissions using decline curves specific to mine status and coal basin. More information on the estimation
methodology and model input data can be found in Chapter 3, Section 3.5, of the national Inventory (EPA
2024).
EPA updated the AMM model to include state-level estimates as a regular output. These state-level
estimates were available for inventory year 2020 and subsequent years of the inventory time series (i.e.,
2021, 2022). Previously, the AMM model included only coal basin identifiers; EPA has added state identifiers.
Under this approach, both national-level and state-level estimates are generated for an inventory year by the
AMM model. The modified model output contains emissions subtotals by state, coal basin, and mine status.
These subtotals are then aggregated to generate state-level estimates. The final model result (i.e., national-
level estimates) is the average of 10,000 model iterations, but the calculated state estimates are not.
Therefore, the sum of the state-level estimates may not exactly equal the final national-level estimate. The
state-level estimates are normalized to the final national-level model result using the difference between the
national-level total and the sum of state-level totals. This approach relies on model simulations using decline
curves based on mine location (state and basin) and mine status, rather than using state allocation factors
(as described below) to develop state-level estimates. Therefore, this approach provides more accurate
state-level estimate.
While state-level estimates for inventory years 2020-2022 are estimated usingApproach 1 as defined in
the Introduction to this report, state-level emissions estimates for the 1990-2019 inventory years are
developed from the national-level emissions estimates usingApproach 2.. Estimates use state values for
mine-level average daily CH4 emissions at the time of abandonment, mine status (i.e., flooded, sealed,
vented, and unknown), date of abandonment, and mine location (basin and state) to disaggregate national
emissions as outlined below.
2.2.2.2.1. Step 1: Develop State Allocation Factors by Basin and Mine Status
For liberated CH4, the estimated mine-level average daily emissions from the AMM model were totaled
by state, mine status, and coal basin (Central Appalachia, Illinois, Northern Appalachia, Warrior, and
Western basins) for each year in the 1990-2019 time series. Using these state-level totals of average daily
emissions and the basin-level totals of average daily emissions by mine status, state allocation factors
(percent) were developed by state, mine status, and coal basin such that allocation factors across all states
within the same coal basin and same mine status total 100% for each year in the time series (see Appendix B,
Tables B-1 through B-4, for these data).
State allocation factors for recovered CH4 were calculated similarly to liberated CH4 state allocation
factors, with the exception that allocation factors were calculated by basin only (not mine status). There are
23 The AMM model is run using @Risk software, which is a stochastic Monte Carlo simulation software.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-42
-------
Section 2 Energy (NIR Chapter 3)
very few CH4 recovery projects for each year in the time series, so the breakdown by coal basin was sufficient
to develop state allocation factors.
For pre-1972 emissions,24 state allocation factors for mines abandoned before 1972 (referred to as
"pre-1972 mines") were developed using 2019 emissions estimates. For these mines, 2019 emissions
estimates serve as a good proxy for the entire time series because the pre-1972 mine estimates are
developed using county-level default percentages built into the AMM model.
As an example, Table 2-5 presents the state allocation factors for liberated CH4 for all states in the
Illinois Basin with sealed abandoned mines for year 2019 in the time series.
Table 2-5. Example State Allocation Factors for the
Illinois Coal Basin (Sealed Mines)
State
Basin
Status
Percent (%) of Emissions
IL
Illinois
Sealed
77%
IN
Illinois
Sealed
6%
KY
Illinois
Sealed
17%
2.2.2.2.2. Step 2: Develop Master Table of Basin and Mine Status-Level Emissions for 1990-2019
EPA compiled data from previous AMM models. The AMM model only estimates annual emissions for a
single inventory year (i.e., for the 1990-2019 time series, there are 29 separate AMM models, each
addressing a single year in the time series). EPA compiled into a master table the time series estimates of
liberated CH4, recovered CH4, and CH4 emissions from previous annual versions of the AMM model for the
1990-2019 time series.
Next, EPA normalized direct calculations to match model iterations. The master table contains the
following AMM model outputs for each year in the time series (under separate categories for liberated
emissions, recovered emissions, and emissions from pre-1972 mines):
1. Annual emissions subtotals by coal basin and by mine status (calculated using in-built decline
curves in the AMM model and input data, such as average daily emissions at the time of
abandonment, date of abandonment, and mine status indicator).
2. Annual national-level total emissions (based on an average of 10,000 stochastic iterations
performed on the AMM model output #1 above and their associated uncertainty ranges).
The master table contains annual subtotals by coal basin and mine status; however, the aggregate of
the annual subtotals by basin and mine status (i.e., sum of AMM model output #1 above) does not match the
annual national-level total emissions estimate (AMM model output #2 above). Model output #2 above is the
average value for 10,000 model iterations. Therefore, there is a very small difference between the two
national-level totals for each year in the time series (typically less than 0.5% in any year of the time series).
For this reason, the annual estimates in the master table (i.e., annual subtotals by coal basin and by mine
24 Because of limited data availability for mines abandoned before 1972, a different approach was used in the AMM model to
estimate emissions from these mines (referred to as "pre-1972 mines") compared with mines abandoned in 1972 and later
years. The AMM model estimates emissions for the pre-1972 mines at the county level and does not use mine-level average
daily emissions at the time of abandonment. Refer to the national Inventory Chapter 3, Section 3.5, for further details.
2-43
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
status; AMM model output #1) must be normalized25 to equal the national-level emissions estimate (AMM
model output #2) that represents the national emissions estimates used in the national Inventory.
2.2.2.2.3. Step 3: Apply State Allocation Factors to Basin- and Mine Status-Level Emissions
The emissions values from the master table generated in Step 2 were multiplied by the state allocation
factors generated in Step 1 to develop 1990-2019 annual state-level CH4 estimates.
For pre-1972 mines, 2019 state allocation factors were applied to the annual pre-1972 national
estimates in the master table.
For mines abandoned after 1972, annual basin and mine status-level state allocation factors were
applied to the normalized basin- and mine status-level emissions estimates in the master table.
2.2.2.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of CH4 emissions from abandoned
coal mines was calculated using the 2006 IPCC Guidelines Approach 2 methodology (IPCC 2006). As
described in Chapter 3 of the national Inventory (EPA 2024), the level of uncertainty in the 2022 national CH4
emission estimate is -21%/+20%.
National-level abandoned mine emissions estimates were developed by predicting the emissions of a
mine since the time of abandonment using basin-level decline curves. Multiple aspects of the estimation
method introduce uncertainty for the emissions estimates. In developing national estimates, because of a
lack of mine-specific data, abandoned mines are grouped by basin with the assumption that they will
generally have the same initial pressures, permeability, and isotherm. Other sources of uncertainty in the
national estimates are mine status (venting, flooded, or sealed) and CH4 liberation rates at the time of
abandonment. These data are not available for all the abandoned mines in the national Inventory.
Abandoned mines with unknown status are assigned a status based on the known status of other mines
located within the same basin. Mine-specific CH4 liberation rates at the time of abandonment are not
available for mines abandoned before 1972 ("pre-1972 mines"). It is assumed that pre-1972 mines are
governed by the same physical, geologic, and hydrologic constraints that apply to post-1971 mines; thus,
their emissions may be characterized by the same decline curves.
State-level estimates have a higher uncertainty because the national emissions estimates were
apportioned to each state based on mine-specific CH4 liberation rates, mine status, and basin information
for all abandoned mines located within the state. Additionally, the number of mines with unknown status in
each state affects the relative uncertainty of state-level estimates. Estimates for states with a greater
number of mines with unknown status are expected to have relatively higher uncertainty compared with
states with fewer abandoned mines with unknown status. Similarly, states with a greater number of pre-1972
abandoned mines are expected to have relatively higher uncertainty compared with states with fewer pre-
1972 mines.
2.2.2.4 Recalculations
No recalculations were applied for this current report.
25 The difference between the national total and summed total of modeled emissions by coal basin and mine status was
allocated to a coal basin and mine status grouping based on their share of the national total (before normalization).
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-44
-------
Section 2 Energy (NIR Chapter 3)
2.2.2.5 Planned Improvements
Any planned improvements for state-Level estimates for abandoned coal mines will align with any
improvements EPA implements for national estimates for abandoned coal mines. There are currently no
improvements planned. For more information, see Chapter 3, Section 3.5, of the national Inventory (EPA
2024).
2.2.2.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventory-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
2.2.3 Petroleum Systems (NIR Section 3.6)
2.2.3.1 Background
This section describes methods used to estimate state-level C02, CH4, and N20 emissions from
petroleum systems. This category includes fugitive emissions from leaks, venting, and flaring. CH4 emissions
from petroleum systems are primarily associated with onshore and offshore crude oil production,
transportation, and refining operations. During these activities, CH4 is released to the atmosphere as
emissions from leaks, venting (including emissions from operational upsets), and flaring. C02 emissions
from petroleum systems are primarily associated with onshore and offshore crude oil production and refining
operations. Note that C02 emissions in petroleum systems exclude all combustion emissions (e.g., engine
combustion) except for flaring C02 emissions. All combustion C02 emissions (except for flaring) are
accounted for in the FFC section. Emissions of N20 from petroleum systems are primarily associated with
flaring.
The methods used to develop the state-level estimates for petroleum systems follow the Hybrid
approach (a combination of Approach 1 and Approach 2), as defined in the Introduction of this report. Most
sources follow Approach 2 and rely on relative differences in basic state activity levels (e.g., petroleum
production), and do not reflect differences between states due to differences in practices, technologies, or
formation types. Approach 1 was used for onshore production emission sources that use a basin-level
approach in the national Inventory, and also for petroleum refining. Petroleum refining emissions are
allocated to states for years after 2010 using facility-level emissions reported to the GHGRP, Subpart Y.
Future state-level inventory reports may incorporate additional state- or region-specific data to improve
estimates and better reflect these differences.
2.2.3.2 Methods/Approach
To compile national Inventory estimates of GHG emissions (CH4, C02, and N20) from petroleum, EPA
compiles emissions estimates for emissions sources in each segment of a petroleum system (e.g.,
exploration, production, transport, refining) into a national total (EPA 2024, Section 3.6). Additional
information on emissions estimates and data used to develop the national-level emissions estimates for
petroleum systems is available at https://www.epa.gov/ghgemissions/natural-gas-and-petroleum-systems-
ghg-inventorv-additional-information-1990-2022-ghg.
The state-level methodology for petroleum systems follows the Hybrid approach. The production
sources that rely on Approach 1 are discussed further in the following Exploration and Production section.
2-45
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
For other industry segments and sources, national emissions from each segment are allocated to all U.S.
states, territories, and federal offshore waters (for the production segment only) using activity data sets that
have information broken out at a state level, such as the number of oil wells or volume of oil production in
each state. Where possible, these data sets are chosen to align with current activity data sets used to
develop national Inventory estimates. See Appendix B for information on the current state-level underlying
proxy data sets (i.e., Tables B-5 to B-7). The specific data sets used to disaggregate national emissions to the
state level vary by segment, as described in the following sections.
2.2.3.2.1. Exploration and Production
For the national Inventory, EPA uses emissions data collected by the GHGRP to quantify emissions for
most exploration and production sources in recent years (i.e., 2010-2022). For sources where recent data
are unavailable, and for earlier years of the time series, estimates are developed using emissions factors
from the Gas Research Institute (GRI)/EPA (1996) and Radian (1999) studies. Other key data sources for the
national estimates include oil well counts and production levels from Enverus, the Bureau of Ocean Energy
Management, and total crude oil production from EIA.
One exploration source (completions with hydraulic fracturing) and five onshore production emission
sources used information available through the updated national Inventory (EPA 2024), to implement
Approach 1 to develop state emissions: pneumatic controllers, storage tanks, equipment leaks (i.e., from
separators, heater/treaters, headers, and wellheads), chemical injection pumps, and workovers with
hydraulic fracturing. These sources relied on basin-specific emission factors and/or activity factors from
GHGRP and basin-level activity data (i.e., well counts, completion counts, and oil production) to estimate
basin emissions across the time series. The basin emissions were then directly allocated to each state using
the same activity data. The state activity data are in Appendix B, Tables B-5 and B-6.
To develop state-level emissions for other petroleum exploration and production emission sources,
national Inventory emissions were allocated to each state, primarily based on the fraction of oil wells in each
state relative to national totals across each year in the time series (Appendix B, Tables B-5 and B-6). Other
key state-level proxy data sets used to disaggregate national emissions include the number of oil well
completions without hydraulic fracturing in each state, as well as the total volume of oil produced in each
state. These state data were derived from time series of oil and gas well data from Enverus, consistent with
the Enverus data set used as activity data to derive total national emissions. For offshore activities,
emissions from state waters in the Gulf of Mexico were allocated based on relative state-level oil production
levels, while emissions from activities in federal waters were retained as a separate category (i.e., not
allocated to states). For both exploration and production segments, the data sets used for state allocation
were consistent across the entire emissions time series.
2.2.3.2.2. Crude Oil Transport
For the national Inventory, EPA estimates emissions of CH4, C02, and N20 from crude oil transport for
petroleum systems using a combination of crude oil transportation and pipeline and crude deliveries data
from EIA, the American Petroleum Institute, and the Oil and Gas Journal.
To develop state-level emissions from crude oil transport, national Inventory emissions were allocated
to each state based on three state proxy data sets. Vented emissions from marine loading were allocated to
states based on oil production from offshore wells in state waters from the Enverus data set (Appendix B,
Table B-6). Similarly, vented emissions from truck loading and rail loading were allocated based on onshore
levels of oil well production in each state. All other transport emissions, including tanks, pump stations, and
floating roof tanks, were allocated based on the relative state counts of oil refineries from GHGRP Subpart Y
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-46
-------
Section 2 Energy (NIR Chapter 3)
data after 2010 and EIA atmospheric crude oil distillation capacity for 1990-2009 (Appendix B, Table B-7), as
described in the next section.
2.2.3.2.3. Refineries
For the national Inventory, EPA uses data from the GHGRP Subpart Y and national-level activity data. All
U.S. refineries have been required to report CH4, C02, and N20 emissions for all major activities starting with
emissions that occurred in 2010. The reported total CH4, C02, and N20 emissions are used for the emissions
in each year from 2010 forward. Certain activities that are not reported to the GHGRP are estimated using
data from Radian (1999). These sources account for a small fraction of refinery emissions. To estimate
emissions for 1990-2009, the emissions data from the GHGRP, alongwith the refinery feed data, are used to
derive emissions factors that are applied to the annual refinery feed in years 1990-2009.
To develop state-level estimates for refineries for 2010-2022, national Inventory emissions from
refineries were apportioned to each state based on that state's share of refinery emissions of each gas, as
reported to GHGRP Subpart Y. This method is consistent with national Inventory estimates for refineries over
these years. For 1990-2009, national Inventory emissions from refineries were apportioned to each state
based on that state's share of national operating atmospheric crude oil distillation capacity (barrels per
calendar day), as shown in Appendix B, Table B-7 (EIA 2024).
2.2.3.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02 and CH4 from petroleum
systems was calculated using the 2006 IPCC Guidelines Approach 2 methodology (IPCC 2006). Uncertainty
estimates for N20 applied the same uncertainty bounds as calculated for C02. As described further in
Chapter 3 and Annex 7 of the national Inventory (EPA 2024), levels of uncertainty in the national estimates in
2022 were -19%/+25% for C02 and N20 and -18%/+23% for CH4.
The uncertainty estimates for the national Inventory largely account for uncertainties in the magnitude
of emissions and activity factors used to develop the national estimates for the largest contributing sources.
State-level estimates of annual emissions and removals have a higher relative uncertainty compared with
these national estimates because of the additional step of apportioning national emissions to each state
using spatial proxy data sets. This allocation method introduces additional uncertainty due to sources of
uncertainty associated with the location information in each underlying data set (e.g., number of oil wells in
each state), as well as the ability of each proxy to accurately represent the point of emission from each
source within the petroleum supply chain. Where possible, this second source of uncertainty was minimized
in the petroleum state-level analysis by selecting proxy data sets that are consistent with activity factors
used in the national Inventory. For example, national C02 and CH4 from vented emissions in the production
segment largely relied on national counts of oil wells and production volumes as activity factors; therefore,
additional uncertainty in the state-level estimates is largely associated with the uncertainty in oil well
locations. The sources of uncertainty for this category, other than refinery emissions, are also consistent
over time because the same proxy data sets were applied across the entire time series. This allocation
method, however, cannot account for state-specific mitigation programs and reduction efforts or state-
specific variations in emissions factors, which each introduce additional uncertainty in the emissions
estimates. As with the national Inventory, the state-level uncertainty estimates for this category may change
as the understanding of the uncertainty and underlying data sets and methodologies improve.
Given the variability of practices and technologies across oil and gas systems and the occurrence of
episodic events, it is possible that EPA's estimates do not include all CH4 emissions from abnormal events.
For many equipment types and activities, EPA's emissions estimates include the full range of conditions,
including "super-emitters." For other situations, where data are available, emissions estimates for abnormal
2-47
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
events were calculated separately and included in the national Inventory (e.g., Aliso Canyon leak event). EPA
continues to work through its stakeholder process to review new data from EPA's GHGRP and research
studies to assess how emissions estimates can be improved.
2.2.3.4 Recalculations
As described in Chapter 3 of the national Inventory report, some emission and sink estimates in the
national Inventory are recalculated and revised with improved methods and/or data. In general,
recalculations are made to incorporate new methodologies, or to update activity and emissions factor data
sets with the most current versions. These improvements are implemented across the previous national
Inventory's entire time series to ensure the national emission trend is accurate. See Section 3.6 of Chapter 3
in the national Inventory report for more details on recalculations in the latest national Inventory estimates.
One exploration source (completions with hydraulic fracturing) and one onshore production emission
source (workovers with hydraulic fracturing) used a new, basin-level methodology for this year's national
Inventory. As such, changes in absolute state-level emissions between this version and the previous state
report for these sources reflect, to some extent, state-specific practices and data.
As the state-level emissions are otherwise estimated using Approach 2 (national emissions are
disaggregated to the state level), changes in absolute state-level emissions between this version and the
previous state report will largely reflect recalculations and improvements implemented in the national
Inventory. Similar to the national Inventory, the calculation of state-level estimates has been updated to
incorporate updates to the underlying state-level proxy data sets. State-level proxy data sets have been
updated across the entire time series, to ensure that the state emission trends are accurate.
2.2.3.5 Planned Improvements
Potential refinements in future state-level inventories include refining state proxies used within each
segment and incorporating additional GHGRP data.
2.2.3.6 References
EIA (U.S. Energy Information Administration) (2024) Crude Oil Production. U.S. Department of Energy.
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
GRI (Gas Research Institute) and EPA (1996) Methane Emissions from the Natural Gas Industry. Available
online at: https://www.epa.gov/natural-gas-star-program/methane-emissions-natural-gas-industrv.
IPCC (Intergovernmental Panel on Climate Change) (2006) 2006IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
Radian (1999) Methane Emissions from the U.S. Petroleum Industry. U.S. Environmental Protection Agency.
2.2.4 Natural Gas Systems (NIR Section 3.7)
2.2.4.1 Background
This section describes methods used to estimate state-level C02, CH4, and N20 emissions from natural
gas systems. Similar to petroleum systems, this category includes fugitive emissions from leaks, venting, and
flaring. The U.S. natural gas system encompasses hundreds of thousands of wells, hundreds of processing
facilities, and over a million miles of gathering, transmission, and distribution pipelines. Methane and C02
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-48
-------
Section 2 Energy (NIR Chapter 3)
emissions from natural gas systems include those resulting from normal operations, routine maintenance,
and system upsets. Emissions from normal operations include natural gas engine and turbine uncombusted
exhaust, flaring, and leak emissions from system components. Routine maintenance emissions originate
from pipelines, equipment, and wells during repair and maintenance activities. Pressure surge relief systems
and accidents can lead to system upset emissions. Emissions of N20 from flaring activities are included in
the national Inventory, with most of the emissions occurring in the processing and production segments.
Note, C02 emissions exclude all combustion emissions (e.g., engine combustion) except for flaring C02
emissions. All combustion C02 emissions (except for flaring) are accounted for in the FFC section.
The methods used to develop the state-level estimates for natural gas systems follow the Hybrid
approach (a combination of Approach 1 and Approach 2), as defined in the Introduction of this report. Most
sources follow Approach 2 and rely on relative differences in basic state activity levels (e.g., gas production),
and do not reflect differences between states due to differences in practices, technologies, or formation
types. Approach 1 was used for onshore production and exploration emission sources that use a basin-level
approach in the national Inventory. Future state-level inventory reports may incorporate additional state-
specific or region-specific data to improve estimates and better reflect these differences.
2.2.4.2 Methods/Approach
To compile national estimates of CH4, C02, and N20 emissions from natural gas systems for the national
Inventory, EPA compiles emissions estimates for emissions sources in each segment of natural gas systems
(i.e., exploration, production, processing, transmission and storage, distribution, and post-meter sources)
into a national total. Additional information on emissions estimates and data used to develop the national-
level emissions estimates for natural gas systems is available online at
https://www.epa.gov/ghgemissions/natural-gas-and-petroleum-systems-ghg-inventory-additional-
information-1990-2022-ghg.
The state-level methodology for natural gas systems follows the Hybrid approach. The exploration and
production sources that rely on Approach 1 are discussed further in the following Exploration and Production
section. For other industry segments and sources, national emissions from each segment are allocated to all
U.S. states, territories, and federal offshore waters (production segment only) using activity data sets that
have information broken out at a state level, such as the number of gas wells or volume of gas produced in
each state. Where possible, these data sets are chosen to align with current activity data sets used to
develop national Inventory estimates. See Appendix B for information underlying the estimates (Tables B-8 to
B-12). The specific data sets used to disaggregate national emissions to the state level vary by segment, as
described in the following sections.
2.2.4.2.1. Exploration and Production
For the national Inventory, EPA uses emissions data collected by the GHGRP to quantify emissions for
most sources in recent years (i.e., 2011-2022) and data from a GRI/EPA 1996 study for earlier years of the
time series or for sources where recent data are unavailable. Other key data sources include data provided in
Zimmerle et al. (2019), production and well count data from Enverus, and offshore production emissions
data from the Bureau of Ocean Energy Management. Each emission source for production in the national
Inventory was generally scaled to the national level using either well counts or gas production.
One exploration emission source and six onshore production emission sources used information
available through the updated national Inventory to implement Approach 1 and develop state emissions:
pneumatic controllers, storage tanks, equipment leaks (i.e., from separators, dehydrators, heaters,
compressors, and meters/piping), liquids unloading, chemical injection pumps, workovers, and completions
(EPA 2024). These sources relied on basin-specific emission factors and/or activity factors from GHGRP and
2-49
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
basin-Level activity data (i.e., well counts, oil production, well completion counts) to estimate basin
emissions across the time series. The basin emissions were then directly allocated to each state using the
same activity data. The state activity data are in Appendix B, Tables B-8 and B-9.
To develop state-level emissions for other natural gas exploration and production emission sources,
national Inventory emissions were generally allocated to states using state-level proxy data sets that align
with the activity data used in the national Inventory (i.e., well counts, gas production). For example, state
counts of gas wells and state-level gas production were derived from time series data from Enverus,
consistent with the Enverus data used as national-level activity data in the national Inventory (see Appendix
B, Tables B-8 and B-9). Proxy data for exploration included the number of wells as well as the total number of
gas wells drilled in each state relative to the national total. Offshore emissions in the Gulf of Mexico and the
state of Alaska were allocated based on natural gas production at each platform. Additional production
emissions from offshore federal waters were not allocated to individual states but were included as a
separate total, and emissions from gathering and boosting were allocated based on the relative emissions in
each state of all other production sources. CH4 emission estimates from one-time well blowout events in
Ohio, Texas, and Louisiana were allocated to each state in the appropriate year (e.g., 60,000 metric tons in
Ohio in 2018,4,800 metric tons in Louisiana in 2019, and 49,000 metric tons in Texas in 2019). In addition,
the allocation of national Inventory estimates from produced water uses produced water volumes from
Enverus to align with the activity data used in the national Inventory. The Enverus gas well counts and
production levels were used to assign basin-level emissions estimates to the appropriate state. For both
exploration and production segments, the sources of proxy data used for state allocation were consistent
across the entire emissions time series.
2.2.4.2.2. Processing
For the national Inventory, EPA uses emissions data collected by GHGRP to quantify emissions for most
sources in recent years (i.e., 2011-2022) and data from GRI/EPA (1996) for earlier years of the time series or
for sources where recent data are unavailable. Key activity data include processing plant counts from Oil and
Gas Journal.
To develop state-level estimates for the processing segment for each year of the time series, EPA
apportioned the total national processing segment emissions to each state based on the fraction of national
onshore marketed natural gas production occurring in each state (EIA 2023), as shown in Appendix B, Table
B-10.
2.2.4.2.3. Transmission and Storage
For the national Inventory, EPA uses emissions data collected by the GHGRP and data from a Zimmerle
et al. (2015) study to quantify emissions from most sources in recent years (i.e., 2011-2022), and GRI/EPA
(1996) data for earlier years of the time series and for sources for which recent data are unavailable. Key
activity data include transmission stations (calculated using PHMSA and FERC data), storage stations
(calculated using Zimmerle et al. and EIA data), and transmission pipeline miles (PHMSA 2024).
To develop state-level estimates for the transmission and storage segment for each year of the time
series, EPA apportioned the total national transmission and storage segment emissions to each state based
on the fraction of national transmission pipeline mileage occurring in each state (Appendix B, Table B-11). In
the national Inventory, CH4 emissions from anomalous events are added to storage emission totals in several
years. In the state-level estimates, these emissions are allocated to the state in which the event occurred,
while remaining emissions from storage wells are allocated based on the relative transmission pipeline
mileage in each state.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-50
-------
Section 2 Energy (NIR Chapter 3)
2.2.4.2.4. Distribution
For the national Inventory, EPA uses data collected by the GHGRP and data from a Lamb et al. (2015)
study to quantify emissions from most sources in recent years (i.e., 2011-2022) and GRI/EPA (1996) data for
earlier years of the time series or for sources for which recent data are unavailable. Key activity data include
pipeline mileage by material from PHMSA, metering and regulation station counts from Subpart W of the
GHGRP, and number of natural gas residential, commercial, and industrial consumers from EIA.
To develop state-level estimates for the distribution segment for each year of the time series, the EPA
national total emissions from pipeline leaks were allocated based on the relative pipeline mileage by
material (cast iron, unprotected/protected steel, plastic) in each state, the relative number of natural gas
residential, commercial, and industrial consumers in each state from EIA, and the number of above- and
below-grade stations in each state as reported to the GHGRP (scaled up by the ratio of PHMSA to GHGRP
pipeline mileage in each state to include non-reporters). Complete PHMSA data are available starting in 2003
and GHGRP data are available for all years starting in 2011. For all earlier years, national emissions were
allocated using the same relative state contributions as those values in the earliest available years (e.g.,
relative state-level pipeline mileage amounts held constant before 2003), as shown in Appendix B, Table B-
12.
2.2.4.2.5. Post-meter Sources
For the national Inventory, post-meter sources include leak emissions from residential and commercial
appliances, industrial facilities and power plants, and natural gas-fueled vehicles. Leak emissions from
residential appliances and industrial facilities and power plants account for the majority of post-meter CH4
emissions. C02 emissions from residential appliances are included in the natural gas residential source
within the energy sector and are not accounted for here. There are no N20 emissions from the post-meter
segment. Key activity data include the counts of homes in the United States with natural gas appliances from
the American Housing Survey national data set, the number of commercial natural gas customers from EIA,
natural gas consumption volumes for industrial and electric generating units from EIA, and counts of
compressed natural gas vehicles from the EPA MOVES model.
To develop state-level estimates for post-meter emissions for each year of the time series, the EPA
national total emissions from residential and commercial appliances were allocated to states usingthe
relative number of residential and commercial natural gas customers in each state from EIA. Industrial and
electric generating unit emissions were allocated based on the relative consumption volumes from the EIA
SEDS, and compressed natural gas vehicles were allocated to the number of compressed natural gas
vehicles in each state, derived from the MOVES model. These proxy data sets are generally consistent with
the activity data sets used in the national Inventory, except for residential emissions, which are allocated
based on data from EIA rather than the American Housing Survey due to the limited state-level information in
the survey data set. The same proxy data sets are used across the entire time series for this segment.
2.2.4.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02 and CH4 from natural gas
systems was calculated usingthe 2006 IPCC Guidelines Approach 2 methodology (IPCC 2006). Uncertainty
estimates for N20 applied the same uncertainty bounds as C02. As described further in Chapter 3 and Annex
7 of the national Inventory (EPA 2024), levels of uncertainty in the national estimates in 2022 were
-12%/+15% for C02 and N20 and 18%/+17% for CH4.
The uncertainty estimates for the national Inventory largely account for uncertainty in the magnitude of
emissions and activity factors used to develop the national estimates for the largest contributing sources.
State-level estimates of annual emissions and removals have a higher relative uncertainty compared with
2-51
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
these national estimates due to the additional step of apportioning national (or basin-level as applicable)
emissions to each state using spatial proxy data sets. This allocation method introduces additional
uncertainty due to sources of uncertainty associated with the location information in each underlying data
set (e.g., number of non-associated gas wells in each state), as well as the ability of each proxy to accurately
represent the point of emission from each source within the natural gas supply chain. Where possible, this
second source of uncertainty is minimized in the natural gas state-level analysis by selecting proxy data sets
that are consistent with activity factors used in the national Inventory. However, this is not always possible
when activity factor data sets only include national aggregate statistics. For example, national C02 and CH4
emissions from natural gas processing largely rely on national-level plant counts. In the state-level
estimates, these emissions are allocated based on the share of national gas production from each state and
will therefore include additional uncertainty associated with the accuracy of the state-specific data in the gas
production data set as well as the accuracy with which relative state-level gas production reflects the relative
state-level emissions from natural gas processing plants. In contrast, the national Inventory estimates for
sources within the natural gas production segment typically use national well counts and production
volumes as activity factors. Therefore, additional uncertainty in the state-level estimates for these sources
will largely be the spatial representation of gas wells in the activity factor data set. The sources of uncertainty
for this category are also consistent over time because the same proxy data sets are applied across the
entire time series. This allocation method, however, cannot account for state-specific mitigation programs
and reduction efforts or state-specific variations in emissions factors, which each introduce additional
uncertainty in the emissions estimates. As with the national Inventory, the state-level uncertainty estimates
for this category may change as the understanding of the uncertainty of estimates and underlying data sets
and methodologies improves.
Given the variability of practices and technologies across oil and gas systems and the occurrence of
episodic events, it is possible that EPA's estimates do not include all methane emissions from abnormal
events. For many equipment types and activities, EPA's emissions estimates include the full range of
conditions, including "super-emitters." For other situations, where data are available, emission estimates
for abnormal events were calculated separately and included in the national Inventory (e.g., Aliso Canyon
leak event and the three well blowout events included for the first time in the 2022 national Inventory). EPA
continues to work through its stakeholder process to review new data from EPA's GHGRP and research
studies to assess how emissions estimates can be improved.
2.2.4.4 Recalculations
As described in Chapter 3 of the national Inventory report, some emission and sink estimates in the
national Inventory are recalculated and revised with improved methods and/or data. In general,
recalculations are made to incorporate new methodologies, or to update activity and emissions factor data
sets with the most current versions. These improvements are implemented across the previous national
Inventory's entire time series to ensure that the national emission trend is accurate. See Section 3.7 of
Chapter 3 in the national Inventory report for more details on recalculations in the latest Inventory estimates.
A new, basin-level methodology was used for completions and workovers for this year's national
Inventory. As such, changes in absolute state-level emissions between this version and the previous state
report for these sources reflect to some extent state-specific practices and data.
As the state-level emissions are otherwise estimated using Approach 2 (national emissions are
disaggregated to the state level), changes in absolute state-level emissions between this version and the
previous state report largely reflect recalculations and improvements implemented in the national Inventory.
See Chapter 3 in the national Inventory report for further details on these updates in the national Inventory.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-52
-------
Section 2 Energy (NIR Chapter 3)
To align with these methodological improvements in the national Inventory, methodological updates to
the state estimates, relative to the previous version, include incorporating the use of the basin-level
emissions estimates developed in the national Inventory for certain exploration and production sources as
described above (Exploration and Production section). These new sources have been allocated to the state
level following the approaches described in the segment-specific sections above.
For other sources, the calculation of state-level estimates has been updated to incorporate updates to
the underlying state-level proxy data sets, following the same procedure as in the national Inventory. State-
level proxy data sets have been updated across the entire time series to ensure that the state-emission
trends are accurate.
2.2.4.5 Planned Improvements
Potential refinements to exploration and production estimates in future state-level inventories include
refining state proxies used for individual sources within each segment and incorporating additional GHGRP
data for allocating emissions within the production segment.
Potential refinements to processing estimates in future state-level inventories include using emissions
levels reported to the GHGRP (along with other data) to apportion emissions to each state. In addition,
information on processing plant locations from other data sets or use of Oil and Gas Journal or EIA data on
gas processing volumes could be incorporated to improve estimates. Potential refinements to transmission
and storage estimates in future state-level inventories include using emissions levels reported to the GHGRP
(along with other data) to apportion emissions to each state. In addition, information on transmission and
storage station locations from other data sets could be incorporated to improve estimates.
2.2.4.6 References
EIA (U.S. Energy Information Administration) (2023) Natural Gas Gross Withdrawals and Production:
Marketed Production. U.S. Department of Energy.
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
GRI (Gas Research Institute) and EPA (1996) Methane Emissions from the Natural Gas Industry. Available
online at: https://www.epa.gov/natural-gas-star-program/methane-emissions-natural-gas-industry.
IPCC (Intergovernmental Panel on Climate Change) (2006) 2006IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
Lamb, B.K., S.L. Edburg, T.W. Ferrara, T. Howard, M.R. Harrison, C.E. Kolb, A. Townsend-Small, W. Dyck, A.
Possolo, and J.R. Whetstone (2015) Direct Measurements Show Decreasing Methane Emissions from
Natural Gas Local Distribution Systems in the United States. Environmental Science and Technology, 49:
5161-5169.
PHMSA (Pipeline and Hazardous Materials Safety Administration) (2023) Gas Distribution, Gas Gathering,
Gas Transmission, Hazardous Liquids, Liquefied Natural Gas (LNG), and Underground Natural Gas
Storage (UNGS) Annual Report Data. U.S. Department of Transportation. Available online at:
https://www.phmsa.dot.gov/data-and-statistics/pipeline/gas-distribution-gas-gathering-gas-
transmission-hazardous-liquids.
2-53
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Zimmerle, D.J., L.L. Williams, T.L. Vaughn, C. Quinn, R. Subramanian, G.P. Duggan, B. Willson, J.D. Opsomer,
A.J. Marchese, D.M. Martinez, and A.L. Robinson (2015) Methane Emissions from the Natural Gas
Transmission and Storage System in the United States. Environmental Science and Technology, 49:
9374-9383.
Zimmerle, D., K. Bennett,T.Vaughn, B. Luck,T. Lauderdale, K. Keen, M. Harrison,A. Marchese, L.Williams,
and D. Allen (2019) Characterization of Methane Emissions from Gathering Compressor Stations: Final
Report. U.S. Department of Energy. Available online at: https://www.osti.gov/servlets/purl/1506681.
2.2.5 Abandoned Oil and Gas Wells (NIR Section 3.8)
2.2.5.1 Background
This section describes methods used to estimate state-level C02 and CH4 emissions from abandoned
oil and gas wells. The term "abandoned wells" encompasses various types of wells, including orphaned
wells and other nonproducing wells such as:
Wells with no recent production, and that are not plugged. Common terms (such as those used in
state databases) might include inactive, temporarily abandoned, shut-in, dormant, and idle.
Wells with no recent production and no responsible operator. Common terms might include
orphaned, deserted, long-term idle, and abandoned.
Wells that have been plugged to prevent migration of gas or fluids.
The U.S. population of abandoned wells, including orphaned wells and other nonproducing wells, is
around 3.9 million (with around 3.0 million abandoned oil wells and 0.9 million abandoned gas wells). The
methods to calculate emissions from abandoned wells involved calculating the total populations of plugged
and unplugged abandoned oil and gas wells in the United States. An estimate of the number of orphaned
wells within this population is not developed as part of the methodology for the national- or state-level
inventories. Other groups have developed estimates of the total national number of orphaned wells. The
Interstate Oil and Gas Compact Commission, for example, estimates 92,198 orphaned wells in the United
States (IOGCC 2021). State applications for grants to plug orphaned wells indicate over 130,000 orphaned
wells in the United States (U.S. Department of the Interior 2022).
The state-level methodology for abandoned oil and gas wells follows Approach 1, as defined in the
Introduction of this report, where emissions from this segment are calculated for each U.S. state in the
methodology used to develop the national Inventory using activity data sets with information broken out at
the state level, includingwell counts, type (e.g., oil, gas), and plugging status. See Appendix B, Table B-13, for
the underlying data sets.
2.2.5.2 Methods/Approach
To compile national estimates of CH4 and C02 emissions from abandoned oil and gas wells for the
national Inventory, EPA develops emissions estimates for plugged and unplugged abandoned wells for each
state and sums to the national level. Key data sources are two research studiesKang et al. (2016) and
Townsend-Small et al. (2016)for emissions factors, as well as the Enverus database and historical state-
level data sets for abandoned well counts.
To develop state-level estimates of GHG emissions from abandoned natural gas and oil wells when
developing the national Inventory, an estimate of the number of abandoned wells in each state (developed
using Enverus and historical data sets), as well as their type (oil versus gas) and plugging status (plugged
versus unplugged) were estimated across the time series. Well type and plugging status were derived from
Enverus. The applicable emission factor was then applied to the state activity data to estimate emissions for
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-54
-------
Section 2 Energy (NIR Chapter 3)
each state. State-Level counts of abandoned oil and natural gas wells (which include all nonproducing wells,
not only orphaned wells) are available in Appendix B, Table B-13.
2.2.5.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of both C02 and CH4 from
abandoned oil and gas wells were each calculated usingthe 2006 IPCC Guidelines Approach 2 methodology
(IPCC 2006). As described further in Chapter 3 and Annex 7 of the national Inventory (EPA 2024), levels of
uncertainty in the national estimates in 2022 for both abandoned oil and gas wells were -83%/+204% for C02
and -83%/+204% for CH4.
The uncertainty estimates for the national Inventory account for uncertainty in the magnitude of
emissions and activity factors used to develop the national estimates. State-level estimates of annual
emissions and removals have a higher relative uncertainty compared with these national estimates, for
example, due to regional emission factors that may not reflect state-specific emissions. The sources of
uncertainty for this category are generally consistent over time, and the same data sets were used across the
entire time series. The uncertainty method cannot account for state-specific variations in emissions factors,
which would introduce additional uncertainty in the emissions estimates. As with the national Inventory, the
state-level uncertainty estimates for this category may change as the understanding of the uncertainty of
estimates and underlying data sets and methodologies improves.
2.2.5.4 Recalculations
As described in Chapter 3 of the national Inventory report, some emission and sink estimates in the
national Inventory are recalculated and revised with improved methods and/or data. In general,
recalculations are made to incorporate new methodologies, or to update activity and emissions factor data
sets with the most current versions. These improvements are implemented across the previous national
Inventory's entire time series to ensure the national emission trend was accurate. See Chapter 3 in the
national Inventory report for more details on recalculations in the latest national Inventory estimates.
The abandoned oil and natural gas well counts were updated using revised data from Enverus. Changes
in absolute state-level emissions between this version and the previous state report will reflect these
recalculations implemented in the national Inventory.
2.2.5.5 Planned Improvements
Potential refinements include incorporating improved state-level abandoned well counts for each year
of the time series.
2.2.5.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
IOGCC (Interstate Oil and Gas Compact Commission) (2021) Idle and Orphan Oil and Gas Wells: State and
Provincial Regulatory Strategies 2021. Available online at:
https://iogcc.ok.goV/sites/g/files/gmc836/f/iogcc idle and orphan wells 2021 final web.pdf.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
2-55
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Kang, M., S. Christian, M.A. Celia, and R.B. Jackson (2016) Identification and Characterization of High
Methane-Emitting Abandoned Oil and Gas Wells. PNAS, 113(48): 13636-13641. Available online at:
https://doi.org/10.1073/pnas.1605913113.
Townsend-Small, A., T.W. Ferrara, D.R. Lyon, A.E. Fries, and B.K. Lamb (2016) Emissions of Coalbed and
Natural Gas Methane from Abandoned Oil and Gas Wells in the United States. Geophysical Research
Letters, 43:1789-1792.
U.S. Department of the Interior (2022) Overwhelming Interest in Orphan Well Infrastructure Investments.
Available online at: https://content.govdelivery.com/accounts/USDQI/bulletins/30416b5.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
2-56
-------
3 Industrial Processes and Product Use (NIR Chapter 4)
For this methodology report, the IPPU sector is organized into four subsectors: minerals, chemicals,
metals, and product use. For more information on IPPU sector emissions, see Chapter 4 of the national
Inventory. Table 3-1 summarizes the different approaches used to estimate state-level IPPU sector emissions
and completeness. Geographic completeness is consistent with the national Inventory. The sections below
provide more detail on each category.
Table 3-1. Overview of Approaches for Estimating State-Level IPPU Sector GHG Emissions
Category
Gas
Approach
Geographic Completeness8
Cement
Production
C02
Hybrid approach
2010-2022: Approach 2
1990-2009: Approach 1
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories (i.e., Puerto Rico) as applicable.
Lime Production
co2
Approach 2
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories (i.e., Puerto Rico) as applicable.
Glass
Production
co2
Approach 2
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories (i.e., Puerto Rico) as applicable.
Other Process
Uses of
Carbonates
co2
Hybrid approach
Non-metallurgical
magnesia production:
Approach 1
All other subcategories:
Approach 2
Includes emissions from all states. Except
for ceramics production, other
subcategories also include emissions from
the District of Columbia, tribal lands, and
territories8 (i.e., American Samoa, Guam,
Northern Mariana Islands, Puerto Rico, and
U.S. Virgin Islands) as applicable.
Carbon Dioxide
Consumption
co2
Approach 2
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories8 (i.e., American Samoa, Guam,
Northern Mariana Islands, Puerto Rico, and
U.S. Virgin Islands) as applicable.
Ammonia
Production
co2
Approach 2
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories8 as applicable.
Urea
Consumption
for
Nonagricultural
Purposes
co2
Approach 2
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories8 (i.e., Puerto Rico, American
Samoa, Guam, Northern Mariana Islands,
and U.S. Virgin Islands) as applicable.
Nitric Acid
Production
N20
Approach 2
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories8 as applicable.
Adipic Acid
Production
n2o
Approach 1
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories8 as applicable.
Caprolactam,
Glyoxal and
Glyoxylic Acid
Production
n2o
Approach 2
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories8 as applicable.
3-1
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Category
Gas
Approach
Geographic Completeness8
Carbide
Production and
Consumption
C02
ch4
Hybrid approach
Production: Approach 1
Consumption: Approach 2
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories8 (i.e., Puerto Rico) as applicable.
Titanium Dioxide
Production
C02
Approach 2
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories8 as applicable.
Soda Ash
Production
C02
Approach 1
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories8 as applicable.
Petrochemical
Production
C02
ch4
Approach 2
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories8 as applicable.
HCFC-22
Production
HFC-
23
Hybrid approach
2010-2021: Approach 1
1990-2009: Approach 2
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories8 as applicable.
Production of
Fluorochemicals
Other than
HCFC-22
HFCs
PFCs
SFs
nf3
Hybrid approach
Large facilities: Approach 1
Small facilities: Approach
2
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories8 as applicable.
Phosphoric Acid
Production
C02
Approach 2
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories8 as applicable.
l&S Production
and
Metallurgical
Coke Production
C02
ch4
Approach 2
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories8 as applicable.
Ferroalloys
Production
co2
ch4
Approach 2
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories8 as applicable.
Aluminum
Production
co2
PFCs
Hybrid approach
2010-2022: Approach 1
1990-2009: Approach 2
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories8 as applicable.
Magnesium
Production and
Processing
C02
SFs
HFCs
Hybrid approach
1999-2022: Approach 1 &
2
1990-1998: Approach 2
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories8 as applicable.
Lead Production
C02
Approach 2
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories8 as applicable.
Zinc Production
C02
Approach 2
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories8 as applicable.
Electronics
Industry
N20
nf3
sf6
HFCs
PFCs
Hybrid approach
2011-2022: Approach 1 &
2
1990-2010: Approach 2
Includes emissions from all states, the
District of Columbia, tribal lands, and
territories8 as applicable.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-2
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
Category
Gas
Approach
Geographic Completeness8
Substitution of
Ozone-
Depleting
Substances
Includes emissions from all states, the
HFCs
PFCs
District of Columbia, tribal lands, and
Hybrid approach
territories8 (i.e., American Samoa, Guam,
Northern Mariana Islands, Puerto Rico, and
U.S. Virgin Islands) as applicable
Electrical
Transmission
and Distribution
Hybrid approach
Includes emissions from all states, the
SFs
2011-2022: Approach 1 &
District of Columbia, and territories8 (i.e.,
2
1990-2010: Approach 2
Puerto Rico, U.S. Virgin Islands, and Guam)
as applicable.
Hybrid approach (varies by
application)
SFs and PFCs
SFs,
PFCs
Military applications:
Includes emissions from all states, the
from Other
Approach 2
District of Columbia, tribal lands, and
Product Use
Scientific applications:
Approach 1 & 2 pending
data availability
territories8 as applicable.
Includes emissions from all states, the
Nitrous Oxide
District of Columbia, tribal lands, and
from Product
N20
Approach 2
territories8 (i.e., American Samoa, Guam,
Uses
Northern Mariana Islands, Puerto Rico, and
U.S. Virgin Islands) as applicable.
a Emissions may be occurring in other U.S. territories; however, due to a lack of available data and the nature of this
category, this analysis includes emissions for only the territories indicated. Territories not listed are not estimated, but in
most instances emissions are likely not occurringfor categories covered in this chapter.
3.1 Minerals
This section presents the methodology used to estimate the minerals portion of IPPU emissions, which
consist of the following sources:
Cement production (C02)
Lime production (C02)
Glass production (C02)
Other process uses of carbonates (C02)
C02 consumption (C02)
3.1.1 Cement Production (NIR Section 4.1)
3.1.1.1 Background
Cement production is an energy- and raw material-intensive process that results in the generation of
C02 both from the energy consumed in making the clinker precursor to cement and from the chemical
process to make the clinker. Emissions from fuels consumed for energy purposes during the production of
cement are accounted for in the energy sector. Process emissions from cement production are based
primarily on clinker production. During the clinker production process, the key reaction occurs when calcium
carbonate, or CaC03, in the form of limestone or similar rocks, is heated in a cement kiln at a temperature
range of about 700 to 1,000 °C (1,300 to 1,800 °F) to form lime (i.e., calcium oxide [CaO]) and C02 in a
process known as calcination or calcining. The quantity of C02 emitted during clinker production is directly
proportional to the lime content of the clinker. During clinker production, some of the raw materials, partially
3-3
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
reacted raw materials, and clinker enter the kiln line's exhaust system as non-calcinated, partially
calcinated, or fully calcinated cement kiln dust (CKD). To the degree that the CKD contains carbonate raw
materials that are returned to the kiln and calcined, there are associated C02 emissions.
Cement is produced in 34 states and Puerto Rico; in descending order, production is most concentrated
in Texas, Missouri, California, and Florida (EPA 2023). In 2022, these four leading cement-producing states
accounted for nearly 43% of U.S. production (USGS 2023).
3.1.1.2 Methods/Approach
To develop state-level estimates of emissions from cement production, national emissions from the
national Inventory were disaggregated using a combination of facility-level emissions data reported to the
GHGRPfrom 2010 to 2022 (EPA 2023) and U.S. Geological Survey (USGS)'s Mineral Commodity Summary
clinker production data for 1990-2009 (EPA 2024), as shown in Table 3-2. See Appendix C, Tables C-1 and C-2
in the "Cement" Tab, for more details on the data used.
This Hybrid approach, as defined in the Introduction chapter of this report, is used due to limitations in
the availability of state-specific activity data for the time series. While GHGRP clinker production data by
state are considered confidential business information (CBI), emissions data by state are not confidential,
and therefore are available for this analysis starting in 2010. State-level emissions of C02 from cement
production were calculated using the Tier 2 method provided by the 2006 IPCC Guidelines (IPCC 2006).
Table 3-2. Summary of Approaches to Disaggregate the National Inventory for Cement Production
Across Time Series
Time Series
Range
Summary of Method
2010-2022
Applied national Inventory emissions factors to clinker production data
estimated using GHGRP emissions data (IPCC 2006 Tier 2).
1990-2009
Applied the national Inventory emissions factors to actual and estimated clinker
production data from USGS (IPCC 2006 Tier 2).
The method used for 2010-2022 (Approach 2) was based on state-level emissions data from the GHGRP
to allocate clinker production by state. Facilities that use the Continuous Emissions Monitoring System
(CEMS) to measure emissions reported combined combustion and process emissions to GHGRP, while
facilities that do not use CEMS reported their process and combustion emissions separately. Usingthe data
from facilities that do not use CEMS, average annual process emissions factors were estimated and applied
to the CEMS emissions data to estimate process-only emissions by state. Those process emissions by state
were converted into a percentage of national process emissions and applied to national clinker production
data to estimate state-level clinker production. Under the GHGRP, any facility that manufactures Portland
cement must report their GHG emissions regardless of the level of emissions.
The method used for 1990-2009 (Approach 1) relied on USGS clinker production data, which is the
same data source for the national Inventory. At the state level, USGS reports clinker production for a few
individual states and combines other states in groups of two to four to protect company proprietary data.
Because of limited information about clinker production or other relevant proxy data by state, production for
grouped states was evenly divided among the states in each group to estimate clinker production.
National emissions factors for C02 from clinker production and cement kiln dust from the national
Inventory were applied to state clinker production to calculate GHG emissions by state.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-4
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
3.1.1.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02 from cement production was
calculated using the 2006 IPCC Guidelines Approach 2 methodology for uncertainty (IPCC 2006). As
described further in Chapter 4 and Annex 7 of the national Inventory (EPA 2024), levels of uncertainty in the
national estimates in 2022 were -4%/+5% for C02.
State-level estimates are expected to have an overall higher uncertainty because the national emissions
estimates were apportioned to each state based on a combination of state-level clinker production data from
the same source used in the national Inventory and GHGRP emissions data by state as a surrogate for clinker
production data. These assumptions were required because of a general lack of more granular state-level
data.
For the 2010-2022 period, GHGRP emissions by state were used to apportion clinker production over
individual states. Over 90% of the cement facilities use CEMS to measure C02 emissions, which includes
combustion emissions as well as process emissions. Using the data from facilities that do not use CEMS,
average annual process emissions factors were estimated and applied to the CEMS emissions data to
estimate process-only emissions by state. Although this approach approximates GHG emissions from
CEMS-monitored kilns, it is not possible to determine whether emissions are overestimated or
underestimated.
While USGS reports the clinker production for a few individual states, most state clinker production is
combined with the clinker production of multiple other states to protect sensitive production data of
individual facilities. For 1990-2009, the method of apportioning the grouped clinker production evenly among
individual states to estimate state GHG emissions likely results in overestimating emissions for some states
and underestimating emissions for others. On a national scale, GHGRP clinker production closely
approximates that reported by USGS.
3.1.1.4 Recalculations
No recalculations were applied for this current report consistent with the national Inventory (see
Section 4.1, page 4-14).
3.1.1.5 Planned Improvements
An important data gap is the production of clinker by each cement-producing state for the full time
series of 1990-2022. The USGS Minerals Yearbook series reports clinker production data for 11 individual
states and Puerto Rico; the remainder of the clinker production data are reported for groups of states to
protect industry-sensitive data. EPA will assess whether industry gross domestic product (GDP) per state or
other state-level data would provide a better way to disaggregate this grouped data. Clinker capacity by
facility for these states was considered, but incomplete data on clinker capacity limited the ability to
estimate clinker production in these groups of states. Additionally, cement kilns do not typically operate at
100% capacity for an entire year, and utilization rates vary from kiln to kiln, facility to facility, and year to year.
Furthermore, EPA is looking to reflect changes occurring in the cement industry to modernize production
methods that affect process emissions (e.g., improve kiln efficiency and capacity). These and other factors
will be examined to identify improvements in the methods used to estimate state-level GHG emissions.
3.1.1.6 References
EPA (U.S. Environmental Protection Agency) (2023) Facility Level Information on GreenHouse gases Tool
(FLIGHT). Data set as of August 18, 2023. Available online at: https://ghgdata.epa.gov/ghgp/.
3-5
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
EPA (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2022. EPA430-R-24-004. Available
online at: https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
USGS (U.S. Geological Survey) (2023) Mineral Commodity Summaries: Cement. Available online at:
https://pubs.usgs.gov/periodicals/mcs2023/mcs2023-cement.pdf.
3.1.2 Lime Production (NIR Section 4.2)
3.1.2.1 Background
Lime is an important manufactured product with many industrial, chemical, and environmental
applications. Lime production involves three main processes: stone preparation, calcination, and hydration.
C02is generated during the calcination stage, when limestoneconsisting of calcium carbonate (CaC03)
and/or magnesium carbonate (MgC03)is roasted at high temperatures in a kiln to produce calcium oxide
(CaO) and C02. The C02 is given off as a gas and is normally emitted into the atmosphere. Emissions are also
generated with the formation of calcined waste produced during lime production, primarily lime kiln dust
(LKD) and also off-spec lime, scrubber sludge, and other miscellaneous waste. Some of the C02 generated
during the production process, however, is recovered at some facilities for use in sugar refining and
precipitated calcium carbonate production. Emissions from fuels consumed for energy purposes during lime
production are included in the energy sector. Lime production emissions from the national Inventory were
disaggregated to 28 states in 2022. Emissions are attributed to only 23 states, as facilities in five of the states
(Colorado, Idaho, Minnesota, North Dakota, and Nebraska) produce beet sugar and all C02 is considered
recovered under the methodology below.
3.1.2.2 Methods/Approach
National estimates were downscaled across states because of limitations in availability of state-
specific data across the time series needed to apply national methods (i.e., IPCC Tier 2 methods) at the state
level. The Approach 2 methodology allocated gross process emissions from lime production to each
producing state using a combination of process emissions reported to the GHGRP and the number of
facilities in a state as surrogates for lime production data. The number of facilities in a state that captured
C02 for use in on-site processes was then used to calculate captured process emissions, which was
subtracted from gross emissions to estimate net process emissions, as shown in Table 3-3. The sum of
emissions by state is consistent with national process emissions as reported in the national Inventory. See
Appendix C, Tables C-3 through C-6 in the "Lime" Tab, for more details on the data used.
Table 3-3. Summary of Approaches to Disaggregate the National Inventory for Lime Production
Across Time Series
Time Series
Range
Summary of Method
2010-2022
GHGRP process emissions data were used to estimate the percentage of gross
emissions by state, multiplied by the national emissions (IPCC 2006 Tier 2).
GHGRP data on number and type of facilities that captured C02 for use in on-site
processes were used to estimate the C02 emissions captured and subtracted
from gross emissions to get net emissions from lime production.
1990-2009
USGS data on number of lime facilities were used to estimate the percentage of
lime production by state, multiplied by the national emissions (IPCC 2006 Tier 2).
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-6
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
Time Series
Range
Summary of Method
GHGRP data on number of facilities that captured C02 for use in on-site
processes from 2010 to 2019 were used to estimate the percentage of emissions
captured, multiplied by national emissions and subtracted from gross emissions
to get net emissions from lime production.
The methodology used for 2010-2022 was based on process emissions data reported to the GHGRP
summed by state (EPA 2010-2022) to calculate a percentage of gross emissions from each state. That
percentage was then applied to the national emissions from lime production per year to calculate
disaggregated gross C02 emissions by state. The GHGRP has a reporting threshold of 25,000 metric tons of
C02 equivalent for lime production, so these emissions data are representative of the larger facilities in the
industry. Using GHGRP emissions data means that emissions from states with smaller facilities were
possibly underestimated.
The methodology used for 1990-2009 was based on dividing the number of facilities in each state by the
number of facilities nationally to calculate a percentage of total U.S. facilities in each state for each year.
This percentage was applied to the gross national C02 emissions from lime production per year (EPA 2024a)
to calculate disaggregated gross C02 emissions by state for each year. The number of facilities per state was
compiled from the USGS Minerals Yearbooks for Lime's "Lime Sold or Used by Producers in the United
States, by State" table (USGS 1991,1992-2010). For some years, USGS aggregated the number of facilities
for some states to avoid disclosing proprietary information related to individual facility production. For those
states and years, the individual state facility counts were estimated based on the knowledge of facility
locations in 2010-2019 and the number of facilities in a state reported in the USGS Minerals Yearbook for
Lime, Table 2, when that state was not aggregated. In the absence of state-specific activity data, using the
number of facilities per state to determine the state allocation percentage assumes that each facility has the
same amount of input and output.
The USGS Mineral Commodity Summaries for lime (1996-2023) only contain U.S. total lime production,
with no breakdown by lime type or state. While the USGS Minerals Yearbooks for Lime (1991-2021) have
hydrated and quicklime production data by region (Northeast, Midwest, South Atlantic, East South Central,
West South Central, and West), additional detail by high-calcium or dolomitic lime or by individual states is
not available, and these data could not be used as activity data in the state disaggregation estimates. Thus,
the following activity data were not available by state from current data sources used to estimate national
emissions (USGS Minerals Yearbooks): lime production data for high-calcium quicklime; dolomitic
quicklime; high-calcium, hydrated; dolomitic, hydrated; dead-burned dolomite; and C02 captured on-site.
As such, these data could not be used as activity data in the state disaggregation estimates.
Although the national Inventory value was adjusted to account for C02 emissions from the production of
LKD, the state disaggregated values do not account for specific facility per state-level C02 emissions from
the production of LKD. The adjustment to the national Inventory value was spread equally across the states
with facilities. In addition, the national Inventory value was not adjusted to account for C02 emissions from
other waste production (e.g., off-spec lime, scrubber sludge, other miscellaneous site-specific waste).
3.1.2.2.1. CEMS Adjustment for2010-2022
In 2010, facilities producing lime started reporting both process and combustion emissions to the
GHGRP. For facilities using a CEMS approach to measure and report C02 emissions, a combined total value
for process and combustion emissions were reported together under Subpart S; otherwise, facilities reported
process emissions under Subpart S and combustion emissions under Subpart C using engineering and
3-7
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
calculation approaches. To disaggregate process emissions for those facilities reporting C02with CEMS, an
industrywide ratio of process emissions to total emissions for facilities that do not report using CEMS was
calculated for each year from 2010 to 2022. While some facilities produce lime as a secondary product,
facilities using CEMS were found to produce lime as a primary product with a primary North American
Industry Classification System (NAICS) code of 327410 for lime manufacturing. Emissions reported to
Subparts S and C were compiled for all facilities with this NAICS code, and the ratio of process emissions to
total emissions for non-CEMS facilities was applied to the total C02 emissions for each CEMS facility to
calculate process emissions for each year that emissions were reported using CEMS. The results were an
estimated process C02 emissions-only value for that CEMS facility.
Because the methodology for 1990-2009 does not use GHGRP emissions data to calculate the state
emissions, there is no need to adjust for CEMS facilities for those years.
3.1.2.2.2. Adjustment for C02 Captured for Use in On-Site Processes
Some facilities recover C02 generated during the lime production process for use in sugar refining and
precipitated calcium carbonate (PCC) production. Emissions from lime use for sugar refining are reported
under Section 3.1.4, Other Process Uses of Carbonates. PCC is used as a filler or coating in the paper, food,
and plastic industries and is derived from reacting hydrated high-calcium quicklime with C02. Per the 2006
IPCC Guidelines, it is assumed that the recovery of C02 for use in the sugar refining process and PCC
production does not result in net emissions of C02 to the atmosphere. Consistent with the national Inventory
methodology, gross emissions per state from lime production were adjusted to subtract the amount of C02
captured for use in on-site processes such as purification.
For 2010-2022, although the quantity of C02 captured on-site at a facility was reported to the GHGRP,
these data are considered confidential business information (CBI) and are not available by facility or state;
they are, however, available at the aggregated national level and are used in the national Inventory.
Information on which facilities captured C02 for on-site use in 2010-2022 and the states where these
facilities are located is publicly available through the GHGRP. The GHGRP indicator of C02 capture on-site,
along with each facility's reported primary NAICS code, were used to identify two types of facilities capturing
C02 on-site: beet sugar manufacturing (NAICS 311313) and lime manufacturing (NACIS 327410). For beet
sugar manufacturing facilities capturing C02 on-site in 2010-2022, all process emissions generated from the
lime kiln were assumed to be captured and used on-site for further beet sugar manufacturing, resulting in net
zero C02 emissions. Note that some states with beet sugar manufacturing facilities that capture C02 also
have additional facilities that do not capture C02, resulting in net C02 emissions greater than zero.
To estimate the quantity of C02 captured for beet manufacturing facilities per state, per year for 2010-
2022, each facility's reported GHGRP process C02 emissions per year were divided by the total annual
GHGRP process C02 value per year. The facility percentage values were summed by state and applied to the
national Inventory gross C02 emissions value. The resulting state quantities of C02 captured for beet
manufacturing facilities were summed for a total value of C02 captured for beet sugar manufacturing
facilities, which was subtracted from the GHGRP national captured C02 value to calculate the quantity of
captured C02 at lime manufacturing plants. The quantity of captured C02 for lime manufacturing facilities
was divided by the total number of lime manufacturing facilities capturing C02 per year to calculate a per-
facility C02 captured value per year. The lime manufacturing per-facility C02 captured value was then
allocated to each lime manufacturing plant that captures C02 per state and year.
For the years 1990-2009, because of a lack of available data on both the quantity of C02 captured on-
site at facilities per state for all years and on the number of facilities that captured C02 on-site in 2009, an
alternative methodology was devised to estimate the quantity of emissions captured, based on available
GHGRP data. The number of facilities that captured C02 for on-site use over the years 2010-2019 and their
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-8
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
Locations were used to estimate the number of facilities in each state that captured C02 for use in on-site
processes in 1990-2009. The number of facilities that captured C02 on-site in a state was divided by the total
number of facilities in the state for each year from 2010-2019 to calculate a percentage of facilities in the
state capturing C02. The annual percentages for 2010-2019 were averaged and then applied to the number
of facilities per state for each year in 1990-2009 to estimate the number of facilities per state that captured
C02 on-site.
In the absence of available state or facility data, the current methodology for the years 1990-2009
distributed annual C02 captured on-site evenly among all facilities that reported capturing C02 on-site to the
GHGRP, assuming that all facilities that captured C02 on-site captured the same quantity of emissions each
year. To estimate the quantity of C02 captured on-site for the yearsl 990-2009 per state, the number of
facilities per state that captured C02 on-site for the years 2010-2019 was divided by the total number of
facilities across the country that captured C02 on-site for each year over the same time period to calculate
state allocation percentages. Each state's percentage was applied to the national data on C02 captured on-
site to estimate the quantity of C02 captured on-site per state, per year. These values were subtracted from
the gross C02 emissions to calculate net C02 emissions by state.
3.1.2.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02 from lime production was
calculated using the 2006 IPCC Guidelines Approach 2 methodology for uncertainty (IPCC 2006). As
described further in Chapter 4 and Annex 7 of the national Inventory (EPA 2024b), levels of uncertainty in the
national estimates in 2022 were -1 %/+1 % for C02.
State-level estimates are expected to have an overall higher uncertainty because the national emissions
estimates were apportioned to each state based on a combination of GHGRP emissions data for 2010-2022
and the estimated number of facilities for 1990-2009. These assumptions were required because of a
general lack of more granular state-level data.
For 1990-2009, the methodology does not differentiate between the type of lime produced at a facility
because of a lack of available data, which increases uncertainty. The chemical composition of the limestone
and dolomite feedstocks is different, resulting in different emissions factors for calculating C02. This
difference has the potential to underestimate or overestimate C02 emissions from a facility, depending on
the types of lime produced.
The diversity of lime manufacturing facility types adds uncertainty to the analysis. The current
methodology for 1990-2009 assumes that each facility has the same amount of inputs and outputs, which
overestimates emissions for smaller facilities (e.g., beet sugar manufacturing) and underestimates
emissions for larger facilities (e.g., lime manufacturing). The 1990-2009 methodology for estimating the
quantity of C02 captured on-site does not differentiate between the type of facility (e.g., beet sugar
manufacturing compared with lime manufacturing), which increases uncertainty. The resulting captured C02
values may overestimate the quantity of C02 captured from beet manufacturing facilities, while
underestimating the quantity of C02 captured from lime manufacturing facilities.
Additionally, some lime facilities go idle for periods of time, and the lack of data on when a facility is in
operation or idle during the year increases uncertainty in the analysis. The GHGRP does not currently acquire
information on whether or for how long plants are idled.
3.1.2.4 Recalculations
No recalculations were applied for this current report, consistent with the national Inventory (see
Section 4.2, page 4-20).
3-9
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
3.1.2.5 Planned Improvements
EPA will consider weighting gross C02 emissions and captured C02 emissions by the type of facility
(primary NAICS code) to better allocate C02 emissions and reduce the uncertainty around overestimating or
underestimating emissions for certain facility types. Of the facilities reporting to the lime Subpart S under the
GHGRP, seven different types of facilities reported using the following primary 2007 NAICS codes: 212312
(Crushed and Broken Limestone Miningand Quarrying), 212391 (Potash, Soda, and Borate Mineral Mining),
311313 (Beet Sugar Manufacturing), 327125 (Nonclay Refractory Manufacturing; also reported as 327120 in
the 2022 NAICS), 327310 (Cement Manufacturing), 327410 (Lime Manufacturing), and 331111 (Iron and Steel
Mills; also reported as331110in the 2022 NAICS).
Further refinements include identifying additional sources of data to confirm facilities within each state
for 1990-2009 and better reflect their associated production (including production by type of lime),
especially for the states that were aggregated in the USGS Minerals Yearbooks. Another potential refinement
includes assessing the range of facilities' production quantity or capacity and improving on the current
underlying assumption associated with using the number of facilities to estimate emissions.
Another potential refinement is to improve the CaO contents and emissions factors used for estimating
C02 emissions from high-calcium lime and dolomitic lime. Consistent with the 2006 IPCC Guidelines, the
current CaO content is assumed to be 95% for both high-calcium and dolomitic lime, which results in
emissions factors of 0.785 metric ton C02 per metric ton CaO for high-calcium lime and 0.913 metric ton C02
per metric ton CaO for dolomitic lime. The average CaO contents and emissions factors per product a re
reported to the GHGRP but are considered CBI. Data aggregation may address CBI concerns.
Potential refinements also include identifying additional information to determine which facilities
captured C02 on-site in 1990-2009, prior to GHGRP reporting. In 2022, all of the beet sugar manufacturing
facilities reporting to the GHGRP captured C02 on-site, and three lime manufacturing facilities that reported
to GHGRP captured C02 on-site. In addition, further research on the use and prevalence of capturing C02 for
use in on-site processes in 1990-2009 is needed. The current methodology assumes that facilities captured
C02 on-site over the full time series and that the quantity of emissions captured is evenly distributed among
those facilities. More research on the range of C02 captured on-site per facility and per year is needed. EPA
plans to initiate a review to understand if precipitated calcium carbonate production practices have changed
and if literature is available since the publication of the 2006 IPCC Guidelines to understand if any C02 is
ultimately emitted from the use of captured C02 in precipitated calcium carbonate production or during the
sugar refining purification processes.
EPA will review time series consistency issues, due to the two methodologies for 1990-2009 and 2010-
2022. Surrogate data (number of facilities per state and number of facilities per state capturing C02 on-site)
were used in place of activity data for the 1990-2009 portion of the time series, and more research is needed
so calculations more closely simulate state trends in emissions.
3.1.2.6 References
EPA (U.S. Environmental Protection Agency) (2010-2022) Envirofacts GHGRP Subpart S and Subpart C Data.
Accessed May 15, 2024. Available online at: https://www.epa.gov/enviro/greenhouse-gas-customized-
search.
EPA (2024a) Aggregation of Reported Facility Level Data Under SubpartsNational Lime Production for
Calendar Years 2010 Through 2022.
EPA (2024b) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2022. EPA430-R-24-004.
Available online at: https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-10
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.jp/public/2006gl/.
USGS (U.S. Geological Survey) (1996-2023) Mineral Commodity Summary: Lime. Available online at:
https://www.usgs.gov/centers/national-minerals-information-center/lime-statistics-and-information.
USGS (1991-2021) Minerals Yearbook: Lime. Available online at: https://www.usgs.gov/centers/national-
minerals-information-center/lime-statistics-and-information.
USGS (1991) Table 4. Lime Sold or Used by Producers in the United States, by State. In: 1990 Minerals
Yearbook: Lime. Available online at:
https://search.librarv.wisc.edu/digital/A5X7AW22D2UR08R/pages/AEH2VMY0UXX4038T.
USGS (1992-2010) Table 2. Lime Sold or Used by Producers in the United States, by State. In: 1991-2009
Minerals Yearbook: Lime. Available online at: https://www.usgs.gov/centers/national-minerals-
information-center/bureau-mines-minerals-yearbook-1932-1993 and
https://www.usgs.gov/centers/national-minerals-information-center/lime-statistics-and-information.
3.1.3 Glass Production (NIR Section 4.3)
3.1.3.1 Background
Glass production is an energy- and raw material-intensive process that results in the generation of C02
from both the energy consumed in making glass and the glass production process itself. Emissions from
fuels consumed for energy purposes during the production of glass are included in the energy sector. The raw
materials (primarily soda ash, limestone, and dolomite) release C02 emissions in a complex high-
temperature chemical reaction during the glass melting process. This process is not directly comparable to
the calcination process used in lime manufacturing, cement manufacturing, and process uses of carbonates
(i.e., limestone/dolomite use) but has the same net effect in terms of C02 emissions. In 2022, glass was
produced in 30 states (EPA 2023).
3.1.3.2 Methods/Approach
The national Inventory method was adapted to calculate state-level GHG emissions from glass
production to ensure consistency with national estimates (EPA 2024). National estimates were downscaled
across states, instead of reapplying the national Tier 3 methodology at the state level, because of limitations
in availability of state-specific data across the time series.
To compile process emissions by state from glass production, an Approach 2 methodology was used to
allocate process emissions to all states with glass production using a combination of process emissions
reported to the GHGRP for 2010-2022 and the number of glass facilities in each state for 1990-2009, as
shown in Table 3-4 below. The sum of emissions by state is consistent with national process emissions as
reported in the national Inventory. See Appendix C, Tables C-7 and C-8 in the "Glass" Tab, for more details on
the data used.
Table 3-4. Summary of Approaches to Disaggregate the National Inventory for Glass Production
Across Time Series
Time Series
Range
Summary of Method
2010-2022
GHGRP process emissions data were used to estimate the percentage of
emissions by state, multiplied by the national emissions (2006 IPCC Guidelines
Tier 3).
3-11
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Time Series
Range
Summary of Method
1990-2009
Data on the number of glass facilities were used to estimate the percentage of
production by state, multiplied by the national emissions (2006 IPCC Guidelines
Tier 3).
The state-Level method used for 2010-2022 was based on process emissions reported to the GHGRP
summed by state (EPA 2023) to calculate a percentage of emissions from each state. That percentage was
then applied to the national emissions from glass production per year to calculate disaggregated C02
emissions by state. GHGRP has a reporting threshold of 25,000 metric tons C02 for glass production, so
these emissions data are representative of the larger glass producers in the industry. The GHGRP threshold
excludes small entities (i.e., artisan facilities). Using GHGRP emissions data means that emissions from
states with smaller facilities were possibly underestimated.
The method used for 1990-2009 was based on the number of glass facilities in each state divided by the
number of facilities nationally to calculate a percentage of glass facilities in each state for each year. This
percentage was applied to the national C02 emissions from glass production per year (EPA 2023) to calculate
disaggregated C02 emissions by state for each year. The number of facilities per state was estimated based
on the knowledge of facility locations in 2010-2022 and research on when these facilities and others began
or ceased operations. Using the number of facilities per state to determine the state allocation percentage
assumes that each facility has the same amount of input and output.
3.1.3.2.1. CEMS Adjustment for2010-2022
Starting in 2010, facilities producing glass and emitting more than 25,000 metric tons of C02 equivalent
per year reported both process and combustion emissions to the GHGRP. For facilities using a CEMS
approach to measure and report C02 emissions, process and combustion emissions were reported together
under Subpart N; otherwise, facilities reported process emissions under Subpart N and combustion
emissions under Subpart C using engineering and calculation approaches.26 To disaggregate process
emissions for those facilities reporting C02with CEMS, the ratio of process emissions to total emissions for
facilities that do not report using CEMS was calculated for each year from 2010 to 2022 and applied to the
total C02 emissions for each CEMS facility to calculate process emissions for each year that emissions were
reported using CEMS. The results were an estimated process C02 emissions-only value for that CEMS
facility.
Because the methodology for 1990-2009 does not use GHGRP emissions data to calculate the state
emissions, there was no need to adjust for CEMS facilities for those years.
3.1.3.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02 from glass production was
calculated using the 2006 IPCC Guidelines Approach 2 methodology for uncertainty (IPCC 2006). As
described further in Chapter 4 and Annex 7 of the national Inventory (EPA 2024), levels of uncertainty in the
national estimates in 2022 were -2%/+2% for C02.
26 For more information on the GFIGRP, see 74 FR 56374, October 30, 2009, available online at
https://www.govinfo.gov/content/pkg/FR-2009-10-3Q/pdf/E9-23315.pdf.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-12
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
State-Level estimates are expected to have an overall higher uncertainty because the national emissions
estimates were apportioned to each state based on a combination of GHGRP emissions data for 2010-2022
and the estimated number of facilities for 1990-2009.
For estimates from 2010-2022, uncertainty is expected to be lower than for 1990-2009 due to the use of
GHGRP emissions data by state to calculate emissions. However, because the sum of GHGRP emissions
from glass production is higher than the national Inventory emissions from glass production, and the GHGRP
does not include emissions from smaller glass production facilities, this methodology could underestimate
emissions in states with smaller facilities and overestimate emissions in states with larger facilities,
potentially increasing the uncertainty of the state-by-state percentage compared with the national Inventory.
For 1990-2009, this allocation method does not address facilities' production capacities or utilization
rates, which vary from facility to facility and from year to year. Because this approach assumes emissions
from all facilities are equal regardless of production capacity or utilization rates, this approach could
overestimate emissions in states with higher shares of smaller facilities and underestimate emissions in
states with larger facilities.
3.1.3.4 Recalculations
Due to GHGRP resubmissions from one facility for 2017 and a second facility for 2021, and a change in
calculations for a facility that was mistakenly identified as a CEMS facility in 2012, recalculations were
performed for 2012, 2017, and 2021. Due to the small changes in emissions, the state-level impacts for the
three years were less than 1 % for all states.
3.1.3.5 Planned Improvements
Potential refinements include identifying data to improve the completeness of state allocation and
reflect smaller facilities. Data gaps to calculate emissions from glass production include partial data sets on
glass production by state and the number of glass facilities by state for the full time series. GHGRP has a
reporting threshold for glass production facilities; facilities emitting more than 25,000 metric tons of C02
equivalent per year must report to the program. Facilities emitting less emissions per year were not captured
in GHGRP data and are not reflected in this state-level estimate. Therefore, it is likely that emissions from
smaller facilities are being attributed to larger facilities that report to GHGRP. Facilities with lower emissions
(e.g., artisan glass production facilities) were not captured in this estimation. EPA could apply other methods
that may improve estimates if more complete activity data are available by state (e.g., glass production,
carbonate consumption used for glass production, glass sales data by state, or GDP related to glass
production by state).
EPA will assess the consistency of the estimates over time, given the use of two approaches to compile
state-level estimates, to ensure that changes in estimates over time are not significantly biased by
methodological and data approaches to the extent possible.
3.1.3.6 References
EPA (U.S. Environmental Protection Agency) (2023) Facility Level Information on GreenHouse gases Tool
(FLIGHT) Data set as of August 18, 2023. Available online at: https://ghgdata.epa.gov/ghgp/.
EPA (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2022. EPA430-R-24-004. Available
online at: https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
3-13
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
3.1.4 Other Process Uses of Carbonates (NIR Section 4.4)
3.1.4.1 Background
Limestone, dolomite, and other carbonates such as soda ash, magnesite, and siderite are basic
materials used by a wide variety of industries, including construction, agriculture, chemical, metallurgy (i.e.,
iron and steel production, ferroalloy production, and magnesium production), glass production,
environmental pollution control, ceramics production, and non-metallurgical magnesia production. This
section addresses only limestone, dolomite, soda ash, and magnesite use. Emissions from the use of these
carbonates are organized into four subcategories: other uses of carbonates (i.e., limestone and dolomite
consumption), ceramics production, other uses of soda ash, and non-metallurgical magnesia production. For
industrial applications, carbonates are heated sufficiently enough to calcine the material and generate C02
as a byproduct. Emissions from limestone and dolomite used in other process sectors, such as the
production of cement, lime, glass, iron and steel, and magnesium, were excluded from this category and are
reported under their respective source sections (e.g., Cement Production). Emissions from soda ash
production are reported under soda ash production. Emissions from soda ash consumption associated with
glass manufacturing are reported under glass production. Emissions from the use of limestone and dolomite
in liming of agricultural soils are included in the agriculture chapter under liming. Emissions from fuels
consumed for energy purposes during these processes are accounted for in the energy sector. Both lime and
limestone can be used as a sorbent for flue gas desulfurization (FGD) systems. Emissions from lime
consumption for FGD systems are reported under lime production.
3.1.4.2 Methods/Approach
For Other Process Uses of Carbonates, a combination of Approach 2 and Approach 1 methodologies
was used. The Approach 2 state-level methodology allocates total national process emissions to all
applicable U.S. states and territories using state-level consumption of limestone and dolomite for other uses
of carbonates, state-level consumption of clay for ceramics production, and state population as a surrogate
for other uses of soda ash, due to limitations in availability of state-specific data. The Approach 1 state-level
methodology utilizes facility-level consumption of magnesite for non-metallurgical magnesia production.
3.1.4.2.1. Other Uses of Carbonates (Limestone and Dolomite Consumption)
National C02 emissions from the consumption of limestone and dolomite for emissive sources,
including flux stone, FGD systems, chemical stone, mine dusting or acid water treatment, acid
neutralization, and sugar refining, were calculated based on USGS data on the national-level consumption of
each carbonate for each end use. USGS does not provide the state-level consumption of limestone and
dolomite for each end use; however, USGS does publish annual state-level data on the total consumption of
each carbonate. Because no other source of data on state-level limestone and dolomite consumption were
identified for any of the emissive sources, the USGS total consumption data by state were used.
For 1991 and 1993-2022, state-level C02 emissions for the national Inventory were estimated using the
USGS annual state-level values for limestone and dolomite sold or used by producers compiled from the
USGS Minerals Yearbook for Crushed Stone (U.S. Bureau of Mines 1991-1995; USGS 1995b-2022b). The
national C02 emissions from limestone and dolomite consumption were disaggregated independently by
calculating the fraction of each state-level consumption for each carbonate and applying that fraction to the
national-level C02 estimated for each of the two carbonates in the national Inventory. The USGS state-level
consumption data exclude the District of Columbia and territories; therefore, their C02 emissions from
limestone and dolomite consumption were not estimated.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-14
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
During 1990 and 1992, USGS did not publish Limestone and dolomite consumption data by state. Data
on consumption by state for 1990 were estimated by applying the 1991 ratios of total limestone and dolomite
consumption by state to total 1990 limestone and dolomite consumption values. Similarly, the 1992
consumption figures were approximated by applying an average of the 1991 and 1993 ratios of total
limestone and dolomite use by state to the 1992 total values.
In 1991 and 1993-2006, certain state-level limestone and dolomite consumption data were withheld
from the USGS publications to avoid disclosing proprietary information. Those limestone and dolomite
values were aggregated and included in a category titled "Other." To ensure that the total reported
consumption values for both limestone and dolomite were accounted for, the "Other" value was equally
distributed to the states for which consumption data were withheld. In 1991, USGS provided an "Other"
value for limestone consumption; however, no states that were included in the state-level table contained an
indication that data were withheld. To account for this limestone usage, the "Other" value was proportionally
allocated to all of the states for which data were reported in 1991 based on their reported usage. See
Appendix C, Tables C-9 through Table C-12 in the "Other Process Uses of Carbonates" Tab, for more details
on the data used.
3.1.4.2.2. Ceramics Production
National C02 emissions from the consumption of clay for emissive sources were calculated based on
USGS data on the national-level consumption of clay for each of the three emissive subcategories (ceramics,
glass, and floor and wall tile; refractories; and heavy clay products). USGS does not provide the state-level
consumption of clay for each end use; however, USGS does publish annual state-level data on the total
consumption of clay. Because no other source of data on state-level clay consumption was identified for any
of the emissive sources, the USGS total clay consumption data by state were used.
For 1990-2022, state-level C02 emissions for the national Inventory were estimated using the USGS
annual state values for clay sold or used by producers, compiled in the USGS Minerals Yearbook for Clay and
Shale (U.S. Bureau of Mines 1991-1995; USGS 1995a-2022a). The national C02 emissions from clay
consumption were disaggregated independently by calculating the fraction of clay consumption for each
state-level consumption and applying that fraction to the estimated national-level C02 emissions for
ceramics production in the national Inventory. The USGS state-level consumption data exclude the District of
Columbia and territories; therefore, their C02 emissions from limestone and dolomite consumption were not
estimated.
For the full time series, certain state-level clay consumption data were withheld from the USGS
publications to avoid disclosing proprietary information. Those values were aggregated and included in a
category titled "Other." To ensure that the total reported consumption values for clay were accounted for,
the "Other" value was equally distributed to the states for which consumption data were withheld. In 2013-
2015, data for additional states were similarly grouped together to avoid disclosing proprietary information.
Those values were also equally distributed to the states in each grouping to ensure that the total reported
consumption values were accounted for. See Appendix C, Table C-13 in the "Other Process Uses of
Carbonates" Tab, for more details on the data used.
3.1.4.2.3. Other Uses of Soda Ash
The national Inventory also estimates national C02 emissions from the consumption of soda ash.
Excluding soda ash consumption for glass manufacturing, most soda ash is consumed in chemical
production, with minor amounts used in soap production, pulp and paper, FGD, and water treatment.
Emissions from soda ash consumption from glass manufacturing are accounted for under Section 4.3, Glass
Production. Data on the consumption of soda ash by state, however, are not available, and due to the
3-15
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
distribution of these end uses across the country and Lack of other surrogate data on end uses by state,
population was used to allocate emissions. To calculate state-level C02 emissions from soda ash
consumption, national C02 estimates from the national Inventory were distributed among the 50 states, the
District of Columbia, Puerto Rico, American Samoa, Guam, the Northern Mariana Islands, and the U.S. Virgin
Islands using U.S. population statistics as a surrogate for data on soda ash consumption not associated with
glass manufacturing (U.S. Census Bureau 2002, 2011, 2021, 2022a, 2022b; Instituto de Estadfsticas de
Puerto Rico 2021). For each year in the 1990-2022 time series, the fraction of the total U.S. population in
each state, the District of Columbia, and territories was calculated by dividing the state population by the
total U.S. population. To estimate C02 emissions for each year by state, national Inventory C02emissions
from soda ash consumption were multiplied by each state's fraction of the total population for that year. See
Appendix G, Table G-1 in the "Population Data" Tab, for more details on the data used.
3.1.4.2.4. Non-Metallurgical Magnesia Production
All national non-metallurgical magnesia production emissions can be attributed to Nevada for the
entirety of the time series. National C02 emissions from the consumption of magnesite for non-metallurgical
magnesia production were calculated based on Nevada Department of Environmental Quality data on the
quantities of magnesium ore extracted and processed at the only non-metallurgical magnesia production
facility in the United States. See Appendix C, Table C-14 in the "Other Process Uses of Carbonates" Tab, for
more details on the data used.
3.1.4.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02 from other process uses of
carbonate was calculated using the 2006 IPCC Guidelines Approach 2 methodology for uncertainty (IPCC
2006). As described further in Chapter 4 and Annex 7 of the national Inventory (EPA 2024), levels of
uncertainty in the national estimates in 2022 were -12%/+15% for C02.
State-level estimates are expected to have a higher uncertainty because the national emissions
estimates were apportioned to each state based on state data of total limestone and dolomite consumption
and state population for soda ash consumption.
3.1.4.4 Recalculations
For the current national Inventory, updated state-level USGS data on limestone and dolomite
consumption were available for 2021, removing the use of 2020 as a proxy, resulting in updated emissions
estimates. Additional recalculations for emissions from soda ash consumption were performed for 2020 and
2021 as updated population data were made available from the U.S. Census Bureau for the time series. The
updated population data had a negligible impact on the emissions estimated for the 50 states, the District of
Columbia, and Puerto Rico due to the low emissions estimated for each state or territory for the sector.
Emissions from ceramics production are being included for the first time this year. The new subcategory
increased national C02 emissions by 756.7 kiloton (kt) C02 equivalent in 1990 and 406.5 kt C02 equivalent in
2022. The states that saw the largest impact to their overall C02 emissions across the full time series include
Georgia, Texas, and Wyoming.
Emissions from non-metallurgical magnesia production are being included for the first time this year.
The new subcategory only impacted the state of Nevada across the full time series because the only non-
metallurgical magnesia production facility in the United States is located in Nevada.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-16
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
3.1.4.5 Planned Improvements
The disaggregation methodology for Limestone and dolomite consumption does not take into account
the consumption of these carbonates from the l&S sector, as is done in the national Inventory C02 emissions
calculations. Given that the methodology for the disaggregation of the l&S sector was developed
concurrently with this sector, EPA was not able to fully assess if the state-level percentages for the l&S sector
could be applied to the l&S limestone and carbonate consumption and then subtracted out from each of the
state-level C02 emissions calculated using the methodology described above. Initial attempts yielded
negative C02 emissions in certain states, thus requiring additional review and likely refinement of
approaches to disaggregate these emissions.
Additionally, further research is needed to determine if data sources may be available to attribute C02
emissions more accurately from each of the emissive sources for limestone and dolomite consumption to
each state. Currently, it is assumed that limestone and dolomite consumption for flux stone, FGD systems,
chemical stone, mine dusting or acid water treatment, acid neutralization, and sugar refining activities is
distributed equally geographically among all states, excluding the District of Columbia and Puerto Rico.
Data gaps for the soda ash consumption category include data on soda ash consumption by state.
3.1.4.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
Instituto de Estadfsticas de Puerto Rico (2021) EstimadosAnuales Poblacionales de los Municipios Desde
1950. Accessed February 2021. Available online at: https://censo.estadisticas.pr/EstimadosPoblacionales.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
U.S. Bureau of Mines (1991 a-1995a) Minerals Yearbook for Clay and Shale. Available online at:
https://www.usgs.gov/centers/national-minerals-information-center/bureau-mines-minerals-vearbook-
1932-1993 and https://www.usgs.gov/centers/national-minerals-information-center/clays-statistics-
and-information.
U.S. Bureau of Mines (1991 b-1995b) Minerals Yearbook for Crushed Stone. Available online at:
https://www.usgs.gov/centers/national-minerals-information-center/bureau-mines-minerals-vearbook-
1932-1993 and https://www.usgs.gov/centers/national-minerals-information-center/crushed-stone-
statistics-and-information.
U.S. Census Bureau (2002) Table CO-EST2001-12-00. In: Time Series of I ntercensal State Population
Estimates: April 1, 1990 to April 1,2000. Release date: April 11, 2002. Available online at:
https://www2.census.gov/programs-survevs/popest/tables/1990-2000/intercensal/st-co/co-est2Q01-12-
OO.pdf.
U.S. Census Bureau (2011) Table ST-EST00INT-01. In: Intercensal Estimates of the Resident Population for
the United States, Regions, States, and Puerto Rico: April 1, 2000 to July 1, 2010. Release date:
September 2011. Available online at: https://www2.census.gov/programs-survevs/popest/datasets/200Q-
2010/intercensal/state/st-est00int-alldata.csv.
3-17
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
U.S. Census Bureau (2021) Table NST-EST2020./Annua/ Estimates of the Resident Population for the United
States, Regions, States, and Puerto Rico: April 1, 2010 to July 1,2020. Release date: July 2021. Available
online at: https://www.census.gov/programs-survevs/popest/technical-
documentation/research/evaluation-estimates/2020-evaluation-estimates/201 Os-state-total.html.
U.S. Census Bureau (2022a) International Database: World Population Estimates and Projections. Accessed
November 23, 2022. Available online at: https://www.census.gov/programs-survevs/international-
programs/about/idb.html.
U.S. Census Bureau (2022b) Table NST-EST2022-POP. In: Annual Estimates of the Resident Population for the
United States, Regions, States, District of Columbia, and Puerto Rico: April 1, 2020 to July 1, 2022.
Release date: December 2022. Available online at: https://www.census.gov/data/tables/time-
series/demo/popest/2020s-state-total.html.
USGS (U.S. Geological Survey) (1995a-2022a) Minerals Yearbook: Clay and Shale Annual Report. Available
online at: https://www.usgs.gov/centers/national-minerals-information-center/clays-statistics-and-
information
USGS (1995b-2022b) Minerals Yearbook: Crushed Stone Annual Report. Available online at:
https://www.usgs.gov/centers/national-minerals-information-center/crushed-stone-statistics-and-
information.
3.1.5 Carbon Dioxide Consumption (NIR Section 4.16)
3.1.5.1 Background
C02 is used for a variety of commercial applications, including food processing, chemical production,
carbonated beverage production, and refrigeration, and is also used in petroleum production for enhanced
oil recovery. C02used for enhanced oil recovery is injected underground to enable additional petroleum to be
produced. For the purposes of this analysis, C02 used in commercial applications other than enhanced oil
recovery is assumed to be emitted to the atmosphere. A further discussion of C02 used in enhanced oil
recovery is described in the national Inventory Energy chapter in Box 3-6, "Carbon Dioxide Transport,
Injection, and Geological Storage," and is not included in this section.
3.1.5.2 Methods/Approach
Data on the consumption of C02 by state are not readily available; therefore, using an Approach 2
method, the state-level methodology for emissions from C02 consumption allocates emissions from C02
consumption across all U.S. states and territories using population as a surrogate. See Appendix G, Table G-
1 in the "Population Data" Tab, for more details on the data used. National estimates were used to
disaggregate emissions by state because of the limitations in the availability of state-specific data for the
time series. The approach is considered reasonable, given many of the sources are end-use categories (e.g.,
carbonated beverage use, dry ice), where per capita use is not likely to vary across states.
To calculate state-level C02 emissions from C02 consumption, national C02 estimates from the
national Inventory were distributed among the 50 states, the District of Columbia, Puerto Rico, American
Samoa, Guam, the Northern Mariana Islands, and the U.S. Virgin Islands using U.S. population statistics as a
surrogate for C02 consumption data (U.S. Census Bureau 2002, 2011, 2021, 2022a, 2022b; Instituto de
Estadfsticas de Puerto Rico 2021). For each year in the 1990-2022 time series, the fraction of the total U.S.
population in each state, the District of Columbia, and each territory was calculated by dividing the state
population by the total U.S. population.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-18
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
3.1.5.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02 consumption was
calculated using the 2006 IPCC Guidelines Approach 2 methodology for uncertainty (IPCC 2006). As
described further in Chapter 4 and Annex 7 of the national Inventory (EPA 2024), levels of uncertainty in the
national estimates in 2022 were -5%/+5% for C02.
State-level estimates are expected to have a higher uncertainty because the national emissions
estimates were apportioned to each state based solely on state population. This assumption was required
because of a general lack of more granular state-level data. This allocation method introduces additional
uncertainty because of limited data on the quantity of C02 consumption by state or nationally for the full time
series. The sources of uncertainty for this category are also consistent over time because the same surrogate
data are applied across the entire time series.
3.1.5.4 Recalculations
Recalculations were performed for 2020 and 2021 due to updated population data, resulting in a
decrease in emissions of 3% for the District of Columbia for 2020. There was no impact on the emissions
estimated for the 50 states and Puerto Rico in 2020 and 2021.
3.1.5.5 Planned Improvements
EPA will explore other sources of data on the consumption of C02 by state for the full time series.
3.1.5.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
Instituto de Estadfsticas de Puerto Rico (2021) EstimadosAnuales Poblacionales de los Municipios Desde
1950. Accessed February 2021. Available online at: https://censo.estadisticas.pr/EstimadosPoblacionales.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
U.S. Census Bureau (2002) Table CO-EST2001-12-00. In: Time Series of Intercensal State Population
Estimates: April 1, 1990 to April 1,2000. Release date: April 11, 2002. Available online at:
https://www2.census.gov/programs-survevs/popest/tables/1990-2000/intercensal/st-co/co-est2Q01-12-
OO.pdf.
U.S. Census Bureau (2011) Table ST-EST00INT-01. In: Intercensal Estimates of the Resident Population for
the United States, Regions, States, and Puerto Rico: April 1, 2000 to July 1, 2010. Release date:
September 2011. Available online at: https://www2.census.gov/programs-survevs/popest/datasets/200Q-
2010/intercensal/state/st-est00int-alldata.csv.
U.S. Census Bureau (2021) Table NST-EST2020. In: Annual Estimates of the Resident Population for the
United States, Regions, States, and Puerto Rico: April 1, 2010 to July 1,2020. Release date: July 2021.
U.S. Census Bureau (2022a) International Database: World Population Estimates and Projections. Accessed
November 23, 2022. Available online at: https://www.census.gov/programs-survevs/international-
programs/about/idb.html.
3-19
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
U.S. Census Bureau (2022b) Table NST-EST2022-POP. In: Annual Estimates of the Resident Population for the
United States, Regions, States, District of Columbia, and Puerto Rico: April 1, 2020 to July 1, 2022.
Release date: December 2022.
3.2 Chemicals
This section presents the methodology used to estimate the chemicals portion of IPPU emissions,
which consist of the following sources:
Ammonia production (C02)
Urea consumption for nonagricultural purposes (C02)
Nitric acid production (N20)
Adipic acid production (N20)
Caprolactam, glyoxal and glyoxylic acid production (N20)
Carbide production and consumption (C02, CH4)
Titanium dioxide production (C02)
Soda ash production (C02)
Petrochemical production (C02)
HCFC-22 production (HFC-23)
Production of fluorochemicals other than HCFC-22 (HFCs, PFCs, SFs, NF3)
Phosphoric acid production (C02)
3.2.1 Ammonia Production (NIR Section 4.5)
3.2.1.1 Background
Emissions of C02 occur duringthe production of synthetic ammonia, primarily through the use of
natural gas, petroleum coke, or naphtha as a feedstock. The processes based on natural gas, naphtha, and
petroleum coke produce C02 and hydrogen, the latter of which is used to produce ammonia. Natural gas is
also used as a fuel in the process. The 2006 IPCC Guidelines recommend including emissions from fuels
consumed for energy purposes duringthe production of ammonia along with feedstock emissions; however,
data on total fuel use (including fuel used for ammonia feedstock and fuel used for energy) for ammonia
production are not known in the United States. National energy use information is only available at the broad
industry sector level and does not provide data broken out by industrial category. Emissions from fuel used
for energy at ammonia plants are accounted for in the energy sector. In 2022,16 companies operated 35
ammonia-producing facilities in 16 states, with approximately 60% of domestic ammonia production
capacity concentrated in Louisiana, Oklahoma, and Texas (USGS 2023).27
27 The number of facilities that report to the GHGRP (29 facilities in 17 states) differs from USGS due to (1) the definition of a
"facility" used by USGS for two locations (Donaldsonville, LA, and Verdigris, OK); (2) the definition of a facility subject to
Subpart G of the GHGRP that requires steam reforming or raw material gasification (see 98.70), which does not appear to be
present at the Freeport, TX, facility in the USGS list; (3) the definition of a facility subject to Subpart G of the GHGRP when a
facility (like the Beaumont, TX, facility in the USGS list) produces methanol, hydrogen, and ammonia (see 98.240[c]); and (4)
an ammonia-producingfacility in Midway, TN, that is not in the USGS list.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-20
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
3.2.1.2 Methods/Approach
To compile emissions by state from ammonia production, the state-Level inventory disaggregated
national emissions from the national Inventory with an Approach 2 method as defined in the Introduction
chapter of this report, using a combination of process emissions reported to the GHGRPfor 2010-2022 and
ammonia production capacity by state and by year for 1990-2009, as shown in Table 3-5. This approach was
taken due to limitations in state-level activity data on ammonia production by feedstock or feedstock
consumption for ammonia production. The sum of emissions by state is consistent with the process
emissions reported in the national Inventory (EPA 2024). See Appendix D, Tables D-1 and D-2 in the
"Ammonia" Tab, for more on the data used.
Table 3-5. Summary of Approaches to Disaggregate the National Inventory for Ammonia
Production Across Time Series
Time Series
Range
Summary of Method
2010-2022
GHGRP (Subpart G) process emissions data (gross C02) were used to estimate
the percentage of emissions by state, multiplied by the national emissions
(IPCC 2006Tier 2).
1990-2009
USGS data on ammonia production capacity were used to estimate the
percentage of production by state, multiplied by the national emissions (IPCC
2006 Tier 2).
The methodology used for 2010-2022 was based on process emissions reported to the GHGRP and
summed by state (EPA 2023) to calculate a percentage of emissions from each state. That state percentage
was then applied to the national Inventory emissions from ammonia production per year to disaggregate C02
emissions by state and by year and ensure emissions are consistent with estimates in the national Inventory.
The GHGRP has no reporting threshold for ammonia production, so all facilities are included, and these
emissions data are, therefore, representative of the industry.
The methodology used for 1990-2009 was based on the total ammonia production capacity in each
state divided by the total ammonia capacity in the United States to calculate a percentage of ammonia
capacity in each state for each year. This percentage was applied to the national C02 emissions from
ammonia production per year to calculate disaggregated C02 emissions by state for each year. The ammonia
capacities per facility per state were compiled from the Minerals Yearbook: Metals and Minerals for Nitrogen,
Table 5, "Domestic Producers of Anhydrous Ammonia" for 1990 and 1991 (U.S. Bureau of Mines 1990-1991);
the Minerals Yearbook: Metals and Minerals for Nitrogen, Table 4, "Domestic Producers of Anhydrous
Ammonia" for 1992 and 1993 (U.S. Bureau of Mines 1992-1993); and the Minerals Yearbook: Nitrogen, Table
4, "Domestic Producers of Anhydrous Ammonia" for 1994-2009 (USGS 1994-2010). Using the ammonia
capacity per state to determine the state allocation percentage assumes that facility utilization rates are
roughly the same from state to state and that production capacity is a reasonable surrogate for production.
3.2.1.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02 from ammonia production
was calculated using the 2006 IPCC Guidelines Approach 2 methodology for uncertainty (IPCC 2006). As
described further in Chapter 4 and Annex 7 of the national Inventory (EPA 2024), levels of uncertainty in the
national estimates in 2022 were -4%/+4% for C02 emissions from ammonia production.
State-level estimates are expected to have an overall higher uncertainty because the national emissions
estimates were apportioned to each state based on a combination of process emissions reported to the
3-21
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
GHGRPfor 2010-2022 and ammonia production capacity by state by year for 1990-2009. These assumptions
were required because of a general Lack of more granular state-level data.
For 2010-2022, uncertainty is expected to be lower due to the use of GHGRP emissions data by state as
a surrogate for ammonia production data by state to calculate emissions; however, because the sum of
GHGRP emissions from ammonia production is higher than the national Inventory emissions from ammonia
production, the uncertainty of the state-by-state percentage may be higher. This may have led to
overestimating or underestimating the percentage of emissions apportioned to each state.
For 1990-2009, this allocation method does not address utilization rates, which vary from facility to
facility and from year to year. While this approach implicitly accounts for the size of a facility in a state, it
could overestimate emissions in states where facilities used less of their capacity and underestimate
emissions in states where facilities used more of their capacity, as a result of the lack of data on utilization
rates.
3.2.1.4 Recalculations
For 2021, the urea consumption value was changed from a rounded value to a more precise unrounded
value. Also, updated ammonia facility-level emissions were obtained from the GHGRP for 2021. Therefore,
recalculations were performed for 2021. A resubmission of GHGRP data for 2021 from one facility in Oregon
occurred after the 1990-2021 state-level inventory was completed. Due to the resubmission and changes to
the urea consumption value, C02 emissions from ammonia production in Oregon for 2021 increased by 3%
(2.2 kt C02), compared to the previous Inventory. Emissions from other states decreased slightly based on
increased allocation to Oregon.
3.2.1.5 Planned Improvements
For the GHGRP emissions data used for 2010-2022, the quantity of C02 that is captured at ammonia
production facilities and used to produce urea has not been subtracted and allocated under Urea
Consumption for Nonagricultural Purposes (Section 3.2.2) and Urea Fertilization (Section 4.2.4) because
these data by state are considered CBI and are not available. Reporters must report all C02 created during
the ammonia production process under Subpart G of the GHGRP. The amount of C02 from the production of
ammonia that is then captured and used to produce urea is reported to the GHGRP. More research on
possible aggregation options is needed.
For the state-level ammonia capacity data used for 1990-2009, additional research is needed to
determine whether the capacities can be adjusted to account for facilities that also produce urea, to be
consistent with the national Inventory.
EPA will review potential time series consistency issues due to the two methodologies for 1990-2009
and for 2010-2022. Surrogate data on production capacity are used in place of activity data for the 1990-
2009 portion of the time series, and more research is needed so calculations during that time period more
closely simulate state trends in emissions.
3.2.1.6 References
EPA (U.S. Environmental Protection Agency) (2023) Facility Level Information on GreenHouse gases Tool
(FLIGHT). Data set as of August 18, 2023. Available online at: https://ghgdata.epa.gov/ghgp/.
EPA (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2022. EPA430-R-24-004. Available
online at: https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-22
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.jp/public/2006gl/.
U.S. Bureau of Mines (1990-1993) Bureau of Mines Minerals Yearbook (1932-1993). Available online at:
https://vwwv.usgs.gov/centers/nmic/bureau-mines-minerals-yearbook-1932-1993.
USGS (U.S. Geological Survey) (1994-2010) Minerals Yearbook: Nitrogen. Available online at:
https://www.usgs.gov/centers/nmic/nitrogen-statistics-and-information.
USGS (2023) Mineral Commodity Summaries: Nitrogen (Fixed)Ammonia. Available online at:
https://pubs.usgs.gov/periodicals/mcs2023/mcs2023-nitrogen.pdf.
3.2.2 Urea Consumption for Nonagricultural Purposes (NIR Section 4.6)
3.2.2.1 Background
Urea is produced using ammonia and C02 as raw materials. All urea produced in the United States was
assumed to be produced at ammonia production facilities where both ammonia and C02 are generated. This
section accounts for C02 emissions associated with urea consumed exclusively for nonagricultural
purposes. Emissions of C02 resulting from agricultural applications of urea are accounted for in the urea
fertilization section of the Agriculture chapter.
3.2.2.2 Methods/Approach
To compile emissions by state from ammonia production, the state-level inventory disaggregated
national emissions from the national Inventory with an Approach 2 method as defined in the Introduction
chapter of this report, using U.S. population statistics as a surrogate for data on nonagricultural applications
of urea due to limitations in the availability of state-specific activity data. See Appendix G, Table G-1 in the
"Population Data" Tab, for more details on the data used.
The national Inventory estimates national C02 emissions from the consumption of urea for
nonagricultural purposes consistent with the Tier 1 method for ammonia production in the 2006 IPCC
Guidelines (IPCC 2006). While data on the consumption of urea by state are not available, due to the
widespread use of urea for nonagricultural purposes, population by state is a reasonable surrogate. To
calculate state-level C02 emissions from urea consumption, national C02 estimates from the national
Inventory were distributed among the 50 states, the District of Columbia, Puerto Rico, American Samoa,
Guam, the Northern Mariana Islands, the U.S. Virgin Islands, and the U.S. Minor Outlying Islands, using U.S.
population statistics as a surrogate (U.S. Census Bureau 2002, 2011, 2021, 2022a, 2022b; Instituto de
Estadfsticas de Puerto Rico 2021). For each year in the time series, the fraction of the total U.S. population in
each state, as well as the District of Columbia and the territories, was calculated by dividing the state
population by the total U.S. population. To estimate C02 emissions for each year by state, national Inventory
C02 emissions from urea consumption were multiplied by each state's fraction of the national population for
that year.
3.2.2.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02 from urea consumption for
nonagricultural purposes was calculated using the 2006 IPCC Guidelines Approach 2 methodology for
uncertainty (IPCC 2006). As described further in Chapter 4 and Annex 7 of the national Inventory {EPA 2024),
levels of uncertainty in the national estimates in 2022 were -4%/+4% for C02.
State-level estimates are expected to have a higher uncertainty because the national emissions
estimates were apportioned to each state based solely on state population. This assumption was required
3-23
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
because of a general Lack of more granular state-level data. This allocation method introduces additional
uncertainty due to limited data on the quantity of urea used for industrial applications by state or nationally
for the full time series. The sources of uncertainty for this category are consistent over time because the
same surrogate data are applied across the entire time series.
3.2.2.4 Recalculations
Based on updated quantities of urea applied for agricultural uses for 2017-2021, updated urea imports
from USGS for 2021, updated urea exports from USGS for 2021, and updated population data for 2020 and
2021, recalculations were performed for 2017-2021 (USGS 2023). Compared to the previous national
Inventory, state-level emissions increased for every state by less than 1 % for 2017, less than 0.05% for 2018,
and less than 0.07% for 2019. For 2020, emissions for the District of Columbia decreased by 3% and
emissions for Massachusetts decreased by 1%, compared to the previous inventory. Compared to the
previous Inventory, state-level emissions for 2021 increased by 33% for Alaska, Connecticut, Florida, Hawaii,
Maine, North Dakota, Oregon, Pennsylvania, and Vermont, and state-level emissions increased by 32% for all
remaining states/territories.
3.2.2.5 Planned Improvements
Data gaps include data on urea consumption for nonagricultural purposes by state for the full 1990-
2022 time series.
3.2.2.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
Instituto de Estadfsticas de Puerto Rico (2021) EstimadosAnuales Poblacionales de los Municipios Desde
1950. Accessed February 2021. Available online at: https://censo.estadisticas.pr/EstimadosPoblacionales.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
U.S. Census Bureau (2002) Table CO-EST2001-12-00. In: Time Series of I ntercensal State Population
Estimates: April 1, 1990 to April 1, 2000. Release date: April 11, 2002. Available online at:
https://www2.census.gov/programs-survevs/popest/tables/1990-2000/intercensal/st-co/co-est2Q01-12-
OO.pdf.
U.S. Census Bureau (2011) Table ST-EST00INT-01. In: Intercensal Estimates of the Resident Population for
the United States, Regions, States, and Puerto Rico: April 1, 2000 to July 1, 2010. Release date:
September 2011. Available online at: https://www2.census.gov/programs-survevs/popest/datasets/200Q-
2010/intercensal/state/st-est00int-alldata.csv.
U.S. Census Bureau (2021) Table NST-EST2020. In: Annual Estimates of the Resident Population for the
United States, Regions, States, and Puerto Rico: April 1, 2010 to July 1,2020. Release date: July 2021.
U.S. Census Bureau (2022a) International Database: World Population Estimates and Projections. Accessed
November 23, 2022. Available online at: https://www.census.gov/programs-survevs/international-
programs/about/idb.html.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-24
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
U.S. Census Bureau (2022b) Table NST-EST2022-POP. In: Annual Estimates of the Resident Population for the
United States, Regions, States, District of Columbia, and Puerto Rico: April 1, 2020 to July 1, 2022.
Release date: December 2022.
USGS (U.S. Geological Survey) (2023) Mineral Commodity Summaries: Nitrogen (Fixed)Ammonia. Available
online at: https://pubs.usgs.gov/periodicals/mcs2023/mcs2023-nitrogen.pdf.
3.2.3 Nitric Acid Production (NIR Section 4.7)
3.2.3.1 Background
N20 is emitted during the production of nitric acid, an inorganic compound used primarily to make
synthetic commercial fertilizers. Nitric acid is also a major component in the production of adipic acida
feedstock for nylonand explosives. Virtually all nitric acid produced in the United States is manufactured
by the high-temperature catalytic oxidation of ammonia. The basic process technology for producing nitric
acid has not changed significantly over time. During this process, N20 is formed as a byproduct and is
released from reactor vents into the atmosphere, unless mitigation measures are put in place. Emissions
from fuels consumed for energy purposes during the production of nitric acid are included in the energy
sector. As of 2022, there were 31 active nitric acid production plants in 20 states (EPA 2024).
3.2.3.2 Methods/Approach
The national Inventory methodology was adapted to calculate state-level GHG emissions from nitric
acid production to ensure consistency with national estimates (EPA 2024). For the national Inventory, the
2006 IPCC Guidelines Tier 2 method was used to estimate emissions from nitric acid production for 1990-
2009, and a country-specific approach similar to the IPCC Tier 3 method was used to estimate N20
emissions for 2010-2022. (IPCC 2006).
To compile emissions by state from nitric acid production, the state-level inventory disaggregated
national emissions from the national Inventory using Approach 2 as defined in the Introduction chapter of
this report and a combination of process emissions reported to the GHGRPfor 2010-2022 and nitric acid
production capacity by state and byyearfor 1990-2009, as shown inTable 3-6 below. Facility production
capacity and location data were updated for 1990-2005 using the SRI Directory of Chemical Producers (SRI
1990-2005) and were updated for 2006 and 2007 using data obtained from Independent Commodity
Intelligence Services (ICIS) (ICIS 2008). The sum of emissions by state is consistent with the national process
emissions reported in the national Inventory.
See Appendix D, Tables D-3 and D-4 in the "Nitric Acid" Tab, for more details on the data used in the
state-level inventory.
Table 3-6. Summary of Approaches to Disaggregate the National Inventory for Nitric Acid
Production Across Time Series
Time Series
Range
Summary of Method
2010-2022
GHGRP process emissions data were used to estimate the percentage of
emissions by state, multiplied by the national emissions (a country-specific
approach similar to IPCC 2006 Tier 3).
1990-2009
SRI Directory data (1990-2005) and ICIS data (2006-2009) on nitric acid
production capacity were used to estimate the percentage of production by
state, multiplied by the national emissions (IPCC 2006Tier 2).
3-25
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
The methodology used for 2010-2022 was based on process emissions reported to the GHGRP and
summed by state (EPA 2023) to calculate a percentage of emissions from each state. That percentage was
then applied to the national Inventory emissions from nitric acid production per year to disaggregate C02
emissions by state and by year. The GHGRP has no reporting threshold for nitric acid production, so these
emissions data are representative of the industry.
The methodology used for 1990-2009 was based on the total nitric acid production capacity in each
state divided by the total nitric acid production capacity in the United States to calculate a percentage of
nitric acid capacity in each state for each year. This percentage was applied to the national C02 emissions
from nitric acid production per year to calculate disaggregated C02 emissions by state for each year. Using
the nitric acid capacity per state to determine the state allocation percentage assumes that facility utilization
rates are roughly the same from state to state. Due to limited data availability, nitric acid capacities per state
for 1990-2005 were estimated using the SRI Directory of Chemical Producers (SR11990-2005). For years
2006-2009, production capacity data were obtained from ICIS at the parent company level, as opposed to
the facility level, necessitating a different approach to estimating state capacity data for 2006-2009 (ICIS
2008). First, GHGRP emissions data were averaged by facility for years 2010-2012. These years were used to
determine the average because that period was deemed to better represent historical nitric acid production
in 2006-2009. These averages were then summed by company to calculate a percentage of total company
emissions from each facility. That percentage was then applied to the total company capacity in 2008 to
disaggregate nitric acid production capacity by facility. Using facility location, the total company capacity in
2008 was disaggregated by state. The capacity data for 2008 were applied to the years 2006-2009. Additional
research included using state-level or region-specific permit websites to determine whether facilities in
operation in 2010, known through the GHGRP, were also in operation each year from 1990-2009; the
research also estimated production data by facility. Because of the lack of permit data available online for all
states and years, this approach was not used.
3.2.3.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of N20 from nitric acid production
was calculated using the 2006 IPCC Guidelines Approach 2 methodology for uncertainty (IPCC 2006). As
described further in Chapter 4 and Annex 7 of the national Inventory (EPA 2024), levels of uncertainty in the
national estimates in 2022 were -5%/+5% for N20.
State-level estimates are expected to have an overall higher uncertainty because the national emissions
estimates were apportioned to each state based on nitric acid production capacity by state and by year for
1990-2009. This assumption was required because of a general lack of more granular state-level data.
For 2010-2022, uncertainty is expected to be lower as a result of the use of GHGRP emissions data by
state as a surrogate for using nitric acid production data by state to calculate emissions. The uncertainty is
also lower because GHGRP emissions account for the use of any abatement technologies at nitric acid
production facilities. The GHGRP emissions are comparable to the national Inventory totals; therefore, the
use of GHGRP emissions to estimate the percentage of emissions by state does not appear to introduce
greater uncertainty for this time period.
For 1990-2009, this allocation method does not address utilization rates, which vary from facility to
facility and from year to year. While this approach implicitly accounts for the size of a facility in a state, it
could overestimate emissions in states where facilities used less of their capacity and underestimate
emissions in states where facilities used more of their capacity as a result of the lack of data on utilization
rates. This approach also does not account for abatement technologies at nitric acid production facilities
because the information is not known for this time period; therefore, this approach could overestimate
emissions in states where abatement technologies were used.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-26
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
3.2.3.4 Recalculations
The use of production capacity data from the SRI Directory of Chemical Producers for 1990-2005, and
the use of ICIS data for 2006-2009, resulted in changes to the total nitric acid capacity per year and per state.
These changes to the distribution of production capacities also resulted in corresponding changes to the
percentages of total national emissions estimated for each state. For years 1990-2005, recalculations show
that the production capacity (and emissions) per state decreased by less than 21 % from the percentages
used in the previous state emission estimates. For years 2006 and 2007, recalculations show that the
production capacity (and emissions) per state increased by 15% from the percentages used in the 1990-
2021 state emissions analysis.
Resubmissions of GHGRP data for 2020 and 2021 from one facility in Texas caused N20 emissions from
nitric acid production to increase by 33% (0.71 kt N20) and 42% (0.82 kt N20), respectively, compared to the
previous Inventory. Due to the resulting change in the overall percentages for all states, emissions from other
states decreased by 2.0% in 2020 and decreased by 3.0% in 2021.
3.2.3.5 Planned Improvements
Data gaps include nitric acid capacity for 2006-2007 and 2009, utilization rates per facility and state,
information about abatement technology installation and use per facility, and nitric acid production per state
for the full time series.
EPA will review time series consistency issues due to the two methodologies for 1990-2009 and 2010-
2022. Incomplete surrogate data on production capacity were used in place of activity data for the 1990-
2009 portion of the time series, and more research is needed to refine the method to enhance accuracy and
consistency of estimated state GHG emissions and trends.
3.2.3.6 References
EPA (U.S. Environmental Protection Agency) (2023) Facility Level Information on GreenHouse gases Tool
(FLIGHT). Data set as of August 18, 2023. Available online at: https://ghgdata.epa.gov/ghgp/.
EPA (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2022. EPA430-R-24-004. Available
online at: https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
ICIS (Independent Commodity Intelligence Services) (2008) Chemical Profile: Nitric Acid. Accessed February
18, 2021. Previously available online at:
https://www.icis.com/explore/resources/news/2008/05/19/9124327/chemical-profile-nitric-acid/.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
SRI (1990-2005). SRI International Directory of Chemical Producers.
3.2.4 Adipic Acid Production (NIR Section 4.8)
3.2.4.1 Background
Adipic acid is produced through a two-stage process during which N20 is generated in the second stage.
Emissions from fuels consumed for energy purposes duringthe production of adipic acid are accounted for
in the energy sector. The first stage of manufacturing usually involves the oxidation of cyclohexane to form a
cyclohexanone/cyclohexanol mixture. The second stage involves oxidizing this mixture with nitric acid to
produce adipic acid. N20 is generated as a byproduct of the nitric acid oxidation stage and, without
mitigation technology, is emitted in the waste gas stream. Process emissions from the production of adipic
3-27
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
acid vary with the types of technologies and Level of emissions controls employed by a facility. The largest
facility producing adipic acid uses an N20 abatement device, but its usage has varied considerably from year
to year over the period 2010-2022, resulting in varying levels of N20 control at that facility and varying levels
of total N20 emissions over that time period. Four adipic acid facilities, located in Florida, Texas, and Virginia,
have produced adipic acid in the United States from 1990 to 2022.
3.2.4.2 Methods/Approach
The national Inventory methodology was used to calculate state-level GHG emissions, using an
Approach 1 method as defined in the Introduction chapter of this report. The methodology for 2010-2022
used facility-level process emissions reported to the GHGRP (EPA 2023). The methodology for 1990-2009
used emissions calculations consistent with Tier 2 methods for two facilities and Tier 3 methods for the other
two facilities, as provided by the 2006 IPCC Guidelines (IPCC 2006). Emissions for each year were summed
by state (EPA 2023) over the full time series to determine disaggregated C02 emissions by state. See
Appendix D, Table D-5 in the "Adipic Acid" Tab, for more details on the data used. The GHGRP has no
reporting threshold for adipic acid production, so these emissions data are representative of the industry.
3.2.4.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of N20 from adipic acid production
was calculated using the 2006 IPCC Guidelines Approach 2 methodology for uncertainty (IPCC 2006). As
described further in Chapter 4 and Annex 7 of the national Inventory (EPA 2024), levels of uncertainty in the
national estimates in 2022 were -4%/+4% for N20.
State-level estimates are expected to have a slightly higher level of uncertainty than the national
Inventory over the full time series as a result of the rounding of the facility-level GHGRP process emissions
used to calculate the percentage of emissions from each state.
3.2.4.4 Recalculations
No recalculations were applied for this current report, consistent with the national Inventory.
3.2.4.5 Planned Improvements
There are no planned methodological refinements for the adipic acid production category.
3.2.4.6 References
EPA (U.S. Environmental Protection Agency) (2023) Facility Level Information on Greenhouse gases Tool
(FLIGHT). Data set as of August 18, 2023. Available online at: https://ghgdata.epa.gov/ghgp/.
EPA (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2022. EPA430-R-24-004. Available
online at: https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
3.2.5 Caprolactam, Glyoxal, and Glyoxylic Acid Production (NIR Section 4.9)
3.2.5.1 Background
Caprolactam is a colorless monomer produced for nylon 6 fibers and plastics. A substantial proportion
of the fiber is used in carpet manufacturing. In the most commonly used caprolactam production process,
benzene is hydrogenated to cyclohexane, which is then oxidized to produce cyclohexanone, which in turn is
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-28
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
used to produce caprolactam. The production of caprolactam can emit N20 from the ammonia oxidation
step. Since 1990, caprolactam has been produced in three states: Virginia, Texas, and Georgia. The facility in
Georgia closed in 2018.
EPA does not currently estimate the emissions associated with the production of glyoxal and glyoxylic
acid because of data availability and a lack of publicly available information on the industry in the United
States.
3.2.5.2 Methods/Approach
To compile emissions by state from caprolactam production, the state-level inventory disaggregated
national emissions from the national Inventory with an Approach 2 method, as defined in the Introduction
chapter of this report, using caprolactam production capacity by state byyearfor 1990-2022 as a surrogate
for caprolactam production data. The GHGRP does not currently cover caprolactam production. See
Appendix D, Table D-6 in the "Caprolactam" Tab, for more details on the data used. State-level emissions for
1990-2022 were estimated as a percentage of total national emissions by state and by year. Emissions of
N20 from the production of caprolactam were calculated using the Tier 1 method provided by the 2006 IPCC
Guidelines.
For 1990-2022, the total caprolactam production capacity in each state was divided by the total
caprolactam capacity in the United States to calculate a percentage of caprolactam capacity in each state
for each year. This percentage was applied to the national N20 emissions from caprolactam production per
year to calculate disaggregated N20 emissions by state for each year.
The caprolactam production capacities per facility, per state, were compiled from the SRI Directory of
Chemical Producers for 1990-1993 and 2004-2005 (SRI 1990-1993 and 2004-2005) and from ICIS for 2006.
The SRI Directory did not list capacity by facility for 1993-2003. The capacity data were applied to each
specific year, where available (1990-1993 and 2004-2006), 1993 SRI capacity data were applied to years
1994-2004, and 2006 ICIS capacity data were applied to years 2006-2022. An additional caprolactam facility
(Evergreen Recycling) was added for 2000 and 2001 (ICIS 2004, Textile World 2000) and for 2007-2015 (U.S.
Department of Energy 2011; Shaw Industries Group, Inc. 2015). Using the caprolactam capacity per state to
determine the state allocation percentage assumes that facility utilization rates are roughly the same from
state to state.
3.2.5.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of N20 from caprolactam
production was calculated using the 2006 IPCC Guidelines Approach 2 methodology for uncertainty (IPCC
2006). As described further in Chapter 4 and Annex 7 of the national Inventory (EPA 2024), levels of
uncertainty in the national estimates in 2022 were 31 %/+31 % for N20.
State-level estimates are expected to have a higher uncertainty because the national emissions
estimates were apportioned to each state based on caprolactam production capacity by state, by year for
1990-2022. This assumption was required because of a general lack of more granular state-level data.
For 1990-2022, this allocation method does not address utilization rates, which vary from facility to
facility and from year to year. While this approach implicitly accounts for the size of a facility in a state, it
could overestimate emissions in states where facilities used less of their capacity and underestimate
emissions in states where facilities used more of their capacity as a result of the lack of data on utilization
rates.
3-29
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
3.2.5.4 Recalculations
Recalculations were performed for 1990-2005 to reflect updated caprolactam capacity data from the
SRI Directory of Chemical Producers (SRI 1990-1993 and 2004-2005). State-level emissions for Georgia
decreased by 4% in 1990 and increased by an average of 4% per year from 1991 to 2005, compared to the
previous national Inventory. State-level emissions for Texas decreased by an average of 13% per year from
1990 to 2003 and decreased by 1 % per year for 2004-2005, compared to the previous national Inventory.
State-level emissions for Virginia increased by an average of 7% per year from 1990 to 2003 and decreased by
1 % per year for 2004-2005, compared to the previous national Inventory.
Recalculations were also performed for 2020 and 2021 to reflect updated caprolactam production data
from the American Chemistry Council's Guide to the Business of Chemistry (ACC 2023). Compared to the
previous Inventory, national annual N20 emissions decreased by 2% in 2020 and 2021, with a corresponding
percent decrease in Texas and Virginia in 2020 and 2021.
3.2.5.5 Planned Improvements
Data gaps to calculate emissions from caprolactam production include caprolactam production by
state for the full time series. Under the current methodology, data gaps include caprolactam capacities per
facility, per state, and utilization rates per facility for the full time series.
EPA will review time series consistency issues resulting from a lack of activity data (caprolactam
production) by state and the use of surrogate data (production capacity) that may not reflect reduced
production before facilities closed. More research is needed to refine the method to enhance accuracy and
consistency of estimated state GHG emissions and trends.
3.2.5.6 References
American Chemistry Council (ACC) (2023) Guide to the Business of Chemistry (Annual Data). Available online
at: https://www.americanchemistry.com/chemistry-in-america/data-industry-
statistics/resources/2023-guide-to-the-business-of-chemistrv
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
ICIS (Independent Commodity Intelligence Services) (2004) Chemical ProfileCaprolactam. Accessed
February 25, 2021. Previously available online at:
https://www.icis.com/explore/resources/news/2005/12/02/547244/chemical-profile-caprolactam/.
ICIS (2006) Chemical ProfileCaprolactam. Accessed February 25, 2021. Previously available online at:
https://www.icis.com/explore/resources/news/2006/10/18/2016832/chemical-profile-caprolactam/.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
Shaw Industries Group, Inc. (2015) Shaw Carpet Recycling Facility Successfully Processes Nylon and
Polyester. Available online at: https://shawinc.com/Newsroom/Press-Releases/Shaw-Carpet-Recvcling-
Facilitv-Successfullv-Proces/.
SRI (1990-1993, 2004-2005) SRI International Directory of Chemical Producers.
Textile World (2000) Evergreen Makes Nylon Live Forever. Textile World. October 1, 2000. Available online at:
https://www.textileworld.com/textile-world/textile-news/2000/10/evergreen-makes-nvlon-live-forever/.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-30
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
U.S. Department of Energy (2011) New Process Recovers and Reuses Nylon from Waste Carpeting Saving
Energy and Costs. Available online at: https://vwvw.energv.gov/eere/amo/nvlon-carpet-recvcling.
3.2.6 Carbide Production and Consumption (NIR Section 4.10)
3.2.6.1 Background
C02 and methane CH4 are emitted from the production of silicon carbide (SiC), a material used for
industrial abrasive, metallurgical, and other nonabrasive applications in the United States. Emissions from
fuels consumed for energy purposes during the production of SiC are accounted for in the energy sector. C02
and CH4 are also emitted duringthe production of calcium carbide, a chemical used to produce acetylene.
C02emissions from producing calcium carbide are implicitly accounted for in the storage factor calculation
for the nonenergy use (NEU) of petroleum coke in the energy sector. Methane emissions from calcium
carbide production are not estimated because data are not available.
3.2.6.2 Methods/Approach
Total emissions for each state are the sum of emissions from SiC production and SiC consumption. A
Hybrid approach, defined in the Introduction chapter of this report, was used to calculate emissions for each
state, as described below. To estimate state-level emissions from SiC production, national SiC production
data were evenly distributed among the two states identified as being home to SiC production facilities:
Illinois and Kentucky. See Appendix D, Table D-7 in the "Carbide Prod" Tab, for more details on the data
used. State-level estimates from SiC consumption were estimated using population statistics as a surrogate
for consumption data and used to disaggregate national SiC consumption emissions. See Appendix G, Table
G-1 in the "Population Data" Tab, for more details on the data used.
The national inventory methodology was adapted to calculate state-level GHG emissions of SiC to
ensure consistency with national estimates. National estimates were used to estimate state-level emissions
across states because of limitations in the availability of state-specific data for the time series.
3.2.6.2.1. SiC Production
Emissions of C02 and CH4from the production of SiC were calculated using Approach 1, as defined in
the Introduction chapter of this report, which is consistent with the Tier 1 method provided by the 2006 IPCC
Guidelines, and the same annual USGS production data (U.S. Bureau of Mines 1990-1993; USGS 1994,1995,
1996-2003, 2004-2017, 2020-2023) used in the national Inventory (EPA 2024). For the period 1990-2001,
reported USGS production data included production from two facilities located in Canada that ceased
operations in 1995 and 2001. U.S. SiC production for 1990-2001 was derived by subtracting SiC production
emissions data from Canada (ECCC 2022). Because of the lack of information on production level by state,
national SiC production data were evenly distributed among the two states identified in the USGS Minerals
Yearbook series as being home to SiC production facilities (Illinois and Kentucky). The state-level SiC
production was multiplied by the national emissions factors for C02 and CH4 to calculate GHG emissions by
state.
3.2.6.2.2. SiC Consumption
Emissions of C02 from the consumption of SiC were calculated using Approach 2, as defined in the
Introduction chapter of this report. SiC is used primarily for abrasive applications but also metallurgical and
other nonabrasive applications. Data on the consumption of SiC by state, however, are not available. To
calculate state-level C02 emissions from SiC consumption, national C02 estimates from the national
Inventory were distributed among the 50 states, the District of Columbia, and Puerto Rico using U.S.
population statistics as a surrogate for SiC consumption data (U.S. Census Bureau 2002, 2011, 2021, 2022;
3-31
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Instituto de Estadfsticas de Puerto Rico 2021). The fraction of the total U.S. population in each state, as well
as the District of Columbia and Puerto Rico, was calculated for each year by dividing the state population by
the total U.S. population. To estimate C02 emissions for each year by state, national Inventory C02 emissions
from SiC consumption were multiplied by each state's fraction of the total population for that year.
3.2.6.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02 from carbide production and
consumption was calculated using the 2006 IPCC Guidelines Approach 2 methodology for uncertainty (IPCC
2006). As described further in Chapter 4 and Annex 7 of the national Inventory (EPA 2024), levels of
uncertainty in the national estimates in 2022 were -10%/+10% for C02and -10%/+11% for CH4.
State-level estimates of production are expected to have a higher uncertainty because the national
emissions estimates were equally apportioned to each of the two states that produce SiC, which assumes
that they produce the same amount of SiC. There is also uncertainty due to the lack of information on
production processes and production levels at the two facilities.
State-level estimates of consumption also have a high uncertainty because national emissions
estimates were apportioned to all 50 states, the District of Columbia, and Puerto Rico using U.S. population
statistics as a surrogate for consumption. These assumptions were required because of a general lack of
more granular state-level data.
3.2.6.4 Recalculations
Recalculations were performed for 2020 and 2021 as updated population data were available from the
U.S. Census Bureau. The updated population data had a negligible impact (less than 0.5%) on the state-level
C02 kt emissions estimated for the 50 states and Puerto Rico for 2020 and 2021 due to the low emissions
estimated for each state or territory for the sector. Compared to the previous inventory, the District of
Columbia 2020 emissions decreased by less than 3% and 2021 emissions decreased by less than 0.5%.
3.2.6.5 Planned Improvements
Data gaps include the production of SiC by state and the consumption of SiC by state for the full time
series. Information to better simulate production at the two SiC facilities is needed and may include
researching state operating permits. EPA will research whether GDP from metal production or a relevant
NAICS code by state is available that would be a better surrogate than population for estimating SiC
consumption emissions.
3.2.6.6 References
ECCC (Environment and Climate Change Canada) (2022) Personal communication between Genevieve
Leblanc-Power, Environment and Climate Change Canada, and Mausami Desai and Amanda Chiu, U.S.
Environmental Protection Agency. April 12, 2022.
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
Instituto de Estadfsticas de Puerto Rico (2021) EstimadosAnuales Poblacionales de los Municipios Desde
1950. Accessed February 2021. Available online at: https://censo.estadisticas.pr/EstimadosPoblacionales.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-32
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
U.S. Bureau of Mines (1990-1991) Table 15. Crude Manufactured Abrasives Produced in the United States
and Canada, by Kind. In: Bureau of Mines Minerals Yearbook: Abrasive Materials. Available online at:
https://www.usgs.gov/centers/national-minerals-information-center/manufactured-abrasives-
statistics-and-information and https://www.usgs.gov/centers/national-minerals-information-
center/bureau-mines-minerals-yearbook-1932-1993.
U.S. Bureau of Mines (1992-1993) Table 12. End Uses of Crude Silicon Carbide and Aluminum Oxide
(Abrasive Grade) in the United States and Canada, as Reported by Producers. In: Bureau of Mines
Minerals Yearbook: Abrasive Materials. Available online at: https://www.usgs.gov/centers/national-
minerals-information-center/bureau-mines-minerals-yearbook-1932-1993.
U.S. Census Bureau (2002) Table CO-EST2001-12-00. In: Time Series of I ntercensal State Population
Estimates: April 1, 1990 to April 1,2000. Release date: April 11, 2002. Available online at:
https://www2.census.gov/programs-survevs/popest/tables/1990-2000/intercensal/st-co/co-est2Q01-12-
OO.pdf.
U.S. Census Bureau (2011) Table ST-EST00INT-01. In: Intercensal Estimates of the Resident Population for
the United States, Regions, States, and Puerto Rico: April 1, 2000 to July 1, 2010. Release date:
September 2011. Available online at: https://www2.census.gov/programs-survevs/popest/datasets/200Q-
2010/intercensal/state/st-est00int-alldata.csv.
U.S. Census Bureau (2021) Table NST-EST2020. In: Annual Estimates of the Resident Population for the
United States, Regions, States, and Puerto Rico: April 1, 2010 to July 1,2019; April 1, 2020; and July 1,
2020. Release date: July 2021. Available online at: https://www.census.gov/data/tables/time-
series/demo/popest/201 Os-state-total.html.
U.S. Census Bureau (2022) Table NST-EST2022-POP. In: Annual Estimates of the Resident Population for the
United States, Regions, States, District of Columbia, and Puerto Rico: April 1, 2020 to July 1, 2022.
Release date: December 2022. Available online at: https://www.census.gov/data/tables/time-
series/demo/popest/2020s-state-total.html.
USGS (U.S. Geological Survey) (1994) Table 2. End Uses of Crude Silicon Carbide and Aluminum Oxide
(Abrasive Grade) in the United States and Canada, as Reported by Producers. In: Minerals Yearbook:
Manufactured Abrasives. Available online at: https://d9-wret.s3-us-west-
2.amazonaws.com/assets/palladium/production/mineral-pubs/abrasives/040494.pdf.
USGS (1995) Table 2. End Uses of Crude Silicon Carbide and Aluminum Oxide (Abrasive Grade) in the United
States and Canada, as Reported by Producers. In: Bureau of Mines Minerals Yearbook: Manufactured
Abrasives. Available online at: https://d9-wret.s3-us-west-
2.amazonaws.com/assets/palladium/production/mineral-pubs/abrasives/040495.pdf.
USGS (1996-2003) Table 2. Production of Crude Silicon Carbide and Fused Aluminum Oxide in the United
States and Canada. In: Minerals Yearbook: Manufactured Abrasives. Available online at:
https://www.usgs.gov/centers/national-minerals-information-center/manufactured-abrasives-
statistics-and-information
USGS (2004-2017) Table 2. Estimated Production of Crude Silicon Carbide and Fused Aluminum Oxide in the
United States and Canada. In: Minerals Yearbook: Manufactured Abrasives. Available online at:
https://www.usgs.gov/centers/national-minerals-information-center/manufactured-abrasives-
statistics-and-information.
3-33
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
USGS (2020-2023) Mineral Commodity Summaries: Manufactured Abrasives. Available online at:
https://www.usgs.gov/centers/national-minerals-information-center/manufactured-abrasives-
statistics-and-information
3.2.7 Titanium Dioxide Production (NIR Section 4.11)
3.2.7.1 Background
Titanium dioxide (Ti02) is manufactured using one of two processes: the chloride process and the
sulfate process. The chloride process uses petroleum coke and chlorine as raw materials and emits process-
related C02. Emissions from fuels consumed for energy purposes during the production of Ti02 are
accounted for in the energy sector. The sulfate process does not use petroleum coke or other forms of
carbon as a raw material and does not emit process C02. Since 2004, all Ti02 produced in the United States
has been produced using the chloride process. Production of Ti02 in 2022 took place in Mississippi, Ohio,
Tennessee, and Louisiana.
3.2.7.2 Methods/Approach
To develop state-level estimates of emissions from Ti02 production, national emissions from the
national Inventory were disaggregated with an Approach 2 method as defined in the Introduction chapter of
this report, using a combination of GHGRP emissions data for 2010-2022 (EPA 2024; EPA 2023) as a
surrogate for Ti02 production data and production capacity for 1990-2009 (see Table 3-7). See Appendix D,
Tables D-8 and D-9 in the "Ti02" Tab, for more details on the data used.
The national Inventory methodology was adapted to calculate state-level GHG emissions of Ti02 to
ensure consistency with national estimates. National estimates were used to estimate state-level emissions
across states because of limitations in availability of state-specific activity data for the time series.
Emissions of C02 from Ti02 production were calculated using the Tier 1 method provided by the 2006
IPCC Guidelines and the same annual USGS production data (USGS 1991-2019, 2014-2022) used in the
national Inventory to calculate national emissions (EPA 2024). NationalTi02 production data were allocated
among the eight states with Ti02 production facilities over the 1990-2022 time series, based on GHGRP
emissions data or production capacity, and multiplied by the national emissions factor.
Table 3-7. Summary of Approaches to Disaggregate the National Inventory for Ti02 Production
Across Time Series
Time Series Range
Summary of Method
2010-2022
GHGRP process emissions data from Ti02 facilities were used to allocate
production by state, multiplied by the national emissions factor to get emissions
(IPCC 2006Tier 1).
1990-2009
USGS data on Ti02 production capacity were used to allocate production by
state, multiplied by the national emissions factor to get emissions (IPCC 2006
Tier 1).
The methodology used for 2010-2022 was based on GHGRP C02 emissions data reported by facilities
summed to state-level totals and used to estimate the fraction of total Ti02 produced in each state. The
GHGRP has no reporting threshold for Ti02, so these emissions data are representative of the industry. The
methodology used for 1990-2009 used USGS production capacity data for each facility to estimate the
fraction of total Ti02 produced in each state.
The estimated state-level Ti02 production was multiplied by the national emissions factor for C02 to
calculate GHG emissions by state (IPCC 2006).
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-34
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
3.2.7.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02 from Ti02was calculated
using the 2006 IPCC Guidelines Approach 2 methodology for uncertainty (IPCC 2006). As described further in
Chapter 4 and Annex 7 of the national Inventory (EPA 2024), levels of uncertainty in the national estimates in
2022 were -12%/+13% for C02.
State-level estimates are expected to have an overall higher uncertainty because the national emissions
estimates were apportioned to each state based on a combination of GHGRP emissions data for 2010-2022
and facility production capacity for 1990-2009. These assumptions were required because of a general lack
of more granular state-level data.
For 2010-2022, uncertainty is expected to be lower because of the use of GHGRP emissions data by
state as a surrogate for usingTi02 production data by state to calculate emissions. For 2010-2022, national
Inventory emissions have exceeded GHGRP emissions from 25% to 35%, possibly indicating that emissions
are overestimated in some states.
For 1990-2009, this allocation method does not address utilization rates, which vary from facility to
facility and from year to year, or differences in the carbon consumption rate for chloride and sulfate
processes. While this approach implicitly accounts for the size of a facility in a state, it could overestimate
emissions in states where facilities used less of their capacity and underestimate emissions in states where
facilities used more of their capacity as a result of the lack of data on utilization rates and production. This
method also does not account for different production processes. The sulfate process does not use
petroleum coke or other forms of carbon as a raw material and does not emit C02. Although the chloride
process has been the only one used in U.S. facilities since 2004, this allocation approach could overestimate
emissions in states where facilities used the sulfate process earlier in the time series.
3.2.7.4 Recalculations
USGS updated the estimated 2019 and 2020 Ti02 production values, and recalculations were performed
for those years. Compared to the previous inventory, C02 from Ti02 production decreased by 9% in 2019 (21
kt C02 for Louisiana, 59 kt C02 for Mississippi, 27 kt C02 for Ohio, and 27 kt C02 for Tennessee) and
increased by 12% in 2020 (20 kt C02 for Louisiana, 70 kt C02 for Mississippi, 28 kt C02 for Ohio, and 29 kt C02
for Tennessee).
3.2.7.5 Planned Improvements
Data gaps include state-level data on Ti02 production for the full time series 1990-2022. GHGRP
emissions data are available for the period 2010-2022 and were used for state inventory calculations, and
these data will be examined for possible use to improve data for the 1990-2009 period.
To address utilization rates that vary from facility to facility and from year to year, or differences in the
carbon consumption rate for chloride and sulfate processes, EPA will research how to account for varying
utilization rates and carbon consumption rate differences for sulfate (non-emissive) and chloride (emissive)
processes.
EPA will review potential time series consistency issues in the two methodologies for 1990-2009 and for
2010-2022. Surrogate data on production capacity were used in place of activity data for the 1990-2009
portion of the time series, and more research on data gaps (e.g., apply overlap technique) is needed to refine
the method to enhance accuracy and consistency of estimated state GHG emissions and trends.
3-35
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
3.2.7.6 References
EPA (U.S. Environmental Protection Agency) (2023) Facility Level Information on GreenHouse gases Tool
(FLIGHT). Data set as of August 18, 2023. Available online at: https://ghgdata.epa.gov/ghgp/.
EPA (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2022. EPA430-R-24-004. Available
online at: https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
USGS (U.S. Geological Survey) (1991-2019) Minerals Yearbook: Titanium.
USGS (2014-2022) Mineral Commodity Summaries: Titanium and Titanium Dioxide. Available online at:
https://pubs.usgs.gov/periodicals/mcs2022/mcs2022-titanium.pdf.
3.2.8 Soda Ash Production (NIR Section 4.12)
3.2.8.1 Background
C02is generated as a byproduct of calcining trona ore to produce soda ash and is eventually emitted
into the atmosphere. In addition, C02 may also be released when soda ash is consumed. Emissions from
soda ash consumption in chemical production processes are reported under other process uses of
carbonates, and emissions from fuels consumed for energy purposes duringthe production and
consumption of soda ash are accounted for in the energy sector.
3.2.8.2 Methods/Approach
All national soda ash production emissions can be attributed to Wyoming for the entirety of the 1990-
2022 time series. See Appendix D, Table D-10 in the "Soda Ash" Tab, for more details on the data used.
The national Inventory methodology was used to calculate state-level GHG emissions to ensure
consistency with national estimates, consistent with an Approach 1 method as defined in the Introduction
chapter of this report. As discussed in the national Inventory (EPA 2024), only two states produce natural
soda ash in the United States: Wyoming and California. Only C02 emissions from Wyoming soda ash
production facilities, which produced soda ash from trona ore, are included in the national estimate for the
1990-2022 time series because no C02 is emitted from the processes used in the California facility, which
produced soda ash from brines rich in sodium carbonate. Additionally, one facility in Colorado produced
soda ash from nahcolite between 2000 and 2004; however, similar to the California facility, the Colorado
facility's production process did not generate C02 emissions. As a result, all national C02 emissions can be
attributed to Wyoming for the entirety of the 1990-2022 time series. Emissions calculations are consistent
with the Tier 1 method provided by the 2006 IPCC Guidelines.
3.2.8.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02 from soda ash production
was calculated using the 2006 IPCC Guidelines Approach 2 methodology (IPCC 2006). As described further
in Chapter 4 and Annex 7 of the national Inventory (EPA 2024), levels of uncertainty in the national estimates
in 2022 were -9%/+8% for C02.
State-level estimates for soda ash production have a similar level of uncertainty as the national
Inventory over the full time series because the same methodology was used, and emissive soda ash
production takes place in one state.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-36
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
3.2.8.4 Recalculations
No recalculations were applied for this current report, consistent with Section 4.12 (page 4-61) of the
national Inventory.
3.2.8.5 Planned Improvements
There are no planned improvements for the soda ash production category. EPA will monitor the U.S.
soda ash production sector to ensure that any new production facilities using emissive processes are
accounted for in the state-level disaggregation.
3.2.8.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
3.2.9 Petrochemical Production (NIR Section 4.13)
3.2.9.1 Background
The production of some petrochemicals results in the release of C02 and CH4 emissions.
Petrochemicals are chemicals isolated or derived from petroleum or natural gas. C02emissions from the
production of acrylonitrile, carbon black, ethylene, ethylene dichloride, ethylene oxide, and methanol, as
well as CH4 emissions from the production of methanol and acrylonitrile, are discussed below. The
petrochemical industry uses primary fossil fuels (i.e., natural gas, coal, and petroleum) for nonfuel purposes
in the production of carbon black and other petrochemicals. Emissions from fuels and feedstocks
transferred out of the system for use in energy purposes (e.g., fuel combustion for indirect or direct process
heat or steam production) are currently accounted for in the energy sector.
In 2022, petrochemicals were produced at 76 facilities in 11 states (EPA 2024). Over 95% of total
production capacity is in Texas and Louisiana.
3.2.9.2 Methods/Approach
To develop state-level estimates of emissions from petrochemical production, EPA disaggregated
national emissions from the national Inventory to all applicable U.S. states and territories using production
capacities by petrochemical process and by state as a surrogate for emissions activity data. This
methodology is consistent with Approach 2, as defined in the Introduction chapter of this report. See
Appendix D, Tables D-11 through D-16 in the "Petrochemical" Tab, for more details on the data used.
The national Inventory methodology was adapted to calculate state-level GHG emissions from
petrochemical production to ensure consistency with national estimates. Consistency with the national
estimates and IPCC Guidelines requires reporting emissions by petrochemical type (i.e., acrylonitrile, carbon
black, ethylene, ethylene dichloride, ethylene oxide, and methanol). State-level emissions were estimated as
a percentage of total national emissions by state and by year.
The national Inventory-derived estimates for carbon black, ethylene, ethylene dichloride, and ethylene
oxide are based on facility-level GHGRP emissions for 2010-2022, and the Inventory-derived estimates for
methanol are based on facility-level GHGRP emissions for 2015-2022. The GHGRP has no reporting
3-37
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
threshold for petrochemical production, so these emissions data are representative of the industry. For all
petrochemicals in 1990-2009, and for methanol in 2010-2014, estimates were based on emissions factors
derived from GHGRP data and production data from the American Chemistry Council and the International
Carbon BlackAssociation (ACC 2023; EPA 2024). For all years, the national emissions estimates for
acrylonitrile were based on emissions factors and production data from the American Chemistry Council
because the national GHGRP data are considered CBI. Similarly, the national GHGRP data for methanol in
2010-2014 are considered CBI.
The method used for the national Inventory cannot be applied to derive state-level petrochemical
emissions due to GHGRP CBI concerns with all the petrochemical types when considering data by state. For
example, all ethylene oxide production facilities are in Louisiana and Texas. For reporting year (RY) 2019
through RY 2022, it appears that GHGRP emissions data could pass the CBI aggregation criteria in both
states; however, for RY 2010-2018, there were only three companies in Louisiana, so data cannot be
aggregated in either state for the same reasons noted below for ethylene, ethylene dichloride, and carbon
black.
GHGRP emissions data for ethylene, ethylene dichloride, and carbon black could also pass CBI
aggregation criteria at the state level in Louisiana and Texas (at least for RY 2019-2022); however, because
there are fewer than four companies making each of these petrochemicals in other states (typically only one
facility per state), it is not possible to aggregate the emissions by petrochemical type in Louisiana and Texas
without revealing the facility-specific emissions at the facilities in other states. Similarly, GHGRP emissions
data for methanol could pass CBI aggregation criteria at the state level in Texas for RY 2015-2022, but it is not
possible to aggregate the methanol emissions in Texas without revealing the facility-specific emissions at the
facilities in other states.
Aggregating total emissions from all types of petrochemical processes, rather than by type of
petrochemical, was also not possible because of CBI concerns, particularly the concern that aggregated
data for one state could reveal, or allow for back calculation of, CBI information about individual facilities in
other states. For example, some states have only one facility producing one type of petrochemical, and
reporting GHGRP emissions by state could disclose facility-specific data considered CBI for those states.
Aggregated GHGRP production data (i.e., the activity data used to calculate emissions when GHGRP
emissions are not available or do not meet CBI aggregation criteria) also have the same CBI concerns as
GHGRP emissions data.
As an alternative, production capacities were used as a surrogate for actual production and emissions
data. In effect, this approach assumes that all facilities producing a particular type of petrochemical have the
same capacity utilization and that emissions are proportional to production. As a result, this approach may
result in overestimating emissions for some states and underestimating emissions for other states.
To calculate emissions, the capacities per year per type of petrochemical per state were summed. The
fraction of the total capacity attributable to each facility in each year per state was determined. This
percentage was multiplied by the annual national Inventory emissions per petrochemical (i.e., the
aggregated GHGRP emissions for ethylene, ethylene dichloride, ethylene oxide, and carbon black in RY
2010-2022, the aggregated GHGRP emissions for methanol in RY 2015-2022, and the calculated nationwide
emissions for other years and for acrylonitrile in all years). For years where production capacity was not
known, data were extrapolated and interpolated to fill in data gaps. Several facilities have opened and closed
over the last 30 years; the precise years of facilities' operations were not always available because capacities
for only a handful of years were known. Details on how capacities were determined for each petrochemical
are described below.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-38
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
3.2.9.2.1. Acrylonitrile
Facility production capacity and Location data were available for 1990-1993, 2004, and 2005 from the
SRI Directory of Chemical Producers (SRI International 1990-2005) and for 2008, 2009, 2011, 2013, and 2017
from the ICIS (ICIS 2008, 2009a, 2011, 2013, 2017). Facility location data and the percentages of the
nationwide capacity held by the two companies with the largest percentage of the total nationwide capacity
were available for 1994-2003 from SRI (SRI International 1990-2005).
Several plants expanded between 1996 and 2001; the estimated capacities in the years prior to the
expansion were assumed to be the same as the previous known capacity in 1993, and the estimated
capacities in the years after the expansion were assumed to be the same as the known capacity in 2004.
Capacities in 2006 and 2007 were estimated using linear interpolation between the known values in 2005 and
2008. The capacities in 2010, 2012, 2014-2016, and 2018-2022 were assumed to be the same as the
previously known capacities. Some further adjustments were made when plant closings were known. For
example, one facility in Texas closed in 2005, and another closed in 2009. Additionally, the capacity for a
facility in Texas that was reported to be idle in 2002 was estimated as zero for 2002.
3.2.9.2.2. Carbon Black
ICIS capacity data were available for 1999, 2002, and 2005. For 1999, only a partial data set was
available; these data were not used because some of the data appeared to be inconsistent with data for
other years (ICIS 1999, 2002a, 2005). SRI data were available for all years between 1990 and 2005, except for
1995 (SRI International 1990-2005). For all years between 1990 and 2005, this analysis used SRI data.
Capacities for 1995 were estimated using linear interpolation between the known 1994 and 1996 values.
Capacities for 2006-2022 were assumed to be the same as in 2005. Five plants closed between 2001 and
2010. One plant in Texas closed early in 2003 and a second closed in 2010. The plant in Arkansas was idled in
2001 and was assumed to not reopen. One plant in West Virginia closed in 2008, and the second closed in
2009. Typically, when a plant was known to have closed during a year, it was assumed that half of the
nameplate capacity was available for that year.
3.2.9.2.3. Ethylene
SRI data on production capacities were available for 1990-1993, 2004, and 2005 (SRI International
1990-2005). The Oil & Gas Journal publishes capacities of ethylene production facilities, and data were
available for 2007, 2013, and 2015 (O&GJ 2007, 2013, 2015).
Because site-specific capacities for 1994-2003 were not known, a linear interpolation of capacities was
assumed between 1993 and 2004, except for known startups and shutdowns. This interpolation resulted in
the total capacity being nearly equal to or slightly less than the total annual production from 1996 through
2000, which suggests some of the more significant expansions must have occurred in the mid-1990s. One
plant in Texas started up in 1992. Due to the data in the 2004 SRI, it was assumed that this facility was
consolidated with a neighboring facility sometime before 2004. One plant in Louisiana started up in 1992.
One plant in Texas started up in 1994 and was expanded in 2002. Several plants closed between 1990 and
2005. One plant in Illinois closed in 1991, and one plant in Kentucky closed in 2000. One plant in Louisiana
closed in 2001. Two plants in Texas closed in 2003, and one plant in Texas closed in 2005.
Capacities for most facilities in 2006 were assumed to be the same as in 2005. However, a linear
interpolation between the known capacities in 2005 and 2007 was assumed for four facilities that had more
than a nominal difference in the known capacities for 2005 and 2007. Capacities for 2008-2012 and 2014
were estimated using linear interpolation between the known values in 2007, 2013, and 2015. Capacities for
2016-2022 were assumed to be the same as in 2015, except for new startups and expansions. One new plant
3-39
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
started up in Texas in 2017 and two in 2022, one new plant started up in Louisiana in 2019 and one in 2020,
one new plant started up in Pennsylvania in 2022, one idled plant was restarted in Louisiana in 2019, one
plant expanded in Texas in 2017, two plants expanded in Texas in 2018, one plant expanded in Louisiana in
2019, and one plant expanded in Texas in 2020 (BIC Magazine 2019; Chevron Phillips Chemical 2018;
ExxonMobil 2018; Indorama Ventures 2015; ExxonMobil 2022; LACC 2016; LyondellBasell 2017; OxyChem
2017; O&GJ 2020, 2022; Petrotahill 2020; TotalEnergies 2022). It was assumed that two plants in Texas
closed in 2013.
3.2.9.2.4. Ethylene Dichloride
The SRI Directory of Chemical Producers production capacity data for ethylene dichloride were
available for 1990,1991,1992,1993, 2004, and 2005 (SRI International 1990-2005). Facility location data
and the percentages of the nationwide capacity held by the companies that accounted for the top 50% of the
total nationwide capacity were available for 1994-2003 from SRI (SRI International 1990-2005). ICIS data on
production capacity are available for the years 2003, 2009, and 2018, although it is not clear whether the
data are complete (ICIS 2003, 2009b, 2018a). The 2003 report has capacities listed for 16 facilities, with two
being idle that year. The 2009 report lists capacities for 14 facilities, the 2018 report lists only 10 facilities,
and the total capacity reporting for 2018 is less than the assumed production in that year.
To maintain consistency, only SRI data were used for 1990-2005. Typically, linear interpolation was
used to estimate capacities for 1994-2003, except for three expansions at unknown dates in the late 1990s.
It was assumed that one facility expanded in 1996, one in 1998, and one in 1999. For one facility in this
Inventory, the linear interpolation values in 1999-2003 were replaced with the 2004 capacity based on new
information documenting that the facility expanded in 1998 (Nemeroff n.d.). In addition, two facilities from
the previous Inventory (one in Texas and one in Louisiana) were removed from the current Inventory because
it was determined that they produce ethylene dichloride using the direct chlorination process, which emits
negligible C02 emissions (SRI International 1990-2005). Making these assumptions resulted in corporate
capacity shares that agreed reasonably well with the SRI percentages.
For most facilities, the ICIS capacities in 2009 matched the SRI International capacities in 2005; thus,
the capacities for these facilities were assumed to be unchanged from 2005 to 2009. For three facilities in
2006-2008, a linear interpolation of capacities was assumed because the known capacities in 2005 and
2009 differed by more than a nominal amount. The capacities in 2010-2022 also were assumed to be the
same as in 2009, except for one facility in Louisiana that closed in 2011 and one newfacility in Louisiana that
started one new unit in 2010, a second new unit in 2011, and a third new unit in 2021.
The capacity utilization (dividing total production from the national Inventory by assumed capacity) was
calculated over the time period as a check on the capacity assumptions used. If production exceeded
assumed capacity, it would indicate the capacity assumptions were too low, while an extremely low-capacity
utilization could indicate that capacity assumptions were too high. The average total capacity utilization over
time was 74%, with a high of 91 % in 1997 and a low of 51 % in 2011. While these statistics indicate there may
be some overestimation or underestimation of capacity in a few years, they were still within the range of
possible values and no further adjustments to capacities were made.
3.2.9.2.5. Ethylene Oxide
SRI data were available for 1990-1993, 2004, and 2005 (SRI International 1990-2005). ICIS data on plant
capacities were available for 2004, 2010, 2012, and 2018 (ICIS 2004, 2010, 2012, 2018b). Facility location
data and the percentages of the nationwide capacity held by the companies that accounted for the top 50%
of the total nationwide capacity were available for 1994-2003 from SRI (SRI International 1990-2005). To
maintain consistency, all capacity estimates for 1990-2005 were based on SRI data, except when ICIS
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-40
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
information for a few facilities on the dates and size of expansions were applied to the SRI data. In the current
Inventory, the estimated capacity of one facility in 2002 was increased to better reflect the combination of
old unit shutdowns and startup of a new unit. Known capacities for 2005 typically were close to the known
capacities for 2010. Thus, in previous Inventories, capacities for 2006-2009 were assumed to be the same as
the previously known capacities in 2005. This approach was applied again for this Inventory, except in the
case of two facilities. In this Inventory, a linear interpolation of capacities was used for 2006-2009 for those
two facilities because the known values in 2005 and 2010 differed by more than a nominal amount.
Capacities for 2011 and 2013-2017 were based on linear interpolation between the known capacities in
2010, 2012, and 2018. All capacities in 2019-2022 were assumed to be the same as the known capacities in
2018, except for three facilities that started up in 2019 and one facility that started up in 2022.
There were several plant openings and closings and capacity changes over the time period. Plant
openings and closings were based on data provided in ICIS writeups, press releases, and other
documentation on company websites (as opposed to extrapolating over time). For example, calculations are
based on the information that one plant expanded in 1997, four in 1999, one in 2001, and one in 2002. The
resulting calculations of corporate capacity shares agreed reasonably well with the SRI percentages.
Capacities for new ethylene oxide units started up by Lotte, Sasol, and MEGlobal in 2019 and by Gulf Coast
Growth Ventures in 2022 were reported directly or could be estimated from other data reported on company
websites (EQUATE 2019; ICIS 2022; LACC 2016; Sasol 2019, 2020).
Capacity utilization was calculated over the time period as a check on the capacity assumptions used.
Assumed total capacity was generally greater than assumed total production from the national Inventory
across the time series, with the exception of 1995 and 2004 where production was 104% of capacity.
Conversely, when using total production from the American Chemistry Council for all years in the times
series, the capacity utilization values of 0.39-0.58 in 2019-2022 appear to be unrealistically low. While this
could mean capacities were overstated in these years, it also appears possible that the American Chemistry
Council production values may not include new on-site captive use production, which would bias the
nationwide production values to be low. Average capacity utilization based on production from the national
Inventory over time was 86%, and average capacity utilization based on production from the American
Chemistry Council for 1990-2022 was 78%. Although the data indicate there may be some overestimation or
underestimation of capacity in a few years, they were still within the range of possible values and no further
adjustments to capacities were made.
3.2.9.2.6. Methanol
SRI data on methanol production capacity were available for 1990- 1993, 2004, and 2005 (SRI
International 1990-2005). ICIS data were available for 2002, 2014, 2016, and 2018 (ICIS 2002b, 2014, 2016,
2018c). Facility location data and the percentages of the nationwide capacity held by the companies that
accounted for the top 50% of the total nationwide capacity were available for 1994-2003 from SRI (SRI
International 1990-2005). To maintain consistency, all capacity estimates for 1990-2005 were based on SRI
data.
Capacities in 1994-2003 typically were assumed to be the same as the preceding known value until a
known or assumed expansion year, and the capacities in years after the expansion were assumed to be the
same as the next known capacity. Capacities in 2006-2009 were assumed to be the same as the known
capacities in 2005, and capacities in 2010-2013 were assumed to be the same as the subsequent known
capacities in 2014. Capacities in 2015 were assumed to be the same as in 2014 and 2016, except for two new
facilities that started up in 2015, one facility that expanded in 2015, and one facility for which a linear
interpolation between the known capacities in 2014 and 2016 was used to estimate the capacity in 2015.
Capacities in 2017 were assumed to be the same as in 2016 and 2018. Capacities in 2019-2022 were
3-41
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
assumed to be the same as in 2018, except for one plant that started up in 2018 (and was not in the ICIS
2018c reference), one plant that started up in 2020, and one plant that started up in 2021. Data on startup
dates for expansions and new plants between 2012 and 2019 were obtained from documentation on
company websites (Celanese 2019; OCI 2018; OCI Partners LP 2016; Methanex 2017; Proman 2023). These
data were used to prorate capacities based on the approximate percentage of the year that they operated
after startup. Capacity for one new unit that started up in 2020 was estimated based on data in the permit to
install and operate (Ohio EPA 2017), and it was assumed to be in operation for 33% of the year based on
information provided in the Toxics Release Inventory Form R (EPA 2022). The capacity for a new plant started
up in 2021 was updated for this Inventory based on new information from the owner's website (Koch
Methanol St. James 2021).
Eight methanol plants closed between 1998 and 2010. Data on plant closures between 1998 and 2005
were from OCI (OCI Partners LP 2016, Appendix D). It was assumed that one plant closed in 2005 and
another in 2009 because that was the latest date for which any information about their operation could be
located, and neither facility reported to the GHGRP in the first year of reporting in 2010.
3.2.9.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02 and CH4 from petrochemical
production was calculated using the 2006 IPCC Guidelines Approach 2 methodology for uncertainty (IPCC
2006). As described further in Chapter 4 and Annex 7 of the national Inventory (EPA 2024), levels of
uncertainty in the national estimates in 2022 were -4%/+4%for C02 and -14%/+14% for CH4.
State-level estimates are expected to have a higher uncertainty because the national emissions
estimates were apportioned to each state based on facility production capacity. These assumptions were
required because the CBI concerns related to GHGRP data and a general lack of other more granular state-
level data.
This allocation method does not address actual utilization rates, which vary from facility to facility and
from year to year. While this approach implicitly accounts for the size of a facility in a state, it could
overestimate emissions in states where facilities used less of their capacity and underestimate emissions in
states where facilities used more of their capacity.
3.2.9.4 Recalculations
The following calculation corrections and changes in assumptions regarding some of the production
capacity data used in the previous Inventory resulted in minor changes to the total capacity per
petrochemical per year for ethylene dichloride, ethylene, and ethylene oxide production, and they also
resulted in changes to the percentage of total capacity in each state. These changes to the distribution of
production capacities also resulted in corresponding changes to the percentages of total national emissions
estimated for each state.
For ethylene dichloride, one facility in Louisiana and one facility in Texas were removed from the
analysis for the current Inventory because these facilities produce ethylene dichloride using the
direct chlorination process, which emits minimal C02 and is not subject to reporting under the
GHGRP. For another facility in Louisiana, a linear interpolation was used in the previous Inventory to
estimate the ethylene dichloride capacity for 1996-2002. New information confirmed that this
facility expanded in 1998 (Nemeroff n.d.). Therefore, in this Inventory, the capacity for this facility in
1996-1998 was assumed to be the same as the known capacity in 1995, and the capacity in 1999-
2002 was assumed to be the same as the known capacity in 2003. For a facility in Kentucky, the
capacity in 2006-2008 was assumed in the previous Inventory to be the same as the known capacity
in 2005. For this Inventory, a linear interpolation between the known capacities in 2005 and 2009
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-42
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
was assumed because the known values differed by more than a nominal amount. Collectively,
these changes resulted in increases in the percentage of total capacity in Kentucky by 21% to 43% in
2006-2008 and 7% to 9% in most other years. The percentage of total capacity in Louisiana typically
decreased by less than 1% per year, except in 2002-2005 where the capacity in Louisiana increased
by 3% to 5% and in 1999, 2000, and 2006-2009 where the capacity decreased by 2% to 3%.
For ethylene oxide, the capacity for one facility in Louisiana was changed in 2002 for this Inventory to
better reflect expansion and partial shutdown of existing units. This resulted in an increase of 4% of
the total capacity in Louisiana and a decrease of 3% in both Delaware and Texas in 2002.
For ethylene, a typographical error in the calculation spreadsheet resulted in the 1990 and 1991
capacities for one facility in Illinois being excluded from the analysis for the previous Inventory. After
correcting the error, the percentage of total capacity in Illinois increased by 40% in 1990 and 39% in
1991 for this Inventory. For a facility in Louisiana, a calculation error in the sum of the previous
capacity and the capacity of an expansion resulted in underestimation of the total expanded
capacity for this facility in 2020 and 2021 in the previous Inventory. Correcting this error resulted in
an increase of 2% in the percentage of total capacity in Louisiana in both years in this Inventory.
Reductions in the percentage of the total capacity for other states were less than 1 % in 1990,1991,
2020, and 2021.
A methodology refinement for calculating emissions from methanol production was implemented in the
national Inventory for 1990-2022. For 2015-2021, these changes resulted in a decrease in the reported C02
emissions, with the size of the decrease ranging from 43% (873 kt) in 2015 to 61% (2,110 kt) in 2018. For
1990-2014, the refinement resulted in a reduction of 61 % each year (287 kt in 2011 to 2,449 kt in 1997). There
were no changes in the estimated capacities per facility or in the percentage of total capacity in each state
for the current Inventory, but as a result of the decrease in nationwide emissions, emissions for each state
decreased by the same percentage as the reduction in emissions in the national Inventory. Additionally, the
methodology refinement reduced CH4 emissions from methanol production in the national Inventory to zero
for all years of the time series because the methodology refinement is based on the assumption that all
carbon input to the process is converted either to primary or secondary products or to C02. Although there
were no changes in the estimated capacities per facility or to the percentage of total capacity in each state,
the reduction of nationwide CH4 emissions to zero means that the CH4 emissions in each state have been
reduced by 100% in this Inventory.
3.2.9.5 Planned Improvements
Continued research is needed for more information on the timing of facility expansions, openings, and
temporary or permanent closures (e.g., permits, permit applications, trade industry data) and on facility
production capacities to address data gaps (e.g., additional versions of SRI International Directory of
Chemical Producers data, annual or biannual Oil & Gas Journal surveys of ethylene steam cracker
capacities).
For 2010-2022, the state-level inventory totals based on production capacity can be compared with the
GHGRP data on total emissions by state to assess how well the estimates represent the industry. Although
petrochemical production emissions by state and petrochemical type are CBI, total petrochemical
production emissions by state across all petrochemical types are not CBI under the GHGRP.
3.2.9.6 References
ACC (American Chemistry Council) (2023) Guide to the Business of Chemistry (Annual Data).
3-43
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
BIC Magazine (2019) Sasol Achieves Beneficial Operation of Louisiana Ethane Cracker. November 18, 2019.
Available online at: https://www.bicmagazine.com/industry/refining-petchem/sasol-achieves-
beneficial-operation-of-louisiana-ethane-cracker/.
Celanese (2019) Celanese Expands Methanol Production at Clear Lake Facility. April 17, 2019. Available online at:
https://investors.celanese.com/websites/celanese/English/30102Q/news-
detail.html?airportNewslD=90ec3195-255e-4178-9c84-6ab8e0d3d874.
Chevron Phillips Chemical (2018) Chevron Phillips Chemical Successfully Starts New Ethane Cracker in Baytown,
Texas. March 12, 2018. Available online at: https://www.cpchem.com/media-events/news/news-
release/chevron-phillips-chemical-successfully-starts-new-ethane-cracker.
EPA (U.S. Environmental Protection Agency) (2022) Form R Reports. Accessed May 16, 2022.
Available online at:
https://enviro.epa.gov/enviro/tri formr v2.fac list?rptvear=2020&facopt=fac name&fvalue=Alpont&fac
search=fac beginning&postal code=&city name=Qregon&county name=&state code=QH&industry t
vpe=&bia code=&tribe Name=&tribe search=fac beginning.
EPA (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2022. EPA430-R-24-004. Available online
at: https://www.epa.gov/ghgemissions/inventory-us-greenhouse-gas-emissions-and-sinks.
EQUATE (2019) EQUATE Group Announces Official Start-Up of MEGlobal Oyster Creek, TXSite. October 14,
2019. Available online at: https://www.meglobal.biz/ovster-creek-start-up/.
ExxonMobil (2018) ExxonMobil Starts Up New Ethane Cracker in Baytown, Texas. July 26, 2018. Available online at:
https://corporate.exxonmobil.com/news/news-releases/2018/0726 exxonmobil-starts-up-new-ethane-
cracker-in-baytown-texas.
ExxonMobil (2022) ExxonMobil, Sabic Start Operations at Gulf Coast Manufacturing Facility. January 20, 2022.
Available online at: https://corporate.exxonmobil.com/news/news-releases/2022/0120 exxonmobil-
and-sabic-start-operations-at-gulf-coast-manufacturing-facility.
ICIS (Independent Commodity Intelligence Services) (1999) Carbon Black. August 29,1999. Accessed February 18,
2021. Previously available online at:
https://www.icis.com/explore/resources/news/1999/08/30/93545/carbon-black/.
ICIS (2002a) Chemical ProfileCarbon Black. June 23, 2002. Accessed May 18, 2021. Previously available online at:
https://www.icis.com/explore/resources/news/2005/12/02/175824/chemical-profile-carbon-black/.
ICIS (2002b) Chemical ProfileMethanol. December 16, 2002. Accessed May 18, 2021. Previously available online
at: https://www.icis.com/explore/resources/news/2005/12/02/186772/chemical-profile-methanol/.
ICIS (2003) Chemical Profile Ethylene Dichloride. November 10, 2010. Accessed May 18, 2021. Previously available
online at: https://www.icis.com/explore/resources/news/2005/12/Q2/532639/chemical-profile-ethylene-
dichloride/.
ICIS (2004) Chemical Profile Ethylene Oxide (EO). October 10, 2004. Accessed May 18, 2021. Previously available
online at: https//www.icis.com/explore/resources/news/2005/12/08/618912/chemical-profile-ethvlene-
oxide-eo-/.
ICIS (2005) Chemical Profile: Carbon Black. June 19, 2005. Accessed May 18, 2021. Previously available online at:
https://www.icis.com/explore/resources/news/2005/06/17/686446/chemical-profile-carbon-black/.
ICIS (2008) Chemical Profile: Acrylonitrile. August 17, 2008. Accessed May 18, 2021. Previously available online at:
https://www.icis.com/explore/resources/news/2008/08/18/9149113/chemical-profile-acrylonitrile/.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-44
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
ICIS (2009a) Chemical Profile: Acrylonitrile. January 27, 2009. Accessed May 18, 2021. Previously available online
at: https://www.icis.com/explore/resources/news/2009/01/27/9188094/chemical-profile-acrvlonitrile/.
ICIS (2009b) US Chemical Profile: Ethylene Dichloride. September 13, 2009. Accessed May 18, 2021. Previously
available online at: https://www.icis.com/explore/resources/news/2009/Q9/14/9246931/us-chemical-
profiLe-ethyLene-dichloride/.
ICIS (2010) US Chemical Profile: Ethylene Oxide. August 1, 2010. Accessed May 18, 2021. Previously available online
at: https://www.icis.com/explore/resources/news/2010/08/02/938Q662/us-chemical-profile-ethvlene-
oxide/.
ICIS (2011) U.S. Chemical Profile: Acrylonitrile. September 4, 2011. Accessed May 18, 2021. Previously available
online at: https://www.icis.com/expLore/resources/news/2011/09/05/9489888/us-chemicaL-profiLe-
acrvLonitriLe/.
ICIS (2012) US Chemical Profile: Ethylene Oxide. March 12, 2012. Accessed May 18, 2021. Previously available
online at: https://www.icis.com/expLore/resources/news/2012/Q3/12/9539954/us-chemicaL-profiLe-
ethvLene-oxide/.
ICIS (2013) Chemical Profile: US Acrylonitrile. May 24, 2013. Accessed May 18, 2021. Previously available online at:
https://www.icis.com/expLore/resources/news/2013/05/26/9672000/chemicaL-profiLe-us-acryLonitriLe/.
ICIS (2014) Chemical Profile: US Methanol. May 2, 2014. Accessed May 18, 2021. Previously available online at:
https://www.icis.com/expLore/resources/news/2014/05/02/9777442/chemicaL-profiLe-us-methanoL/.
ICIS (2016) Chemical Profile: US Methanol. April 7, 2016. Accessed May 18, 2021. Previously available online at:
https://www.icis.com/expLore/resources/news/2016/04/07/9986081/chemicaL-profiLe-us-methanoL.
ICIS (2017) Chemical Profile: US Acrylonitrile. January 12, 2017. Accessed May 18, 2021. Previously available online
at: https://www.icis.com/expLore/resources/news/2017/01/12/10Q69751/chemicaL-profiLe-us-
acrvLonitriLe/.
ICIS (2018a) Chemical Profile: US Ethylene Dichloride. August 10, 2018. Accessed May 18, 2021. Previously available
online at: https://www.icis.com/expLore/resources/news/2018/08/09/1Q249213/chemicaL-profiLe-us-
ethvLene-dichLoride/?redirect=engLish.
ICIS (2018b) Chemical Profile: US Ethylene Oxide. April 13, 2018. Accessed May 18, 2021. Previously available
online at: https://www.icis.com/expLore/resources/news/2018/04/12/10211482/chemicaL-profiLe-us-
ethyLene-oxide/.
ICIS (2018c) Chemical Profile: US Methanol. September 13, 2018. Accessed May 18, 2021. Previously available
online at: https://www.icis.com/expLore/resources/news/2018/09/13/1Q259297/chemicaL-profiLe-us-
methanoL/.
ICIS (2022) ExxonMobil, Sabic JVExpects to Start US EG PE Complex in Q4 '21. PreviousLy avaiLabLe onLine at:
https://www.icis.eom/expLore/resources/news/2020/11/13/10574908/exxonmobiL-sabic-jv-expects-to-
start-us-eg-pe-compLex-in-q4-21/.
Indorama Ventures (2015) Indorama Ventures Olefins: Acquisition in 2015. Available online at:
https://www.indoramaventures.com/en/worLdwide/1214/indorama-ventures-oLefins.
IPCC (Intergovernmental Panel on Climate Change) (2006) 2006IPCC Guidelines for National Greenhouse Gas
Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. AvaiLabLe onLine at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
3-45
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Koch Methanol St. James (2021) Who We Are. Available online at: https://www.kochmethanol.com/koch-
methanol-st-james/who-we-are/.
LACC (2016) Axiall and Lotte Chemical Hold LACC Ethane Cracker Groundbreaking Ceremony. June 14, 2016.
Available online at: https://www.westlake.com/axiall-and-lotte-chemical-hold-lacc-ethane-cracker-
groundbreaking-ceremony.
LyondellBasell (2017) LyondellBasell Corpus Christi Complex Expansion Complete. January 19, 2017. Available
online at: https://www.lvondellbasell.com/en/news-events/corporate--financial-news/lvondellbasell-
corpus-christi-complex-expansion-complete/.
Methanex (2017) Methanex: Corporate History. Available online at:
https://www.methanex.com/sites/default/files/news/media-
resources/MX%20Corporate%20Historv 2017.pdf.
Nemeroff (n.d.) Vulcan Occidental Geismar. Available online at:
https://www.nemerofflaw.com/asbestos/asbestos-job-sites/louisiana/vulcan-occidental-geismar/.
OCI (2018) Natgasoline LLC Begins Production at Largest Methanol Facility in the United States. June 25, 2018.
Available online at: https://www.oci.nl/media/1335/natgasoline-begins-methanol-production-final.pdf.
OCI Partners LP (2016) OCI Partners LP: 3Q 2016 Results Presentation. Available online at:
http://ocipartnerslp.investorroom.com/presentations.
O&GJ (Oil & Gas Journal) (2007) International Survey of Ethylene from Steam Crackers. July 16, 2007.
O&GJ (2013) International Survey of Ethylene from Steam Crackers.
O&GJ (2015) International Survey of Ethylene from Steam Crackers. July 6, 2015.
O&GJ (2020) Shintech Commissions Louisiana Ethylene Plant. February 14, 2020. Available online at:
https://www.ogj.com/refining-processing/petrochemicals/article/14167793/shintech-commissions-
louisiana-ethvlene-plant.
O&GJ (2022) Shell Commissions Pennsylvania Petrochemical Complex. November 15, 2022. Available online
at: https://www.ogj.com/refining-processing/petrochemicals/article/14285782/shell-commissions-
pennsvlvania-petrochemical-complex.
Ohio EPA (Environmental Protection Agency) (2017) Final Air Pollution Permit-to-lnstall and Operate. Permit
number P0122449. Previously available online at:
http://wwwapp.epa.ohio.gov/dapc/permits issued/1595429.pdf.
OxyChem (2017) OxyChem and Mexichem Announce Startup of Their Joint Venture Ethylene Cracker in Ingleside,
Texas. February 27, 2017. Available online at:
https://www.businesswire.com/news/home/20170227006515/en/QxvChem-and-Mexichem-Announce-
Startup-of-their-Joint-Venture-Ethylene-Cracker-in-lngleside-Texas.
Petrotahill (2020) Formosa Plastics' New Texas Cracker Starts Operating. January 16, 2020. Available online at:
http://www.petrotahlil.com/Section-news-2/43843-formosa-plastics-new-texas-cracker-starts-
operating.
Proman (2023) Proman USA (Pampa). Available online at: https://www.proman.org/companies/pampa-fuels/.
Sasol (2019) Ethylene Oxide/Ethylene Glycol (EO/EG) Unit: Fast Facts. Accessed July 16, 2021. Available online at:
https://web.archive.org/web/20210622162954/https://3lkev1w1rsfaewp8njfu43u6-wpengine.netdna-
ssl.com/wp-content/uploads/2019/07/fact sheet eoeg.pdf.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-46
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
Sasol (2020) Project Update. June 1, 2020. Available online at:
https://web.archive.org/web/20220520135142/sasolnorthamerica.com/proiectupdate/.
SRI International (1990-2005) Directory of Chemical Producers: United States of America.
TotalEnergies (2022) United States: TotalEnergies Announces the Start-up of New Ethane Cracker in Port
Arthur. July 21, 2022. Available online at: https://totalenergies.com/media/news/press-releases/united-
states-totalenergies-announces-start-new-ethane-cracker-port.
3.2.10 HCFC-22 Production (NIR Section 4.14)
3.2.10.1 Background
Trifluoromethane (HFC-23 or CHF3) is generated as a byproduct during when manufacturing
chlorodifluoromethane (HCFC-22), which is used as a feedstock for several fluoropolymers. Before 2010,
HCFC-22 was widely used as a refrigerant, but its production and import for this application in the United
States were phased out between 2010 and 2020 under Title VI of the Clean Air Act, which controls production
and consumption of HCFCs and other compounds that deplete stratospheric ozone. Production of HCFC-22
for use as a feedstock is allowed to continue indefinitely.
3.2.10.2 Methods/Approach
As discussed on page 4-74 of the national Inventory, methods comparable to the Tier 3 methods in the
2006 IPCC Guidelines (IPCC 2006) were used to estimate HFC-23 emissions for five of the eight HCFC-22
plants that have operated in the United States since 1990. For the other three plants, the last of which closed
in 1993, methods comparable to the Tier 1 method in the 2006 IPCC Guidelines were used. However, as
discussed further below, EPA does not have access to the individual plant estimates for 1990-2009; for those
years, EPA has access only to national totals aggregated across the plants.
To develop state-level estimates of HFC-23 emissions from HCFC-22 production, EPA disaggregated
national emissions from the national Inventory using a combination of facility-level reporting to the GHGRP
from 2010-2022, reports verifying emissions by facility for earlier years, and production capacity data, as
shown in Table 3-8 below. The sum of emissions by state is consistent with national process emissions as
reported in the national Inventory over the time series.
Table 3-8. Summary of Approaches to Disaggregate the National Inventory for HCFC-22
Production Across Time Series
Time Series Range
Summary of Method
1990-2009
Facility-specific information on emissions control efforts and production
capacities, in combination with facility-specific GHGRP data for 2010, were
used to estimate emissions by state (Approach 2).
2010-2022
Facility-specific GHGRP data on HFC-23 emissions were compiled by state
(Approach 1).
For each state, HFC-23 emissions from 2010-2022 were drawn from facility-level reporting to the
GHGRP. The same data were used for the national Inventory.
Facility-level reports of HFC-23 emissions are not available foryears before 2010, which was the first
year of GHGRP reporting. As described in the national Inventory, national totals for 1990-2009 were based on
totals provided to EPA by the Alliance for Responsible Atmospheric Policy, which aggregated the HFC-23
emissions and HCFC-22 production reported to the Alliance by each HCFC-22 production facility and HFC-
23 destruction facility. (A list of the nine facilities that have operated in the United States since 1990, their
locations, and dates of opening or closure is shown in Table 3-9 below.) These totals, as well as the individual
3-47
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
facility reports, were reviewed and corrected, as necessary, by an EPA contractor in 1997 and 2008. The
totals and qualitative information on each plant's emissions estimation methods, trends, and control
measures were summarized in two reports. EPA used the second of these reports, Verification of Emission
Estimates ofHFC-23 from the Production of HCFC-22: Emissions from 1990 through 2006 (RTI International
2008), hereinafter referred to as the 2008 Verification Report, to estimate facility-level emissions and develop
state-level estimates for 1990-2009. EPA also used GHGRP data from 2010-2022 and the estimated 2003
HCFC-22 production capacity of each facility from the 2004 edition of the Chemical and Economics
Handbook (CEH) Research Report: Fluorocarbons (SRI Consulting 2004).
In combination with two key trends seen at the national level, these resources provide some insight into
the magnitudes and trends of emissions of the various facilities. The two key national trends are a steady
decrease in the HFC-23/HCFC-22 emissions factor from 1990 to 2010 and a slow increase in HCFC-22
production from 1990 to 2000, followed by fluctuating production through 2007, and then a decline in later
years. The 2008 Verification Report indicates that the downward trend in the emissions factor was at least
partially driven by (1) the closure during the early 1990s of four HCFC-22 production facilities whose
emissions were uncontrolled and whose production was replaced by a facility that opened in 1993 in
Alabama with tight emissions controls and (2) actions taken by a production facility in Kentucky to
significantly reduce its emissions rate beginning in 2000. While HCFC-22 production and production capacity
data were not available for all the plants operating before 2003, the generally upward trend in national
production seen between 1990 and 2003 indicates that the closure of the four plants in the early 1990s, in
combination with the opening of the Alabama plant in 1993, likely did not result in a significant net loss of
production capacity in the United States as a whole during that period. Thus, EPA estimated production at
the four plants by equating their joint production capacity to that of the Alabama plant, which was available
from the CEH report.
To allocate national emissions to each facility, EPA first back-cast the relatively small emissions
reported by the HCFC-22 production facility in Alabama and one HFC-23 destruction facility in West Virginia.
As noted above, the Alabama HCFC-22 production facility was known to have tightly controlled HFC-23
emissions since it began operating in 1993; thus, emissions from 1996-2009 were assumed to equal the
average of the emissions reported by this facility from 2010-2014, a period during which emissions were
relatively flat before they began to decline in 2015. (Emissions from 1993-1996 were assumed to rise
gradually as the plant replaced HCFC-22 production from closing plants.) The HFC-23 destruction facility in
West Virginia is understood to have begun destroying HFC-23 in 2000 when an HCFC-22 production facility
owned by the same company began capturing byproduct HFC-23 and shipping some of it to the West Virginia
facility for destruction. Emissions from 2000-2009 were equated to the average emissions reported by the
West Virginia facility under Subpart O of the GHGRP from 2010-2013 (about 3 kg per year), after which
emissions dropped.
To estimate the 2003-2009 emissions from the other two HCFC-22 production facilities that operated
during that period (in Kentucky and Louisiana), the emissions estimated for the Alabama and West Virginia
facilities were subtracted from the national total, and the remaining emissions were then allocated to the
Kentucky and Louisiana facilities based on each facility's estimated production and estimated emissions
rate. The production of each facility throughout the time series was estimated based on the 2003 capacity
reported in the CEH report. The 1999 emissions rates of both facilities were assumed to be equal to the
national emissions rate in that year after subtracting out the estimated emissions and production of the
controlled Alabama facility; the resulting emissions rate was 0.018 kg HFC-23/kg HCFC-22. The emissions
rate of the Louisiana facility was assumed to have remained constant at this level based on the
characterization of that facility's emissions control efforts in the 2008 Verification Report. The emissions rate
for the Kentucky facility was assumed to have declined linearly to 0.005 kg HFC-23/kg HCFC-22 as the facility
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-48
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
implemented the emissions reduction efforts documented in the 2008 Verification Report.28 To estimate the
share of national emissions attributable to each facility, each facility's estimated production was multiplied
by its estimated emissions rate, resulting in a provisional emissions estimate for each facility for each year.
Each facility's provisional emissions estimate was then divided by the sum of the provisional emissions
estimates for both facilities. The resulting fraction was multiplied by the national emissions (minus the
emissions of the Alabama and West Virginia facilities) to obtain the final estimate of emissions for each
facility.
To estimate facility-level emissions from 1990 to 2002, it was necessary to account for the emissions of
the five HCFC-22 production facilities that ceased production before 2003. These facilities, which operated
through 1991-1993,1995, and 2002, did not have production capacities listed in the CEH report and did not
control their emissions, based on the 2008 Verification Report. The production capacity of the facility that
operated through 2002, in Kansas, was estimated as the difference between the total U.S. HCFC-22
production in 2000 and the sum of the CEH-estimated production capacities for the other three plants in
operation during that year. (U.S. HCFC-22 production reached a peak in 2000.) This plant was assumed to
have linearly decreased production to zero between 2000 and 2003. Its emissions factor was assumed to
equal the value calculated for uncontrolled plants in 1999, at 0.018 kg HFC-23/kg HCFC-22. U.S. emissions
from 2000-2002 were then allocated to this plant and to the Kentucky and Louisiana plants as described
above.
As noted earlier, the production capacities of the four facilities that closed in the early 1990s were each
assumed to equal one-fourth of the production capacity of the Alabama facility that opened in 1993.
Because none of the four plants controlled their emissions, their emissions factors were assumed to be
equal to those of the Kansas, Kentucky, and Louisiana plants from 1990 to 1999. U.S. emissions (minus
those of the Alabama plant) from 1990-1999 were therefore allocated to each facility based on its estimated
share of U.S. HCFC-22 production capacity.
Table 3-9. Facilities Producing HCFC-22 or Destroying HFC-23 Generated During
HCFC-22 Production from 1990 to 2022
Years When HCFC-22 Was
Company
Plant Location
Produced or HFC-23 Was
Destroyed
Arkema
Calvert City, KY
1990-1991
Wichita, KS
1990-2002
Clean Harbors
El Dorado, AR
2019
Montague, Ml
1990-1995
DuPont/Chemours
Louisville, KY
1990-2022
Washington, WV
2000-2022
Honeywell
El Segundo, CA
1990-1992
Baton Rouge, LA
1990-2012
LaRoche Industries
Gramercy, LA
1990-1993
MDA
Manufacturing/Daikin
Decatur, AL
1993-2022
28 The 0.005 emissions factor was estimated by subtracting the 2010 HFC-23 emissions reported by the other facilities from
the national emissions total, subtracting the 2010 production estimated for the other facilities (based on their production
capacities and national production) from the 2010 national production total, and dividing the first by the second.
3-49
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
3.2.10.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of HFC-23 from HCFC-22
production was calculated using the 2006 IPCC Guidelines Approach 2 methodology (IPCC 2006). As
described further on page 4-75 of the national Inventory (EPA 2024), the uncertainty in the national estimate
in 2022 was estimated at -7%/+10%. Based on an uncertainty analysis that was performed for the 2008
Verification Report, the uncertainties in the emissions of the individual plants that have accounted for most
of the emissions since 2010 (i.e., the plants in Kentucky and Louisiana) were comparable to this uncertainty
in 2006 (-5%/+11 % and -9%/+11 %, respectively). The 2006 uncertainty in the much smaller emissions from
the plant in Alabama was estimated at -48%/+47%. Because the methods used to estimate emissions at
these plants are not believed to have changed significantly since 2006, and because plant-level emissions
data are available for these plants for 2010 and later years, the uncertainties in the emissions of the
Kentucky, Louisiana, and Alabama plants for 2010 and later years are believed to be similar to those
estimated in the 2008 Verification Report.
For the years 1990-2009, plant-level data are not available, significantly increasing the uncertainty of
emissions estimates for individual facilities and states. This is particularly true for the five HCFC-22
production facilities that closed before 2003, for which production capacity data are therefore not available.
The uncertainties of the emissions of these five facilities also increased the uncertainties of the 1990-2002
emissions of the three HCFC-22 production facilities for which production capacity data are available,
because the (unknown) production at the five facilities probably affected the capacity utilization of the other
three. Capacity utilization can vary significantly across plants and from year to year.
3.2.10.4 Recalculations
The 2019 emissions estimate for Arkansas increased from 0 to 0.05 kg of HFC-23 to reflect newly
reported emissions from a facility that destroys HFC-23.
3.2.10.5 Planned Improvements
During the 2007-2008 review of the HFC-23 emissions estimates provided to EPA by the Alliance for
Responsible Atmospheric Policy, RTI International (EPA's contractor) was able to review the annual
estimates of individual HCFC-22 production facilities, but under the confidentiality agreements in place at
the time of the review, EPA did not have direct access to the individual plant- or facility-level estimates. If one
or more HCFC-22 production facilities were able to share their 1990-2009 emissions estimates with EPA,
this would considerably reduce the uncertainty of EPA's 1990-2009 state-level estimates.
3.2.10.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
RTI International (2008) Verification of Emission Estimates of HFC-23 from the Production of HCFC-22:
Emissions from 1990 Through 2006. U.S. Environmental Protection Agency.
SRI Consulting (2004) CEH Market Research Report: Fluorocarbons.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-50
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
3.2.11 Production of Fluorochemicals Other than HCFC-22 (NIR Section 4.15)
3.2.11.1 Background
Fluorochemical production includes processes that produce or transform saturated and unsaturated (HFCs,
PFCs, SFs NF3, hydrofluoroethers (HFEs), perfluoroalkylamines, and other fluorinated compounds. Emissions may
include reactants, products, and byproducts from the production or transformation process; residual gas vented
from containers; and residual emissions from destruction of previously produced fluorinated GHGs. Most
saturated HFCs were developed for use as replacements for or alternatives to ozone-depleting substances such as
CFCs and HCFCs that have been phased out under the Montreal Protocol, and many saturated HFCs are now
themselves being phased out under the Kigali Agreement and U.S. AIM program. PFCs are commonly used in the
semiconductor industry. SF6 is used for electric power systems, magnesium production, and electronics
manufacturing, and NF3 is also used in the semiconductor industry. Other fluorinated GHGs are used for a variety
of purposes (e.g., for firefighting, as anesthesia, and as feedstocks for fluoropolymer production). Fluorinated GHG
emissions from the national Inventory were disaggregated across states in 2023 using facility-level reporting to the
GHGRP from 2011 to 2022 and production data, emission factors, and facility-provided emissions data for earlier
years.
3.2.11.2 Methods/Approach
As discussed on page 4-81 of the national Inventory, methods comparable to the Tier 3 methods in the
2006 IPCC Guidelines (as elaborated by the 2019 Refinement) were used to estimate fluorinated GHG
emissions from most U.S. facilities producing fluorinated compounds, while the Tier 1 method was used to
estimate fluorinated GHG emissions from U.S. production facilities for which there are fewer data. For the
facilities for which Tier 3 methods were used, facility-specific estimates had been developed and summed to
arrive at the estimates in the national Inventory. For this analysis, therefore, those facility-specific estimates
were readily available to disaggregate to the states where the facilities are located. The same was true for
one facility for which the Tier 1 method was used, relying on publicly available production capacity data. For
the other facilities for which the Tier 1 method was used, confidentiality concerns prohibit the publication of
facility-specific emissions estimates because facility-specific production can be back-calculated from the
emissions and the Tier 1 emission factors. Thus, for these facilities, the total emissions calculated for the
facilities were divided by the number of the facilities operating in each year, and the results were allocated to
the states where those facilities are located. The sum of emissions by state is consistent with national
process emissions as reported in the national Inventory over the time series. Table 3-10 summarizes the
approaches used to disaggregate the national Inventory for fluorochemical production across the time
series.
Table 3-10. Summary of Approaches to Disaggregate the National Inventory for Fluorochemical
Production Across Time Series
Time Series Range
Summary of Method
1990-2010
For 17 facilities, facility-specific estimates from the national Inventory were
compiled by state (Approach 1). For five facilities, the national estimate for all
five facilities was divided by five and allocated to each state where the
facilities were located (Approach 2).
2011-2022
For 17 facilities, facility-specific estimates from the national Inventory were
compiled by state (Approach 1). For five to seven facilities, the national
estimate for all five to seven facilities was divided by five to seven, as
applicable in that year, and allocated to each state where facilities were
located (Approach 2).
3-51
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
3.2.11.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of fluorinated GHG emissions from
fluorochemical production was calculated using the 2006 IPCC Guidelines Approach 1 methodology for uncertainty
(IPCC 2006). As described further in Chapter 4 (pages 4-90 to 4-95) and in Annex 7 of the national Inventory (EPA
2024), the uncertainty in the national estimates in 2022 was estimated at -19%/+19% for fluorinated GHGs (HFCs,
PFCs, SFs, NFs).
Emissions uncertainties at the state level are higher than emissions uncertainties at the national level
because most states contain only one or two facilities, providing less of an opportunity for facility-level
uncertainties to "cancel out" over a large number of facilities. Uncertainties at the state level are likely to be only
slightly smaller than uncertainties at the facility level. For fluorochemical production facilities that reported their
emissions to the GHGRP in 2022, the relative uncertainties of facility-level emissions are estimated to have ranged
from ±17% to ±89%, depending on the shares of emissions coming from process vents whose emission factors
have been measured, process vents whose emission factors have been calculated, leaks, and container venting, all
of which have different uncertainties. The average relative uncertainty of emissions from facilities that reported
their emissions under the GHGRP was estimated to be ±47%. For fluorochemical production facilities that reported
only production under the GHGRP, the relative uncertainties of facility-level emissions are estimated at ±98%, but
the actual uncertainties for the estimates for these facilities in this analysis are higher because the facility-specific
emissions are calculated by dividing the total emissions across these facilities by the number of facilities. This
approach is likely to underestimate the emissions of some facilities while overestimating the emissions of others.
These quantitative uncertainty estimates capture only some of the uncertainties in the emissions estimates.
The sources of uncertainty in both the 1990-2010 estimates and the 2011-2022 estimates are described in detail
in the national Inventory. These sources of uncertainty also apply to the state estimates, and like the quantified
uncertainty estimates, are likely to have a larger impact on the uncertainties of the state-level estimates than on
the uncertainties of the national estimates.
3.2.11.4 Recalculations
This is a new category included for the current (i.e., 1990-2022) Inventory; thus, no recalculations were
performed.
3.2.11.5 Planned Improvements
EPA is planning to refine its estimates of emissions from facilities that do not report their emissions to
the GHGRP after confirming with the facilities that their actual per-facility uncontrolled emissions fall below
25,000 metric tons C02 Eq. EPA is also planning to refine its estimates of emissions for other facilities for
1990-2009 (e.g., by comparing these against emissions inferred from atmospheric measurements).
Moreover, EPA is continuing to seek data sets that can be used to improve and/or QA/QC emissions
estimates, particularly for the years 1990-2009. These data sets may include, for example, real-time facility-
specific estimates or additional global "top-down," atmosphere-based emissions estimates that could be
used to establish an upper limit on emissions of certain compounds.
3.2.11.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-52
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
Full citations of references included in Chapter 4.15 (Production of Fluorochemicals Other than HCFC-
22 [CRT Source Category 2B9b]) of the national Inventory are available online here:
https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-inventorv-2Q24-chapter-10-references O.pdf.
3.2.12 Phosphoric Acid Production (NIR Section 4.17)
3.2.12.1 Background
Phosphoric acid, or H3P04, is a basic raw material used in the production of phosphate-based fertilizers.
Phosphoric acid production from natural phosphate rock is a source of C02 emissions, due to the chemical
reaction of the inorganic carbon (calcium carbonate) component of the phosphate rock.. Emissions from
fuels consumed for energy purposes during the production of phosphoric acid are accounted for as part of
fossil fuel combustion in the industrial end-use sector reported under the Energy chapter. In 2022,
phosphoric acid was produced in Florida, Idaho, Louisiana, North Carolina, and Wyoming.
3.2.12.2 Methods/Approach
To develop state-level estimates of emissions from phosphoric acid production, EPA disaggregated
national emissions from the national Inventory to all applicable U.S. states using an Approach 2 method, as
defined in the Introduction chapter of this report, using a combination of process emissions reported to the
GHGRPfor 2010-2022 and estimated phosphoric acid production capacity by state for 1990-2009, as shown
in Table 3-11. The national Inventory methodology was adapted to calculate state-level GHG emissions from
phosphoric acid production to ensure consistency with national estimates. The sum of emissions by state
are consistent with national process emissions as reported in the national Inventory. See Appendix D, Tables
D-17 through D-22 in the "Phosphoric Acid" Tab, for more details on the data used.
Table 3-11. Summary of Approaches to Disaggregate the National Inventory for Phosphoric Acid
Production Across Time Series
Time Series Range
Summary of Method
2010-2022
GHGRP process emissions data were used to estimate the percentage of
emissions by state, multiplied by the national emissions (consistent with IPCC
2006 Tier 1).
1990-2009
Phosphoric acid production capacity data were used to estimate the percentage
of production by state, multiplied by the national emissions (consistent with
IPCC 2006Tier 1).
The methodology used for 2010-2022 used a combination of process emissions reported to the GHGRP
for each phosphoric acid facility and their assumed use of phosphate rock by origin. The GHGRP has no
reporting threshold for phosphoric acid production, so these emissions data are representative of the
industry. Consistent with national C02 emissions calculations in the national Inventory, state-level
emissions from phosphoric acid production were estimated using the C02 content and usage of three
categories of phosphate rock origin, where rocks sourced from each category were assumed to have
consistent C02 content: (1) Florida and North Carolina (FL/NC), (2) Idaho and Utah (ID/UT), and (3) Morocco
and Peru (imported).
Phosphoric acid production facilities operated in Florida, Idaho, Louisiana, Mississippi, North Carolina,
Texas, and Wyoming over the time series. As noted in the national Inventory, all phosphate rock mining
companies in the United States are vertically integrated, with fertilizer plants that produce phosphoric acid
located near the mines. Based on the location of mines, all phosphoric acid produced in Florida and North
Carolina was attributed to the FL/NC rock type, and the phosphoric acid produced in Idaho and Wyoming
was attributed to the ID/UT rock type. For production facilities in Louisiana, Mississippi, and Texas, USGS
3-53
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Minerals Yearbook information was used to assign the phosphate rock origin for each year from 1990-2022
(USGS 1994-2023). Where the USGS Minerals Yearbook did not discuss the rock origin for a facility in a given
year, EPA made assumptions regarding the rock origin based on information available in prior or subsequent
year publications. Because the rock usage by origin was not available for facilities, it was assumed that when
domestic phosphate rock and imported rock were both used at a facility, they were used in equal amounts
such that half of the plant capacity used each rock type. One facility in Louisiana was assumed to use half
FL/NC phosphate rock and half imported phosphate rock, whereas another was assumed to use only
imported rock. The facilities in Mississippi and Texas were assumed to only use imported phosphate rock.
For each of the three rock origin categories, the aggregated phosphoric acid production capacities for
each state were calculated and then used to allocate percentages of national emissions to each facility on
an annual basis. The estimated emissions from each facility for each rock type were then used to calculate a
percentage of emissions from each state for each rock type. That percentage was then applied to the
national Inventory emissions for each rock type per year to disaggregate national C02 emissions by state and
by year.
The methodology used for 1990-2009 attributes annual national phosphate rock usage to states based
on the production capacities of phosphoric acid production facilities and their assumed use of phosphate
rock by origin. Using location, estimated annual production capacity information, and operational status on
phosphoric acid production facilities for 1990-2005, EPA identified facilities operatingwet process
phosphoric acid production in each state (SRI International 1990-2005). For 2006-2009, EPA proxied using
2005 annual plant capacity information. Based on USGS Minerals Yearbook information on the operations of
each facility, the rock origins for each facility were identified on an annual basis. State-level emissions from
phosphoric acid production were estimated using the C02 content and usage of the same FL/NC, ID/UT, and
imported phosphate rock origin categories described above. For each of the three rock origin categories, the
aggregated phosphoric acid production capacities for each state were calculated and then used to allocate
percentages of national emissions to each state on an annual basis.
3.2.12.3 Uncertainty
The overall uncertainty associated with the 2020 national estimates of C02 from phosphoric acid
production was calculated using the 2006 IPCC Guidelines Approach 2 methodology for uncertainty (IPCC
2006). As described further in Chapter 4 and Annex 7 of the national Inventory (EPA 2024), levels of
uncertainty in the national estimates in 2022 were -18%/+20% for C02.
State-level estimates are expected to have an overall higher uncertainty because the national emissions
estimates were apportioned to each state based on a combination of GHGRP process emissions data for
2010-2021 and facility production capacity for 1990-2009. These assumptions were required because of a
general lack of more granular state-level data.
For 2010-2022, uncertainty is expected to be lower because GHGRP emissions data will be used by
state as a surrogate for using phosphoric acid production data by state to calculate emissions.
For 1990-2009, this allocation method does not address actual utilization or production rates, which
vary from facility to facility and from year to year. While this approach implicitly accounts for the size of a
facility in a state, it could overestimate emissions in states where facilities used less of their capacity and
underestimate emissions in states where facilities used more of their capacity as a result of the lack of data
on utilization rates and production data.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-54
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
3.2.12.4 Recalculations
The 2021 value for the total U.S. production of phosphate rock was updated based on updated USGS
data. These updates resulted in an overall decrease of 35 kt C02 in 2021 at the national level. State-level
changes include a 5% decrease for Florida (446.0 to 424.7 kt C02), a 5% decrease for Idaho (94.9 to 90.4 kt
C02), a 1% increase for Louisiana (149.1 to 150.7 kt C02), and a 5% decrease for North Carolina (161.8 to
154.1 kt C02).
3.2.12.5 Planned Improvements
For the facility-level phosphoric acid production capacity data used for 2006-2009, additional research
is needed to more accurately represent the level of production and emissions associated with each state.
EPA was able to locate the reference publication for the 1990-2005 time series but was not able to obtain the
2006-2009 publication before publishing this state-level inventory. Other data gaps include the origin of
phosphate rock used in some facilities and some years.
3.2.12.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
SRI International (1990-2005) Directory of Chemical Producers: United States of America.
USGS (U.S. Geological Survey) (1994-2023) Minerals Yearbook. Phosphate Rock Annual Report. Available
online at: https://www.usgs.gov/centers/national-minerals-information-center/phosphate-rock-
statistics-and-information.
3.3 Metals
This section presents the methodology used to estimate the metals portion of IPPU emissions, which
consist of the following sources:
Iron and steel production (C02, CH4)
Ferroalloy production (C02, CH4)
Aluminum production (C02, PFCs)
Magnesium production and processing (C02, HFCs, SFs)
Lead production (C02)
Zinc production (C02)
3.3.1 Iron & Steel Production and Metallurgical Coke Production (NIR Section 4.18)
3.3.1.1 Background
Iron and steel (l&S) production is a multistep process that generates process-related emissions of C02
and CH4 as raw materials are refined into iron and then transformed into crude steel. Emissions from
conventional fuels (e.g., natural gas, fuel oil) consumed for energy purposes (fuel combustion) duringthe
production of l&S are accounted for in the energy sector. I&S production includes seven distinct production
processes: metallurgical coke production, sinter production, direct reduced iron production, pellet
3-55
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
production, pig iron29 production, electric arc furnace (EAF) steel production, and BOF steel production. In
addition to the production processes, C02 is also generated at l&S mills through the consumption of process
byproducts (e.g., blast furnace gas, coke oven gas) used for various purposes, including heating, annealing,
and generating electricity. In general, C02 emissions are generated in these production processes through
the reduction and consumption of various carbon-containing inputs (e.g., ore, scrap, flux, coke byproducts).
Fugitive CH4 emissions can also be generated from these processes, as well as from sinter, direct iron, and
pellet production.
In 2022, l&S production occurred in 37 states, with seven states accounting for roughly 62% of total raw
steel production: Indiana, Alabama, Tennessee, Kentucky, Mississippi, Arkansas, and Ohio (AISI 2023).
3.3.1.2 Methods/Approach
To compile emissions by state from l&S and metallurgical coke production using available data,
national emissions were disaggregated from the national Inventory with an Approach 2 method as defined in
the Introduction chapter of this report, using a combination of coking coal consumption data, process
emissions reported to the GHGRP, and data on steel production and employment as a surrogate for steel
production data. The sum of emissions by state is consistent with the national total process emissions
reported in the national Inventory. See Appendix H, Tables H-1 through H-4 in the "l&S" Tab, for more details
on the data used.
The national Inventory methodology was adapted to calculate state-level GHG emissions to ensure
consistency with national estimates, which were downscaled across states because of limitations in the
availability of state-specific data across the time series to use national methods at the state level (i.e., IPCC
Tier 1 and 2 methods).
The emissions from l&S and metallurgical coke production were broken into the following categories for
national emissions calculations in the national Inventory and also as part of the state-level breakout:
Metallurgical coke production
Steel productionBOF
Steel productionEAF
Sinter production
Iron production
Pellet production
Other activities
The methodologies for calculating state emissions from each category are detailed below.
3.3.1.2.1. Metallurgical Coke Production
National emissions from metallurgical coke production used for l&S are estimated based on the amount
of coke used in l&S and a carbon balance around the amount of coking coal used to produce the coke, while
accounting for any coproducts produced. Specific state-level data on coke production for l&S are not readily
available; however, state-level data on coking coal consumption are available from ElA's SEDS. Those data
29 "Pig iron" is the common industry term to describe what should technically be called crude iron. Pig iron is a subset of
crude iron that has lost popularity overtime as industry trends have shifted. Throughout this report and consistent with the
national Inventory, "pig iron" will be used interchangeably with "crude iron," but it should be noted that other data sets or
reports may not use "pig iron" and "crude iron" interchangeably and may provide different values for the two.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-56
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
are broken out byfueltype and energy consumption sector (i.e., residential, commercial, industrial,
transportation, and electric power) and available for 1960-2021 (EIA 2023). Energy consumption estimates
from SEDS use data from surveys of energy suppliers that report consumption, sales, or distribution of
energy at the state level, and most SEDS estimates rely directly on collected state-level consumption data.
The sums of the state estimates equal the national totals as closely as possible for each energy type and
end-use sector, and energy consumption estimates are generally comparable to national energy statistics.
National-level metallurgical coke production emissions from l&S were allocated to the state level based on
the percentage of total coking coal consumed per state. This approach assumes that emissions from
metallurgical coke production are directly proportional to the amount of coking coal consumed in a state. As
discussed in the Energy chapter, state-level coking coal use is based on coke production in a given state,
which is not necessarily equal to coke use. Given the lack of specific data, however, coking coal production
was determined to be a reasonable surrogate for coke use within a given state because coke production is
often integrated with l&S production where the coke is used.
3.3.1.2.2. Steel Production
National emissions from steel production (BOF and EAF) were estimated based on a carbon balance
around carbon-containing inputs and outputs. State-level data on all the process inputs and outputs were
not readily available; therefore, surrogate data on steel production by state were used to allocate national-
level steel production emissions to the state level.
For 2010-2022, process emissions reported to the GHGRP under Subpart Q (l&S facilities) were
summed by state (EPA 2024a) to calculate a percentage of emissions from each state. Fuel combustion
emissions from l&S facilities reporting to the GHGRP are reported separately under Subpart C (combustion
units). Generally, fuel combustion emissions are reported under the energy portion of the national Inventory;
however, some of these emissions were included in l&S national Inventory calculations, specifically blast
furnace emissions. Portions of fuel consumption data for several fuel categories were included in the IPPU
calculations (e.g., l&S) because they are consumed during nonenergy-related industrial process activity. A
consistent approach to avoid double counting emissions from l&S was taken for state-level emissions,
subtracting state-level l&S process emissions from each state's energy sector emissions. More information
on this allocation process is available in the Energy chapter of this report.
A combination of Subpart Q and Subpart C data were used when estimating state emissions
percentages from l&S facilities in 2010-2022. Because emissions are reported by unit type in the GHGRP,
EPA was able to disaggregate state-level emissions at the process level, including steel production by type,
iron, sinter, pellet, metallurgical coke, and other activities. For steel production, GHGRP data were available
by process type for BOF and EAF. The percentage of total emissions by steel type per state from the GHGRP
data was then applied to the national emissions of steel production by type from the national Inventory per
year to calculate disaggregated C02 emissions by state.
GHGRP has a reporting threshold of 25,000 metric tons of C02 equivalent for l&S production, so these
emissions data are representative of the larger facilities in the industry. Using GHGRP emissions data means
that emissions from states with smaller facilities were possibly underestimated.
For the years 1990-2009, a combination of employment data from the U.S. Census and production data
from the American Iron and Steel Institute (AISI) was used to allocate national emissions from steel
production to states (U.S. Census Bureau 1992,1997, 2002, 2007; AIS11997-2021). AISI total steel
production data were available at the state level for the top five l&S-producing states) for each year, and data
for the other states were combined into regions. Percentages of steel production for these lower producing
states were approximated using U.S. Census Bureau industry employment data. It was assumed steel
production was directly proportional to the number of employees in the state.
3-57
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Census data were available for the years 1992,1997, 2002, and 2007. Data for the years 1990 and 1991
were proxied based on 1992, and data for the years 2008 and 2009 were proxied based on 2007. Data for
interim years were interpolated. For 1992, data were pulled by state for the NAICS codes Subsector 331:
Primary Metal Manufacturing and Subsector 332: Fabricated Metal Product Manufacturing. For 1997, 2002,
and 2007, state data were pulled for NAICS codes 331111 Iron and Steel Mills and Ferroalloy Manufacturing,
331210 Iron and Steel Pipe and Tube Manufacturing from Purchased Steel, 331221 Rolled Steel Shape
Manufacturing, 331222 Steel Wire Drawing, 331511 Iron Foundries, 331512 Steel Investment Foundries,
331513 Steel Foundries (except Investment), and 332111 Iron and Steel Forging. For some states, the NAICS
code had a low number of employees or low number of facilities to the point where it was not reported
because of anonymity concerns; therefore, these states were excluded from this analysis. For some cases,
states were included if data were available at a higher NAICS code. One exception was Maryland, where data
were withheld to maintain anonymity, but the state is known to have had sizable steel production; it was
assumed Maryland had 2,000 employees in the steel sector in the latest year of Census data (2007).30 The
percentage of employees and steel production across the region aggregated with Maryland in the AISI data
(Rhode Island, Connecticut, New Jersey, New York, Delaware, and Maryland) based on the 2007 data were
applied across the entire time series.
Furthermore, steel production by state was broken out into BOF and EAF steel production based on the
national totals of each type of steel produced from AISI data. Steel production in each state by type was
assumed to be proportional to the national totals by type for each year. Once data on steel production by
type were determined for each state and year, the total national emissions by steel type was attributed to
each state based on steel production in each state. This approach assumes that emissions from steel
production are directly proportional to the amount of steel produced in a state. This assumption could lead
to overestimations or underestimations of emissions per state depending on the type of steel production and
relative emissions profile of steel production in a given state. Furthermore, basing the state-level split of BOF
and EAF on the national averages could lead to overestimation or underestimation of a specific type of steel
production in a given state. Given the lack of data, this approach is considered reasonable. However, this is
an area for future improvement based on consideration of any available state-level steel production data.
3.3.1.2.3. Sinter Production, Iron Production, Pellet Production, and Other Activities
For 2010-2022, emissions from sinter production, iron production, pellet production, and other
activities were allocated based on the GHGRP data for the process types. The GHGRP reporting threshold of
25,000 metric tons of C02 equivalent for l&S production is applicable for these process types as well.
For 1990-2009, emissions from sinter production, iron production, pellet production, and other
activities were allocated to states based on the percentage of BOF steel production by state from U.S.
Census employment data and AISI production data (U.S. Census Bureau 1992,1997, 2002, 2007; AIS11997-
2021), as described above. It was assumed that emissions from sinter production, iron production, pellet
production, and other activities would be most closely aligned with BOF steel production.
3.3.1.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02 and CH4 from l&S production
was calculated using the 2006 IPCC Guidelines Approach 2 methodology for uncertainty (IPCC 2006). As
described further in Chapter 4 and Annex 7 of the national Inventory (EPA 2024b), levels of uncertainty in the
national estimates in 2022 were -16%/+16% for C02 and -7%/+7% for CH4.
30 Based on https://millstories.umbc.edu/sparrows-point/.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-58
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
State-Level estimates are expected to have a higher uncertainty because the national emissions
estimates were apportioned to each state based on a combination of coking coal consumption data and
process emissions reported to GHGRP. These assumptions were required because of a general lack of more
granular state-level data.
Emissions from metallurgical coke production for l&S were assumed to be directly proportional to the
amount of coking coal consumed in a state, and metallurgical coke was assumed to be used in the same
state it was produced. While industry trends suggest mostly on-site use, this method could overestimate
emissions from coking coal for states where facilities transfer coking coal off-site and underestimate
emissions for states where facilities transfer coking coal for metallurgical coke production across state
boundaries.
For 2010-2022, GHGRP data were used to disaggregate national Inventory emissions to the state level
for steel, sinter, iron, pellet, and other activities. Because GHGRP receives detailed data down to the process
unit level, uncertainty is lower. While the GHGRP data have a reporting threshold of 25,000 metric tons of
C02 equivalent, GHGRP estimates that 99.8% of industry emissions are accounted for (EPA 2009), and the
GHGRP data are likely representative of the whole industry.
For 1990-2009, U.S. Census data were used as a surrogate for production data for steel, sinter, iron,
pellet, and other activities to disaggregate national Inventory data by state. Because this method assumes
that all facilities produce the same amount of emissions regardless of production capacities, it could
overestimate emissions in states with smaller facilities and underestimate emissions in states with larger
facilities. Additionally, for sinter, iron, pellet, and other activities, emissions are based on BOF steel
production for the state, which may overestimate or underestimate state-level emissions for these activities.
Byproduct fuels are assumed to be used on-site in this method. Although industry trends show facilities
using byproduct fuels such as coke oven gas or blast furnace gas on-site, if these byproducts are shipped off-
site, this adds an additional level of uncertainty to state-level estimates. If these byproducts are shipped
across state lines for energy use, emissions may be overestimated for states where facilities transfer
byproducts off-site and across state boundaries and underestimated for states where facilities use
byproducts on-site from across state boundaries.
3.3.1.4 Recalculations
Recalculations in the national Inventory were performed for the year 2021 using updated USGS values for DRI, pig
iron, and scrap steel consumption for both BOF and EAF steel production. Additionally, revisions to GHGRP data for
2020 and 2021 resulted in minor changes to activity data that were adjusted using GHGRP data. Compared to the
previous Inventory, CO2 emissions from steel production increased by less than 1% (11 kt CO2) in 2020 and by less
than 1% (216 kt CO2) in 2021. The largest changes in emissions by state occurred in Alabama, which saw a
13.7% increase in C02 emissions from steel production. Finally, the heat content of coal was updated from
23.89 million Btu/ton to 23.91 million Btu/ton in the national Inventory, which resulted in a minor increase in
C02 emissions from pig iron production.
3.3.1.5 Planned Improvements
AISI production data were only available for the years 1997-2020 (AIS11997-2021), so data are
incomplete for earlier years of the time series. This is an area for future improvement based on consideration
of any available state-level production data.
Census employment data are released every five years, and employment estimates were based on
NAICS codes. The NAICS codes used might not encompass the whole industry, and generally as a method,
3-59
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
the number of employees may not correlate well to emissions. One area of future improvement is to better
understand the completeness of employment data and make adjustments as necessary.
Combustion emissions from GHGRP data are not entirely consistent across reporting facilities because
some facilities report under Subpart C and some report combined emissions using CEMS. Also, fuel use data
from the GHGRP might not be equivalent to data included in the national Inventory calculations under l&S
because the GHGRP data do not specifically indicate if fuel is used in nonenergy applications. One area of
future improvement is to examine the GHGRP energy use estimates in comparison to what is assumed in the
national Inventory calculations and adjust as needed.
EPA plans to compare coking coal consumption data from EIA SEDS to the data from the GHGRP
reporting program for the years 2010-2022 as a QA/QC check.
EPA also plans to compare BOF and EAF data by state from the GHGRP to the AISI national percentage
breakout of EAF and BOF by state to see if there is a better approach to allocating BOF and EAF production by
state for 1990-2009. In general, EPA plans to compare the industry data to the GHGRP program data across
time to see how close they are and if using the industry data is a reasonable approach.
EPA will review time series consistency issues related particularly to steel production. Surrogate data
on industry employment were used in place of activity data for all but the top five producing states for the
1990-2009 portion of the time series, and more research will be undertaken to identify potential
methodological refinements to enhance the accuracy and consistency of estimated state GHG emissions
and trends.
3.3.1.6 References
AISI (American Iron and Steel Institute) (1997-2021, 2023) Annual Statistical Report.
EIA (U.S. Energy Information Administration) (2023) State Energy Data System (SEDS): 1960-2021
(Complete). U.S. Department of Energy. Accessed May 2024. Available online at:
https://www.eia.gov/state/seds/seds-data-complete.php.
EPA (U.S. Environmental Protection Agency) (2009) Technical Support Document for the Iron and Steel
Sector: Proposed Rule for Mandatory Reporting of Greenhouse Gases. Available online at:
https://www.epa.gov/sites/default/files/2015-02/documents/tsd iron and steel epa 9-8-08.pdf.
EPA (2024a) Envirofacts GHGRP Subpart Q, Subpart C, and Common Data. Accessed May 13, 2024. Available
online at: https://enviro.epa.gov/query-builder/ghg.
EPA (2024b) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2022. EPA430-R-24-004.
Available online at: https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
U.S. Census Bureau (1992,1997, 2002, 2007) Geographic Area Series, Manufacturing. Available online at:
https://www.census.gov/librarv/publications/1995/econ/mc92-a.html (1992),
https://www.census.gov/library/publications/1997/econ/census/manufacturing-reports.html
(1997), https://www.census.gov/library/publications/2002/econ/census/manufacturing-reports.html
(2002), https://www.census.gov/data/tables/2007/econ/census/manufacturing-reports.html (2007).
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-60
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
3.3.2 Ferroalloys Production (NIR Section 4.19)
3.3.2.1 Background
C02 and CH4 are emitted from the production of several ferroalloys. Ferroalloys are composites of iron
and other elements such as silicon, manganese, and chromium. Emissions from fuels consumed for energy
purposes during the production of ferroalloys are accounted for in the energy sector. Emissions from the
production of two types of ferrosilicon (25% to 55% and 56% to 95% silicon by mass), silicon metal (96% to
99% silicon by mass), and miscellaneous alloys (32% to 65% silicon by mass) have been calculated.
Consistent with the national Inventory, emissions from the production of ferrochromium and
ferromanganese are not included because of the small number of manufacturers of these materials in the
United States. Government information disclosure rules prevent the publication of production data for these
production facilities. Additionally, production of ferrochromium in the United States ceased in 2009.
Similar to emissions from the production of l&S, C02 is emitted when metallurgical coke is oxidized
during a high-temperature reaction with iron and the selected alloying element. Although most of the carbon
contained in the process materials is released to the atmosphere as C02, a percentage is also released as
CH4 and other volatiles. The amount of CH4 that is released depends on furnace efficiency, operation
technique, and control technology.
In 2022, ferroalloy production occurred in six states: Ohio, Kentucky, Pennsylvania, Alabama, West
Virginia, and Michigan.
3.3.2.2 Methods/Approach
To compile emissions by state from ferroalloy production, the state-level inventory disaggregated
national emissions from the national Inventory with an Approach 2 method as defined in the Introduction
chapter of this report, using a combination of process emissions reported to the GHGRP and the number of
facilities in a state (see Table 3-12). See Appendix H, Tables H-5 and H-6 in the "Ferroalloy" Tab, for more
details on the data used.
The national Inventory methodology was adapted to calculate state-level GHG emissions to ensure
consistency with national estimates. National estimates were downscaled across states because of
limitations in the availability of state-specific data across the time series to use national methods (i.e., IPCC
Tier 1 methods) at the state level. The sum of emissions by state is consistent with the national process
emissions reported in the national Inventory.
Table 3-12. Summary of Approaches to Disaggregate the National Inventory for Ferroalloys
Production Across Time Series
Time Series Range
Summary of Method
2010-2022
GHGRP facility process emissions data were used.
Remaining emissions reported in the national Inventory were allocated evenly
across remaining known facilities (IPCC 2006 Tier 1).
1990-2009
Data on number of facilities that reported to the GHGRP were used to allocate
emissions for those facilities.
Remaining emissions reported in the national Inventory were allocated evenly
across remaining known facilities (IPCC 2006 Tier 1).
To identify all ferroalloy-producing facilities for 1990-2022, the number of facilities in each state was
compiled from the USGS Minerals Yearbooks for ferroalloys as available (USGS 2008-2018) and compared
with the facilities reporting to the GHGRP. The GHGRP has a reporting threshold of 25,000 metric tons of C02
3-61
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
equivalent for ferroalloy production, so these emissions data are representative of the larger facilities in the
industry. Combining GHGRP emissions data with the number of facilities in each state includes smaller
facilities and improves the completeness of the state-level inventory. The total number of facilities from the
2008 USGS Minerals Yearbook for ferroalloys was used for the years 1990-2007 because the Minerals
Yearbooks for years before 2008 did not contain the number of facilities. Additionally, facilities were not
included in years that EPA determined the facility was not operational. EPA used internet searches to
determine the opening dates of ferroalloys facilities and to determine whether they were operational during
all inventory years (AMG Vanadium 2017; Bloomberg 2021a, 2021b; Businesswire 2020, 2017; Centerra Gold
2021; Flessner 2015; D&B 2021; Ferroglobe 2020; Global Titanium Inc. 2010; RTI International Metals 2007;
Vanadium Price 2019).
Five of the facilities listed in the USGS Minerals Yearbook also reported to the GHGRP in 2010-2022, and
the reported process emissions data were used for these facilities. To improve the completeness of this
state-level inventory and estimate emissions from the remaining known facilities in 2010-2022, process
emissions reported to the GHGRP were summed (EPA 2010-2022) for each year and subtracted from the
national Inventory total emissions for each year. The remaining balance was distributed equally among the
facilities listed in the USGS Minerals Yearbook that did not report to the GHGRP.
For 1990-2009, the average GHGRP emissions from each GHGRP facility for the years 2010-2012 were
applied to each year, and the remaining emissions were evenly distributed among the remaining facilities.
Values for the years 2010-2012 were used because these were expected to be a more accurate
representation of emissions in 1990-2009.
Once facility-level emissions were calculated, the emissions were summed by state to calculate C02
and CH4 emissions by state for each year.
3.3.2.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02 and CH4 from ferroalloy
production was calculated using the 2006 IPCC Guidelines Approach 2 methodology for uncertainty (IPCC
2006). As described further in Chapter 4 and Annex 7 of the national Inventory (EPA 2024), levels of
uncertainty in the national estimates in 2022 were -13%/+13% for C02 and -12%/+13% for CH4.
State-level estimates are expected to have a higher uncertainty because the national emissions
estimates were apportioned to each state based on process emissions reported to the GHGRP and the
number of facilities in a state. These assumptions were required because of a general lack of more granular
state-level data.
For 2010-2022, this allocation method relies partially on GHGRP emissions data, which have a lower
uncertainty for states where those reporting facilities are located but have a higher uncertainty for states
where smaller facilities that did not report to the GHGRP are located. This method could underestimate
emissions from larger facilities and overestimate emissions from smaller facilities.
For 1990-2009, this allocation method does not fully address facilities' production capacities or
utilization rates, which vary from facility to facility and from year to year. Because this approach implicitly
assumes that emissions from facilities that did not report to the GHGRP are equal regardless of production
capacity or utilization rates and that facilities that did report to the GHGRP had the same annual emissions
levels for these years, this approach could overestimate emissions in some states and underestimate
emissions in others.
Emissions for ferromanganese and ferrochromium are not included in the national Inventory estimate
because of the small number of manufacturers in the United States. The facilities producing these
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-62
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
ferroalloys, however, are included in the state Inventory disaggregation; thus, state-level estimates are likely
an underestimate.
3.3.2.4 Recalculations
No recalculations were performed for the 1990-2021 portion of the time series.
3.3.2.5 Planned Improvements
There are significant differences between USGS and GHGRP data regarding which facilities are included
in the ferroalloys industry. Six facilities reported to the GHGRP but were not listed by USGS, and six facilities
were listed by USGS but did not report to the GHGRP. The GHGRP has a reporting threshold for ferroalloys
production, which may contribute to the difference in the latter group of facilities. Clarifying why this
discrepancy exists would improve inventory data accuracy both at the national and disaggregated state
levels.
Because USGS does not list ferroalloy production at the state level, EPA estimated that all facilities that
did not report to the GHGRP produced equal emissions. Data on the size and capacity of each facility would
allow EPA to distribute emissions more accurately. As a future improvement, EPA may use Title V or state-
level permits to look for capacity data for each facility to better estimate emissions by state.
While production of ferrochromium in the United States ceased in 2009, EPA will assess whether data
are available to incorporate emissions from facilities producing ferromanganese and ferrochromium in the
national- and state-level inventories over the time series.
3.3.2.6 References
AMG Vanadium (2017) Our History. Available online at: https://amg-v.com/timeline amg v/.
Bloomberg (2021 a) Bear Metallurgical Co. Available online at:
https://www.bloomberg.eom/profile/companv/0589837D:US.
Bloomberg (2021 b) Eramet Marietta Inc. Available online at:
https://www.bloomberg.eom/profile/companv/0205877D:US.
Businesswire (2017) Felman Production Reports on Temporary Shut Down of Its New Haven, IV. Va. Facility.
July 25, 2017. Available online at:
https://www.businesswire.com/news/home/20170725006161/en/Felman-Production-Reports-on-
Temporarv-Shut-Down-of-lts-New-Haven-W.-Va.-Facility.
Businesswire (2020) CC Metals and Alloys, LLC Is Shutting Down Its Operations on July 1 Due to Poor Market
Conditions. June 24, 2020. Available online at:
https://www.businesswire.eom/news/home/20200624005217/en/CC-Metals-and-Allovs-LLC-is-Shutting-
Down-its-Operations-on-Julv-l-Due-to-Poor-Market-Conditions.
Centerra Gold (2021) Molybdenum Business Unit: Langeloth Metallurgical Facility. Available online at:
https://www.centerragold.com/operations/molvbdenum-business-unit/.
D&B (2021) Reading Alloys, Inc. Accessed July 15, 2021. Available online at: https://www.dnb.com/business-
directorv/companv-profiles.reading alloys Ilc.21d7f6ff83866c7a6c9580cl45fb46e5.html.
EPA (U.S. Environmental Protection Agency) (2010-2022) Envirofacts GHGRP Subpart Kand Subpart C Data.
Accessed May 13, 2024. Available online at: https://www.epa.gov/enviro/greenhouse-gas-customized-
search.
3-63
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
EPA (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2022. EPA430-R-24-004. Available
online at: https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
Ferroglobe (2020) Beverly: History. Available online at: https://www.ferroglobe.com/about-
ferroglobe/industrial-footprint/beverlv.
Flessner, D. (2015) Focus on the Worker: Lesson Learned in Bridgeport Aids Growing Silicon Industry.
Chattanooga Times Free Press. August 16, 2015. Available online at:
https://www.timesfreepress.com/news/business/aroundregion/storv/2015/aug/16/lessons-learned-
bridgeport-inform-merger/319744/?bcsubid=81b8962b-36f8-4876-9b5f-77e1a0bbf54b&pbdialog=reg-
wall-login-created-tfp.
Global Titanium Inc. (2010) History. Accessed July 15, 2021. Available online at:
https://web.archive.Org/web/20221228055337/http://www.globaltitanium.net/historv.htm.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
RTI International Metals, Inc. (2007) Form 10-K: Annual Report Pursuant to Section 13 or 15(D) of the
Securities Exchange Act of 1934 for the Fiscal Year Ended December 31,2006. Accessed May 10, 2023.
Available online at:
https://www.sec.gov/Archives/edgar/data/1068717/000095015207001633/l24102aelQvk.htm.
USGS (U.S. Geological Survey) (2008-2018) Minerals Yearbook: Ferroalloys Annual Report. Available online
at: https://www.usgs.gov/centers/national-minerals-information-center/ferroallovs-statistics-and-
information.
Vanadium Price (2019) US Vanadium LLC Announces Agreement to Acquire Evraz Stratcor, Inc. August 12,
2019. Available online at: https://www.vanadiumprice.com/us-vanadium-llc-announces-agreement-to-
acquire-evraz-stratcor-inc/.
3.3.3 Aluminum Production (NIR Section 4.20)
3.3.3.1 Background
The production of primary aluminumin addition to consuming large quantities of electricityresults in
process-related emissions of C02 and two perfluorocarbons: perfluoromethane (CF4) and perfluoroethane
(C2F6). Aluminum Production occurs or has occurred in the past in the following 14 states: Indiana, Kentucky,
Maryland, Missouri, Montana, North Carolina, New York, Ohio, Oregon, South Carolina, Tennessee, Texas,
Washington, and West Virginia.
C02 is emitted during the aluminum smelting process when alumina (aluminum oxide, Al203) is reduced
to aluminum using the Hall-Heroult reduction process. The reduction of the alumina occurs through
electrolysis in a molten bath of natural or synthetic cryolite (Na3AlF6). The reduction cells contain a carbon
liningthat serves as the cathode. Carbon is also contained in the anode, which can be a carbon mass of
paste, coke briquettes, or prebaked carbon blocks from petroleum coke. During reduction, most of this
carbon is oxidized and released to the atmosphere as C02.
In addition to C02 emissions, the aluminum production industry is also a source of PFC emissions.
During the smelting process, when the alumina ore content of the electrolytic bath falls below critical levels
required for electrolysis, rapid voltage increases occur, which are termed high-voltage anode effects (HVAEs)
HVAEs cause carbon from the anode and fluorine from the dissociated molten cryolite bath to combine,
thereby producing fugitive emissions of CF4 and C2F6. In general, the magnitude of emissions for a given
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-64
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
smelter and Level of production depends on the frequency and duration of these anode effects. As the
frequency and duration of the anode effects increase, emissions increase. Another type of anode effect, Low-
voltage anode effects (LVAEs), became a concern in the early 2010s as the aluminum industry increasingly
began to use cell technologies with higher amperage and additional anodes (IPCC 2019). LVAEs emit CF4 and
are included in PFC emissions totals from 2006 forward.
3.3.3.2 Methods/Approach
National emissions of C02 and PFCs from aluminum production are estimated using a combination of
IPCC Tier 1, Tier 2 and Tier 3 methods (i.e., EPA GHGRP data) over the time series as discussed in Chapter 4,
Section 4.20 (on pages 4-121 through 4-127) of the national Inventory. IPCC Tier 1 methods were used only to
estimate PFC emissions from LVAEs.
Aluminum production emissions calculated nationally were allocated to the state level using a Hybrid
approach due to lack of facility-level and/or state-level production data for earlier years of the time series.
For 2010 and later, EPA used the same underlying methods that were used for the national Inventory (i.e.,
facility-specific process emissions reported to EPA's GHGRP under subpart F: Aluminum Production were
used to estimate state-level emissions); for 1990-2009, EPA used the ratio of each state's smelter capacity
to the U.S. total capacity to allocate national emissions to each state. The approach summarized in Table
3-13 was taken to compile aluminum production estimates by state consistent with national totals.
Table 3-13. Summary of Approaches to Disaggregate the National Inventory for Aluminum
Production Across Time Series
Time Series Range
Summary of Method
2010-2022
GHGRP process emissions data were used to get emissions by state (i.e.,
Approach 1).
1990-2009
Data on smelter capacity were used to get percentage of production by state,
which was then multiplied by national emissions (Approach 2).
For 2010-2022, EPA used facility-specific emissions reported to the GHGRP and facility locations to
allocate estimated emissions to each state. All aluminum production facilities in the United States report
their emissions to EPA. CF4 emissions from LVAEs were estimated by allocating total U.S. LVAE emissions
according to each state's yearly percentage of total HVAE CF4 emissions. The percentages were calculated on
a yearly basis (state total/yearly total) to account for non-reporting years.
For 1990-2009, EPA allocated national totals to each state using the ratio of each state's smelter capacity to
the U.S. total capacity, on a yearly basis (i.e., state X emissions = national emissions x [ratio = state X smelter
capacity/national smelter capacity]). Capacity data for the years 1990,1993, 2001, and 2004-2009 were collected
from the respective years' USGS Aluminum yearbook, and capacities for other years were interpolated from the
aforementioned USGS Aluminum yearbooks' capacity data trends (USGS 1996-2022). Information on idle facilities
and shutdowns was incorporated in determining state smelter capacities based on USGS Aluminum yearbook
notes and additional sources (including public articles and expert reviewers' feedback). National emissions during
this time period were developed using smelter capacity data and the USAA U.S. primary aluminum production
estimates (USAA 2020), combined with the process emissions and activity data reported under EPA's Voluntary
Aluminum Industrial Partnership Program (VAIP). Facilities under the parent company Alcoa had certain
production data aggregated within the 1990-2009 time series; these data were allocated by building
percentage assumptions based on all the data and information described above.
3-65
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
3.3.3.3 Uncertainty
The overall uncertainties associated with the 2022 national estimates of C02 and PFC emissions from
Aluminum production were calculated using the 2019 Refinement to the 2006IPCC Guidelines. As described
further in Chapter 4 of the national Inventory, levels of uncertainty in the national estimates in 2022
surrounding the reported C02, CF4, and C2F6 emission values were determined to have a normal distribution
with uncertainty ranges of approximately 3% below to 3% above, 8% below to 8% above, and 9% below to 9%
above their 2020 emission estimates, respectively.
For the 2010 to 2022 time series, the uncertainties associated with the state-level estimates are
expected to be lower than those for the 1990-2009 time series because emissions are estimated and
reported at the facility level. Nevertheless, the 2010 to 2022 state-level uncertainties are somewhat higher
than 2010 to 2022 national-level uncertainties because, for each gas, the uncertainty of each smelter's
emissions is higher than the uncertainty of the emissions across all smelters.31 The uncertainty of each
smelter's C02 emissions is estimated at -/+6%; the uncertainty of each smelter's HVAE CF4 emissions is
estimated to range from -/+16%; and the uncertainty of each smelter's HVAE C2F6 emissions is estimated to
range from -/+20%. The uncertainty associated with LVAE emissions is estimated based on the smelter
technology type and is estimated to range from -/+99% for each smelter. Because LVAE emissions make up a
small share of total PFC emissions, this uncertainty does not have a large impact on the overall uncertainty
of PFC emissions at either the smelter or the US level. For more details on national-level uncertainty, see the
Uncertainty discussion in Chapter 4 of the national Inventory.
State-level estimates are expected to have significantly higher uncertainties for 1990-2009 than more
recent years due to the methods used to apportion the national emission estimates to each state based on
the capacity data from the USGS Aluminum yearbooks. This approach does not reflect the volatility in actual
aluminum production activities in each smelter (and thus in the different states) from year to year, and the
estimated emissions in each state may therefore differ from the actual emissions resulting from aluminum
production activities in that state.
3.3.3.4 Recalculations
Refer to Section 4.20 (page 4-127) of the national Inventory report (EPA 2024) for a complete list of
recalculations for the national Inventory.
3.3.3.5 Planned Improvements
EPA identified a potential refinement in the approach used to compile annual state estimates over
1990-2009. The refinement would allocate emissions based on emissions data reported under EPA's VAIP.
Where facility-specific data are not reported under VAIP, additional data, including technology type and
estimated production, could be used to allocate data to the states from the VAIP data.
EPA will further investigate the sources of historical total primary aluminum production estimates for the
earlier years in the time series and potentially update historical estimates to aim for increased consistency
throughout the time series. As part of this planned improvement, EPA will review whether historical estimates are
broken down into smelter specific production estimates, which are the basis for calculating smelter, and therefore
state, PFC (for non-partners) and CO2 emissions (for all facilities) for the 1990 through 2009 time series (years
preceding GHGRP reporting). Additional improvements include evaluating the LVAE emissions calculations method
by state for the 2010-2022 time series. Currently, the LVAE CF4 emissions are based on each state's yearly
31 Note that this holds true generally for the sum of variables with independent errors: the error of the sum tends to be lower
than the error of each variable.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-66
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
percentage of total HVAE CF4 emissions. Future iterations of the state disaggregation estimates of LVAE CF4
emissions will be based on estimates of aluminum production, consistent with the Tier 1 LVAE method and the
national Inventory.
3.3.3.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventory-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2019)2079 Refinement to the 2006IPCC Guidelines for
National Greenhouse Gas Inventories. E.C. Buendia, K. Tanabe, A. Kranjc, J. Baasansuren, M. Fukuda, S.
Ngarize A. Osako, Y. Pyrozhenko, P. Shermanau, and S. Federici (eds.). Available online at:
https://www.ipcc.ch/report/2019-refinement-to-the-2006-ipcc-guidelines-for-national-greenhouse-gas-
inventories/.
USAA (U.S. Aluminum Association). (2020). U.S. Primary Aluminum Production: Report for August 2020.
USGS (U.S. Geological Survey). (1996-2022). Minerals Yearbook: Aluminum.
3.3.4 Magnesium Production and Processing (NIR Section 4.21)
3.3.4.1 Background
The magnesium metal production and casting industry uses sulfur hexafluoride (SFs) and other
greenhouse gases (i.e., HFC-134a and Novec 612) to prevent the rapid oxidation of molten magnesium in the
presence of air. A dilute gaseous mixture of these gases with dry air and/or C02 is blown over molten
magnesium metal to induce and stabilize the formation of a protective crust. A small portion of the cover gas
reacts with the magnesium to form a thin molecular film of mostly magnesium oxide and magnesium
fluoride. The amount of cover gas reacting in magnesium production and processing is considered to be
negligible; thus, all cover gas used is assumed to be emitted into the atmosphere. Magnesium production
occurs or has occurred previously in the following states: California, Illinois, Indiana, Michigan, Minnesota,
Missouri, Ohio, Tennessee, Utah, and Washington.
3.3.4.2 Methods/Approach
National emissions of SFs, HFC-134a, Novec 612, and C02 from magnesium production and processing
are estimated using a combination of IPCC Tier 2 and Tier 3 methods over the time series as discussed in
Chapter 4, Section 4.21 (on pages 4-127 through 4-133) of the national Inventory (EPA 2024).
National magnesium processing and production emissions were allocated to the state level using a
Hybrid approach due to a lack of facility-level data for some years and for some facilities. For 2011-2022,
EPA used facility-specific emissions data from its GHGRP for primary and secondary production, die casting,
and sand casting. For these same years estimates of national emissions from permanent mold, wrought, and
anode production were allocated to the state level based on state emissions percentages developed using
data reported to the GHGRP. No producers of permanent mold, wrought, and anode magnesium products
report to the GHGRP. EPA assumed that non-reporting facilities were located in the same states as reporting
facilities.
For 1999-2010, EPA used company-specific reported cover gas emissions data reported to EPA through
the SFs Emission Reduction Partnership for the Magnesium Industry to both allocate emissions to the states
and process types with reporting partner companies, as well as derive a percentage of emissions by state.
These percentages by state were applied to the remaining non-Partner emissions such that the full
complement of national magnesium emission could be apportioned to the state level, similar to the
3-67
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
approach used for Later years when GHGRP data became available. For 1990-1998, where GHGRP and
Partnership data are not available, a simplified assumption of national to state-level apportionment based on
1999 data was used to estimate emissions from all magnesium production and processes.
Table 3-14 provides additional specifics on the approaches taken to compile state-level estimates of
emissions for magnesium production consistent with national totals.
Table 3-14. Summary of Approaches to Disaggregate the National Inventory for Magnesium
Production Across Time Series
Time Series
Range
Summary of Method
2011-2022
For primary, secondary, die casting, and sand casting, emissions were allocated
by facility locations based on information reported to the GHGRP (Approach 1).
For permanent, wrought, and anode, emissions were allocated proportionally to
states with reported emissions (Approach 2).
1999-2010
For primary, secondary, die casting, and sand casting, emissions were allocated
by company and facility locations based on cover gas usage reported to the EPA
Partnership Program (Approach 1).
For permanent, wrought, and anode, emissions were allocated proportionally to
states with reported emissions for secondary, die casting, and sand casting,
excluding the primary production company (Approach 2).
1990-1998
Percentage of emissions by state and process type in 1999 was used to allocate
national emissions across states from 1990 to 1998 and included all process
types (Approach 2; please refer to the national Inventory for more details).
3.3.4.2.1. All Processes
The methodology used for all process for 1990-1998 is based on disaggregating 1999 national
emissions by process type and by state and then using that to develop shares of state emissions as a portion
of total national emissions. These 1999 state emissions shares by process type were used to allocate
estimated total U.S. emissions by process type to states for 1990-1998.
3.3.4.2.2. Primary, Secondary, Die Casting, and Sand Casting
The methodology used for 2011-2022 relied on GHGRP-reported emissions (EPA 2024b). EPA allocated
emissions from GHGRP reporting facilities to the states in which the reporting facilities are located. For non-
reported estimated emissions or emissions estimated from smaller casting facilities falling under the GHGRP
reporting threshold, EPA allocated emissions associated with the non-reporting population proportionally to
states with reported emissions. For example, if state A had X% of total reported GHGRP emissions for a
particular process type, state A got X% of total U.S. estimated non-reported emissions for that particular
process type.
The methodology used for 1999-2010 relied on emissions reported to EPA as under EPA's SFs Emission
Reduction Partnership for the Magnesium Industry. EPA allocated emissions from partners to the state in
which facilities are located as reported through the GHGRP or identified through online research. Note that
the national Inventory assumes that all U.S. emissions from primary and secondary production in 1999-2010
were from partners. This is not the case for die casting and sand casting. For non-reported estimated
emissions, EPA allocated emissions associated with the non-reporting population proportionally to states
with reported emissions for the appropriate process type.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-68
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
3.3.4.2.3. Permanent, Wrought, and Anode
For 2011-2022 emissions associated with these processes are not reported through the GHGRP. Total
U.S. production is reported through the USGS Yearbook (USGS 2002, 2003, 2005-2017, 2020-2022).
Therefore, EPA used a similar methodology that is used for the non-reported emissions state allocation for
primary, secondary, die, and sand casting. Emissions associated with these types of processes were
allocated proportionally to states with reported emissions, with the exclusion of primary production facilities
because there is only one facility and it is not in a state that has other magnesium facilities.
For 1999-2010, emissions associated with these processes were not reported through the Partnership
Program. Total U.S. production is reported through the USGS Yearbook. Therefore, EPA used a methodology
similar to the methodology for allocating non-reported emissions for primary, secondary, die, and sand
casting to the states. EPA allocated total U.S. emissions associated with these types of processes
proportionally to states with reported emissions for secondary, die casting, and sand casting, excluding the
primary production facility, assuming that these states were the most likely to contain facilities that
produced magnesium products via permanent, wrought, and anode processes; however, it is possible that
other states have emissions from these production processes.
3.3.4.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of SFs, HFC-134a, and C02
emissions from magnesium production and processing were calculated using the were calculated using the
2019 Refinement to the 2006IPCC Guidelines. As described further in Chapter 4 of the national Inventory,
levels of uncertainty in the national estimates in 2022 for all gases in aggregate were -9%/+9%.
Overall, the state-level estimates of emissions for magnesium are expected to have a higher uncertainty
than the national estimates; however, the variability in uncertainty levels between state-level estimates and
national estimates differs throughout the time series. For the 2011-2022 time series, the uncertainties
associated with the state-level estimates are expected to be low because emissions are estimated and
reported at the facility level for the most part. Nevertheless, the 2011 -2022 state-level uncertainties are
somewhat higher than 2011-2022 national-level uncertainties because for some process types facility-
reported data are not available (i.e., permanent, wrought, and anode). For 1999-2010, state-level estimates
have a higher uncertainty that national estimates in the same time period, as well as more uncertainty than
that of the state-level estimates for 2011-2022. This is due to a higher proportion of facility data being
available through the GHGRP as compared to the EPA Partnership for each year. Allocation of estimated but
unreported emissions for specific process types (i.e., sand casting, die casting, permanent, wrought, and
anode) is also done within this time period based on the state proportions of reported emissions, leadingto
increased uncertainty due to the assumption that unreported emissions occur in the same proportion across
states as reported emissions. For 1990-1998, state-level estimates are expected to have a significantly
higher level of uncertainty than that of more recent years because no facility-specific emissions are available
and because emissions have been allocated to states based on a single year of state-level data, which does
not account for changes in emitters over the time period, such as plant openings and closures or process
changes. These assumptions were required due to lack of available state- or regional-level data. For more
details on national-level uncertainty, see the Uncertainty discussion in Chapter 4 of the national Inventory.
3.3.4.4 Recalculations
Additional data and new information became available through the GHGRP that affected state
estimates:
3-69
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Updates to back casting methodology for a die casting facility based on an earlier confirmed
opening on the facility. Updates to values previously held constant for 2001 -2013 by shifting to
interpolation between the new confirmed opening year and the year of first reported data.
Updates to the estimation methodology of sand casting non-partner GHGRP volumes and updates
to the emission factor for sand casting from 1990 to 2011 changed the amount of nonreported sand
emissions and the distribution of those emissions to states.
Refer to Section 4.21 (page 4-109) of the national Inventory report for a complete list of recalculations
for the national Inventory.
3.3.4.5 Planned Improvements
One planned improvement would be to investigate information that could be used to update the factors
used to allocate emissions from non-reporters. Currently, this is based on the fraction of GHGRP-reported
emissions in each state.
Planned improvements are the same as those planned for improving national estimates, given that the
underlying methods for state GHG estimates are the same as those in the national Inventory, and given that
improvements in the national Inventory will lead directly to improvements in the quality of state-level
estimates as well. For more information, see Chapter 4, Section 4.20, of the national Inventory.
3.3.4.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990 -2022. EPA 430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventory-us-
greenhouse-gas-emissions-and-sinks.
EPA (2024b) Envirofacts. Subpart T: Magnesium Production. Available online at:
http ://www.epa.gov/enviro/facts/ghg/search.html.
U.S. Geological Survey (2002, 2003, 2005-2017, 2020-2022) Minerals Yearbook: Magnesium. Available
online at: http://minerals.usgs.gOv/minerals/pubs/commoditv/magnesium/index.html#mis.
3.3.5 Lead Production (NIR Section 4.22)
3.3.5.1 Background
Primary production of lead through the direct smelting of lead concentrate produces C02 emissions as
the lead concentrates are reduced in a furnace using metallurgical coke. Similar to primary lead production,
C02 emissions from secondary lead production result when a reducing agent, usually metallurgical coke, is
added to the smelter to aid in the reduction process. C02 emissions from secondary lead production also
occur through the treatment of secondary raw materials. Emissions from fuels consumed for energy
purposes during the production of lead are accounted for in the energy sector. In 2022, emissive lead
production occurred in eight states: Alabama, Minnesota, Indiana, Missouri, New York, Florida, California,
and Pennsylvania. The last primary lead production facility in the United States closed at the end of 2013.
3.3.5.2 Methods/Approach
To compile emissions by state from lead production using available data, this state-level inventory
disaggregated national emissions from the national Inventory with an Approach 2 method as defined in the
Introduction chapter, using a combination of process emissions reported to the GHGRP to calculate process
emissions and the number of facilities in a state (see Table 3-15). See Appendix H, Tables H-7 through H-9 in
the "Lead" Tab, for more details on the data used.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-70
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
The national Inventory methodology was adapted to calculate state-level GHG emissions to ensure
consistency with national estimates. National estimates were downscaled across states because of
limitations in availability of state-specific data across the time series to use when applying national methods
(i.e., IPCC Tier 1 methods) at the state level. The sum of emissions by state is consistent with national
process emissions as reported in the national Inventory.
Table 3-15. Summary of Approaches to Disaggregate the National Inventory for Lead Production
Across Time Series
Time Series
Range
Summary of Method
2010-2022
GHGRP process emissions data were used to estimate the percentage of
emissions by state, multiplied by the national emissions (IPCC 2006 Tier 1).
1990-2009
Data on number of lead facilities were used to estimate the percentage of
production by state, multiplied by the national emissions (IPCC 2006 Tier 1).
The methodology used for 2010-2022 was based on process emissions reported to the GHGRP
summed by state (EPA 2010-2022) to calculate a percentage of emissions from each state. The GHGRP has a
reporting threshold of 25,000 metric tons of C02 equivalent for lead production, so these emissions data are
representative of the larger facilities in the industry. Using GHGRP emissions data means that emissions
from states with smaller facilities were possibly underestimated. That percentage was then applied to the
national emissions from lead production peryear to calculate disaggregated gross C02 emissions by state.
The methodology used for 1990-2009 was based on the number of facilities in each state divided by the
number of facilities nationally to calculate a percentage of facilities in each state for each year. This
percentage was applied to the national C02 emissions from lead production per year (EPA 2024) to
disaggregate C02 emissions by state for each year. For 1995-2009, the number of facilities per state was
compiled from the USGS Minerals Yearbooks for lead, as available (USGS 1995-2009), and locations were
estimated based on available information. For 1990-1994, the number of facilities from the 1995 USGS
Minerals Yearbook for lead was used because the Minerals Yearbooks for those years did not contain the
number of facilities.
The USGS Mineral Commodity Summaries for lead (USGS 1995-2022) only provide primary and
secondary lead production as total national values, with no breakdown by state. The USGS Minerals
Yearbooks for lead also did not have any state-specific production data. As such, these sources could not be
used for state-level data in the state disaggregation estimates.
3.3.5.2.1. Primary Versus Secondary Production Adjustment
In general, C02 emissions from primary lead production facilities are about two times the C02
emissions from secondary lead facilities on a per-unit or production basis. To account for the difference
between primary and secondary lead facilities for the years 1990-2013, when primary lead production took
place in the United States, an adjustment was made to the state primary and secondary facility counts. The
GHGRP C02 emissions for the one primary facility and the secondary facilities for RYs 2010-2013 were
compiled. Next, the production for the primary facility and secondary facilities from the USGS Minerals
Yearbooks was compiled for 2010-2013. The ratio of C02 emissions to production for each year for the
primary facility and secondary facilities was calculated and then averaged across those years. Primary
facilities have, on average, a 1:1 ratio of C02 emissions to production tons. Secondary facilities have, on
average, a 1:2 ratio of C02 emissions to production tons. The average ratios for primary and secondary
facilities were applied to each state's primary and secondary facility count to calculate a weighted
percentage of emissions per state for primary and secondary facilities.
3-71
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
3.3.5.2.2. CEMS Adjustment for2010-2022
Starting in 2010, Lead-producing facilities with emissions over the GHGRP reporting threshold reported
both process and combustion emissions to the GHGRP. One facility started using a CEMS to measure and
report C02 emissions in 2016. For this facility starting in 2016, process and combustion emissions were
reported together under Subpart C per the GHGRP requirements. All other facilities not using a CEMS
reported process emissions under Subpart R and combustion emissions under Subpart C.32 To disaggregate
process emissions for the facility using a CEMS, a facility-specific default ratio of process emissions to total
emissions was calculated for each year from 2010 to 2015 and averaged. Emissions reported to Subparts R
and C were compiled for the one facility, and the percentage of process emissions to total emissions for the
non-CEMS years was applied to the total C02 emissions for each year the facility used CEMS in order to
calculate process emissions for each year. The results were an estimated process C02 emissions value for
that CEMS facility for 2016-2022.
Because the methodology for 1990-2009 does not use GHGRP emissions data to calculate the state
emissions and the facility did not begin using a CEMS to report emissions until 2016, there is no need to
adjust for CEMS facilities for those years.
3.3.5.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02 from lead production was
calculated using the 2006 IPCC Guidelines Approach 2 methodology for uncertainty (IPCC 2006). As
described further in Chapter 4 and Annex 7 of the national Inventory (EPA 2024), levels of uncertainty in the
national estimates in 2022 were -15%/+16% for C02.
State-level estimates are expected to have an overall higher uncertainty because the national emissions
estimates were apportioned to each state based on a combination of GHGRP emissions data for 2010-2022
and the estimated number and location of facilities for 1990-2009.
For 2010-2022, uncertainty is expected to be lower because of the use of GHGRP emissions data by
state to allocate national GHG emissions by state, which is a surrogate for using lead production data by
state to calculate emissions. National Inventory estimates, however, have been 7% to 36% lower than
GHGRP estimates for 2010-2022. State-level inventory estimates are derived from the national Inventory
figures and, therefore, are lower than the corresponding totals for facilities from a given state that reports to
the GHGRP.
For 1990-2009, this allocation method does not address facilities' production capacities or utilization
rates, which vary from facility to facility and from year to year. While this approach does assume differences
in primary and secondary production processes, it implicitly assumes emissions from those primary and
secondary facilities, respectively, are equal regardless of production capacity or utilization rates, which
could overestimate emissions in states with smaller facilities and underestimate emissions in states with
larger facilities.
Primary lead production occurred in the United States from 1990 to 2013. To minimize uncertainty,
methods were adjusted to account for differences in emissions from primary and secondary lead production.
32 For more information on the GHGRP, see 74 FR 56374, October 30, 2009, available online at
https://www.govinfo.gov/content/pkg/FR-2009-10-3Q/pdf/E9-23315.pdf.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-72
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
3.3.5.4 Recalculations
Minor recalculations were performed in this report for 2020 and 2021 due to updates to the national
Inventory data set, based upon revised USGS data for secondary lead production. Compared to prior
estimates, estimated C02 emissions decreased by approximately 3% for 2020 and 2% for 2021.
3.3.5.5 Planned Improvements
More information on combustion C02 emissions from smelting furnaces is needed to disaggregate
combustion and process emissions from the facility reporting C02 with a CEMS to the GHGRP in 2016-2022.
Additionally, because the GHGRP data set is available starting with 2010, EPA is assessing the feasibility to
review and update lead production data by state for earlier parts of the time series. For example, the
estimated number and location of facilities producing lead per state for 1990-2009 still need to be
confirmed, especially for 1990-1994.
EPA will review time series consistency issues due to the two methodologies for 1990-2009 and 2010-
2022. Surrogate data on the number of primary and secondary lead production facilities were used in place
of activity data for the 1990-2009 portion of the time series, and more research is needed so calculations
more closely reflect state trends in emissions.
3.3.5.6 References
EPA (U.S. Environmental Protection Agency) (2010-2022) Envirofacts GHGRP Subpart R and Subpart C Data.
Accessed May 13, 2024. Available online at: https://enviro.epa.gov/query-
builder/ghghttps://www.epa.gov/enviro/greenhouse-gas-customized-search.
EPA (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2022. EPA430-R-24-004. Available
online at: https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
USGS (U.S. Geological Survey) (1995-2009) Minerals Yearbook: Lead. Available online at:
https://www.usgs.gov/centers/national-minerals-information-center/lead-statistics-and-information.
USGS (1995-2022) Mineral Commodity Summary: Lead. Available online at:
https://www.usgs.gov/centers/national-minerals-information-center/lead-statistics-and-information.
3.3.6 Zinc Production (NIR Section 4.23)
3.3.6.1 Background
Zinc production in the United States consists of both primary and secondary processes. Of the primary
and secondary processes currently in use in the United States, only the Waelz kiln secondary process results
in nonenergy C02 emissions. For earlier years in the time series, the emissive electrothermic process was
utilized from before 1990 to 2014, the pig iron zinc oxide furnace process from 2009 to 2012, and the flame
reactor process from 1993 to 2013. Emissions from fuels consumed for energy purposes during the
production of zinc are accounted for in the energy sector. In 2022, emissive zinc production occurred in five
states: Alabama, Pennsylvania, South Carolina, Tennessee, and Illinois.
3.3.6.2 Methods/Approach
To compile emissions by state from zinc production using available data, this state-level inventory
disaggregated national emissions from the national Inventory with an Approach 2 method as defined in the
3-73
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Introduction chapter, using a combination of process emissions reported to the GHGRP and the number of
facilities in a state (see Table 3-16). See Appendix H, Tables H-10 through H-14 in the "Zinc" Tab, for more
details on the data used.
The national Inventory methodology was adapted to calculate state-level GHG emissions to ensure
consistency with national estimates. National estimates were downscaled across states because of
limitations in the availability of state-specific data across the time series to use when applying national
methods (e.g., IPCC Tier 2 methods) at the state level. The sum of emissions by state is consistent with
national process emissions as reported in the national Inventory.
Table 3-16. Summary of Approaches to Disaggregate the National Inventory for Zinc Production
Across Time Series
Time Series Range
Summary of Method
2010-2022
GHGRP process emissions data were used to estimate the percentage of
emissions by state, multiplied by the national emissions (IPCC 2006Tier 2).
1990-2009
Data on number of zinc facilities were used to estimate the percentage of
production by state, multiplied by the national emissions (IPCC 2006 Tier 2).
The methodology for 1990-2009 used the number of facilities in each state divided by the number of
facilities nationally to calculate a percentage of facilities in each state for each year. This percentage was
applied to the national C02 emissions from zinc production per year (EPA 2024) to calculate disaggregated
C02 emissions by state for each year. The number of facilities per state was determined from reviewing the
number of facilities reporting to the GHGRP and using company websites to confirm when facilities opened
and closed, as well as the number of electrothermic furnaces, Waelz kilns, other furnaces, and flame reactor
units.
The methodology for 2010-2022 used process emissions reported to the GHGRP summed by state and
nationally (EPA 2010-2022) to calculate a percentage of emissions from each state. That percentage was
then applied to the national emissions from zinc production per year to calculate disaggregated gross C02
emissions by state. The GHGRP has a reporting threshold of 25,000 metric tons of C02 equivalent for zinc
production, so these emissions data are representative of the larger facilities in the industry. Using GHGRP
emissions data means emissions from states with smaller facilities were possibly underestimated.
The USGS Mineral Commodity Summaries for zinc (USGS 1990-2021) only had U.S. zinc production as
total national values with no breakdown by state. The USGS Minerals Yearbooks for zinc also did not have any
state-specific production data. As such, these sources could not be used for state-level data in the state
disaggregation estimates.
3.3.6.2.1. EAF Dust Consumption Facility Accounting for2010-2022
Since 2010, the GHGRP has required zinc manufacturing facilities that operate electrothermic furnaces
or Waelz kilns to report C02 emissions. The national Inventory includes emissive facilities that operate
electrothermic furnaces or Waelz kilns and other facilities that process EAF dust. The one facility utilizing an
electrothermic furnace was in operation from before 1990-2014. Two additional facilities that process EAF
dust do not have electrothermic furnaces or Waelz kilns and do not report to the GHGRP, but they are
accounted for in the national Inventory: PIZO Operating Co. in Blytheville, Arizona, and American Zinc
Recycling Corp. (AZR; formerly Horsehead Holding Corp.) in Beaumont, Texas.
The PIZO Blytheville facility was in operation from 2009 to 2012 (ADEQ 2021). The national Inventory
methodology of using estimated EAF dust consumed values and an emissions factor of 1.24 metric ton C02
per metric ton EAF dust consumed was used to calculate C02 emissions for each year.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-74
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
The AZR facility in Beaumont was in operation from around 1993-2009 (AZR 2021). The EAF dust
recycling and processing capacity for the AZR facility for 2009 was obtained from the U.S. Securities and
Exchange Commission (Horsehead Holding Corp. 2010). The C02 emissions for the AZR facility were
calculated using the national Inventory methodology, using estimated EAF dust consumed values and an
emissions factor of 1.24 metric ton C02 per metric ton EAF dust consumed.
3.3.6.2.2. Electrothermic Furnace, Waelz Kiln, Other Furnaces, and Flame Reactor Unit
Adjustment for 1990-2009
Emissions data reported to GHGRP show that per-unit production C02 emissions from Waelz kilns are
about two times the C02 emissions from electrothermic furnaces (EPA 2010-2012). The 2010-2019 GHGRP
C02 emissions for electrothermic furnaces and Waelz kilns and number of units by type (i.e., electrothermic
furnaces and Waelz kilns) per facility were compiled to calculate the average C02 emissions per facility and
average C02 emissions per unit per facility. Note that 2020 through 2022 GHGRP emissions data were not
included in calculating these averages, as 2020 and future year data may not be as representative to apply to
1990-2009 emissions estimates. Only one facility had electrothermic furnaces. The average C02 emissions
per unit per facility were calculated across the five facilities with Waelz kilns. To account for the difference in
the quantity of C02 emissions from electrothermic furnaces and Waelz kilns, an adjustment was made to the
number of electrothermic furnaces and Waelz kilns per state for the years 1990-2009.
The 2009 C02 emissions value for the PIZO facility was used to estimate C02 emissions for other
furnaces, while the 2009 C02 emissions value for the AZR facility was used to estimate C02 emissions for
flame reactor units.
The average C02 emissions per unit for electrothermic furnaces and Waelz kilns and the 2009 C02
emissions per unit value for other furnaces and flame reactor units were applied to calculate a weighted
percentage of emissions per state for electrothermic furnaces, Waelz kilns, other furnaces, and flame
reactor units. Each percentage of emissions per state was applied to the national C02 emissions from the
national Inventory to calculate C02 emissions per state.
3.3.6.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02 from zinc production was
calculated using the 2006 IPCC Guidelines Approach 2 methodology for uncertainty (IPCC 2006). As
described further in Chapter 4 and Annex 7 of the national Inventory (EPA 2024), levels of uncertainty in the
national estimates in 2022 were -18%/+20% for C02.
State-level estimates are expected to have an overall higher uncertainty because the national emissions
estimates were apportioned to each state based on the number of facilities and production processes for
1990-2009 and GHGRP emissions data for 2010-2022.
For 1990-2009, this allocation method does not address production capacity or utilization rate at a
facility-specific level. This approach could overestimate emissions in states with smaller capacity or less
used production units and underestimate emissions in states with larger capacity or high utilization
production units.
For 2010-2022, uncertainty is expected to be lower than for the period 1990-2009 due to the use of
GHGRP emissions data by state to calculate emissions. Smaller facilities do not report to GHGRP, however,
and were excluded from these estimates, affecting the completeness of the estimates.
3-75
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
3.3.6.4 Recalculations
Minor recalculations were performed in this report for 2021 state-level inventory estimates due to a
revision to the national Inventory based on updated EAF dust consumption data. The 2021 national Inventory
revised estimate for emissions from zinc production increased by 4% as a result. This update results in a
corresponding increase in estimated state-level emissions for 2021.
3.3.6.5 Planned Improvements
Data gaps to calculate emissions from zinc production include zinc production by unit type by state for
the full time series. The estimated number of facilities producing zinc per state for 1990-2009 needs to be
confirmed, including the zinc production methodology (e.g., electrothermic furnaces, Waelz kilns, other
facilities processing EAF dust).
3.3.6.6 References
ADEQ (Arkansas Division of Environmental Quality) (2021) Personal communication between Thomas
Rheaume, Arkansas Division of Environmental Quality, and Amanda Chiu, U.S. Environmental Protection
Agency. February 16, 2021.
AZR (American Zinc Recycling) (2021) Summary of Company History. Accessed March 3, 2021. Available
online at: https://web.archive.Org/web/20210620033241/https://azr.com/our-historv/.
EPA (U.S. Environmental Protection Agency) (2010-2022) Envirofacts GHGRP Subpart GG and Subpart C
Data. Accessed May 13, 2024. Available online at: https://www.epa.gov/enviro/greenhouse-gas-
customized-search
EPA (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2022. EPA430-R-24-004. Available
online at: https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
Horsehead Holding Corp. (2010) Form 10-K: Annual Report Pursuant to Section 13 or 15(d) of the Securities
Exchange Act of 1934 forthe Fiscal Year Ended December 31, 2009. Available online at:
https://lastl0k.eom/sec-filings/zincq/0000950123-10-025167.htm#link fullReport.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
USGS (U.S. Geological Survey) (1990-2021) Mineral Commodity Summary: Zinc. Available online at:
https://www.usgs.gov/centers/national-minerals-information-center/zinc-statistics-and-information.
3.4 Product Use (Fluorinated Sources, N20)
The product use portion of IPPU emissions is a catch-all category that consists of the following:
Electronics industry (HFCs, PFCs, SFs, NF3, N20)
Substitution of ozone-depleting substances (ODSs) (HFCs, PFCs)
Electrical transmission and distribution (SFs)
SFs and PFCs from other product use
N20 from product uses (N20)
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-76
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
3.4.1 Electronics Industry (NIR Section 4.24)
3.4.1.1 Background
The electronics industry uses multiple greenhouse gases in its manufacturing processes. In
semiconductor manufacturing, these include long-lived fluorinated greenhouse gases used for plasma
etching and chamber cleaning, fluorinated heat transfer fluids used for temperature control and other
applications, and nitrous oxide (N20) used to produce thin films through chemical vapor deposition. Similar
to semiconductor manufacturing, the manufacturing of micro-electro-mechanical systems (MEMS) devices
and photovoltaic cells requires the use of multiple long-lived fluorinated greenhouse gases for various
processes. Electronics manufacturing occurs in the following states: Arizona, California, Colorado, Florida,
Georgia, Hawaii, Idaho, Indiana, Maine, Maryland, Massachusetts, Minnesota, Mississippi, Missouri, New
Jersey, New Mexico, New York, North Carolina, Oregon, Pennsylvania, Texas, Utah, Vermont, Virginia, and
Washington.
For semiconductors, a single 300 mm silicon wafer that yields between 400 to 600 semiconductor
products (devices or chips) may require more than 100 distinct fluorinated-gas-using process steps,
principally to deposit and pattern dielectric films. Plasma etching (or patterning) of dielectric films, such as
silicon dioxide and silicon nitride, is performed to provide pathways for conducting material to connect
individual circuit components in each device. The patterning process uses plasma-generated fluorine atoms,
which chemically react with exposed dielectric film to selectively remove the desired portions of the film. The
material removed as well as undissociated fluorinated gases flow into waste streams and, unless emission
abatement systems are employed, into the atmosphere. Plasma enhanced chemical vapor deposition
chambers, used for depositing dielectric films, are cleaned periodically using fluorinated and other gases.
During the cleaning cycle the gas is converted to fluorine atoms in plasma, which etches away residual
material from chamber walls, electrodes, and chamber hardware. Undissociated fluorinated gases and other
products pass from the chamber to waste streams and, unless abatement systems are employed, into the
atmosphere.
In addition to emissions of unreacted gases, some fluorinated compounds can also be transformed in
the plasma processes into different fluorinated compounds which are then exhausted, unless abated, into
the atmosphere. For example, when C2F6 is used in cleaning or etching, CF4 is typically generated and
emitted as a process byproduct. In some cases, emissions of the byproduct gas can rival or even exceed
emissions of the input gas, as is the case for NF3 used in remote plasma chamber cleaning, which often
generates CF4 as a byproduct.
Nitrous oxide is used in manufacturing semiconductor devices to produce thin films by CVD and
nitridation processes as well as for N-doping of compound semiconductors and reaction chamber
conditioning (Doering and Nishi 2000).
Liquid perfluorinated compounds are also used as heat transfer fluids (F-HTFs) for temperature control,
device testing, cleaning substrate surfaces and other parts, and soldering in certain types of semiconductor
manufacturing production processes. Leakage and evaporation of these fluids during use is a source of
fluorinated gas emissions (EPA 2006).
3.4.1.2 Methods/Approach
Emissions associated with the electronics industry include emissions from manufacturing of
semiconductors, MEMS, and PV. National emissions were estimated using IPCC Tier 2 methods as
discussed further in Chapter 4, Section 4.24 (on page 4-143) of the national Inventory (EPA 2024). In general,
EPA used a Hybrid approach to disaggregate national estimates.
3-77
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
3.4.1.2.1. Semiconductor and Micro-Electro-Mechanical Systems (MEMS) Manufacturing
To disaggregate emissions by state for semiconductors and MEMS, EPA used data from the GHGRP and
the World Fab Forecast (WFF).33 A Hybrid approach was used to estimate emissions from semiconductor and
MEMS manufacturing, relying on a mix of state-level data derived from the GHGRP and disaggregation of
national-level emission estimates where facility-level data were not available. For years before 2011, when
data gathering under the GHGRP began, each state's estimated share of U.S. total manufactured layer area
(TMLA) was multiplied by the national semiconductor emissions estimate to calculate that state's
semiconductor emissions. To calculate each state's MEMS emissions, a linear interpolation was used
between 1990 (assuming zero emissions from MEMS manufacturing in the state in that year) and 2011, the
first year of available GHGRP data. Table 3-17 summarizes methods used to compile emissions of CF4, C2F6,
C3F8, CHF3, SFs, NF3, C4F8, C4F6, C4F80, C5F8, CH2F2, CH3F, CH2FCF3, C2H2F4, and N20 from semiconductor
and MEMS manufacturing.
Table 3-17. Summary of Approaches to Disaggregate the National Inventory for Semiconductor and
MEMS Manufacturing Across Time Series
Time Series Range
Summary of Method
2015-2022
Emissions from reported fabs were allocated to the state in which the reporting
facility was located as reported through the GHGRP (Approach 1).
Emissions from non-reporting facilities were allocated by calculating the total
TMLA estimated for non-reporting facilities in each state using the WFF data set
and multiplying by the total emission factor of each gas in MT of gas per TMLA.
These emission factors were derived by performing a linear regression of the
MT emissions per gas from reporter facilities via GHGRP (regression y-axis
values) with the associated total TMLA of these facilities from the proprietary
WFF data (regression x-axis values).
Emissions from non-reporting MEMS facilities were not estimated, which is
consistent with the national Inventory.
2014
Emissions from reported fabs were allocated to the state in which the reporting
facility was located as reported through the GHGRP.
Emissions from non-reporting facilities were allocated by calculating the
percentage of TMLA estimated for non-reporting facilities in each state using
the WFF data set and multiplying by the total estimate of non-reported
emissions in the national Inventory. The unreported emissions were scaled up
by 0.017% to account for time series consistency (Approach 2).
Emissions from non-reporting MEMS facilities were not estimated, which is
consistent with the national Inventory.
2013
Emissions from reported fabs, adjusted for time series consistency in the
national Inventory, were allocated based on the location of the GHGRP facility.
The reported emissions were scaled up by 0.017% to account for time series
consistency (Approach 1).
Emissions from non-reporting facilities were allocated by calculating the
percentage of TMLA estimated for non-reporting facilities in each state using
the WFF data set and multiplying by the total estimate of non-reported
emissions in the national Inventory. The unreported emissions were scaled up
by 0.017% to account for time series consistency (Approach 2).
33 EPA periodically purchases the World Fab Forecast from SEMI (https://www.semi.org/en/products-services/market-
data/world-fab-forecast).
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-78
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
Time Series Range
Summary of Method
Emissions from non-reporting MEMS facilities were not estimated, which is
consistent with the national Inventory.
2011-2012
Emissions from reported fabs, adjusted for time series consistency in the
national Inventory, were allocated based on the location of the GHGRP facility
(Approach 1).
Emissions from non-reporting facilities were allocated by calculating the
percentage of TMLA estimated for non-reportingfacilities in each state using
the WFF data set and multiplying by the total estimate of non-reported
emissions in the national Inventory (Approach 2).
Emissions from non-reporting MEMS facilities were not estimated, which is
consistent with the national Inventory.
2008-2010
Emissions were allocated to states using the proportional state-level TMLA
breakdowns for the respective year, which were applied to total estimates from
the national Inventory (Approach 2).
1990-2007
Emissions from semiconductor manufacturing were allocated between states
from the national Inventory in the same proportion as they were in 2008
(Approach 2).
Emissions from MEMS were assumed to be zero in 1990. Emissions from MEMS
facilities from 1991 to 2010 were then estimated by interpolating between 1990
emissions and the emissions estimated for 2011 for each state (Approach 2).
N20 emissions data were first reported in 2015, so emissions from MEMS
facilities from 1991 to 2014 were interpolated for N20 (Approach 2).
From 2014 to 2022, emissions from reported fabs were allocated to the state in which the reporting
facility was located as reported through the GHGRP. From 2015 to 2022, emissions from non-reporters were
allocated to each state as described above. For 2014, emissions from non-reportingfacilities that
manufactured semiconductors were estimated by calculating the percentage of TMLA estimated for non-
reportingfacilities in each state using the WFF data set; the state's percentage of total non-reporter TMLA
was then used to allocate the non-reporter portion of national emissions as calculated in the national
Inventory. Non-reporter emissions from 2014 were scaled up by 0.017% to account for the differences in
emissions factor utilized. Emissions from non-reporting MEMs fabs are not estimated, which is consistent
with the national Inventory.
From 2011 to 2013, fluorinated GHGs (F-GHG) and N20 emissions from reported fabs, adjusted for time
series consistency in the national Inventory, were allocated based on the location of the GHGRP facility.
Emissions from non-reporters were allocated to each state as described above. Emissions from non-
reporting facilities that manufactured semiconductors were estimated using the same approach described
above for non-reporter emissions from 2014. Both reporter and non-reporter emissions from 2013 were
scaled up by 0.017% to account for the differences in emissions factor utilized. Emissions from non-
reporting MEMS facilities are not estimated, which is consistent with the national Inventory.
From 2008 to 2010, F-GHG and N20 emissions from semiconductor manufacturing were allocated to
states using the proportional state-level TMLA breakdowns for the respective year, which were applied to
total estimates from the national Inventory.
From 1990 to 2007, F-GHG and N20 from semiconductor manufacturing emissions were allocated
between states in the same proportion as they were in 2008.
From 1990 to 2011, emissions from MEMS facilities were estimated by interpolating between 1990
emissions and the emissions estimated for 2011. Emissions from MEMS were assumed to be zero in 1990.
3-79
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
N20 emissions from MEMS facilities were first reported in 2015 and assumed to be zero in 1990. Emissions
from 1991 to 2014 were interpolated between 1990 emissions and the emissions estimate for 2015. Only one
facility in New York, GE Global Research Center, reported N20 emissions, so all N20 emissions in the time
series were attributed to New York.
Only 26 states were identified as containing semiconductor fabs, six of which also reported emissions
from the production of MEMS.
3.4.1.2.2. Fluorinated Heat Transfer Fluids (F-HTFs)
To estimate state-level emissions of F-HTFs, EPA used a Hybrid approach to disaggregate national
emissions. For the national Inventory, for years when GHGRP data were available, EPA estimated state-level
emissions based on facility location. For earlier years, EPA allocated national F-HTF emissions to each state
based on that state's share of national F-GHG emissions from semiconductor manufacturing. This Hybrid
approach was used due to a lack of available data on reported HTF emissions or HTF consumption at the
facility or state level for years prior to GHGRP's availability. Table 3-18 summarizes methods used to compile
HTF emissions.
Table 3-18. Summary of Approaches to Disaggregate the National Inventory for Fluorinated Heat
Transfer Fluids Across Time Series
Time Series
Range
Summary of Method
2011-2022
National F-HTF emissions were allocated to the states in the same proportion as
emissions from reported fabs were allocated to the states in which the reporting
facilities were located, as reported through the GHGRP (Approach 1).
Emissions from non-reporters were added to each state's emissions from HTFs
by multiplying state emissions of HTFs by the estimated non-reporter GHGRP
emissions percentage taken from the national Inventory (Approach 2).
2000-2010
National F-HTF emissions were allocated to states in the same proportion as F-
GHG emissions associated with semiconductor manufacturing (Approach 2).
1990-1999
F-HTF emissions do not occur and are not estimated in the national Inventory
during 1990-1999 and thus are estimated to not occur at state levels.
From 2011 to 2022, emissions from reported fabs were allocated to the state in which the reporting
facility was located as reported through the GHGRP. Emissions from non-reporters were added to each
state's emissions from HTFs by multiplying state emissions of HTFs by the estimated non-reporter GHGRP
emissions percentage taken from the national Inventory.
For emissions from 2000 to 2010, F-HTF emissions were allocated between states in the same
proportion as F-GHG emissions associated with semiconductor manufacturing. Emissions data were taken
directly from the national Inventory and the allocation was only applied to the HTF emissions that were
included in the national Inventory totals. HTF emissions were assumed to not occur during or before 2000. A
total of 23 states were identified as reporting emissions of F-HTFs.
Emissions from 1990 to 1999 are assumed not to have occurred. Fluorinated HTF use in semiconductor
manufacturing is assumed to have begun in the early 2000s.
Additionally, the state-level HTF emissions estimates utilize GWPs as published in the latest version of
40 CFR part 98 Table A-1, which is comprised of GWPs from the IPCC Fifth Assessment Report (AR5) (and
Sixth Assessment Report [AR6] where 100-year GWPs are not available in AR5). This approach is consistent
with the rest of the state-level emissions estimates and the national Inventory, with the exception of the
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-80
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
national HTF emissions estimates from the electronics sector, which apply GWPs as published in the IPCC
Fourth Assessment Report (AR4). The HTF GWPs utilized in the national lnventoryw\[[ be updated to reflect
those in the latest version of 40 CFR part 98 Table A-1 in the next national Inventory cycle (see Section
1.4.1.5).
3.4.1.2.3. Photovoltaics
To estimate state-level emissions from photovoltaics (PV) manufacturing, EPA used a Hybrid approach,
applying a GHGRP-derived emissions factor to state-level manufacturing capacity data. Two different
emissions factors were developed: one for fluorinated GHGs and one for N20. For years with available
GHGRP data, Approach 1 was used for manufacturers that reported PV emissions at the state level. This Hybrid
approach was used due to a lack of available data on reported emissions at the state level for years prior to the
GHGRP's availability. Table 3-19 summarizes methods used to compile state-level emissions from C2F6, C3F8,
CF4, CHF3, SFs, NF3, C4F8, and N20.
Table 3-19. Summary of Approaches to Disaggregate the National Inventory for Photovoltaics
Across Time Series
Time Series
Range
Summary of Method
2011-2022
State-level estimates of manufacturing capacity were used to allocate emissions
for non-reporters (Approach 2).
Reported facility data were allocated to the state where the facility was located
(Approach 1).
2000-2010
State-level estimates of manufacturing capacity based on facility-level
manufacturing capacity data were used to allocate emissions. Capacity was
interpolated for years in which capacity data were unavailable (Approach 2).
1998-1999
State-level emissions were interpolated for 1998 and 1999 (Approach 2).
1990-1997
Capacity was assumed to be zero during 1990-1997 (Approaches 1 and 2).
For 2011-2022, reported state-level emissions from photovoltaics (PV) manufacturing were estimated
by allocating emissions from GHGRP reporters to the state in which the reporting facility is located. Two PV
facilities, Micron Technology and Mission Solar, reported to the GHGRP, during this time period (neither for
the full period of 2011 through 2020). Therefore, all the reported emissions were allocated to Idaho and
Texasthe states in which Micron Technology and Mission Solar are located, respectivelyfor the years for
which reported data are available. Non-reporter emissions were estimated using manufacturing capacity
data from DisplaySearch (2010), which provides facility-specific data, including the facility's state. Emissions
from non-reporters were calculated by multiplying the manufacturing capacity of each state by emissions
factors in million metric tons C02e per megawatt (MW) (two emissions factors were developed, one for F-
GHGs and one for N20) based on reported emissions from Mission Solar.
For 2000-2010, non-reporter emissions were estimated using the proportion of each state's
manufacturing capacity in 2009 (the most recent year of DisplaySearch data purchased) to the overall non-
reporter estimate used in the national Inventory.
Manufacturing capacity was interpolated between 1997 and 2000 and used to estimate emissions in
1998 and 1999 using the same emissions factor described above. Manufacturing capacity was assumed to
be zero in 1997 and before based on an assessment of available industry manufacturing data (Platzer 2015).
Manufacturing capacity was interpolated between 1997 and 2000 and used to estimate emissions in 1998
and 1999 using the same emissions factor described above.
3-81
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
3.4.1.3 Uncertainty
The overall uncertainty associated with the national emissions estimates for the electronics industry
was calculated using the 2019 Refinement to the 2006IPCC Guidelines. As described further in Chapter 4 of
the national Inventory, levels of uncertainty in the national estimates in 2022 were -6%/+6% across the
electronics industry.
State-level estimates are expected to have a higher uncertainty than national estimates because the
uncertainty of each facility's emissions is higher than the uncertainty of emissions across all facilities, or in
other words the uncertainty of a sum of independent variables is lower than the uncertainty of the variables.
For years with state- and facility-level GHGRP data, state-level estimates will still be higher than national
totals due to the uncertainty of many additional independent variables. State-level estimates will have the
most uncertainty for years where state-level activity data were not available, namely years before the start of
GHGRP data. Pre-2011 estimates are generated by apportioning the national totals by state-level TMLA
estimates, which come from various sources including World Fab Watch and WFF. State-level estimates for
1990-2007 are apportioned using the most recent year of state-level TMLA data (2008), which will add
significant uncertainty to those estimates. For more details on national-level uncertainty, see the Uncertainty
discussion in Section 4.24 of the national Inventory.
3.4.1.4 Recalculations
The list of non-reporting semiconductor manufacturing facilities in 2015 was updated to remove one
facility that had been inadvertently included, addressing an error in the national Inventory. In addition, state-
level estimates for HTF emissions were updated to use AR5 and AR6 GWPs, addressing an error in the
national Inventory where HTF estimates were still using AR4 GWPs. Thus, overall semiconductor emissions
might not sum to estimates published in the national Inventory. The error will be addressed in the next
national Inventory published in April 2025.
Refer to the national Inventory report for a complete list of recalculations for the national Inventory.
3.4.1.5 Planned Improvements
Planned improvements are consistent with those planned for improving national estimates, given that
the underlying methods for state GHG estimates are the same as those in the national Inventory. For more
information, see Chapter 4, Section 4.24, of the national Inventory.
3.4.1.6 References
DisplaySearch (2010) DisplaySearch Q4 ' 09 Quarterly FPD Supply/Demand and Capital Spending Report.
Robert Doeringand Yoshiro Nishi (2000) Handbook of Semiconductor Manufacturing Technology. CRC Press.
EPA (U.S. Environmental Protection Agency) (2006) Uses and Emissions of Liquid PFC Heat Transfer Fluids
from the Electronics Sector. EPA-430-R-06-901.
EPA (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 -2022. EPA 430-R-24-004. Available
online at: https://www.epa.gov/ghgemissions/inventory-us-greenhouse-gas-emissions-and-sinks.
EPA (2024) Envirofacts. Subpart I: Electronics Manufacture. Available online at:
https://enviro.epa.gov/querv-builder/ghg
Platzer, M. (2015) U.S. Solar Photovoltaic Manufacturing: Industry Trends, Global Competition, Federal
Support. Congressional Research Service.
SEMI (Semiconductor Equipment and Materials International) (2012) World Fab Forecast, August2012
Edition.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-82
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
SEMI (2013) World Fab Forecast, May2013 Edition.
SEMI (2016) World Fab Forecast, May2017Edition.
SEMI (2017) World Fab Forecast, August2018 Edition.
SEMI (2018) World Fab Forecast, June 2018 Edition.
SEMI (2021a) World Fab Forecast, December2021 Update Edition.
SEMI (2021 b) World Fab Forecast, June 2021 Edition.
SEMI (2022) World Fab Forecast, December2022 Edition.
3.4.2 Substitution of Ozone-Depleting Substances (NIR Section 4.25)
3.4.2.1 Background
HFCs, PFCs, and C02 are used as alternatives to several classes of ODSs that are being phased out
under the terms of the Montreal Protocol and the Clean Air Act Amendments of 1990.34 ODSs such as
chlorofluorocarbons (CFCs), halons, carbon tetrachloride, methyl chloroform, and
hydrochlorofluorocarbons (HCFCs), are used in a variety of industrial applications, including refrigeration
and air conditioning equipment, solvent cleaning, foam production, sterilization, fire suppression, and
aerosols. HFCs, PFCs, and C02 are not harmful to the stratospheric ozone layer; they are GHGs with GWPs
ranging from 1 for C02 to tens of thousands for HFC-23 and some PFCs (EPA 2024).
3.4.2.2 Methods/Approach
As described in the national Inventory report (EPA 2024), EPA employs its Vintaging Model to estimate
national use, banks, emissions, and transition of ODS-containing equipment and products to substitutes,
including HFCs, PFCs, C02, and blends that contain such substances. The Vintaging Model estimates ODS
and ODS substitute trends in the United States based on modeled estimates of the quantity of equipment or
products sold each year that contain these chemicals and the amount of the chemical required to
manufacture or maintain equipment and products over time. Emissions for each end use were estimated by
applying annual leak rates and release profiles, which account for the lag in emissions from equipment as it
leaks over time. The model uses a Tier 2 bottom-up modeling methodology to estimate emissions and hence
requires extensive research, data, assumptions, and expert judgment to develop the activity levels and
emissions profiles over the time series for each of the 80 end uses modeled. See Section 4.25 and Annex 3.9
of the national Inventory for an additional description of the Vintaging Model and further details such as the
end uses modeled (EPA 2024).
An approach similar to the Vintaging Model can be used to develop state-level emissions estimates.
California, for example, uses this approach (CARB 2016). Doing so, however, requires the same extensive
data gathering and may be difficult to monitor given the interstate commerce that occurs for many of the
products involved.
Another approach to estimate a state's emissions would be to assume the state's proportion of national
emissions is the same as the state's proportion of national population. For many ODS substitute equipment
types, this is a reasonable approach. For instance, the number of supermarkets, home refrigerators, and
light-duty vehicles with air conditioning, per person, is not expected to vary significantly from state to state.
For some other end uses, however, that is not the case. For instance, EIA (2023) statistics confirm that the
34 42 U.S.C. § 7671, CM Title VI.
3-83
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
use of air conditioning varies by region, which could Lead to a significant difference that is not directly related
to population. As noted in the national Inventory, EPA estimates that residential unitary air conditioning is the
largest emitting (in C02 equivalent terms) end use within the refrigeration and air conditioning sector, which
accounts for 81% of national emissions (EPA 2024).
The disaggregation approach used here is a combination of using population as a proxy for emissions
(i.e., "Approach 2") while incorporating data provided at a finer geographical distribution than the national
emissions estimates (i.e., "Approach 1").
Analysis by NOAA further points to the varying nature of emissions across the United States (Hu et al.
2017, 2022, 2024; Montzka et al. 2023). The analysis incorporated data from a variety of ground- and air-level
measurements of various fluorocarbons. By applying Lagrangian atmospheric transport models and a
Bayesian inverse modeling technique, Hu et al. estimated emissions on a 1° * 1°grid across the contiguous
states and District of Columbia. The papers estimated emissions of various fluorocarbons (ODS and HFCs)
over six regions of the United States through this approach. The authors observed that spatial patterns for
individual compounds agree well with qualitative expectations, pointing to examples of higher per capita
emissions of chemicals used as blowing agents in building insulation foams (CFC-11, HCFC-142b, and HFC-
365mfc) in the northern states and higher per capita emissions of HCFC-22, HFC-125, and HFC-32 used in
residential and commercial air conditioning in southeastern and central south states. These results agreed
with recommendations for thermal insulation (U.S. Department of Energy 2016) in northern regions and the
higher percentage of homes with air conditioning (EIA 2018a, 2018b) in southern regions. Derived per capita
emissions of HFC-134a displayed similar regional patterns as refrigerants used in residential air conditioning,
except in the Central North region where the per capita emissions were comparable to that in southern
regions. The authors surmised that this distribution may stem from additional use of HFC-134a in
refrigeration, which may correspond to the higher use of a second refrigerator or a separate freezer in the
midwest (EIA 2023), and as a foam-blowing agent in building insulation in northern regions.
A population distribution was modified with data from Hu et al. (2017, 2022, 2024) to disaggregate
national emissions to individual states, territories, and the District of Columbia. For this exercise, data from
the U.S. Census were used to gather population estimates to distribute national-level emissions to the
regions incorporated into the national emissions estimates (i.e., for the 50 states, the District of Columbia,
Puerto Rico, American Samoa, Guam, the Northern Mariana Islands, and the U.S. Virgin Islands) (U.S.
Census Bureau 2021). Population estimates across the time series were not available for the Federated
states of Micronesias, the Marshall Islands, and Palau; therefore, none of the U.S. national emissions
estimates was attributed to those territories. For years in which a population estimate was not provided,
linear interpolation was used.
Annual emissions per capita for the six regions analyzed in Hu et al. (2017, 2022, 2024) were used.
Specifically, emissions for HFC-32, HFC-125, HFC-134a, and HFC-143a from 2008-2021 were available. The
six regions described in the paper are West (California, Oregon, and Washington), Mountain (Montana to New
Mexico), Central North (North Dakota to Kansas to Ohio), Central South (Texas to Alabama to Kentucky),
Southeast (North Carolina to Florida), and Northeast (West Virginia to Maine).
Because the Hu et al. (2017, 2022, 2024) estimates cover the 48 contiguous states and the District of
Columbia, emissions estimates from the remaining states (Alaska and Hawaii) and the five other territories
were derived strictly based on the state's or territory's population compared to the national population for
the full 1990-2022 time series. Likewise, the emissions of HFCs other than the four listed above were
distributed to all states and territories by population. The emissions of HFC-32, HFC-125, HFC-134a, and
HFC-143a were distributed to the six regions in the same ratio as the best estimate of such distribution
shown in Hu et al. (2017, 2022, 2024). Uncertainty ranges from Hu et al. were not applied or analyzed here.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-84
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
Because these data ended in 2021, the ratio from that year was used for 2022 as well. Likewise, ratios from
2008 were used for 1990-2008. Once regional distributions were made in this way, each region's emissions
were distributed to the states within the region by population.
3.4.2.3 Uncertainty
The overall uncertainty associated with the 2020 national estimates of HFC emissions as ODS
substitutes was calculated using a Monte Carlo analysis. As described further in Chapter 4, Section 4.25 of
the national Inventory (EPA 2024), the uncertainty of national emissions was -4.1%/+15.1% for a 95%
confidence interval. State-level estimates are expected to have a higher uncertainty because of the use of
population by state or territory during certain steps of the methodology, as described above, and from the
use of atmospheric inversions to apportion emissions of four HFCs by state.
This analysis did not calculate the specific activity data and emissions factor (and importantly for this
category, the reuse of chemicals not emitted) at each state and how the national activity data and emissions
factors could vary based on conditions other than population for the different end uses that comprise the
sector. For this reason, the division of emissions by sector (e.g., refrigeration and air conditioning, foams) are
provided at the state level under the same apportionment as used in the national emission estimates. The Hu
et al. (2017, 2022, 2024) papers used in these state-level emissions estimates show that certain HFC
emissions do not distribute evenly by population; hence, the steps of this methodology that use population
distributions introduce uncertainty. In addition to the uncertainty introduced from population distributions,
use of the Hu et al. work introduces uncertainty into the state-level estimates in two basic ways. First, there
is uncertainty in the regional emissions estimated from atmospheric inversions, as described in the papers;
such uncertainties would extrapolate through to the regional apportionment of HFC-32, HFC-125, HFC-134a,
and HFC-143a calculated duringthe state-level estimate approach. Secondly, the Hu et al. analyses are
limited in scope in both geography and time. Because their results cover only the contiguous 48 states and
the District of Columbia, uncertainty from the population distribution described above exists outside that
area and again when distributing emissions to states within each of the six regions from the Hu et al. work.
The time frame of the Hu et al. analysis is 2008-2021, so extrapolation before and after that time frame
introduces additional uncertainty.
3.4.2.4 Recalculations
No recalculations were applied to the state disaggregation method for this current report. Changes that
resulted from recalculations to the state-level estimates are the same as those presented in Section 4.25 of
the national Inventory, given that improvements in the national Inventory will lead directly to improvements in
the quality of state-level estimates as well.
3.4.2.5 Planned Improvements
This approach of combining population and atmospheric measurement information can be improved in
several ways in future publications of this annual data. First, atmospherically derived emissions estimates
similar to those from Hu et al. (2017) for additional years, primarily after 2014, were incorporated using data
from Hu et al. (2022, 2024), and similar updates are anticipated. Further extension of these data, when
available, can then be used to redistribute the annual emissions after 2021. Also, although emissions derived
from atmospheric measurements were not available before 2008, looking at the trends, if any, in the data can
show if a back-year extrapolation of the data would give better results than applying the earliest year ratios
back to 1990. The Hu et al. (2017, 2022, 2024) data also include information for HFC-227ea and HFC-365mfc.
While the emissions of these chemicals are much lower than the four HFCs used here, the same approach
could be used. It might also be appropriate to use ODS information as a proxy for other HFCs. For instance,
the Hu et al. (2017) paper found that emissions of CFC-11, HCFC-141b, HCFC-142b and HFC-365mfc
3-85
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
showed regional distributions expected based on their primary use as a blowing agent for insulating foam.
These data sets could be used to distribute HFC-245fa and HCFO-1233zd(E) emissions, because these two
chemicals are also used primarily in foams, noting that such foam use in household refrigerator foam and
commercial refrigeration foam is unlikely to be affected by regional weather patterns.
Other improvements could be made by combining more bottom-up information to distribute national
emissions to states or to derive separate state-level emissions estimates. Data on the number of
supermarkets, car registrations, and air conditioning use, or value-added data in representative sectors,
could all apply directly to modeled end uses. Other data could be used as a proxy for end uses, such as
commercial real estate square footage as a proxy for commercial air conditioning.
3.4.2.6 References
CARB (California Air Resources Board) (2016) California's High Global Warming Potential Gases Emission
Inventory: Emission Inventory Methodology and Technical Support Document. Available online at:
https://ww3.arb.ca.gov/cc/inventorv/slcp/doc/hfc inventory tsd 20160411.pdf.
EIA (U.S. Energy Information Administration) (2018a) Table HC7.7. Air Conditioning in Homes in the Northeast
and Midwest Regions, 2015. U.S. Department of Energy. Available online at:
https://www.eia.gOv/consumption/residential/data/2015/hc/php/hc7.7.php.
EIA (2018b) Table HC7.8. Air Conditioning in Homes in the South and l/l/esf Regions, 2015. U.S. Department of
Energy. Available online at: https://www.eia.gOv/consumption/residential/data/2015/hc/php/hc7.8.php.
EIA (2023) Use of Energy Explained: Energy Use in Homes. U.S. Department of Energy. Available online at:
https://www.eia.gov/energyexplained/use-of-energy/electricity-use-in-homes.php.
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
Hu, L., S.A. Montzka, S.J. Lehman, D.S. Godwin, B.R. Miller, A.E. Andrews, K. Thoning, J.B. Miller, C. Sweeney,
C. Siso, J.W. Elkins, B.D. Hall, D.J. Mondeel, D. Nance, T. Nehrkorn, M. Mountain, M.L. Fischer, S.C.
Biraud, H. Chen, and P.P.Tans (2017) Considerable Contribution ofthe Montreal Protocol to Declining
Greenhouse Gas Emissions from the United States. Geophysical Research Letters, 44(15): 8075-8083.
Available online at: https://doi.org/10.1002/2017GLQ74388.
Hu, L., S.A. Montzka, E.J. Dlugokencky, C. Sweeney, K.W. Thoning, B.D. Hall, D. Ottinger, D. Godwin, L.
Western, L. Bruhwiler, A. Andrews, I.J. Vimont, J.D. Nance, S.I. Espinosa, S.M. Miller, S. Bogle, P. DeCola,
and C.D. Nevison (2022) U.S. Non-C02 Greenhouse Gas Emissions for 2007-2020 Derived from
Atmospheric Observations. American Geophysical Union Fall Meeting, Chicago, Illinois, December 12-
16, 2022. Abstract available online at: https://agu.confex.com/agu/fm22/meetingapp.cgi/Paper/1112748.
Hu, L, S. Montzka, I. Vimont, I., and G. Dutton (2024) US Emission Tracker for Potent GHGs. NOAA Global
Monitoring Laboratory. Available online at: https://doi.org/10.15138/ECY2-RX35.
Montzka, S., L. Hu, P. DeCola, D. Godwin, I. Vimont, B. Croes, T. Kuwayama, G. Dutton, D. Nance, B. Hall, C.
Sweeney, and A. Andrews (2023) Making Best Use of Atmosphere- and Inventory-Based Approaches for
Quantifying and Understanding Emissions of Greenhouse Gases and Ozone-Depleting Substances on a
Range of Spatial Scales. Abstract EGU23-10714. EGU General Assembly 2023, Vienna, Austria, April 24-
28, 2023. Available online at: https://doi.org/10.5194/egusphere-egu23-10714.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-86
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
U.S. Census Bureau (2021) Table A. Apportionment Population, Resident Population, and Overseas
Population: 2020 Census and2010 Census. Accessed April 27, 2021. Available online at:
https://en.wikipedia.org/wiki/List of states and territories of the United States by population#cite note
Census2020-8.
U.S. Department of Energy (2016) Insulation. Available online at: http://energy.gov/energysaver/insulation.
3.4.3 Electrical Equipment (NIR Section 4.26)
3.4.3.1 Background
The section describes methods used to estimate state-level SFs emissions consistent with the national
Inventory. Fugitive emissions of SFs can escape from gas-insulated substations and switchgear through
seals, especially from older equipment. The gas can also be released during equipment manufacturing,
installation, servicing, and disposal. These emissions occur in all 50 states and have also been estimated for
three territories (Puerto Rico, the Virgin Islands, and Guam).
3.4.3.2 Methods/Approach (Electrical Equipment)
As discussed in Chapter 4, Section 4.26 (on page 4-168) of the national Inventory {EPA 2024), EPA used a
combination of IPCCTier 2, Tier 3, and country-specific methods to estimate national SFs emissions from
Electrical Equipment.
The national Inventory uses facility-level data reported to the GHGRP or the SFs Emission Reduction
Partnership for Electric Power Systems combined with information on total transmission miles in the US to
develop SFs emission estimates from electrical equipment used for electricity transmission and distribution.
However, facilities, as defined in the GHGRP or the Partnership, in the electrical equipment sector, often
cross multiple states. Thus, Approach 2 as described in the Introduction was used to estimate emissions
from electrical equipment. To disaggregate emissions by state for electrical equipment, EPA used data
sources from the GHGRP and Homeland Infrastructure Foundation Level Data (HIFLD) (U.S. Department of
Homeland Security 2019, 2020, 2021, 2022). For years prior to 2011 before GHGRP data were available,
state-level SFs emissions from electrical equipment were determined by applying the percentage of the total
U.S. transmission miles for each state to the total U.S. emissions estimate for the entire time series,
modified to include additional state-level or facility-level information in the years it is available. For 2011 and
later, the method was modified as described below to first allocate emissions to states as reported to the
GHGRP if the facility only reported one state or if the facility reported multiple states and there was a
reasonable match between the states and total transmission miles reported to the GHGRP and reported by
HIFLD, before applyingto the above method to remainingtransmission miles. See Table 3-20 for a summary
of methods across the time series.
Table 3-20. Summary of Approaches to Disaggregate the National Inventory for Electrical
Equipment Across Time Series
Time Series
Range
Summary of Method
2011-2022
For all GHGRP reporters that reported having transmission miles in only one
state (according to RY 2017-RY 2022 reports, excluding California), their facility-
reported emissions and transmission miles were allocated to that state
(Approach 1).
For GHGRP reporters that had transmission miles in multiple states and had a
reasonable match between the states and total transmission miles reported to
the GHGRP and reported by HIFLD, facility-reported emissions and transmission
3-87
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Time Series
Range
Summary of Method
miles were allocated to each state in which their facility lies by the percentage of
their transmission miles in each state according to HIFLD (Approach 2).
Emissions for California were obtained from the California Air Resources Board
California (CARB) GHG Emission Inventory for 2011-2021 (Approach 2).
However, for 2022 and in cases where CARB's estimate is smaller than the
GHGRP reported emissions plus emissions estimated for non-reporting
facilities, EPA used the GHGRP reported emissions plus the non-reporting
facilities estimate.
The remaining emissions from the national Inventory were allocated to states by
calculating the percentage of remaining transmission miles by state (adjusted
state transmission miles/adjusted national transmission miles). These state
percentages were then applied to the adjusted national emissions estimate
(national emissions excluding GHGRP single-state emissions, emissions from
matched multi-state facilities and California emissions). State transmission
miles were obtained from HIFLD data (2022) and scaled using the transmission
mile growth rate from UDI data sets (Approach 2).
1990-2010
Emissions from the national Inventory were allocated to states by calculating
the percentage of transmission miles by state. These state percentages were
then applied to the national emissions estimate. State transmission miles were
obtained from HIFLD data (2019) for all states. State percentages of the total
transmission were held constant at the 2019 percentage for all states (Approach
2).
For disaggregating national ET&D estimates, state emissions (gas) were determined by multiplying the
percentage of the total U.S. transmission miles for each state by the total national estimate from the
Inventory for the entire time series. U.S. transmission miles were obtained from the U.S. Department of
Homeland Security data from Homeland Infrastructure FoundationLevel Data (HIFLD) (U.S. Department of
Homeland Security 2023), which was last updated September 2023. The data set includes mileage of
transmission lines operated at relatively high voltages varying from 3 kV up to 765 kV. Geographic coverage
includes the United States and the U.S. territories.35
The fraction of transmission miles greater than 34.5 kV in each state was calculated using geographic
information system (GIS) mapping. Figure 3-1 displays the GIS mapping of the transmission lines by state.
Geographic software that identifies lines within state boundaries was used for the disaggregation because it
removed the task of identifying and addressing changes to ownership of service territories as part of this
methodology.
35 Transmission miles greater than 34.5 kv in 2020 totaled 734,291 miles based on the HIFLD data set and 749,847 miles
based on the UDI data set and GHGRP-reported transmission mileage. Despite the discrepancy, HIFLD data provide the
closest match of total miles compared to other data sets previously examined, which gives us reasonable confidence on
using the percentage breakdown by state that can be obtained using GIS mapping.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-88
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
Figure 3-1. U.S. Transmission Lines Separated by State Using GIS Processing Tool
As described below, this method was modified to include additional state-level or facility-level
information in the years for which it was available.
For 2011-2021, CARB provides emissions of SFs from California's electric power systems as reported
through the Regulation for Reducing Sulfur Hexafluoride Emissions from Gas Insulated Switchgear for 2011
2021 (CARB 2021, 2023). EPA concluded that these reported values were a more accurate representation of
state-level emissions from California. However, CARB estimates are not used in two cases: (1) for 2022,
because they are not available yet, and (2) for 2015 and 2016, when CARB's estimates are lower than
estimates from GHGRP and for non-reportingfacilities, as it is assumed that the GHGRP plus non-reporting
facilities estimates better capture emissions from non-reporting facilities in these cases. To estimate
emissions for all other states and territories, EPA removed California from the total transmission miles and
adjusted the percentage breakdown of transmission miles by state accordingly. State and territory emissions
were then disaggregated using the revised percentages.
For 2011-2022, for all GHGRP reporters that reported having transmission miles in only one state
(according to RY 2017-RY 2022 reports), their facility-reported emissions and transmission miles were
allocated to that state. Approximately 72% of reporting facilities had transmission miles in only one state
during RY 2017-RY 2022. On average, these facilities constituted approximately 15% of the national
emissions between 2011 and 2022. Emissions from GHGRP reporters that reported having transmission
miles in multiple states were allocated to the states reported by percentage of transmission miles in each
state according to HIFLD if the GHGRP facility could be cross-walked to the HIFLD data by state and total
transmission miles. Approximately 11 % of reporting facilities had transmission miles in multiple states
during RY 2017-RY 2022 that were successfully cross-walked and matched to the HIFLD data. On average,
these facilities constituted an additional 20% of the national emissions between RY 2017 and RY 2022.
For states where this scenario applied, the GHGRP-reported transmission miles for these facilities were
subtracted from the state transmission mile total, as determined by the HIFLD data, to arrive at an adjusted
3-89
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
total of state transmission miles.36 The sum of GHGRP-reported transmission miles in only one state and the
cross-walked multi-state facilities was also deducted from the total national transmission miles. Because
the HIFLD data represent 2020 transmission miles, transmission mileage was scaled down using UDI's
transmission mile growth rate for 2011-2020 (UDI 2010, 2013, 2017).
Total facility-reported emissions for cases where a facility's transmission miles are reported in only one
state and for multi-state facilities that were cross-walked with the HIFLD data were summed and subtracted
from the national emissions estimate.37 To allocate the remaining national emissions by state, the
percentage transmission miles by state was calculated (adjusted state transmission miles/adjusted national
transmission miles). These state percentages were then applied to the adjusted national emissions estimate
(national emissions excluding GHGRP-only one state emissions and California emissions).
Finally, state-level emissions for GHGRP-reported facilities that reported as being located in only one
state (where applicable) were summed with the calculated state-level emissions based on the calculation
above to arrive at a total state emissions estimate for electric power systems.
The approach taken to disaggregate national emissions enables EPA to use facility-level emissions data
from the reporting program starting in 2011. While this approach has limitations, it also sets up the emissions
estimations for future improvements as more data become available (e.g., additional facility-level
information on state locations of transmission lines obtained through research or additional reportingwould
facilitate greater use of GHGRP data). Additionally, using reported data for California better represents
impacts of regulations on emissions in that state (e.g., California). Similarly, using data reported to EPA can
help account for any state-influenced actions (e.g., climate action planning at state and local levels).
Total emissions from 1990-1999 were disaggregated using the percentage breakdown of transmission
miles by state from the HIFLD data.
3.4.3.3 Methods/Approach (Manufacture of Electrical Equipment)
Emissions were reported by facility for 2011-2022. EPA determined state-level emissions using
Approach 1 based on reported facility locations, which included Connecticut, Illinois, Mississippi, and
Pennsylvania. In the absence of additional industry information, EPA used Approach 2 and assumed that all
non-reporting facilities are located in the same states as reporting facilities. EPA estimates that GHGRP
reporters represent about 50% of all original equipment manufacturers (OEM) emissions and for state-level
estimates, applied the national scale-up factor at the state level.
Foryears prior to when GHGRP data were reported, usingApproach 2, an average percentage state
breakdown across the reporting time series (RY 2011-2022) was applied to emissions in each year to
calculate state emissions from OEMs before 2011. The methods used are summarized in Table 3-21.
Additional research is required to understand (1) if EPA's assumption about the portion of OEM
emissions covered is accurate and (2) in what states these non-reporting emissions occur. Additionally,
further research is necessary to determine whether the reporting facilities were in operation in all years
before 2011.
36 California transmission miles were removed from the HIFLD transmission miles because the state-reported emissions
were used in lieu of this approach. Therefore, state percentages were calculated out of the total national transmission miles
minus California.
37 The national emissions estimate was adjusted by deducting California's emissions (either CARB-reported or estimates for
GHGRP reporters and non-reporters, whichever was used in a given year).
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-90
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
Table 3-21. Summary of Approaches to Disaggregate the National Inventory for Manufacture of
Electrical Equipment Across Time Series
Time Series
Range
Summary of Method
2011-2022
Emissions reported to the GHGRP were allocated based on reported facility
locations (Approach 1). Non-reporters were assumed to be located in the same
states with emissions allocated at the same state percentage of the total non-
reporting emissions as for the emissions reported to the GHGRP (Approach 2).
1990-2010
Emissions from the national Inventory were allocated to states by applying the
average percentage state breakdown across the GHGRP reportingyears (2011
2020) to national estimate for each year between 1990 and 2010 in the Inventory
(Approach 2).
3.4.3.4 Uncertainty
The overall uncertainty associated with the national Inventory of SFs emissions from electrical
equipment source category were calculated using the 2019 Refinement to the 2006IPCC Guidelines. Partner
reported emissions uncertainty was estimated to be -/+ 10% and GHGRP reporter emissions uncertainty was
estimated to be -/+ 20%. As described further in Chapter 4 of the national Inventory (EPA 2024), levels of
uncertainty in the national estimates in 2022 of the source category were -25%/+25%.
State-level estimates are expected to have a higher uncertainty across the time series due to the use of
HIFLD transmission mileage data to apportion the emissions of facilities that either do not report to the
GHGRP or that operate in multiple states. This allocation method introduces additional uncertainty due to
the potential inaccuracy of transmission mile locations and the variability of emission rates per transmission
mile across reporting facilities. As with the national Inventory, the state-level uncertainty estimates for this
category may change as the understanding of the uncertainty of estimates and underlying data sets and
methodologies improve.
3.4.3.5 Recalculations
No recalculations were applied to the state disaggregation method for this current report. Changes that
resulted from recalculations to the state-level estimates are the same as those presented in Section 4.26 of
the national Inventory (page 4-177), given that improvements in the national Inventory will lead directly to
improvements in the quality of state-level estimates as well.
3.4.3.6 Planned Improvements
EPA plans to incorporate facility-specific reported data from the SFs Emission Reduction Partnership
into the inventory for 1999-2010 based on historical emissions estimates collected under EPA's SFs Emission
Reduction Partnership for Electric Power Systems. EPA will consider smoothing emissions for states where
reported emissions cause an unexpected trend in overall state emissions of SFs. Improvements will be
incorporated as more data becomes available (e.g., additional facility-level information on state locations of
transmission lines obtained through research or additional reporting would facilitate greater use of GHGRP
and/or Partnership data). Additional research into regional or state-level trends will also be conducted to
refine the estimates where possible. Finally, EPA plans to incorporate estimates for additional U.S. territories
and estimate emissions for Guam for all years in the time series.
3-91
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
3.4.3.7 References
CARB (California Air Resources Board) (2021) Regulation for Reducing Sulfur Hexafluoride Emissions from
Gas Insulated Switchgear. California Code of Regulations, Title 17, 95350-95359. Available online at:
https://govt.westlaw.com/calregs/Browse/Home/California/CaliforniaCodeofRegulations7guicM076988205A2
111EC8227000D3A7C4BC3.
CARB (2023) 2000 -2021 California Greenhouse Gas Emission Inventory (2023 Edition). Available at
https://ww2.arb.ca.gov/ghg-inventorv-data.
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
UDI (2010) 2010 UDI Directory of Electric Power Producers and Distributors, 118th Edition. Platts.
U Dl (2013) 2013 UDI Directory of Electric Power Producers and Distributors, 121st Edition. Platts.
UDI (2017) 2017 UDI Directory of Electric Power Producers and Distributors, 125th Edition. Platts.
U.S. Department of Homeland Security (2019) Homeland Infrastructure Foundation-Level Data (HIFLD).
Accessed March 2021. Available online at: https://hifld-
geoplatform.hub.arcgis.com/maps/bd24d1 a282c54428b024988d32578e59
U.S. Department of Homeland Security (2020) Homeland Infrastructure Foundation-Level Data (HIFLD).
Accessed October 2021. Available online at: https://hifld-
geoplatform.hub.arcgis.com/maps/bd24d1 a282c54428b024988d32578e59
U.S. Department of Homeland Security (2021) Homeland Infrastructure Foundation-Level Data (HIFLD).
Accessed September 2022. Available online at: https://hifld-
geoplatform.hub.arcgis.com/maps/bd24d1 a282c54428b024988d32578e59
U.S. Department of Homeland Security (2022) Homeland Infrastructure Foundation-Level Data (HIFLD).
Accessed September 2023. Available online at: https://hifld-
geoplatform.hub.arcgis.com/maps/bd24d1 a282c54428b024988d32578e59
3.4.4 SF6 and PFCs from Other Product Use (NIR Section 4.27)
3.4.4.1 Background
SFs and PFC emissions result from other product use and other processes, including military and
scientific applications. Many of these applications utilize SFs or PFCs to exploit their unique chemical
properties, such as the high dielectric strength of SFs and the stability of PFCs. Emission profiles from these
processes may vary greatly, ranging from immediate and unavoidable release of all the chemical to largely
avoidable, delayed release from leak-tight products after decades of use.
Military applications employ SFs and PFCs in many processes. For example, SFs is used in the radar
systems commonly known as Airborne Warning and Control Systems (AWACS)of military
reconnaissance planes of the Boeing E-3A type. Other uses of SFs in military applications include the
oxidation of lithium in navel torpedoes and infrared decoys. SFs has also been documented for use in the
quieting of torpedo propellers, and it is also a byproduct of the processing of nuclear material for the
production of fuel and nuclear warheads.
Military electronics are believed to be a key application for PFC heat transfer fluids, particularly in areas
such as ground and airborne radar avionics, missile guidance systems, and sonar. PFCs may also be used to
cool electric motors, especially for equipment where noise reduction is a priority (e.g., submarines).
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-92
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
SFs and PFCs are also employed in several scientific applications, such as for use in particle
accelerators. Particle accelerators can be found in university and research settings, as well as in industrial
and medical applications. SFs is typically used as an insulatinggas and is operated in a vessel exceeding
atmospheric pressure. PFCs (particularly PFC-14) may also be used in particle accelerators as particle
detectors or counters (Workman et al. 2022).SFs may also be employed in other high-voltage scientific
equipment, including lasers, x-rays, and electron microscopes.
There is a range of unidentified processes (such as R&D activities) that also use SFs and PFCs. PFCs are
likely used primarily as HTFs. Emissions are reported for these unknown activities under "Other Scientific
Applications."
3.4.4.2 Methods/Approach
National emissions were based primarily on data reported though the Federal Energy Management
Program (FEMP) by the U.S. Department of Energy (DOE) and Department of Defense (DOD), with
methodologies from the IPCC used to make additional emission estimates where FEMP data were not readily
available(DOE 2022, IPCC 2006). Military application and scientific application emissions were estimated
separately using different approaches as discussed in Chapter 4, Section 4.27 (on pages 4-178 through 4-
183) of the national Inventory (EPA 2024). In general, EPA used a hybrid approach to disaggregate national
estimates.
3.4.4.2.1. Military Applications
AWACS emissions from the national Inventory were allocated to states based on the distribution of the
U.S. AWACS fleet of 33 planes. Alaska and Oklahoma were the only two states assumed to have E-3 planes in
the U.S. AWACS fleet, with four planes and 29 planes, respectively, throughout the entire time series.
National emissions from other military applications throughout the time series were disaggregated by
equal allocation to all states due to a lack of state-level data.
3.4.4.2.2. Scientific Applications
National Inventory particle accelerator emissions were allocated to states in which particle
accelerators are operating. State-level emissions from non-DOE research and industrial particle
accelerators in the United States were calculated using facility-level emissions estimated by applying an
average SFs charge and emission factor based on the particle accelerator type.
Reported emissions from DOE particle accelerators were disaggregated equally among the nine states
in which they are operating (i.e., California, Illinois, Maryland, Massachusetts, New Mexico, New York,
Tennessee, Virginia, and Washington). Emissions from DOE tandem accelerators were disaggregated equally
amongthe states (i.e., New Mexico, California, New York, and Washington) with tandem accelerators
located at their facility, and emissions from DOE ion beam accelerators and gas purging (i.e., at Argonne National
Lab, Oak Ridge National Lab, and Brookhaven National Lab) were disaggregated equally among the states in which
those particle accelerators are located (i.e., Illinois, Tennessee, and New York, respectively).
Emissions from other scientific applications reported by DOE were similarly allocated equally to each of
the nine states with DOE particle accelerators listed above.
3.4.4.3 Uncertainty
The overall uncertainty associated with the national emissions estimates of SFs and PFCs from other
product use was calculated using the 2019 Refinement to the 2006 IPCC Guidelines (IPPC 2019). As
described further in Chapter 4 of the national Inventory, levels of uncertainty in the national estimates in
2022 were -36%/+38% across the industry.
3-93
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
State-Level estimates are expected to have a higher uncertainty because, in some cases, the national
estimates were apportioned to each state equally. This assumption was required because of a general lack
of more granular state-level data.
3.4.4.4 Recalculations
This is a new category included for both the current (i.e., 1990-2022) national Inventory and state-level
estimates, and therefore no recalculations were performed.
3.4.4.5 Planned Improvements
EPA plans to revisit the methodology for determining emissions of SFs and PFCs from other product
usein particular, the assumptions that emissions from other military applications (i.e., non-AWACS) are
consistent across all states and that emissions from DOE particle accelerators are consistent across all nine
states with DOE particle accelerators. Planned improvements also include developing a more complete list
of states with DOE facilities for purposes of disaggregating emissions from other scientific applications
reported by DOE. Additional collaboration with DOE and DOD will be required to confirm or modify the
assumptions regardingthe distribution of emissions across states.
3.4.4.6 References
DOE (U.S. Department of Energy) (2022) Federal Comprehensive Annual Energy Reporting Requirements.
Available online at: https://www.energv.gov/femp/federal-comprehensive-annual-energy-reporting-
requirements.
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-
2022. EPA 430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-
gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
IPCC (2019) 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories. E.C.
Buendia, K. Tanabe, A. Kranjc, J. Baasansuren, M. Fukuda, S. Ngarize A. Osako, Y. Pyrozhenko, P.
Shermanau, and S. Federici (eds.). Available online at: https://www.ipcc.ch/report/2019-refinement-to-
the-2006-ipcc-guidelines-for-national-greenhouse-gas-inventories/.
Workman, R.L. et al. (Particle Data Group) (2022) 35. Particle Detectors at Accelerators. In: The Review of
Particle Physics. Progress in Theoretical and Experimental Physics, 2022(8): 083C01. Available online at:
https://pdg.lbl.gov/2022/reviews/rpp2022-rev-particle-detectors-accel.pdf.
3.4.5 Nitrous Oxide from Product Uses (NIR Section 4.28)
3.4.5.1 Background
N2O is primarily used in carrier gases with oxygen to administer more potent inhalation anesthetics for general
anesthesia, and as an anesthetic in various dental and veterinary applications. The second main use of N2O is as a
propellant in pressure and aerosol products, the largest application being pressure-packaged whipped cream.
Smaller quantities of N2O also are used in the following applications: oxidizing agent and etchant used in
semiconductor manufacturing, oxidizing agent used with acetylene in atomic absorption spectrometry, production
of sodium azide for use in airbags, fuel oxidant in auto racing, and oxidizing agent in blowtorches used by jewelers
and others. The amount of N2O that is actually emitted depends on the specific product use or application. Only
the medical/dental and food propellant subcategories were assumed to release emissions into the atmosphere
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-94
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
that are not captured under another source category; therefore, these subcategories were the only usage
subcategories with emissions rates. N2O product use emissions from the national Inventory were disaggregated
across all 50 states, the District of Columbia, and U.S. territories in 2022.
3.4.5.2 Methods/Approach
The state-Level methodology for N20 emissions from product usage is to allocate emissions to all
applicable U.S. states and territories using population statistics as a surrogate for state-specific N20 usage,
consistent with Approach 2 as defined in the Introduction to this report. See Appendix I, Table 1-1 in the "N20
Use" Tab, for more details on the N20 product use categories and their assumed emissions factors and
Appendix G, Table G-1 in the "Population Data" Tab, for details on the population data used. The national
Inventory methodology was adapted to calculate state-level GHG emissions of N20 to ensure consistency
with national estimates. National estimates were used to disaggregate emissions by state because of
limitations in the availability of state-specific data for the time series. Total emissions for each state are the
sum of emissions from N20 product use.
State-level emissions of N20 usage for medicine/dental anesthesia, sodium azide production, food
processing propellant and aerosols, and other applications (e.g., fuel oxidant in auto racing, oxidizing agent
in blowtorches) were calculated using the same methodology in the national Inventory to calculate national
emissions (EPA 2024). Data on the usage of N20 by state, however, are not available. To calculate N20
product usage by state, national N20 usage and emissions were distributed among the 50 states, the District
of Columbia, and U.S. territories (including Puerto Rico, American Samoa, Guam, the Northern Mariana
Islands, and the U.S. Virgin Islands) using U.S. population statistics as a surrogate for state-specific N20
usage (U.S. Census Bureau 2002, 2011, 2021, 2022a, 2022b; Instituto de Estadfsticas de Puerto Rico 2021).
For each year in the 1990-2022 time series, the fraction of the total U.S. population in each state, as well as
the District of Columbia and U.S. territories, was calculated by dividing the state population by the total U.S.
population.
To estimate N20 emissions for each year by state, total national Inventory N20 production was
multiplied by the share of the national usage and emissions rate for each respective application and then
multiplied by each state's fraction of the total population for that year. The calculated emissions by
application and by state were then summed by state. Using state populations to calculate the N20 use and
emissions by state assumed that N20 use is consistent across all states.
3.4.5.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of N2O from N2O product use was
calculated using the 2006 IPCC Guidelines Approach 2 methodology for uncertainty (IPCC 2006). As described
further in Chapter 4 and Annex 7 of the national Inventory (EPA 2024), levels of uncertainty in the national
estimates in 2022 were -24%/+24% for N2O.
State-level estimates are expected to have a higher uncertainty because the national emissions estimates
were apportioned to each state based solely on state population for some subcategories. This assumption was
required because of a general lack of more granular state-level data. Using state population for medical/dental
anesthesia and for food propellant in the state-level estimates may have lower uncertainty because these uses
tend to be related to population. Using state population for other uses (e.g., fuel oxidant in auto racing, oxidizing
agent in blowtorches) introduces higher uncertainty because state-level activities are not known and less likely to
be related to population. This allocation method introduces additional uncertainty due to limited data on the
quantity of N2O used by state or nationally for the full time series. The sources of uncertainty for this category are
also consistent over time because the same surrogate data are applied across the entire time series.
3-95
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
3.4.5.4 Recalculations
Recalculations were performed for 2020-2022 as updated population data for those years were made
available from the U.S. Census Bureau. The updated population data had a negligible impact on the
emissions estimated for the 50 states, the District of Columbia, and Puerto Rico due to the low emissions
estimated for each state or territory for the sector.
3.4.5.5 Planned Improvements
EPA recently initiated an evaluation of alternative production statistics for cross-verification and
updating time series activity data, emission factors, assumptions, and more, and a reassessment of N20
product use subcategories that accurately represent trends. This evaluation includes conducting a literature
review of publications and research that may provide additional details on the industry. This work remains
ongoing, and thus far no additional data sources have been found to update this category.
Pending additional resources and planned improvement prioritization, EPA may also evaluate
production and use cycles, and potentially need to incorporate a time lag between production and ultimate
product use and resulting release of N20. Additionally, planned improvements include considering imports
and exports of N20 for product uses.
Finally, for future inventories, EPA will examine data from the GHGRP to improve the emission estimates
for the N20 product use subcategory. Particular attention will be made to ensure aggregated information can
be published without disclosing CBI and time series consistency, as the facility-level reporting data from
EPA's GHGRP are not available for all inventory years as required in this state-level inventory. This is a lower
priority improvement, and EPA is still assessing the possibility of incorporating aggregated GHGRP CBI data
to estimate emissions; therefore, this planned improvement is still in development and not incorporated in
the current Inventory report.
3.4.5.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
Instituto de Estadfsticas de Puerto Rico (2021) EstimadosAnuales Poblacionales de los Municipios Desde
1950. Accessed February 2021. Available online at: https://censo.estadisticas.pr/EstimadosPoblacionales.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
U.S. Census Bureau (2002) Table CO-EST2001-12-00. In: Time Series of I ntercensal State Population
Estimates: April 1, 1990 to April 1,2000. Release date: April 11, 2002. Available online at:
https://www2.census.gov/programs-survevs/popest/tables/1990-2000/intercensal/st-co/co-est2Q01-12-
OO.pdf.
U.S. Census Bureau (2011) Table ST-EST00INT-01. In: Intercensal Estimates of the Resident Population for
the United States, Regions, States, and Puerto Rico: April 1, 2000 to July 1, 2010. Release date:
September 2011. Available online at: https://www2.census.gov/programs-survevs/popest/datasets/200Q-
2010/intercensal/state/st-est00int-alldata.csv.
U.S. Census Bureau (2021) Table NST-EST2020. In: Annual Estimates of the Resident Population for the
United States, Regions, States, the District of Columbia, and Puerto Rico: April 1,2010 to July 1,2019;
April 1,2020; and July 1, 2020. Release date: July 2021.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
3-96
-------
Section 3 Industrial Processes and Product Use (NIR Chapter 4)
U.S. Census Bureau (2022a) International Database: World Population Estimates and Projections.
Population of U.S. territories. Accessed November 23, 2022. Available online at:
https://www.census.gov/programs-survevs/international-programs/about/idb.html.
U.S. Census Bureau (2022b) Table NST-EST2022-POP. In: Annual Estimates of the Resident Population for the
United States, Regions, States, District of Columbia, and Puerto Rico: April 1, 2020 to July 1, 2022.
Release date: December 2022.
3-97
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
4 Agriculture (NIR Chapter5)
For this methodology report, the Agriculture chapter consists of two subsectors: livestock management
and other agriculture activities. More information on national-level emissions and methods is available in
Chapter 5 of the national Inventory, available online at: https://www.epa.gov/svstem/files/documents/2024-
04/us-ghg-inventorv-2024-chapter-5-agriculture.pdf. Table 4-1 summarizes the different approaches used to
estimate state-level agriculture emissions. The sections below provide more detail on each category.
Table 4-1. Overview of Approaches for Estimating State-Level Agriculture Sector GHG Emissions
Category
Gas
Approach
Completeness8
Enteric Fermentation
ch4
Approach 1
Includes emissions from all states
and tribal lands.8
Manure Management
ch4,
n2o
Approach 1
Includes emissions from all states
and tribal lands.8
Agricultural Soil
Management
n2o
Hybrid: 4.2.2.2
1990-2017: Approach 1
2018-2022: Approach 2
Includes emissions from all states,
the District of Columbia, tribal
lands, and territories.8 Some
components of Alaska and Hawaii
were not estimated.
Rice Cultivation
ch4
Hybrid:
1990-2020: Approach 1
2021-2022: Approach 2
Includes emissions from all 13
states (and tribal lands) cultivating
rice.8
Liming
co2
Hybrid:
1990-2021: Approach 1
2022: Approach 2
Includes emissions from all states
(and tribal lands) for which USGS
(through Minerals Yearbook and the
Mineral Industry Survey) reports
limestone and dolomite
consumption for agriculture in
current and historical yearbooks
and surveys.8
Urea
co2
Approach 1
Includes emissions from all states
and territories8 (i.e., Puerto Rico).
Field Burning of
Agricultural Residues
ch4,
n2o
Hybrid:
1990-2014: Approach 1
2015-2022: Approach 2
Sugarcane: 1990-2020
(Approach 1)
Sugarcane: 2021-2022
(Approach 2)
Includes emissions from all states
except Alaska and Hawaii.8
" Emissions are likely occurring in other U.S. territories; however, due to a lack of available data and the nature of this
category, this analysis includes emissions for only the territories indicated. Territories not listed are not estimated. See
planned improvements discussions across Chapter 5 of the national Inventory. Includes tribal areas in the conterminous
United States.
4.1 Livestock Management
This section presents the methodology applied to estimate the livestock management emissions, which
consist of the following sources:
Enteric fermentation (CH4)
4-1
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Manure management (CH4, N20)
4.1.1 Enteric Fermentation (NIR Section 5.1)
4.1.1.1 Background
Methane is produced as part of normal digestive processes in animals. During digestion, microbes that
reside in an animal's digestive system ferment food consumed by the animal. This microbial fermentation
process, referred to as enteric fermentation, produces CH4as a byproduct, which can be exhaled or
eructated by the animal. The amount of CH4produced and emitted by an individual animal depends primarily
upon the animal's digestive system, and the amount and type of feed it consumes.
4.1.1.2 Methods/Approach
EPA compiles state-level CH4 emissions from enteric fermentation using the same methods applied in
the national Inventory. The methods applied in the national Inventory are summarized below in Table 4-2.
Estimates are available for all 50 states. Territories are not currently estimated, and tribal lands are not
explicitly included based on USDA survey practices, which depend on the presence of the animal on the
farm/operation, not the geographic area.
Table 4-2. Approaches to Estimate Enteric Fermentation Methane Across Time Series
Time Series
Range
Method
1990-2022
Cattle: IPCC Tier 2 (Cattle Enteric Fermentation Model)
Non-cattle: IPCCTier 1 (population * default emissions factor)
Please refer to Section 5.1 and Annex 3.10 the national Inventory on enteric fermentation for details on
the methods applied to estimate state-level emissions for the years 1990-2022 (EPA 2024). Below is a
summary:
For cattle, the Cattle Enteric Fermentation Model (CEFM) was used to estimate CH4 emissions using
the IPCCTier 2 method. The CEFM utilizes the IPCCTier 2 method and other analyses of cattle
population, feeding practices, diet data, and production characteristics.
For non-cattle animals, USDA state population estimates (from USDA QuickStats and the U.S.
Census of Agriculture) were multiplied by the corresponding default IPCC emissions factors (IPCC
2006).
Data Appendix E-1 to this report provides state-level non-cattle livestock population numbers for all
inventory years. These population data serve as the activity data that are multiplied by default IPCC
emission factors to estimate CH4 emissions from enteric fermentation.
Data Appendix E-2 to this report provides state-level cattle population numbers disaggregated by
animal type for all inventory years.
To allow for greater exploration of the underlying data that support cattle enteric fermentation
emissions estimates, state-level implied emission factors for all cattle types across the time series
are provided in Data Appendix E-3 to this report. These implied emission factors are calculated post-
hoc from the CEFM output where emissions estimates are modeled based on data inputs regarding
livestock populations, diet attributes, feeding practices, and production characteristics. The
resulting enteric fermentation emissions estimates were divided by cattle population numbers to
calculate the implied emission factor that describes average CH4 produced per head of cattle in
each state in a given year.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
4-2
-------
Section 4 Agriculture (NIR Chapter 5)
4.1.1.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of CH4 from enteric fermentation
was calculated using the 2006 IPCC Guidelines Approach 2 methodology (IPCC 2006). As described further
in Chapter 5 of the national Inventory (EPA 2024), levels of uncertainty in the national estimates in 2022 were
-11%/+18% for CH4. State-level estimates have a higher uncertainty due to apportioning the national or
default emission estimates to each state. This approach does not address state-level differences in
uncertainty when applying regional diet data or factors. It is important to note that beef and dairy cattle diets
can vary significantly even between states that are in similar regions because of the wide variety of forage
types being grown on range and pasture land. Additionally, producers often develop unique feed for their
livestock based on the availability of specific feed inputs in their area. Regionally derived data were applied at
the state level because state-level data were limited or unavailable for many parameters. For more details on
national-level uncertainty, see the uncertainty discussion in Section 5.1 of the national Inventory.
4.1.1.4 Recalculations
Changes that resulted from recalculations to the state-level estimates are the same as those presented
in Section 5.1 of the national Inventory (page 5-10), given that improvements in the national Inventory will
lead directly to improvements in the quality of state-level estimates as well. In particular, consistent with the
national Inventory, EPA updated the 2021 estimates that had been previously calculated using a simplified
method to use the complete method consistent with the full time series.
4.1.1.5 Planned Improvements
Planned improvements to the state-level estimates are the same as those presented in Section 5.1 of
the national Inventory (page 5-10), given that improvements in the national Inventory will lead directly to
improvements in the quality of state-level estimates as well. In particular, state-level livestock diet data
would be of value for improving estimates of enteric fermentation.
4.1.1.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventory-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
Full citations of references included in Chapter 5.1 (Enteric Fermentation) and Annex 3.10 of the
national Inventory are available online here: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-
inventorv-2024-chapter-10-references O.pdf and https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-
inventorv-2024-annex-3-additional-source-or-sink-categories-part-b.pdf.
4.1.2 Manure Management (NIR Section 5.2)
4.1.2.1 Background
The treatment, storage, and transportation of livestock manure can produce anthropogenic CH4 and
N20 emissions. Methane is produced by the anaerobic decomposition of manure and N20 is produced from
direct and indirect pathways through the processes of nitrification and denitrification, volatilization, and
runoff and leaching. In addition, there are many underlying factors that can affect these resulting emissions
from manure management. For CH4, the type of manure management system, ambient temperature,
4-3
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
moisture, and residency (storage) time of the manure affect bacteria growth and therefore subsequent
emissions. For N20, the composition of the manure (manure includes both feces and urine), the type of
bacteria involved in the process, and the amount of oxygen and liquid in the manure system affect the
resulting emissions.
4.1.2.2 Methods/Approach
EPA compiles state-level emissions from manure management using the same methods applied in the
national Inventory as summarized in Table 4-3. Estimates are available for all 50 states. Territories are not
currently estimated, and tribal lands are not explicitly included based on USDA survey practices, which
depend on the presence of the animal on the farm/operation, not the geographic area.
Table 4-3. Approaches to Estimate Manure Management Methane and N20 Across Time Series
Time Series
Range
Method
1990-2022
Combination of IPCCTier 1 and 2 approaches as described in the national
Inventory.
For 1990-2022, please refer to the national Inventory Chapter 5, Section 5.2 and Annex 3.11, which
provides additional detail on the methods to estimate state-level manure management emissions (EPA
2024). As noted in that section, the basic approach applies a combination of IPCCTier 1 and Tier 2
methodologies. EPA applies Tier 1 default N20 emissions factors and CH4 conversion factors for dry systems
from the IPCC (2006), U.S.-specific CH4 conversion factors for liquid systems, and U.S.-specific values for
the volatile solids production rate and the nitrogen excretion rate for some animal types, including cattle
values from the CEFM (see Section 4.1.1 Enteric Fermentation).
4.1.2.3 Uncertainty
The overall uncertainty associated with the 2022 national estimates of CH4 and N20 from manure
management were calculated using the 2006 IPCC Guidelines Approach 2 methodology (IPCC 2006). As
described further in Chapter 5 of the national Inventory (EPA 2024), levels of uncertainty in the national
estimates in 2022 were -18%/+20% for CH4 and -16%/+24%for N20. State-level estimates have a higher
uncertainty due to apportioning the national or default emission estimates to each state. This approach does
not address state-level differences in uncertainty when applying regional waste management system
distributions or factors. These assumptions were applied because state-level data are limited or unavailable
for many parameters. For more details on national-level uncertainty, see the uncertainty discussion in
Section 5.2 of the national Inventory (EPA 2024).
4.1.2.4 Recalculations
Changes that resulted from recalculations to the state-level estimates are the same as those presented
in Section 5.2 of the national Inventory (page 5-19), given that improvements in the national Inventory lead
directly to improvements in the quality of state-level estimates as well. In particular, consistent with the
national Inventory, EPA updated the 2021 estimates that had been previously calculated using a simplified
method to use the complete method consistent with the full time series. EPA updated waste management
system distribution data for poultry broilers and layers and for beef feedlot animal types and also updated
the direct N20 emission factor for solid storage waste management systems pursuant to guidance in the
IPCC 2019 Refinement to the 2006 IPCC Guidelines. EPA also updated anaerobic digester usage for poultry
manure management and for swine manure management and improved the representation of livestock
characteristics such as calf typical animal mass and urinary energy for feedlot cattle within the CEFM.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
4-4
-------
Section 4 Agriculture (NIR Chapter 5)
4.1.2.5 Planned Improvements
Planned improvements to the state-Level estimates are the same as those presented in Chapter 5,
Section 5.2 of the national Inventory (page 5-20), given that improvements in the national Inventory will lead
directly to improvements in the quality of state-level estimates as well.
4.1.2.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventory-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
Full citations of references included in Chapter 5.2 (Manure Management) and Annex 3.11 of the
national Inventory are available online here: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-
inventorv-2024-chapter-10-references O.pdf and https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-
inventorv-2024-annex-3-additional-source-or-sink-categories-part-b.pdf.
4.2 Other (Agriculture)
This section presents the methodology applied to estimate the other agricultural activity emissions,
which consist of the following source categories:
Rice cultivation (CH4)
Agricultural soil management (N20)
Liming (C02)
Urea fertilization (C02)
Field burning of agricultural residues (CH4, N20)
4.2.1 Rice Cultivation (NIR Section 5.3)
4.2.1.1 Background
Most of the world's rice is grown on flooded fields that create anaerobic conditions, leading to CH4
production through a process known as methanogenesis. Approximately 60% to 90% of the CH4 produced by
methanogenic bacteria in flooded rice fields is oxidized in the soil and converted to C02 by methanotrophic
bacteria. The remainder is emitted to the atmosphere or transported as dissolved CH4 into groundwater and
waterways. Methane is transported to the atmosphere primarily through the rice plants, but some CH4 also
escapes via ebullition (i.e., bubbling through the water) and to a much lesser extent by diffusion through the
water.
4.2.1.2 Methods/Approach
EPA compiles state-level CH4 emissions from rice cultivation using the same methods applied in the
national Inventory. Rice is currently cultivated in 13 states: Arkansas, California, Florida, Illinois, Kentucky,
Louisiana, Minnesota, Mississippi, Missouri, New York, South Carolina, Tennessee, and Texas. This is
described in Chapter 5, Section 5.3 (pages 5-21 through 5-28), of the national Inventory. Additional
information on the methodologies and data is also provided in Annex 3.12.
4-5
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
As described in the national Inventory, the methodology used to estimate CH4 emissions from rice
cultivation is based on a combination of IPCCTier 1 and 3 approaches. The IPCC Tier 3 method utilizes the
DayCent process-based model to estimate CH4 emissions from rice cultivation. DayCent is used to simulate
hydrological conditions and thermal regimes, organic matter decomposition, root exudation, rice plant
growth and its influence on oxidation of CH4, as well as CH4 transport through the plant and via ebullition
(Cheng et al. 2013). This method captures the influence of organic amendments and rice straw management
on methanogenesis in the flooded soils, and ratooning of rice crops with a second harvest during the growing
season. In addition to CH4 emissions, DayCent simulates soil carbon stock changes and N20 emissions and
allows for a seamless set of simulations for crop rotations that include both rice and non-rice crops (EPA
2024).
The IPCC Tier 1 method is applied to estimate CH4 emissions from rice when grown in rotation with
crops that are not simulated by DayCent, such as vegetable crops. The Tier 1 method is also used for areas
converted between agriculture (i.e., cropland and grassland) and other land uses such as forest land,
wetland, and settlements. In addition, the Tier 1 method is used to estimate CH4 emissions from organic
soils (i.e., Histosols) and from areas with very gravelly, cobbly, or shaley soils (greater than 35% by volume).
The Tier 3 method using DayCent has not been fully tested for estimating emissions associated with these
conditions (EPA 2024). The most recent national Inventory includes state-level emissions for the 13 states
mentioned above for the years 1990-2018, which were used for this report (Approach 1). Cultivated rice
areas for the 13 rice-cultivating states are determined from land-use and cropping history information
derived from the National Resources Inventory (NRI) for the 1990-2017 period (USDA 2020), and the time
series is extended from 2018 to 2020 using crop data provided in the USDA National Agricultural Statistics
Service Cropland Data Layer (USDA-NASS CDL) (USDA 2021). Within the national Inventory, EPA does not
currently directly estimate state-level emissions from rice cultivation for the years 2021-2022, so it is not
possible to develop state-level estimates for those years using the same approach. The national-level
emissions for 2021-2022 are estimated using a surrogate data method, and were disaggregated to the state
level in a two-step process for this report (Approach 2). First, the average proportion of the total national
emissions was computed for each state for the years 2018-2020, which are the last three years for which
state-level emissions have been estimated. Second, the state-level proportions were multiplied by the total
national emissions to approximate the emissions occurring in each state from 2021 to 2022. Data Appendix
E-4 to this report lists the total rice cultivated areas of each of the 13 states with rice cultivation across the
1990-2020 time period. State-level rice cultivated areas are disaggregated to show the land area in each
state for which the Tier 3 and Tier 1 methods were used to estimate CH4 emissions from rice cultivation.
State-level total rice harvested areas, which account for land area on which a second rice crop is harvested,
are also provided in Data Appendix E-4 to this report.
4.2.1.3 Uncertainty
The overall uncertainty associated with national estimates of CH4 from rice cultivation was calculated
using the IPCC Approach 2 (i.e., Monte Carlo simulation). As described in Chapter 5 of the national Inventory
(EPA 2024), sources of uncertainty include incomplete information on management practices, uncertainties
in model structure (i.e., algorithms and parameterization), emissions factors, and variance associated with
the NRI sample. Levels of uncertainty in the national CH4 rice cultivation estimates in 2022 were -34%/+34%
of total emissions estimated using the Tier 1 method and -86%/+86% of total emissions estimated using the
Tier 3 method, with a combined uncertainty of -73%/+73% of national CH4 emissions from rice cultivation.
Uncertainty will be greater for the years 2021-2022, where a surrogate data method is used to extend the
time series past the period over which NRI data and direct emissions estimates are available.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
4-6
-------
Section 4 Agriculture (NIR Chapter 5)
4.2.1.4 Recalculations
Changes that resulted from recalculations to the state-level estimates are the same as those presented
in Section 5.3 of the national Inventory (page 5-28), given that improvements in the national Inventory will
lead directly to improvements in the quality of state-level estimates as well. Included in the latest Inventory
were improvements to the characterization of rice cultivated land areas in the land representation activity
data, an extension of crop history data using CDL as described above, and improvements to the
characterization of rice cultivation practices (e.g., ratooning, winter flooding) and inputs (e.g., fertilizer and
organic amendment additions, crop residue inputs).
4.2.1.5 Planned Improvements
Planned improvements to the state-level estimates are anticipated to be the same as those presented
in Section 5.3 of the national Inventory, given that improvements in the national Inventory will lead directly to
improvements in the quality of state-level estimates as well.
4.2.1.6 References
Cheng, K., S.M. Ogle, W.J. Parton, and G. Pan (2013) Predicting Methanogenesis from Rice Paddies Using the
DAYCENT Ecosystem Model. Ecological Modelling, 261-262(Suppl.): 19-31. Available online at:
https://doi.Org/10.1016/i.ecolmodel.2013.04.003.
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990 -2022. EPA 430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
USDA(U.S. Department of Agriculture) (2020) 2017 National Resources Inventory: Summary Report. USDA
Natural Resources Conservation Service and Center for Survey Statistics and Methodology, Iowa State
University. Available online at: https://www.nrcs.usda.gov/sites/default/files/2022-
10/2017NRISummarv Final.pdf.
USDA (2021) CropScape-Cropland Data Layer. Accessed July 2021. Available online at:
https://nassgeodata.gmu.edu/CropScape/.
Full citations of references included in Chapter 5.3 (Rice Cultivation) and Annex 3.12 of the national
Inventory are available online here: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-inventorv-
2024-chapter-10-references O.pdf and https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-inventorv-
2024-annex-3-additional-source-or-sink-categories-part-b.pdf.
4.2.2 Agricultural Soil Management (NIR Section 5.4)
4.2.2.1 Background
N20 is naturally produced in soils through the microbial processes of nitrification and denitrification
that are driven by the availability of mineral nitrogen. Mineral nitrogen is made available in soils through
decomposition of soil organic matter and plant litter, asymbiotic fixation of nitrogen from the atmosphere,
and agricultural management practices, which are discussed below.
Several agricultural activities increase mineral nitrogen availability in soils that lead to direct N20
emissions at the site of a management activity. These activities include synthetic nitrogen fertilization;
application of managed livestock manure; application of other organic materials such as biosolids (i.e.,
treated sewage sludge); deposition of manure on soils by domesticated animals in pastures, range, and
paddocks (PRP) (i.e., unmanaged manure); retention of crop residues (nitrogen-fixing legumes and non-
legume crops and forages); and drainage of organic soils (i.e., Histosols) (IPCC 2006). Additionally,
4-7
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
agricultural soil management activities, including irrigation, drainage, tillage practices, cover crops, and
fallowing of land, can influence nitrogen mineralization from soil organic matter and plant litter, as well as
levels of asymbiotic nitrogen fixation.
Indirect emissions of N20 occur when nitrogen is transported from a site and is subsequently converted
to N20. There are two pathways for indirect emissions: (1) volatilization and subsequent atmospheric
deposition of applied/mineralized nitrogen and (2) surface runoff and leaching of applied/mineralized
nitrogen into groundwater and surface water.
4.2.2.2 Methods/Approach
EPA compiles state-level N20 emissions from Agricultural Soil Management using the same methods
applied in the national Inventory. Please see the methodologies described in Chapter 5, Section 5.4 (pages 5-
28 through 5-47), of the national Inventory.
For this report, a hybrid of Approach 1 and 2 was applied in developing state-level estimates. Estimates
are available for all 50 states and the District of Columbia, with emissions occurring on tribal lands implicitly
captured in the estimates for the states within which these lands occur; however, some components of this
category are not estimated for Alaska and Hawaii, as described in the national Inventory. Specifically, Alaska
and Hawaii only have estimates of N20 emissions that result from applying nitrogen to soils in the form of
biosolid waste and livestock manure, including managed manure and manure deposited onto
pasture/range/paddock, which is not managed. Soil N20 emissions in Hawaii are also estimated from crop
residue additions to soils. Therefore, soil N20 emissions associated with synthetic fertilization,
mineralization of nitrogen from soil organic matter, and drainage of organic soils (i.e., Histosols) are not
estimated for Alaska or Hawaii.
Estimates of N20 emissions from managed croplands and grasslands are not available for Alaska and
Hawaii except for managed manure nitrogen, PRP nitrogen, and biosolid additions for Alaska and managed
manure and PRP nitrogen, biosolid additions, and crop residue for Hawaii.
Additional information on methodologies and data is also provided in Annex 3.12 of the national
Inventory.
4.2.2.3 Uncertainty
The overall uncertainty associated with national estimates of N20 from agricultural soil management is
described in Chapter 5 of the national Inventory (EPA 2024. Uncertainty is estimated for each of the following
five components of N20 emissions from agricultural soil management: (1) direct emissions simulated by
DayCent, (2) the components of indirect emissions (nitrogen volatilized and leached or runoff) simulated by
DayCent, (3) direct emissions estimated with the IPCCTier 1 method, (4) the components of indirect
emissions (nitrogen volatilized and leached or runoff) estimated with the IPCC (2006) Tier 1 method, and (5)
indirect emissions estimated with the IPCCTier 1 method.
Levels of uncertainty in the national N20 agricultural soil management emissions estimates in 2022
were -28%/+28% of the emissions estimate for direct N20 and -51 %/+123% of the emissions estimate for
indirect N20 across all methodologies at the national scale.
4.2.2.4 Recalculations
Changes that resulted from recalculations to the state-level estimates are the same as those presented
in Section 5.4 of the national Inventory (pages 5-46 and 5-47), given that improvements in the national
lnventory\N\[[ lead directly to improvements in the quality of state-level estimates as well. These
improvements include an updated time series of land representation data; re-calibration of the soil carbon
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
4-8
-------
Section 4 Agriculture (NIR Chapter 5)
module in the DayCent model (see Annex 3.12); a more accurate output variable to estimate asymbiotic
nitrogen fixation in the Tier 3 method; corrections associated with manure deposited on pasture, range, and
paddock; and estimation of leaching based on irrigation status.
4.2.2.5 Planned Improvements
Planned improvements to the state-level estimates are anticipated to be the same as those presented
in Section 5.4 (page 5-47) of the national Inventory, given that improvements in the national Inventory will
lead directly to improvements in the quality of state-level estimates as well.
4.2.2.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990 -2022. EPA 430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
Full citations of references included in Chapter 5.4 (Agricultural Soil Management) and Annex 3.12 of
the national Inventory are available online here: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-
inventorv-2024-chapter-10-references O.pdf and https://www.epa.gov/system/files/documents/2024-04/us-
ghg-inventorv-2024-annex-3-additional-source-or-sink-categories-part-b.pdf.
4.2.3 Liming (NIR Section 5.5)
4.2.3.1 Background
Crushed limestone (calcium carbonate) and dolomite (CaMg[C03]2) are added to soils by land
managers to increase soil pH (i.e., to reduce acidification). C02 emissions occur as these compounds react
with hydrogen ions in soils. The rate of degradation of applied limestone and dolomite depends on the soil
conditions, soil type, climate regime, and whether limestone or dolomite is applied. Emissions from
limestone and dolomite that are used in industrial processes (e.g., cement production, glass production) are
reported under the IPPU chapter.
4.2.3.2 Methods/Approach
EPA compiles state-level C02 emissions from liming using the same methods applied in the national
Inventory. The national method is a Tier 2 approach based on the amount of limestone and dolomite applied
to agricultural soils, multiplied by a country-specific emissions factor. This is described in Chapter 5, Section
5.5 (pages 5-47 through 5-50), of the national Inventory.
The current national Inventory includes state-level emissions for the years 1990-2021. For this report, a
hybrid Approach 1 and Approach 2 was used to extend state-level estimates across the time series. The
national estimates for 2022, which were estimated using a linear extrapolation method, are disaggregated to
the state level based on the proportion of total C02 emissions from carbonate lime application occurring in
each state for 2021. Estimates are currently available for all 50 states as well as the District of Columbia and
implicitly include emissions from liming occurring on tribal lands.
Within the national activity data that leverage statistics on the application rates of crushed limestone
and dolomite for agricultural purposes, a portion of total limestone and dolomite applied nationally are
"withheld" and not allocated to specific states to avoid the disclosure of company proprietary data related to
poultry grit and mineral food. In order to allocate this withheld pool of limestone and dolomite to states so
4-9
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
that the sum of all Limestone and dolomite applied to all states and the District of Columbia, the withheld
pools of limestone and dolomite were allocated to states relative to the proportion of total
limestone/dolomite consumed by each state.
Data Appendix E-5 to this report provides state-level limestone and dolomite agricultural application
rates for all 50 states as well as the District of Columbia across the time series. Separate tables are provided
where withheld pools of limestone and dolomite are retained as discrete categories and where the withheld
pools of limestone and dolomite are allocated to states using the assumptions and methodology described
above.
4.2.3.3 Uncertainty
The overall uncertainty associated with national estimates of C02 from liming is described in Chapter 5
of the national Inventory (EPA 2024). A Monte Carlo uncertainty analysis was applied, and the analysis was
performed on the amount of limestone and dolomite applied to soils. The emissions factors included the
fraction of lime dissolved by nitric acid versus the fraction that reacts with carbonic acid, as well as the
portion of bicarbonate that leaches through the soil and is transported to the ocean. Uncertainty regarding
the time associated with leaching and transport is not addressed in the national Inventory uncertainty
analysis. The overall level of uncertainty in the national C02 liming estimates in 2022 was -85%/+89% of
national emissions estimates.
4.2.3.4 Recalculations
Limestone and dolomite application data for 2020 and 2021 were updated with the recently acquired
data from the U.S. Geological Survey. Changes that resulted from recalculations to the state-level estimates
are the same as those presented in Section 5.5 of the national Inventory (page 5-50), given that
improvements in the national Inventory will lead directly to improvements in the quality of state-level
estimates as well.
4.2.3.5 Planned Improvements
Planned improvements to the state-level estimates are anticipated to be the same as those presented
in Section 5.5 (page 5-50) of the national Inventory, given that improvements in the national Inventory will
lead directly to improvements in the quality of state-level estimates as well. As noted, there are no specific
improvements identified at this time for C02 emissions from liming.
4.2.3.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990 -2022. EPA 430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
Full citations of references included in Chapter 5.5 (Liming) of the national Inventory are available online
here: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-inventorv-2Q24-chapter-10-
references O.pdf.
4.2.4 Urea Fertilization (NIR Section 5.6)
4.2.4.1 Background
The use of urea, or CO(NH2)2, as a fertilizer leads to GHG emissions through the release of C02 that was
fixed during the production of urea. In the presence of water and urease enzymes, urea that is applied to soils
as fertilizer is converted into ammonium, hydroxyl ion, and bicarbonate. The bicarbonate then evolves into
C02 and water.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
4-10
-------
Section 4 Agriculture (NIR Chapter 5)
4.2.4.2 Methods/Approach
EPA compiles state-Level C02 emissions from urea fertilization using the same IPCCTier 1 methods
applied in the national Inventory (Approach 1). With this approach, state-level fertilizer sales data are
multiplied by the default IPCC emissions factor. This approach is described in Chapter 5, Section 5.6 (pages
5-50 through 5-52), of the national Inventory. Estimates are currently available for all 50 states and Puerto
Rico and implicitly include emissions from urea applied on tribal lands. Data Appendix E-6 to this report
provides seasonal and annual urea fertilizer consumption data by state across the time series, which serve
as the underlying activity data used to calculate state-level C02 emissions from urea application.
4.2.4.3 Uncertainty
The overall uncertainty associated with national estimates of C02 from urea fertilization is described in
Chapter 5 of the national Inventory (EPA 2024). A Monte Carlo uncertainty analysis was applied. The largest
source of uncertainty is the default emissions factor, which assumes that 100% of the carbon in C0(NH2)2
applied to soils is emitted as C02. The overall level of uncertainty in the national C02 urea fertilization
estimates in 2022 was -43%/+3%.
4.2.4.4 Recalculations
Changes that resulted from recalculations to the state-level estimates are the same as those presented
in Section 5.6 of the national Inventory (page 5-52), given that improvements in the national Inventory will
lead directly to improvements in the quality of state-level estimates as well. Updated fertilizer consumption
statistics led to time series recalculations at the state level.
4.2.4.5 Planned Improvements
Planned improvements to the state-level estimates are anticipated to be the same as those presented
in Section 5.6 (page 5-52) of the national Inventory, given that improvements in the national Inventory will
lead directly to improvements in the quality of state-level estimates as well.
4.2.4.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-
2022. EPA 430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-
gas-emissions-and-sinks.
Full citations of references included in Chapter 5.6 (Urea Fertilization) of the national Inventory are
available online here: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-inventorv-2Q24-chapter-10-
references O.pdf.
4.2.5 Field Burning of Agricultural Residues (NIR Section 5.7)
4.2.5.1 Background
Crop production creates large quantities of agricultural crop residues, which farmers manage in a
variety of ways. For example, crop residues can be left in the field and possibly incorporated into the soil with
tillage; collected and used as fuel, animal bedding material, supplemental animal feed, or construction
material; composted and applied to soils; transported to landfills; or burned in the field. Field burning of crop
residues is not considered a net source of C02 emissions because the carbon released to the atmosphere as
C02 during burning is reabsorbed during the next growing season by the crop. However, crop residue burning
is a net source of CH4, N20, carbon monoxide, and nitrogen oxide, which are released during combustion.
4-11
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
In the United States, field burning of agricultural residues is more common in southeastern states, the
Great Plains, and the Pacific Northwest. The primary crops that are managed with residue burning include
corn, cotton, lentils, rice, soybeans, sugarcane, and wheat.
4.2.5.2 Methods/Approach
EPA compiles state-level CH4 and N20 emissions from field burning of agricultural residues using the
same methods applied in the national Inventory. The national Inventory applies a country-specific Tier 2
methodology. This is described in Chapter 5, Section 5.7 (pages 5-53 through 5-62), of the national Inventory.
The most recent national Inventory includes state-level emissions for 1990-2014, but not for 2015-
2022. The exception is sugarcane, for which emissions have been estimated for 1990-2020, with 2021 to
2022 emissions estimated using a data splicing method. State estimates were developed using Approach 1
for 1990-2014 and Approach 2 for disaggregating 2015-2022 national estimates. National-level emissions for
2015-2022 are estimated using a linear extrapolation of the pattern from the previous years in the national
Inventory. For this report, these national totals were disaggregated to the state level in a two-step process.
First, the average proportion of the total national emissions was computed for each state for the years 2012-
2014, which are the last three years in which state-level emissions had been estimated. Second, the state-
level proportions were multiplied by the total national emissions to approximate the amount of emissions
occurring in each state from 2015 to 2022. Estimates are currently available for all states excludingAlaska
and Hawaii, consistent with the national Inventory, because these two states are not captured in the current
analysis. Field burning of agricultural residues does not occur in the District of Columbia and as such is not
estimated for this area. Emissions from field burning of agricultural residues occurring on tribal lands located
in the conterminous United States are implicitly captured in national and state-level estimates. No estimates
are included for U.S. territories. See Data Appendix E-7 to this report for the underlying state-level activity
data detailing the mass of residue burned and the agricultural area burned by crop type from 1990 to 2014.
4.2.5.3 Uncertainty
The overall uncertainty associated with national estimates of CH4 and N20 from field burning of
agricultural residues is described in Chapter 5 of the national Inventory (EPA 2024). As described in the
national Inventory, emissions are estimated using a linear regression model with autoregressive moving-
average errors for the 2015-2021 period. The linear regression autoregressive moving-average model also
produced estimates of the upper and lower bounds to quantify uncertainty.
Because of data limitations, there are additional uncertainties in agricultural residue burning,
particularly the potential omission of burning associated with Kentucky bluegrass (produced on farms for turf
grass installation). EPA is aware that some agricultural residue burning is not currently captured in the
national Inventory analysis; please see national Inventory planned improvements information. Overall levels
of uncertainty in the national CH4 and N20 field burning of agricultural residue estimates in 2020 were
-11%/+11% for CH4 and -13%/+13% for N20.
4.2.5.4 Recalculations
Changes that resulted from recalculations to the state-level estimates are the same as those presented
in Section 5.7 of the national Inventory (page 5-62), given that improvements in the national Inventory will
lead directly to improvements in the quality of state-level estimates as well. Recalculations have been
conducted for this Inventory to account for field burning of sugarcane residue, which was not included in the
previous Inventory.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
4-12
-------
Section 4 Agriculture (NIR Chapter 5)
4.2.5.5 Planned Improvements
Planned improvements to the state-Level estimates are anticipated to be the same as those presented
in Section 5.7 (page 5-62) of the national Inventory, given that improvements in the national Inventory will
lead directly to improvements in the quality of state-level estimates as well.
4.2.5.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990 -2022. EPA 430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
Full citations of references included in Chapter 5.7 (Field Burning of Agricultural Residues) of the
national Inventory are available online here: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-
inventorv-2024-chapter-10-references O.pdf.
4-13
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
5 Land Use, Land-Use Change, and Forestry (NIR Chapter 6)
This chapter describes the methods applied to estimate state-Level GHG fluxes resulting from land use
and land-use change within states according to changes within and conversions between all land use types,
including forest land, cropland, grassland, wetlands, and settlements (as well as other land). More
information on national-level emissions and removals and associated methods is available in Chapter 6 of
the national Inventory, available online at: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-
inventorv-2024-chapter-6-land-use-land-use-change-and-forestrv O.pdf. Table 5-1 summarizes the different
approaches used to estimate state-level LULUCF emissions and sinks completeness. State completeness is
consistent with the national Inventory. The sections below provide more detail on each category.
See also Chapter 6.1 in the national Inventory for a description of how the U.S. land base is represented
to identify land areas consistent with IPCC Guidelines. Work is underway to provide additional spatial and
temporal resolution to the representation of the U.S. land base and will help refine methods for state-level
estimates in subsequent annual publications of these data.
Table 5-1. Overview of Approaches for Estimating State-Level LULUCF Sector GHG Emissions and
Sinks
Category
Gas
Approach
Geographic Completeness8
Forest Land Remaining
Forest Land and Lands
Converted to Forest Land
Carbon,
ch4, n2o
Approach 1
Includes estimates from all
states, U.S. Territories8, and
tribal lands. For Alaska, Lands
Converted to Forest are
included in the Forest Land
Remaining Forest Land data.
Cropland and Lands
Converted to Cropland
Carbon
Hybrid:
1990-2017: Approach 1
2018-2022: Approach 2
Includes estimates from all
states (except Alaska) and
tribal lands.8
Grassland and Lands Converted to Grassland
C Stock Changes
Carbon
Hybrid:
1990-2017: Approach 1
2018-2022: Approach 2
Includes estimates from all
states (except Alaska) and
tribal lands.8
Non-C02 Emissions from
Grassland Fires
ch4, n2o
Hybrid:
1990-2020: Approach 1
2021-2022: Approach 2
Includes estimates from all
states (except Alaska) and
tribal lands.8
Wetlands and Lands Converted to Wetlands
Coastal Wetlands
Carbon,
ch4
Approach 1
Includes estimates from all
states, the District of
Columbia, and tribal lands
with coastal wetlands (except
Alaska and Hawaii).8
Peatlands
C02, ch4,
n2o
Approach 2
Includes estimates from all
states (except Hawaii) and
tribal lands.8
Flooded Lands
C02, ch4
Approach 1
Includes estimates from all
states, the District of
Columbia, tribal lands, and
territories (i.e., Puerto Rico).8
5-1
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Category
Gas
Approach
Geographic Completeness8
Settlements and Land Converted to Settlements
Soil Carbon
Carbon
Hybrid:
1990-2017: Approach 1
2018-2022: Approach 2
Includes estimates from all
states (except Alaska) and
tribal lands.8
Settlement Trees
Carbon
Hybrid:
1990-2017: Approach 1
2018-2022: Approach 2
Includes estimates from all
states, the District of
Columbia, and tribal lands.8
N20 from Settlement Soils
N20
Hybrid:
1990-2017: Approach 1
2018-2022: Approach 2
Estimates from all states
(except Alaska) and tribal
lands.8
Landfilled Yard Trimmings
and Food Scrap
Carbon
Approach 2
Estimates from all states, the
District of Columbia, tribal
lands, and territories (i.e.,
Puerto Rico).8
a Emissions are likely occurring in other U.S. territories; however, due to a lack of available data and the nature of this
category, this analysis includes emissions for only the territories indicated. Territories not listed are not estimated. Tribal
lands are included for estimates within the conterminous United States. See planned improvements of the national
Inventory.
5.1.1 Forest Land Remaining Forest Land (NIR Section 6.2)
5.1.1.1 Background
Carbon is continuously cycled among the forest ecosystem carbon storage pools (i.e., aboveground
biomass, belowground biomass, dead wood, litter, and soil organic carbon) and the atmosphere because of
biogeochemical processes in forests (e.g., photosynthesis, respiration, decomposition, disturbances such
as fires or pest outbreaks) and anthropogenic activities (e.g., harvesting, thinning, replanting). The net change
in forest carbon, however, is not equivalent to the net flux between forests and the atmosphere because timber
harvests do not cause an immediate flux of all harvested biomass carbon to the atmosphere. Instead, harvesting
transfers a portion of the carbon stored in wood to a "product pool." Once in a product pool, the carbon is emitted
over time as CO2 in the case of decomposition and as CO2, Cm, N2O, carbon monoxide, and nitrogen oxide when
the wood product combusts.
Emissions of non-CC>2 gases from forest fires, both wild and prescribed, also occur, along with N2O emissions
from nitrogen additions to the soil and CO2, Cm, and N2O emissions from drained organic soils.
5.1.1.2 Methods/Approach
To compile national estimates for the national Inventory of C stock changes from forest ecosystem
carbon pools on forest land remaining forest land, as well as non-C02 emissions from fires and non-C02
emissions from drained organic soils on forest land remaining forest land and land converted to forest land,
estimates for each state were produced and summed into a national total. This is described in Chapter 6,
Section 6.2 (pages 6-25 through 6-53), of the national Inventory. Additional information on the methodologies
and data is also provided in Annex 3.13.
Estimates are included for all U.S. states, including tribal and trust lands, as well as DC and U.S.
territories. For this Inventory, estimates for Hawaii and U.S. territories are included. Emissions of non-CC>2
gases from forest fires and non-C02 emissions from drained organic soils include emissions from both forest
land remaining forest land and land converted to forest land because it is not possible to report them separately at
this time. Additionally, the estimates of the C stock change in harvested wood are not currently available at
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
5-2
-------
Section 5 Land Use, Land-Use Change, and Forestry (NIR Chapter 6)
the state Level. Finally, the estimate of N inputs (direct and indirect N20) are not currently estimated at the
state level. Work is underway to develop an approach for disaggregating the national estimates down to state
level.
5.1.1.3 Uncertainty
The subcategories included in this state-level report include the C stock changes in forest ecosystem
carbon storage pools, non-CC>2 gases from forest fires, and non-CC>2 emissions from drained organic soils. A brief
overview of the uncertainty analyses for each of the subcategories included in the national Inventory is provided
below. Uncertainty analyses for the subcategories are:
C stock changes in forest ecosystem carbon storage pools. The overall uncertainty associated with
the 2022 national estimate of C stock changes in forest ecosystem carbon storage pools was
calculated through a combination of sample-based and model-based approaches to uncertainty for
forest ecosystem C02 flux using the IPCC Approach 1 (IPCC 2006). As described further in Chapter
6.2 of the national Inventory (EPA 2024), levels of uncertainty in the national estimates in 2022 were
-10.1%/+10.0%. State-level estimates of uncertainty vary significantly among the states but, in
general, tend to be higher than those provided for the United States in the national Inventory. These
higher uncertainties can occur when the models and factors developed from studies done at a larger
geographical scale are used to generate estimates at smaller geographic scales, such as the state
level. The potential for unique circumstances occurring within a state can reduce the accuracy and
precision of the flux estimates and increase the overall uncertainty. For more details on national-
level uncertainty, see the uncertainty discussion in Section 6.2 and Annex 3.13 of the national
Inventory.
Non-CCh gases from forest fires (includes both forest land remaining forest land and land converted to
forest land). The overall uncertainty associated with the 2022 national estimate of non-CC>2 gases from
forest fires was calculated through a Monte Carlo sampling approach, per IPCC Approach 2 (IPCC
2006), employed to propagate uncertainty based on the model and data applied for U.S. forest land.
As shown in Chapter 6 of the national Inventory, levels of uncertainty in the national estimates in
2022 were -32%/+32% for CH4 and -36%/+37%for N20. State-level estimates of uncertainty vary
significantly among the states but, in general, tend to be higher than those provided in the national
Inventory. These higher uncertainties can occur when the models and factors developed from
studies done at a larger geographical scale are used to generate estimates at smaller geographic
scales, such as the state level. The potential for unique circumstances occurring within a state can
reduce the accuracy and precision of the flux estimates and increase the overall uncertainty. For
more details on national-level uncertainty and the quantities and assumptions employed to define
and propagate uncertainty, see the uncertainty discussion in Section 6.2 and Annex 3.13 of the
national Inventory.
Non-C02 gases from drained organic soils (includes both forest land remaining forest land and
land converted to forest land). The overall uncertainty associated with the 2022 national estimate
of non-C02 gases from drained organic soils was calculated through IPCC Approach 1 (IPCC 2006).
As described further in Chapter 6 of the national Inventory, levels of uncertainty in the national
estimates in 2022 were -69%/+82% for CH4 and -118%/+132% N20. State-level estimates of
uncertainty vary significantly among the states but, in general, tend to be higher than those provided
in the national Inventory. For more details on national-level uncertainty and the quantities and
assumptions employed to define and propagate uncertainty, see the uncertainty discussion in
Section 6.2 and Annex 3.13 of the national Inventory.
5-3
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
5.1.1.4 Recalculations
Changes that resulted from recalculations to the state-level estimates are the same as those presented
in Section 6.2 of the national Inventory (pages 6-38 through 6-39, 6-42, 6-45, and 6-48), given that
improvements in the national Inventory will lead directly to improvements in the quality of state-level
estimates as well. As described in the national Inventory, a number of key improvements were made,
including estimation of Hawaii and U.S. territories, and implementation of new methods for estimating
standing live and dead aboveground biomass carbon in the FIA program (Westfall et al. 2024). These new
methods, leveraging the newly developed national-scale volume and biomass (NSVB) framework, represent
nearly a decade of research and development in the FIA program. The new methods: (1) greatly simplify
predictions of aboveground biomass because only five model specifications are used nationally instead of
dozens of species-specific and species-group-specific models used in each region and/or state, (2) eliminate
administrative boundaries (e.g., regions or states) in favor of ecologically based regions (i.e., ecodivisions) to
capture variation in tree size and volume (or biomass) within species or species groups, (3) are based on tree
measurements from in-situ data, which also facilitates more accurate quantification of model uncertainty,
(4) result in consistent model behavior for all tree species and sizes, and (5) use species-specific carbon
fractions for biomass-to-carbon conversions (a departure from the previous method, which assumed a
default 50% biomass-to-carbon fraction). These improvements resulted in significant recalculations to both
the national and state-level estimates. More information on the NSVB framework can be found here:
https://www.fs.usda.gov/research/programs/fia/nsvb.
5.1.1.5 Planned Improvements
The planned improvements are consistent with those planned for improving the national estimates
given that the underlying methods for the state-level GHG estimates are the same as those in the national
Inventory. To view the planned improvements to the methods and data for estimating emissions and
removals from forest land remaining forest land, see the planned improvements discussion on pages 6-44
through 6-45, 6-47, and 6-53 of Chapter 6.2 in the national Inventory for a description of future work to
improve these estimates. In addition, as noted by the USFS 2023 Resource Bulletin (Domke et al. 2023),
investments are being made to leverage existing state-level forest products information to allow for the
disaggregation of harvested wood product estimates at the state level in the future.
5.1.1.6 References
Domke, G.M., B.F. Walters, C.L. Giebink, E.J. Greenfield, J.E. Smith, M.C. Nichols, J.A. Knott, S.M. Ogle, J.W.
Coulston, and J. Steller (2023) Greenhouse Gas Emissions and Removals from Forest Land, Woodlands,
Urban Trees, and Harvested Wood Products in the United States, 1990-2021. Resource Bulletin WO-101.
U.S. Department of Agriculture. Available online at: https://doi.org/10.2737/WQ-RB-101.
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventory-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
Westfall, J.A., J.W. Coulston, A.N. Gray, J.D. Shaw, P.J. Radtke, D.M. Walker, A.R. Weiskittel, D.W.
MacFarlane, D.L.R. Affleck, D. Zhao, H. Temesgen, K.P. Poudel, J.M. Frank, S.P. Prisley, Y. Wang, A.J.
Sanchez Meador, D. Auty, and G.M. Domke (2024) A National-Scale Tree Volume, Biomass, and Carbon
Modeling System for the United States. General Technical Report WO-104. U.S. Department of Agriculture,
Forest Service. Available online at: https://doi.org/10.2737/WQ-GTR-104.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
5-4
-------
Section 5 Land Use, Land-Use Change, and Forestry (NIR Chapter 6)
Full citations of all references included in Chapter 6.2 (Forest Land Remaining Forest Land) of the
national Inventory are available online here: https://www.epa.gov/system/files/documents/2024-04/us-ghg-
inventorv-2024-chapter-10-references O.pdf.
5.1.2 Land Converted to Forest Land (NIR Section 6.3)
5.1.2.1 Background
Land use conversions into forest land can result in C stock changes to all forest ecosystem carbon
pools (i.e., aboveground biomass, belowground biomass, dead wood, litter, and soil organic carbon). Section
5.1.2 provides estimates of C stock changes resulting from conversion of cropland, grassland, wetlands,
settlements, and other lands to forest land (Domke et al. 2023).
5.1.2.2 Methods/Approach
The methods applied for estimating C stock changes in land converted to forest land are the same as
those applied for forest land remaining forest land. This is described in Chapter 6, Section 6.3 (pages 6-53
through 6-61), of the national Inventory. Additional information on the methodologies and data is also
provided in Annex 3.13 of the national Inventory. Please note that estimates for Hawaii or U.S. territories are
not included in the national total or available at the state level at this time. Forest ecosystem C stock
changes from land conversion in Alaska are currently included in the forest land remaining forest land
chapter because there are insufficient data to separate the changes at this time.
5.1.2.3 Uncertainty
The overall uncertainty associated with the 2022 national estimate of the C stock changes in forest
ecosystem carbon storage pools for land converted to forest land is described in Chapter 6.3 of the national
Inventory (EPA 2024). The uncertainty estimates were calculated through a combination of sample-based
and model-based approaches to uncertainty for non-soil forest ecosystem C02 flux using IPCC Approach 1
(IPCC 2006), in combination with IPCC Approach 2 for mineral soils (described in Section 6.4, Cropland
Remaining Cropland, of the Inventory report). Uncertainty estimates are provided for each land conversion
category and carbon pool. The combined level of uncertainty in the national estimates in 2022 was
-11 %/+11 %. State-level estimates of uncertainty are not available but are likely to vary significantly from the
national estimates and, in general, tend to be higher than those provided for the United States in the national
Inventory. These higher uncertainties can occur when the models and factors developed from studies done
at a larger geographical scale are used to generate estimates at smaller geographic scales, such as the state
level, the potential for unique circumstances occurring within a state can reduce the accuracy and precision
of the flux estimates and increase the overall uncertainty. For more details on national-level uncertainty, see
the uncertainty discussion in Section 6.4 and Annex 3.13 of the national Inventory.
5.1.2.4 Recalculations
Changes that resulted from recalculations to the state-level estimates are the same as those presented
in Section 6.3 of the national Inventory (pages 6-59 through 6-60), given that improvements in the national
lnventory\N\[[ lead directly to improvements in the quality of state-level estimates as well.
5.1.2.5 Planned Improvements
The planned improvements are consistent with those planned for improving the national estimates
given that the underlying methods for state GHG estimates are the same as those in the national Inventory.
To review the planned improvements to the methods and data for estimating emissions and removals from
land converted to forest land, see the planned improvements discussion on page 6-61 of Chapter 6.3 in the
national Inventory.
5-5
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
5.1.2.6 References
Domke, G.M., B.F. Walters, C.L. Giebink, E.J. Greenfield, J.E. Smith, M.C. Nichols, J.A. Knott, S.M. Ogle, J.W.
Coulston, and J. Steller (2023) Greenhouse Gas Emissions and Removals from Forest Land, Woodlands,
Urban Trees, and Harvested Wood Products in the United States, 1990 -2021. Resource Bulletin WO-101.
U.S. Department of Agriculture. Available online at: https://doi.org/10.2737/WQ-RB-101.
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventory-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006) 2006IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
Full citations of references included in Chapter 6.3 (Land Converted to Forest Land) of the national
Inventory are available online here: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-inventorv-
2024-chapter-10-references O.pdf.
5.1.3 Cropland Remaining Cropland (NIR Section 6.4)
5.1.3.1 Background
Carbon in cropland ecosystems occurs in biomass, dead organic matter, and soils. However, carbon
storage in cropland biomass and dead organic matter is relatively ephemeral and does not need to be
reported according to IPCC (2006), with the exception of carbon stored in perennial woody crop biomass,
such as citrus groves and apple orchards, in addition to the biomass, downed wood, and dead organic
matter in agroforestry systems. Within soils, carbon is found in organic and inorganic forms, but soil organic
carbon is the main source and sink for atmospheric C02 in most soils.
IPCC (2006) recommends reporting changes in soil organic C stocks due to agricultural land use and
management activities for mineral and organic soils. Management of croplands and cropland soils has an
impact on organic matter inputs and microbial decomposition, and thereby results in a net C stock change.
Cropland remaining cropland includes all cropland in an inventory year that has been cropland for a
continuous time period of at least 20 years. This determination is based on the USDA NRI for nonfederal
lands and the National Land Cover Database for federal lands. Cropland includes all land that is used to
produce food and fiber, forage that is harvested and used as feed (e.g., hay and silage), and cropland that has
been enrolled in the Conservation Reserve Program (i.e., considered set-aside cropland).
5.1.3.2 Methods/Approach
EPA compiles state-level emissions from cropland remaining cropland using the same methods
applied in the national Inventory. Please see the methodologies described in Chapter 6, Section 6.4 (pages 6-
61 through 6-74), of the national Inventory. For this report, estimates were developed using a hybrid of
Approach 1 and Approach 2. The current national Inventory includes state-level emissions for the years
1990-2017 for soil organic carbon stock changes. The remaining years in the time series were only estimated
at the national scale using a surrogate data method, and a two-step process was used to approximate the
state-level emissions for the remainingyears. First, the average proportion of the total national emissions
was computed for each state for the years from 2015-2017. Second, the state-level proportions were
multiplied by the total national emissions to approximate the amount of emissions occurring in each state for
2018-2022. Estimates are included for all states except Alaska. The USDA National Resources Inventory
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
5-6
-------
Section 5 Land Use, Land-Use Change, and Forestry (NIR Chapter 6)
covers tribal and trust Lands. Emissions from cropland remaining cropland do not occur in the District of
Columbia, and emissions for U.S. territories (including Puerto Rico) are not estimated at this time.
Additional information on methodologies and data is also provided in Annex 3.12 of the national
Inventory.
5.1.3.3 Uncertainty
The overall uncertainty associated with national estimates from Cropland Remaining Cropland is
described in Chapter 6 of the national Inventory (EPA 2024) and in further detail in Annex 3.12. Uncertainty for
the Tier 2 and 3 approaches is derived using a Monte Carlo approach. The combined uncertainty for soil
organic carbon stocks in cropland remaining cropland in 2022 is -212%/+212%.
5.1.3.4 Recalculations
Recalculations were applied for this current report consistent with the national Inventory (see Section
6.4, pages 6-72 and 6-73).
5.1.3.5 Planned Improvements
The planned improvements are anticipated to be the same as those planned for improving the national
estimates given that the underlying methods for state GHG estimates are the same as those in the national
Inventory and will lead directly to improvements in the quality of state-level estimates as well. To review the
planned improvements to the methods and data for estimating emissions and removals from cropland
remaining cropland, see the planned improvements discussion on pages 6-73 and 6-74 of Chapter 6.4 in the
national Inventory.
5.1.3.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventory-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006) 2006IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
Full citations of references included in Chapter 6.4 (Cropland Remaining Cropland) and Annex 3.12 of
the national Inventory are available online here: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-
inventorv-2024-chapter-10-references O.pdf and https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-
inventorv-2024-annex-3-additional-source-or-sink-categories-part-b.pdf.
5.1.4 Land Converted to Cropland (NIR Section 6.5)
5.1.4.1 Background
Land use change can lead to large losses of carbon to the atmosphere, particularly conversions from
forest land. Moreover, conversion of forests to another land use (i.e., deforestation) is one of the largest
anthropogenic sources of emissions to the atmosphere globally.
Land converted to cropland includes all cropland in an inventory year that (1) had been in at least one
other land use during the previous 20 years and (2) is used to produce food, fiber or forage that is harvested
and used as feed (e.g., hay and silage). For example, grassland or forest land converted to cropland during
the past 20 years would be reported in this category. Recently converted lands are retained in this category
for 20years as recommended by IPCC (2006).
5-7
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
5.1.4.2 Methods/Approach
EPA compiles state-Level emissions from Land converted to cropland using the same methods applied
in the national Inventory. Please see the methodologies described in Chapter 6, Section 6.5 (pages 6-74
through 6-81), of the national Inventory. For this report, estimates were developed using a hybrid of Approach
1 and Approach 2. The current national Inventory includes state-level fluxes for the years 1990-2022 for
biomass, standing dead, dead wood, and litter and for the years 1990-2017 for soil organic carbon stock
changes. The remaining years in the time series for soil organic carbon stock changes were only estimated at
the national scale using a surrogate data method, and a two-step process was used to approximate the
state-level emissions for the remainingyears. First, the average proportion of the total national emissions
was computed for each state for the years 2015-2017. Second, the state-level proportions were multiplied by
the total national emissions to approximate the amount of emissions occurring in each state for 2018-2022.
Estimates are included for all states except Alaska. The USDA National Resources Inventory covers tribal and
trust lands. Emissions from land converted to cropland do not occur in the District of Columbia, and
emissions for U.S. territories (including Puerto Rico) are not estimated at this time.
Additional information on methodologies and data is also provided in Annex 3.12 of the national
Inventory.
5.1.4.3 Uncertainty
The overall uncertainty associated with national estimates from land converted to cropland is described
in Chapter 6 of the national Inventory (EPA 2024) and in further detail in Annex 3.12 and Annex 3.13
(Forestland Converted to Cropland). The uncertainty analyses for mineral soil organic C stock changes using
the Tier 3 and Tier 2 methodologies are based on a Monte Carlo approach that is used in the cropland
remaining cropland analysis. The combined uncertainty for total carbon stocks in land converted to cropland
in 2021 was -93%/+93%.
5.1.4.4 Recalculations
Changes that resulted from recalculations to the state-level estimates are the same as those presented
in Section 6.5 of the national Inventory (pages 6-80 and 6-81), given that improvements in the national
lnventory\N\[[ lead directly to improvements in the quality of state-level estimates as well.
5.1.4.5 Planned Improvements
The planned improvements are anticipated to be the same as those planned for improving the national
estimates, given that the underlying methods for state GHG estimates are the same as those in the national
Inventory and will lead directly to improvements in the quality of state-level estimates as well. To review the
planned improvements to the methods and data for estimating emissions and removals from land converted
to cropland, see the planned improvements discussion on page 6-81 of Chapter 6.5 in the national Inventory.
5.1.4.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006) 2006IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
Full citations of references included in Chapter 6.5 (Land Converted to Cropland) and Annex 3.12 of the
national Inventory are available online here: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
5-8
-------
Section 5 Land Use, Land-Use Change, and Forestry (NIR Chapter 6)
inventorv-2024-chapter-10-references O.pdf and https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-
inventorv-2024-annex-3-additional-source-or-sink-categories-part-b.pdf.
5.1.5 Grassland Remaining Grassland (NIR Section 6.6)
5.1.5.1 Background
Carbon in grassland ecosystems occurs in biomass, dead organic matter, and soils. Soils are the
largest pool of carbon in grasslands and have the greatest potential for longer-term storage or release of
carbon. Biomass and dead organic matter carbon pools are relatively ephemeral compared to the soil
carbon pool, with the exception of carbon stored in tree and shrub biomass that occurs on grasslands.
The 2006 IPCC Guidelines recommend reporting changes in biomass, dead organic matter, and soil
organic C stocks with land use and management. C stock changes for aboveground and belowground
biomass, dead wood, and litter pools are reported for woodlands (i.e., a subcategory of grasslands), and may
be extended to include agroforestry management associated with grasslands in the future. For soil organic
carbon, the 2006 IPCC Guidelines (IPCC 2006) recommend reporting changes due to (1) agricultural land use
and management activities on mineral soils and (2) agricultural land use and management activities on
organic soils.
Grassland remaininggrassland includes all grassland in an inventory year that had been grassland fora
continuous time period of at least 20 years. Grassland includes pasture and rangeland that are primarily, but
not exclusively, used for livestock grazing. Rangelands are typically extensive areas of native grassland that
are not intensively managed, while pastures are typically seeded grassland (possibly following tree removal)
that may also have additional management, such as irrigation or inter-seeding of legumes. Woodlands are
also considered grassland and are areas of continuous tree cover that do not meet the definition of forest
land.
Non-C02 emissions from grassland fires are also included for all U.S. states but are not estimated for
DC and U.S. territories. These emissions do not currently include emissions from burning perennial biomass
(a national Inventory planned improvement).
5.1.5.2 Methods/Approach
EPA compiles state-level emissions from grassland remaining grassland using the same methods
applied in the national Inventory. Please see the methodologies described in Chapter 6.6 (pages 6-81 through
6-92) of the national Inventory For this report, estimates were developed using a hybrid of Approach land
Approach 2. The current national Inventory includes state-level emissions for the years 1990-2022 for
biomass, standing dead, dead wood, and litter, as well as for the years 1990-2017 for soil organic carbon
stock changes. The remaining years in the time series for soil organic C stock changes were only estimated at
the national scale using a surrogate data method, and a two-step process was used to approximate the
state-level emissions for the remaining years. First, the average proportion of the total national emissions
was computed for each state for the years 2015-2017. Second, the state-level proportions were multiplied by
the total national emissions to approximate the amount of emissions occurring in each state for 2018-2022.
Estimates are included for all states except Alaska. The USDA National Resources Inventory covers tribal and
trust lands. Emissions from grasssland remaininggrassland do not occur in the District of Columbia, and
emissions for U.S. territories (including Puerto Rico) are not estimated at this time.
Additional information on national Inventory methodologies and data is also provided in Annex 3.12 of
the national Inventory.
5-9
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
5.1.5.3 Uncertainty
The overall uncertainty associated with national estimates from grassland remaining grassland is
described in Chapter 6 of the national Inventory (EPA 2024) and in further details in Annex 3.12. The
uncertainty analyses for mineral soil organic carbon stock changes using the Tier 3 and Tier 2 methodologies
are based on a Monte Carlo approach that is used in cropland remaining cropland analysis. Uncertainty
estimates are also developed for biomass burning in grassland using a linear regression autoregressive
moving-average model to estimate the upper and lower bounds of the emissions estimate. The combined
uncertainty for flux associated with C stock changes occurring in grassland remaining grassland in 2022 was
-926%/+926%. The uncertainty for Non-C02 emissions from grassland fires in 2021 was -100%/+137% for
both CH4 and N20.
5.1.5.4 Recalculations
Changes that resulted from recalculations to the state-level estimates are the same as those presented
in Section 6.6 of the national Inventory (pages 6-88 and 6-92), given that improvements in the national
lnventory\N\[[ lead directly to improvements in the quality of state-level estimates as well.
5.1.5.5 Planned Improvements
The planned improvements are anticipated to be the same as those planned for improving the national
estimates given that the underlying methods for state GHG estimates are the same as those in the national
Inventory and will lead directly to improvements in the quality of state-level estimates as well. To review the
planned improvements to the methods and data for estimating emissions and removals from grassland
remaining grassland, see the planned improvements discussion on pages 6-89, 6-90 and 6-92 of Chapter 6.6
in the national Inventory.
5.1.5.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006) 2006IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
Full citations of references included in Chapter 6.6 (Grassland Remaining Grassland) and Annex 3.12 of
the national Inventory are available online here: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-
inventorv-2024-chapter-10-references O.pdf and https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-
inventorv-2024-annex-3-additional-source-or-sink-categories-part-b.pdf.
5.1.6 Land Converted to Grassland (NIR Section 6.7)
5.1.6.1 Background
Land use change can lead to large losses of carbon to the atmosphere, particularly conversions from
forest land. Moreover, conversion of forests to another land use (i.e., deforestation) is one of the largest
anthropogenic sources of emissions to the atmosphere globally.
Land converted to grassland includes all grassland in an inventory year that had been in at least one
other land use during the previous 20 years. For example, cropland or forest land converted to grassland
during the past 20 years would be reported in this category. Recently converted lands are retained in this
category for 20 years as recommended by IPCC (2006).
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
5-10
-------
Section 5 Land Use, Land-Use Change, and Forestry (NIR Chapter 6)
5.1.6.2 Methods/Approach
EPA compiles state-Level emissions from Land converted to grassland using the same methods applied
in the national Inventory. Please see the methodologies described in Chapter 6, Section 6.7 (pages 6-93
through 6-101), of the national Inventory. For this report, estimates were developed using a hybrid of
Approach 1 and Approach 2. The current national Inventory includes state-level emissions for the years
1990-2022 for biomass, standing dead, dead wood, and litter, and for the years 1990-2017 for soil organic
carbon stock changes. The remainingyears in the time series for soil organic carbon stock changes were only
estimated at the national scale using a surrogate data method, and a two-step process was used to
approximate the state-level emissions for the remaining years. First, the average proportion of the total
national emissions was computed for each state for the years 2015-2017. Second, the state-level
proportions were multiplied by the total national emissions to approximate the amount of emissions
occurring in each state for 2018-2022. Estimates are included for all states except Alaska. The USDA
National Resources Inventory covers tribal and trust lands. Emissions from land converted to cropland do
not occur in the District of Columbia, and emissions estimates for U.S. territories (including Puerto Rico) are
not estimated at this time.
Additional information on methodologies and data is also provided in Annex 3.12 of the national
Inventory.
5.1.6.3 Uncertainty
The overall uncertainty associated with national estimates from Land Converted to Grassland is
described in Chapter 6 of the national Inventory (EPA 2024) and in further details in Annex 3.12. The
uncertainty analyses for mineral soil organic C stock changes using the Tier 3 and Tier 2 methodologies are
based on a Monte Carlo approach that is used in cropland remaining cropland analysis. The combined
uncertainty for total carbon stocks in land converted to grassland in 2021 was -156%/+156%.
5.1.6.4 Recalculations
Changes that resulted from recalculations to the state-level estimates are the same as those presented
in Section 6.7 of the national Inventory (page 6-100), given that improvements in the national Inventory will
lead directly to improvements in the quality of state-level estimates as well.
5.1.6.5 Planned Improvements
The planned improvements are anticipated to be the same as those planned for improving the national
estimates given that the underlying methods for state GHG estimates are the same as those in the national
Inventory. To review the planned improvements to the methods and data for estimating emissions and
removals from land converted to grassland, see the planned improvements discussion on pages 6-100 and
6-101 of Chapter 6.7 in the national Inventory.
5.1.6.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006) 2006IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
Full citations of references included in Chapter 6.7 (Land Converted to Grassland) and Annex 3.12 of the
national Inventory are available online here: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-
5-11
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
inventorv-2024-annex-3-additional-source-or-sink-categories-part-b.pdf and
https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-inventorv-2Q24-chapter-10-references O.pdf.
5.1.7 Wetlands Remaining Wetlands (NIR Section 6.8)
This section presents methods for estimating state-Level C02, CH4, and N20 emissions and removals
from management of wetlands consistent with the national Inventory, specifically:
Coastal wetlands remaining coastal wetlands (C02, CH4)
Peatlands remaining peatlands (C02, CH4, and N20)
Flooded land remaining flooded land (CH4)
5.1.7.1 Coastal Wetlands Remaining Coastal Wetlands
5.1.7.1.1. Background
Consistent with ecological definitions of wetlands, the United States has historically included under the
category of wetlands those coastal shallow water areas of estuaries and bays that lie within the extent of the
wetland representation. The national Inventory includes all privately owned and publicly owned coastal
wetlands (i.e., mangroves and tidal marsh) alongthe oceanic and gulf shores on the conterminous United
States, including tribal lands, but does not include coastal wetlands remaining coastal wetlands in Alaska,
Hawaii, or U.S. territories.
Soil and biomass carbon stocks from seagrasses are not currently included in the national Inventory
because of insufficient data on distribution, change through time, and carbon stocks or carbon stock
changes as a result of anthropogenic influence. Additionally, the estimates of N20 emissions from
aquaculture are only available at the national level because of data limitations and have not been included in
the current state estimates.
Under the coastal wetlands remaining coastal wetlands category, the following emissions and removals
subcategories are quantified at the state level:
C stock changes and CH4 emissions on vegetated coastal wetlands remaining vegetated coastal
wetlands.
C stock changes on vegetated coastal wetlands converted to unvegetated open water coastal
wetlands.
C stock changes on unvegetated open water coastal wetlands converted to vegetated coastal
wetlands.
5.1.7.1.2. Methods/Approach
To compile national estimates of C stock changes and CH4 emissions from coastal wetlands remaining
coastal wetlands for the national Inventory, estimates for each state and the District of Columbia with
coastal wetlands were produced and summed into a national total. A description of the methods and data
used to apply methodology provided in the IPCC 2013 Wetlands Supplement (IPCC 2013) to estimate state-
level emissions is provided in Chapter 6, Section 6.8 (pages 6-108 through 6-126).
States (plus the District of Columbia) with coastal wetlands currently included in the national Inventory
are Alabama, California, Connecticut, Delaware, Florida, Georgia, Louisiana, Maine, Maryland,
Massachusetts, Mississippi, New Hampshire, New Jersey, New York, North Carolina, Oregon, Pennsylvania,
Rhode Island, South Carolina, Texas, Virginia, and Washington. Please note that estimates for Hawaii and
Alaska, and U.S. territories, are not included in the national total or available at the state level at this time.
See Planned Improvements.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
5-12
-------
Section 5 Land Use, Land-Use Change, and Forestry (NIR Chapter 6)
5.1.7.1.3. Uncertainty
Uncertainty estimates for each of the emissions and removals categories are only available at the
national level. A brief overview of the uncertainty analyses for each of the subcategories included in the
national Inventory is provided below:
C stock changes and CH4 emissions on vegetated coastal wetlands remaining vegetated
coastal wetlands. Underlying uncertainties in the estimates of soil and biomass C stock changes
and CH4emissions include uncertainties associated with Tier 2 literature values of soil C stocks,
biomass C stocks, and CH4 flux; assumptions that underlie the methodological approaches applied;
and uncertainties linked to interpretation of remote sensing data. Uncertainty specific to vegetated
coastal wetlands remaining vegetated coastal wetlands include differentiation of palustrine and
estuarine community classes, which determines the soil C stock and CH4 flux applied. Uncertainties
for soil and biomass C stock data for all subcategories are not available and thus assumptions were
applied using expert judgment about the most appropriate assignment of a C stock to a
disaggregation of a community class. IPCC Approach 1 (IPCC 2006) was used to calculate these
uncertainties. As described further in Chapter 6.8 of the national Inventory (EPA 2024), levels of
uncertainty in the national estimates in 2022 are -24.1%/+24.1% for biomass C stock change,
-17.7%/+17.7% for soil C stock change, and -29.9%/+29.9% for CH4 emissions. The combined
uncertainty across all sub-sources is -36.5%/+36.5%, which is primarily driven by the uncertainty in
the CH4 estimates because there is high variability in CH4 emissions, which is accounted for when
the salinity is less than 18 parts per thousand. State-level estimates of uncertainty will vary
significantly among the states but, in general, tend be higher than those provided for the United
States in the national Inventory. For more details on national-level uncertainty, see the uncertainty
discussion in Section 6.8 of the national Inventory.
C stock changes on vegetated coastal wetlands converted to unvegetated open water coastal
wetlands. Underlying uncertainties in the estimates of soil and biomass C stock changes are
associated with country-specific (Tier 2) literature values of these stocks, while the uncertainties
with the Tier 1 estimates are associated with subtropical estuarine forested wetland dead organic
matter stocks. Assumptions that underlie the methodological approaches applied and uncertainties
linked to interpretation of remote sensing data are also included in this uncertainty assessment.
IPCC Approach 1 (IPCC 2006) was used to calculate these uncertainties. As described further in
Chapter 6.8 of the national Inventory (EPA 2024), levels of uncertainty in the national estimates in
2022 are -24.1%/+24.1% for biomass C stock change, -25.8%/+25.8%for dead organic matter C
stock change, and -15.0%/+15.0% for soil C stock change. The combined uncertainty across all sub-
sources is -32.0%/+32.0%, which is primarily driven by the uncertainty in the soil C stock change
estimates. State-level estimates of uncertainty will vary significantly among the states but, in
general, tend to be higher than those provided for the United States in the national Inventory. For
more details on national-level uncertainty, see the uncertainty discussion in Section 6.8 of the
national Inventory.
C stock changes on unvegetated open water coastal wetlands converted to vegetated coastal
wetlands. Underlying uncertainties in estimates of soil and biomass C stock changes include
uncertainties associated with country-specific (Tier 2) literature values of these C stocks and
assumptions that underlie the methodological approaches applied and uncertainties linked to
interpreting remote sensing data. Uncertainty specific to coastal wetlands includes differentiation of
palustrine and estuarine community classes that determine the soil C stock applied. IPCC Approach
1 (IPCC 2006) was used to calculate these uncertainties. As described further in Chapter 6.8 of the
national Inventory (EPA 2024), levels of uncertainty in the national estimates in 2022 are -20%/+20%
5-13
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
for biomass C stock change, -25.8%/+25.8% dead organic matter C stock change, and
-17.7%/+17.7% for soil C stock change. The combined uncertainty across all sub-sources is
-33.3%/+33.3%. State-Level estimates of uncertainty will vary significantly among the states but, in
general, tend to be higher than those provided for the United States in the national Inventory. For
more details on national-level uncertainty, see the uncertainty discussion in Section 6.8 of the
national Inventory.
5.1.7.1.4. Recalculations
Changes that resulted from recalculations to the state-level estimates are the same as those presented
in Section 6.8 of the national Inventory (pages 6-115,6-120, and 6-124), given that improvements in the
national Inventory will lead directly to improvements in the quality of state-level estimates as well.
5.1.7.1.5. Planned Improvements
The planned improvements are consistent with those planned for improving the national estimates
given that the underlying methods for the state GHG estimates are the same as those in the national
Inventory. To review the planned improvements to the methods and data for estimating emissions and
removals from coastal wetlands remaining coastal wetlands, see the planned improvements discussions on
pages 6-116, 6-120, and 6-124 of Chapter 6.8 in the national Inventory.
While the N20 flux from aquaculture has not been estimated for this initial version of the national
Inventory by state, EPA intends to include these data in future annual publications.
5.1.7.1.6. References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventory-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006) 2006IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
IPCC (2013) 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories:
Wetlands. T. Hiraishi, T. Krug, K. Tanabe, N. Srivastava, J. Baasansuren, M. Fukuda, and T.G. Troxler
(eds.). Available online at: https://www.ipcc.ch/publication/2013-supplement-to-the-2006-ipcc-
guidelines-for-national-greenhouse-gas-inventories-wetlands/.
Full citations of the references included in Chapter 6.8 (Wetlands Remaining Wetlands) of the national
Inventory are available online here: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-inventorv-
2024-chapter-10-references O.pdf.
5.1.7.2 Peatlands Remaining Peatlands
5.1.7.2.1. Background
This section describes methods to estimate state-level C02, CH4, and N20 emissions from peatlands
remaining peatlands (managed peatlands).
Managed peatlands are peatlands that have been cleared and drained for peat production. The
production cycle of a managed peatland has three phases: land conversion in preparation for peat extraction
(e.g., clearing surface biomass, draining); extraction (which results in the emissions reported under
peatlands remaining peatlands); and abandonment, restoration, rewetting, or conversion of the peatland to
another use. On-site and off-site emissions also result from managed peatlands. On-site emissions from
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
5-14
-------
Section 5 Land Use, Land-Use Change, and Forestry (NIR Chapter 6)
managed peatlands occur as the Land is cleared of vegetation and the underlying peat is exposed to sun and
weather. Off-site C02 emissions from managed peatlands occur from waterborne carbon losses and the
horticultural and landscaping use of peat.
5.1.7.2.2. Methods/Approach
State-level estimates were compiled using Approach 2 and are based on the national-level methods
included in Chapter 6.8, Wetlands Remaining Wetlands, of the national Inventory. State-level peat
production was estimated using Bureau of Mines and USGS Minerals Yearbooks from 1990-2020, covering
the contiguous 48 states and the District of Columbia. For Alaska, the method is the same as the national-
level method; the national Inventory historically breaks out peat production and emissions separately for
Alaska. For the years 2021 and 2022, peat production was estimated as described below. Hawaii and Puerto
Rico are not estimated because peat production data were not available, and regional data provided in the
USGS yearbooks did not include these states as peat producers. There are no estimates for other U.S.
territories. EPA calculates state-level estimates of emissions from peatlands based on voluntary survey data
from peat-producing companies in the contiguous 48 states and Alaska. Data from these surveys do not
distinguish peat produced from tribal and non-tribal lands.
For annual state-level peat production for 1990-2022, the primary activity data used to estimate
emissions were calculated as follows given that no single data source covers all years:
For 1990-1993, state-level annual peat production data were obtained from the Bureau of Mines
Minerals Yearbooks (U.S. Bureau of Mines 1990,1991,1992,1993). These data were available for
only select states and the Bureau of Mines also reported a total national production value. The
Bureau of Mines state peat production data were summed by year to obtain total known state peat
production. States and territories with no individual peat production data and that are not within a
peat-producing region are assumed to not be producing peat. State production values were
normalized to sum to the national production value.
For 1994-1997, state-level annual peat production data were obtained from the USGS Minerals
Yearbooks for those years (USGS 2020). Regional total data became available in 1994. To determine
peat production for states within a "peat-producing region" (i.e., Northeast, Great Lakes, Southeast,
West) but no individual reported peat production data, individual state values were summed and
then subtracted from the available regional total peat production value to determine the peat
production not accounted for in the regional data. The peat production for states with individual
reported peat production data and peat production estimated from region-based peat production
data were then summed. This value was subtracted from the most recent available total national
peat production of the contiguous 48 states available from the appropriate year's USGS annual
Minerals Commodities Summary (USGS 2023c). States and territories with no individual peat
production data and that are not within a peat-producing region are assumed to not be producing
peat. State production values were normalized to sum to the national production value.
For 1998-2020, state-level annual peat production data were obtained from the USGS Minerals
Yearbooks (USGS 2020, 2023a, 2023b) from the respective years. To determine peat production for
states within a peat-producing region (i.e., East, Great Lakes, West) but with no individual reported
peat production data, individual state values were summed and then subtracted from the available
regional total peat production value to determine the peat production not accounted for in the
regional data. Note that between 1997 and 1998, peat-producing regions changed from Northeast,
Great Lakes, Southeast, and West to East, Great Lakes, and West. States placed within these
regions varied from year to year. The peat production for states with individual reported peat
production data and peat production estimated from region-based peat production data were then
5-15
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
summed. States and territories with no individual peat production data and that are not within a
peat-producing region are assumed to not be producing peat. State production values were
normalized to sum to the national production value.
State-level peat production in 2021 and 2022 were estimated as an average fraction of total peat
production for the previous 10 years because 2021 and 2022 USGS Minerals Yearbook data were not
available when the national Inventory was developed. There is annual variability in the peat
production values, which lends itself to using an average, rather than relying solely on the previous
year, 2020, to estimate peat production. An average percentage was estimated by calculating the
average fraction of total U.S. peat production over the past 10 years for a given state. This average
fraction was then multiplied by the 2021 and 2022 total U.S. peat production of the conterminous 48
states available from the respective USGS annual Minerals Commodities Summary (USGS 2023c).
States and territories with no individual peat production data and not within a peat-producing region
were assumed to not be producing peat.
Data Appendix E-8 of this report provides state-level peat production data as well as state-level
estimated peat area across the time series for all 50 states and the District of Columbia.
Following peat production estimation, peat production area was calculated using a standard
conversion factor from mass of peat production to land area required for that mass of peat production: 100
metric tons of peat per hectare per year (Vacuum method, Canada) (Cleary et a I. 2005).
To estimate state-level emissions from peatlands remaining peatlands, national assumptions were
applied to estimate the percentage of nutrient-rich versus the percentage of nutrient-poor peat soil, which
affects emissions. Six separate calculations were then performed to yield C02, CH4, and N20 emissions
estimates:
Emissions factors for off-site C02emissions from horticulture use (which differentiates between rich
and poor peat) and dissolved organic carbon were applied to peat production, and the areas of peat
production were calculated to yield off-site C02 emissions. Because of a lack of peat application
data, off-site peat was assumed to be applied proportionally to U.S. domestic state population in
two separate components: horticulture use (which includes peat application in Hawaii, Alaska, and
Puerto Rico) and dissolved organic carbon. Off-site C02 emissions were distributed proportionally by
the percentage of the total U.S. population (1990-1999: U.S. Census Bureau 2002; 2000-2009: U.S.
Census Bureau 2011; 2010-2021: U.S. Census Bureau 2021, 2023, Instituto de Estadfsticas de
Puerto Rico 2022), as it is assumed that horticulture use is positively correlated to population. Off-
site C02 emissions are not estimated for U.S. territories. EPA intends to continue reviewing this
assumption; see the planned improvements below.
An IPCC (2013) emissions factor for on-site C02 emissions of drained organic soils was applied to
peat production to yield on-site C02 emissions.
IPCC (2013) emissions factors for direct CH4 emissions for drained land surfaces and drainage
ditches created from peat extraction were applied to the peat production area to yield on-site CH4
emissions.
An IPCC (2013) emissions factor for on-site N20 emissions was applied to the peat production area of
nutrient-rich peat soil only to yield on-site N20 emissions.
5.1.7.2.3. Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02, CH4, and N20 from
peatlands remaining peatlands were calculated using the 2006 IPCC Guidelines Approach 2 methodology
(IPCC 2006). As described further in Chapter 6 of the national Inventory (EPA 2024), levels of uncertainty in
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
5-16
-------
Section 5 Land Use, Land-Use Change, and Forestry (NIR Chapter 6)
the national estimates in 2022 were 16%/+16% for C02, -59%/+79% for CH4, and -52%/+53% for N20.
State-Level estimates have a higher uncertainty due to apportioning national data to the state level and due to
the assumption that any state without data on peat production is a non-producing state. These assumptions
were required due to a general lack of data confirming that states are either producing or non-producing. For
more details on national-level uncertainty, see the uncertainty discussion in Section 6.8 of the national
Inventory.
5.1.7.2.4. Recalculations
Changes that resulted from recalculations to the state-level estimates are the same as those presented
in the national Inventory, given that improvements in the national Inventory will lead directly to improvements
in the quality of state-level estimates (see Section 6.8, page 6-108, of the national Inventory). In particular,
the lower 48 states' peat production estimates were updated using the peat section of the Mineral
Commodity Summaries 2023. The 2023 edition updated 2018 through 2021 national peat estimates (which
are used to estimate state peat production). Changes also occurred in estimates for state peat production
for on-site and off-site C02 emissions due to revised population data for 2020 through 2022.
5.1.7.2.5. Planned Improvements
The planned improvements are consistent with those planned for improving the national estimates,
given that the underlying methods for state GHG estimates are based on those used in the national Inventory.
In addition, the methodology used to estimate state-level emissions will be reviewed and revised over time to
identify other data and update assumptions (e.g., data on consumption, data and approaches for proxy peat
production to better refine where peat is produced). Planned improvements include:
EPA plans to investigate estimating emissions for Hawaii, Puerto Rico, and applicable territories,
pending data availability. Emissions from off-site horticulture use are currently not estimated in non-
conterminous states and territories, even though peat spreading is not limited to conterminous
states.
EPA will continue monitoring for data sources to reduce or eliminate the disparity between
estimated state peat production and the national peat production estimate, especially for
production values in 1990-2000. Some amount of normalization is currently performed for most
years throughout the time series.
To find information on planned improvements to refine methods for estimating emissions and removals
from wetlands remaining wetlands (coastal wetlands remaining coastal wetlands and peatlands remaining
peatlands), see the planned improvements discussion on pages 6-108 described in the national Inventory at
the link provided above.
5.1.7.2.6. References
Cleary, J., N. Roulet, and T.R. Moore (2005) Greenhouse Gas Emissions from Canadian Peat Extraction,
1990-2000: A Life-Cycle Analysis. Ambio, 34:456-461.
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
Instituto de Estadfsticas de Puerto Rico (2021) EstimadosAnuales Poblacionales de los Municipios Desde
1950. Accessed February 2021. Available online at:
https://censo.estadisticas.pr/EstimadosPoblacionales.
5-17
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
IPCC (Intergovernmental Panel on Climate Change) (2006) 2006IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.jp/public/2006gl/.
IPCC (2013) 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories:
Wetlands. T. Hiraishi, T. Krug, K. Tanabe, N. Srivastava, J. Baasansuren, M. Fukuda, and T.G. Troxler
(eds.). Available online at: https://www.ipcc.ch/publication/2013-supplement-to-the-2006-ipcc-guidelines-
for-national-greenhouse-gas-inventories-wetlands/.
U.S. Bureau of Mines (1990) Peat. In: Minerals Yearbook. U.S. Department of the Interior. Available online at:
https://digital.library.wisc.edU/1711 .dl/EZRI27J2VYVCG8G.
U.S. Bureau of Mines (1991) Peat. In: Minerals Yearbook. U.S. Department of the Interior. Available online at:
https://digital.librarv.wisc.edu/1711.dl/5X7AVV22D2URQ8R.
U.S. Bureau of Mines (1992) Peat. In: Minerals Yearbook. U.S. Department of the Interior. Available online at:
https://digital.library.wisc.edu/1711.dl/FYIUSH2IKTZTI8Q.
U.S. Bureau of Mines (1993) Peat. In: Minerals Yearbook. U.S. Department of the Interior. Available online at:
https://digital.librarv.wisc.edU/1711 .dl/2YIJA2GUJDK0B86.
U.S. Census Bureau (2002) Table CO-EST2001-12-00. In: Time Series of I ntercensal State Population
Estimates: April 1, 1990 to April 1,2000. Release date: April 11, 2002. Available online at:
https://www2.census.gov/programs-surveys/popest/tables/1990-2000/intercensal/st-co/co-est20Q1-
12-OO.pdf.
U.S. Census Bureau (2011) Table ST-EST00INT-01. In: Intercensal Estimates of the Resident Population for
the United States, Regions, States, and Puerto Rico: April 1, 2000 to July 1, 2010. Release date:
September 2011. Available online at: https://www2.census.gov/programs-
survevs/popest/datasets/2000-2010/intercensal/state/st-est00int-alldata.csv.
U.S. Census Bureau (2021) Table NST-EST2020. In: Annual Estimates of the Resident Population for the United
States, Regions, States, and Puerto Rico: April 1, 2010 to July 1, 2020. Release date: July 2021.
U.S. Census Bureau (2023) Table NST-EST2023-POP. In: Annual Estimates of the Resident Population for the
United States, Regions, States, District of Columbia, and Puerto Rico: April 1, 2020 to July 1, 2023.
Release date: December 2023.
USGS (U.S. Geological Survey) (2020) Minerals Yearbook: Peat (1994-2018). Available online at:
https://www.usgs.gov/centers/national-minerals-information-center/peat-statistics-and-information.
USGS (2023a) 2019 Minerals Yearbook: Peat. Tables-only release. Available online at
https://www.usgs.gov/centers/national-minerals-information-center/peat-statistics-and-information.
USGS (2023b) 2020 Minerals Yearbook: Peat. Tables-only release. Available online at
https://www.usgs.gov/centers/national-minerals-information-center/peat-statistics-and-information.
USGS (2023c) Mineral Commodity Summaries. Available online at: https://www.usgs.gov/centers/national-
minerals-information-center/mineral-commoditv-summaries.
Full citations of the references included in Chapter 6.8 (Wetlands Remaining Wetlands) of the national
Inventory are available online here: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-inventorv-
2024-chapter-10-references O.pdf.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
5-18
-------
Section 5 Land Use, Land-Use Change, and Forestry (NIR Chapter 6)
5.1.8 Flooded Land Remaining Flooded Land (NIR Section 6.8)
5.1.8.1 Background
Flooded Lands are defined as (1) water bodies where human activities have caused changes in the
amount of surface area covered by water, typically through water level regulation, such as constructing a
dam; (2) water bodies where human activities have changed the hydrology of existing natural water bodies,
thereby altering water residence times and/or sedimentation rates, and in turn causing changes to the
natural emission of GHGs; and (3) water bodies that have been created by excavation, such as canals,
ditches, and ponds (IPCC 2019). Flooded lands include water bodies with seasonally variable degrees of
inundation, but these water bodies would be expected to retain some inundated area throughout the year
under normal conditions.
Flooded lands are broadly classified as "reservoirs" or "other constructed water bodies" (IPCC 2019).
Reservoirs are defined as flooded land greater than 8 hectares and include the seasonally flooded land on
the perimeter of permanently flooded land (i.e., inundation areas). IPCC guidance (IPCC 2019) provides
default emissions factors for reservoirs and several types of other constructed water bodies, including
freshwater ponds and canals/ditches.
Land that has been flooded for more than 20 years is defined as flooded land remaining flooded land
and land flooded for 20 years or less is defined as land converted to flooded land. The distinction is based on
literature reports that CH4 and C02 emissions are high immediately following flooding (as labile organic
matter is rapidly degraded) but decline to a steady background level approximately 20 years after flooding.
Emissions of CH4 are estimated for flooded land remainingflooded land, but C02 emissions are not included
as they are primarily the result of decomposed organic matter entering the waterbody from the catchment or
contained in inundated soils and are included elsewhere in the IPCC guidelines (IPCC 2006).
5.1.8.2 Methods/Approach
EPA compiles state-level emissions from flooded land remaining flooded land using the same methods
applied in the national Inventory. Please see the methodologies described in Chapter 6.8 (pages 6-121
through 6-129) of the national Inventory. For this report, the state-level estimates were developed using
Approach 1. Estimates of emissions from reservoirs and associated inundation areas and other constructed
waterbodies that include freshwater ponds and canals/ditches include all states, tribal lands, the District of
Columbia, and Puerto Rico. Other U.S. territories are not estimated at this time.
5.1.8.3 Uncertainty
The overall uncertainty associated with national estimates from reservoirs and other constructed water
bodies is described in Chapter 6 of the national Inventory (EPA 2024). Uncertainty for both reservoirs and
other constructed waterbodies is developed using IPCC Approach 2. The total uncertainty for reservoirs is
-1.6%/+2.9%, and the total uncertainty for other constructed water bodies is -0.8%/+1.0%. State-level
estimates of uncertainty will vary significantly among the states but, in general, tend to be higher than those
provided for the United States in the national Inventory. For more details on national-level uncertainty, see
the uncertainty discussion in Section 6.8 of the national Inventory.
5.1.8.4 Recalculations
Changes that resulted from recalculations to the state-level estimates are the same as those presented
in Section 6.8 of the national Inventory (pages 6-134 and 6-143), given that improvements in the national
lnventory\N\[[ lead directly to improvements in the quality of state-level estimates as well.
5-19
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
5.1.8.5 Planned Improvements
The planned improvements are anticipated to be the same as those planned for improving the national
estimates, given that the underlying methods for state GHG estimates are the same as those in the national
Inventory and will lead directly to improvements in the quality of state-level estimates as well. To review the
planned improvements to the methods and data for estimating emissions from flooded land remaining
flooded land, see the planned improvements discussion on pages 6-134 and 6-144 of Chapter 6.8 in the
national Inventory.
5.1.8.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
IPCC (2019) 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories. E.C.
Buendia, K. Tanabe, A. Kranjc, J. Baasansuren, M. Fukuda, S. Ngarize A. Osako, Y. Pyrozhenko, P.
Shermanau, and S. Federici (eds.). Available online at: https://www.ipcc.ch/report/2019-refinement-to-
the-2006-ipcc-guidelines-for-national-greenhouse-gas-inventories/.
Full citations of references included in Chapter 6.8 (for flooded land remaining flooded land) of the
national Inventory are available online here: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-
inventorv-2024-chapter-10-references O.pdf.
5.1.9 Land Converted to Wetlands (NIR Section 6.9)
This section describes methods for estimating state-level C02 and CH4 emissions from managing
wetlands, as consistent with the national Inventory, specifically:
Land converted to coastal wetlands (C02 and CH4)
Land converted to flooded land (C02 and CH4)
5.1.9.1 Land Converted to Coastal Wetlands
5.1.9.1.1. Background
Land converted to vegetated coastal wetlands occurs as a result of inundation of unprotected low-lying
coastal areas with gradual sea-level rise, flooding of previously drained land behind hydrological barriers,
and active restoration and creation of coastal wetlands through removing hydrological barriers. Land use
conversions into coastal wetlands can result in C stock changes to all coastal wetland carbon pools (i.e.,
aboveground biomass, belowground biomass, dead wood, litter, and soil organic carbon) and emissions of
CH4 if inundated with fresh water (IPCC 2013). This section provides estimates of C02 and CH4 emissions
and removals resulting from converting cropland, grassland, wetlands, settlements, and other lands to
vegetated coastal wetlands.
5.1.9.1.2. Methods/Approach
To compile national estimates of C stock changes and CH4 emissions from land converted to vegetated
coastal wetlands for the national Inventory, estimates for each state with coastal wetlands and the District of
Columbia, includingtribal lands in these states, were produced and summed into a national total. A
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
5-20
-------
Section 5 Land Use, Land-Use Change, and Forestry (NIR Chapter 6)
description of the methods and data used to estimate state-Level emissions is provided in Chapter 6, Section
6.9 (pages 6-144 through 6-150) of the national Inventory. Please note that estimates for Hawaii, Alaska and
U.S. territories are not included in the national total or available at the state level at this time.
States (plus the District of Columbia) with coastal wetlands currently included in the national Inventory
are Alabama, California, Connecticut, Delaware, Florida, Georgia, Louisiana, Maine, Maryland,
Massachusetts, Mississippi, New Hampshire, New Jersey, New York, North Carolina, Oregon, Pennsylvania,
Rhode Island, South Carolina, Texas, Virginia, and Washington.
5.1.9.1.3. Uncertainty
Underlying uncertainties in estimates of soil carbon removal factors, biomass change, dissolved
organic matter, and CH4 emissions include error in uncertainties associated with Tier 2 literature values of
soil carbon removal estimates, biomass stocks, dissolved organic matter, and IPCC default CH4 emissions
factors; uncertainties linked to interpretating remote sensing data; and assumptions that underlie the
methodological approaches applied. IPCC Approach 1 (IPCC 2006) was used to calculate these
uncertainties. As described further in Chapter 6.9 of the national Inventory (EPA 2024), levels of uncertainty
in the national estimates in 2022 are -20%/+20% for biomass C stock change, -25.8%/+25.8% for dead
organic matter C stock change, -17.7%/+17.7% for soil C stock change, and -29.9%/+29.9% CH4 emissions.
The combined uncertainty across all subcategories is -42.2%/+42.2%. State-level estimates of uncertainty
will vary significantly among the states but, in general, tend to have a higher uncertainty than those provided
for the United States in the national Inventory. For more details on national-level uncertainty see the
uncertainty discussion in Section 6.9 of the national Inventory.
5.1.9.1.4. Recalculations
Changes that resulted from recalculations to the state-level estimates are the same as those presented
in Section 6.9 of the national Inventory (page 6-149), given that improvements in the national Inventory will
lead directly to improvements in the quality of state-level estimates as well.
5.1.9.1.5. Planned Improvements
The planned improvements are consistent with those planned for improving the national estimates
given that the underlying methods for the state GHG estimates are the same as those in the national
Inventory. To review the planned improvements to the methods and data for estimating emissions and
removals from land converted to vegetated coastal wetlands, see the planned improvements discussions on
page 6-150 of Chapter 6.9 in the national Inventory.
5.1.9.1.6. References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006) 2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
IPCC (2013) 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories:
Wetlands. T. Hiraishi, T. Krug, K. Tanabe, N. Srivastava, J. Baasansuren, M. Fukuda, and T.G. Troxler
(eds.). Available online at: https://www.ipcc.ch/publication/2013-supplement-to-the-2006-ipcc-guidelines-
for-national-greenhouse-gas-inventories-wetlands/.
5-21
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Full citations of the references included in Chapter 6.9 (Lands Converted to Coastal Wetlands) of the
national Inventory are available online here: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-
inventorv-2024-chapter-10-references O.pdf.
5.1.9.2 Land Converted to Flooded Land
5.1.9.2.1. Background
Land that has been flooded for less than 20 years is defined as land converted to flooded land. The
distinction is based on literature reports that C02 and CH4 emissions are high immediately following flooding
(as labile organic matter is rapidly degraded) but decline to a steady background level approximately 20 years
after flooding. Both C02 and CH4 emissions are inventoried for both reservoirs and associated inundation
areas and freshwater ponds within the other constructed waterbodies subcategory of land converted to
flooded land.
5.1.9.2.2. Methods/Approach
To compile national estimates of C stock changes and CH4 emissions from land converted to flooded
land for the national Inventory, estimates for each state, tribal lands, the District of Columbia, and Puerto
Rico were produced and summed into a national total. Estimates are not available for other U.S. territories at
this time. A description of the methods and data used to estimate state-level emissions is provided in
Chapter 6, Section 6.9 (pages 6-150 through 6-165) of the national Inventory.
5.1.9.2.3. Uncertainty
The overall uncertainty associated with national estimates of C02 and CH4 from reservoirs and other
constructed water bodies on flooded land remaining flooded land is described in Chapter 6.9 of the national
Inventory (EPA 2024). Uncertainty for both reservoirs and other constructed water bodies is developed using
IPCC Approach 2. The total uncertainty for C02 and CH4 emissions from reservoirs is 12.2%/+18.8%, and
the total uncertainty for C02 and CH4 emissions from other constructed waterbodies is -1.8%/+2.1%. State-
level estimates of uncertainty will vary significantly among the states but, in general, tend to be higher than
those provided for the United States in the national Inventory. For more details on national-level uncertainty,
see the uncertainty discussion in Section 6.9 of the national Inventory.
5.1.9.2.4. Recalculations
Changes that resulted from recalculations to the state-level estimates are the same as those presented
in Section 6.9 of the national Inventory (pages 6-158 and 6-165), given that improvements in the national
lnventory\N\[[ lead directly to improvements in the quality of state-level estimates as well.
5.1.9.2.5. Planned Improvements
The planned improvements are consistent with those planned for improving the national estimates
given that the underlying methods for the state GHG estimates are the same as those in the national
Inventory. To review the planned improvements to the methods and data for estimating emissions and
removals from land converted to flooded land, see the planned improvements discussions on pages 6-158
and 6-165 of Chapter 6.9 in the national Inventory.
5.1.9.2.6. References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventory-us-
greenhouse-gas-emissions-and-sinks.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
5-22
-------
Section 5 Land Use, Land-Use Change, and Forestry (NIR Chapter 6)
Full citations of the references included in Chapter 6.9 (for Land converted to flooded land) of the
national Inventory are available online here: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-
inventorv-2024-chapter-10-references O.pdf.
5.1.10 Settlements Remaining Settlements (NIR Section 6.10)
This section presents methods used to estimate state-level C02, CH4, and N20 emissions and removals
from settlements remaining settlements consistent with the national Inventory. Settlements are land uses
where human populations and activities are concentrated. The section is organized to address the following
subcategories:
C02 emissions from drained organic soils (C02)
Changes in C stocks in settlement trees (C02)
N20 emissions from settlement soils (N20)
C stock changes in landfilled yard trimmings and food scraps (C02)
5.1.10.1 Soil C Stock Changes
5.1.10.1.1. Background
Soil organic C stock changes for settlements remaining settlements occur in both mineral and organic
soils. However, the United States does not estimate changes in soil organic C stocks for mineral soils in
settlements remaining settlements. This approach is consistent with the assumption of the Tier 1 method in
the 2006 IPCC Guidelines (IPCC 2006) that inputs equal outputs, and therefore the soil organic C stocks do
not change. In contrast, drainage of organic soils can lead to continued losses of carbon for an extended
period of time.
Drainage of organic soils is common when wetland areas have been developed for settlements. Organic
soils, also referred to as Histosols, include all soils with more than 12%-20% organic carbon by weight,
depending on clay content. The organic layer of these soils can be very deep (i.e., several meters), and form
under inundated conditions that result in minimal decomposition of plant residues. Drainage of organic soils
leads to aeration of the soil that accelerates decomposition rate and C02 emissions. Due to the depth and
richness of the organic layers, carbon loss from drained organic soils can continue over long periods of time,
which vary depending on climate and composition (i.e., decomposability) of the organic matter. See Chapter
6 of the national Inventory for more information (EPA 2024).
5.1.10.1.2. Methods/Approach
EPA compiles state-level estimates of soil C stock changes using the same methods applied in the
national Inventory. Please see the methodologies described in Chapter 6, Section 6.10 (pages 6-166 through
6-169) of the national Inventory. EPA used a hybrid of Approach 1 and Approach 2 for state-level estimates.
The current national Inventory includes state-level emissions for the years 1990-2017 for soil organic C stock
changes. The remaining years in the time series were only estimated at the national scale using a linear
extrapolation method, and a two-step process was used to approximate the state-level emissions for the
remaining years. First, the average proportion of the total national emissions was computed for each state
for the years 2015-2017. Second, the state-level proportions were multiplied by the total national emissions
to approximate the amount of emissions occurring in each state for 2018-2022. Estimates are included for all
states, but no estimates for the District of Columbia and U.S. territories are available for this report. The
USDA National Resources Inventory covers tribal and trust lands.
5-23
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
5.1.10.1.3. Uncertainty
The overall uncertainty associated with national estimates from soil C stock changes is described in
Chapter 6 of the national Inventory (EPA 2024). Uncertainty for the Tier 2 approach is derived using a Monte
Carlo approach. The uncertainty for total soil C stock changes in 2022 is -50%/+50%.
5.1.10.1.4. Recalculations
Recalculations were applied consistent with the national Inventory (see Section 6.01, page 6-168).
5.1.10.1.5. Planned Improvements
The planned improvements are consistent with those planned for improving the national estimates
given that the underlying methods for the state GHG estimates are the same as those in the national
Inventory. To review the planned improvements to the methods and data for estimating emissions and
removals from soil C stock changes, see the planned improvements discussions on page 6-168 of Chapter
6.10 in the national Inventory.
5.1.10.1.6. References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventory-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006) 2006IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
Full citations of the references included in Chapter 6.10 (for soil C stock changes) of the national
Inventory are available online here: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-inventorv-
2024-chapter-10-references O.pdf.
5.1.10.2 Changes in C Stocks in Settlement Trees
5.1.10.2.1. Background
In settlement areas, the anthropogenic impacts on tree growth, stocking, and mortality are particularly
pronounced (Nowak2012) in comparison to forest lands where non-anthropogenic forces can have more
significant impacts. Trees in settlement areas of the United States are a significant sink over the time series.
Dominant factors affecting carbon flux trends for settlement trees are changes in the amount of settlement
area (increasing sequestration due to more settlement lands and trees) and net changes in tree cover (e.g.,
tree losses versus tree gains through planting and natural regeneration), with percent tree cover trending
downward recently. In addition, changes in species composition, tree sizes, and tree densities affect base
carbon flux estimates. Annual sequestration increased by 43% between 1990 and 2022 due to increases in
settlement area and changes in tree cover. Trees in settlements often grow faster than forest trees because
of their relatively open structure (Nowak and Crane 2002). Because tree density in settlements is typically
much lower than in forested areas, the C storage per hectare of land is in fact smaller for settlement areas
than for forest areas. Also, percent tree cover in settlement areas is less than in forests, and this tree cover
varies significantly across the United States (e.g., Nowak and Greenfield 2018).
5.1.10.2.2. Methods/Approach
To compile national estimates of C02 emissions and removals from C stock changes from settlement
trees for the national Inventory, estimates for all 50 states and the District of Columbia were produced and
summed into a national total. Estimates for U.S. territories are not available at this time. In this case, EPA is
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
5-24
-------
Section 5 Land Use, Land-Use Change, and Forestry (NIR Chapter 6)
applying an Approach 1 method (i.e., using methods consistent with the national Inventory). A description of
the methods and data used to estimate changes in C stocks in settlement trees is found in Chapter 6,
Section 6.10 (pages 6-169 through 6-177), of the national Inventory (EPA 2024).
5.1.10.2.3. Uncertainty
Uncertainty associated with changes in C stocks in settlement trees includes the uncertainty
associated with settlement area, percent tree cover in developed land and how well it represents percent
tree cover in settlement areas, and estimates of gross and net carbon sequestration for each of the 50 states
and the District of Columbia. Additional uncertainty is associated with the biomass models, conversion
factors, and decomposition assumptions used to calculate carbon sequestration and emission estimates
(Nowak et al. 2002). These results also exclude changes in soil C stocks, and there is likely some overlap
between the settlement tree carbon estimates and the forest tree carbon estimates (e.g., Nowak et al. 2013).
IPCC Approach 2 (IPCC 2006) was used to calculate these uncertainties. As described further in Chapter
6.10 of the national Inventory (EPA 2024), levels of uncertainty in the national estimates in 2022 for C stock
change are -51%/+52%. State-level estimates of uncertainty will vary significantly among the states but, in
general, tend to have a higher uncertainty than those provided for the United States in the national Inventory.
For more details on national-level uncertainty see the uncertainty discussion in Section 6.10 of the national
Inventory.
5.1.10.2.4. Recalculations
Changes that resulted from recalculations to the state-level estimates are the same as those presented
in Section 6.10 of the national Inventory (pages 6-176 and 6-177), given that improvements in the national
lnventory\N\[[ lead directly to improvements in the quality of state-level estimates as well.
5.1.10.2.5. Planned Improvements
The planned improvements are consistent with those planned for improving national estimates given
that the underlying methods for state GHG estimates are the same as those in the national Inventory. To
review planned improvements to refine methods for estimating changes in settlement tree C stocks, see the
planned improvements discussion on page 6-177 of Section 6.10 in the national Inventory for a description of
future work to further refine these estimates.
5.1.10.2.6. References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventory-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006) 2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
Nowak, D.J. (2012) Contrasting Natural Regeneration and Tree Planting in 14 North American Cities. Urban
Forestry and Urban Greening, 11: 374-382.
Nowak, D.J., and D.E. Crane (2002) Carbon Storage and Sequestration by Urban Trees in the United States.
Environmental Pollution, 116(3): 381-389.
Nowak, D.J., and E.J. Greenfield (2018) U.S. Urban Forest Statistics, Values and Projections. Journal of
Forestry, 116(2): 164-177.
5-25
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Nowak, D.J., D.E. Crane, J.C. Stevens, and M. Ibarra (2002) Brooklyn ' s Urban Forest. General Technical
Report NE290. U.S. Department of Agriculture, Forest Service.
Nowak, D.J., E.J. Greenfield, R.E. Hoehn, and E. Lapoint (2013) Carbon Storage and Sequestration by Trees in
Urban and Community Areas of the United States. Environmental Pollution, 178: 229-236.
Full citations of the references included in Chapter 6.10 (for changes in C stocks in settlement trees) of
the national Inventory are available online here: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-
inventorv-2024-chapter-10-references O.pdf.
5.1.10.3 N20 Emissions from Settlement Soils
5.1.10.3.1. Background
Of the synthetic nitrogen fertilizers applied to soils in the United States, approximately 1-2% are
currently applied to lawns, golf courses, and other landscaping within settlement areas, and contribute to
soil N20 emissions. The area of settlements is considerably smaller than other land uses that are managed
with fertilizer, particularly cropland soils, and therefore, settlements account for a smaller proportion of total
synthetic fertilizer application in the United States. In addition to synthetic nitrogen fertilizers, a portion of
surface-applied biosolids (i.e., treated sewage sludge) is used as an organic fertilizer in settlement areas,
and drained organic soils (i.e., soils with high organic matter content, known as Histosols) also contribute to
emissions of soil N20.
Nitrogen additions to soils result in direct and indirect N20 emissions. Direct emissions occur on-site
due to the nitrogen additions in the form of synthetic fertilizers and biosolids, as well as enhanced
mineralization of nitrogen in drained organic soils. Indirect emissions result from fertilizer and biosolids
nitrogen that is transformed and transported to another location in a form other than N20 (i.e., NH3 and
nitrogen oxide volatilization, nitrate leaching and runoff), and later converted into N20 at the off-site location.
The indirect emissions are assigned to settlements because the management activity leading to the
emissions occurred in settlements (EPA 2024).
5.1.10.3.2. Methods/Approach
EPA compiles state-level estimates of N20 emissions from settlement soils using the same methods
applied in the national Inventory. Please see the methodologies described in Chapter 6, Section 6.10 (pages
6-177 through 6-180) of the national Inventory. EPA applied a hybrid Approach 1 and Approach 2 for state-
level estimates. The current national Inventory includes state-level emissions for the years 1990-2017 for
synthetic nitrogen and nitrogen inputs from drained organic soils. The remainingyears in the time series were
only estimated at the national scale using a surrogate data method, and a two-step process was used to
approximate the state-level emissions for the remaining years. First, the average proportion of the total
national emissions was computed for each state for the years 2015-2017. Second, the state-level
proportions were multiplied by the total national emissions to approximate the amount of emissions
occurring in each state for 2018-2022. Soil N20 emissions for additions of biosolid nitrogen are only
estimated at the national scale for the entire time series. For this source of nitrogen, soil N20 emissions were
disaggregated to the state level based on the proportion of the U.S. population occurring in each state.
Estimates are included for all states except Alaska, DC, and U.S. territories.
5.1.10.3.3. Recalculations
Changes that resulted from recalculations to the state-level estimates are the same as those presented
in Section 6.10 of the national Inventory (page 6-180), given that improvements in the national Inventory will
lead directly to improvements in the quality of state-level estimates as well.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
5-26
-------
Section 5 Land Use, Land-Use Change, and Forestry (NIR Chapter 6)
5.1.10.3.4. Uncertainty
The overall uncertainty associated with national estimates from N20 emissions from settlement soils is
described in Chapter 6 of the national Inventory (EPA 2024). As described:
The amount of N20 emitted from settlement soils depends not only on nitrogen inputs and area of
drained organic soils, but also on a large number of variables that can influence rates of nitrification
and denitrification, including organic carbon availability; rate, application method, and timing of
nitrogen input; oxygen gas partial pressure; soil moisture content; pH; temperature; and
irrigation/watering practices. The effect of the combined interaction of these variables on N20
emissions is complex and highly uncertain. The IPCC default methodology does not explicitly
incorporate any of these variables, except variation in the total amount of fertilizer nitrogen and
biosolids application, which then leads to uncertainty in the results.
Uncertainties exist in both the fertilizer nitrogen and biosolids application rates in addition to the
emissions factors. Uncertainty in the area of drained organic soils is based on the estimated
variance from the NRI survey. For biosolids, there is uncertainty in the amounts of biosolids applied
to nonagricultural lands and used in surface disposal. These uncertainties are derived from
variability in several factors, including nitrogen content of biosolids, total sludge applied in 2000,
wastewater existing flow in 1996 and 2000, and the biosolids disposal practice distributions to
nonagricultural land application and surface disposal. In addition, there is uncertainty in the direct
and indirect emissions factors that are provided by IPCC (2006).
Uncertainty is propagated through the calculations of N20 emissions from fertilizer nitrogen and
drainage of organic soils based on a Monte Carlo analysis. The overall levels of uncertainty for national
Inventory direct N20 emissions from soils and indirect N20 emissions are -47%/+54% and -76%/+218%,
respectively.
5.1.10.3.5. Planned Improvements
The planned improvements are consistent with those planned for improving the national estimates
given that the underlying methods for the state GHG estimates are the same as those in the national
Inventory. To review the planned improvements to the methods and data for estimating N20 emissions from
settlement soils, see the planned improvements discussions on page 6-180 of Chapter 6.10 in the national
Inventory.
5.1.10.3.6. References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventory-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006) 2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
Full citations of the references included in Chapter 6.10 (for N20 emissions from soils) of the national
Inventory are available online here: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-inventorv-
2024-chapter-10-references O.pdf.
5-27
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
5.1.10.4 Changes in Yard Trimmings and Food Scrap C Stocks in Landfills
5.1.10.4.1. Background
In the United States, yard trimmings (i.e., grass clippings, Leaves, and branches) and food scraps
account for a significant portion of the municipal waste stream, and a Large fraction of the collected yard
trimmings and food scraps are put in landfills. Carbon contained in landfilled yard trimmings and food scraps
can be stored for very long periods. C stock changes in yard trimmings and food scraps and associated C02
emissions and removals are reported under settlements remaining settlements because the bulk of the
carbon, which comes from yard trimmings, originates from settlement areas. While the majority of food
scraps originate from cropland and grassland, in the national Inventory they are reported with the yard
trimmings in the settlements remaining settlements section. Additionally, landfills are considered part of the
managed land base under settlements and reporting these C stock changes that occur entirely within
landfills fits most appropriately in the settlements remaining settlements section.
5.1.10.4.2. Methods/Approach
State-level C stocks were compiled using Approach 2 by allocating net national changes in C stocks and
associated emissions and removals to all 50 states and the District of Columbia based on their fraction of
total U.S. land area classified as urban area. U.S. territories are not estimated usingthis method. "Urban
area" is defined by USDA as land area containing densely populated areas with at least 50,000 people
(urbanized areas) and densely populated areas with 2,500 to 50,000 people (urban clusters). EPA assumed
"urban area" matched the definition of "settlements area" for the purpose of state-level estimates. This
approach was applied due to unavailability of state-level activity data on mass of yard trimmings and food
scraps discarded to managed landfills, and the assumption that most yard trimmings and food scraps would
be generated in densely populated areas. EPA used settlement area estimates from the USDA Economic
Research Service's Major Land Uses data. The total settlements area and changes in carbon stock estimates
in the United States includes all U.S. states and the District of Columbia but excludes territories such as
Puerto Rico at this time. EPA does not have disaggregated tribal versus state urban land area available within
its data sources for the estimation of state-level emissions.
State emissions were calculated using the following stepwise process:
1. EPA obtained U.S. settlements area data from USDA (2017). For years without U.S. settlements area
data (2013-2022), settlements area data were forecast using 2002-2012 data to capture the most
recent available trends.
2. The fraction of total settlements area was calculated for each state, including the District of
Columbia, by dividing the state settlements area by the U.S. total settlements area.
3. The state fraction of settlements area was multiplied by the total national yard trimmings and food
scraps C stocks from the 1990-2022 national Inventory to estimate state-level yard trimming and
food scrap C stocks. This calculation was also performed for grass, leaves, branches, and food
scraps to yield state-level C stocks for each subcategory.
Data Appendix E-9 to this report provides activity data related to total land in urban areas and percent of
total land area that occurs in urban areas by state (including the District of Columbia) across the time series.
These data are used in the calculations of carbon storage in landfilled yard trimmings and food scraps in
each state.
5.1.10.4.3. Uncertainty
The overall uncertainty associated with the 2022 national estimates of C02 from changes in yard
trimmings and food scraps C stocks were calculated using the Approach 2 methodology (IPCC 2006). As
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
5-28
-------
Section 5 Land Use, Land-Use Change, and Forestry (NIR Chapter 6)
described further in Chapter 6 of the national Inventory (EPA 2024), pages 6-185 and 6-186, Levels of
uncertainty in the national estimates in 2022 were -47%/+58% for C02. State-level estimates have a higher
uncertainty due to apportioning the national estimates to states based on their fraction of the settlements
area. These assumptions were required because of a general lack of available state-level data on yard
trimmings and food scraps. For more details on national-level uncertainty, see the uncertainty discussion in
Section 6.10 of the national Inventory.
5.1.10.4.4. Recalculations
Changes that resulted from recalculations to the state-level estimates in 2021 were due to expected
forecasted data changes and are reflected in the national Inventory; see Section 6.10, page 6-186.
5.1.10.4.5. Planned Improvements
EPA will review and revise the state-level methodology over the time series, and as appropriate, assess
if other information would better reflect state-level activity (e.g., mass of yard trimmings and food scraps
discarded to managed landfills) to improve the accuracy of the estimates. EPA plans to incorporate updated
Urban Area data in the next Inventory year. Sources of settlements area data for Puerto Rico and other U.S.
territories are also needed to provide a more accurate estimate of net C stock changes in the United States.
Additional planned improvements are consistent with those planned for improving national estimates given
that the underlying methods for state GHG estimates are derived from those in the national Inventory. For
example, updated data are expected in a new release of the Advancing Sustainable Materials Management:
Facts and Figures report for 2019, 2020, 2021, and 2022. The discussion of planned improvements to refine
methods for estimating changes in C stocks in landfilled yard trimmings at the national level starts on page 6-
186 of Chapter 6.10 in the national Inventory.
5.1.10.4.6. References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. U.S. Environmental Protection Agency. EPA 430-R-24-004. Available online at:
https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006) 2006IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
USDA (U.S. Department of Agriculture) (2017) Urban Area, 1945-2012, by State: Densely-Populated Areas
with At Least 50,000 People (Urbanized Areas) and Densely-Populated Areas with 2,500 to 50,000 People
(Urban Clusters). Available online at: https://www.ers.usda.gov/data-products/major-land-uses/.
Full citations of all other references relevant to estimating landfilled yard trimmings and food scraps C
stock changes included in Chapter 6.10 (Settlements Remaining Settlements) are available online here:
https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-inventorv-2024-chapter-10-references O.pdf.
5.1.11 Land Converted to Settlements (NIR Section 6.11)
5.1.11.1 Background
Land converted to settlements includes all settlements in an inventory year that had been in at least one
other land use during the previous 20 years. For example, cropland, grassland, or forest land converted to
settlements during the past 20 years would be reported in this category. Converted lands are retained in this
category for 20 years as recommended by IPCC (2006). The national Inventory includes all settlements in the
5-29
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
United States except Alaska. Areas of drained organic soils on settlements in federal lands are also not
included in the national Inventory.
Land use change can lead to large losses of carbon to the atmosphere, particularly conversions from
forest land. Moreover, conversion of forests to another land use (i.e., deforestation) is one of the largest
anthropogenic sources of emissions to the atmosphere globally, although this source may be declining
globally according to a recent assessment.
IPCC (2006) recommends reporting changes in biomass, dead organic matter, and soil organic C stocks
due to land use change. All soil organic C stock changes are estimated and reported for land converted to
settlements, but there is limited reporting of other pools in the national Inventory. Loss of aboveground and
belowground biomass, dead wood, and litter carbon are reported for forest land converted to settlements,
but not for other land use conversions to settlements (EPA 2024).
5.1.11.2 Methods/Approach
EPA compiles state-level estimates of land converted to settlements using the same methods applied in
the national Inventory. Please see the methodologies described in Chapter 6, Section 6.11 (pages 6-187
through 6-194), of the national Inventory. EPA used a hybrid Approach 1 and Approach 2 for state-level
estimates. The current national Inventory includes state-level emissions for the years 1990-2022 for
biomass, standing dead, dead wood, and litter, and for the years 1990-2017 for soil organic C stock changes.
The remainingyears in the time series for soil organic C stock changes were only estimated at the national
scale using a surrogate data method, and a two-step process was used to approximate the state-level
emissions for the remaining years. First, the average proportion of the total national emissions was
computed for each state for the years 2015-2017. Second, the state-level proportions were multiplied by the
total national emissions to approximate the amount of emissions occurring in each state for 2018-2022.
Estimates are included for all states except Alaska. The USDA National Resources Inventory covers tribal and
trust lands. Emissions from land converted to settlements do not occur in the District of Columbia, and
emissions for U.S. territories (including Puerto Rico) are not estimated at this time.
5.1.11.3 Uncertainty
The overall uncertainty associated with national estimates from land converted to settlements is
described in Chapter 6 of the national Inventory (EPA 2024). As described:
The uncertainty analysis for carbon losses for forest land converted to settlements is conducted in
the same way as the uncertainty assessment for forest ecosystem carbon flux in the forest land
remaining forest land category. For additional details, see the uncertainty analysis in Annex 3.13.
The uncertainty analysis for mineral soil organic C stock changes and annual carbon emission
estimates from drained organic soils in land converted to settlements is estimated using a Monte
Carlo approach, which is also described in the cropland remaining cropland section of the national
Inventory.
The overall level of uncertainty for national Inventory land converted to settlements estimates is -36%/+36%.
5.1.11.4 Recalculations
Changes that resulted from recalculations to the state-level estimates are the same as those presented
in Section 6.11 of the national Inventory (page 6-193), given that improvements in the national Inventory will
lead directly to improvements in the quality of state-level estimates as well.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
5-30
-------
Section 5 Land Use, Land-Use Change, and Forestry (NIR Chapter 6)
5.1.11.5 Planned Improvements
The planned improvements are consistent with those planned for improving the national estimates
given that the underlying methods for the state GHG estimates are the same as those in the national
Inventory. To review the planned improvements to the methods and data for estimating emissions and
removals from land converted to settlements, see the planned improvements discussions on pages 6-193
and 6-194 of Chapter 6.11 in the national Inventory.
5.1.11.6 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006) 2006IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
Full citations of the references included in Chapter 6.11 (Land Converted to Settlements) of the national
Inventory are available online here: https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-inventorv-
2024-chapter-10-references O.pdf.
5.1.12 Other Land Remaining Other Land (NIR Section 6.12) and Land Converted to
Other Land (NIR Section 6.13)
Other Land is a land use category that includes bare soil, rock, ice, and all land areas that do not fall into
any of the other five land use categories (i.e., forest land, cropland, grassland, wetlands, and settlements).
Followingthe guidance provided by IPCC (2006), C stock changes and non-C02 emissions are not estimated
for other land because these areas are largely devoid of biomass, litter, and soil carbon pools. However, C
stock changes and non-C02 emissions are estimated for land converted to other land during the first 20 years
following conversion to account for legacy effects. While the magnitude of these area changes is known (see
national Inventory, page 6-11, Tables 6-4 and 6-5), research is ongoing to track carbon across other land
remaining other land and land converted to other land. Until reliable and comprehensive estimates of carbon
across these land use and land use change categories can be produced, it is not possible to separate C02,
CH4, or N20 fluxes on land converted to other land from fluxes on other land remaining other land. Emissions
and removals from other lands and lands converted to other lands will be included in future versions of this
publication when they are integrated into the national Inventory. See Chapters 6.12 and 6.13 on page 6-194
of the national Inventory (EPA 2024).
5.1.12.1 References
EPA (U.S. Environmental Protection Agency) (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006) 2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
5-31
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
6 Waste (NIR Chapter 7)
For this methodology report, the Waste chapter consists of two subsectors: solid waste disposal and
wastewater treatment and discharge. More information on national-level emissions and methods is available
in Chapter 7 of the national Inventory. Note that emissions from waste incineration are discussed in Chapter
2, Section 2.1.4, of this methodology report. Table 6-1 summarizes the different approaches used to estimate
state-level waste emissions and completeness across states. Geographic completeness is consistent with
the national Inventory. The sections below provide more detail on each category.
Table 6-1. Overview of Approaches for Estimating State-Level Waste Sector GHG Emissions and Sinks
Category
Gas
Approach
Geographic Completeness8
Landfills
ch4
Approach 2
Includes emissions from
industrial and municipal
waste landfills from all states,
the District of Columbia,
tribal lands, and some
territories (i.e., Guam, Puerto
Rico) as applicable.
Wastewater
ch4; n2o
Approach 2
Includes emissions from all
states, the District of
Columbia, tribal lands, and
some territories (i.e.,
American Samoa, Guam,
Northern Mariana Islands,
Puerto Rico, and U.S. Virgin
Islands for domestic
wastewater) as applicable.
Composting
ch4)N20
Approach 2
Includes emissions from
commercial composting
facilities from all states, the
District of Columbia, tribal
lands, and some territories
(Puerto Rico) as applicable.
Anaerobic Digestion at
(Stand-Alone) Biogas
Facilities
ch4
Approach 2
Includes emissions from all
states, the District of
Columbia, tribal lands, and
territories.
a Emissions are likely occurring in other U.S. territories; however, due to a lack of available data and the nature of this
category, this analysis includes emissions for only the territories indicated. Territories not listed are not estimated.
6.1 Solid Waste Disposal
This section presents the methodology used to estimate the emissions from solid waste disposal
management activities, which consist of the following sources:
Landfills (MSW and industrial waste) (CH4)
Composting (CH4, N20)
Anaerobic digestion at biogas facilities (stand-alone) (CH4)
6-1
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
6.1.1 Landfills (NIR Section 7.1)
6.1.1.1 Background
After being placed in a Landfill, organic waste such as paper, food scraps, and yard trimmings is initially
decomposed by aerobic bacteria. After the oxygen has been depleted, the remaining waste is available for
consumption by anaerobic bacteria, which break down organic matter into substances such as cellulose,
amino acids, and sugars. These substances are further broken down through fermentation into gases and
short-chain organic compounds that form the substrates for the growth of methanogenic bacteria. These
CH4-producing anaerobic bacteria convert the fermentation products into stabilized organic materials and
biogas consisting of approximately 50% biogenic C02 and 50% CH4 by volume. CH4 and C02are the primary
constituents of landfill gas generation and emissions. Consistent with the 2006 IPCC Guidelines, net C02 flux
from C stock changes in landfills are estimated and reported under the LULUCF sector (see Section 3.4 of
this report) (IPCC 2006).
More information on emission pathways and national-level emissions from landfills and associated
methods can be found in the Waste chapter (Chapter 7), Section 7.1, of the national Inventory (EPA 2024),
available online at https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-inventorv-2Q24-chapter-7-
waste 04-17-2024.pdf.
6.1.1.2 Methods/Approach (Municipal Solid Waste Landfills)
The MSW landfill state emissions inventories apply Approach 2 for disaggregating national estimates
and rely heavily on the Subpart HH data collected through the GHGRP. As explained in the methodology
discussion of Section 7.1 of the national Inventory, EPA uses an IPCC Tier 2 approach and several data
sources, methods, and assumptions to estimate emissions (see pages 7-8 through 7-9 for details on the
inputs and equations). The state inventories apply a state percentage of either waste landfilled or net CH4
emissions by state as reported to Subpart HH (EPA 2022) as a proxy for each state's share of CH4 net
emissions over the time series. Table 6-2 summarizes the methodology used to develop the state-level
estimates, followed by additional detail. The annual state percentages were applied to the national
estimates to retain an IPCC Tier 2 approach consistent with the national Inventory.
Table 6-2. Summary of Approaches to Disaggregate the National Inventory for MSW Landfills
Across Time Series
Time Series
Range
Summary of Method
1990-2009
Applied the percentage of waste landfilled by state (aggregated total as reported
by landfills in each state to Subpart HH for historical years) to the national CH4
net emissions for each year (IPCC 2006Tier 2)
The state percentage approach accounts for all emissions, including those
calculated in the national Inventory through back-casting Subpart HH data and
scaling up emissions to account for smaller landfills that do not report through
Subpart HH.
2010-2022
Applied the percentage of net CH4 emissions by state (aggregated total as
reported by landfills in each state to Subpart HH) to the national CH4 net
emissions for each year.
The state percentage approach accounts for all emissions, including those
calculated by scaling up emissions to account for smaller landfills that do not
report through Subpart HH.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
6-2
-------
Section 6 Waste (NIR Chapter 7)
Historical waste disposed of since a facility began operating is reported using prescribed methods in the
rule to maintain consistency across the facility data. The quantity of waste landfilled by Subpart HH reporters
was assumed to be representative of the universe of MSW landfills in the United States because Subpart HH
reporters include each state's highest emitting MSW landfills, which is directly tied to the quantity of waste
landfilled. The national Inventory methodology back-casts Subpart HH net CH4 emissions and uses a scale-
up factor to account for lower-emitting MSW landfills (e.g., non-reporters). The intent of the scale-up factor is
to estimate CH4 emissions from MSW landfills that do not report to the GHGRP. EPA has put significant effort
into identifying landfills that do not report to the GHGRP, most recently in 2021. Basic landfill characteristics
such as the landfill's name and location, first year of operation, current operational status, and waste-in-
place data have been compiled for these landfills when available. Disaggregating the Subpart HH data by
state was determined to be a reasonable assumption considering the lack of historical data for landfills that
do not report to the GHGRP.
The methodology used for 1990-2009 applies a state percentage of waste landfilled for this time frame
as reported by landfills under Subpart HH of the GHGRP to the national estimates of CH4 emissions.
Approximately 1,200 MSW landfills have reported to the GHGRP since reporting began in 2010. This approach
disaggregates national net emissions values by applying the state percentage as a proxy of net emissions.
The methodology for 2010-2022 applies a state percentage of net CH4 emissions reported by landfills
under Subpart HH to the national estimates of CH4 emissions. Using net CH4 emissions is consistent with the
recent methodological refinements in the national Inventory to incorporate reported Subpart HH net CH4
emissions. Unlike the national Inventory, scale-up factors for each state were not developed since these
would require significant effort; instead, the national emissions values are disaggregated by a proxy that is
assumed to be generally representative of state-by-state emissions.
Emissions from managed landfills located in Puerto Rico and Guam are included because facilities in
these territories report to Subpart HH.
6.1.1.3 Methods/Approach (Industrial Landfills)
The state inventories estimate CH4 emissions from industrial waste landfills for two industry categories
consistent with the national Inventory: (1) pulp and paper and (2) food and beverage. Data reported to the
GHGRP on industrial waste landfills suggest that most of the organic waste that would result in CH4
emissions is disposed of at pulp and paper and food processing facilities. Information on both industry
categories with respect to the amount of waste generated and disposed of in a dedicated industrial waste
landfill is limited; thus, EPA uses a Tier 1 approach to estimate CH4 emissions. Additionally, no
comprehensive list of industrial waste landfills exists. While the information is available in the Waste
Business Journal (WBJ), the date of data related to each waste management facility included is unknown.
Therefore, EPA does not have information on the number of industrial waste landfills that were operational
over the time series and information regarding the number of industrial waste landfills located in each state.
The types and amounts of waste disposed of in the operational industrial waste landfills are also limited.
A portion of pulp and paper mills in the United States report to Subpart TT (Industrial Waste Landfills) of
the GHGRP. Previous analyses of the 2016 pulp and paper emissions from the GHGRP (RTI International
2018) showed that total Subpart TT emissions from facilities associated with a pulp and paper NAICS code
generally align (within approximately 10-20%) with the national Inventory's national estimate of emissions
from the pulp and paper manufacturing sector. On the other hand, a small number of facilities associated
with a food and beverage NAICS code report to Subpart TT, and these emissions are vastly different between
Subpart TT and the national Inventory.
6-3
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Because of the data [imitations described above, Approach 2 was used to disaggregate the national
Inventory CH4 emissions for both industry categories, rather than a more detailed facility-specific, bottom-up
approach.
6.1.1.3.1. Pulp and Paper Manufacturing
For the pulp and paper source category, EPA extracted a state-by-state count of mills in the United
States from two sources: Data Basin for 2008 and Mills OnLine for 2015-2016 (Conservation Biology Institute
2008; Ga Tech Center for Paper Business and Industry Studies n.d.). The count of facilities is approximately
233 and 332 from Data Basin and Mills OnLine, respectively. The count and percentage of mills by state are
shown in Appendix F (Table F-1). Accordingto the Industrial Resources Council, mills are located in 41
states, not includingAlaska, Colorado, North Dakota, Nebraska, Nevada, Rhode Island, South Dakota, Utah,
and Wyoming. For comparison, the Subpart TT pulp and paper facilities across RYs 2011-2019 represent a
maximum of 92 facilities located across 21 states.
To estimate CH4 generation and emissions, the Data Basin 2008 percentages by state were applied to
the national Inventory estimate for the pulp and paper manufacturing sector for 1990-2010, and the Mills
OnLine 2015-2016 percentages by state were applied for 2011-2022. This approach assumes broadly that
each facility is generating an equal amount of waste that is landfilled and, therefore, an equal amount of CH4
emissions. Consistent with the national Inventory, this assumption and this approach were used in an
attempt to ensure complete coverage of industrial waste landfills in the United States because the Subpart
TT pulp and paper facilities may not equal the total number of pulp and paper facilities disposing of waste in
dedicated industrial waste landfills. The exact number of pulp and paper manufacturingfacilities that
dispose of waste in industrial waste landfills is unknown.
CH4 emissions from the pulp and paper sector were disaggregated by applying the percentage of the
mills by state as a proxy for facilities generating and disposing of waste in industrial waste landfills. No
additional calculations were performed, and the IPCC Tier 1 methodology (IPCC 2006) used to generate the
national emissions estimates was applied by default.
6.1.1.3.2. Food and Beverage Manufacturing
Minimal data are available to characterize the amounts and types of waste generated nationally from
food and beverage manufacturers and disposed of in industrial waste landfills. Less is known about the
number of facilities in each state that dispose of waste in a dedicated industrial landfill.
A similar approach using a count of assumed industrial food and beverage manufacturing facilities that
dispose of waste in an industrial waste landfill by state was applied to the national food and beverage
category estimates. The list of food and beverage manufacturing facilities consists of 13 NAICS codes as
shown in Appendix F (Table F-2) comprising 9,175 facilities. This list (which can be shared on request) was
extracted from the 2021 update to the EPA Excess Food Opportunities database (EPA 2021].
The EPA Excess Food Opportunities database includes a low- and high-end estimate of the amount of
excess food generated (tons/year). These data were not used in the methodology. Rather, the average
percentage of the amount of excess food generated by each state across the selected NAICS codes was
used as a proxy for the share of CH4 generation and emissions estimates. The same approach used for the
pulp and paper manufacturing sector was applied whereby the average percentage of excess food by state
was applied to the national total amount of CH4 generation and CH4 emissions for each year of the time
series. This is a broad assumption but allows for the calculation of emissions with limited knowledge on the
locations of facilities disposing of food waste into industrial waste landfills.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
6-4
-------
Section 6 Waste (NIR Chapter 7)
The percentage of excess food generated by state is presented in Appendix F (see Table F-3). Note that
the Excess Food Opportunities database and map do not indicate the management pathway for the excess
food. The EPA Facts and Figures methodology (EPA 2020) also does not include an estimation of the amount
of excess food being disposed of in industrial waste landfills. Therefore, the percentage of waste disposed of
is likely overrepresented for some states and is why the estimates for the District of Columbia, the Virgin
Islands, and Puerto Rico have been zeroed out.
6.1.1.4 Recalculations
No recalculations were applied for this current report.
6.1.1.5 Uncertainty
The overall uncertainty associated with the 2022 national estimates of CH4 from MSW and industrial
waste landfills was calculated using the Approach 2 methodology (IPCC 2006). As described further in
Chapter 7 of the national Inventory, levels of uncertainty in the national estimates in 2022 were -2%/+20% of
the estimated CH4 emissions for MSW landfills and -31%/+25% for industrial waste landfills.
State-level estimates likely have a higher uncertainty due to (1) apportioning the national emissions
estimates to each state based on assumptions made to disaggregate the national emissions estimates,
which are based on state percentages as reported to the GHGRP, and (2) the application of the scale-up
factor to nationally compiled landfill gas recovery databases used in the national Inventory. Additionally,
state-level estimates before the GHGRP began (i.e., before 2010) may have more uncertainty than state-level
estimates after the GHGRP began (i.e., 2010 and afterward). For more details on national level uncertainty,
see the uncertainty discussion in Section 7.1 of the national Inventory.
6.1.1.6 Planned Improvements
Potential refinements to landfill estimation methods include the following:
MSW landfills. Planned improvements to the state-level estimates are consistent with those
presented in Section 7.1 of the national Inventory. In particular, EPA plans to improve completeness
of emissions from all waste management practices (i.e., open dumpsites) in U.S. territories by
identifying data and applying methods to include emissions from open dumpsites in territories.
Industrial waste landfills. A more complete and comprehensive list of pulp and paper facilities in
the United States will be identified, includingyears of operation since 1990. Further QC on this
inventory will be performed by comparing the counts of industrial waste landfills by state in available
data sets.
6.1.1.7 References
Center for Paper Business and Industry Studies (n.d.) Mills Online 2015-2016. Available online at:
https://epay.gatech.edu/C20793_ustores/web/product_detail.jsp?PRODUCTID=2466.
Conservation Biology Institute (2008) Data Basin: U.S. Pulp and Paper Mills. Available online at:
https://databasin.org/maps/new/#datasets=1f2a22ee1aa441568cbf5bea1b275c88.
EPA (U.S. Environmental Protection Agency) (2020) Advancing Sustainable Materials Management: 2018
Tables and Figures. Available online at: https://www.epa.gov/sites/production/files/2020-
11/documents/2018 tables and figures fnl 508.pdf.
EPA (2021) Excess Food Opportunities Map, Available online at: https://www.epa.gov/sustainable-
management-food/excess-food-opportunities-map.
6-5
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
EPA (2022) Envirofacts Data. Subpart HH: Municipal Solid Waste Landfills.
EPA (2024) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2022. EPA430-R-24-004. Available
online at: https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
IPCC (Intergovernmental Panel on Climate Change) (2006) 2006IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
RTI International (2018) Comparison of Industrial Waste Data Reported Under Subpart TT and the Solid Waste
chapter of the GHG Inventory. Memorandum prepared by K. Bronstein, B. Jackson, and M. McGrath for R.
Schmeltz (EPA).
6.1.2 Composting (NIR Section 7.3)
6.1.2.1 Background
This section presents methods used to estimate state-Level GHGs from large-scale commercial
composting facilities that typically include sections of the waste that operate in an anaerobic environment
where degradable organic carbon in the waste material is broken down, generating CH4 and N20. Even
though C02 emissions are generated, they are not included in net GHG emissions for composting. Consistent
with the national Inventory, emissions from residential (backyard) composting are not included in the scope.
Additionally, the national Inventory assumes windrow is the composting method used, and the waste mixture
is homogeneous, consisting primarily of yard waste and some food. Annual throughput data on static and in-
vessel commercial composting methods were not identified in secondary (published) data. Consistent with
the 2006 IPCC Guidelines, net C02fluxfrom C stock changes in waste material is estimated and reported
under the LULUCF sector (see Chapter 3.4 of this report) (IPCC 2006).
More information on emission pathways and national-level emissions from composting and associated
methods can be found in the Waste chapter (Chapter 7), Section 7.3 of the national Inventory, available
online at https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-inventorv-2024-chapter-7-waste 04-17-
2024.pdf.
6.1.2.2 Methods/Approach
EPA compiles national CH4 and N20 emissions estimates for commercial composting facilities in the
United States using an IPCC Tier 1 method by which an IPCC default emissions factor is applied to the
national quantity of material composted. No facility-specific information is used because it is generally
unavailable over the time series.
The national Inventory was disaggregated to the state level using Approach 2 on the basis of data
available for the proportion of material composted by state for select years. Table 6-3 summarizes published
state-level estimates of composted material used in this inventory. Years where published data are not
available are either interpolated or extrapolated using population growth and published estimates.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
6-6
-------
Section 6 Waste (NIR Chapter 7)
Table 6-3. Summary of Availability and Sources for Composting Data
Year Composting Data Available for
Reference Citation
2000
Goldstein and Madtes 2001
2002
Kaufman et al. 2004
2004
Goldstein et al. 2006
2006
Arsova et al. 2008
2008
Arsova et al. 2010
2010
EREF 2016
2011
Shin 2014
2012
Piatt 2014
2013
EREF 2016
2016
WBJ 2016
2020
WBJ 2020
The state-Level data were Largely compiled from voluntary surveys of state agencies that reported MSW
generated and estimates by relevant management pathways (e.g., landfill, recycling, composting).
Composting estimates may be directly reported by the state agencies or estimated or adjusted by the report
authors using the best available information for available years. Occasionally, data for some states are not
available and are indicated as such in the data sources. The WBJ is an annually updated database of which
the quality is unknown, but it is used because there is a general lack of data. Both the WBJ 2016 and 2020
were used to estimate state data for 2017-2019. Completeness is one limitation with the available state data
used.
The general methodology to estimate the annual quantity of waste composted per year is as follows:
Composteds = %s x Nc
where:
Composteds = the mass of material composted by state (tons/year)
%s = the state percentage of material composted, calculated using available
state data (%)
Nc = the national estimate of material composted as reported in the EPA
Advancing Sustainable Materials Facts and Figures reports (tons/year) (EPA
2020)
The state percentages of material composted were calculated by dividing each state-reported amount
of waste composted by the total of all material composted for that year. The sum of all state-reported data is
referred to as national estimates by the report authors, but to avoid confusion with the Facts and Figures
data published by EPA, this methodology report refers to this as "the sum of state-reported data." .
Limitations with the state-reported survey data include its voluntary nature and the occasional lack of data
for states that did not provide a survey response. The report authors noted they made assumptions to
estimate and adjust data to the extent possible. For years where no state data were reported in a specific
survey, EPA estimated the data using the prior or next year of available data. These gaps were minimal (i.e.,
five or fewer states for each survey year).
Because state data are only available for select years, interpolation and extrapolation were required to
generate estimates for each year of the time series. State proportions applied to 1990-1999 are the same as
those for 2000 (Goldstein and Madtes 2001). No state data exist for this portion of the time series, and there
6-7
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
is a Large amount of uncertainty surrounding the number of facilities and amount of material composted. This
is a conservative approach since it is unknown when a state began compositing operations, so it is assumed
if they had operations in 2000 that they did in 1990 as well. Data in between the survey data were interpolated
using the prior year's and next year's survey data (the state proportion of material composted). Annual state
data were interpolated for 2001, 2003, 2005, 2007, 2009, 2014, 2017, 2018, and 2019. Annual state data for
2021 and 2022 were extrapolated using population growth (U.S. Census Bureau 2022) and estimates of
material composted (WBJ 2020). State percentages for each year are presented in Appendix F (Table F-4).
The formula used for interpolation of the state percentage for the year in question is as follows:
y = ^r=y)xM~'t
where:
y
= state percentage of waste composted
for the year without data, %
y^
= state percentage of waste composted
for the prior year with data, %
Y2
= state percentage of waste composted
for the next year with data, %
X
= the year without data
Xi
= the prior year with data
X2
= the next year with data
The state percentage data were multiplied by the national estimate of material composted from the EPA
Facts and Figures reports to cap the total quantity composted across the states and match the state totals to
the national Inventory. The EPA Facts and Figures national estimates were directly used to estimate the
national Inventory. The IPCCTier 1 method used in the national Inventory estimates (IPCC 2006) is the
product of an emissions factor and the mass of organic waste composted.
The final step in developing the state inventory was estimating the CH4 and N20 emissions. For
simplicity, the state percentages were multiplied by the annual national emissions estimates.
6.1.2.3 Recalculations
No recalculations were applied for this current report.
6.1.2.4 Uncertainty
The overall uncertainty associated with the 2022 national estimates of CH4 and N20 from composting
(specifically large-scale, commercial composting facilities) was calculated using the 2006 IPCC Guidelines
Approach 1 methodology (IPCC 2006). As described further in Chapter 7 of the national Inventory, levels of
uncertainty in the national estimates in 2022 were -58%/+58% for CH4 and for N20. State-level estimates will
have a higher uncertainty than the national estimates because of apportioning the national quantity of
material composted (sourced from the EPA Sustainable Materials Management reports and calculated with a
mass balance methodology) to each state based on sporadically published waste management data from a
voluntary state agency survey for select years. The national methodology also assumes most composting in
the United States uses the windrow method and treats a homogeneous mixture of primarily yard trimmings
and some food waste. For more details on national-level uncertainty, see the uncertainty discussion in
Section 7.3 of the national Inventory, available online at https://www.epa.gov/svstem/files/documents/2024-
04/us-ghg-inventorv-2024-chapter-7-waste 04-17-2024.pdf.
6.1.2.5 Planned Improvements
In future annual publications, EPA plans to investigate state volumes of material composted where the
report authors (from referenced composting data sources) indicated potentially combined volumes of waste
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
6-8
-------
Section 6 Waste (NIR Chapter 7)
sent to composting, recycling, and anaerobic digestion. EPA will continue to identify annual quantities of
material composted in states where data are lacking (e.g., Alaska, Guam). For example, a 2021 desk-based
investigation into composting facilities in Alaska revealed operational aerated composting facilities, but the
annual capacity and throughput were not identified. EPA will continue to search for relevant data for
commercial composting facilities in these states. Planned improvements to the national estimates for
composting outlined in Section 7.3 (page 7-58) of the national lnventoryw\[[ lead directly to improvements in
the quality of state-level estimates as well.
6.1.2.6 References
Arsova, L., R. van Haaren, N. Goldstein, S.M. Kaufman, and N.J. Themelis (2008) The State of Garbage in
America. BioCycle. Available online at: https://www.biocvcle.net/the-state-of-garbage-in-america-3/.
Arsova, L., R. Van Haaren, N. Goldstein, S. Kaufman, and N. Themelis (2010) The State of Garbage in
America. BioCycle, 51(10): 16. Available online at: https://www.biocvcle.net/2010/10/26/the-State-of-
garbage-in-america-4/.
EPA (U.S. Environmental Protection Agency) (2020) Advancing Sustainable Materials Management: 2018
Tables and Figures. Available online at: https://www.epa.gov/sites/production/files/2020-
11/documents/2018 tables and figures fnl 508.pdf.
EREF (Environmental Research & Education Foundation) (2016) Municipal Solid Waste Management in the
United States: 2010 & 2013.
Goldstein, N., and C. Madtes (2001) The State of Garbage in America. BioCycle.
Goldstein, N., S. Kaufman, N. Themelis, and J. Thompson Jr. (2006) The State of Garbage in America.
BioCycle, 47(4): 26. Available online at: https://www.biocvcle.net/2006/04/21/the-state-of-garbage-in-
america-2/.
IPCC (Intergovernmental Panel on Climate Change) (2006) 2006IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.ip/public/2006gl/.
Kaufman, S.M., N. Goldstein, K. Millrath, and N.J. Themelis (2004) The State of Garbage in America. Biocycle,
45(1): 31. Available online at: https://www.biocvcle.net/the-state-of-garbage-in-america/.
Piatt, B. (2014) State of Composting in the US: What, Why, Where & How. Institute for Local Self-Reliance.
Available online at: https://ilsr.org/articles/state-of-composting/.
Shin, D. (2014) Generation and Disposition of Municipal Solid Waste (MSW) in the United StatesA National
Survey. Thesis. Columbia University, Department of Earth and Environmental Engineering, January 3,
2014.
U.S. Census Bureau (2022) Table NST-EST2022-POP. In: Annual Estimates of the Resident Population for the
United States, Regions, States, District of Columbia, and Puerto Rico: April 1, 2020 to July 1, 2022.
Release date: December 2022.WBJ (Waste Business Journal) (2016) Directory of Waste Processing &
Disposal Sites 2016.
WBJ (2020) Directory of Waste Processing & Disposal Sites 2020.
6-9
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
6.1.3 Anaerobic Digestion at Biogas Facilities (Stand-Alone) (NIR Section 7.4)
6.1.3.1 Background
Anaerobic digestion is a series of biological processes in the absence of oxygen in which
microorganisms break down organic matter, producing biogas and soil. Stand-alone digestion is one of three
main categories of anaerobic digestion facilities, which also includes on-farm digesters and digesters at
water resource recovery facilities. This section focuses exclusively on stand-alone digesters, which typically
manage food waste from different sources, including food and beverage processing industries. Emissions
from on-farm digesters and digesters at water resource recovery facilities are reflected under Sections 4.1.2
(Manure Management) and 6.2.1 (Wastewater Treatment and Discharge) of this report. Based on available
data, anaerobic digestion of food waste occurs in 31 states, listed in Appendix F (Table F-5).
At stand-alone digestors, CH4 emissions may result from a fraction of the biogas lost during the process
due to leakages and other unexpected events (0-10% of the amount of CH4 generated; IPCC 2006). The
remaining biogas (90-100% of gas generated) is flared or used beneficially, often combusted to produce heat
and power, or further processed into renewable natural gas or for use as a transportation fuel. C02 emissions
are biogenic in origin and should be reported as an informational item in the energy sector (IPCC 2006).
More information on emission pathways and national-level emissions and methods can be found in
Section 7.4 of the national Inventory.
6.1.3.2 Methods/Approach
EPA compiles national CH4 emissions estimates for stand-alone anaerobic digester facilities in the
United States using an IPCC Tier 1 method, which applies an IPCC default leakage factor of 5% to the CH4
generated. The amount of CH4 generated is the product of an emission factor and the mass of organic waste
processed. The weighted average annual quantity of material processed is estimated from voluntary EPA AD
Survey data (EPA 2018, 2019, 2021, 2023) and an estimated number of operating facilities per year (see Table
7-46 and Table 7-47, respectively, of the national Inventory). No facility-specific quantities of material
digested were directly used because of a general lack of facility-specific data over the time series. The
methodology applied to generate the national Inventory was based on two large assumptionsthe number of
operational facilities and the weighted average of material digested for two of the 30 years in the time series
(1990-2022). The state inventory further takes these assumptions to a state level by assuming that the same
percentage of total operational facilities is the same for each year of the time series because of a general
lack of data on total operational facilities by state across the time series. Therefore, the state-level
inventories are a gross estimate that may be refined in future years if available information by state is
obtained.
In the national Inventory, EPA calculated a weighted average of material digested using masked survey
data from available survey reports for 2015 to 2019 (EPA 2018, 2019, 2021, 2023). The weighted average was
applied to an estimated number of operational facilities per year to estimate the annual quantity of material
digested. The first step to calculating the state inventory was to disaggregate the annual estimates of the
material digested. This was disaggregated by applying a state percentage of operational facilities as reported
to the two published EPA survey reports (EPA 2018, 2019). The state proportions of operational facilities in
2015 and 2016 are presented in Appendix F (Table F-5).
The state proportions were multiplied by the national quantity digested for each year in the time series,
which forced the total quantities across the states to match the national Inventory estimates. The same state
percentage was used for each year in the time series because of a lack of compiled data on the number of
stand-alone digesters by state between 1990 and 2022.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
6-10
-------
Section 6 Waste (NIR Chapter 7)
6.1.3.3 Recalculations
For the current Inventory, a methodological change was made whereby the CH4 emissions are
considered equal to leakage from the digester network of pipes. A leakage factor of 5% (as recommended in
IPCC 2006) is applied to the CH4 generation estimate for all years in the time series. This methodological
change applies to every year in the time series and significantly reduces annual CH4 emissions estimates.
Previously, the EPA AD Survey data reporting the amount of biogas produced at AD facilities was used for the
amount of gas recovered, with the remaining gas assumed to be leaked or emitted. This method calculated
higher emissions estimates, which showed the majority of the gas generated at an AD being emitted instead
of being used in biogas projects. This was inconsistent with the EPA AD Survey findings that approximately
95% of stand-alone AD facilities use some or all biogas on-site, and it was also inconsistent with the IPCC
guidance on default leakage from AD facilities. EPA will further investigate the survey data for the total
biogas-produced, since they indicate very low gas utilized as compared to this revised methodology. See the
recalculations discussion (page 7-63) in Section 7.4 of the national Inventory.
6.1.3.4 Uncertainty
The overall uncertainty associated with the 2022 national estimates of CH4 from stand-alone anaerobic
digesters was calculated using the 2006 IPCC Guidelines Approach 1 methodology (IPCC 2006). As
described further in Chapter 7 of the national Inventory, levels of uncertainty in the national estimates in
2019 were -54%/+54% CH4. State-level estimates will have a higher uncertainty because of apportioning the
national emissions estimates to each state based solely on the number of stand-alone anerobic digester
facilities from EPA survey data collected between 2015 and 2018. Emissions estimates before 2015 will have
a higher uncertainty than those in 2015 and later years. These assumptions were required because of limited
facility-specific data presented in secondary resources. For more details on national level uncertainty, see
the uncertainty discussion in Section 7.4 of the national Inventory.
6.1.3.5 Planned Improvements
The planned improvements are consistent with those planned for improving national estimates given
that the underlying methods for state GHG estimates are the same as those in the national Inventory. To find
information on planned improvements to refine methods for estimating emissions from stand-alone
anaerobic digestion, see the planned improvements discussion starting on page 7-64 of Section 7.4 in the
national Inventory.
6.1.3.6 References
EPA (U.S. Environmental Protection Agency) (2018) Anaerobic Digestion Facilities Processing Food Waste in
the United States in 2015: Survey Results. EPA/903/S-18/001. Available online at:
https://www.epa.gov/sites/default/files/2018-
08/documents/ad data report final 508 compliant no password.pdf.
EPA (2019) Anaerobic Digestion Facilities Processing Food Waste in the United States in 2016: Survey
Results. EPA/903/S-19/001. Available online at: https://www.epa.gov/sites/default/files/2019-
09/documents/ad data report vlO - 508 comp vl.pdf.
EPA (2021) Anaerobic Digestion Facilities Processing Food Waste in the United States (2017 & 2018): Survey
Results. EPA/903/S-21/001. Available online at: https://www.epa.gov/sites/default/files/2021-
02/documents/2021 final ad report feb 2 with links.pdf.
EPA (2023) Anaerobic Digestion Facilities Processing Food Waste in the United States (2019): Survey Results.
EPA 530-R-23-003. Available online at: https://www.epa.gov/svstem/files/documents/2023-
04/Anaerobic Digestion Facilities Processing Food Waste in the United States 2019 20230404 508.pdf.
6-11
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
IPCC (Intergovernmental Panel on Climate Change) (2006) 2006IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.jp/public/2006gl/.
6.2 Wastewater Management
This section presents the methodology used to estimate the emissions from domestic and industrial
wastewater treatment and discharge (CH4, N20).
6.2.1 Wastewater Treatment and Discharge (NIR Section 7.2)
6.2.1.1 Background
Consistent with the national Inventory and international guidance, EPA has developed disaggregated
state estimates for both domestic and industrial wastewater treatment and discharge, as discussed below:
Domestic wastewater CH4 and N20 emissions originate from both septic systems and centralized
treatment plants. Within these centralized plants, CH4 emissions can arise from aerobic systems
that liberate dissolved CH4 that formed within the collection system or that are (1) designed to have
periods of anaerobic activity, (2) from anaerobic systems, and (3) from anaerobic sludge digesters
when the captured biogas is not completely combusted. N20 emissions can result from aerobic
systems as a byproduct of nitrification, or as an intermediate product of denitrification. Methane
emissions will also result from the discharge of treated effluent from centralized treatment plants to
water bodies where carbon accumulates in sediments, while N20 emissions will also result from
discharge of centrally treated wastewater to water bodies with nutrient-impacted or eutrophic
conditions.
Industrial wastewater CH4 emissions originate from in-plant treatment systems, typically comprising
biological treatment operations in which some operations are designed to have anaerobic activity or
may periodically form anaerobic conditions. N20 emissions are primarily expected to occur from
aerobic treatment systems as a byproduct of nitrification, or as an intermediate product of
denitrification. Emissions will also result from discharge of treated effluent to waterbodies.
6.2.1.2 Methods/Approach (Domestic Wastewater)
EPA estimated state-level domestic wastewater treatment and discharge emissions (CH4) using a
simplified approach to apportion the national emission estimates to each state based on population (i.e.,
Approach 2 as defined in the Introduction to this report) and state-level septic data. In this method, EPA
accessed historical U.S. Census data to compile state-level population data for each year of the inventory
(1990-1999: U.S. Census Bureau 2002; 2000-2009: U.S. Census Bureau 2011; 2010-2021: U.S. Census
Bureau 2021a, 2021b, 2022, 2023; Instituto de Estadfsticas de Puerto Rico 2021). The U.S. Census Bureau
(1990) and NEBRA (2022) reported the percentage of the population associated with septic systems in each
state for 1990 and 2018, respectively. These percentages were multiplied by the 1990 and 2018 state-level
population and then divided by the total summed national population to estimate the percentage of the
national population with a septic system in each state and territory in 1990 and 2018. The state-level
percentages for 1991-2017 were linearly interpolated between 1990 and 2018, and the remainder of the time
series was set equal to 2018, as shown in Appendix F, Table F-6.
EPA calculated state- and territory-level emissions by multiplying the proportion of the U.S. population
on centralized treatment or septic systems in each state or territory by the national CH4 and N20 emissions
for each year of the time series.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
6-12
-------
Section 6 Waste (NIR Chapter 7)
This simplified approach assumes the following:
Every state has the same wastewater treatment system usage as the national Inventory.
Every state has same distribution of discharge to various waterbody types as the national Inventory.
Kitchen disposal usage is the same in every state, and wastewater biochemical oxygen demand
(BOD) produced per capita, with and without kitchen scraps, is the same in every state (i.e.,
assumes total wastewater BOD produced per capita is the same as national production).
Per capita protein consumption in the United States is the same in every state (i.e., assumes per
capita consumption is the same as national consumption).
EPA did not perform a more detailed approach that would account for the specific types of treatment at
centralized systems, such as anaerobic reactors or activated sludge, used in each state (see planned
improvements below in Section 6.2.1.6). Similarly, there are insufficient readily available data sources to
allow classification of the type of specific water bodies within each state, so EPA did not consider the type of
water body receiving wastewater discharges within each state.
6.2.1.3 Methods/Approach (Industrial Wastewater)
Consistent with the national Inventory and national estimates, both CH4 and N20 emissions were
estimated for treating industrial wastewater from pulp and paper manufacturing, meat and poultry
processing, petroleum refining, and breweries, while CH4 emissions were also estimated for treating
industrial wastewater from vegetables, fruits, and juices processing, and for starch-based ethanol
production. These are the industry categories that are likely to produce significant GHG emissions from
wastewater treatment. Data on industrial production by state are available or can be estimated from other
readily available data for at least some of the time series of the inventory.
EPA estimated state-level emissions by estimating the percentage of the industry production that
occurs in each state (i.e., using Approach 2 as described in the Introduction to this report). Where data were
readily available, EPA estimated the distribution of production for each year of the time series and multiplied
that by the national emissions estimate for each year of the time series. In some cases, due to time and
resources, EPA was able to estimate the distribution of production for a subset of years in the time series, as
discussed below by industry.
For pulp and paper manufacturing, state-level production data are not available, so EPA estimated
state- level emissions by estimating the percentage of wastewater directly discharged in that state compared
to the total flow of wastewater directly discharged for that industry, using data reported to EPA's ICIS
National Pollutant Discharge Elimination System (NPDES) database. EPA acknowledges that this
methodology ignores production at mills that either do not discharge wastewater or that discharge to a publicly
owned treatment works. In both cases, these mills could be performing on-site treatment and emitting GHGs that
cannot be captured.
EPA then multiplied that percentage by the national emissions estimate to obtain a state-level
emissions estimate. Because of the limitation of data resources for this effort, EPA accepted most ICIS-
NPDES data as is, but some outliers were determined and handled as described below (see planned
improvements below in Section 6.2.1.6).
Both approaches assume the following:
All facilities in an industry within a state have the same distribution of wastewater treatment
operations as the national distribution.
6-13
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Every state has the same BOD and total nitrogen in untreated industry wastewater as the national-
Level estimates.
Every state has the same nitrogen removal factor as the national-level estimates.
The percentage of wastewater directly discharged by the state represents the distribution of all pulp
and paper production by the state.
Further details on methods and data sources assumptions for each industry treating wastewater are
described below.
6.2.1.3.1. Pulp and Paper Manufacturing
Industrial production data for pulp and paper are highly confidential and are not available by state.
EPA used the amount of wastewater directly discharged by pulp mills by statereported to both
ICIS- NPDES from Enforcement Compliance History Online (EPA 2024b) and the Washington
Department of Ecology's Permitting and Reporting Information System, or PARIS (Washington
Department of Ecology 2022, 2024)to proportion U.S. national emissions estimates to a state (as
shown in Appendix F, Table F-7). Because wastewater flow data housed in ECHO changed in 2016,
using older data may cause discontinuities in the time series. EPA determined the distribution of
discharge flow by state for 2019-2022 using ECHO and PARIS data and applied the 2019 distribution
to all prior years of the national Inventory. There was no wastewater flow reported for the District of
Columbia or U.S. territories for this industry.
o Pulp and paper mills were determined in ECHO using Standard Industrial Classification codes
2611, 2621, and 2631.
For facilities in states other than Washington, EPA:
o Downloaded the total pulp and paper permit universe in ECHO, including permits that have
discharge monitoring report data (252 facilities in 2022), and permits with information only (e.g.,
facility address) (322 facilities in 2022).
¦ Stormwater, construction, or non-mill-related permits that were reported for a facility that
also reported using another permit (such as a major or general permit) were removed from
the analysis to prevent overestimating flow (see Table 1). If a facility only reported using
stormwater permits, a single stormwater permit was retained so that it would be counted in
the universe and flow could be estimated for the facility.
¦ In four cases, it was discovered that a permit for an operational facility was missing from the
data download for certain years, so those permit numbers were manually added to the
facility universe for the missingyears and the flow data for those facilities were estimated
following the methodology (see Table 1).
o Downloaded 2019-2022 flow data where available (EPA 2024b). Not all facilities report total flow
if it is not required by their permit. Total flow was summed by state.
¦ EPA determined four state flow outliers, one for Missouri in 2020, West Virginia in 2021 and
2022, and one for Minnesota in 2022. Outliers, determined as values that are at least an
order of magnitude larger (10 times) than other years'values for the state, were removed. It
is assumed these values are data entry errors in ECHO. An average of the other available
values was used as a surrogate for removed values.
o For permits without flow data, total flow was estimated by using average flow by state, or
average national total flow for that year if no state data were available, multiplied by the number
of permits without flow data for that state.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
6-14
-------
Section 6 Waste (NIR Chapter 7)
Facilities Located in the state of Washington are not currently reported within ECHO due to Lagging
electronic reporting. To fill this known gap, EPA investigated a separate source for these data and:
o Downloaded and reviewed permit data for known pulp mills determined from the Washington
Department of Ecology's Industrial Facility Permits website.
o Downloaded 2019 flow data where available (Washington Department of Ecology 2022) for
monitoring locations that are associated with process wastewater, per the facility permit. During
the 1990-2022 Inventory update cycle, a new Washington permit that had since become active
(WA0002925) was added to the 2022 estimate (Washington Department of Ecology 2024).
o Multiplied the daily flow rate by 365.25 days to estimate a total yearly flow, then multiplied by
number of months data were reported (to prevent overestimating annual flow, which was done
to better match the methodology in ECHO),
o Integrated into the other state data for all years.
EPA calculated the percentage of national flow by state:
o As with Washington, some states are missing from ECHO (e.g., Montana, Colorado). EPA
assumed some of these states have nonzero emissions, but they do not have the data to
determine whether there are facilities present or to estimate emissions, so they are reported as
not applicable.
EPA calculated the state-level emissions by multiplying national emissions by the percentage of
national flow by state.
Example: 2022 Georgia emissions
o Georgia has 18 facilities in the facility universe, of which 14 have reported annual flow data.
o The total flow based on the sum of reported flows (14 facilities) and calculated flows (4 facilities)
from the state average flow of 8,042 million gallons (MMGal) for all facilities was 144,760 MMGal
in 2022.
o Georgia's flow was 8.59% of the total national total flow of (1,684,263 MMGal).
o Pulp and paper's national CH4 emissions in 2022 was 30 Gg CH4, so Georgia's 2022 emissions
were estimated to be (30 Gg CH4 X 8.59% = 2.6 Gg CH4).
6-15
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
Table 6-4. Pulp and Paper Permits Manually Removed From or Added to Analysis
NPDES Stormwater/Construction or Non-Mill
Permit Numbers Removed
ALG060506, ALG060521, ALG141038, ARG160040,
ARR001954, ARR00A499, ARR00A634, ARR00A771,
ARR00A776, ARR156061, ARR157067, ARR157281,
ARR157554, FLG071465, FLR05A517, FLR05B628,
FLR05F649, FLR05G876, FLR05H761, FLR20CA56,
FLR20CJ24, GAIS00705, GAIS00742, GAIS01355,
GAIS01421, GAIS01678, GAIS02742, GAIS02751,
GAIS02887, GAIS03756, GAIS04381, GAIS12083,
GAIS13069, GAIS13100, GAIS13110, GAIS13795,
GAIS14256, IDR053113, INRM01785, LAR05M630,
LAR05P618, LAR100640, LAR10O712, MANOE3652,
MAR053165, MAR053218, MDR003165, MDR003388,
MER05B433, MER05B451, MER05B451, MER05B608,
MER05B983, MER05B984, MER05C163, MER05C172,
MER05C178, MER05C178, MER05C200, MER05C269,
MER05C269, MERNEB567, MERNEB567, MI0001210,
MIS111133, MIS210982, MSR000044, MSR000382,
MSR001256, MSR002038, MSR107486, MSR110045,
MSR110077, MSR110077, MSR110099, MSR110099,
MSR110118, MSR110118, MSR321403, MSR321403,
NCS000101, NCS000105, NCS000106, NCS000106,
NCS000211, NCS000211, NHG360002, NHNOEJ036,
NHR053059, NHR053105, NJG224901, NMNOE3331,
NYR00A955, NYR00B038, NYR00B199, NYR00B504,
NYR00C573, NYR00E290, NYR00F582, NYR00F629,
NYR00F629, NYR00G104, NYR00G109, ORR109393,
ORR10A484, ORR10A546, ORR10E944, ORR10F683,
ORR10F683, ORR10F806, ORR10F806, ORR220121,
ORR221181, ORR240152, ORR240152, ORR240316,
ORR241051, ORR241051, ORR241118, ORR241300,
ORR241300, ORR241524, ORR241524, SCR004687,
SCR005224, SCR006095, SCR006186, SCR006186,
TNR050417, TNR050850, TNR050850, TNR051032,
TNR051032, TNR052093, TNR053851, TNR054024,
TNR054024, TNR055894, TNR056437, TNR058104,
TXR05DP51, TXR1505FH, TXR15193U, TXR1547FX,
TXR15691U, TXR1569FF, VAN030049, VAN030133,
VAN040066, VAN040070, VAN040073, VAR052485
NPDES Permit Numbers Added (Consistent
with previous Inventories)
GA0001988, KYR003292, MN0001431,
MN0001643
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
6-16
-------
Section 6 Waste (NIR Chapter 7)
6.2.1.3.2. Meat and Poultry Processing
Annual U.S. and state-Level production data for red meat processing and poultry processing data are
available from USDA-NASS (as shown in Appendix F, Table F-8). Depending on the commodity,
limited state-level data are available. Typically, USDA reports only break out the primary states
where the commodity is processed and then present production in "other states."
For red meat processing:
o EPA gathered state-level 2022 and 1990 average live weight and total head slaughtered for the
following commodities: beef, calves, hogs, and lamb/mutton (USDA 2023a, 1991a). EPA
retained 2021 and 2012 data from the 1990-2021 state-level production data, 2019 data from
the 1990-2019 state-level production data, and 2020 and 2004 data from the 1990-2020 state-
level production data.
¦ U.S. territories and the District of Columbia are not included in USDA-reported data,
o For total head slaughtered (thousand head):
¦ To populate states for which specific production data are not disclosed by USDA ("D"
states), EPA evenly divided the difference between the sum of the state-level data and the
reported national-level total to those D states.
¦ Similarly, USDA provided a total for New England states that was evenly distributed to those
states noted (Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, and
Vermont).
¦ USDA provided a combined total for Delaware and Maryland, which was evenly distributed
between the two states.
o For average live weight (pounds):
¦ EPA used the average of available state-level data and the national average to determine the
appropriate average live weight for the remaining states (D states). This calculated value
was applied to all D states.
¦ Similarly, the reported average live weight value for New England states was applied to
those states.
¦ USDA provided a combined total for Delaware and Maryland, which was evenly distributed
between the two states.
o As with the national Inventory, EPA determined live weight killed (LWK) by multiplying the average
live weight by the total head/1,000 to get to million pounds LWK.
o EPA added the disaggregated red meat processing data by state and divided the data by the
reported national production to determine the proportion distributed to states. Because of the
estimated nature of the calculated values, the total state-level LWK is estimated at about 95% of
the national total, so the percentages were normalized to 100%.
For poultry processing:
o EPA gathered state-level 2022 and 1990 poultry live weight data. EPA retained 2021 and 2012
data from the 1990-2021 state-level production data, 2019 data from the 1990-2019 state-level
production data, and 2020 and 2004 data from the 1990-2020 state-level production data. Only
young chickens, or broilers, had state-level data available. Turkeys and mature chickens did not.
¦ Young turkey data were available by state. EPA assumed that states with young turkeys
would be representative of turkey processing production; therefore, young turkey data were
used as a proxy for total turkeys (USDA 2023b, 1991 b).
6-17
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
¦ Young chickens were used to represent mature chicken processing production by state
(USDA 2023b, 1991b).
o To populate D states for 2022, EPA evenly divided the difference between the sum of the state-
level data and the reported national-level total to those D states.
o To populate D states for 1990, EPA first proxied the reported D states for 2012 because the
individual states for 1990 were not available or reported by USDA. This was done to encourage
time series consistency and avoid showing states known to have poultry processing as having no
emissions for the industry. EPA acknowledges this method could attribute minor emissions to
states without poultry in 1990. Then, as with 2022, EPA evenly divided the difference between
the sum of the state-level data and the reported national-level total to those D states.
o For turkeys and mature chickens, the proportion of young turkeys and young chickens,
respectively, was multiplied by the national-level value to determine the pounds of processing
production per state.
o Those values were added together and then divided by the total poultry (young chickens, mature
chickens, turkeys) values to determine the proportion of poultry LWK for states.
To calculate CH4emissions, EPA:
o Multiplied national red meat plant CH4 emissions by the percentage of U.S. total meat
processing and added that to the national poultry plant CH4 emissions multiplied by the
percentage of U.S. total poultry processing by state.
o Multiplied the 2004 (from the 1990-2020 inventory), 2012, 2019 (from the 1990-2019 inventory),
2020 (from the 1990-2020 inventory), 2021 (from the 1990-2021 inventory), and 2022 state-level
proportion of U.S. meat and poultry BOD treated on-site by the national effluent CH4 emissions
from meat and poultry.
o For 1991-2003, used linear interpolation of 1990 and 2004 state-level proportions, for 2005-
2011, used linear interpolation of 2004 and 2012 state-level proportions, and for and 2013-2018,
used the 2012 and 2019 proportions. Multiplied those values by the national effluent CH4
emissions from meat and poultry.
o Added plant and effluent emissions for total state-level emissions.
To calculate N20 emissions, EPA:
o Multiplied the 2004 (from the 1990-2020 inventory), 2012, 2019 (from the 1990-2019 inventory),
2020 (from the 1990-2020 inventory), 2021 (from the 1990-2021 inventory), and 2022 state-level
proportion of U.S. total nitrogen in both
¦ 1) aerobically treated meat and poultry wastewater by the N20 emissions from meat and
poultry processing wastewater treatment for each year in the time series and
¦ 2) discharged meat and poultry wastewater by the N20 emissions from meat and poultry
processing wastewater treatment effluent for each year in the time series.
o For 1991-2003, used linear interpolation of 1990 and 2003, for 2005-2011 and 2013-2018, EPA
used linear interpolation of 2004 and 2012, and 2012 and 2019 state-level proportions,
respectively. Multiplied those values by the national effluent N20 emissions from meat and
poultry.
o Added plant and effluent emissions for total state-level emissions.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
6-18
-------
Section 6 Waste (NIR Chapter 7)
6.2.1.3.3. Vegetables, Fruits, and Juices Processing
Annual U.S. production data for vegetables, fruits, and juices processing are available from USDA.
Depending on the commodity, state-level data are available (as shown in Appendix F, Table F-9).
Typically, USDA reports only identify the primary states where the commodity is processed. For
example, production data on broccoli are provided for California and "other states," while
production data on asparagus are provided for Michigan, Washington, and "other states."
o U.S. territories and the District of Columbia are not included in the USDA-reported data.
EPA determined that the most recent year with complete state-level production values is 2017
because USDA suspended the reporting of some state-level production values in 2018 and more
notably in 2019-2022.
Previously, to better inform the time series, EPA investigated an earlier year, determined 2012 to be
complete, and subsequently determined the state-level production values for 2012. EPA previously
investigated and included 2004 during the 1990-2020 Inventory.
For processing production data:
o State-level data for potato processingwere not available. Instead, EPA used state-level potato
production (i.e., the production of potatoes grown not processed) as a proxy to determine the
states to include (USDA 2014).
o For other vegetables, EPA gathered data for asparagus, broccoli, carrots, cauliflower, sweet
corn, cucumber (for pickles), lima beans, green peas, snap beans, spinach, and tomatoes
(USDA 2015a). Where USDA reported data for "other states," those data were distributed
equally among the commodities. EPA added the production for these commodities to determine
the percentage of the U.S. total for all "other vegetables," which is the production value used in
the national Inventory (not the individual commodities).
o Processed apples, grapes used for wine, and citrus fruits were also determined at a state level.
For apples, where USDA reported data for "other states," those data were distributed equally
(USDA2015b, 2015c).
o Noncitrus fruits are split out into separate commodities (e.g., blueberries, sweet cherries38); no
state-level data are available for the aggregated "noncitrus fruit" category. Therefore, EPA
gathered the state-level "utilized production" data for these separate commodities to determine
the appropriate states and relative percentage of utilized production for noncitrus fruits (USDA
2015c).
o Processed noncitrus fruit data are typically calculated in the national Inventory as utilized
production minus fresh minus apples minus grapes for wine; however, because of the intensive
nature of gathering data for the separate commodities, "utilized production" was used as a
proxy for processed production data.
To calculate emissions, EPA calculated the 2004, 2012, and 2017 percentage of U.S. total BOD by
state and multiplied that by the national vegetables and fruits emissions for each year in the time
series.
38 EPA gathered 2004 and 2017 production for apricots; avocados (2012 values reported as "not available"); blueberries,
cultivated blueberries (2004 only), and wild blueberries; boysenberries (2004 only); sweet and tart cherries; coffee (2017
only); cranberries; dates; loganberries (2004 only); nectarines; olives; papaya (2012 Hawaii crop reported as "not available"),
including guavas and pineapples (Hawaii crops, 2004 only); peaches; pears; plums; prunes (combined with plums in 2004);
raspberries; and strawberries.
6-19
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
For 2005-2011 and 2013-2016, EPA determined state-Level proportions by Linear interpoLation of
2004 and 2012, and 2012 and 2017 vaLues, respectiveLy. Proportions for 2018-2022 were assumed to
be the same as 2017.
6.2.1.3.4. Petroleum Refining
AnnuaL production data are avaiLabLe from EIA within the Department of Energy (EIA 2024a), as
shown in Appendix F (TabLe F-10).
Because state-LeveL data may reveaL confidentiaL data, production data are aggregated by PetroLeum
Administration for Defense Districts (PADDs). Production data for the foLLowing PADDs and
subdistricts are avaiLabLe:
o PADD I (East Coast)
¦ Subdistrict A (New EngLand): Connecticut, Maine, Massachusetts, New Hampshire, Rhode
IsLand, and Vermont
¦ Subdistrict B (CentraL AtLantic): DeLaware, District of CoLumbia, MaryLand, New Jersey, New
York, and PennsyLvania
¦ Subdistrict C (Lower AtLantic): FLorida, Georgia, North CaroLina, South CaroLina, Virginia, and
West Virginia
o PADD II (Midwest): ILLinois, Indiana, Iowa, Kansas, Kentucky, Michigan, Minnesota, Missouri,
Nebraska, North Dakota, South Dakota, Ohio, OkLahoma, Tennessee, and Wisconsin
o PADD III (GuLf Coast): ALabama, Arkansas, Louisiana, Mississippi, New Mexico, and Texas
o PADD IV (Rocky Mountain): CoLorado, Idaho, Montana, Utah, and Wyoming
o PADD V (West Coast): ALaska, Arizona, CaLifornia, Hawaii, Nevada, Oregon, and Washington
Operating capacity by state is avaiLabLe from EIA (2024b) for 1990-2022.
EPA created state-LeveL annuaL production data for each year of the time series (1990-2022) by
dividing the annuaL production for each PADD subdistrict by the percentage of operating capacity
each state provided in that year.
PetroLeum operating capacity vaLues were not avaiLabLe for 1996 and 1998. These vaLues were
LinearLy interpoLated.
ExampLe: 2022 CaLifornia emissions
o CaLifornia data are incLuded in PADD V.
o PADD V has a totaL of 27 refineries with an operating capacity of 2,659,271 barreLs.
o CaLifornia has a totaL of 15 refineries with an operating capacity of 1,749,871 barreLs (or 65.8% of
PADD V capacity),
o PADD V produced 1,000,921 barreLs in 2022.
o Estimate CaLifornia production as 1,000,921 barreLs * 65.8% = 658,633 barreLs.
o CaLcuLate CaLifornia's percentage of nationaL production (658,633 barreLs/7,079,773 barreLs =
9%).
o CaLcuLate CaLifornia emissions as nationaL emissions * percentage of nationaL production (4.4
Gg CH4 x 9% = 0.4 Gg CH4).
6.2.1.3.5. Starch-based Ethanol Production
State-LeveL ethanoL production data are avaiLabLe from ElA's State Energy Data System (SEDS) (EIA
2023) (as shown in Appendix F, TabLe F-11).
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
6-20
-------
Section 6 Waste (NIR Chapter 7)
o Fuel ethanol production data, including denaturant, in thousand barrels are available for 1960-
2021 (EIA2023).
o EPA checked the difference between SEDS national production and the reported production in
the national Inventory and found small differenceson average, a 0.9% difference for the time
seriesfurther confirming SEDS is a good source of state-level production.
o Typically, the most recent year of data is used as a surrogate for the last year of available
production data. For example, during the 1990-2022 Inventory by State, 2021 production values
were used for 2022. This is due to the timing of when production data are released versus to
publication of the Inventory by State.
Calculated the percentage of national production by state for every year, using the production data
noted above.
Calculated the state-level emissions by multiplying national emissions by percentage production by
state.
Example: 2021 California emissions
o 2021 California production value is 2,293 thousand barrels,
o National production for 2021 is 375,517 thousand barrels,
o California produced 1.2% of the national production in 2021.
o Calculate 2021 California emissions as national emissions * percentage of national production
(5.9 Gg CH4 x 0.6% = 0.04 Gg CH4).
6.2.1.3.6. Breweries
Annual production data by state are available from the Alcohol and Tobacco Tax and Trade Bureau
(TTB 2024) (as shown in Appendix F, Table F-12).
o Quarterly state-level production data are available for 2015-2022. Annual, state-level taxable
production values are available for 2008-2020. The quarterly state-level production values are
preferred and were used for 2015-2022. Data for earlier years of the time series are still not
available, so the calculated percentage of national production for 2008 was used for 1990-2007.
o In cases where one or two quarters of data were not disclosed due to confidentiality reasons,
EPA averaged the remaining quarters and added that average to the annual sum to estimate the
total annual production for that state,
o In cases where no quarterly state-level data were available for 2015-2020 (Delaware, Florida,
Missouri, and New Hampshire), EPA used the annual taxable removals data to proportion the
unaccounted-for production data (the difference between the U.S. total production and the sum
of the state-level quarterly production data). For 2021 and 2022, EPA forecasted the available
2015-2020 data to estimate production data for those years, due to annual taxable removals
data not being available,
o Data are not available broken out between craft and noncraft production, so the approach
assumes each state has the same distribution of craft and noncraft production as the national
distribution.
Calculated the percentage of national production by state.
Calculated the state-level emissions by multiplying national emissions by percentage production by
state.
Example: 2019 California emissions
o California production is 20,948,150 barrels.
6-21
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
o National production is 180,866,990 barrels,
o California produces 11.2% of national production.
o Calculate California emissions as national emissions * percentage of national production (5.2
GgCH4x 11.2% = 0.581 Gg CH4).
6.2.1.4 Recalculations
Recalculations discussed here are specific to state-level production or disaggregated data. To see
impacts from updates to national-level data, see the recalculations discussion in Section 7.2 of the Waste
chapter (Chapter 7) in the national Inventory, available online at
https://vwwv.epa.gov/system/files/documents/2024-04/us-ghg-inventory-2024-chapter-7-waste 04-17-
2024.pdf.
EPA updated the domestic methodology to include state-level proportions of septic versus centralized
treatment based on available data (U.S. Census Bureau 1990. These updates, in conjunction with the
changes to the national Inventory,39 resulted in changes for 1990-2017 for all state-level domestic CH4 and
N20 emission estimates.
Updates to the following state-level industrial production data, in conjunction with national-level
updates, resulted in changes for the entire time series for every state-level total industrial CH4 and N20
emission estimates:
Pulp and paper. Including 2019, 2020, and 2021 flow estimates for all available state data due to an
updated methodology to determine/download flow data from ECHO, affecting all years.
Meat and poultry processing. Including 1990 production data, affecting 1990-2003.
Breweries. Updating to use state-level production data (rather than taxable removals), affecting
2015-2021.
6.2.1.5 Uncertainty
The overall uncertainty associated with the 2022 national estimates of CH4 and N20 from wastewater
treatment and discharge were calculated using the 2006 IPCC Guidelines Approach 2 methodology (IPCC
2006). As described further in Chapter 7 of the national Inventory (EPA 2024a), levels of uncertainty in the
national estimates in 2022 were -29%/+33% for CH4 and -36%/+192% for N20. State-level estimates have a
higher uncertainty due to apportioning the national emissions estimates to each state based solely on state
population (for domestic) or state industry sector production (for industrial). This approach does not address
state-level differences in the type of wastewater treatment systems in use or in the conditions of the state's
receiving waterbodies. State-level emissions for the time series were estimated based on limited years of
state-level data, which also results in higher uncertainty for the state estimates. These assumptions were
required due to the general lack of readily available state- or regional-level data. For more details on national-
level uncertainty, see the uncertainty discussion in Section 7.2 of the Waste chapter (Chapter 7) in the
national Inventory, available online at https://www.epa.gov/svstem/files/documents/2024-04/us-ghg-
inventory-2024-chapter-7-waste 04-17-2024.
6.2.1.6 Planned Improvements
Generally, EPA plans to review feedback from reviews of the state-level inventory methods and assess
potential to use data sets identified in comments to see if they provide comparable data for all states or
39 See Section 7.2, page 7-52, of the national Inventory.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
6-22
-------
Section 6 Waste (NIR Chapter 7)
most states. The steps outlined below may inform the potential improvements for both domestic and
industrial state-level emissions estimates. EPA plans to undertake the following assessments as resources
allow:
Determine state-level sources for the type of wastewater treatment systems in use for municipal or
domestic or for industrial wastewater (by industrial sector).
Determine state-level sources for BOD or total nitrogen data in municipal or domestic wastewater or
industrial wastewater (by industrial sector).
As stated in Section 7.2 of the national Inventory, investigate additional sources for estimating
wastewater volume discharged and discharge location for both domestic and industrial sources.
For individual industries, EPA notes the following potential improvements.
6.2.1.6.1. Pulp and Paper Manufacturing
Investigate state-level sources for the production of pulp, paper, and paperboard.
Investigate additional years of ECHO data to improve the time series. Part of this includes evaluating
the facilities present year to year to confirm time series consistency.
Investigate states where data are reported as not applicable and confirm emissions estimates do not
apply. Pending findings, determine another source to estimate wastewater flow for these states.
Refine criteria for evaluating stormwater permits and identifying duplicate permits that should be
removed from analysis.
6.2.1.6.2. Meat and Poultry Processing
Continue to investigate additional years of available USDA data for inclusion to improve the time
series.
Investigate the presence of meat and poultry processing in the U.S. territories or the District of
Columbia and, pending findings, additional sources for estimating those emissions. For the District
of Columbia, reach out to USDA-NASS to confirm if the District of Columbia is already included in
reporting.
6.2.1.6.3. Vegetables, Fruits, and Juices Processing
Continue to investigate other years of available USDA data for inclusion. EPA investigated 1990 and
2008 during the 1990-2022 Inventory and determined that those data sets are inconsistent with the
current data sets.
Investigate the presence of vegetables, fruits, and juices processing in the U.S. territories or the
District of Columbia and, pending findings, additional sources for estimating those emissions. For
the District of Columbia, reach out to USDA-NASS to confirm if the District of Columbia is already
included in reporting.
6.2.1.6.4. Starch-Based Ethanol Production
Investigate sources to break down wet and dry milling by state over the time series.
6.2.1.6.5. Breweries
Investigate sources to break down craft and noncraft breweries by state over the time series.
6-23
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
6.2.1.7 References
EIA (U.S. Energy Information Administration) (2023) State Energy Data System (SEDS): 1960-2021
(Complete). U.S. Department of Energy. Available online at: https://www.eia.gov/state/seds/seds-data-
complete.php.
EIA (2024a) Refinery and Blender Net Production. U.S. Department of Energy. Available online at:
https://www.eia.gov/dnav/pet/pet pnp refp a epOO ypr mbbl a.htm.
EIA (2024b) Number and Capacity of Petroleum Refineries. U.S. Department of Energy. Available online at:
https://www.eia.gov/state/seds/seds-data-complete.php.
EPA (U.S. Environmental Protection Agency) (2024a) Inventory of U.S. Greenhouse Gas Emissions and Sinks:
1990-2022. EPA430-R-24-004. Available online at: https://www.epa.gov/ghgemissions/inventorv-us-
greenhouse-gas-emissions-and-sinks.
EPA (2024b) Water Pollution Search. Available online at: https://echo.epa.gov/trends/loading-tool/water-
pollution-search.
Instituto de Estadfsticas de Puerto Rico (2021) EstimadosAnuales Poblacionales de los Municipios Desde
1950. Accessed February 2021. Available online at:
https://censo.estadisticas.pr/EstimadosPoblacionales.
IPCC (Intergovernmental Panel on Climate Change) (2006)2006 IPCC Guidelines for National Greenhouse
Gas Inventories. H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Institute for Global
Environmental Strategies. Available online at: https://www.ipcc-nggip.iges.or.jp/public/2006gl/.
NEBRA (North East Biosolids and Residuals Association) (2022) U.S. National Biosolids Data. Available online
at:
https://static1 .squarespace.com/static/601837d1 c67bcc4e1 b11862f/t/62f4f5fbae32804dd9f51 ef6/166
022092535 6/National BiosolidsDataSummary NBDP 20220811.pdf.
TTB (Alcohol and Tobacco Tax and Trade Bureau) (2024) Beer Statistics. Available online at:
https://www.ttb.gov/beer/statistics.
U.S. Census Bureau (1990). Historical Census of Housing Tables: Sewage Disposal, 1990. Available online:
https://www2.census.gov/programs-survevs/decennial/tables/time-series/coh-
sewage/sewage1990.txt?itid=lk inline enhanced-template
U.S. Census Bureau (2002) Table CO-EST2001-12-00. In: Time Series of I ntercensal State Population
Estimates: April 1, 1990 to April 1, 2000. Release date: April 11, 2002. Available online at:
https://www2.census.gov/programs-survevs/popest/tables/1990-2000/intercensal/st-co/co-est20Q1-
12-OO.pdf.
U.S. Census Bureau (2011) Table ST-EST00INT-01. In: Intercensal Estimates of the Resident Population for
the United States, Regions, States, and Puerto Rico: April 1, 2000 to July 1, 2010. Release date:
September 2011. Available online at: https://www2.census.gov/programs-
surveys/popest/datasets/2000-2010/intercensal/state/st-est00int-alldata.csv.
U.S. Census Bureau (2021a) Table NST-EST2020. In: Annual Estimates of the Resident Population for the
United States, Regions, States, and Puerto Rico: April 1, 2010 to July 1, 2020. Release date: July 2021.
U.S. Census Bureau (2021b) Table NST-EST2021-POP. In: Annual Estimates of the Resident Population for the
United States, Regions, States, District of Columbia, and Puerto Rico: April 1, 2020 to July 1, 2021.
Release date: December 2021.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
6-24
-------
Section 6 Waste (NIR Chapter 7)
U.S. Census Bureau (2022) International Database: World Population Estimates and Projections. Accessed
January 2022. Available online at: https://www.census.gov/programs-surveys/international-
programs/about/idb.html.
U.S. Census Bureau (2023) International Database. Accessed September 2023. Available online at:
https://www.census.gov/data-
tools/demo/idb/#/trends?YR ANIM=2020&dashPages=DASH&FIPS SINGLE=US&COUNTRY YEAR=2022
&menu=trendsViz&TREND RANGE=1990.2022&TREND STEP=5&TREND ADD YRS=&FIPS=AO.GO.CO.
RO.VO&measures=POP&CCODE=AS.GU.MP.PR.US.VI&CCODE SINGLE=US&COUNTRY YR ANIM=202
2.
USDA (U.S. Department of Agriculture) (1991a) Livestock Slaughter: 1990 Summary. Available online at:
https://downloads.usda.librarv.cornell.edu/usda-
esmis/files/r207tp32d/00000397q/bn999b62f/LiveSlauSu-03-00-1991.pdf.
USDA (1991b) Poultry Slaughter. Available online at: https://downloads.usda.library.cornell.edu/usda-
esmis/files/3197xm04j/x920fz52j/cj82k9125/PoulSlau-07-06-1990.pdf.
USDA (2014) Potatoes: 2013 Summary. Available online at:
https://downloads.usda.librarv.cornell.edu/usda-esmis/files/fx719m44h/h128nh39g/g445cg71m/Pota-
09-18-2014.pdf.
USDA (2015a) Vegetables: 2014 Summary. Available online at:
https://downloads.usda.librarv.cornell.edu/usda-
esmis/files/02870v86p/b2773z217/br86b639p/VegeSumm-01 -29-2015.pdf.
USDA (2015b) Citrus Fruits: 2015 Summary. Available online at:
https://downloads.usda.library.cornell.edu/usda-esmis/files/j9602060k/gt54kq48n/xk81jn69p/CitrFrui-
09-17-2015.pdf.
USDA (2015c) Noncitrus Fruits and Nuts: 2014 Summary. Available online at:
https://downloads.usda.library.cornell.edu/usda-
esmis/files/zs25x846c/6108vd86r/3r074x570/NoncFruiNu-07-17-2015.pdf.
USDA (U.S. Department of Agriculture) (2023a) Livestock Slaughter: 2022 Summary. Available online at:
https://downloads.usda.librarv.cornell.edu/usda-
esmis/files/r207tp32d/8p58qs65g/g445dv089/lsan0423.pdf
USDA (U.S. Department of Agriculture) (2023b) Poultry Slaughter: 2022 Summary. Available online at:
https://downloads.usda.librarv.cornell.edu/usda-
esmis/files/pg15bd88s/m613p944x/ht24xx05j/pslaan23.pdf.
Washington Department of Ecology (2022) Discharge Monitoring Reports (DMR) Data. Available online at:
https://apps.ecologv.wa.gov/paris/DischargeMonitoringData.aspx.
Washington Department of Ecology (2024) Facility Summary: Nippon Paper Industries USA Co. Available
online at: https://apps.ecology.wa.gov/paris/FacilitySummary.aspx?Facilityld=18.
6-25
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Appendix A Data Appendices
The data appendices include underlying data used to estimate state-level emissions and sinks (e.g., activity
data/factors, etc.).
A: Energy Sector Combustion Estimates
Please see separate xlsx file.
B: Energy Sector Fugitive Estimates
Please see separate xlsx file.
C: IPPU Minerals Sector Estimates
Please see separate xlsx file.
D: IPPU Chemicals Sector Estimates
Please see separate xlsx file.
E: Agriculture LULUCF Estimates
Please see separate xlsx file.
F: Waste Estimates
Please see separate xlsx file.
G: US Population Data Used in Estimates
Please see separate xlsx file.
H: IPPU Metals Sector Estimates
Please see separate xlsx file.
I: IPPU Product Use Sector Estimates
Please see separate xlsx file.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
A-1
-------
Appendix B State-Level GHG Data Caveats
The state-Level estimates were developed to be consistent with the national Inventory, meaning they
were compiled to avoid double counting or gaps in emissions coverage between states. This was done to
ensure that the state totals, when summed, would equal totals in the national Inventory.
However, there were some instances where either lack of data or updates in the data sources used resulted
in state-level totals that did not add up to the national totals for the categories listed. This was true for the
following source and/or sink categories:
Table B-1. State Level-GHG Data Differences with National GHG Data
^ Years % Difference in Sum of
Sector/Emission and/or ^ ^
^ Where State Totals vs. Reason
Sink Category Di((erent NatlonaiTotal
EnergyFFC C02
2022
-0.0001% (% differences
within FFC subsectors
are higher)
The state-level estimates are
based on updated energy use
data that will be incorporated
into the next version of the
national Inventory.
EnergyNEU C02
All
Max 0.015%
Rounding, differences in
territories data, and
adjustments made to match up
state-level and national-level
NEU values.
EnergyCoal Mines C02
All
Averages <0.01% below
national estimates
across time series
State-level estimates currently
do not include C02from
methane flaring and recovered
coal bed methane. These
estimates are currently only
estimated at the national level
but may be included in the next
annual publication of this data,
potentially in August 2025.
IPPUElectronics
2011-2022
Averages 0.7% lower
from 2011-2022
The list of non-reporting
semiconductor manufacturing
facilities in 2015 was updated
to remove one facility that had
been inadvertently included,
addressing an error in the
national Inventory. In addition,
state-level estimates for HTF
emissions are updated to use
AR5 and AR6 GWPs (where no
value is available in AR5),
addressing an error in the
national Inventory where HTF
estimates were still using AR4
GWPs from 2011 -2022. Thus,
sum of state-level
semiconductor emissions
B-1
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
-------
Methodology Documentation
_ Years % Difference in Sum of
Sector/Emission and/or
, _ Where State Totals vs. Reason
Sink Category ... ,
Different National Total
might not match total
estimates published in the
national Inventory.
LULUCF
Forest Land (harvested
wood pools)
Coastal Wetlands (N20
from aquaculture)
All years
Averages ~12% higher in
the net LULUCF sector
total. While a
percentage is provided, it
is a percentage of net
emissions and sinks in
the LULUCF sector, so it
may not accurately
reflect the relative
sectoral contribution in a
year, including2022.
State-level estimates do not
include (1) emissions and
removals from carbon stock
changes associated with
harvested wood products
(HWP) and (2) N20 emissions
from aquaculture.
Disaggregation of these
sources to the state level will
require further assessment of
potential methods and/or
appropriate surrogate data to
allocate national estimates to
states.
Methodology Report: Inventory of U.S. GHG Emissions and Sinks by State
B-2
------- |