1990-2021
4Rk United States
Environmental Protection
La I » m Agency
EPA430-D-23-001
DRAFT Inventory of
U.S. Greenhouse Gas
Emissions and Sinks
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Front cover photo credit for cow and digester: Vanguard Renewables.
-------
1
HOW TO OBTAIN COPIES
2 You can electronically download this document on the U.S. EPA's homepage at
3 https://www.epa.gov/ghgemissions/inventory-us-greenhouse-gas-emissions-and-sinks.
4 All data tables of this document for the full time series 1990 through 2021, inclusive, will be made available with
5 the final report published by April 15, 2023 at the internet site mentioned above.
6
7 RECOMMENDED CITATION
8 EPA (2023) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021. U.S. Environmental Protection
9 Agency, EPA 430-D-23-001. https://www.epa.gov/ghgemissions/draft-inventorv-us-greenhouse-gas-emissions-
10 and-sinks-1990-2021.
11
12 FOR FURTHER INFORMATION
13 Contact Ms. Mausami Desai, Environmental Protection Agency, (202) 343-9381, desai.mausami@epa.gov,
14 or Mr. Vincent Camobreco, Environmental Protection Agency, (202) 564-9043, camobreco.vincent@epa.gov.
15 For more information regarding climate change and greenhouse gas emissions, see the EPA web site at
16 https://www.epa.gov/ghgemissions.
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Acknowledgments
The Environmental Protection Agency would like to acknowledge the many individual and organizational
contributors to this document, 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 directed the work to compile estimates of emissions
from mobile sources. Work on fugitive methane emissions from the Energy sector was directed by Melissa Weitz
and Chris Sherry. Development and compilation of emissions estimates for the Waste sector were led by Lauren
Aepli and Mausami Desai. Tom Wirth and John Steller directed work to compile estimates for the Agriculture and
the Land Use, Land-Use Change, and Forestry chapters with support from Jake Beaulieu (ORD) on compiling the
inventories for CO2 and Cm associated with flooded lands. Development and compilation of Industrial Processes
and Product Use (IPPU) CO2, CH4, and N2O emissions was directed by Amanda Chiu and Vincent Camobreco.
Development and compilation of emissions of HFCs, PFCs, SF6, and NF3 from the IPPU sector was directed by
Deborah Ottinger, Dave Godwin, Stephanie Bogle, and Kersey Manliclic. Cross-cutting work was directed by
Mausami Desai. We thank Bill Irving for general advice, guidance, and cross-cutting review.
We also thank Kenna Rewcastle and Martin Wolf (AAAS Science & Technology Policy Fellows hosted by the Office
of Atmospheric Protection) for their advice and review in areas such as agricultural, land use, land use change and
forestry, and industrial processes and product use estimates, respectively.
Other EPA offices and programs also contributed data, analysis, and technical review for this report. The Office of
Atmospheric Protection's Greenhouse Gas Reporting Program (OAP) staff facilitated aggregation and review of
facility-level data for use in the Inventory, in particular aggregation of confidential business information data. The
Office of Transportation and Air Quality and the Office of Air Quality Planning and Standards provided analysis (i.e.,
precursors) and review for several of the source categories (i.e., natural gas and petroleum systems) included in
this report. The Office of Research and Development conducted field research and developed estimates associated
with flooded lands. The Office of Land and Emergency Management also contributed analysis and research.
The Energy Information Administration and the Department of Energy contributed invaluable data and analysis on
numerous energy-related topics. William Sanchez at EIA provided annual energy data that are used in fossil fuel
combustion estimates. Other government agencies have contributed data as well, including the U.S. Geological
Survey, the Federal Highway Administration, the Department of Transportation, the Bureau of Transportation
Statistics, the Department of Commerce, the Mine Safety and Health Administration, and the National Agricultural
Statistics Service.
We thank the Department of Defense (David Asiello, DoD and Matthew Cleaver of Leidos) for compiling the data
on military bunker fuel use.
We thank the Federal Aviation Administration (Ralph Lovinelli and Jeetendra Upadhyay) for compiling the
inventory of emissions from commercial aircraft jet fuel consumption.
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
We thank the U.S. Forest Service (Grant Domke, Brian Walters, Jim Smith and Mike Nichols) for compiling the
inventories for CO2, CH4, and N2O fluxes associated with forest land.
We thank the Department of Agriculture's Agricultural Research Service (Stephen Del Grosso) and the Natural
Resource Ecology Laboratory and Department of Statistics at Colorado State University (Stephen Ogle, Bill Parton,
F. Jay Breidt, Shannon Spencer, Alisa Keyser, Teng Liu, Ryan Scheiderer, Veronica Thompson, Doug Vander Wilt,
Stephen Williams, Guhan Dheenadayalan Sivakami) for compiling the inventories for CH4 emissions, N2O emissions,
and CO2 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 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 N2O
emissions from aquaculture in coastal wetlands. We also thank NOAA (Stephen Montzka and Lei Hu) for
information on atmospheric measurements and derived emissions of HFCs.
We thank Marian Martin Van Pelt, Leslie Chinery, Alexander Lataille, and the full Inventory team at ICF including
Diana Pape, Robert Lanza, Mollie Averyt, Larry O'Rourke, Deborah Harris, Rebecca Ferenchiak, Fiona Wissell,
Bikash Acharya, Mollie Carroll, Kyle Herdegen, Hazelle Tomlin, Deep Shah, Lou Browning, Sarah Whitlock, David
Landolfi, Emily Carr, Abby Wiseman, Georgia Kerkezis, Katie O'Malley, Emily Adkins, Zeyu Hu, Alex Da Silva, Shubh
Jain, Alida Monaco, Kenny Yerardi, Hannah Krauss, Sophie Johnson, Leah Hartung, Molly Rickles, Audrey Ichida,
Seth Hartley, Max Kaffel, Sam Pournazer, and Ajo Rabemiarisoa for technical support in compiling synthesis
information across the report and preparing many of the individual analyses for specific report chapters including
fluorinated emissions and fuel combustion.
We thank Eastern Research Group for their analytical support. Deborah Bartram, Kara Edquist and Tara Stout
support the development of emissions estimates for wastewater. Kara Edquist, Cortney Itle, Amber Allen, Spencer
Sauter, Tara Stout, and Sarah Wagner support the development of emission estimates for Manure Management,
Enteric Fermentation, Peatlands (included in Wetlands Remaining Wetlands^, and Landfilled Yard Trimmings and
Food Scraps (included in Settlements Remaining Settlements). Brandon Long, Gopi Manne, Marty Wolf, and Sarah
Downes, develop estimates for Natural Gas and Petroleum Systems. Gopi Manne and Tara Stout support the
development of emission estimates for coal mine methane.
Finally, we thank the RTI International team: Kate Bronstein, Emily Thompson, Jeff Coburn, and Keith Weitz for
their analytical support in development of the estimates of emissions from landfills; Jason Goldsmith, Melissa
Icenhour, Michael Laney, David Randall, Gabrielle Raymond, Karen Schaffner, Riley Vanek, Ricky Strott, Libby
Robinson, Matt Hakos, and Jeremy Kaelin for their analytical support in development of IPPU CO2, CH4, and N2O
emissions; and Tiffany Moore and Matt Hakos for their analytical support on disaggregating industrial sector fossil
fuel combustion emissions.
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Preface
The United States Environmental Protection Agency (EPA) prepares the official U.S. Inventory of Greenhouse Gas
Emissions and Sinks to fulfill annual existing commitments under the United Nations Framework Convention on
Climate Change (UNFCCC). Under UNFCCC Article 4 and decisions at the First, Second, Fifth and Nineteenth
Conference of Parties, national inventories for UNFCCC Annex I parties should be provided to the UNFCCC
Secretariat each year by April 15.
In an effort to engage the public and researchers across the country, the EPA has instituted an annual public
review and comment process for this document. The availability of the draft document on the EPA Greenhouse
Gas Emissions web site was announced via Federal Register Notice. The public comment period covers a 30-day
period from February 15 through March 17, 2023, and comments received during the public review period will be
posted to the docket EPA-HQ-OAR-2023-0001. Comments received after the closure of the public comment period
are accepted and will be considered for the next edition of this annual report. Responses to comments are typically
posted to EPA's website 2-4 weeks following publication of the final report in April 2023.
-------
1 Table of Contents
2 TABLE OF CONTENTS VI
3 LIST OF TABLES, FIGURES, BOXES, AND EQUATIONS IX
4 EXECUTIVE SUMMARY ES-1
5 ES.l Background Information ES-2
6 ES.2 Recent Trends in U.S. Greenhouse Gas Emissions and Sinks ES-4
7 ES.3 Overview of Sector Emissions and Trends ES-16
8 ES.4 Other Information ES-20
9 1. INTRODUCTION 1-1
10 1.1 Background Information 1-3
11 1.2 National Inventory Arrangements 1-11
12 1.3 Inventory Process 1-13
13 1.4 Methodology and Data Sources 1-17
14 1.5 Key Categories 1-18
15 1.6 Quality Assurance and Quality Control (QA/QC) 1-24
16 1.7 Uncertainty Analysis of Emission Estimates 1-29
17 1.8 Completeness 1-32
18 1.9 Organization of Report 1-32
19 2. TRENDS IN GREENHOUSE GAS EMISSIONS AND REMOVALS 2-1
20 2.1 Overview of U.S. Greenhouse Gas Emissions and Sinks Trends 2-1
21 2.2 Emissions by Economic Sector 2-28
22 2.3 Precursor Greenhouse Gas Emissions (CO, NOx, NMVOCs, and SO2) - TO BE UPDATED FOR FINAL
23 INVENTORY REPORT 2-40
24 3. ENERGY 3-1
25 3.1 Fossil Fuel Combustion (CRF Source Category 1A) 3-7
26 3.2 Carbon Emitted from Non-Energy Uses of Fossil Fuels (CRF Source Category 1A) 3-51
27 3.3 Incineration of Waste (CRF Source Category 1A5) 3-59
28 3.4 Coal Mining (CRF Source Category lBla) 3-63
vi DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 3.5 Abandoned Underground Coal Mines (CRF Source Category lBla) 3-71
2 3.6 Petroleum Systems (CRF Source Category lB2a) 3-75
3 3.7 Natural Gas Systems (CRF Source Category lB2b) 3-93
4 3.8 Abandoned Oil and Gas Wells (CRF Source Categories lB2a and lB2b) 3-112
5 3.9 International Bunker Fuels (CRF Source Category 1: Memo Items) 3-117
6 3.10 Biomass and Biofuels Consumption (CRF Source Category 1A) 3-122
7 3.11 Energy Sources of Precursor Greenhouse Gas -TO BE UPDATED FOR FINAL INVENTORY REPORT 3-127
8 4. INDUSTRIAL PROCESSES AND PRODUCT USE 4-1
9 4.1 Cement Production (CRF Source Category 2A1) 4-10
10 4.2 Lime Production (CRF Source Category 2A2) 4-14
11 4.3 Glass Production (CRF Source Category 2A3) 4-20
12 4.4 Other Process Uses of Carbonates (CRF Source Category 2A4) 4-24
13 4.5 Ammonia Production (CRF Source Category 2B1) 4-28
14 4.6 Urea Consumption for Non-Agricultural Purposes 4-33
15 4.7 Nitric Acid Production (CRF Source Category 2B2) 4-36
16 4.8 Adipic Acid Production (CRF Source Category 2B3) 4-41
17 4.9 Caprolactam, Glyoxal and Glyoxylic Acid Production (CRF Source Category 2B4) 4-45
18 4.10 Carbide Production and Consumption (CRF Source Category 2B5) 4-48
19 4.11 Titanium Dioxide Production (CRF Source Category 2B6) 4-52
20 4.12 Soda Ash Production (CRF Source Category 2B7) 4-55
21 4.13 Petrochemical Production (CRF Source Category 2B8) 4-58
22 4.14 HCFC-22 Production (CRF Source Category 2B9a) 4-67
23 4.15 Carbon Dioxide Consumption (CRF Source Category 2B10) 4-70
24 4.16 Phosphoric Acid Production (CRF Source Category 2B10) 4-73
25 4.17 Iron and Steel Production (CRF Source Category 2C1) and Metallurgical Coke Production 4-77
26 4.18 Ferroalloy Production (CRF Source Category 2C2) 4-89
27 4.19 Aluminum Production (CRF Source Category 2C3) 4-93
28 4.20 Magnesium Production and Processing (CRF Source Category 2C4) 4-100
29 4.21 Lead Production (CRF Source Category 2C5) 4-106
30 4.22 Zinc Production (CRF Source Category 2C6) 4-109
31 4.23 Electronics Industry (CRF Source Category 2E) 4-115
32 4.24 Substitution of Ozone Depleting Substances (CRF Source Category 2F) 4-132
33 4.25 Electrical Transmission and Distribution (CRF Source Category 2G1) 4-141
34 4.26 Nitrous Oxide from Product Uses (CRF Source Category 2G3) 4-153
35 4.27 Industrial Processes and Product Use Sources of Precursor Gases 4-156
vii
-------
1 5. AGRICULTURE 5-1
2 5.1 Enteric Fermentation (CRF Source Category 3A) 5-4
3 5.2 Manure Management (CRF Source Category 3B) 5-12
4 5.3 Rice Cultivation (CRF Source Category 3C) 5-21
5 5.4 Agricultural Soil Management (CRF Source Category 3D) 5-28
6 5.5 Liming (CRF Source Category 3G) 5-48
7 5.6 Urea Fertilization (CRF Source Category 3H) 5-51
8 5.7 Field Burning of Agricultural Residues (CRF Source Category 3F) 5-53
9 6. LAND USE, LAND-USE CHANGE, AND FORESTRY 6-1
10 6.1 Representation of the U.S. Land Base 6-9
11 6.2 Forest Land Remaining Forest Land (CRF Category 4A1) 6-24
12 6.3 Land Converted to Forest Land (CRF Source Category 4A2) 6-48
13 6.4 Cropland Remaining Cropland (CRF Category 4B1) 6-56
14 6.5 Land Converted to Cropland (CRF Category 4B2) 6-67
15 6.6 Grassland Remaining Grassland (CRF Category 4C1) 6-75
16 6.7 Land Converted to Grassland (CRF Category 4C2) 6-87
17 6.8 Wetlands Remaining Wetlands (CRF Category 4D1) 6-95
18 6.9 Land Converted to Wetlands (CRF Source Category 4D2) 6-138
19 6.10 Settlements Remaining Settlements (CRF Category 4E1) 6-160
20 6.11 Land Converted to Settlements (CRF Category 4E2) 6-181
21 6.12 Other Land Remaining Other Land (CRF Category 4F1) 6-188
22 6.13 Land Converted to Other Land (CRF Category 4F2) 6-188
23 7. WASTE 7-1
24 7.1 Landfills (CRF Source Category 5A1) 7-4
25 7.2 Wastewater Treatment and Discharge (CRF Source Category 5D) 7-21
26 7.3 Composting (CRF Source Category 5B1) 7-55
27 7.4 Anaerobic Digestion at Biogas Facilities (CRF Source Category 5B2) 7-60
28 7.5 Waste Incineration (CRF Source Category 5C1) 7-66
29 7.6 Waste Sources of Precursor Greenhouse Gases -TO BE UPDATED FOR FINAL INVENTORY REPORT 7-67
30 8. OTHER 8-1
31 9. RECALCULATIONS AND IMPROVEMENTS 9-1
32 REFERENCES AND ABBREVIATIONS 9-1
33
34
viii DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 List of Tables, Figures, Boxes, and
2 Equations
3 Tables
4 Table ES-1: Global Warming Potentials (100-Year Time Horizon) Used in this Report ES-3
5 Table ES-2: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks (MMT CO2 Eq.) ES-4
6 Table ES-3: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks by IPCC Sector/Category (MMT CO2 Eq.)
7 ES-16
8 Table ES-4: U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use, Land-Use Change, and
9 Forestry (MMT C02 Eq.) ES-19
10 Table ES-5: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO2 Eq.) ES-21
11 Table ES-6: U.S. Greenhouse Gas Emissions with Electricity-Related Emissions Distributed by Economic Sector
12 (MMT CO2 Eq.) ES-22
13 Table ES-7: Recent Trends in Various U.S. Data (Index 1990 = 100) ES-23
14 Table 1-1: Global Atmospheric Concentration, Rate of Concentration Change, and Atmospheric Lifetime of
15 Selected Greenhouse Gases 1-5
16 Table 1-2: Global Warming Potentials and Atmospheric Lifetimes (Years) Used in this Report 1-10
17 Table 1-3: Comparison of 100-Year GWP values 1-11
18 Table 1-4: Summary of Key Categories for the United States (1990 and 2021) by Sector 1-19
19 Table 1-5: Estimated Overall Inventory Quantitative Uncertainty for 1990 (MMT CO2 Eq. and Percent) - TO BE
20 UPDATED FOR FINAL INVENTORY REPORT 1-29
21 Table 1-6: Estimated Overall Inventory Quantitative Uncertainty for 2020 (MMT CO2 Eq. and Percent) - TO BE
22 UPDATED FOR FINAL INVENTORY REPORT 1-30
23 Table 1-7: Quantitative Assessment of Trend Uncertainty (MMT CO2 Eq. and Percent) 1-31
24 Table 1-8: IPCC Sector Descriptions 1-32
25 Table 1-9: List of Annexes 1-33
26 Table 2-1: RecentTrends in U.S. Greenhouse Gas Emissions and Sinks (MMT CO2 Eq.) 2-3
27 Table 2-2: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks (kt) 2-6
28 Table 2-3: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks by IPCC Sector/Category (MMT CO2 Eq.).. 2-9
29 Table 2-4: Emissions from Energy (MMT CO2 Eq.) 2-11
30 Table 2-5: CO2 Emissions from Fossil Fuel Combustion by End-Use Sector (MMT CO2 Eq.) 2-14
31 Table 2-6: Emissions from Industrial Processes and Product Use (MMT CO2 Eq.) 2-19
32 Table 2-7: Emissions from Agriculture (MMT CO2 Eq.) 2-22
33 Table 2-8: U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use, Land-Use Change, and Forestry
34 (MMT CO2 Eq.) 2-24
35 Table 2-9: Emissions from Waste (MMT CO2 Eq.) 2-27
ix
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Table 2-10: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO2 Eq. and Percent of Total in
2021) 2-29
Table 2-11: Electric Power-Related Greenhouse Gas Emissions (MMT CO2 Eq.) 2-33
Table 2-12: U.S. Greenhouse Gas Emissions by Economic Sector and Gas with Electricity-Related Emissions
Distributed (MMT CO2 Eq.) and Percent of Total in 2021 2-34
Table 2-13: Transportation-Related Greenhouse Gas Emissions (MMT CO2 Eq.) 2-37
Table 2-14: Recent Trends in Various U.S. Data (Index 1990 = 100) 2-39
Table 2-15: Emissions of NOx, CO, NMVOCs, and S02 (kt) 2-41
Table 3-1: CO2, CFU, and N2O Emissions from Energy (MMT CO2 Eq.) 3-3
Table 3-2: CO2, CFU, and N2O Emissions from Energy (kt) 3-4
Table 3-3: CO2, CFU, and N2O Emissions from Fossil Fuel Combustion (MMT CO2 Eq.) 3-7
Table 3-4: CO2, CFU, and N2O Emissions from Fossil Fuel Combustion (kt) 3-7
Table 3-5: CO2 Emissions from Fossil Fuel Combustion by Fuel Type and Sector (MMT CO2 Eq.) 3-8
Table 3-6: Annual Change in CO2 Emissions and Total 2021 CO2 Emissions from Fossil Fuel Combustion for Selected
Fuels and Sectors (MMT CO2 Eq. and Percent) 3-9
Table 3-7: CO2 Emissions from Stationary Fossil Fuel Combustion (MMT CO2 Eq.) 3-13
Table 3-8: CFU Emissions from Stationary Combustion (MMT CO2 Eq.) 3-14
Table 3-9: N2O Emissions from Stationary Combustion (MMT CO2 Eq.) 3-14
Table 3-10: CO2, CFU, and N2O Emissions from Fossil Fuel Combustion by Sector (MMT CO2 Eq.) 3-15
Table 3-11: CO2, CFU, and N2O Emissions from Fossil Fuel Combustion by End-Use Sector with Electricity Emissions
Distributed (MMT C02 Eq.) 3-16
Table 3-12: Electric Power Generation by Fuel Type (Percent) 3-17
Table 3-13: CO2 Emissions from Fossil Fuel Combustion in Transportation End-Use Sector (MMT CO2 Eq.) 3-27
Table 3-14: CFU Emissions from Mobile Combustion (MMT CO2 Eq.) 3-30
Table 3-15: N2O Emissions from Mobile Combustion (MMT CO2 Eq.) 3-31
Table 3-16: Carbon Intensity from Direct Fossil Fuel Combustion by Sector (MMT CO2 Eq./QBtu) 3-36
Table 3-17: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Energy-Related Fossil Fuel
Combustion by Fuel Type and Sector (MMT CO2 Eq. and Percent) 3-39
Table 3-18: Comparison of Electric Power Sector Emissions (MMT CO2 Eq. and Percent) 3-40
Table 3-19: Comparison of Emissions Factors (MMT Carbon/QBtu) 3-40
Table 3-20: Approach 2 Quantitative Uncertainty Estimates for CFU and N2O Emissions from Energy-Related
Stationary Combustion, Including Biomass (MMT CO2 Eq. and Percent) 3-45
Table 3-21: Approach 2 Quantitative Uncertainty Estimates for CFU and N2O Emissions from Mobile Sources (MMT
CO2 Eq. and Percent) 3-49
Table 3-22: CO2 Emissions from Non-Energy Use Fossil Fuel Consumption (MMT CO2 Eq. and Percent C) 3-52
Table 3-23: Adjusted Consumption of Fossil Fuels for Non-Energy Uses (TBtu) 3-53
Table 3-24: 2021 Adjusted Non-Energy Use Fossil Fuel Consumption, Storage, and Emissions 3-54
x DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Table 3-25: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Non-Energy Uses of Fossil Fuels
(MMT CO2 Eq. and Percent) 3-56
Table 3-26: Approach 2 Quantitative Uncertainty Estimates for Storage Factors of Non-Energy Uses of Fossil Fuels
(Percent) 3-57
Table 3-27: CO2, CFU, and N2O Emissions from the Combustion of Waste (MMT CO2 Eq.) 3-60
Table 3-28: CO2, CFU, and N2O Emissions from the Combustion of Waste (kt) 3-60
Table 3-29: Municipal Solid Waste Combusted (Short Tons) 3-60
Table 3-30: Calculated Fossil CO2 Content per Ton Waste Combusted (kg C02/Short Ton Combusted) 3-61
Table 3-31: CO2 Emissions from Combustion of Tires (MMT CO2 Eq.) 3-61
Table 3-32: Approach 2 Quantitative Uncertainty Estimates for CO2 and N2O from the Incineration of Waste (MMT
CO2 Eq. and Percent) 3-62
Table 3-33: Coal Production (kt) 3-63
Table 3-34: CFU Emissions from Coal Mining (MMT CO2 Eq.) 3-64
Table 3-35: CFU Emissions from Coal Mining (kt) 3-64
Table 3-36: CO2 Emissions from Coal Mining (MMT CO2 Eq.) 3-67
Table 3-37: CO2 Emissions from Coal Mining (kt) 3-68
Table 3-38: Approach 2 Quantitative Uncertainty Estimates for CFU and CO2 Emissions from Coal Mining (MMT CO2
Eq. and Percent) 3-70
Table 3-39: CFU Emissions from Abandoned Coal Mines (MMT CO2 Eq.) 3-72
Table 3-40: CFU Emissions from Abandoned Coal Mines (kt) 3-72
Table 3-41: Number of Gassy Abandoned Mines Present in U.S. Basins in 2021, Grouped by Class According to
Post-Abandonment State 3-73
Table 3-42: Approach 2 Quantitative Uncertainty Estimates for CFU Emissions from Abandoned Underground Coal
Mines (MMT CO2 Eq. and Percent) 3-75
Table 3-43: Total Greenhouse Gas Emissions (CO2, CFU, and N2O) from Petroleum Systems (MMT CO2 Eq.) 3-77
Table 3-44: CFU Emissions from Petroleum Systems (MMT CO2 Eq.) 3-78
Table 3-45: CFU Emissions from Petroleum Systems (kt CH4) 3-78
Table 3-46: CO2 Emissions from Petroleum Systems (MMT CO2) 3-78
Table 3-47: CO2 Emissions from Petroleum Systems (kt CO2) 3-78
Table 3-48: N2O Emissions from Petroleum Systems (Metric Tons CO2 Eq.) 3-79
Table 3-49: N2O Emissions from Petroleum Systems (Metric Tons N2O) 3-79
Table 3-50: Approach 2 Quantitative Uncertainty Estimates for CFU and CO2 Emissions from Petroleum Systems
(MMT CO2 Eq. and Percent) 3-82
Table 3-51
Table 3-52
Table 3-53
Table 3-54
Table 3-55
Recalculations of CO2 in Petroleum Systems (MMT CO2) 3-85
Recalculations of CH4 in Petroleum Systems (MMT CO2 Eq.) 3-86
Pneumatic Controllers National CFU Emissions (Metric Tons CH4) 3-87
Production Equipment Leaks National CH4 Emissions (Metric Tons CH4) 3-87
Chemical Injection Pumps National CFU Emissions (Metric Tons CH4) 3-88
xi
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Table 3-56: Storage Tanks National Cm Emissions (Metric Tons CFU) 3-88
Table 3-57: Storage Tanks National CO2 Emissions (kt CO2) 3-89
Table 3-58: Associated Gas Flaring National CO2 Emissions (kt CO2) 3-89
Table 3-59: Miscellaneous Production Flaring National CO2 Emissions (kt CO2) 3-90
Table 3-60: Miscellaneous Production Flaring National CH4 Emissions (Metric Tons CH4) 3-90
Table 3-61: Offshore Production National CFU Emissions (Metric Tons CH4) 3-90
Table 3-62: Refining National CO2 Emissions (kt CO2) 3-91
Table 3-63: Quantity of CO2 Captured and Extracted for EOR Operations (kt CO2) 3-92
Table 3-64: Geologic Sequestration Information Reported Under GHGRP Subpart RR 3-92
Table 3-65: Total Greenhouse Gas Emissions (CH4, CO2, and N2O) from Natural Gas Systems (MMT CO2 Eq.) 3-96
Table 3-66: CFU Emissions from Natural Gas Systems (MMT CO2 Eq.) 3-96
Table 3-67: CFU Emissions from Natural Gas Systems (kt) 3-96
Table 3-68: CO2 Emissions from Natural Gas Systems (MMT) 3-97
Table 3-69: CO2 Emissions from Natural Gas Systems (kt) 3-97
Table 3-70: N2O Emissions from Natural Gas Systems (MetricTons CO2 Eq.) 3-97
Table 3-71: N2O Emissions from Natural Gas Systems (Metric Tons N2O) 3-97
Table 3-72: Approach 2 Quantitative Uncertainty Estimates for CH4 and Non-combustion CO2 Emissions from
Natural Gas Systems (MMT CO2 Eq. and Percent) 3-100
Table 3-73: Recalculations of CO2 in Natural Gas Systems (MMT CO2) 3-103
Table 3-74: Recalculations of CH4 in Natural Gas Systems (MMT CO2 Eq.) 3-104
Table 3-75: Pneumatic Controllers National CFU Emissions (Metric Tons CH4) 3-105
Table 3-76: Storage Tanks National CH4 Emissions (Metric Tons CH4) 3-105
Table 3-77: Storage Tanks National CO2 Emissions (kt CO2) 3-106
Table 3-78: Production Equipment Leaks National CH4 Emissions (Metric Tons CH4) 3-106
Table 3-79: Chemical Injection Pumps National CFU Emissions (Metric Tons CH4) 3-107
Table 3-80: Liquids Unloading National CFU Emissions (Metric Tons CH4) 3-108
Table 3-81: Liquids Unloading National CO2 Emissions (MetricTons CO2) 3-108
Table 3-82: Miscellaneous Production Flaring National Emissions (kt CO2) 3-108
Table 3-83: Tanks National Emissions (Metric Tons CH4) 3-109
Table 3-84: Station Blowdowns National Emissions (Metric Tons CH4) 3-109
Table 3-85: Dehydrator Vents National Emissions (Metric Tons CH4) 3-109
Table 3-86: Dehydrator Vents National Emissions (kt CO2) 3-109
Table 3-87: Production Storage Tanks National Emissions (kt CO2) 3-110
Table 3-88: Processing Segment Flares National CO2 Emissions (kt CO2) 3-110
Table 3-89: AGR Vents National CO2 Emissions (kt CO2) 3-110
Table 3-90: Mains - Unprotected Steel National CFU Emissions (MetricTons CH4) 3-111
xii DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Table 3-91: Post-Meter National Cm Emissions (Metric Tons Cm) 3-111
Table 3-92: Cm Emissions from Abandoned Oil and Gas Wells (MMT CO2 Eq.) 3-112
Table 3-93: Cm Emissions from Abandoned Oil and Gas Wells (kt) 3-112
Table 3-94: CO2 Emissions from Abandoned Oil and Gas Wells (MMT CO2) 3-113
Table 3-95: CO2 Emissions from Abandoned Oil and Gas Wells (kt) 3-113
Table 3-96: Abandoned Oil Wells Activity Data, CH4 and CO2 Emissions (kt) 3-114
Table 3-97: Abandoned Gas Wells Activity Data, CH4 and CO2 Emissions (kt) 3-114
Table 3-98: Approach 2 Quantitative Uncertainty Estimates for Cm and CO2 Emissions from Petroleum and Natural
Gas Systems (MMT CO2 Eq. and Percent) 3-115
Table 3-99: CO2, Cm, and N2O Emissions from International Bunker Fuels (MMT CO2 Eq.) 3-118
Table 3-100: CO2, CH4, and N2O Emissions from International Bunker Fuels (kt) 3-118
Table 3-101: Aviation Jet Fuel Consumption for International Transport (Million Gallons) 3-119
Table 3-102: Marine Fuel Consumption for International Transport (Million Gallons) 3-120
Table 3-103: CO2 Emissions from Wood Consumption by End-Use Sector (MMT CO2 Eq.) 3-122
Table 3-104: CO2 Emissions from Wood Consumption by End-Use Sector (kt) 3-123
Table 3-105: CO2 Emissions from Biogenic Components of MSW (MMT CO2 Eq.) 3-123
Table 3-106: CO2 Emissions from Biogenic Components of MSW (kt) 3-123
Table 3-107: CO2 Emissions from Ethanol Consumption (MMT CO2 Eq.) 3-123
Table 3-108: CO2 Emissions from Ethanol Consumption (kt) 3-123
Table 3-109: CO2 Emissions from Biodiesel Consumption (MMT CO2 Eq.) 3-124
Table 3-110: CO2 Emissions from Biodiesel Consumption (kt) 3-124
Table 3-111: Calculated Biogenic CO2 Content per Ton Waste (kg C02/Short Ton Combusted) 3-124
Table 3-112: Woody Biomass Consumption by Sector (Trillion Btu) 3-125
Table 3-113: Ethanol Consumption by Sector (Trillion Btu) 3-125
Table 3-114: Biodiesel Consumption by Sector (Trillion Btu) 3-125
Table 3-115: NOx, CO, NMVOC, and SO2 Emissions from Energy-Related Activities (kt) 3-127
Table 4-1: Emissions from Industrial Processes and Product Use (MMT CO2 Eq.) 4-4
Table 4-2: Emissions from Industrial Processes and Product Use (kt) 4-5
Table 4-3: CO2 Emissions from Cement Production (MMT CO2 Eq. and kt) 4-10
Table 4-4: Clinker Production (kt) 4-12
Table 4-5: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Cement Production (MMT CO2
Eq. and Percent) 4-12
Table 4-6: CO2 Emissions from Lime Production (MMT CO2 Eq. and kt) 4-15
Table 4-7: Gross, Recovered, and Net CO2 Emissions from Lime Production (kt) 4-15
Table 4-8: High-Calcium- and Dolomitic-Quicklime, High-Calcium- and Dolomitic-Hydrated, and Dead-Burned-
Dolomite Lime Production (kt) 4-17
xiii
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Table 4-9: Adjusted Lime Production (kt) 4-17
Table 4-10: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Lime Production (MMT CO2 Eq.
and Percent) 4-18
Table 4-11: CO2 Emissions from Glass Production (MMT CO2 Eq. and kt) 4-21
Table 4-12: Limestone, Dolomite, Soda Ash, and Other Carbonates Used in Glass Production (kt) and Average
Annual Production Index for Glass and Glass Product Manufacturing 4-22
Table 4-13: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Glass Production (MMT CO2 Eq.
and Percent) 4-23
Table 4-14: CO2 Emissions from Other Process Uses of Carbonates (MMT CO2 Eq.) 4-25
Table 4-15: CO2 Emissions from Other Process Uses of Carbonates (kt) 4-25
Table 4-16: Limestone and Dolomite Consumption (kt) 4-26
Table 4-17: Soda Ash Consumption Not Associated with Glass Manufacturing (kt) 4-27
Table 4-18: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Other Process Uses of
Carbonates (MMT CO2 Eq. and Percent) 4-28
Table 4-19: CO2 Emissions from Ammonia Production (MMT CO2 Eq.) 4-30
Table 4-20: CO2 Emissions from Ammonia Production (kt) 4-30
Table 4-21: Total Ammonia Production, Total Urea Production, and RecoveredC02 Consumed for Urea Production
(kt) 4-31
Table 4-22: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Ammonia Production (MMT
CO2 Eq. and Percent) 4-32
Table 4-23: CO2 Emissions from Urea Consumption for Non-Agricultural Purposes (MMT CO2 Eq.) 4-34
Table 4-24: CO2 Emissions from Urea Consumption for Non-Agricultural Purposes (kt) 4-34
Table 4-25: Urea Production, Urea Applied as Fertilizer, Urea Imports, and Urea Exports (kt) 4-35
Table 4-26: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Urea Consumption for Non-
Agricultural Purposes (MMT CO2 Eq. and Percent) 4-35
Table 4-27: N2O Emissions from Nitric Acid Production (MMT CO2 Eq. and kt N2O) 4-37
Table 4-28: Nitric Acid Production (kt) 4-39
Table 4-29: Approach 2 Quantitative Uncertainty Estimates for N2O Emissions from Nitric Acid Production (MMT
CO2 Eq. and Percent) 4-40
Table 4-30: N2O Emissions from Adipic Acid Production (MMT CO2 Eq. and kt N2O) 4-42
Table 4-31: Adipic Acid Production (kt) 4-43
Table 4-32: Approach 2 Quantitative Uncertainty Estimates for N2O Emissions from Adipic Acid Production (MMT
CO2 Eq. and Percent) 4-43
Table 4-33: N2O Emissions from Caprolactam Production (MMT CO2 Eq. and kt N2O) 4-46
Table 4-34: Caprolactam Production (kt) 4-47
Table 4-35: Approach 2 Quantitative Uncertainty Estimates for N2O Emissions from Caprolactam, Glyoxal and
Glyoxylic Acid Production (MMT CO2 Eq. and Percent) 4-47
Table 4-36: CO2 and CFU Emissions from Silicon Carbide Production and Consumption (MMT CO2 Eq.) 4-49
Table 4-37: CO2 and CFU Emissions from Silicon Carbide Production and Consumption (kt) 4-49
xiv DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Table 4-38: Production and Consumption of Silicon Carbide (Metric Tons) 4-51
Table 4-39: Approach 2 Quantitative Uncertainty Estimates for Cm and CO2 Emissions from Silicon Carbide
Production and Consumption (MMT CO2 Eq. and Percent) 4-51
Table 4-40: CO2 Emissions from Titanium Dioxide (MMT CO2 Eq. and kt) 4-53
Table 4-41: Titanium Dioxide Production (kt) 4-54
Table 4-42: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Titanium Dioxide Production
(MMT CO2 Eq. and Percent) 4-55
Table 4-43: CO2 Emissions from Soda Ash Production (MMT CO2 Eq. and kt CO2) 4-56
Table 4-44: Trona Ore Used in Soda Ash Production (kt) 4-57
Table 4-45: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Soda Ash Production (MMT
CO2 Eq. and Percent) 4-58
Table 4-46: CO2 and Cm Emissions from Petrochemical Production (MMT CO2 Eq.) 4-60
Table 4-47: CO2 and Cm Emissions from Petrochemical Production (kt) 4-60
Table 4-48: Production of Selected Petrochemicals (kt) 4-63
Table 4-49: Approach 2 Quantitative Uncertainty Estimates for Cm Emissions from Petrochemical Production and
CO2 Emissions from Petrochemical Production (MMT CO2 Eq. and Percent) 4-64
Table 4-50: HFC-23 Emissions from HCFC-22 Production (MMT C02 Eq. and kt HFC-23) 4-67
Table 4-51: HCFC-22 Production (kt) 4-68
Table 4-52: Approach 2 Quantitative Uncertainty Estimates for HFC-23 Emissions from HCFC-22 Production (MMT
CO2 Eq. and Percent) 4-69
Table 4-53: CO2 Emissions from CO2 Consumption (MMT CO2 Eq. and kt) 4-70
Table 4-54: CO2 Production (kt CO2) and the Percent Used for Non-EOR Applications 4-71
Table 4-55: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from CO2 Consumption (MMT CO2
Eq. and Percent) 4-72
Table 4-56: CO2 Emissions from Phosphoric Acid Production (MMT CO2 Eq. and kt) 4-74
Table 4-57: Phosphate Rock Domestic Consumption, Exports, and Imports (kt) 4-75
Table 4-58: Chemical Composition of Phosphate Rock (Percent by Weight) 4-75
Table 4-59: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Phosphoric Acid Production
(MMT CO2 Eq. and Percent) 4-76
Table 4-60: CO2 Emissions from Metallurgical Coke Production (MMT CO2 Eq.) 4-78
Table 4-61: CO2 Emissions from Metallurgical Coke Production (kt) 4-78
Table 4-62: CO2 Emissions from Iron and Steel Production (MMT CO2 Eq.) 4-79
Table 4-63: CO2 Emissions from Iron and Steel Production (kt) 4-79
Table 4-64: CH4 Emissions from Iron and Steel Production (MMT CO2 Eq.) 4-79
Table 4-65: CH4 Emissions from Iron and Steel Production (kt) 4-80
Table 4-66: Material Carbon Contents for Metallurgical Coke Production 4-81
Table 4-67: Production and Consumption Data for the Calculation of CO2 Emissions from Metallurgical Coke
Production (Thousand Metric Tons) 4-82
xv
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Table 4-68: Production and Consumption Data for the Calculation of CO2 Emissions from Metallurgical Coke
Production (Million ft3) 4-82
Table 4-69: Material Carbon Contents for Iron and Steel Production 4-83
Table 4-70: Cm Emission Factors for Sinter and Pig Iron Production 4-84
Table 4-71: CO2 Emission Factors for Sinter Production, Direct Reduced Iron Production, and Pellet Production 4-84
Table 4-72: Production and Consumption Data for the Calculation of CO2 and CFU Emissions from Iron and Steel
Production (Thousand Metric Tons) 4-85
Table 4-73: Production and Consumption Data for the Calculation of CO2 Emissions from Iron and Steel Production
(Million ft3 unless otherwise specified) 4-86
Table 4-74: Approach 2 Quantitative Uncertainty Estimates for CO2 and CFU Emissions from Iron and Steel
Production and Metallurgical Coke Production (MMT CO2 Eq. and Percent) 4-87
Table 4-75: CO2 and CFU Emissions from Ferroalloy Production (MMT CO2 Eq.) 4-89
Table 4-76: CO2 and CFU Emissions from Ferroalloy Production (kt) 4-89
Table 4-77: Production of Ferroalloys (Metric Tons) 4-91
Table 4-78: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Ferroalloy Production (MMT
CO2 Eq. and Percent) 4-92
Table 4-79: CO2 Emissions from Aluminum Production (MMT CO2 Eq. and kt) 4-94
Table 4-80: PFC Emissions from Aluminum Production (MMT CO2 Eq.) 4-94
Table 4-81: PFC Emissions from Aluminum Production (kt) 4-94
Table 4-82: Summary of HVAE Emissions 4-97
Table 4-83: Summary of LVAE Emissions 4-98
Table 4-84: Production of Primary Aluminum (kt) 4-98
Table 4-85: Approach 2 Quantitative Uncertainty Estimates for CO2 and PFC Emissions from Aluminum Production
(MMT CO2 Eq. and Percent) 4-99
Table 4-86: SF6, HFC-134a, FK 5-1-12 and CO2 Emissions from Magnesium Production and Processing (MMT CO2
Eq.) 4-100
Table 4-87: SF6, HFC-134a, FK 5-1-12 and CO2 Emissions from Magnesium Production and Processing (kt) 4-100
Table 4-88: SF6 Emission Factors (kg SF6 per metric ton of magnesium) 4-103
Table 4-89: Approach 2 Quantitative Uncertainty Estimates for SF6, HFC-134a and CO2 Emissions from Magnesium
Production and Processing (MMT CO2 Eq. and Percent) 4-104
Table 4-90: CO2 Emissions from Lead Production (MMT CO2 Eq. and kt) 4-107
Table 4-91: Lead Production (Metric Tons) 4-107
Table 4-92: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Lead Production (MMT CO2 Eq.
and Percent) 4-108
Table 4-93: CO2 Emissions from Zinc Production (MMT CO2 Eq. and kt) 4-110
Table 4-94: Zinc Production (Metric Tons) 4-111
Table 4-95: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Zinc Production (MMT CO2 Eq.
and Percent) 4-114
Table 4-96: PFC, HFC, SF6, NF3, and N2O Emissions from Electronics Industry (MMT CO2 Eq.) 4-117
xvi DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Table 4-97: PFC, HFC, SF6, NF3, and N2O Emissions from Semiconductor Manufacture (Metric Tons) 4-118
Table 4-98: F-HTF Emissions from Electronics Manufacture by Compound Group (kt CO2 Eq.) 4-118
Table 4-99: Approach 2 Quantitative Uncertainty Estimates for HFC, PFC, SF6, NF3 and N2O Emissions from
Electronics Manufacture (MMT CO2 Eq. and Percent) 4-130
Table 4-100: Emissions of HFCs, PFCs, and CO2 from ODS Substitutes (MMT CO2 Eq.) 4-132
Table 4-101: Emissions of HFCs, PFCs, and CO2 from ODS Substitution (Metric Tons) 4-133
Table 4-102: Emissions of HFCs, PFCs, and CO2 from ODS Substitutes (MMT CO2 Eq.) by Sector 4-134
Table 4-103: Approach 2 Quantitative Uncertainty Estimates for HFC and PFC Emissions from ODS Substitutes
(MMT CO2 Eq. and Percent) 4-137
Table 4-104: SF6 and CF4 Emissions from Electric Power Systems and Electrical Equipment Manufacturers (MMT
CO2 Eq.) 4-142
Table 4-105: SF6 and CF4 Emissions from Electric Power Systems and Electrical Equipment Manufacturers (kt) 4-142
Table 4-106: GHGRP-only Average Emission Rate (kg per mile) 4-146
Table 4-107: Categorization of Utilities and Timeseries for Application of Corresponding Emission Estimation
Methodologies 4-146
Table 4-108: Approach 2 Quantitative Uncertainty Estimates for SF6 and CF4 Emissions from Electrical Transmission
and Distribution (MMT CO2 Eq. and Percent) 4-148
Table 4-109: N20 Production (kt) 4-153
Table 4-110: N2O Emissions from N2O Product Usage (MMT CO2 Eq. and kt) 4-153
Table 4-111: Approach 2 Quantitative Uncertainty Estimates for N2O Emissions from N2O Product Usage (MMT
CO2 Eq. and Percent) 4-155
Table 4-112: NOx, CO, NMVOC, and SO2 Emissions from Industrial Processes and Product Use (kt) 4-156
Table 5-1: Emissions from Agriculture (MMT CO2 Eq.) 5-3
Table 5-2: Emissions from Agriculture (kt) 5-3
Table 5-3: CH4 Emissions from Enteric Fermentation (MMT CO2 Eq.) 5-5
Table 5-4: CH4 Emissions from Enteric Fermentation (kt) 5-5
Table 5-5: Cattle Sub-Population Categories for 2021 Population Estimates 5-8
Table 5-6: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from Enteric Fermentation (MMT CO2
Eq. and Percent) 5-10
Table 5-7: CH4 and N2O Emissions from Manure Management (MMT CO2 Eq.) 5-13
Table 5-8: CH4 and N2O Emissions from Manure Management (kt) 5-14
Table 5-9: Approach 2 Quantitative Uncertainty Estimates for CH4 and N2O (Direct and Indirect) Emissions from
Manure Management (MMT CO2 Eq. and Percent) 5-19
Table 5-10: IPCC (2006) Implied Emission Factor Default Values Compared with Calculated Values for CH4 from
Manure Management (kg/head/year) 5-20
Table 5-11
Table 5-12
Table 5-13
CH4 Emissions from Rice Cultivation (MMT CO2 Eq.) 5-22
CH4 Emissions from Rice Cultivation (kt) 5-23
Rice Area Harvested (1,000 Hectares) 5-25
xvii
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Table 5-14: Average Ratooned Area as Percent of Primary Growth Area (Percent) 5-26
Table 5-15: Approach 2 Quantitative Uncertainty Estimates for Cm Emissions from Rice Cultivation (MMT CO2 Eq.
and Percent) 5-27
Table 5-16: N2O Emissions from Agricultural Soils (MMT CO2 Eq.) 5-31
Table 5-17: N2O Emissions from Agricultural Soils (kt) 5-31
Table 5-18: Direct N2O Emissions from Agricultural Soils by Land Use Type and N Input Type (MMT CO2 Eq.).... 5-31
Table 5-19: Indirect N2O Emissions from Agricultural Soils (MMT CO2 Eq.) 5-32
Table 5-20: Quantitative Uncertainty Estimates of N2O Emissions from Agricultural Soil Management in 2021
(MMT CO2 Eq. and Percent) 5-46
Table 5-21: Emissions from Liming (MMT CO2 Eq.) 5-48
Table 5-22: Emissions from Liming (MMT C) 5-48
Table 5-23: Applied Minerals (MMT) 5-50
Table 5-24: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Liming (MMT CO2 Eq. and
Percent) 5-50
Table 5-25: CO2 Emissions from Urea Fertilization (MMT CO2 Eq.) 5-51
Table 5-26: CO2 Emissions from Urea Fertilization (MMT C) 5-51
Table 5-27: Applied Urea (MMT) 5-52
Table 5-28: Quantitative Uncertainty Estimates for CO2 Emissions from Urea Fertilization (MMT CO2 Eq. and
Percent) 5-52
Table 5-29: CFU and N2O Emissions from Field Burning of Agricultural Residues (MMT CO2 Eq.) 5-54
Table 5-30: CFU, N2O, CO, and NOx Emissions from Field Burning of Agricultural Residues (kt) 5-55
Table 5-31: Agricultural Crop Production (kt of Product) 5-58
Table 5-32: U.S. Average Percent Crop Area Burned by Crop (Percent) 5-59
Table 5-33: Parameters for Estimating Emissions from Field Burning of Agricultural Residues 5-59
Table 5-34: Greenhouse Gas Emission Ratios and Conversion Factors 5-60
Table 5-35: Approach 2 Quantitative Uncertainty Estimates for CFU and N2O Emissions from Field Burning of
Agricultural Residues (MMT CO2 Eq. and Percent) 5-61
Table 6-1: Emissions and Removals (Net Flux) from Land Use, Land-Use Change, and Forestry (MMT CO2 Eq.).... 6-4
Table 6-2: Emissions and Removals from Land Use, Land-Use Change, and Forestry by Gas (MMT CO2 Eq.) 6-6
Table 6-3: Emissions and Removals from Land Use, Land-Use Change, and Forestry by Gas (kt) 6-7
Table 6-4: Managed and Unmanaged Land Area by Land-Use Categories for All 50 States (Thousands of Hectares)
6-11
Table 6-5: Land Use and Land-Use Change for the U.S. Managed Land Base for All 50 States (Thousands of
Hectares) 6-11
Table 6-6: Data Sources Used to Determine Land Use and Land Area for the Conterminous United States, Hawaii,
and Alaska 6-17
Table 6-7: Total Land Area (Hectares) by Land Use Category for U.S. Territories 6-24
xviii DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Table 6-8: Net CO2 Flux from Forest Ecosystem Pools in Forest Land Remaining Forest Land and Harvested Wood
Pools (MMT CO2 Eq.) 6-28
Table 6-9: Net C Flux from Forest Ecosystem Pools in Forest Land Remaining Forest Land and Harvested Wood
Pools (MMT C) 6-29
Table 6-10: Forest Area (1,000 ha) and C Stocks in Forest Land Remaining Forest Land and Harvested Wood Pools
(MMT C) 6-30
Table 6-11: Estimates of CO2 (MMT per Year) Emissions3 from Forest Fires in the Conterminous 48 States and
Alaska 6-32
Table 6-12: Quantitative Uncertainty Estimates for Net CO2 Flux from Forest Land Remaining Forest Land: Changes
in Forest C Stocks (MMT CO2 Eq. and Percent) 6-36
Table 6-13: Recalculations of Forest Area (1,000 ha) and C Stocks in Forest Land Remaining Forest Land and
Harvested Wood Pools (MMT C) 6-38
Table 6-14: Recalculations of Net C Flux from Forest Ecosystem Pools in Forest Land Remaining Forest Land and
Harvested Wood Pools (MMT C) 6-38
Table 6-15: Non-CC>2 Emissions from Forest Fires (MMT CO2 Eq.)a 6-40
Table 6-16: Non-CC>2 Emissions from Forest Fires (kt)a 6-40
Table 6-17: Quantitative Uncertainty Estimates of Non-C02 Emissions from Forest Fires (MMT CO2 Eq. and
Percent)3 6-41
Table 6-18: N2O Fluxes from Soils in Forest Land Remaining Forest Land and Land Converted to Forest Land (MMT
CO2 Eq. and kt N20) 6-42
Table 6-19: Quantitative Uncertainty Estimates of N2O Fluxes from Soils in Forest Land Remaining Forest Land and
Land Converted to Forest Land (MMT CO2 Eq. and Percent) 6-43
Table 6-20: Non-C02 Emissions from Drained Organic Forest Soilsa b (MMT CO2 Eq.) 6-45
Table 6-21: Non-C02 Emissions from Drained Organic Forest Soilsa b (kt) 6-45
Table 6-22: States identified as having Drained Organic Soils, Area of Forest on Drained Organic Soils, and
Sampling Error 6-46
Table 6-23: Quantitative Uncertainty Estimates for Non-C02 Emissions on Drained Organic Forest Soils (MMT CO2
Eq. and Percent)3 6-47
Table 6-24: Net CO2 Flux from Forest C Pools in Land Converted to Forest Land by Land Use Change Category
(MMT CO2 Eq.) 6-49
Table 6-25: Net C Flux from Forest C Pools in Land Converted to Forest Land by Land Use Change Category (MMT
C) 6-50
Table 6-26: Quantitative Uncertainty Estimates for Forest C Pool Stock Changes (MMT CO2 Eq. per Year) in 2021
from Land Converted to Forest Land by Land Use Change 6-53
Table 6-27: Recalculations of the Net C Flux from Forest C Pools in Land Converted to Forest Land by Land Use
Change Category (MMT C) 6-55
Table 6-28: Net CO2 Flux from Soil C Stock Changes in Cropland Remaining Cropland (MMT CO2 Eq.) 6-57
Table 6-29: Net CO2 Flux from Soil C Stock Changes in Cropland Remaining Cropland (MMT C) 6-57
Table 6-30: Approach 2 Quantitative Uncertainty Estimates for Soil C Stock Changes occurring within Cropland
Remaining Cropland (MMT CO2 Eq. and Percent) 6-65
xix
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Table 6-31: Comparison of Managed Land Area in Cropland Remaining Cropland and Area in the Current Cropland
Remaining Cropland Inventory (Thousand Hectares) 6-66
Table 6-32: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes in Land Converted to
Cropland by Land Use Change Category (MMT CO2 Eq.) 6-68
Table 6-33: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes in Land Converted to
Cropland (MMT C) 6-69
Table 6-34: Approach 2 Quantitative Uncertainty Estimates for Soil, Dead Organic Matter and Biomass C Stock
Changes occurring within Land Converted to Cropland (MMT CO2 Eq. and Percent) 6-72
Table 6-35: Comparison of Managed Land Area in Land Converted to Cropland and the Area in the current Land
Converted to Cropland Inventory (Thousand Hectares) 6-74
Table 6-36: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes in Grassland Remaining
Grassland (MMT C02 Eq.) 6-76
Table 6-37: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes in Grassland Remaining
Grassland (MMT C) 6-76
Table 6-38: Approach 2 Quantitative Uncertainty Estimates for C Stock Changes Occurring Within Grassland
Remaining Grassland (MMT CO2 Eq. and Percent) 6-81
Table 6-39: Comparison of Managed Land Area in Grassland Remaining Grassland and the Area in the current
Grassland Remaining Grassland Inventory (Thousand Hectares) 6-83
Table 6-40: CH4 and N2O Emissions from Biomass Burning in Grassland (MMT CO2 Eq.) 6-84
Table 6-41: CH4, N2O, CO, and NOx Emissions from Biomass Burning in Grassland (kt) 6-84
Table 6-42: Thousands of Grassland Hectares Burned Annually 6-85
Table 6-43: Uncertainty Estimates for Non-C02 Greenhouse Gas Emissions from Biomass Burning in Grassland
(MMT CO2 Eq. and Percent) 6-86
Table 6-44: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for Land Converted to
Grassland (MMT C02 Eq.) 6-88
Table 6-45: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for Land Converted to
Grassland (MMT C) 6-88
Table 6-46: Approach 2 Quantitative Uncertainty Estimates for Soil, Dead Organic Matter and Biomass C Stock
Changes occurring within Land Converted to Grassland (MMT CO2 Eq. and Percent) 6-92
Table 6-47: Comparison of Managed Land Area in Land Converted to Grassland and Area in the current Land
Converted to Grassland Inventory (Thousand Hectares) 6-94
Table 6-48: Emissions from Peatlands Remaining Peatlands (MMT CO2 Eq.) 6-97
Table 6-49: Emissions from Peatlands Remaining Peatlands (kt) 6-97
Table 6-50: Peat Production of Conterminous 48 States (kt) 6-98
Table 6-51: Peat Production of Alaska (Thousand Cubic Meters) 6-98
Table 6-52: Peat Production Area of Conterminous 48 States (Hectares) 6-99
Table 6-53: Peat Production Area of Alaska (Hectares) 6-99
Table 6-54: Peat Production (Hectares) 6-99
Table 6-55: Approach 2 Quantitative Uncertainty Estimates for CO2, CH4, and N2O Emissions from Peatlands
Remaining Peatlands (MMT CO2 Eq. and Percent) 6-101
xx DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Table 6-56: Emissions and Removals from Coastal Wetlands Remaining Coastal Wetlands (MMT CO2 Eq.) 6-104
Table 6-57: Net CO2 Flux from C Stock Changes in Vegetated Coastal Wetlands Remaining Vegetated Coastal
Wetlands (MMT C02 Eq.) 6-105
Table 6-58: Net CO2 Flux from C Stock Changes in Vegetated Coastal Wetlands Remaining Vegetated Coastal
Wetlands (MMT C) 6-105
Table 6-59: CFU Emissions from Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands (MMT CO2
Eq. and kt CH4) 6-105
Table 6-60: Area of Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands, Vegetated Coastal
Wetlands Converted to Unvegetated Open Water Coastal Wetlands, and Unvegetated Open Water Coastal
Wetlands Converted to Vegetated Coastal Wetlands (ha) 6-106
Aboveground Biomass Carbon Stocks for Vegetated Coastal Wetlands (t C ha_1) 6-106
Root to Shoot Ratios for Vegetated Coastal Wetlands 6-107
Annual Soil Carbon Accumulation Rates for Vegetated Coastal Wetlands (t C ha 1 yr 6-107
Table 6-61
Table 6-62
Table 6-63
Table 6-64: IPCC Approach 1 Quantitative Uncertainty Estimates for C Stock Changes and CFU Emissions occurring
within Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands in 2021 (MMT CO2 Eq. and Percent).. 6-
108
Table 6-65: Net CO2 Flux from C Stock Changes in Vegetated Coastal Wetlands Converted to Unvegetated Open
Water Coastal Wetlands (MMT C02 Eq.) 6-110
Table 6-66: Net CO2 Flux from C Stock Changes in Vegetated Coastal Wetlands Converted to Unvegetated Open
Water Coastal Wetlands (MMT C) 6-111
Table 6-67: Approach 1 Quantitative Uncertainty Estimates for CO2 Flux Occurring within Vegetated Coastal
Wetlands Converted to Unvegetated Open Water Coastal Wetlands in 2020 (MMT CO2 Eq. and Percent) 6-113
Table 6-68: CO2 Flux from C Stock Changes from Unvegetated Open Water Coastal Wetlands Converted to
Vegetated Coastal Wetlands (MMT CO2 Eq.) 6-114
Table 6-69: CO2 Flux from C Stock Changes from Unvegetated Open Water Coastal Wetlands Converted to
Vegetated Coastal Wetlands (MMT C) 6-115
Table 6-70: Approach 1 Quantitative Uncertainty Estimates for C Stock Changes Occurring within Unvegetated
Open Water Coastal Wetlands Converted to Vegetated Coastal Wetlands in 2021 (MMT CO2 Eq. and Percent) 6-117
Table 6-71: N2O Emissions from Aquaculture in Coastal Wetlands (MMT CO2 Eq. and kt N2O) 6-118
Table 6-72: Approach 1 Quantitative Uncertainty Estimates for N2O Emissions from Aquaculture Production in
Coastal Wetlands in 2021 (MMT CO2 Eq. and Percent) 6-119
Table 6-73: CFU Emissions from Flooded Land Remaining Flooded Land—Reservoirs (MMT CO2 Eq.) 6-121
Table 6-74: CFU Emissions from Flooded Land Remaining Flooded Land —Reservoirs (kt CH4) 6-121
Table 6-75: Surface and Downstream CH4 Emissions from Reservoirs in Flooded Land Remaining Flooded Land in
2021 (kt CH4) 6-122
Table 6-76: IPCC (2019) Default CFU Emission Factors for Surface Emission from Reservoirs in Flooded Land
Remaining Flooded Land 6-124
Table 6-77: National Totals of Reservoir Surface Area in Flooded Land Remaining Flooded Land (millions of ha)... 6-
125
Table 6-78: State Breakdown of Reservoir Surface Area in Flooded Land Remaining Flooded Land (millions of ha) 6-
125
xx i
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Table 6-79: Approach 2 Quantitative Uncertainty Estimates for CFU Emissions from Reservoirs in Flooded Land
Remaining Flooded Land 6-127
Table 6-80: CFU Emissions from Other Constructed Waterbodies in Flooded Land Remaining Flooded Land (MMT
C02 Eq.) 6-128
Table 6-81: CFU Emissions from Other Constructed Waterbodies in Flooded Land Remaining Flooded Land (kt CH4)
6-129
Table 6-82: CFU Emissions from Other Constructed Waterbodies in Flooded Land Remaining Flooded Land in 2021
(kt CH4) 6-129
Table 6-83: IPCC (2019) Default CH4 Emission Factors for Surface Emissions from Other Constructed Waterbodies in
Flooded Land Remaining Flooded Land 6-131
Table 6-84: Predictors used in Decision Tree to Identify Canal/Ditches 6-132
Table 6-85: Validation Results for Ditch/Canal Classification Decision Tree 6-132
Table 6-86: National Surface Area Totals in Flooded Land Remaining Flooded Land - Other Constructed
Waterbodies (ha) 6-133
Table 6-87: State Totals of Surface Area in Flooded Land Remaining Flooded Land— Canals and Ditches (ha). 6-134
Table 6-88: State Totals of Surface Area in Flooded Land Remaining Flooded Land— Freshwater Ponds (ha)... 6-135
Table 6-89: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from Other Constructed
Waterbodies in Flooded Land Remaining Flooded Land 6-136
Table 6-90: Net CO2 Flux from C Stock Changes in Land Converted to Vegetated Coastal Wetlands (MMT CO2 Eq.)6-
139
Table 6-91: Net CO2 Flux from C Stock Changes in Land Converted to Vegetated Coastal Wetlands (MMT C)... 6-139
Table 6-92: CH4 Emissions from Land Converted to Vegetated Coastal Wetlands (MMT CO2 Eq. and kt CH4).... 6-140
Table 6-93: Approach 1 Quantitative Uncertainty Estimates for C Stock Changes occurring within Land Converted
to Vegetated Coastal Wetlands in 2021 (MMT CO2 Eq. and Percent) 6-142
Table 6-94: CH4 Emissions from Land Converted to Flooded Land - Reservoirs (MMT CO2 Eq.) 6-145
Table 6-95: CH4 Emissions from Land Converted to Flooded Land—Reservoirs (kt CH4) 6-145
Table 6-96: CO2 Emissions from Land Converted to Flooded Land —Reservoirs (MMT CO2) 6-146
Table 6-97: CO2 Emissions from Land Converted to Flooded Land —Reservoirs (MMT C) 6-146
Table 6-98: Methane and CO2 Emissions from Reservoirs in Land Converted to Flooded Land in 2021 (kt CH4; kt
CO2) 6-146
Table 6-99: IPCC (2019) Default CH4 and CO2 Emission Factors for Surface Emissions from Reservoirs in Land
Converted to Flooded Land 6-148
Table 6-100: National Totals of Reservoir Surface Area in Land Converted to Flooded Land (thousands of ha).. 6-149
Table 6-101: State Breakdown of Reservoir Surface Area in Land Converted to Flooded Land (thousands of ha)
6-150
Table 6-102: Approach 2 Quantitative Uncertainty Estimates for CH4 and CO2 Emissions from Reservoirs in Land
Converted to Flooded Land 6-151
Table 6-103: CH4 Emissions from Other Constructed Waterbodies in Land Converted to Flooded Land (MMT CO2
Eq.) 6-153
Table 6-104: CH4 Emissions from Other Constructed Waterbodies in Land Converted to Flooded Land (kt CH4) 6-153
xxii DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Table 6-105: CO2 Emissions from Other Constructed Waterbodies in Land Converted to Flooded Land (MMT CO2)..
6-153
Table 6-106: CO2 Emissions from Other Constructed Waterbodies in Land Converted to Flooded Land (MMT C).... 6-
154
Table 6-107: CFU and CO2 Emissions from Other Constructed Waterbodies in Land Converted to Flooded Land in
2021 (MT CO2 Eq.) 6-154
Table 6-108: IPCC Default Methane and CO2 Emission Factors for Other Constructed Waterbodies in Land
Converted to Flooded Land 6-155
Table 6-109: National Surface Area Totals of Other Constructed Waterbodies in Land Converted to Flooded Land
(ha) 6-156
Table 6-110: State Surface Area Totals of Other Constructed Waterbodies in Land Converted to Flooded Land (ha)
6-157
Table 6-111: Approach 2 Quantitative Uncertainty Estimates for CH4 and CO2 Emissions from Other Constructed
Waterbodies in Land Converted to Flooded Land 6-158
Table 6-112: Net CO2 Flux from Soil C Stock Changes in Settlements Remaining Settlements (MMT CO2 Eq.)... 6-160
Table 6-113: Net CO2 Flux from Soil C Stock Changes in Settlements Remaining Settlements (MMT C) 6-161
Table 6-114: Thousands of Hectares of Drained Organic Soils in Settlements Remaining Settlements 6-161
Table 6-115: Uncertainty Estimates for CO2 Emissions from Drained Organic Soils in Settlements Remaining
Settlements (MMT CO2 Eq. and Percent) 6-162
Table 6-116: Area of Managed Land in Settlements Remaining Settlements that is not included in the current
Inventory (Thousand Hectares) 6-162
Table 6-117: Net Flux from Trees in Settlements Remaining Settlements (MMT CO2 Eq. and MMT C)a 6-164
Table 6-118: Carbon Storage (kg C/m2 tree cover), Gross and Net Sequestration (kg C/m2 tree cover/year) and Tree
Cover (percent) among Sampled U.S. Cities (see Nowak et al. 2013) 6-166
Table 6-119: Estimated Annual C Sequestration, Tree Cover, and Annual C Sequestration per Area of Tree Cover
for settlement areas in the United States by State and the District of Columbia (2021) 6-168
Table 6-120: Approach 2 Quantitative Uncertainty Estimates for Net CO2 Flux from Changes in C Stocks in
Settlement Trees (MMT CO2 Eq. and Percent) 6-170
Table 6-121: Recalculations of the Settlement Tree Categories 6-170
Table 6-122: N2O Emissions from Soils in Settlements Remaining Settlements (MMT CO2 Eq. and kt N2O) 6-171
Table 6-123: Quantitative Uncertainty Estimates of N2O Emissions from Soils in Settlements Remaining
Settlements (MMT CO2 Eq. and Percent) 6-173
Table 6-124: Net Changes in Yard Trimmings and Food Scrap Carbon Stocks in Landfills (MMT CO2 Eq.) 6-175
Table 6-125: Net Changes in Yard Trimmings and Food Scrap Carbon Stocks in Landfills (MMT C) 6-175
Table 6-126: Moisture Contents, C Storage Factors (Proportions of Initial C Sequestered), Initial C Contents, and
Decay Rates for Yard Trimmings and Food Scraps in Landfills 6-178
Table 6-127: C Stocks in Yard Trimmings and Food Scraps in Landfills (MMT C) 6-179
Table 6-128: Approach 2 Quantitative Uncertainty Estimates for CO2 Flux from Yard Trimmings and Food Scraps in
Landfills (MMT CO2 Eq. and Percent) 6-179
Table 6-129: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for Land Converted to
Settlements (MMT C02 Eq.) 6-182
xxiii
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Table 6-130: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for Land Converted to
Settlements (MMTC) 6-182
Table 6-131: Approach 2 Quantitative Uncertainty Estimates for Soil, Dead Organic Matter and Biomass C Stock
Changes occurring within Land Converted to Settlements (MMT CO2 Eq. and Percent) 6-185
Table 6-132: Area of Managed Land in Land Converted to Settlements that is not included in the current Inventory
(Thousand Hectares) 6-187
Table 7-1: Emissions from Waste (MMT CO2 Eq.) 7-2
Table 7-2: Emissions from Waste (kt) 7-2
Table 7-3: CFU Emissions from Landfills (MMT CO2 Eq.) 7-7
Table 7-4: CFU Emissions from Landfills (kt) 7-7
Table 7-5: Approach 2 Quantitative Uncertainty Estimates for CFU Emissions from Landfills (MMT CO2 Eq. and
Percent) 7-15
Table 7-6: Materials Discarded in the Municipal Waste Stream by Waste Type from 1990 to 2018 (Percent) 7-20
Table 7-7: CFU and N2O Emissions from Domestic and Industrial Wastewater Treatment (MMT CO2 Eq.) 7-23
Table 7-8: CFU and N2O Emissions from Domestic and Industrial Wastewater Treatment (kt) 7-23
Table 7-9: Domestic Wastewater CFU Emissions from Septic and Centralized Systems (2021, kt, MMT CO2 Eq. and
Percent) 7-25
Table 7-10: Variables and Data Sources for CFU Emissions from Septic Systems 7-25
Table 7-11: Variables and Data Sources for Organics in Domestic Wastewater 7-26
Table 7-12: U.S. Population (Millions) and Domestic Wastewater TOW (kt) 7-27
Table 7-13: Variables and Data Sources for Organics in Centralized Domestic Wastewater 7-28
Table 7-14: Variables and Data Sources for CFU Emissions from Centrally Treated Aerobic Systems (Other than
Constructed Wetlands) 7-28
Table 7-15: Variables and Data Sources for CFU Emissions from Centrally Treated Aerobic Systems (Constructed
Wetlands) 7-30
Table 7-16: Variables and Data Sources for CFU Emissions from Centrally Treated Anaerobic Systems 7-31
Table 7-17: Variables and Data Sources for Emissions from Anaerobic Sludge Digesters 7-31
Table 7-18: Variables and Data Sources for CH4 Emissions from Centrally Treated Systems Discharge 7-32
Table 7-19: Total Industrial Wastewater CH4 Emissions by Sector (2021, MMT CO2 Eq. and Percent) 7-34
Table 7-20: U.S. Pulp and Paper, Meat, Poultry, Vegetables, Fruits and Juices, Ethanol, Breweries, and Petroleum
Refining Production (MMT) 7-36
U.S. Industrial Wastewater Characteristics Data (2021) 7-36
U.S. Industrial Wastewater Treatment Activity Data 7-37
Sludge Variables for Aerobic Treatment Systems 7-37
Fraction of TOW Removed During Treatment by Industry 7-38
Wastewater Outflow (m3/ton) for Pulp, Paper, and Paperboard Mills 7-39
Wastewater Outflow (m3/ton) and BOD Production (g/L) for U.S. Vegetables, Fruits, and Juices
Production 7-40
Table 7-21
Table 7-22
Table 7-23
Table 7-24
Table 7-25
Table 7-26
xxiv DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Table 7-27: Domestic Wastewater N2O Emissions from Septic and Centralized Systems (2021, kt, MMT CO2 Eq. and
Percent) 7-42
Table 7-28: Variables and Data Sources for Protein Consumed 7-43
Table 7-29: Variables and Data Sources for N2O Emissions from Septic System 7-44
Table 7-30: Variables and Data Sources for Non-Consumed Protein and Nitrogen Entering Centralized Systems 7-45
Table 7-31: Variables and Data Sources for N2O Emissions from Centrally Treated Aerobic Systems (Other than
Constructed Wetlands) 7-46
Table 7-32: Variables and Data Sources for N2O Emissions from Centrally Treated Aerobic Systems (Constructed
Wetlands) 7-47
Table 7-33: Variables and Data Sources for N2O Emissions from Centrally Treated Anaerobic Systems 7-48
Table 7-34: U.S. Population (Millions) Fraction of Population Served by Centralized Wastewater Treatment
(percent), Protein Supply (kg/person-year), and Protein Consumed (kg/person-year) 7-48
Table 7-35: Variables and Data Sources for N2O Emissions from Centrally Treated Systems Discharge 7-49
Table 7-36: Total Industrial Wastewater N2O Emissions by Sector (2021, MMT CO2 Eq. and Percent) 7-50
Table 7-37: U.S. Industrial Wastewater Nitrogen Data 7-51
Table 7-38: Industrial Wastewater Nitrogen Discharged in 2018 by Sector (kg N) 7-52
Table 7-39: Approach 2 Quantitative Uncertainty Estimates for 2021 Emissions from Wastewater Treatment (MMT
CO2 Eq. and Percent) 7-53
Table 7-40: CFU and N2O Emissions from Composting (MMT CO2 Eq.) 7-57
Table 7-41: CFU and N2O Emissions from Composting (kt) 7-57
Table 7-42: U.S. Waste Composted (kt) 7-58
Table 7-43: Tier 1 Quantitative Uncertainty Estimates for Emissions from Composting (MMT CO2 Eq. and Percent)
7-58
Table 7-44: CFU Emissions from Anaerobic Digestion at Biogas Facilities (MMT CO2 Eq.) from 1990-2021 7-61
Table 7-45: CFU Emissions from Anaerobic Digestion at Biogas Facilities (kt) from 1990-2021 7-61
Table 7-46: U.S. Waste Digested (kt) from 1990-2021 7-63
Table 7-47: Estimated Number of Stand-Alone AD Facilities Operating3 from 1990-2021 7-64
Table 7-48: Estimated Biogas Produced and Methane Recovered from Anaerobic Digestion at Biogas Facilities
Operating from 1990-20213 7-64
Table 7-49: Approach 1 Quantitative Uncertainty Estimates for Emissions from Anaerobic Digestion (MMT CO2 Eq.
and Percent) 7-65
Table 7-50: Emissions of NOx, CO, NMVOC, and SO2 from Waste (kt) 7-67
Table 9-1: Revisions to the U.S. Greenhouse Gas Emissions, Including Quantitative Change Related to the Use of
AR5 GWP Values (MMT CO2 Eq.) 9-5
Table 9-2: Revisions to U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use, Land-Use Change,
and Forestry, Including Quantitative Change Related to the Use of AR5 GWP Values (MMT CO2 Eq.) 9-8
Table 9-3: Revisions to U.S. Greenhouse Gas Emissions, Excluding Quantitative Change Related to the Use of AR5
GWP Values (MMT CO2 Eq.) 9-9
XXV
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Table 9-4: Revisions to U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use, Land-Use Change,
and Forestry, Excluding Quantitative Change Related to the Use of AR5 GWP Values (MMT CO2 Eq.) 9-11
Figures
Figure ES-1: U.S. Greenhouse Gas Emissions and Sinks by Gas ES-5
Figure ES-2: Annual Percent Change in Gross U.S. Greenhouse Gas Emissions and Sinks Relative to the Previous
Year ES-6
Figure ES-3: 2021 Total Gross U.S. Greenhouse Gas Emissions by Gas (Percentages based on MMT CO2 Eq.) ES-7
Figure ES-4: 2021 Sources of CO2 Emissions ES-8
Figure ES-5: 2021 CO2 Emissions from Fossil Fuel Combustion by Sector and Fuel Type ES-9
Figure ES-6: 2021 End-Use Sector Emissions of CO2 from Fossil Fuel Combustion ES-10
Figure ES-7: Electric Power Generation and Emissions ES-12
Figure ES-8: 2021 Sources of CFU Emissions ES-13
Figure ES-9: 2021 Sources of N2O Emissions ES-14
Figure ES-10: 2021 Sources of HFCs, PFCs, SF6, and NF3 Emissions ES-15
Figure ES-11: U.S. Greenhouse Gas Emissions and Sinks by IPCC Sector/Category ES-16
Figure ES-12: 2021 U.S. Energy Consumption by Energy Source (Percent) ES-17
Figure ES-13: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors ES-21
Figure ES-14: U.S. Greenhouse Gas Emissions with Electricity-Related Emissions Distributed to Economic Sectors
ES-23
Figure ES-15: U.S. Greenhouse Gas Emissions Per Capita and Per Dollar of Gross Domestic Product (GDP) ES-24
Figure ES-16: 2021 Key Categories (Approach 1 including LULUCF)3 ES-25
Figure 1-1: National Inventory Arrangements and Process Diagram 1-12
Figure 1-2: U.S. QA/QC Plan Summary 1-26
Figure 2-1: U.S. Greenhouse Gas Emissions and Sinks by Gas 2-2
Figure 2-2: Annual Percent Change in Gross U.S. Greenhouse Gas Emissions Relative to the Previous Year 2-2
Figure 2-3: 2021 Gross Total U.S. Greenhouse Gas Emissions by Gas (Percentages based on MMT CO2 Eq.) 2-3
Figure 2-4: U.S. Greenhouse Gas Emissions and Sinks by IPCC Sector 2-9
Figure 2-5: Trends in Energy Sector Greenhouse Gas Sources 2-11
Figure 2-6: Trends in CO2 Emissions from Fossil Fuel Combustion by End-Use Sector and Fuel Type 2-15
Figure 2-7: Trends in End-Use Sector Emissions of CO2 from Fossil Fuel Combustion 2-16
Figure 2-8: Electric Power Generation (Billion kWh) and Emissions (MMT CO2 Eq.) 2-17
Figure 2-9: Trends in Industrial Processes and Product Use Sector Greenhouse Gas Sources 2-19
Figure 2-10: Trends in Agriculture Sector Greenhouse Gas Sources 2-21
Figure 2-11: Trends in Emissions and Removals (Net CO2 Flux) from Land Use, Land-Use Change, and Forestry.. 2-24
Figure 2-12: Trends in Waste Sector Greenhouse Gas Sources 2-27
Figure 2-13: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors 2-28
xxvi DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Figure 2-14: U.S. Greenhouse Gas Emissions with Electricity-Related Emissions Distributed to Economic Sectors.
. 2-34
Figure 2-15: Trends in Transportation-Related Greenhouse Gas Emissions 2-37
Figure 2-16: U.S. Greenhouse Gas Emissions Per Capita and Per Dollar of Gross Domestic Product 2-40
Figure 3-1: 2021 Energy Sector Greenhouse Gas Sources 3-2
Figure 3-2: Trends in Energy Sector Greenhouse Gas Sources 3-2
Figure 3-3: 2021 U.S. Fossil Carbon Flows 3-2
Figure 3-4: 2021 U.S. Energy Use by Energy Source 3-10
Figure 3-5: Annual U.S. Energy Use 3-11
Figure 3-6: 2021CO2 Emissions from Fossil Fuel Combustion by Sector and Fuel Type 3-11
Figure 3-7: Annual Deviations from Normal Heating Degree Days for the United States (1950-2021, Index Normal =
100) 3-12
Figure 3-8: Annual Deviations from Normal Cooling Degree Days for the United States (1950-2021, Index Normal =
100) 3-12
Figure 3-9: Fuels Used in Electric Power Generation and Total Electric Power Sector CO2 Emissions 3-18
Figure 3-10: Electric Power Retail Sales by End-Use Sector 3-19
Figure 3-11: Industrial Production Indices (Index 2017=100) 3-20
Figure 3-12: Fuels and Electricity Used in Industrial Sector, Industrial Output, and Total Sector CO2 Emissions
(Including Electricity) 3-21
Figure 3-13: Fuels and Electricity Used in Residential and Commercial Sectors, Heating and Cooling Degree Days,
and Total Sector CO2 Emissions (Including Electricity) 3-22
Figure 3-14: Fuels Used in Transportation Sector, On-road VMT, and Total Sector CO2 Emissions 3-25
Figure 3-15: Sales-Weighted Fuel Economy of New Passenger Cars and Light-Duty Trucks, 1990-2021 3-27
Figure 3-16: Sales of New Passenger Cars and Light-Duty Trucks, 1990-2021 3-27
Figure 3-17: Mobile Source CH4 and N2O Emissions 3-30
Figure 3-18: U.S. Energy Consumption and Energy-Related CO2 Emissions Per Capita and Per Dollar GDP 3-37
Figure 4-1: 2021 Industrial Processes and Product Use Sector Greenhouse Gas Sources 4-2
Figure 4-2: Trends in Industrial Processes and Product Use Sector Greenhouse Gas Sources 4-3
Figure 4-3: U.S. Emissions of SF6 Comparison3 4-150
Figure 5-1: 2021 Agriculture Sector Greenhouse Gas Emission Sources 5-1
Figure 5-2: Trends in Agriculture Sector Greenhouse Gas Emission Sources 5-2
Figure 5-3: Annual CH4 Emissions from Rice Cultivation, 2015 5-24
Figure 5-4: Sources and Pathways of N that Result in N2O Emissions from Agricultural Soil Management 5-30
Figure 5-5: Croplands, 2020 Annual Direct N2O Emissions Estimated Using the Tier 3 DayCent Model 5-33
Figure 5-6: Grasslands, 2020 Annual Direct N2O Emissions Estimated Using the Tier 3 DayCent Model 5-33
Figure 5-7: Croplands, 2020 Annual Indirect N2O Emissions from Volatilization Using the Tier 3 DayCent Model 5-34
xxvii
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Figure 5-8: Grasslands, 2020 Annual Indirect N2O Emissions from Volatilization Using the Tier 3 DayCent Model
5-35
Figure 5-9: Croplands, 2020 Annual Indirect N2O Emissions from Leaching and Runoff Using the Tier 3 DayCent
Model 5-35
Figure 5-10: Grasslands, 2020 Annual Indirect N2O Emissions from Leaching and Runoff Using the Tier 3 DayCent
Model 5-36
Figure 6-1: 2021 LULUCF Chapter Greenhouse Gas Sources and Sinks 6-3
Figure 6-2: Trends in Emissions and Removals (Net CO2 Flux) from Land Use, Land-Use Change, and Forestry 6-3
Figure 6-3: Percent of Total Land Area for Each State in the General Land Use Categories for 2021 6-13
Figure 6-4: Changes in Forest Area by Region for Forest Land Remaining Forest Land in the conterminous United
States and Alaska (1990-2021) 6-27
Figure 6-5: Estimated Net Annual Changes in C Stocks for All C Pools in Forest Land Remaining Forest Land in the
Conterminous United States and Alaska (1990-2021) 6-31
Figure 6-6: Total Net Annual Soil C Stock Changes for Mineral Soils under Agricultural Management within States,
2015, Cropland Remaining Cropland 6-58
Figure 6-7: Total Net Annual Soil C Stock Changes for Organic Soils under Agricultural Management within States,
2015, Cropland Remaining Cropland 6-59
Figure 6-8: Total Net Annual Soil C Stock Changes for Mineral Soils under Agricultural Management within States,
2015, Grassland Remaining Grassland 6-77
Figure 6-9: Total Net Annual Soil C Stock Changes for Organic Soils under Agricultural Management within States,
2015, Grassland Remaining Grassland 6-78
Figure 6-10: U.S. Reservoirs (black polygons) in the Flooded Land Remaining Flooded Land Category in 2021.. 6-121
Figure 6-11: Total CH4 Emissions (Downstream + Surface) from Reservoirs in Flooded Land Remaining Flooded Land
in 2021 (kt CH4) 6-122
Figure 6-12: Selected Features from NWI that Meet Flooded Lands Criteria 6-125
Figure 6-13: 2021CFU Emissions from A) Ditches and Canals and B) Freshwater Ponds in Flooded Land Remaining
Flooded Land (kt CH4) 6-130
Figure 6-14: Left: NWI Features Identified as Canals/Ditches (pink) by Unique Narrow, Linear/Angular Morphology.
Right: Non-Canal/Ditches with More Natural Morphology (blue) 6-132
Figure 6-15: Structure of Decision Tree Used to Identify Canals/Ditches 6-133
Figure 6-16: 2021 Surface Area of A) Ditches and Canals and B) Freshwater Ponds in Flooded Land Remaining
Flooded Land (hectares) 6-134
Figure 6-17: U.S. Reservoirs (black polygons) in the Land Converted to Flooded Land Category in 2021 6-145
Figure 6-18: 2021 A) CFU and B) CO2 Emissions from U.S. Reservoirs in Land Converted to Flooded Land 6-146
Figure 6-19: Selected Features from NWI that meet Flooded Lands Criteria 6-149
Figure 6-20: Number of Dams Built per Year from 1990 through 2021 6-150
Figure 6-21: 2021 A) CFU and B) CO2 Emissions from Other Constructed Waterbodies (Freshwater Ponds) in Land
Converted to Flooded Land (MT CO2 Eq.) 6-155
Figure 6-22: Surface Area of Other Constructed Waterbodies in Land Converted to Flooded Land (ha) 6-157
Figure 7-1: 2021 Waste Sector Greenhouse Gas Sources 7-1
xxviii DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Figure 7-2: Trends in Waste Sector Greenhouse Gas Sources 7-2
Figure 7-3: Methodologies Used Across the Time Series to Compile the U.S. Inventory of Emission Estimates for
MSW Landfills 7-9
Figure 7-4: Management of Municipal Solid Waste in the United States, 2018 7-18
Figure 7-5: MSW Management Trends from 1990 to 2018 7-19
Figure 7-6: Percent of Degradable Materials Diverted from Landfills from 1990 to 2018 (Percent) 7-20
Figure 9-1: Impacts from Recalculations to U.S. Greenhouse Gas Emissions by Sector, Including Quantitative
Change Related to the Use of AR5 GWP Values 9-5
Boxes
Box ES-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals, including
Relationship to EPA's Greenhouse Gas Reporting Program ES-2
Box ES-2: Trends in Various U.S. Greenhouse Gas Emissions-Related Data ES-23
Box ES-3: Use of Ambient Measurements Systems for Validation of Emission Inventories ES-26
Box 1-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals, including
Relationship to EPA's Greenhouse Gas Reporting Program 1-2
Box 1-2: The IPCCSixth Assessment Report and Global Warming Potentials 1-10
Box 1-3: Examples of Verification Activities 1-27
Box 2-1: Methodology for Aggregating Emissions by Economic Sector 2-31
Box 2-2: Trends in Various U.S. Greenhouse Gas Emissions-Related Data 2-39
Box 3-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals, including
Relationship to EPA's Greenhouse Gas Reporting Program 3-6
Box 3-2: Weather and Non-Fossil Energy Effects on CO2 Emissions from Fossil Fuel Combustion Trends 3-11
Box 3-3: Uses of Greenhouse Gas Reporting Program Data and Improvements in Reporting Emissions from
Industrial Sector Fossil Fuel Combustion 3-21
Box 3-4: Carbon Intensity of U.S. Energy Consumption 3-35
Box 3-5: Reporting of Lubricants, Waxes, and Asphalt and Road Oil Product Use in Energy Sector 3-55
Box 3-6: Carbon Dioxide Transport, Injection, and Geological Storage 3-91
Box 4-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals 4-7
Box 4-2: Industrial Process and Product Use Data from EPA's Greenhouse Gas Reporting Program 4-9
Box 5-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals 5-4
Box 5-2: Surrogate Data Method 5-26
Box 5-3: Tier 1 vs. Tier 3 Approach for Estimating N2O Emissions 5-37
Box 5-4: Data Splicing Method 5-39
Box 5-5: Comparison of the Tier 2 U.S. Inventory Approach and IPCC (2006) Default Approach 5-49
Box 5-6: Comparison of Tier 2 U.S. Inventory Approach and IPCC (2006) Default Approach 5-57
Box 6-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals 6-9
Box 6-2: Preliminary Estimates of Land Use in U.S. Territories 6-23
xx ix
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Box 6-3: CO2 Emissions from Forest Fires 6-31
Box 6-4: Surrogate Data Method 6-61
Box 6-5: Tier 3 Approach for Soil C Stocks Compared to Tier 1 or 2 Approaches 6-62
Box 7-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals, including
Relationship to Greenhouse Gas Reporting Data 7-3
Box 7-2: Description of a Modern, Managed Landfill in the United States 7-5
Box 7-3: Nationwide Municipal Solid Waste Data Sources 7-13
Box 7-4: Overview of U.S. Solid Waste Management Trends 7-18
Equations
Equation 1-1: Calculating CO2 Equivalent Emissions 1-9
Equation 3-1: Estimating Fugitive CO2 Emissions from Underground Mines 3-68
Equation 3-2: Estimating CO2 Emissions from Drained Methane Flared or Catalytically Oxidized 3-69
Equation 3-3: Decline Function to Estimate Venting Abandoned Mine Methane Emissions 3-72
Equation 3-4: Decline Function to Estimate Flooded Abandoned Mine Methane Emissions 3-73
Equation 4-1: 2006IPCC Guidelines Tier 1 Emission Factor for Clinker (precursor to Equation 2.4) 4-11
Equation 4-2: 2006 IPCC Guidelines Tier 2 Emission Factor for Lime Production, High-Calcium Lime (Equation 2.9) 4-
16
Equation 4-3: 2006 IPCC Guidelines Tier 2 Emission Factor for Lime Production, Dolomitic Lime (Equation 2.9).. 4-16
Equation 4-4: 2006 IPCC Guidelines Tier 3: N2O Emissions From Nitric Acid Production (Equation 3.6) 4-39
Equation 4-5: 2006 IPCC Guidelines Tier 2: N2O Emissions From Adipic Acid Production (Equation 3.8) 4-42
Equation 4-6: 2006 IPCC Guidelines Tier 1: N2O Emissions From Caprolactam Production (Equation 3.9) 4-46
Equation 4-7: 2006 IPCC Guidelines Tier 1: Emissions from Carbide Production (Equation 3.11) 4-50
Equation 4-8: 2006 IPCC Guidelines Tier 1: CO2 Emissions from Titanium Production (Equation 3.12) 4-53
Equation 4-9: CO2 Emissions from Phosphoric Acid Production 4-74
Equation 4-10: CO2 Emissions from Coke, Pig Iron, EAF Steel, and BOF Steel Production, based on 2006 IPCC
Guidelines Tier 2 Methodologies 4-80
Equation 4-11: 2006 IPCC Guidelines Tier 1: Emissions from Sinter, Direct Reduced Iron, and Pellet Production
(Equations 4.6, 4.7, and 4.8) 4-80
Equation 4-12: 2006 IPCC Guidelines Tier 1: CO2 Emissions for Ferroalloy Production (Equation 4.15) 4-90
Equation 4-13: 2006 IPCC Guidelines Tier 1: CFU Emissions for Ferroalloy Production (Equation 4.18) 4-90
Equation 4-14: CF4 Emissions Resulting from Low Voltage Anode Effects 4-97
Equation 4-15: 2006 IPCC Guidelines Tier 1: CO2 Emissions From Lead Production (Equation 4.32) 4-107
Equation 4-16: 2006 IPCC Guidelines Tier 1: CO2 Emissions From Zinc Production (Equation 4.33) 4-111
Equation 4-17: Waelz Kiln CO2 Emission Factor for Zinc Produced 4-112
Equation 4-18: Waelz Kiln CO2 Emission Factor for EAF Dust Consumed 4-112
Equation 4-19: Total Emissions from Electronics Industry 4-127
xxx DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Equation 4-20: Total Emissions from Semiconductor Manufacturing 4-127
Equation 4-21: Total Emissions from MEMS Manufacturing 4-129
Equation 4-22: Total Emissions from PV Manufacturing 4-129
Equation 4-23: Estimation for SF6 Emissions from Electric Power Systems 4-143
Equation 4-24: Regression Equation for Estimating SF6 Emissions of Non-Reporting Facilities in 1999 4-145
Equation 4-25: Regression Equation for Estimating SF6 Emissions of GHGRP-Only Reporters in 2011 4-145
Equation 4-26: N2O Emissions from Product Use 4-153
Equation 5-1: Elemental C or N Released through Oxidation of Crop Residues 5-56
Equation 5-2: Emissions from Crop Residue Burning 5-57
Equation 5-3: Estimation of Greenhouse Gas Emissions from Fire 5-57
Equation 6-1: Net State Annual Carbon Sequestration 6-168
Equation 6-2: Total C Stock for Yard Trimmings and Food Scraps in Landfills 6-177
Equation 6-3: C Stock Annual Flux for Yard Trimmings and Food Scraps in Landfills 6-178
Equation 7-1: Landfill Methane Generation 7-8
Equation 7-2: Net Methane Emissions from MSW Landfills 7-8
Equation 7-3: Net Methane Emissions from Industrial Waste Landfills 7-12
Equation 7-4: Total Domestic CFU Emissions from Wastewater Treatment and Discharge 7-25
Equation 7-5: CFU Emissions from Septic Systems 7-25
Equation 7-6: Total Wastewater BODs Produced per Capita (U.S.-Specific [ERG 2018a]) 7-26
Equation 7-7: Total Organically Degradable Material in Domestic Wastewater (IPCC 2019 [Eq. 6.3]) 7-26
Equation 7-8: Total Domestic CFU Emissions from Centrally Treated Aerobic Systems 7-28
Equation 7-9: Total Organics in Centralized Wastewater Treatment [IPCC 2019 (Eq. 6.3A)] 7-28
Equation 7-10: Organic Component Removed from Aerobic Wastewater Treatment (IPCC 2019 [Eq. 6.3B]) 7-28
Equation 7-11: Emissions from Centrally Treated Aerobic Systems (other than Constructed Wetlands) (IPCC 2019
[Eq. 6.1]) 7-28
Equation 7-12: Emissions from Centrally Treated Aerobic Systems (Constructed Wetlands Only) [IPCC 2014 (Eq.
6.1)] 7-30
Equation 7-13: Emissions from Centrally Treated Aerobic Systems (Constructed Wetlands used as Tertiary
Treatment) (U.S. Specific) 7-30
Equation 7-14: Emissions from Centrally Treated Anaerobic Systems [IPCC 2019 (Eq. 6.1)] 7-31
Equation 7-15: Emissions from Anaerobic Sludge Digesters (U.S. Specific) 7-31
Equation 7-16: Emissions from Centrally Treated Systems Discharge (U.S.-Specific) 7-32
Equation 7-17: Total Organics in Centralized Treatment Effluent (IPCC 2019 [Eq. 6.3D]) 7-32
Equation 7-18: Total Organics in Effluent Discharged to Reservoirs, Lakes, or Estuaries (U.S.-Specific) 7-32
Equation 7-19: Total Organics in Effluent Discharged to Other Waterbodies (U.S.-Specific) 7-32
Equation 7-20: Total CFU Emissions from Industrial Wastewater 7-34
xxxi
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Equation 7-21: TOW in Industry Wastewater Treatment Systems 7-34
Equation 7-22: Organic Component Removed from Aerobic Wastewater Treatment - Pulp, Paper, and Paperboard
7-35
Equation 7-23: Organic Component Removed from Aerobic Treatment Plants 7-35
Equation 7-24: Raw Sludge Removed from Wastewater Treatment as Dry Mass 7-35
Equation 7-25: Cm Emissions from Industrial Wastewater Treatment Discharge 7-37
Equation 7-26: TOW in Industrial Wastewater Effluent 7-38
Equation 7-27: Emissions from Pulp and Paper Discharge (U.S. Specific) 7-39
Equation 7-28: Total Organics in Pulp and Paper Effluent Discharged to Reservoirs, Lakes, Or Estuaries (U.S.
Specific) 7-39
Equation 7-29: Total Organics in Pulp and Paper Effluent Discharged to Other Waterbodies (U.S. Specific) 7-39
Equation 7-30: Total Domestic N2O Emissions from Wastewater Treatment and Discharge 7-42
Equation 7-31: Annual per Capita Protein Supply (U.S. Specific) 7-43
Equation 7-32: Consumed Protein [IPCC 2019 (Eq. 6.10A)] 7-43
Equation 7-33: Total Nitrogen Entering Septic Systems (IPCC 2019 [Eq. 6.10]) 7-43
Equation 7-34: Emissions from Septic Systems (IPCC 2019 [Eq. 6.9]) 7-44
Equation 7-35: Total Nitrogen Entering Centralized Systems (IPCC 2019 [Eq. 10]) 7-45
Equation 7-36: Total Domestic N2O Emissions from Centrally Treated Aerobic Systems 7-46
Equation 7-37: Emissions from Centrally Treated Aerobic Systems (other than Constructed Wetlands) (IPCC 2019
[Eq. 6.9]) 7-46
Equation 7-38: Emissions from Centrally Treated Aerobic Systems (Constructed Wetlands Only) (IPCC 2014 [Eq.
6.9]) 7-46
Equation 7-39: Emissions from Centrally Treated Aerobic Systems (Constructed Wetlands used as Tertiary
Treatment) (U.S.-Specific) 7-46
Equation 7-40: Emissions from Centrally Treated Anaerobic Systems (IPCC 2019 [Eq. 6.9]) C (kt N20/year) 7-47
Equation 7-41: Emissions from Centrally Treated Systems Discharge (U.S.-Specific) 7-49
Equation 7-42: Total Organics in Centralized Treatment Effluent (IPCC 2019 [Eq. 6.8]) 7-49
Equation 7-43: Total Nitrogen in Effluent Discharged to Impaired Waterbodies (U.S.-Specific) 7-49
Equation 7-44: Total Nitrogen in Effluent Discharged to Nonimpaired Waterbodies (U.S.-Specific) 7-49
Equation 7-45: Total Nitrogen in Industrial Wastewater 7-51
Equation 7-46: N2O Emissions from Indsutrial Wastewater Treatment Plants 7-51
Equation 7-47: N2O Emissions from Industrial Wastewater Treatment Effluent 7-52
Equation 7-48: Greenhouse Gas Emission Calculation for Composting 7-57
Equation 7-49: Methane Emissions Calculation for Anaerobic Digestion 7-61
Equation 7-50: Recovered Methane Estimation for Anaerobic Digestion 7-62
Equation 7-51: Weighted Average of Waste Processed 7-62
xxxii DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Executive Summary
An inventory that identifies and quantifies a country's anthropogenic1 sources and sinks of greenhouse gas
emissions and removals is essential for addressing climate change. This Inventory adheres to both (1) a
comprehensive and detailed set of methodologies for estimating national sources and sinks of anthropogenic
greenhouse gases, and (2) a common and consistent format that enables Parties to the United Nations Framework
Convention on Climate Change (UNFCCC) to compare the relative contribution of different emission sources and
greenhouse gases to climate change.
In 1992, the United States signed and ratified the UNFCCC. As stated in Article 2 of the UNFCCC, "The ultimate
objective of this Convention and any related legal instruments that the Conference of the Parties may adopt is to
achieve, in accordance with the relevant provisions of the Convention, stabilization of greenhouse gas
concentrations in the atmosphere at a level that would prevent dangerous anthropogenic interference with the
climate system. Such a level should be achieved within a time-frame sufficient to allow ecosystems to adapt
naturally to climate change, to ensure that food production is not threatened and to enable economic
development to proceed in a sustainable manner."2
As a signatory to the UNFCCC, consistent with Article 43 and decisions at the First, Second, Fifth, and Nineteenth
Conference of Parties,4 the United States is committed to submitting a national inventory of anthropogenic
sources and sinks of greenhouse gases to the UNFCCC by April 15 of each year. The United States views this report,
in conjunction with Common Reporting Format (CRF) reporting tables that accompany this report, as an
opportunity to fulfill this annual commitment under the UNFCCC.
This executive summary provides the latest information on U.S. anthropogenic greenhouse gas emission trends
from 1990 through 2021. The structure of this report is consistent with the UNFCCC guidelines for inventory
reporting, as discussed in Box ES-1.5
1 The term "anthropogenic," in this context, refers to greenhouse gas emissions and removals that are a direct result of human
activities or are the result of natural processes that have been affected by human activities (IPCC 2006).
2 Article 2 of the Framework Convention on Climate Change published by the UNEP/WMO Information Unit on Climate Change.
See http://unfccc.int.
3 Article 4(l)(a) of the United Nations Framework Convention on Climate Change (also identified in Article 12) and subsequent
decisions by the Conference of the Parties elaborated the role of Annex I Parties in preparing national inventories. Article 4
states "Parties to the Convention, by ratifying, shall develop, periodically update, publish and make available...national
inventories of anthropogenic emissions by sources and removals by sinks of all greenhouse gases not controlled by the
Montreal Protocol, using comparable methodologies..." See http://unfccc.int for more information.
4 See UNFCCC decisions 3/CP.l, 9/CP.2, 3/CP.5, and 24/CP.19 at https://unfccc.int/documents.
5 See http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
ES-1
-------
1
2
In following the UNFCCC requirement under Article 4.1 and related decisions to develop and submit annual
national greenhouse gas emission inventories, the emissions and removals presented in this report and this
chapter are organized by source and sink categories and calculated using internationally accepted methods
provided by the IPCC in the 2006IPCC Guidelines for National Greenhouse Gas Inventories (2006IPCC
Guidelines) and where appropriate, its supplements and refinements. Additionally, the calculated emissions and
removals in a given year for the United States are presented in a common manner in line with the UNFCCC
reporting guidelines for the reporting of inventories under this international agreement. The use of consistent
methods to calculate emissions and removals by all nations providing their inventories to the UNFCCC ensures
that these reports are comparable. The presentation of emissions and removals provided in this Inventory does
not preclude alternative examinations, but rather this Inventory presents emissions and removals in a common
format consistent with how countries are to report inventories under the UNFCCC. The report itself, and this
chapter, follows this standardized format, and provides an explanation of the application of methods used to
calculate emissions and removals.
EPA also collects greenhouse gas emissions data from individual facilities and suppliers of certain fossil fuels and
industrial gases through its Greenhouse Gas Reporting Program (GHGRP), which is complementary to the U.S.
Inventory.6 The GHGRP applies to direct greenhouse gas emitters, fossil fuel suppliers, industrial gas suppliers,
and facilities that inject carbon dioxide (CO2) underground for sequestration or other reasons and requires
reporting by over 8,000 sources or suppliers in 41 industrial categories.7 Annual reporting is at the facility level,
except for certain suppliers of fossil fuels and industrial greenhouse gases. In general, the threshold for
reporting is 25,000 metric tons or more of CO2 Eq. per year. Facilities in most source categories subject to
GHGRP began reporting for the 2010 reporting year while additional types of industrial operations began
reporting for reporting year 2011. Methodologies used in EPA's GHGRP are consistent with the 2006 IPCC
Guidelines. While the GHGRP does not provide full coverage of total annual U.S. greenhouse gas emissions and
sinks (e.g., the GHGRP excludes emissions from the agricultural, land use, and forestry sectors), it is an
important input to the calculations of national-level emissions in this Inventory.
The GHGRP dataset provides not only annual emissions information, but also other annual information such as
activity data and emission factors that can improve and refine national emission estimates over time. GHGRP
data also allow EPA to disaggregate national inventory estimates in new ways that can highlight differences
across regions and sub-categories of emissions, along with enhancing the application of QA/QC procedures and
assessment of uncertainties. See Annex 9 for more information on specific uses of GHGRP data in the Inventory
(e.g., use of Subpart W data in compiling estimates for natural gas systems).
5 Greenhouse gases absorb infrared radiation, trapping heat in the atmosphere and making the planet warmer. The
6 most important greenhouse gases directly emitted by humans include carbon dioxide (CO2), methane (CH4),
7 nitrous oxide (N2O), and several fluorine-containing halogenated substances (HFCs, PFCs, SF6 and NF3). Although
8 CO2, CH4, and N2O occur naturally in the atmosphere, human activities have changed their atmospheric
6 On October 30, 2009 the EPA promulgated a rule requiring annual reporting of greenhouse gas data from large greenhouse
gas emissions sources in the United States. Implementation of the rule, codified at 40 CFR Part 98, is referred to as EPA's
Greenhouse Gas Reporting Program (GHGRP).
7 See http://www.epa.gov/ghgreporting and http://ghgdata.epa.gov/ghgp/main.do.
2 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
3
4
ES.l Background Information
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
concentrations. From the pre-industrial era (i.e., ending about 1750) to 2021, concentrations of these greenhouse
gases have increased globally by 48.1,170.8, and 23.8 percent, respectively (IPCC 2013; NOAA/ESRL 2023a, 2023b,
2023c). This annual report estimates the total national greenhouse gas emissions and removals associated with
human activities across the United States.
Global Warming Potentials
The IPCC developed the global warming potential (GWP) concept to compare the ability of a greenhouse gas to
trap heat in the atmosphere relative to another gas. The GWP of a greenhouse gas is defined as the ratio of the
accumulated radiative forcing within a specific time horizon caused by emitting 1 kilogram of the gas, relative to
that of the reference gas CO2 (IPCC 2013); therefore, CC>2-equivalent emissions are provided in million metric tons
of CO2 equivalent (MMT CO2 Eq.) for non-CC>2 greenhouse gases.8,9 All estimates are provided throughout the
main report in both CO2 equivalents and unweighted units, while estimates for all gases in this Executive Summary
are presented in units of MMT CO2 Eq. Emissions by gas in unweighted mass kilotons are also provided in the
Trends and sector chapters of this report and in the Common Reporting Format (CRF) tables that are included in
the submission to the UNFCCC.
Recent decisions under the UNFCCC10 require Parties to use 100-year GWP values from the IPCC Fifth Assessment
Report (AR5) for calculating CC>2-equivalents in their national reporting (IPCC 2013) by the end of 2024. This
reflects updated science and ensures that national greenhouse gas inventories reported by all nations are
comparable. In preparation for upcoming UNFCCC requirements,11 this report reflects CC>2-equivalent greenhouse
gas emission totals using 100-year AR5 GWP values. A comparison of emission values with the previously used 100-
year GWP values from the IPCC Fourth Assessment Report (AR4) (IPCC 2007), and the IPCC Sixth Assessment Report
(AR6) (IPCC 2021) values can be found in Annex 6.1 of this report. The 100-year GWP values used in this report are
listed below in Table ES-1.
Table ES-1: Global Warming Potentials (100-Year Time Horizon) Used in this Report
Gas GWP
C02 1
CH4a 28
N20 265
MFCs up to 12,400
PFCs up to 11,100
SF6 23,500
NFs 16,100
Other Fluorinated Gases See Annex 6
a The GWP of CH4 includes the direct effects
and those indirect effects due to the
production of tropospheric ozone and
stratospheric water vapor. The indirect effect
due to production of C02 is not included. See
Annex 6 for additional information.
Source: IPCC (2013).
8 Carbon comprises 12/44 of carbon dioxide by weight.
9 One million metric ton is equal to 1012 grams or one teragram.
10 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/sbsta2022 L25a01E.pdf.
11 See Annex to decision 18/CMA.l, available online at https://unfccc.int/sites/default/files/resource/CMA2018 03a02E.pdf.
ES-3
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
ES.2 Recent Trends in U.S. Greenhouse Gas
Emissions and Sinks
In 2021, total gross U.S. greenhouse gas emissions were 6,347.7 million metric tons of carbon dioxide equivalent
(MMT CO2 Eq).12 Total U.S. emissions have decreased by 2.0 percent from 1990 to 2021, down from a high of 15.8
percent above 1990 levels in 2007. Emissions increased from 2020 to 2021 by 5.5 percent (333.2 MMT CO2 Eq.).
Net emissions (including sinks) were 5,593.5 MMT CO2 Eq. in 2021. Overall, net emissions increased 6.8 percent
from 2020 to 2021 and decreased 16.3 percent from 2005 levels as shown in Table ES-2. From 2019 to 2020, there
was a sharp decline in emissions largely due to the impacts of the coronavirus (COVID-19) pandemic on travel and
other economic activity. Between 2020 and 2021, the increase in total greenhouse gas emissions was driven
largely by an increase in CO2 emissions from fossil fuel combustion due to economic activity rebounding after the
COVID-19 pandemic. In 2021, CO2 emissions from fossil fuel combustion increased by 7.0 percent relative to the
previous year. Carbon dioxide emissions from natural gas use increased by 8.3 MMT CO2 Eq., a 0.5 percent
increase from 2020. In a shift from recent trends, CO2 emissions from coal consumption increased by 122.1 MMT
CO2 Eq., a 14.6 percent increase from 2020. The increase in natural gas consumption and emissions in 2021 is
observed across all sectors except the Electric Power sector and U.S. Territories, while the coal increase is primarily
in the Electric Power sector. Emissions from petroleum use also increased by 175.8 MMT CO2 Eq. (9.3 percent)
from 2020 to 2021. In 2021, CO2 emissions from fossil fuel combustion were 4,651.0 MMT CO2 Eq., or 1.6 percent
below emissions in 1990.
Figure ES-1, Figure ES-2, and Figure ES-3 illustrate the overall trends in total U.S. emissions by gas, annual percent
changes, and relative change since 1990 for each year of the time series, and Table ES-2 provides information on
trends in gross U.S. greenhouse gas emissions and sinks for 1990 through 2021. Unless otherwise stated, all tables
and figures provide total gross emissions and exclude the greenhouse gas fluxes from the Land Use, Land-Use
Change, and Forestry (LULUCF) sector. For more information about the LULUCF sector see Section ES.3 Overview
of Sector Emissions and Trends.
Table ES-2: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks (MMT CO2 Eq.)
Gas/Source
1990
2005
2017
2018
2019
2020
2021
C02
5,121.4
6,132.4
5,212.1
5,378.0
5,259.8
4,714.4
5,048.2
CH4c
868.7
791.2
762.8
774.2
767.8
742.3
727.4
N2Oc
396.7
405.1
402.8
418.5
399.1
377.7
384.8
HFCs
39.0
116.4
160.8
160.9
165.4
168.2
175.1
PFCs
21.8
6.1
3.8
4.3
4.0
3.9
3.5
sf6
30.5
15.5
7.2
7.1
7.8
7.5
8.0
nf3
+
0.4
0.5
0.5
0.5
0.6
0.6
Total Gross Emissions (Sources)
6,478.3
7,466.9
6,550.0
6,743.4
6,604.4
6,014.5
6,347.7
LULUCF Emissions3
57.9
72.4
68.3
64.4
64.2
76.4
77.8
ch4
53.5
61.3
60.1
57.3
56.9
65.4
66.0
n2o
4.4
11.1
8.3
7.0
7.3
11.0
11.8
LULUCF Carbon Stock Changeb
(938.9)
(853.5)
(842.5)
(829.5)
(768.2)
(852.5)
(832.0)
LULUCF Sector Net Totalc
(881.0)
(781.1)
(774.2)
(765.1)
(704.0)
(776.2)
(754.2)
Net Emissions (Sources and Sinks)
5,597.3
6,685.8
5,775.8
5,978.3
5,900.3
5,238.3
5,593.5
+ Does not exceed 0.05 MMT C02 Eq.
a LULUCF emissions of CH4 and N20 are reported separately from gross emissions totals. LULUCF emissions include the CH4
and N20 emissions reported for Peatlands Remaining Peatlands, forest fires, drained organic soils, grassland fires, and
12 The gross emissions total presented in this report for the United States excludes emissions and removals from Land Use,
Land-Use Change, and Forestry (LULUCF). The net emissions total presented in this report for the United States includes
emissions and removals from LULUCF.
4 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Coastal Wetlands Remaining Coastal Wetlands; CH4 emissions from Land Converted to Coastal Wetlands, Flooded Land
Remaining Flooded Land, and Land Converted to Flooded Land; and N20 emissions from forest soils and settlement soils.
b LULUCF Carbon Stock Change is the net C stock change from the following categories: Forest Land Remaining Forest Land,
Land Converted to Forest Land, Cropland Remaining Cropland, Land Converted to Cropland, Grassland Remaining Grassland,
Land Converted to Grassland, Wetlands Remaining Wetlands, Land Converted to Wetlands, Settlements Remaining
Settlements, and Land Converted to Settlements.
c The LULUCF Sector Net Total is the net sum of all LULUCF CH4 and N20 emissions to the atmosphere plus LULUCF net C stock
changes.
Notes: Total (gross) are emissions presented without LULUCF. Net emissions are presented with LULUCF. Totals may not sum
due to independent rounding. Parentheses indicate negative values or sequestration.
l Figure ES-1: U.S. Greenhouse Gas Emissions and Sinks by Gas
¦ HFCs, PFCs, SFe and NFa — Net Emissions (including LULUCF sinks)
9,000 m Nitrous Oxide
¦ Methane
8,000 B Carbon Dioxide
| ¦ Net CO2 Flux from LULUCF^
7,000
6,000 . • "
& 5,000
LU
N
S 4,000
I-
2:
E 3,000
2,000
1,000
0
-1,000
o • vjDrv»coc7^0'-
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Figure ES-2: Annual Percent Change in Gross U.S. Greenhouse Gas Emissions and Sinks
Relative to the Previous Year
8%
6%
4%
2%
0%
-2%
-4%
-6%
-8%
-10%
5.5%
1.6% 2.9%
1.6% °-8%
2.8%
CM CO TJ- LD UD
CTi CTi CTi CTi O'i CTi
CTi CTi CTi CTi CTi CTi
00 0s! O t—i fM
CT\ CTi O O O
CTiCTiOOOoooooooaoooocpcjoooocj
i-H-^HfMrsjfNrMrsjfMrMrMfNtNjrsjrsjrsJtNrvjfNrMfNrMrvjfNj(Nj
Improvements and Recalculations Relative to the Previous
Inventory
Each year, some emission and sink estimates in the Inventory are recalculated and revised to incorporate
improved methods and/or data. The most common reason for recalculating U.S. greenhouse gas emission
estimates is to update recent historical data. Changes in historical data are generally the result of changes in data
supplied by other U.S. government agencies or organizations, as they continue to make refinements and
improvements. These improvements are implemented consistently across the previous Inventory's time series, as
necessary, (i.e., 1990 to 2020) to ensure that the trend is accurate. In addition, for the current Inventory, CO2-
equivalent emission estimates have been updated to reflect the 100-year GWP values provided in the IPCC Fifth
Assessment Report (AR5) (IPCC 2013).
Below are categories with methodological and data-related recalculations13 resulting in an average change of
greater than 2.5 MMT CO2 Eq. over the time series.
• Forest Land Remaining Forest Land: Changes in Forest Carbon Stocks (CO2)
• Wetlands Remaining Wetlands: Emissions from Flooded Land Remaining Flooded Land (CH4)
• Petroleum Systems (CH4)
• Land Converted to Grassland: Changes in all Ecosystem Carbon Stocks (CO2)
• Land Converted to Cropland: Changes in all Ecosystem Carbon Stocks (CO2)
• Natural Gas Systems (CH4)
In addition, the Inventory includes two new categories not included in the previous Inventory that improve
completeness of the national estimates: CO2 emissions from the Substitution of Ozone Depleting Substances and
CO2 from the biogenic components of municipal solid waste combustion (reported as a memo item in the
Inventory).
In each Inventory, the results of all methodological changes and historical data updates and the inclusion of new
sources and sink estimates are summarized in the Recalculations and Improvements chapter (Chapter 9). For more
detailed descriptions of each recalculation including references for data, please see the respective source or sink
13 This does not include the recalculations related to the update from AR4 to AR5 GWP values. For more information on the
impact of that update, please see Chapter 9, Recalculations and Improvements.
6 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
category description(s) within the relevant report chapter (i.e., the Energy chapter [Chapter 3], the Industrial
Processes and Product Use [IPPU] chapter [Chapter 4] the Agriculture chapter [Chapter 5], the Land Use, Land Use
Change and Forestry [LULUCF] chapter [Chapter 6], and the Waste chapter [Chapter 7]). In implementing
improvements, the United States follows the 2006IPCC Guidelines (IPCC 2006), which states,
"Both methodological changes and refinements over time are an essential part of improving inventory
quality. It is good practice to change or refine methods when: available data have changed; the previously
used method is not consistent with the IPCC guidelines for that category; a category has become key; the
previously used method is insufficient to reflect mitigation activities in a transparent manner; the capacity for
inventory preparation has increased; new inventory methods become available; and for correction of errors."
Emissions by Gas
Figure ES-3 illustrates the relative contribution of the greenhouse gases to total gross U.S. emissions in 2021,
weighted by global warming potential. The primary greenhouse gas emitted by human activities in the United
States is CO2, representing 79.5 percent of total greenhouse gas emissions. The largest source of CO2 and of overall
greenhouse gas emissions is fossil fuel combustion, primarily from transportation and power generation. Methane
(CH4) emissions account for 11.5 percent of emissions. The major sources of methane include enteric fermentation
associated with domestic livestock, natural gas systems, and decomposition of wastes in landfills. Agricultural soil
management, wastewater treatment, stationary sources of fuel combustion, and manure management are the
major sources of N2O emissions. Ozone depleting substance substitute emissions are the primary contributor to
aggregate hydrofluorocarbon (HFC) emissions. Perfluorocarbon (PFC) emissions are primarily attributable to
electronics manufacturing and primary aluminum production. Electrical transmission and distribution systems
account for most sulfur hexafluoride (SFs) emissions. The electronics industry is the only source of nitrogen
trifluoride (NF3) emissions.
Figure ES-3: 2021 Total Gross U.S. Greenhouse Gas Emissions by Gas (Percentages based on
MMT COz Eq.)
3.0%
HFCs, PFCs, SFe and NF3
Note: Emissions and removals from Land Use, Land-Use Change, and Forestry are excluded from figure above.
From 1990 to 2021, total emissions of CO2 decreased by 73.3 MMT CO2 Eq. (1.4 percent), total emissions of CH4
decreased by 141.3 MMT CO2 Eq. (16.3 percent), and emissions of N2O decreased by 11.8 MMT CO2 Eq. (3.0
percent). During the same period, emissions of fluorinated greenhouse gases including HFCs, PFCs, SF6, and NF3
rose by 95.9 MMT CO2 Eq. (104.8 percent). From 1990 to 2021, emissions of HFCs increased by 136.1 MMT CO2 Eq.
ES-7
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
(348.6 percent) and NF3 emissions increased by 0.6 MMT CO2 Eq. (1,318.9 percent), while emissions of PFCs
decreased by 18.3 MMT CO2 Eq. (83.8 percent) and SF6 emissions decreased by 22.5 MMT CO2 Eq. (73.7 percent).
Despite being emitted in smaller quantities relative to the other principal greenhouse gases, emissions of HFCs,
PFCs, SFs and NF3 are significant because many of these gases have extremely high global warming potentials and,
in the cases of PFCs and SF6, long atmospheric lifetimes. Conversely, U.S. greenhouse gas emissions were partly
offset by carbon (C) sequestration in forests, trees in urban areas, agricultural soils, landfilled yard trimmings and
food scraps, and coastal wetlands, which together offset 13.1 percent of gross total emissions in 2021 (as reflected
in Figure ES-1). The following sections describe each gas's contribution to total U.S. greenhouse gas emissions in
more detail.
Carbon Dioxide Emissions
The global carbon cycle is made up of large carbon flows and reservoirs. Billions of tons of carbon in the form of
CO2 are absorbed by oceans and living biomass (i.e., sinks) and are emitted to the atmosphere annually through
natural processes (i.e., sources). When in equilibrium, global carbon fluxes among these various reservoirs are
roughly balanced.14
Since the Industrial Revolution (i.e., about 1750), global atmospheric concentrations of CO2 have risen 48.1 percent
(IPCC 2013; NOAA/ESRL 2023a), principally due to the combustion of fossil fuels for energy. Globally, an estimated
33,000 MMT of CO2 were added to the atmosphere through the combustion of fossil fuels in 2021, of which the
United States accounted for approximately 14 percent.15
Within the United States, fossil fuel combustion accounted for 92.1 percent of CO2 emissions in 2021. Nationally,
the fossil fuel combustion transportation subsector was the largest emitter of CO2 in 2021 followed by the electric
power generation subsector. There are 27 additional sources of CO2 emissions included in the Inventory (see Table
2-1). Although not illustrated in Table ES-4, changes in land use and forestry practices can also lead to net CO2
emissions (e.g., through conversion of forest land to agricultural or urban use) or to a net sink for CO2 (e.g.,
through net additions to forest biomass). See more on these emissions and removals in Table ES-4.
Figure ES-4: 2021 Sources of CO2 Emissions
Fossil Fuel Combustion
Non-Energy Use of Fuels
Iron and Steel Production
Cement Production
Natural Gas Systems
Petrochemical Production
Other Industrial Processes
Petroleum Systems
Incineration of Waste
Ammonia Production
Lime Production
Other Energy
Net Carbon Stock Change from LULUCF
-100 -75 -50 -25 0 25 50 75 100 125 150
MMT CO2 Eq.
14 The term "flux" is used to describe the exchange of C02 to and from the atmosphere, with net flux being either positive or
negative depending on the overall balance. Removal and long-term storage of C02 from the atmosphere is also referred to as
"carbon sequestration."
15 Global C02 emissions from fossil fuel combustion were taken from International Energy Agency Global energy-related C02
emissions, 1990-2021 - Charts Available at: https://www.iea.org/data-and-statistics/charts/global-energy-related-co2-
emissions-1990-2021 (IEA 2022).
8 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Note: Other Industrial Processes includes emissions from Aluminum Production, Carbide Production, Carbon Dioxide
Consumption, Ferroalloy Production, Lead Production, Magnesium Production, Other Process Uses of Carbonates,
Phosphoric Acid Production, Soda Ash, Titanium Dioxide, Urea Consumption, and Zinc Production. Other Energy includes
emissions from Abandoned Oil and Gas Wells and Coal Mining.
As the largest source of U.S. greenhouse gas emissions, CO2 from fossil fuel combustion has accounted for an
average of 74.9 percent of CCh-equivalent total gross U.S. emissions across the time series. Between 1990 and
2021, CO2 emissions from fossil fuel combustion decreased from 4,728.2 MMT CO2 Eq. to 4,651.0 MMT CO2 Eq., a
1.6 percent total decrease. Conversely, CO2 emissions from fossil fuel combustion decreased by 1,096.4 MMT CO2
Eq. from 2005 levels, a decrease of 19.1 percent. From 2020 to 2021, these emissions increased by 306.1 MMT CO2
Eq. (7.0 percent).
Historically, changes in emissions from fossil fuel combustion have been the driving factor affecting U.S. emission
trends. Changes in CO2 emissions from fossil fuel combustion are influenced by many long-term and short-term
factors. Important drivers include: (1) changes in demand for energy; and (2) a general decline in the carbon
intensity of fuels combusted for energy in recent years by non-transport sectors of the economy. Long-term
factors affecting energy demand include population and economic trends, technological changes including energy
efficiency, shifting energy fuel choices, and various policies at the national, state, and local level. In the short term,
the overall consumption and mix of fossil fuels in the United States fluctuates primarily in response to changes in
general economic conditions, overall energy prices, the relative price of different fuels, weather, and the
availability of non-fossil alternatives. For example, between 2019 and 2021, changes in economic activity and
travel due to the COVID-19 pandemic and the subsequent recovery have had significant impacts on energy use and
fossil fuel combustion emissions.
The five major fuel-consuming economic sectors are transportation, electric power, industrial, residential, and
commercial and are described below. Carbon dioxide emissions are produced by the electric power sector as fossil
fuel is consumed to provide electricity to one of the other four sectors, or "end-use" sectors, see Figure ES-5. Note
that this Figure reports emissions from U.S. Territories as their own end-use sector due to incomplete data for
their individual end-use sectors. Fossil fuel combustion for electric power also includes emissions of less than 0.5
MMT CChEq. from geothermal-based generation.
Figure ES-5: 2021 CO2 Emissions from Fossil Fuel Combustion by Sector and Fuel Type
2,500
2,000
d"
^ 1,500
8
1-
i 1,000
500
0
Table ES-6 summarizes CO2 emissions from fossil fuel combustion by end-use sector including electric power
emissions. For Figure ES-6, electric power emissions have been distributed to each end-use sector on the basis of
each sector's share of aggregate electricity use (i.e., indirect fossil fuel combustion). This method of distributing
emissions assumes that each end-use sector uses electricity that is generated from the national average mix of
fuels according to their carbon intensity. Emissions from electric power are also addressed separately after the
end-use sectors are discussed.
Relative Contribution by Fuel Type
<0.05%
(Geothermal)
224
23
Coal
Natural Gas
I Geothermal
I Petroleum
U.S. Territories
Commercial
Residential
1,789
1,542
Industrial Electricity Generation Transportation
ES-9
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Figure ES-6: 2021 End-Use Sector Emissions of CO2 from Fossil Fuel Combustion
Transportation End-Use Sector. Transportation activities accounted for 38.6 percent of U.S. CO2 emissions from
fossil fuel combustion in 2021, with the largest contributors being light-duty trucks (37.0 percent), followed by
freight trucks (23.5 percent) and passenger vehicles (20.6 percent), and. Annex 3.2 presents the total emissions
from all transportation and mobile sources, including CO2, Cm, N2O, and HFCs.
In terms of the overall trend, from 1990 to 2021, total transportation CO2 emissions increased due, in large part, to
increased demand for travel a result of a confluence of factors including population growth, economic growth,
urban sprawl, and low fuel prices during the beginning of this period. From 2020 to 2021, transportation CO2
emissions increased 13.8 percent, largely reflective of a rebound in travel activity as COVID-19 pandemic
restrictions were eased. While an increased demand for travel has led to generally increasing CO2 emissions since
1990, improvements in average new vehicle fuel economy since 2005 have slowed the rate of increase of CO2
emissions. In 2021, petroleum-based products supplied 94.6 percent of the energy consumed for transportation,
primarily from gasoline consumption in automobiles and other highway vehicles (53.2 percent), diesel fuel for
freight trucks (24.5 percent), jet fuel for aircraft (10.2 percent), and natural gas, residual fuel, aviation gasoline,
and liquefied petroleum gases (6.7 percent). The remaining 5.5 percent is associated with renewable fuels (i.e.,
biofuels).
Industrial End-Use Sector. Industrial CO2 emissions, resulting both directly from the combustion of fossil fuels and
indirectly from the generation of electricity that is used by industry, accounted for 24.3 percent of CO2 emissions
from fossil fuel combustion in 2021. Approximately 63.4 percent of these emissions resulted from direct fossil fuel
combustion to produce steam and/or heat for industrial processes. The remaining emissions resulted from the use
of electricity for motors, electric furnaces, ovens, lighting, and other applications. Total direct and indirect
emissions from the industrial sector have declined by 22.6 percent since 1990. This decline is due to structural
changes in the U.S. economy (i.e., shifts from a manufacturing-based to a service-based economy), fuel switching,
and efficiency improvements. From 2020 to 2021, total energy use in the industrial sector increased by 2.2 percent
due to an increase in total industrial production and manufacturing output.
Residential and Commercial End-Use Sectors. The residential and commercial end-use sectors accounted for 19.1
and 16.0 percent, respectively, of CO2 emissions from fossil fuel combustion in 2021. The residential and
commercial sectors relied heavily on electricity for meeting energy demands, with 65.1 and 69.9 percent,
respectively, of their emissions attributable to electricity use for lighting, heating, cooling, and operating
appliances. The remaining emissions were due to the consumption of natural gas and petroleum for heating and
cooking. Total direct and indirect emissions from the residential sector have decreased by 4.7 percent since 1990.
Total direct and indirect emissions from the commercial sector have decreased by 3.0 percent since 1990. From
2020 to 2021, an increase in heating degree days (0.5 percent) increased energy demand for heating in the
residential and commercial sectors, however, a 1.9 percent decrease in cooling degree days compared to 2020
reduced demand for air conditioning in the residential and commercial sectors. This resulted in a 0.7 percent
10 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
increase in residential sector electricity use. From 2020 to 2021 energy use in the commercial sector increased by
2.9 percent, due in part to the gradual recovery from the COVID-19 pandemic, which had reduced economic and
manufacturing activity in 2020.
Electric Power. The United States relies on electricity to meet a significant portion of its energy demands.
Electricity generators used 30.7 percent of U.S. energy from fossil fuels and emitted 33.2 percent of the CO2 from
fossil fuel combustion in 2021. The type of energy source used to generate electricity is the main factor influencing
emissions.16 The mix of fossil fuels used also impacts emissions. The electric power sector is the largest consumer
of coal in the United States. The coal used by electricity generators accounted for 91.9 percent of all coal
consumed for energy in the United States in 2021.17 However, the amount of coal and the percent of total
electricity generation from coal has been decreasing overtime. Coal-fired electric generation (in kilowatt-hours
[kWh]) decreased from 54.2 percent of generation in 1990 to 22.5 percent in 2021.18 This corresponded with an
increase in natural gas generation and non-fossil fuel renewable energy generation, largely from wind and solar
energy. Natural gas generation (in kWh) represented 10.7 percent of electric power generation in 1990 and
increased over the thirty-two-year period to represent 37.2 percent of electric power generation in 2021. Wind
and solar generation (in kWh) represented 0.1 percent of electric power generation in 1990 and increased over the
thirty-two-year period to represent 11.0 percent of electric power generation in 2021. The recovery from the
COVID-19 pandemic led to an increase in electricity use of about 1.9 percent from 2020 to 2021. Between 2020
and 2021, coal electricity generation increased by 13.1 percent, natural gas generation decreased by 5.9 percent,
and renewable energy generation increased by 2.8 percent.
Across the time series, changes in electricity generation and the carbon intensity of fuels used for electric power
have a significant impact on CO2 emissions. While CO2 emissions from fossil fuel combustion from the electric
power sector have decreased by 15.3 percent since 1990, the carbon intensity of the electric power sector, in
terms of CO2 Eq. per QBtu input, has significantly decreased during that same timeframe by 24.9 percent. This
decoupling of the level of electric power generation and the resulting CO2 emissions is shown in Figure ES-7.
16 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 included in the electricity sector totals
and trends discussed in this section. Net carbon fluxes from changes in biogenic carbon reservoirs are accounted for in the
estimates for Land Use, Land-Use Change and Forestry.
17 See Table 6.2 Coal Consumption by Sector of EIA (2022a).
18 Values represent electricity net generation from the electric power sector. See Table 7.2b Electricity Net Generation: Electric
Power Sector of EIA (2022a).
ES-11
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Figure ES-7: Electric Power Generation and Emissions
I Nuclear Generation (Billion kWh)
Renewable Generation (Billion kWh)
I Petroleum Generation (Billion kWh)
Coal Generation (Billion kWh)
Natural Gas Generation (Billion kWh)
-Total Emissions (MMT CO2 Eq.) [Right Axis]
500
OTHH
OIO^O^O^O^O^O^OIOSO^OOOOOOOOOOt—It—It-Hi—It-Hi—I t-H t-H t—It-H(N(N
cncricncrtcT.o^criChcjic^oooooooooooooooooooooo
HrHHrHrHHHrHHrH(N(NlN(N(N(N(N(N(N(NfMrv|(N(NtN(N(N(N(N(NN(N
3,500
3,000
2,500 S
(N
o
u
I-
2,000 I
to
c
o
1,500 I
LU
ro
1,000 H
Other significant CO2 trends included the following:
• Carbon dioxide emissions from natural gas and petroleum systems were 36.8 and 24.7 MMT CO2 Eq.,
respectively, and combined accounted for 1.2 percent of CO2 emissions and 1.0 percent of total gross
emissions in 2021. These emissions increased by 19.6 MMT CO2 Eq. (46.9 percent) from 1990 to 2021.
This increase is due primarily to increases in the production segment, where flaring emissions from
associated gas flaring, tanks, and miscellaneous production flaring have increased over time.
• Carbon dioxide emissions from iron and steel production and metallurgical coke production were 42 MMT
CO2 Eq. in 2021 and accounted for less than 1 percent of CO2 and total gross emissions. Emissions have
decreased by 62.7 MMT CO2 Eq. (59.9 percent) from 1990 through 2021. This decrease is primarily due to
restructuring of the industry, technological improvements, and increased scrap steel utilization.
• Total C stock change (i.e., net CO2 removals) in the LULUCF sector decreased by 11.4 percent between
1990 and 2021. This decrease was primarily due to a decrease in the rate of net C accumulation in forest C
stocks and Cropland Remaining Cropland, as well as an increase in emissions from Land Converted to
Settlements.
Methane Emissions
Methane (CH4) is significantly more effective than CO2 at trapping heat in the atmosphere-by a factor of 28 over a
100-year time frame based on the IPCC Fifth Assessment Report estimate (IPCC 2013). Over the last two hundred
and fifty years, the concentration of CH4 in the atmosphere increased by 170.8 percent (IPCC 2013; NOAA/ESRL
2023b). Within the United States, the main anthropogenic sources of CH4 include enteric fermentation from
domestic livestock, natural gas systems, landfills, domestic livestock manure management, coal mining, and
petroleum systems (see Figure ES-8).
12 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Figure ES-8: 2021 Sources of ChU Emissions
Wastewater Treatment
Stationaiy Combustion
Manure Management
Enteric Fermentation
Natural Gas Systems
LULUCF Emissions
Petroleum Systems
Coal Mining
Other Energy
Rice Cultivation
Other Waste
Landfills
¦ CO*
¦ ChU
¦ N2O
11.5°/
CH4 as a Portion of All
Emissions
195
Field Burning of Agricultural Residues
Other Industrial Processes < °-5
¦ HFCs, PFCs, SFe and NF3
0 20 40 60
80 100 120 140 160 180 200
MMTCO2 Eq.
Note: Other Energy includes CH4 emissions from Abandoned Oil and Gas Wells, Underground Coal Mines, Incineration of Waste,
and Mobile Combustion. Other Waste includes CH4 emissions from anaerobic digestion at biogas facilities and composting.
Methane emissions from Carbide Production and Consumption, Ferroalloy Production, Iron and Steel Production, and Other
Industrial Processes includes Petrochemical Production. LULUCF emissions include the CH4 reported for Peatlands Remaining
Peatlands, forest fires, drained organic soils, grassland fires, and Coastal Wetlands Remaining Coastal Wetlands, Land
Converted to Coastal Wetlands, Flooded Land Remaining Flooded Land, and Land Converted to Flooded Land.
Significant trends for the largest sources of U.S. CH4 emissions include the following:
• Enteric fermentation was the largest anthropogenic source of CH4 emissions in the United States in 2021,
accounting for 194.9 MMT CO2 Eq. of Cm (26.8 percent of total Cm emissions and 3.1 percent of total
gross emissions) and representing an increase of 11.9 MMT CO2 Eq. (6.5 percent) since 1990. This increase
in emissions from 1990 to 2021 generally follows the increasing trends in cattle populations.
• Natural gas systems were the second largest anthropogenic source category of CH4 emissions in the
United States in 2021, accounting for 181.4 MMT CO2 Eq. of Cm (24.9 percent of total CH4 emissions and
2.9 percent of total gross emissions). Emissions decreased by 33.7 MMT CO2 Eq. (15.7 percent) since 1990
largely due to decreases in emissions from distribution, transmission, and storage.
• Landfills were the third largest anthropogenic source of CH4 emissions in the United States in 2021,
accounting for 122.6 MMT CO2 Eq. (16.9 percent of total CH4 emissions and 1.9 percent of total gross
emissions) and representing a decrease of 75.1 MMT CO2 Eq. (38.0 percent) since 1990, with small year-
to-year increases. This downward trend in emissions coincided with increased landfill gas collection and
control systems, and a reduction of decomposable materials (i.e., paper and paperboard, food scraps, and
yard trimmings) discarded in MSW landfills over the time series.19
Nitrous oxide (N2O) is produced by biological processes that occur in soil and water and by a variety of
anthropogenic activities in the agricultural, energy, industrial, and waste management fields. While total N2O
emissions are much lower than CO2 emissions, N2O is 265 times more powerful than CO2 at trapping heat in the
atmosphere over a 100-year time frame (IPCC 2013). Since 1750, the global atmospheric concentration of N2O has
risen by 23.8 percent (IPCC 2013; NOAA/ESRL 2023c). The main anthropogenic activities producing N2O in the
19 Carbon dioxide emissions from landfills are not included specifically in summing waste sector totals. Net carbon fluxes from
changes in biogenic carbon reservoirs and decay of disposed wood products are accounted for in the estimates for LULUCF.
Nitrous Oxide Emissions
ES-13
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
United States are agricultural soil management, wastewater treatment, stationary fuel combustion, manure
management, fuel combustion in motor vehicles, and nitric acid production (see Figure ES-9).
Figure ES-9: 2021 Sources of N2O Emissions
Management 285
Stationary Combustion
Wastewater T reatment I
Manure Management I
Mobile Combustion
LULUCF Emissions
Nitric Acid Production
Adipic Acid Production
Other Industrial Processes ® CO2
¦ r h„
Composting I
Other Energy I , PFCs, SFe and NF3
Field Burning of Agricultural Residues
20
MMT CO2 Eq.
Note: Other Industrial Processes includes N20 emissions from Caprolactam, Glyoxal, and Glyoxylic Acid Production, Electronics
Industry, and Product Uses. Other Energy includes N20 emissions from Petroleum Systems, Natural Gas Systems, and
Incineration of Waste. LULUCF emissions include N20 emissions reported for Peatlands Remaining Peatlands, forest fires,
drained organic soils, grassland fires, Coastal Wetlands Remaining Coastal Wetlands, forest soils and settlement soils.
Significant trends for the largest sources of U.S. emissions of N2O include the following:
• Agricultural soils were the largest anthropogenic source of N2O emissions in 2021, accounting for 285.2
MMT CO2 Eq., 74.1 percent of N2O emissions and 4.5 percent of total gross greenhouse gas emissions in
the United States. These emissions increased by 6.8 MMT CO2 Eq. (2.5 percent) from 1990 to 2021, but
have fluctuated during that period due to annual variations in weather patterns, fertilizer use, and crop
production.
• Stationary combustion was the second largest source of anthropogenic N2O emissions in 2021, accounting
for 22.1 MMT CO2 Eq. (5.7 percent of N2O emissions) and 0.3 percent of total gross U.S. greenhouse gas
emissions in 2021. Stationary combustion emissions peaked in 2007, and have steadily decreased since
then.
• Wastewater treatment, both domestic and industrial, was the third largest anthropogenic source of N2O
emissions in 2021, accounting for 20.9 MMT CO2 Eq., 5.4 percent of N2O emissions and 0.3 percent of
total gross greenhouse gas emissions in the United States in 2021. Emissions from wastewater treatment
increased by 6.1 MMT CO2 Eq. (41.6 percent) since 1990 as a result of growing U.S. population and protein
consumption. Nitrous oxide emissions from industrial wastewater treatment sources fluctuated
throughout the time series with production changes associated with the treatment of wastewater from
the pulp and paper manufacturing, meat and poultry processing, fruit and vegetable processing, starch-
based ethanol production, petroleum refining, and brewery industries.
• Nitrous oxide emissions from manure management accounted for 17.4 MMT CO2 Eq., 4.5 percent of N2O
emissions and 0.3 percent of total gross greenhouse gas emissions in the United States in 2021. These
emissions increased by 5.0 MMT CO2 Eq. (40.5 percent) from 1990 to 2021. While the industry trend has
been a shift toward liquid systems, driving down the emissions per unit of nitrogen excreted (dry manure
handling systems have greater aerobic conditions that promote N2O emissions), increases in specific
animal populations have driven an increase in overall manure management N2O emissions over the time
series.
14 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
• Nitrous oxide emissions from mobile combustion, the fifth largest source of N2O emissions in 2021,
decreased by 21.3 MMT CO2 Eq. (55.4 percent) from 1990 to 2021, primarily as a result of national vehicle
emissions standards and emission control technologies for on-road vehicles.
HFC, PFC, SF6, and NF3 Emissions
Hydrofluorocarbons (HFCs) are synthetic chemicals that are used as alternatives to ozone depleting substances
(ODS), which are being phased out under the Montreal Protocol and Clean Air Act Amendments of 1990.
Hydrofluorocarbons do not deplete the stratospheric ozone layer and therefore have been used as alternatives
under the Montreal Protocol on Substances that Deplete the Ozone Layer.
Perfluorocarbons (PFCs) are emitted from the production of electronics and aluminum and also (in smaller
quantities) from their use as alternatives to ozone depleting substances. Sulfur hexafluoride (SFs) is emitted from
the manufacturing and use of electrical transmission and distribution equipment as well as the production of
electronics and magnesium. NF3 is emitted from electronics production. HFCs are also emitted during production
of HCFC-22 and electronics (see Figure ES-10).
HFCs, PFCs, SFs, and NF3 are potent greenhouse gases. In addition to having very high global warming potentials,
SFs, NF3, and PFCs have extremely long atmospheric lifetimes, resulting in their essentially irreversible
accumulation in the atmosphere once emitted. Sulfur hexafluoride is the most potent greenhouse gas the IPCC has
evaluated (IPCC 2021).
Figure ES-10: 2021 Sources of HFCs, PFCs, SFe, and NF3 Emissions
HFCs, PFCs, SFe, and NFa as a
Portion of All Emissions
of Ozone Depleting Substances | 172
Electrical Transmission and Distribution
Electronics Industry
HCFC-22 Production
¦ CO2
Magnesium Production and Processing ¦ CH4
¦ N2O
Aluminum Production ¦ HFCs, PFCs, SFe and NFa
0 2 4 6 8 10 12 14 16 18 20
MMT CO2 Eq.
Some significant trends for the largest sources of U.S. HFC, PFC, SF6, and NF3 emissions include the following:
• Hydrofluorocarbon and perfluorocarbon emissions resulting from their use as substitutes for ODS (e.g.,
chlorofluorocarbons [CFCs]) are the largest share of fluorinated emissions (92.1 percent) in 2021 and have
been consistently increasing, from small amounts in 1990 to 172.5 MMT CO2 Eq. in 2021. This increase
was in large part the result of efforts to phase out CFCs and other ODS in the United States.
• Sulfur hexafluoride emissions from electric power transmission and distribution systems decreased by
18.7 MMT CO2 Eq. (75.7 percent) from 1990 to 2021. There are two factors contributing to this decrease:
(1) a sharp increase in the price of SF6 during the 1990s and (2) a growing awareness of the environmental
impact of SF6 emissions through programs such as EPA's SF6 Emission Reduction Partnership for Electric
Power Systems.
• HFC-23 emissions from HCFC-22 production decreased by 36.4 MMT C02 Eq. (94.2 percent) from 1990 to
2021. The decrease from 1990 emissions was caused primarily by a reduction in the HFC-23 emission rate
(kg HFC-23 emitted/kg HCFC-22 produced). The emission rate was lowered by optimizing the production
process and capturing much of the remaining HFC-23 for use or destruction.
• PFC emissions from aluminum production decreased by 18.4 MMT CO2 Eq. (95.3 percent) from 1990 to
2021, due to both industry emission reduction efforts and lower domestic aluminum production.
ES-15
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
ES.3 Overview of Sector Emissions and Trends
Figure ES-11 and Table ES-3 aggregate emissions and sinks by the sectors defined by the UNFCCC reporting
guidelines and methodological framework in the IPCC Guidelines to promote comparability across countries. Over
the thirty-two-year period of 1990 to 2021, total emissions from the Industrial Processes and Product Use and
Agriculture sectors grew by 41.1 MMT CO2 Eq. (12.2 percent), and 50.8 MMT CO2 Eq. (9.4 percent), respectively.
Emissions from the Energy and Waste sectors decreased by 155.6 MMT CO2 Eq. (2.9 percent) and 66.8 MMT CO2
Eq. (28.3 percent) respectively. Over the same period, net carbon (C) sequestration in the LULUCF sector
decreased by 106.8 MMT CO2 (11.4 percent decrease in total net C sequestration), while emissions from the
LULUCF sector (i.e., CH4 and N2O) increased by 19.9 MMT CO2 Eq. (34.4 percent).
Figure ES-11: U.S. Greenhouse Gas Emissions and Sinks by IPCC Sector/Category
¦ LULUCF (emissions) ¦ Agriculture
9,000 ¦ Waste ¦ Energy
¦ Industrial Processes and Product Use ¦ LULUCF (removals)
8,000 — Met Emissions (including LULUCF sinks)
7,000
6,000
5,000
o
£ 4,000
2
s
3,000
2,000
1,000
0
-1,000
Oi-irMn^-mv£)rN.cx5(T»Oi-irMro^-LOvDrvvcz)cr*o->-irMro^m^DrvooCT*Oi-i
O\O^O>O^CT>O^O^O^C*0^OOOOOOOOOO'—1 1—1 1—Ii—It—Ii—li—Ii—1 1—1 1—ifNCM
o^cricricricricncricricricrioooooooooooooooooooooo
1—I i—I »-H '—I ^—I '—I 1—I I i—I rvJ CNl fN fN PJ C\J CM CNl fN CM f\l C\l fN CM fN CM (N fN fN CM ("M CM
Table ES-3: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks by IPCC
Sector/Category (MMT CO2 Eq.)
IPCC Sector/Category
1990
2005
2017
2018
2019
2020
2021
Energy
5,368.2
6,351.8
5,418.8
5,589.7
5,458.3
4,893.8
5,212.5
Industrial Processes and Product Use
335.7
356.1
359.1
362.2
366.8
363.2
376.8
Agriculture
538.4
567.0
601.2
617.8
603.3
586.0
589.3
Waste
236.0
192.1
170.9
173.7
176.0
171.5
169.2
Total Gross Emissions3 (Sources)
6,478.3
7,466.9
6,550.0
6,743.4
6,604.4
6,014.5
6,347.7
LULUCF Sector Net Totalb
(881.0)
(781.1)
(774.2)
(765.1)
(704.0)
(776.2)
(754.2)
Net Emissions (Sources and Sinks)c
5,597.3
6,685.8
5,775.8
5,978.3
5,900.3
5,238.3
5,593.5
a Total emissions without LULUCF.
b The LULUCF Sector Net Total is the sum of all LULUCF CH4 and N20 emissions to the atmosphere plus LULUCF net carbon
stock changes in units of MMT C02 Eq.
c Net emissions with LULUCF.
Notes: Total emissions presented without LULUCF. Net emissions are presented with LULUCF. Totals may not sum due to
independent rounding. Parentheses indicate negative values or sequestration.
16 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Energy
2 The Energy chapter contains emissions of all greenhouse gases resulting from stationary and mobile energy
3 activities including fuel combustion and fugitive fuel emissions, and the use of fossil fuels for non-energy purposes.
4 Energy-related activities, primarily fossil fuel combustion, accounted for the vast majority of U.S. CO2 emissions for
5 the period of 1990 through 2021. Energy-related activities are also responsible for CFU and N2O emissions (41.6
6 percent and 10.3 percent of total U.S. emissions of each gas, respectively). Overall, emission sources in the Energy
7 chapter account for a combined 82.1 percent of total gross U.S. greenhouse gas emissions in 2021.
8 In 2021, 79.3 percent of the energy used in the United States (on a Btu basis) was produced through the
9 combustion of fossil fuels. The remaining 20.7 percent came from other energy sources, such as hydropower,
10 biomass, nuclear, wind, and solar energy (see Figure ES-12).
11 Figure ES-12: 2021 U.S. Energy Consumption by Energy Source (Percent)
12
Nuclear Electric Power
13
14 Industrial Processes and Product Use
15 The Industrial Processes and Product Use (IPPU) chapter contains greenhouse gas emissions generated and
16 emitted as the byproducts of non-energy-related industrial processes, which involve the chemical or physical
17 transformation of raw materials and can release waste gases such as CO2, Cm, N2O, and fluorinated gases (e.g.,
18 HFC-23). These processes include iron and steel production and metallurgical coke production, cement production,
19 petrochemical production, ammonia production, lime production, other process uses of carbonates (e.g., flux
20 stone, flue gas desulfurization, and soda ash consumption not associated with glass manufacturing), nitric acid
21 production, adipic acid production, urea consumption for non-agricultural purposes, aluminum production, HCFC-
22 22 production, glass production, soda ash production, ferroalloy production, titanium dioxide production,
23 caprolactam production, zinc production, phosphoric acid production, lead production, and silicon carbide
24 production and consumption. Most of these industries also emit CO2 from fossil fuel combustion which, in line with
25 IPCC sectoral definitions, is included in the Energy Sector.
26 This chapter also contains emissions resulting from the release of HFCs, PFCs, SF6, and NF3 and other man-made
27 compounds used in industrial manufacturing processes and by end-consumers (e.g., residential and mobile air
28 conditioning). These industries include electronics manufacturing, electric power transmission and distribution,
29 and magnesium metal production and processing. In addition, N2O is used in and emitted by electronics industry
30 and anesthetic and aerosol applications, and CO2 is consumed and emitted through various end-use applications.
31 In 2021, emissions resulting from use of the substitution of ODS (e.g., chlorofluorocarbons [CFCs]) by end-
32 consumers was the largest source of IPPU emissions and accounted for 172.5 MMT CO2 Eq, or 45.8 percent of total
33 IPPU emissions.
ES-17
-------
1 IPPU activities are responsible for 3.4, 0.1, and 5.1 percent of total U.S. CO2, CFU, and N2O emissions respectively as
2 well as for all U.S. emissions of fluorinated gases including HFCs, PFCs, SF6 and NF3. Overall, emission sources in the
3 IPPU chapter accounted for 5.9 percent of U.S. greenhouse gas emissions in 2021.
4 Agriculture
5 The Agriculture chapter contains information on anthropogenic emissions from agricultural activities (except fuel
6 combustion, which is addressed in the Energy chapter, and some agricultural CO2, CFU, and N2O fluxes, which are
7 addressed in the Land Use, Land-Use Change, and Forestry chapter).
8 Agricultural activities contribute directly to emissions of greenhouse gases through a variety of processes,
9 including the following sources: agricultural soil management, enteric fermentation in domestic livestock, livestock
10 manure management, rice cultivation, urea fertilization, liming, and field burning of agricultural residues.
11 In 2021, agricultural activities were responsible for emissions of 589.3 MMT CO2 Eq., or 9.3 percent of total gross
12 U.S. greenhouse gas emissions. Methane, N2O, and CO2 are greenhouse gases emitted by agricultural activities.
13 Methane emissions from enteric fermentation and manure management represented 35.8 percent of total CH4
14 emissions from anthropogenic activities in 2021. Agricultural soil management activities, such as application of
15 synthetic and organic fertilizers, deposition of livestock manure, and growing N-fixing plants, were the largest
16 contributors to U.S. N2O emissions in 2021, accounting for 74.1 percent of total N2O emissions. Carbon dioxide
17 emissions from the application of crushed limestone and dolomite (i.e., soil liming) and urea fertilization
18 represented 0.2 percent of total CO2 emissions from anthropogenic activities.
19 Land Use, Land-Use Change, and Forestry
20 The LULUCF chapter contains emissions and removals of CO2 and emissions of CFU and N2O from managed lands in
21 the United States. Consistent with the 2006IPCC Guidelines, emissions and removals from managed lands are
22 considered to be anthropogenic, while emissions and removals from unmanaged lands are considered to be
23 natural.20 The share of managed land in the U.S. is approximately 95 percent of total land included in the
24 Inventory.21 More information on the definition of managed land used in the Inventory is provided in Chapter 6.
25 Overall, the Inventory results show that managed land is a net sink for CO2 (C sequestration). The primary drivers
26 of fluxes on managed lands include forest management practices, tree planting in urban areas, the management of
27 agricultural soils, lands remaining and lands converted to reservoirs and other constructed waterbodies, landfilling
28 of yard trimmings and food scraps, and activities that cause changes in C stocks in coastal wetlands. The main
29 drivers for forest C sequestration include forest growth and increasing forest area (i.e., afforestation), as well as a
30 net accumulation of C stocks in harvested wood pools. The net sequestration in Settlements Remaining
31 Settlements, which occurs predominantly from urban forests (i.e., Settlement Trees) and landfilled yard trimmings
32 and food scraps, is a result of net tree growth and increased urban forest area, as well as long-term accumulation
33 of yard trimmings and food scraps carbon in landfills.
34 The LULUCF sector in 2021 resulted in a net increase in C stocks (i.e., net CO2 removals) of 832.0 MMT CO2 Eq.22
35 The removals of C offset 13.1 percent of total gross greenhouse gas emissions in 2021. Emissions of CH4 and N2O
36 from LULUCF activities in 2021 were 77.8 MMT CO2 Eq. and represent 1.4 percent of net greenhouse gas
20 See http://www.ipcc-negip.iges.or.ip/public/2006el/pdf/4 Volume4/V4 01 Chi lntroduction.pdf.
21 The current land representation does not include land in U.S. Territories, but there are planned improvements to include
these regions in future Inventories. U.S. Territories represent approximately 0.1 percent of the total land base for the United
States. See Box 6-2 in Chapter 6 of this report.
22 LULUCF Carbon Stock Change is the net C stock change from the following categories: Forest Land Remaining Forest Land,
Land Converted to Forest Land, Cropland Remaining Cropland, Land Converted to Cropland, Grassland Remaining Grassland,
Land Converted to Grassland, Wetlands Remaining Wetlands, Land Converted to Wetlands, Settlements Remaining
Settlements, and Land Converted to Settlements.
18 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 emissions.23 Carbon dioxide removals from C stock changes are presented in Table ES-4 along with CH4 and N2O
2 emissions for LULUCF source categories.
3 Between 1990 and 2021, total C sequestration in the LULUCF sector decreased by 11.4 percent, primarily due to a
4 decrease in the rate of net C accumulation in forests and Cropland Remaining Cropland, as well as an increase in
5 CO2 emissions from Land Converted to Settlements. The overall net flux from LULUCF (i.e., net sum of all CH4 and
6 N2O emissions to the atmosphere plus LULUCF net carbon stock changes in units of MMT CO2 eq.) resulted in a
7 removal of 754.2 MMT C02 Eq. in 2021.
8 Flooded lands were the largest source of CH4 emissions from the LULUCF sector in 2021, totaling 45.4 MMT CO2
9 Eq. (1,623 kt of CH4). Forest fires were the second largest source and resulted in CH4 emissions of 15.5 MMT CO2
10 Eq. (554 kt of CH4), followed by Coastal Wetlands Remaining Coastal Wetlands with CH4 emissions of 4.3 MMT CO2
11 Eq. (154 kt of CH4).
12 Forest fires were the largest source of N2O emissions from the LULUCF sector in 2021, totaling 8.9 MMT CO2 Eq.
13 (34 kt of N2O). Nitrous oxide emissions from fertilizer application to settlement soils in 2021 totaled 2.1 MMT CO2
14 Eq. (8 kt of N2O).
15 Table ES-4: U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use, Land-
16 Use Change, and Forestry (MMT CO2 Eq.)
Land-Use Category
1990
2005
2017
2018
2019
2020
2021
Forest Land Remaining Forest Land3
(815.8)
(695.4)
(695.2)
(692.9)
(638.1)
(684.0)
(670.5)
Land Converted to Forest Landb
(98.5)
(98.4)
(98.3)
(98.3)
(98.3)
(98.3)
(98.3)
Cropland Remaining Cropland
(23.2)
(29.0)
(22.3)
(16.6)
(14.5)
(23.3)
(18.9)
Land Converted to Cropland0
54.8
54.7
56.6
56.3
56.3
56.7
56.5
Grassland Remaining Grasslandd
00
00
11.7
11.6
11.9
14.6
6.7
10.6
Land Converted to Grassland0
(6.7)
(40.1)
(24.5)
(24.2)
(23.3)
(25.9)
(24.7)
Wetlands Remaining Wetlandse
41.5
43.1
41.8
41.8
41.8
41.8
41.8
Land Converted to Wetlandse
3.3
1.4
0.8
0.8
0.8
0.6
0.6
Settlements Remaining Settlements'
(107.8)
(113.9)
(125.6)
(125.0)
(124.5)
(131.6)
(132.5)
Land Converted to Settlements0
62.5
85.0
80.9
81.0
81.1
81.0
81.0
LULUCF Carbon Stock Change^ (938.9) (853.5) (842.5) (829.5) (768.2) (852.5) (832.0)
LULUCF Emissions11 57.9 72.4 68.3 64.4 64.2 76.4 77.8
ch4
53.5
61.3
60.1
57.3
56.9
65.4
66.0
n2o
4.4
11.1
8.3
7.0
7.3
11.0
11.8
LULUCF Sector Net Total1
(881.0)
(781.1)
(774.2)
(765.1)
(704.0)
(776.2)
(754.2)
a Includes the net changes to carbon stocks stored in all forest ecosystem pools and harvested wood products, emissions
from fires on both Forest Land Remaining Forest Land and Land Converted to Forest Land, emissions from N fertilizer
additions on both Forest Land Remaining Forest Land and Land Converted to Forest Land, and CH4 and N20 emissions from
drained organic soils on both Forest Land Remaining Forest Land and Land Converted to Forest Land.
b Includes the net changes to carbon stocks stored in all forest ecosystem pools.
c Includes changes in mineral and organic soil carbon stocks for all land use conversions to cropland, grassland, and
settlements, respectively. Also includes aboveground/belowground biomass, dead wood, and litter carbon stock changes
for conversion of forest land to cropland, grassland, and settlements, respectively.
d Estimates include CH4 and N20 emissions from fires on both Grassland Remaining Grassland and Land Converted to
Grassland.
e Estimates include CH4 emissions from Flooded Land Remaining Flooded Land and Land Converted to Flooded Land.
f Estimates include N20 emissions from N fertilizer additions on both Settlements Remaining Settlements and Land
Converted to Settlements because it is not possible to separate the activity data at this time.
g LULUCF Carbon Stock Change includes any C stock gains and losses from all land use and land use conversion categories.
h LULUCF emissions include the CH4 and N20 emissions reported for Peatlands Remaining Peatlands, forest fires, drained
organic soils, grassland fires, and Coastal Wetlands Remaining Coastal Wetlands; CH4 emissions from Land Converted to
23 LULUCF emissions include the CH4 and N20 emissions reported for Peatlands Remaining Peatlands, forest fires, drained
organic soils, grassland fires, and Coastal Wetlands Remaining Coastal Wetlands; CH4 emissions from Land Converted to Coastal
Wetlands; and N20 emissions from forest soils and settlement soils.
ES-19
-------
Coastal Wetlands, Flooded Land Remaining Flooded Land, and Land Converted to Flooded Land; and N20 emissions from
forest soils and settlement soils.
' The LULUCF Sector Net Total is the net sum of all LULUCF CH4 and N20 emissions to the atmosphere plus LULUCF net
carbon stock changes in units of MMT C02 Eq.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
1 Waste
2 The Waste chapter contains emissions from waste management activities (except the incineration of waste, which
3 is addressed in the Energy chapter). Landfills were the largest source of anthropogenic greenhouse gas emissions
4 from waste management activities, generating 122.6 MMT CO2 Eq. and accounting for 72.5 percent of total
5 greenhouse gas emissions from waste management activities, and 16.9 percent of total U.S. CFU emissions.24
6 Additionally, wastewater treatment generated emissions of 42.0 MMT CO2 Eq. and accounted for 24.8 percent of
7 total Waste sector greenhouse gas emissions, 2.9 percent of U.S. CFU emissions, and 5.4 percent of U.S. N2O
8 emissions in 2021. Emissions of CH4 and N2O from composting are also accounted for in this chapter, generating
9 emissions of 2.6 MMT CO2 Eq., and 1.8 MMT CO2 Eq., respectively. Anaerobic digestion at biogas facilities
10 generated CH4 emissions of 0.2 MMT CO2 Eq., accounting for 0.1 percent of emissions from the waste sector.
11 Overall, emission sources accounted for in the Waste chapter generated 169.2 MMT CO2 Eq., or 2.7 percent of
12 total gross U.S. greenhouse gas emissions in 2021.
13 ES.4 Other Information
14 Emissions by Economic Sector
15 Throughout the Inventory of U.S. Greenhouse Gas Emissions and Sinks report, emission estimates are grouped into
16 five sectors (i.e., chapters) defined by the IPCC: Energy, IPPU, Agriculture, LULUCF, and Waste. It is also useful to
17 characterize emissions according to commonly used economic sector categories: residential, commercial, industry,
18 transportation, electric power, and agriculture. Emissions from U.S. Territories are reported as their own end-use
19 sector due to a lack of specific consumption data for the individual end-use sectors within U.S. Territories. For
20 more information on trends in the Land use, Land Use Change and Forestry sector, see Section ES.2 Recent Trends
21 in U.S. Greenhouse Gas Emissions and Sinks.
22 Figure ES-13 shows the trend in emissions by economic sector from 1990 to 2021, and Table ES-5 summarizes
23 emissions from each of these economic sectors.
24 Landfills also store carbon, due to incomplete degradation of organic materials such as harvest wood products, yard
trimmings, and food scraps, as described in the Land Use, Land-Use Change, and Forestry chapter of the Inventory report.
20 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Figure I
2,500
2,000
^ 1,500
o
I-
£ 1,000
500
0
Note: Emissions and removals from Land Use, Land-Use Change, and Forestry are excluded from figure above. Excludes U.S.
Territories.
Table ES-5: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO2 Eq.)
Economic Sectors
1990
2005
2017
2018
2019
2020
2021
Transportation
1,521.4
1,966.0
1,841.6
1,871.3
1,871.7
1,624.9
1,841.7
Electric Power Industry
1,879.7
2,456.9
1,779.2
1,799.1
1,650.5
1,481.8
1,585.4
Industry
1,677.8
1,574.7
1,494.7
1,558.3
1,568.4
1,464.9
1,474.9
Agriculture
583.2
619.5
642.3
658.9
644.2
626.3
630.2
Commercial
447.0
418.9
437.6
453.7
462.0
436.0
429.9
Residential
345.6
371.2
328.4
375.8
382.4
356.9
362.3
U.S. Territories
23.4
59.7
26.3
26.3
25.1
23.5
23.3
Total Gross Emissions (Sources)
6,478.3
7,466.9
6,550.0
6,743.4
6,604.4
6,014.5
6,347.7
LULUCF Sector Net Total3
(881.0)
(781.1)
(774.2)
(765.1)
(704.0)
(776.2)
(754.2)
Net Emissions (Sources and Sinks)
5,597.3
6,685.8
5,775.8
5,978.3
5,900.3
5,238.3
5,593.5
a The LULUCF Sector Net Total is the net sum of all LULUCF CH4 and N20 emissions to the atmosphere plus LULUCF
net carbon stock changes.
Notes: Total (gross) emissions are presented without LULUCF. Total net emissions are presented with LULUCF. Totals
may not sum due to independent rounding. Parentheses indicate negative values or sequestration.
Using this categorization, emissions from transportation activities accounted for the largest portion (29.0 percent)
of total gross U.S. greenhouse gas emissions in 2021. Electric power accounted for the second largest portion (25.0
percent) of U.S. greenhouse gas emissions in 2021, while emissions from industry accounted for the third largest
portion (23.2 percent). Emissions from industry have in general declined over the past decade, due to a number of
factors, including structural changes in the U.S. economy (i.e., shifts from a manufacturing-based to a service-
based economy), fuel switching, and energy efficiency improvements.
The remaining 22.8 percent of U.S. greenhouse gas emissions were contributed by, in order of magnitude, the
agriculture, commercial, and residential sectors, plus emissions from U.S. Territories. Activities related to
agriculture accounted for 9.9 percent of U.S. emissions; unlike other economic sectors, agricultural sector
emissions were dominated by N2O emissions from agricultural soil management and CH4 emissions from enteric
fermentation. An increasing amount of carbon is stored in agricultural soils each year, but this CO2 sequestration is
assigned to the LULUCF sector rather than the agriculture economic sector. The commercial and residential sectors
accounted for 6.8 percent and 5.7 percent of emissions, respectively, and U.S. Territories accounted for 0.4
:-13: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors
ES-21
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
percent of emissions; emissions from these sectors primarily consisted of CO2 emissions from fossil fuel
combustion. Carbon dioxide was also emitted and sequestered by a variety of activities related to forest
management practices, tree planting in urban areas, the management of agricultural soils, landfilling of yard
trimmings, and changes in C stocks in coastal wetlands.
Electricity is ultimately used in the economic sectors described above. Table ES-6 presents greenhouse gas
emissions from economic sectors with emissions related to electric power distributed into end-use categories (i.e.,
emissions from electric power generation are allocated to the economic sectors in which the electricity is used). To
distribute electricity emissions among end-use sectors, emissions from the source categories assigned to electric
power were allocated to the residential, commercial, industry, transportation, and agriculture economic sectors
according to retail sales of electricity for each end-use sector (EIA 20 22).25 These source categories include CO2
from fossil fuel combustion and the use of limestone and dolomite for flue gas desulfurization, CO2 and N2O from
incineration of waste, CFU and N2O from stationary sources, and SF6 from electrical transmission and distribution
systems.
When emissions from electricity use are distributed among these end-use sectors, industrial activities and
transportation account for the largest shares of U.S. greenhouse gas emissions (29.8 percent and 29.1 percent,
respectively) in 2021. The commercial and residential sectors contributed the next largest shares of total gross U.S.
greenhouse gas emissions in 2021 (15.2 and 15.1 percent, respectively). Emissions from the commercial and
residential sectors increase substantially when emissions from electricity use are included, due to their relatively
large share of electricity use for energy (e.g., lighting, cooling, appliances). Figure ES-14 shows the trend in these
emissions by sector from 1990 to 2021.
Table ES-6: U.S. Greenhouse Gas Emissions with Electricity-Related Emissions Distributed
by Economic Sector (MMT CO2 Eq.)
Economic Sectors
1990
2005
2017
2018
2019
2020
2021
Industry
2,351.6
2,290.2
1,974.0
2,033.6
2,011.4
1,852.4
1,894.5
Transportation
1,524.6
1,970.9
1,846.0
1,876.2
1,876.7
1,629.2
1,846.9
Commercial
1,002.4
1,241.0
1,060.4
1,074.5
1,029.7
930.5
963.9
Residential
957.8
1,247.5
962.3
1,034.9
982.0
918.3
955.7
Agriculture
618.4
657.8
681.0
698.1
679.4
660.7
663.4
U.S. Territories
23.4
59.7
26.3
26.3
25.1
23.5
23.3
Total Gross Emissions (Sources)
6,478.3
7,466.9
6,550.0
6,743.4
6,604.4
6,014.5
6,347.7
LULUCF Sector Net Total3
(881.0)
(781.1)
(774.2)
(765.1)
(704.0)
(776.2)
(754.2)
Net Emissions (Sources and Sinks)
5,597.3
6,685.8
5,775.8
5,978.3
5,900.3
5,238.3
5,593.5
a The LULUCF Sector Net Total is the net sum of all LULUCF CH4 and N20 emissions to the atmosphere plus LULUCF net
carbon stock changes.
Notes: Emissions from electric power are allocated based on aggregate electricity use in each end-use sector. Totals
may not sum due to independent rounding. Parentheses indicate negative values or sequestration.
25 U.S. Territories consumption data that are obtained from EIA are only available at the aggregate level and cannot be broken
out by end-use sector. The distribution of emissions to each end-use sector for the 50 states does not apply to territories data.
22 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
Figure ES-14: U.S. Greenhouse Gas Emissions with Electricity-Related Emissions Distributed
to Economic Sectors
2,500
2,000
w 1,500
IN
8
1,000
500
Industry
Transportation
Commercial (Orange)
Residential (Blue)
Agriculture
£
0"i q\ o* a* O"* o &
o^fNm^-mu3rsvcoa>0'^HCMro'id_-. _ . - _
f ) CD t* ^ M f *> CD CD f ^ CD i-H i-H t-H i-H t-H i-H t-H i-H t-H
OOOOOOOOOOOOOOOOOOOOOO
CN(NfNOJ(NfN(NfN(N(N(N(NCN{NtNfN(NtNfN(NfN(N
Note: Emissions and removals from Land Use, Land-Use Change, and Forestry are excluded from figure above. Excludes U.S.
Territories.
Box ES-2: Trends in Various U.S. Greenhouse Gas Emissions-Related Data
Total (gross) greenhouse gas emissions can be compared to other economic and social indices to highlight
changes over time. These comparisons include: (1) emissions per unit of aggregate energy use, because energy-
related activities are the largest sources of emissions; (2) emissions per unit of fossil fuel consumption, because
almost all energy-related emissions involve the combustion of fossil fuels; (3) emissions per unit of total gross
domestic product as a measure of national economic activity; and (4) emissions per capita.
Table ES-7 provides data on various statistics related to U.S. greenhouse gas emissions normalized to 1990 as a
baseline year. These values represent the relative change in each statistic since 1990. Greenhouse gas emissions
in the United States have declined at an average annual rate of 0.02 percent since 1990, although changes from
year to year have been significantly larger. This growth rate is slightly slower than that for total energy use and
fossil fuel consumption, and overall gross domestic product (GDP), and national population (see Figure ES-15).
The direction of these trends started to change after 2005, when greenhouse gas emissions, total energy use,
and fossil fuel consumption began to peak. Greenhouse gas emissions in the United States have decreased at an
average annual rate of 0.9 percent since 2005. Fossil fuel consumption has decreased at a slower rate than
emissions since 2005, while total energy use, GDP, and national population, generally, continued to increase
noting 2020 was impacted by COVID-19 pandemic.
Variable
1990
2005
2017
2018
2019
2020
2021
Avg. Annual
Growth Rate
Since 1990a
Avg. Annual
Growth Rate
Since 2005a
Greenhouse Gas
100
115
101
104
102
93
98
(+)%
-0.9%
Energy Usec
100
119
116
120
119
109
115
0.5%
-0.1%
GDPd
100
159
193
199
203
198
209
2.4%
1.8%
Population6
100
118
¦
130
130
131
133
134
0.9%
0.8%
ES-23
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
+ Absolute value does not exceed 0.05 percent.
a Average annual growth rate.
b Gross total GWP-weighted values.
c Energy content-weighted values (EIA 2022).
d GDP in chained 2009 dollars (BEA 2022).
e U.S. Census Bureau (2021).
Figure ES-15: U.S. Greenhouse Gas Emissions Per Capita and Per Dollar of Gross Domestic
Product (GDP)
Source: BEA (2022), U.S. Census Bureau (2021), and emission estimates in this report.
Key Categories
The 2006IPCC Guidelines (IPCC 2006) defines a key category as a "[category] that is prioritized within the national
inventory system because its estimate has a significant influence on a country's total inventory of greenhouse
gases in terms of the absolute level, the trend, or the uncertainty in emissions and removals."26 A key category
analysis identifies priority source or sink categories for focusing efforts to improve overall Inventory quality. In
addition, a qualitative review of key categories and non-key categories can also help identify additional source and
sink categories to consider for improvement efforts, including reducing uncertainty.
Figure ES-16 presents the 2021 key categories identified by the Approach 1 level assessment, including the LULUCF
sector. A level assessment using Approach 1 identifies all source and sink categories that cumulatively account for
95 percent of total (i.e., gross) emissions in a given year when assessed in descending order of absolute magnitude.
For a complete list of key categories and more information regarding the overall key category analysis, including
approaches accounting for uncertainty and the influence of trends of individual source and sink categories, see the
Introduction chapter, Section 1.5 - Key Categories and Annex 1.
26 See Chapter 4 "Methodological Choice and Identification of Key Categories" in IPCC (2006). See http://www.ipcc-
nggip.iges.or.ip/public/2006gl/voll.html.
24 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
Figure ES-16: 2021 Key Categories (Approach 1 including LULUCF)8
CO2 Emissions from Mobile Combustion: Road
CO2 Emissions from Stationary Combustion - Coal - Electricity Generation
Net Carbon Stock Change from Forest Land Remaining Forest Land
CO2 Emissions from Stationary Combustion - Gas - Electricity Generation
CO2 Emissions from Stationary Combustion - Gas - Industrial
CO2 Emissions from Stationary Combustion - Gas - Residential
Direct N2O Emissions from Agricultural Soil Management
CO2 Emissions from Stationary Combustion - Oil - Industrial
Cm Emissions from Enteric Fermentation: Cattle
CH4 Emissions from Natural Gas Systems
CO2 Emissions from Stationary Combustion - Gas - Commercial
CO2 Emissions from Mobile Combustion: Aviation
CO2 Emissions from Non-Energy Use of Fuels
Emissions from ODS Substitutes: Refrigeration and Air Conditioning
Net Carbon Stock Change from Settlements Remaining Settlements
CH4 Emissions from MSW Landfills
Net Carbon Stock Change from Land Converted to Forest Land
Net Carbon Stock Change from Land Converted to Settlements
CO2 Emissions from Mobile Combustion: Other
Net Carbon Stock Change from Land Converted to Cropland
CO2 Emissions from Stationary Combustion - Oil - Residential
CH4 Emissions from Petroleum Systems
CH4 Emissions from Flooded Land Remaining Flooded Land
Fugitive Emissions from Coal Mining
CO2 Emissions from Stationary Combustion - Coal - Industrial
CO2 Emissions from Iron and Steel Production & Metallurgical Coke Production
CO2 Emissions from Stationaiy Combustion - Oil - Commercial
CO2 Emissions from Cement Production
CO2 Emissions from Mobile Combustion: Marine
CH4 Emissions from Manure Management: Cattle
CO2 Emissions from Natural Gas Systems
CO2 Emissions from Petrochemical Production
CO2 Emissions from Mobile Combustion: Railways
CH4 Emissions from Manure Management: Other Livestock
Indirect N2O Emissions from Applied Nitrogen
Net Carbon Stock Change from Land Converted to Grassland
CO2 Emissions from Petroleum Systems
N2O Emissions from Domestic Wastewater Treatment
Net Carbon Stock Change from Cropland Remaining Cropland
CH4 Emissions from Industrial Landfills
Emissions from Substitutes for Ozone Depleting Substances: Aerosols
CO2 Emissions from Stationary Combustion - Oil - U.S. Territories
CFU Emissions from Rice Cultivation
Net Carbon Stock Change from Grassland Remaining Grassland
CFU Emissions from Abandoned Oil and Gas Wells
CFf# Emissions from Stationary Combustion - Residential
Key Categories as a Portion of
All Emissions
96.0%
I Key Categories LULUCF
Other Categories
I Key Categories
0 200 400 600 800 1,000 1,200 1,400
2021 Emissions (MMT CO2 Eq.)
1 For a complete list of key categories and detailed discussion of the underlying key category analysis, see Annex 1. Bars indicate
key categories identified using Approach 1 level assessment including the LULUCF sector. The absolute values of net C02
emissions from LULUCF are presented in this figure but reported separately from gross emissions totals. Refer to Table ES-4
for a breakout of emissions and removals for LULUCF by source/sink category.
Quality Assurance and Quality Control (QA/QC)
The United States seeks continuous improvements to the quality, transparency, and usability of the Inventory of
U.S. Greenhouse Gas Emissions and Sinks. To assist in these efforts, the United States implemented a systematic
approach to QA/QC. The procedures followed for the Inventory have been formalized in accordance with the U.S.
Inventory QA/QC plan for the Inventory, and the UNFCCC reporting guidelines and 2006IPCC Guidelines. The QA
process includes expert and public reviews for both the Inventory estimates and the Inventory report.
ES-25
-------
1
Box ES-3: Use of Ambient Measurements Systems for Validation of Emission Inventories
In following the UNFCCC requirement under Article 4.1 to develop and submit national greenhouse gas emission
inventories, the emissions and sinks presented in this report are organized by source and sink categories and
calculated using internationally accepted methods provided by the IPCC.27 Several recent studies have
estimated emissions at the national or regional level with estimated results that sometimes differ from EPA's
estimate of emissions. EPA has engaged with researchers on how remote sensing, ambient measurement, and
inverse modeling techniques for estimating greenhouse gas emissions could assist in improving the
understanding of inventory estimates. In working with the research community to improve national greenhouse
gas inventories, EPA follows guidance from the IPCC on the use of measurements and modeling to validate
emission inventories.28 An area of particular interest in EPA's outreach efforts is how ambient measurement
data can be used to assess estimates or potentially be incorporated into the Inventory in a manner consistent
with this Inventory report's transparency of its calculation methodologies, and the ability of inverse modeling
techniques to attribute emissions and removals from remote sensing to anthropogenic sources, as defined by
the IPCC for this report, versus natural sources and sinks.
The 2019 Refinement to the IPCC 2006 Guidelines for National Greenhouse Gas Inventories (IPCC 2019) Volume
1 General Guidance and Reporting, Chapter 6: Quality Assurance, Quality Control and Verification notes that
atmospheric concentration measurements can provide independent data sets as a basis for comparison with
inventory estimates. The 2019 Refinement provides guidance on conducting such comparisons (as summarized
in Table 6.2 of IPCC [2019] Volume 1, Chapter 6) and provides guidance on using such comparisons to identify
areas of improvement in national inventories (as summarized in Box 6.5 of IPCC [2019] Volume 1, Chapter 6)
given the technical complexity of such comparisons. Further, it identified fluorinated gases as particularly
suitable for such comparisons. The 2019 Refinement makes this conclusion for fluorinated gases based on their
lack of significant natural sources, their generally long atmospheric lifetimes, their well-known loss mechanisms,
and the potential uncertainties in bottom-up inventory methods for some of their sources. Unlike emissions of
CO2, Cm, and N2O, emissions of fluorinated greenhouse gases are almost exclusively anthropogenic, meaning
that the fluorinated GHG emission sources included in this Inventory account for the majority of the total U.S.
emissions of these gases detectable in the atmosphere.
In this Inventory, EPA presents the results of two comparisons between fluorinated gas emissions inferred from
atmospheric measurements and fluorinated gas emissions estimated based on bottom-up measurements and
modeling. These comparisons, performed for HFCs and SF6 respectively, are described under the QA/QC and
Verification discussions in Chapter 4, Sections 4.24 Substitution of Ozone Depleting Substances and 4.25
Electrical Transmission and Distribution in the IPPU chapter of this report.
Consistent with the 2019 Refinement, a key element to facilitate such comparisons is a gridded prior inventory
as an input to inverse modeling. To improve the ability to compare the national-level greenhouse gas inventory
with measurement results that may be at other scales, a team at Harvard University along with EPA and other
coauthors developed a gridded inventory of U.S. anthropogenic methane emissions with 0.1° x 0.1° spatial
resolution, monthly temporal resolution, and detailed scale-dependent error characterization. The gridded
inventory is designed to be consistent with the 1990 to 2014 U.S. EPA Inventory of U.S. Greenhouse Gas
Emissions and Sinks estimates for the year 2012, which presents national totals for different source types.29 This
gridded inventory is consistent with the recommendations contained in two National Academies of Science
reports examining greenhouse gas emissions data (National Research Council 2010; National Academies of
Sciences, Engineering, and Medicine 2018).
27 See http://www.ipcc-negip.iges.or.jp/public/index.html.
28 See http://www.ipcc-nggip.iges.or.jp/meeting/pdfiles/1003 Uncertaintv%20meeting report.pdf.
29 See https://www.epa.gov/ghgemissions/gridded-2012-methane-emissions.
26 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Finally, in addition to use of atmospheric concentration measurement data for comparison with Inventory data,
information from top-down studies is directly incorporated in the Natural Gas Systems calculations to quantify
emissions from certain well blowout events.
Uncertainty Analysis of Emission and Sink Estimates
Uncertainty assessment is an essential element of a complete inventory of greenhouse gas emissions and removals
because it helps to inform and prioritize inventory improvements. Recognizing the benefit of conducting an
uncertainty analysis, the UNFCCC reporting guidelines follow the recommendations of the 2006IPCC Guidelines
(IPCC 2006), Volume 1, Chapter 3 and require that countries provide single estimates of uncertainty for source and
sink categories. In addition to quantitative uncertainty assessments, a qualitative discussion of uncertainty is
presented for each source and sink category identifying specific factors affecting the uncertainty surrounding the
estimates provided in accordance with UNFCCC reporting guidelines. Some of the current estimates, such as those
for CO2 emissions from energy-related combustion activities, are considered to have low uncertainties. This is
because the amount of CO2 emitted from energy-related combustion activities is directly related to the amount of
fuel consumed, the fraction of the fuel that is oxidized, and the carbon content of the fuel, and for the United
States, the uncertainties associated with estimating those factors is relatively small. For some other categories of
emissions and sinks, however, inherent variability or a lack of data increases the uncertainty or systematic error
associated with the estimates presented. Finally, an analysis is conducted to assess uncertainties associated with
the overall emissions, sinks and trends estimates. The overall uncertainty surrounding total net greenhouse gas
emissions is estimated to be -5 to +6 percent in 1990 and -6 to +6 percent in 2020. When the LULUCF sector is
excluded from the analysis the uncertainty is estimated to be -2 to +5 percent in 1990 and -3 to +3 percent in 2020.
ES-27
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
1.
Introduction
This report presents an inventory of U.S. anthropogenic greenhouse gas emissions and sinks for the years 1990
through 2021 compiled by the United States government. A summary of these estimates is provided in Table 2-1
and Table 2-2 by gas and source category in the Trends in Greenhouse Gas Emissions chapter. The emission and
sink estimates in these tables are presented on both a full mass basis and on a global warming potential (GWP)-
weighted basis1 in order to show the relative contribution of each gas to global average radiative forcing. This
report also discusses the methods and data used to calculate the emission and sink estimates.
In 1992, the United States signed and ratified the United Nations Framework Convention on Climate Change
(UNFCCC). As stated in Article 2 of the UNFCCC, 'The ultimate objective of this Convention and any related legal
instruments that the Conference of the Parties may adopt is to achieve, in accordance with the relevant provisions
of the Convention, stabilization of greenhouse gas concentrations in the atmosphere at a level that would prevent
dangerous anthropogenic interference with the climate system. Such a level should be achieved within a time-
frame sufficient to allow ecosystems to adapt naturally to climate change, to ensure that food production is not
threatened and to enable economic development to proceed in a sustainable manner."2'3
As a signatory to the UNFCCC, consistent with Article 44 and decisions at the First, Second, Fifth, and Nineteenth
Conference of Parties,5 the U.S. is committed to submitting a national inventory of anthropogenic sources and
sinks of greenhouse gases to the UNFCCC by April 15 of each year. This Inventory provides a national estimate of
sources and sinks for the United States, including all states, the District of Columbia and U.S. Territories.6 The
United States views this report, in conjunction with Common Reporting Format (CRF) reporting tables that
accompany this report, as an opportunity to fulfill this annual commitment under the UNFCCC. Overall, this
1 More information provided in the Global Warming Potentials section of this chapter on the use of IPCC Fifth Assessment
Report (AR5) GWP values.
2 The term "anthropogenic," in this context, refers to greenhouse gas emissions and removals that are a direct result of human
activities or are the result of natural processes that have been affected by human activities (IPCC 2006).
3 Article 2 of the Framework Convention on Climate Change published by the UNEP/WMO Information Unit on Climate Change
(UNEP/WMO 2000). See http://unfccc.int.
4 Article 4(l)(a) of the United Nations Framework Convention on Climate Change (also identified in Article 12) and subsequent
decisions by the Conference of the Parties elaborated the role of Annex I Parties in preparing national inventories. Article 4
states "Parties to the Convention, by ratifying, shall develop, periodically update, publish and make available...national
inventories of anthropogenic emissions by sources and removals by sinks of all greenhouse gases not controlled by the
Montreal Protocol, using comparable methodologies..." See http://unfccc.int for more information.
5 See UNFCCC decisions 3/CP.l, 9/CP.2, 3/CP.5, and 24/CP.19 at https://unfccc.int/documents.
6 U.S. Territories include American Samoa, Guam, Commonwealth of the Northern Mariana Islands, Puerto Rico, U.S. Virgin
Islands, and other outlying U.S. Pacific Islands which are not permanently inhabited such as Wake Island. See
https://www.usgs.Eov/faas/how-are-us-states-territories-and-commonwealths-designated-geographic-names-information-
svstem?qt-news science products=Q#qt-news science products. See more information on completeness in Section 1.8.
Introduction 1-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Inventory of anthropogenic greenhouse gas emissions and sinks provides a common and consistent mechanism
through which Parties to the UNFCCC can compare the relative contribution of individual sources, gases, and
nations to climate change. The structure of this report is consistent with the current UNFCCC Guidelines on Annual
Inventories (UNFCCC 2014) for Parties included in Annex I of the Convention.
In 1988, preceding the creation of the UNFCCC, the World Meteorological Organization (WMO) and the United
Nations Environment Programme (UNEP) jointly established the Intergovernmental Panel on Climate Change
(IPCC). The role of the IPCC is to assess on a comprehensive, objective, open and transparent basis the scientific,
technical and socio-economic information relevant to understanding the scientific basis of risk of human-induced
climate change, its potential impacts and options for adaptation and mitigation (IPCC 2021). Under Working Group
1 of the IPCC, nearly 140 scientists and national experts from more than thirty countries collaborated in the
creation of the Revised 1996 IPCC Guidelines for National Greenhouse Gas Inventories (IPCC/UNEP/OECD/IEA 1997)
to ensure that the emission inventories submitted to the UNFCCC are consistent and comparable between nations.
The IPCC Good Practice Guidance and Uncertainty Management in National Greenhouse Gas Inventories and the
IPCC Good Practice Guidance for Land Use, Land-Use Change, and Forestry further expanded upon the
methodologies in the Revised 1996 IPCC Guidelines. In 2006, the IPCC accepted the 2006 Guidelines for National
Greenhouse Gas Inventories at its Twenty-Fifth Session (Mauritius, April 2006). The 2006 IPCC Guidelines built upon
the previous bodies of work and include new sources and gases, "...as well as updates to the previously published
methods whenever scientific and technical knowledge have improved since the previous guidelines were issued."
The UNFCCC adopted the 2006 IPCC Guidelines as the standard methodological approach for Annex I countries and
encouraged countries to gain experience in using the 2013 Supplement to the 2006 IPCC Guidelines for National
Greenhouse Gas Inventories: Wetlands at the Nineteenth Conference of the Parties (Warsaw, November 11-23,
2013). The IPCC has recently released the 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse
Gas Inventories to clarify and elaborate on the existing guidance in the 2006 IPCC Guidelines, along with providing
updates to default values of emission factors and other parameters based on updated science. This report applies
both the 2013 Supplement and updated guidance in the 2019 Refinement to improve accuracy and completeness
of the Inventory. For more information on specific uses see Section 1.4 of this chapter on Methodology and Data
Sources.
Box 1-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals, including
Relationship to EPA's Greenhouse Gas Reporting Program
In following the UNFCCC requirement under Article 4.1 and decision 24/CP.19 to develop and submit annual
national greenhouse gas emission inventories, the emissions and removals presented in this report and this
chapter are organized by source and sink categories and calculated using internationally-accepted methods
provided by the IPCC in the 2006 IPCC Guidelines for National Greenhouse Gas Inventories (2006 IPCC
Guidelines) and where appropriate, its supplements and refinements. Additionally, the calculated emissions and
removals in a given year for the United States are presented in a common format in line with the UNFCCC
reporting guidelines for the reporting of inventories under this international agreement. The use of consistent
methods to calculate emissions and removals by all nations reporting their inventories to the UNFCCC ensures
that the estimates are comparable. The presentation of emissions and removals provided in this Inventory does
not preclude alternative examinations, but rather this Inventory presents emissions and removals in a common
format consistent with how countries are to report Inventories under the UNFCCC. The report itself, and this
chapter, follows this standardized format, and provides an explanation of the application of methods used to
calculate emissions and removals.
EPA also collects greenhouse gas emissions data from individual facilities and suppliers of certain fossil fuels and
industrial gases through its Greenhouse Gas Reporting Program (GHGRP).7 The GHGRP applies to direct
greenhouse gas emitters, fossil fuel suppliers, industrial gas suppliers, and facilities that inject carbon dioxide
7 On October 30, 2009, the EPA promulgated a rule requiring annual reporting of greenhouse gas data from large greenhouse
gas emissions sources in the United States. Implementation of the rule, codified at 40 CFR Part 98, is referred to as EPA's
Greenhouse Gas Reporting Program (GHGRP).
1-2 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
(C02) underground for sequestration or other reasons and requires reporting by over 8,000 sources or suppliers
in 41 industrial categories.8 Annual reporting is at the facility level, except for certain suppliers of fossil fuels and
industrial greenhouse gases. In general, the threshold for reporting is 25,000 metric tons or more of C02 Eq. per
year. Facilities in most source categories subject to GHGRP began reporting for the 2010 reporting year while
additional types of industrial operations began reporting for reporting year 2011. While the GHGRP does not
provide full coverage of total annual U.S. greenhouse gas emissions and sinks (e.g., the GHGRP excludes
emissions from the agricultural, land use, and forestry sectors), it is an important input to the calculations of
national-level emissions in the Inventory.
Data presented in this Inventory report and EPA's GHGRP are complementary. The GHGRP dataset continues to
be an important resource for the Inventory, providing not only annual emissions information, but also other
annual information such as activity data and emission factors that can improve and refine national emission
estimates and trends over time. Methodologies used in EPA's GHGRP are consistent with the 2006IPCC
Guidelines (e.g., higher tier methods). GHGRP data also allow EPA to disaggregate national inventory estimates
in new ways that can highlight differences across regions and sub-categories of emissions, along with enhancing
the application of QA/QC procedures and assessment of uncertainties. EPA uses annual GHGRP data in a
number of categories to improve the national estimates presented in this Inventory consistent with IPCC
methodological guidance. See Annex 9 for more information on specific uses of GHGRP data in the Inventory
(e.g., natural gas systems).
1
2 1.1 Background Information
3 Science
4 For over the past 200 years, the burning of fossil fuels such as coal and oil, along with deforestation, land-use
5 changes, and other activities have caused the concentrations of heat-trapping "greenhouse gases" to increase
6 significantly in our atmosphere (IPCC 2021). These gases in the atmosphere absorb some of the energy being
7 radiated from the surface of the Earth that would otherwise be lost to space, essentially acting like a blanket that
8 makes the Earth's surface warmer than it would be otherwise.
9 Greenhouse gases are necessary to life as we know it. Without greenhouse gases to create the natural heat-
10 trapping properties of the atmosphere, the planet's surface would be about 60 degrees Fahrenheit cooler than
11 present (USGCRP 2017). Carbon dioxide is also necessary for plant growth. With emissions from biological and
12 geological sources, there is a natural level of greenhouse gases that is maintained in the atmosphere. Human
13 emissions of greenhouse gases and subsequent changes in atmospheric concentrations alter the balance of energy
14 transfers between space and the earth system (IPCC 2021). A gauge of these changes is called radiative forcing,
15 which is a measure of a substance's total net effect on the global energy balance for which a positive number
16 represents a warming effect, and a negative number represents a cooling effect (IPCC 2021). IPCC concluded in its
17 most recent scientific assessment report that it is "unequivocal that human influence has warmed the atmosphere,
18 ocean and land" (IPCC 2021).
19 As concentrations of greenhouse gases continue to increase in from man-made sources, the Earth's temperature is
20 climbing above past levels. The Earth's average land and ocean surface temperature has increased by about 2.0
21 degrees Fahrenheit from the 1850 to 1900 period to the decade of 2011 to 2020 (IPCC 2021). The last four decades
22 have each been the warmest decade successively at the Earth's surface since at least 1850 (IPCC 2021). Other
23 aspects of the climate are also changing, such as rainfall patterns, snow and ice cover, and sea level. If greenhouse
24 gas concentrations continue to increase, climate models predict that the average temperature at the Earth's
8 See http://www.epa.gov/ghgreporting and http://ghgdata.epa.gov/ghgp/main.do.
Introduction 1-3
-------
1 surface is likely to increase by up to 8.3 degrees Fahrenheit above 2011 to 2020 levels by the end of this century,
2 depending on future emissions and the responsiveness of the climate system (IPCC 2021), though the lowest
3 emission scenario would limit future warming to an additional 0.5 degrees (best estimate).
4 For further information on greenhouse gases, radiative forcing, and implications for climate change, see the recent
5 scientific assessment reports from the IPCC,9 the U.S. Global Change Research Program (USGCRP),10 and the
6 National Academies of Sciences, Engineering, and Medicine (NAS).11
7 Greenhouse Gases
8 Although the Earth's atmosphere consists mainly of oxygen and nitrogen, neither plays a significant role in
9 enhancing the greenhouse effect because both are essentially transparent to terrestrial radiation. The greenhouse
10 effect is primarily a function of the concentration of water vapor, carbon dioxide (CO2), methane (CH4), nitrous
11 oxide (N2O), and other trace gases in the atmosphere that absorb the terrestrial radiation leaving the surface of
12 the Earth (IPCC 2021).
13 Naturally occurring greenhouse gases include water vapor, CO2, CH4, N2O, and ozone (O3). Several classes of
14 halogenated substances that contain fluorine, chlorine, or bromine are also greenhouse gases, but they are, for the
15 most part, solely a product of industrial activities. Chlorofluorocarbons (CFCs) and hydrochlorofluorocarbons
16 (HCFCs) are halocarbons that contain chlorine, while halocarbons that contain bromine are referred to as
17 bromofluorocarbons (i.e., halons). As stratospheric ozone depleting substances, CFCs, HCFCs, and halons are
18 covered under the Montreal Protocol on Substances that Deplete the Ozone Layer. The UNFCCC defers to this
19 earlier international treaty. Consequently, Parties to the UNFCCC are not required to include these gases in
20 national greenhouse gas inventories.12 Some other fluorine-containing halogenated substances—
21 hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), sulfur hexafluoride (SFs), and nitrogen trifluoride (NF3)—do
22 not deplete stratospheric ozone but are potent greenhouse gases. These latter substances are addressed by the
23 UNFCCC and accounted for in national greenhouse gas inventories.
24 There are also several other substances that influence the global radiation budget but are short-lived and
25 therefore not well-mixed, leading to spatially variable radiative forcing effects. These substances include carbon
26 monoxide (CO), nitrogen dioxide (NO2), sulfur dioxide (SO2), and tropospheric (ground level) ozone (O3).
27 Tropospheric ozone is formed from chemical reactions in the atmosphere of precursor pollutants, which include
28 volatile organic compounds (VOCs, including CH4) and nitrogen oxides (NOx), in the presence of ultraviolet light
29 (sunlight).
30 Aerosols are extremely small particles or liquid droplets suspended in the Earth's atmosphere that are often
31 composed of sulfur compounds, carbonaceous combustion products (e.g., black carbon), crustal materials (e.g.,
32 dust) and other human-induced pollutants. They can affect the absorptive characteristics of the atmosphere (e.g.,
33 scattering incoming sunlight away from the Earth's surface, or, in the case of black carbon, absorb sunlight) and
34 can play a role in affecting cloud formation and lifetime, as well as the radiative forcing of clouds and precipitation
35 patterns.
36 Carbon dioxide, CFU, and N2O are continuously emitted to and removed from the atmosphere by natural processes
37 on Earth. Anthropogenic activities (such as fossil fuel combustion, cement production, land-use, land-use change,
38 and forestry, agriculture, or waste management), however, can cause additional quantities of these and other
39 greenhouse gases to be emitted or sequestered, thereby changing their global average atmospheric
40 concentrations. Natural activities such as respiration by plants or animals and seasonal cycles of plant growth and
9 See https://www.ipcc.ch/report/ar6/wgl/.
10 See https://nca2018.globalchange.gov/.
11 See https://www.nationalacademies.org/topics/climate.
12 Emissions estimates of CFCs, HCFCs, halons and other ozone-depleting substances are included in this document for
informational purposes.
1-4 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
decay are examples of processes that only cycle carbon or nitrogen between the atmosphere and organic biomass.
Such processes, except when directly or indirectly perturbed out of equilibrium by anthropogenic activities,
generally do not alter average atmospheric greenhouse gas concentrations over decadal timeframes. Climatic
changes resulting from anthropogenic activities, however, could have positive or negative feedback effects on
these natural systems. Atmospheric concentrations of these gases, along with their rates of growth and
atmospheric lifetimes, are presented in Table 1-1.
Table 1-1: Global Atmospheric Concentration, Rate of Concentration Change, and
Atmospheric Lifetime of Selected Greenhouse Gases
Atmospheric Variable
C02
ch4
n2o
sf6
cf4
Pre-industrial atmospheric concentration
280 ppm
0.730 ppm
0.270 ppm
Oppt
40 ppt
Atmospheric concentration
419 ppma
1.895 ppmb
0.334 ppmc
11.08 pptd
85.5 ppte
Rate of concentration change
2.38 ppm/yrf
18.21 ppb/yrf'g
1.29 ppb/yrf
0.39 ppt/yrf
0.81 ppt/yrf
Atmospheric lifetime (years)
See footnote11
11.8
1091
About l,000i
50,000
a The atmospheric C02 concentration is the 2021 annual average at the Mauna Loa, HI station (NOAA/ESRL 2023a). The global
atmospheric C02 concentration, computed using an average of sampling sites across the world, was 415 ppm in 2021.
b The values presented are global 2022 annual average mole fractions (NOAA/ESRL 2023b).
c The values presented are global 2022 annual average mole fractions (NOAA/ESRL 2023c).
d The values presented are global 2022 annual average mole fractions (NOAA/ESRL 2023d).
e The 2019 CF4 global mean atmospheric concentration is from the Advanced Global Atmospheric Gases Experiment (IPCC 2021).
f The rate of concentration change for C02 is an average of the rates from 2007 through 2021 and has fluctuated between 1.5 to
3.0 ppm per year over this period (NOAA/ESRL 2023a). The rate of concentration change for CH4, N20, and SF6, is the average
rate of change between 2007 and 2021 (NOAA/ESRL 2023b; NOAA/ESRL 2023c; NOAA/ESRL 2023d). The rate of concentration
change for CF4 is the average rate of change between 2011 and 2019 (IPCC 2021).
s The growth rate for atmospheric CH4 decreased from over 10 ppb/year in the 1980s to nearly zero in the early 2000s; recently,
the growth rate has been about 18.21 ppb/year (NOAA/ESRL 2023b).
h For a given amount of C02 emitted, some fraction of the atmospheric increase in concentration is quickly absorbed by the
oceans and terrestrial vegetation, some fraction of the atmospheric increase will only slowly decrease over a number of years,
and a small portion of the increase will remain for many centuries or more.
' This lifetime has been defined as an "adjustment time" that takes into account the indirect effect of the gas on its own
residence time.
i The lifetime for SF6 was revised from 3,200 years to about 1,000 years based on recent studies (IPCC 2021).
Source: Pre-industrial atmospheric concentrations and atmospheric lifetimes for CH4, N20, SF6, and CF4 are from IPCC (2021).
A brief description of each greenhouse gas, its sources, and its role in the atmosphere is given below. The following
section then explains the concept of GWPs, which are assigned to individual gases as a measure of their relative
average global radiative forcing effect.
Water Vapor (H2O). Water vapor is the largest contributor to the natural greenhouse effect. Water vapor is
fundamentally different from other greenhouse gases in that it can condense and rain out when it reaches high
concentrations, and the total amount of water vapor in the atmosphere is in part a function of the Earth's
temperature. While some human activities such as evaporation from irrigated crops or power plant cooling release
water vapor into the air, these activities have been determined to have a negligible effect on global climate (IPCC
2021). The lifetime of water vapor in the troposphere is on the order of 10 days. Water vapor can also contribute
to cloud formation, and clouds can have both warming and cooling effects by either trapping or reflecting heat.
Because of the relationship between water vapor levels and temperature, water vapor and clouds serve as a
feedback to climate change, such that for any given increase in other greenhouse gases, the total warming is
greater than would happen in the absence of water vapor. Aircraft emissions of water vapor can create contrails,
which may also develop into contrail-induced cirrus clouds, with complex regional and temporal net radiative
forcing effects that currently have a low level of scientific certainty (IPCC 2021).
Carbon Dioxide (CCh). In nature, carbon is cycled between various atmospheric, oceanic, land biotic, marine biotic,
and mineral reservoirs. The largest fluxes occur between the atmosphere and terrestrial biota, and between the
atmosphere and surface water of the oceans. In the atmosphere, carbon predominantly exists in its oxidized form
as CO2. Atmospheric CO2 is part of this global carbon cycle, and therefore its fate is a complex function of
Introduction 1-5
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
geochemical and biological processes. Carbon dioxide concentrations in the atmosphere increased from
approximately 280 parts per million by volume (ppmv) in pre-industrial times to 415 ppmv in 2021, a 48 percent
increase (IPCC 2021; NOAA/ESRL 2023a).1314 The IPCC states that "Observed increases in well-mixed greenhouse
gas (GHG) concentrations since around 1750 are unequivocally caused by human activities" (IPCC 2021). The
predominant source of anthropogenic C02 emissions is the combustion of fossil fuels. Forest clearing, other
biomass burning, and some non-energy production processes (e.g., cement production) also emit notable
quantities of C02. In its Sixth Assessment Report, the IPCC determined that of the 2.0 degrees of observed warming,
the best estimate is that 1.9 degrees of that are due to human influence, with elevated CO2 concentrations being
the most important contributor to that warming (IPCC 2021).
Methane (CHa). Methane is primarily produced through anaerobic decomposition of organic matter in biological
systems. Agricultural processes such as wetland rice cultivation, enteric fermentation in animals, and the
decomposition of animal wastes emit CH4, as does the decomposition of municipal solid wastes and treatment of
wastewater. Methane is also emitted during the production and distribution of natural gas and petroleum, and is
released as a byproduct of coal mining and incomplete fossil fuel combustion. Atmospheric concentrations of CH4
have increased by about 162 percent since 1750, from a pre-industrial value of about 730 ppb to 1,895 ppb in
202115 although the rate of increase decreased to near zero in the early 2000s, and has recently increased again to
about 18.12 ppb/year. The IPCC has estimated that about half of the current CH4 flux to the atmosphere (and the
entirety of the increase in concentration) is anthropogenic, from human activities such as agriculture, fossil fuel
production and use, and waste disposal (IPCC 2021).
Methane is primarily removed from the atmosphere through a reaction with the hydroxyl radical (OH) and is
ultimately converted to CO2. Minor removal processes also include reaction with chlorine in the marine boundary
layer, a soil sink, and stratospheric reactions. Increasing emissions of CH4 reduce the concentration of OH, a
feedback that increases the atmospheric lifetime of CH4 (IPCC 2021). Methane's reactions in the atmosphere also
lead to production of tropospheric ozone and stratospheric water vapor, both of which also contribute to climate
change. Tropospheric ozone also has negative effects on human health and plant productivity.
Nitrous Oxide (N2O). Anthropogenic sources of N20 emissions include agricultural soils, especially production of
nitrogen-fixing crops and forages, the use of synthetic and manure fertilizers, and manure deposition by livestock;
fossil fuel combustion, especially from mobile combustion; adipic (nylon) and nitric acid production; wastewater
treatment and waste incineration; and biomass burning. The atmospheric concentration of N20 has increased by
24 percent since 1750, from a pre-industrial value of about 270 ppb to 334 ppb in 2021,16 a concentration that has
not been exceeded during at least the last 800 thousand years. Nitrous oxide is primarily removed from the
atmosphere by the photolytic action of sunlight in the stratosphere (IPCC 2021).
Ozone (O3). Ozone is present in both the upper stratosphere,17 where it shields the Earth from harmful levels of
ultraviolet radiation, and at lower concentrations in the troposphere,18 where it is the main component of
anthropogenic photochemical "smog." During the last two decades, emissions of anthropogenic chlorine and
bromine-containing halocarbons, such as CFCs, have depleted stratospheric ozone concentrations. This loss of
13 The pre-industrial period is considered as the time preceding the year 1750 (IPCC 2013).
14 Carbon dioxide concentrations during the last 1,000 years of the pre-industrial period (i.e., 750 to 1750), a time of relative
climate stability, fluctuated by about +10 ppmv around 280 ppmv (IPCC 2013).
15 This value is the global 2021 annual average mole fraction (NOAA/ESRL 2023b).
16 This value is the global 2021 annual average (NOAA/ESRL 2023c).
17 The stratosphere is the layer from the troposphere up to roughly 50 kilometers. In the lower regions the temperature is
nearly constant but in the upper layer the temperature increases rapidly because of sunlight absorption by the ozone layer. The
ozone-layer is the part of the stratosphere from 19 kilometers up to 48 kilometers where the concentration of ozone reaches
up to 10 parts per million.
18 The troposphere is the layer from the ground up to 11 kilometers near the poles and up to 16 kilometers in equatorial
regions (i.e., the lowest layer of the atmosphere where people live). It contains roughly 80 percent of the mass of all gases in
the atmosphere and is the site for most weather processes, including most of the water vapor and clouds.
1-6 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
ozone in the stratosphere has resulted in negative radiative forcing, representing an indirect effect of
anthropogenic emissions of chlorine and bromine compounds (IPCC 2021). The depletion of stratospheric ozone
and its radiative forcing remained relatively unchanged since 2000 for the last two decades and is starting to
decline; recovery is expected to occur shortly after the middle of the twenty-first century (WMO/UNEP 2018).
The past increase in tropospheric ozone, which is also a greenhouse gas, is estimated to provide the third largest
increase in direct radiative forcing since the pre-industrial era, behind CO2 and Cm. Tropospheric ozone is
produced from complex chemical reactions of volatile organic compounds (including CH4) mixing with NOx in the
presence of sunlight. The tropospheric concentrations of ozone and these other pollutants are short-lived and,
therefore, spatially variable (IPCC 2021).
Halocarbons, Sulfur Hexafluoride, and Nitrogen Triftuoride. Halocarbons are, for the most part, man-made
chemicals that have direct radiative forcing effects and could also have an indirect effect. Halocarbons that contain
chlorine (CFCs, HCFCs, methyl chloroform, and carbon tetrachloride) and bromine (halons, methyl bromide, and
hydrobromofluorocarbons) result in stratospheric ozone depletion and are therefore controlled under the
Montreal Protocol on Substances that Deplete the Ozone Layer. Although most CFCs and HCFCs are potent global
warming gases, their net radiative forcing effect on the atmosphere is reduced because they cause stratospheric
ozone depletion, which itself is a greenhouse gas but which also shields the Earth from harmful levels of ultraviolet
radiation. Under the Montreal Protocol, the United States phased out the production and importation of halons by
1994 and of CFCs by 1996. Under the Copenhagen Amendments to the Protocol, a cap was placed on the
production and importation of HCFCs by non-Article 5 countries, including the United States,19 beginning in 1996,
and then followed by intermediate requirements and a complete phase-out by the year 2030. While ozone
depleting gases covered under the Montreal Protocol and its Amendments are not covered by the UNFCCC, they
are reported in this Inventory under Annex 6.2 for informational purposes.
Hydrofluorocarbons, PFCs, SF6, and NF3 are not ozone depleting substances. The most common HFCs are, however,
powerful greenhouse gases. Hydrofluorocarbons are primarily used as replacements for ozone depleting
substances but also emitted as a byproduct of the HCFC-22 (chlorodifluoromethane) manufacturing process.
Currently, they have a small aggregate radiative forcing impact, but it is anticipated that without further controls
their contribution to overall radiative forcing will increase, the ERF (effective radiative forcing) of halogenated
gases increased by 3.5 percent between 2011 and 2019 primarily due to a decrease in atmospheric mixing-ratios of
CFCs and an increase in their replacements (IPCC 2021). On December 27, 2020, the American Innovation and
Manufacturing (AIM) Act was enacted by Congress and which gives EPA authority to phase down HFC production
and consumption (i.e., production plus import, minus export), through an allowance allocation program,
promulgate certain regulations for purposes of maximizing reclamation and minimizing releases of HFCs and their
substitutes from equipment, and facilitating the transition to next-generation technologies through sector-based
restrictions, which will lead to lower HFC emissions over time. Perfluorocarbons, SF6, and NF3 are predominantly
emitted from various industrial processes including aluminum smelting, semiconductor manufacturing, electric
power transmission and distribution, and magnesium casting. Currently, the radiative forcing impact of PFCs, SF6,
and NF3 is also small, but they have a significant growth rate, extremely long atmospheric lifetimes, and are strong
absorbers of infrared radiation, and therefore have the potential to influence climate far into the future (IPCC
2021).
Carbon Monoxide (CO). Carbon monoxide has an indirect radiative forcing effect by elevating concentrations of CH4
and tropospheric ozone through chemical reactions with other atmospheric constituents (e.g., the hydroxyl radical,
OH) that would otherwise assist in destroying CH4 and tropospheric ozone. Carbon monoxide is created when
carbon-containing fuels are burned incompletely. Through natural processes in the atmosphere, it is eventually
oxidized to CO2. Carbon monoxide concentrations are both short-lived in the atmosphere and spatially variable.
19 Article 5 of the Montreal Protocol covers several groups of countries, especially developing countries, with low consumption
rates of ozone depleting substances. Developing countries with per capita consumption of less than 0.3 kg of certain ozone
depleting substances (weighted by their ozone depleting potential) receive financial assistance and a grace period of ten
additional years in the phase-out of ozone depleting substances.
Introduction 1-7
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Nitrogen Oxides (NOx). The primary climate change effects of nitrogen oxides (i.e., NO and NO2) are indirect.
Warming effects can occur due to reactions leading to the formation of ozone in the troposphere, but cooling
effects can occur due to the role of NOx as a precursor to nitrate particles (i.e., aerosols) and due to destruction of
stratospheric ozone when emitted from very high-altitude aircraft.20 Additionally, NOx emissions are also likely to
decrease CH4 concentrations, thus having a negative radiative forcing effect (IPCC 2021). Nitrogen oxides are
created from lightning, soil microbial activity, biomass burning (both natural and anthropogenic fires) fuel
combustion, and, in the stratosphere, from the photo-degradation of N20. Concentrations of NOx are both
relatively short-lived in the atmosphere and spatially variable.
Non-methane Volatile Organic Compounds (NMVOCs). Non-methane volatile organic compounds include
substances such as propane, butane, and ethane. These compounds participate, along with NOx, in the formation
of tropospheric ozone and other photochemical oxidants. NMVOCs are emitted primarily from transportation and
industrial processes, as well as biomass burning and non-industrial consumption of organic solvents.
Concentrations of NMVOCs tend to be both short-lived in the atmosphere and spatially variable.
Aerosols. Aerosols are extremely small particles or liquid droplets found in the atmosphere that are either directly
emitted into or are created through chemical reactions in the Earth's atmosphere. Aerosols or their chemical
precursors can be emitted by natural events such as dust storms, biogenic or volcanic activity, or by anthropogenic
processes such as transportation, coal combustion, cement manufacturing, waste incineration, or biomass burning.
Various categories of aerosols exist from both natural and anthropogenic sources, such as soil dust, sea salt,
biogenic aerosols, sulfates, nitrates, volcanic aerosols, industrial dust, and carbonaceous21 aerosols (e.g., black
carbon, organic carbon). Aerosols can be removed from the atmosphere relatively rapidly by precipitation or
through more complex processes under dry conditions.
Aerosols affect radiative forcing differently than greenhouse gases. Their radiative effects occur through direct and
indirect mechanisms: directly by scattering and absorbing solar radiation (and to a lesser extent scattering,
absorption, and emission of terrestrial radiation); and indirectly by increasing cloud droplets and ice crystals that
modify the formation, precipitation efficiency, and radiative properties of clouds (IPCC 2021). Despite advances in
understanding of cloud-aerosol interactions, the contribution of aerosols to radiative forcing are difficult to
quantify because aerosols generally have short atmospheric lifetimes, and have number concentrations, size
distributions, and compositions that vary regionally, spatially, and temporally (IPCC 2021).
The net effect of aerosols on the Earth's radiative forcing is believed to be negative (i.e., net cooling effect on the
climate). In fact, aerosols contributed a cooling influence of up to 1.4 degrees, offsetting a substantial portion of
greenhouse gas warming (IPCC 2021). Because aerosols remain in the atmosphere for only days to weeks, their
concentrations respond rapidly to changes in emissions.22 Not all aerosols have a cooling effect. Current research
suggests that another constituent of aerosols, black carbon, has a positive radiative forcing by heating the Earth's
atmosphere and causing surface warming when deposited on ice and snow (IPCC 2021). Black carbon also
influences cloud development, but the direction and magnitude of this forcing is an area of active research.
Global Warming Potentials
A global warming potential (GWP) is a quantified measure of the globally averaged relative radiative forcing
impacts of a particular greenhouse gas (see Table 1-2). It is defined as the accumulated radiative forcing within a
specific time horizon caused by emitting 1 kilogram (kg) of the gas, relative to that of the reference gas CO2 (IPCC
2021). Direct radiative effects occur when the gas itself absorbs radiation. Indirect radiative forcing occurs when
20 NOx emissions injected higher in the stratosphere, primarily from fuel combustion emissions from high altitude supersonic
aircraft, can lead to stratospheric ozone depletion.
21 Carbonaceous aerosols are aerosols that are comprised mainly of organic substances and forms of black carbon (or soot)
(IPCC 2013).
22 Volcanic activity can inject significant quantities of aerosol producing sulfur dioxide and other sulfur compounds into the
stratosphere, which can result in a longer lasting negative forcing effect (i.e., a few years) (IPCC 2013).
1-8 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
chemical transformations involving the original gas produce a gas or gases that are greenhouse gases, or when a
gas influences other radiatively important processes such as the atmospheric lifetimes of other gases. The
reference gas used is CO2, and therefore GWP-weighted emissions are measured in million metric tons of CO2
equivalent (MMT CO2 Eq.).23 The relationship between kilotons (kt) of a gas and MMT CO2 Eq. can be expressed as
follows:
Equation 1-1: Calculating CO2 Equivalent Emissions
GWP values allow for a comparison of the impacts of emissions and reductions of different gases. According to the
IPCC, GWPs typically have an uncertainty of ±40 percent.
All estimates are provided throughout the report in both MMT CO2 equivalents and unweighted units. Recent
decisions under the UNFCCC24 require Parties to use 100-year GWP values from the IPCC Fifth Assessment Report
(AR5) for calculating CC>2-equivalence in their national reporting (IPCC 2013) by the end of 2024.
...Decides that, until it adopts a further decision on the matter, the global warming potential values
used by Parties in their reporting under the Convention to calculate the carbon dioxide equivalence of
anthropogenic greenhouse gas emissions by sources and removals by sinks shall be based on the
effects of greenhouse gases over a 100-year time horizon as listed in table 8.A.1 in appendix 8.A to the
contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on
Climate Change,25 excluding the value for fossil methane; ,26
This reflects updated science and ensures that national GHG inventories reported by all nations are comparable. In
preparation for upcoming UNFCCC requirement,27 this report reflects CC>2-equivalent greenhouse gas totals using
100-year AR5 GWP values. A comparison of emission values with the previously used 100-year GWP values from
IPCC Fourth Assessment Report (AR4) (IPCC 2007), and the IPCC Sixth Assessment Report (AR6) (IPCC 2021) values
can be found in Annex 6.1 of this report. The 100-year GWP values used in this report are listed below in Table 1-2.
Greenhouse gases with relatively long atmospheric lifetimes (e.g., CO2, CFU, N2O, HFCs, PFCs, SF6, NF3) tend to be
evenly distributed throughout the atmosphere, and consequently global average concentrations can be
determined. The short-lived gases such as water vapor, carbon monoxide, tropospheric ozone, ozone precursors
(e.g., NOx, and NMVOCs), and tropospheric aerosols (e.g., SO2 products and carbonaceous particles), however, vary
regionally, and consequently it is difficult to quantify their global radiative forcing impacts. Parties to the UNFCCC
have not agreed upon GWP values for these gases that are short-lived and spatially inhomogeneous in the
23 Carbon comprises 12/44ths of carbon dioxide by weight.
24 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/sbsta2022 L25a01E.pdf. The UNFCCC reporting guidelines
require use of the 100-year GWPs listed in table 8.A.1 in Annex 8.A of Chapter 8 of the Fifth Assessment Report (AR5) of the
Intergovernmental Panel on Climate Change, excluding the value for fossil methane.
25 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. TF Stocker, D Qin, G-K
Plattner, et al. (eds.). Cambridge and New York: Cambridge University Press. Available at http://www.ipcc.ch/report/ar5/wgl.
26 United Nations Framework Convention on Climate Change, see
https://unfccc.int/sites/default/files/resource/sbsta2022 L25a01E.pdf.
27 See Annex to decision 18/CMA.l, available online at https://unfccc.int/sites/default/files/resource/CMA2018 03a02E.pdf
where,
MMT CO2 Eq.
kt
GWP
MMT
= Million metric tons of C02 equivalent
= kilotons (equivalent to a thousand metric tons)
= Global warming potential
= Million metric tons
atmosphere.
Introduction 1-9
-------
Table 1-2: Global Warming Potentials and Atmospheric Lifetimes (Years) Used in this Report
Gas
Atmospheric Lifetime
GWPa
C02
See footnote15
1
CH4c
12.4
28
n2o
121
265
HFC-23
222
12,400
HFC-32
5.2
677
HFC-41d
2.8
116
HFC-125
28.2
3,170
HFC-134a
13.4
1,300
HFC-143a
47.1
4,800
HFC-152a
1.5
138
HFC-227ea
38.9
3,350
HFC-236fa
242
8,060
cf4
50,000
6,630
c2f6
10,000
11,100
C3Fs
2,600
8,900
c-C4Fs
3,200
9,540
sf6
3,200
23,500
nf3
500
16,100
Other Fluorinated Gases
See Annex 6
a 100-year time horizon.
b For a given amount of C02 emitted, some fraction of the atmospheric
increase in concentration is quickly absorbed by the oceans and
terrestrial vegetation, some fraction of the atmospheric increase will
only slowly decrease over a number of years, and a small portion of the
increase will remain for many centuries or more.
c The GWP of CH4 includes the direct effects and those indirect effects
due to the production of tropospheric ozone and stratospheric water
vapor. The indirect effect due to the production of C02 is not included.
d See Table A-l of 40 CFR Part 98
Source: IPCC(2013).
Box 1-2: The IPCC Sixth Assessment Report and Global Warming Potentials
In 2021, the IPCC published its Sixth Assessment Report (AR6), which updated its comprehensive scientific
assessment of climate change. Within the AR6 report, the GWP values of gases were revised relative to previous
IPCC reports, namely the IPCC Second Assessment Report (SAR) (IPCC 1996), the IPCC Third Assessment Report
(TAR) (IPCC 2001), the IPCC Fourth Assessment Report (AR4) (IPCC 2007), and the IPCC Fifth Assessment Report
(AR5) (IPCC 2014). Although the AR5 GWP values are used throughout this report, consistent with UNFCCC
reporting requirements, it is straight-forward to review the changes to the GWP values and their impact on
estimates of the total GWP-weighted emissions of the United States. In the AR6, the IPCC used more recent
estimates of the atmospheric lifetimes and radiative efficiencies of some gases and updated background
concentrations. The AR6 now includes climate-carbon feedback effects for non-CC>2 gases, improving the
consistency between treatment of CO2 and non-CC>2 gases. Indirect effects of gases on other atmospheric
constituents (such as the effect of methane on ozone) have also been updated to match more recent science.
Table 1-3 presents the new GWP values, relative to those presented in the AR4 and AR5, using the 100-year
time horizon common to UNFCCC reporting. For consistency with international reporting standards under the
UNFCCC, official emission estimates are reported by the United States using AR4 100-year GWP values, as
1-10 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
required by the 2013 revision to the UNFCCC reporting guidelines for national inventories.28 Updated reporting
guidelines under the Paris Agreement which require the United States and other countries to shift to use of the
IPCC Fifth Assessment Report (AR5) (IPCC 2013) 100-year GWP values (without feedbacks) take effect for
national inventory reporting in 2024.29 All estimates provided throughout this report are also presented in
unweighted units. For informational purposes, emission estimates that use 100-year GWPs from other recent
IPCC Assessment Reports are presented in detail in Annex 6.1 of this report.
Table 1-3: Comparison of 100-Year GWP values
100-Year GWP Values
Comparisons to AR5
AR5 with
AR5 with
Gas
AR4 AR5a
feedbacks'1
AR6C
AR4
feedbacks'1 AR6C
C02
1
1
1
1
NC
NC
NC
CH4d
25
28
34
27
(3)
6
1
N20
298
265
298
273
33
33
8
HFC-23
14,800
12,400
13,856
14,600
2,400
1,456
2,200
HFC-32
675
677
817
771
(2)
140
94
HFC-41
92
116
141
135
(24)
25
19
HFC-125
3,500
3,170
3,691
3,740
330
521
570
HFC-134a
1,430
1,300
1,549
1,530
130
249
230
HFC-143a
4,470
4,800
5,508
5,810
(330)
708
1,010
HFC-152a
124
138
167
164
(14)
29
26
HFC-227ea
3,220
3,350
3,860
3,600
(130)
510
250
HFC-236fa
9,810
8,060
8,998
8,690
1,750
938
630
cf4
7,390
6,630
7,349
7,380
760
719
750
c2f6
12,200
11,100
12,340
12,400
1,100
1,240
1,300
C3Fs
8,830
8,900
9,878
9,290
(70)
978
390
c-C4Fs
10,300
9,540
10,592
10,200
(760)
1,052
660
sf6
22,800
23,500
26,087
24,300
700
2,587
800
nf3
17,200
16,100
17,885
17,400
(1,100)
1,785
1,300
NC (No Change)
a The GWP values in this column reflect values used in this report from AR5 excluding climate-carbon feedbacks and
the value for fossil methane.
b The GWP values in this column are from the AR5 report but include climate-carbon feedbacks for the non-C02 gases
in order to be consistent with the approach used in calculating the C02 lifetime.
c The GWP values in this column are from the AR6 report.
d The GWP of CH4 includes the direct effects and those indirect effects due to the production of tropospheric ozone
and stratospheric water vapor. Including the indirect effect due to the production of C02 resulting from methane
oxidation would lead to an increase in AR5 methane GWP values by 2 for fossil methane and is not shown in this table.
Note: Parentheses indicate negative values.
Sources: IPCC (2021), IPCC (2013), IPCC (2007), IPCC (2001), IPCC (1996).
1.2 National Inventory Arrangements
The U.S. Environmental Protection Agency (EPA), in cooperation with other U.S. government agencies, prepares
the Inventory of U.S. Greenhouse Gas Emissions and Sinks. A wide range of agencies and individuals are involved in
supplying data to, planning methodological approaches and improvements, reviewing, or preparing portions of the
28 See http://unfccc.int/resource/docs/2013/copl9/ene/10a03.pdf.
29 See https://unfccc.int/process-and-meetings/transparencv-and-reporting/reporting-and-review-under-the-paris-agreement.
Introduction 1-11
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Inventory—including federal and state government authorities, research and academic institutions, industry
associations, and private consultants.
Within EPA, the Office of Atmospheric Protection (OAP) is the lead office responsible for the emission and removal
calculations provided in the Inventory, as well as the completion of the National Inventory Report and the
Common Reporting Format (CRF) tables. EPA's Office of Transportation and Air Quality (OTAQ) and Office of
Research and Development (ORD) are also involved in calculating emissions and removals for the Inventory. The
U.S. Department of State (DOS) serves as the overall national focal point to the UNFCCC, and EPA's OAP serves as
the National Inventory Focal Point for this report, including responding to technical questions and comments on
the U.S. Inventory. EPA staff coordinate the annual methodological choice, activity data collection, emission and
removal calculations, uncertainty assessment, QA/QC processes, and improvement planning at the individual
source and sink category level. EPA, the inventory coordinator, compiles the entire Inventory into the proper
reporting format for submission to the UNFCCC, and is responsible for the synthesis of information and for the
consistent application of cross-cutting IPCCgood practice across the Inventory.
Several other government agencies contribute to the collection and analysis of the underlying activity data used in
the Inventory calculations via formal (e.g., interagency agreements) and informal relationships, in addition to the
calculation of estimates integrated in the report (e.g., U.S. Department of Agriculture's U.S. Forest Service and
Agricultural Service, National Oceanic and Atmospheric Administration, Federal Aviation Administration, and
Department of Defense). Other U.S. agencies provide official data for use in the Inventory. The U.S. Department of
Energy's Energy Information Administration provides national fuel consumption data and the U.S. Department of
Defense provides data on military fuel consumption and use of bunker fuels. Other U.S. agencies providing activity
data for use in EPA's emission calculations include: the U.S. Department of Agriculture, National Oceanic and
Atmospheric Administration, the U.S. Geological Survey, the Federal Highway Administration, the Department of
Transportation, the Bureau of Transportation Statistics, the Department of Commerce, and the Federal Aviation
Administration. Academic and research centers also provide activity data and calculations to EPA, as well as
individual companies participating in voluntary outreach efforts with EPA. Finally, EPA as the National Inventory
Focal Point, in coordination with the U.S. Department of State, officially submits the Inventory to the UNFCCC each
April.
Figure 1-1: National Inventory Arrangements and Process Diagram
United States Greenhouse Gas Inventory Institutional Arrangements
1. Data Collection
Energy Data Sources
2. Emissions And
Removals
Calculations
Agriculture and
LULUCF Data Sources
Industrial Processes
and Product Use Data
Sources
U.S. Environmental
Protection Agency
Other U.S.
Government Agencies
(USFS, NOAA,
DOD, FAA)
Waste Data Sources
3. Inventory
Compilation
(including report and
reporting table
compilation)
U.S. Environmental
Protection Agency
Inventory Compiler
4. Inventory
U.S. Department
of State
1-12 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Overview of Inventory Data Sources by Source and Sink Category
Energy
Agriculture and LULUCF
IPPU
Waste I
U.S. Energy Information
Administration
USDA U.S. Forest Service Forest
Inventory and Analysis Program
(FIA)
EPA Greenhouse Gas Reporting EPA Greenhouse Gas
Program (GHGRP) Reporting Program (GHGRP)
U.S. Department of Commerce USDA Natural Resource
- Bureau of the Census Conservation Service (NRCS)
U.S. Geological Survey (USGS)
National Minerals Information
Center
EPA Office of Land and
Emergency Management
(OLEM)
U.S. Department of Defense -
Defense Logistics Agency
USDA National Agricultural
Statistics Service (NASS) and
Agricultural Research Service
(ARS)
American Chemistry Council
(ACC)
EPA Clean Watershed Needs
Survey (CWNS)
U.S. Department of Homeland
Security
EPA Office of Research and
Development (ORD)
American Iron and Steel
Institute (AISI)
American Housing Survey
U.S. Department of
Transportation - Federal
Highway Administration
U.S. Fish and Wildlife Service
U.S. International Trade
Commission (USITC)
Data from research studies,
trade publications, and
industry associations
U.S. Department of
Transportation - Federal
Aviation Administration
U.S. Department of Agriculture
(USDA) Animal and Plant Health
Inspection Service (APHIS)
Air-Conditioning, Heating, and
Refrigeration Institute
U.S. Department of
Transportation & Bureau of
Transportation Statistics
Association of American Plant
Food Control Officials (AAPFCO)
Data from other U.S.
government agencies, research
studies, trade publications, and
industry association
U.S. Department of Labor -
Mine Safety and Health
Administration
National Oceanic and
Atmospheric Administration
(NOAA)
UNEP Technology and
Economic Assessment Panel
U.S. Department of Energy and EPA Office of Land and Emergency
its National Laboratories Management (OLEM)
EPA Acid Rain Program
USDA Farm Service Agency
EPA MOVES Model
U.S. Geological Survey (USGS)
EPA Greenhouse Gas Reporting U.S. Department of the Interior
Program (GHGRP) (DOI), Bureau of Land
Management (BLM)
U.S. Department of Labor - EPA Office of Land and Emergency
Mine Safety and Health Management (OLEM)
Administration
American Association of Alaska Department of Natural
Railroads Resources
American Public Transportation U.S. Census Bureau
Association
Data from research studies, Data from research studies, trade
trade publications, and industrypublications, and industry
associations associations
2 Note: This table is not an exhaustive list of all data sources.
s 1.3 Inventory Process
4 This section describes EPA's approach to preparing the annual U.S. Inventory, which consists of the National
5 Inventory Report (NIR) and Common Reporting Format (CRF) tables. The inventory coordinator at EPA, with
6 support from the cross-cutting compilation staff, is responsible for aggregating all emission and removal estimates,
Introduction 1-13
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
conducting the overall uncertainty analysis of Inventory emissions and trends over time, and ensuring consistency
and quality throughout the NIR and CRF tables. Emission and removal calculations, including associated
uncertainty analysis for individual sources and/or sink categories are the responsibility of individual source and
sink category leads, who are most familiar with each category, underlying data, and the unique national
circumstances relevant to its emissions or removals profile. Using IPCC good practice guidance, the individual leads
determine the most appropriate methodology and collect the best activity data to use in the emission and removal
calculations, based upon their expertise in the source or sink category, as well as coordinating with researchers
and expert consultants familiar with the sources and sinks. Each year, the coordinator oversees a multi-stage
process for collecting information from each individual source and sink category lead to compile all information
and data for the Inventory.
Methodology Development, Data Collection, and Emissions
and Sink Estimation
Source and sink category leads at EPA collect input data and, as necessary, evaluate or develop the estimation
methodology for the individual source and/or sink categories. Because EPA has been preparing the Inventory for
many years, for most source and sink categories, the methodology for the previous year is applied to the new
"current" year of the Inventory, and inventory analysts collect any new data or update data that have changed
from the previous year. If estimates for a new source or sink category are being developed for the first time, or if
the methodology is changing for an existing category (e.g., the United States is implementing improvement efforts
to apply a higher tiered approach for that category), then the source and/or sink category lead will develop and
implement the new or refined methodology, gather the most appropriate activity data and emission factors (or in
some cases direct emission measurements) for the entire time series, and conduct any further category-specific
review with involvement of relevant experts from industry, government, and universities (see Chapter 9 and Box
ES-3 on EPA's approach to recalculations).
Once the methodology is in place and the data are collected, the individual source and sink category leads
calculate emission and removal estimates. The individual leads then update or create the relevant report text and
accompanying annexes for the Inventory. Source and sink category leads are also responsible for completing the
relevant sectoral background tables of the CRF, conducting quality control (QC) checks, preparing relevant
category materials for QA, or expert reviews, category-level uncertainty assessments, and reviewing data for
publication in EPA's GHG Data Explorer.
The treatment of confidential business information (CBI) in the Inventory is based on EPA internal guidelines, as
well as regulations30 applicable to the data used. EPA has specific procedures in place to safeguard CBI during the
inventory compilation process. When information derived from CBI data is used for development of inventory
calculations, EPA procedures ensure that these confidential data are sufficiently aggregated to protect
confidentiality while still providing useful information for analysis. For example, within the Energy and Industrial
Processes and Product Use (IPPU) sectors, EPA has used aggregated facility-level data from the Greenhouse Gas
Reporting Program (GHGRP) to develop, inform, and/or quality-assure U.S. emission estimates. In 2014, EPA's
GHGRP, with industry engagement, compiled criteria that would be used for aggregating its confidential data to
shield the underlying CBI from public disclosure.31 In the Inventory, EPA is publishing only data values that meet
the GHGRP aggregation criteria.32 Specific uses of aggregated facility-level data are described in the respective
30 40 CFR part 2, Subpart B titled "Confidentiality of Business Information" which is the regulation establishing rules governing
handling of data entitled to confidentiality treatment. See https://www.ecfr.gov/cgi-bin/text-
idx?SID=a764235c9eadf9afe05fe04c07a28939&mc=true&node=sp40.1.2.b&rgn=div6.
31 Federal Register Notice on "Greenhouse Gas Reporting Program: Publication of Aggregated Greenhouse Gas Data." See pp.
79 and 110 of notice at https://www.gpo.gov/fdsys/pkg/FR-2014-06-09/pdf/2014-13425.pdf.
32 U.S. EPA Greenhouse Gas Reporting Program. Developments on Publication of Aggregated Greenhouse Gas Data, November
25, 2014. See http://www.epa.gov/ghgreporting/confidential-business-information-ghg-reporting.
1-14 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
methodological sections within those chapters. In addition, EPA uses historical data reported voluntarily to EPA via
various voluntary initiatives with U.S. industry (e.g., EPA Voluntary Aluminum Industrial Partnership (VAIP)) and
follows guidelines established under the voluntary programs for managing CBI.
Data Compilation and Archiving
The inventory coordinator at EPA with support from the data/document manager collects the source and sink
categories' descriptive text and annexes, and also aggregates the emission and removal estimates into a summary
data file that links the individual source and sink category data files together. This summary data file contains all of
the essential data in one central location, in formats commonly used in the Inventory document. In addition to the
data from each source and sink category, other national trend and related data are also gathered in the summary
sheet for use in the Executive Summary, Introduction, and Trends sections of the Inventory report (e.g., GDP,
population, energy use). Similarly, the recalculation analysis and key category analysis are completed in a separate
data file based on output from the summary data file. The uncertainty estimates for each source and sink category
are also aggregated into uncertainty summary data files that are used to conduct the overall Inventory uncertainty
analysis (see Section 1.7). Microsoft SharePoint, kept on a central server at EPA under the jurisdiction of the
inventory coordinator, provides a platform for facilitating collaboration during each compilation phase, but also
the efficient storage and archiving of electronic files each annual cycle. Previous final published inventories are
also maintained on a report archive page on EPA's Greenhouse Gas website.33
National Inventory Report (NIR) Preparation
The NIR is compiled from the sections developed by each individual source or sink category lead. In addition, the
inventory coordinator prepares a brief overview of each chapter that summarizes the emissions and removals from
all sources and sinks discussed in the chapters. Also at this time, the Executive Summary, Introduction,, Trends in
Greenhouse Gas Emissions and Removals, and Recalculations and Improvements chapters are drafted, to reflect
the trends and impact from improvements for the time series of the current Inventory. The analysis of trends
necessitates gathering supplemental data, including weather and temperature conditions, economic activity and
gross domestic product, population, atmospheric conditions, and the annual consumption of electricity, energy,
and fossil fuels. Changes in these data are used to explain the trends observed in greenhouse gas emissions in the
United States. Furthermore, specific factors that affect individual sectors are researched and discussed. Many of
the factors that affect emissions are included in the Inventory document as separate analyses or side discussions in
boxes within the text. Finally, the uncertainty analysis and key category analysis are compiled and updated in the
report as part of final analysis steps. Throughout the report text boxes are also created to provide additional
documentation (e.g., definitions) and/or examine the data aggregated in different ways than in the remainder of
the document, such as a focus on transportation activities or emissions from electricity generation. The document
is prepared to align with the specification of the UNFCCC reporting guidelines for National Inventory Reports while
also reflecting national circumstances.
Common Reporting Format Table (CRF) Compilation
The CRF tables are compiled from individual time series input data sheets completed by each individual source or
sink category lead, which contain emissions and/or removals and activity data, estimates, methodological and
completeness notations and associated explanations. The inventory coordinator and cross-cutting compilation
staff import the category data into the UNFCCC's "CRF Reporter" for the United States, assuring consistency and
completeness across all sectoral tables. The summary reports for emissions and removals, methods, and emission
factors used, the summary tables indicating completeness of estimates (i.e., notation key NE/IE tables), the
recalculation tables, and the emission and removal trends tables automatically compiled by the CRF Reporter and
reviewed by the inventory coordinator with support from the cross-cutting compilation staff. Internal automated
33 See https://www.epa.gov/ghgemissions/us-greenhouse-gas-inventory-report-archive.
Introduction 1-15
-------
1 quality checks within the CRF Reporter, as well as reviews by the cross-cutting and category leads, are completed
2 for the entire time series of CRF tables before submission.
3 QA/QC and Uncertainty
4 Quality assurance and quality control (QA/QC) and uncertainty analyses are guided by the QA/QC and Inventory
5 coordinators, who help maintain the QA/QC plan and the overall uncertainty analysis procedures (see sections on
6 QA/QC and Uncertainty, below) in collaboration with the broader inventory compilation team. The QA/QC
7 coordinator works closely with the Inventory coordinator and source and sink category leads to ensure that a
8 consistent QA/QC plan is implemented across all inventory categories. Similarly, the Inventory coordinator ensures
9 the uncertainty analysis is implemented consistently across all categories. The inventory QA/QC plan, outlined in
10 Section 1.7 and Annex 8, is consistent with the quality assurance procedures outlined by EPA and IPCC good
11 practices. The QA/QC and uncertainty findings also inform overall improvement planning, and specific
12 improvements are noted in the Planned Improvements sections of respective categories. QA processes are
13 outlined below.
14 Expert, Public, and UNFCCC Reviews
15 The compilation of the inventory includes a two-stage review or QA process, in addition to international technical
16 expert review following submission of the report to the UNFCCC. During the first stage (the 30-day Expert Review
17 period), a first draft of sectoral chapters of the document are sent to a select list of technical experts outside of
18 EPA who are not directly involved in preparing estimates. The purpose of the Expert Review is to provide an
19 objective review, encourage feedback on the methodological and data sources used in the current Inventory,
20 especially for sources and sinks which have experienced any changes since the previous Inventory.
21 Once comments are received and addressed, the second stage, or second draft of the document is released for
22 public review by publishing a notice in the U.S. Federal Register and posting the entire draft Inventory document
23 on the EPA website. The Public Review period allows for a 30-day comment period and is open to the entire U.S.
24 public. Comments received may require further discussion with experts and/or additional research, and specific
25 Inventory improvements requiring further analysis as a result of comments are noted in the relevant category's
26 Planned Improvement section. EPA publishes responses to comments received during both reviews with the
27 publication of the final report on its report website.
28 Following completion and submission of the report to the UNFCCC, the report also undergoes review by an
29 international team of independent experts for adherence to UNFCCC reporting guidelines and IPCC methodological
30 guidance.34 Feedback from all review processes that contribute to improving inventory quality over time are
31 described within each planned improvement section and further in Annex 8. See also the Improvement Planning
32 process discussed below.
33 Final Submittal to UNFCCC, Document and Data Publication
34 After the final revisions to incorporate any comments from the Expert Review and Public Review periods, EPA
35 prepares the final NIR and the accompanying CRF tables for electronic reporting. EPA, as the National Inventory
36 focal point, sends the official submission of the U.S. Inventory to the UNFCCC using the CRF Reporter software,
37 coordinating with the U.S. Department of State, the overall UNFCCC focal point. Concurrently, for timely public
38 access, the report is also published on EPA's website.35 On EPA's website, users can also visualize and download
34 See https://unfccc.int/process-and-meetings/transparencv-and-reporting/reporting-and-review-under-the-
convention/greenhouse-gas-inventories-annex-i-parties/review-process.
35 See https://www.epa.gov/ghgemissions/inventory-us-greenhouse-gas-emissions-and-sinks.
1-16 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
the current time-series estimates from the GHG Inventory Data Explorer Tool,36 and also download more detailed
data presented in tables within the report and report annex in CSV format.
Improvement Planning
Each year, many emission and sink estimates in the Inventory of U.S. Greenhouse Gas Emissions and Sinks are
recalculated and revised, 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. In this effort, the United States follows the 2006IPCC Guidelines (IPCC 2006), which state, "Both
methodological changes and refinements over time are an essential part of improving inventory quality. It is good
practice to change or refine methods when available data have changed; the previously used method is not
consistent with the IPCC guidelines for that category; a category has become key; the previously used method is
insufficient to reflect mitigation activities in a transparent manner; the capacity for inventory preparation has
increased; improved inventory methods become available; and/or for correction of errors." The EPA's OAP
coordinates improvement planning across all sectors and also cross-cutting analyses based on annual review and
input from the technical teams leading compilation of each sector's estimates, including continuous improvements
to the overall data and document compilation processes. Planned improvements are identified through QA/QC
processes (including completeness checks), the key category analysis, and the uncertainty analysis. The inventory
coordinator, with input from EPA source and sink category leads, maintains a log of all planned improvements, by
sector and cross-cutting, tracking the category significance, specific category improvement, prioritization,
anticipated time frame for implementation of each proposed improvement, and status of progress in
implementing improvement. Improvements for significant or key categories are usually prioritized across all
improvements unless effort would require disproportionate levels of effort and resources relative to
improvements for other key categories to address.
1.4 Methodology and Data Sources
Emissions and removals of greenhouse gases from various source and sink categories have been estimated using
methodologies that are consistent with the 2006 IPCC Guidelines for National Greenhouse Gas Inventories (IPCC
2006) and its supplements and refinements. To a great extent, this report makes use of published official economic
and physical statistics for activity data, emission factors and other key parameters. Depending on the category,
activity data can include fuel consumption or deliveries, vehicle-miles traveled, raw material processed, etc.
Emission factors are factors that relate quantities of emissions to an activity. For more information on data sources
see Section 1.2 above, Box 1-1 on use of GHGRP data, and categories' methodology sections for more information
on other data sources. In addition to official statistics, the report utilizes findings from academic studies, trade
association surveys and statistical reports, along with expert judgment, consistent with the 2006 IPCC Guidelines.
The methodologies provided in the 2006 IPCC Guidelines represent foundational methodologies for a variety of
source and sink categories, and many of these methodologies continue to be improved and refined as new
research and data become available. This report uses those IPCC methodologies when applicable, and supplements
them with refined guidance, other available country-specific methodologies and data where possible (e.g., EPA's
GHGRP). For example, as noted earlier in this chapter, this report does apply recent supplements and refinements
to 2006 IPCC Guidelines in estimating emissions and removals from coal mining, wastewater treatment and
discharge, low voltage anode effects (LVAE) during aluminum production, drained organic soils, and management
of wetlands, including flooded lands. Choices made regarding the methodologies and data sources used are
provided in the Methodology and Time Series Consistency discussion of each category within each sectoral chapter
of the report. Where additional detail is helpful and necessary to explain methodologies and data sources used to
estimate emissions, complete documentation is provided in the annexes as indicated in the methodology sections
36 See https://cfpub.epa.gov/ghgdata/inventoryexplorer/.
Introduction 1-17
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
of those respective source categories (e.g., Annex 3.13 for Forest Land Remaining Forest Land and Land Converted
to Forest Land).
1.5 Key Categories
The 2006IPCC Guidelines (IPCC 2006) defines a key category as a "[category] that is prioritized within the national
inventory system because its estimate has a significant influence on a country's total inventory of greenhouse
gases in terms of the absolute level, the trend, or the uncertainty in emissions and removals."37 A key category
analysis identifies source or sink categories for focusing efforts to improve overall Inventory quality.
The 2006 IPCC Guidelines (IPCC 2006) defines several approaches, both quantitative and qualitative, to conduct a
key category analysis and identify key categories both in terms of absolute level and trend, along with
consideration of uncertainty. This report employs all approaches to identify key categories for the United States.
The first approach, Approach 1, identifies significant or key categories without considering uncertainty in its
calculations. An Approach 1 level assessment identifies all source and sink categories that cumulatively account for
95 percent of total level, i.e., total emissions (gross) in a given year when assessed in descending order of absolute
magnitude. The level analysis was performed twice, including and excluding sources and sinks from the Land Use,
Land-Use Change, and Forestry (LULUCF) sector categories. Similarly, an Approach 1 trend analysis can identify
categories with trends that differ significantly from overall trends by identifying all source and sink categories that
cumulatively account for 95 percent of the sum all the trend assessments (e.g., percent change relative to national
trend) when sorted in descending order of absolute magnitude.
The next method, Approach 2, was then implemented to identify any additional key categories not already
identified from the Approach 1 level and trend assessments by considering uncertainty. The Approach 2 analysis
differs from Approach 1 by incorporating each category's uncertainty assessments in its calculations and was also
performed twice, including and excluding LULUCF categories. An Approach 2 level assessment identifies all sources
and sink categories that cumulatively account for 90 percent of the sum of all level assessments when sorted in
descending order of magnitude. Similarly, an Approach 2 trend analysis can identify categories that whose trends
differ significantly from overall trends and also weighting the relative trend difference with the category's
uncertainty assessment for 2020.
For 2021, based on the key category analysis, excluding the LULUCF sector and uncertainty, 34 categories
accounted for 95 percent of emissions. Four categories account for 55 percent of emissions: CO2 from road
transport-related fuel combustion, CO2 from coal-fired electricity generation, CO2 from gas fired electricity
generation, and CO2 from gas-fired industrial combustion. When considering uncertainties, additional categories
such as Cm from abandoned oil and gas wells were also identified as a key category. In the trend analysis, 32
categories were identified as key categories, and when considering uncertainties, 7 additional categories were
identified as key. The trend analysis shows that HFC and PFC emissions from Substitutes of Ozone Depleting
Substances, in addition to CO2 from coal-fired electricity generation and CO2 from gas fired electricity generation,
and CO2 from road transport related combustion are also significant with respect to trends over the time series.
When considering the contribution of the LULUCF sector to 2021 emissions and sinks, 42 categories accounted for
95 percent of emissions and sinks, with the most significant category from LULUCF being net CO2 emission from
Forest Land Remaining Forest Land. When considering uncertainties and the contribution of the LULUCF sector,
additional categories such as CO2 emissions from Grasslands Remaining Grasslands were also identified as a key
category. In the trend analysis, 39 categories were identified as key, and when considering uncertainties, 8
additional categories were identified as key. The trend analysis includes additional categories such as non-CC>2
emissions from forest fires as key categories in the LULUCF sector.
37 See Chapter 4 Volume 1, "Methodological Choice and Identification of Key Categories" in IPCC (2006). See http://www.ipcc-
nggip.iges.or.jp/public/2006gl/index. html.
1-18 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Finally, in addition to conducting Approach 1 and 2 level and trend assessments as described above, a qualitative
2 assessment of the source and sinks categories was conducted to capture any additional key categories that were
3 not identified using the previously described quantitative approaches. For this Inventory, no additional categories
4 were identified using qualitative criteria recommend by IPCC, but EPA continues to review its qualitative
5 assessment on an annual basis. Find more information on the key category analysis, including the approach to
6 disaggregation of inventory estimates, see Annex 1 to this report.
7 Table 1-4: Summary of Key Categories for the United States (1990 and 2021) by Sector
CRF Code and
Source/Sink
Categories
Gas
Approach 1
Approach 2 (includes uncertainty)
2021
Emissions
(MMT
C02 Eq.)
Level Trend Level Trend
Without Without With With
LULUCF LULUCF LULUCF LULUCF
Level Trend Level Trend
Without Without With With
LULUCF LULUCF LULUCF LULUCF
Energy
l.A.3.b CO2
Emissions from
Transportation: Road
C02
• • • •
• • • •
1,482.3
l.A.l C02 Emissions
from Stationary
Combustion - Coal -
Electricity
Generation
C02
• • • •
• • • •
909.7
l.A.l C02 Emissions
from Stationary
Combustion - Gas -
Electricity
Generation
C02
• • • •
• • • •
615.1
1.A.2 C02 Emissions
from Stationary
Combustion - Gas -
Industrial
C02
• • • •
• • • •
498.4
l.A.4.b C02
Emissions from
Stationary
Combustion - Gas -
Residential
C02
• • • •
• •
258.6
1.A.2 C02 Emissions
from Stationary
Combustion - Oil -
Industrial
C02
• • • •
• • • •
220.3
l.A.4.a C02
Emissions from
Stationary
Combustion - Gas -
Commercial
C02
• • • •
• • • •
180.9
l.A.3.a C02
Emissions from
Transportation:
Aviation
C02
• • • •
• •
164.5
1.A.5 C02 Emissions
from Non-Energy
Use of Fuels
C02
• • • •
• • • •
143.2
l.A.3.e C02
Emissions from
Transportation:
Other
C02
• • • •
•
64.2
Introduction 1-19
-------
CRF Code and
Source/Sink
Categories
Gas
Approach 1
Approach 2 (includes uncertainty)
2021
Emissions
(MMT
C02 Eq.)
Level Trend Level Trend
Without Without With With
LULUCF LULUCF LULUCF LULUCF
Level Trend Level Trend
Without Without With With
LULUCF LULUCF LULUCF LULUCF
l.A.4.b C02
Emissions from
Stationary
Combustion - Oil -
Residential
C02
• • • •
• •
51.5
1.A.2 C02 Emissions
from Stationary
Combustion - Coal -
Industrial
C02
• • • •
• • • •
43.7
l.A.4.a C02
Emissions from
Stationary
Combustion - Oil -
Commercial
C02
• • • •
•
41.6
l.A.3.d C02
Emissions from
Transportation:
Domestic Navigation
C02
• •
41.0
1.B.2 C02 Emissions
from Natural Gas
Systems
C02
• •
•
36.8
1.A.3.C C02
Emissions from
Transportation:
Railways
C02
• •
32.1
1.B.2 C02 Emissions
from Petroleum
Systems
C02
• • • •
• •
24.7
1.A.5 C02 Emissions
from Stationary
Combustion - Oil -
U.S. Territories
C02
• •
17.5
l.A.l C02 Emissions
from Stationary
Combustion - Oil -
Electricity
Generation
C02
• • • •
• • •
17.1
l.A.5.b C02
Emissions from
Transportation:
Military
C02
• •
5.2
l.A.4.a C02
Emissions from
Stationary
Combustion - Coal -
Commercial
C02
• •
1.4
l.A.4.b C02
Emissions from
Stationary
Combustion - Coal -
Residential
C02
• •
NO
1-20 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
CRF Code and
Source/Sink
Categories
Gas
Approach 1
Approach 2 (includes uncertainty)
2021
Emissions
(MMT
C02 Eq.)
Level Trend Level Trend
Without Without With With
LULUCF LULUCF LULUCF LULUCF
Level Trend Level Trend
Without Without With With
LULUCF LULUCF LULUCF LULUCF
1.B.2 CH4 Emissions
from Natural Gas
Systems
ch4
• • • •
• • • •
181.4
1.B.2 CH4 Emissions
from Petroleum
Systems
ch4
• •
• •
50.2
l.B.l Fugitive
Emissions from Coal
Mining
ch4
• • • •
• • • •
44.7
1.B.2 CH4 Emissions
from Abandoned Oil
and Gas Wells
ch4
• •
8.2
l.A.4.b CH4
Emissions from
Stationary
Combustion -
Residential
ch4
• • • •
4.6
l.A.l N20 Emissions
from Stationary
Combustion - Coal -
Electricity
Generation
n2o
•
15.1
l.A.3.b N20
Emissions from
Transportation: Road
n2o
• • • •
• • •
9.6
l.A.l N20 Emissions
from Stationary
Combustion - Gas -
Electricity
Generation
n2o
•
3.9
Industrial Processes and Product Use
2.C.1 C02 Emissions
from Iron and Steel
Production &
Metallurgical Coke
Production
co2
• • • •
• • • •
42.0
2.A.1 C02 Emissions
from Cement
Production
co2
• • • •
41.3
2.B.8 C02 Emissions
from Petrochemical
Production
co2
• • • •
33.2
2.B.3 N20 Emissions
from Adipic Acid
Production
n2o
•
6.6
2.F.1 Emissions from
Substitutes for
Ozone Depleting
Substances:
Refrigeration and Air
conditioning
HFCs,
PFCs
• • • •
• • • •
139.1
Introduction 1-21
-------
CRF Code and
Source/Sink
Categories
Gas
Approach 1
Approach 2 (includes uncertainty)
2021
Emissions
(MMT
C02 Eq.)
Level Trend Level Trend
Without Without With With
LULUCF LULUCF LULUCF LULUCF
Level Trend Level Trend
Without Without With With
LULUCF LULUCF LULUCF LULUCF
2.F.4 Emissions from
Substitutes for
Ozone Depleting
Substances: Aerosols
HFCs,
PFCs
• • • •
• • • •
17.7
2.F.2 Emissions from
Substitutes for
Ozone Depleting
Substances: Foam
Blowing Agents
HFCs,
PFCs
• •
10.8
2.G SF6 and CF4
Emissions from
Electrical
Transmission and
Distribution
sf6,
cf4
• •
•
6.0
2.E PFC, HFC, SF6,
and NFs Emissions
from Electronics
Industry
PFCs,
HFCs,
SFe,
NF3
• •
•
4.5
2.B.9 HFC-23
Emissions from
HCFC-22 Production
HFCs
• • • •
• •
2.2
2.C.4 SF6 Emissions
from Magnesium
Production and
Processing
sf6
• •
1.1
2.C.3 PFC Emissions
from Aluminum
Production
PFCs
• •
0.9
Agriculture
3.A.1 CH4 Emissions
from Enteric
Fermentation: Cattle
ch4
• • • •
• • •
188.2
3.B.1 CH4 Emissions
from Manure
Management: Cattle
ch4
• • • •
• • •
37.9
3.B.4 CH4 Emissions
from Manure
Management: Other
Livestock
ch4
• • •
28.1
3.C CH4 Emissions
from Rice Cultivation
ch4
•
• •
16.8
3.D.1 Direct N20
Emissions from
Agricultural Soil
Management
n2o
• •
• • • •
257.7
3.D.2 Indirect N20
Emissions from
Applied Nitrogen
n2o
• •
• • • •
27.5
Waste
5.A CH4 Emissions
from MSW Landfills
ch4
• • • •
• • • •
103.7
1-22 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
CRF Code and
Source/Sink
Categories
Gas
Approach 1
Level Trend Level Trend
Without Without With With
LULUCF LULUCF LULUCF LULUCF
Approach 2 (includes uncertainty)
Level Trend
Without Without
LULUCF LULUCF
Level Trend
With With
LULUCF LULUCF
2021
Emissions
(MMT
C02 Eq.)
5.A CH4 Emissions
from Industrial
Landfills
5.D CH4 Emissions
from Domestic
Wastewater
Treatment
CH4
ch4
18.9
13.9
5.D N20 Emissions
from Domestic
Wastewater
Treatment
N20
20.4
Land Use, Land-Use Change, and Forestry
CO
• •
• •
81.0
co2
•
•
56.5
CO
• •
10.0
co2
•
• •
(18.9)
CO
• •
• •
(24.7)
co2
•
•
(98.3)
CO
• •
• •
(134.5)
co2
• •
• •
(695.4)
CH-
•
45.4
4.E.2 Net C02
Emissions from Land
Converted to
Settlements
4.B.2 Net C02
Emissions from Land
Converted to
Cropland
4.C.1 Net C02
Emissions from
Grassland Remaining
Grassland
4.B.1 Net C02
Emissions from
Cropland Remaining
Cropland
4.C.2 Net C02
Emissions from Land
Converted to
Grassland
4.A.2 Net C02
Emissions from Land
Converted to Forest
Land
4.E.1 Net C02
Emissions from
Settlements
Remaining
Settlements
4.A.1 Net C02
Emissions from
Forest Land
Remaining Forest
Land
4.D.1 CH4 Emissions
from Flooded Land
Remaining Flooded
Land
Introduction 1-23
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Approach 1 Approach 2 (includes uncertainty) 2021
CRF Code and Level Trend Level Trend Level Trend Level Trend Emissions
Source/Sink Without Without With With Without Without With With (MMT
Categories Gas LULUCF LULUCF LULUCF LULUCF LULUCF LULUCF LULUCF LULUCF CP2 Eq.)
4.A.1 CH4 Emissions
from Forest Fires
4.A.1 N20 Emissions
from Forest Fires
Subtotal of Key Categories Without LULUCFb
6,172.6
Total Gross Emissions Without LULUCF
6,347.7
Percent of Total Without LULUCF
97%
Subtotal of Key Categories With LULUCFC
5,393.2
Total Net Emissions With LULUCF
5,593.5
Percent of Total With LULUCF
96%
NO (Not Occurring)
a Other includes emissions from pipelines.
b Subtotal includes key categories from Level Approach 1 Without LULUCF, Level Approach 2 Without LULUCF, Trend Approach
1 Without LULUCF, and Trend Approach 2 Without LULUCF.
c Subtotal includes key categories from Level Approach 1 With LULUCF, Level Approach 2 With LULUCF, Trend Approach 1 With
LULUCF, and Trend Approach 2 With LULUCF.
Note: Parentheses indicate negative values (or sequestration).
CH4
n2o
15.5
8.9
1.6 Quality Assurance and Quality Control
(QA/QC)
As part of efforts to achieve its stated goals for inventory quality, transparency, and credibility, the United States
has developed a quality assurance and quality control plan designed to check, document, and improve the quality
of its inventory over time. QA/QC activities on the Inventory are undertaken within the framework of the U.S.
Quality Assurance/Quality Control and Uncertainty Management Plan (QA/QC plan)/or the U.S. Greenhouse Gas
Inventory: Procedures Manual for QA/QC and Uncertainty Analysis.
Key attributes of the QA/QC plan are summarized in Figure 1-2. These attributes include:
• Procedures and Forms: detailed and specific systems that serve to standardize the process of
documenting and archiving information, as well as to guide the implementation of QA/QC and the
analysis of uncertainty
• Implementation of Procedures: application of QA/QC procedures throughout the whole Inventory
development process from initial data collection, through preparation of the emission and removal
estimates, to publication of the Inventory
• Quality Assurance (QA): expert and public reviews for both the inventory estimates and the Inventory
report (which is the primary vehicle for disseminating the results of the inventory development process).
The expert technical review conducted by the UNFCCC supplements these QA processes, consistent with
the QA good practice and the 2006IPCC Guidelines (IPCC 2006)
• Quality Control (QC): application of General (Tier 1) and Category-specific (Tier 2) quality controls and
checks, as recommended by 2006 IPCC Guidelines (IPCC 2006), along with consideration of secondary data
and category-specific checks (additional Tier 2 QC) in parallel and coordination with the uncertainty
assessment; the development of protocols and templates, which provides for more structured
communication and integration with the suppliers of secondary information
• General (Tier 1) and Category-specific (Tier 2) Checks: quality controls and checks, as recommended by
1-24 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
IPCC Good Practice Guidance and 2006IPCC Guidelines (IPCC 2006)
• Record Keeping: provisions to track which procedures have been followed, the results of the QA/QC,
uncertainty analysis, and feedback mechanisms for corrective action based on the results of the
investigations which provide for continual data quality improvement and guided research efforts.
• Multi-Year Implementation: a schedule for coordinating the application of QA/QC procedures across
multiple years, especially for category-specific QC, prioritizing key categories
• Interaction and Coordination: promoting communication within the EPA, across Federal agencies and
departments, state government programs, and research institutions and consulting firms involved in
supplying data or preparing estimates for the Inventory. The QA/QC Management Plan itself is intended
to be revised and reflect new information that becomes available as the program develops, methods are
improved, or additional supporting documents become necessary.
Introduction 1-25
-------
1 Figure 1-2: U.S. QA/QC Plan Summary
Data
Data
Calculating
Gathering
4
Documentation (
*
Emissions
• Obtain data in
•
Contact reports
• Clearly label
electronic
for non-electronic
parameters, units,
format (if
communications
and conversion
possible)
•
Provide cell
factors
• Review data
references for
• Review data
input/calculation
primary data
input/calculation
workbooks
elements
workbooks
o Avoid
•
Obtain copies of
integrity
hardwiring
all data sources
o Equations
o Use data
•
List and location
o Units
+-»
{/)
>
validation
of any
o Inputs and
~ru
o Protect cells
working/external
outputs
c
<
• Develop
data or
• Develop
>.
automatic
input/calculation
automated
o
-1—>
checkers for:
workbooks
checkers for:
c
(U
o Outliers,
•
Document
o Input ranges
>
c
negative
assumptions
o Calculations
values, or
missing data
•
Complete QA/QC
checklists
o Emission
aggregation
o Variable types
•
CRF and summary
o Trend and IEF
match values
tab links
checks
o Time series
consistency
• Maintain
tracking tab for
status of
gathering efforts
• Check input data
•
Check citations in
• Reproduce
for transcription
data
calculations
errors
input/calculation
• Review time
• Inspect
workbooks and
series
automatic
text for accuracy
consistency
checkers
and style
• Review changes
+->
ISl
• Identify data
•
Check reference
in
03
input/calculation
docket for new
data/consistency
c
<
workbooks
citations
with IPCC
u
modifications
•
Review
methodology
a
§
that could
provide
additional
QA/QC checks
•
•
documentation
for any data /
methodology
changes
Complete QA/QC
checklists
CRF and summary
tab links
Cross-Cutting
Coordination
• Common starting
versions for each
inventory year
• Utilize
unalterable
summary and
CRF tab for each
source data
input/calculation
workbook for
linking to a
master summary
workbook
• Follow strict
version control
procedures
• Document
QA/QC
procedures
1-26 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
Box 1-3: Examples of Verification Activities
Consistent with IPCC guidance for national GHG inventories, verification activities include comparisons with
emission or removal estimates prepared by other bodies and comparisons with estimates derived from fully
independent assessments, e.g., atmospheric concentration measurements. Verification activities provide
information to improve inventories and are part of the overall QA/QC system.
Use of lowertier methods. The UNFCCC reporting guidelines require countries to complete a "top-down"
reference approach for estimating CO2 emissions from fossil fuel combustion in addition to their "bottom-up"
sectoral methodology for purposes of verification. This estimation method uses alternative methodologies and
different data sources than those contained in that section of the Energy chapter. The reference approach
estimates fossil fuel consumption by adjusting national aggregate fuel production data for imports, exports, and
stock changes rather than relying on end-user consumption surveys (see Annex 4 of this report). The reference
approach assumes that once carbon-based fuels are brought into a national economy, they are either saved in
some way (e.g., stored in products, kept in fuel stocks, or left unoxidized in ash) or combusted, and therefore
the carbon in them is oxidized and released into the atmosphere. Accounting for actual consumption of fuels at
the sectoral or sub-national level is not required.
Use of Ambient Measurements Systems for Validation of Emission Inventories. In following the UNFCCC
requirement under Article 4.1 to develop and submit national greenhouse gas emission inventories, the
emissions and sinks presented in this report are organized by source and sink categories and calculated using
internationally accepted methods provided by the IPCC.38 Several recent studies have estimated emissions at
the national or regional level with estimated results that sometimes differ from EPA's estimate of emissions.
EPA has engaged with researchers on how remote sensing, ambient measurement, and inverse modeling
techniques for estimating greenhouse gas emissions could assist in improving the understanding of inventory
estimates. In working with the research community to improve national greenhouse gas inventories, EPA
follows guidance from the IPCC on the use of measurements and modeling to validate emission inventories.39
An area of particular interest in EPA's outreach efforts is how ambient measurement data can be used to assess
estimates or potentially be incorporated into the Inventory in a manner consistent with this Inventory report's
transparency of its calculation methodologies, and the ability of inverse modeling to attribute emissions and
removals from remote sensing to anthropogenic sources, as defined by the IPCC for this report, versus natural
sources and sinks.
The 2019 Refinement to the IPCC 2006 Guidelines for National Greenhouse Gas Inventories (IPCC 2019) Volume
1 General Guidance and Reporting, Chapter 6: Quality Assurance, Quality Control and Verification notes that
atmospheric concentration measurements can provide independent data sets as a basis for comparison with
inventory estimates. The 2019 Refinement provides guidance on conducting such comparisons (as summarized
in Table 6.2 of IPCC [2019] Volume 1, Chapter 6) and provides guidance on using such comparisons to identify
areas of improvement in national inventories (as summarized in Box 6.5 of IPCC [2019] Volume 1, Chapter 6)
given the technical complexity of such comparisons. Further, it identified fluorinated gases as particularly
suitable for such comparisons. The 2019 Refinement makes this conclusion for fluorinated gases based on their
lack of significant natural sources, their generally long atmospheric lifetimes, their well-known loss mechanisms,
and the potential uncertainties in bottom-up inventory methods for some of their sources. Unlike emissions of
CO2, Cm, and N2O, emissions of fluorinated greenhouse gases are almost exclusively anthropogenic, meaning
that the fluorinated GHG emission sources included in this Inventory account for the majority of the total U.S.
emissions of these gases detectable in the atmosphere.
In this Inventory, EPA presents the results of two comparisons between fluorinated gas emissions inferred from
atmospheric measurements and fluorinated gas emissions estimated based on bottom-up measurements and
modeling consistent with guidance from the 2019 Refinement. These comparisons, performed for HFCs and SF6
respectively, are described under the QA/QC and Verification discussions in Chapter 4, Sections 4.24
38 See http://www.ipcc-negip.iges.or.jp/public/index.html.
39 See http://www.ipcc-nggip.iges.or.jp/meeting/pdfiles/1003 Uncertaintv%20meeting report.pdf.
Introduction 1-27
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Substitution of Ozone Depleting Substances and 4.25 Electrical Transmission and Distribution in the IPPU
chapter of this report.
Consistent with the 2019 Refinement, a key element to facilitate such comparisons is a gridded prior inventory
as an input to inverse modeling. To improve the ability to compare the national-level greenhouse gas inventory
with measurement results that may be at other scales, a team at Harvard University along with EPA and other
coauthors developed a gridded inventory of U.S. anthropogenic methane emissions with 0.1° x 0.1° spatial
resolution, monthly temporal resolution, and detailed scale-dependent error characterization. The gridded
inventory is designed to be consistent with the 1990 to 2014 U.S. EPA Inventory of U.S. Greenhouse Gas
Emissions and Sinks estimates for the year 2012, which presents national totals for different source types.40 This
gridded inventory is consistent with the recommendations contained in two National Academies of Science
reports examining greenhouse gas emissions data (National Research Council 2010; National Academies of
Sciences, Engineering, and Medicine 2018).
Finally, in addition to use of atmospheric concentration measurement data for comparison with Inventory data,
information from top-down studies is directly incorporated in the Natural Gas Systems calculations to quantify
emissions from certain well blowout events.
In addition, based on the national QA/QC plan for the Inventory, some sector, subsector and category-specific
QA/QC and verification checks have been developed. These checks follow the procedures outlined in the national
QA/QC plan, tailoring the procedures to the specific documentation and data files associated with individual
sources. For each greenhouse gas emissions source or sink category included in this Inventory, a minimum of
general or Tier 1 QC analysis has been undertaken. Where QC activities for a particular category go beyond the
minimum Tier 1 level, and include category-specific checks (Tier 2) or include verification, further explanation is
provided within the respective source or sink category text. Similarly, responses or updates based on comments
from the expert, public and the international technical expert reviews (e.g., UNFCCC) are also addressed within the
respective source or sink category sections in each sectoral chapter and Annex 8.
The quality control activities described in the U.S. QA/QC plan occur throughout the inventory process; QA/QC is
not separate from, but is an integral part of, preparing the Inventory. Quality control—in the form of both good
practices (such as documentation procedures) and checks on whether good practices and procedures are being
followed—is applied at every stage of inventory development and document preparation. In addition, quality
assurance occurs during the Expert Review and the Public Review, in addition to the UNFCCC expert technical
review. While all phases significantly contribute to improving inventory quality, the public review phase is also
essential for promoting the openness of the inventory development process and the transparency of the inventory
data and methods.
The QA/QC plan guides the process of ensuring inventory quality by describing data and methodology checks,
developing processes governing peer review and public comments, and developing guidance on conducting an
analysis of the uncertainty surrounding the emission and removal estimates. The QA/QC procedures also include
feedback loops and provide for corrective actions that are designed to improve the inventory estimates over time.
40 See https://www.epa.gov/ghgemissions/gridded-2012-methane-emissions.
1-28 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
1.7 Uncertainty Analysis of Emission
Estimates
Emissions and removals calculated for the U.S. Inventory reflect best estimates for greenhouse gas source and sink
categories in the United States and are continuously revised and improved as new information becomes available.
Uncertainty assessment is an essential element of a complete and transparent emissions inventory because it
helps inform and prioritize Inventory improvements. For the U.S. Inventory, uncertainty analyses are conducted for
each source and sink category as well as for the uncertainties associated with the overall emission (current and
base year) and trends estimates. These analyses reflect the quantitative uncertainty in the emission (and removal)
estimates associated with uncertainties in their input parameters (e.g., activity data and EFs) and serve to evaluate
the relative contribution of individual input parameter uncertainties to the overall Inventory, its trends, and each
source and sink category.
The overall level and trend uncertainty estimates for total U.S. greenhouse gas emissions was developed using the
IPCC Approach 2 uncertainty estimation methodology (assuming a Normal distribution for Approach 1 estimates),
which employs a Monte Carlo Stochastic Simulation technique. The IPCC provides good practice guidance on two
approaches—Approach 1 and Approach 2—to estimating uncertainty for both individual and combined source
categories. Approach 2 quantifies uncertainties based on a distribution of emissions (or removals), built-up from
repeated calculations of emission estimation models and the underlying input parameters, randomly selected
according to their known distributions. Approach 2 methodology is applied to each individual source and sink
category wherever data and resources are permitted and is also used to quantify the uncertainty in the overall
Inventory and its Trends. Source and sink chapters in this report provide additional details on the uncertainty
analysis conducted for each source and sink category. See Annex 7 of this report for further details on the U.S.
process for estimating uncertainty associated with the overall emission (base and current year) and trends
estimates. Consistent with IPCC (IPCC 2006), the United States has ongoing efforts to continue to improve the
overall Inventory uncertainty estimates presented in this report.
The United States has also implemented many improvements over the last several years to reduce uncertainties
across the source and sink categories and improve Inventory estimates. These improvements largely result from
new data sources that provide more accurate data and/or increased data coverage, as well as methodological
improvements. Following IPCC good practice, additional efforts to reduce Inventory uncertainties can occur
through efforts to incorporate excluded emission and sink categories (see Annex 5), improve estimation methods,
and collect more detailed, measured, and representative data. Individual category chapters and Annex 7 both
describe current ongoing and planned Inventory and uncertainty analysis improvements. Consistent with IPCC
(2006), the United States has ongoing efforts to continue to improve the category-specific uncertainty estimates
presented in this report, largely prioritized by considering improvements categories identified as significant by the
Key Category Analysis.
Estimates of quantitative uncertainty for the total U.S. greenhouse gas emissions in 1990 (base year) and 2020 are
shown below in Table 1-5 and Table 1-6, respectively. The overall uncertainty surrounding the Total Net Emissions
is estimated to be -5 to +6 percent in 1990 and -6 to +6 percent in 2020. When the LULUCF sector is excluded from
the analysis the uncertainty is estimated to be -2 to +5 percent in 1990 and -3 to +3 percent in 2020.
Table 1-5: Estimated Overall Inventory Quantitative Uncertainty for 1990 (MMT CO2 Eq. and
Percent) - TO BE UPDATED FOR FINAL INVENTORY REPORT
1990 Emission Uncertainty Range Relative to Greenhouse Gas
Standard
Estimate Estimate3
Meanb Deviation'5
Gas
(MMT C02
Eq.) (MMT C02 Eq.) (%)
(MMT C02 Eq.)
Lower Upper Lower Upper
Introduction 1-29
-------
Bound11 Bound11 Bound Bound
co2
5,122.5
5,017.3
5,357.6
-2%
5%
5,186.5
88.0
CH4d
780.8
720.1
871.5
-8%
12%
794.9
38.8
N2Od
450.5
365.6
574.9
-19%
28%
457.8
54.1
PFC, HFC, SF6, and NF3d
99.7
90.2
112.5
-9%
13%
100.4
5.6
Total Gross Emissions
6,453.5
6,330.2
6,761.5
-2%
5%
6,539.5
110.6
LULUCF Emissions6
31.4
29.3
33.8
-7%
8%
31.5
1.1
LULUCF Carbon Stock Change Fluxf
(892.0)
(1,183.9)
(709.3)
33%
-20%
(944.1)
119.3
LULUCF Sector Net Totals
(860.6)
(1,152.7)
(677.7)
34%
-21%
(912.6)
119.3
Net Emissions (Sources and Sinks)
5,592.8
5,306.8
5,953.6
-5%
6%
5,626.9
163.9
a The lower and upper bounds for emission estimates correspond to a 95 percent confidence interval, with the lower bound
corresponding to 2.5th percentile and the upper bound corresponding to 97.5th percentile.
b Mean value indicates the arithmetic average of the simulated emission estimates; standard deviation indicates the extent of
deviation of the simulated values from the mean.
c The lower and upper bound emission estimates for the sub-source categories do not sum to total emissions because the low
and high estimates for total emissions were calculated separately through simulations.
d The overall uncertainty estimates did not take into account the uncertainty in the GWP values for CH4, N20 and high GWP
gases used in the Inventory emission calculations for 1990.
e LULUCF emissions include the CH4 and N20 emissions reported for Peatlands Remaining Peatlands, forest fires, drained
organic soils, grassland fires, and Coastal Wetlands Remaining Coastal Wetlands; CH4 emissions from Land Converted to
Coastal Wetlands, Land Converted to Flooded Land, and Flooded Land Remaining Flooded Land; and N20 emissions from
forest soils and settlement soils.
f LULUCF Carbon Stock Change is the net C stock change from the following categories: Forest Land Remaining Forest Land,
Land Converted to Forest Land, Cropland Remaining Cropland, Land Converted to Cropland, Grassland Remaining Grassland,
Land Converted to Grassland, Wetlands Remaining Wetlands, Land Converted to Wetlands, Settlements Remaining
Settlements, and Land Converted to Settlements. Since the resulting flux is negative the signs of the resulting lower and upper
bounds are reversed.
g The LULUCF Sector Net Total is the net sum of all CH4 and N20 emissions to the atmosphere plus net carbon stock changes.
Notes: Total emissions (excluding emissions for which uncertainty was not quantified) are presented without LULUCF. Net
emissions are presented with LULUCF. Totals may not sum due to independent rounding. Parentheses indicate net
sequestration.
1 Table 1-6: Estimated Overall Inventory Quantitative Uncertainty for 2020 (MMT CO2 Eq. and
2 Percent) - TO BE UPDATED FOR FINAL INVENTORY REPORT
2020
Emission
Uncertainty Range Relative to Greenhouse
Standard
Estimate
Gas Estimate3
Mean'5
Deviation1*
Gas
(MMT C02
Eq.)
(MMT C02
Eq.)
(%)
(MMT C02 Eq.)
Lower
Upper
Lower
Upper
Boundc
Boundc
Bound
Bound
C02
4,715.7
4,610.6
4,908.0
-3%
3%
4,759.8
76.4
CH4d
650.4
595.9
723.6
-10%
10%
659.7
32.6
N2Od
426.1
342.4
551.1
-21%
27%
436.1
53.3
PFC, HFC, SF6, and NF3d
189.2
182.6
213.7
-8%
8%
198.2
7.9
Total Gross Emissions
5,981.4
5,863.8
6,253.0
-3%
3%
6,053.7
98.2
LULUCF Emissions6
53.2
44.4
62.9
-17%
18%
53.5
4.9
LULUCF Carbon Stock Change Fluxf
(812.2)
(1,075.7)
(647.8)
25%
-25%
(860.2)
109.4
LULUCF Sector Net Totals
(758.9)
(1,023.2)
(594.5)
27%
-26%
(806.7)
109.6
Net Emissions (Sources and Sinks)
5,222.4
4,956.9
5,540.9
-6%
6%
5,247.0
148.1
1-30 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
a The lower and upper bounds for emission estimates correspond to a 95 percent confidence interval, with the lower bound
corresponding to 2.5th percentile and the upper bound corresponding to 97.5th percentile.
b Mean value indicates the arithmetic average of the simulated emission estimates; standard deviation indicates the extent of
deviation of the simulated values from the mean.
c The lower and upper bound emission estimates for the sub-source categories do not sum to total emissions because the low
and high estimates for total emissions were calculated separately through simulations.
d The overall uncertainty estimates did not take into account the uncertainty in the GWP values for CH4, N20 and high GWP
gases used in the Inventory emission calculations for 2020.
e LULUCF emissions include the CH4 and N20 emissions reported for Peatlands Remaining Peatlands, forest fires, drained organic
soils, grassland fires, and Coastal Wetlands Remaining Coastal Wetlands; CH4 emissions from Land Converted to Coastal
Wetlands, Land Converted to Flooded Land, and Flooded Land Remaining Flooded Land; and N20 emissions from forest soils
and settlement soils.
f LULUCF Carbon Stock Change is the net C stock change from the following categories: Forest Land Remaining Forest Land, Land
Converted to Forest Land, Cropland Remaining Cropland, Land Converted to Cropland, Grassland Remaining Grassland, Land
Converted to Grassland, Wetlands Remaining Wetlands, Land Converted to Wetlands, Settlements Remaining Settlements, and
Land Converted to Settlements. Since the resulting flux is negative the signs of the resulting lower and upper bounds are
reversed.
s The LULUCF Sector Net Total is the net sum of all CH4 and N20 emissions to the atmosphere plus net carbon stock changes.
Notes: Total emissions (excluding emissions for which uncertainty was not quantified) are presented without LULUCF. Net
emissions are presented with LULUCF. Totals may not sum due to independent rounding. Parentheses indicate net
sequestration.
In addition to the estimates of uncertainty associated with the current and base year estimates, Table 1-7 presents
the estimates of inventory trend uncertainty. The 2006IPCC Guidelines defines trend as the difference in emissions
between the base year (i.e., 1990) and the current year (i.e., 2020) Inventory estimates. However, for purposes of
understanding the concept of trend uncertainty, the trend is defined in this Inventory as the percentage change in
the gross emissions (or net emissions) estimated for the current year, relative to the gross emission (or net
emissions) estimated for the base year. The uncertainty associated with this trend is referred to as trend
uncertainty and is reported as between -14 and 1 percent at the 95 percent confidence level between 1990 and
2020. This indicates a range of approximately -7 percent below and 8 percent above the trend estimate of -7
percent. See Annex 7 for trend uncertainty estimates for individual source and sink categories by gas.
Table 1-7: Quantitative Assessment of Trend Uncertainty (MMT CO2 Eq. and Percent)
Base Year
2020
Emissions
Gas/Source
Emissions3
Emissions
Trend
Trend Rangeb
(MMT C02 Eq.)
(%)
1
[%)
Lower
Upper
Bound
Bound
C02
5,122.5
4,715.7
-8%
-12%
-4%
ch4
780.8
650.4
-17%
-28%
-5%
n2o
450.5
426.1
-5%
-31%
32%
HFCs, PFCs, SF6, and NF3
99.7
189.2
90%
73%
125%
Total Gross Emissionsc
6,453.5
5,981.4
-7%
-12%
-3%
LULUCF Emissions'1
31.4
53.2
70%
39%
103%
LULUCF Carbon Stock Change Fluxe
(892.0)
(812.2)
-9%
-37%
30%
LULUCF Sector Net Total'
(860.6)
(758.9)
-12%
-40%
28%
Net Emissions (Sources and Sinks)c
5,592.8
5,222.4
-7%
-14%
1%
a Base Year is 1990 for all sources.
bThe trend range represents a 95 percent confidence interval for the emission trend, with the lower bound corresponding to
2.5th percentile value and the upper bound corresponding to 97.5th percentile value.
c Totals exclude emissions for which uncertainty was not quantified.
d LULUCF emissions include the CH4 and N20 emissions reported for Peatlands Remaining Peatlands, forest fires, drained
organic soils, grassland fires, and Coastal Wetlands Remaining Coastal Wetlands; CH
4 emissions from Land Converted to Coastal
Wetlands, Land Converted to Flooded Land, and Flooded Land Remaining Flooded Land; and N20 emissions from forest soils
and settlement soils.
Introduction 1-31
-------
1
2
3
4
5
6
7
8
e LULUCF Carbon Stock Change is the net C stock change from the following categories: Forest Land Remaining Forest Land,
Land Converted to Forest Land, Cropland Remaining Cropland, Land Converted to Cropland, Grassland Remaining Grassland,
Land Converted to Grassland, Wetlands Remaining Wetlands, Land Converted to Wetlands, Settlements Remaining
Settlements, and Land Converted to Settlements.
f The LULUCF Sector Net Total is the net sum of all CH4 and N20 emissions to the atmosphere plus net carbon stock changes.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration. Total emissions (excluding
emissions for which uncertainty was not quantified) are presented without LULUCF. Net emissions are presented with LULUCF.
9 1.8 Completeness
10 This report, along with its accompanying CRF tables, serves as a thorough assessment of the anthropogenic
11 sources and sinks of greenhouse gas emissions for the United States for the time series 1990 through 2021. This
12 report is intended to be comprehensive and includes the vast majority of emissions and removals identified as
13 anthropogenic, consistent with IPCC and UNFCCC guidelines. In general, sources or sink categories not accounted
14 for in this Inventory are excluded because they are not occurring in the United States and its territories, or because
15 data are unavailable to develop an estimate and/or the categories were determined to be insignificant41 in terms
16 of overall national emissions per UNFCCC reporting guidelines.
17 The United States is continually working to improve upon the understanding of such sources and sinks currently
18 not included and seeking to find the data required to estimate related emissions and removals, focusing on
19 categories that are anticipated to be significant. As such improvements are implemented, new emission and
20 removal estimates are quantified and included in the Inventory, improving completeness of national estimates. For
21 a list of sources and sink categories not included and more information on significance of these categories, see
22 Annex 5 and the respective category sections in each sectoral chapter of this report.
23 1.9 Organization of Report
24 In accordance with the revision of the UNFCCC reporting guidelines agreed to at the nineteenth Conference of the
25 Parties (UNFCCC 2014), this Inventory of U.S. Greenhouse Gas Emissions and Sinks is grouped into five sector-
26 specific chapters consistent with the UN Common Reporting Framework, listed below in Table 1-8. In addition,
27 chapters on Trends in Greenhouse Gas Emissions, Other information, and Recalculations and Improvements to be
28 considered as part of the U.S. Inventory submission are included.
29 Table 1-8: IPCC Sector Descriptions
Chapter (IPCC Sector) Activities Included
Energy Emissions of all greenhouse gases resulting from stationary and mobile energy
activities including fuel combustion and fugitive fuel emissions, and non-energy
use of fossil fuels.
Industrial Processes and Emissions resulting from industrial processes and product use of greenhouse
Product Use gases.
41 See paragraph 32 of Decision 24/CP.19, the UNFCCC reporting guidelines on annual inventories for Parties included in Annex
1 to the Convention. Paragraph notes that "...An emission should only be considered insignificant if the likely level of emissions
is below 0.05 per cent of the national total GHG emissions, and does not exceed 500 kt C02 Eq. The total national aggregate of
estimated emissions for all gases and categories considered insignificant shall remain below 0.1 percent of the national total
GHG emissions."
1-32 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Agriculture Emissions from agricultural activities except fuel combustion, which is
addressed under Energy.
Land Use, Land-Use Emissions and removals of C02, and emissions of CH4, and N20 from land use,
Change, and Forestry land-use change and forestry.
Waste Emissions from waste management activities.
1 Within each chapter, emissions are identified by the anthropogenic activity that is the source or sink of the
2 greenhouse gas emissions being estimated (e.g., coal mining). Overall, the following organizational structure is
3 consistently applied throughout this report:
4 Chapter/IPCC Sector: Overview of emissions and trends for each IPCC defined sector.
5 CRF Source or Sink Category: Description of category pathway and emission/removal trends based on IPCC
6 methodologies, consistent with UNFCCC reporting guidelines.
7 Methodology: Description of analytical methods (e.g., from 2006 IPCC Guidelines, or country-specific methods)
8 employed to produce emission estimates and identification of data references, primarily for activity data and
9 emission factors.
10 Uncertainty and Time-Series Consistency: A discussion and quantification of the uncertainty in emission estimates
11 and a discussion of time-series consistency.
12 QA/QC and Verification: A discussion on steps taken to QA/QC and verify the emission estimates, consistent with
13 the U.S. QA/QC plan, and any key QC findings.
14 Recalculations Discussion: A discussion of any data or methodological changes that necessitate a recalculation of
15 previous years' emission estimates, and the impact of the recalculation on the emission estimates, if applicable.
16 Planned Improvements: A discussion on any category-specific planned improvements, if applicable.
17 Special attention is given to C02 from fossil fuel combustion relative to other sources because of its share of
18 emissions and its dominant influence on emission trends. For example, each energy consuming end-use sector
19 (i.e., residential, commercial, industrial, and transportation), as well as the electricity generation sector, is
20 described individually. Additional information for certain source categories and other topics is also provided in
21 several Annexes listed in Table 1-9.
22 Table 1-9: List of Annexes
ANNEX 1 Key Category Analysis
ANNEX 2 Methodology and Data for Estimating C02 Emissions from Fossil Fuel Combustion
2.1. Methodology for Estimating Emissions of C02 from Fossil Fuel Combustion
2.2. Methodology for Estimating the Carbon Content of Fossil Fuels
2.3. Methodology for Estimating Carbon Emitted from Non-Energy Uses of Fossil Fuels
ANNEX 3 Methodological Descriptions for Additional Source or Sink Categories
3.1. Methodology for Estimating Emissions of CH4, N20, and Indirect Greenhouse Gases from Stationary
Combustion
3.2. Methodology for Estimating Emissions of CH4, N20, and Indirect Greenhouse Gases from Mobile
Combustion and Methodology for and Supplemental Information on Transportation-Related Greenhouse
Gas Emissions
3.3. Methodology for Estimating Emissions from Commercial Aircraft Jet Fuel Consumption
3.4. Methodology for Estimating CH4 Emissions from Coal Mining
3.5. Methodology for Estimating CH4 and C02 Emissions from Petroleum Systems
3.6. Methodology for Estimating CH4 Emissions from Natural Gas Systems
3.7. Methodology for Estimating C02 and N20 Emissions from Incineration of Waste
3.8. Methodology for Estimating Emissions from International Bunker Fuels used by the U.S. Military
3.9. Methodology for Estimating HFC and PFC Emissions from Substitution of Ozone Depleting Substances
3.10. Methodology for Estimating CH4 Emissions from Enteric Fermentation
3.11. Methodology for Estimating CH4 and N20 Emissions from Manure Management
3.12. Methodology for Estimating N20 Emissions, CH4 Emissions and Soil Organic C Stock Changes from
Introduction 1-33
-------
Agricultural Lands (Cropland and Grassland)
3.13. Methodology for Estimating Net Carbon Stock Changes in Forest Land Remaining Forest Land and Land
Converted to Forest Land
3.14. Methodology for Estimating CH4 Emissions from Landfills
ANNEX 4 IPCC Reference Approach for Estimating C02 Emissions from Fossil Fuel Combustion
ANNEX 5 Assessment of the Sources and Sinks of Greenhouse Gas Emissions Not Included
ANNEX 6 Additional Information
6.1. Global Warming Potential Values
6.2. Ozone Depleting Substance Emissions
6.3. Greenhouse Gas Precursors: Cross-Walk of NEI categories to the Inventory
6.4. Constants, Units, and Conversions
6.5. Chemical Formulas
ANNEX 7 Uncertainty
7.1. Overview
7.2. Methodology and Results
7.3. Reducing Uncertainty
7.4. Planned Improvements
7.5. Additional Information on Uncertainty Analyses by Source
ANNEX 8 QA/QC Procedures
8.1. Background
8.2. Purpose
8.3. Assessment Factors
8.4. Responses During the Review Process
ANNEX 9 Use of Greenhouse Gas Reporting Program (GHGRP) in Inventory
1-34 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
i 2. Trends in Greenhouse Gas Emissions
and Removals
3 2.1 Overview of U.S. Greenhouse Gas
4 Emissions and Sinks Trends
5 In 2021, total gross U.S. greenhouse gas emissions were 6,347.7 million metric tons of carbon dioxide equivalent
6 (MMT CO2 Eq).1 Total U.S. emissions have decreased by 2.0 percent from 1990 to 2021, down from a high of 15.8
7 percent above 1990 levels in 2007. Emissions increased from 2020 to 2021 by 5.5 percent (333.2 MMT CO2 Eq.).
8 Net emissions (i.e., including sinks) were 5,593.5 MMT CO2 Eq. in 2021. Overall, net emissions increased 6.8
9 percent from 2020 to 2021 and decreased 16.3 percent from 2005 levels, as shown in Table 2-1. Between 2020
10 and 2021, the increase in total greenhouse gas emissions was driven largely by an increase in CO2 emissions from
11 fossil fuel combustion due to economic activity rebounding after the COVID-19 pandemic. The CO2 emissions from
12 fossil fuel combustion increased by 7.0 percent from 2020 to 2021, including a 13.8 percent increase in
13 transportation sector emissions and a 7.1 percent increase in the electric power sector emissions. The increase in
14 electric power sector emissions was due to an increase in electricity demand of 2.1 percent since 2020, although
15 the overall decrease in electric power sector emissions from 1990 through 2021 reflects the combined impacts of
16 long-term trends in many factors, including population, economic growth, energy markets, technological changes
17 including energy efficiency, and the carbon intensity of energy fuel choices. Between 2019 and 2021, there was still
18 a decrease of 1.3 percent and 4.0 percent in CO2 emissions from fossil fuel combustion from the transportation
19 and electric power sectors, respectively.
20 Figure 2-1 and Figure 2-2 illustrate the overall trend in total U.S. emissions and sinks by gas and annual percent
21 changes relative to the previous year since 1990.
22
1 The gross emissions total presented in this report for the United States excludes emissions and sinks from removals from Land
Use, Land-Use Change, and Forestry (LULUCF). The net emissions total presented in this report for the United States includes
emissions and sinks from removals from LULUCF.
Trends 2-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
Figure 2-1; U.S. Greenhouse Gas Emissions and Sinks by Gas
9,000
8,000
7,000
6,000
ri- 5,000
8 4,000
3,000
2,000
1,000
0
-1,000
HFCs, PFCs, SFc and NFa
I Nitrous Oxide
I Methane
I Carbon Dioxide
I Net COz Flux from LULUCF=
¦ Net Emissions (including LULUCF sinks)
S a
r\j 00
cn
&
in vo
CT*
CT.
ooovo-rnrNjooTrLOvor^coc^Oir-irsifo^j-mvo
C^OOOOOOOOOOO*-<"^H-r-<-i-i^T-HT-H
ChC^OOOOOO 00 000000000
HHiNfMfMlNlNlNNrMfMlNfMlNfMINfNfMlNlNlNlNNlN
3 The term "flux" is used to describe the exchange of C02 to and from the atmosphere, with net flux being either positive or
negative depending on the overall balance. Removal and long-term storage of C02 from the atmosphere is also referred to as
"carbon sequestration."
Figure 2-2: Annual Percent Change in Gross U.S. Greenhouse Gas Emissions Relative to the
Previous Year
5.5%
-10%
1.6% 2.9%
1.6% 0.8%
i.8% 1.2%
CT> CTi CTi CTi CTi CTi CTiCTi
OiaOiOiOiOlOiOiOl
N(N(\l\fM(\fNfM(\(NNfMfM(NfMMfM(NNrMNN
Emissions and Sinks by Gas
Figure 2-3 illustrates the relative contribution of the greenhouse gases to total gross U.S. emissions in 2021, in CO2-
equivalents, i.e., weighted by global warming potential. The primary greenhouse gas emitted by human activities in
the United States is CO2, representing 79.5 percent of total greenhouse gas emissions. The largest source of CO2,
2-2 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 and of overall greenhouse gas emissions, is fossil fuel combustion primarily from transportation and power
2 generation. Methane (Cm) emissions account for 11.5 percent of emissions. The major sources of methane include
3 enteric fermentation associated with domestic livestock, natural gas systems, and decomposition of wastes in
4 landfills. Agricultural soil management, wastewater treatment, stationary sources of fuel combustion, and manure
5 management are the major sources of N2O emissions. Ozone depleting substance (ODS) substitute emissions were
6 the primary contributor to aggregate hydrofluorocarbon (HFC) emissions. Perfluorocarbon (PFC) emissions were
7 primarily attributable to electronics manufacturing and primary aluminum production. Electrical transmission and
8 distribution systems accounted for most sulfur hexafluoride (SFs) emissions. The electronics industry is the only
9 source of nitrogen trifluoride (NF3) emissions.
10 Figure 2-3: 2021 Gross Total U.S. Greenhouse Gas Emissions by Gas (Percentages based on
11 MMT COz Eq.)
3.0%
HFCs, PFCs, SFe and NFb
12
13 Note: Emissions and removals from Land Use, Land-Use Change, and Forestry are excluded from figure above.
14 Overall from 1990 to 2021, total emissions of CO2decreased by 73.3 MMT CO2 Eq. (1.4 percent), total emissions of
15 methane (CH4) decreased by 141.3 MMT CO2 Eq. (16.3 percent), and total emissions of nitrous oxide (N2O)
16 decreased by 11.8 MMT CO2 Eq. (3.0 percent). During the same period, emissions of fluorinated gases including
17 hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), sulfur hexafluoride (SFs), and nitrogen trifluoride (NF3) rose
18 by 95.9 MMT CO2 Eq. (104.8 percent). Despite being emitted in smaller quantities relative to the other principal
19 greenhouse gases, emissions of HFCs, PFCs, SF6, and NF3 are significant because many of them have extremely high
20 global warming potentials (GWPs), and, in the cases of PFCs, SF6, and NF3, very long atmospheric lifetimes.
21 Conversely, U.S. greenhouse gas emissions were partly offset by carbon (C) sequestration in managed forests,
22 trees in urban areas, agricultural soils, landfilled yard trimmings, and coastal wetlands. These were estimated to
23 offset 13.1 percent (832.0 MMT CO2 Eq.) of total gross emissions in 2021.
24 Table 2-1 provides information on trends in emissions and sinks from all U.S. anthropogenic sources in weighted
25 units of MMT CO2 Eq., while unweighted gas emissions and sinks in kilotons (kt) are provided in Table 2-2.
26 Table 2-1: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks (MMT CO2 Eq.)
Gas/Source
1990
2005
2017
2018
2019
2020
2021
CO,
5,121.4
6,132.4
5,212.1
5,378.0
5,259.8
4,714.4
5,048.2
Fossil Fuel Combustion
4,728.2
5,747.3
4,852.5
4,989.8
4,853.4
4,344.8
4,651.0
Transportation
1,468.9
1,858.6
1,780.1
1,812.9
1,813.9
1,572.5
1,789.4
Electric Power Sector
1,820.0
2,400.1
1,732.0
1,753.4
1,606.7
1,439.6
1,542.2
Industrial
852.4
850.8
789.0
813.5
815.9
767.9
762.4
Trends 2-3
-------
Residential
338.6
358.9
293.4
338.2
341.4
313.2
310.1
Commercial
228.3
227.1
232.0
245.8
250.7
228.5
223.9
U.S. Territories
20.0
51.9
25.9
25.9
24.8
23.2
23.0
Non-Energy Use of Fuels
112.4
128.9
112.8
129.4
121.6
119.2
143.2
Iron and Steel Production &
Metallurgical Coke Production
104.7
70.1
40.8
42.9
43.1
37.7
42.0
Cement Production
33.5
46.2
40.3
39.0
40.9
40.7
41.3
Natural Gas Systems
32.4
25.2
31.8
33.0
38.7
36.3
36.8
Petrochemical Production
21.6
27.4
28.9
29.3
30.7
29.8
33.2
Petroleum Systems
9.5
10.2
24.5
36.1
46.9
29.1
24.7
Incineration of Waste
12.9
13.3
13.2
13.3
12.9
12.9
12.5
Ammonia Production
14.4
10.2
12.5
12.7
12.4
13.0
12.2
Lime Production
11.7
14.6
12.9
13.1
12.1
11.3
11.9
Other Process Uses of Carbonates
6.2
7.5
9.9
7.4
8.4
8.4
8.0
Urea Fertilization
2.4
3.5
4.9
4.9
5.0
5.1
5.2
Carbon Dioxide Consumption
1.5
1.4
4.6
4.1
4.9
5.0
5.0
Urea Consumption for Non-
Agricultural Purposes
3.8
3.7
5.2
6.1
6.2
5.8
5.0
Liming
4.7
4.4
3.1
2.2
2.2
2.9
3.0
Coal Mining
4.6
4.2
3.2
3.1
3.0
2.2
2.5
Glass Production
2.3
2.4
2.0
2.0
1.9
1.9
2.0
Soda Ash Production
1.4
1.7
1.8
1.7
1.8
1.5
1.7
Ferroalloy Production
2.2
1.4
2.0
2.1
1.6
1.4
1.6
Aluminum Production
6.8
4.1
1.2
1.5
1.9
1.7
1.5
Titanium Dioxide Production
1.2
1.8
1.7
1.5
1.5
1.2
1.5
Zinc Production
0.6
1.0
0.9
1.0
1.0
1.0
1.0
Phosphoric Acid Production
1.5
1.3
1.0
0.9
0.9
0.9
0.9
Lead Production
0.5
0.6
0.5
0.5
0.5
0.5
0.4
Carbide Production and
Consumption
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Abandoned Oil and Gas Wells
+
+
+
+
+
+
+
Substitution of Ozone Depleting
Substances
+
+
+
+
+
+
+
Magnesium Production and
Processing
0.1
+
+
+
+
+
+
Biomass and Biofuel Consumptiona
237.9
245.4
328.9
336.0
333.1
305.6
313.3
International Bunker Fuelsb
103.6
113.3
120.2
122.2
116.1
69.6
69.3
CH4c
868.7
791.2
762.8
774.2
767.8
742.3
727A
Enteric Fermentation
183.1
188.2
195.9
196.8
197.3
196.2
194.9
Natural Gas Systems
215.1
203.4
186.4
194.4
193.6
185.4
181.4
Landfills
197.8
147.7
123.9
126.7
129.0
124.8
122.6
Manure Management
39.0
54.9
64.4
66.5
65.7
66.7
66.0
Petroleum Systems
51.3
50.9
61.9
60.6
59.9
54.5
50.2
Coal Mining
108.1
71.8
61.4
59.1
53.0
46.2
44.7
Wastewater Treatment
22.7
22.7
21.5
21.4
21.2
21.3
21.1
Rice Cultivation
17.9
20.2
16.7
17.4
16.9
17.6
16.8
Stationary Combustion
9.6
00
00
8.6
9.6
9.8
00
00
8.9
Abandoned Oil and Gas Wells
7.7
8.1
8.3
8.3
8.3
8.2
8.2
Abandoned Underground Coal
Mines
8.1
7.4
7.2
6.9
6.6
6.5
6.4
Mobile Combustion
7.2
4.4
2.9
2.9
2.9
2.6
2.6
Composting
0.4
2.1
2.7
2.5
2.5
2.6
2.6
Field Burning of Agricultural
Residues
0.4
0.5
0.5
0.5
0.5
0.5
0.5
Petrochemical Production
0.2
0.1
0.3
0.3
0.4
0.3
0.4
2-4 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Anaerobic Digestion at Biogas
Facilities
+
+
0.2
0.2
0.2
0.2
0.2
Ferroalloy Production
+
+
+
+
+
+
+
Carbide Production and
Consumption
+
+
+
+
+
+
+
Iron and Steel Production &
Metallurgical Coke Production
+
+
+
+
+
+
+
Incineration of Waste
+
+
+
+
+
+
+
International Bunker Fuelsb
0.2
0.1
0.1
0.1
0.1
0.1
0.1
N2Oc
396.7
405.1
402.8
418.5
399.1
377.7
384.8
Agricultural Soil Management
278.4
280.8
298.7
312.1
298.2
279.3
285.2
Stationary Combustion
22.3
30.5
25.3
25.1
22.2
20.7
22.1
Wastewater Treatment
14.8
18.1
20.6
21.2
21.3
20.9
20.9
Manure Management
12.4
14.5
16.9
17.2
17.4
17.5
17.4
Mobile Combustion
38.4
37.0
18.5
17.5
19.0
16.1
17.1
Nitric Acid Production
10.8
10.1
8.3
8.5
8.9
8.3
7.9
Adipic Acid Production
13.5
6.3
6.6
9.3
4.7
7.4
6.6
N20 from Product Uses
3.8
3.8
3.8
3.8
3.8
3.8
3.8
Composting
0.3
1.5
1.9
1.8
1.8
1.8
1.8
Caprolactam, Glyoxal, and Glyoxylic
Acid Production
1.5
1.9
1.3
1.3
1.2
1.2
1.2
Incineration of Waste
0.4
0.3
0.4
0.4
0.4
0.3
0.4
Electronics Industry
+
0.1
0.2
0.2
0.2
0.3
0.3
Field Burning of Agricultural
Residues
0.1
0.2
0.2
0.2
0.2
0.2
0.2
Petroleum Systems
+
+
+
+
+
+
+
Natural Gas Systems
+
+
+
+
+
+
+
International Bunker Fuelsb
0.8
0.9
0.9
1.0
0.9
0.5
0.5
HFCs
39.0
116.4
160.8
160.9
165.4
168.2
175.1
Substitution of Ozone Depleting
Substancesd
0.3
99.4
156.1
157.7
161.9
166.1
172.4
HCFC-22 Production
38.6
16.8
4.3
2.7
3.1
1.8
2.2
Electronics Industry
0.2
0.2
0.3
0.3
0.3
0.3
0.4
Magnesium Production and
Processing
NO
NO
0.1
0.1
0.1
0.1
+
PFCs
21.8
6.1
3.8
4.3
4.0
3.9
3.5
Electronics Industry
2.5
3.0
2.7
2.8
2.5
2.4
2.6
Aluminum Production
19.3
3.1
1.0
1.4
1.4
1.4
0.9
Substitution of Ozone Depleting
Substancesd
NO
+
+
+
+
+
+
Electrical Transmission and
Distribution
NO
+
+
NO
+
+
+
SF,
30.5
15.5
7.2
7.1
7.8
7.5
8.0
Electrical Transmission and
Distribution
24.7
11.8
5.5
5.2
6.1
5.9
6.0
Magnesium Production and
Processing
0.5
0.8
0.7
0.8
0.8
0.8
0.9
Electronics Industry
5.4
2.9
1.0
1.1
0.9
0.9
1.1
NF,
+
0.4
0.5
0.5
0.5
0.6
0.6
Electronics Industry
+
0.4
0.5
0.5
0.5
0.6
0.6
Total Gross Emissions (Sources)
6,478.3
7,466.9
6,550.0
6,743.4
6,604.4
6,014.5
6,347.7
LULUCF Emissions'1
57.9
72.4
68.3
64.4
64.2
76.4
77.8
ch4
53.5
61.3
60.1
57.3
56.9
65.4
66.0
n2o
4.4
11.1
8.3
7.0
7.3
11.0
11.8
LULUCF Carbon Stock Changed
(938.9)
(853.5)
(842.5)
(829.5)
(768.2)
(852.5)
(832.0)
LULUCF Sector Net Totale
(881.0)
(781.1)
(774.2)
(765.1)
(704.0)
(776.2)
(754.2)
Trends 2-5
-------
Net Emissions (Sources and Sinks) 5,597.3 6,685.8 5,775.8 5,978.3 5,900.3 5,238.3 5,593.5
+ Does not exceed 0.05 MMT C02 Eq.
NO (Not Occurring)
a Emissions from Biomass and Biofuel Consumption are not included specifically in Energy sector totals. Net carbon
fluxes from changes in biogenic carbon reservoirs are accounted for in the estimates for LULUCF.
b Emissions from International Bunker Fuels are not included in totals.
c LULUCF emissions of CH4 and N20 are reported separately from gross emissions totals. LULUCF emissions include the
CH4 and N20 emissions reported for Peatlands Remaining Peatlands, forest fires, drained organic soils, grassland fires,
and Coastal Wetlands Remaining Coastal Wetlands; CH4 emissions from Land Converted to Coastal Wetlands, Flooded
Land Remaining Flooded Land, and Land Converted to Flooded Land; and N20 emissions from forest soils and
settlement soils. Refer to Table 2-8 for a breakout of emissions and removals for LULUCF by gas and source category.
d Small amounts of PFC emissions also result from this source.
e LULUCF Carbon Stock Change is the net C stock change from the following categories: Forest Land Remaining Forest
Land, Land Converted to Forest Land, Cropland Remaining Cropland, Land Converted to Cropland, Grassland
Remaining Grassland, Land Converted to Grassland, Wetlands Remaining Wetlands, Land Converted to Wetlands,
Settlements Remaining Settlements, and Land Converted to Settlements. Refer to Table 2-8 for a breakout of
emissions and removals for LULUCF by gas and source category.
f The LULUCF Sector Net Total is the net sum of all LULUCF CH4 and N20 emissions to the atmosphere plus LULUCF net
carbon stock changes.
Notes: Total (gross) emissions are presented without LULUCF. Net emissions are presented with LULUCF. Totals may not
sum due to independent rounding. Parentheses indicate negative values or sequestration.
l Table 2-2: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks (kt)
Gas/Source
1990
2005
2017
2018
2019
2020
2021
CO,
5,121,447
6,132,355
5,212,068
5,377,950
5,259,759
4,714,391
5,048,172
Fossil Fuel Combustion
4,728,194
5,747,307
4,852,515
4,989,843
4,853,402
4,344,837
4,650,953
Transportation
1,468,944
1,858,552
1,780,103
1,812,903
1,813,869
1,572,525
1,789,400
Electric Power Sector
1,819,951
2,400,057
1,732,033
1,753,432
1,606,721
1,439,563
1,542,206
Industrial
852,413
850,812
789,024
813,528
815,894
767,899
762,364
Residential
338,578
358,898
293,410
338,218
341,400
313,175
310,113
Commercial
228,298
227,130
231,999
245,838
250,703
228,463
223,854
U.S. Territories
20,011
51,858
25,947
25,924
24,815
23,211
23,016
Non-Energy Use of Fuels
112,407
128,920
112,841
129,441
127,621
119,208
143,209
Iron and Steel Production &
Metallurgical Coke
Production
104,737
70,076
40,810
42,858
43,090
37,712
42,041
Cement Production
33,484
46,194
40,324
38,971
40,896
40,688
41,312
Natural Gas Systems
32,363
25,206
31,770
32,974
38,705
36,296
36,846
Petrochemical Production
21,611
27,383
28,890
29,314
30,702
29,780
33,170
Petroleum Systems
9,519
10,221
24,462
36,102
46,874
29,081
24,667
Incineration of Waste
12,900
13,254
13,161
13,339
12,948
12,921
12,476
Ammonia Production
14,404
10,234
12,481
12,669
12,401
13,006
12,207
Lime Production
11,700
14,552
12,882
13,106
12,112
11,299
11,870
Other Process Uses of
Carbonates
6,233
7,459
9,869
7,351
8,422
8,399
7,951
Urea Fertilization
2,417
3,504
4,862
4,939
5,030
5,122
5,214
Carbon Dioxide Consumption
1,472
1,375
4,580
4,130
4,870
4,970
4,990
Urea Consumption for Non-
Agricultural Purposes
3,784
3,653
5,161
6,111
6,154
5,814
4,989
Liming
4,690
4,351
3,069
2,240
2,203
2,915
3,047
Coal Mining
4,606
4,170
3,153
3,141
2,992
2,198
2,456
Glass Production
2,262
2,401
1,984
1,989
1,940
1,858
1,969
Soda Ash Production
1,431
1,655
1,753
1,714
1,792
1,461
1,714
Ferroalloy Production
2,152
1,392
1,975
2,063
1,598
1,377
1,567
Aluminum Production
6,831
4,142
1,205
1,455
1,880
1,748
1,541
2-6 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Titanium Dioxide Production
1,195
1,755
1,688
1,541
1,474
1,193
1,474
Zinc Production
632
1,030
900
999
1,026
977
969
Phosphoric Acid Production
1,529
1,342
1,025
937
909
901
909
Lead Production
516
553
513
527
531
464
446
Carbide Production and
Consumption
243
213
181
184
175
154
172
Abandoned Oil and Gas Wells
7
7
7
7
8
7
7
Substitution of Ozone
Depleting Substances
+
1
3
3
3
4
4
Magnesium Production and
Processing
128
3
3
2
2
3
3
Biomass and Biofuela
237,946
245,421
328,888
335,973
333,059
305,562
313,346
International Bunker Fuelsb
103,634
113,328
120,192
122,179
116,132
69,638
69,280
CH4c
31,025
28,255
27,243
27,649
27,421
26,509
25,980
Enteric Fermentation
6,539
6,722
6,998
7,028
7,046
7,007
6,962
Natural Gas Systems
7,682
7,263
6,657
6,943
6,915
6,620
6,479
Landfills
7,063
5,275
4,424
4,525
4,607
4,456
4,379
Manure Management
1,394
1,960
2,300
2,375
2,348
2,383
2,358
Petroleum Systems
1,833
1,819
2,209
2,165
2,138
1,945
1,791
Coal Mining
3,860
2,566
2,192
2,110
1,893
1,648
1,595
Wastewater T reatment
811
809
770
763
755
761
753
Rice Cultivation
640
720
596
623
602
630
600
Stationary Combustion
344
313
307
344
351
313
316
Abandoned Oil and Gas Wells
274
289
295
296
297
295
295
Abandoned Underground
Coal Mines
288
264
257
247
237
232
228
Mobile Combustion
258
158
105
102
103
92
94
Composting
15
75
98
90
91
92
92
Field Burning of Agricultural
Residues
15
17
17
17
17
17
17
Petrochemical Production
9
3
10
12
13
12
15
Anaerobic Digestion at Biogas
Facilities
1
2
6
6
6
6
6
Ferroalloy Production
1
+
1
1
+
+
+
Carbide Production and
Consumption
1
+
+
+
+
+
+
Iron and Steel Production &
Metallurgical Coke
Production
1
1
+
+
+
+
+
Incineration of Waste
+
+
+
+
+
+
+
International Bunker Fuelsb
7
5
4
4
4
3
3
N2Oc
1,497
1,529
1,520
1,579
1,506
1,425
1,452
Agricultural Soil Management
1,050
1,060
1,127
1,178
1,125
1,054
1,076
Stationary Combustion
84
115
95
95
84
78
83
Wastewater T reatment
56
68
78
80
80
79
79
Manure Management
47
55
64
65
65
66
66
Mobile Combustion
145
140
70
66
72
61
65
Nitric Acid Production
41
38
31
32
34
31
30
AdipicAcid Production
51
24
25
35
18
28
25
N20 from Product Uses
14
14
14
14
14
14
14
Composting
1
6
7
7
7
7
7
Caprolactam, Glyoxal, and
Glyoxylic Acid Production
6
7
5
5
5
4
5
Incineration of Waste
2
1
1
1
1
1
1
Electronics Industry
+
+
1
1
1
1
1
Trends 2-7
-------
Field Burning of Agricultural
Residues
1
1
1
1
1
1
1
Petroleum Systems
+
+
+
+
+
+
+
Natural Gas Systems
+
+
+
+
+
+
+
International Bunker Fuelsb
3
3
4
4
3
2
2
HFCs
M
M
M
M
M
M
M
Substitution of Ozone
Depleting Substances
M
M
M
M
M
M
M
HCFC-22 Production
3
1
+
+
+
+
+
Electronics Industry
M
M
M
M
M
M
M
Magnesium Production and
Processing
NO
NO
+
+
+
+
+
PFCs
M
M
M
M
M
M
M
Electronics Industry
M
M
M
M
M
M
M
Aluminum Production
M
M
M
M
M
M
M
Substitution of Ozone
Depleting Substancesd
+
+
+
+
+
+
+
Electrical Transmission and
Distribution
+
+
+
+
+
+
+
sf6
1
1
+
+
+
+
+
Electrical Transmission and
Distribution
1
1
+
+
+
+
+
Magnesium Production and
Processing
+
+
+
+
+
+
+
Electronics Industry
+
+
+
+
+
+
+
nf3
+
+
+
+
+
+
+
Electronics Industry
+
+
+
+
+
+
+
CO
130,085
66,912
34,752
32,827
32,279
31,496
30,713
NOx
21,700
17,176
8,285
7,726
7,176
6,719
6,424
so,
20,935
13,193
2,302
2,210
1,798
1,615
1,706
NMVOCs
20,923
13,310
9,483
9,173
8,751
8,650
8,549
+ Does not exceed 0.5 kt.
M (Mixture of multiple gases)
NO (Not Occurring)
a Emissions from Biomass and Biofuel Consumption are not included specifically in Energy sector totals. Net carbon fluxes from
changes in biogenic carbon reservoirs are accounted for in the estimates for LULUCF.
b Emissions from International Bunker Fuels are not included in totals.
c LULUCF emissions of LULUCF CH4 and N20 are reported separately from gross emissions totals. Refer to Table 2-8 for a
breakout of emissions and removals for LULUCF by gas and source category.
d Small amounts of PFC emissions also result from this source.
Notes: Totals by gas may not sum due to independent rounding. Parentheses indicate negative values or sequestration.
1 Emissions by IPCC Sector
2 Emissions and removals of all gases can be summed from each source and sink category into a set of five sectors
3 defined by the UNFCCC Reporting Guidelines and methodological framework provided by the Intergovernmental
4 Panel on Climate Change (IPCC). Figure 2-4 and Table 2-3 illustrate that over the thirty-two-year period of 1990 to
5 2021, total emissions from the Energy and Waste sectors decreased by 155.6 MMT CO2 Eq. (2.9 percent) and 66.8
6 MMT CO2 Eq. (28.3 percent), respectively. Emissions from Industrial Processes and Product Use and Agriculture
7 grew by 41.1 MMT CO2 Eq. (12.2 percent) and 50.8 MMT CO2 Eq. (9.4 percent), respectively. Over the same period,
8 total C sequestration in the Land Use, Land-Use Change, and Forestry (LULUCF) sector decreased by 106.8 MMT
9 CO2 (11.4 percent decrease in total C sequestration), and emissions from the LULUCF sector increased by 19.9
10 MMT CO2 Eq. (34.4 percent).
2-8 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
Figure 2-4: U.S. Greenhouse Gas Emissions and Sinks by IPCC Sector
¦ LULUCF (emissions) ¦ Agriculture
9,000 ¦ Waste ¦ Energy
¦ Industrial Processes and Product Use ¦ LULUCF (removals)
8,000 ~ Net Emissions (including LULUCF sinks)
Table 2-3: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks by IPCC
Sector/Category (MMT CO2 Eq.)
IPCC Sector/Category
1990
2005
2017
2018
2019
2020
2021
Energy
5,368.2
6,351.8
5,418.8
5,589.7
5,458.3
4,893.8
5,212.5
Fossil Fuel Combustion
4,728.2
5,747.3
4,852.5
4,989.8
4,853.4
4,344.8
4,651.0
Natural Gas Systems
247.5
228.6
218.2
227.4
232.3
221.7
218.3
Non-Energy Use of Fuels
112.4
128.9
112.8
129.4
127.6
119.2
143.2
Petroleum Systems
60.8
61.2
86.4
96.8
106.8
83.6
74.8
Coal Mining
112.7
76.0
64.5
62.2
56.0
48.3
47.1
Stationary Combustion3
31.9
39.3
33.9
34.7
32.0
29.4
30.9
Mobile Combustion3
45.6
41.4
21.5
20.4
21.9
18.7
19.8
Incineration of Waste
13.3
13.6
13.5
13.7
13.3
13.3
12.8
Abandoned Oil and Gas Wells
7.7
8.1
8.3
8.3
8.3
8.3
8.3
Abandoned Underground Coal Mines
8.1
7.4
7.2
6.9
6.6
6.5
6.4
Biomass and Biofuel Consumptionb
237.9
245.4
328.9
336.0
333.1
305.6
313.3
International Bunker Fuelsc
104.6
114.3
121.2
123.2
117.1
70.3
69.9
Industrial Processes and Product Use
335.7
356.1
359.1
362.2
366.8
363.2
376.8
Substitution of Ozone Depleting
Substances
0.3
99.4
156.1
157.8
162.0
166.1
172.5
Iron and Steel Production &
Metallurgical Coke Production
104.8
70.1
40.8
42.9
43.1
37.7
42.0
Cement Production
33.5
46.2
40.3
39.0
40.9
40.7
41.3
Petrochemical Production
21.9
27.5
29.2
29.7
31.1
30.1
33.6
Ammonia Production
14.4
10.2
12.5
12.7
12.4
13.0
12.2
Lime Production
11.7
14.6
12.9
13.1
12.1
11.3
11.9
Other Process Uses of Carbonates
6.2
7.5
9.9
7.4
8.4
8.4
8.0
Nitric Acid Production
10.8
10.1
8.3
8.5
8.9
8.3
7.9
Adipic Acid Production
13.5
6.3
6.6
9.3
4.7
7.4
6.6
Electrical Transmission and
Distribution
24.7
11.8
5.5
5.2
6.1
5.9
6.0
Trends 2-9
-------
Carbon Dioxide Consumption
1.5
1.4
4.6
4.1
4.9
5.0
5.0
Urea Consumption for Non-
Agricultural Purposes
3.8
3.7
5.2
6.1
6.2
5.8
5.0
Electronics Industry
3.3
4.5
4.6
4.7
4.3
4.4
4.8
N20 from Product Uses
3.8
3.8
3.8
3.8
3.8
3.8
3.8
Aluminum Production
26.1
7.2
2.2
2.9
3.3
3.2
2.5
HCFC-22 Production
38.6
16.8
4.3
2.7
3.1
1.8
2.2
Glass Production
2.3
2.4
2.0
2.0
1.9
1.9
2.0
Soda Ash Production
1.4
1.7
1.8
1.7
1.8
1.5
1.7
Ferroalloy Production
2.2
1.4
2.0
2.1
1.6
1.4
1.6
Titanium Dioxide Production
1.2
1.8
1.7
1.5
1.5
1.2
1.5
Caprolactam, Glyoxal, and Glyoxylic
Acid Production
1.5
1.9
1.3
1.3
1.2
1.2
1.2
Magnesium Production and
Processing
5.5
2.9
1.1
1.1
1.0
0.9
1.2
Zinc Production
0.6
1.0
0.9
1.0
1.0
1.0
1.0
Phosphoric Acid Production
1.5
1.3
1.0
0.9
0.9
0.9
0.9
Lead Production
0.5
0.6
0.5
0.5
0.5
0.5
0.4
Carbide Production and Consumption
0.3
0.2
0.2
0.2
0.2
0.2
0.2
Agriculture
538.4
567.0
601.2
617.8
603.3
586.0
589.3
Agricultural Soil Management
278.4
280.8
298.7
312.1
298.2
279.3
285.2
Enteric Fermentation
183.1
188.2
195.9
196.8
197.3
196.2
194.9
Manure Management
51.4
69.4
81.3
83.7
83.1
84.2
83.4
Rice Cultivation
17.9
20.2
16.7
17.4
16.9
17.6
16.8
Urea Fertilization
2.4
3.5
4.9
4.9
5.0
5.1
5.2
Liming
4.7
4.4
3.1
2.2
2.2
2.9
3.0
Field Burning of Agricultural Residues
0.6
0.7
0.7
0.6
0.6
0.6
0.6
Waste
236.0
192.1
170.9
173.7
176.0
171.5
169.2
Landfills
197.8
147.7
123.9
126.7
129.0
124.8
122.6
Wastewater Treatment
37.5
40.7
42.2
42.5
42.5
42.2
42.0
Composting
0.7
3.6
4.7
4.3
4.3
4.4
4.4
Anaerobic Digestion at Biogas
Facilities
+
+
0.2
0.2
0.2
0.2
0.2
Total Gross Emissions'1 (Sources)
6,478.3
7,466.9
6,550.0
6,743.4
6,604.4
6,014.5
6,347.7
LULUCF Sector Net Totale
(881.0)
(781.1)
(774.2)
(765.1)
(704.0)
(776.2)
(754.2)
Forest Land
(914.2)
(793.9)
(793.5)
(791.2)
(736.3)
(782.2)
(768.7)
Cropland
31.6
25.6
34.3
39.7
41.7
33.4
37.6
Grassland
2.2
(28.4)
(12.9)
(12.3)
(8.7)
(19.3)
(14.0)
Wetlands
44.8
44.4
42.7
42.6
42.6
42.4
42.4
Settlements
(45.3)
(28.9)
(44.7)
(44.0)
(43.4)
(50.5)
(51.4)
Net Emission (Sources and Sinks)f
5,597.3
6,685.8
5,775.8
5,978.3
5,900.3
5,238.3
5,593.5
+ Does not exceed 0.05 MMT C02 Eq.
a Includes CH4 and N20 emissions from fuel combustion.
b Emissions from Biomass and Biodiesel Consumption are not included specifically in summing Energy sector totals. Net
carbon fluxes from changes in biogenic carbon reservoirs are accounted for in the estimates for LULUCF.
c Emissions from International Bunker Fuels are not included in totals.
d Total emissions without LULUCF.
e LULUCF emissions of CH4 and N20 are reported separately from gross emissions totals. LULUCF emissions include the CH4
and N20 emissions reported for Peatlands Remaining Peatlands, forest fires, drained organic soils, grassland fires, and
Coastal Wetlands Remaining Coastal Wetlands; CH4 emissions from Land Converted to Coastal Wetlands, Flooded Land
Remaining Flooded Land, and Land Converted to Flooded Land; and N20 emissions from forest soils and settlement soils.
Refer to Table 2-8 for a breakout of emissions and removals for LULUCF by gas and source category.
f Net emissions with LULUCF.
Notes: Total (gross) emissions are presented without LULUCF. Net emissions are presented with LULUCF. Totals may not
sum due to independent rounding. Parentheses indicate negative values or sequestration.
2-10 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Energy
Emissions from energy-related activities come from two main categories, including direct emissions associated
with fuel use (i.e., fossil fuel combustion, non-energy use of fossil fuels and waste combustion) and fugitive
emissions mainly from coal, natural gas, and oil production. Energy emissions also include some categories that are
not added to energy sector totals but are instead presented as memo items, including international bunker fuels
and biomass emissions. Energy-related activities, primarily fossil fuel combustion, accounted for the vast majority
of U.S. CO2 emissions for the period of 1990 through 2021. Fossil fuel combustion is the largest source of energy-
related emissions, with CO2 being the primary gas emitted (see Figure 2-5). Due to their relative importance, fossil
fuel combustion-related CO2 emissions are considered in detail in the Energy chapter (see Chapter 3).
In 2021, 79.3 percent of the energy used in the United States on a Btu basis was produced through the combustion
of fossil fuels. The remaining 20.7 percent came from other energy sources such as hydropower, biomass, nuclear,
wind, and solar energy. A discussion of specific trends related to CO2 and other greenhouse gas emissions from
energy use is presented here with more detail in the Energy chapter. Energy-related activities are also responsible
for Cm and N2O emissions (41.6 percent and 10.3 percent of total U.S. emissions of each gas, respectively). Table
2-4 presents greenhouse gas emissions from the Energy chapter by source and gas.
Figure 2-5: Trends in Energy Sector Greenhouse Gas Sources
0
U
8,000
7,000
6,000
5,000
4,000
3,000
2,000
1,000
0
] Incineration of Waste
I U.S Territories Fossil Fuel Combustion
Non-Energy Use of Fuels
Commerical Fossil Fuel Combustion
Residential Fossil Fuel Combustion
I Fugitive Emissions
I Industrial Fossil Fuel Combustion
I Transportation Fossil Fuel Combustion
I Electric Power Fossil Fuel Combustion
0 *—« cm m u~>
C?\ Q\ C?\ 0\ On CTv
CJ%
CO
C-
CT*
C-
O H (N
OOO
no
C3
in
o o o
r-.
0
CO
c
CTi O
O *-t
N n TT
in vo r-
CO O* O
*-* *-* <-nj rvj
^
*-¦
M M M
rvj
rvl CM CM
CM
<"M
f"M CM CM
rN
S ^ ^
(N (N (N
ible 2-4: Emissions from Energy (MMT CO2 Eq.)2
Gas/Source
1990
2005
2017
2018
2019
2020
2021
C02
4,900.0
5,929.1
5,037.9
5,204.8
5,082.5
4,544.5
4,870.6
Fossil Fuel Combustion
4,728.2
5,747.3
4,852.5
4,989.8
4,853.4
4,344.8
4,651.0
Transportation
1,468.9
1,858.6
1,780.1
1,812.9
1,813.9
1,572.5
1,789.4
Electricity Generation
1,820.0
2,400.1
1,732.0
1,753.4
1,606.7
1,439.6
1,542.2
Industrial
852.4
850.8
789.0
813.5
815.9
767.9
762.4
Residential
338.6
358.9
293.4
338.2
341.4
313.2
310.1
2 The full time series data is available in Common Reporting Format (CRF) Tables included in the U.S. UNFCCC submission and in
CSV format available at https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
Trends 2-11
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Commercial
228.3
227.1
232.0
245.8
250.7
228.5
223.9
U.S. Territories
20.0
51.9
25.9
25.9
24.8
23.2
23.0
Non-Energy Use of Fuels
112.4
128.9
112.8
129.4
121.6
119.2
143.2
Natural Gas Systems
32.4
25.2
31.8
33.0
38.7
36.3
36.8
Petroleum Systems
9.5
10.2
24.5
36.1
46.9
29.1
24.7
Incineration of Waste
12.9
13.3
13.2
13.3
12.9
12.9
12.5
Coal Mining
4.6
4.2
3.2
3.1
3.0
2.2
2.5
Abandoned Oil and Gas Wells
+
+
+
+
+
+
+
Biomass-Wooda
215.2
206.9
212.0
220.0
217.7
200.4
202.8
Biofuels-Ethanola
4.2
22.9
82.1
81.9
82.6
71.8
79.1
International Bunker Fuelsb
103.6
113.3
120.2
122.2
116.1
69.6
69.3
Biofuels-Biodiesela
0.0
0.9
18.7
17.9
17.1
17.7
16.1
Biofuels-MSWa
18.5
14.7
16.1
16.1
15.7
15.6
15.3
ch4
407.0
354.8
336.7
341.8
334.2
312.0
302.3
Natural Gas Systems
215.1
203.4
186.4
194.4
193.6
185.4
181.4
Petroleum Systems
51.3
50.9
61.9
60.6
59.9
54.5
50.2
Coal Mining
108.1
71.8
61.4
59.1
53.0
46.2
44.7
Stationary Combustion
9.6
00
00
8.6
9.6
9.8
00
00
8.9
Abandoned Oil and Gas Wells
in
8.1
8.3
8.3
8.3
8.2
8.2
Abandoned Underground Coal
8.1
7.4
7.2
6.9
6.6
6.5
6.4
Mines
Mobile Combustion
7.2
4.4
2.9
2.9
2.9
2.6
2.6
Incineration of Waste
+
+
+
+
+
+
+
International Bunker Fuelsb
0.2
0.1
0.1
0.1
0.1
0.1
0.1
n2o
61.1
67.9
44.2
43.1
41.5
37.2
39.6
Stationary Combustion
22.3
30.5
25.3
25.1
22.2
20.7
22.1
Mobile Combustion
38.4
37.0
18.5
17.5
19.0
16.1
17.1
Incineration of Waste
0.4
0.3
0.4
0.4
0.4
0.3
0.4
Petroleum Systems
+
+
+
+
+
+
+
Natural Gas Systems
+
+
+
+
+
+
+
International Bunker Fuelsb
0.8
0.9
0.9
1.0
0.9
0.5
0.5
Total
5,368.2
6,351.8
5,418.8
5,589.7
5,458.3
4,893.8
5,212.5
+ Does not exceed 0.05 MMT C02 Eq.
a Emissions from Biomass and Biofuel Consumption are not included specifically in Energy sector totals. Net carbon fluxes
from changes in biogenic carbon reservoirs are accounted for in the estimates for LULUCF.
b Emissions from International Bunker Fuels are not included in totals.
Note: Totals may not sum due to independent rounding.
Fossil Fuel Combustion CO2 Emissions
As the largest contributor to U.S. greenhouse gas emissions, CO2 from fossil fuel combustion has accounted for
approximately 74.9 percent of CC>2-equivalent total gross emissions on average across the time series. Within the
United States, fossil fuel combustion accounted for 92.1 percent of CO2 emissions in 2021. Emissions from this
source category include CO2 associated with the combustion of fossil fuels (coal, natural gas, and petroleum) for
energy use. Fossil fuel combustion CO2 emissions decreased by 1.6 percent (77.2 MMT CO2 Eq.) from 1990 to 2021
and were responsible for most of the decrease in national emissions during this period. Similarly, CO2 emissions
from fossil fuel combustion have decreased by 1,096.4 MMT CO2 Eq. since 2005, representing a decrease of 19.1
percent. From 2020 to 2021, these emissions increased by 7.0 percent (306.1 MMT CO2 Eq.). Historically, changes
in emissions from fossil fuel combustion have been the main factor influencing U.S. emission trends.
Changes in CO2 emissions from fossil fuel combustion since 1990 are affected by many long-term and short-term
factors, including population and economic growth, energy price fluctuations and market trends, technological
changes, carbon intensity of energy fuel choices, and seasonal temperatures. Carbon dioxide emissions from coal
combustion gradually increased between 1990 and 2007, then began to decrease at a faster rate from 2008 to
2-12 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
2021. Carbon dioxide emissions from natural gas combustion remained relatively constant, with a slight increase
between 1990 and 2009, then began to consistently increase between 2010 and 2019. The replacement of coal
combustion with natural gas combustion was largely driven by new discoveries of natural gas fields and
advancements in drilling technologies, which led to more competitive natural gas prices. On an annual basis, the
overall consumption and mix of fossil fuels in the United States fluctuates primarily in response to changes in
general economic conditions, overall energy prices, the relative price of different fuels, weather, and the
availability of non-fossil alternatives. For example, coal consumption for electric power is influenced by a number
of factors, including the relative price of coal and alternative sources, the ability to switch fuels, and longer-term
trends in coal markets. Between 2020 and 2021, coal consumption for electric power increased 15.4 percent, a
reversal of the overall trend since 2008. However, this followed a 19.2 percent reduction in coal generation
between 2019 and 2020 due in part to the COVID-19 pandemic reducing overall demand for fossil fuels across all
sectors. There has been a 35.7 percent reduction in overall CO2 emissions from electric power generation from
2005 to 2021 (see Figure 2-7), reflecting the continued shift in the share of electric power generation from coal to
natural gas and renewables since 2005.
Fossil fuel combustion CO2 emissions also depend on the type of fuel consumed or energy used and its carbon
intensity. Producing a unit of heat or electricity using natural gas instead of coal, for example, reduces CO2
emissions because of the lower C content of natural gas (see Table 3-12 in Chapter 3 for more detail on electricity
generation by source and see Table A-22 in Annex 2.1 for more detail on the C content coefficient of different fossil
fuels).
Petroleum use is another major driver of CO2 emissions from fossil fuel combustion, particularly in the
transportation sector, which has represented the largest source of CO2 emissions from fossil fuel combustion since
2017. Emissions from petroleum consumption for transportation (including bunker fuels) have increased by 13.9
percent since 2020. This trend can be primarily attributed to a 11.2 percent increase in vehicle miles traveled
(VMT) from 2020 to 2021 due to the gradual recovery from the COVID-19 pandemic, which limited travel in 2020.
From 2019 to 2021, emissions from petroleum consumption for transportation (including bunker fuels) decreased
by 1.7 percent following a decrease of 1.0 percent in VMT over that time period. Fuel economy of light-duty
vehicles is another important factor. The decline in new light-duty vehicle fuel economy between 1990 and 2004
reflected the increasing market share of light-duty trucks, which grew from about 30 percent of new vehicle sales
in 1990 to 48 percent in 2004. Starting in 2005, average new vehicle fuel economy began to increase while light-
duty VMT grew only modestly for much of the period and has slowed the rate of increase of CO2 emissions.
Overall, across all sectors, there was a 7.0 percent increase in total CO2 emissions from fossil fuel combustion from
2020 to 2021.
Trends in CO2 emissions from fossil fuel combustion, separated by end-use sector, are presented in Table 2-5 and
Figure 2-6 based on the underlying U.S. energy consumer data collected by the U.S. Energy Information
Administration (EIA). Figure 2-7 further describes trends in direct and indirect CO2 emissions from fossil fuel
combustion, separated by end-use sector. Estimates of CO2 emissions from fossil fuel combustion are calculated
from these EIA "end-use sectors" based on total fuel consumption and appropriate fuel properties described
below. (Any additional analysis and refinement of the EIA data is further explained in the Energy chapter of this
report.)
• Transportation. ElA's fuel consumption data for the transportation sector consists of all vehicles whose
primary purpose is transporting people and/or goods from one physical location to another.
• Electric Power. ElA's fuel consumption data for the electric power sector are comprised of electricity-only
and combined-heat-and-power (CHP) plants within the North American Industry Classification System
(NAICS) 22 category whose primary business is to sell electricity, or electricity and heat, to the public.
(Non-utility power producers are included in this sector as long as they meet the electric power sector
definition.)
• Industry. EIA statistics for the industrial sector include fossil fuel consumption that occurs in the fields of
manufacturing, agriculture, mining, and construction. ElA's fuel consumption data for the industrial sector
consist of all facilities and equipment used for producing, processing, or assembling goods. (EIA includes
Trends 2-13
-------
1 generators that produce electricity and/or useful thermal output primarily to support on-site industrial
2 activities in this sector.)
3 • Residential. ElA's fuel consumption data for the residential sector consist of living quarters for private
4 households.
5 • Commercial. ElA's fuel consumption data for the commercial sector consist of service-providing facilities
6 and equipment from private and public organizations and businesses. (EIA includes generators that
7 produce electricity and/or useful thermal output primarily to support the activities at commercial
8 establishments in this sector.)
9 Table 2-5: CO2 Emissions from Fossil Fuel Combustion by End-Use Sector (MMT CO2 Eq.)
End-Use Sector
1990
2005
2017
2018
2019
2020
2021
Transportation
1,472.0
1,863.3
1,784.4
1,817.7
1,818.7
1,576.6
1,794.5
Combustion
1,468.9
1,858.6
1,780.1
1,812.9
1,813.9
1,572.5
1,789.4
Electricity
3.0
4.7
4.3
4.8
4.8
4.1
5.1
Industrial
1,538.8
1,587.1
1,293.4
1,314.9
1,281.4
1,177.7
1,202.8
Combustion
852.4
850.8
789.0
813.5
815.9
767.9
762.4
Electricity
686.4
736.3
504.4
501.3
465.5
409.8
440.5
Residential
931.3
1,214.9
910.5
980.5
925.1
858.5
887.3
Combustion
338.6
358.9
293.4
338.2
341.4
313.2
310.1
Electricity
592.7
856.0
617.1
642.3
583.7
545.3
577.2
Commercial
766.0
1,030.1
838.2
850.9
803.4
708.8
743.3
Combustion
228.3
227.1
232.0
245.8
250.7
228.5
223.9
Electricity
537.7
803.0
606.2
605.0
552.7
480.3
519.5
U.S. Territories3
20.0
51.9
25.9
25.9
24.8
23.2
23.0
Total
4,728.2
5,747.3
4,852.5
4,989.8
4,853.4
4,344.8
4,651.0
Electric Power
1,820.0
2,400.1
1,732.0
1,753.4
1,606.7
1,439.6
1,542.2
a Fuel consumption by U.S. Territories (i.e., American Samoa, Guam, Puerto Rico, U.S. Virgin Islands, Wake
Island, and other outlying U.S. Pacific Islands) is included in this report.
Notes: Combustion-related emissions from electric power are allocated based on aggregate national
electricity use by each end-use sector. Totals may not sum due to independent rounding.
10
2-14 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Figure 2-6: Trends in CO2 Emissions from Fossil Fuel Combustion by End-Use Sector and Fuel
2 Type
¦ Coal ¦ Geothermal E Natural Gas ¦ Petroleum
2,000
O
u
: 1,000
2,000
O
u
: 1,000
2,000
o
u
: 1,000
U.S. Territories
Commercial
Residential
2,000
o
u
1,000
2,000
o
u
1,000
2,000
o
u
1,000
1991 1996 2001 2006 2011 2016 2021
Industrial
Electric Power
Transportation
1991 1996 2001 2006 2011 2016 2021
Note: Fossil Fuel Combustion for electric power also includes emissions of less than 0.5 MMT C02 Eq. from geothermai-based
generation.
Trends 2-15
-------
1
2
3
4
5
6
7
8
9
10
Figure 2-7: Trends in End-Use Sector Emissions of CO2 from Fossil Fuel Combustion
¦ Direct Fossil Fuel Combustion ¦ Indirect Fossil Fuel Combustion
2,000
ct 1,500
S 1,000
500
0
2,000
1,500
8 1,000
500
0
2,000
1,500
S 1,000
500
0
U.S. Territories
Commercial
Residential
2,000
1,500
8 1,000
500
0
2,000
1,500
8 1,000
500
Industrial
Transportation
1991 1996 2001 2006 2011 2016 2021
1991 1996 2001 2006 2011 2016 2021
Electric power was the second largest emitter of CO2 in 2021 (surpassed by transportation in 2017); electric power
generators used 30.7 percent of U.S. energy from fossil fuels and emitted 33.2 percent of the CO2 from fossil fuel
combustion in 2021. Changes in electricity demand and the carbon intensity of fuels used for electric power
generation have a significant impact on CO2 emissions. Carbon dioxide emissions from fossil fuel combustion from
the electric power sector have decreased by 15.3 percent since 1990, and the carbon intensity of the electric
power sector, in terms of CO2 Eq. per QBtu input, has significantly decreased by 24.9 percent during that same
timeframe. This decoupling of electric power generation and the resulting CO2 emissions is shown below in Figure
2-8.
2-16 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Figure 2-8: Electric Power Generation (Billion kWh) and Emissions (MMT CO2 Eq.)
I Nuclear Generation (Billion kWh)
Renewable Generation (Billion kWh)
I Petroleum Generation (Billion kWh)
Coal Generation (Billion kWh)
Natural Gas Generation (Billion kWh)
-Total Emissions (MMT CO2 Eq.) [Right Axis]
3,500
3,000
2,500 w
rM
o
u
1-
2,000 |
1,500
1,000
500
Electric power CO2 emissions can also be allocated to the end-use sectors that use electricity, as presented in Table
2-5. With electricity CO2 emissions allocated to end-use sectors, the transportation end-use sector represents the
largest source of fossil fuel combustion emissions accounting for 1,794.5 MMT CO2 Eq. in 2021 or 38.6 percent of
total CO2 emissions from fossil fuel combustion. The industrial end-use sector accounted for 24.3 percent of CO2
emissions from fossil fuel combustion when including allocated electricity emissions. The residential and
commercial end-use sectors accounted for 19.1 and 16.0 percent, respectively, of CO2 emissions from fossil fuel
combustion when including allocated electricity emissions. Both of these end-use sectors were heavily reliant on
electricity for meeting energy needs, with electricity use for lighting, heating, air conditioning, and operating
appliances contributing 65.1 and 69.9 percent of emissions from the residential and commercial end-use sectors,
respectively.
Other Significant Energy Sector Trends
Other significant trends in emissions from energy source categories (Figure 2-6 and Figure 2-7) over the thirty-two-
year period from 1990 through 2021 included the following:
• Methane emissions from natural gas systems and petroleum systems (combined here) decreased 34.8
MMT CO2 Eq. (13.1 percent) from 1990 to 2021, from 266.4 MMT CO2 Eq. in 1990 to 231.6 MMT CO2 Eq.
in 2021. Natural gas systems CFU emissions have decreased by 33.7 MMT CO2 Eq. (15.7 percent) since
1990, largely due to a decrease in emissions from distribution, transmission and storage, processing, and
exploration. The decrease in distribution is largely due to decreased emissions from pipelines and
distribution station leaks, and the decrease in transmission and storage emissions is largely due to
reduced compressor station emissions (including emissions from compressors and leaks). At the same
time, emissions from the natural gas production segment increased. Methane emissions from natural gas
systems decreased 2.1 percent between 2020 and 2021. Petroleum systems CFU emissions decreased by
1.2 MMT CO2 Eq. (or 2.2 percent) since 1990 and 7.9 percent between 2020 and 2021. This decrease is
due primarily to decreases in emissions from offshore platforms, tanks, and pneumatic controllers.
Carbon dioxide emissions from natural gas and petroleum systems increased by 19.6 MMT CO2 Eq. (46.9
percent) from 1990 to 2021. This increase is due primarily to increases in the production segment, where
Trends 2-17
-------
1
2
flaring emissions from associated gas flaring, tanks, and miscellaneous production flaring have increased
overtime.
3
4
5
• Methane emissions from coal mining decreased by 63.4 MMT CO2 Eq. (58.7 percent) from 1990 through
2021 and by 3.2 percent between 2020 and 2021 primarily due to a decrease in the number of active
mines and annual coal production over this time period.
6
7
8
9
10
• Nitrous oxide emissions from mobile combustion decreased by 21.3 MMT CO2 Eq. (55.4 percent) from
1990 through 2021, primarily as a result of national vehicle criteria pollutant emissions standards and
emission control technologies for on-road vehicles. Emissions increased by 1.0 MMT CO2 Eq. (6.0 percent)
between 2020 and 2021 due to a gradual rebound in travel activity since the reduced travel seen in 2020
due to the COVID-19 pandemic.
11
12
13
14
• Carbon dioxide emissions from non-energy uses of fossil fuels increased by 30.8 MMT CO2 Eq. (27.4
percent) from 1990 through 2021, and 20 percent (24.0 MMT CO2 Eq.) between 2020 and 2021.Emissions
from non-energy uses of fossil fuels were 143.2 MMT CO2 Eq. in 2021, which constituted 2.8 percent of
total national CO2 emissions, approximately the same proportion as in 1990.
15
16
17
18
• Carbon dioxide emissions from incineration of waste (12.5 MMT CO2 Eq. in 2021) decreased slightly by 0.4
MMT CO2 Eq. (3.3 percent) from 1990 through 2021, as the volume of scrap tires and other fossil C-
containing materials in waste decreased. Emissions decreased 0.4 MMT CO2 Eq. (3.3 percent) between
2020 and 2021, consistent with trends across the time series.
19 Industrial Processes and Product Use
20 Greenhouse gases can be generated and emitted by industry in two different ways. First, they are generated and
21 emitted as the byproducts of many non-energy-related industrial activities. For example, industrial processes can
22 chemically or physically transform raw materials, which often release waste gases such as CO2, CFU, N2O, and
23 fluorinated gases (e.g., HFC-23). In the case of byproduct emissions, the emissions are generated by an industrial
24 process itself, and are not directly a result of energy consumed during the process.
25 Second, industrial manufacturing processes and use by end-consumers also release HFCs, PFCs, SF6, and NF3 and
26 other man-made compounds. In addition to the use of HFCs and some PFCs as substitutes for ozone depleting
27 substances (ODS), fluorinated compounds such as HFCs, PFCs, SF6, NF3, and others are also emitted through use by
28 a number of other industrial sources in the United States. These industries include the electronics industry, electric
29 power transmission and distribution, and magnesium metal production and processing. In addition, N2O is used in
30 and emitted by the electronics industry and anesthetic and aerosol applications, and CO2 is consumed and emitted
31 through various end-use applications.
32 Emission sources in the Industrial Processes and Product Use (IPPU) chapter accounted for 5.9 percent of U.S.
33 greenhouse gas emissions in 2021. Emissions from the IPPU sector increased by 12.2 percent from 1990 to 2021.
34 The use of HFCs and PFCs as substitutes for ODS is the largest source of emissions in this sector, contributing 45.8
35 percent of IPPU emissions in 2021. Total emissions from IPPU increased 3.7 percent between 2020 and 2021,
36 reversing the emissions reduction trend in 2020 from reduced industrial activity due to the COVID-19 pandemic.
37 Despite the sectoral increase in emissions, emissions from aluminum, ammonia, lead, zinc, adipic acid, and nitric
38 acid production all decreased from 2020 to 2021, along with emissions from other process uses of carbonates and
39 urea consumption. Figure 2-9 presents greenhouse gas emissions from IPPU by source category.
40
2-18 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Figure 2-9: Trends in Industrial Processes and Product Use Sector Greenhouse Gas Sources
500
450
400
350
. 300
8 250
200
150
100
50
I Electronics Industry
Other Product Manufacture and Use
I Mineral Industry
I Metal Industry
I Chemical Industry
] Substitution of Ozone Depleting Substances
Table 2-6: Emissions from Industrial Processes and Product Use (MMT CO2 Eq.)
Gas/Source
1990
2005
2017
2018
2019
2020
2021
C02
214.3
195.4
166.2
165.9
170.0
161.8
169.3
Iron and Steel Production & Metallurgical Coke
Production
104.7
70.1
40.8
42.9
43.1
37.7
42.0
Iron and Steel Production
99.1
66.2
38.8
41.6
40.1
35.4
38.8
Metallurgical Coke Production
5.6
3.9
2.0
1.3
3.0
2.3
3.2
Cement Production
33.5
46.2
40.3
39.0
40.9
40.7
41.3
Petrochemical Production
21.6
27.4
28.9
29.3
30.7
29.8
33.2
Ammonia Production
14.4
10.2
12.5
12.7
12.4
13.0
12.2
Lime Production
11.7
14.6
12.9
13.1
12.1
11.3
11.9
Other Process Uses of Carbonates
6.2
7.5
9.9
7.4
8.4
8.4
8.0
Carbon Dioxide Consumption
1.5
1.4
4.6
4.1
4.9
5.0
5.0
Urea Consumption for Non-Agricultural
Purposes
3.8
3.7
5.2
6.1
6.2
5.8
5.0
Glass Production
2.3
2.4
2.0
2.0
1.9
1.9
2.0
Soda Ash Production
1.4
1.7
1.8
1.7
1.8
1.5
1.7
Ferroalloy Production
2.2
1.4
2.0
2.1
1.6
1.4
1.6
Aluminum Production
6.8
4.1
1.2
1.5
1.9
1.7
1.5
Titanium Dioxide Production
1.2
1.8
1.7
1.5
1.5
1.2
1.5
Zinc Production
0.6
1.0
0.9
1.0
1.0
1.0
1.0
Phosphoric Acid Production
1.5
1.3
1.0
0.9
0.9
0.9
0.9
Lead Production
0.5
0.6
0.5
0.5
0.5
0.5
0.4
Carbide Production and Consumption
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Substitution of Ozone Depleting Substances
+
+
+
+
+
+
+
Magnesium Production and Processing
0.1
+
+
+
+
+
+
ch4
0.3
0.1
0.3
0.4
0.4
0.4
0.4
Petrochemical Production
0.2
0.1
0.3
0.3
0.4
0.3
0.4
Ferroalloy Production
+
+
+
+
+
+
+
Carbide Production and Consumption
+
+
+
+
+
+
+
Iron and Steel Production & Metallurgical Coke
Production
+
+
+
+
+
+
+
Trends 2-19
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Iron and Steel Production
+
+
+
+
+
+
+
Metallurgical Coke Production
NO
NO
NO
NO
NO
NO
NO
n2o
29.6
22.2
20.2
23.1
18.7
20.8
19.7
Nitric Acid Production
10.8
10.1
8.3
8.5
8.9
8.3
7.9
AdipicAcid Production
13.5
6.3
6.6
9.3
4.7
7.4
6.6
N20 from Product Uses
3.8
3.8
3.8
3.8
3.8
3.8
3.8
Caprolactam, Glyoxal, and Glyoxylic Acid
Production
1.5
1.9
1.3
1.3
1.2
1.2
1.2
Electronics Industry
+
0.1
0.2
0.2
0.2
0.3
0.3
HFCs
39.0
116.4
160.8
160.9
165.4
168.2
175.1
Substitution of Ozone Depleting Substances3
0.3
99.4
156.1
157.7
161.9
166.1
172.4
HCFC-22 Production
38.6
16.8
4.3
2.7
3.1
1.8
2.2
Electronics Industry
0.2
0.2
0.3
0.3
0.3
0.3
0.4
Magnesium Production and Processing
NO
NO
0.1
0.1
0.1
0.1
+
PFCs
21.8
6.1
3.8
4.3
4.0
3.9
3.5
Electronics Industry
2.5
3.0
2.7
2.8
2.5
2.4
2.6
Aluminum Production
19.3
3.1
1.0
1.4
1.4
1.4
0.9
Substitution of Ozone Depleting Substances
NO
+
+
+
+
+
+
Electrical Transmission and Distribution
NO
+
+
NO
+
+
+
sf6
30.5
15.5
7.2
7.1
7.8
7.5
8.0
Electrical Transmission and Distribution
24.7
11.8
5.5
5.2
6.1
5.9
6.0
Magnesium Production and Processing
5.4
2.9
1.0
1.1
0.9
0.9
1.1
Electronics Industry
0.5
0.8
0.7
0.8
0.8
0.8
0.9
nf3
+
0.4
0.5
0.5
0.5
0.6
0.6
Electronics Industry
+
0.4
0.5
0.5
0.5
0.6
0.6
Total
335.7
356.1
359.1
362.2
366.8
363.2
376.8
+ Does not exceed 0.05 MMT C02 Eq.
NO (Not Occurring)
a Small amounts of PFC emissions also result from this source.
Note: Totals may not sum due to independent rounding.
Some significant trends in U.S. emissions from IPPU source categories over the thirty-two-year period from 1990
through 2021 included the following:
• HFC and PFC emissions resulting from the substitution of ODS (e.g., chlorofluorocarbons [CFCs]) have
been increasing from small amounts in 1990 to 172.5 MMT CO2 Eq. in 2021 (68,134.2 percent).
• Combined CO2 and CFU emissions from iron and steel production and metallurgical coke production
decreased by 11.5 percent from 2020 to 2021 to 42.0 MMT CO2 Eq. and have declined overall by 62.7
MMT CO2 Eq. (59.9 percent) from 1990 through 2021, due to restructuring of the industry. The trend in
the United States has been a shift towards fewer integrated steel mills and more electric arc furnaces
(EAFs). EAFs use scrap steel as their main input and generally have less on-site emissions.
• Carbon dioxide emissions from petrochemicals increased by 53.5 percent between 1990 and 2021 from
21.6 MMT CO2 Eq. to 33.2 MMT CO2 Eq. The increase in emissions is largely driven by a doubling of
production of ethylene over that time period.
• Carbon dioxide emissions from ammonia production (12.2 MMT CO2 Eq. in 2021) decreased by 15.2
percent (2.2 MMT CO2 Eq.) since 1990. Ammonia production relies on natural gas as both a feedstock and
a fuel, and as such, market fluctuations and volatility in natural gas prices affect the production of
ammonia from year to year. Emissions from ammonia production have increased since 2016, due to the
addition of new ammonia production facilities and new production units at existing facilities. Agricultural
demands continue to drive demand for nitrogen fertilizers and the need for new ammonia production
capacity.
• Carbon dioxide emissions from cement production increased by 23.4 percent (7.8 MMT CO2 Eq.) from
1990 through 2021. They rose from 1990 through 2006 and then fell until 2009, due to a decrease in
2-20 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
demand for construction materials during the economic recession. Since 2010, CO2 emissions from
cement production have risen 31.4 percent (9.9 MMT CO2 Eq.).
HFC-23 emissions from HCFC-22 production decreased by 36.4 MMT CO2 Eq. (94.2 percent) from 1990 to
2021 due to a reduction in the HFC-23 emission rate (kg HFC-23 emitted/kg HCFC-22 produced).
PFC emissions from aluminum production decreased by 18.4 MMT CO2 Eq. (95.3 percent) from 1990 to
2021, due to both industry emission reduction efforts and lower domestic aluminum production.
• SFs emissions from electrical transmission and distribution decreased by 18.7 MMT CO2 Eq. (75.7 percent)
from 1990 to 2021 due to industry emission reduction efforts.
Agriculture
Agricultural activities contribute directly to emissions of greenhouse gases through a variety of processes,
including the following source categories: enteric fermentation in domestic livestock, livestock manure
management, rice cultivation, agricultural soil management, liming, urea fertilization, and field burning of
agricultural residues. Methane, N2O, and CO2 are the primary greenhouse gases emitted by agricultural activities.
Carbon stock changes from agricultural soils are included in the LULUCF sector.
In 2021, agricultural activities were responsible for emissions of 589.3 MMT CO2 Eq., or 9.3 percent of total U.S.
greenhouse gas emissions. Agricultural soil management activities, such as the application of synthetic and organic
fertilizers, deposition of livestock manure, and growing N-fixing plants, were the largest contributors to
agricultural-related emissions (48.4 percent) and were the largest source of U.S. N2O emissions in 2021, accounting
for 74.1 percent. Methane emissions from enteric fermentation and manure management represented 26.8
percent and 9.1 percent of total CH4 emissions from anthropogenic activities, respectively, in 2021. Carbon dioxide
emissions from the application of crushed limestone and dolomite (i.e., soil liming) and urea fertilization
represented 0.2 percent of total CO2 emissions from anthropogenic activities. Figure 2-10 and Table 2-7 illustrate
agricultural greenhouse gas emissions by source and gas.
Figure 2-10: Trends in Agriculture Sector Greenhouse Gas Sources
700
650
600
550
500
450
d"
UJ
400
rM
O
<_>
350
1—
s:
5;
300
250
200
150
100
50
0
I Field Burning of Agricultural Residues
Urea Fertilization
I Liming
Rice Cultivation
I Manure Management
I Enteric Fermentation
Agricultural Soil Management
Ch CTi
_ LO
LO U">
,
Trends 2-21
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Table 2-7: Emissions from Agriculture (MMT CO2 Eq.)
Gas/Source
1990
2005
2017
2018
2019
2020
2021
CO?
7.1
7.9
7.9
7.2
7.2
8.0
8.3
Urea Fertilization
2.4
3.5
4.9
4.9
5.0
5.1
5.2
Liming
4.7
4.4
3.1
2.2
2.2
2.9
3.0
ch4
240.4
263.7
277.5
281.2
280.4
281.0
278.2
Enteric Fermentation
183.1
188.2
195.9
196.8
197.3
196.2
194.9
Manure Management
39.0
54.9
64.4
66.5
65.7
66.7
66.0
Rice Cultivation
17.9
20.2
16.7
17.4
16.9
17.6
16.8
Field Burning of Agricultural
Residues
0.4
0.5
0.5
0.5
0.5
0.5
0.5
n2o
290.9
295.4
315.7
329.4
315.7
297.0
302.8
Agricultural Soil Management
278.4
280.8
298.7
312.1
298.2
279.3
285.2
Manure Management
12.4
14.5
16.9
17.2
17.4
17.5
17.4
Field Burning of Agricultural
Residues
0.1
0.2
0.2
0.2
0.2
0.2
0.2
Total
538.4
567.0
601.2
617.8
603.3
586.0
589.3
Note: Totals may not sum due to independent rounding.
Some significant trends in U.S. emissions from Agriculture source categories (Figure 2-10) over the thirty-two-year
period from 1990 through 2021 included the following:
• Agricultural soils are the largest anthropogenic source of agriculture-related emissions and also N2O
emissions in the United States, accounting for 74.1 percent of N2O emissions and 4.5 percent of total
emissions in the United States in 2021. Estimated emissions from this source in 2021 were 285.2 MMT
CO2 Eq. Annual N2O emissions from agricultural soils fluctuated between 1990 and 2021, and overall
emissions were 6.8 MMT CO2 Eq. or 2.5 percent higher in 2021 than in 1990. Year-to-year fluctuations are
largely a reflection of annual variation in weather patterns, synthetic fertilizer use, and crop production.
• Enteric fermentation is the largest anthropogenic source of CH4 emissions in the United States. In 2021,
enteric fermentation CFU emissions were 26.8 percent of total CFU emissions (194.9 MMT CO2 Eq.), which
represents an increase of 11.9 MMT CO2 Eq. (6.5 percent) since 1990. This increase in emissions from
enteric fermentation from 1990 to 2021 generally follows the increasing trends in cattle populations. For
example, from 1990 to 1995, emissions increased and then generally decreased from 1996 to 2004,
mainly due to fluctuations in beef cattle populations and increased digestibility of feed for feedlot cattle.
Emissions increased from 2005 to 2007, as both dairy and beef populations increased. Research indicates
that the feed digestibility of dairy cow diets decreased during this period. Emissions decreased again from
2008 to 2014 as beef cattle populations again decreased. Emissions increased from 2014 to 2021,
consistent with an increase in beef cattle population over those same years.
• Manure management emissions increased 62.3 percent between 1990 and 2021. This encompassed an
increase of 69.2 percent for CFU, from 39.0 MMT CO2 Eq. in 1990 to 66.0 MMT CO2 Eq. in 2021; and an
increase of 40.5 percent for N2O, from 12.4 MMT CO2 Eq. in 1990 to 17.4 MMT CO2 Eq. in 2021. The
majority of the increase observed in CFU resulted from swine and dairy cattle manure, where emissions
increased 38.3 and 124.3 percent, respectively, from 1990 to 2021. From 2020 to 2021, there was a 1.1
percent decrease in total CH4 emissions from manure management, mainly due to minor shifts in the
animal populations and the resultant effects on manure management system allocations.
• Liming and urea fertilization are the only sources of CO2 emissions reported in the Agriculture sector. All
other CO2 emissions and removals (e.g., carbon stock changes from the management of croplands, etc.)
are characterized in the LULUCF sector. Estimated emissions from these sources were 3.0 and 5.2 MMT
CO2 Eq., respectively. Liming emissions increased by 4.5 percent relative to 2020 and decreased 1.6 MMT
CO2 Eq. or 35.0 percent relative to 1990, while urea fertilization emissions increased by 1.8 percent
relative to 2020 and 2.8 MMT CO2 Eq. or 115.7 percent relative to 1990.
2-22 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Land Use, Land-Use Change, and Forestry
When humans alter the terrestrial biosphere through land use, changes in land use, and land management
practices, they also influence the carbon (C) stock fluxes on these lands and cause emissions of Cm and N2O.
Overall, managed land is a net sink for CO2 (C sequestration) in the United States. The primary driver of fluxes on
managed lands is from management of forest lands, but also includes trees in settlements (i.e., urban areas),
afforestation, conversion of forest lands to settlements and croplands, the management of croplands and
grasslands, flooded lands, and the landfilling of yard trimmings and food scraps. The main drivers for net forest
sequestration include net forest growth, increasing forest area, and a net accumulation of C stocks in harvested
wood pools. The net sequestration in Settlements Remaining Settlements, is driven primarily by C stock gains in
urban forests (i.e., Settlement Trees) through net tree growth and increased urban area, as well as long-term
accumulation of C in landfills from additions of yard trimmings and food scraps.
The LULUCF sector in 2021 resulted in a net increase in C stocks (i.e., net CO2 removals) of 832.0 MMT CO2 Eq.
(Table 2-8).3 This represents an offset of 13.1 percent of total (i.e., gross) greenhouse gas emissions in 2021.
Emissions of CFU and N2O from LULUCF activities in 2021 were 77.8 MMT CO2 Eq. and represented 1.4 percent of
net greenhouse gas emissions.4 Between 1990 and 2021, total net C sequestration in the LULUCF sector decreased
by 11.4 percent, primarily due to a decrease in the rate of net C accumulation in forests and Cropland Remaining
Cropland, as well as an increase in CO2 emissions from Land Converted to Settlements.
Flooded Land Remaining Flooded Land was the largest source of CH4 emissions from LULUCF in 2021, totaling 45.4
MMT CO2 Eq. (1,623 kt of CH4). Forest fires were the second largest source of CFU emissions from LULUCF in 2021,
totaling 15.5 MMT CO2 Eq. (554 kt of CH4). Coastal Wetlands Remaining Coastal Wetlands resulted in CFU emissions
of 4.3 MMT CO2 Eq. (154 kt of CH4). Grassland fires resulted in CH4 emissions of 0.3 MMT CO2 Eq. (12 kt of CH4).
Land Converted to Wetlands, drained organic soils, and Peatlands Remaining Peatlands resulted in CFU emissions
of less than 0.05 MMT CO2 Eq. each.
Forest fires were the largest source of N2O emissions from LULUCF in 2021, totaling 8.9 MMT CO2 Eq. (34 kt of
N2O). Nitrous oxide emissions from fertilizer application to settlement soils in 2021 totaled to 2.1 MMT CO2 Eq. (8
kt of N2O). Additionally, the application of synthetic fertilizers to forest soils in 2021 resulted in N2O emissions of
0.4 MMT CO2 Eq. (2 kt of N2O). Grassland fires resulted in N2O emissions of 0.3 MMT CO2 Eq. (1 kt of N2O). Coastal
Wetlands Remaining Coastal Wetlands and drained organic soils resulted in N2O emissions of 0.5 MMT CO2 Eq.
each (0.5 kt of N2O). Peatlands Remaining Peatlands resulted in N2O emissions of less than 0.05 MMT CO2 Eq.
Figure 2-11 and Table 2-8 along with CFU and N2O emissions (purple) for LULUCF source categories.
3 LULUCF Carbon Stock Change is the net C stock change from the following categories: Forest Land Remaining Forest Land,
Land Converted to Forest Land, Cropland Remaining Cropland, Land Converted to Cropland, Grassland Remaining Grassland,
Land Converted to Grassland, Wetlands Remaining Wetlands, Land Converted to Wetlands, Settlements Remaining
Settlements, and Land Converted to Settlements.
4 LULUCF emissions include the CH4 and N20 emissions reported for Peatlands Remaining Peatlands, forest fires, drained
organic soils, grassland fires, and Coastal Wetlands Remaining Coastal Wetlands; CH4 emissions from Flooded Land Remaining
Flooded Land, Land Converted to Flooded Land, and Land Converted to Coastal Wetlands; and N20 emissions from forest soils
and settlement soils.
Trends 2-23
-------
Figure 2-11: Trends in Emissions and Removals (Net CO2 Flux) from Land Use, Land-Use
Change, and Forestry
400
300
Wetlands Remaining Wetlands
Land Converted to Wetlands
I Land Converted to Settlements
Land Converted to Grassland
Land Converted to Forest Land
I Land Converted to Cropland
Grassland Remaining Grassland
I Cropland Remaining Cropland
I Settlements Remaining Settlements
I Forest Land Remaining Forest Land
¦ Net Emissions (Sources and Sinks)
o
u
200 en ™
100
0
-100
-200
-300
-400
-500
-600
-700
-800
-900
-1,000
Gl Ol C> G"! &i CT\
ui ffi (Ji (Ji
oi en o-* o o o o
Oi O O O O
irNrNfNrvJfNfNrslfNfNJrsJfNJfNfNJfNfNJfNjfNfNfNrNjfNfN
Table 2-8: U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use, Land-
Use Change, and Forestry (MMT CO2 Eq.)
Land-Use Category
Forest Land Remaining Forest Land
Changes in Forest Carbon Stocks3
Non-C02 Emissions from Forest Firesb
N20 Emissions from Forest Soilsc
Non-C02 Emissions from Drained Organic
Soilsd
Land Converted to Forest Land
Changes in Forest Carbon Stockse
Cropland Remaining Cropland
Changes in Mineral and Organic Soil
Carbon Stocks
Land Converted to Cropland
Changes in all Ecosystem Carbon Stocks'
Grassland Remaining Grassland
Changes in Mineral and Organic Soil
Carbon Stocks
Non-C02 Emissions from Grassland Fires^
Land Converted to Grassland
Changes in all Ecosystem Carbon Stocks'
2018
2019
2020
2021
10.9
0.6
(24.5)
(24.5)
(692.9)
(704.4)
11.0
0.4
0.1
(98.3)
(98.3)
(16.6)
(16.6)
56.3
56.3
11.9
11.3
0.6
(24.2)
(24.2)
(638.1)
(649.3)
10.8
0.4
0.1
(98.3)
(98.3)
(14.5)
(14.5)
56.3
56.3
14.6
14.0
0.6
(23.3)
(23.3)
(684.0)
(707.4)
23.0
0.4
0.1
(98.3)
(98.3)
(23.3)
(23.3)
56.7
56.7
6.7
6.0
0.6
(25.9)
(25.9)
(670.5)
(695.4)
24.4
0.4
0.1
(98.3)
(98.3)
(18.9)
(18.9)
56.5
56.5
10.6
10.0
0.6
(24.7)
(24.7)
2-24 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Wetlands Remaining Wetlands
41.5
43.1
41.8
41.8
41.8
41.8
41.8
Changes in Organic Soil Carbon Stocks in
Peatlands
1.1
1.1
0.8
0.8
0.8
0.7
0.7
Non-C02 Emissions from Peatlands
Remaining Peatlands
+
+
+
+
+
+
+
Changes in Biomass, DOM, and Soil
Carbon Stocks in Coastal Wetlands
(8.4)
(7.7)
(8.8)
(8.8)
(8.8)
(8.8)
(8.8)
CH4 Emissions from Coastal Wetlands
Remaining Coastal Wetlands
4.2
4.2
4.3
4.3
4.3
4.3
4.3
N20 Emissions from Coastal Wetlands
Remaining Coastal Wetlands
0.1
0.2
0.1
0.1
0.1
0.1
0.1
CH4 Emissions from Flooded Land
Remaining Flooded Land
44.6
45.3
45.4
45.4
45.4
45.4
45.4
Land Converted to Wetlands
3.3
1.4
0.8
0.8
0.8
0.6
0.6
Changes in Biomass, DOM, and Soil
Carbon Stocks in Land Converted to
Coastal Wetlands
0.5
0.5
(+)
(+)
(+)
(+)
(+)
CH4 Emissions from Land Converted to
Coastal Wetlands
0.3
0.3
0.2
0.2
0.2
0.2
0.2
Changes in Land Converted to Flooded
Land
1.4
0.4
0.4
0.4
0.4
0.3
0.3
CH4 Emissions from Land Converted to
Flooded Land
1.1
0.3
0.3
0.3
0.3
0.2
0.2
Settlements Remaining Settlements
(107.8)
(113.9)
(125.6)
(125.0)
(124.5)
(131.6)
(132.5)
Changes in Organic Soil Carbon Stocks
11.3
12.2
16.0
15.9
15.9
15.9
15.9
Changes in Settlement Tree Carbon
Stocks
(96.4)
(117.4)
(129.6)
(129.5)
(129.3)
(136.7)
(137.8)
N20 Emissions from Settlement Soilsh
1.8
2.8
1.9
2.0
2.0
2.0
2.1
Changes in Yard Trimming and Food
Scrap Carbon Stocks in Landfills
(24.5)
(11.4)
(13.8)
(13.4)
(13.1)
(12.8)
(12.6)
Land Converted to Settlements
62.5
85.0
80.9
81.0
81.1
81.0
81.0
Changes in all Ecosystem Carbon Stocks'
62.5
85.0
80.9
81.0
81.1
81.0
81.0
LULUCF Emissions'
57.9
72.4
68.3
64.4
64.2
76.4
77.8
ch4
53.5
61.3
60.1
57.3
56.9
65.4
66.0
n2o
4.4
11.1
8.3
7.0
7.3
11.0
11.8
LULUCF Carbon Stock Change'
(938.9)
(853.5)
(842.5)
(829.5)
(768.2)
(852.5)
(832.0)
LULUCF Sector Net Totalk
(881.0)
(781.1)
(774.2)
(765.1)
(704.0)
(776.2)
(754.2)
+ Absolute value does not exceed 0.05 MMT C02 Eq.
a Includes the net changes to carbon stocks stored in all forest ecosystem pools (estimates include C stock changes from
drained organic soils from both Forest Land Remaining Forest Land and Land Converted to Forest Land.) and harvested
wood products.
b Estimates include emissions from fires on both Forest Land Remaining Forest Land and Land Converted to Forest Land.
c Estimates include emissions from N fertilizer additions on both Forest Land Remaining Forest Land and Land Converted to
Forest Land.
d Estimates include CH4 and N20 emissions from drained organic soils on both Forest Land Remaining Forest Land and Land
Converted to Forest Land. Carbon stock changes from drained organic soils are included with the Forest Land Remaining
Forest Land forest ecosystem pools.
e Includes the net changes to carbon stocks stored in all forest ecosystem pools.
f Includes changes in mineral and organic soil carbon stocks for all land use conversions to cropland, grassland, and
settlements. Also includes aboveground/belowground biomass, dead wood, and litter carbon stock changes for
conversion of forest land to cropland, grassland, and settlements.
s Estimates include CH4 and N20 emissions from fires on both Grassland Remaining Grassland and Land Converted to
Grassland.
h Estimates include N20 emissions from N fertilizer additions on both Settlements Remaining Settlements and Land
Converted to Settlements because it is not possible to separate the activity data at this time.
' LULUCF Carbon Stock Change includes any C stock gains and losses from all land use and land-use conversion categories.
i LULUCF emissions include the CH4 and N20 emissions reported for Peatlands Remaining Peatlands, forest fires, drained
organic soils, grassland fires, and Coastal Wetlands Remaining Coastal Wetlands; CH4 emissions from Flooded Land
Trends 2-25
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Remaining Flooded Land, and Land Converted to Flooded Land, and Land Converted to Coastal Wetlands; and N20
emissions from forest soils and settlement soils.
k The LULUCF Sector Net Total is the net sum of all LULUCF CH4 and N20 emissions to the atmosphere plus LULUCF net
carbon stock changes in units of MMT C02 Eq.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Other significant trends from 1990 to 2021 in emissions from LULUCF categories (Figure 2-11) over the thirty-two-
year period included the following:
• Annual carbon (C) sequestration by forest land (i.e., annual C stock accumulation in the five ecosystem C
pools and harvested wood products for Forest Land Remaining Forest Land and Land Converted to Forest
Land) has decreased by 13.7 percent since 1990. This is primarily due to decreased C stock gains in Land
Converted to Forest Land and the harvested wood products pools within Forest Land Remaining Forest
Land.
• Annual C sequestration from Settlements Remaining Settlements (which includes organic soils, settlement
trees, and landfilled yard trimmings and food scraps) has increased by 22.8 percent over the period from
1990 to 2021. This is primarily due to an increase in urbanized land area in the United States with trees
growing on it.
• Annual emissions from Land Converted to Settlements increased by 29.7 percent from 1990 to 2021 due
primarily to C stock losses from Forest Land Converted to Settlements and mineral soils C stocks from
Grassland Converted to Settlements.
Waste management and treatment activities are sources of CFU and N2O emissions (see Figure 2-12 and Table 2-9)
In 2021, landfills were the largest source of waste emissions, accounting for 72.5 percent of waste-related
emissions. Landfills are also the third-largest source of U.S. anthropogenic CH4 emissions, generating 122.6 MMT
CO2 Eq. and accounting for 16.9 percent of total U.S. CH4 emissions in 2021.5 Additionally, wastewater treatment
generated emissions of 42.0 MMT CO2 Eq. and accounted for 24.8 percent of waste emissions, 2.9 percent of U.S.
Cm emissions, and 5.4 percent of U.S. N2O emissions in 2021. Emissions of CH4 and N2O from composting are also
accounted for in this chapter, generating emissions of 2.6 MMT CO2 Eq. and 1.8 MMT CO2 Eq., respectively.
Anaerobic digestion at biogas facilities generated CH4 emissions of 0.2 MMT CO2 Eq., accounting for 0.1 percent of
emissions from the Waste sector. Overall, emission sources accounted for in the Waste chapter generated 169.2
MMT C02Eq„ or 2.7 percent of total U.S. greenhouse gas emissions in 2021.
5 Landfills also store carbon, due to incomplete degradation of organic materials such as wood products and yard trimmings, as
described in the Land Use, Land-Use Change, and Forestry chapter.
2-26 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Figure 2-12: Trends in Waste Sector Greenhouse Gas Sources
¦ Anaerobic Digestion at Biogas Facilities
Oi-irMroTj-mvDr^ooa^Oi-HfvjfOTrmvDr^cocr>Oi-H(%jf»*)T3-Lnorvoocrio-^H
CTicriCTicriCT'icricriCTiCT'icrioooooooooo-'-ti—ii—ii—ii—ii-Hi—ii—ir-Njr\j
CTiCTiCT^CT^CT^CTiCT^CT^C^CT^OOOOOOOOOOOOOOOOOOOOOO
Table 2-9: Emissions from Waste (MMT CO2 Eq.)
Gas/Source
1990
2005
2017
2018
2019
2020
2021
ch4
220.9
172.5
148.3
150.8
152.9
148.8
146.4
Landfills
197.8
147.7
123.9
126.7
129.0
124.8
122.6
Wastewater Treatment
22.7
22.7
21.5
21.4
21.2
21.3
21.1
Composting
0.4
2.1
2.7
2.5
2.5
2.6
2.6
Anaerobic Digestion at
Biogas Facilities
+
+
0.2
0.2
0.2
0.2
0.2
n2o
15.1
19.5
22.6
22.9
23.1
22.7
22.7
Wastewater Treatment
14.8
18.1
20.6
21.2
21.3
20.9
20.9
Composting
0.3
1.5
1.9
1.8
1.8
1.8
1.8
Total
236.0
192.1
170.9
173.7
176.0
171.5
169.2
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Some significant trends in U.S. emissions from waste source categories (Figure 2-12) over the thirty-two-year
period from 1990 through 2021 included the following:
• Net Cm emissions from landfills decreased by 75.1 MMT CO2 Eq. (38.0 percent), with small increases
occurring in interim years. This downward trend in emissions coincided with increased landfill gas
collection and control systems, and a reduction of decomposable materials (i.e., paper and paperboard,
food scraps, and yard trimmings) discarded in municipal solid waste (MSW) landfills over the time series.
• Methane and N2O emissions from wastewater treatment decreased by 1.6 MMT CO2 Eq. (7.2 percent) and
increased by 6.1 MMT CO2 Eq. (41.6 percent), respectively. Methane emissions from domestic wastewater
treatment have decreased since 1999 due to decreasing percentages of wastewater being treated in
anaerobic systems, including reduced use of on-site septic systems and central anaerobic treatment
systems. Nitrous oxide emissions from wastewater treatment processes gradually increased across the
time series as a result of increasing U.S. population and protein consumption.
• Combined CH4 and N2O emissions from composting have increased by 3.7 MMT CO2 Eq. since 1990, from
0.7 MMT CO2 Eq. to 4.4 MMT CO2 Eq. in 2021, which represents more than a six-fold increase over the
time series. The growth in composting since the 1990s is attributable to primarily four factors: (1) the
enactment of legislation by state and local governments that discouraged the disposal of yard trimmings
Trends 2-27
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
and food waste in landfills; (2) an increase in yard trimming collection and yard trimming drop off sites
provided by local solid waste management districts; (3) an increased awareness of the environmental
benefits of composting; and (4) loans or grant programs to establish or expand composting infrastructure.
2.2 Emissions by Economic Sector
Throughout this report, emission estimates are grouped into five sectors (i.e., chapters) defined by the IPCC and
detailed above: Energy, IPPU, Agriculture, LULUCF, and Waste. It is also useful to characterize emissions according
to commonly used economic sector categories: residential, commercial, industry, transportation, electric power,
and agriculture. Emissions from U.S. Territories are reported as their own end-use sector due to a lack of specific
consumption data for the individual end-use sectors within U.S. Territories. See Box 2-1 for more information on
how economic sectors are defined. For more information on trends in the Land Use, Land Use Change, and
Forestry sector, see Section 2.1.
Using this categorization, transportation activities accounted for the largest portion (29.0 percent) of total U.S.
greenhouse gas emissions in 2021. Emissions from electric power accounted for the second largest portion (25.0
percent), while emissions from industry accounted for the third largest portion (23.2 percent) of total U.S.
greenhouse gas emissions in 2021. Emissions from industry have in general declined over the past decade due to a
number of factors, including structural changes in the U.S. economy (i.e., shifts from a manufacturing-based to a
service-based economy), fuel switching, and efficiency improvements.
The remaining 22.8 percent of U.S. greenhouse gas emissions were contributed by, in order of magnitude, the
agriculture, commercial, and residential sectors, plus emissions from U.S. Territories. Activities related to
agriculture accounted for roughly 9.9 percent of emissions; unlike other economic sectors, agricultural sector
emissions were dominated by N2O emissions from agricultural soil management and CH4 emissions from enteric
fermentation, rather than CO2 from fossil fuel combustion. An increasing amount of carbon is stored in agricultural
soils each year, but this C sequestration is assigned to the LULUCF sector rather than the agriculture economic
sector. The commercial and residential sectors accounted for roughly 6.8 percent and 5.7 percent of greenhouse
gas emissions, respectively, and U.S. Territories accounted for 0.4 percent of emissions; emissions from these
sectors primarily consisted of CO2 emissions from fossil fuel combustion. Carbon dioxide was also emitted and
sequestered (in the form of C) by a variety of activities related to forest management practices, tree planting in
urban areas, the management of agricultural soils, landfilling of yard trimmings, and changes in C stocks in coastal
wetlands. Table 2-10 presents a detailed breakdown of emissions from each of these economic sectors by source
category, as they are defined in this report. Figure 2-13 shows the trend in emissions by sector from 1990 to 2021.
Figure 2-13: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors
2-28 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Note: Emissions and removals from Land Use, Land Use Change, and Forestry are excluded from figure above. Excludes U.S.
2 Territories.
3 Table 2-10: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO2 Eq. and
4 Percent of Total in 2021)
Sector/Source
1990
2005
2017
2018
2019
2020
2021
Percent3
Transportation
1,521.4
1,966.0
1,841.6
1,871.3
1,871.7
1,624.9
1,841.7
29.0%
C02 from Fossil Fuel Combustion
1,468.9
1,858.6
1,780.1
1,812.9
1,813.9
1,572.5
1,789.4
28.2%
Substitution of Ozone Depleting
Substances
+
63.1
37.0
35.5
34.0
32.5
31.2
0.5%
Mobile Combustion15
40.6
34.2
14.9
13.6
15.0
12.1
13.0
0.2%
Non-Energy Use of Fuels
11.8
10.2
9.6
9.2
8.8
7.8
8.0
0.1%
Electric Power Industry
1,879.7
2,456.9
1,779.2
1,799.1
1,650.5
1,481.8
1,585.4
25.0%
C02 from Fossil Fuel Combustion
1,820.0
2,400.1
1,732.0
1,753.4
1,606.7
1,439.6
1,542.2
24.3%
Stationary Combustion15
18.7
27.7
23.2
23.1
20.2
18.9
20.4
0.3%
Incineration of Waste
13.3
13.6
13.5
13.7
13.3
13.3
12.8
0.2%
Electrical Transmission and
Distribution
24.7
11.8
5.5
5.2
6.1
5.9
6.0
0.1%
Other Process Uses of Carbonates
3.1
3.7
4.9
3.7
4.2
4.2
4.0
0.1%
Industry
1,677.8
1,574.7
1,494.7
1,558.3
1,568.4
1,464.9
1,474.9
23.2%
C02 from Fossil Fuel Combustion
809.0
800.0
749.2
773.7
776.2
728.8
722.7
11.4%
Natural Gas Systems
247.5
228.6
218.2
227.4
232.3
221.7
218.3
3.4%
Non-Energy Use of Fuels
97.2
111.2
103.1
120.0
118.5
111.2
135.0
2.1%
Petroleum Systems
60.8
61.2
86.4
96.8
106.8
83.6
74.8
1.2%
Coal Mining
112.7
76.0
64.5
62.2
56.0
48.3
47.1
0.7%
Iron and Steel Production
104.8
70.1
40.8
42.9
43.1
37.7
42.0
0.7%
Cement Production
33.5
46.2
40.3
39.0
40.9
40.7
41.3
0.7%
Petrochemical Production
21.9
27.5
29.2
29.7
31.1
30.1
33.6
0.5%
Substitution of Ozone Depleting
Substances
+
7.9
30.2
32.1
33.3
34.1
32.4
0.5%
Landfills (Industrial)
12.2
16.1
18.4
18.5
18.6
18.8
18.9
0.3%
Ammonia Production
14.4
10.2
12.5
12.7
12.4
13.0
12.2
0.2%
Lime Production
11.7
14.6
12.9
13.1
12.1
11.3
11.9
0.2%
Abandoned Oil and Gas Wells
7.7
8.1
8.3
8.3
8.3
8.3
8.3
0.1%
Nitric Acid Production
10.8
10.1
8.3
8.5
8.9
8.3
7.9
0.1%
Wastewater Treatment
6.6
7.1
7.4
7.5
7.6
7.6
7.6
0.1%
Adipic Acid Production
6.5
6.8
6.9
6.9
6.9
7.0
6.9
0.1%
Abandoned Underground Coal
Mines
3.8
3.7
5.3
5.2
6.0
6.0
6.0
0.1%
Mobile Combustion15
7.2
6.6
6.7
6.4
6.2
5.9
5.8
0.1%
Carbon Dioxide Consumption
3.9
6.1
5.6
5.8
6.0
6.1
5.7
0.1%
Urea Consumption for Non-
Agricultural Purposes
1.5
1.4
4.6
4.6
4.1
4.9
5.0
0.1%
Electronics Industry
3.1
3.7
5.4
4.9
3.7
4.9
4.9
0.1%
Other Process Uses of Carbonates
3.6
4.8
5.0
4.9
5.1
4.7
4.7
0.1%
N20 from Product Uses
4.2
4.2
4.2
4.2
4.2
4.2
4.2
0.1%
Stationary Combustion15
4.9
4.7
4.2
4.1
4.0
4.0
3.7
+%
Aluminum Production
28.3
7.6
2.7
2.3
3.1
3.6
3.4
0.1%
HCFC-22 Production
46.1
20.0
2.8
5.2
3.3
3.7
2.1
0.1%
Glass Production
2.3
2.4
2.1
2.0
2.0
1.9
1.9
+%
Soda Ash Production
1.4
1.7
1.7
1.8
1.7
1.8
1.5
+%
Ferroalloy Production
2.2
1.4
1.8
2.0
2.1
1.6
1.4
+%
Titanium Dioxide Production
1.2
1.8
1.7
1.7
1.5
1.5
1.3
+%
Trends 2-29
-------
Caprolactam, Glyoxal, and
Glyoxylic Acid Production
1.7
2.1
1.7
1.5
1.4
1.4
1.2
+%
Magnesium Production and
Processing
0.6
1.0
0.8
0.9
1.0
1.0
1.0
+%
Zinc Production
1.5
1.3
1.0
1.0
0.9
0.9
0.9
+%
Phosphoric Acid Production
5.3
2.7
1.2
1.1
1.1
0.9
0.9
+%
Lead Production
0.5
0.6
0.5
0.5
0.5
0.5
0.5
+%
Carbide Production and
Consumption
0.3
0.2
0.2
0.2
0.2
0.2
0.2
+%
Agriculture
583.2
619.5
642.3
658.9
644.2
626.3
630.2
9.9%
N20 from Agricultural Soil
Management
278.4
280.8
298.7
312.1
298.2
279.3
285.2
4.5%
Enteric Fermentation
183.1
188.2
195.9
196.8
197.3
196.2
194.9
3.1%
Manure Management
51.4
69.4
81.3
83.7
83.1
84.2
83.4
1.3%
C02 from Fossil Fuel Combustion
43.4
50.8
39.8
39.8
39.7
39.1
39.7
0.6%
Rice Cultivation
17.9
20.2
16.7
17.4
16.9
17.6
16.8
0.3%
Urea Fertilization
2.4
3.5
4.9
4.9
5.0
5.1
5.2
0.1%
Liming
4.7
4.4
3.1
2.2
2.2
2.9
3.0
+%
Mobile Combustion15
1.4
1.6
1.2
1.2
1.2
1.2
1.2
+%
Field Burning of Agricultural
Residues
0.6
0.7
0.7
0.6
0.6
0.6
0.6
+%
Stationary Combustion15
+
+
+
+
+
+
+
+%
Commercial
447.0
418.9
437.6
453.7
462.0
436.0
429.9
6.8%
C02 from Fossil Fuel Combustion
228.3
227.1
232.0
245.8
250.7
228.5
223.9
3.5%
Landfills (Municipal)
185.5
131.6
105.5
108.2
110.4
106.0
103.7
1.6%
Substitution of Ozone Depleting
Substances
+
21.4
58.9
58.5
59.8
60.8
61.9
1.0%
Wastewater Treatment
30.9
33.6
34.7
35.0
34.8
34.6
34.3
0.5%
Composting
0.7
3.6
4.7
4.3
4.3
4.4
4.4
0.1%
Stationary Combustion15
1.5
1.5
1.6
1.7
1.7
1.6
1.6
+%
Anaerobic Digestion at Biogas
Facilities
+
+
0.2
0.2
0.2
0.2
0.2
+%
Residential
345.6
371.2
328.4
375.8
382.4
356.9
362.3
5.7%
C02 from Fossil Fuel Combustion
338.6
358.9
293.4
338.2
341.4
313.2
310.1
4.9%
Substitution of Ozone Depleting
Substances
0.2
7.0
30.0
31.7
34.8
38.7
46.9
0.7%
Stationary Combustion15
6.8
5.3
4.9
5.9
6.2
5.1
5.3
0.1%
U.S. Territories
23.4
59.7
26.3
26.3
25.1
23.5
23.3
0.4%
C02 from Fossil Fuel Combustion
20.0
51.9
25.9
25.9
24.8
23.2
23.0
0.4%
Non-Energy Use of Fuels
3.4
7.6
0.2
0.2
0.2
0.2
0.2
+%
Stationary Combustion15
0.1
0.2
0.1
0.1
0.1
0.1
0.1
+%
Total Gross Emissions (Sources)
6,478.3
7,466.9
6,550.0
6,743.4
6,604.4
6,014.5
6,347.7
100.0%
LULUCF Sector Net Total0
(881.0)
(781.1)
(774.2)
(765.1)
(704.0)
(776.2)
(754.2)
(11.9%)
Net Emissions (Sources and Sinks)
5,597.3
6,685.8
5,775.8
5,978.3
5,900.3
5,238.3
5,593.5
88.1%
+ Does not exceed 0.05 MMT C02 Eq. or 0.05 percent.
a Percent of total (gross) emissions excluding emissions from LULUCF for 2021.
b Includes CH4 and N20 emissions from fuel combustion.
c The LULUCF Sector Net Total is the net sum of all LULUCF CH4 and N20 emissions to the atmosphere plus LULUCF net carbon
stock changes.
Notes: Total gross emissions presented are without LULUCF. Total net emissions are presented with LULUCF. Totals may not
sum due to independent rounding. Parentheses indicate negative values or sequestration.
2-30 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
Box 2-1: Methodology for Aggregating Emissions by Economic Sector
In presenting the Economic Sectors in the annual Inventory of U.S. Greenhouse Gas Emissions and Sinks, the
Inventory expands upon the standard IPCC sectors common for UNFCCC reporting. Discussing greenhouse gas
emissions relevant to U.S.-specific economic sectors improves communication of the report's findings.
The Electric Power economic sector includes CO2, CFU and N2O emissions from the combustion of fossil fuels
that are included in the EIA electric power sector. Carbon dioxide, CH4, and N2O emissions from waste
incineration are included in the Electric Power economic sector, as the majority of municipal solid waste is
combusted in plants that produce electricity. The Electric Power economic sector also includes SF6 from
Electrical Transmission and Distribution, and a portion of CO2 from Other Process Uses of Carbonates (from
pollution control equipment installed in electric power plants).
The Transportation economic sector includes CO2 emissions from the combustion of fossil fuels that are
included in the EIA transportation fuel-consuming sector. (Additional analyses and refinement of the EIA data
are further explained in the Energy chapter of this report.) Emissions of CH4 and N2O from mobile combustion
are also apportioned to the Transportation economic sector based on the EIA transportation fuel-consuming
sector. Substitution of Ozone Depleting Substances emissions are apportioned to the Transportation economic
sector based on emissions from refrigerated transport and motor vehicle air-conditioning systems. Finally, CO2
emissions from Non-Energy Uses of Fossil Fuels identified as lubricants for transportation vehicles are included
in the Transportation economic sector.
The Industry economic sector includes CO2 emissions from the combustion of fossil fuels that are included in the
EIA industrial fuel-consuming sector, minus the agricultural use of fuel explained below. The CFU and N2O
emissions from stationary and mobile combustion are also apportioned to the Industry economic sector based
on the EIA industrial fuel-consuming sector, minus emissions apportioned to the Agriculture economic sector.
Substitution of Ozone Depleting Substances emissions are apportioned based on their specific end-uses within
the source category, with most emissions falling within the Industry economic sector. Finally, CH4 emissions
from industrial landfills and CFU and N2O from industrial wastewater treatment are included in the Industry
economic sector.
Additionally, all process-related emissions from sources with methods considered within the IPCC IPPU sector
are apportioned to the Industry economic sector. This includes the process-related emissions (i.e., emissions
from the actual process to make the material, not from fuels to power the plant) from activities such as Cement
Production, Iron and Steel Production and Metallurgical Coke Production, and Ammonia Production.
Additionally, fugitive emissions from energy production sources, such as Natural Gas Systems, Coal Mining, and
Petroleum Systems are included in the Industry economic sector. A portion of CO2 from Other Process Uses of
Carbonates (from pollution control equipment installed in large industrial facilities) is also included in the
Industry economic sector. Finally, all remaining CO2 emissions from Non-Energy Uses of Fossil Fuels are assumed
to be industrial in nature (besides the lubricants for transportation vehicles specified above) and are attributed
to the Industry economic sector.
The Agriculture economic sector includes CO2 emissions from the combustion of fossil fuels that are based on
supplementary sources of agriculture fuel use data, because EIA includes agriculture equipment in the industrial
fuel-consuming sector. Agriculture fuel use estimates are obtained from U.S. Department of Agriculture survey
data, in combination with EIA Fuel Oil and Kerosene Sales (FOKS) data (EIA 1991 through 2021). Agricultural
operations are based on annual energy expense data from the Agricultural Resource Management Survey
(ARMS) conducted by the National Agricultural Statistics Service (NASS) of the USDA. NASS collects information
on farm production expenditures including expenditures on diesel fuel, gasoline, LP gas, natural gas, and
electricity use on the farm with the annual ARMS. A USDA publication (USDA/NASS 2020) shows national totals,
as well as select States and ARMS production regions. These supplementary data are subtracted from the
industrial fuel use reported by EIA to obtain agriculture fuel use. Carbon dioxide emissions from fossil fuel
combustion, and CH4 and N2O emissions from stationary and mobile combustion, are then apportioned to the
Agriculture economic sector based on agricultural fuel use.
Trends 2-31
-------
The other IPCC Agriculture emission source categories apportioned to the Agriculture economic sector include
N2O emissions from Agricultural Soils, Cm from Enteric Fermentation, CH4 and N2O from Manure Management,
Cm from Rice Cultivation, CO2 emissions from Liming and Urea Application, and CH4 and N2O from Field Burning
of Agricultural Residues.
The Residential economic sector includes CO2 emissions from the combustion of fossil fuels that are included in
the EIA residential fuel-consuming sector. Stationary combustion emissions of CH4 and N2O are also based on
the EIA residential fuel-consuming sector. Substitution of Ozone Depleting Substances are apportioned to the
Residential economic sector based on emissions from residential air-conditioning systems. Nitrous oxide
emissions from the application of fertilizers to developed land (termed "settlements" by the IPCC) are also
included in the Residential economic sector.
The Commercial economic sector includes CO2 emissions from the combustion of fossil fuels that are included in
the EIA commercial fuel-consuming sector. Emissions of CH4 and N2O from mobile combustion are also
apportioned to the Commercial economic sector based on the EIA commercial fuel-consuming sector.
Substitution of Ozone Depleting Substances emissions are apportioned to the Commercial economic sector
based on emissions from commercial refrigeration/air-conditioning systems. Public works sources, including
direct CH4 from municipal landfills, CH4 from anaerobic digestion at biogas facilities, CH4 and N2O from domestic
wastewater treatment, and composting, are also included in the Commercial economic sector.
1
2 Emissions with Electricity Distributed to Economic Sectors
3 It is also useful to view greenhouse gas emissions from economic sectors with emissions related to electric power
4 distributed into end-use categories (i.e., emissions from electric power are allocated to the economic sectors in
5 which the electricity is used).
6 The generation, transmission, and distribution of electricity accounted for 25.0 percent of total U.S. greenhouse
7 gas emissions in 2021. Electric power-related emissions decreased by 15.7 percent since 1990 mainly due to fuel
8 switching in the electric power sector. From 2020 to 2021, electric power-related emissions increased by 7.0
9 percent due to in part to electricity use rebounding after the COVID-19 pandemic. Between 2020 to 2021, the
10 consumption of natural gas for electric power generation decreased by 3.1 percent, while the consumption of coal
11 and petroleum increased by 15.4 and 6.8 percent, respectively. However, even with the increase in 2021, electric
12 power-related emissions are still lower than pre-pandemic 2019 levels.
13 From 2020 to 2021, electricity sales to the residential end-use sector increased by 0.8 percent. Electricity sales to
14 the commercial end-use and industrial sectors both increased by 2.9. Overall, from 2020 to 2021, the amount of
15 electricity retail sales (in kWh) increased by 2.1 percent. Table 2-11 provides a detailed summary of emissions from
16 electric power-related activities.
2-32 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Table 2-11: Electric Power-Related Greenhouse Gas Emissions (MMT CO2 Eq.)
Gas/Fuel Type or Source
1990
2005
2017
2018
2019
2020
2021
CO?
1,836.0
2,417.0
1,750.1
1,770.4
1,623.9
1,456.7
1,558.7
Fossil Fuel Combustion
1,820.0
2,400.1
1,732.0
1,753.4
1,606.7
1,439.6
1,542.2
Coal
1,546.5
1,982.8
1,207.1
1,152.9
973.5
788.2
909.7
Natural Gas
175.4
318.9
505.6
577.9
616.6
634.8
615.1
Petroleum
97.5
98.0
18.9
22.2
16.2
16.2
17.1
Geothermal
0.5
0.5
0.4
0.4
0.4
0.4
0.4
Incineration of Waste
12.9
13.3
13.2
13.3
12.9
12.9
12.5
Other Process Uses of Carbonates
3.1
3.7
4.9
3.7
4.2
4.2
4.0
ch4
0.5
1.0
1.2
1.4
1.4
1.4
1.4
Stationary Sources3
0.5
1.0
1.2
1.4
1.4
1.4
1.4
Incineration of Waste
+
+
+
+
+
+
+
n2o
18.6
27.1
22.4
22.1
19.1
17.9
19.4
Stationary Sources3
18.2
26.7
22.0
21.7
18.8
17.5
19.0
Incineration of Waste
0.4
0.3
0.4
0.4
0.4
0.3
0.4
sf6
24.7
11.8
5.5
5.2
6.1
5.9
6.0
Electrical Transmission and Distribution
24.7
11.8
5.5
5.2
6.1
5.9
6.0
PFCs
+
+
+
+
+
+
+
Electrical Transmission and Distribution
+
+
+
+
+
+
+
Total
1,879.7
2,456.9
1,779.2
1,799.1
1,650.5
1,481.8
1,585.4
+ Does not exceed 0.05 MMT C02 Eq.
3 Includes only stationary combustion emissions related to the generation of electricity.
Note: Totals may not sum due to independent rounding.
To distribute electricity emissions among economic end-use sectors, emissions from the source categories
assigned to the electric power sector were allocated to the residential, commercial, industry, transportation, and
agriculture economic sectors according to each economic sector's share of retail sales of electricity (EIA 2020b;
USDA/NASS 2020). These source categories include CO2 from Fossil Fuel Combustion, CFU and N2O from Stationary
Combustion, Incineration of Waste, Other Process Uses of Carbonates, and SF6 from Electrical Transmission and
Distribution Systems. Note that only 50 percent of the Other Process Uses of Carbonates emissions were
associated with electric power and distributed as described; the remainder of Other Process Uses of Carbonates
emissions were attributed to the industry economic end-use sector.6
When emissions from electricity use are distributed among these economic end-use sectors, emissions from
industrial activities account for the largest share of total U.S. greenhouse gas emissions (29.8 percent), followed
closely by emissions from transportation (29.1 percent). Emissions from the commercial and residential sectors
also increase substantially when emissions from electricity are included (15.2 and 15.1 percent, respectively). In all
economic end-use sectors except agriculture, CO2 accounts for more than 75 percent of greenhouse gas emissions,
primarily from the combustion of fossil fuels. Table 2-12 presents a detailed breakdown of emissions from each of
these economic sectors, with emissions from electric power distributed to them. Figure 2-14 shows the trend in
these emissions by sector from 1990 to 2021.
6 Emissions were not distributed to U.S. Territories, since the electric power sector only includes emissions related to the
generation of electricity in the 50 states and the District of Columbia.
Trends 2-33
-------
1
2
Figure 2-14: U.S. Greenhouse Gas Emissions with Electricity-Related Emissions Distributed
to Economic Sectors
3
4 Note: Emissions and removals from Land Use, Land-Use Change, and Forestry are excluded from figure above. Excludes U.S.
5 Territories.
6 Table 2-12: U.S. Greenhouse Gas Emissions by Economic Sector and Gas with Electricity-
7 Related Emissions Distributed (MMT CO2 Eq.) and Percent of Total in 2021
Sector/Gas
1990
2005
2017
2018
2019
2020
2021
Percent3
Industry
2,351.6
2,290.2
1,974.0
2,033.6
2,011.4
1,852.4
1,894.5
29.8%
Direct Emissions
1,677.8
1,574.7
1,494.7
1,558.3
1,568.4
1,464.9
1,474.9
23.2%
C02
1,164.0
1,142.4
1,072.9
1,128.2
1,149.1
1,065.2
1,087.0
17.1%
ch4
411.7
367.4
353.5
358.0
350.4
329.6
319.9
5.0%
n2o
35.6
29.9
27.3
30.3
26.0
27.8
26.8
0.4%
HFCs, PFCs, SF6 and NF3
66.5
35.0
40.9
41.8
42.9
42.3
41.3
0.7%
Electricity-Related
673.8
715.5
479.3
475.2
442.9
387.5
419.7
6.6%
C02
658.1
703.8
471.5
467.7
435.8
380.9
412.6
6.5%
ch4
0.2
0.3
0.3
0.4
0.4
0.4
0.4
+%
n2o
6.7
7.9
6.0
5.8
5.1
4.7
5.1
0.1%
sf6
00
00
3.4
1.5
1.4
1.6
1.5
1.6
+%
Transportation
1,524.6
1,970.9
1,846.0
1,876.2
1,876.7
1,629.2
1,846.9
29.1%
Direct Emissions
1,521.4
1,966.0
1,841.6
1,871.3
1,871.7
1,624.9
1,841.7
29.0%
C02
1,480.8
1,868.7
1,789.7
1,822.1
1,822.7
1,580.3
1,797.4
28.3%
ch4
6.4
3.2
1.8
1.7
1.7
1.5
1.6
+%
n2o
34.3
31.0
13.1
11.9
13.3
10.7
11.5
0.2%
HFCsb
+
63.1
37.0
35.5
34.0
32.5
31.2
0.5%
Electricity-Related
3.1
4.8
4.4
4.9
5.0
4.2
5.2
0.1%
C02
3.1
4.8
4.4
4.8
4.9
4.2
5.1
0.1%
ch4
+
+
+
+
+
+
+
+%
n2o
+
0.1
0.1
0.1
0.1
0.1
0.1
+%
sf6
+
+
+
+
+
+
+
+%
Residential
957.8
1,247.5
962.3
1,034.9
982.0
918.3
955.7
15.1%
Direct Emissions
345.6
371.2
328.4
375.8
382.4
356.9
362.3
5.7%
C02
338.6
358.9
293.4
338.2
341.4
313.2
310.1
4.9%
ch4
5.9
4.5
4.2
5.1
5.3
4.4
4.6
0.1%
n2o
0.9
0.8
0.7
0.8
0.8
0.7
0.7
+%
2-34 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
sf6
0.2
7.0
30.0
31.7
34.8
38.7
46.9
0.7%
Electricity-Related
612.2
876.3
633.9
659.0
599.6
561.3
593.4
9.3%
C02
598.0
862.1
623.6
648.5
590.0
551.8
583.4
9.2%
ch4
0.2
0.3
0.4
0.5
0.5
0.5
0.5
+%
n2o
6.1
9.7
8.0
8.1
6.9
6.8
7.3
0.1%
sf6
8.0
4.2
2.0
1.9
2.2
2.2
2.2
+%
Commercial
1,002.4
1,241.0
1,060.4
1,074.5
1,029.7
930.5
963.9
15.2%
Direct Emissions
447.0
418.9
437.6
453.7
462.0
436.0
429.9
6.8%
C02
228.3
227.1
232.0
245.8
250.7
228.5
223.9
3.5%
ch4
203.7
150.9
124.3
126.6
128.5
124.2
121.6
1.9%
n2o
15.1
19.4
22.4
22.8
23.0
22.5
22.5
0.4%
HFCs
+
21.4
58.9
58.5
59.8
60.8
61.9
1.0%
Electricity-Related
555.4
822.0
622.8
620.8
567.7
494.5
534.0
8.4%
C02
542.5
808.7
612.6
610.9
558.6
486.1
525.0
8.3%
ch4
0.1
0.3
0.4
0.5
0.5
0.5
0.5
+%
n2o
5.5
9.1
7.8
7.6
6.6
6.0
6.5
0.1%
sf6
7.3
4.0
1.9
1.8
2.1
2.0
2.0
+%
Agriculture
618.4
657.8
681.0
698.1
679.4
660.7
663.4
10.5%
Direct Emissions
583.2
619.5
642.3
658.9
644.2
626.3
630.2
9.9%
C02
50.5
58.7
47.8
47.0
46.9
47.1
48.0
0.8%
ch4
240.6
263.9
277.7
281.4
280.5
281.2
278.4
4.4%
n2o
292.1
296.9
316.9
330.5
316.8
298.0
303.9
4.8%
Electricity-Related
35.2
38.3
38.7
39.2
35.2
34.4
33.1
0.5%
C02
34.3
37.7
38.1
38.5
34.6
33.8
32.6
0.5%
ch4
+
+
+
+
+
+
+
+%
n2o
0.3
0.4
0.5
0.5
0.4
0.4
0.4
+%
sf6
0.5
0.2
0.1
0.1
0.1
0.1
0.1
+%
U.S. Territories
23.4
59.7
26.3
26.3
25.1
23.5
23.3
0.4%
Total Gross Emissions
(Sources)
6,478.3
7,466.9
6,550.0
6,743.4
6,604.4
6,014.5
6,347.7
100.0%
LULUCF Sector NetTotalc
(881.0)
(781.1)
(774.2)
(765.1)
(704.0)
(776.2)
(754.2)
(11.9%)
Net Emissions (Sources and
Sinks)
5,597.3
6,685.8
5,775.8
5,978.3
5,900.3
5,238.3
5,593.5
88.1%
+ Does not exceed 0.05 MMT C02 Eq. or 0.05 percent.
a Percent of total (gross) emissions excluding emissions from LULUCF for year 2021.
b Includes primarily HFC-134a.
c The LULUCF Sector Net Total is the net sum of all LULUCF CH4 and N20 emissions to the atmosphere plus LULUCF net carbon
stock changes.
Notes: Total gross emissions are presented without LULUCF. Net emissions are presented with LULUCF. Emissions from electric
power are allocated based on aggregate electricity use in each end-use sector. Totals may not sum due to independent
rounding.
Industry
The industry end-use sector includes CO2 emissions from fossil fuel combustion from all manufacturing facilities, in
aggregate, and with the distribution of electricity-related emissions, accounts for 29.8 percent of U.S. greenhouse
gas emissions in 2021. This end-use sector also includes emissions that are produced as a byproduct of the non-
energy-related industrial process activities. The variety of activities producing these non-energy-related emissions
includes Cm emissions from petroleum and natural gas systems, fugitive CH4 and CO2 emissions from coal mining,
byproduct CO2 emissions from cement manufacture, and HFC, PFC, SF6, and NF3 byproduct emissions from the
electronics industry, to name a few.
Since 1990, industrial sector emissions have declined by 22.6 percent. The decline has occurred both in direct
emissions and indirect emissions associated with electricity use. Structural changes within the U.S. economy that
led to shifts in industrial output away from energy-intensive manufacturing products to less energy-intensive
products (e.g., from steel to computer equipment) have had a significant effect on industrial emissions.
Trends 2-35
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Transportation
When electricity-related emissions are distributed to economic end-use sectors, transportation activities
accounted for 29.1 percent of U.S. greenhouse gas emissions in 2021. The largest sources of transportation
greenhouse gas emissions in 2021 were light-duty trucks, which include sport utility vehicles, pickup trucks, and
minivans (36.8 percent); medium- and heavy-duty trucks (23.3 percent); passenger cars (20.5 percent); commercial
aircraft (5.0 percent); other aircraft (4.0 percent); pipelines (3.5 percent); ships and boats (2.7 percent); and rail
(1.9 percent). These figures include direct CO2, Cm, and N2O emissions from fossil fuel combustion used in
transportation, indirect emissions from electricity use, and emissions from non-energy use (i.e., lubricants) used in
transportation, as well as HFC emissions from mobile air conditioners and refrigerated transport allocated to these
vehicle types.
From 1990 to 2021, total transportation emissions from fossil fuel combustion increased by approximately 21.8
percent. From 2020 to 2021, emissions increased by 13.8 percent, which followed a decline of 13.3 percent from
2019 to 2020 due to reduced travel demand during the COVID-19 pandemic. The increase in transportation
emissions from 1990 to 2021 was due, in large part, to increased demand for travel. The number of VMT by light-
duty motor vehicles (passenger cars and light-duty trucks) increased 48.4 percent from 1990 to 2021 as a result of
a confluence of factors including population growth, economic growth, urban sprawl, and periods of low fuel
prices.
The decline in new light-duty vehicle fuel economy between 1990 and 2004 reflected the increasing market share
of light-duty trucks, which grew from approximately 29.6 percent of new vehicle sales in 1990 to 48.0 percent in
2004. Starting in 2005, average new vehicle fuel economy began to increase while light-duty VMT grew only
modestly for much of the period. Light-duty VMT grew by less than one percent or declined each year between
2005 and 2013, then grew at a faster rate until 2016 (2.6 percent from 2014 to 2015, and 2.5 percent from 2015 to
2016). Since 2016, the rate of light-duty VMT growth has slowed to at or less than one percent each year. Average
new vehicle fuel economy has increased almost every year since 2005, while light-duty truck market share
decreased to 33.0 percent in 2009 and has since varied from year to year between 35.6 and 62.9 percent. Light-
duty truck market share was about 62.9 percent of new vehicles in model year 2021 (EPA 2022b).
Table 2-13 provides a detailed summary of greenhouse gas emissions from transportation-related activities with
electricity-related emissions included in the totals. Historically, the majority of electricity use in the transportation
sector was for rail transport. However, more recently there has been increased electricity use in on-road electric
and plug-in hybrid vehicles. For a more detailed breakout of emissions by fuel type by vehicle see Table A-99 in
Annex 3.
Almost all of the energy used for transportation was supplied by petroleum-based products, with more than half
being related to gasoline consumption in automobiles and other highway vehicles. Other fuel uses, especially
diesel fuel for freight trucks and jet fuel for aircraft, accounted for the remainder. The primary driver of
transportation-related emissions was CO2 from fossil fuel combustion, which increased by 21.9 percent from 1990
to 2021. This rise in CO2 emissions, combined with an increase in HFCs from close to zero emissions in 1990 to 31.2
MMT CO2 Eq. in 2021, led to an increase in overall greenhouse gas emissions from transportation activities of 21.1
percent.
2-36 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Figure 2-15: Trends in Transportation-Related Greenhouse Gas Emissions
2,800
2,600
2,400
2,200
2,000
1,800
lB" 1,600
8 1,400
I-
1 1,200
1,000
800
600
400
200
0
Lubricants ¦ Ships and Boats
Motorcycles ¦ Aircraft
Buses ¦ Medium- and Heavy-Duty Trucks
I Pipelines ¦ Light-Duty Trucks
Rail ¦ Passenger Cars
(N cn in 10
CJi Oi Gt
Gi © Ot Ol Oi
coctiOt-ic\im<3-m^r'-.coa^OT-i(Nro'<3-ir)<£ir-..ooo^OTH
G""i O^i O O OOOOOOOOt—It—I t—I t—It—It—It—It—It—It—lf\J (N
O^OiOOOOOOOOOOOOOOOOOOOOOO
Table 2-13: Transportation-Related Greenhouse Gas Emissions (MMT CO2 Eq.)
Gas/Vehicle
1990
2005
2017
2018
2019
2020
2021
Passenger Cars
648,4
564,4
392.7
398.7
395.5
341.7
378.5
C02
622.2
521.1
379.0
386.5
384.2
331.9
369.2
ch4
3.8
1.2
0.3
0.3
0.3
0.3
0.3
n2o
22.5
13.3
3.0
2.5
2.6
2.0
2.0
HFCs
0.0
28.8
10.4
9.4
8.4
7.6
7.0
Light-Duty Trucks
302,5
659,5
716.2
720.6
711.8
615.4
680.1
C02
292.2
614.2
692.7
699.1
690.2
596.3
662.2
ch4
1.5
1.0
0.6
0.6
0.6
0.5
0.6
n2o
8.7
14.0
5.4
4.6
5.6
4.4
4.3
HFCs
0.0
30.2
17.5
16.4
15.4
14.2
13.0
Medium- and Heavy-Duty Trucks
234,3
391,3
395.6
406.7
409.5
386.7
430.1
C02
232.8
386.5
387.5
398.2
400.6
377.9
420.7
ch4
0.5
0.2
0.1
0.1
0.1
0.1
0.1
n2o
1.0
1.5
2.6
2.7
3.0
2.7
3.0
HFCs
0.0
3.2
5.4
5.6
5.8
6.1
6.3
Buses
13.4
17.7
23.4
24.4
24.8
23.6
26.5
C02
13.3
17.2
22.8
23.7
24.2
23.0
25.9
ch4
+
0,1
0.1
0.1
0.1
+
+
n2o
+
0.1
0,2
0.2
0.2
0,2
0.2
HFCs
0.0
0.2
0.4
0.4
0.4
0.4
0.4
Motorcycles
3.4
5.0
7.2
7.4
7.5
6.7
7.6
C02
3.4
4.9
7.0
7.3
7.4
6.6
7.5
ch4
+
+
+
+
+
+
+
N20
+
+
0.1
0.1
0.1
0,1
0.1
Commercial Aircraft3
110.8
133.8
129.0
130.7
135.3
92.0
92.0
Trends 2-37
-------
co2
109.9
132.7
128.0
129.6
134.2
91.3
91.3
ch4
0.0
0.0
0.0
0.0
0.0
0.0
0.0
n2o
0.9
1.1
1.0
1.1
1.1
0.7
0.7
Other Aircraftb
78.0
59.5
45.5
44.6
45.6
31.0
74.7
C02
77.3
59.0
45.1
44.2
45.2
30.7
74.1
ch4
0.1
0.1
+
+
+
+
+
n2o
0.6
0.5
0.4
0.4
0.4
0.2
0.6
Ships and Boatsc
47.0
45.5
43.8
41.1
40.0
32.4
50.0
C02
46.3
44.3
39.9
36.9
35.5
27.6
44.8
ch4
0.4
0.5
0.5
0.5
0.4
0.4
0.5
n2o
0.2
0.2
0.2
0.2
0.2
0.1
0.3
HFCs
0.0
0.5
3.2
3.6
3.9
4.2
4.5
Rail
39.0
51.4
41.3
42.5
39.7
34.0
35.2
C02
38.5
50.8
40.7
41.9
39.1
33.5
34.6
ch4
0.1
0.1
0.1
0.1
0.1
0.1
0.1
n2o
0.3
0.4
0.3
0.3
0.3
0.3
0.3
HFCs
0.0
0.1
0.1
0.1
0.1
0.1
0.1
Other Emissions from Electric
Powerd
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Pipelines8
36.0
32.6
41.6
50.2
58.2
57.9
64.2
C02
36.0
32.6
41.6
50.2
58.2
57.9
64.2
Lubricants
11.8
10.2
9.6
9.2
8.8
7.8
8.0
C02
11.8
10.2
9.6
9.2
00
00
7.8
8.0
Total Transportation
1,524.6
1,970.9
1,846.0
1,876.2
1,876.7
1,629.2
1,846.9
International Bunker Fuels1
54.7
44.6
34.5
32.4
26.3
22.7
22.6
Ethanol C029
4.1
21.6
77.7
78.6
78.7
68.1
76.3
Biodiesel C02a
0.0
0.9
18.7
17.9
17.1
17.7
16.1
+ Does not exceed 0.05 MMT C02 Eq.
a Consists of emissions from jet fuel consumed by domestic operations of commercial aircraft (no bunkers).
b Consists of emissions from jet fuel and aviation gasoline consumption by general aviation and military aircraft.
c Fluctuations in emission estimates are associated with fluctuations in reported fuel consumption and may reflect
issues with data sources.
d Other emissions from electric power are a result of waste incineration (as the majority of municipal solid waste is
combusted in "trash-to-steam" electric power plants), electrical transmission and distribution, and a portion of Other
Process Uses of Carbonates (from pollution control equipment installed in electric power plants).
e C02 estimates reflect natural gas used to power pipelines, but not electricity. While the operation of pipelines
produces CH4 and N20, these emissions are not directly attributed to pipelines in the Inventory.
f Emissions from International Bunker Fuels include emissions from both civilian and military activities; these emissions
are not included in the transportation totals,
s Ethanol and biodiesel C02 estimates are presented for informational purposes only. See Section 3.11 and the
estimates in Land Use, Land-Use Change, and Forestry (see Chapter 6), in line with IPCC methodological guidance and
UNFCCC reporting obligations, for more information on ethanol and biodiesel.
Notes: Passenger cars and light-duty trucks include vehicles typically used for personal travel and less than 8,500 lbs;
medium- and heavy-duty trucks include vehicles larger than 8,500 lbs. HFC emissions primarily reflect HFC-134a. Totals
may not sum due to independent rounding.
1 Residential
2 The residential end-use sector, with electricity-related emissions distributed, accounts for 15.1 percent of U.S.
3 greenhouse gas emissions in 2021 and similarly, is heavily reliant on electricity for meeting energy needs, with
4 electricity use for lighting, heating, air conditioning, and operating appliances. The remaining emissions were
5 largely due to the direct consumption of natural gas and petroleum products, primarily for heating and cooking
6 needs. Emissions from the residential sector have generally been increasing since 1990, and annual variations are
7 often correlated with short-term fluctuations in energy use caused by weather conditions, rather than prevailing
2-38 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
economic conditions. In the long term, the residential sector is also affected by population growth, migration
trends toward warmer areas, and changes in housing and building attributes (e.g., larger sizes and improved
insulation). A shift toward energy-efficient products and more stringent energy efficiency standards for household
equipment has also contributed to recent trends in energy demand in households (EIA 2018).
Commercial
The commercial end-use sector, with electricity-related emissions distributed, accounts for 15.2 percent of U.S.
greenhouse gas emissions in 2021 and is heavily reliant on electricity for meeting energy needs, with electricity use
for lighting, heating, air conditioning, and operating appliances. The remaining emissions were largely due to the
direct consumption of natural gas and petroleum products, primarily for heating and cooking needs. Energy-
related emissions from the commercial sector have generally been increasing since 1990, and annual variations are
often correlated with short-term fluctuations in energy use caused by weather conditions, rather than prevailing
economic conditions. Decreases in energy-related emissions in the commercial sector in recent years can be
largely attributed to an overall reduction in energy use driven by a reduction in heating degree days and increases
in energy efficiency.
Municipal landfills and wastewater treatment are included in the commercial sector, with landfill emissions
decreasing since 1990 and wastewater treatment emissions increasing slightly.
Agriculture
The agriculture end-use sector accounts for 10.5 percent of U.S. greenhouse gas emissions in 2021 when
electricity-related emissions are distributed, and includes a variety of processes, including enteric fermentation in
domestic livestock, livestock manure management, and agricultural soil management. In 2021, agricultural soil
management was the largest source of N2O emissions, and enteric fermentation was the largest source of CFU
emissions in the United States. This sector also includes small amounts of CO2 emissions from fossil fuel
combustion by motorized farm equipment such as tractors.
Box 2-2: Trends in Various U.S. Greenhouse Gas Emissions-Related Data
Total (gross) greenhouse gas emissions can be compared to other economic and social indices to highlight
changes over time. These comparisons include: (1) emissions per unit of aggregate energy use, because energy-
related activities are the largest sources of emissions; (2) emissions per unit of fossil fuel consumption, because
almost all energy-related emissions involve the combustion of fossil fuels; (3) emissions per unit of total gross
domestic product as a measure of national economic activity; and (4) emissions per capita.
Table 2-14 provides data on various statistics related to U.S. greenhouse gas emissions normalized to 1990 as a
baseline year. These values represent the relative change in each statistic since 1990. Greenhouse gas emissions
in the United States have decreased at an average annual rate of 0.02 percent since 1990, although changes
from year to year have been significantly larger. This growth rate is slightly slower than that for total energy use,
overall gross domestic product (GDP) and national population (see Table 2-14 and Figure 2-16). The direction of
these trends started to change after 2005, when greenhouse gas emissions, total energy use and associated
fossil fuel consumption began to peak. Greenhouse gas emissions in the United States have decreased at an
average annual rate of 0.9 percent since 2005. Fossil fuel consumption has also decreased at a slower rate than
emissions since 2005, while total energy use, GDP, and national population, generally continued to increase,
noting 2020 was impacted by the COVID-19 pandemic.
Table 2-14: Recent Trends in Various U.S. Data (Index 1990 = 100)
Avg.
Avg.
Annual
Annual
Variable
1990
2005
2017
2018
2019
2020
2021
Change
Change
Greenhouse Gas Emissions'5
100
115
101
104
102
93
98
(+)%
-0.9%
Energy Usec
100
119
116
120
119
109
115
0.5%
-0.1%
Trends 2-39
-------
1
2
3
4
5
6
7
8
9
10
11
12
GDPd
Population6
100 159 193 199 203 198 209 2.4% 1.8%
100 118 130 130 131 133 134 0.9% 0.8%
+ Absolute value does not exceed 0.05 percent.
a Average annual growth rate.
b Gross total GWP-weighted values.
c Energy-content-weighted values (EIA 2022).
d GDP in chained 2009 dollars (BEA 2022).
e U.S. Census Bureau (2021).
Figure 2-16: U.S. Greenhouse Gas Emissions Per Capita and Per Dollar of Gross Domestic
Product
Source: BEA (2022), U.S. Census Bureau (2021), and gross emission estimates in this report.
2.3 Precursor Greenhouse Gas Emissions
(CO, NOx, NMVOCs, and S02) - TO BE
UPDATED FOR FINAL INVENTORY REPORT
The reporting requirements of the UNFCCC7 request that information be provided on emissions of compounds that
are precursors to greenhouse gases, which include carbon monoxide (CO), nitrogen oxides (NOx), non-methane
volatile organic compounds (NMVOCs), and sulfur dioxide (SO2). These gases are not direct greenhouse gases, but
can indirectly impact Earth's radiative balance, by altering the concentrations of other greenhouse gases (e.g.,
tropospheric ozone) and atmospheric aerosol (e.g., particulate sulfate). Carbon monoxide is produced when
carbon-containing fuels are combusted incompletely in energy, transportation, and industrial processes, and is also
emitted from practices such as agricultural burning and waste disposal and treatment. Anthropogenic sources of
nitrogen oxides (i.e., NO and NO2) are primarily fossil fuel combustion (for energy, transportation, industrial
7 See http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
2-40 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
process) and agricultural burning. Anthropogenic sources of NMVOCs, which include hundreds of organic
compounds that participate in atmospheric chemical reactions (i.e., propane, butane, xylene, toluene, ethane, and
many others)—are emitted primarily from transportation, industrial processes, oil and natural gas production,
waste practices, agricultural burning, and non-industrial consumption of organic solvents. In the United States, SO2
is primarily emitted from coal combustion for electric power generation and the metals industry.
As noted above and summarized in Chapter 6 of IPCC (2021), these compounds can have important indirect effects
of Earth's radiative balance. For example, reactions between NMVOCs and NOx in the presence of sunlight lead to
tropospheric ozone formation, a greenhouse gas. Concentrations of NMVOCs, NOx, and CO can also impact the
abundance and lifetime of primary greenhouse gases. This largely occurs by altering the atmospheric
concentrations of the hydroxyl radical (OH), which is the main sink for atmospheric CH4. For example, NOx
emissions can lead to increases in O3 concentrations and subsequent OH production, which will increase the
amount of OH molecules that are available to destroy CH4. In contrast, NMVOCs and CO can both react directly
with OH, leading to lower OH concentrations, a longer atmospheric lifetime of CH4, and a decrease in CO2
production (i.e., CO+OH-> CO2). Changes in atmospheric CH4 can also feedback on background concentrations of
tropospheric O3. Other indirect impacts include the formation of sulfate and nitrate aerosol from emissions of NOx
and SO2, both of which have a net negative impact on radiative forcing.
Since 1970, the United States has published triennial estimates of emissions of CO, NOx, NMVOCs, and SO2 (EPA
2021b), which are regulated under the Clean Air Act. Emissions of each of these precursor greenhouse gases has
decreased significantly since 1990 as a result of implementation of Clean Air Act programs, as well as technological
improvements.8 Precursor emission estimates for this report for 1990 through 2021 were obtained from data
published on EPA's National Emissions Inventory (NEI) Air Pollutants Emissions Trends Data website (EPA 2021b).
For Table 2-15, NEI-reported emissions of CO, NOx, SO2, and NMVOCs are recategorized from NEI Tier 1/Tier 2
source categories to those more closely aligned with IPCC categories, based on EPA (2022a) and detailed in Annex
6. Table 2-15 shows that fuel combustion accounts for the majority of emissions of these precursors. Industrial
processes—such as the manufacture of chemical and allied products, metals processing, and industrial uses of
solvents—are also significant sources of CO, NOx, and NMVOCs. Precursor emissions from Agriculture and LULUCF
categories are estimated separately and therefore are not taken from EPA (2021b); see Sections 5.7, 6.2, and 6.6.
Table 2-15: Emissions of NOx, CO, NMVOCs, and SO2 (kt)
Gas/Activity
1990
2005
2017
2018
2019
2020
2021
NOx
21,764
17,333
8,792
8,483
8,008
7,425
7,128
Energy
21,106
16,602
8,268
7,883
7,456
6,962
6,471
IPPU
592
572
402
397
397
397
397
LULUCF
52
142
107
188
139
50
244
Agriculture
13
15
14
14
14
14
14
Waste
+
2
1
1
1
1
1
CO
132,759
74,553
39,981
43,688
39,531
34,170
43,799
Energy
125,640
64,985
34,461
33,401
32,392
31,384
30,376
LULUCF
2,673
7,642
4,099
8,936
5,789
1,436
12,074
IPPU
4,129
1,557
1,075
1,007
1,007
1,007
1,007
Agriculture
315
363
340
339
338
337
336
Waste
1
7
6
5
5
5
5
NMVOCs
20,923
13,309
9,855
9,483
9,310
9,136
8,963
Energy
12,612
7,345
6,022
5,664
5,491
5,318
5,145
IPPU
7,638
5,849
3,776
3,767
3,767
3,767
3,767
Waste
673
114
57
52
52
52
52
Agriculture
NA
NA
NA
NA
NA
NA
NA
LULUCF
NA
NA
NA
NA
NA
NA
NA
8 More information is available online at: https://www.epa.gov/clean-air-act-overview/progress-cleaning-air-and-improving-
peoples-health and https://gispub.epa.gov/neireport/2017/.
Trends 2-41
-------
so.
20,935
13,196
2,906
2,303
2,211
1,943
1,780
Energy
19,628
12,364
2,439
1,794
1,701
1,433
1,270
IPPU
1,307
831
466
509
509
509
509
Waste
+
1
1
1
1
1
1
Agriculture
NA
NA
NA
NA
NA
NA
NA
LULUCF
NA
NA
NA
NA
NA
NA
NA
+ Does not exceed 0.5 kt.
NA (Not Available)
Note: Totals by gas may not sum due to independent rounding.
Source: (EPA 2021b) except for estimates from forest fires, grassland fires, and Field Burning of Agricultural Residues.
Emission categories from EPA (2021b) are aggregated into IPCC categories following as shown in Table ES-3.
2-42 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
3. Energy
if
Energy-related activities were the primary sources of U.S. anthropogenic greenhouse gas emissions, accounting for
82.1 percent of total greenhouse gas emissions on a carbon dioxide (CO2) equivalent basis in 2021.1 This included
96.5, 41.6, and 10.3 percent of the nation's CO2, methane (CH4), and nitrous oxide (N2O) emissions, respectively.
Energy-related CO2 emissions alone constituted 76.7 percent of U.S. greenhouse gas emissions from all sources on
a CC>2-equivalent basis, while the non-CC>2 emissions from energy-related activities represented a much smaller
portion of total national emissions (5.4 percent collectively).
Emissions from fossil fuel combustion comprise the vast majority of energy-related emissions, with CO2 being the
primary gas emitted (see Figure 3-1 and Figure 3-2). Globally, approximately 33,000 million metric tons (MMT) of
CO2 were added to the atmosphere through the combustion of fossil fuels in 2021, of which the United States
accounted for approximately 14 percent.2 Due to their relative importance over time (see Figure 3-2), fossil fuel
combustion-related CO2 emissions are considered in more detail than other energy-related emissions in this report
(see Figure 3-3).
Fossil fuel combustion also emits CFU and N2O. Stationary combustion of fossil fuels was the second largest source
of N2O emissions in the United States and mobile fossil fuel combustion was the fifth largest source. Energy-related
activities other than fuel combustion, such as the production, transmission, storage, and distribution of fossil fuels,
also emit greenhouse gases. These emissions consist primarily of fugitive CH4 emissions from natural gas systems,
coal mining, and petroleum systems.
1 Estimates are presented in units of million metric tons of carbon dioxide equivalent (MMT C02 Eq.), which weight each gas by
its global warming potential, or GWP, value. See section on global warming potentials in the Executive Summary.
2 Global C02 emissions from fossil fuel combustion were taken from International Energy Agency Global energy-related C02
emissions, 1990-2021 - Charts Available at: https://www.iea.org/data-and-statistics/charts/global-energy-related-co2-
emissions-1990-2021 (IEA 2022).
Energy 3-1
-------
Figure 3-1; 2021 Energy Sector Greenhouse Gas Sources
CO2 Emissions from Fossil Fuel Combustion
Natural Gas Systems
Non-Energy Use of Fuels
Petroleum Systems
Coal Mining
Non-CCh Emissions from Stationary Combustion
Non-CCh Emissions from Mobile Combustion
Incineration of Waste
Abandoned Oil and Gas Wells
Abandoned Underground Coal Mines
150
MMT CO2 Eq.
Figure 3-2: Trends in Energy Sector Greenhouse Gas Sources
8,000
7,000
6,000
5,000
4,000
3,000
2,000
1,000
0
Incineration of Waste
U.S Territories Fossil Fuel Combustion
I Non-Energy Use of Fuels
Commerical Fossil Fuel Combustion
Residential Fossil Fuel Combustion
I Fugitive Emissions
I Industrial Fossil Fuel Combustion
I Transportation Fossil Fuel Combustion
I Electric Power Fossil Fuel Combustion
M(NrN(NrM(NrMr\ilN(N(N(NNl\(NlM(N(NrvJ(NrN(N
Figure 3-3: 2021 U.S. Fossil Carbon Flows
3-2 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
International
2 Table 3-1 summarizes emissions from the Energy sector in units of MMT CO2 Eq., while unweighted gas emissions
3 in kilotons (kt) are provided in Table 3-2. Overall, emissions due to energy-related activities were 5,212.5 MMT CO2
4 Eq. in 2021,3 a decrease of 2.9 percent since 1990 and an increase of 6.5 percent since 2020. The increase in 2021
5 emissions was due to rebounding activity levels after the coronavirus (COVID-19) pandemic reduced overall
6 demand for fossil fuels across all sectors in 2020. Longer term trends are driven by a number of factors including a
7 shift from coal to natural gas and renewables in the electric power sector.
8 Table 3-1: CO2, ChU, and N2O Emissions from Energy (MMT CO2 Eq.)
Gas/Source
1990
2005
2017
2018
2019
2020
2021
C02
4,900.0
5,929.1
5,037.9
5,204.8
5,082.5
4,544.5
4,870.6
Fossil Fuel Combustion
4,728.2
5,747.3
4,852.5
4,989.8
4,853.4
4,344.8
4,651.0
Transportation
1,468.9
1,858.6
1,780.1
1,812.9
1,813.9
1,572.5
1,789.4
Electricity Generation
1,820.0
2,400.1
1,732.0
1,753.4
1,606.7
1,439.6
1,542.2
Industrial
852.4
850.8
789.0
813.5
815.9
767.9
762.4
Residential
338.6
358.9
293.4
338.2
341.4
313.2
310.1
Commercial
228.3
227.1
232.0
245.8
250.7
228.5
223.9
U.S. Territories
20.0
51.9
25.9
25.9
24.8
23.2
23.0
Non-Energy Use of Fuels
112.4
128.9
112.8
129.4
121.6
119.2
143.2
Natural Gas Systems
32.4
25.2
31.8
33.0
38.7
36.3
36.8
Petroleum Systems
9.5
10.2
24.5
36.1
46.9
29.1
24.7
Incineration of Waste
12.9
13.3
13.2
13.3
12.9
12.9
12.5
Coal Mining
4.6
4.2
3.2
3.1
3.0
2.2
2.5
Abandoned Oil and Gas Wells
+
+
+
+
+
+
+
Biomass-Wooda
215.2
206.9
212.0
220.0
217.7
200.4
202.8
Biofuels-Ethanola
4.2
22.9
82.1
81.9
82.6
71.8
79.1
International Bunker Fuelsb
103.6
113.3
120.2
122.2
116.1
69.6
69.3
Biofuels-Biodiesela
0.0
0.9
18.7
17.9
17.1
17.7
16.1
Biomass-MSWa
18.5
14.7
16.1
16.1
15.7
15.6
15.3
ch4
407.0
354.8
336.7
341.8
334.2
312.0
302.3
Natural Gas Systems
215.1
203.4
186.4
194.4
193.6
185.4
181.4
3 Following the current reporting requirements under the UNFCCC, this Inventory report presents C02 equivalent values based
on the IPCC Fifth Assessment Report (AR5) GWP values. See Chapter 1, Introduction for more information.
Energy 3-3
-------
Petroleum Systems
51.3
50.9
61.9
60.6
59.9
54.5
50.2
Coal Mining
108.1
71.8
61.4
59.1
53.0
46.2
44.7
Stationary Combustion
9.6
8.8
8.6
9.6
9.8
8.8
8.9
Abandoned Oil and Gas Wells
7.7
8.1
8.3
8.3
8.3
8.2
8.2
Abandoned Underground Coal
Mines
8.1
7.4
7.2
6.9
6.6
6.5
6.4
Mobile Combustion
7.2
4.4
2.9
2.9
2.9
2.6
2.6
Incineration of Waste
+
+
+
+
+
+
+
International Bunker Fuelsb
0.2
0.1
0.1
0.1
0.1
0.1
0.1
n2o
61.1
67.9
44.2
43.1
41.5
37.2
39.6
Stationary Combustion
22.3
30.5
25.3
25.1
22.2
20.7
22.1
Mobile Combustion
38.4
37.0
18.5
17.5
19.0
16.1
17.1
Incineration of Waste
0.4
0.3
0.4
0.4
0.4
0.3
0.4
Petroleum Systems
+
+
+
+
+
+
+
Natural Gas Systems
+
+
+
+
+
+
+
International Bunker Fuelsb
0.8
0.9
0.9
1.0
0.9
0.5
0.5
Total
5,368.2
6,351.8
5,418.8
5,589.7
5,458.3
4,893.8
5,212.5
+ Does not exceed 0.05 MMT C02 Eq.
a Emissions from Biomass and Biofuel Consumption are not included specifically in summing energy sector totals. Net
carbon fluxes from changes in biogenic carbon reservoirs are accounted for in the estimates for LULUCF.
b Emissions from International Bunker Fuels are not included in totals. These values are presented for informational
purposes only, in line with the 2006 IPCC Guidelines and UNFCCC reporting obligations.
Note: Totals may not sum due to independent rounding.
Table 3-2: CO2, ChU, and N2O Emissions from Energy (kt)
Gas/Source
1990
2005
2017
2018
2019
2020
2021
CO?
Fossil Fuel
Combustion
Non-Energy Use of
Fuels
Natural Gas
Systems
Petroleum Systems
Incineration of
Waste
Coal Mining
Abandoned Oil and
Gas Wells
Biomass-Wood"
Biofuels-Ethanola
International
Bunker Fuelsb
Biofuels-Biodiesela
Biomass-MSWa
CH4
Natural Gas
Systems
Petroleum Systems
Coal Mining
Stationary
Combustion
4,899,997
4,728,194
112,407
32,363
9,519
12,900
4,606.5
7
215,186
4,227
103,634
0
18,534
14,537
7,682
1,833
3,860
344
5,929,084
5,747,307
128,920
25,206
10,221
13,254
4,169.7
7
206,901
22,943
113,328
856
14,722
12,671
7,263
1,819
2,566
313
5,037,909 5,204,849 5,082,550 4,544,547 4,870,614
4,852,515 4,989,843 4,853,402 4,344,837 4,650,953
112,841
31,770
24,462
13,161
3,153.1
7
211,965
82,088
120,192
18,705
16,130
12,024
6,657
2,209
2,192
307
129,441
32,974
36,102
13,339
3,141.4
7
220,005
81,917
122,179
17,936
16,115
12,208
6,943
2,165
2,110
344
127,621
38,705
46,874
12,948
2,992.3
217,692
82,578
116,132
17,080
15,709
11,934
6,915
2,138
1,893
351
119,208 143,209
36,296
29,081
12,921
2,197.6
7
200,421
71,848
69,638
17,678
15,614
11,145
6,620
1,945
1,648
313
36,846
24,667
12,476
2,456.0
7
202,841
79,064
69,280
16,112
15,329
10,798
6,479
1,791
1,595
316
3-4 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Abandoned Oil and
Gas Wells
Abandoned
Underground
Coal Mines
Mobile
Combustion
Incineration of
Waste
International
Bunker Fuelsb
n2o
Stationary
Combustion
Mobile
Combustion
Incineration of
Waste
Petroleum Systems
Natural Gas
Systems
International
Bunker Fuelsb
274
288
258
7
231
84
145
2
+
289
264
158
5
256
115
140
1
+
295
257
105
4
167
95
70
1
+
296
247
102
4
163
95
66
1
+
297
237
103
4
157
84
72
1
+
295
232
92
3
140
78
61
1
+
295
228
94
3
149
83
65
1
+
+ Does not exceed 0.5 kt.
a Emissions from Biomass and Biofuel Consumption are not included specifically in summing Energy sector totals. Net
carbon fluxes from changes in biogenic carbon reservoirs are accounted for in the estimates for LULUCF.
b Emissions from International Bunker Fuels are not included in totals. These values are presented for informational
purposes only, in line with the 2006 IPCC Guidelines and UNFCCC reporting obligations.
Note: Totals by gas may not sum due to independent rounding.
Emissions estimates reported in the Energy chapter from fossil fuel combustion and fugitive sources include those
from all 50 states, including Hawaii and Alaska, and the District of Columbia. Emissions are also included from U.S.
Territories to the extent they are known to occur (e.g., coal mining does not occur in U.S. Territories). For some
sources there is a lack of detailed information on U.S. Territories including some non-CC>2 emissions from biomass
combustion. As part of continuous improvement efforts, EPA reviews this on an ongoing basis to ensure emission
sources are included across all geographic areas including U.S. Territories if they are occurring. See Annex 5 for
more information on EPA's assessment of the sources not included in this Inventory.
Each year, some emission and sink estimates in the Inventory are recalculated and revised with improved methods
and/or data. In general, recalculations are made to the U.S. greenhouse gas emission estimates either to
incorporate new methodologies or, most commonly, to update recent historical data. These improvements are
implemented consistently across the previous Inventory's time series (i.e., 1990 to 2020) to ensure that the trend
is accurate. Key updates in this year's Inventory include, updates to the transportation methodology which use
distributions ofehicle miles traveled (VMT) and fuel use from EPA's MOVES3 model to estimate vehicle emissions
by vehicle class, updates to the Cm and N2O emission factors for alternative fuel vehicles based on the GREET2022
model,, revisions to methods for estimating CH4 from both Natural Gas Systems and Petroleum Systems now
incorporate additional basin-level data from GHGRP Subpart W for several emission sources in the onshore
production segment, changes to the Non-Energy Use of Fossil Fuel methodology (e.g., updated energy
consumption statistics, updated polyester fiber and acetic acid production data, updated import and export data,
and updated shipment data from the U.S census Bureau), and accounting for biogenic emissions from combusted
MSW within Biomass estimates. In addition, the GWPs for calculating CC>2-equivalent totals emissions of CFU and
N2O have been revised to reflect the 100-year global warming potentials (GWPs) provided in the IPCC Fifth
Assessment Report (AR5) (IPCC 2013). AR5 GWP values differ slightly from those presented in the IPCC Fourth
Assessment Report (AR4) (IPCC 2007) (used in the previous Inventories). The combined impact of these
Energy 3-5
-------
1 recalculations averaged 9.6 MMT CO2 Eq. (+0.2 percent) per year across the time series. For more information on
2 specific methodological updates, please see the Recalculations Discussion section for each category in this chapter.
3
4
Box 3-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals, including
Relationship to EPA's Greenhouse Gas Reporting Program
In following the United Nations Framework Convention on Climate Change (UNFCCC) requirement under Article
4.1 to develop and submit national greenhouse gas emission inventories, the emissions and removals presented
in this report and this chapter are organized by source and sink categories and calculated using internationally-
accepted methods provided by the Intergovernmental Panel on Climate Change (IPCC) in the 2006IPCC
Guidelines for National Greenhouse Gas Inventories (2006 IPCC Guidelines). Additionally, the calculated
emissions and removals in a given year for the United States are presented in a common format in line with the
UNFCCC reporting guidelines for the reporting of inventories under this international agreement. The use of
consistent methods to calculate emissions and removals by all nations providing their inventories to the
UNFCCC ensures that these reports are comparable. The presentation of emissions and removals provided in
the Energy chapter do not preclude alternative examinations, but rather, this chapter presents emissions and
removals in a common format consistent with how countries are to report Inventories under the UNFCCC. The
report itself, and this chapter, follows this standardized format, and provides an explanation of the application
of methods used to calculate emissions and removals from energy-related activities.
Energy Data from EPA's Greenhouse Gas Reporting Program
EPA's Greenhouse Gas Reporting Program (GHGRP)4 dataset and the data presented in this Inventory are
complementary. The Inventory was used to guide the development of the GHGRP, particularly in terms of scope
and coverage of both sources and gases. The GHGRP dataset continues to be an important resource for the
Inventory, providing not only annual emissions information, but also other annual information, such as activity
data and emission factors that can improve and refine national emission estimates and trends over time.
GHGRP data also allow EPA to disaggregate national inventory estimates in new ways that can highlight
differences across regions and sub-categories of emissions, along with enhancing application of QA/QC
procedures and assessment of uncertainties.
EPA uses annual GHGRP data in a number of Energy sector categories to improve the national estimates
presented in this Inventory consistent with IPCC guidelines (see Box 3-3 of this chapter, and Sections 3.3
Incineration of Waste, 3.4 Coal Mining, 3.6 Petroleum Systems, and 3.7 Natural Gas Systems).5 Methodologies
used in EPA's GHGRP are consistent with IPCC guidelines, including higher tier methods. Under EPA's GHGRP,
facilities collect detailed information specific to their operations according to detailed measurement standards.
It should be noted that the definitions and provisions for reporting fuel types in EPA's GHGRP may differ from
those used in the Inventory in meeting the UNFCCC reporting guidelines. In line with the UNFCCC reporting
guidelines, the Inventory report is a comprehensive accounting of all emissions from fuel types identified in the
IPCC guidelines and provides a separate reporting of emissions from biomass.
In addition to using GHGRP data to estimate emissions (Sections 3.3 Incineration of Waste, 3.4 Coal Mining, 3.6
Petroleum Systems, and 3.7 Natural Gas Systems), EPA also uses the GHGRP fuel consumption activity data in
the Energy sector to disaggregate industrial end-use sector emissions in the category of CO2 Emissions from
Fossil Fuel Combustion, for use in reporting emissions in Common Reporting Format (CRF) tables (See Box 3-3).
The industrial end-use sector activity data collected for the Inventory (EIA 2022) represent aggregated data for
the industrial end-use sector. EPA's GHGRP collects industrial fuel consumption activity data by individual
categories within the industrial end-use sector. Therefore, GHGRP data are used to provide a more detailed
breakout of total emissions in the industrial end-use sector within that source category.
4 On October 30, 2009, the U.S. Environmental Protection Agency (EPA) published a rule requiring annual reporting of
greenhouse gas data from large greenhouse gas emission sources in the United States. Implementation of the rule, codified at
40 CFR Part 98, is referred to as EPA's Greenhouse Gas Reporting Program (GHGRP).
5 See http://www.ipcc-nggip.iges.or.jp/public/tb/TFI Technical Bulletin l.pdf.
3-6 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
As indicated in the respective Planned Improvements sections for source categories in this chapter, EPA
continues to examine the uses of facility-level GHGRP data to improve the national estimates presented in this
Inventory. See Annex 9 for more information on use of EPA's GHGRP in the Inventory.
3.1 Fossil Fuel Combustion (CRF Source
Category 1A)
Emissions from the combustion of fossil fuels for energy include the greenhouse gases CO2, CH4, and N2O. Given
that CO2 is the primary gas emitted from fossil fuel combustion and represents the largest share of U.S. total
emissions, CO2 emissions from fossil fuel combustion are discussed at the beginning of this section. An overview of
Cm and N2O emissions from the combustion of fuels in stationary sources is then presented, followed by fossil fuel
combustion emissions for all three gases by sector: electric power, industrial, residential, commercial, U.S.
Territories, and transportation.
Methodologies for estimating CO2 emissions from fossil fuel combustion differ from the estimation of CH4 and N2O
emissions from stationary combustion and mobile combustion. Thus, three separate descriptions of
methodologies, uncertainties, recalculations, and planned improvements are provided at the end of this section.
Total CO2, CH4, and N2O emissions from fossil fuel combustion are presented in Table 3-3 and Table 3-4.
Table 3-3: CO2, ChU, and N2O Emissions from Fossil Fuel Combustion (MMT CO2 Eq.)
Gas
1990
2005
2017
2018
2019
2020
2021
C02
4,728.2
5,747.3
4,852.5
4,989.8
4,853.4
4,344.8
4,651.0
ch4
16.8
13.2
11.5
12.5
12.7
11.3
11.5
n2o
60.7
67.6
43.8
42.6
41.1
36.8
39.2
Total
4,805.7 |
5,828.0 |
4,907.9
5,045.0
4,907.3
4,392.9
4,701.7
Note: Totals may not sum due to independent rounding.
ible 3-4: CO2, ChU, and N2O Emissions from Fossil Fuel Combustion (kt)
Gas
1990
2005
2017
2018
2019
2020
2021
C02
4,728,194
5,747,307
4,852,515
4,989,843
4,853,402
4,344,837
4,650,953
ch4
601
471
412
446
454
405
410
n2o
229
255
165
161
155
139
148
CO2 from Fossil Fuel Combustion
Carbon dioxide is the primary gas emitted from fossil fuel combustion and represents the largest share of U.S. total
greenhouse gas emissions. Carbon dioxide emissions from fossil fuel combustion are presented in Table 3-5. In
2021, CO2 emissions from fossil fuel combustion increased by 7.0 percent relative to the previous year (as shown in
Table 3-6). The increase in CO2 emissions from fossil fuel consumption was a result of a 5.9 percent increase in
fossil fuel energy use. This increase in fossil fuel consumption was due primarily rebounding economic activity after
the COVID-19 pandemic. Carbon dioxide emissions from natural gas increased by 8.3 MMT CO2 Eq., a 0.5 percent
increase from 2020. In a shift from recent trends, CO2 emissions from coal consumption increased by 122.1 MMT
CO2 Eq., a 14.6 percent increase from 2020. The increase in natural gas consumption and emissions in 2021 is
observed across all sectors except the Electric Power sector and U.S. Territories, while the coal increase is primarily
in the Electric Power sector. Emissions from petroleum use also increased 175.8 MMT CO2 Eq. (9.3 percent) from
Energy 3-7
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
2020 to 2021. In 2021, CO2 emissions from fossil fuel combustion were 4,651.0 MMT CO2 Eq., or 1.6 percent below
emissions in 1990 (see Table 3-5).6
Table 3-5: CO2 Emissions from Fossil Fuel Combustion by Fuel Type and Sector (MMT CO2
Eq.)
Fuel/Sector
1990
2005
2017
2018
2019
2020
2021
Coal
1,719.8
2,113.7
1,270.0
1,211.6
1,028.2
835.6
957.7
Residential
3.0
0.8
NO
NO
NO
NO
NO
Commercial
12.0
9.3
2.0
1.8
1.6
1.4
1.4
Industrial
157.8
117.8
58.7
54.4
49.5
43.0
43.7
Transportation
NO
NO
NO
NO
NO
NO
NO
Electric Power
1,546.5
1,982.8
1,207.1
1,152.9
973.5
788.2
909.7
U.S. Territories
0.5
3.0
2.3
2.6
3.6
3.1
2.9
Natural Gas
998.6
1,166.2
1,433.2
1,592.0
1,649.3
1,612.4
1,620.7
Residential
237.8
262.2
241.5
273.8
275.5
256.4
258.6
Commercial
142.0
162.9
173.2
192.5
192.9
173.8
180.9
Industrial
407.4
387.8
468.1
493.5
501.5
486.1
498.4
Transportation
36.0
33.1
42.3
50.9
58.9
58.7
65.1
Electric Power
175.4
318.9
505.6
577.9
616.6
634.8
615.1
U.S. Territories
NO
1.3
2.5
3.3
3.8
2.6
2.6
Petroleum
2,009.2
2,467.0
2,148.8
2,185.8
2,175.6
1,896.4
2,072.2
Residential
97.8
95.9
51.9
64.4
65.9
56.8
51.5
Commercial
74.3
54.9
56.8
51.5
56.2
52.8
41.4
Industrial
287.1
345.2
262.2
265.6
264.9
238.9
220.3
Transportation
1,432.9
1,825.5
1,737.8
1,762.0
1,754.9
1,513.9
1,724.3
Electric Power
97.5
98.0
18.9
22.2
16.2
16.2
17.1
U.S. Territories
19.5
47.6
21.1
20.1
17.5
17.5
17.5
Geothermal3
0.5
0.5
0.4
0.4
0.4
0.4
0.4
Electric Power
0.5
0.5
0.4
0.4
0.4
0.4
0.4
Total
4,728.2
5,747.3
4,852.5
4,989.8
4,853.4
4,344.8
4,651.0
NO (Not Occurring)
a Although not technically a fossil fuel, geothermal energy-related C02 emissions are included for reporting purposes.
The source of C02 is non-condensable gases in subterranean heated water.
Note: Totals may not sum due to independent rounding.
Trends in CO2 emissions from fossil fuel combustion are influenced by many long-term and short-term factors. On
a year-to-year basis, the overall demand for fossil fuels in the United States and other countries generally
fluctuates in response to changes in general economic conditions, energy prices, weather, and the availability of
non-fossil alternatives. For example, in a year with increased consumption of goods and services, low fuel prices,
severe summer and winter weather conditions, nuclear plant closures, and lower precipitation feeding
hydroelectric dams, there would likely be proportionally greater fossil fuel consumption than a year with poor
economic performance, high fuel prices, mild temperatures, and increased output from nuclear and hydroelectric
plants. The 2020 to 2021 trends were particularly impacted by the COVID-19 pandemic which generally led to a
reduction in demand for fossil fuels in 2020, but an increase in demand as activities rebounded in 2021.
Longer-term changes in energy usage patterns, however, tend to be more a function of aggregate societal trends
that affect the scale of energy use (e.g., population, number of cars, size of houses, and number of houses), the
efficiency with which energy is used in equipment (e.g., cars, HVAC systems, power plants, steel mills, and light
bulbs), and social planning and consumer behavior (e.g., walking, bicycling, or telecommuting to work instead of
driving).
6 An additional discussion of fossil fuel emission trends is presented in the Trends in U.S. Greenhouse Gas Emissions chapter.
3-8 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Carbon dioxide emissions also depend on the source of energy and its carbon (C) intensity. The amount of C in
fuels varies significantly by fuel type. For example, coal contains the highest amount of C per unit of useful energy.
Petroleum has roughly 75 percent of the C per unit of energy as coal, and natural gas has only about 55 percent.7
Table 3-6 shows annual changes in emissions during the last five years for coal, petroleum, and natural gas in
selected sectors.
Table 3-6: Annual Change in CO2 Emissions and Total 2021 CO2 Emissions from Fossil Fuel
Combustion for Selected Fuels and Sectors (MMT CO2 Eq. and Percent)
Sector
Fuel Type
2017 to 2018
2018 to 2019
2019 to 2020
2020 to 2021
Total 2021
Transportation
Petroleum
24.1
1.4%
-7.0
-0.4%
-241.1
-13.7%
210.4
13.9%
1,724.3
Electric Power
Coal
-54.2
-4.5%
-179.3
-15.6%
-185.4
-19.0%
121.6
15.4%
909.7
Electric Power
Natural Gas
72.3
14.3%
38.7
6.7%
18.2
3.0%
-19.8
-3.1%
634.8
Industrial
Natural Gas
25.3
5.4%
8.0
1.6%
-15.5
-3.1%
12.3
2.5%
498.4
Residential
Natural Gas
32.3
13.4%
1.7
0.6%
-19.1
-6.9%
2.3
0.9%
258.6
Commercial
Natural Gas
19.3
11.2%
0.4
0.2%
-19.1
-9.9%
7.0
4.0%
180.9
Transportation
All Fuels3
32.8
1.8%
1.0
0.1%
-241.3
-13.3%
216.9
13.8%
1,789.4
Electric Power
All Fuels3
21.4
1.2%
-146.7
-8.4%
-167.2
-10.4%
102.6
7.1%
1,542.2
Industrial
All Fuels3
24.5
3.1%
2.4
0.3%
-48.0
-5.9%
-5.5
-0.7%
762.4
Residential
All Fuels3
44.8
15.3%
3.2
0.9%
-28.2
-8.3%
-3.1
-1.0%
310.1
Commercial
All Fuels3
13.8
6.0%
4.9
2.0%
-22.2
-8.9%
-4.6
-2.0%
223.9
All Sectorsa'b
All Fuels3
137.3
2.8%
-136.4
-2.7%
-508.6
-10.5%
306.1
7.0%
4,651.0
a Includes sector and fuel combinations not shown in this table.
b Includes U.S. Territories.
As shown in Table 3-6, recent trends in CO2 emissions from fossil fuel combustion show a 2.8 percent increase
from 2017 to 2018, a 2.7 percent decrease from 2018 to 2019, a 10.5 percent decrease from 2019 to 2020, and a
7.0 percent increase from 2020 to 2021. These changes contributed to an overall 4.2 percent decrease in CO2
emissions from fossil fuel combustion from 2017 to 2021.
Recent trends in CO2 emissions from fossil fuel combustion are largely driven by the electric power sector, which
until 2017 has accounted for the largest portion of these emissions. The types of fuels consumed to produce
electricity have changed in recent years. Electric power sector consumption of natural gas primarily increased due
to increased production capacity as natural gas-fired plants replaced coal-fired plants and increased electricity
demand related to heating and cooling needs (EIA 2018; EIA 2022e). Total net electric power generation from all
fossil and non-fossil sources increased by 3.6 percent from 2017 to 2018, decreased by 1.3 percent from 2018 to
2019, decreased by 2.9 percent from 2019 to 2020, and increased by 2.8 percent from 2020 to 2021 (EIA 2022a).
Carbon dioxide emissions from the electric power sector increased from 2020 to 2021 by 7.1 percent due to
increased production and the increased use of coal for electric power generation. Carbon dioxide emissions from
coal consumption for electric power generation decreased by 24.6 percent overall since 2017, but increased by
15.4 percent from 2020 to 2021.
The recent trends in CO2 emissions from fossil fuel combustion also follow changes in heating degree days (see Box
3-2). Emissions from natural gas consumption in the residential and commercial sectors increased by 7.1 percent
and 4.4 percent from 2017 to 2021, respectively. This trend can be partially attributed to a 2.6 percent increase in
heating degree days from 2017 to 2021, which led to an increased demand for heating fuel and electricity for heat
in these sectors. Industrial consumption of natural gas is dependent on market effects of supply and demand in
addition to weather-related heating needs.
Petroleum use in the transportation sector is another major driver of emissions, representing the largest source of
CO2 emissions from fossil fuel combustion in 2021. Emissions from petroleum consumption for transportation have
decreased by 0.8 percent since 2017 and are primarily attributed to a 0.5 percent decrease in VMT over the same
7 Based on national aggregate carbon content of all coal, natural gas, and petroleum fuels combusted in the United States. See
Annex 2.2 for more details on fuel carbon contents.
Energy 3-9
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
time period. Beginning with 2017, the transportation sector is the largest source of national CO2 emissions-
whereas in prior years, electric power was the largest source sector.
The overall 2020 to 2021 trends were largely driven by the gradual recovery from the COVID-19 pandemic, which
saw reduced economic activity in 2020 and caused changes in energy demand and supply patterns across different
sectors. The recovery from the COVID-19 pandemic generally led to increased energy use and emissions across all
economic sectors from 2020 to 2021. The increase in emissions from 2020 to 2021 was also due to a reversal in
recent trends in coal use. In recent years the trend has been one of decreased coal use however, from 2020 to
2021 overall use of coal increased by 14.6 percent (EIA 2022a).
In the United States, 79.3 percent of the energy used in 2021 was produced through the combustion of fossil fuels
such as petroleum, natural gas, and coal (see Figure 3-4 and Figure 3-5). Specifically, petroleum supplied the
largest share of domestic energy demands, accounting for 36 percent of total U.S. energy used in 2021. Natural gas
and coal followed in order of fossil fuel energy demand importance, accounting for approximately 32 percent and
11 percent of total U.S. energy used, respectively. Petroleum was consumed primarily in the transportation end-
use sector and the majority of coal was used in the electric power sector. Natural gas was broadly consumed in all
end-use sectors except transportation (see Figure 3-6) (EIA 2021c). The remaining portion of energy used in 2021
was supplied by nuclear electric power (8 percent) and by a variety of renewable energy sources (12 percent),
primarily wind energy, hydroelectric power, solar, geothermal and biomass (EIA 2021c).8
Figure 3-4: 2021 U.S. Energy Use by Energy Source
Nuclear Electric Power
8 Renewable energy, as defined in ElA's energy statistics, includes the following energy sources: hydroelectric power,
geothermal energy, biomass, solar energy, and wind energy.
3-10 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Figure 3-5; Annual U.S. Energy Use
4
5
6
7
8
9
10
120
§, 100
o
Cl
E
E
3
i/>
80
60
40
20
Total Energy
Fossil Fuels
Renewable & Nuclear
OT-HrMmTj-m^or-.coc^
OiOiaOiO^OiCJiOlOiOi
fM(NNfMfM(MrN(NlM(N(NfM(NfM(N(NlN(\(NfMfMlN
Figure 3-6: 2021 CO2 Emissions from Fossil Fuel Combustion by Sector and Fuel Type
2,500
2,000
1,500
1,000
500
Relative Contribution by Fuel Type
<0.05%
(Geothermal)
1,789
Coal
Natural Gas
I Geothermal
I Petroleum
1,542
224
23
U.S. Territories
Commercial
Residential
Industrial
Electricity Generation Transportation
Fossil fuels are generally combusted for the purpose of producing energy for useful heat and work. During the
combustion process, the C stored in the fuels is oxidized and emitted as CO2 and smaller amounts of other gases,
including CH4, carbon monoxide (CO), and non-methane volatile organic compounds (NMVOCs).9 These other C-
containing non-CCh gases are emitted as a byproduct of incomplete fuel combustion, but are, for the most part,
eventually oxidized to CO2 in the atmosphere. Therefore, as per IPCC guidelines it is assumed all of the C in fossil
fuels used to produce energy is eventually converted to atmospheric CO2.
11
Box 3-2: Weather and Non-Fossil Energy Effects on CO2 Emissions from Fossil Fuel Combustion Trends
The United States in 2021 experienced a colder winter overall compared to 2020, as heating degree days
increased 0.5 percent. Colder winter conditions compared to 2020 impacted the amount of energy required for
heating. In 2021 heating degree days in the United States were 9.3 percent below normal (see Figure 3-7).
Cooling degree days decreased by 1.9 percent compared to 2020, which decreased demand for air conditioning
in the residential and commercial sector. Cooler summer conditions compared to 2020 impacted the amount of
y See the sections entitled Stationary Combustion and Mobile Combustion in this chapter for information on non-C02 gas
emissions from fossil fuel combustion.
Energy 3-11
-------
energy required for cooling. 2020 cooling degree days in the United States were 11.8 percent above normal (see
Figure 3-8) (EIA 2022a).10 The combination of colder winter and summer conditions led to overall residential and
commercial energy consumption decrease of 1.0 and 2.0 percent, respectively relative to 2020.
Figure 3-7: Annual Deviations from Normal Heating Degree Days for the United States
(1950-2021, Index Normal = 100)
30
20
§ 10
[O
'§
Q
-10
-20
Normal
(4,313 Heating Degree Days)
99% Confidence
Note: Climatological normal data are highlighted in dark red. Statistical confidence interval for "normal" climatology period of
-30 1991 through 2021.
T-HmLnr^cri-rHmLnrxO^T-ifoinisHC>^romr^.c^-^Hmmr^.cr>-^ifOLnr^CT>-rHtriLnr,«.CT^'-H
iDLnLnLnLn^vDvcvD^r^r^-r^r^r^cocococococric^c^criCTiOoooo-rH-^-t-^H-i-i-^-ifNj
o>CTiciCT*oiCTiO'icricricriCT*o^CTic>cricyiCT>cr>CT>c^cj'iooooooooooo
¦»—ItHt—iT-H-i—It-I-tH-tH*—It-l-rHi-l'r-l-r-4-t-l-n-l'r-l.r-l-r—!¦»—I*—tTH-rHrMfNrsJfNfNrMrMrMrMrMtN
Figure 3-8: Annual Deviations from Normal Cooling Degree Days for the United States
(1950-2021, Index Normal = 100)
-20
-30
-40
Normal
(1,099 Cooling Degree Days)
99% Confidence ¦
P*P"I'
Note: Climatological normal data are highlighted dark blue. Statistical confidence interval for "normal" climatology period of
1991 through 2021.
in n oi
Lnr-.CTiTHmLni-N.CT>
b-l N d
OsiOsiO>OsiCT>CTiCT>CT>CT^CTiCT>CTiCT>CT>CriO>iCTiCXiCT'iC7>
uir».criT-r-.CT>
- - - OOOOO-i-H-i-Hi-t-f-Hi-H
oooooooooo _
(NfMlNfMININlNfMlNlNfM
1
10 Degree days are relative measurements of outdoor air temperature. Heating degree days are deviations of the mean daily
temperature below 65 degrees Fahrenheit, while cooling degree days are deviations of the mean daily temperature above 65
degrees Fahrenheit. Heating degree days have a considerably greater effect on energy demand and related emissions than do
cooling degree days. Excludes Alaska and Hawaii. Normals are based on data from 1991 through 2020. The variation in these
normals during this time period was ±16 percent and ±27 percent for heating and cooling degree days, respectively (99 percent
confidence interval).
3-12 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
The carbon intensity of the electric power sector is impacted by the amount of non-fossil energy sources of
electricity. The utilization (i.e., capacity factors)11 of nuclear power plants in 2021 remained high at 93 percent. In
2021, nuclear power represented 20 percent of total electricity generation. Since 1990, the wind and solar power
sectors have shown strong growth and have become relatively important electricity sources. Between 1990 and
2021, renewable energy generation (in kWh) from solar and wind energy have increased from 0.1 percent in 1990
to 12 percent in 2021 of total electricity generation, which helped drive the decrease in the carbon intensity of the
electricity supply in the United States.
Stationary Combustion
The direct combustion of fuels by stationary sources in the electric power, industrial, commercial, and residential
sectors represent the greatest share of U.S. greenhouse gas emissions. Table 3-7 presents CO2 emissions from
fossil fuel combustion by stationary sources. The CO2 emitted is closely linked to the type of fuel being combusted
in each sector (see Methodology section of CO2 from Fossil Fuel Combustion). In addition to the CO2 emitted from
fossil fuel combustion, CH4 and N2O are emitted as well. Table 3-8 and Table 3-9 present CFU and N2O emissions
from the combustion of fuels in stationary sources. The CFU and N2O emissions are linked to the type of fuel being
combusted as well as the combustion technology (see Methodology section for CH4 and N2O from Stationary
Combustion).
Table 3-7: CO2 Emissions from Stationary Fossil Fuel Combustion (MMT CO2 Eq.)
Sector/Fuel Type
1990
2005
2017
2018
2019
2020
2021
Electric Power
1,820.0
2,400.1
1,732.0
1,753.4
1,606.7
1,439.6
1,542.2
Coal
1,546.5
1,982.8
1,207.1
1,152.9
973.5
788.2
909.7
Natural Gas
175.4
318.9
505.6
577.9
616.6
634.8
615.1
Fuel Oil
97.5
98.0
18.9
22.2
16.2
16.2
17.1
Geothermal
0.5
0.5
0.4
0.4
0.4
0.4
0.4
Industrial
852.4
850.8
789.0
813.5
815.9
767.9
762.4
Coal
157.8
117.8
58.7
54.4
49.5
43.0
43.7
Natural Gas
407.4
387.8
468.1
493.5
501.5
486.1
498.4
Fuel Oil
287.1
345.2
262.2
265.6
264.9
238.6
220.2
Commercial
228.3
227.1
232.0
245.8
250.7
228.5
223.9
Coal
12.0
9.3
2.0
1.8
1.6
1.4
1.4
Natural Gas
142.0
162.9
173.2
192.5
192.9
173.8
180.9
Fuel Oil
74.3
54.9
56.8
51.5
56.2
52.8
41.4
Residential
338.6
358.9
293.4
338.2
341.4
313.2
310.1
Coal
3.0
0.8
NO
NO
NO
NO
NO
Natural Gas
237.8
262.2
241.5
273.8
275.5
256.4
258.6
Fuel Oil
97.8
95.9
51.9
64.4
65.9
56.8
51.5
U.S. Territories
20.0
51.9
25.9
25.9
24.8
23.2
23.0
Coal
0.5
3.0
2.3
2.6
3.6
3.1
2.9
Natural Gas
NO
1.3
2.5
3.3
3.8
2.6
2.6
Fuel Oil
19.5
47.6
21.1
20.1
17.5
17.5
17.5
Total
3,259.3
3,888.8
3,072.4
3,176.9
3,039.5
2,772.3
2,861.6
NO (Not Occurring)
Note: Totals may not sum due to independent rounding.
11 The capacity factor equals generation divided by net summer capacity. Summer capacity is defined as "The maximum output
that generating equipment can supply to system load, as demonstrated by a multi-hour test, at the time of summer peak
demand (period of June 1 through September 30)" (EIA 2020a). Data for both the generation and net summer capacity are from
EIA (2022a).
Energy 3-13
-------
l Table 3-8: ChU Emissions from Stationary Combustion (MMT CO2 Eq.)
Sector/Fuel Type
1990
2005
2017
2018
2019
2020
2021
Electric Power
0.5
1.0
1.2
1.4
1.4
1.4
1.4
Coal
0.4
0.4
0.3
0.3
0.2
0.2
0.2
Fuel Oil
+
+
+
+
+
+
+
Natural gas
0.1
0.5
1.0
1.1
1.2
1.2
1.2
Wood
+
+
+
+
+
+
+
Industrial
2.0
1.9
1.7
1.7
1.7
1.6
1.6
Coal
0.5
0.3
0.2
0.2
0.1
0.1
0.1
Fuel Oil
0.2
0.2
0.2
0.2
0.2
0.2
0.1
Natural gas
0.2
0.2
0.2
0.2
0.3
0.2
0.3
Wood
1.2
1.2
1.2
1.1
1.1
1.1
1.1
Commercial
1.2
1.2
1.3
1.4
1.4
1.3
1.3
Coal
+
+
+
+
+
+
+
Fuel Oil
0.3
0.2
0.2
0.2
0.2
0.2
0.2
Natural gas
0.4
0.4
0.4
0.5
0.5
0.4
0.5
Wood
0.5
0.6
0.7
0.7
0.7
0.7
0.7
Residential
5.9
4.5
4.2
5.1
5.3
4.4
4.6
Coal
0.3
0.1
NO
NO
NO
NO
NO
Fuel Oil
0.4
0.4
0.2
0.3
0.3
0.2
0.2
Natural Gas
0.6
0.7
0.6
0.7
0.7
0.6
0.6
Wood
4.6
3.4
3.4
4.2
4.4
3.5
3.7
U.S. Territories
+
0.1
+
+
+
+
+
Coal
+
+
+
+
+
+
+
Fuel Oil
+
0.1
+
+
+
+
+
Natural Gas
0.0
+
+
+
+
+
+
Wood
NE
NE
NE
NE
NE
NE
NE
Total
9.6
8.8
8.6
9.6
9.8
8.8
8.9
+ Does not exceed 0.05 MMT C02 Eq.
NO (Not Occurring)
NE (Not Estimated)
Note: Totals may not sum due to independent rounding.
2 Table 3-9: N2O Emissions from Stationary Combustion (MMT CO2 Eq.)
Sector/Fuel Type
1990
2005
2017
2018
2019
2020
2021
Electric Power
18.2
26.7
22.0
21.7
18.8
17.5
19.0
Coal
17.9
24.9
18.8
18.1
14.8
13.5
15.1
Fuel Oil
0.1
0.1
+
+
+
+
+
Natural Gas
0.3
1.7
3.2
3.6
3.9
4.0
3.9
Wood
+
+
+
+
+
+
+
Industrial
2.7
2.6
2.3
2.2
2.2
2.1
2.0
Coal
0.7
0.5
0.2
0.2
0.2
0.2
0.2
Fuel Oil
0.4
0.5
0.3
0.3
0.3
0.3
0.2
Natural Gas
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Wood
1.5
1.5
1.5
1.4
1.4
1.4
1.4
Commercial
0.3
0.3
0.3
0.3
0.3
0.3
0.3
Coal
+
+
+
+
+
+
+
Fuel Oil
0.2
0.1
0.1
0.1
0.1
0.1
0.1
Natural Gas
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Wood
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Residential
0.9
0.8
0.7
0.8
0.8
0.7
0.7
Coal
+
+
NO
NO
NO
NO
NO
Fuel Oil
0.2
0.2
0.1
0.2
0.2
0.1
0.1
3-14 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Natural Gas
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Wood
0.6
0.4
0.4
0.5
0.5
0.4
0.5
U.S. Territories
+
0.1
0.1
0.1
0.1
0.1
0.1
Coal
+
+
+
+
+
+
+
Fuel Oil
+
0.1
+
+
+
+
+
Natural Gas
0.0
+
+
+
+
+
+
Wood
NE
NE
NE
NE
NE
NE
NE
Total
22.3
30.5
25.3
25.1
22.2
20.7
22.1
+ Does not exceed 0.05 MMT C02 Eq.
NO (Not Occurring)
NE (Not Estimated)
Note: Totals may not sum due to independent rounding.
1 Fossil Fuel Combustion Emissions by Sector
2 Table 3-10 provides an overview of the CO2, CH4, and N2O emissions from fossil fuel combustion by sector,
3 including transportation, electric power, industrial, residential, commercial, and U.S. Territories.
4 Table 3-10: CO2, ChU, and N2O Emissions from Fossil Fuel Combustion by Sector (MMT CO2
5 Eq.)
End-Use Sector
1990
2005
2017
2018
2019
2020
2021
Transportation
1,514.6
1,900.0
1,801.6
1,833.3
1,835.7
1,591.2
1,809.2
C02
1,468.9
1,858.6
1,780.1
1,812.9
1,813.9
1,572.5
1,789.4
ch4
7.2
4.4
2.9
2.9
2.9
2.6
2.6
n2o
38.4
37.0
18.5
17.5
19.0
16.1
17.1
Electric Power
1,838.7
2,427.8
1,755.3
1,776.5
1,626.9
1,458.5
1,562.6
C02
1,820.0
2,400.1
1,732.0
1,753.4
1,606.7
1,439.6
1,542.2
ch4
0.5
1.0
1.2
1.4
1.4
1.4
1.4
n2o
18.2
26.7
22.0
21.7
18.8
17.5
19.0
Industrial
857.2
855.4
793.0
817.5
819.8
771.6
765.9
C02
852.4
850.8
789.0
813.5
815.9
767.9
762.4
ch4
2.0
1.9
1.7
1.7
1.7
1.6
1.6
n2o
2.7
2.6
2.3
2.2
2.2
2.1
2.0
Residential
345.4
364.2
298.3
344.2
347.6
318.3
315.4
C02
338.6
358.9
293.4
338.2
341.4
313.2
310.1
ch4
5.9
4.5
4.2
5.1
5.3
4.4
4.6
n2o
0.9
0.8
0.7
0.8
0.8
0.7
0.7
Commercial
229.8
228.6
233.6
247.5
252.4
230.1
225.4
C02
228.3
227.1
232.0
245.8
250.7
228.5
223.9
ch4
1.2
1.2
1.3
1.4
1.4
1.3
1.3
n2o
0.3
0.3
0.3
0.3
0.3
0.3
0.3
U.S. Territories3
20.1
52.1
26.0
26.0
24.9
23.3
23.1
Total
4,805.7
5,828.0
4,907.9
5,045.0
4,907.3
4,392.9
4,701.7
a U.S. Territories are not apportioned by sector, and emissions shown in the table are total greenhouse gas emissions from
all fuel combustion sources.
Note: Totals may not sum due to independent rounding.
6 Other than greenhouse gases CO2, Cm, and N2O, gases emitted from stationary combustion include the
7 greenhouse gas precursors nitrogen oxides (NOx), CO, NMVOCs, and SO2. Methane and N2O emissions from
8 stationary combustion sources depend upon fuel characteristics, size and vintage of combustion device, along with
9 combustion technology, pollution control equipment, ambient environmental conditions, and operation and
10 maintenance practices. Nitrous oxide emissions from stationary combustion are closely related to air-fuel mixes
11 and combustion temperatures, as well as the characteristics of any pollution control equipment that is employed.
Energy 3-15
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Methane emissions from stationary combustion are primarily a function of the Cm content of the fuel and
combustion efficiency.
Mobile combustion also produces emissions of CFU, N2O, and greenhouse gas precursors including NOx, CO, and
NMVOCs. As with stationary combustion, N2O and NOx emissions from mobile combustion are closely related to
fuel characteristics, air-fuel mixes, combustion temperatures, and the use of pollution control equipment. Nitrous
oxide from mobile sources, in particular, can be formed by the catalytic processes used to control NOx, CO, and
hydrocarbon emissions. Carbon monoxide emissions from mobile combustion are significantly affected by
combustion efficiency and the presence of post-combustion emission controls. Carbon monoxide emissions are
highest when air-fuel mixtures have less oxygen than required for complete combustion. These emissions occur
especially in vehicle idle, low speed, and cold start conditions. Methane and NMVOC emissions from motor
vehicles are a function of the CH4 content of the motor fuel, the amount of hydrocarbons passing uncombusted
through the engine, and any post-combustion control of hydrocarbon emissions (such as catalytic converters).
An alternative method of presenting combustion emissions is to allocate emissions associated with electric power
to the sectors in which it is used. Four end-use sectors are defined: transportation, industrial, residential, and
commercial. In Table 3-11 below, electric power emissions have been distributed to each end-use sector based
upon the sector's share of national electricity use, with the exception of CH4 and N2O from transportation
electricity use.12 Emissions from U.S. Territories are also calculated separately due to a lack of end-use-specific
consumption data.13 This method assumes that emissions from combustion sources are distributed across the four
end-use sectors based on the ratio of electricity use in that sector. The results of this alternative method are
presented in Table 3-11.
Table 3-11: CO2, ChU, and N2O Emissions from Fossil Fuel Combustion by End-Use Sector
with Electricity Emissions Distributed (MMT CO2 Eq.)
End-Use Sector
1990
2005
2017
2018
2019
2020
2021
Transportation
1,517.6
1,904.7
1,805.9
1,838.1
1,840.6
1,595.3
1,814.2
C02
1,472.0
1,863.3
1,784.4
1,817.7
1,818.7
1,576.6
1,794.5
ch4
7.2
4.4
2.9
2.9
2.9
2.6
2.6
n2o
38.4
37.0
18.5
17.5
19.0
16.1
17.1
Industrial
1,550.7
1,600.2
1,304.2
1,325.5
1,291.1
1,186.8
1,212.3
C02
1,538.8
1,587.1
1,293.4
1,314.9
1,281.4
1,177.7
1,202.8
ch4
2.2
2.2
2.1
2.1
2.1
2.0
2.0
n2o
9.6
10.8
8.7
8.5
7.6
7.1
7.5
Residential
944.2
1,230.1
923.8
994.9
938.6
870.8
900.3
C02
931.3
1,214.9
910.5
980.5
925.1
858.5
887.3
ch4
6.0
4.9
4.7
5.6
5.8
4.9
5.1
n2o
6.9
10.3
8.5
00
00
7.7
7.4
7.9
Commercial
773.1
1,040.9
848.0
860.5
812.0
716.8
751.8
C02
766.0
1,030.1
838.2
850.9
803.4
708.8
743.3
ch4
1.3
1.5
1.8
1.8
1.9
1.8
1.8
n2o
5.7
9.3
8.0
7.8
6.8
6.2
6.7
U.S. Territories3
20.1
52.1
26.0
26.0
24.9
23.3
23.1
Total
4,805.7
5,828.0
4,907.9
5,045.0
4,907.3
4,392.9
4,701.7
a U.S. Territories are not apportioned by sector, and emissions are total greenhouse gas emissions from all fuel
combustion sources.
Notes: Totals may not sum due to independent rounding. Emissions from fossil fuel combustion by electric power are
allocated based on aggregate national electricity use by each end-use sector.
12 Separate calculations are performed for transportation-related CH4 and N20. The methodology used to calculate these
emissions is discussed in the Mobile Combustion section.
13 U.S. Territories (including American Samoa, Guam, Puerto Rico, U.S. Virgin Islands, Wake Island, and other outlying U.S.
Pacific Islands) consumption data obtained from EIA are only available at the aggregate level and cannot be broken out by end-
use sector. The distribution of emissions to each end-use sector for the 50 states does not apply to territories data.
3-16 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Electric Power Sector
The process of generating electricity is the largest stationary source of CO2 emissions in the United States,
representing 28.5 percent of total CO2 emissions from all CO2 emissions sources across the United States. Methane
and N2O accounted for a small portion of total greenhouse gas emissions from electric power, representing 0.1
percent and 1.2 percent, respectively. Electric power also accounted for 33.2 percent of CO2 emissions from fossil
fuel combustion in 2021. Methane and N2O from electric power represented 12.2 and 48.6 percent of total CH4
and N2O emissions from fossil fuel combustion in 2021, respectively.
For the underlying energy data used in this chapter, the Energy Information Administration (EIA) places electric
power generation into three functional categories: the electric power sector, the commercial sector, and the
industrial sector. The energy use and emissions associated with the electric power sector are included here. The
electric power sector consists of electric utilities and independent power producers whose primary business is the
production of electricity. This includes both regulated utilities and non-utilities (e.g., independent power
producers, qualifying co-generators, and other small power producers). Energy use and emissions associated with
electric generation in the commercial and industrial sectors is reported in those other sectors where the producer
of the power indicates that its primary business is something other than the production of electricity.14
Total greenhouse gas emissions from the electric power sector have decreased by 15.0 percent since 1990. From
1990 to 2007, electric power sector emissions increased by 33 percent, driven by a significant increase in electricity
demand (39 percent) while the carbon intensity of electricity generated showed a modest decline (3.2 percent).
From 2008 to 2021, as electricity demand increased by 1.6 percent, electric power sector emissions decreased by
35 percent, driven by a significant drop (22 percent) in the carbon intensity of electricity generated. Overall, the
carbon intensity of the electric power sector, in terms of CO2 Eq. per QBtu, decreased by 25 percent from 1990 to
2020 with additional trends detailed in Box 3-4. This decoupling of electric power generation and the resulting CO2
emissions is shown in Figure 3-9. This recent decarbonization of the electric power sector is a result of several key
drivers.
Coal-fired electric generation (in kilowatt-hours [kWh]) decreased from 54 percent of generation in 1990 to 23
percent in 2021.15 This corresponded with an increase in natural gas generation and renewable energy generation,
largely from wind and solar energy. Natural gas generation (in kWh) represented 11 percent of electric power
generation in 1990 and increased over the 32-year period to represent 37 percent of electric power sector
generation in 2021 (see Table 3-12). Natural gas has a much lower carbon content than coal and is generated in
power plants that are generally more efficient in terms of kWh produced per Btu of fuel combusted, which has led
to lower emissions as natural gas replaces coal-powered electricity generation. Natural gas and coal used in the
United States in 2021 had an average carbon content of 14.43 MMT C/QBtu and 26.13 MMT C/QBtu respectively.
Table 3-12: Electric Power Generation by Fuel Type (Percent)
Fuel Type
1990
2005
2017
2018
2019
2020
2021
Coal
54.1%
51.1%
30.9%
28.4%
24.2%
19.9%
22.5%
Natural Gas
10.7%
17.5%
30.9%
34.0%
37.3%
39.5%
37.2%
Nuclear
19.9%
20.0%
20.8%
20.1%
20.4%
20.5%
19.6%
Renewables
11.3%
8.3%
16.8%
16.8%
17.6%
19.5%
20.1%
Petroleum
4.1%
3.0%
0.5%
0.6%
0.4%
0.4%
0.4%
Other Gases3
0.0%
0.1%
0.1%
0.1%
0.1%
0.1%
0.1%
Net Electricity Generation
(Billion kWh)b
2,905
3,902
3,878
4,020
3,966
3,851
3,961
+ Does not exceed 0.05 percent.
14 Utilities primarily generate power for the U.S. electric grid for sale to retail customers. Non-utilities typically generate
electricity for sale on the wholesale electricity market (e.g., to utilities for distribution and resale to retail customers). Where
electricity generation occurs outside the ElA-defined electric power sector, it is typically for the entity's own use.
15 Values represent electricity net generation from the electric power sector (EIA 2022a).
Energy 3-17
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
a Other gases include blast furnace gas, propane gas, and other manufactured and waste gases derived from fossil
fuels.
b Represents net electricity generation from the electric power sector. Excludes net electricity generation from
commercial and industrial combined-heat-and-power and electricity-only plants. Does not include electricity
generation from purchased steam as the fuel used to generate the steam cannot be determined.
In 2021, CO2 emissions from the electric power sector increased by 7.1 percent relative to 2020. This increase in
CO2 emissions was primarily driven by an increase in coal consumed to produce electricity in the electric power
sector. Consumption of coal for electric power increased by 15.4 percent while consumption of natural gas
decreased 3.1 percent from 2020 to 2021, leading to an overall increase in emissions. There has also been a rapid
increase in renewable energy electricity generation in the electric power sector in recent years. Electricity
generation from renewable sources increased by 6 percent from 2020 to 2021 (see Table 3-12). A decrease in coal-
powered electricity generation and increase in natural gas and renewable energy electricity generation
contributed to a decoupling of emissions trends from electric power generation trends over the recent time series
(see Figure 3-9).
Decreases in natural gas prices and the associated increase in natural gas generation, particularly between 2005
and 2021, was one of the main drivers of the recent fuel switching and decrease in electric power sector carbon
intensity. During this time period, the cost of natural gas (in $/MMBtu) decreased by 25 percent while the cost of
coal (in $/MMBtu) increased by 71 percent (EIA 2021c). Also, between 1990 and 2021, renewable energy
generation (in kWh) from wind and solar energy increased from 0.1 percent of total generation in 1990 to 12
percent in 2021, which also helped drive the decrease in electric power sector carbon intensity. This decrease in
carbon intensity occurred even as total electricity retail sales increased 40 percent, from 2,713 billion kWh in 1990
to 3,795 billion kWh in 2021.
Figure 3-9: Fuels Used in Electric Power Generation and Total Electric Power Sector CO2
Emissions
50,000
40,000
D
4—1
H 30,000
O)
U)
ZD
>*
I? 20,000
10,000
0
I Nuclear (TBtu)
Renewable Energy Sources (TBtu)
I Petroleum (TBtu)
Matural Gas (TBtu)
Coal (TBtu)
¦ Net Generation (Index from 1990) [Right Axis]
¦ Sector CO2 Emissions (Index from 1990) [Right Axis]
160
140
120
100
80
60
40
20
0
01
en
0000000000*-l*-H-»H*-H*-H*-l'»H»-H*-H*-«CMfN
0000000000000000000000
Electricity was used primarily in the residential, commercial, and industrial end-use sectors for lighting, heating,
electric motors, appliances, electronics, and air conditioning (see Figure 3-10). Note that transportation is an end-
use sector as well but is not shown in Figure 3-10 due to the sector's relatively low percentage of electricity use.
Table 3-13 provides a break-out of CO2 emissions from electricity use in the transportation end-use sector.
3-18 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Figure 3-10: Electric Power Retail Sales by End-Use Sector
1,600
1,500
1,400
^ 1,300
J 1,200
s 1,100
1,000
900
800
o-^HrNjmTrLnor^oocr>o->-irvjm^rLnvor^oocriO-^HfNmTrLnvors».ooa>o-^H
Ql 01" CT" O O O O O O O O O O ¦»—I I t-H i—I -i—I i—I ¦»—I i—I t—I t-H ("NJ f"s]
CTiCT^Cr>CT^CTia>CTiCriCT^Cr.0000000000000000000000
T-HT-H-^n-^H-T-HT-iT-HT-iT-iT-irMrMrNjrNjrNjrMrMrNjrMrvjrsirNjrMrNjrsjrNjrsirMrMrNjrNjrNj
In 2021, electricity sales to the residential and commercial end-use sectors, as presented in Figure 3-10, increased
by 0.8 percent and 2.9 percent relative to 2020, respectively. Electricity sales to the industrial sector in 2021
increased by approximately 2.9 percent relative to 2020. The sections below describe end-use sector energy use in
more detail. Overall, in 2021, the amount of electricity retail sales (in kWh) increased by 2.1 percent relative to
2020.
Industrial Sector
Industrial sector CO2, CFU, and N2O emissions accounted for 16,14, and 5 percent of CO2, CFU, and N2O emissions
from fossil fuel combustion, respectively in 2021. Carbon dioxide, CFU, and N2O emissions resulted from the direct
consumption of fossil fuels for steam and process heat production.
The industrial end-use sector, per the underlying energy use data from EIA, includes activities such as
manufacturing, construction, mining, and agriculture. The largest of these activities in terms of energy use is
manufacturing, of which six industries—Petroleum Refineries, Chemicals, Paper, Primary Metals, Food, and
Nonmetallic Mineral Products—represent the majority of the energy use (EIA 2021c; EIA 2009b).
There are many dynamics that impact emissions from the industrial sector including economic activity, changes in
the make-up of the industrial sector, changes in the emissions intensity of industrial processes, and weather-
related impacts on heating and cooling of industrial buildings.16 Structural changes within the U.S. economy that
lead to shifts in industrial output away from energy-intensive manufacturing products to less energy-intensive
products (e.g., from steel to computer equipment) have had a significant effect on industrial emissions.
From 2020 to 2021, total industrial production and manufacturing output increased by 4.9 percent (FRB 2022).
Over this period, output increased slightly across production indices for Food, Nonmetallic Mineral Products,
Paper, Petroleum Refineries, and Primary Metals. Production of chemicals declined slightly between 2020 and
2021 (see Figure 3-11). From 2020 to 2021, energy use from fossil fuels in the industrial sector decreased by less
than half a percent. Total energy use in the industrial sector increased by 0.7 percent, driven mainly by a 2.9
percent increase in the consumption of renewables. Due to the relative increases and decreases of individual
indices there was an increase in natural gas and an increase in electricity used by the sector (see Figure 3-12). In
16
Some commercial customers are large enough to obtain an industrial price for natural gas and/or electricity and are
consequently grouped with the industrial end-use sector in U.S. energy statistics. These misclassifications of large commercial
customers likely cause the industrial end-use sector to appear to be more sensitive to weather conditions.
Energy 3-19
-------
1 2021, CCh, CH4, and N:0 emissions from fossil fuel combustion and electricity use within the industrial end-use
2 sector totaled 1,212.3 MMT CO2 Eq., a 2.1 percent increase from 2020 emissions.
3 Through EPA's Greenhouse Gas Reporting Program (GHGRP), specific industrial sector trends can be discerned
4 from the overall total EIA industrial fuel consumption data used for these calculations. For example, from 2020 to
5 2021, the underlying EIA data showed increased consumption of coal and natural gas in the industrial sector. The
6 GHGRP data highlights that several industries contributed to these trends, including chemical manufacturing; pulp,
7 paper and print; food processing, beverages and tobacco; minerals manufacturing; and agriculture-forest-
8 fisheries.17
9 Figure 3-11: Industrial Production Indices (Index 2017=100)
140
120
100
80
60
Paper
17 Further details on industrial sector combustion emissions are provided by EPA's GHGRP. See
http://ghedata.epa.gov/ghgp/main.do.
3-20 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Figure 3-12: Fuels and Electricity Used in Industrial Sector, Industrial Output, and Total
2 Sector CO2 Emissions (Including Electricity)
35,000
30,000
25,000
s
~ 20,000
ID
>.
o3 15,000
S
10,000
5,000
0
3
4 Despite the growth in industrial output (60 percent) and the overall U.S. economy (109 percent) from 1990 to
5 2021, direct CO2 emissions from fossil fuel combustion in the industrial sector decreased by 10.6 percent over the
6 same time series. A number of factors are assumed to result in decoupling of growth in industrial output from
7 industrial greenhouse gas emissions, for example: (1) more rapid growth in output from less energy-intensive
8 industries relative to traditional manufacturing industries, and (2) energy-intensive industries such as steel are
9 employing new methods, such as electric arc furnaces, that are less carbon intensive than the older methods.
10
11
Oi-HrMm^rmvor^cooiOi-HrMm^rLnvor^cocTiOi-HrMfn^rLnvor^cocriOi-i
CTiCTiO'iCriCTiCT'iO'iO'iCriCTiOOOOOOOOOO-'—t-i—1 1—1 i—1 -i—It—It—iPMfM
cria^cr
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
that are submitted to the UNFCCC along with this report.19 The efforts in reconciling fuels focus on standard,
common fuel types (e.g., natural gas, distillate fuel oil) where the fuels in ElA's national statistics aligned well
with facility-level GHGRP data. For these reasons, the current information presented in the Common Reporting
Format (CRF) tables should be viewed as an initial attempt at this exercise. Additional efforts will be made for
future Inventory reports to improve the mapping of fuel types and examine ways to reconcile and coordinate
any differences between facility-level data and national statistics. The current analysis includes the full time
series presented in the CRF tables. Analyses were conducted linking GHGRP facility-level reporting with the
information published by EIA in its MECS data in order to disaggregate the full 1990 through 2021 time period in
the CRF tables. It is believed that the current analysis has led to improvements in the presentation of data in the
Inventory, but further work will be conducted, and future improvements will be realized in subsequent
Inventory reports. This includes incorporating the latest MECS data as it becomes available.
Residential and Commercial Sectors
Emissions from the residential and commercial sectors have generally decreased since 2005. Short-term trends are
often correlated with seasonal fluctuations in energy use caused by weather conditions, rather than prevailing
economic conditions. Population growth and a trend towards larger houses has led to increasing energy use over
the time series, while population migration to warmer areas and improved energy efficiency and building
insulation have slowed the increase in energy use in recent years. Starting in around 2014, energy use and
emissions begin to decouple due to decarbonization of the electric power sector (see Figure 3-13).
Figure 3-13: Fuels and Electricity Used in Residential and Commercial Sectors, Heating and
Cooling Degree Days, and Total Sector CO2 Emissions (Including Electricity)
>.
pi
1
25,000
20,000
15,000
10,000
5,000
Coal (TBtu)
Renewable Energy Sources (TBtu)
I Petroleum (TBtu)
Natural Gas
I Electricity Use (TBtu)
¦ Sector CO2 Emissions (Index vs. 1990) [Right Axis]
¦ Heating and Cooling Degree Days (Index vs. 1990) [Right Axis]
180
160
140
120
100
80
60
40
20
0
0
-
01
>
X
oj
—
HfMfOt^'ONW^OHiNnfin^NODO'iOHiNfOt^'DNCOffiOH
0^0"*CF>0"\0"*0^0"\0^0"\0000 OOOOO 0->—t-Ht—(t-Ht—It—< t-h t—1 oJ fN
ChO^O^C^O^O^CnChOOOOOOOOOOOOOOOOOOOOOOO
*-H t-h t-h HHHHH(N(M(N(N(N(M(NMN(N(N(N(M(N(NNNNOJ(NN(N
In 2021 the residential and commercial sectors accounted for 7 and 5 percent of CO2 emissions from fossil fuel
combustion, respectively; 40 and 11 percent of CH4 emissions from fossil fuel combustion, respectively; and 2 and
1 percent of N2O emissions from fossil fuel combustion, respectively. Emissions from these sectors were largely
due to the direct consumption of natural gas and petroleum products, primarily for heating and cooking needs.
Coal consumption was a minor component of energy use in the commercial sector and did not contribute to any
energy use in the residential sector. In 2021, total emissions (CO2, CH4, and N2O) from fossil fuel combustion and
19 See https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
3-22 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 electricity use within the residential and commercial end-use sectors were 900.3 MMT CO2 Eq. and 751.8 MMT CO2
2 Eq., respectively. Total CO2, CH4, and N2O emissions from combined fossil fuel combustion and electricity use
3 within the residential and commercial end-use sectors increased by 3.4 and 4.9 percent from 2020 to 2021,
4 respectively. An increase in heating degree days (0.5 percent) increased energy demand for heating in the
5 residential and commercial sectors. This was partially offset by a 1.9 percent decrease in cooling degree days
6 compared to 2020, which impacted demand for air conditioning in the residential and commercial sectors. This
7 resulted in a 0.8 percent increase in residential sector electricity use. From 2020 to 2021 there was a 0.7 percent
8 lower direct energy use in the commercial sector. In addition, a shift toward energy efficient products and more
9 stringent energy efficiency standards for household equipment has contributed to a decrease in energy demand in
10 households (EIA 2022g), resulting in a decrease in energy-related emissions. In the long term, the residential sector
11 is also affected by population growth, migration trends toward warmer areas, and changes in total housing units
12 and building attributes (e.g., larger sizes and improved insulation).
13 In 2021, combustion emissions from natural gas consumption represented 83 and 81 percent of the direct fossil
14 fuel CO2 emissions from the residential and commercial sectors, respectively. Carbon dioxide emissions from
15 natural gas combustion in the residential and commercial sectors in 2021 increased by 0.9 percent and increased
16 by 4.0 percent from 2020 to 2021, respectively.
17 U.S. Territories
18 Emissions from U.S. Territories are based on the fuel consumption in American Samoa, Guam, Puerto Rico, U.S.
19 Virgin Islands, Wake Island, and other outlying U.S. Pacific Islands. As described in the Methodology section of CO2
20 from Fossil Fuel Combustion, this data is collected separately from the sectoral-level data available for the general
21 calculations. As sectoral information is not available for U.S. Territories, CO2, CFU, and N2O emissions are not
22 presented for U.S. Territories in the tables above by sector, though the emissions will occur across all sectors and
23 sources including stationary, transportation and mobile combustion sources. Due to data availability limitations,
24 2021 and 2020 energy consumption for U.S. Territories for petroleum is proxied to 2019 consumption data.
25 Transportation Sector and Mobile Combustion
26 This discussion of transportation emissions follows the alternative method of presenting combustion emissions by
27 allocating emissions associated with electricity generation to the transportation end-use sector, as presented in
28 Table 3-11. Table 3-10 presents direct CO2, CFU, and N2O emissions from all transportation sources (i.e., excluding
29 emissions allocated to electricity consumption in the transportation end-use sector).
30 The transportation end-use sector and other mobile combustion accounted for 1,814 MMT CO2 Eq. in 2021, which
31 represented 37 percent of CO2 emissions, 26 percent of CFU emissions, and 39 percent of N2O emissions from fossil
32 fuel combustion, respectively.20 Fuel purchased in the United States for international aircraft and marine travel
33 accounted for an additional 69.9 MMT CO2 Eq. in 2021;21 these emissions are recorded as international bunkers
34 and are not included in U.S. totals according to UNFCCC reporting protocols.
35 Transportation End-Use Sector
36 From 1990 to 2019, transportation emissions from fossil fuel combustion rose by 21 percent, followed by a 13
37 percent reduction from 2019 to 2020. Overall, from 1990 to 2021, transportation emissions from fossil fuel
38 combustion increased by 20 percent. The increase in transportation emissions from fossil fuel combustion from
39 1990 to 2021 was due, in large part, to increased demand for travel (see Figure 3-14). The number of vehicle miles
40 traveled (VMT) by light-duty motor vehicles (passenger cars and light-duty trucks) increased 48 percent from 1990
20 Note that these totals include C02, CH4 and N20 emissions from some sources in the U.S. Territories (ships and boats,
recreational boats, non-transportation mobile sources) and CH4 and N20 emissions from transportation rail electricity.
21 Some bunker fuels data are not yet available and has been proxied for 2021. This value will be updated for the Final Report
published in April 2023.
Energy 3-23
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
to 2021,22 as a result of a confluence of factors including population growth, economic growth, urban sprawl, and
periods of low fuel prices. Between 2019 and 2020, emissions from light-duty vehicles fell by 11 percent, primarily
the result of the COVID-19 pandemic and associated restrictions, such as people working from home and traveling
less. Light-duty vehicle VMT rebounded in 2021 but is still estimated to be 1 percent below 2019 levels.23
Emissions from commercial aircraft for 2021 will be estimated in the Final Report published in April 2023, which
will incorporate the latest data from FAA and other data sources. Here, commercial aircraft emissions are proxied
to remain the same between 2020 and 2021. Commercial aircraft emissions have decreased 35 percent since 2007
(FAA 2022 and DOT 1991 through 2021).24 Decreases in jet fuel emissions (excluding bunkers) started in 2007 due
in part to improved operational efficiency that results in more direct flight routing, improvements in aircraft and
engine technologies to reduce fuel burn and emissions, and the accelerated retirement of older, less fuel-efficient
aircraft; however, the sharp decline in commercial aircraft emissions from 2019 to 2020 is primarily due to COVID-
19 impacts on scheduled passenger air travel.
Almost all of the energy consumed for transportation was supplied by petroleum-based products, with more than
half being related to gasoline consumption in automobiles and other highway vehicles. Other fuel uses, especially
diesel fuel for freight trucks and jet fuel for aircraft, accounted for the remainder. The primary driver of
transportation-related emissions was CO2 from fossil fuel combustion, which increased by 22 percent from 1990 to
2021. Annex 3.2 presents the total emissions from all transportation and mobile sources, including CO2, N2O, CFU,
and HFCs.
22 VMT estimates are based on data from FHWA Highway Statistics Table VM-1 (FHWA 1996 through 2020). VMT estimates
from FHWA are allocated to vehicle type using ratios of VMT per vehicle type to total VMT, derived from EPA's MOVES3 model
(see Annex 3.2 for information about the MOVES model). Data for 2021 has been proxied using FHWA Traffic Volume Trends.
23 2021 VMT is estimated based on FHWA Traffic Volume Trends data and will be updated when the 2021 data are released by
FHWA.
24 Commercial aircraft consists of passenger aircraft, cargo, and other chartered flights.
3-24 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Figure 3-14: Fuels Used in Transportation Sector, On-road VMT, and Total Sector CO2
Emissions
40,000
35,000
30,000
25,000
20,000
15,000
10,000
5,000
I Other Fuels (TBtu)
I Residual Fuel (TBtu)
Natural Gas (TBtu)
Renewable Energy (TBtu)
Jet Fuel (TBtu)
I Distillate Fuel (TBtu)
I Motor Gasoline (TBtu)
Onroad VMT (Index vs. 1990) [Right Axis]
¦ Sector CO2 Emissions (Index vs. 1990) [Right Axis]
200
180
160
140
120
100
80
60
40
20
0
iniDNfflOiOHMr
o O O O O t—i H t
o-\c->cr>o-vcr*cr\crvcr>avavoooooooooooooc
HHHHHHHHHH{\fMrMtN(N(N(N(N(N(N(N(N(N(N(N(N(N[N(NtN(NrM
Notes: Distillate fuel, residual fuel, and jet fuel include adjustments for international bunker fuels. Distillate fuel and motor
gasoline include adjustments for the sectoral allocation of these fuels. Other Fuels includes aviation gasoline and propane.
Source: Information on fuel consumption was obtained from EIA (2022).
Transportation Fossil Fuel Combustion CO2 Emissions
Domestic transportation CO2 emissions increased by 22 percent (323 MMT CO2) between 1990 and 2021, an
annualized increase of 0.7 percent. This includes a 24 percent increase in CO2 emissions between 1990 and 2019,
followed by a 13 percent decrease in 2020. Carbon dioxide emissions then increased by 14 percent between 2020
and 2021. Among domestic transportation sources in 2021, light-duty vehicles (including passenger cars and light-
duty trucks) represented 57 percent of CO2 emissions from fossil fuel combustion, medium- and heavy-duty trucks
and buses 25 percent, commercial aircraft 5 percent, and other sources 13 percent. See Table 3-13 for a detailed
breakdown of transportation CO2 emissions by mode and fuel type.
Almost all of the energy consumed by the transportation sector is petroleum-based, including motor gasoline,
diesel fuel, jet fuel, and residual oil. Carbon dioxide emissions from the combustion of ethanol and biodiesel for
transportation purposes, along with the emissions associated with the agricultural and industrial processes
involved in the production of biofuel, are captured in other Inventory sectors.25 Ethanol consumption by the
transportation sector has increased from 0.7 billion gallons in 1990 to 13.2 billion gallons in 2021, while biodiesel
consumption has increased from 0.01 billion gallons in 2001 to 1.7 billion gallons in 2021. For additional
information, see Section 3.10 on biofuel consumption at the end of this chapter and Table A-76 in Annex 3.2.
25 Biofuel estimates are presented in the Energy chapter for informational purposes only, in line with IPCC methodological
guidance and UNFCCC reporting obligations. Net carbon fluxes from changes in biogenic carbon reservoirs in croplands are
accounted for in the estimates for Land Use, Land-Use Change, and Forestry (see Chapter 6). More information and additional
analyses on biofuels are available at EPA's Renewable Fuels Standards website. See https://www.epa.gov/renewable-fuel-
standard-program.
Energy 3-25
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Carbon dioxide emissions from passenger cars and light-duty trucks totaled 1,031.4 MMT CO2 in 2021, an increase
of 13 percent (117 MMT CO2) from 1990 to 2021. The increase in CO2 emissions from passenger cars and light-duty
trucks from 1990 to 2021 was due, in large part, to increased demand for travel as fleet-wide light-duty vehicle fuel
economy was relatively stable (average new vehicle fuel economy declined slowly from 1990 through 2004 and
then increased more rapidly from 2005 through 2021). Carbon dioxide emissions from passenger cars and light-
duty trucks peaked at 1,145.7 MMT CO2 in 2004, and since then have declined about 10 percent. The decline in
new light-duty vehicle fuel economy between 1990 and 2004 (Figure 3-15) reflects the increasing market share of
light-duty trucks, which grew from about 30 percent of new vehicle sales in 1990 to 48 percent in 2004. Starting in
2005, average new vehicle fuel economy began to increase while light-duty vehicle VMT grew only modestly for
much of the period. Light-duty vehicle VMT grew by less than one percent or declined each year between 2005
and 2013, and again between 2017 and 2019.26 VMT grew at faster rates of 2.6 percent from 2014 to 2015 and 2.5
percent from 2015 to2016. From 2019 to 2020, light-duty vehicle VMT declined by 11 percent due to the COVID-19
pandemic; from 2020 to 2021 light-duty vehicle VMT rebounded, increasing by 11.2 percent.
Average new vehicle fuel economy has increased almost every year since 2005, while the light-duty truck share of
new vehicle sales decreased to about 33 percent in 2009 and has since varied from year to year between 36 and 61
percent. Since 2014, the light-duty truck share has steadily increased, reaching 61 percent of new vehicles sales in
model year 2021 (EPA 2022b). See Annex 3.2 for data by vehicle mode and information on VMT and the share of
new vehicles (in VMT).
Medium- and heavy-duty truck CO2 emissions increased by 81 percent from 1990 to 2021. This increase was largely
due to a substantial growth in medium- and heavy-duty truck VMT, which increased by 71 percent between 1990
and 2021.
Carbon dioxide emissions from the domestic operation of commercial aircraft decreased by 17 percent (18.6 MMT
CO2) from 1990 to 2021. Across all categories of aviation, excluding international bunkers, CO2 emissions
decreased by 12 percent (21.8 MMT CO2) between 1990 and 2021.27 Emissions from military aircraft decreased 68
percent between 1990 and 2021. Commercial aircraft emissions increased 27 percent between 1990 and 2007,
dropped 4 percent from 2007 to 2019, and then dropped 32 percent from 2019 to 2020, a change of
approximately 17 percent between 1990 and 2021. Commercial aircraft emissions are proxied to remain the same
between 2020 and 2021 and will be updated in the Final Report published in April 2023.
Transportation sources also produce CH4 and N2O; these emissions are included in Figure 3-14 and Table 3-15 and
in the CFU and N2O from Mobile Combustion section. Annex 3.2 presents total emissions from all transportation
and mobile sources, including CO2, CFU, N2O, and HFCs.
26 VMT estimates are based on data from FHWA Highway Statistics Table VM-1 (FHWA 1996 through 2020). VMT estimates
from FHWA are allocated to vehicle type using ratios of VMT per vehicle type to total VMT, derived from EPA's MOVES3 model
(see Annex 3.2 for information about the MOVES model). Data for 2021 has been proxied using FHWA Traffic Volume Trends.
27 Includes consumption of jet fuel and aviation gasoline. Does not include aircraft bunkers, which are not included in national
emission totals, in line with IPCC methodological guidance and UNFCCC reporting obligations.
3-26 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Figure 3-15: Sales-Weighted Fuel Economy of New Passenger Cars and Light-Duty Trucks,
2 1990-2021
3
4 Source: EPA (2022a).
5
6 Figure 3-16: Sales of New Passenger Cars and Light-Duty Trucks, 1990-2021
7
8 Source: EPA (2022b).
9
10 Table 3-13: CO2 Emissions from Fossil Fuel Combustion in Transportation End-Use Sector
11 (MMT CO2 Eq.)
Fuel/Vehicle Type
1990
2005
2017
2018
2019
2020
2021
Gasoline3
958,9
1,150.1
1,081.8
1,097.0
1,086.5
937.0
1,040.3
Passenger Cars
612.8
518.9
375.2
382.5
380.0
328.0
364.6
Energy 3-27
-------
Light-Duty Trucks
283.6
583.4
661.5
667.6
658.6
565.8
627.0
Medium- and Heavy-Duty
Trucks'5
42.8
28.1
24.9
26.2
27.0
24.1
un
Buses
2.1
1.1
2.5
2.7
2.8
2.5
2.9
Motorcycles
3.4
4.9
7.0
7.3
7.4
6.6
7.5
Recreational Boatsc
14.3
13.7
10.6
10.7
10.7
10.1
10.6
Distillate Fuel Oil (Diesel)a
274.6
472.1
474.9
486.6
484.1
455.0
502.2
Passenger Cars
9.4
2.2
3.0
2.8
2.7
2.5
2.8
Light-Duty Trucks
8.4
30.4
31.1
31.2
31.2
30.2
34.4
Medium- and Heavy-Duty
Trucks'5
189.0
357.2
362.0
371.5
373.0
353.4
392.6
Buses
11.1
15.5
19.7
20.4
20.7
19.8
22.1
Rail
35.5
46.1
37.4
38.5
36.0
31.0
32.1
Recreational Boatsc
2.7
2.9
2.8
2.8
2.9
2.7
2.8
Ships and Non-Recreational
Boatsc
6.8
8.4
10.0
9.3
7.5
7.4
in
International Bunker Fuelse
11.7
9.5
9.0
10.0
10.1
7.8
1A
Jet Fuel
222.3
249.5
249.4
253.1
258.5
160.4
203.8
Commercial Aircraft'
109.9
132.7
128.0
129.6
134.2
91.3
91.3
Military Aircraft
35.7
19.8
12.5
12.1
12.1
10.7
11.3
General Aviation Aircraft
38.5
36.8
31.2
30.6
31.4
18.6
61.3
International Bunker Fuelse
38.2
60.2
77.8
80.9
80.8
39.8
39.9
International Bunker Fuels
from Commercial Aviation
30.0
55.6
74.5
77.7
77.6
36.7
36.7
Aviation Gasoline
3.1
2.4
1.4
1.5
1.6
1.4
1.5
General Aviation Aircraft
3.1
2.4
1.4
1.5
1.6
1.4
1.5
Residual Fuel Oil
76.3
62.9
49.9
45.4
39.7
29.4
45.5
Ships and Non-Recreational
Boatse
22.6
19.3
16.5
14.0
14.5
7.3
23.6
International Bunker Fuelse
53.7
43.6
33.4
31.4
25.2
22.1
21.9
Natural Gas'
36.0
33.1
42.3
50.9
58.9
58.7
65.1
Passenger Cars
+
+
+
+
+
+
+
Light-Duty Trucks
+
+
+
+
+
+
+
Medium- and Heavy-Duty
Trucks
+
0.1
0.1
0.1
0.1
0.1
0.1
Buses
+
0.3
0.5
0.6
0.6
0.6
0.7
Pipeline5
36.0
32.6
41.6
50.2
58.2
57.9
64.2
LPG'
1.4
1.8
0.6
0.6
0.5
0.3
0.3
Passenger Cars
+
+
+
+
+
+
+
Light-Duty Trucks
0.2
0.3
+
0.1
0.1
+
0.1
Medium- and Heavy-Duty
Trucks'5
1.1
1.1
0.4
0.4
0.4
0.2
0.2
Buses
0.2
0.3
0.1
0.1
0.1
+
+
Electricity'
3.0
4.7
4.3
4.8
4.8
4.1
5.1
Passenger Cars
+
+
0.8
1.2
1.4
1.3
1.8
Light-Duty Trucks
+
+
0.1
0.2
0.2
0.3
0.7
Buses
+
+
+
+
+
0.1
0.1
Rail
3.0
4.7
3.4
3.4
3.1
2.4
2.5
Total (Excluding Bunkers)8
1,472.0
1,863.3
1,784.4
1,817.7
1,818.7
1,576.6
1,794.5
Total (Including Bunkers)'
1,575.6
1976.6
1,904.6
1,939.8
1,934.8
1,646.3
1,863.7
Biofuels-Ethanolh
4.1
21.6
77.7
78.6
78.7
68.1
76.3
Biofuels-Biodieselh
+
0.9
18.7
17.9
17.1
17.7
16.1
+ Does not exceed 0.05 MMT C02 Eq.
3-28 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
a On-road fuel consumption data from FHWA Table MF-21 and MF-27 were used to determine total on-road use of motor
gasoline and diesel fuel (FHWA 1996 through 2020). Data for 2021 is proxied using FHWA Traffic Volume Travel Trends.
Ratios developed from MOVES3 output are used to apportion FHWA fuel consumption data to vehicle type and fuel type
(see Annex 3.2 for information about the MOVES model).
b Includes medium- and heavy-duty trucks over 8,500 lbs.
c In 2014, EPA incorporated the NONROAD2008 model into the MOVES model framework. The current Inventory uses the
Nonroad component of MOVES3 for years 1999 through 2021. See Annex 3.2 for information about the MOVES model.
d Note that large year over year fluctuations in emission estimates partially reflect nature of data collection for these
sources.
e Official estimates exclude emissions from the combustion of both aviation and marine international bunker fuels;
however, estimates including international bunker fuel-related emissions are presented for informational purposes.
f Commercial aircraft, as modeled in FAA's Aviation Environmental Design Tool (AEDT), consists of passenger aircraft,
cargo, and other chartered flights.
g Pipelines reflect C02 emissions from natural gas-powered pipelines transporting natural gas.
h Ethanol and biodiesel estimates are presented for informational purposes only. See Section 3.10 of this chapter and the
estimates in Land Use, Land-Use Change, and Forestry (see Chapter 6), in line with IPCC methodological guidance and
UNFCCC reporting obligations, for more information on ethanol and biodiesel.
' Transportation sector natural gas and LPG consumption are based on data from EIA (2021b). Prior to the 1990 to 2015
Inventory, data from DOE TEDB were used to estimate each vehicle class's share of the total natural gas and LPG
consumption. Since TEDB does not include estimates for natural gas use by medium and heavy-duty trucks or LPG use by
passenger cars, EIA Alternative Fuel Vehicle Data (Browning 2017) is now used to determine each vehicle class's share of
the total natural gas and LPG consumption. These changes were first incorporated in the 1990 to 2016 Inventory and
apply to the 1990 to 2021 time period.
j Includes emissions from rail electricity.
k Electricity consumption by passenger cars, light-duty trucks (SUVs), and buses is based on plug-in electric vehicle sales
and engine efficiency data, as outlined in Browning (2018a). In prior Inventory years, C02 emissions from electric vehicle
charging were allocated to the residential and commercial sectors. They are now allocated to the transportation sector.
These changes apply to the 2010 through 2021 time period.
Notes: This table does not include emissions from non-transportation mobile sources, such as agricultural equipment and
construction/mining equipment; it also does not include emissions associated with electricity consumption by pipelines
or lubricants used in transportation. In addition, this table does not include C02 emissions from U.S. Territories, since
these are covered in a separate chapter of the Inventory. Totals may not sum due to independent rounding.
1 Mobile Fossil Fuel Combustion CH4 and N2O Emissions
2 Mobile combustion includes emissions of CH4 and N2O from all transportation sources identified in the U.S.
3 Inventory with the exception of pipelines and electric locomotives;28 mobile sources also include non-
4 transportation sources such as construction/mining equipment, agricultural equipment, vehicles used off-road,
5 and other sources (e.g., snowmobiles, lawnmowers, etc.).29 Annex 3.2 includes a summary of all emissions from
6 both transportation and mobile sources. Table 3-14 and Table 3-15 provide mobile fossil fuel CH4 and N2O emission
7 estimates in MMT CO2 Eq.30
28 Emissions of CH4 from natural gas systems are reported separately. More information on the methodology used to calculate
these emissions are included in this chapter and Annex 3.4.
29 See the methodology sub-sections of the C02 from Fossil Fuel Combustion and CH4 and N20 from Mobile Combustion
sections of this chapter. Note that N20 and CH4 emissions are reported using different categories than C02. C02 emissions are
reported by end-use sector (Transportation, Industrial, Commercial, Residential, U.S. Territories), and generally adhere to a top-
down approach to estimating emissions. C02 emissions from non-transportation sources (e.g., lawn and garden equipment,
farm equipment, construction equipment) are allocated to their respective end-use sector (i.e., construction equipment C02
emissions are included in the Industrial end-use sector instead of the Transportation end-use sector). CH4 and N20 emissions
are reported using the "Mobile Combustion" category, which includes non-transportation mobile sources. CH4 and N20
emission estimates are bottom-up estimates, based on total activity (fuel use, VMT) and emissions factors by source and
technology type. These reporting schemes are in accordance with IPCC guidance. For informational purposes only, C02
emissions from non-transportation mobile sources are presented separately from their overall end-use sector in Annex 3.2.
30 See Annex 3.2 for a complete time series of emission estimates for 1990 through 2021.
Energy 3-29
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Mobile combustion was responsible for a small portion of national Cm emissions (0.4 percent) and was the fifth
largest source of national N2O emissions (4.5 percent). From 1990 to 2021, mobile source CFU emissions declined
by 64 percent, to 2.6 MMT CO2 Eq. (94 kt CH4), due largely to emissions control technologies employed in on-road
vehicles since the mid-1990s to reduce CO, NOx, NMVOC, and CFU emissions. Mobile source emissions of N2O
decreased by 55 percent from 1990 to 2021, to 17.1 MMT CO2 Eq. (65 kt N2O). Earlier generation emissions control
technologies initially resulted in higher N2O emissions, causing a 31 percent increase in N2O emissions from mobile
sources between 1990 and 1997. Improvements in later-generation emissions control technologies have reduced
N2O emissions, resulting in a 66 percent decrease in mobile source N2O emissions from 1997 to 2021 (Figure 3-17).
Overall, CFU and N2O emissions were predominantly from gasoline-fueled passenger cars and light-duty trucks and
non-highway sources. See Annex 3.2 for data by vehicle mode and information on VMT and the share of new
vehicles.
Figure 3-17: Mobile Source ChU and N2O Emissions
50
40
o
u
30
20
10
N2O
ChU
CTi CTi CTi CTi CTi
0s! CTi CTi CTi CTi
LD vO
CT-i Ch
Ch
CTi CTi CTi
CTi CT< d
(MrMrMfM(M(NNfMrM(N(NfMfM(NfMfMrM(NrM(N(MrM
Table 3-14: ChU Emissions from Mobile Combustion (MMT CO2 Eq.)
Fuel Type/Vehicle Type3
1990
2005
2017
2018
2019
2020
2021
Gasoline On-Roadb
5.8
2.4
1.0
0.9
1.0
0.8
0.8
Passenger Cars
3.8
1.2
0.3
0.3
0.3
0.2
0.3
Light-Duty Trucks
1.5
1.0
0.6
0.5
0.6
0.5
0.5
Medium- and Heavy-Duty Trucks
and Buses
0.5
0.1
+
+
+
+
+
Motorcycles
+
+
+
+
+
+
+
Diesel On-Roadb
+
+
0.1
0.1
0.1
0.1
0.1
Passenger Cars
+
+
+
+
+
+
+
Light-Duty Trucks
+
+
+
+
+
+
+
Medium- and Heavy-Duty Trucks
+
+
0.1
0.1
0.1
0.1
0.1
Medium- and Heavy-Duty Buses
+
+
+
+
+
+
+
Alternative Fuel On-Road
+
0.2
0.1
0.1
0.1
0.1
0.1
Non-Roadc
1.4
1.8
1.7
1.7
1.7
1.6
1.6
Ships and Boats
0.4
0.5
0.5
0.5
0.4
0.4
0.5
Raild
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Aircraft
0.1
0.1
+
+
+
+
+
Agricultural Equipment6
0.2
0.2
0.1
0.1
0.1
0.1
0.1
Construction/Mining Equipment'
0.2
0.3
0.2
0.2
0.2
0.2
0.2
Others
0.5
0.7
0.8
0.8
0.8
0.8
0.7
Total
7.2
4.4
2.9
2.9
2.9
2.6
2.6
+ Does not exceed 0.05 MMT C02 Eq.
a See Annex 3.2 for definitions of on-road vehicle types.
3-30 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
b Gasoline and diesel highway vehicle mileage estimates are based on data from FHWA Highway Statistics Table VM-1.
VMT estimates from FHWA are allocated to vehicle type using ratios of VMT per vehicle type to total VMT, derived from
EPA's MOVES3 model (see Annex 3.2 for information about the MOVES model). Data for 2021 is proxied using FHWA
Traffic Volume Trends Data.
c Nonroad fuel consumption estimates for 2020 are adjusted to account for the COVID-19 pandemic and associated
restrictions. For agricultural equipment and airport equipment, sector specific adjustment factors were applied to the
2019 data. For all other sectors, a 7.7 percent reduction factor is used, based on transportation diesel use (EIA 2022
dRail emissions do not include emissions from electric powered locomotives. Class II and Class III diesel consumption data
for 2014 to 2021 is estimated by applying the historical average fuel usage per carload factor to the annual number of
carloads.
e Includes equipment, such as tractors and combines, as well as fuel consumption from trucks that are used off-road in
agriculture.
f Includes equipment, such as cranes, dumpers, and excavators, as well as fuel consumption from trucks that are used
off-road in construction.
g "Other" includes snowmobiles and other recreational equipment, logging equipment, lawn and garden equipment,
railroad equipment, airport equipment, commercial equipment, and industrial equipment, as well as fuel consumption
from trucks that are used off-road for commercial/industrial purposes.
l Table 3-15: N2O Emissions from Mobile Combustion (MMT CO2 Eq.)
Fuel Type/Vehicle Type3
1990
2005
2017
2018
2019
2020
2021
Gasoline On-Roadb
32.0
28.5
8.4
7.0
8.1
6.4
6.2
Passenger Cars
22.4
13.3
2.9
2.5
2.5
2.0
2.0
Light-Duty Trucks
8.7
14.0
5.2
4.3
5.4
4.2
4.0
Medium- and Heavy-Duty Trucks
and Buses
0.8
1.2
0.2
0.2
0.2
0.1
0.1
Motorcycles
+
+
0.1
0.1
0.1
0.1
0.1
Diesel On-Roadb
0.2
0.4
2.8
3.0
3.2
3.0
3.4
Passenger Cars
+
+
+
+
+
+
+
Light-Duty Trucks
+
+
0.2
0.3
0.2
0.2
0.3
Medium- and Heavy-Duty Trucks
0.2
0.3
2.5
2.8
3.0
2.8
3.2
Medium- and Heavy-Duty Buses
+
+
0.2
0.2
0.3
0.2
0.3
Alternative Fuel On-Road
+
+
+
+
+
+
+
Non-Roadc
6.2
8.1
7.4
7.5
7.6
6.8
7.5
Ships and Boats
0.2
0.2
0.2
0.2
0.2
0.1
0.3
Raild
0.2
0.3
0.3
0.3
0.2
0.2
0.2
Aircraft
1.5
1.6
1.4
1.4
1.5
1.0
1.3
Agricultural Equipment6
1.2
1.4
1.1
1.1
1.1
1.1
1.
Construction/Mining Equipment'
1.2
1.9
1.6
1.6
1.7
1.6
1.7
Others
1.8
2.8
2.8
2.9
2.9
2.8
2.9
Total
38.4
37.0
18.5
17.5
19.0
16.1
17.1
+ Does not exceed 0.05 MMT C02 Eq.
a See Annex 3.2 for definitions of on-road vehicle types.
b Gasoline and diesel highway vehicle mileage estimates are based on data from FHWA Highway Statistics Table VM-1.
VMT estimates from FHWA are allocated to vehicle type using ratios of VMT per vehicle type to total VMT, derived from
EPA's MOVES3 model (see Annex 3.2 for information about the MOVES model). Data for 2021 is proxied using FHWA
Traffic Volume Trends Data.
c Nonroad fuel consumption estimates for 2020 are adjusted to account for the COVID-19 pandemic and associated
restrictions. For agricultural equipment and airport equipment, sector specific adjustment factors were applied to the
2019 data. For all other sectors, a 7.7 percent reduction factor is used, based on transportation diesel use (EIA 2022).
d Rail emissions do not include emissions from electric powered locomotives. Class II and Class III diesel consumption data
for 2014 through 2021 is estimated by applying the historical average fuel usage per carload factor to the annual number
of carloads.
e Includes equipment, such as tractors and combines, as well as fuel consumption from trucks that are used off-road in
agriculture.
f Includes equipment, such as cranes, dumpers, and excavators, as well as fuel consumption from trucks that are used off-
road in construction.
Energy 3-31
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
g "Other" includes snowmobiles and other recreational equipment, logging equipment, lawn and garden equipment,
railroad equipment, airport equipment, commercial equipment, and industrial equipment, as well as fuel consumption
from trucks that are used off-road for commercial/industrial purposes.
Note: Totals may not sum due to independent rounding.
C02 from Fossil Fuel Combustion
Methodology and Time-Series Consistency
CO2 emissions from fossil fuel combustion are estimated in line with a Tier 2 method described by the IPCC in the
2006 IPCC Guidelines for National Greenhouse Gas Inventories (IPCC 2006) with some exceptions as discussed
below.31 A detailed description of the U.S. methodology is presented in Annex 2.1, and is characterized by the
following steps:
1. Determine total fuel consumption by fuel type and sector. Total fossil fuel consumption for each year is
estimated by aggregating consumption data by end-use sector (e.g., commercial, industrial), primary fuel
type (e.g., coal, petroleum, gas), and secondary fuel category (e.g., motor gasoline, distillate fuel oil). Fuel
consumption data for the United States were obtained directly from the EIA of the U.S. Department of
Energy (DOE), primarily from the Monthly Energy Review (EIA 2022a). EIA data include fuel consumption
statistics from the 50 U.S. states and the District of Columbia, including tribal lands. The EIA does not
include territories in its national energy statistics, so fuel consumption data for territories were collected
separately from ElA's International Energy Statistics (EIA 2022b).32
For consistency of reporting, the IPCC has recommended that countries report energy data using the
International Energy Agency (IEA) reporting convention and/or IEA data. Data in the IEA format are
presented "top down"—that is, energy consumption for fuel types and categories are estimated from
energy production data (accounting for imports, exports, stock changes, and losses). The resulting
quantities are referred to as "apparent consumption." The data collected in the United States by EIA on
an annual basis and used in this Inventory are predominantly from mid-stream or conversion energy
consumers such as refiners and electric power generators. These annual surveys are supplemented with
end-use energy consumption surveys, such as the Manufacturing Energy Consumption Survey, that are
conducted on a periodic basis (every four years). These consumption datasets help inform the annual
surveys to arrive at the national total and sectoral breakdowns for that total.33
Also, note that U.S. fossil fuel energy statistics are generally presented using gross calorific values (GCV)
(i.e., higher heating values). Fuel consumption activity data presented here have not been adjusted to
correspond to international standards, which are to report energy statistics in terms of net calorific values
(NCV) (i.e., lower heating values).34
Subtract uses accounted for in the Industrial Processes and Product Use chapter. Portions of the fuel
consumption data for seven fuel categories—coking coal, distillate fuel, industrial other coal, petroleum
coke, natural gas, residual fuel oil, and other oil—were reallocated to the Industrial Processes and Product
Use chapter, as they were consumed during non-energy-related industrial activity. To make these
31 The IPCC Tier 3B methodology is used for estimating emissions from commercial aircraft.
32 Fuel consumption by U.S. Territories (i.e., American Samoa, Guam, Puerto Rico, U.S. Virgin Islands, Wake Island, and other
U.S. Pacific Islands) is included in this report and contributed total emissions of 23.0 MMT C02 Eq. in 2020. Data is only available
for ElA's International Energy Statistics through 2020 for coal and natural gas consumption and through 2019 for petroleum
consumption. For this reason, data for the 2020 U.S. Territories emission estimates is proxied to the most recent data available.
33 See IPCC Reference Approach for Estimating C02 Emissions from Fossil Fuel Combustion in Annex 4 for a comparison of U.S.
estimates using top-down and bottom-up approaches.
34 A crude convention to convert between gross and net calorific values is to multiply the heat content of solid and liquid fossil
fuels by 0.95 and gaseous fuels by 0.9 to account for the water content of the fuels. Biomass-based fuels in U.S. energy
statistics, however, are generally presented using net calorific values.
3-32 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
adjustments, additional data were collected from AISI (2004 through 2021), Coffeyville (2012), U.S. Census
Bureau (2001 through 2011), EIA (2022a, 2022d, 2022f), USAA (2008 through 2021), USGS (1991 through
2020), (USGS 2019), USGS (2014 through 2021a), USGS (2014 through 2021b), USGS (1995 through 2013),
USGS (1995,1998, 2000, 2001, 2002, 2007), USGS (2021a), USGS (1991 through 2015a), USGS (1991
through 2020), USGS (2014 through 2021a), USGS (1991 through 2015b), USGS (2021b), USGS (1991
through 2020).35
2. Adjust for biofuels and petroleum denaturant. Fossil fuel consumption estimates are adjusted downward
to exclude fuels with biogenic origins and avoid double counting in petroleum data statistics. Carbon
dioxide emissions from ethanol added to motor gasoline and biodiesel added to diesel fuel are not
included specifically in summing energy sector totals. Net carbon fluxes from changes in biogenic carbon
reservoirs are accounted for in the estimates for LULUCF, therefore, fuel consumption estimates are
adjusted to remove ethanol and biodiesel.36 For the years 1993 through 2008, petroleum denaturant is
currently included in EIA statistics for both natural gasoline and finished motor gasoline. To avoid double
counting, petroleum denaturant is subtracted from finished motor gasoline for these years.37
3. Adjust for exports ofCC>2. Since October 2000, the Dakota Gasification Plant has been exporting CO2
produced in the coal gasification process to Canada by pipeline. Because this CO2 is not emitted to the
atmosphere in the United States, the associated fossil fuel (lignite coal) that is gasified to create the
exported CO2 is subtracted from EIA (2022f) coal consumption statistics that are used to calculate
greenhouse gas emissions from the Energy Sector. The associated fossil fuel is the total fossil fuel burned
at the plant with the CO2 capture system multiplied by the fraction of the plant's total site-generated CO2
that is recovered by the capture system. To make these adjustments, data for CO2 exports were collected
from Environment and Climate Change Canada (2022). A discussion of the methodology used to estimate
the amount of CO2 captured and exported by pipeline is presented in Annex 2.1.
4. Adjust sectoral allocation of distillate fuel oil and motor gasoline. EPA conducted a separate bottom-up
analysis of transportation fuel consumption based on data from the Federal Highway Administration that
indicated that the amount of distillate and motor gasoline consumption allocated to the transportation
sector in the EIA statistics should be adjusted. Therefore, for these estimates, the transportation sector's
distillate fuel and motor gasoline consumption were adjusted to match the value obtained from the
bottom-up analysis. As the total distillate and motor gasoline consumption estimate from EIA are
considered to be accurate at the national level, the distillate and motor gasoline consumption totals for
the residential, commercial, and industrial sectors were adjusted proportionately. The data sources used
in the bottom-up analysis of transportation fuel consumption include AAR (2008 through 2022), Benson
(2002 through 2004), DOE (1993 through 2020), EIA (2007), EIA (1991 through 2022), EPA (2022c), and
FHWA (1996 through 2021).38
5. Adjust for fuels consumed for non-energy uses. U.S. aggregate energy statistics include consumption of
fossil fuels for non-energy purposes. These are fossil fuels that are manufactured into plastics, asphalt,
lubricants, or other products. Depending on the end-use, this can result in storage of some or all of the C
contained in the fuel for a period of time. As the emission pathways of C used for non-energy purposes
are vastly different than fuel combustion (since the C in these fuels ends up in products instead of being
combusted), these emissions are estimated separately in Section 3.2 - Carbon Emitted and Stored in
Products from Non-Energy Uses of Fossil Fuels. Therefore, the amount of fuels used for non-energy
35 See sections on Iron and Steel Production and Metallurgical Coke Production, Ammonia Production and Urea Consumption,
Petrochemical Production, Titanium Dioxide Production, Ferroalloy Production, Aluminum Production, and Silicon Carbide
Production and Consumption in the Industrial Processes and Product Use chapter.
36 Natural gas energy statistics from EIA (2022e) are already adjusted downward to account for biogas in natural gas.
37 These adjustments are explained in greater detail in Annex 2.1.
38 Bottom-up gasoline and diesel highway vehicle fuel consumption estimates are based on data from FHWA Highway Statistics
Table MF-21, MF-27, and VM-1 (FHWA 1996 through 2021).
Energy 3-33
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
purposes was subtracted from total fuel consumption. Data on non-fuel consumption were provided by
EIA (2022d).
6. Subtract consumption of international bunker fuels. According to the UNFCCC reporting guidelines
emissions from international transport activities, or bunker fuels, should not be included in national
totals. U.S. energy consumption statistics include these bunker fuels (e.g., distillate fuel oil, residual fuel
oil, and jet fuel) as part of consumption by the transportation end-use sector, however, so emissions from
international transport activities were calculated separately following the same procedures used to
calculate emissions from consumption of all fossil fuels (i.e., estimation of consumption, and
determination of carbon content).39 The Office of the Under Secretary of Defense (Installations and
Environment) and the Defense Logistics Agency Energy (DLA Energy) of the U.S. Department of Defense
(DoD) (DLA Energy 2022) supplied data on military jet fuel and marine fuel use. Commercial jet fuel use
was estimated based on data from FAA (2022) and DOT (1991 through 2022); residual and distillate fuel
use for civilian marine bunkers was obtained from DOC (1991 through 2022) for 1990 through 2001 and
2007 through 2020, and DHS (2008) for 2003 through 2006.40 Consumption of these fuels was subtracted
from the corresponding fuels totals in the transportation end-use sector. Estimates of international
bunker fuel emissions for the United States are discussed in detail in Section 3.9 - International Bunker
Fuels.
7. Determine the total carbon content of fuels consumed. Total C was estimated by multiplying the amount
of fuel consumed by the amount of C in each fuel. This total C estimate defines the maximum amount of C
that could potentially be released to the atmosphere if all of the C in each fuel was converted to CO2. A
discussion of the methodology and sources used to develop the C content coefficients are presented in
Annexes 2.1 and 2.2.
8. Estimate CO2 Emissions. Total CO2 emissions are the product of the adjusted energy consumption (from
the previous methodology steps 1 through 6), the carbon content of the fuels consumed, and the fraction
of C that is oxidized. The fraction oxidized was assumed to be 100 percent for petroleum, coal, and
natural gas based on guidance in IPCC (2006) (see Annex 2.1). Carbon emissions were multiplied by the
molecular-to-atomic weight ratio of CO2 to C (44/12) to obtain total CO2 emitted from fossil fuel
combustion in million metric tons (MMT).
9. Allocate transportation emissions by vehicle type. This report provides a more detailed accounting of
emissions from transportation because it is such a large consumer of fossil fuels in the United States. 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. Heat contents and
densities were obtained from EIA (2022d) and USAF (1998).41
• For on-road vehicles, annual estimates of combined motor gasoline and diesel fuel consumption by
vehicle category were obtained from FHWA (1996 through 2021); for each vehicle category, the
39 See International Bunker Fuels section in this chapter for a more detailed discussion.
40 Data for 2002 were interpolated due to inconsistencies in reported fuel consumption data.
41 For a more detailed description of the data sources used for the analysis of the transportation end use sector see the Mobile
Combustion (excluding C02) and International Bunker Fuels sections of the Energy chapter, Annex 3.2, and Annex 3.8,
respectively.
3-34 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
percent gasoline, diesel, and other (e.g., CNG, LPG) fuel consumption are estimated using data from
EPA's MOVES model and DOE (1993 through 2022).42'43
• For non-road vehicles, activity data were obtained from AAR (2008 through 2022), APTA (2007
through 2021), APTA (2006), BEA (1991 through 2015), Benson (2002 through 2004), DLA Energy
(2022), DOC (1991 through 2022), DOE (1993 through 2022), DOT (1991 through 2022), EIA (2009a),
EIA (2022e), EIA (2002), EIA (1991 through 2022), EPA (2022c),44 and Gaffney (2007).
• For jet fuel used by aircraft, CO2 emissions from commercial aircraft were developed by the U.S.
Federal Aviation Administration (FAA) using a Tier 3B methodology, consistent IPCC (2006) (see
Annex 3.3). Carbon dioxide emissions from other aircraft were calculated directly based on reported
consumption of fuel as reported by EIA. Allocation to domestic military uses was made using DoD
data (see Annex 3.8). General aviation jet fuel consumption is calculated as the remainder of total jet
fuel use (as determined by EIA) nets all other jet fuel use as determined by FAA and DoD. For more
information, see Annex 3.2.
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2021. Due to data availability and sources, some adjustments outlined in the methodology above are not
applied consistently across the full 1990 to 2021 time series. As described in greater detail in Annex 2.1, to align
with ElA's methodology for calculating motor gasoline consumption, petroleum denaturant adjustments are
applied to motor gasoline consumption only for the period 1993 through 2008. In addition to ensuring time-series
consistency, to ensure consistency in reporting between the Inventory and the Canadian National Greenhouse Gas
Inventory, the amount of associated fossil fuel (lignite coal) that is gasified to create the exported CO2 from the
Dakota Gasification Plant is adjusted to align with the Canadian National Greenhouse Gas Inventory (Environment
and Climate Change Canada 2022). This adjustment is explained in greater detail in Annex 2.1. As discussed in
Annex 5, data are unavailable to include estimates of CO2 emissions from any liquid fuel used in pipeline transport
or non-hazardous industrial waste incineration, but those emissions are assumed to be insignificant.
Box 3-4: Carbon Intensity of U.S. Energy Consumption
The amount of C emitted from the combustion of fossil fuels is dependent upon the carbon content of the fuel
and the fraction of that C that is oxidized. Fossil fuels vary in their average carbon content, ranging from about
53 MMT CO2 Eq./QBtu for natural gas to upwards of 95 MMT CO2 Eq./QBtu for coal and petroleum coke (see
Tables A-42 and A-43 in Annex 2.1 for carbon contents of all fuels). In general, the carbon content per unit of
energy of fossil fuels is the highest for coal products, followed by petroleum, and then natural gas. The overall
carbon intensity of the U.S. economy is thus dependent upon the quantity and combination of fuels and other
energy sources employed to meet demand.
Table 3-16 provides a time series of the carbon intensity of direct emissions for each sector of the U.S. economy.
The time series incorporates only the energy from the direct combustion of fossil fuels in each sector. For
42 On-road fuel consumption data from FHWA Table MF-21 and MF-27 were used to determine total on-road use of motor
gasoline and diesel fuel (FHWA 1996 through 2020). Data for 2021 is proxied using FHWA Traffic Volume Travel Trends. Ratios
developed from MOVES3 output are used to apportion FHWA fuel consumption data to vehicle type and fuel type (see Annex
3.2 for information about the MOVES model).
43 Transportation sector natural gas and LPG consumption are based on data from EIA (2022a). In previous Inventory years,
data from DOE (1993 through 2022) TEDB was used to estimate each vehicle class's share of the total natural gas and LPG
consumption. Since TEDB does not include estimates for natural gas use by medium- and heavy-duty trucks or LPG use by
passenger cars, EIA Alternative Fuel Vehicle Data (Browning 2017) is now used to determine each vehicle class's share of the
total natural gas and LPG consumption. These changes were first incorporated in the 1990 through 2015 Inventory and apply to
the time period from 1990 to 2015.
44 In 2014, EPA incorporated the NONROAD2008 model into the MOVES model framework (EPA 2022c). The current Inventory
uses the Nonroad component of MOVES3 for years 1999 through 2021.
Energy 3-35
-------
example, the carbon intensity for the residential sector does not include the energy from or emissions related to
the use of electricity for lighting, as it is instead allocated to the electric power sector. For the purposes of
maintaining the focus of this section, renewable energy and nuclear energy are not included in the energy totals
used in Table 3-16 in order to focus attention on fossil fuel combustion as detailed in this chapter. Looking only
at this direct consumption of fossil fuels, the residential sector exhibited the lowest carbon intensity, which is
related to the large percentage of its energy derived from natural gas for heating. The carbon intensity of the
commercial sector has predominantly declined since 1990 as commercial businesses shift away from petroleum
to natural gas. The industrial sector was more dependent on petroleum and coal than either the residential or
commercial sectors, and thus had higher C intensities over this period. The carbon intensity of the
transportation sector was closely related to the carbon content of petroleum products (e.g., motor gasoline and
jet fuel, both around 70 MMT CO2 Eq./QBtu), which were the primary sources of energy. Lastly, the electric
power sector had the highest carbon intensity due to its heavy reliance on coal for generating electricity.
Table 3-16: Carbon Intensity from Direct Fossil Fuel Combustion by Sector (MMT CO2
Eq./QBtu)
Sector
1990
2005
2017
2018
2019
2020
2021
Residential3
57.4
56.8
55.1
55.3
55.2
55.1
54.8
Commercial3
59.7
57.8
56.6
56.0
56.1
56.2
55.5
Industrial3
64.6
64.7
60.8
60.5
60.3
59.8
59.4
Transportation3
71.1
71.5
71.2
71.0
70.9
70.8
70.9
Electric Powerb
87.3
85.8
77.3
75.5
72.9
70.5
72.3
U.S. Territories0
73.1
73.4
71.0
70.4
70.8
71.6
71.5
All Sectors0
73.1
73.6
69.1
68.3
67.3
66.3
67.0
a Does not include electricity or renewable energy consumption.
b Does not include electricity produced using nuclear or renewable energy.
c Does not include nuclear or renewable energy consumption.
Note: Excludes non-energy fuel use emissions and consumption.
For the time period of 1990 through about 2008, the carbon intensity of U.S. energy consumption was fairly
constant, as the proportion of fossil fuels used by the individual sectors did not change significantly over that
time. Starting in 2008 the carbon intensity has decreased, reflecting the shift from coal to natural gas in the
electric power sector during that time period. Per capita energy consumption fluctuated little from 1990 to
2007, but then started decreasing after 2007 and, in 2021, was approximately 13.1 percent below levels in 1990
(see Figure 3-18). To differentiate these estimates from those of Table 3-16, the carbon intensity trend shown in
Figure 3-18 and described below includes nuclear and renewable energy EIA data to provide a comprehensive
economy-wide picture of energy consumption. Due to a general shift from a manufacturing-based economy to a
service-based economy, as well as overall increases in efficiency, energy consumption and energy-related CO2
emissions per dollar of gross domestic product (GDP) have both declined since 1990 (BEA 2022).
3-36 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Figure 3-18: U.S. Energy Consumption and Energy-Related CO2 Emissions Per Capita and
Per Dollar GDP
120
110
100
I 90
i—l
> 80
§5
1 70
60
50
40
Carbon intensity estimates were developed using nuclear and renewable energy data from EIA (2022d), EPA
(2010), and fossil fuel consumption data as discussed above and presented in Annex 2.1.
Uncertainty
For estimates of CO2 from fossil fuel combustion, the amount of CO2 emitted is directly related to the amount of
fuel consumed, the fraction of the fuel that is oxidized, and the carbon content of the fuel. Therefore, a careful
accounting of fossil fuel consumption by fuel type, average carbon contents of fossil fuels consumed, and
production of fossil fuel-based products with long-term carbon storage should yield an accurate estimate of CO2
emissions.
Nevertheless, there are uncertainties in the consumption data, carbon content of fuels and products, and carbon
oxidation efficiencies. For example, given the same primary fuel type (e.g., coal, petroleum, or natural gas), the
amount of carbon contained in the fuel per unit of useful energy can vary. For the United States, however, the
impact of these uncertainties on overall CO2 emission estimates is believed to be relatively small. See, for example,
Marland and Pippin (1990). See also Annex 2.2 for a discussion of uncertainties associated with fuel carbon
contents. Recent updates to carbon factors for natural gas and coal utilized the same approach as previous
Inventories with updated recent data, therefore, the uncertainty estimates around carbon contents of the
different fuels as outlined in Annex 2.2 were not impacted and the historic uncertainty ranges still apply.
Although national statistics of total fossil fuel and other energy consumption are relatively accurate, the allocation
of this consumption to individual end-use sectors (i.e., residential, commercial, industrial, and transportation) is
less certain. For example, for some fuels the sectoral allocations are based on price rates (i.e., tariffs), but a
commercial establishment may be able to negotiate an industrial rate or a small industrial establishment may end
up paying an industrial rate, leading to a misallocation of emissions. Also, the deregulation of the natural gas
industry and the more recent deregulation of the electric power industry have likely led to some minor challenges
in collecting accurate energy statistics as firms in these industries have undergone significant restructuring.
To calculate the total CO2 emission estimate from energy-related fossil fuel combustion, the amount of fuel used in
non-energy production processes were subtracted from the total fossil fuel consumption. The amount of CO2
emissions resulting from non-energy related fossil fuel use has been calculated separately and reported in the
Carbon Emitted from Non-Energy Uses of Fossil Fuels section of this report (Section 3.2). These factors all
contribute to the uncertainty in the CO2 estimates. Detailed discussions on the uncertainties associated with C
emitted from Non-Energy Uses of Fossil Fuels can be found within that section of this chapter.
Energy 3-37
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Various sources of uncertainty surround the estimation of emissions from international bunker fuels, which are
subtracted from the U.S. totals (see the detailed discussions on these uncertainties provided in Section 3.9 -
International Bunker Fuels). Another source of uncertainty is fuel consumption by U.S. Territories. The United
States does not collect energy statistics for its territories at the same level of detail as for the fifty states and the
District of Columbia. Therefore, estimating both emissions and bunker fuel consumption by these territories is
difficult.
Uncertainties in the emission estimates presented above also result from the data used to allocate CO2 emissions
from the transportation end-use sector to individual vehicle types and transport modes. In many cases, bottom-up
estimates of fuel consumption by vehicle type do not match aggregate fuel-type estimates from EIA. Further
research is planned to improve the allocation into detailed transportation end-use sector emissions.
The uncertainty analysis was performed by primary fuel type for each end-use sector, using the IPCC-
recommended Approach 2 uncertainty estimation methodology, Monte Carlo Stochastic Simulation technique,
with @ RISK software. For this uncertainty estimation, the inventory estimation model for CO2 from fossil fuel
combustion was integrated with the relevant variables from the inventory estimation model for International
Bunker Fuels, to realistically characterize the interaction (or endogenous correlation) between the variables of
these two models. About 170 input variables were modeled for CO2 from energy-related Fossil Fuel Combustion
(including about 20 for non-energy fuel consumption and about 20 for International Bunker Fuels).
In developing the uncertainty estimation model, uniform distributions were assumed for all activity-related input
variables and emission factors, based on the SAIC/EIA (2001) report.45 Triangular distributions were assigned for
the oxidization factors (or combustion efficiencies). The uncertainty ranges were assigned to the input variables
based on the data reported in SAIC/EIA (2001) and on conversations with various agency personnel.46
The uncertainty ranges for the activity-related input variables were typically asymmetric around their inventory
estimates; the uncertainty ranges for the emissions factors were symmetric. Bias (or systematic uncertainties)
associated with these variables accounted for much of the uncertainties associated with these variables (SAIC/EIA
2001).47 For purposes of this uncertainty analysis, each input variable was simulated 10,000 times through Monte
Carlo sampling.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3-17. Fossil fuel
combustion CO2 emissions in 2021 were estimated to be between 4,553.9 and 4,856.1 MMT CO2 Eq. at a 95
percent confidence level. This indicates a range of 2 percent below to 4 percent above the 2021 emission estimate
of 4,651.0 MMTCO2 Eq.
45 SAIC/EIA (2001) characterizes the underlying probability density function for the input variables as a combination of uniform
and normal distributions (the former to represent the bias component and the latter to represent the random component).
However, for purposes of the current uncertainty analysis, it was determined that uniform distribution was more appropriate to
characterize the probability density function underlying each of these variables.
46 In the SAIC/EIA (2001) report, the quantitative uncertainty estimates were developed for each of the three major fossil fuels
used within each end-use sector; the variations within the sub-fuel types within each end-use sector were not modeled.
However, for purposes of assigning uncertainty estimates to the sub-fuel type categories within each end-use sector in the
current uncertainty analysis, SAIC/EIA (2001)-reported uncertainty estimates were extrapolated.
47 Although, in general, random uncertainties are the main focus of statistical uncertainty analysis, when the uncertainty
estimates are elicited from experts, their estimates include both random and systematic uncertainties. Hence, both these types
of uncertainties are represented in this uncertainty analysis.
3-38 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Table 3-17: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Energy-
2 Related Fossil Fuel Combustion by Fuel Type and Sector (MMT CO2 Eq. and Percent)
2021 Emission Estimate Uncertainty Range Relative to Emission Estimate3
Fuel/Sector (MMT CP2 Eq.) (MMT CP2 Eq.) (%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Coalb
957.7
925.0
1,048.0
-3%
9%
Residential
NO
NO
NO
NO
NO
Commercial
1.4
1.4
1.7
-5%
15%
Industrial
43.7
41.6
50.6
-5%
16%
Transportation
NO
NO
NO
NO
NO
Electric Power
909.7
874.9
997.4
-4%
10%
U.S. Territories
2.9
2.5
3.4
-12%
19%
Natural Gasb
1,620.7
1,600.4
1,695.0
-1%
5%
Residential
258.6
251.3
276.8
-3%
7%
Commercial
180.9
175.8
193.6
-3%
7%
Industrial
498.4
482.0
535.2
-3%
7%
Transportation
65.1
63.3
69.7
-3%
7%
Electric Power
615.1
597.2
646.4
-3%
5%
U.S. Territories
2.6
2.3
3.1
-12%
17%
Petroleumb
2,072.2
1,947.0
2,195.5
-6%
6%
Residential
51.5
48.4
54.5
-6%
6%
Commercial
41.6
39.4
43.5
-5%
5%
Industrial
220.3
166.7
272.0
-24%
23%
Transportation
1,724.3
1,616.1
1,831.2
-6%
6%
Electric Power
17.1
16.2
18.5
-5%
8%
U.S. Territories
17.5
16.3
19.4
-7%
11%
Total (excluding Geothermal)b
4,650.6
4,553.4
4,855.6
-2%
4%
Geothermal
0.4
NE
NE
NE
NE
Electric Power
0.4
NE
NE
NE
NE
Total (including Geothermal)b'c
4,651.0
4,553.9
4,856.1
-2%
4%
NO (Not Occurring)
NE (Not Estimated)
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
b The low and high estimates for total emissions were calculated separately through simulations and, hence, the low and
high emission estimates for the sub-source categories do not sum to total emissions.
c Geothermal emissions added for reporting purposes, but an uncertainty analysis was not performed for C02 emissions
from geothermal production.
Note: Totals may not sum due to independent rounding.
3 QA/QC and Verification
4 In order to ensure the quality of the CO2 emission estimates from fossil fuel combustion, general (IPCC Tier 1) and
5 category-specific (Tier 2) Quality Assurance/Quality Control (QA/QC) procedures were implemented consistent
6 with the U.S. Inventory QA/QC plan outlined in Annex 8. The Tier 2 procedures that were implemented involved
7 checks specifically focusing on the activity data and methodology used for estimating CO2 emissions from fossil fuel
8 combustion in the United States. Emission totals for the different sectors and fuels were compared and trends
9 were investigated to determine whether any corrective actions were needed. Minor corrective actions were taken.
10 One area of QA/QC and verification is to compare the estimates and emission factors used in the Inventory with
11 other sources of CO2 emissions reporting. Two main areas and sources of data were considered. The first is a
12 comparison with the EPA GHGRP combustion data (subpart C) for stationary combustion sources excluding the
Energy 3-39
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
electric power sector. This mainly focused on considering carbon factors for natural gas. The second comparison is
with the EPA Air Markets Program data for electric power production. This considered carbon factors for coal and
natural gas used in electric power production.
The EPA GHGRP collects greenhouse gas emissions data from large emitters including information on fuel
combustion. This excludes emissions from mobile sources and smaller residential and commercial sources, those
emissions are covered under supplier reporting (subparts MM and NN) and are areas for further research. Fuel
combustion CO2 data reported in 2021 was 2,084.0 MMT CO2. Of that, 1,581.4 MMT CO2 was from electricity
production. Therefore, the non electric power production fuel combustion reporting was a fraction of the total
covered by the Inventory under fossil fuel combustion. Furthermore, reporters under the GHGRP can use multiple
methods of calculating emissions; one method is to use the default emission factors provided in the rule, while
another is based on a tier 3 approach using their own defined emission factors. Based on data from reporters on
approach used, it was determined that only about 10 percent of natural gas combustion emissions were based on
a tier 3 approach. Given the small sample size compared to the overall Inventory calculations for natural gas
combustion EPA determined it was not reasonable to consider the GHGRP tier 3 natural gas factors at this time.
EPA collects detailed sulfur dioxide (SO2), nitrogen oxides (NOx), and carbon dioxide (CO2) emissions data and other
information from power plants across the country as part of the Acid Rain Program (ARP), the Cross-State Air
Pollution Rule (CSAPR), the CSAPR Update, and the Revised CSAPR Update (RCU). The CO2 data from these Air
Market Programs (AMP) can be compared to the electric power sector emissions calculated from the Inventory as
shown in Table 3-18 for the three most recent years of data.
Table 3-18: Comparison of Electric Power Sector Emissions (MMT CO2 Eq. and Percent)
CO?
Emissions (MMT C02
Eq.)
% Change
Fuel/Sector
2019
2020
2021
19-20
20-21
Inventory Electric Power Sector
1,606.7
1,439.0
1,542.22
-10.4%
7.1%
Coal
973.5
788.2
909.7
-19.0%
15.4%
Natural Gas
616.6
634.3
615.1
3.0%
-3.1%
Petroleum
16.2
16.2
17.1
0.0%
5.6%
AMP Electric Power Sector
1,605.4
1,437.7
1,538.6
-10.4%
7.0%
Coal
980.9
796.3
917.2
-18.8%
15.2%
Natural Gas
616.4
632.6
612.7
2.6%
-3.2%
Petroleum
8.1
8.8
8.7
7.8%
-0.6%
Note: Totals may not sum due to independent rounding.
In general the emissions and trends from the two sources line up well. There are differences expected based on
coverage and scope of each source. The Inventory covers all emissions from the electric power sector as defined
above. The EPA AMP data covers emissions from electricity generating units of a certain size so in some respects it
could cover more sources (like electric power units at industrial facilities that would be covered under the
industrial sector in the Inventory) and not as many sources (since smaller units are excluded). The EPA AMP data
also includes heat input for different fuel types. That data can be combined with emissions to calculate implied
emission factors.48 The following Table 3-19 shows the implied emissions factors for coal and natural gas from the
EPA AMP data compared to the factors used in the Inventory for the three most recent years of data.
Table 3-19: Comparison of Emissions Factors (MMT Carbon/QBtu)
Fuel Type 2019 2020 2021
EPA AMP
Coal 25.52 25.52 25.55
Natural Gas 14.43 14.47 14.50
48 These emission factors can be converted from MMT Carbon/QBtu to MMT C02 Eq./QBtu by multiplying the emission factor
by 44/12, the molecular-to-atomic weight ratio of C02 to C. This would assume the fraction oxidized to be 100 percent, which is
the guidance in IPCC (2006) (see Annex 2.1).
3-40 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
EPA Inventory
Electric Power Coal
Natural Gas
26.08
14.43
26.12
14.43
26.13
14.43
The factors for natural gas line up reasonably well. For coal the EPA emissions factors are roughly 2 percent higher
than those calculated from the EPA AMP data. One possible reason for the difference is that the EPA Inventory
factors are based on all coal used in electric power production while the factors from the EPA AMP data are based
on only units where coal is the only source of fuel used. There are units that use coal and other fuel sources but
emissions for each foul type could not be calculated. This is an area of further research but given current data
available the approach to develop carbon factors as outlined in Annex 2 is still felt to be the most appropriate to
represent total fuel combustion in the United States.
The UNFCCC reporting guidelines also require countries to complete a "top-down" reference approach for
estimating CO2 emissions from fossil fuel combustion in addition to their "bottom-up" sectoral methodology. The
reference approach (detailed in Annex 4) uses alternative methodologies and different data sources than those
contained in this section of the report. The reference approach estimates fossil fuel consumption by adjusting
national aggregate fuel production data for imports, exports, and stock changes rather than relying on end-user
consumption surveys. The reference approach assumes that once carbon-based fuels are brought into a national
economy, they are either saved in some way (e.g., stored in products, kept in fuel stocks, or left unoxidized in ash)
or combusted, and therefore the carbon in them is oxidized and released into the atmosphere. In the reference
approach, accounting for actual consumption of fuels at the sectoral or sub-national level is not required. One
difference between the two approaches is that emissions from carbon that was not stored during non-energy use
of fuels are subtracted from the sectoral approach and reported separately (see Section 3.2). These emissions,
however, are not subtracted in the reference approach. As a result, the reference approach emission estimates are
comparable to those of the sectoral approach, with the exception that the Non-Energy Use (NEU) source category
emissions are included in the reference approach (see Annex 4 for more details).
Recalculations Discussion
Several updates to activity data and emission factors lead to recalculations of previous year results. The major
updates are as follows:
• EIA (2022a) updated energy consumption statistics across the time series relative to the previous
Inventory. This includes an update to transportation sector propane consumption data post 2010.
• EIA (2022a) updated industrial energy sector activity data post 2010 relative to the previous Inventory.
This caused the annually variable carbon contents for HGL (energy use) and HGL (non-energy use) to be
updated across the time series, because post 2010 data is used to back-cast data for prior years. EIA
(2022a) updated petroleum statistics in coordination with its Petroleum Supply Annual 2021. This
impacted the HGL category across the time series.
• EPA revised territories data to correct for an error in how LPG data was pulled. The values for LPG were
previously referencing the values for Other Petroleum from the ElA's International Energy Statistics (EIA
2022b) and have been corrected to reflect the values for Liquified Petroleum Gas from the same source.
• Natural gas consumption data from ElA's Monthly Energy Review (EIA 2022a) Table 10b was updated,
which impacted years 2018-2020.
• The carbon content for propylene was updated from 65.95 kg C02/MMBtu to 67.77 kg C02/MMBtu to
reflect values used in the EPA Greenhouse Gas Emission Factors Hub.
• Fuel consumption changes for the U.S. Territories provided by ElA's International Energy Statistics (EIA
2022b) was updated across the time series.
• Updated values of natural gas used for ammonia production across the time series relative to the previous
Inventory.
All of the revisions discussed above resulted in the following impacts on emissions over time:
• From 1990 to 2020, petroleum emissions from the residential sector decreased by an average annual
amount of 0.09 MMT CO2 Eq. (less than half a percent). Petroleum emissions from the commercial,
Energy 3-41
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
industrial, and transportation sectors increased by an average annual amount of 0.05 MMT CO2 Eq. (less
than half a percent), 0.15 MMT CO2 Eq. (less than half a percent), and 0.01 MMT CO2 Eq. (less than half a
percent), respectively. These changes are due to changes in EIA consumption statistics for petroleum,
changes in EIA industrial energy sector activity data, and the change in carbon content for propylene.
• Petroleum emissions from U.S. Territories decreased by an average annual amount of 1.82 MMT CO2 Eq.
(5.51 percent) due to the correction in data pulled for LPG from 1990 to 2020, change in carbon content
for propylene, and change in fuel consumption data for U.S. Territories.
• Natural gas emissions across the residential, commercial, transportation, and electric power sectors for
years 2018-2020 increased by an average annual amount of 0.19 MMT CO2 Eq. (less than half a percent)
due to an update in natural gas consumption for these sectors in ElA's Monthly Energy Review (EIA 2022a)
Table 10b.
• Natural gas emissions for the industrial sector from 1990-2017 decreased by an average annual amount of
1.00 MMT CO2 Eq. (less than half a percent) due to an update in the correction for natural gas used for
ammonia production. Natural gas emissions for the industrial sector from 2018-2020 decreased by 0.03
MMT CO2 Eq. (less than half a percent) due to updates to both ammonia production and MER table 10b.
• Coal emissions from U.S. Territories decreased by an average annual amount of less than 0.01 MMT CO2
Eq. (less than half a percent) due to the change in fuel consumption data for U.S. Territories.
Overall, these changes resulted in an average annual decrease of 2.5 MMT CO2 Eq. (less than 0.05 percent) in CO2
emissions from fossil fuel combustion for the period 1990 through 2020, relative to the previous Inventory.
However, there were bigger absolute changes across the time series as discussed above.
Planned Improvements
To reduce uncertainty of CO2 from fossil fuel combustion estimates for U.S. Territories, further expert elicitation
may be conducted to better quantify the total uncertainty associated with emissions from U.S. Territories.
Additionally, although not technically a fossil fuel, since geothermal energy-related CO2 emissions are included for
reporting purposes, further expert elicitation may be conducted to better quantify the total uncertainty associated
with CO2 emissions from geothermal energy use.
The availability of facility-level combustion emissions through EPA's GHGRP will continue to be examined to help
better characterize the industrial sector's energy consumption in the United States and further classify total
industrial sector fossil fuel combustion emissions by business establishments according to industrial economic
activity type. Most methodologies used in EPA's GHGRP are consistent with IPCC methodologies, though for EPA's
GHGRP, facilities collect detailed information specific to their operations according to detailed measurement
standards, which may differ with the more aggregated data collected for the Inventory to estimate total, national
U.S. emissions. In addition, and unlike the reporting requirements for this chapter under the UNFCCC reporting
guidelines, some facility-level fuel combustion emissions reported under the GHGRP may also include industrial
process emissions.49 In line with UNFCCC reporting guidelines, fuel combustion emissions are included in this
chapter, while process emissions are included in the Industrial Processes and Product Use chapter of this report. In
examining data from EPA's GHGRP that would be useful to improve the emission estimates for the CO2 from fossil
fuel combustion category, particular attention will also be made to ensure time-series consistency, as the facility-
level reporting data from EPA's GHGRP are not available for all inventory years as reported in this Inventory.
Additional analyses will be conducted to align reported facility-level fuel types and IPCC fuel types per the national
energy statistics. For example, additional work will look at CO2 emissions from biomass to ensure they are
separated in the facility-level reported data and maintaining consistency with national energy statistics provided
by EIA. In implementing improvements and integration of data from EPA's GHGRP, the latest guidance from the
IPCC on the use of facility-level data in national inventories will continue to be relied upon.50
49 See https://unfccc.int/resource/docs/2013/copl9/ene/10a03.pdf#paee=2.
50 See http://www.ipcc-nggip.iges.or.lp/public/tb/TFI Technical Bulletin l.pdf.
3-42 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 An ongoing planned improvement is to develop improved estimates of domestic waterborne fuel consumption.
2 The Inventory estimates for residual and distillate fuel used by ships and boats is based in part on data on bunker
3 fuel use from the U.S. Department of Commerce. Domestic fuel consumption is estimated by subtracting fuel sold
4 for international use from the total sold in the United States. It may be possible to more accurately estimate
5 domestic fuel use and emissions by using detailed data on marine ship activity. The feasibility of using domestic
6 marine activity data to improve the estimates will continue to be investigated.
7 EPA is also evaluating the methods used to adjust for conversion of fuels and exports of CO2. EPA is exploring the
8 approach used to account for CO2 transport, injection, and geologic storage, as part of this there may be changes
9 made to accounting for CO2 exports.
10 Finally, another ongoing planned improvement is to evaluate data availability to update the carbon and heat
11 content of more fuel types accounted for in this Inventory. This update will impact consumption and emissions
12 across all sectors and will improve consistency with EIA data as carbon and heat contents of fuels will be accounted
13 for as annually variable and therefore improve accuracy across the time series. Some of the fuels considered in this
14 effort include petroleum coke, residual fuel, and woody biomass.
15 CH4 and N20 from Stationary Combustion
16 Methodology and Time-Series Consistency
17 Methane and N2O emissions from stationary combustion were estimated by multiplying fossil fuel and wood
18 consumption data by emission factors (by sector and fuel type for industrial, residential, commercial, and U.S.
19 Territories; and by fuel and technology type for the electric power sector). The electric power sector utilizes a Tier
20 2 methodology, whereas all other sectors utilize a Tier 1 methodology. The activity data and emission factors used
21 are described in the following subsections.
22 More detailed information on the methodology for calculating emissions from stationary combustion, including
23 emission factors and activity data, is provided in Annex 3.1.
24 Industrial, Residential, Commercial, and U.S. Territories
25 National coal, natural gas, fuel oil, and wood consumption data were grouped by sector: industrial, commercial,
26 residential, and U.S. Territories. For the CFU and N2O emission estimates, consumption data for each fuel were
27 obtained from ElA's Monthly Energy Review (EIA 2022a). Because the United States does not include territories in
28 its national energy statistics, fuel consumption data for territories were provided separately by ElA's International
29 Energy Statistics (EIA 2022b).51 Fuel consumption for the industrial sector was adjusted to subtract out mobile
30 source construction and agricultural use, which is reported under mobile sources. Construction and agricultural
31 mobile source fuel use was obtained from EPA (2022b) and FHWA (1996 through 2022). Estimates for wood
32 biomass consumption for fuel combustion do not include municipal solid waste, tires, etc., that are reported as
33 biomass by EIA. Non-CC>2 emissions from combustion of the biogenic portion of municipal solid waste and tires is
34 included under waste incineration (Section 3.2). Estimates for natural gas combustion do not include biogas, and
35 therefore non-CC>2 emissions from biogas are not included (see the Planned Improvements section, below). Tier 1
36 default emission factors for the industrial, commercial, and residential end-use sectors were provided by the 2006
37 IPCC Guidelines for National Greenhouse Gas Inventories (IPCC 2006). U.S. Territories' emission factors were
38 estimated using the U.S. emission factors for the primary sector in which each fuel was combusted.
51 U.S. Territories data also include combustion from mobile activities because data to allocate territories' energy use were
unavailable. For this reason, CH4 and N20 emissions from combustion by U.S. Territories are only included in the stationary
combustion totals.
Energy 3-43
-------
1 Electric Power Sector
2 The electric power sector uses a Tier 2 emission estimation methodology as fuel consumption for the electric
3 power sector by control-technology type was based on EPA's Acid Rain Program Dataset (EPA 2022a). Total fuel
4 consumption in the electric power sector from EIA (2022a) was apportioned to each combustion technology type
5 and fuel combination using a ratio of fuel consumption by technology type derived from EPA (2022a) data. The
6 combustion technology and fuel use data by facility obtained from EPA (2022a) were only available from 1996 to
7 2020, so the consumption estimates from 1990 to 1995 were estimated by applying the 1996 consumption ratio by
8 combustion technology type from EPA (2022a) to the total EIA (2022a) consumption for each year from 1990 to
9 1995.
10 Emissions were estimated by multiplying fossil fuel and wood consumption by technology-, fuel-, and country-
11 specific Tier 2 emission factors. The Tier 2 emission factors used are based in part on emission factors published by
12 EPA, and EPA's Compilation of Air Pollutant Emission Factors, AP-42 (EPA 1997) for coal wall-fired boilers, residual
13 fuel oil, diesel oil and wood boilers, natural gas-fired turbines, and combined cycle natural gas units.52
14 Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
15 through 2021 as discussed below. As discussed in Annex 5, data are unavailable to include estimates of Cm and
16 N2O emissions from biomass use in Territories, but those emissions are assumed to be insignificant.
17 Uncertainty
18 Methane emission estimates from stationary sources exhibit high uncertainty, primarily due to difficulties in
19 calculating emissions from wood combustion (i.e., fireplaces and wood stoves). The estimates of CFU and N2O
20 emissions presented are based on broad indicators of emissions (i.e., fuel use multiplied by an aggregate emission
21 factor for different sectors), rather than specific emission processes (i.e., by combustion technology and type of
22 emission control).
23 An uncertainty analysis was performed by primary fuel type for each end-use sector, using the IPCC-recommended
24 Approach 2 uncertainty estimation methodology, Monte Carlo Stochastic Simulation technique, with @RISK
25 software.
26 The uncertainty estimation model for this source category was developed by integrating the CH4 and N2O
27 stationary source inventory estimation models with the model for CO2 from fossil fuel combustion to realistically
28 characterize the interaction (or endogenous correlation) between the variables of these three models. About 55
29 input variables were simulated for the uncertainty analysis of this source category (about 20 from the CO2
30 emissions from fossil fuel combustion inventory estimation model and about 35 from the stationary source
31 inventory models).
32 In developing the uncertainty estimation model, uniform distribution was assumed for all activity-related input
33 variables and N2O emission factors, based on the SAIC/EIA (2001) report.53 For these variables, the uncertainty
34 ranges were assigned to the input variables based on the data reported in SAIC/EIA (2001).54 However, the CFU
52 Several of the U.S. Tier 2 emission factors were used in IPCC (2006) as Tier 1 emission factors. See Table A-69 in Annex 3.1 for
emission factors by technology type and fuel type for the electric power sector.
53 SAIC/EIA (2001) characterizes the underlying probability density function for the input variables as a combination of uniform
and normal distributions (the former distribution to represent the bias component and the latter to represent the random
component). However, for purposes of the current uncertainty analysis, it was determined that uniform distribution was more
appropriate to characterize the probability density function underlying each of these variables.
54 In the SAIC/EIA (2001) report, the quantitative uncertainty estimates were developed for each of the three major fossil fuels
used within each end-use sector; the variations within the sub-fuel types within each end-use sector were not modeled.
However, for purposes of assigning uncertainty estimates to the sub-fuel type categories within each end-use sector in the
current uncertainty analysis, SAIC/EIA (2001)-reported uncertainty estimates were extrapolated.
3-44 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 emission factors differ from those used by EIA. These factors and uncertainty ranges are based on IPCC default
2 uncertainty estimates (IPCC 2006).
3 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3-20. Stationary
4 combustion CFU emissions in 2021 (including biomass) were estimated to be between 5.8 and 20.3 MMT CO2 Eq. at
5 a 95 percent confidence level. This indicates a range of 35 percent below to 129 percent above the 2021 emission
6 estimate of 8.9 MMT CO2 Eq.55 Stationary combustion N2O emissions in 2021 (including biomass) were estimated
7 to be between 16.3 and 33.2 MMT CO2 Eq. at a 95 percent confidence level. This indicates a range of 26 percent
8 below to 50 percent above the 2021 emission estimate of 22.1 MMT CO2 Eq.
9 Table 3-20: Approach 2 Quantitative Uncertainty Estimates for ChU and N2O Emissions from
10 Energy-Related Stationary Combustion, Including Biomass (MMT CO2 Eq. and Percent)
Source
Gas
2021 Emission Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Stationary Combustion
Stationary Combustion
ch4
n2o
8.9
22.1
5.8 20.3
16.3 33.2
-35% 129%
-26% 50%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
11 The uncertainties associated with the emission estimates of CH4 and N2O are greater than those associated with
12 estimates of CO2 from fossil fuel combustion, which mainly rely on the carbon content of the fuel combusted.
13 Uncertainties in both Cm and N2O estimates are due to the fact that emissions are estimated based on emission
14 factors representing only a limited subset of combustion conditions. For the indirect greenhouse gases,
15 uncertainties are partly due to assumptions concerning combustion technology types, age of equipment, emission
16 factors used, and activity data projections.
17 QA/QC and Verification
18 In order to ensure the quality of the non-CC>2 emission estimates from stationary combustion, general (IPCC Tier 1)
19 and category-specific (Tier 2) Quality Assurance/Quality Control (QA/QC) procedures were implemented consistent
20 with the U.S. Inventory QA/QC plan outlined in Annex 8. The Tier 2 procedures that were implemented involved
21 checks specifically focusing on the activity data and emission factor sources and methodology used for estimating
22 Cm, N2O, and the greenhouse gas precursors from stationary combustion in the United States. Emission totals for
23 the different sectors and fuels were compared and trends were investigated.
24 Recalculations Discussion
25 EIA (2022a) updated petroleum statistics in coordination with its Petroleum Supply Annual 2021. This impacted the
26 HGL category across the time series.
27 Fuel consumption data for U.S. Territories provided by ElA's International Energy Statistics (EIA 2022b) was
28 updated across the timeseries. Non-C02 emissions from U.S Territories decreased by an average annual amount of
29 less than 0.01 MMT CO2 Eq. (less than half a percent) for coal and less than 0.01 MMT CO2 Eq. (5.76 percent) for
30 fuel oil due to the update in fuel consumption data for U.S. Territories.
55 The low emission estimates reported in this section have been rounded down to the nearest integer values and the high
emission estimates have been rounded up to the nearest integer values.
Energy 3-45
-------
1 Wood and natural gas consumption data from ElA's Monthly Energy Review (EIA 2022a) Table 10b was updated,
2 which impacted years 2018-2020. Non-CCh emissions across the residential, commercial, industrial, and electric
3 power sectors decreased by an average annual amount of less than 0.04 MMT CO2 Eq. (less than half a percent) for
4 wood and increased by an average annual amount of less than 0.03 MMT CO2 Eq. (less than half a percent) for
5 natural gas due to the update in MER table 10b.
6 In addition, for the current Inventory, CCh-equivalent emissions of CFU and N2O from stationary combustion have
7 been revised to reflect the 100-year global warming potentials (GWPs) provided in the IPCC Fifth Assessment
8 Report (AR5) (IPCC 2013). AR5 GWP values differ slightly from those presented in the IPCC Fourth Assessment
9 Report (AR4), used in previous Inventories (IPCC 2007). The AR5 GWPs have been applied across the entire time
10 series for consistency. Prior inventories used GWPs of 25 and 298 for CFU and N2O, respectively. These values have
11 been updated to 28 and 265, respectively. Compared to the previous Inventory which applied 100-year GWP
12 values from AR4, the average annual change in CC>2-equivalent CFU emissions was a 12 percent increase and the
13 average annual change in CC>2-equivalent N2O emissions was an 11 percent decrease for the time series. As a result
14 of the change in methodology, total emissions across the timeseries changed by an average annual decrease of 2.3
15 MMT CO2 Eq. (6.1 percent) relative to emissions results calculated using the prior GWPs. Further discussion on this
16 update and the overall impacts of updating the Inventory GWP values to reflect the IPCC AR5 can be found in
17 Chapter 9, Recalculations and Improvements.
is Planned Improvements
19 Several items are being evaluated to improve the CH4 and N2O emission estimates from stationary combustion and
20 to reduce uncertainty for U.S. Territories. Efforts will be taken to work with EIA and other agencies to improve the
21 quality of the U.S. Territories data. Because these data are not broken out by stationary and mobile uses, further
22 research will be aimed at trying to allocate consumption appropriately. In addition, the uncertainty of biomass
23 emissions will be further investigated because it was expected that the exclusion of biomass from the estimates
24 would reduce the uncertainty; and in actuality the exclusion of biomass increases the uncertainty. These
25 improvements are not all-inclusive but are part of an ongoing analysis and efforts to continually improve these
26 stationary combustion estimates from U.S. Territories.
27 Other forms of biomass-based gas consumption include biogas. As an additional planned improvement, EPA will
28 examine EIA and GHGRP data on biogas collected and burned for energy use and determine if CH4 and N2O
29 emissions from biogas can be included in future Inventories. EIA (2022a) natural gas data already deducts biogas
30 used in the natural gas supply, so no adjustments are needed to the natural gas fuel consumption data to account
31 for biogas.
32 CH4 and N20 from Mobile Combustion
33 Methodology and Time-Series Consistency
34 Estimates of CFU and N2O emissions from mobile combustion were calculated by multiplying emission factors by
35 measures of activity for each fuel and vehicle type (e.g., light-duty gasoline trucks). Activity data included vehicle
36 miles traveled (VMT) for on-road vehicles and fuel consumption for non-road mobile sources. The activity data and
37 emission factors used in the calculations are described in the subsections that follow. A complete discussion of the
38 methodology used to estimate CFU and N2O emissions from mobile combustion and the emission factors used in the
39 calculations is provided in Annex 3.2.
40 On-Road Vehicles
41 Estimates of CFU and N2O emissions from gasoline and diesel on-road vehicles are based on VMT and emission
42 factors (in grams of CFU and N2O per mile) by vehicle type, fuel type, model year, and emission control technology.
3-46 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Emission estimates for alternative fuel vehicles (AFVs) are based on VMT and emission factors (in grams of Cm and
N2O per mile) by vehicle and fuel type.56
Cm and N2O emissions factors by vehicle type and emission tier for newer (starting with model year 2004) on-road
gasoline vehicles were calculated by Browning (2019) from annual vehicle certification data compiled by EPA. CH4
and N2O emissions factors for older (model year 2003 and earlier) on-road gasoline vehicles were developed by ICF
(2004). These earlier emission factors were derived from EPA, California Air Resources Board (CARB) and
Environment and Climate Change Canada (ECCC) laboratory test results of different vehicle and control technology
types. The EPA, CARB and ECCC tests were designed following the Federal Test Procedure (FTP). The procedure
covers three separate driving segments, since vehicles emit varying amounts of greenhouse gases depending on
the driving segment. These driving segments are: (1) a transient driving cycle that includes cold start and running
emissions, (2) a cycle that represents running emissions only, and (3) a transient driving cycle that includes hot
start and running emissions. For each test run, a bag was affixed to the tailpipe of the vehicle and the exhaust was
collected; the content of this bag was then analyzed to determine quantities of gases present. The emissions
characteristics of driving segment 2 tests were used to define running emissions. Running emissions were
subtracted from the total FTP emissions to determine start emissions. These were then recombined to
approximate average driving characteristics, based upon the ratio of start to running emissions for each vehicle
class from MOBILE6.2, an EPA emission factor model that predicts grams per mile emissions of CO2, CO, HC, NOx,
and PM from vehicles under various conditions.57
Diesel on-road vehicle emission factors were developed by ICF (2006a). CH4 and N2O emissions factors for newer
(starting with model year 2007) on-road diesel vehicles (those using engine aftertreatment systems) were
calculated from annual vehicle certification data compiled by EPA.
CH4 and N2O emission factors for AFVs were developed based on the 2021 Greenhouse gases, Regulated
Emissions, and Energy use in Transportation (GREET) model (ANL 2022). For light-duty trucks, EPA used travel
fractions for LDT1 and LDT2 (MOVES Source Type 31 for LDT1 and MOVES Source Type 32 for LDT2; see Annex 3.2
for information about the MOVES model) to determine light-duty truck emission factors. For medium-duty
vehicles, EPA used emission factors for light heavy-duty vocational trucks. For heavy-duty vehicles, EPA used
emission factors for long-haul combination trucks. For buses, EPA used emission factors for transit buses. These
values represent vehicle operations only (tank-to-wheels); upstream well-to-tank emissions are calculated
elsewhere in the Inventory. Biodiesel CH4 emission factors were corrected from GREET values to be the same as
CH4 emission factors for diesel vehicles. GREET overestimated biodiesel CH4 emission factors based upon an
incorrect CI-U-to-THC ratio for diesel vehicles with aftertreatment technology.
Annual VMT data for 1990 through 2020 were obtained from the Federal Highway Administration's (FHWA)
Highway Performance Monitoring System database as reported in Highway Statistics (FHWA 1996 through 2020).58
VMT estimates were then allocated to vehicle type using ratios of VMT per vehicle type to total VMT, derived from
EPA's MOVES3 model (see Annex 3.2 for information about the MOVES model). This corrects time series
inconsistencies in FHWA definitions of vehicle types (Browning 2022a). VMT for alternative fuel vehicles (AFVs)
were estimated based on Browning (2022b). The age distributions of the U.S. vehicle fleet were obtained from EPA
(2004, 2021b), and the average annual age-specific vehicle mileage accumulation of U.S. vehicles were obtained
from EPA (2021b).
56 Alternative fuel and advanced technology vehicles are those that can operate using a motor fuel other than gasoline or
diesel. This includes electric or other bi-fuel or dual-fuel vehicles that may be partially powered by gasoline or diesel.
57 Additional information regarding the MOBILE model can be found at https://www.epa.gov/moves/description-and-historv-
mobile-highway-vehicle-emission-factor-model.
58 Note that VMT for 2021 is estimated with FHWA Traffic Volume Trends data for this public review, but actual data for 2021
will be included in the Final Report when it is released.
Energy 3-47
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Control technology and standards data for on-road vehicles were obtained from EPA's Office of Transportation and
Air Quality (EPA 2021c, 2021d, and 1998) and Browning (2005). These technologies and standards are defined in
Annex 3.2, and were compiled from EPA (1994a, 1994b, 1998,1999a) and IPCC (2006) sources.
Non-Road Mobile Sources
The nonroad mobile category for Cm and N2O includes ships and boats, aircraft, locomotives, and other mobile
non-road sources (e.g., construction or agricultural equipment). For locomotives, aircraft, ships and non-
recreational boats, fuel-based emission factors are applied to data on fuel consumption, following the IPCC Tier 1
approach, The Tier 2 approach for these sources would require separate fuel-based emissions factors by
technology, for which data are not currently available. For other non-road sources, EPA uses the Nonroad
component of the MOVES model to estimate fuel use. Emission factors by horsepower bin are estimated from EPA
engine certification data. Because separate emission factors are applied to specific engine technologies; these non-
road sources utilize a Tier 2 approach.
To estimate CFU and N2O emissions from non-road mobile sources, fuel consumption data were employed as a
measure of activity and multiplied by fuel-specific emission factors (in grams of N2O and CFU per kilogram of fuel
consumed). 59 Activity data were obtained from AAR (2008 through 2022), APTA (2007 through 2022), Rail Inc
(2014 through 2022), APTA (2006), BEA (1991 through 2015), Benson (2002 through 2004), DLA Energy (2022),
DOC (1991 through 2022), DOE (1993 through 2022), DOT (1991 through 2022), EIA (2002, 2007, 2022), EIA
(2022f), EIA (1991 through 2022), EPA (2022b), Esser (2003 through 2004), FAA (2022), FHWA (1996 through
2022),60 Gaffney (2007), and Whorton (2006 through 2014). Emission factors for non-road modes were taken from
IPCC (2006) and Browning (2020a and 2018b).
Uncertainty
A quantitative uncertainty analysis was conducted for the mobile source sector using the IPCC-recommended
Approach 2 uncertainty estimation methodology, Monte Carlo Stochastic Simulation technique, using @RISK
software. The uncertainty analysis was performed on 2021 estimates of CH4 and N2O emissions, incorporating
probability distribution functions associated with the major input variables. For the purposes of this analysis, the
uncertainty was modeled for the following four major sets of input variables: (1) VMT data, by on-road vehicle and
fuel type, (2) emission factor data, by on-road vehicle, fuel, and control technology type, (3) fuel consumption,
data, by non-road vehicle and equipment type, and (4) emission factor data, by non-road vehicle and equipment
type.
Uncertainty analyses were not conducted for NOx, CO, or NMVOC emissions. Emission factors for these gases have
been extensively researched because emissions of these gases from motor vehicles are regulated in the United
States, and the uncertainty in these emission estimates is believed to be relatively low. For more information, see
Section 3.11. However, a much higher level of uncertainty is associated with CH4 and N2O emission factors due to
limited emission test data, and because, unlike CO2 emissions, the emission pathways of CFU and N2O are highly
complex.
59 The consumption of international bunker fuels is not included in these activity data, but emissions related to the
consumption of international bunker fuels are estimated separately under the International Bunker Fuels source category.
60 This Inventory uses FHWA's Agriculture, Construction, and Commercial/Industrial MF-24 fuel volumes along with the MOVES
model gasoline volumes to estimate non-road mobile source CH4 and N20 emissions for these categories. For agriculture, the
MF-24 gasoline volume is used directly because it includes both non-road trucks and equipment. For construction and
commercial/industrial category gasoline estimates, the 2014 and older MF-24 volumes represented non-road trucks only;
therefore, the MOVES gasoline volumes for construction and commercial/industrial categories are added to the respective
categories in the Inventory. Beginning in 2015, this addition is no longer necessary since the FHWA updated its methods for
estimating on-road and non-road gasoline consumption. Among the method updates, FHWA now incorporates MOVES
equipment gasoline volumes in the construction and commercial/industrial categories.
3-48 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Based on the uncertainty analysis, mobile combustion Cm emissions from all mobile sources in 2021 were
estimated to be 2.5 and 3.4 MMT CO2 Eq. at a 95 percent confidence level. This indicates a range of 4 percent
below to 30 percent above the corresponding 2021 emission estimate of 2.6 MMT CO2 Eq. Mobile combustion N2O
emissions from mobile sources in 2021 were estimated to be between 16.0 and 20.8 MMT CO2 Eq. at a 95 percent
confidence level. This indicates a range of 7 percent below to 21 percent above the corresponding 2021 emission
estimate of 17.2 MMT CO2 Eq.
Table 3-21: Approach 2 Quantitative Uncertainty Estimates for ChU and N2O Emissions from
Mobile Sources (MMT CO2 Eq. and Percent)
Source
Gas
2021 Emission Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (Percent)
Lower
Bound
Upper
Bound
Lower Upper
Bound Bound
Mobile Sources
ch4
2.6
2.5
3.4
-4% +30%
Mobile Sources
n2o
17.2
16.0
20.8
-7% +21%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
This uncertainty analysis is a continuation of a multi-year process for developing quantitative uncertainty estimates
for this source category using the IPCC Approach 2 uncertainty estimation methodology. As a result, as new
information becomes available, uncertainty characterization of input variables may be improved and revised. For
additional information regarding uncertainty in emission estimates for Cm and N2O please refer to the Uncertainty
Annex. As discussed in Annex 5, data are unavailable to include estimates of CH4 and N2O emissions from any liquid
fuel used in pipeline transport or some biomass used in transportation sources, but those emissions are assumed
to be insignificant.
QA/QC and Verification
In order to ensure the quality of the emission estimates from mobile combustion, general (IPCC Tier 1) and
category-specific (Tier 2) Quality Assurance/Quality Control (QA/QC) procedures were implemented consistent
with the U.S. Inventory QA/QC plan outlined in Annex 8. The specific plan used for mobile combustion was
updated prior to collection and analysis of this current year of data. The Tier 2 procedures focused on the emission
factor and activity data sources, as well as the methodology used for estimating emissions. These procedures
included a qualitative assessment of the emission estimates to determine whether they appear consistent with the
most recent activity data and emission factors available. A comparison of historical emissions between the current
Inventory and the previous Inventory was also conducted to ensure that the changes in estimates were consistent
with the changes in activity data and emission factors.
Recalculations Discussion
In previous inventories, on-highway greenhouse gas emissions were calculated using FHWA fuel consumption and
vehicle miles traveled (VMT) data delineating by FHWA vehicle classes. These fuel consumption estimates were
then combined with estimates of fuel shares by vehicle type from Oak Ridge National Laboratory's Transportation
Energy Data Book (TEDB), to develop an estimate of fuel consumption for each vehicle type in the Inventory (i.e.,
passenger cars, light-duty trucks, buses, medium- and heavy-duty trucks, motorcycles). However, in 2011, FHWA
changed its methods for estimating VMT and related data. These methodological changes included how vehicles
are classified, moving from a system based on body-type to one that is based on wheelbase. These changes were
first incorporated in the 1990 through 2008 Inventory and applied to the time series beginning in 2007. The FHWA
methodology update resulted in large changes in VMT and fuel consumption by vehicle class, leading to a shift in
emissions among vehicle classes. For example, FHWA replaced the vehicle category "Passenger Cars" with "Light-
duty Vehicles-Short Wheelbase" and the "Other 2 axle-4 Tire Vehicles" category was replaced by "Light-duty
Vehicles, Long Wheelbase." FHWA changed the definition of light-duty vehicles to less than 10,000 lbs. GVWR
instead of 8,500 lbs. GVWR pushed some single-unit heavy-duty trucks to the light-duty class. This change in
vehicle classification also moved some smaller trucks and sport utility vehicles from the light truck category to the
Energy 3-49
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
passenger cars category in this Inventory. These updates resulted in a disconnect in FHWA VMT and fuel
consumption data in the 2006 to 2007 timeframe, generating a large drop in the light-duty truck VMT and fuel
consumption trend lines between 2006 and 2007, and a corresponding increase in the passenger cars trend lines.
To address this inconsistency in the time series, EPA updated the methodology to divide FHWA VMT data into
vehicle classes and fuel type using distributions from EPA's Motor Vehicle Emission Simulator, MOVES. The MOVES
model is a nationally recognized model based on vehicle registration, travel activity, and emission rates that are
updated with each model release. MOVES3 is the latest version of MOVES and uses forecast growth factors which
provide EPA's best estimate of likely future activity based on historical data (see Annex 3.2 for more information
about the MOVES model). Thus, dividing FHWA total VMT data into vehicle class and fuel type using MOVES3
ratios provides a more consistent estimate of vehicle activity over the Inventory time series. MOVES3 ratios are
also used to reallocate FHWA gasoline and diesel fuel use data (Browning 2022a). For this update, the MOVES3
model was run for calendar years 1990 and 1999 through 2021 for all vehicle types. Calendar years 1991 through
1998 were linearly interpolated from 1990 and 1999 calendar year MOVES3 outputs. Model outputs of VMT and
fuel consumption were binned by calendar year, MOVES vehicle type, and fuel type; MOVES vehicle types were
then mapped to the vehicle types used in the Inventory. Only outputs of gasoline and diesel fuel consumption from
MOVES3 were used; alternative fuel VMT and fuel consumption outputs are ignored because they are calculated
for the Inventory under a separate methodology. Total gasoline and diesel fuel consumption values from FHWA
were then allocated to Inventory vehicle types using gasoline and diesel fuel consumption ratios by vehicle type
from MOVES3. Similarly, VMT by vehicle type and fuel type was calculated by multiplying the total VMT from
FHWA by VMT ratios by vehicle and fuel type generated by MOVES3. Overall, because total fuel consumption and
VMT values are conserved, the changes in total emissions are small, within 0.1 percent. Observed differences in
total emissions are due to changes in CH4 and N2O emissions, as the methodology for calculating these non-C02
emissions utilizes more detailed activity data and is therefore sensitive to the re-allocation of activity data. While
total emissions estimates are not significantly impacted by this methodology update, there are significant changes
in the allocation of emissions by vehicle type. The share of emissions allocated to passenger cars now generally
decline through the time series while the share of emissions allocated to light-duty trucks increase over time.
In addition, the methodology for estimating emissions from alternative fuel vehicles was revised. In previous
Inventories, EPA used Energy Information Administration (EIA) surveys of fleet vehicles used by electricity
providers, federal agencies, natural gas providers, propane providers, state agencies and transit agencies to
determine fuel use and vehicle counts for most alternative fuel vehicles. However, EIA stopped conducting these
surveys in 2017. To address this data void, EPA used various methods to determine vehicle counts. Beginning with
the 1990 through 2018 Inventory, electric, plug-in electric, and fuel cell vehicle counts were determined from
vehicles sales data published by Wards Intelligence. Beginning with this Inventory, electric and fuel cell heavy-duty
bus counts are determined from Zukowski, D. (2022) for calendar years 2018 through 2021. Vehicle counts for
other fuels (methanol, ethanol, natural gas, and LPG) for 2018 onward were estimated via regression analysis
(Browning 2022b).
In addition, the latest version of Argonne National Laboratory's Greenhouse Gas, Regulated Emissions, and Energy
Use in Transportation Model (GREET2022) provided updated emission factors for all alternative fuel vehicle classes
(ANL 2022). Updated emission factors from GREET2022 were implemented in this Inventory, across the entire time
series.
The updated vehicle counts and emission factors resulted in a 16 percent reduction in CO2, a 51 percent reduction
in CH4, and a 92 percent reduction in N2O in calendar year 2020 for alternative fuel vehicles compared with the
previous methodology. This resulted in a 21 percent overall reduction in CO2 Eq. for alternative fuel vehicles
compared with the previous methodology.
In addition, for the current Inventory, C02-equivalent estimates of CH4and N2O emissions from transportation and
mobile combustion have been revised to reflect the 100-year global warming potentials (GWPs) provided in the
IPCC Fifth Assessment Report (AR5) (IPCC 2013). AR5 GWP values differ slightly from those presented in the IPCC
3-50 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Fourth Assessment Report (AR4) (IPCC 2007) (used in the previous inventories). The AR5 GWPs have been applied
across the entire time series for consistency.
The GWP of Cm increased, leading to an overall increase in CH4 emissions reported in CO2 equivalent. The GWP of
N2O decreased, leading to a decrease in emissions from N2O reported in CO2 equivalent. Compared to the previous
Inventory which applied 100-year GWP values from AR4, the average annual change in CC>2-equivalent CH4
emissions was a 12 percent increase and the average annual change in CC>2-equivalent N2O emissions was 11
percent decrease for the time series. The net impact from these updates was an average annual 0.1 percent
decrease in total CO2 Eq. emissions for the time series in recent years. Further discussion on this update and the
overall impacts of updating the Inventory GWP values to reflect the IPCC AR5 can be found in Chapter 9,
Recalculations and Improvements.
Planned Improvements
While the data used for this report represent the most accurate information available, several areas for
improvement have been identified.
• Update emission factors for ships and non-recreational boats using residual fuel and distillate fuel,
emission factors for locomotives using ultra low sulfur diesel, and emission factors for aircraft using jet
fuel. The Inventory currently uses IPCC default values for these emission factors.
• Continue to explore potential improvements to estimates of domestic waterborne fuel consumption for
future Inventories. The Inventory estimates for residual and distillate fuel used by ships and boats is based
in part on data on bunker fuel use from the U.S. Department of Commerce. Domestic fuel consumption is
estimated by subtracting fuel sold for international use from the total sold in the United States. Since
2015, all ships travelling within 200 nautical miles of the U.S. coastlines must use distillate fuels thereby
overestimating the residual fuel used by U.S. vessels and underestimating distillate fuel use in these ships.
3.2 Carbon Emitted from Non-Energy Uses
of Fossil Fuels (CRF Source Category 1A)
In addition to being combusted for energy, fossil fuels are also consumed for non-energy uses (NEU) in the United
States. The fuels used for these purposes are diverse, including natural gas, hydrocarbon gas liquids (HGL),61
asphalt (a viscous liquid mixture of heavy crude oil distillates), petroleum coke (manufactured from heavy oil), and
coal (metallurgical) coke (manufactured from coking coal). The non-energy applications of these fuels are equally
diverse, including feedstocks for the manufacture of plastics, rubber, synthetic fibers and other materials; reducing
agents for the production of various metals and inorganic products; and products such as lubricants, waxes, and
asphalt (IPCC 2006). Emissions from non-energy use of lubricants, paraffin waxes, bitumen / asphalt, and solvents
are reported in the Energy sector, as opposed to the Industrial Processes and Product Use (IPPU) sector, to reflect
national circumstances in its choice of methodology and to increase transparency of this source category's unique
country-specific data sources and methodology (see Box 3-5). In addition, estimates of non-energy use emissions
included here do not include emissions already reflected in the IPPU sector, e.g., fuels used as reducing agents. To
avoid double counting, the "raw" non-energy fuel consumption data reported by EIA are reduced to account for
these emissions already included under IPPU.
61 HGL (formerly referred to as liquefied petroleum gas, or LPG) are hydrocarbons that occur as gases at atmospheric pressure
and as liquids under higher pressures. HGLs include paraffins, such as ethane, propane, butanes, isobutane, and natural
gasoline (formerly referred to as pentanes plus), and HGLs include olefins, such as ethylene, propylene, butylene and
isobutylene.
Energy 3-51
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Carbon dioxide emissions arise from non-energy uses via several pathways. Emissions may occur during the
manufacture of a product, as is the case in producing plastics or rubber from fuel-derived feedstocks. Additionally,
emissions may occur during the product's lifetime, such as during solvent use. Overall, throughout the time series
and across all uses, about 62 percent of the total C consumed for non-energy purposes was stored in products
(e.g., plastics), and not released to the atmosphere; the remaining 38 percent was emitted.
There are several areas in which non-energy uses of fossil fuels are closely related to other parts of this Inventory.
For example, some of the non-energy use products release CO2 at the end of their commercial life when they are
combusted after disposal; these emissions are reported separately within the Energy chapter in the Incineration of
Waste source category. There are also net exports of petrochemical intermediate products that are not completely
accounted for in the EIA data, and the Inventory calculations adjust for the effect of net exports on the mass of C in
non-energy applications.
As shown in Table 3-22, fossil fuel emissions in 2021 from the non-energy uses of fossil fuels were 143.2 MMT CO2
Eq., which constituted approximately 2.8 percent of overall fossil fuel emissions. In 2021, the consumption of fuels
for non-energy uses (after the adjustments described above) was 5,938.1 TBtu (see Table 3-23). A portion of the C
in the 5,938.1 TBtu of fuels was stored (234.4 MMT CO2 Eq.), while the remaining portion was emitted (143.2 MMT
CO2 Eq.). Non-energy use emissions increased by 20.1 percent from 2020 to 2021, mainly due to an increase in HGL
and industrial coking coal fuel consumption, which contributed to an 18.3 MMT CO2 Eq. increase in emissions from
2020 to 2021. Although a rise in consumption of some fuels was potentially due to a bounce back in production
following the early effects of the COVID-19 pandemic (e.g., naphtha and special naphtha production returned
closer to pre-2020 levels), the overall increase in 2021 emissions for select industries exceeds pre-pandemic levels.
See Annex 2.3 for more details.
Table 3-22: CO2 Emissions from Non-Energy Use Fossil Fuel Consumption (MMT CO2 Eq. and
Percent C)
Year
1990
2005
2017
2018
2019
2020
2021
Potential Emissions
305.8
366.9
332.4
352.6
355.9
350.2
377.6
C Stored
193.4
238.0
219.6
223.2
228.2
231.0
234.4
Emissions as a % of Potential
37%
35%
34%
37%
36%
34%
38%
C Emitted
112.4
128.9
112.8
129.4
127.6
119.2
143.2
Note: NEU emissions presented in this table differ from the NEU emissions presented in CRF table l.A(a)s4 as the CRF
NEU emissions do not include NEU of lubricants and other petroleum in U.S. Territories. NEU emissions from U.S.
Territories are reported under U.S. Territories in the CRF table l.A(a)s4.
Methodology and Time-Series Consistency
The first step in estimating C stored in products was to determine the aggregate quantity of fossil fuels consumed
for non-energy uses. The C content of these feedstock fuels is equivalent to potential emissions, or the product of
consumption and the fuel-specific C content values. Both the non-energy fuel consumption and C content data
were supplied by the EIA (2022) (see Annex 2.1). Consumption values for industrial coking coal, petroleum coke,
other oils, and natural gas in Table 3-23 and Table 3-24 have been adjusted to subtract non-energy uses that are
included in the source categories of the Industrial Processes and Product Use chapter.62 Consumption of natural
gas, HGL, naphthas, other oils, and special naphtha were adjusted to subtract out net exports of these products
that are not reflected in the raw data from EIA. Consumption values were also adjusted to subtract net exports of
HGL components (e.g., propylene, ethane).
62 These source categories include Iron and Steel Production, Lead Production, Zinc Production, Ammonia Manufacture, Carbon
Black Manufacture (included in Petrochemical Production), Titanium Dioxide Production, Ferroalloy Production, Silicon Carbide
Production, and Aluminum Production.
3-52 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 For the remaining non-energy uses, the quantity of C stored was estimated by multiplying the potential emissions
2 by a storage factor.
3 • For several fuel types—petrochemical feedstocks (including natural gas for non-fertilizer uses, HGL,
4 naphthas, other oils, still gas, special naphtha, and industrial other coal), asphalt and road oil, lubricants,
5 and waxes—U.S. data on C stocks and flows were used to develop C storage factors, calculated as the
6 ratio of (a) the C stored by the fuel's non-energy products to (b) the total C content of the fuel consumed.
7 A lifecycle approach was used in the development of these factors in order to account for losses in the
8 production process and during use. Because losses associated with municipal solid waste management
9 are handled separately in the Energy sector under the Incineration of Waste source category, the storage
10 factors do not account for losses at the disposal end of the life cycle.
11 • For industrial coking coal and distillate fuel oil, storage factors were taken from Marland and Rotty (1984).
12 • For the remaining fuel types (petroleum coke, miscellaneous products and other petroleum), IPCC (2006)
13 does not provide guidance on storage factors, and assumptions were made based on the potential fate of
14 C in the respective non-energy use products. Carbon dioxide emissions from carbide production are
15 implicitly accounted for in the storage factor calculation for the non-energy use of petroleum coke.
16
17 Table 3-23: Adjusted Consumption of Fossil Fuels for Non-Energy Uses (TBtu)
Year
1990
2005
2017
2018
2019
2020
2021
Industry
4,317.8
5,115.1
5,089.8
5,448.0
5,484.1
5,444.8
5,815.9
Industrial Coking Coal
NO
80.4
113.0
124.7
112.8
70.0
160.3
Industrial Other Coal
7.6
11.0
9.5
9.5
9.5
9.5
9.5
Natural Gas to Chemical Plants
282.4
260.9
588.0
676.4
667.6
663.3
667.3
Asphalt & Road Oil
1,170.2
1,323.2
849.2
792.8
843.9
832.3
898.1
HGLa
1,218.0
1,610.1
2,193.7
2,506.9
2,550.7
2,658.0
2,819.6
Lubricants
186.3
160.2
124.9
122.0
118.3
111.1
113.9
Natural Gasolineb
117.5
95.4
81.7
105.3
155.0
163.7
202.4
Naphtha (<401 °F)
327.0
679.5
413.0
421.2
369.5
329.4
331.1
Other Oil (>401 °F)
663.6
499.5
242.9
219.1
212.1
195.6
196.3
Still Gas
36.7
67.7
163.8
166.9
158.7
145.4
152.8
Petroleum Coke
29.1
104.2
NO
NO
NO
NO
NO
Special Naphtha
101.1
60.9
95.3
87.0
89.5
80.8
76.1
Distillate Fuel Oil
7.0
16.0
5.8
5.8
5.8
5.8
5.8
Waxes
33.3
31.4
10.2
12.4
10.4
9.2
11.8
Miscellaneous Products
137.8
112.8
198.8
198.0
180.2
170.7
170.8
Transportation
176.0
151.3
142.0
137.0
131.3
115.6
118.6
Lubricants
176.0
151.3
142.0
137.0
131.3
115.6
118.6
U.S. Territories
50.8
114.9
3.5
3.6
3.6
3.6
3.6
Lubricants
0.7
4.6
1.0
1.0
1.0
1.0
1.0
Other Petroleum (Misc. Prod.)
50.1
110.3
2.4
2.5
2.6
2.6
2.6
Total
4,544.6
5,379.4
5,235.3
5,588.5
5,619.1
5,564.0
5,938.1
NO (Not Occurring)
a Excludes natural gasoline.
b Formerly referred to as "Pentanes Plus." This source has been adjusted and is reported separately from HGL to align with
historic data and revised EIA terminology.
Energy 3-53
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Table 3-24: 2021 Adjusted Non-Energy Use Fossil Fuel Consumption, Storage, and Emissions
Adjusted
Non-Energy Carbon Content
Potential
Storage
Carbon
Carbon
Carbon
Use3
Coefficient
Carbon
Factor
Stored
Emissions
Emissions
Sector/Fuel Type
(TBtu)
(MMT C/QBtu)
(MMT C)
(MMT C)
(MMT C) (MMTCOz Eq.)
Industry
5,815.9
NA
100.5
NA
63.7
36.8
135.0
Industrial Coking Coal
160.3
25.60
4.1
0.10
0.4
3.7
13.5
Industrial Other Coal
9.5
26.10
0.2
0.59
0.1
0.1
0.4
Natural Gas to
Chemical Plants
667.3
14.47
9.6
0.59
5.7
3.9
14.4
Asphalt & Road Oil
898.1
20.55
18.5
1.00
18.4
0.1
0.3
HGLb
2,819.6
16.83
47.4
0.59
28.0
19.4
71.1
Lubricants
113.9
20.20
2.3
0.09
0.2
2.1
7.7
Natural Gasolinec
202.4
18.24
3.7
0.59
2.2
1.5
5.5
Naphtha (<401° F)
331.1
18.55
6.1
0.59
3.6
2.5
9.2
Other Oil (>401° F)
196.3
20.17
4.0
0.59
2.3
1.6
5.9
Still Gas
152.8
17.51
2.7
0.59
1.6
1.1
4.0
Petroleum Coke
NO
27.85
NO
0.30
NO
NO
NO
Special Naphtha
76.1
19.74
1.5
0.59
0.9
0.6
2.3
Distillate Fuel Oil
5.8
20.22
0.1
0.50
0.1
0.1
0.2
Waxes
11.8
19.80
0.2
0.58
0.1
0.1
0.4
Miscellaneous
Products
170.8
NO
NO
NO
NO
NO
NO
Transportation
118.6
NA
2.4
NA
0.2
2.2
8.0
Lubricants
118.6
20.20
2.4
0.09
0.2
2.2
8.0
U.S. Territories
3.6
NA
0.1
NA
+
0.1
0.2
Lubricants
1.0
20.20
+
0.09
+
+
0.1
Other Petroleum
(Misc. Prod.)
2.6
20.00
0.1
0.10
+
+
0.2
Total
5,938.1
103.0
63.9
39.1
143.2
+ Does not exceed 0.05 TBtu, MMT C, or MMT C02 Eq.
NA (Not Applicable)
NO (Not Occurring)
a To avoid double counting, net exports have been deducted.
b Excludes natural gasoline.
c Formerly referred to as "Pentanes Plus." This source has been adjusted and is reported separately from HGL to align with
historic data and revised EIA terminology.
Note: Totals may not sum due to independent rounding.
Lastly, emissions were estimated by subtracting the C stored from the potential emissions (see Table 3-22). More
detail on the methodology for calculating storage and emissions from each of these sources is provided in Annex
2.3.
Where storage factors were calculated specifically for the United States, data were obtained on (1) products such
as asphalt, plastics, synthetic rubber, synthetic fibers, cleansers (soaps and detergents), pesticides, food additives,
antifreeze and deicers (glycols), and silicones; and (2) industrial releases including energy recovery (waste gas from
chemicals), Toxics Release Inventory (TRI) releases, hazardous waste incineration, and volatile organic compound,
solvent, and non-combustion CO emissions. Data were taken from a variety of industry sources, government
reports, and expert communications. Sources include EPA reports and databases such as compilations of air
emission factors (EPA 2001), National Emissions Inventory (NEI) Air Pollutant Emissions Trends Data (EPA 2022),
Toxics Release Inventory, 1998 (EPA 2000b), Biennial Reporting System (EPA 2000a, 2009), Resource Conservation
and Recovery Act Information System (EPA 2013b, 2015, 2016b, 2018b, 2021), pesticide sales and use estimates
(EPA 1998,1999, 2002, 2004, 2011, 2017), and the Chemical Data Access Tool (EPA 2014b); the EIA Manufacturer's
Energy Consumption Survey (MECS) (EIA 1994,1997, 2001, 2005, 2010, 2013, 2017, 2021); the National
3-54 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Petrochemical & Refiners Association (NPRA 2002); the U.S. Census Bureau (1999, 2004, 2009, 2014, 2021); Bank
of Canada (2012, 2013, 2014, 2016, 2017, 2018, 2019, 2020, 2021, 2022); Financial Planning Association (2006);
INEGI (2006); the United States International Trade Commission (2022); Gosselin, Smith, and Hodge (1984); EPA's
Municipal Solid Waste (MSW) Facts and Figures (EPA 2013, 2014a, 2016a, 2018a, 2019); the U.S. Tire
Manufacturers Association (USTMA2012, 2013, 2014, 2016, 2018, 2020, 2022); the International Institute of
Synthetic Rubber Products (IISRP 2000, 2003); the Fiber Economics Bureau (FEB 2001, 2003, 2005, 2007, 2009,
2010, 2011, 2012, 2013); the Independent Chemical Information Service (ICIS 2008, 2016); the EPA Chemical Data
Access Tool (CDAT) (EPA 2014b); the American Chemistry Council (ACC 2003 through 2011, 2013, 2014, 2015,
2016, 2017, 2018, 2019, 2020, 2021, 2022a); the Guide to the Business of Chemistry (ACC 2022b); and the
Chemistry Industry Association of Canada (CIAC 2022). Specific data sources are listed in full detail in Annex 2.3.
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2021 as discussed below.
Box 3-5: Reporting of Lubricants, Waxes, and Asphalt and Road Oil Product Use in Energy Sector
IPCC (2006) provides methodological guidance to estimate emissions from the first use of fossil fuels as a
product for primary purposes other than combustion for energy purposes (including lubricants, paraffin waxes,
bitumen / asphalt, and solvents) under the IPPU sector.63 In this Inventory, C storage and C emissions from
product use of lubricants, waxes, and asphalt and road oil are reported under the Energy sector in the Carbon
Emitted from Non-Energy Uses of Fossil Fuels source category (CRF Source Category 1A5).64
The emissions are reported in the Energy sector, as opposed to the IPPU sector, to reflect national
circumstances in its choice of methodology and to increase transparency of this source category's unique
country-specific data sources and methodology. Although emissions from these non-energy uses are reported in
the Energy chapter the methodologies used to determine emissions are compatible with the 2006 IPCC
Guidelines. The country-specific methodology used for the Carbon Emitted from Non-Energy Uses of Fossil Fuels
source category is based on a carbon balance (i.e., C inputs-outputs) calculation of the aggregate amount of
fossil fuels used for non-energy uses, including inputs of lubricants, waxes, asphalt and road oil (see Table 3-24).
For those inputs, U.S. country-specific data on C stocks and flows are used to develop carbon storage factors,
which are calculated as the ratio of the C stored by the fossil fuel non-energy products to the total C content of
the fuel consumed, taking into account losses in the production process and during product use.65 The country-
specific methodology to reflect national circumstances starts with the aggregate amount of fossil fuels used for
non-energy uses and applies a C balance calculation, breaking out the C emissions from non-energy use of
lubricants, waxes, and asphalt and road oil. The emissions are reported under the Energy chapter to improve
transparency, report a more complete carbon balance and to avoid double counting. Due to U.S. national
circumstances, reporting these C emissions separately under IPPU would involve making artificial adjustments
to allocate both the C inputs and C outputs of the non-energy use C balance. For example, only the emissions
from the first use of lubricants and waxes are to be reported under the IPPU sector, emissions from use of
lubricants in 2-stroke engines and emissions from secondary use of lubricants and waxes in waste incineration
with energy recovery are to be reported under the Energy sector. Reporting these non-energy use emissions
from only first use of lubricants and waxes under IPPU would involve making artificial adjustments to the non-
energy use C carbon balance and could potentially result in double counting of emissions. These artificial
adjustments would also be required for asphalt and road oil and solvents (which are captured as part of
petrochemical feedstock emissions) and could also potentially result in double counting of emissions. To avoid
63 See for example Volume 3: Industrial Processes and Product Use, and Chapter 5: Non-Energy Products from Fuels and
Solvent Use of the 2006 IPCC Guidelines for National Greenhouse Gas Inventories (IPCC 2006).
64 Non-methane volatile organic compound (NMVOC) emissions from solvent use are reported separately in the IPPU sector,
following Chapter 5 of the 2006 IPCC Guidelines.
65 Data and calculations for lubricants and waxes and asphalt and road oil are in Annex 2.3 - Methodology for Estimating
Carbon Emitted from Non-Energy Uses of Fossil Fuels.
Energy 3-55
-------
presenting an incomplete C balance and a less transparent approach for the Carbon Emitted from Non-Energy
Uses of Fossil Fuels source category calculation, the entire calculation of C storage and C emissions is therefore
conducted in the Non-Energy Uses of Fossil Fuels category calculation methodology, and both the C storage and
C emissions for lubricants, waxes, and asphalt and road oil are reported under the Energy sector.
However, emissions from non-energy uses of fossil fuels as feedstocks or reducing agents (e.g., petrochemical
production, aluminum production, titanium dioxide, and zinc production) are reported in the IPPU chapter,
unless otherwise noted due to specific national circumstances.
1
2 Uncertainty
3 An uncertainty analysis was conducted to quantify the uncertainty surrounding the estimates of emissions and
4 storage factors from non-energy uses. This analysis, performed using @RISK software and the IPCC-recommended
5 Approach 2 methodology (Monte Carlo Stochastic Simulation technique), provides for the specification of
6 probability density functions for key variables within a computational structure that mirrors the calculation of the
7 inventory estimate. The results presented below provide the 95 percent confidence interval, the range of values
8 within which emissions are likely to fall, for this source category.
9 As noted above, the non-energy use analysis is based on U.S.-specific storage factors for (1) feedstock materials
10 (natural gas, HGL, natural gasoline, naphthas, other oils, still gas, special naphthas, and other industrial coal), (2)
11 asphalt, (3) lubricants, and (4) waxes. For the remaining fuel types (the "other" category in Table 3-23 and Table
12 3-24) the storage factors were taken directly from IPCC (2006), where available, and otherwise assumptions were
13 made based on the potential fate of carbon in the respective NEU products. To characterize uncertainty, five
14 separate analyses were conducted, corresponding to each of the five categories. In all cases, statistical analyses or
15 expert judgments of uncertainty were not available directly from the information sources for all the activity
16 variables; thus, uncertainty estimates were determined using assumptions based on source category knowledge.
17 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3-25 (emissions) and Table
18 3-26 (storage factors). Carbon emitted from non-energy uses of fossil fuels in 2021 was estimated to be between
19 84.0 and 205.5 MMT CO2 Eq. at a 95 percent confidence level. This indicates a range of 41 percent below to 43
20 percent above the 2021 emission estimate of 143.2 MMT CO2 Eq. The uncertainty in the emission estimates is a
21 function of uncertainty in both the quantity of fuel used for non-energy purposes and the storage factor.
22 Table 3-25: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Non-
23 Energy Uses of Fossil Fuels (MMT CO2 Eq. and Percent)
2021 Emission Estimate Uncertainty Range Relative to Emission Estimate3
Source Gas
(MMT CO? Eq.) (MMT C02 Eq.) (%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Feedstocks
C02
112.9
59.3
178.6
-48%
58%
Asphalt
C02
0.3
0.1
0.7
-58%
+125%
Lubricants
C02
15.7
13.0
18.2
-17%
+16%
Waxes
C02
0.4
0.3
0.7
-24%
+83%
Other
C02
13.9
2.5
16.2
-82%
+16%
Total
C02
143.2
84.0
205.5
-41%
+43%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence
interval.
Note: Totals may not sum due to independent rounding.
3-56 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Table 3-26: Approach 2 Quantitative Uncertainty Estimates for Storage Factors of Non-
Energy Uses of Fossil Fuels (Percent)
2021 Storage Factor Uncertainty Range Relative to Emission Estimate3
Source Gas
(%) (%) (%, Relative)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Feedstocks
C02
59.1%
47.5%
72.2%
-20%
+22%
Asphalt
C02
99.6%
99.0%
99.8%
-0.5%
+0.3%
Lubricants
C02
9.2%
4.0%
17.4%
-57%
+90%
Waxes
C02
57.8%
47.3%
67.5%
-18%
+17%
Other
C02
11.1%
6.4%
83.3%
-42%
+650%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent
confidence interval, as a percentage of the inventory value (also expressed in percent terms).
As shown in Table 3-26, waxes and asphalt contribute least to overall storage factor uncertainty on a percentage
basis. Although the feedstocks category—the largest use category in terms of total carbon flows—also appears to
have relatively tight confidence limits, this is to some extent an artifact of the way the uncertainty analysis was
structured. As discussed in Annex 2.3, the storage factor for feedstocks is based on an analysis of six fates that
result in long-term storage (e.g., plastics production), and eleven that result in emissions (e.g., volatile organic
compound emissions). Rather than modeling the total uncertainty around all of these fate processes, the current
analysis addresses only the storage fates, and assumes that all C that is not stored is emitted. As the production
statistics that drive the storage values are relatively well-characterized, this approach yields a result that is
probably biased toward understating uncertainty.
As is the case with the other uncertainty analyses discussed throughout this document, the uncertainty results
above address only those factors that can be readily quantified. More details on the uncertainty analysis are
provided in Annex 2.3.
QA/QC and Verification
In order to ensure the quality of the emission estimates from non-energy uses of fossil fuels, general (IPCC Tier 1)
and category-specific (Tier 2) Quality Assurance/Quality Control (QA/QC) procedures were implemented consistent
with the U.S. Inventory QA/QC plan outlined in Annex 8. This effort included a general analysis, as well as portions
of a category specific analysis for non-energy uses involving petrochemical feedstocks and for imports and exports.
The Tier 2 procedures that were implemented involved checks specifically focusing on the activity data and
methodology for estimating the fate of C (in terms of storage and emissions) across the various end-uses of fossil
C. Emission and storage totals for the different subcategories were compared, and trends across the time series
were analyzed to determine whether any corrective actions were needed. Corrective actions were taken to rectify
minor errors and to improve the transparency of the calculations, facilitating future QA/QC.
For petrochemical import and export data, special attention was paid to NAICS numbers and titles to verify that
none had changed or been removed. Import and export totals were compared with 2020 totals as well as their
trends across the time series.
It is important to ensure no double counting of emissions between fuel combustion, non-energy use of fuels and
industrial process emissions. For petrochemical feedstock production, our review of the categories suggests this is
not a significant issue since the non-energy use industrial release data includes different categories of sources and
sectors than those included in the Industrial Processes and Product Use (IPPU) emissions category for
petrochemicals. Further data integration is not available at his time because feedstock data from the EIA used to
estimate non-energy uses of fuels are aggregated by fuel type, rather than disaggregated by both fuel type and
particular industries. Also, GHGRP-reported data on quantities of fuel consumed as feedstocks by petrochemical
producers are unable to be used due to the data failing GHGRP CBI aggregation criteria.
Energy 3-57
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Recalculations Discussion
Several updates to activity data factors lead to recalculations of previous year results. The major updates are as
follows:
• ACC (2022b) updated polyester, polyolefin and nylon fiber, ethylene glycol, maleic anhydride, adipic acid,
and acetic acid production in 2020, which resulted in a slight decrease in emissions relative to the
previous Inventory.
• U.S. International Trade Commission (2022) updated historical import and export data from 1996 to 2020,
resulting in fewer net exports relative to the previous Inventory.
• Updates to the petrochemical feedstock production and stocks led to an increase to the annually variable
storage factor from 1996 to 2020 for feedstocks, leading to less carbon emitted and a decrease in
emissions, most notably from HGL
• CIAC (2022) revised shipments for years 2017 to 2020, which resulted in a slight increase in emissions
from plastics from 2017 to 2020.
Overall, these changes resulted in an average annual decrease of 0.2 MMT CO2 Eq. (0.2 percent) in carbon
emissions from non-energy uses of fossil fuels for the period 1990 through 2020, relative to the previous
Inventory.
Planned Improvements
There are several future improvements planned:
• More accurate accounting of C in petrochemical feedstocks. EPA has worked with EIA to determine the
cause of input/output discrepancies in the C mass balance contained within the NEU model. In the future,
two strategies to reduce or eliminate this discrepancy will continue to be pursued as part of quality
control procedures. First, accounting of C in imports and exports will be improved. The import/export
adjustment methodology will be examined to ensure that net exports of intermediaries such as ethylene
and propylene are fully accounted for. Second, the use of top-down C input calculation in estimating
emissions will be reconsidered. Alternative approaches that rely more substantially on the bottom-up C
output calculation will be considered instead.
• Improving the uncertainty analysis. Most of the input parameter distributions are based on professional
judgment rather than rigorous statistical characterizations of uncertainty.
• Better characterizing flows of fossil C. Additional fates may be researched, including the fossil C load in
organic chemical wastewaters, plasticizers, adhesives, films, paints, and coatings. There is also a need to
further clarify the treatment of fuel additives and backflows (especially methyl tert-butyl ether, MTBE).
• Reviewing the trends in fossil fuel consumption for non-energy uses. Annual consumption for several fuel
types is highly variable across the time series, including industrial coking coal and other petroleum. A
better understanding of these trends will be pursued to identify any mischaracterized or misreported fuel
consumption for non-energy uses.
• Updating the average C content of solvents was researched, since the entire time series depends on one
year's worth of solvent composition data. The data on C emissions from solvents that were readily
available do not provide composition data for all categories of solvent emissions and also have conflicting
definitions for volatile organic compounds, the source of emissive C in solvents. Additional sources of
solvents data will be investigated in order to update the C content assumptions.
• Updating the average C content of cleansers (soaps and detergents) was researched; although production
and consumption data for cleansers are published every 5 years by the Census Bureau, the composition (C
content) of cleansers has not been recently updated. Recently available composition data sources may
3-58 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
facilitate updating the average C content for this category.
• Revising the methodology for consumption, production, and C content of plastics was researched;
because of recent changes to the type of data publicly available for plastics, the NEU model for plastics
applies data obtained from personal communications. Potential revisions to the plastics methodology to
account for the recent changes in published data will be investigated.
• Although U.S.-specific storage factors have been developed for feedstocks, asphalt, lubricants, and waxes,
default values from IPCC are still used for two of the non-energy fuel types (industrial coking coal,
distillate oil), and broad assumptions are being used for miscellaneous products and other petroleum.
Over the long term, there are plans to improve these storage factors by analyzing C fate similar to those
described in Annex 2.3 or deferring to more updated default storage factors from IPCC where available.
• Reviewing the storage of carbon black across various sectors in the Inventory; in particular, the carbon
black abraded and stored in tires.
• Assess the current method and/or identify new data sources (e.g., EIA) for estimating emissions from
ammonia/fertilizer use of natural gas.
• Investigate EIA NEU and MECS data to update, as needed, adjustments made for ammonia production
and "natural gas to chemical plants, other uses" and "natural gas to other" non-energy uses, including
iron and steel production, in energy uses and IPPU.
3.3 Incineration of Waste (CRF Source
Category 1A5)
Combustion is used to manage about 7 to 19 percent of the solid wastes generated in the United States,
depending on the source of the estimate and the scope of materials included in the definition of solid waste (EPA
2000; EPA 2020; Goldstein and Madtes 2001; Kaufman et al. 2004; Simmons et al. 2006; van Haaren et al. 2010). In
the context of this section, waste includes all municipal solid waste (MSW) as well as scrap tires. In the United
States, combustion of MSW tends to occur at waste-to-energy facilities or industrial facilities where useful energy
is recovered, and thus emissions from waste combustion are accounted for in the Energy chapter. Similarly, scrap
tires are combusted for energy recovery in industrial and utility boilers, pulp and paper mills, and cement kilns.
Combustion of waste results in conversion of the organic inputs to CO2. According to the 2006 IPCC Guidelines,
when the CO2 emitted is of fossil origin, it is counted as a net anthropogenic emission of CO2 to the atmosphere.
Thus, the emissions from waste combustion are calculated by estimating the quantity of waste combusted and the
fraction of the waste that is C derived from fossil sources.
Most of the organic materials in MSW are of biogenic origin (e.g., paper, yard trimmings), and have their net C
flows accounted for under the Land Use, Land-Use Change, and Forestry chapter. However, some components of
MSW and scrap tires—plastics, synthetic rubber, synthetic fibers, and carbon black—are of fossil origin. Plastics in
the U.S. waste stream are primarily in the form of containers, packaging, and durable goods. Rubber is found in
durable goods, such as carpets, and in non-durable goods, such as clothing and footwear. Fibers in MSW are
predominantly from clothing and home furnishings. As noted above, scrap tires (which contain synthetic rubber
and carbon black) are also considered a "non-hazardous" waste and are included in the waste combustion
estimate, though waste disposal practices for tires differ from MSW. Estimates on emissions from hazardous waste
combustion can be found in Annex 2.3 and are accounted for as part of the C mass balance for non-energy uses of
fossil fuels.
Approximately 27.8 million metric tons of MSW were combusted in 2021 (EPA 2021). Carbon dioxide emissions
from combustion of waste decreased 3.3 percent since 1990, to an estimated 12.5 MMT CO2 (12,476 kt) in 2021.
Emissions across the time series are shown in Table 3-27Error! Reference source not found, and Table 3-28Error!
Reference source not found..
Energy 3-59
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Waste combustion is also a source of Cm and N2O emissions (De Soete 1993; IPCC 2006). Methane emissions from
the combustion of waste were estimated to be less than 0.05 MMT CO2 Eq. (less than 0.05 kt CH4) in 2021 and
have remained steady since 1990. Nitrous oxide emissions from the combustion of waste were estimated to be 0.4
MMT CO2 Eq. (1.3 kt N2O) in 2021 and have decreased by 13 percent since 1990. This decrease is driven by the
decrease in total MSW combusted.
Table 3-27: CO2, ChU, and N2O Emissions from the Combustion of Waste (MMT CO2 Eq.)
Gas 1990 2005 2017 2018 2019 2020 2021
C02 12.9 13.3 13.2 13.3 12.9 12.9 12.5
CH4 + + + + + + +
N2Q 04 03 OA OA OA 03 0.4
Total 133 13^6 13.5 13.7 13.3 13.3 12.8
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Table 3-28: CO2, ChU, and N2O Emissions from the Combustion of Waste (kt)
Gas 1990 2005 2017 2018 2019 2020 2021
C02 12,900 13,254 13,161 13,339 12,948 12,921 12,476
CH4 + + + + + + +
N2Q 2 1 1 1 1 1 1_
+ Does not exceed 0.05 kt.
Methodology and Time-Series Consistency
Municipal Solid Waste Combustion
To determine both CO2 and non-CC>2 emissions from the combustion of waste, the tonnage of waste combusted
and an estimated emissions factor are needed. Emission estimates from the combustion of tires are discussed
separately. Data for total waste combusted was derived from BioCycle (van Haaren et al. 2010), EPA Facts and
Figures Report, Energy Recovery Council (ERC), EPA's Greenhouse Gas Reporting Program (GHGRP), and the U.S.
Energy Information Administration (EIA). Multiple sources were used to ensure a complete, quality dataset, as
each source encompasses a different timeframe.
EPA determined the MSW tonnages based on data availability and accuracy throughout the time series.
• 1990-2006: MSW combustion tonnages are from Biocycle combustion data. Tire combustion data from
the U.S. Tire Manufacturers Association (USTMA) are removed to arrive at MSW combusted without tires
• 2006-2010: MSW combustion tonnages are an average of Biocycle (with USTMA tire data tonnage
removed), U.S. EPA Facts and Figures, EIA, and Energy Recovery Council data (with USTMA tire data
tonnage removed).
• 2011-2021: MSW combustion tonnages are from EPA's GHGRP data.
Table 3-29 provides the estimated tons of MSW combusted including and excluding tires.
Table 3-29: Municipal Solid Waste Combusted (Short Tons)
Waste Combusted Waste Combusted
Year (excluding tires) (including tires)
1990 33,344,839 33,766,239
2005 26,486,414 28,631,054
3-60 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
2017 28,574,258 30,310,598
2018 29,162,364 30,853,949
2019 28,174,311 29,821,141
2020 27,586,271 29,106,686
2021 27,867,446 29,261,446
Sources: BioCycle, EPA Facts and Figures, ERC, GHGRP, EIA,
USTMA.
1 CO2 Emissions from MSW Excluding Scrap Tires
2 Fossil CO2 emission factors were calculated from EPA's GHGRP data for non-biogenic sources. Using GHGRP-
3 reported emissions for CHUand N2O and assumed emission factors, the tonnage of waste combusted, excluding
4 tires, was derived. Methane and N2O emissions and assumed emission factors were used to estimate the amount
5 of MSW combusted in terms of energy content. The energy content of MSW combusted was then converted into
6 tonnage based on assumed MSW heating value. Two estimates were generated (one for CH4 and one for N2O) and
7 the two were averaged together. Dividing fossil CO2 emissions from GHGRP FLIGHT data for MSW combustors by
8 this estimated tonnage yielded an annual CO2 emission factor. As this data was only available following 2011, all
9 years prior use an average of the emission factors from 2011 through 2015. See Annex 3.7 for more detail on how
10 MSW C factors were calculated.
11 Finally, CO2 emissions were calculated by multiplying the annual tonnage estimates, excluding tires, by the
12 calculated emissions factor. Calculated fossil CO2 emission factors are shown in Table 3-30.
13 Table 3-30: Calculated Fossil CO2 Content per Ton Waste Combusted (kg C02/Short Ton
14 Combusted)
1990 2005
2017
2018
2019
2020
2021
C02 Emission Factors
366 366
360
361
363
377
365
15 CO2 Emissions from Scrap Tires
16 Scrap tires contain several types of synthetic rubber, carbon black, and synthetic fibers. Each type of synthetic
17 rubber has a discrete C content, and carbon black is 100 percent C. For synthetic rubber and carbon black in scrap
18 tires, information was obtained biannually from U.S. Scrap Tire Management Summary for 2005 through 2021 data
19 (USTMA 2022). Information about scrap tire composition was taken from the Rubber Manufacturers' Association
20 internet site (USTMA 2012a). Emissions of CO2 were calculated based on the amount of scrap tires used for fuel
21 and the synthetic rubber and carbon black content of scrap tires. The mass of combusted material is multiplied by
22 its C content to calculate the total amount of carbon stored. More detail on the methodology for calculating
23 emissions from each of these waste combustion sources is provided in Annex 3.7. Table 3-31 provides CO2
24 emissions from combustion of waste tires.
25 Table 3-31: CO2 Emissions from Combustion of Tires (MMT CO2 Eq.)
1990
2005
2017
2018
2019
2020
2021
Synthetic Rubber
0.3
1.6
1.3
1.3
1.2
1.1
1.0
C Black
0.4
2.0
1.6
1.5
1.5
1.4
1.3
Total
0.7
3.6
2.9
2.8
2.7
2.5
2.3
Energy 3-61
-------
1 Non-CCh Emissions
2 Combustion of waste also results in emissions of Cl-Uand N2O. These emissions were calculated by multiplying the
3 total estimated mass of waste combusted, including tires, by the respective emission factors. The emission factors
4 for Cm and N2O emissions per quantity of MSW combusted are default emission factors for the default
5 continuously-fed stoker unit MSW combustion technology type and were taken from IPCC (2006).
6 Uncertainty
7 An Approach 2 Monte Carlo analysis was performed to determine the level of uncertainty surrounding the
8 estimates of CO2 emissions and N2O emissions from the incineration of waste (given the very low emissions for
9 CH4, no uncertainty estimate was derived). IPCC Approach 2 analysis allows the specification of probability density
10 functions for key variables within a computational structure that mirrors the calculation of the Inventory estimate.
11 Statistical analyses or expert judgments of uncertainty were not available directly from the information sources for
12 most variables; thus, uncertainty estimates for these variables were determined using assumptions based on
13 source category knowledge and the known uncertainty estimates for the waste generation variables.
14 The uncertainties in the waste incineration emission estimates arise from both the assumptions applied to the data
15 and from the quality of the data. Key factors include reported CO2 emissions; N2O and CH4emissions factors, and
16 tire synthetic rubber and black carbon contents. The highest levels of uncertainty surround the reported emissions
17 from GHGRP; the lowest levels of uncertainty surround variables that were determined by quantitative
18 measurements (e.g., combustion efficiency, C content of C black).
19 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3-32. Waste incineration
20 CO2 emissions in 2021 were estimated to be between 10.4 and 14.9 MMT CO2 Eq. at a 95 percent confidence level.
21 This indicates a range of 17 percent below to 19 percent above the 2021 emission estimate of 12.5 MMT CO2 Eq.
22 Also at a 95 percent confidence level, waste incineration N2O emissions in 2021 were estimated to be between 0.2
23 and 0.9 MMT CO2 Eq. This indicates a range of 54 percent below to 163 percent above the 2021 emission estimate
24 of 0.4 MMT CO2 Eq.
25 Table 3-32: Approach 2 Quantitative Uncertainty Estimates for CO2 and N2O from the
26 Incineration of Waste (MMT CO2 Eq. and Percent)
2021 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
Source
Gas
(MMT CO? Eq.)
(MMT CO?
Eq.)
(%)
Lower
Upper
Lower Upper
Bound
Bound
Bound Bound
Incineration of Waste
C02
12.5
10.4
14.9
-17% 19%
Incineration of Waste
N20
0.4
0.2
0.9
-54% 163%
a Range of emission estimates predicted by Monte Carlo Simulation for a 95 percent confidence interval.
27 QA/QC and Verification
28 In order to ensure the quality of the emission estimates from waste combustion, general (IPCC Tier 1) and
29 category-specific (Tier 2) Quality Assurance/Quality Control (QA/QC) procedures were implemented consistent
30 with the U.S. Inventory QA/QC plan outlined in Annex 8. The Tier 2 procedures that were implemented involved
31 checks specifically focusing on the activity data and specifically focused on the emission factor and activity data
32 sources and methodology used for estimating emissions from combustion of waste. Trends across the time series
33 were analyzed to determine whether any corrective actions were needed. Corrective actions were taken to rectify
34 minor errors in the use of activity data.
3-62 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Recalculations Discussion
2 For the current Inventory, CCh-equivalent emissions of CFU and N2O from waste incineration have been revised to
3 reflect the 100-year global warming potentials (GWPs) provided in the IPCC Fifth Assessment Report (AR5) (IPCC
4 2013). AR5 GWP values differ slightly from those presented in the IPCC Fourth Assessment Report (AR4), used in
5 the previous inventories (IPCC 2007). The AR5 GWPs have been applied across the entire time series for
6 consistency. Prior inventories used GWPs of 25 and 298 for CH4 and N2O, respectively. These values have been
7 updated to 28 and 265, respectively. Compared to the previous Inventory which applied 100-year GWP values
8 from AR4, the average annual change in CCh-equivalent CFU emissions was a 12 percent increase and the average
9 annual change in CCh-equivalent N2O emissions was an 11 percent decrease for the time series. As a result of the
10 change in methodology, total emissions across the timeseries changed by an average annual decrease of less than
11 0.05 MMT CO2 Eq. (0.3 percent) relative to emissions results calculated using the prior GWPs. Further discussion
12 on this update and the overall impacts of updating the Inventory GWP values to reflect the IPCC Fifth Assessment
13 Report can be found in Chapter 9, Recalculations and Improvements. All other recalculations described in this
14 section are compared using the prior GWPs.
15 Recalculations were performed for CO2 estimates from 1990 through 2010. In previous Inventories, for years prior
16 to 2011, fossil CO2 content per ton of waste was calculated based on the average of 2011 to the current year of
17 data. For this cycle the calculation was updated to be an average of estimates from 2011 - 2015. Earlier data is
18 assumed to more closely approximate the MSW composition for historic years. As a result of the change in
19 methodology, CO2 emissions in 1990 decreased by less than 0.05 MMT CO2 Eq. relative to the previous Inventory
20 and there was an average annual decrease by less than 0.05 MMT CO2 Eq. from 1990 through 2010.
21 Recalculations were performed on the estimate of combusted scrap tires in 2020. 2020 estimates for the scrap tire
22 market were previously proxied from the 2019 U.S. Scrap Tire Management Summary (USTMA 2020). The 2021
23 U.S. Scrap Tire Management Summary was released in October 2022, allowing 2020 estimates to now be
24 calculated by linear interpolation between 2019 and 2021 data. As a result of the change in methodology, CO2
25 emissions in 2020 decreased by 0.2 MT CO2 Eq. relative to the previous Inventory.
26 Planned Improvements
27 No planned improvements for waste combustion were identified.
28 3.4 Coal Mining (CRF Source Category
29 lBla)
30 Three types of coal mining-related activities release CH4 and CO2 to the atmosphere: underground mining, surface
31 mining, and post-mining (i.e., coal-handling) activities. While surface coal mines account for the majority of U.S.
32 coal production, underground coal mines contribute the largest share of fugitive CFU emissions (see Table 3-34 and
33 Table 3-35) due to the higher CFU content of coal in the deeper underground coal seams. In 2021,174
34 underground coal mines and 332 surface mines were operating in the United States (EIA 2022). In recent years, the
35 total number of active coal mines in the United States has declined. In 2021, the United States was the fourth
36 largest coal producer in the world (539 MMT), after China (3,685 MMT), India (771 MMT), and Indonesia (545
37 MMT) (IEA 2022).
38 Table 3-33: Coal Production (kt)
Year
Underground
Surface
Total
Number of Mines Production
Number of Mines Production
Number of Mines Production
1990 1,683 384,244 1,656 546,808 3,339 931,052
Energy 3-63
-------
2005 586 334,399 789 691,447 1,398 1,025,846
2017 237 247,778 434 454,301 671 702,080
2018 236 249,804 430 435,521 666 685,325
2019 226 242,557 432 397,750 658 640,307
2020 196 177,380 350 307,944 546 485,324
2021 174 200,122 332 323,142 506 523,264
1 Fugitive CH4 Emissions
2 Underground coal mines liberate Cm from ventilation systems and from degasification systems. Ventilation
3 systems pump air through the mine workings to dilute noxious gases and ensure worker safety; these systems can
4 exhaust significant amounts of Cm to the atmosphere in low concentrations. Degasification systems are wells
5 drilled from the surface or boreholes drilled inside the mine that remove large, often highly concentrated volumes
6 of Cm before, during, or after mining. Some mines recover and use CH4 generated from ventilation and
7 degasification systems, thereby reducing emissions to the atmosphere.
8 Surface coal mines liberate CH4 as the overburden is removed and the coal is exposed to the atmosphere. Methane
9 emissions are normally a function of coal rank (a classification related to the percentage of carbon in the coal) and
10 depth. Surface coal mines typically produce lower-rank coals and remove less than 250 feet of overburden, so their
11 level of emissions is much lower than from underground mines.
12 In addition, Cm is released during post-mining activities, as the coal is processed, transported, and stored for use.
13 Total Cm emissions in 2021 were estimated to be 1,595 kt (44.7 MMT CO2 Eq.), a decline of approximately 59
14 percent since 1990 (see Table 3-34 and Table 3-35). In 2021, underground mines accounted for approximately 74
15 percent of total emissions, surface mines accounted for 13 percent, and post-mining activities accounted for 13
16 percent. In 2021, total CH4 emissions from coal mining decreased by approximately 3 percent relative to the
17 previous year. Total coal production in 2021 increased by 8 percent compared to 2020. This resulted in an increase
18 of 7 percent in CH4 emissions from surface mining and post-mining activities in 2021. However, surface mining and
19 post-mining activities have a lower impact on total CH4 compared to underground mining (74 percent of total
20 emissions in 2021). The number of operating underground mines decreased in 2021 resulting in a slight decrease
21 in overall CH4 emissions (3 percent), compared to 2020. Additionally, the amount of CH4 recovered and used in
22 2021 decreased by less than 0.5 percent compared to 2020 levels.
23 Table 3-34: ChU Emissions from Coal Mining (MMT CO2 Eq.)
Activity
1990
2005
2017
2018
2019
2020
2021
Underground (UG) Mining
83.1
47.1
45.6
43.6
38.5
35.2
32.9
Liberated
90.5
66.9
65.1
64.6
56.3
53.5
51.2
Recovered & Used
(7.4)
(19.8)
(19.5)
(21.0)
(17.8)
(18.3)
(18.3)
Surface Mining
12.0
13.3
CO
7.8
7.2
5.4
5.7
Post-Mining (UG)
10.3
8.6
6.0
5.9
5.8
4.3
4.8
Post-Mining (Surface)
2.6
2.9
1.8
1.7
1.5
1.2
1.2
Total
108.1
71.8
61.4
59.1
53.0
46.2
44.7
Note: Parentheses indicate negative values.
24 Table 3-35: ChU Emissions from Coal Mining (kt)
Activity
1990
2005
2017
2018
2019
2020
2021
Underground (UG) Mining
2,968
1,682
1,627
1,557
1,376
1,257
1,176
Liberated
3,231
2,388
2,324
2,308
2,012
1,912
1,828
Recovered & Used
(263)
(706)
(697)
(751)
(636)
(654)
(652)
Surface Mining
430
475
290
280
255
194
205
3-64 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Post-Mining (UG)
Post-Mining (Surface)
368 306
93 103
213 212 206 155 170
63 61 55 42 44
Total
3,860 2,566
2,192 2,110 1,893 1,648 1,595
Note: Parentheses indicate negative values.
Methodology and Time-Series Consistency
EPA uses an IPCC Tier 3 method for estimating Cm emissions from underground coal mining and an IPCC Tier 2
method for estimating Cm emissions from surface mining and post-mining activities (for coal production from both
underground mines and surface mines). The methodology for estimating Cm emissions from coal mining consists
of two steps:
• Estimate Cm emissions from underground mines. These emissions have two sources: ventilation systems
and degasification systems. They are estimated using mine-specific data, then summed to determine total
Cm liberated. The CH4 recovered and used is then subtracted from this total, resulting in an estimate of
net emissions to the atmosphere.
• Estimate Cl-U emissions from surface mines and post-mining activities. Unlike the methodology for
underground mines, which uses mine-specific data, the methodology for estimating emissions from
surface mines and post-mining activities consists of multiplying basin-specific coal production by basin-
specific gas content and an emission factor.
Step 1: Estimate CH4 Liberated and CH4 Emitted from Underground Mines
Underground mines generate CH4 from ventilation systems and degasification systems. Some mines recover and
use the liberated CH4, thereby reducing emissions to the atmosphere. Total CH4 emitted from underground mines
equals the CH4 liberated from ventilation systems, plus the CH4 liberated from degasification systems, minus the
CH4 recovered and used.
Step 1.1: Estimate CH4 Liberated from Ventilation Systems
To estimate CH4 liberated from ventilation systems, EPA uses data collected through its Greenhouse Gas Reporting
Program (GHGRP)66 (Subpart FF, "Underground Coal Mines"), data provided by the U.S. Mine Safety and Health
Administration (MSHA) (MSHA 2022), 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 MT CO2 Eq.)—have been required to
report to EPA's GHGRP (EPA 2022).67 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.68
Since 2013, ventilation CH4 emission estimates have been calculated based on both quarterly GHGRP data
submitted by underground mines and on quarterly measurement data obtained directly from MSHA. Because not
all mines report under EPA's GHGRP, the emissions of the mines that do not report must be calculated using MSHA
data. The MSHA data also serves as a quality assurance tool for validating GHGRP data. For GHGRP data, reported
66 In implementing improvements and integrating data from EPA's GHGRP, EPA followed the latest guidance from the IPCC on
the use of facility-level data in national inventories (IPCC 2011).
67 Underground coal mines report to EPA under Subpart FF of the GHGRP (40 CFR Part 98). In 2021, 60 underground coal mines
reported to the program.
68 MSHA records coal mine CH4 readings with concentrations of greater than 50 ppm (parts per million) CH4. Readings below
this threshold are considered non-detectable.
Energy 3-65
-------
1 quarterly ventilation methane emissions (metric tons) are summed for each mine to develop mine-specific annual
2 ventilation emissions. For MSHA data, the average daily Cm emission rate for each mine is determined using the
3 Cm total for all data measurement events conducted during the calendar year and total duration of all data
4 measurement events (in days). The calculated average daily CH4 emission rate is then multiplied by 365 days to
5 estimate annual ventilation Cm emissions for the MSHA dataset.
6 Step 1.2: Estimate CH4 Liberated from Degasification Systems
1 Particularly gassy underground mines also use degasification systems (e.g., wells or boreholes) to remove CH4
8 before, during, or after mining. This CH4 can then be collected for use or vented to the atmosphere. Twenty mines
9 used degasification systems in 2021 and all of these mines reported the CH4 removed through these systems to
10 EPA's GHGRP under Subpart FF (EPA 2022). Based on the weekly measurements reported to EPA's GHGRP,
11 degasification data summaries for each mine are added to estimate the CH4 liberated from degasification systems.
12 Twelve of the 20 mines with degasification systems had operational CH4 recovery and use projects, including two
13 mines with two recovery and use projects each (see step 1.3 below).69
14 Degasification data reported to EPA's GHGRP by underground coal mines is the primary source of data used to
15 develop estimates of CH4 liberated from degasification systems. Data reported to EPA's GHGRP were used
16 exclusively to estimate CH4 liberated from degasification systems at 15 of the 20 mines that used degasification
17 systems in 2021. Data from state gas well production databases were used to supplement GHGRP degasification
18 data for the remaining five mines (DMME 2022; GSA 2022; WVGES 2022).
19 For pre-mining wells, cumulative degasification volumes that occur prior to the well being mined through are
20 attributed to the mine in the inventory year in which the well is mined through.70 EPA's GHGRP does not require
21 gas production from virgin coal seams (coalbed methane) to be reported by coal mines under Subpart FF.71 Most
22 pre-mining wells drilled from the surface are considered coalbed methane wells prior to mine-through and
23 associated CH4 emissions are reported under another subpart of the GHGRP (Subpart W, "Petroleum and Natural
24 Gas Systems"). As a result, GHGRP data must be supplemented to estimate cumulative degasification volumes that
25 occurred prior to well mine-through. There were four mines with degasification systems that include pre-mining
26 wells that were mined through in 2021. For all of these mines, GHGRP data were supplemented with historical
27 data from state gas well production databases (DMME 2022; ERG 2022; GSA 2022; WVGES 2022), as well as with
28 mine-specific information regarding the locations and dates on which the pre-mining wells were mined through
29 (JWR 2010; El Paso 2009; ERG 2022).
30 Step 1.3: Estimate CH4 Recovered from Ventilation and Degasification Systems, and Utilized or
31 Destroyed (Emissions Avoided)
32 Twelve mines had a total of fourteen CH4 recovery and use projects in place in 2021, including two mines that each
33 have two recovery and use projects. Thirteen of these projects involved degasification systems with one mine
34 having a ventilation air methane abatement project (VAM). Ten of these mines sold the recovered CH4 to a
35 pipeline, including one that also used CH4 to fuel a thermal coal dryer. One mine destroyed recovered CH4 using
36 flares. One mine destroyed the recovered CH4 (VAM) using regenerative thermal oxidation (RTO) without energy
37 recovery and using enclosed flares.
38 The CH4 recovered and used (or destroyed) at the twelve mines described above are estimated using the following
39 methods:
69 Several of the mines venting CH4from degasification systems use a small portion of the gas to fuel gob well blowers in
remote locations where electricity is not available. However, this CH4 use is not considered to be a formal recovery and use
project.
70 A well is "mined through" when coal mining development or the working face intersects the borehole or well.
71 This applies for pre-drainage in years prior to the well being mined through. Beginning with the year the well is mined
through, the annual volume of CH4 liberated from a pre-drainage well is reported under Subpart FF of EPA's GHGRP.
3-66 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
• EPA's GHGRP data was exclusively used to estimate the Cm recovered and used from six of the 12 mines
that deployed degasification systems in 2021. Based on weekly measurements, the GHGRP degasification
destruction data summaries for each mine are added together to estimate the Cm recovered and used
from degasification systems.
• State sales data were used to supplement GHGRP data to estimate CH4 recovered and used from five
mines that deployed degasification systems in 2021 (DMME 2022, ERG 2022, GSA 2022, and WVGES
2022). Four of these mines intersected pre-mining wells in 2021. Supplemental information is used for
these mines because estimating CH4 recovery and use from pre-mining wells requires additional data not
reported under Subpart FF of EPA's GHGRP (see discussion in step 1.2 above) to account for the emissions
avoided prior to the well being mined through. The supplemental data is obtained from state gas
production databases as well as mine-specific information on the location and timing of mined-through
pre-mining wells.
• For the single mine that employed VAM for CH4 recovery and use, the estimates of CH4 recovered and
used were obtained from the mine's offset verification statement (OVS) submitted to the California Air
Resources Board (CARB) (McElroy OVS 2022). This mine also reported CH4 reductions from flaring. GHGRP
data were used to estimate CH4 recovered and flared in 2021.
Step 2: Estimate CH4 Emitted from Surface Mines 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 the Energy Information Administration's
Annual Coal Report (EIA 2022) is multiplied by basin-specific CH4 contents (EPA 1996, 2005) and a 150 percent
emission factor (to account for CH4from over- and under-burden) to estimate CH4 emissions (King 1994; Saghafi
2013). For post-mining activities, basin-specific coal production is multiplied by basin-specific CH4 contents and a
mid-range 32.5 percent emission factor for CH4 desorption during coal transportation and storage (Creedy 1993).
Basin-specific in situ gas content data were compiled from AAPG (1984) and USBM (1986).
Fugitive C02 Emissions
Methane and CO2 are naturally occurring in coal seams and are collectively referred to as coal seam gas. These
gases remain trapped in the coal seam until coal is mined (i.e., coal seam is exposed and fractured during mining
operations). Fugitive CO2 emissions occur during underground coal mining, surface coal mining, and post-mining
activities. Methods and data to estimate fugitive CO2 emissions from underground and surface coal mining are
presented in the sections below. Fugitive CO2 emissions from post-mining activities were not estimated due to the
lack of an IPCC method and unavailability of data.
Total fugitive CO2 emissions in 2021 were estimated to be 2,456 kt (2.5 MMT CO2 Eq.), a decline of approximately
47 percent since 1990. In 2021, underground mines accounted for approximately 89 percent of total fugitive CO2
emissions. In 2021, total fugitive CO2 emissions from coal mining increased by approximately 12 percent relative to
the previous year. This increase was due to an increase in annual coal production.
Table 3-36: CO2 Emissions from Coal Mining (MMT CO2 Eq.)
Activity
1990
2005
2017
2018
2019
2020
2021
Underground (UG) Mining
4.2
3.6
2.8
2.8
2.7
1.9
2.2
Liberated
4.2
3.6
2.7
2.7
2.6
1.9
2.2
Recovered & Used
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Flaring
NO
NO
0.1
0.1
0.1
+
+
Surface Mining
0.4
0.6
0.4
0.4
0.3
0.2
0.3
Total
4.6
4.2
3.2
3.1
3.0
2.2
2.5
+ Does not exceed 0.05 MMT C02 Eq.
NO (Not Occurring)
Note: Parentheses indicate negative values.
Energy 3-67
-------
l Table 3-37: CO2 Emissions from Coal Mining (kt)
Activity
1990
2005
2017
2018
2019
2020
2021
Underground (UG) Mining
4,164
3,610
2,785
2,789
2,670
1,948
2,194
Liberated
4,171
3,630
2,690
2,712
2,633
1,926
2,173
Recovered & Used
(8)
(20)
(19)
(21)
(18)
(18)
(18)
Flaring
NO
NO
114
97
55
41
40
Surface Mining
443
560
368
353
322
249
262
Total
4,606
4,170
3,153
3,141
2,992
2,198
2,456
NO (Not Occurring)
Notes: Parentheses indicate negative values. Totals may not sum due to independent
rounding.
2 Methodology and Time-Series Consistency
3 EPA uses an IPCC Tier 1 method for estimating fugitive CO2 emissions from underground coal mining and surface
4 mining (IPCC 2019). IPCC methods and data to estimate fugitive CO2 emissions from post-mining activities (for both
5 underground and surface coal mining) are currently not available.
6 Step 1: Underground Mining
7 EPA used the following overarching IPCC equation to estimate fugitive CO2 emissions from underground coal mines
8 (IPCC 2019):
9 Equation 3-1: Estimating Fugitive CO2 Emissions from Underground Mines
10 Total C02 from Underground Mines
11 = C02 from underground mining — Amount of C02 in gas recovered
12 + C02 from methane flaring
13 Step 1.1: Estimate Fugitive CO2 Emissions from Underground Mining
14 EPA estimated fugitive CO2 emissions from underground mining using the IPCC Tier 1 emission factor (5.9
15 m3/metric ton) and annual coal production from underground mines (EIA 2022). The underground mining default
16 emission factor accounts for all the fugitive CO2 likely to be emitted from underground coal mining. Therefore, the
17 amount of CO2 from coal seam gas recovered and utilized for energy is subtracted from underground mining
18 estimates in Step 2, below. Under IPCC methods, the CO2 emissions from gas recovered and utilized for energy use
19 (e.g., injected into a natural gas pipeline) are reported under other sectors of the Inventory (e.g., stationary
20 combustion of fossil fuel or oil and natural gas systems) and not under the coal mining sector.
21 Step 1.2: Estimate Amount of CO2 In Coal Seam Gas Recovered for Energy Purposes
22 EPA estimated fugitive CO2 emissions from coal seam gas recovered and utilized for energy purposes by using the
23 IPCC Tier 1 default emission factor (19.57 metric tons CCh/million cubic meters of coal bed methane (CBM)
24 produced) and quantity of coal seam gas recovered and utilized. Data on annual quantity of coal seam gas
25 recovered and utilized are available from GHGRP and state sales data (GHGRP 2022; DMME 2022; ERG 2022; GSA
26 2022; WVGES 2022). The quantity of coal seam gas recovered and destroyed without energy recovery (e.g., VAM
27 projects) is deducted from the total coal seam gas recovered quantity (McElroy OVS 2022).
28 Step 1.3: Estimate Fugitive CO2 Emissions from Flaring
29 The IPCC method includes combustion CO2 emissions from gas recovered for non-energy uses (i.e., flaring, or
30 catalytic oxidation) under fugitive CO2 emission estimates for underground coal mining. In effect, these emissions,
31 though occurring through stationary combustion, are categorized as fugitive emissions in the Inventory. EPA
32 estimated CO2 emissions from methane flaring using the following equation:
3-68 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Equation 3-2: Estimating CO2 Emissions from Drained Methane Flared or Catalytically
2 Oxidized
3
4
5
C02 from flaring
= 0.98 x Volume of methane flared x Conversion Factor
x Stoichiometric Mass Factor
6 Currently there are three mines that report catalytic oxidation of recovered methane through flaring without
7 energy use. Annual data for 2021 were obtained from one mine's offset verification statement (OVS) submitted to
8 the California Air Resources Board (CARB) and the GHGRP for the remaining two mines (McElroy OVS 2022; GHGRP
9 2022).
10 Step 2: Surface Mining
11 EPA estimated fugitive CO2 emissions from surface mining using the IPCC Tier 1 emission factor (0.44 m3/metric
12 ton) and annual coal production from surface mines (EIA 2022).
14 A quantitative uncertainty analysis was conducted for the coal mining source category using the IPCC-
15 recommended Approach 2 uncertainty estimation methodology. Because emission estimates of Cm from
16 underground ventilation systems were based on actual measurement data from EPA's GHGRP or from MSHA,
17 uncertainty is relatively low. A degree of imprecision was introduced because the ventilation air measurements
18 used were not continuous but rather quarterly instantaneous readings that were used to determine the average
19 annual emission rates. Additionally, the measurement equipment used can be expected to have resulted in an
20 average of 10 percent overestimation of annual CH4 emissions (Mutmansky & Wang 2000). Equipment
21 measurement uncertainty is applied to GHGRP data.
22 Estimates of CH4 liberated and recovered by degasification systems are relatively certain for utilized CH4 because of
23 the availability of EPA's GHGRP data and state gas sales information. Many of the liberation and recovery
24 estimates use data on wells within 100 feet of a mined area. However, uncertainty exists concerning the radius of
25 influence of each well. The number of wells counted, and thus the liberated CH4 and avoided emissions, may vary if
26 the drainage area is found to be larger or smaller than estimated.
27 EPA's GHGRP requires weekly CH4 monitoring of mines that report degasification systems, and continuous CH4
28 monitoring is required for CH4 utilized on- or off-site. Since 2012, GHGRP data have been used to estimate CH4
29 emissions from vented degasification wells, reducing the uncertainty associated with prior MSHA estimates used
30 for this sub-source. Beginning in 2013, GHGRP data were also used for determining CH4 recovery and use at mines
31 without publicly available gas usage or sales records, which has reduced the uncertainty from previous estimation
32 methods that were based on information from coal industry contacts.
33 Surface mining and post-mining emissions are associated with considerably more uncertainty than underground
34 mines, because of the difficulty in developing accurate emission factors from field measurements. However, since
35 underground coal mining, as a general matter, results in significantly larger CH4 emissions due to production of
36 higher-rank coal and greater depth, and estimated emissions from underground mining constitute the majority of
37 estimated total coal mining CH4 emissions, the uncertainty associated with underground emissions is the primary
38 factor that determines overall uncertainty.
39 The major sources of uncertainty for estimates of fugitive CO2 emissions are the Tier 1 IPCC default emission
40 factors used for underground mining (-50 percent to +100 percent) and surface mining (-67 percent to +200
41 percent) (IPCC 2019). Additional sources of uncertainty for fugitive CO2 emission estimates include ElA's annual
42 coal production data and data used for gas recovery projects, such as GHGRP data, state gas sales data, and VAM
43 estimates for the single mine that operates an active VAM project. Uncertainty ranges for these additional data
44 sources are already available, as these are the same data sources used for CH4 emission estimates.
13 Uncertainty
Energy 3-69
-------
1 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3-38. Coal mining Cm
2 emissions in 2021 were estimated to be between 40.1 and 54.3 MMT CO2 Eq. at a 95 percent confidence level. This
3 indicates a range of 10.2 percent below to 21.5 percent above the 2021 emission estimate of 44.7 MMT CO2 Eq.
4 Coal mining fugitive CO2 emissions in 2021 were estimated to be between 0.8 and 4.3 MMT CO2 Eq. at a 95 percent
5 confidence level. This indicates a range of 67.6 percent below to 75.8 percent above the 2021 emission estimate of
6 2.5 MMT CO2 Eq.
7 Table 3-38: Approach 2 Quantitative Uncertainty Estimates for ChU and CO2 Emissions from
8 Coal Mining (MMT CO2 Eq. and Percent)
Source
Gas
2021 Emission Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Emission Estimate
(MMT C02 Eq.) (%)
a
Lower Upper
Lower
Upper
Bound Bound
Bound
Bound
Coal Mining
ch4
44.7
40.1 54.3
-10.2%
+21.5%
Coal Mining
C02
2.5
CO
'sT
00
O
-67.6%
+75.8%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
9 QA/QC and Verification
10 In order to ensure the quality of the emission estimates for coal mining, general (IPCC Tier 1) and category-specific
11 (Tier 2) Quality Assurance/Quality Control (QA/QC) procedures were implemented consistent with the U.S.
12 Inventory QA/QC plan outlined in Annex 8. The Tier 2 procedures that were implemented involved checks
13 specifically focusing on the activity data and reported emissions data used for estimating fugitive emissions from
14 coal mining. Trends across the time series were analyzed to determine whether any corrective actions were
15 needed.
16 Emission estimates for coal mining rely in large part on data reported by coal mines to EPA's GHGRP. EPA verifies
17 annual facility-level reports through a multi-step process to identify potential errors and ensure that data
18 submitted to EPA are accurate, complete, and consistent. All reports submitted to EPA are evaluated by electronic
19 validation and verification checks. If potential errors are identified, EPA will notify the reporter, who can resolve
20 the issue either by providing an acceptable response describing why the flagged issue is not an error or by
21 correcting the flagged issue and resubmitting their annual report. Additional QA/QC and verification procedures
22 occur for each GHGRP subpart. No QA/QC issues or errors were identified in the 2021 Subpart FF data.
23 Recalculations Discussion
24 State gas sales production values were updated for five mines, as part of normal updates. This update impacted
25 Cm emissions for 1998-2020. As a result of this update, CH4 emissions from degasification systems and CH4
26 emissions avoided increased across the time-series. Degasification CH4 emissions increased slightly by an average
27 of 0.4 percent and CH4 emissions avoided increased by an average of 1.6 percent over the 1998 to 2020 time
28 series, compared to the previous Inventory.
29 Fugitive CO2 emissions from flaring were recalculated for 2014 through 2020 as a result of adding two flaring
30 projects to the Inventory, as part of normal updates. One of the flaring projects was operational from 2014
31 onwards and the other one started in 2020. As a result of this update, flaring CO2 emissions for 2014 to 2020
32 increased by an average of 230 percent, compared to the previous Inventory, with 2020 emissions increasing by
33 277 percent. However, as flaring CO2 emissions only contribute 2 percent of total fugitive CO2 emissions, this
34 update resulted in a slight increase of overall fugitive CO2 emissions for 2014 to 2020 by an average of 2 percent,
35 compared to the previous Inventory.
36 In addition to the above-mentioned updates, for the current Inventory, estimates of C02-equivalent CH4 emissions
37 from coal mining have been revised to reflect the 100-year global warming potentials (GWPs) provided in the IPCC
38 Fifth Assessment Report (AR5) (IPCC 2013). AR5 GWP values differ slightly from those presented in the IPCC Fourth
3-70 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Assessment Report (AR4) (IPCC 2007) (used in the previous inventories). The AR5 GWPs have been applied across
the entire time series for consistency. The GWP of CFU increased from 25 to 28, leading to an overall increase in
CCh-equivalent Cm emissions. Compared to the previous Inventory which applied 100-year GWP values from AR4,
the average annual change in CC>2-equivalent CFU emissions was a 12 percent increase for each year of the time
series. Further discussion on this update and the overall impacts of updating the inventory GWPs to reflect the
IPCC Fifth Assessment Report can be found in Chapter 9, Recalculations and Improvements.
The net impact from the updates listed above was an average annual 12 percent increase in CFU emissions and an
average annual 0.4 percent increase in CO2 emissions for the time series.
Planned Improvements
EPA is assessing planned improvements for future reports, but at this time has no specific planned improvements
for estimating CH4 and CO2 emissions from underground and surface mining and CH4 emissions from post-mining.
3.5 Abandoned Underground Coal Mines
(CRF Source Category lBla)
Underground coal mines contribute the largest share of coal mine methane (CMM) emissions, with active
underground mines the leading source of underground emissions. However, mines also 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 through intrusion of groundwater or surface water into the 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 Cm 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. The following factors influence abandoned mine emissions:
• Time since abandonment;
• Gas content and adsorption characteristics of coal;
• CH4 flow capacity of the mine;
• Mine flooding;
• Presence of vent holes; and
• Mine seals.
Annual gross abandoned mine CH4 emissions ranged from 8.1 to 12.1 MMT CO2 Eq. from 1990 to 2021, varying, in
general, by less than 1 percent to approximately 19 percent from year to year. Fluctuations were due mainly to the
number of mines closed during a given year as well as the magnitude of the emissions from those mines when
active. Gross abandoned mine emissions peaked in 1996 (12.1 MMT CO2 Eq.) due to the large number of gassy
mine72 closures from 1994 to 1996 (72 gassy mines closed during the three-year period). In spite of this rapid rise,
abandoned mine emissions have been generally on the decline since 1996. Since 2002, there have been fewer than
twelve gassy mine closures each year. In 2021 there were two gassy mine closures. Gross abandoned mine
emissions decreased slightly from 9.4 MMT CO2 Eq. (335 kt CH4) in 2020 to 9.2 (330 kt CH4) MMT CO2 Eq. in 2021
72 A mine is considered a "gassy" mine if it emits more than 100 thousand cubic feet of CH4 per day (100 Mcfd).
Energy 3-71
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
(see Table 3-39 and Table 3-40). Gross emissions are reduced by Cm recovered and used at 47 mines, resulting in
net emissions in 2021 of 6.4 MMT CO2 Eq. (228 kt CH4).
Table 3-39: ChU Emissions from Abandoned Coal Mines (MMT CO2 Eq.)
Activity
1990
2005
2017
2018
2019
2020
2021
Abandoned Underground Mines
8.1
9.3
10.3
9.9
9.6
9.4
9.2
Recovered & Used
NO
(2.0)
(3.1)
(3.0)
(2.9)
(2.9)
(2.9)
Total
8.1
7.4
7.2
6.9
6.6
6.5
6.4
NO (Not Occurring)
Notes: Parentheses indicate negative values. Totals may not sum due to independent rounding.
Table 3-40: ChU Emissions from Abandoned Coal Mines (kt)
Activity
1990
2005
2017
2018
2019
2020
2021
Abandoned Underground Mines
288
334
367
355
341
335
330
Recovered & Used
NO
(70)
(109)
(107)
(104)
(103)
(103)
Total
288
264
257
247
237
232
228
NO (Not Occurring)
Notes: Parentheses indicate negative values. Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
Estimating Cm emissions from an abandoned coal mine requires predicting the emissions of a mine 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 or sealed. The CH4
emission rate before abandonment reflects the gas content of the coal, the rate of coal mining, and the flow
capacity of the mine in much the same way as the initial rate of a water-free conventional gas well reflects the gas
content of the producing formation and the flow capacity of the well. A well or a mine that produces gas from a
coal seam and the surrounding strata will produce less gas through time as the reservoir of gas is depleted.
Depletion of a reservoir will follow a predictable pattern depending on the interplay of a variety of natural physical
conditions imposed on the reservoir. The depletion of a reservoir is commonly modeled by mathematical
equations and mapped as a type curve. Type curves, which are referred to as decline curves, have been developed
for abandoned coal mines. Existing data on abandoned mine emissions through time, although sparse, appear to
fit the hyperbolic type of decline curve used in forecasting production from natural gas wells.
To estimate Cm emissions over time for a given abandoned mine, it is necessary to apply a decline function,
initiated upon abandonment, to that mine. In the analysis, mines were grouped by coal basin with the assumption
that they will generally have the same initial pressures, permeability, and isotherm. As CH4 leaves the system, the
reservoir pressure (Pr) declines as described by the isotherm's characteristics. The emission rate declines because
the mine pressure (Pw) is essentially constant at atmospheric pressure for a vented mine, and the productivity
index (PI), which is expressed as the flow rate per unit of pressure change, is essentially constant at the pressures
of interest (atmospheric to 30 psia). The CH4 flow rate is determined by the laws of gas flow through porous media,
such as Darcy's Law. A rate-time equation can be generated that can be used to predict future emissions. This
decline through time is hyperbolic in nature and can be empirically expressed as:
Equation 3-3: Decline Function to Estimate Venting Abandoned Mine Methane Emissions
q = qt (1 + bDit)(-1/h)
where,
q = Gas flow rate at time t in million cubic feet per day (mmcfd)
q, = Initial gas flow rate at time zero (t0), mmcfd
b = The hyperbolic exponent, dimensionless
Di = Initial decline rate, 1/year
3-72 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
t = Elapsed time from t0 (years)
This equation is applied to mines of various initial emission rates that have similar initial pressures, permeability,
and adsorption isotherms (EPA 2004).
The decline curves created to model the gas emission rate of coal mines must account for factors that decrease the
rate of emissions after mining activities cease, such as sealing and flooding. Based on field measurement data, it
was assumed that most U.S. mines prone to flooding will become completely flooded within eight years and
therefore will no longer have any measurable Cm emissions. Based on this assumption, an average decline rate for
flooded mines was established by fitting a decline curve to emissions from field measurements. An exponential
equation was developed from emissions data measured at eight abandoned mines known to be filling with water
located in two of the five basins. Using a least squares, curve-fitting algorithm, emissions data were matched to
the exponential equation shown below. For this analysis of flooded abandoned mines, there was not enough data
to establish basin-specific equations, as was done with the vented, non-flooding mines (EPA 2004). This decline
through time can be empirically expressed as:
Equation 3-4: Decline Function to Estimate Flooded Abandoned Mine Methane Emissions
where,
q = qie{-Dt)
q = Gas flow rate at time t in mmcfd
q, = Initial gas flow rate at time zero (t0), mmcfd
D = Decline rate, 1/year
t = Elapsed time from t0 (years)
Seals have an inhibiting effect on the rate of flow of Cm into the atmosphere compared to the flow rate that
would exist if the mine had an open vent. The total volume emitted will be the same, but emissions will occur over
a longer period of time. The methodology, therefore, treats the emissions prediction from a sealed mine similarly
to the emissions prediction from a vented mine, but uses a lower initial rate depending on the degree of sealing. A
computational fluid dynamics simulator was used with the conceptual abandoned mine model to predict the
decline curve for inhibited flow. The percent sealed is defined as 100 x (1 - [initial emissions from sealed mine /
emission rate at abandonment prior to sealing]). Significant differences are seen between 50 percent, 80 percent,
and 95 percent closure. These decline curves were therefore used as the high, middle, and low values for
emissions from sealed mines (EPA 2004).
For active coal mines, those mines producing over 100 thousand cubic feet per day (Mcfd) of Cm account for
about 98 percent of all CFU emissions. This same relationship is assumed for abandoned mines. It was determined
that the 530 abandoned mines closed since 1972 produced CFU emissions greater than 100 Mcfd when active.
Further, the status of 306 of the 530 mines (or 58 percent) is known to be either: 1) vented to the atmosphere; 2)
sealed to some degree (either earthen or concrete seals); or 3) flooded (enough to inhibit CFU flow to the
atmosphere). The remaining 42 percent of the mines whose status is unknown were placed in one of these three
categories by applying a probability distribution analysis based on the known status of other mines located in the
same coal basin (EPA 2004). Table 3-41 presents the count of mines by post-abandonment state, based on EPA's
probability distribution analysis.
Table 3-41: Number of Gassy Abandoned Mines Present in U.S. Basins in 2021, Grouped by
Class According to Post-Abandonment State
Total
Basin
Sealed
Vented
Flooded
Known
Unknown
Total Mines
Central Appl.
43
25
50
118
144
262
Illinois
35
3
14
52
31
83
Northern Appl.
48
23
15
86
39
125
Warrior Basin
0
0
16
16
0
16
Western Basins
28
4
2
34
10
44
Energy 3-73
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Total 154 55 97 306 224 530
Inputs to the decline equation require the average CFU emission rate prior to abandonment and the date of
abandonment. Generally, these data are available for mines abandoned after 1971; however, such data are largely
unknown for mines closed before 1972. Information that is readily available, such as coal production by state and
county, is helpful but does not provide enough data to directly employ the methodology used to calculate
emissions from mines abandoned before 1972. 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.
During the 1970s, 78 percent of Cm emissions from coal mining came from seventeen counties in seven states.
Mine closure dates were obtained for two states, Colorado and Illinois, for the hundred-year period extending
from 1900 through 1999. The data were used to establish a frequency of mine closure histogram (by decade) and
applied to the other five states with gassy mine closures. As a result, basin-specific decline curve equations were
applied to the 145 gassy coal mines estimated to have closed between 1920 and 1971 in the United States,
representing 78 percent of the emissions. State-specific, initial emission rates were used based on average coal
mine CFU emission rates during the 1970s (EPA 2004).
Abandoned mine emission estimates are based on all closed mines known to have active mine Cm ventilation
emission rates greater than 100 Mcfd at the time of abandonment. For example, for 1990 the analysis included
145 mines closed before 1972 and 258 mines closed between 1972 and 1990. Initial emission rates based on MSHA
reports, time of abandonment, and basin-specific decline curves influenced by a number of factors were used to
calculate annual emissions for each mine in the database (MSHA 2022). Coal mine degasification data are not
available for years prior to 1990, thus the initial emission rates used reflect only ventilation emissions for pre-1990
closures. Methane degasification amounts were added to the quantity of Cm vented to determine the total CFU
liberation rate for all mines that closed between 1992 and 2021. Since the sample of gassy mines described above
is assumed to account for 78 percent of the pre-1972 and 98 percent of the post-1971 abandoned mine emissions,
the modeled results were multiplied by 1.22 and 1.02, respectively, to account for all U.S. abandoned mine
emissions.
From 1993 through 2021, emission totals were downwardly adjusted to reflect Cm emissions avoided from those
abandoned mines with Cm recovery and use or destruction systems. The Inventory totals were not adjusted for
abandoned mine Cm emissions avoided from 1990 through 1992, because no data was reported for abandoned
coal mine Cm recovery and use or destruction projects during that time.
Uncertainty
A quantitative uncertainty analysis was conducted for the abandoned coal mine source category using the IPCC-
recommended Approach 2 uncertainty estimation methodology. The uncertainty analysis provides for the
specification of probability density functions for key variables within a computational structure that mirrors the
calculation of the Inventory estimate. The results provide the range within which, with 95 percent certainty,
emissions from this source category are likely to fall.
As discussed above, the parameters for which values must be estimated for each mine to predict its decline curve
are: 1) the coal's adsorption isotherm; 2) CFU flow capacity as expressed by permeability; and 3) pressure at
abandonment. Because these parameters are not available for each mine, a methodological approach to
estimating emissions was used that generates a probability distribution of potential outcomes based on the most
likely value and the probable range of values for each parameter. The range of values is not meant to capture the
extreme values, but rather values that represent the highest and lowest quartile of the cumulative probability
density function of each parameter. Once the low, mid, and high values are selected, they are applied to a
probability density function.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3-42. Annual abandoned
coal mine Cm emissions in 2021 were estimated to be between 5.0 and 7.7 MMT CO2 Eq. at a 95 percent
3-74 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 confidence level. This indicates a range of 22 percent below to 21 percent above the 2021 emission estimate of 6.4
2 MMT CO2 Eq. One of the reasons for the relatively narrow range is that mine-specific data is available for use in the
3 methodology for mines closed in 1972 and later years. Emissions from mines closed prior to 1972 have the largest
4 degree of uncertainty because no mine-specific Cm liberation rates at the time of abandonment exist.
5 Table 3-42: Approach 2 Quantitative Uncertainty Estimates for ChU Emissions from
6 Abandoned Underground Coal Mines (MMT CO2 Eq. and Percent)
Source Gas
2021 Emission Estimate
(MMT C02 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower
Bound
Upper
Bound
Lower Upper
Bound Bound
Abandoned Underground
CH4
Coal Mines
6.4
5.0
7.7
-21.7% +20.6%
a Range of emission estimates predicted by Monte Carlo Simulation for a 95 percent confidence interval.
7 QA/QC and Verification
8 In order to ensure the quality of the emission estimates for abandoned coal mines, general (IPCC Tier 1) and
9 category-specific (Tier 2) Quality Assurance/Quality Control (QA/QC) procedures were implemented consistent
10 with the U.S. Inventory QA/QC plan outlined in Annex 8. The Tier 2 procedures that were implemented involved
11 checks specifically focusing on the activity data and reported emissions data used for estimating emissions from
12 abandoned coal mines. Trends across the time series were analyzed to determine whether any corrective actions
13 were needed.
14 Recalculations Discussion
15 For the current Inventory, estimates of C02-equivalent Cm emissions from abandoned coal mines have been
16 revised to reflect the 100-year global warming potentials (GWPs) provided in the IPCC Fifth Assessment Report
17 (AR5) (IPCC 2013). AR5 GWP values differ slightly from those presented in the IPCC Fourth Assessment Report
18 (AR4) (IPCC 2007) (used in the previous inventories). The AR5 GWPs have been applied across the entire time
19 series for consistency. The GWP of Cm increased from 25 to 28, leading to an overall increase in C02-equivalent
20 Cm emissions. Compared to the previous Inventory which applied 100-year GWP values from AR4, the average
21 annual change in C02-equivalent CH4 emissions was 12 percent increase for each year of the time series. Further
22 discussion on this update and the overall impacts of updating the inventory GWPs to reflect the IPCC Fifth
23 Assessment Report can be found in Chapter 9, Recalculations and Improvements.
24 3.6 Petroleum Systems (CRF Source
25 Category lB2a)
26 Note that this draft of the Inventory does not yet incorporate updated activity data products for the following data
TJ inputs, due to a data base subscription lapse: oil well counts, wells drilled, wells completed, and production. Year
28 2020 values for activity data are used in place of year 2021. The Final Inventory (to be published April 2023) will
29 incorporate the latest activity data.
30 This IPCC category (lB2a) is for fugitive emissions from petroleum systems, which per IPCC guidelines include
31 emissions from leaks, venting, and flaring. Methane emissions from petroleum systems are primarily associated
32 with onshore and offshore crude oil exploration, production, transportation, and refining operations. During these
33 activities, CH4 is released to the atmosphere as emissions from leaks, venting (including emissions from operational
34 upsets), and flaring. Carbon dioxide emissions from petroleum systems are primarily associated with onshore and
Energy 3-75
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
offshore crude oil production and refining operations. Note, CO2 emissions in petroleum systems exclude all
combustion emissions (e.g., engine combustion) except for flaring CO2 emissions. All combustion CO2 emissions
(except for flaring) are accounted for in the fossil fuel combustion chapter (see Section 2). Emissions of N2O from
petroleum systems are primarily associated with flaring. Total greenhouse gas emissions (CH4, CO2, and N2O) from
petroleum systems in 2021 were 74.8 MMT CO2 Eq., an increase of 23 percent from 1990, primarily due to
increases in CO2 emissions. Total emissions increased by 10 percent from 2010 levels and have decreased by 10
percent since 2020. Total CO2 emissions from petroleum systems in 2021 were 24.67 MMT CO2 (24,667 kt CO2), 2.6
times higher than in 1990. Total CO2 emissions in 2021 were 1.8 times higher than in 2010 and 15 percent lower
than in 2020. Total CH4 emissions from petroleum systems in 2021 were 50.2 MMT CO2 Eq. (1,791 kt CH4), a
decrease of 2 percent from 1990. Since 2010, total Cm emissions decreased by 8 percent; and since 2020, Cm
emissions decreased by 8 percent. Total N2O emissions from petroleum systems in 2021 were 0.022 MMT CO2 Eq.
(0.082 kt N2O), 1.7 times higher than in 1990,1.2 times higher than in 2010, and 34 percent lower than in 2020.
Since 1990, U.S. oil production has increased by 46 percent. In 2021, U.S. oil production was 105 percent higher
than in 2010 and 1 percent lower than in 2020.
Each year, some estimates in the Inventory are recalculated with improved methods and/or data. These
improvements are implemented consistently across the entire Inventory's time series (i.e., 1990 to 2021) to ensure
that the trend is representative of changes in emissions levels. Recalculations in petroleum systems in this year's
Inventory include:
• Updates to oil and gas production volumes using the most recent data from the United States Energy
Information Administration (EIA)
• Recalculations due to Greenhouse Gas Reporting Program (GHGRP) submission revisions
• Recalculations due to methodological updates to four onshore production segment sources - pneumatic
controllers, equipment leaks, chemical injection pumps, and storage tanks.
• Recalculations due to updating the global warming potential (GWP) for Cm and N2O to use AR5 values.
Updated well counts and produced water volumes were not available for Public Review estimates, and 2021 data
were set equal to 2020. The latest data will be incorporated into the final Inventory.
The Recalculations Discussion section below provides more details on the updated methods.
Exploration. Exploration includes well drilling, testing, and completion. Exploration accounts for less than 0.5
percent of total CH4 emissions (including leaks, vents, and flaring) from petroleum systems in 2021. The
predominant sources of Cm emissions from exploration are hydraulically fractured oil well completions. Other
sources include well testing, well drilling, and well completions without hydraulic fracturing. Since 1990,
exploration CH4 emissions have decreased 96 percent, and while the number of hydraulically fractured wells
completed increased 64 percent, there were decreases in the fraction of such completions without reduced
emissions completions (RECs) or flaring. Emissions of CH4 from exploration were highest in 2012, over 60 times
higher than in 2021; and lowest in 2021. Emissions of CH4 from exploration decreased 52 percent from 2020 to
2021, due to a decrease in emissions from hydraulically fractured oil well completions without RECs, as well as due
to hydraulically fractured oil well completions with RECs and venting. Exploration accounts for 2 percent of total
CO2 emissions (including leaks, vents, and flaring) from petroleum systems in 2021. Emissions of CO2 from
exploration in 2021 were 28 percent higher than in 1990, and decreased by 44 percent from 2020, largely due to a
decrease in the number of hydraulically fractured oil well completions without RECS or flaring (by 36 percent from
2020). Emissions of CO2 from exploration were highest in 2014, over 8 times higher than in 2021. Exploration
accounts for 1 percent of total N2O emissions from petroleum systems in 2021. Emissions of N2O from exploration
in 2021 are 35 percent higher than in 1990, and 39 percent lower than in 2020, due to the above-mentioned
changes in hydraulically fractured oil well completions with flaring.
Production. Production accounts for 98 percent of total CH4 emissions (including leaks, vents, and flaring) from
petroleum systems in 2021. The predominant sources of emissions from production field operations are pneumatic
controllers, offshore oil platforms, equipment leaks, chemical injection pumps, gas engines, produced water, and
associated gas flaring. In 2021, these seven sources together accounted for 94 percent of the CH4 emissions from
3-76 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
production. Since 1990, Cm emissions from production have increased by 6 percent primarily due to increases in
emissions from pneumatic devices. Overall, production segment CFU emissions decreased by 8 percent from 2020
levels due primarily to lower pneumatic controller emissions. The number of high- and intermittent-bleed
pneumatic controllers decreased from 2020 to 2021 whereas, the number of low-bleed pneumatic controllers
increased from 2020 to 2021. Production emissions account for 81 percent of the total CO2 emissions (including
leaks, vents, and flaring) from petroleum systems in 2021. The principal sources of CO2 emissions are associated
gas flaring, miscellaneous production flaring, and oil tanks with flares. In 2021, these three sources together
accounted for 97 percent of the CO2 emissions from production. In 2021, CO2 emissions from production were 3.4
times higher than in 1990, due to increases in flaring emissions from associated gas flaring, miscellaneous
production flaring, and tanks. Overall, in 2021, production segment CO2 emissions decreased by 17 percent from
2020 levels primarily due to decreases in associated gas flaring and miscellaneous production flaring in the
Permian and Williston Basins. Production emissions accounted for 48 percent of the total N2O emissions from
petroleum systems in 2021. The principal sources of N2O emissions are associated gas flaring, oil tanks with flares,
miscellaneous production flaring, and offshore flaring. In 2021, N2O emissions from production were 115 percent
higher than in 1990 and were 51 percent lower than in 2020.
Crude Oil Transportation. Emissions from crude oil transportation account for a very small percentage of the total
emissions (including leaks, vents, and flaring) from petroleum systems. Crude oil transportation activities account
for 0.4 percent of total CH4 emissions from petroleum systems. Emissions from tanks, marine loading, and truck
loading operations account for 78 percent of CH4 emissions from crude oil transportation. Since 1990, CH4
emissions from transportation have increased by 21 percent. In 2021, CFU emissions from transportation
decreased by 3 percent from 2020 levels. Crude oil transportation activities account for less than 0.01 percent of
total CO2 emissions from petroleum systems. Emissions from tanks, marine loading, and truck loading operations
account for 78 percent of CO2 emissions from crude oil transportation.
Crude Oil Refining. Crude oil refining processes and systems account for 2 percent of total fugitive (including leaks,
vents, and flaring) CH4 emissions from petroleum systems. This low share is because most of the CFU in crude oil is
removed or escapes before the crude oil is delivered to the refineries. There is a negligible amount of CH4 in all
refined products. Within refineries, flaring accounts for 52 percent of the CH4 emissions, while delayed cokers,
uncontrolled blowdowns, and equipment leaks account for 16,13 and 9 percent, respectively. Fugitive CH4
emissions from refining of crude oil have increased by 12 percent since 1990, and decreased 5 percent from 2020;
however, like the transportation subcategory, this increase has had little effect on the overall emissions of CH4
from petroleum systems. Crude oil refining processes and systems account for 17 percent of total fugitive
(including leaks, vents, and flaring) CO2 emissions from petroleum systems. Of the total fugitive CO2 emissions
from refining, almost all (about 99 percent) of it comes from flaring.73 Since 1990, refinery fugitive CO2 emissions
increased by 28 percent and have decreased by less than 1 percent from the 2020 levels, due to a decrease in
flaring. Flaring occurring at crude oil refining processes and systems accounts for 51 percent of total fugitive N2O
emissions from petroleum systems. In 2021, refinery fugitive N2O emissions increased by 37 percent since 1990,
and decreased by less than 1 percent from 2020 levels.
Table 3-43: Total Greenhouse Gas Emissions (CO2, ChU, and N2O) from Petroleum Systems
(MMT COz Eq.)
Activity
1990
2005
2017
2018
2019
2020
2021
Exploration
4.7
6.3
2.3
3.8
2.9
1.2
0.6
Production
52.0
50.1
79.3
88.2
97.6
77.1
68.9
Transportation
0.2
0.1
0.2
0.2
0.3
0.2
0.2
Crude Refining
4.0
4.6
4.5
4.6
6.0
5.1
5.1
Total
60.8
61.2
86.4
96.8
106.8
83.6
74.8
73 Petroleum Systems includes fugitive emissions (leaks, venting, and flaring). In many industries, including petroleum
refineries, the largest source of onsite C02 emissions is often fossil fuel combustion, which is covered in Section 3.1 of this
chapter.
Energy 3-77
-------
l Table 3-44: Cm Emissions from Petroleum Systems (MMT CO2 Eq.)
Activity
1990
2005
2017
2018
2019
2020
2021
Exploration
4.3
5.9
0.5
0.6
0.5
0.3
0.2
Production
46.1
44.0
60.3
59.0
58.2
53.0
48.9
Pneumatic Controllers
21.3
23.3
38.1
35.3
24.8
31.7
28.4
Offshore Production
9.9
7.2
5.7
5.5
5.5
5.3
5.5
Equipment Leaks
2.3
2.9
3.3
3.7
3.9
3.2
3.3
Gas Engines
2.3
2.0
2.5
2.6
2.6
2.5
2.5
Produced Water
2.6
1.7
2.4
2.6
2.7
2.5
2.5
Chemical Injection Pumps
1.3
3.0
3.4
3.9
10.8
3.3
3.2
Assoc Gas Flaring
0.5
0.4
1.1
1.9
2.3
1.2
0.8
Other Sources
5.9
3.5
3.7
3.6
5.5
3.5
2.8
Crude Oil Transportation
0.2
0.1
0.2
0.2
0.3
0.2
0.2
Refining
0.7
0.8
0.9
0.8
1.0
0.9
0.8
Total
51.3
50.9
61.9
60.6
59.9
54.5
50.2
2 Table 3-45: ChU Emissions from Petroleum Systems (kt ChU)
Activity
1990
2005
2017
2018
2019
2020
2021
Exploration
154
211
17
20
16
12
6
Production
1,646
1,573
2,152
2,107
2,078
1,894
1,748
Pneumatic Controllers
760
833
1,362
1,260
886
1,131
1,015
Offshore Production
353
259
205
197
196
188
195
Equipment Leaks
82
102
120
132
138
115
117
Gas Engines
82
71
89
92
94
89
89
Produced Water
91
62
84
93
98
89
89
Chemical Injection Pumps
47
105
121
139
387
116
116
Assoc Gas Flaring
20
14
38
66
82
43
28
Other Sources
211
125
133
128
197
124
99
Crude Oil Transportation
7
5
8
8
9
8
8
Refining
26
30
33
30
36
31
30
Total
1,833
1,891
2,209
2,165
2,138
1,945
1,791
)ble 3-46: CO2 Emissions from Petroleum Systems (MMT CO2)
Activity
1990
2005
2017
2018
2019
2020
2021
Exploration
0.4
0.4
1.9
3.2
2.4
0.8
0.5
Production
5.9
6.1
19.0
29.2
39.4
24.0
20.0
Transportation
+
+
+
+
+
+
+
Crude Refining
3.3
3.7
3.6
3.7
5.0
4.2
4.2
Total
9.5
10.2
24.5
36.1
46.9
29.1
24.7
+ Does not exceed 0.05 MMT C02 Eq.
4 Table 3-47: CO2 Emissions from Petroleum Systems (kt CO2)
Activity
1990
2005
2017
2018
2019
2020
2021
Exploration
364
395
1,853
3,208
2,434
838
467
Production
5,869
6,097
19,025
29,187
39,429
24,000
19,985
Transportation
0.9
0.7
1.1
1.2
1.3
1.2
1.1
Crude Refining
3,284
3,728
3,582
3,706
5,009
4,242
4,214
Total
9,519
10,221
24,462
36,102
46,874
29,081
24,667
3-78 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
Table 3-48: N2O Emissions from Petroleum Systems (Metric Tons CO2 Eq.)
Activity
1990
2005
2017
2018
2019
2020
2021
Exploration
161
174
722
1,338
894
361
219
Production
4,907
5,465
13,450
25,638
26,522
21,665
10,539
Transportation
NE
NE
NE
NE
NE
NE
NE
Crude Refining
8,096
9,189
9,286
9,351
13,127
11,149
11,083
Total
13,164
14,827
23,458
36,327
40,542
33,175
21,841
NE (Not Estimated)
2 Table 3-49: N2O Emissions from Petroleum Systems (Metric Tons N2O)
Activity
1990
2005
2017
2018
2019
2020
2021
Exploration
0.6
0.7
2.7
5.0
3.4
1.4
0.8
Production
18.5
20.6
50.8
96.7
100.1
81.8
39.8
Transportation
NE
NE
NE
NE
NE
NE
NE
Crude Refining
30.5
34.7
35.0
35.3
49.5
42.1
41.8
Total
49.7
56.0
88.5
137.1
153.0
125.2
82.4
NE (Not Estimated)
3 Methodology and Time-Series Consistency
4 See Annex 3.5 for the full time series of emissions data, activity data, emission factors, and additional information
5 on methods and data sources.
6 Petroleum systems includes emission estimates for activities occurring in petroleum systems from the oil wellhead
7 through crude oil refining, including activities for crude oil exploration, production field operations, crude oil
8 transportation activities, and refining operations. Generally, emissions are estimated for each activity by
9 multiplying emission factors (e.g., emission rate per equipment or per activity) by corresponding activity data (e.g.,
10 equipment count or frequency of activity). Certain sources within petroleum refineries are developed using an
11 IPCC Tier 3 approach (i.e., all refineries in the nation report facility-level emissions data to the GHGRP, which are
12 included directly in the national emissions estimates here). Other estimates are developed with a Tier 2 approach.
13 Tier 1 approaches are not used.
14 EPA received stakeholder feedback on updates in the Inventory through EPA's stakeholder process on oil and gas
15 in the Inventory. Stakeholder feedback is noted below in Recalculations Discussion and Planned Improvements.
16 More information on the stakeholder process can be found online.74
17 Emission Factors. Key references for emission factors include Methane Emissions from the Natural Gas Industry by
18 the Gas Research Institute and EPA (GRI/EPA 1996), Estimates of Methane Emissions from the U.S. Oil Industry (EPA
19 1999), Compilation of Air Pollutant Emission Factors, AP-42 (EPA 1997), Global Emissions of Methane from
20 Petroleum Sources (API 1992), consensus of industry peer review panels, Bureau of Ocean Energy Management
21 (BOEM) reports, Nonpoint Oil and Gas Emission Estimation Tool (EPA 2017), and analysis of GHGRP data (EPA
22 2022).
23 Emission factors for hydraulically fractured (HF) oil well completions and workovers (in four control categories)
24 were developed using EPA's GHGRP data; year-specific data were used to calculate emission factors from 2016-
25 forward and the year 2016 emission factors were applied to all prior years in the time series. The emission factors
74 See https://www.epa.gov/ghgemissions/natural-gas-and-petroleum-systems.
Energy 3-79
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
for well testing and associated gas venting and flaring were developed using year-specific GHGRP data for years
2015 forward; earlier years in the time series use 2015 emission factors. For miscellaneous production flaring,
year-specific emission factors were developed for years 2015 forward from GHGRP data, an emission factor of 0
(assumption of no flaring) was assumed for 1990 through 1992, and linear interpolation was applied to develop
emission factors for 1993 through 2014. For more information, please see memoranda available online.75 For
offshore oil production, emission factors were calculated using BOEM data for offshore facilities in federal waters
of the Gulf of Mexico (and these data were also applied to facilities located in state waters of the Gulf of Mexico)
and GHGRP data for offshore facilities off the coasts of California and Alaska. For many other sources, emission
factors were held constant for the period 1990 through 2021, and trends in emissions reflect changes in activity
levels. Emission factors from EPA 1999 are used for all other production and transportation activities.
For associated gas venting and flaring and miscellaneous production flaring, emission factors were developed on a
production basis (i.e., emissions per unit oil produced). Additionally, for these two sources, basin-specific activity
and emission factors were developed for each basin that in any year from 2011 forward contributed at least 10
percent of total source emissions (on a CO2 Eq. basis) in the GHGRP. For associated gas venting and flaring, basin-
specific factors were developed for four basins: Williston, Permian, Gulf Coast, and Anadarko. For miscellaneous
production flaring, basin-specific factors were developed for three basins: Williston, Permian, and Gulf Coast. For
each source, data from all other basins were combined, and activity and emission factors were developed for the
other basins as a single group.
For pneumatic controllers and tanks, basin-specific emission factors were calculated for all the basins reporting to
the GHGRP. These emission factors were calculated for all the years with applicable GHGRP data (i.e., 2011 - 2021
or 2015 - 2021). For the remaining basins (i.e., basins not reporting to the GHGRP), subpart W average emission
factors were used. For more information, please see memoranda available online.3
For the exploration and production segments, in general, CO2 emissions for each source were estimated with
GHGRP data or by multiplying CO2 content factors by the corresponding CH4 data, as the CO2 content of gas relates
to its CH4 content. Sources with CO2 emission estimates calculated using GHGRP data include HF completions and
workovers, associated gas venting and flaring, tanks, well testing, pneumatic controllers, chemical injection pumps,
miscellaneous production flaring, and certain offshore production facilities (those located off the coasts of
California and Alaska). For these sources, CO2 was calculated using the same methods as used for CH4. Carbon
dioxide emission factors for offshore oil production in the Gulf of Mexico were derived using data from BOEM,
following the same methods as used for CH4 estimates. For other sources, the production field operations emission
factors for CO2 are generally estimated by multiplying the CH4 emission factors by a conversion factor, which is the
ratio of CO2 content and CH4 content in produced associated gas.
For the exploration and production segments, N2O emissions were estimated for flaring sources using GHGRP or
BOEM OGOR-B data and the same method used for CO2. Sources with N2O emissions in the exploration segment
include well testing and HF completions with flaring. Sources with N2O emissions in the production segment
include associated gas flaring, tank flaring, miscellaneous production flaring, HF workovers with flaring, and flaring
from offshore production sources.
For crude oil transportation, emission factors for CH4 were largely developed using data from EPA (1997), API
(1992), and EPA (1999). Emission factors for CO2 were estimated by multiplying the CH4 emission factors by a
conversion factor, which is the ratio of CO2 content and CH4 content in whole crude post-separator.
For petroleum refining activities, year-specific emissions from 2010 forward were directly obtained from EPA's
GHGRP. All U.S. refineries have been required to report CH4, CO2, and N2O emissions for all major activities starting
with emissions that occurred in 2010. The reported total CH4, CO2, and N2O emissions for each activity was used
for the emissions in each year from 2010 forward. To estimate emissions for 1990 to 2009, the 2010 to 2013
emissions data from GHGRP along with the refinery feed data for 2010 to 2013 were used to derive CH4 and CO2
emission factors (i.e., sum of activity emissions/sum of refinery feed) and 2010 to 2017 data were used to derive
75 See https://www.epa.gov/ghgemissions/natural-gas-and-petroleum-systems.
3-80 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
N2O emission factors; these emission factors were then applied to the annual refinery feed in years 1990 to 2009.
GHGRP delayed coker CH4 emissions for 2010 through 2017 were increased using the ratio of certain reported
emissions for 2018 to 2017, to account for a more accurate GHGRP calculation methodology that was
implemented starting in reporting year 2018.
A complete list of references for emission factors and activity data by emission source is provided in Annex 3.5.
Activity Data. References for activity data include Enverus data (Enverus 2021), Energy Information Administration
(EIA) reports, Methane Emissions from the Natural Gas Industry by the Gas Research Institute and EPA (EPA/GRI
1996), Estimates of Methane Emissions from the U.S. Oil Industry (EPA 1999), consensus of industry peer review
panels, BOEM reports, the Oil & Gas Journal, the Interstate Oil and Gas Compact Commission, the United States
Army Corps of Engineers, and analysis of GHGRP data (EPA 2022).Enverus data for 2021 are not currently available;
this version of the Inventory uses 2020 data as proxy for 2021.
For pneumatic controllers, equipment leaks, chemical injection pumps, and tanks, basin-specific activity factors
were calculated for all the basins reporting to the GHGRP. These factors were calculated for all the years with
applicable GHGRP data (i.e., 2011 through 2021 or 2015 through 2021). For the remaining basins (i.e., basins not
reporting to the GHGRP), GHGRP average activity factors were used. For more information, please see memoranda
available online.76
For many sources, complete activity data were not available for all years of the time series. In such cases, one of
three approaches was employed to estimate values, consistent with IPCC good practice. Where appropriate, the
activity data were calculated from related statistics using ratios developed based on EPA/GRI (1996) and/or GHGRP
data. In some cases, activity data are developed by interpolating between recent data points (such as from GHGRP)
and earlier data points, such as from EPA/GRI (1996). Lastly, in limited instances the previous year's data were
used if current year data were not yet available.
A complete list of references for emission factors and activity data by emission source is provided in Annex 3.5. The
United States reports data to the UNFCCC using this Inventory report along with Common Reporting Format (CRF)
tables. This note is provided for those reviewing the CRF tables: The notation key "IE" is used for CO2 and CH4
emissions from venting and flaring in CRF table l.B.2. Disaggregating flaring and venting estimates across the
Inventory would involve the application of assumptions and could result in inconsistent reporting and, potentially,
decreased transparency. Data availability varies across segments within oil and gas activities systems, and emission
factor data available for activities that include flaring can include emissions from multiple sources (flaring, venting
and leaks).
As noted above, EPA's GHGRP data, available starting in 2010 for refineries and in 2011 for other sources, have
improved estimates of emissions from petroleum systems. Many of the previously available datasets were
collected in the 1990s. To develop a consistent time series for sources with new data, EPA reviewed available
information on factors that may have resulted in changes over the time series (e.g., regulations, voluntary actions)
and requested stakeholder feedback on trends as well. For most sources, EPA developed annual data for 1993
through 2009 or 2014 by interpolating activity data or emission factors or both between 1992 (when GRI/EPA data
are available) and 2010 or 2015 data points. Information on time-series consistency for sources updated in this
year's Inventory can be found in the Recalculations Discussion below, with additional detail provided in supporting
memos (relevant memos are cited in the Recalculations Discussion). For information on other sources, please see
the Methodology Discussion above and Annex 3.5.
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2021.
76 See https://www.epa.gov/ghgemissions/natural-gas-and-petroleum-systems.
Energy 3-81
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Uncertainty— TO BE UPDATED FOR FINAL INVENTORY REPORT
EPA conducted a quantitative uncertainty analysis using the IPCC Approach 2 methodology (Monte Carlo
Simulation technique) to characterize uncertainty for petroleum systems. For more information on the approach,
please see the memoranda Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2016: Natural Gas and
Petroleum Systems Uncertainty Estimates and Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2019:
Update for Natural Gas and Petroleum Systems CO2 Uncertainty EstimatesP
EPA used Microsoft Excels @RISK add-in tool to estimate the 95 percent confidence bound around CH4 and CO2
emissions from petroleum systems for the current Inventory. For the CH4 uncertainty analysis, EPA focused on the
six highest methane-emitting sources for the year 2020, which together emitted 76 percent of methane from
petroleum systems in 2020, and extrapolated the estimated uncertainty for the remaining sources For the CO2
uncertainty analysis, EPA focused on the 3 highest-emitting sources for the year 2020 which together emitted 80
percent of CO2 from petroleum systems in 2020, and extrapolated the estimated uncertainty for the remaining
sources. The @RISK add-in provides for the specification of probability density functions (PDFs) for key variables
within a computational structure that mirrors the calculation of the inventory estimate. The IPCC guidance notes
that in using this method, "some uncertainties that are not addressed by statistical means may exist, including
those arising from omissions or double counting, or other conceptual errors, or from incomplete understanding of
the processes that may lead to inaccuracies in estimates developed from models." As a result, the understanding
of the uncertainty of emission estimates for this category evolves and improves as the underlying methodologies
and datasets improve. The uncertainty bounds reported below only reflect those uncertainties that EPA has been
able to quantify and do not incorporate considerations such as modeling uncertainty, data representativeness,
measurement errors, misreporting or misclassification. To estimate uncertainty for N2O, EPA applied the
uncertainty bounds calculated for CO2. EPA will seek to refine this estimate in future Inventories.
The results presented below provide the 95 percent confidence bound within which actual emissions from this
source category are likely to fall for the year 2020, using the recommended IPCC methodology. The results of the
Approach 2 uncertainty analysis are summarized in Table 3-50. Petroleum systems CFU emissions in 2020 were
estimated to be between 29.0 and 53.1 MMT CO2 Eq., while CO2 emissions were estimated to be between 23.5
and 38.0 MMT CO2 Eq. at a 95 percent confidence level. Petroleum systems N2O emissions in 2020 were estimated
to be between 0.03 and 0.05 MMT CO2 Eq. at a 95 percent confidence level.
Uncertainty bounds for other years of the time series have not been calculated, but uncertainty is expected to vary
over the time series. For example, years where many emission sources are calculated with interpolated data would
likely have higher uncertainty than years with predominantly year-specific data. In addition, the emission sources
that contribute the most to CFU and CO2 emissions are different over the time series, particularly when comparing
recent years to early years in the time series. For example, associated gas venting emissions were higher and
flaring emissions were lower in early years of the time series, compared to recent years. Technologies also
changed over the time series (e.g., reduced emissions completions were not used early in the time series).
Table 3-50: Approach 2 Quantitative Uncertainty Estimates for ChU and CO2 Emissions from
Petroleum Systems (MMT CO2 Eq. and Percent)
2020 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.)b
(MMT C02
Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Petroleum Systems CH4
40.2
29.0
53.1
-28%
+32%
Petroleum Systems C02
30.2
23.5
38.0
-22%
+26%
Petroleum Systems n20
0.04
0.03
0.05
-22%
+26%
77 See https://www.epa.gov/ghgemissions/natural-gas-and-petroleum-svstems.
3-82 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
a Range of emission estimates estimated by applying the 95 percent confidence intervals obtained from the Monte Carlo
Simulation analysis conducted for the year 2020 CH4 and C02 emissions.
b All reported values are rounded after calculation. As a result, lower and upper bounds may not be duplicable from other
rounded values as shown in table.
QA/QC and Verification Discussion
The petroleum systems emission estimates in the Inventory are continually being reviewed and assessed to
determine whether emission factors and activity factors accurately reflect current industry practices. A QA/QC
analysis was performed for data gathering and input, documentation, and calculation. QA/QC checks are
consistently conducted to minimize human error in the emission calculations. EPA performs a thorough review of
information associated with new studies, GHGRP data, regulations, public webcasts, and the Natural Gas STAR
Program to assess whether the assumptions in the Inventory are consistent with current industry practices. EPA
has a multi-step data verification process for GHGRP data, including automatic checks during data-entry, statistical
analyses on completed reports, and staff review of the reported data. Based on the results of the verification
process, EPA follows up with facilities to resolve mistakes that may have occurred.78
As in previous years, EPA conducted early engagement and communication with stakeholders on updates prior to
public review of the current Inventory. EPA held stakeholder webinars on greenhouse gas data for oil and gas in
September and November of 2022. EPA released memos detailing updates under consideration and requesting
stakeholder feedback. Stakeholder feedback received through these processes is discussed in the Recalculations
Discussion and Planned Improvements sections below.
In recent years, several studies have measured emissions at the source level and at the national or regional level
and calculated emission estimates that may differ from the Inventory. There are a variety of potential uses of data
from new studies, including replacing a previous estimate or factor, verifying or QA of an existing estimate or
factor, and identifying areas for updates. In general, there are two major types of studies related to oil and gas
greenhouse gas data: studies that focus on measurement or quantification of emissions from specific activities,
processes, and equipment, and studies that use tools such as inverse modeling to estimate the level of overall
emissions needed to account for measured atmospheric concentrations of greenhouse gases at various scales. The
first type of study can lead to direct improvements to or verification of Inventory estimates. In the past few years,
EPA has reviewed, and in many cases, incorporated data from these data sources. The second type of study can
provide general indications on potential over- and under-estimates.
A key challenge in using these types of studies to assess Inventory results is having a relevant basis for comparison
(e.g., the two data sets should have comparable time frames and geographic coverage, and the independent study
should assess data from the Inventory and not another data set, such as the Emissions Database for Global
Atmospheric Research, or "EDGAR"). In an effort to improve the ability to compare the national-level Inventory
with measurement results that may be at other spatial and temporal scales, a team at Harvard University along
with EPA and other coauthors developed a gridded inventory of U.S. anthropogenic methane emissions with 0.1
degree x 0.1 degree spatial resolution, monthly temporal resolution, and detailed scale-dependent error
characterization.79 The gridded methane inventory is designed to be consistent with the U.S. EPA's Inventory of
U.S. Greenhouse Gas Emissions and Sinks: 1990-2014 estimates for the year 2012, which presents national totals.80
An updated version of the gridded inventory is being developed and will improve efforts to compare results of the
inventory with atmospheric studies.
78 See https://www.epa.eov/sites/production/files/2015-07/documents/eherp verification factsheet.pdf.
79 See https://www.epa.eov/eheemissions/eridded-2012-methane-emissions.
80 See https://www.epa.eov/eheemissions/us-ereenhouse-eas-inventorv-report-1990-2014.
Energy 3-83
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
As discussed above, refinery emissions are quantified by using the total emissions reported to GHGRP for the
refinery emission categories included in Petroleum Systems. Subpart Y has provisions that refineries are not
required to report under Subpart Y if their emissions fall below certain thresholds. Each year, a review is conducted
to determine whether an adjustment is needed to the Inventory emissions to include emissions from refineries
that stopped reporting to the GHGRP. Based on the review of the most recent GHGRP data, EPA identified a
refinery last reported annual emissions data to the GHGRP for reporting year 2012, due to meeting the criteria for
cessation of reporting. EPA used the 2012 reported emissions for the refinery as proxy to gap fill annual emissions
for 2013 through 2020 for this refinery.
Recalculations Discussion
EPA received information and data related to the emission estimates through GHGRP reporting and stakeholder
feedback on updates under consideration. In October 2022, EPA released a draft memorandum that discussed
changes under consideration and requested stakeholder feedback on those changes.81 EPA did not receive written
feedback on the memorandum. Memoranda cited in the Recalculations Discussion below are: Inventory of U.S.
Greenhouse Gas Emissions and Sinks 1990-2021: Updates Under Consideration for Incorporating Additional
Geographically Disaggregated Data (Disaggregation memo) and Inventory of U.S. Greenhouse Gas Emissions and
Sinks 1990-2021: Updates Under Consideration for Incorporating Additional Geographically Disaggregated Data for
the Production Segment (Production Disaggregation memo).
In this Inventory, an update that incorporates additional basin-level data from GHGRP subpart W was implemented
for several emission sources in the onshore production segment. The update seeks to improve the ability of EPA's
gridded and state inventories to reflect variation due to differences in formation types, technologies and practices,
regulations, or voluntary initiatives, and not only the differences in key activity levels that are reflected in the
current gridded and state inventories. This would allow EPA to use the gridded inventory for improved
comparisons of the national Inventory with various atmospheric observation studies (since regions will better
reflect the local differences in emissions rates as reported to GHGRP) and would allow the state-level inventory to
reflect differences in state-level programs, formation type mixes, and varying technologies and practices. For many
sources, an approach that develops estimates using geographically disaggregated data may not be possible or
preferable to a national level approach based on the currently available data. For some emission sources in the
Inventory, emission factor data come from research studies and are applied at the national level. For example,
many of the emission factors used to quantify emissions in the Inventory for the gathering and boosting,
transmission and storage, distribution, and post-meter segments are from research studies and do not have a level
of detail or total population comparable to GHGRP. For petroleum refineries, because there is no reporting
threshold for GHGRP Subpart Y, facility-level data are generally available for all refineries in the United States, and
these site-specific data are already used to develop the gridded and state-level greenhouse gas estimates. Even in
cases where geographically disaggregated data are available, such an approach may not always be preferable. In
cases with limited variation between areas, such an approach would have limited impact on emissions estimates
regionally or nationally. In cases with limited data in certain areas, disaggregated approaches might substantially
increase the uncertainty of estimates and basin-specific calculations would not be an improvement over use of a
national average. EPA continues to seek stakeholder feedback on the draft approach in this Inventory.
EPA evaluated relevant information available and made several updates to the Inventory, including for pneumatic
controllers, equipment leaks, chemical injection pumps, and storage tanks. For each of these emission sources,
EPA modified the calculation methodology to use GHGRP data to develop basin-specific activity factors and/or
emission factors. General information for these source specific recalculations are presented below and details
(including the basin-specific emissions estimates) are available in the Disaggregation memo and Production
Disaggregation memo.
81 Stakeholder materials including draft memoranda for the current (i.e., 1990 to 2021) Inventory are available at
https://www.epa.gov/ghgemissions/natural-gas-and-petroleum-svstems.
3-84 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 In addition to the updates to production segment sources mentioned above, for certain sources, Cm and/or CO2
2 emissions changed by greater than 0.05 MMT CO2 Eq., comparing the previous estimate for 2020 to the current
3 (recalculated) estimate for 2020. The emissions changes were mostly due to GHGRP data submission revisions.
4 These sources are discussed below and include associated gas flaring, miscellaneous production flaring, offshore
5 production, and refinery flaring.
6 In addition, for the current Inventory, CC>2-equivalent emissions totals have been revised to reflect the 100-year
7 global warming potentials (GWPs) provided in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). AR5 GWP
8 values differ slightly from those presented in the IPCC Fourth Assessment Report (AR4) (IPCC 2007) used in the
9 previous inventories. The AR5 GWPs have been applied across the entire time series for consistency. The GWP of
10 Cm has increased from 25 to 28, leading to an increase in the calculated CC>2-equivalent emissions of CH4, while
11 the GWP of N2O has decreased from 298 to 265, leading to a decrease in the calculated CC>2-equivalent emissions
12 of N2O. Further discussion on this update and the overall impacts of updating the Inventory GWP values to reflect
13 the IPCC Fifth Assessment Report can be found in Chapter 9, Recalculations and Improvements.
14 The combined impact of revisions to 2020 petroleum systems CH4 emission estimates on a CC>2-equivalent basis,
15 compared to the previous Inventory, is an increase from 40.2 to 54.5 MMT CO2 Eq. (14.2 MMT CO2 Eq., or 35
16 percent). The recalculations resulted in higher CH4 emission estimates on average across the 1990 through 2020
17 time series, compared to the previous Inventory, by 11.0 MMT CO2 Eq., or 25 percent.
18 The combined impact of revisions to 2020 petroleum systems CO2 emission estimates, compared to the previous
19 Inventory, is a decrease from 30.2 to 29.1 MMT CO2 (1.1 MMT CO2, or 4 percent). The recalculations resulted in
20 lower emission estimates on average across the 1990 through 2020 time series, compared to the previous
21 Inventory, by 1.2 MMT CO2 Eq., or 9 percent.
22 The combined impact of revisions to 2020 petroleum systems N2O emission estimates on a CC>2-equivalent basis,
23 compared to the previous Inventory, is a decrease of 0.004 MMT CO2, Eq. or 12 percent. The recalculations
24 resulted in an average decrease in emission estimates across the 1990 through 2020 time series, compared to the
25 previous Inventory, of 0.002 MMT CO2 Eq., or 11 percent.
26 In Table 3-51 and Table 3-52 below are categories in Petroleum Systems with updated methodologies or with
27 recalculations resulting in a change of greater than 0.05 MMT CO2 Eq., comparing the previous estimate for 2020
28 to the current (recalculated) estimate for 2020. For more information, please see the discussion below.
29 For certain sources, CH4 emissions for 2020 changed by greater than 0.05 MMT CO2 Eq., compared to the previous
30 Inventory due to the use of an updated GWP value (AR5). These sources are not discussed below and include
31 associated gas venting and flaring, produced water, gas engines, heaters, and refineries.
32 Table 3-51: Recalculations of CO2 in Petroleum Systems (MMT CO2)
Segment/Source
Previous Estimate
Year 2020,
2022 Inventory
Current Estimate
Year 2020,
2023 Inventory
Current Estimate
Year 2021,
2023 Inventory
Exploration
0.9
0.8
0.5
Production
25.0
24.0
20.0
Tanks
6.5
5.3
5.4
Pneumatic Controllers
0.1
0.1
0.1
Equipment Leaks
+
+
+
Chemical Injection Pumps
+
+
+
Associated Gas Flaring
13.0
13.3
9.6
Miscellaneous Production Flaring
4.6
4.7
4.2
Transportation
+
+
+
Refining
4.3
4.2
4.2
Flares
4.3
4.2
4.2
Petroleum Systems Total
30.2
29.1
24.7
+ Does not exceed 0.05 MMT C02.
Energy 3-85
-------
1
Table 3-52: Recalculations of ChU in Petroleum Systems (MMT CO2 Eq.)
Segment/Source
Previous Estimate
Year 2020,
2022 Inventory
Current Estimate
Year 2020,
2023 Inventory
Current Estimate
Year 2021,
2023 Inventory
Exploration
0.3
0.3
0.2
Production
38.9
53.0
48.9
Tanks
0.7
0.8
0.6
Pneumatic Controllers
21.3
31.7
28.4
Equipment Leaks
2.4
3.2
3.3
Chemical Injection Pumps
1.9
3.3
3.2
Miscellaneous Production Flaring
0.4
0.6
0.5
Offshore Production
4.8
5.3
5.5
Transportation
0.2
0.2
0.2
Refining
0.8
0.9
0.8
Petroleum Systems Total
40.2
54.5
50.2
2 Exploration
3 Recalculations for the exploration segment have resulted in lower calculated CFU and CO2 emissions over the time
4 series (less than 0.1 percent), compared to the previous Inventory.
5 Production
6 Pneumatic Controllers (Methodological Update)
1 EPA updated the calculation methodology for pneumatic controllers to use basin-specific activity factors and
8 emission factors calculated from subpart W data for each type of controller (i.e., high, intermittent, and low
9 bleed). Previously, national average activity and emission factors calculated using subpart W data were applied to
10 estimate pneumatic controller emissions. In this methodological update, EPA summed basin-level emissions
11 together to develop national emissions. The Disaggregation memo and Production Disaggregation memo present
12 additional information and considerations for this update.
13 EPA calculated basin-specific activity factors and CH4 emission factors for all basins that reported subpart W data.
14 The factors were year-specific for RY2011 through RY2021. EPA retained the previous Inventory's activity factor
15 assumptions for 1990 through 1993 and applied linear interpolation between the 1993 and 2011 activity factors at
16 the basin-level. Year 2011 emission factors were applied to all prior years for each basin. For basins without
17 subpart W data available, EPA applied national average activity and emission factors.
18 The estimation methodology for CO2 emissions was not updated to use the basin-specific approach for the public
19 review version of the Inventory. CO2 emissions were estimated by applying a CO2 to CFU ratio to the estimated CH4
20 emissions. EPA will calculate pneumatic controller CO2 emissions in the same manner as CFU emissions for the final
21 Inventory.
22 As a result of this methodological update, CH4 emissions estimates are an average of 22 percent higher across the
23 time-series and 32 percent higher in 2020, compared to the previous Inventory. The most significant changes are in
24 recent years, 2013 through 2020, due specifically to changes in intermittent bleed controller emissions estimates.
25 Certain basins (e.g., Anadarko Basin, Appalachian, Appalachian Basin (Eastern Overthrust), Bend Arch, Fort Worth
26 Syncline, Gulf Coast, and Sedgwick) have higher activity factors (mainly the average number of controllers per
27 well) and/or emission factors for intermittent bleed pneumatic controllers, compared to the national average.
28 Some of these basins also exhibit large changes in emissions over these recent years.
3-86 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
Table 3-53: Pneumatic Controllers National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2017
2018
2019
2020
2021
High Bleed Controllers
708,800
493,011
89,472
73,438
73,278
87,884
48,202
Low Bleed Controllers
51,170
63,773
20,104
31,779
50,456
36,752
46,360
Intermittent Bleed Controllers
0
276,145
1,252,028
1,155,041
762,647
1,006,263
920,518
Total Emissions
759,970
832,929
1,361,605
1,260,259
886,382
1,130,899
1,015,080
Previous Estimate
736,447
708,680
835,129
727,365
732,092
853,562
NA
NA (Not Applicable)
2 Equipment Leaks (Methodological Update)
3 EPA updated the calculation methodology for onshore production equipment leaks to use basin-specific
4 equipment-level activity factors (e.g., separators/well) from GHGRP data. Previously, national average equipment
5 activity factors developed using RY2014 GHGRP data were used in the Inventory for all years. In this
6 methodological update, EPA summed basin-level emissions together to develop national emissions. The
7 Disaggregation memo and Production Disaggregation memo present additional information and considerations
8 for this update.
9 EPA calculated basin-specific equipment-level activity factors for all basins that reported subpart W data. The
10 factors were year-specific for RY2015 through RY2021. EPA retained the previous Inventory's activity factors for
11 1990 through 1993 and used linear interpolation between the 1993 and 2015 activity factors at the basin-level. For
12 basins without subpart W data available, EPA applied national average activity factors using all subpart W data.
13 This methodological update applies only for activity factors. The previous Inventory's Cm emission factors for
14 onshore production segment equipment leaks (by equipment type) were retained and used to develop Cm
15 estimates.
16 The calculation methodology for CO2 emissions was not updated for the public review version of the Inventory.
17 The previous Inventory's methodology was retained to develop CO2 estimates. EPA will calculate equipment leak
18 CO2 emissions in the same manner as CH4 emissions for the final Inventory.
19 This update resulted in CH4 emissions an average of 18 percent higher across the time-series compared with the
20 previous Inventory and a 21 percent higher estimate for 2020, compared to the previous Inventory. The emissions
21 increase is due to certain basins having higher activity factors compared to the national average activity factors
22 (e.g., Anadarko, Appalachian, Appalachian Basin (Eastern Overthrust), and Gulf Coast).
23 Table 3-54: Production Equipment Leaks National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2017
2018
2019
2020
2021
Oil Wellheads
56,524
51,563
60,557
59,195
60,877
58,632
60,029
Separators
10,970
17,514
30,021
42,001
38,510
29,356
27,107
Heater/Treaters
11,119
20,741
16,245
17,492
22,706
18,734
21,307
Headers
3,323
12,434
12,754
13,217
15,595
8,075
8,444
Total Emissions
81,936
102,251
119,577
131,904
137,688
114,797
116,887
Previous Estimate
81,874
86,248
100,450
99,287
98,459
94,921
NA
NA (Not Applicable)
24 Chemical injection Pumps (Methodological Update)
25 EPA updated the calculation methodology for chemical injection pumps to use basin-specific activity factors from
26 GHGRP data. Previously, a national average activity factor developed using RY2014 GHGRP data was used in the
27 Inventory for all years. In this methodological update, EPA summed basin-level emissions together to develop
28 national emissions. The Disaggregation memo and Production Disaggregation memo present additional
29 information and considerations for this update.
Energy 3-87
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
EPA calculated basin-specific activity factor for all basins that reported subpart W data. The factors were year-
specific for RY2015 through RY2021. EPA also retained the previous Inventory's activity factor for 1990 through
1993 and used linear interpolation between the 1993 and 2015 activity factors at the basin-level. For basins
without subpart W data available, EPA applied the national average unweighted activity factor from all subpart W
data. This methodological update applies only to activity factors. The previous Inventory's CFU emission factor for
chemical injection pumps was retained and used to develop Cm estimates.
The estimation methodology for CO2 emissions was not updated for the public review version of the Inventory. The
previous Inventory's methodology was retained to develop CO2 estimates. EPA will calculate chemical injection
pump CO2 emissions in the same manner as CH4 emissions for the final Inventory.
This update resulted in calculated CH4 emissions an average of 63 percent higher across the time-series compared
with the previous Inventory and 52 percent higher in 2020, compared to the previous Inventory. The emissions
increase is due to certain basins having a higher activity factor compared to the national average activity factor
(e.g., Anadarko Basin, Appalachian, Appalachian Basin (Eastern Overthrust), Bend Arch, Fort Worth Syncline, Green
River, and Gulf Coast).
Table 3-55: Chemical Injection Pumps National ChU Emissions (Metric Tons ChU)
Source 1990 2005 2017 2018 2019 2020 2021
Chemical Injection Pumps 47,401 105,458 121,469 138,866 387,416 116,080 115,678
Previous Estimate 46,758 67,685 80,728 79,793 79,128 76,284 NA
NA (Not Applicable)
Storage Tanks (Methodological Update)
EPA updated the calculation methodology for production segment storage tanks to use basin-specific activity
factors and emission factors, calculated from Ssubpart W data for each storage tank category. Previously, national
annual average activity and emission factors calculated using subpart W data were applied to estimate storage
tank emissions. In this update, EPA developed national emission estimates by summing calculated basin-level total
emission estimates, using basin-level data emission and activity factors developed from Subpart W. The Production
Disaggregation memo presents additional information and considerations for this update.
EPA calculated basin-specific activity factors and CH4 and CO2 emission factors for all basins that reported subpart
W data. The factors were year-specific for reporting year (RY) 2015 through RY2021. EPA also retained the previous
Inventory's activity factor assumptions (i.e., all oil tanks were uncontrolled in 1990) and used linear interpolation
between the 1990 and 2015 activity factors at the basin-level. Year 2015 emission factors were applied to all prior
years for each basin. For basins without Subpart W data available, EPA applied national average activity and
emission factors (unweighted average of all Subpart W reported data).
This update resulted in oil tank CFU emission estimates that are on average 16 percent lower across the time series
than in the previous Inventory. The CH4 estimates for 2020 are 2 percent lower than in the previous Inventory. Oil
tank CO2 emissions are on average 55 percent lower across the time series than in the previous Inventory and 2020
emissions estimates are 20 percent lower than in the previous Inventory. The CH4 emissions estimate decrease
occurs mainly from 1990 through 2005, where there is an average decrease in calculated emissions of 39 percent,
compared to the previous inventory. Oil tank CO2 emissions have a similarly large decrease in that time frame.
The Arctic Coastal Plains Province Basin has a large impact on these earlier time series year emissions, when this
basin accounts for a large percentage of total liquids production, but very little of the production in that basin is
stored in tanks. Oil tank CO2 emissions decreased in recent years of the time series due to certain basins with
higher production (e.g., Denver Basin, Gulf Coast, Permian) having lower activity factors and emission factors than
the national average.
Table 3-56: Storage Tanks National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2017
2018
2019
2020
2021
Large Tanks w/Flares 0 993 5,142 6,330 4,226 3,715 3,108
3-88 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
721
513
Large Tanks w/o Control
105,668
40,150
42,112
42,679
26,491
21,294
12,290
Small Tanks w/Flares
0
15
45
16
23
29
68
Small Tanks w/o Flares
7,438
3,448
2,991
3,326
2,755
2,709
3,598
Malfunctioning Separator Dump
Valves
2,397
1,472
4,247
785
428
338
320
Total Emissions
115,503
46,799
63,871
55,546
36,243
29,112
19,896
Previous Estimate
218,419
60,186
61,098
57,412
35,266
29,613
NA
NA (Not Applicable)
Table 3-57: Storage Tanks National CO2 Emissions (kt CO2)
Source
1990
2005
2017
2018
2019
2020
2021
Large Tanks w/Flares
0
716
3,771
5,348
5,974
5,212
5,381
Large Tanks w/VRU
0
3
4
4
6
2
1
Large Tanks w/o Control
24
8
5
4
5
6
5
Small Tanks w/Flares
0
3
11
7
9
10
9
Small Tanks w/o Flares
12
5
4
5
4
4
5
Malfunctioning Separator Dump
Valves
12
13
32
30
26
20
37
Total Emissions
47
748
3,828
5,398
6,024
5,255
5,439
Previous Estimate
115
2,505
4,313
6,189
6,682
6,537
NA
NA (Not Applicable)
2 Associated Gas Flaring (Recalculation with Updated Data)
3 Associated gas flaring CO2 emission estimates are on average of 0.1 percent higher across the time series
4 compared with the previous Inventory and in 2020 are 2 percent higher than in the previous Inventory. The
5 emission changes were due to GHGRP data submission revisions.
6 Table 3-58: Associated Gas Flaring National CO2 Emissions (kt CO2)
Source
1990
2005
2017
2018
2019
2020
2021
220 - Gulf Coast Basin (LA, TX)
225
124
749
645
712
801
410
360 - Anadarko Basin
102
63
62
79
18
10
8
395 - Williston Basin
969
1,243
6,954
10,698
15,334
8,257
6,772
430 - Permian Basin
2,844
1,971
3,141
6,700
7,333
3,605
1,942
"Other" Basins
944
507
384
633
1,006
619
486
Total Emissions
5,084
3,908
11,291
18,756
24,403
13,293
9,619
220 - Gulf Coast Basin (LA, TX)
225
124
749
651
713
798
NA
360 - Anadarko Basin
102
63
62
79
18
10
NA
395 - Williston Basin
969
1,243
6,909
11,140
14,762
8,052
NA
430 - Permian Basin
2,844
1,971
3,141
6,711
7,227
3,558
NA
"Other" Basins
944
507
384
624
990
624
NA
Previous Estimate
5,084
3,908
11,245
19,206
23,710
13,041
NA
NA (Not Applicable)
7 Miscellaneous Production Flaring
Energy 3-89
-------
1 Miscellaneous production flaring CO2 emission estimates are on average 0.3 percent higher across the time series
2 than in the previous Inventory and in 2020 are 2 percent higher than in the previous Inventory. The emission
3 estimate changes were due to GHGRP data submission revisions.
4 Table 3-59: Miscellaneous Production Flaring National CO2 Emissions (kt CO2)
Source
1990
2005
2017
2018
2019
2020
2021
220 - Gulf Coast Basin (LA, TX)
0
105
509
584
616
651
787
395 - Williston Basin
0
72
537
1,701
2,643
852
882
430 - Permian Basin
0
209
1,465
1,406
4,320
2,798
2,216
"Other" Basins
0
400
551
615
646
378
270
Total Emissions
0
786
3,063
4,307
8,225
4,679
4,154
Previous Estimate
0
786
3,031
4,166
7,989
4,589
NA
NA (Not Applicable)
5 Miscellaneous production flaring Cm emission estimates are on average 2 percent higher across the time series
6 compared with the previous inventory and in 2020 are 31 percent higher than calculated in the previous Inventory.
7 The emission changes were due to GHGRP data submission revisions.
8 Table 3-60: Miscellaneous Production Flaring National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2017
2018
2019
2020
2021
220 - Gulf Coast Basin (LA, TX)
0
440
2,119
1,978
2,506
2,452
2,989
395 - Williston Basin
0
179
1,618
3,031
3,503
1,670
1,396
430 - Permian Basin
0
1,097
5,389
5,296
21,296
16,712
11,305
"Other" Basins
0
1,291
1,904
1,816
1,731
1,249
961
Total Emissions
0
3,008
11,030
12,121
29,036
22,082
16,650
Previous Estimate
0
3,008
10,928
11,669
22,994
16,807
NA
NA (Not Applicable)
9 Offshore Production (Recalculation with Updated Data)
10 Offshore production CH4 emission estimates are on average less than 0.05 percent lower across the time series
11 than in the previous Inventory. The 2020 value is 3 percent lower than in the previous Inventory. The emission
12 changes were due to updated offshore complex counts.
13 Table 3-61: Offshore Production National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2017
2018
2019
2020
2021
GOM Federal Waters
302,936
219,285
187,433
183,236
181,488
173,336
179,891
GOM State Waters
5,657
665
96
60
71
60
59
Pacific Waters
22,609
17,659
5,052
3,794
3,370
4,262
4,554
Alaska State Waters
21,936
21,191
12,163
9,834
10,711
10,366
10,664
Total Emissions
353,138
258,801
204,745
196,924
195,640
188,024
195,168
Previous Estimate
353,138
258,801
203,917
196,349
195,626
192,943
NA
NA (Not Applicable)
14 Transportation
15 Recalculations for the transportation segment have resulted in calculated CH4 and CO2 emissions over the time
16 series from this segment that are lower (by less than 0.2 percent) than in the previous Inventory.
3-90 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Refining
Recalculations due to resubmitted GHGRP data in the refining segment have resulted in average calculated Cm
emissions over the time series 3 percent lower than in the previous Inventory, and 2020 Cm emissions 0.9 lower
than in the previous Inventory.
Refining CO2 emission estimates are on average 0.3 percent lower across the time series than in the previous
Inventory and 2 percent lower in 2020 than in the previous Inventory. This change is due to GHGRP resubmissions
and was largely due to a change in reported flaring CO2 emissions.
Table 3-62: Refining National CO2 Emissions (kt CO2)
Source
1990
2005
2017
2018
2019
2020
2021
Flares
3,134
3,557
3,509
3,643
4,961
4,208
4,182
Total Refining
3,284
3,728
3,582
3,706
5,009
4,242
4,214
Previous Estimate
3,284
3,728
3,725
3,820
5,080
4,326
NA
NA (Not Applicable)
Planned Improvements
Planned Improvements for 2023 Inventory
This draft of the Inventory does not yet incorporate updated activity data products for the following data inputs,
due to a data base subscription lapse: oil well counts, wells drilled, wells completed, and production. For these
inputs, year 2020 values for activity data are used in place of year 2021. The Final Inventory (to be published April
2023) will incorporate the latest activity data.
Basin-level approaches for pneumatic controllers, equipment leaks, and chemical injection pumps were applied to
calculate Cm emissions for public review. For the final Inventory, EPA would apply consistent methods for both
CO2 and Cm emissions calculations.
Additional information on the update and specific requests for stakeholder feedback can be found in the
Disaggregation memo and Production Disaggregation memos. Feedback EPA has received in response to the
memo include that basin-level data from GHGRP can improve accuracy of estimates when applied appropriately,
and that EPA should consider application of the approach to only basins with 50 percent coverage or more, EPA
will consider this feedback and any additional feedback received and may revise the calculations in the Inventory
based.
Upcoming Data, and Additional Data that Could Inform the Inventory
EPA will assess new data received by the Greenhouse Gas Reporting Program, the Methane Challenge Program and
other relevant programs on an ongoing basis, which may be used to confirm or improve existing estimates and
assumptions.
EPA continues to track studies that contain data that may be used to update the Inventory. EPA will also continue
to assess studies that include and compare both top-down and bottom-up estimates, and which could lead to
improved understanding of unassigned high emitters (e.g., identification of emission sources and information on
frequency of high emitters) as recommended in previous stakeholder comments.
Box 3-6: Carbon Dioxide Transport, Injection, and Geological Storage
Carbon dioxide is produced, captured, transported, and used for Enhanced Oil Recovery (EOR) as well as
commercial and non-EOR industrial applications, or is stored geologically. This CO2 is produced from both
naturally-occurring CO2 reservoirs and from industrial sources such as natural gas processing plants and
ammonia plants. In the Inventory, emissions of CO2 from naturally-occurring CO2 reservoirs are estimated based
Energy 3-91
-------
on the specific application.
In the Inventory, CChthat is used in non-EOR industrial and commercial applications (e.g., food processing,
chemical production) is assumed to be emitted to the atmosphere during its industrial use. These emissions are
discussed in the Carbon Dioxide Consumption section, 4.15.
For EOR CO2, as noted in the 2006IPCC Guidelines, "At the Tier 1 or 2 methodology levels [EOR CO2 is]
indistinguishable from fugitive greenhouse gas emissions by the associated oil and gas activities." In the U.S.
estimates for oil and gas fugitive emissions, the Tier 2 emission factors for CO2 include CO2 that was originally
injected and is emitted along with other gas from leak, venting, and flaring pathways, as measurement data
used to develop those factors would not be able to distinguish between CO2 from EOR and CO2 occurring in the
produced natural gas. Therefore, EOR CO2 emitted through those pathways is included in CO2 estimates in 1B2.
IPCC includes methodological guidance to estimate emissions from the capture, transport, injection, and
geological storage of CO2. The methodology is based on the principle that the carbon capture and storage
system should be handled in a complete and consistent manner across the entire Energy sector. The approach
accounts for CO2 captured at natural and industrial sites as well as emissions from capture, transport, and use.
For storage specifically, a Tier 3 methodology is outlined for estimating and reporting emissions based on site-
specific evaluations. However, IPCC (IPCC 2006) notes that if a national regulatory process exists, emissions
information available through that process may support development of CO2 emission estimates for geologic
storage.
In the United States, facilities that produce CO2 for various end-use applications (including capture facilities such
as acid gas removal plants and ammonia plants), importers of CO2, exporters of CO2, facilities that conduct
geologic sequestration of CO2, and facilities that inject CO2 underground, are required to report greenhouse gas
data annually to EPA through its GHGRP. Facilities reporting geologic sequestration of CO2 to the GHGRP
develop and implement an EPA-approved site-specific monitoring, reporting and verification plan, and report
the amount of CO2 sequestered using a mass balance approach.
GHGRP data relevant for this Inventory estimate consists of national-level annual quantities of CO2 captured and
extracted for EOR applications for 2010 to 2021 and data reported for geologic sequestration from 2016 to
2021.
The amount of CO2 captured and extracted from natural and industrial sites for EOR applications in 2020 is
35,090 kt (35.1 MMT CO2 Eq.) (see 6). The quantity of CO2 captured and extracted is noted here for information
purposes only; CO2 captured and extracted from industrial and commercial processes is generally assumed to be
emitted and included in emissions totals from those processes.
Table 3-63: Quantity of CO2 Captured and Extracted for EOR Operations (kt CO2)
Stage
2017
2018
2019
2020
2021
Quantity of C02 Captured
and Extracted for EOR
Operations
49,600
48,400
52,100
35,210
35,090
Several facilities are reporting under GHGRP Subpart RR (Geologic Sequestration of Carbon Dioxide). See Table
3-64 for the number of facilities reporting under Subpart RR, the reported CO2 sequestered in subsurface
geologic formations in each year, and of the quantity of CO2 emitted from equipment leaks in each year. The
quantity of CO2 sequestered and emitted is noted here for information purposes only; EPA is considering
updates to its approach in the Inventory for this source for future Inventories.
Table 3-64: Geologic Sequestration Information Reported Under GHGRP Subpart RR
Stage 2017 2018 2019 2020 2021
Number of Reporting Facilities 3 5 5 6 9
Reported Annual C02
Sequestered (kt) 5,958 7,662 8,332 6,802 6,947
3-92 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Reported Annual C02
Emissions from Equipment
Leaks (kt)
10
11
16
13
37
1
2
3 3.7 Natural Gas Systems (CRF Source
4 Category lB2b)
5 Note that this draft of the Inventory does not yet incorporate updated activity data products for the following data
6 inputs, due to a data base subscription lapse: gas well counts, wells drilled, wells completed, and production. Year
1 2020 values for activity data are used in place of year 2021. The Final Inventory (to be published April 2023) will
8 incorporate the latest activity data.
9 The U.S. natural gas system encompasses hundreds of thousands of wells, hundreds of processing facilities, and
10 over a million miles of transmission and distribution pipelines. This IPCC category (lB2b) is for fugitive emissions
11 from natural gas systems, which per IPCC guidelines include emissions from leaks, venting, and flaring. Total
12 greenhouse gas emissions (Cm, CO2, and N2O) from natural gas systems in 2021 were 218.3 MMT CO2 Eq., a
13 decrease of 12 percent from 1990 and a decrease of 2 percent from 2020, both primarily due to decreases in CH4
14 emissions. From 2010, emissions decreased by 3 percent, primarily due to decreases in Cm emissions. National
15 total dry gas production in the United States increased by 94 percent from 1990 to 2021, increased by 3 percent
16 from 2020 to 2021, and increased by 62 percent from 2010 to 2021. Of the overall greenhouse gas emissions
17 (218.3 MMT CO2 Eq.), 83 percent are Cm emissions (181.4 MMT CO2 Eq.), 17 percent are CO2 emissions (36.8
18 MMT), and less than 0.01 percent are N2O emissions (0.01 MMT CO2 Eq.).
19 Overall, natural gas systems emitted 181.4 MMT CO2 Eq. (6,479 kt CH4) of Cm in 2021, a 16 percent decrease
20 compared to 1990 emissions, and 2 percent decrease compared to 2020 emissions (see Table 3-66 and Table 3-67).
21 For non-combustion CO2, a total of 36.8 MMT CO2 Eq. (36,846 kt) was emitted in 2021, a 14 percent increase
22 compared to 1990 emissions, and a 2 percent increase compared to 2020 levels. The 2021 N2O emissions were
23 estimated to be 0.01 MMT CO2 Eq. (0.03 kt N2O), a 75 percent increase compared to 1990 emissions, and an 8
24 percent decrease compared to 2020 levels.
25 The 1990 to 2021 emissions trend is not consistent across segments or gases. Overall, the 1990 to 2021 decrease in
26 Cm emissions is due primarily to the decrease in emissions from the following segments: distribution (70 percent
27 decrease), transmission and storage (30 percent decrease), processing (40 percent decrease), and exploration (94
28 percent decrease). Over the same time period, the production segment saw increased CH4 emissions of 45 percent
29 (with onshore production emissions increasing 27 percent, offshore production emissions decreasing 86 percent,
30 and gathering and boosting [G&B] emissions increasing 110 percent), and post-meter emissions increasing by 60
31 percent. The 1990 to 2021 increase in CO2 emissions is primarily due to an increase in CO2 emissions in the
32 production segment, where emissions from flaring have increased over time.
33 Methane and CO2 emissions from natural gas systems include those resulting from normal operations, routine
34 maintenance, and system upsets. Emissions from normal operations include natural gas engine and turbine
35 uncombusted exhaust, flaring, and leak emissions from system components. Routine maintenance emissions
36 originate from pipelines, equipment, and wells during repair and maintenance activities. Pressure surge relief
37 systems and accidents can lead to system upset emissions. Emissions of N2O from flaring activities are included in
38 the Inventory, with most of the emissions occurring in the processing and production segments. Note, CO2
39 emissions exclude all combustion emissions (e.g., engine combustion) except for flaring CO2 emissions. All
40 combustion CO2 emissions (except for flaring) are accounted for in Section 3.1 CO2 from Fossil Fuel Combustion.
Energy 3-93
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Each year, some estimates in the Inventory are recalculated with improved methods and/or data. These
improvements are implemented consistently across the previous Inventory's time series (i.e., 1990 to 2020) to
ensure that the trend is representative of changes in emissions. Recalculations in natural gas systems in this year's
Inventory include:
• Methodological updates to five onshore production segment sources - pneumatic controllers, equipment
leaks, chemical injection pumps, storage tanks, and liquids unloading
• Recalculations due to Greenhouse Gas Reporting Program (GHGRP) submission revisions
• Recalculations due to updating the global warming potential (GWP) for Cm and N2O to use AR5 values.
Updates to well counts and produced water volumes were not available for Public Review estimates, and 2021
data were set equal to 2020.
The Recalculations Discussion section below provides more details on the updated methods.
Below is a characterization of the six emission subcategories of natural gas systems: exploration, production
(including gathering and boosting), processing, transmission and storage, distribution, and post-meter. Each of the
segments is described and the different factors affecting CH4, CO2, and N2O emissions are discussed.
Exploration. Exploration includes well drilling, testing, and completion. Emissions from exploration accounted for
less than 0.2 percent of CH4 emissions and of CO2 emissions from natural gas systems in 2021. Well completions
accounted for approximately 88 percent of CFU emissions from the exploration segment in 2021, with the rest
resulting from well testing and drilling. Well completion flaring emissions account for most of the CO2 emissions.
Methane emissions from exploration decreased by 94 percent from 1990 to 2021, with the largest decreases
coming from hydraulically fractured gas well completions without reduced emissions completions (RECs). Methane
emissions decreased 17 percent from 2020 to 2021 due to decreases in emissions from non-hydraulically fractured
well completions with venting. Methane emissions were highest from 2005 to 2008. Carbon dioxide emissions
from exploration decreased by 94 percent from 1990 to 2021 primarily due to decreases in hydraulically fractured
gas well completions. Carbon dioxide emissions from exploration decreased by 83 percent from 2020 to 2021 due
to decreases in emissions from hydraulically fractured gas well completions with flaring. Carbon dioxide emissions
were highest from 2006 to 2008. Nitrous oxide emissions decreased 98 percent from 1990 to 2021 and decreased
86 percent from 2020 to 2021.
Production (including gathering and boosting). In the production segment, wells are used to withdraw raw gas
from underground formations. Emissions arise from the wells themselves, and from well-site equipment and
activities such as pneumatic controllers, tanks and separators, and liquids unloading. Gathering and boosting
emission sources are included within the production sector. The gathering and boosting sources include gathering
and boosting stations (with multiple emission sources on site) and gathering pipelines. The gathering and boosting
stations receive natural gas from production sites and transfer it, via gathering pipelines, to transmission pipelines
or processing facilities (custody transfer points are typically used to segregate sources between each segment).
Boosting processes include compression, dehydration, and transport of gas to a processing facility or pipeline.
Emissions from production (including gathering and boosting) accounted for 52 percent of CH4 emissions and 25
percent of CO2 emissions from natural gas systems in 2021. Emissions from gathering and boosting and pneumatic
controllers in onshore production accounted for most of the production segment CH4 emissions in 2021. Within
gathering and boosting, the largest sources of CFU are compressor exhaust slip, compressor venting and leaks, and
tanks. Flaring emissions account for most of the CO2 emissions from production, with the highest emissions coming
from flare stacks at gathering stations, miscellaneous onshore production flaring, and tank flaring. Methane
emissions from production increased by 45 percent from 1990 to 2021, due primarily to increases in emissions
from pneumatic controllers (due to an increase in the number of controllers, particularly in the number of
intermittent bleed controllers) and increases in emissions from compressor exhaust slip in gathering and boosting.
Methane emissions decreased 3 percent from 2020 to 2021 due to decreases in emissions from pneumatic
controllers and liquids unloading. Carbon dioxide emissions from production increased by approximately a factor
of 2.7 from 1990 to 2021 due to increases in emissions at flare stacks in gathering and boosting and miscellaneous
onshore production flaring and increased 3 percent from 2020 to 2021 due primarily to increases in emissions
from tanks and acid gas removal units at gathering and boosting stations. Nitrous oxide emissions decreased by 28
3-94 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
percent from 1990 to 2021 and decreased 16 percent from 2020 to 2021. The decrease in N2O emissions from
1990 to 2021 and from 2020 to 2021 is primarily due to decreases in emissions from flaring at gathering and
boosting stations.
Processing. In the processing segment, natural gas liquids and various other constituents from the raw gas are
removed, resulting in "pipeline quality" gas, which is injected into the transmission system. Methane emissions
from compressors, including compressor seals, are the primary emission source from this stage. Most of the CO2
emissions come from acid gas removal (AGR) units, which are designed to remove CO2 from natural gas. Processing
plants accounted for 8 percent of Cm emissions and 71 percent of CO2 emissions from natural gas systems.
Methane emissions from processing decreased by 40 percent from 1990 to 2021 as emissions from compressors
(leaks and venting) and equipment leaks decreased; and increased 3 percent from 2020 to 2021 due to increased
emissions from gas engines. Carbon dioxide emissions from processing decreased by 8 percent from 1990 to 2021,
due to a decrease in AGR emissions, and increased 3 percent from 2020 to 2021 due to increased emissions from
AGR. Nitrous oxide emissions decreased 1 percent from 2020 to 2021.
Transmission and Storage. Natural gas transmission involves high pressure, large diameter pipelines that transport
gas long distances from field production and processing areas to distribution systems or large volume customers
such as power plants or chemical plants. Compressor station facilities are used to move the gas throughout the
U.S. transmission system. Leak CH4 emissions from these compressor stations and venting from pneumatic
controllers account for most of the emissions from this stage. Uncombusted compressor engine exhaust and
pipeline venting are also sources of Cm emissions from transmission. Natural gas is also injected and stored in
underground formations, or liquefied and stored in above ground tanks, during periods of low demand (e.g.,
summer), and withdrawn, processed, and distributed during periods of high demand (e.g., winter). Leak and
venting emissions from compressors are the primary contributors to CH4 emissions from storage. Emissions from
liquefied natural gas (LNG) stations and terminals are also calculated under the transmission and storage segment.
Methane emissions from the transmission and storage segment accounted for approximately 25 percent of
emissions from natural gas systems, while CO2 emissions from transmission and storage accounted for 4 percent of
the CO2 emissions from natural gas systems. Cm emissions from this source decreased by 30 percent from 1990 to
2021 due to reduced pneumatic device and compressor station emissions (including emissions from compressors
and leaks) and decreased 2 percent from 2020 to 2021 due to decreased emissions from pipeline venting
transmission compressors. CO2 emissions from transmission and storage were 4.7 times higher in 2021 than in
1990, due to increased emissions from LNG export terminals, and decreased by 16 percent from 2020 to 2021, also
due to LNG export terminals and flaring (both transmission and storage). The quantity of LNG exported from the
United States increased by a factor of 68 from 1990 to 2021, and by 49 percent from 2020 to 2021. LNG emissions
are about 1 percent of Cm and 89 percent of CO2 emissions from transmission and storage in year 2021. Nitrous
oxide emissions from transmission and storage increased by 165 percent from 1990 to 2021 and decreased 12
percent from 2020 to 2021.
Distribution. Distribution pipelines take the high-pressure gas from the transmission system at "city gate" stations,
reduce the pressure and distribute the gas through primarily underground mains and service lines to individual end
users. There were 1,337,012 miles of distribution mains in 2021, an increase of 392,855 miles since 1990 (PHMSA
2021). Distribution system emissions, which accounted for 8 percent of CH4 emissions from natural gas systems
and less than 1 percent of CO2 emissions, result mainly from leak emissions from pipelines and stations. An
increased use of plastic piping, which has lower emissions than other pipe materials, has reduced both CH4 and
CO2 emissions from this stage, as have station upgrades at metering and regulating (M&R) stations. Distribution
system Cm emissions in 2021 were 70 percent lower than 1990 levels and 1 percent lower than 2020 emissions.
Distribution system CO2 emissions in 2021 were 70 percent lower than 1990 levels and 1 percent lower than 2020
emissions. Annual CO2 emissions from this segment are less than 0.1 MMT CO2 Eq. across the time series.
Post-Meter. Post-meter includes 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 Cm emissions. Methane emissions from the
post-meter segment accounted for approximately 7 percent of emissions from natural gas systems in 2021. Post-
meter Cm emissions increased by 60 percent from 1990 to 2021 and increased by less than 1 percent from 2020 to
2021, due to increases in the number of residential houses using natural gas and increased natural gas
Energy 3-95
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
consumption at industrial facilities and power plants. CO2 emissions from post-meter account for less than 0.01
percent of total CO2 emissions from natural gas systems.
Total greenhouse gas emissions from the six subcategories within natural gas systems are shown in MMT CO2 Eq.
in Table 3-65. Total CH4 emissions for these same segments of natural gas systems are shown in MMT CO2 Eq.
(Table 3-66) and kt (Table 3-67). Most emission estimates are calculated using a net emission approach. However,
a few sources are still calculated with a potential emission approach. Reductions data are applied to those sources.
In 2021, 2.6 MMT CO2 Eq. CH4 is subtracted from production segment emissions, 4.3 MMT CO2 Eq. Cm is
subtracted from the transmission and storage segment, and 0.1 MMT CO2 Eq. Cm is subtracted from the
distribution segment to calculate net emissions. More disaggregated information on potential emissions, net
emissions, and reductions data is available in Annex 3.6, Methodology for Estimating CH4 and CO2 Emissions from
Natural Gas Systems.
Table 3-65: Total Greenhouse Gas Emissions (CH4, CO2, and N2O) from Natural Gas Systems
(MMT COz Eq.)
Segment
1990
2005
2017
2018
2019
2020
2021
Exploration
3.6
11.5
1.8
3.0
2.3
0.3
0.2
Production
68.1
102.4
111.5
116.2
115.5
106.2
103.2
Processing
52.2
31.8
35.8
36.3
40.4
39.3
40.4
Transmission and Storage
64.4
44.7
41.4
43.9
45.7
47.4
46.2
Distribution
51.0
28.5
15.7
15.6
15.5
15.5
15.3
Post-Meter
8.1
9.6
11.9
12.5
12.8
13.0
13.0
Total
247.5
228.6
218.2
227.4
232.3
221.7
218.3
Note: Totals may not sum due to independent rounding.
Table 3-66: ChU Emissions from Natural Gas Systems (MMT CO2 Eq.)
Segment
1990
2005
2017
2018
2019
2020
2021
Exploration
3.3
10.0
1.4
2.6
2.1
0.2
0.2
Production
64.7
97.9
103.5
107.0
104.7
97.3
94.0
Onshore Production
39.3
69.0
59.9
62.9
59.4
53.8
50.0
Gathering and Boosting
20.7
26.8
42.9
43.3
44.6
42.6
43.4
Offshore Production
4.8
2.0
0.7
0.8
0.7
0.9
0.7
Processing
23.9
13.0
12.9
13.5
14.2
13.9
14.3
Transmission and Storage
64.1
44.3
41.0
43.2
44.3
45.5
44.6
Distribution
50.9
28.5
15.7
15.6
15.5
15.5
15.3
Post-Meter
8.1
9.6
11.9
12.5
12.8
13.0
13.0
Total
215.1
203.4
186.4
194.4
193.6
185.4
181.4
Note: Totals may not sum due to independent rounding.
ible 3-67: ChU Emissions from Natural Gas Systems (kt)
Segment
1990
2005
2017
2018
2019
2020
2021
Exploration
119
358
49
94
75
9
7
Production
2,311
3,495
3,697
3,823
3,739
3,475
3,359
Onshore Production
1,403
2,464
2,139
2,246
2,122
1,923
1,786
Gathering and Boosting
739
958
1,533
1,547
1,591
1,520
1,548
Offshore Production
170
73
26
30
25
32
24
Processing
853
463
460
483
506
495
510
Transmission and Storage
2,289
1,584
1,465
1,542
1,584
1,625
1,592
3-96 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Distribution
Post-Meter
1,819 1,018 561 557 554 553 548
290 344 424 445 457 463 463
Total 7,682 7,263 6,657 6,943 6,915 6,620 6,479
Note: Totals may not sum due to independent rounding.
l Table 3-68: CO2 Emissions from Natural Gas Systems (MMT)
Segment
1990
2005
2017
2018
2019
2020
2021
Exploration
0.3
1.4
0.4
0.3
0.2
0.1
+
Production
3.3
4.6
8.0
9.1
10.9
8.9
9.1
Processing
28.3
18.8
22.9
22.8
26.2
25.4
26.1
Transmission and Storage
0.3
0.3
0.4
0.7
1.4
1.9
1.6
Distribution
0.1
+
+
+
+
+
+
Post-Meter
+
+
+
+
+
+
+
Total
32.4
25.2
31.8
33.0
38.7
36.3
36.8
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
2 Table 3-69: CO2 Emissions from Natural Gas Systems (kt)
Segment
1990
2005
2017
2018
2019
2020
2021
Exploration
297
1,434
444
336
220
96
17
Production
3,337
4,556
7,967
9,147
10,857
8,878
9,141
Processing
28,338
18,836
22,935
22,766
26,225
25,419
26,096
Transmission and Storage
336
349
405
707
1,384
1,884
1,574
Distribution
54
30
17
17
16
16
16
Post-Meter
1
1
2
2
2
2
2
Total
32,363
25,206
31,770
32,974
38,705
36,296
36,846
NE (Not Estimated)
Note: Totals may not sum due to independent rounding.
3 Table 3-70: N2O Emissions from Natural Gas Systems (Metric Tons CO2 Eq.)
Segment
1990
2005
2017
2018
2019
2020
2021
Exploration
355
1,090
217
156
103
45
6
Production
3,840
5,153
3,730
4,061
4,774
3,310
2,779
Processing
NO
2,977
2,643
2,998
5,081
4,349
4,300
Transmission and Storage
298
351
364
290
636
903
791
Distribution
NO
NO
NO
NO
NO
NO
NO
Post-Meter
NO
NO
NO
NO
NO
NO
NO
Total
4,494
9,572
6,953
7,506
10,594
8,608
7,877
NO (Not Occurring)
Note: Totals may not sum due to independent rounding.
4 Table 3-71: N2O Emissions from Natural Gas Systems (Metric Tons N2O)
Segment
1990
2005
2017
2018
2019
2020
2021
Exploration
1.3
4.1
0.8
0.6
0.4
0.2
0.0
Production
14.5
19.4
14.1
15.3
18.0
12.5
10.5
Processing
NO
11.2
10.0
11.3
19.2
16.4
16.2
Transmission and Storage
1.1
1.3
1.4
1.1
2.4
3.4
3.0
Distribution
NO
NO
NO
NO
NO
NO
NO
Post-Meter
NO
NO
NO
NO
NO
NO
NO
Total
17.0
36.1
26.2
28.3
40.0
32.5
29.7
NO (Not Occurring)
Note: Totals may not sum due to independent rounding.
Energy 3-97
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
Methodology and Time-Series Consistency
See Annex 3.6 for the full time series of emissions data, activity data, and emission factors, and additional
information on methods and data sources—for example, the specific years of reporting data from EPA's GHGRP
that are used to develop certain factors.
This section provides a general overview of the methodology for natural gas system emission estimates in the
Inventory, which involves the calculation of Cm, CO2, and N2O emissions for over 100 emissions sources (i.e.,
equipment types or processes), and then the summation of emissions for each natural gas segment.
The approach for calculating emissions for natural gas systems generally involves the application of emission
factors to activity data. For most sources, the approach uses technology-specific emission factors or emission
factors that vary over time and take into account changes to technologies and practices, which are used to
calculate net emissions directly. For others, the approach uses what are considered "potential methane factors"
and emission reduction data to calculate net emissions. The estimates are developed with an IPCC Tier 2 approach.
Tier 1 approaches are not used.
Emission Factors. Key references for emission factors for CH4 and CO2 emissions from the U.S. natural gas industry
include a 1996 study published by the Gas Research Institute (GRI) and EPA (GRI/EPA 1996), EPA's GHGRP (EPA
2022), and others.
The 1996 GRI/EPA study developed over 80 CH4 emission factors to characterize emissions from the various
components within the operating segments of the U.S. natural gas system. The GRI/EPA study was based on a
combination of process engineering studies, collection of activity data, and measurements at representative
natural gas facilities conducted in the early 1990s. Year-specific natural gas CFU compositions are calculated using
U.S. Department of Energy's Energy Information Administration (EIA) annual gross production data for National
Energy Modeling System (NEMS) oil and gas supply module regions in conjunction with data from the Gas
Technology Institute (GTI, formerly GRI) Unconventional Natural Gas and Gas Composition Databases (GTI 2001).
These year-specific CFU compositions are applied to emission factors, which therefore may vary from year to year
due to slight changes in the CFU composition of natural gas for each NEMS region.
GHGRP Subpart W data were used to develop CH4, CO2, and N2O emission factors for many sources in the
Inventory. In the exploration and production segments, GHGRP data were used to develop emission factors used
for all years of the time series for well testing, gas well completions and workovers with and without hydraulic
fracturing, pneumatic controllers and chemical injection pumps, condensate tanks, liquids unloading,
miscellaneous flaring, gathering and boosting pipelines, and certain sources at gathering and boosting stations. In
the processing segment, for recent years of the times series, GHGRP data were used to develop emission factors
for leaks, compressors, flares, dehydrators, and blowdowns/venting. In the transmission and storage segment,
GHGRP data were used to develop factors for all years of the time series for LNG stations and terminals and
transmission pipeline blowdowns, and for pneumatic controllers for recent years of the times series.
Other data sources used for CH4 emission factors include Zimmerle et al. (2015) for transmission and storage
station leaks and compressors, GTI (2009 and 2019) for commercial and industrial meters, Lamb et al. (2015) for
recent years for distribution pipelines and meter/regulator stations, Zimmerle et al. (2019) for gathering and
boosting stations, Bureau of Ocean Energy Management (BOEM) reports, and Fischer et al. (2019) and IPCC (2019)
for post-meter emissions.
For CO2 emissions from sources in the exploration, production and processing segments that use emission factors
not directly calculated from GHGRP data, data from the 1996 GRI/EPA study and a 2001 GTI publication were used
to adapt the CH4 emission factors into related CO2 emission factors. For sources in the transmission and storage
segment that use emission factors not directly calculated from GHGRP data, and for sources in the distribution
segment, data from the 1996 GRI/EPA study and a 1993 GTI publication were used to adapt the CH4 emission
factors into non-combustion related CO2 emission factors. CO2 emissions from post-meter sources (commercial,
industrial and vehicles) were estimated using default emission factors from IPCC (2019). Carbon dioxide emissions
from post-meter residential sources are included in fossil fuel combustion data.
3-98 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Flaring N2O emissions were estimated for flaring sources using GHGRP data.
2 See Annex 3.6 for more detailed information on the methodology and data used to calculate CH4, CO2, and N2O
3 emissions from natural gas systems.
4 Activity Data. Activity data were taken from various published data sets, as detailed in Annex 3.6. Key activity data
5 sources include data sets developed and maintained by EPA's GHGRP (EPA 2022); Enverus (Enverus 2021); BOEM;
6 Federal Energy Regulatory Commission (FERC); EIA; the Natural Gas STAR and Methane Challenge Programs annual
7 data; Oil and Gas Journal; and PHMSA. Enverus data for 2021 are not currently available; this public review version
8 of the Inventory uses 2020 data as proxy for 2021.
9 For a few sources, recent direct activity data are not available. For these sources, either 2020 data were used as a
10 proxy for 2021 data, or a set of industry activity data drivers was developed and used to calculate activity data over
11 the time series. Drivers include statistics on gas production, number of wells, system throughput, miles of various
12 kinds of pipe, and other statistics that characterize the changes in the U.S. natural gas system infrastructure and
13 operations. More information on activity data and drivers is available in Annex 3.6.
14 A complete list of references for emission factors and activity data by emission source is provided in Annex 3.6.
15 Calculating Net Emissions. For most sources, net emissions are calculated directly by applying emission factors to
16 activity data. Emission factors used in net emission approaches reflect technology-specific information, and take
17 into account regulatory and voluntary reductions. However, for production, transmission and storage, and
18 distribution, some sources are calculated using potential emission factors, and CH4 that is not emitted is deducted
19 from the total CH4 potential estimates. To take into account use of such technologies and practices that result in
20 lower emissions but are not reflected in "potential" emission factors, data are collected on both regulatory and
21 voluntary reductions. Regulatory actions addressed using this method include EPA National Emission Standards for
22 Hazardous Air Pollutants (NESHAP) regulations for dehydrator vents. Voluntary reductions included in the
23 Inventory are those reported to Natural Gas STAR and Methane Challenge for certain sources. Natural Gas STAR
24 and Methane Challenge reductions were reassessed for this Inventory, see the Recalculations Discussion for more
25 information.
26 Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
27 through 2020. GHGRP data available (starting in 2011) and other recent data sources have improved estimates of
28 emissions from natural gas systems. To develop a consistent time series, for sources with new data, EPA reviewed
29 available information on factors that may have resulted in changes over the time series (e.g., regulations, voluntary
30 actions) and requested stakeholder feedback on trends as well. For most sources, EPA developed annual data for
31 1993 through 2010 by interpolating activity data or emission factors or both between 1992 and 2011 data points.
32 Information on time-series consistency for sources updated in this year's Inventory can be found in the
33 Recalculations Discussion below, with additional detail provided in supporting memos (relevant memos are cited in
34 the Recalculations Discussion). For detailed documentation of methodologies, please see Annex 3.5.
35 Through EPA's stakeholder process on oil and gas in the Inventory, EPA received stakeholder feedback on updates
36 under consideration for the Inventory. Stakeholder feedback is noted below in Recalculations Discussion and
37 Planned Improvements.
38 The United States reports data to the UNFCCC using this Inventory report along with Common Reporting Format
39 (CRF) tables. This note is provided for those reviewing the CRF tables: The notation key "IE" is used for CO2 and CH4
40 emissions from venting and flaring in CRF table l.B.2. Disaggregating flaring and venting estimates across the
41 Inventory would involve the application of assumptions and could result in inconsistent reporting and, potentially,
42 decreased transparency. Data availability varies across segments within oil and gas activities systems, and emission
43 factor data available for activities that include flaring can include emissions from multiple sources (flaring, venting
44 and leaks).
45 Uncertainty - TO BE UPDATED FOR FINAL INVENTORY REPORT
46 EPA has conducted a quantitative uncertainty analysis using the IPCC Approach 2 methodology (Monte Carlo
Energy 3-99
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Simulation technique) to characterize the uncertainty for natural gas systems. For more information on the
approach, please see the memoranda Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2016: Natural
Gas and Petroleum Systems Uncertainty Estimates and Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-
2019: Update for Natural Gas and Petroleum Systems CO2 Uncertainty Estimates,82
EPA used Microsoft Excels @RISK add-in tool to estimate the 95 percent confidence bound around CH4 and CO2
emissions from natural gas systems for the current Inventory. For the CH4 uncertainty analysis, EPA focused on the
16 highest-emitting sources for the year 2020, which together emitted 76 percent of methane from natural gas
systems in 2020, and extrapolated the estimated uncertainty for the remaining sources. For the CO2 uncertainty
analysis, EPA focused on the 3 highest-emitting sources for the year 2020, which together emitted 80 percent of
CO2 from natural gas systems in 2020, and extrapolated the estimated uncertainty for the remaining sources. To
estimate uncertainty for N2O, EPA applied the uncertainty bounds calculated for CO2. EPA will seek to refine this
estimate in future Inventories. The @ RISK add-in provides for the specification of probability density functions
(PDFs) for key variables within a computational structure that mirrors the calculation of the inventory estimate.
The IPCC guidance notes that in using this method, "some uncertainties that are not addressed by statistical means
may exist, including those arising from omissions or double counting, or other conceptual errors, or from
incomplete understanding of the processes that may lead to inaccuracies in estimates developed from models."
The uncertainty bounds reported below only reflect those uncertainties that EPA has been able to quantify and do
not incorporate considerations such as modeling uncertainty, data representativeness, measurement errors,
misreporting or misclassification. The understanding of the uncertainty of emission estimates for this category
evolves and improves as the underlying methodologies and datasets improve.
The results presented below provide the 95 percent confidence bound within which actual emissions from this
source category are likely to fall for the year 2020, using the IPCC methodology. The results of the Approach 2
uncertainty analysis are summarized in Table 3-72. Natural gas systems CFU emissions in 2020 were estimated to
be between 135.2 and 194.6 MMT CO2 Eq. at a 95 percent confidence level. Natural gas systems CO2 emissions in
2020 were estimated to be between 29.7 and 42.2 MMT CO2 Eq. at a 95 percent confidence level. Natural gas
systems N2O emissions in 2020 were estimated to be between 0.009 and 0.012 MMT CO2 Eq. at a 95 percent
confidence level.
Uncertainty bounds for other years of the time series have not been calculated, but uncertainty is expected to vary
over the time series. For example, years where many emission sources are calculated with interpolated data would
likely have higher uncertainty than years with predominantly year-specific data. In addition, the emission sources
that contribute the most to CFU and CO2 emissions are different over the time series, particularly when comparing
recent years to early years in the time series. For example, venting emissions were higher and flaring emissions
were lower in early years of the time series, compared to recent years. Technologies also changed over the time
series (e.g., liquids unloading with plunger lifts and reduced emissions completions were not used early in the time
series and cast iron distribution mains were more prevalent than plastic mains in early years). Transmission and
gas processing compressor leak and vent emissions were also higher in the early years of the time series.
Table 3-72: Approach 2 Quantitative Uncertainty Estimates for CH4 and Non-combustion CO2
Emissions from Natural Gas Systems (MMT CO2 Eq. and Percent)
2020 Emission Estimate Uncertainty Range Relative to Emission Estimate3
source uas
(MMT C02 Eq.)b
(MMT C02
Eq.)
(%)
Lower
Boundb
Upper
Boundb
Lower
Boundb
Upper
Boundb
Natural Gas Systems CH4 164.9
135.2
194.6
-18%
+18%
Natural Gas Systems C02 35.4
29.7
42.3
-16%
+19%
Natural Gas Systems N20 +
+
+
-16%
+19%
+ Less than 0.05 MMT C02 Eq.
82 See https://www.epa.gov/ghgemissions/natural-gas-and-petroleum-svstems.
3-100 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
a Range of emission estimates estimated by applying the 95 percent confidence intervals obtained from the Monte Carlo
Simulation analysis conducted for the year 2020 CH4 and C02 emissions.
b All reported values are rounded after calculation. As a result, lower and upper bounds may not be duplicable from other
rounded values as shown in Table 3-66 and Table 3-67.
QA/QC and Verification Discussion
The natural gas systems emission estimates in the Inventory are continually being reviewed and assessed to
determine whether emission factors and activity factors accurately reflect current industry practices. A QA/QC
analysis was performed for data gathering and input, documentation, and calculation. QA/QC checks are
consistently conducted to minimize human error in the model calculations. EPA performs a thorough review of
information associated with new studies, GHGRP data, regulations, public webcasts, and the Natural Gas STAR
Program to assess whether the assumptions in the Inventory are consistent with current industry practices. The
EPA has a multi-step data verification process for GHGRP data, including automatic checks during data-entry,
statistical analyses on completed reports, and staff review of the reported data. Based on the results of the
verification process, the EPA follows up with facilities to resolve mistakes that may have occurred.83
As in previous years, EPA conducted early engagement and communication with stakeholders on updates prior to
public review of the current Inventory. EPA held stakeholder webinars in September and November of 2022. EPA
released memos detailing updates under consideration and requesting stakeholder feedback.
In recent years, several studies have measured emissions at the source level and at the national or regional level
and calculated emission estimates that may differ from the Inventory. There are a variety of potential uses of data
from new studies, including replacing a previous estimate or factor, verifying or QA of an existing estimate or
factor, and identifying areas for updates. In general, there are two major types of studies related to oil and gas
greenhouse gas data: studies that focus on measurement or quantification of emissions from specific activities,
processes and equipment, and studies that use tools such as inverse modeling to estimate the level of overall
emissions needed to account for measured atmospheric concentrations of greenhouse gases at various scales. The
first type of study can lead to direct improvements to or verification of Inventory estimates. In the past few years,
EPA has reviewed and in many cases, incorporated data from these data sources. The second type of study can
provide general indications of potential over- and under-estimates. In addition, in recent years information from
top-down studies has been directly incorporated to quantify emissions from well blowouts.
A key challenge in using these types of studies to assess Inventory results is having a relevant basis for comparison
(e.g., the two data sets should have comparable time frames and geographic coverage, and the independent study
should assess data from the Inventory and not another data set, such as the Emissions Database for Global
Atmospheric Research, or "EDGAR"). In an effort to improve the ability to compare the national-level Inventory
with measurement results that may be at other spatial or temporal scales, a team at Harvard University along with
EPA and other coauthors developed a gridded inventory of U.S. anthropogenic methane emissions with 0.1 degree
x 0.1 degree spatial resolution, monthly temporal resolution, and detailed scale-dependent error
characterization.84 The gridded methane inventory is designed to be consistent with the U.S. EPA's Inventory of
U.S. Greenhouse Gas Emissions and Sinks: 1990-2014 estimates for the year 2012, which presents national totals 85
An updated version of the gridded inventory is being developed and will improve efforts to compare results of the
Inventory with atmospheric studies.
83 See https://www.epa.eov/sites/production/files/2015-07/documents/eherp verification factsheet.pdf.
84 See https://www.epa.eov/eheemissions/eridded-2012-methane-emissions.
85 See https://www.epa.eov/eheemissions/us-ereenhouse-eas-inventorv-report-1990-2014.
Energy 3-101
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Recalculations Discussion
EPA received information and data related to the emission estimates through GHGRP reporting and stakeholder
feedback on updates under consideration. In October 2022, EPA released a draft memorandum that discussed
changes under consideration and requested stakeholder feedback on those changes.86 EPA did not receive written
feedback on the memorandum. Memoranda cited in the Recalculations Discussion below are: Inventory of U.S.
Greenhouse Gas Emissions and Sinks 1990-2021: Updates Under Consideration for Incorporating Additional
Geographically Disaggregated Data (Disaggregation memo) and Inventory of U.S. Greenhouse Gas Emissions and
Sinks 1990-2021: Updates Under Consideration for Incorporating Additional Geographically Disaggregated Data for
the Production Segment (Production Disaggregation memo).
In this Inventory, an update that incorporates additional basin-level data from GHGRP subpart W was implemented
for several emission sources in the onshore production segment. The update seeks to improve the ability of EPA's
gridded and state inventories to reflect variation due to differences in formation types, technologies and practices,
regulations, or voluntary initiatives, and not only the differences in key activity levels that are reflected in the
current gridded and state inventories. This would allow EPA to use the gridded inventory for improved
comparisons of the national Inventory with various atmospheric observation studies (since regions will better
reflect the local differences in emissions rates as reported to GHGRP) and would allow the state-level inventory to
reflect differences in state-level programs, formation type mixes, and varying technologies and practices. For many
sources, an approach that develops estimates using geographically disaggregated data may not be possible or
preferable to a national level approach based on the currently available data. For some emission sources in the
Inventory, emission factor data come from research studies and are applied at the national level. For example,
many of the emission factors used to quantify emissions in the Inventory for the gathering and boosting,
transmission and storage, distribution, and post-meter segments are from research studies and do not have a level
of detail or total population comparable to GHGRP. Even in cases where geographically disaggregated data are
available, such an approach may not always be preferable. In cases with limited variation between areas, such an
approach would have limited impact on emissions estimates regionally or nationally. In cases with limited data in
certain areas, disaggregated approaches might substantially increase the uncertainty of estimates and basin-
specific calculations would not be an improvement over use of a national average. EPA continues to seek
stakeholder feedback on the draft approach in this Inventory.
EPA evaluated relevant information available and made several updates to the Inventory, including for pneumatic
controllers, equipment leaks, chemical injection pumps, storage tanks, and liquids unloading. For each of these
emission sources, EPA modified the calculation methodology to use GHGRP data to develop basin-specific activity
factors and/or emission factors. General information for these source specific recalculations are presented below
and details are available in the Disaggregation memo and Production Disaggregation memo, including additional
considerations for the updates.
In addition to the production segment sources mentioned above, for certain sources, CH4 and/or CO2 emissions
changed by greater than 0.05 MMT CO2 Eq., comparing the previous estimate for 2020 to the current
(recalculated) estimate for 2020. The emissions changes were mostly due to GHGRP data submission revisions.
These sources are discussed below and include miscellaneous production flaring, offshore production, distribution
pipelines, and post-meter emissions.
In addition, for the current Inventory, CC>2-equivalent emissions totals have been revised to reflect the 100-year
global warming potentials (GWPs) provided in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). AR5 GWP
values differ slightly from those presented in the IPCC Fourth Assessment Report (AR4) (IPCC 2007) used in the
previous inventories. The AR5 GWPs have been applied across the entire time series for consistency. The GWP of
CH4 has increased from 25 to 28, leading to an increase in the calculated CC>2-equivalent emissions of CH4, while
the GWP of N2O has decreased from 298 to 265, leading to a decrease in the calculated CC>2-equivalent emissions
86 Stakeholder materials including draft memoranda for the current (i.e., 1990 to 2021) Inventory are available at
https://www.epa.gov/ehgemissions/natural-gas-and-petroleum-systems.
3-102 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 of N2O. Further discussion on this update and the overall impacts of updating the Inventory GWP values to reflect
2 the IPCC Fifth Assessment Report can be found in Chapter 9, Recalculations and Improvements.
3 The combined impact of revisions to 2020 natural gas systems CH4 emissions, compared to the previous Inventory,
4 is an increase from 164.9 to 185.4 MMT CO2 Eq. (20.5 MMT CO2 Eq., or 12 percent). The recalculations resulted in
5 an average increase in the annual CH4 emission estimates across the 1990 through 2020 time series, compared to
6 the previous Inventory, of 24.1 MMT CO2 Eq., or 14 percent.
7 The combined impact of revisions to 2020 natural gas systems CO2 emissions, compared to the previous Inventory,
8 is an increase from 35.4 MMT to 36.3 MMT, or 2.7 percent. The recalculations resulted in an average increase in
9 emission estimates across the 1990 through 2020 time series, compared to the previous Inventory, of 0.4 MMT
10 CO2 Eq., or 1.3 percent.
11 The combined impact of revisions to 2020 natural gas systems N2O emissions, compared to the previous Inventory,
12 is a decrease from 10.2 kt CO2 Eq. to 8.6 kt CO2 Eq., or 15 percent. The recalculations resulted in an average
13 decrease in emission estimates across the 1990 through 2020 time series, compared to the previous Inventory, of
14 11 percent.
15 In Table 3-73 and Table 3-74 below are categories in Natural Gas Systems with recalculations resulting in a change
16 of greater than 0.05 MMT CO2 Eq., comparing the previous estimate for 2019 to the current (recalculated)
17 estimate for 2019. No changes made to N2O estimates resulted in a change greater than 0.05 MMT CO2 Eq. For
18 more information, please see the Recalculations Discussion below.
19 For certain sources, the change in GWP for CH4 alone (i.e., not the results of other recalculations) resulted in
20 calculated CFU CC>2-equivalent emissions for 2020 changing by greater than 0.05 MMT CO2 Eq., compared to the
21 previous Inventory. These sources are not discussed below. The production segment sources impacted by the GWP
22 update are: wellhead leaks, produced water, dehydrator kimray pumps, gas engine exhaust, G&B compressors,
23 G&B pneumatic controllers, G&B pneumatic pumps, G&B combustion slip, G&B yard piping, and G&B pipeline
24 leaks. The natural gas processing sources impacted by the GWP update are: reciprocating compressors, gas engine
25 exhaust, and blowdowns. The transmission and storage sources impacted by the GWP update are: compressor
26 station leaks, reciprocating compressors, centrifugal compressors, M&R, gas engine exhaust, pneumatic
27 controllers, pipeline venting, and compressor station venting. The distribution sources impacted by the GWP
28 update are distribution main and service leaks, customer meters, and mishaps.
29 Table 3-73: Recalculations of CO2 in Natural Gas Systems (MMT CO2)
Segment and Emission Sources with
Changes of Greater than 0.05 MMT C02
due to Recalculations
Previous Estimate
Year 2020,
2022 Inventory
Current Estimate
Year 2020,
2023 Inventory
Current Estimate
Year 2021,
2023 Inventory
Exploration
0.1
0.1
+
Production
7.7
8.9
9.1
Misc. Onshore Production Flaring
1.1
1.3
1.0
Large Tanks with Flares
0.6
0.8
0.8
Liquids Unloading
+
+
+
G&B Station Sources
5.8
6.5
7.1
Processing
25.5
25.4
26.1
Flares
7.9
8.1
7.4
Transmission and Storage
2.0
1.9
1.6
Distribution
+
+
+
Post-Meter
+
+
+
Total
35.4
36.3
36.8
+ Does not exceed 0.05 MMT C02.
Energy 3-103
-------
1
Table 3-74: Recalculations of CH4 in Natural Gas Systems (MMT CO2 Eq.)
Segment and Emission Sources with Changes of
Greater than 0.05 MMT C02 due to
Recalculations
Previous
Estimate Year
2020,
2022 Inventory
Current Estimate
Year 2020, 2023
Inventory
Current Estimate
Year 2021, 2023
Inventory
Exploration
0.2
0.2
0.2
Production
86.4
97.3
94.0
Well pad Equipment Leaks
6.6
10.3
9.6
Chemical Injection Pumps
2.8
2.4
2.1
Pneumatic Controllers
23.8
22.8
21.3
Tanks
0.4
1.5
1.2
Liquids Unloading
3.2
4.5
3.4
G&B Station Sources
34.1
38.7
39.8
Processing
12.4
13.9
14.3
Transmission and Storage
40.6
45.5
44.6
Distribution
13.9
15.5
15.3
Pipeline Mains - Unprotected Steel
1.0
1.1
1.0
Post-Meter
11.5
13.0
13.0
Total
164.9
185.4
181.4
2 Exploration
3 There were no methodological updates to the exploration segment, and recalculations due to updated data
4 resulted in average decreases in calculated CFU and CO2 emissions over the time series of less than 1 percent.
5 Production
6 Pneumatic Controllers (Methodological Update)
1 EPA updated the calculation methodology for pneumatic controllers to use basin-specific activity factors and
8 emission factors calculated from subpart W data for each type of controller (i.e., high, intermittent, and low
9 bleed). Previously, national average activity and emission factors calculated using subpart W data were applied to
10 estimate pneumatic controller emissions. In this methodological update, EPA summed basin-level emissions
11 together to develop national emissions. The Disaggregation memo and Production Disaggregation memo present
12 additional information and considerations for this update.
13 EPA calculated basin-specific activity factors and CH4 emission factors were calculated for all basins that reported
14 subpart W data. The factors were year-specific for RY2011 through RY2021. EPA retained the previous Inventory's
15 activity factor assumptions for 1990 through 1992 and applied linear interpolation between the 1992 and 2011
16 activity factors at the basin-level. Year 2011 emission factors were applied to all prior years for each basin. For
17 basins without subpart W data available, EPA applied national average activity and emission factors.
18 The estimation methodology for CO2 emissions was not updated to use the basin-specific approach for the public
19 review version of the Inventory. CO2 emissions were estimated by applying a CO2 to CFU ratio to the estimated CH4
20 emissions. EPA will calculate pneumatic controller CO2 emissions in the same manner as CFU emissions for the final
21 Inventory.
22 As a result of this methodological update, CFU emissions estimates are on average 4 percent higher across the
23 time-series than in the previous Inventory. The estimate for 2020 is 14 percent lower than in the previous
24 Inventory. Pneumatic controller CH4 emissions were higher for all years between 1990 through 2011 by an average
25 of 8 percent and CFU emissions were lower for 2011 through 2020 by an average of 6 percent, compared to the
26 previous inventory. Emissions were lower in recent years due to some basins having slightly lower activity factors
27 and/or emission factors for intermittent bleed pneumatic controllers, compared to the national average. Emissions
3-104 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 were higher in early years of the time series due to basins having higher emission factors than the national
2 average. Multiple basins impact the emissions changes for pneumatic controllers at gas wells.
3 Table 3-75: Pneumatic Controllers National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2017
2018
2019
2020
2021
Low Bleed Controllers
High Bleed Controllers
Intermittent Bleed Controllers
Total Emissions
0
350,535
230,504
581,039
22,745
483,375
569,592
1,075,712
32,360
108,533
873,015
1,013,908
33,805
87,071
835,249
956,125
31,475
53,233
874,372
959,080
27,364
42,332
744,622
814,318
25,609
42,828
692,097
760,534
Previous Estimate
510,354
1,041,503
1,104,896
1,072,874
1,024,678
950,718
NA
NA (Not Applicable)
4 Storage Tanks (Methodological Update)
5 EPA updated the calculation methodology for production segment storage tanks to use basin-specific activity
6 factors and emission factors calculated from Subpart W data for each storage tank category. Previously, national
7 annual average activity and emission factors calculated using Subpart W data were applied to estimate storage
8 tank emissions. In this methodological update, EPA summed basin-level emissions together to develop national
9 emissions. The calculation methodology was updated to estimate Cm and CO2 emissions using basin-level data
10 from subpart W. The Production Disaggregation memo presents additional information and considerations for this
11 update.
12 EPA calculated basin-specific activity factors and CH4 and CO2 emission factors for all basins that reported subpart
13 W data. The factors were year-specific for reporting year (RY) 2015 through RY2021. EPA also retained the previous
14 Inventory's activity factor assumptions for 1990 and used linear interpolation between the 1990 and 2015 activity
15 factors at the basin-level. Year 2015 emission factors were applied to all prior years for each basin. For basins
16 without Subpart W data available, EPA applied national average activity and emission factors.
17 This update resulted in CH4 emission estimates an average of 276 percent higher across the time series compared
18 with the previous Inventory. The estimate for 2020 is 210 percent higher than in the previous Inventory. Storage
19 tank CO2 emissions are an average of 43 percent higher across the time series compared to the previous Inventory.
20 The 2020 emission estimate is 50 percent higher than in the previous Inventory.
21 The basin-level approach's emissions increased because certain basins with high liquids production and storage
22 tank throughput had higher emission factors and/or activity factors than the national average. The time-series is
23 also impacted as the basin-level approach reflects changing levels of liquids production, and hence storage tank
24 throughput, for basins across the time-series; basins with more production and storage tank throughput in the
25 early 90s also corresponded to basins with higher emission factors and/or activity factors than the national
26 average. For CH4, this is particularly noticeable for basins with small tanks without flares (e.g., Arkoma Basin, Bend
27 Arch, Central Western Overthrust, East Texas, Piceance) and for CO2 emissions this is noticeable for basins using
28 large tanks with flares (e.g., Anadarko Basin, Appalachian, Chautauqua Platform, Denver, Gulf Coast, Permian,
29 South Oklahoma Folded Belt).
30 Table 3-76: Storage Tanks National Cm Emissions (Metric Tons Cm)
Source
1990
2005
2017
2018
2019
2020
2021
Large Tanks w/Flares
505
336
1,016
1,273
789
600
606
Large Tanks w/VRU
0
27
205
143
905
525
371
Large Tanks w/o Control
16,161
6,867
6,622
15,416
2,446
4,284
4,916
Small Tanks w/Flares
0
51
249
237
208
201
168
Small Tanks w/o Flares
89,757
31,176
40,152
43,448
63,168
47,749
37,959
Malfunctioning Separator Dump
Valves
7
4
648
40
80
254
197
Total Emissions
106,429
38,461
48,892
60,556
67,595
53,613
44,217
Previous Estimate
16,421
11,331
21,493
24,435
21,194
17,294
NA
Energy 3-105
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
NA (Not Applicable)
Table 3-77: Storage Tanks National CO2 Emissions (kt CO2)
Source
1990
2005
2017
2018
2019
2020
2021
Large Tanks w/Flares
579
422
1,804
1,356
840
795
825
Large Tanks w/VRU
0
2
0
0
1
1
1
Large Tanks w/o Control
2
1
1
37
1
1
1
Small Tanks w/Flares
0
13
72
87
82
41
28
Small Tanks w/o Flares
47
18
23
26
33
24
18
Malfunctioning Separator Dump
Valves
0
0
2
0
0
1
0
Total Emissions
628
456
1,902
1,507
956
862
873
Previous Estimate
298
380
1,131
844
634
574
NA
NA (Not Applicable)
Equipment Leaks (Methodological Update)
EPA updated the calculation methodology for onshore production equipment leaks to use basin-specific
equipment-level activity factors (e.g., separators per well) from GHGRP data. Previously, national average
equipment activity factors developed using RY2014 GHGRP data were used in the Inventory for all years. In this
methodological update, EPA summed basin-level emissions together to develop national emissions. The
Disaggregation memo and Production Disaggregation memo present additional information and considerations
for this update.
EPA calculated basin-specific equipment-level activity factors for all basins that reported Subpart W data. The
factors were year-specific for RY2015 through RY2021. EPA also retained the previous Inventory's activity factors
for 1990 through 1992 and used linear interpolation between the 1992 and 2015 activity factors at the basin-level.
For basins without subpart W data available, EPA applied national average activity factors. This methodological
update applies only for activity factors. The previous Inventory's CH4 emission factors for onshore production
segment equipment leaks (by equipment type) were retained and used to develop CFU estimates. Since the CFU
emission factors were not updated, EPA also retained the Gas STAR reductions that are applicable to equipment
leaks.
The calculation methodology for CO2 emissions was not updated for the public review version of the Inventory.
The previous Inventory's methodology was retained to develop CO2 estimates. EPA will calculate equipment leak
CO2 emissions in the same manner as CH4 emissions for the final Inventory.
This update resulted in CH4 emission estimates an average of 8 percent higher across the time series compared to
the previous Inventory. The 2020 emission estimate is 39 percent higher than in the previous Inventory. The early
years of the time series are minimally impacted by the update, with average CH4 emissions 1 percent lower for
years 1990 through 2002, compared to the previous Inventory. Methane emissions are an average of 14 percent
higher for 2002 through 2020, compared to the previous Inventory. These recent years of the time series relied on
the basin-specific activity factors and certain basins had higher activity factors compared to the national average
factors (e.g., Anadarko Basin, Arkla, Fort Worth Syncline, Gulf Coast, Powder River, San Juan, Strawn).
Table 3-78: Production Equipment Leaks National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2017
2018
2019
2020
2021
Heaters
12,116
20,307
20,068
80,312
16,421
19,223
17,694
Separators
40,746
92,060
129,978
124,339
128,675
132,409
112,425
Dehydrators
12,722
12,796
4,485
5,552
3,739
3,133
4,128
Meters/Piping
42,205
72,148
78,403
81,139
85,625
154,544
135,476
Compressors
29,858
64,877
73,000
72,026
64,471
60,157
73,963
Gas STAR Reductions for Leaks
0
20,908
2,748
71
133
133
133
Total Emissions
137,647
239,280
303,187
363,296
298,797
369,333
343,553
3-106 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Previous Estimate 138,844 220,489 273,028 274,664 270,662 265,657 NA
NA (Not Applicable)
1 Chemical injection Pumps (Methodological Update)
2 EPA updated the calculation methodology for chemical injection pumps to use basin-specific activity factors from
3 GHGRP data. Previously, national average activity factors developed using RY2014 GHGRP data were used in the
4 Inventory for all years. In this methodological update, EPA summed basin-level emissions together to develop
5 national emissions. The Disaggregation memo and Production Disaggregation memo present additional
6 information and considerations for this update.
7 EPA calculated basin-specific activity factors for all basins that reported subpart W data. The factors were year-
8 specific for RY2015 through RY2021. EPA also retained the previous Inventory's activity factors for 1990 through
9 1992 and applied linear interpolation between the 1992 and 2015 activity factors at the basin-level. For basins
10 without subpart W data available, EPA applied national average activity factors. This methodological update
11 applies only to activity factors. The previous Inventory's Cm emission factor for chemical injection pumps was
12 retained and used to develop Cm estimates.
13 The estimation methodology for CO2 emissions was not updated for the public review version of the Inventory. The
14 previous Inventory's methodology was retained to develop CO2 estimates. EPA will calculate chemical injection
15 pump CO2 emissions in the same manner as CH4 emissions for the final Inventory.
16 This update resulted in CH4 emission estimates an average of 86 percent higher across the time-series. The 2020
17 emission estimate is 24 percent lower than in the previous Inventory. The emissions increase across the time-
18 series is predominantly due to the Bend Arch, which has a very high RY2015 activity factor (chemical injection
19 pumps per well), which then impacts prior years because it's used in the linear interpolation back to the 1992
20 activity factor.
21 Table 3-79: Chemical Injection Pumps National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2017
2018
2019
2020
2021
Chemical Injection Pumps
25,345
183,832
113,726
120,984
108,546
84,002
76,315
Previous Estimate
27,158
84,573
116,107
115,140
113,538
110,785
NA
NA (Not Applicable)
22 Liquids Unloading (Methodological Update)
23 EPA updated the calculation methodology for liquids unloading to use basin-specific activity factors and emission
24 factors calculated from subpart W data for each type of liquids unloading (i.e., with and without plunger lifts).
25 Previously, national average activity and emission factors calculated using Subpart W data were applied to
26 estimate liquids unloading emissions. In this methodological update, EPA summed basin-level emissions together
27 to develop national emissions. The Disaggregation memo and Production Disaggregation memo present additional
28 information and considerations for this update.
29 EPA calculated basin-specific activity factors, and CFU and CO2 emission factors for all basins that reported subpart
30 W data. The factors were also year-specific for RY2011 through RY2021. EPA also revised the previous Inventory's
31 activity factor and emission factor assumptions for 1990 through 1992. Previously, Year 2011 emission factors
32 were applied to all prior years of the time series and activity factors were derived by linear interpolation between
33 Year 2011 data and API/ANGA data (collected in 2011) for 1990. In the current Inventory, EPA used activity and
34 emission factors developed using GRI data for 1990 through 1992 (GRI/EPA 1996). The 1996 GRI study did not
35 include CO2 data for liquids unloading. EPA used RY2011CO2 emission factors for the earlier years in the time
36 series (i.e., 1990 through 2010). The same activity and emission factors derived from the GRI data were used for all
37 basins for 1990 through 1992. For the remaining time series years (i.e., 1993-2010), EPA applied linear
38 interpolation between the 1992 and 2011 factors at the basin-level. For basins without subpart W data available,
39 EPA applied national average activity and emission factors.
Energy 3-107
-------
1 This update resulted in Cm and CO2 emission estimates an average of 15 percent lower across the time series than
2 in the previous Inventory. In the earlier years of the time series (i.e., 1990 through 2006), CH4 emissions are lower
3 than in the previous Inventory by an average of 43 percent. CO2 emissions over the same time period are lower
4 than in the previous Inventory by an average of 38 percent. For the time series years with reported GHGRP data
5 (i.e., 2011 through 2020), CH4 emissions increased by an average of 21 percent, compared to the previous
6 Inventory. Similarly, CO2 emissions also increased by an average of 17 percent during 2011 through 2020. The
7 basin-level approach's emissions were higher than the previous Inventory's because certain basins with high gas
8 well counts (e.g., Appalachian and Anadarko basins) had higher emission factors than the national average.
9 Table 3-80: Liquids Unloading National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2017
2018
2019
2020
2021
Liquids Unloading With Plunger
Lifts
0
144,856
68,633
99,159
85,536
60,280
39,456
Liquids Unloading Without Plunger
Lifts
76,815
214,070
116,012
166,014
124,428
98,687
80,690
Total Emissions
76,815
358,925
184,645
265,173
209,964
158,968
120,145
Previous Estimate
373,528
379,184
155,178
207,603
175,156
129,831
NA
NA (Not Applicable)
able 3-81: Liquids Unloading National CO2 Emissions (Metric Tons CO2)
Source
1990
2005
2017
2018
2019
2020
2021
Liquids Unloading With Plunger
Lifts
0
11,926
3,376
4,212
2,864
2,606
1,967
Liquids Unloading Without Plunger
Lifts
44,810
40,806
5,390
7,227
7,270
3,562
3,733
Total Emissions
44,810
52,733
8,767
11,439
10,134
6,168
5,700
Previous Estimate
83,155
67,087
7,487
9,181
8,284
5,491
NA
NA (Not Applicable)
11 Miscellaneous Production Flaring (Recalculation with Updated Data)
12 Miscellaneous production flaring CO2 emissions estimates are on average 0.2 percent higher across the 1990 to
13 2020 time series compared with the previous Inventory and the 2020 estimate is 23 percent higher, compared to
14 the previous Inventory. These changes were due to GHGRP submission revisions.
15 Table 3-82: Miscellaneous Production Flaring National Emissions (kt CO2)
Source
1990
2005
2017
2018
2019
2020
2021
Miscellaneous Flaring-Gulf
Coast Basin
NO
166
209
137
398
250
267
Miscellaneous Flaring-
Williston Basin
NO
+
10
6
3
4
4
Miscellaneous Flaring-
Permian Basin
NO
260
622
707
889
831
483
Miscellaneous Flaring-Other
Basins
NO
117
306
476
305
213
236
Total Emissions
NO
543
1,148
1,326
1,595
1,298
991
Previous Estimate
NO
543
1,145
1,344
1,904
1,060
NA
+ Does not exceed 0.5 kt.
NO (Not Occurring)
NA (Not Applicable)
3-108 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Gathering and Boosting - Tanks (Recalculation with Updated Data)
2 Methane emission estimates for gathering and boosting tanks are on average 0.1 percent lower across the 1990 to
3 2020 time series than in the previous Inventory. The 2020 estimate is 2 percent lower than in the previous
4 Inventory. These changes were due to GHGRP submission revisions.
5 Table 3-83: Tanks National Emissions (Metric Tons ChU)
Source
1990
2005
2017
2018
2019
2020
2021
Tanks
Previous Estimate
129,829
129,829
165,236
165,236 .
255,244
255,244
249,489
249,489
295,914
300,169
239,623
244,257
276,748
NA
NA (Not Applicable)
6 Gathering and Boosting - Station Blowdowns
1 Methane emissions estimates for gathering and boosting station blowdowns are on average 0.7 percent lower
8 across the 1990 to 2020 time series than in the previous Inventory. The 2020 estimate is 10 percent lower than in
9 the previous Inventory. These changes were due to GHGRP submission revisions.
10 Table 3-84: Station Blowdowns National Emissions (Metric Tons ChU)
Source
1990
2005
2017
2018
2019
2020
2021
Station Blowdowns
Previous Estimate
20,517
20,517
26,113
26,113
63,852
63,852
78,548
78,548
38,412
43,865
40,468
44,881
42,231
NA
NA (Not Applicable)
11 Gathering and Boosting - Dehydrator Vents (Large Units)
12 Methane emissions for dehydrator vents at large units are on average of 4 percent higher across the 1990 to 2020
13 time series compared with the previous Inventory. The 2020 estimate is 115 percent higher compared to the
14 previous Inventory. The dehydrator vents at large units CO2 emissions estimate increased by an average of 10
15 percent across the time series and by 292 percent in 2020, compared to the previous Inventory. These changes
16 were due to GHGRP submission revisions.
17 Table 3-85: Dehydrator Vents National Emissions (Metric Tons ChU)
Source 1990 2005 2017 2018 2019 2020 2021
Dehydrator Vents 35,716 45,457 61,754 56,543 56,405 52,323 59,207
Previous Estimate 35,716 45,457 61,386 56,381 55,967 24,345 NA
NA (Not Applicable)
18 Table 3-86: Dehydrator Vents National Emissions (kt CO2)
"Source 1990 2005 2017 2018 2019 2020 2021
Dehydrator Vents 371 472 771 820 1,039 1,048 995
Previous Estimate 371 472 772 820 907 267 NA
NA (Not Applicable)
19 Gathering and Boosting - Flare Stacks (Recalculation with Updated Data)
20 The flare stacks CO2 emissions estimate are an average of 0.3 percent lower across the time series compared with
21 the previous Inventory. The 2020 estimate is 4 percent lower, compared to the previous Inventory. These changes
22 were due to GHGRP submission revisions.
Energy 3-109
-------
l Table 3-87: Production Storage Tanks National Emissions (kt CO2)
Source
1990
2005
2017
2018
2019
2020
2021
Flare Stacks
1,355
1,725
2,256
3,696
4,777
2,822
2,631
Previous Estimate
1,355
1,725
2,256
3,695
5,028
2,926
NA
NA (Not Applicable)
2 Processing
3 Flares (Recalculation with Updated Data)
4 Processing segment flare CO2 emission estimates are on average of less than 1 percent higher across the 1993 to
5 2020 time series than in the previous Inventory. The estimate for 2020 is 3 percent higher than in the previous
6 Inventory. These changes were due to GHGRP submission revisions.
7 Table 3-88: Processing Segment Flares National CO2 Emissions (kt CO2)
Source
1990
2005
2017
2018
2019
2020
2021
Flares
NO
3,517
5,587
5,945
9,859
8,120
7,381
Previous Estimate
NO
3,517
5,590
6,176
9,837
7,879
NA
NA (Not Applicable)
NO (Not Occurring)
8 AGR Vents (Recalculation with Updated Data)
9 AGR vents CO2 emission estimates are on average lower than the previous Inventory by less than 1 percent across
10 the 1990 to 2020 time series. Emission estimates for 2020 are 2 percent lower than in the previous Inventory.
11 These changes were due to GHGRP submission revisions.
12 Table 3-89: AGR Vents National CO2 Emissions (kt CO2)
Source
1990
2005
2017
2018
2019
2020
2021
AGR Vents
Previous Estimate
28,282
28,2S2
15,281
15,281
17,313
17,364
16,788
16,792
16,325
16,505
17,258
17,559
18,658
NA
NA (Not Applicable)
NO (Not Occurring)
13 Transmission and Storage
14 There were no methodological updates to the transmission and storage segment, and recalculations resulted in an
15 average increase in calculated CH4 emissions over the time series of 0.2 percent. CO2 emissions will be updated for
16 the Final Inventory; see Planned Improvements.
17 Distribution
18 Mains - Unprotected Steel (Recalculation with Updated Data)
19 Methane emissions estimates for unprotected steel distribution mains are on average 0.6 percent lower across the
20 1990 to 2020 time series compared to the previous Inventory and 6 percent lower in 2020, compared to the
21 previous Inventory. The emission changes were due to updated PHMSA pipeline mileage data.
3-110 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
Table 3-90: Mains - Unprotected Steel National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2017
2018
2019
2020
2021
Mains - Unprotected Steel
Previous Estimate
231,201
231,201
91,262
91,262
44,574
47,236
42,581
45,213
40,732
43,369
39.261
41,554
37,488
NA
NA (Not Applicable)
2 Post-Meter
3 Post-Meter (Recalculation with Updated Data)
4 Post-Meter Cm emissions estimates are higher by an average of 0.1 percent across the 1990 to 2020 time series
5 compared with the previous Inventory, and 1 percent higher in 2020, compared to the previous Inventory. The
6 emission changes were due to changes in residential and industrial natural gas consumption data.
7 Table 3-91: Post-Meter National Cm Emissions (Metric Tons Cm)
Source
1990
2005
2017
2018
2019
2020
2021
Post-Meter
Previous Estimate
289,951
289,951
344,464
344,464
424,492
424,492
445,323
445,220
456,679
456,551
462,751
459,072
463,072
NA
NA (Not Applicable)
8 Planned Improvements
9 Planned Improvements for 2023 Inventory
10 This draft of the Inventory does not yet incorporate updated activity data products for the following data inputs,
11 due to a data base subscription lapse: gas well counts, wells drilled, wells completed, and production. For these
12 inputs, year 2020 values for activity data are used in place of year 2021. The Final Inventory (to be published April
13 2023) will incorporate the latest activity data.
14 The CO2 emissions estimates for LNG export terminals will be updated for the Final Inventory to correct an error in
15 the emission factor calculations in this draft Inventory. The recalculation will result in average annual CO2
16 emissions estimates for 1990 through 2015 decreasing from 122 kt to 23 kt, consistent with the prior Inventory,
17 and annual average CO2 emissions for 2016 through 2021 will increase by 69 kt.
18 Basin-level approaches for pneumatic controllers, equipment leaks, and chemical injection pumps were applied to
19 calculate CFU emissions for public review. For the final Inventory, EPA would apply consistent methods for both
20 CO2 and CFU emissions calculations.
21 Additional information on the update and specific requests for stakeholder feedback can be found in the
22 Disaggregation memo and Production Disaggregation memos. Feedback EPA has received in response to the
23 memo include that basin-level data from GHGRP can improve accuracy of estimates when applied appropriately,
24 that EPA should consider application of the approach to only basins with 50 percent coverage or more, and that
25 liquids unloading is a source that may be well-suited to a basin-level approach, EPA will consider this feedback and
26 any additional feedback received and may revise the calculations in the Inventory.
27 Upcoming Data, and Additional Data that Could Inform the Inventory
28 EPA will assess new data received by EPA's Greenhouse Gas Reporting Program, Methane Challenge Program on
29 an ongoing basis, which may be used to validate or improve existing estimates and assumptions.
30 EPA continues to track studies that contain data that may be used to update the Inventory. EPA will also continue
31 to assess studies that include and compare both top-down and bottom-up emission estimates, which could lead to
Energy 3-111
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
improved understanding of unassigned high emitters (e.g., identification of emission sources and information on
frequency of high emitters) as recommended in previous stakeholder comments.
3.8 Abandoned Oil and Gas Wells (CRF
Source Categories lB2a and lB2b)
Note that this draft of the Inventory does not yet incorporate updated activity data for the following data inputs,
due to a data base subscription lapse: abandoned well counts, and fractions of plugged and unplugged abandoned
wells. Year 2020 values for activity data are used in place of year 2021. The Final Inventory (to be published April
2023) will incorporate the latest activity data.
The term "abandoned wells", as used in the Inventory, encompasses various types of oil and gas wells, including
orphaned wells and other non-producing wells:
• Wells with no recent production, and 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 oil and gas wells (including orphaned wells and other non-producing wells) is
around 3.7 million (with around 2.9 million abandoned oil wells and 0.8 million abandoned gas wells). The methods
to calculate emissions from abandoned wells involve calculating the total populations of plugged and unplugged
abandoned oil and gas wells in the United States and the application of emission factors. An estimate of the
number of orphaned wells within this population is not developed as part of the methodology. Wells that are
plugged have much lower average emissions than wells that are unplugged (less than 1 kg Cm per well per year,
versus over 100 kg Cm per well per year). Around 42 percent of the abandoned well population in the United
States is plugged. This fraction has increased over the Inventory time series (from around 22 percent in 1990) as
more wells fall under regulations and programs requiring or promoting plugging of abandoned wells.
Abandoned oil wells. Abandoned oil wells emitted 231 kt Cm and 5 kt CO2 in 2021. Emissions of both gases
increased by 3 percent from 1990, while the total population of abandoned oil wells increased 37 percent.
Abandoned gas wells. Abandoned gas wells emitted 63 kt Cm and 3 kt CO2 in 2021. Emissions of both gases
increased by 25 percent from 1990, while the total population of abandoned gas wells increased 75 percent.
Table 3-92: ChU Emissions from Abandoned Oil and Gas Wells (MMT CO2 Eq.)
Activity
1990
2005
2017
2018
2019
2020
2021
Abandoned Oil Wells
6.3
6.5
6.5
6.5
6.5
6.5
6.5
Abandoned Gas Wells
1.4
1.6
1.8
1.8
1.8
1.8
1.8
Total
7.7
8.1
00
W
8.3
8.3
8.2
8.2
able 3-93: ChU Emissions from Abandoned Oil and Gas Wells (kt)
Activity
1990
2005
2017
2018
2019
2020
2021
Abandoned Oil Wells
223
232
232
232
233
231
231
Abandoned Gas Wells
51
57
63
63
64
63
63
Total
274
289
295
296
297
295
295
3-112 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Table 3-94: CO2 Emissions from Abandoned Oil and Gas Wells (MMT CO2)
Activity 1990 2005 2017 2018 2019 2020 2021
Abandoned Oil Wells + + + + + + +
Abandoned Gas Wells + + + + + + +
Total + + + + + + +
+ Does not exceed 0.05 MMT C02 Eq.
Table 3-95: CO2 Emissions from Abandoned Oil and Gas Wells (kt)
Activity 1990 2005 2017 2018 2019 2020 2021
Abandoned Oil Wells 5 5 5 5 5 5 5
Abandoned Gas Wells 2 2 3 3 3 3 3
Total 7 7 7 7 7 7 8^
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
EPA uses a Tier 2 method from IPCC (2019) to quantify emissions from abandoned oil and gas wells. EPA's
approach is based on the number of plugged and unplugged abandoned wells in the Appalachian region and in the
rest of the U.S., and emission factors for plugged and unplugged abandoned wells in Appalachia and the rest of the
U.S. Methods for abandoned wells are unavailable in IPCC (2006). The details of this approach and of the data
sources used are described in the memorandum Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2016:
Abandoned Wells in Natural Gas and Petroleum Systems (2018 Abandoned Wells Memo).
EPA developed abandoned well Cm emission factors using data from Kang et al. (2016) and Townsend-Small et al.
(2016). Plugged and unplugged abandoned well CFU emission factors were developed at the national-level (using
emission data from Townsend-Small et al.) and for the Appalachia region (using emission data from measurements
in Pennsylvania and Ohio conducted by Kang et al. and Townsend-Small et al., respectively). The Appalachia region
emissions factors were applied to abandoned wells in states in the Appalachian basin region, and the national-level
emission factors were applied to abandoned wells in all other states. EPA developed abandoned well CO2 emission
factors using the CH4 emission factors and an assumed ratio of CCh-to-Cm gas content, similar to the approach
used to calculate CO2 emissions for many sources in Petroleum Systems and Natural Gas Systems. For abandoned
oil wells, EPA used the Petroleum Systems default production segment associated gas ratio of 0.020 MT CO2/MT
Cm, which was derived through API TankCalc modeling runs. For abandoned gas wells, EPA used the Natural Gas
Systems default production segment CH4 and CO2 gas content values (GRI/EPA 1996, GTI 2001) to develop a ratio
of 0.044 MT CO2/MT CH4. The same respective emission factors are applied for each year of the time series.
EPA developed state-level annual counts of abandoned wells for 1990 through 2020 by summing together an
annual estimate of abandoned wells in the Enverus data set (Enverus 2021), and an estimate of total abandoned
wells not included the Enverus dataset (see 2018 Abandoned Wells Memo for additional information on how the
value was calculated) for each state. References reviewed to develop the number of abandoned wells not included
in the Enverus dataset include historical records collected by state agencies and by USGS.
The total abandoned well population was then split into plugged and unplugged wells by applying an assumption
that all abandoned wells were unplugged in 1950 and using Enverus data to calculate the fraction of plugged
abandoned wells in 2020 in that data set, which was then applied to the total population of abandoned wells for
2020 and 2021. Linear interpolation was applied between the 1950 value and 2020 value to calculate the plugged
fraction for intermediate years. See the memorandum Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-
2016: Abandoned Wells in Natural Gas and Petroleum Systems (2018 Abandoned Wells Memo) for details.87 State-
level plugged and unplugged fractions were developed for the time-series using state-level Enverus data for 2020
87 See https://www.epa.gov/ghgemissions/natural-gas-and-petroleum-systems.
Energy 3-113
-------
1 and linear interpolation between 1950 and 2020 plugged and unplugged fractions. Abandoned wells in all states
2 were assumed to be unplugged in 1950.
3 Abandoned Oil Wells
4 Table 3-96: Abandoned Oil Wells Activity Data, ChU and CO2 Emissions (kt)
Source
1990
2005
2017
2018
2019
2020
2021
Plugged abandoned oil wells
474,432
799,331
1,105,366
1,139,476
1,175,867
1,192,907
1,192,907
Unplugged abandoned oil
wells
1,664,717
1,749,329
1,749,813
1,751,999
1,756,573
1,739,533
1,739,533
Total Abandoned Oil Wells
2,139,149
2,548,660
2,855,179
2,891,475
2,932,440
2,932,440
2,932,440
Abandoned oil wells in
Appalachia
23%
21%
19%
19%
19%
19%
19%
Abandoned oil wells outside
of Appalachia
77%
79%
81%
81%
81%
81%
81%
CH4 from plugged
abandoned oil wells (kt)
0.20
0.30
0.39
0.40
0.41
0.42
0.42
CH4from unplugged
abandoned oil wells(kt)
223.1
231.3
231.5
231.8
232.5
230.7
230.7
Total ChUfrom Abandoned
oil wells (kt)
223.3
231.6
231.9
232.2
232.9
231.1
231.1
Total C02 from Abandoned
oil wells (kt)
4.5
4.7
4.7
4.7
4.7
4.7
4.7
5 Abandoned Gas Wells
6 Table 3-97: Abandoned Gas Wells Activity Data, ChU and CO2 Emissions (kt)
Source
1990
2005
2017
2018
2019
2020
2021
Plugged abandoned gas wells
107,292
206,413
332,743
342,495
353,746
358,871
358,871
Unplugged abandoned gas
wells
349,041
397,844
440,367
442,014
444,532
439,407
439,407
Total Abandoned Gas Wells
456,333
604,257
773,110
784,509
798,278
798,278
798,278
Abandoned gas wells in
Appalachia
29%
26%
24%
24%
25%
25%
25%
Abandoned gas wells outside
of Appalachia
71%
74%
76%
76%
75%
75%
75%
CH4from plugged abandoned
gas wells (kt)
0.07
0.12
0.17
0.18
0.19
0.19
0.19
CH4from unplugged
abandoned gas wells (kt)
50.9
56.8
62.8
63.2
63.8
63.2
63.2
Total CH4 from Abandoned
gas wells (kt)
50.9
56.9
63.0
63.4
64.0
63.4
63.4
Total C02 from Abandoned
gas wells (kt)
2.2
2.5
2.8
2.8
2.8
2.8
2.8
7
8
9
10
11
Uncertainty—TO BE UPDATED FOR FINAL INVENTORY REPORT
To characterize uncertainty surrounding estimates of abandoned well emissions, EPA conducted a quantitative
uncertainty analysis using the IPCC Approach 2 methodology (Monte Carlo simulation technique). See the 2018
Abandoned Wells Memo for details of the uncertainty analysis methods. EPA used Microsoft Excel's @ RISK add-in
tool to estimate the 95 percent confidence bound around total methane emissions from abandoned oil and gas
3-114 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
wells in year 2019, then applied the calculated bounds to both Cm and CO2 emissions estimates for each
population. The @RISK add-in provides for the specification of probability density functions (PDFs) for key variables
within a computational structure that mirrors the calculation of the inventory estimate. EPA used measurement
data from the Kang et al. (2016) and Townsend-Small et al. (2016) studies to characterize the CFU emission factor
PDFs. For activity data inputs (e.g., total count of abandoned wells, split between plugged and unplugged), EPA
assigned default uncertainty bounds of ± 10 percent based on expert judgment.
The IPCC guidance notes that in using this method, "some uncertainties that are not addressed by statistical means
may exist, including those arising from omissions or double counting, or other conceptual errors, or from
incomplete understanding of the processes that may lead to inaccuracies in estimates developed from models." As
a result, the understanding of the uncertainty of emission estimates for this category evolves and improves as the
underlying methodologies and datasets improve. The uncertainty bounds reported below only reflect those
uncertainties that EPA has been able to quantify and do not incorporate considerations such as modeling
uncertainty, data representativeness, measurement errors, misreporting or misclassification.
The results presented below in Table 3-98 provide the 95 percent confidence bound within which actual emissions
from abandoned oil and gas wells are likely to fall for the year 2019, using the recommended IPCC methodology.
Abandoned oil well CFU emissions in 2019 were estimated to be between 0.9 and 16.5 MMT CO2 Eq., while
abandoned gas well CFU emissions were estimated to be between 0.2 and 4.3 MMT CO2 Eq. at a 95 percent
confidence level. Uncertainty bounds for other years of the time series have not been calculated, but uncertainty is
expected to vary over the time series.
Table 3-98: Approach 2 Quantitative Uncertainty Estimates for ChU and CO2 Emissions from
Petroleum and Natural Gas Systems (MMT CO2 Eq. and Percent)
2019 Emission Estimate Uncertainty Range Relative to Emission Estimate3
Source Gas
(MMT CP2 Eq.)b (MMT CP2 Eq.) (%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Abandoned Oil Wells
ch4
5.2
0.9
16.5
-83%
+217%
Abandoned Gas Wells
ch4
1.4
0.2
4.3
-83%
+217%
Abandoned Oil Wells
C02
0.004
0.001
0.013
-83%
+217%
Abandoned Gas Wells
C02
0.002
0.0004
0.008
-83%
+217%
a Range of emission estimates estimated by applying the 95 percent confidence intervals obtained from the Monte Carlo
Simulation analysis conducted for total abandoned oil and gas well CH4 emissions in year 2019.
b All reported values are rounded after calculation. As a result, lower and upper bounds may not be duplicable from other
rounded values as shown in table.
QA/QC and Verification Discussion
The emission estimates in the Inventory are continually reviewed and assessed to determine whether emission
factors and activity factors accurately reflect current industry practices. A QA/QC analysis was performed for data
gathering and input, documentation, and calculation. QA/QC checks are consistently conducted to minimize
human error in the model calculations. EPA performs a thorough review of information associated with new
studies to assess whether the assumptions in the Inventory are consistent with industry practices and whether
new data is available that could be considered for updates to the estimates. As in previous years, EPA conducted
early engagement and communication with stakeholders on updates prior to public review. EPA held stakeholder
webinars on greenhouse gas data for oil and gas in September and November of 2022.
Energy 3-115
-------
1 Recalculations Discussion
2 EPA updated the Inventory methodology to estimate abandoned well emissions at the state-level as an
3 intermediate step to calculating national emissions. Previously, well counts were developed for the Appalachian
4 region and for all other regions as a total, and plugged and unplugged fractions were developed at the national-
5 level. In the current Inventory, EPA used abandoned well counts and plugged and unplugged fractions at the state-
6 level to estimate emissions. The incorporation of disaggregated, state-level data will improve future versions of
7 both the gridded and state-level greenhouse gas inventories as geographic differences in plugging rates can now
8 be reflected. This will allow EPA to use the gridded greenhouse gas inventory for improved comparisons with
9 atmospheric observation studies, because regions will reflect local differences. In addition, this update will
10 improve the ability of the state-level Inventory to reflect impacts of state-level programs.
11 The emission factors from the previous Inventory were retained and used to estimate state-level emissions, with
12 Appalachia-specific factors applied to states in Appalachia. The state-level emissions were then summed up to the
13 national level. As an outcome of these revisions, total calculated abandoned well CH4 emissions across the time
14 series are an average of 6 percent higher than in the previous Inventory. The calculated value for 2020 is 7 percent
15 higher than in the previous Inventory.
16 The main cause of increased emission estimate across the time series is the application of state-specific fractions
17 of plugged wells, which resulted in a larger fraction of unplugged wells in Appalachia (which has a higher
18 unplugged well emission factor than other regions) than in the previous inventory, which applied a national
19 average plugging fraction to the entire U.S. abandoned well population.
20 In the previous Inventory, abandoned dry wells were proportionally allocated between abandoned oil and gas
21 wells at the national level. In the current Inventory, dry wells are proportionally allocated to abandoned oil and gas
22 wells at the state level. The total counts of abandoned wells changed by 0.02 percent (decrease), compared with
23 the previous inventory. The counts of abandoned oil wells are about 1.6 percent lower across the time series
24 compared to the previous Inventory and gas wells are about 7 percent higher.
25 In addition, for the current Inventory, CC>2-equivalent emissions totals have been revised to reflect the 100-year
26 global warming potentials (GWPs) provided in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). AR5 GWP
27 values differ slightly from those presented in the IPCC Fourth Assessment Report (AR4) (IPCC 2007) used in the
28 previous inventories. The AR5 GWPs have been applied across the entire time series for consistency. The GWP of
29 Cm has increased from 25 to 28, leading to an overall increase in the calculated CC>2-equivalent emissions of Cl-U.
30 Compared to the previous Inventory which applied 100-year GWP values from AR4, in the current Inventory
31 (including other recalculations noted above), CC>2-equivalent Cl-U emissions increased by 16 percent on average
32 over the time series. Further discussion on this update and the overall impacts of updating the Inventory GWP
33 values to reflect the IPCC AR5 can be found in Chapter 9, Recalculations and Improvements.
34 Planned Improvements
35 This draft of the Inventory does not yet incorporate updated activity data for the following data inputs, due to a
36 data base subscription lapse: abandoned well counts, and fractions of plugged and unplugged abandoned wells.
37 Year 2020 values for activity data are used in place of year 2021. The Final Inventory (to be published April 2023)
38 will incorporate the latest activity data.
39 EPA will continue to assess new data and stakeholder feedback on considerations (such as potential use of
40 emission factor data from regions not included in the measurement studies on which current emission factors are
41 based) to improve the abandoned well count estimates and emission factors. In future Inventories, EPA will assess
42 data that become available from Department of Interior and Department of Energy orphan well plugging
43 programs.
3-116 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
3.9 International Bunker Fuels (CRF Source
Category 1: Memo Items)
Emissions resulting from the combustion of fuels used for international transport activities, termed international
bunker fuels under the UNFCCC, are not included in national emission totals, but are reported separately based
upon location of fuel sales. The decision to report emissions from international bunker fuels separately, instead of
allocating them to a particular country, was made by the Intergovernmental Negotiating Committee in establishing
the Framework Convention on Climate Change.88 These decisions are reflected in the IPCC methodological
guidance, including IPCC (2006), in which countries are requested to report emissions from ships or aircraft that
depart from their ports with fuel purchased within national boundaries and are engaged in international transport
separately from national totals (IPCC 2006).89
Two transport modes are addressed under the IPCC definition of international bunker fuels: aviation and marine.90
Greenhouse gases emitted from the combustion of international bunker fuels, like other fossil fuels, include CO2,
Cm and N2O for marine transport modes, and CO2 and N2O for aviation transport modes. Emissions from ground
transport activities—by road vehicles and trains—even when crossing international borders are allocated to the
country where the fuel was loaded into the vehicle and, therefore, are not counted as bunker fuel emissions.
The 2006 IPCC Guidelines distinguish between three different modes of air traffic: civil aviation, military aviation,
and general aviation. Civil aviation comprises aircraft used for the commercial transport of passengers and freight,
military aviation comprises aircraft under the control of national armed forces, and general aviation applies to
recreational and small corporate aircraft. The 2006 IPCC Guidelines further define international bunker fuel use
from civil aviation as the fuel combusted for civil (e.g., commercial) aviation purposes by aircraft arriving or
departing on international flight segments. However, as mentioned above, and in keeping with the 2006 IPCC
Guidelines, only the fuel purchased in the United States and used by aircraft taking-off (i.e., departing) from the
United States are reported here. The standard fuel used for civil and military aviation is kerosene-type jet fuel,
while the typical fuel used for general aviation is aviation gasoline.91
Emissions of CChfrom aircraft are essentially a function of fuel consumption. Nitrous oxide emissions also depend
upon engine characteristics, flight conditions, and flight phase (i.e., take-off, climb, cruise, decent, and landing).
Recent data suggest that little or no CH4 is emitted by modern engines (Anderson et al. 2011), and as a result, CH4
emissions from this category are reported as zero. In jet engines, N2O is primarily produced by the oxidation of
atmospheric nitrogen, and the majority of emissions occur during the cruise phase.
International marine bunkers comprise emissions from fuels burned by ocean-going ships of all flags that are
engaged in international transport. Ocean-going ships are generally classified as cargo and passenger carrying,
military (i.e., U.S. Navy), fishing, and miscellaneous support ships (e.g., tugboats). For the purpose of estimating
greenhouse gas emissions, international bunker fuels are solely related to cargo and passenger carrying vessels,
which is the largest of the four categories, and military vessels. Two main types of fuels are used on sea-going
vessels: distillate diesel fuel and residual fuel oil. Carbon dioxide is the primary greenhouse gas emitted from
marine shipping.
88 See report of the Intergovernmental Negotiating Committee for a Framework Convention on Climate Change on the work of
its ninth session, held at Geneva from 7 to 18 February 1994 (A/AC.237/55, annex I, para. lc).
89 Note that the definition of international bunker fuels used by the UNFCCC differs from that used by the International Civil
Aviation Organization.
90 Most emission related international aviation and marine regulations are under the rubric of the International Civil Aviation
Organization (ICAO) or the International Maritime Organization (IMO), which develop international codes, recommendations,
and conventions, such as the International Convention of the Prevention of Pollution from Ships (MARPOL).
91 Naphtha-type jet fuel was used in the past by the military in turbojet and turboprop aircraft engines.
Energy 3-117
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Overall, aggregate greenhouse gas emissions in 2021 from the combustion of international bunker fuels from both
aviation and marine activities were 69.9 MMT CO2 Eq., or 33.2 percent below emissions in 1990 (see Table 3-99
and Table 3-100). Emissions from international flights and international shipping voyages departing from the
United States have increased by 4.5 percent and decreased by 55.1 percent, respectively, since 1990. The majority
of these emissions were in the form of CO2; however, small amounts of CH4 (from marine transport modes) and
N2O were also emitted. Commercial aviation bunker fuel data for 2021 were not yet available and were proxied
based on 2020 data.
Table 3-99: CO2, ChU, and N2O Emissions from International Bunker Fuels (MMT CO2 Eq.)
Gas/Mode
1990
2005
2017
2018
2019
2020
2021
C02
103.6
113.3
120.2
122.2
116.1
69.6
69.3
Aviation
38.2
60.2
77.8
80.9
80.8
39.8
39.9
Commercial
30.0
55.6
74.5
77.7
77.6
36.7
36.7
Military
8.2
4.6
3.3
3.2
3.2
3.1
3.2
Marine
65.4
53.1
42.4
41.3
35.4
29.9
29.4
ch4
0.2
0.1
0.1
0.1
0.1
0.1
0.1
Aviation
NO
NO
NO
NO
NO
NO
NO
Marine
0.2
0.1
0.1
0.1
0.1
0.1
0.1
n2o
0.8
0.9
0.9
1.0
0.9
0.5
0.5
Aviation
0.3
0.5
0.7
0.7
0.7
0.3
0.3
Marine
0.4
0.4
0.3
0.3
0.2
0.2
0.2
Total
104.6
114.3
121.2
123.2
117.1
70.3
69.9
NO (Not Occurring)
Notes: Totals may not sum due to independent rounding. Includes aircraft cruise altitude emissions.
2021 commercial aviation data were not yet available and were proxied based on 2020 data.
Table 3-100: CO2, ChU, and N2O Emissions from International Bunker Fuels (kt)
Gas/Mode
1990
2005
2017
2018
2019
2020
2021
C02
103,634
113,328
120,192
122,179
116,132
69,638
69,280
Aviation
38,205
60,221
77,764
80,853
80,780
39,781
39,912
Marine
65,429
53,107
42,428
41,325
35,351
29,857
29,369
ch4
7
5
4
4
4
3
3
Aviation
NO
NO
NO
NO
NO
NO
NO
Marine
7
5
4
4
4
3
3
NzO
3
3
4
4
3
2
2
Aviation
1
2
2
3
3
1
1
Marine
2
1
1
1
1
1
1
NO (Not Occurring)
Notes: Totals by gas may not sum due to independent rounding. Includes aircraft cruise altitude
emissions. 2021 commercial aviation data were not yet available and were proxied based on 2020
data.
Methodology and Time-Series Consistency
Emissions of CO2 were for the most part estimated by applying C content and fraction oxidized factors to fuel
consumption activity data. This approach is analogous to that described under Section 3.1 - CO2 from Fossil Fuel
Combustion. Carbon content and fraction oxidized factors for jet fuel (except for commercial aviation as per
below), distillate fuel oil, and residual fuel oil are the same as used for CO2 from Fossil Fuel Combustion and are
presented in Annex 2.1, Annex 2.2, and Annex 3.8 of this Inventory. Density conversions were taken from ASTM
(1989) and USAF (1998). Heat content for distillate fuel oil and residual fuel oil were taken from EIA (2022) and
USAF (1998), and heat content for jet fuel was taken from EIA (2022). See below for details on how emission
estimates for commercial aviation were determined.
A complete description of the methodology and a listing of the various factors employed can be found in Annex
2.1. See Annex 3.8 for a specific discussion on the methodology used for estimating emissions from international
3-118 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
bunker fuel use by the U.S. military.
Emission estimates for Cm and N2O were calculated by multiplying emission factors by measures of fuel
consumption by fuel type and mode. Emission factors used in the calculations of CH4 and N2O emissions were
obtained from the Revised 1996IPCC Guidelines (IPCC/UNEP/OECD/IEA 1997), which is also referenced in the 2006
IPCC Guidelines (IPCC 2006). For aircraft emissions, the following value, in units of grams of pollutant per kilogram
of fuel consumed (g/kg), was employed: 0.1 for N2O (IPCC 2006). For marine vessels consuming either distillate
diesel or residual fuel oil the following values (g/MJ), were employed: 0.315 for CFU and 0.08 for N2O. Activity data
for aviation included solely jet fuel consumption statistics, while the marine mode included both distillate diesel
and residual fuel oil.
Activity data on domestic and international aircraft fuel consumption were developed by the U.S. Federal Aviation
Administration (FAA) using radar-informed data from the FAA Enhanced Traffic Management System (ETMS) for
1990 and 2000 through 2020 as modeled with the Aviation Environmental Design Tool (AEDT). This bottom-up
approach is built from modeling dynamic aircraft performance for each flight occurring within an individual
calendar year. The analysis incorporates data on the aircraft type, date, flight identifier, departure time, arrival
time, departure airport, arrival airport, ground delay at each airport, and real-world flight trajectories. To generate
results for a given flight within AEDT, the radar-informed aircraft data is correlated with engine and aircraft
performance data to calculate fuel burn and exhaust emissions. Information on exhaust emissions for in-
production aircraft engines comes from the International Civil Aviation Organization (ICAO) Aircraft Engine
Emissions Databank (EDB). This bottom-up approach is in accordance with the Tier 3B method from the 2006 IPCC
Guidelines (IPCC 2006).
International aviation CO2 estimates for 1990 and 2000 through 2020 were obtained directly from FAA's AEDT
model (FAA 2022). The radar-informed method that was used to estimate CO2 emissions for commercial aircraft
for 1990 and 2000 through 2020 was not possible for 1991 through 1999 because the radar dataset was not
available for years prior to 2000. FAA developed Official Airline Guide (OAG) schedule-informed inventories
modeled with AEDT and great circle trajectories for 1990, 2000, and 2010. Because fuel consumption and CO2
emission estimates for years 1991 through 1999 are unavailable, consumption estimates for these years were
calculated using fuel consumption estimates from the Bureau of Transportation Statistics (DOT 1991 through
2013), adjusted based on 2000 through 2005 data. See Annex 3.3 for more information on the methodology for
estimating emissions from commercial aircraft jet fuel consumption. Data for 2021 are not yet available so 2021
data were proxied based on 2020 data and scaled by the percent difference of 2020 and 2021 jet fuel consumption
for commercial aviation reported by the Bureau of Transportation (DOT 1991 through 2021).
Data on U.S. Department of Defense (DoD) aviation bunker fuels and total jet fuel consumed by the U.S. military
was supplied by the Office of the Under Secretary of Defense (Installations and Environment), DoD. Estimates of
the percentage of each Service's total operations that were international operations were developed by DoD.
Military aviation bunkers included international operations, operations conducted from naval vessels at sea, and
operations conducted from U.S. installations principally over international water in direct support of military
operations at sea. Military aviation bunker fuel emissions were estimated using military fuel and operations data
synthesized from unpublished data from DoD's Defense Logistics Agency Energy (DLA Energy 2022). Together, the
data allow the quantity of fuel used in military international operations to be estimated. Densities for each jet fuel
type were obtained from a report from the U.S. Air Force (USAF 1998). Final jet fuel consumption estimates are
presented in Table 3-101. See Annex 3.8 for additional discussion of military data.
Table 3-101: Aviation Jet Fuel Consumption for International Transport (Million Gallons)
Nationality
1990
2005
2017
2018
2019
2020
2021
U.S. and Foreign Carriers
3,155
5,858
7,844
8,178
8,170
3,859
3,859
U.S. Military
862
462
326
315
318
308
321
Total
4,017
6,321
8,171
8,493
8,488
4,167
4,180
Note: Totals may not sum due to independent rounding. U.S. and Foreign Carriers 2021 data are not available, so
data were proxied based on 2020 data.
Energy 3-119
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
In order to quantify the civilian international component of marine bunker fuels, activity data on distillate diesel
and residual fuel oil consumption by cargo or passenger carrying marine vessels departing from U.S. ports were
collected for individual shipping agents on a monthly basis by the U.S. Customs and Border Protection. This
information was then reported in unpublished data collected by the Foreign Trade Division of the U.S. Department
of Commerce's Bureau of the Census (DOC 1991 through 2022) for 1990 through 2001, 2007 through 2021, and
the Department of Homeland Security's Bunker Report for 2003 through 2006 (DHS 2008). Fuel consumption data
for 2002 was interpolated due to inconsistencies in reported fuel consumption data. Activity data on distillate
diesel consumption by military vessels departing from U.S. ports were provided by DLA Energy (2022). The total
amount of fuel provided to naval vessels was reduced by 21 percent to account for fuel used while the vessels
were not-underway (i.e., in port). Data on the percentage of steaming hours underway versus not underway were
provided by the U.S. Navy. These fuel consumption estimates are presented in Table 3-102.
Table 3-102: Marine Fuel Consumption for International Transport (Million Gallons)
Fuel Type
1990
2005
2017
2018
2019
2020
2021
Residual Fuel Oil
4,781
3,881
2,975
2,790
2,246
1,964
1,953
Distillate Diesel Fuel & Other
617
444
568
684
702
461
437
U.S. Military Naval Fuels
522
471
307
285
281
296
285
Total
5,920
4,796
3,850
3,759
3,229
2,721
2,674
Note: Totals may not sum due to independent rounding.
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2021.
Uncertainty
Emission estimates related to the consumption of international bunker fuels are subject to the same uncertainties
as those from domestic aviation and marine mobile combustion emissions; however, additional uncertainties
result from the difficulty in collecting accurate fuel consumption activity data for international transport activities
separate from domestic transport activities.92 For example, smaller aircraft on shorter routes often carry sufficient
fuel to complete several flight segments without refueling in order to minimize time spent at the airport gate or
take advantage of lower fuel prices at particular airports. This practice, called tankering, when done on
international flights, complicates the use of fuel sales data for estimating bunker fuel emissions. Tankering is less
common with the type of large, long-range aircraft that make many international flights from the United States,
however. Similar practices occur in the marine shipping industry where fuel costs represent a significant portion of
overall operating costs and fuel prices vary from port to port, leading to some tankering from ports with low fuel
costs.
Uncertainties exist with regard to the total fuel used by military aircraft and ships. Total aircraft and ship fuel use
estimates were developed from DoD records, which document fuel sold to the DoD Components (e.g., Army,
Department of Navy and Air Force) from the Defense Logistics Agency Energy. These data may not include fuel
used in aircraft and ships as a result of a Service procuring fuel from, selling fuel to, trading fuel with, or giving fuel
to other ships, aircraft, governments, or other entities.
Additionally, there are uncertainties in historical aircraft operations and training activity data. Estimates for the
quantity of fuel actually used in Navy and Air Force flying activities reported as bunker fuel emissions had to be
estimated based on a combination of available data and expert judgment. Estimates of marine bunker fuel
emissions were based on Navy vessel steaming hour data, which reports fuel used while underway and fuel used
while not underway. This approach does not capture some voyages that would be classified as domestic for a
commercial vessel. Conversely, emissions from fuel used while not underway preceding an international voyage
are reported as domestic rather than international as would be done for a commercial vessel. There is uncertainty
92 See uncertainty discussions under section 3.1 C02 from Fossil Fuel Combustion.
3-120 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
associated with ground fuel estimates for 1997 through 2021, including estimates for the quantity of jet fuel
allocated to ground transportation. Small fuel quantities may have been used in vehicles or equipment other than
that which was assumed for each fuel type.
There are also uncertainties in fuel end-uses by fuel type, emissions factors, fuel densities, diesel fuel sulfur
content, aircraft and vessel engine characteristics and fuel efficiencies, and the methodology used to back-
calculate the data set to 1990 using the original set from 1995. The data were adjusted for trends in fuel use based
on a closely correlating, but not matching, data set. All assumptions used to develop the estimate were based on
process knowledge, DoD data, and expert judgments. The magnitude of the potential errors related to the various
uncertainties has not been calculated but is believed to be small. The uncertainties associated with future military
bunker fuel emission estimates could be reduced through revalidation of assumptions based on data regarding
current equipment and operational tempo, however, it is doubtful data with more fidelity exist at this time.
Although aggregate fuel consumption data have been used to estimate emissions from aviation, the recommended
method for estimating emissions of gases other than CO2 in the 2006IPCC Guidelines (IPCC 2006) is to use data by
specific aircraft type, number of individual flights and, ideally, movement data to better differentiate between
domestic and international aviation and to facilitate estimating the effects of changes in technologies. The IPCC
also recommends that cruise altitude emissions be estimated separately using fuel consumption data, while
landing and take-off (LTO) cycle data be used to estimate near-ground level emissions of gases other than CO2.93
There is also concern regarding the reliability of the existing DOC (1991 through 2022) data on marine vessel fuel
consumption reported at U.S. customs stations due to the significant degree of inter-annual variation.
QA/QC and Verification
In order to ensure the quality of the emission estimates from international bunker fuels, General (IPCC Tier 1) and
category-specific (Tier 2) Quality Assurance/Quality Control (QA/QC) procedures were implemented consistent
with the U.S. Inventory QA/QC plan outlined in Annex 8. The Tier 2 procedures that were implemented involved
checks specifically focusing on the activity data and emission factor sources and methodology used for estimating
CO2, CH4, and N2O emissions from international bunker fuels in the United States. Emission totals for the different
sectors and fuels were compared and trends were investigated. No corrective actions were necessary.
Recalculations Discussion
For the current Inventory, C02-equivalent emissions of CH4 and N2O from international bunker fuels have been
revised to reflect the 100-year global warming potentials (GWPs) provided in the IPCC Fifth Assessment Report
(AR5) (IPCC 2013). AR5 GWP values differ slightly from those presented in the IPCC Fourth Assessment Report
(AR4), which was used in the previous inventories (IPCC 2007). The AR5 GWPs have been applied across the entire
time series for consistency. Prior inventories used GWPs of 25 and 298 for CH4 and N2O, respectively. These values
have been updated to 28 and 265, respectively. Compared to the previous Inventory which applied 100-year GWP
values from AR4, the average annual change in C02-equivalent CH4 emissions was a 12 percent increase and the
average annual change in C02-equivalent N2O emissions was an 11 percent decrease for the time series. As a result
of the change in methodology, total emissions across the time series changed by an average annual decrease of 0.1
MMT CO2 Eq. (less than half a percent) relative to emissions results calculated using the prior GWPs. Further
93 U.S. aviation emission estimates for CO, NOx, and NMVOCs are reported by EPA's National Emission Inventory (NEI) Air
Pollutant Emission Trends website, and reported under the Mobile Combustion section. It should be noted that these estimates
are based solely upon LTO cycles and consequently only capture near ground-level emissions, which are more relevant for air
quality evaluations. These estimates also include both domestic and international flights. Therefore, estimates reported under
the Mobile Combustion section overestimate IPCC-defined domestic CO, NOx, and NMVOC emissions by including landing and
take-off (LTO) cycles by aircraft on international flights, but underestimate because they do not include emissions from aircraft
on domestic flight segments at cruising altitudes.
Energy 3-121
-------
1 discussion on this update and the overall impacts of updating the Inventory GWP values to reflect the IPCC AR5 can
2 be found in Chapter 9, Recalculations and Improvements.
3 Planned Improvements
4 EPA will evaluate data availability to update the sources for densities, energy contents, and emission factors
5 applied to estimate emissions from aviation and marine fuels. Many are from sources from the late 1990s, such as
6 IPCC/UNEP/OECD/IEA (1997). Potential sources with more recent data include the International Maritime
7 Organization (IMO) greenhouse gas emission inventory, International Air Transport Association (IATA)/ICAO
8 greenhouse gas reporting system (CORSIA), and the EPA Greenhouse Gas Reporting Program (GHGRP) Technical
9 Support Document for Petroleum Products. Specifically, EPA will evaluate data availability to support updating the
10 heat contents and carbon contents of jet fuel with input from EIA.
11 A longer-term effort is underway to consider the feasibility of including data from a broader range of domestic and
12 international sources for bunker fuels. Potential sources include the IMO greenhouse gas emission inventory, data
13 from the U.S. Coast Guard on vehicle operation currently used in criteria pollutant modeling, data from the
14 International Energy Agency (IEA), relevant updated FAA models to improve aviation bunker fuel estimates, and
15 researching newly available marine bunker data.
16 3.10 Biomass and Biofuels Consumption
I? (CRF Source Category 1A)
18 The combustion of biomass fuels—such as wood, charcoal, the biogenic portions of MSW, and wood waste and
19 biomass-based fuels such as ethanol, biogas, and biodiesel—generates CO2 in addition to CFU and N2O already
20 covered in this chapter. In line with the reporting requirements for inventories submitted under the UNFCCC, CO2
21 emissions from biomass combustion have been estimated separately from fossil fuel CO2 emissions and are not
22 directly included in the energy sector contributions to U.S. totals. In accordance with IPCC methodological
23 guidelines, any such emissions are calculated by accounting for net carbon fluxes from changes in biogenic C
24 reservoirs in wooded or crop lands. For a more complete description of this methodological approach, see the
25 Land Use, Land-Use Change, and Forestry chapter (Chapter 6), which accounts for the contribution of any resulting
26 CO2 emissions to U.S. totals within the Land Use, Land-Use Change, and Forestry sector's approach.
27 Therefore, CO2 emissions from biomass and biofuel consumption are not included specifically in summing energy
28 sector totals. However, they are presented here for informational purposes and to provide detail on biomass and
29 biofuels consumption.
30 In 2021, total CO2 emissions from the burning of woody biomass in the industrial, residential, commercial, and
31 electric power sectors were approximately 202.8 MMT CO2 Eq. (202,841 kt) (see Table 3-103 and Table 3-104). As
32 the largest consumer of woody biomass, the industrial sector was responsible for 62.1 percent of the CO2
33 emissions from this source. The residential sector was the second largest emitter, constituting 23.6 percent of the
34 total, while the electric power and commercial sectors accounted for the remainder.
35 Table 3-103: CO2 Emissions from Wood Consumption by End-Use Sector (MMT CO2 Eq.)
End-Use Sector
1990
2005
2017
2018
2019
2020
2021
Industrial
135.3
136.3
135.4
134.4
132.1
127.3
126.0
Residential
59.8
44.3
44.3
54.1
56.3
45.5
47.8
Commercial
6.8
7.2
8.6
8.7
8.7
8.6
8.5
Electric Power
13.3
19.1
23.6
22.8
20.7
19.1
20.5
Total
215.2
206.9
212.0
220.0
217.7
200.4
202.8
3-122 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Table 3-104: CO2 Emissions from Wood Consumption by End-Use Sector (kt)
End-Use Sector
1990
2005
2017
2018
2019
2020
2021
Industrial
135,348
136,269
135,386
134,417
132,069
127,301
125,970
Residential
59,808
44,340
44,298
54,124
56,253
45,452
47,823
Commercial
6,779
7,218
8,634
8,669
8,693
8,554
8,528
Electric Power
13,252
19,074
23,647
22,795
20,677
19,115
20,519
Total
215,186
206,901
211,965
220,005
217,692
200,421
202,841
Note: Totals may not sum due to independent rounding.
Carbon dioxide emissions from combustion of the biogenic components of MSW by the electric power sector were
an estimated 15.3 MMT CO2 (15,329 kt) in 2021. Emissions across the time series are shown in Table 3-105 and
Table 3-106. As discussed in Section 3.3, MSW is combusted to produce electricity and the CO2 emissions from the
fossil portion of the MSW (e.g., plastics, textiles, etc.) are included in the energy sector FFC estimates. The MSW
also includes biogenic components (e.g., food waste, yard trimmings, natural fibers) and the CO2 emissions
associated with that biogenic portion is included here.
Table 3-105: CO2 Emissions from Biogenic Components of MSW (MMT CO2 Eq.)
End-Use Sector 1990 2005 2017 2018 2019 2020 2021
Electric Power 18.5 14.7 16.1 16.1 15.7 15.6 15.3
Table 3-106: CO2 Emissions from Biogenic Components of MSW (kt)
End-Use Sector 1990 2005 2017 2018 2019 2020 2021
Electric Power 18,534 14,722 16,130 16,115 15,709 15,614 15,329
The transportation sector is responsible for most of the fuel ethanol consumption in the United States. Ethanol
used for fuel is currently produced primarily from corn grown in the Midwest, but it can be produced from a
variety of biomass feedstocks. Most ethanol for transportation use is blended with gasoline to create a 90 percent
gasoline, 10 percent by volume ethanol blend known as E-10 or gasohol.
In 2021, the United States transportation sector consumed an estimated 1,114.3 trillion Btu of ethanol (96 percent
of total), and as a result, produced approximately 76.3 MMT CO2 Eq. (76,279 kt) (see Table 3-107 and Table 3-108)
of CO2 emissions. Smaller quantities of ethanol were also used in the industrial and commercial sectors. Ethanol
fuel production and consumption has grown significantly since 1990 due to the favorable economics of blending
ethanol into gasoline and federal policies that have encouraged use of renewable fuels.
Table 3-107: CO2 Emissions from Ethanol Consumption (MMT CO2 Eq.)
End-Use Sector
1990
2005
2017
2018
2019
2020
2021
Transportation3
4.1
21.6
77.7
78.6
78.7
68.1
76.3
Industrial
0.1
1.2
1.9
1.4
1.6
1.6
1.2
Commercial
0.1
0.2
2.5
1.9
2.2
2.2
1.6
Total
4.2
22.9
82.1
81.9
82.6
71.8
79.1
a See Annex 3.2, Table A-76 for additional information on transportation consumption of these fuels.
Note: Totals may not sum due to independent rounding.
ible 3-108: CO2 Emissions from Ethanol Consumption (kt)
End-Use Sector
1990
2005
2017
2018
2019
2020
2021
Transportation3
4,059
21,616
77,671
78,603
78,739
68,085
76,279
Industrial
105
1,176
1,868
1,404
1,610
1,582
1,171
Commercial
63
151
2,550
1,910
2,229
2,182
1,615
Total
4,227
22,943
82,088
81,917
82,578
71,848
79,064
Energy 3-123
-------
a See Annex 3.2, Table A-76 for additional information on transportation consumption of these fuels.
Note: Totals may not sum due to independent rounding.
1 The transportation sector is assumed to be responsible for all of the biodiesel consumption in the United States
2 (EIA 2022a). Biodiesel is currently produced primarily from soybean oil, but it can be produced from a variety of
3 biomass feedstocks including waste oils, fats, and greases. Biodiesel for transportation use appears in low-level
4 blends (less than 5 percent) with diesel fuel, high-level blends (between 6 and 20 percent) with diesel fuel, and 100
5 percent biodiesel (EIA 2022b).
6 In 2021, the United States consumed an estimated 218.2 trillion Btu of biodiesel, and as a result, produced
7 approximately 16.1 MMT CO2 Eq. (16,112 kt) (see Table 3-109 and Table 3-110) of CO2 emissions. Biodiesel
8 production and consumption has grown significantly since 2001 due to the favorable economics of blending
9 biodiesel into diesel and federal policies that have encouraged use of renewable fuels (EIA 2022b). There was no
10 measured biodiesel consumption prior to 2001 EIA (2022a).
11 Table 3-109: CO2 Emissions from Biodiesel Consumption (MMT CO2 Eq.)
End-Use Sector
1990
2005
2017
2018
2019
2020
2021
Transportation3
NO
0.9
18.7
17.9
17.1
17.7
16.1
NO (Not Occurring)
a See Annex 3.2, Table A-76 for additional information on transportation consumption of these fuels.
12 Table 3-110: CO2 Emissions from Biodiesel Consumption (kt)
End-Use Sector
1990
2005
2017
2018
2019
2020
2021
Transportation3
NO
856
18,705
17,936
17,080
17,678
16,112
NO (Not Occurring)
a See Annex 3.2, Table A-76 for additional information on transportation consumption of these fuels.
13 Methodology and Time-Series Consistency
14 Woody biomass emissions were estimated by applying two gross heat contents from EIA (Lindstrom 2006) to U.S.
15 consumption data (EIA 2022a) (see Table 3-112), provided in energy units for the industrial, residential,
16 commercial, and electric power sectors. One heat content (16.95 MMBtu/MT wood and wood waste) was applied
17 to the industrial sector's consumption, while the other heat content (15.43 MMBtu/MT wood and wood waste)
18 was applied to the consumption data for the other sectors. An EIA emission factor of 0.434 MT C/MT wood
19 (Lindstrom 2006) was then applied to the resulting quantities of woody biomass to obtain CO2 emission estimates.
20 The woody biomass is assumed to contain black liquor and other wood wastes, have a moisture content of 12
21 percent, and undergo complete combustion to be converted into CO2.
22 Data for total waste incinerated, excluding tires, from 1990 to 2021 was derived following the methodology
23 described in Section 3.3. Biogenic CO2 emissions associated with MSW combustion were obtained from EPA's
24 GHGRP FLIGHT data for MSW combustion sources (EPA 2022b). Dividing biogenic CO2 emissions from GHGRP
25 FLIGHT data for MSW combustors by estimated MSW tonnage combusted yielded an annual biogenic CO2 emission
26 factor. This approach follows the same approach used to develop the fossil CO2 emissions from MSW combustion
27 as discussed in Section 3.3. As this data was only available following 2011, all years prior use an average of the
28 emission factors from 2011 through 2015.
29 Biogenic CO2 emissions from MSW combustion were calculated by multiplying the annual tonnage estimates,
30 excluding tires, by the calculated emissions factor. Calculated biogenic CO2 emission factors are shown in Table
31 3-111.
32 Table 3-111: Calculated Biogenic CO2 Content per Ton Waste (kg C02/Short Ton
33 Combusted)
1990
2005
2017
2018
2019
2020
2021
C02 Emission Factors
556
556
564
553
558
566
550
3-124 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 The amount of ethanol allocated across the transportation, industrial, and commercial sectors was based on the
2 sector allocations of ethanol-blended motor gasoline. The sector allocations of ethanol-blended motor gasoline
3 were determined using a bottom-up analysis conducted by EPA, as described in the Methodology section of Fossil
4 Fuel Combustion. Total U.S. ethanol consumption from EIA (2022a) was allocated to individual sectors using the
5 same sector allocations as ethanol-blended motor gasoline. The emissions from ethanol consumption were
6 calculated by applying an emission factor of 18.67 MMT C/Qbtu (EPA 2010) to adjusted ethanol consumption
7 estimates (see Table 3-113). The emissions from biodiesel consumption were calculated by applying an emission
8 factor of 20.1 MMT C/Qbtu (EPA 2010) to U.S. biodiesel consumption estimates that were provided in energy units
9 (EIA 2022a) (see Table 3-114).94
10 Table 3-112: Woody Biomass Consumption by Sector (Trillion Btu)
End-Use Sector
1990
2005
2017
2018
2019
2020
2021
Industrial
1,441.9
1,451.7
1,442.3
1,432.0
1,407.0
1,356.2
1,342.0
Residential
580.0
430.0
429.6
524.9
545.5
440.8
463.8
Commercial
65.7
70.0
83.7
84.1
84.3
83.0
82.7
Electric Power
128.5
185.0
229.3
221.1
200.5
185.4
199.0
Total
2,216.2
2,136.7
2,185.0
2,262.0
2,237.3
2,065.3
2,087.5
Note: Totals may not sum due to independent rounding.
11 Table 3-113: Ethanol Consumption by Sector (Trillion Btu)
End-Use Sector
1990
2005
2017
2018
2019
2020
2021
Transportation
59.3
315.8
1,134.6
1,148.2
1,150.2
994.6
1,114.3
Industrial
1.5
17.2
27.3
20.5
23.5
23.1
17.1
Commercial
0.9
2.2
37.2
27.9
32.6
31.9
23.6
Total
61.7
335.1
1,199.1
1,196.6
1,206.3
1,049.5
1,155.0
Note: Totals may not sum due to independent rounding.
12 Table 3-114: Biodiesel Consumption by Sector (Trillion Btu)
End-Use Sector 1990 2005 2017 2018 2019 2020 2021
Transportation NO VL6 253.3 242.9 231.3 239.4 218.2
NO (Not Occurring)
13 Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
14 through 2021.
is Uncertainty
16 It is assumed that the combustion efficiency for biomass is 100 percent, which is believed to be an overestimate of
17 the efficiency of biomass combustion processes in the United States. Decreasing the combustion efficiency would
18 decrease emission estimates for CO2. Additionally, the heat content applied to the consumption of woody biomass
19 in the residential, commercial, and electric power sectors is unlikely to be a completely accurate representation of
20 the heat content for all the different types of woody biomass consumed within these sectors. Emission estimates
21 from ethanol and biodiesel production are more certain than estimates from woody biomass consumption due to
22 better activity data collection methods and uniform combustion techniques.
94 C02 emissions from biodiesel do not include emissions associated with the C in the fuel that is from the methanol used in the
process. Emissions from methanol use and combustion are assumed to be accounted for under Non-Energy Use of Fuels. See
Annex 2.3 - Methodology for Estimating Carbon Emitted from Non-Energy Uses of Fossil Fuels.
Energy 3-125
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Recalculations Discussion
The CO2 emissions associated with the biogenic components of MSW were added to this year's report. The
emissions were calculated based on the same approach used to develop fossil CO2 emissions from the fossil
components of MSW as described in Section 3.3.
Planned Improvements
Future research will investigate the availability of data on woody biomass heat contents and carbon emission
factors to see if there are newer, improved data sources available for these factors.
Currently, emission estimates from biomass and biomass-based fuels included in this Inventory are limited to
woody biomass, biogenic components of MSW, ethanol, and biodiesel. Additional forms of biomass-based fuel
consumption include biogas, and other renewable diesel fuels. EPA will investigate additional forms of biomass-
based fuel consumption, research the availability of relevant emissions factors, and integrate these into the
Inventory as feasible. EPA will examine EIA data on biogas and other renewable diesel fuels to see if these fuel
types can be included in future Inventories. EIA (2022a) natural gas data already deducts biogas used in the natural
gas supply, so no adjustments are needed to the natural gas fuel consumption data to account for biogas. Distillate
fuel statistics are adjusted in this Inventory to remove other renewable diesel fuels as well as biodiesel.
Additionally, options for including "Other Renewable Fuels," as defined by EIA, will be evaluated.
The availability of facility-level combustion emissions through EPA's GHGRP will be examined to help better
characterize the industrial sector's energy consumption in the United States and further classify woody biomass
consumption by business establishments according to industrial economic activity type. Most methodologies used
in EPA's GHGRP are consistent with IPCC, although for EPA's GHGRP, facilities collect detailed information specific
to their operations according to detailed measurement standards, which may differ with the more aggregated data
collected for the Inventory to estimate total, national U.S. emissions. In addition, and unlike the reporting
requirements for this chapter under the UNFCCC reporting guidelines, some facility-level fuel combustion
emissions reported under EPA's GHGRP may also include industrial process emissions.95
In line with UNFCCC reporting guidelines, fuel combustion emissions are included in this chapter, while process
emissions are included in the Industrial Processes and Product Use chapter of this report. In examining data from
EPA's GHGRP that would be useful to improve the emission estimates for the CO2 from biomass combustion
category, particular attention will also be made to ensure time-series consistency, as the facility-level reporting
data from EPA's GHGRP are not available for all inventory years as reported in this Inventory. Additionally, analyses
will focus on aligning reported facility-level fuel types and IPCC fuel types per the national energy statistics,
ensuring CO2 emissions from biomass are separated in the facility-level reported data, and maintaining consistency
with national energy statistics provided by EIA. In implementing improvements and integration of data from EPA's
GHGRP, the latest guidance from the IPCC on the use of facility-level data in national inventories will be relied
upon.96
95 See https://unfccc.int/resource/docs/2013/copl9/ene/10a03.pdf#paee=2.
96 See http://www.ipcc-nggip.iges.or.lp/public/tb/TFI Technical Bulletin l.pdf.
3-126 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
3.11 Energy Sources of Precursor
Greenhouse Gases-TO BE UPDATED FOR
FINAL INVENTORY REPORT
In addition to the main greenhouse gases addressed above, energy-related activities are also sources of
greenhouse gas precursors. The reporting requirements of the UNFCCC97 request that information be provided
on
precursor emissions, which include carbon monoxide (CO), nitrogen oxides (NOx), non-methane volatile organic
compounds (NMVOCs), and sulfur dioxide (SO2). These gases are not direct greenhouse gases, but indirectly impact
Earth's radiative balance by altering the concentrations of greenhouse gases (e.g., tropospheric ozone) and
atmospheric aerosol (e.g., particulate sulfate). Total emissions of NOx, CO, NMVOCs, and SO2 from energy-related
activities from 1990 to 2020 are reported in Table 3-115.
Table 3-115: NOx, CO, NMVOC, and SO2 Emissions from Energy-Related Activities (kt)
Gas/Activity
1990
2005
2017
2018
2019
2020
2021
NOx
21,106
16,602
7,883
7,318
6,792
6,334
6,039
Fossil Fuel Combustion
20,885
16,153
7,246
6,622
6,225
5,768
5,473
Transportation
10,862
10,295
4,519
3,903
3,790
3,502
3,214
Industrial
2,559
1,515
859
898
864
864
864
Electric Power Sector
6,045
3,434
1,049
1,025
886
717
710
Commercial
671
490
537
512
402
402
402
Residential
749
418
283
283
284
284
284
Petroleum and Natural Gas Systems
137
301
530
586
465
465
465
Incineration of Waste
82
128
71
71
71
71
71
Other Energy
2
20
35
39
31
31
31
International Bunker Fuelsa
1,953
1,699
1,475
1,456
1,290
977
965
CO
125,640
64,985
33,401
31,455
30,959
30,177
29,394
Fossil Fuel Combustion
124,360
63,263
31,634
29,639
29,176
28,393
27,611
Transportation
119,360
58,615
27,942
25,957
25,580
24,798
24,015
Residential
3,668
2,856
2,291
2,286
2,286
2,286
2,286
Industrial
797
1,045
736
758
753
753
753
Electric Power Sector
329
582
532
505
424
424
424
Commercial
205
166
133
133
133
133
133
Petroleum and Natural Gas Systems
299
294
546
592
561
561
561
Incineration of Waste
978
1,403
1,175
1,175
1,175
1,175
1,175
Other Energy
3
24
46
49
47
47
47
International Bunker Fuelsa
102
131
153
158
154
83
83
NMVOCs
12,612
7,345
5,664
5,395
5,168
5,067
4,966
Fossil Fuel Combustion
11,836
6,594
3,293
2,686
2,635
2,534
2,433
Transportation
10,932
5,724
2,728
2,114
2,064
1,963
1,862
Residential
686
518
319
319
319
319
319
Commercial
10
188
116
116
116
116
116
Industrial
165
120
101
106
107
107
107
Electric Power Sector
43
44
29
30
28
28
28
Petroleum and Natural Gas Systems
552
497
2,205
2,534
2,362
2,362
2,362
Incineration of Waste
222
241
109
109
109
109
109
Other Energy
2
13
57
66
62
62
62
International Bunker Fuelsa
57
54
50
50
46
32
31
97 See http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
Energy 3-127
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
so2
19,628
12,364
1,793
1,724
1,405
1,222
1,313
Fossil Fuel Combustion
19,200
12,159
1,686
1,606
1,299
1,116
1,206
Electric Power Sector
14,433
9,439
1,257
1,193
921
739
831
Industrial
3,221
1,574
342
330
301
301
301
Transportation
793
619
48
45
40
38
37
Commercial
589
370
28
26
26
26
26
Residential
165
158
12
11
11
11
11
Petroleum and Natural Gas Systems
387
174
83
93
79
79
79
Incineration of Waste
38
25
22
22
24
24
24
Other Energy
3
5
2
3
2
2
2
International Bunker Fuelsa
NA
NA
NA
NA
NA
NA
NA
NA (Not Applicable)
a ¦
These values are presented for informational purposes only and are not included in totals.
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
Emission estimates for 1990 through 2020 were obtained from data published on the National Emissions Inventory
(NEI) Air Pollutant Emissions Trends Data website (EPA 2021a). For Table 3-117, NEI reported emissions of CO, NOx,
NMVOCs, and SO2 are recategorized from NEI Tier 1/Tier 2 source categories to those more closely aligned with
IPCC categories, based on EPA (2022).98 NEI Tier 1 emission categories related to the energy sector categories in
this report include: fuel combustion for electric utilities, industrial, and other; petroleum and related industries;
highway vehicles; off-highway; and waste disposal and recycling (incineration, open burning). As described in detail
in the NEI Technical Support Documentation (TSD) (EPA 2021b), NEI emissions are estimated through a
combination of emissions data submitted directly to the EPA by state, local, and tribal air agencies, as well as
additional information added by the Agency from EPA emissions programs, such as the emission trading program,
Toxics Release Inventory (TRI), and data collected during rule development or compliance testing.
Methodological approaches were applied to the entire time series to ensure time-series consistency from 1990
through 2020, which are described in detail in the NEI'sTSD and on EPA's Air Pollutant Emission Trends website
(EPA; EPA 2021). Updates to historical activity data are documented in NEI's TSD (EPA 2021). No quantitative
estimates of uncertainty were calculated for this source category.
98 The NEI estimates and reports emissions from six criteria air pollutants (CAPs) and 187 hazardous air pollutants (HAPs) in
support of National Ambient Air Quality Standards. Reported NEI emission estimates are grouped into 60 sectors and 15 Tier 1
source categories, which broadly cover similar source categories to those presented in this chapter. For this report, EPA has
mapped and regrouped emissions of greenhouse gas precursors (CO, NOx, S02, and NMVOCs) from NEI Tier 1/Tier 2 categories
to better align with NIR source categories, and to ensure consistency and completeness to the extent possible. See Annex 6.6
for more information on this mapping.
3-128 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
4.
Industrial Processes
and Product
Use
The Industrial Processes and Product Use (IPPU) chapter includes greenhouse gas emissions occurring from
industrial processes and from the use of greenhouse gases in products. The industrial processes and product use
categories included in this chapter are presented in Figure 4-1 and Figure 4-2. Greenhouse gas emissions from
industrial processes can occur in two different ways. First, they may be generated and emitted as the byproducts
of various non-energy-related industrial activities. Second, they may be emitted due to their use in manufacturing
processes or by end-consumers. Combustion-related energy use emissions from industry are reported in Chapter
3, Energy.
In the case of byproduct emissions, the emissions are generated by an industrial process itself and are not directly
a result of energy consumed during the process. For example, raw materials can be chemically or physically
transformed from one state to another. This transformation can result in the release of greenhouse gases such as
carbon dioxide (CO2), methane (CH4), nitrous oxide (N2O), and fluorinated greenhouse gases (e.g., HFC-23). The
greenhouse gas byproduct generating processes included in this chapter include iron and steel production and
metallurgical coke production, cement production, petrochemical production, ammonia production, lime
production, other process uses of carbonates (e.g., flux stone, flue gas desulfurization, and soda ash consumption
not associated with glass manufacturing), nitric acid production, adipic acid production, urea consumption for non-
agricultural purposes, aluminum production, HCFC-22 production, glass production, soda ash production,
ferroalloy production, titanium dioxide production, caprolactam production, zinc production, phosphoric acid
production, lead production, and silicon carbide production and consumption.
Greenhouse gases that are used in manufacturing processes or by end-consumers include man-made compounds
such as hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), sulfur hexafluoride (SFs), and nitrogen trifluoride
(NF3). The present contribution of HFCs, PFCs, SF6, and NF3 gases to the radiative forcing effect of all anthropogenic
greenhouse gases is small; however, because of their extremely long lifetimes, many of them will continue to
persist in the atmosphere long after they were first released. In addition, many of these gases have high global
warming potentials; SF6 is the most potent greenhouse gas the Intergovernmental Panel on Climate Change (IPCC)
has evaluated. Use of HFCs is growing rapidly since they are the primary substitutes for ozone depleting substances
(ODS), which are being phased-out under the Montreal Protocol on Substances that Deplete the Ozone Layer.
Hydrofluorocarbons, PFCs, SF6, and NF3 are employed and emitted by a number of other industrial sources in the
United States, such as electronics industry, electric power transmission and distribution, aluminum production,
and magnesium metal production and processing. Carbon dioxide is also consumed and emitted through various
end-use applications. In addition, nitrous oxide is used in and emitted by the electronics industry and anesthetic
and aerosol applications.
In 2021, IPPU generated emissions of 376.8 million metric tons of CO2 equivalent (MMT CO2 Eq.), or 5.9 percent of
total U.S. greenhouse gas emissions.1 Carbon dioxide emissions from all industrial processes were 169.3 MMT CO2
1 Emissions reported in the IPPU chapter include those from all 50 states, including Hawaii and Alaska, as well as from U.S.
Territories.
Industrial Processes and Product Use 4-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Eq. (169,298 kt CO2) in 2021, or 3.4 percent of total U.S. CO2 emissions. Methane emissions from industrial
processes resulted in emissions of approximately 0.4 MMT CO2 Eq. (16 kt CH4) in 2021, which was 0.1 percent of
U.S. Cm emissions. Nitrous oxide emissions from IPPU were 19.7 MMT CO2 Eq. (74 kt N2O) in 2021, or 5.1 percent
of total U.S. N2O emissions. In 2021 combined emissions of HFCs, PFCs, SF6, and NF3 totaled 187.3 MMT CO2 Eq.
Total emissions from IPPU in 2021 were 12.2 percent more than 1990 emissions. Total emissions from IPPU
remained relatively constant between 2020 and 2021, increasing by 3.7 percent due to offsetting trends within the
sector. More information on emissions of greenhouse gas precursors emissions that also result from IPPU are
presented in Section 4.27 of this chapter.
The largest source of IPPU-related emissions is the Substitution of Ozone Depleting Substances, which accounted
for 45.8 percent of sector emissions in 2021. These emissions have increased by 73.5 percent since 2005, and 3.8
percent between 2020 and 2021. Iron and Steel Production and Metallurgical Coke Production was the second
largest source of IPPU emissions in 2021, accounting for 11.2 percent of IPPU emissions in 2021. Cement
Production was the third largest source of IPPU emissions, accounting for 11.0 percent of the sector total in 2021.
Figure 4-1: 2021 Industrial Processes and Product Use Sector Greenhouse Gas Sources
Substitution of Ozone Depleting Substances
Iron and Steel Production & Metallurgical Coke Production
Cement Production
Petrochemical Production
Ammonia Production
Lime Production
Other Process Uses of Carbonates
Nitric Acid Production
Adipic Acid Production
Electrical Transmission and Distribution
Carbon Dioxide Consumption
Urea Consumption for Non-Agricultural Purposes
Electronics Industry
N2O from Product Uses
Aluminum Production
HCFC-22 Production
Glass Production
Soda Ash Production
Ferroalloy Production
Titanium Dioxide Production
Caprolactam, Glyoxal, and Glyoxylic Acid Production
Magnesium Production and Processing
Zinc Production
Phosphoric Acid Production
Lead Production
Carbide Production and Consumption
1172
Industrial Processes and Product Use as a
Portion of All Emissions
< 0.5
< 0.5
I Energy
I Agriculture
I IPPU
Waste
10
20
30 40
MMT CO2 Eq.
50
60
70
The increase in overall IPPU emissions since 1990 reflects a range of emission trends among the emission sources,
as shown in Figure 4-2. Emissions resulting from most types of metal production have declined significantly since
1990, largely due to production shifting to other countries, but also due to transitions to less-emissive methods of
production (in the case of iron and steel) and to improved practices (in the case of PFC emissions from aluminum
production). Carbon dioxide and CH4 emissions from some chemical production sources (e.g., petrochemical
production, urea consumption for non-agricultural purposes) have increased since 1990, while emissions from
other chemical production sources (e.g., ammonia production, phosphoric acid production) have decreased.
Emissions from mineral sources have either increased (e.g., cement production) or not changed significantly (e.g.,
lime production) since 1990 and largely follow economic cycles. Hydrofluorocarbon emissions from the
substitution of ODS have increased drastically since 1990 and are the largest source of IPPU emissions (45.8
percent in 2021), while the emissions of HFCs, PFCs, SF6, and NF3 from other sources have generally declined.
Nitrous oxide emissions from the production of nitric acid have decreased. Some emission sources (e.g., adipic
acid) exhibit varied interannual trends. Trends are explained further within each emission source category
throughout the chapter.
4-2 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Figure 4-2: Trends in Industrial Processes and Product Use Sector Greenhouse Gas Sources
500
450
400
350
. 300
8 250
I Electronics Industry
Other Product Manufacture and Use
I Mineral Industry
I Metal Industry
^ <•£>
° CO
N S "
m cn
200
150
100
50
0
I Chemical Industry
] Substitution of Ozone Depleting Substances
m *sD ud
Table 4-1 summarizes emissions for the IPPU chapter in MMT CO2 Eq. using IPCC Fifth Assessment Report (AR5)
GWP values, following the requirements of the current United Nations Framework Convention on Climate Change
(UNFCCC) reporting guidelines for national inventories (IPCC 2007).2 Unweighted gas emissions in kt are also
provided in Table 4-2. The source descriptions that follow in the chapter are presented in the order as reported to
the UNFCCC in the Common Reporting Format (CRF) tables, corresponding generally to: mineral industry, chemical
industry, metalindustry, and emissions from the uses of HFCs, PFCs, SF6, and NF3.
Each year, some emission and sink estimates in the IPPU sector of the Inventory are recalculated and revised with
improved methods and/or data. In general, recalculations are made to the U.S. greenhouse gas emission estimates
either to incorporate new methodologies or, most commonly, to update recent historical data. These
improvements are implemented consistently across the previous Inventory's time series (i.e., 1990 to 2020) to
ensure that the trend is accurate. Key updates to this year's inventory include revisions to the Ammonia
Production methodology to use GHGRP activity data for 2010 through 2021; Glass Production methodology to use
additional GHGRP activity data for the years 2010 through 2020; updates to emission estimates from Urea
Consumption for Non-Agricultural Purposes driven by revisions to quantities of urea applied, urea imports, and
urea exports; and revisions to method for electrical equipment for estimating historical emissions for non-Partners
based on the comparison with atmospheric data . In addition, estimates of CCh-equivalent emissions totals of Cm,
N2O, HFCs, PFCs, SFs and NF3 have been revised to reflect the 100-year global warming potentials (GWPs) provided
in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). AR5 GWP values differ slightly from those presented in the
IPCC Fourth Assessment Report (AR4) (IPCC 2007) (used in the previous inventories). Together, these updates
increased greenhouse gas emissions an average of 2.4 MMT CO2 Eq. (1 percent) across the time series. For more
information on specific methodological updates, please see the Recalculations Discussion section for each category
in this chapter.
2 See http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
Industrial Processes and Product Use 4-3
-------
l Table 4-1: Emissions from Industrial Processes and Product Use (MMT CO2 Eq.)
Gas/Source
1990
2005
2017
2018
2019
2020
2021
CO?
214.3
195.4
166.2
165.9
170.0
161.8
169.3
Iron and Steel Production &
Metallurgical Coke Production
104.7
70.1
40.8
42.9
43.1
37.7
42.0
Iron and Steel Production
99.1
66.2
38.8
41.6
40.1
35.4
38.8
Metallurgical Coke Production
5.6
3.9
2.0
1.3
3.0
2.3
3.2
Cement Production
33.5
46.2
40.3
39.0
40.9
40.7
41.3
Petrochemical Production
21.6
27.4
28.9
29.3
30.7
29.8
33.2
Ammonia Production
14.4
10.2
12.5
12.7
12.4
13.0
12.2
Lime Production
11.7
14.6
12.9
13.1
12.1
11.3
11.9
Other Process Uses of Carbonates
6.2
7.5
9.9
7.4
8.4
8.4
8.0
Carbon Dioxide Consumption
1.5
1.4
4.6
4.1
4.9
5.0
5.0
Urea Consumption for Non-
Agricultural Purposes
3.8
3.7
5.2
6.1
6.2
5.8
5.0
Glass Production
2.3
2.4
2.0
2.0
1.9
1.9
2.0
Soda Ash Production
1.4
1.7
1.8
1.7
1.8
1.5
1.7
Ferroalloy Production
2.2
1.4
2.0
2.1
1.6
1.4
1.6
Aluminum Production
6.8
4.1
1.2
1.5
1.9
1.7
1.5
Titanium Dioxide Production
1.2
1.8
1.7
1.5
1.5
1.2
1.5
Zinc Production
0.6
1.0
0.9
1.0
1.0
1.0
1.0
Phosphoric Acid Production
1.5
1.3
1.0
0.9
0.9
0.9
0.9
Lead Production
0.5
0.6
0.5
0.5
0.5
0.5
0.4
Carbide Production and
Consumption
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Substitution of Ozone Depleting
Substances3
+
+
+
+
+
+
+
Magnesium Production and
Processing
0.1
+
+
+
+
+
+
ch4
0.3
0.1
0.3
0.4
0.4
0.4
0.4
Petrochemical Production
0.2
0.1
0.3
0.3
0.4
0.3
0.4
Ferroalloy Production
+
+
+
+
+
+
+
Carbide Production and
Consumption
+
+
+
+
+
+
+
Iron and Steel Production &
Metallurgical Coke Production
+
+
+
+
+
+
+
Iron and Steel Production
+
+
+
+
+
+
+
Metallurgical Coke Production
NO
NO
NO
NO
NO
NO
NO
n2o
29.6
22.2
20.2
23.1
18.7
20.8
19.7
Nitric Acid Production
10.8
10.1
8.3
8.5
8.9
8.3
7.9
Adipic Acid Production
13.5
6.3
6.6
9.3
4.7
7.4
6.6
N20 from Product Uses
3.8
3.8
3.8
3.8
3.8
3.8
3.8
Caprolactam, Glyoxal, and Glyoxylic
Acid Production
1.5
1.9
1.3
1.3
1.2
1.2
1.2
Electronics Industry
+
0.1
0.2
0.2
0.2
0.3
0.3
HFCs
39.0
116.4
160.8
160.9
165.4
168.2
175.1
Substitution of Ozone Depleting
Substances3
0.3
99.4
156.1
157.7
161.9
166.1
172.4
HCFC-22 Production
38.6
16.8
4.3
2.7
3.1
1.8
2.2
Electronics Industry
0.2
0.2
0.3
0.3
0.3
0.3
0.4
Magnesium Production and
Processing
NO
NO
0.1
0.1
0.1
0.1
+
PFCs
21.8
6.1
3.8
4.3
4.0
3.9
3.5
Electronics Industry
2.5
3.0
2.7
2.8
2.5
2.4
2.6
4-4 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Aluminum Production
19.3
3.1
1.0
1.4
1.4
1.4
0.9
Substitution of Ozone Depleting
Substances
NO
+
+
+
+
+
+
Electrical Transmission and
Distribution
NO
+
+
NO
+
+
+
sf6
30.5
15.5
7.2
7.1
7.8
7.5
8.0
Electrical Transmission and
Distribution
24.7
11.8
5.5
5.2
6.1
5.9
6.0
Magnesium Production and
Processing
5.4
2.9
1.0
1.1
0.9
0.9
1.1
Electronics Industry
0.5
0.8
0.7
0.8
0.8
0.8
0.9
nf3
+
0.4
0.5
0.5
0.5
0.6
0.6
Electronics Industry
+
0.4
0.5
0.5
0.5
0.6
0.6
Totalb
335.7
356.1
359.1
362.2
366.8
363.2
376.8
+ Does not exceed 0.05 MMT C02 Eq.
NO (Not Occurring)
a Small amounts of PFC emissions also result from this source.
bTotal does not include other fluorinated gases, such as HFEs and PFPEs, which are reported separately in section 4.23.
Note: Totals may not sum due to independent rounding. Emissions of F-HTFs that are not HFCs, PFCs or SF6 are not included in
inventory totals and are included for informational purposes only in section 4.23. Emissions presented for informational
purposes include HFEs, PFPMIEs, perfluoroalkylmorpholines, and perfluorotrialkylamines.
l Table 4-2: Emissions from Industrial Processes and Product Use (kt)
Gas/Source
1990
2005
2017
2018
2019
2020
2021
co2
214,344
195,415
166,228
165,924
169,976
161,807
169,298
Iron and Steel Production &
Metallurgical Coke Production
104,737
70,076
40,810
42,858
43,090
37,712
42,041
Iron and Steel Production
99,129
66,156
38,832
41,576
40,084
35,387
38,817
Metallurgical Coke Production
5,608
3,921
1,978
1,282
3,006
2,325
3,224
Cement Production
33,484
46,194
40,324
38,971
40,896
40,688
41,312
Petrochemical Production
21,611
27,383
28,890
29,314
30,702
29,780
33,170
Ammonia Production
14,404
10,234
12,481
12,669
12,401
13,006
12,207
Lime Production
11,700
14,552
12,882
13,106
12,112
11,299
11,870
Other Process Uses of
Carbonates
6,233
7,459
9,869
7,351
8,422
8,399
7,951
Carbon Dioxide Consumption
1,472
1,375
4,580
4,130
4,870
4,970
4,990
Urea Consumption for Non-
Agricultural Purposes
3,784
3,653
5,161
6,111
6,154
5,814
4,989
Glass Production
2,262
2,401
1,984
1,989
1,940
1,858
1,969
Soda Ash Production
1,431
1,655
1,753
1,714
1„792
1,461
1,714
Ferroalloy Production
2,152
1,392
1,975
2,063
1,598
1,377
1,567
Aluminum Production
6,831
4,142
1,205
1,455
1,880
1,748
1,541
Titanium Dioxide Production
1,195
1,755
1,688
1,541
1,474
1,193
1,474
Zinc Production
632
1,030
900
999
1,026
977
969
Phosphoric Acid Production
1,529
1,342
1,025
937
909
901
909
Lead Production
516
553
513
527
531
464
446
Carbide Production and
Consumption
243
213
181
184
175
154
172
Substitution of Ozone Depleting
Substances3
+
1
3
3
3
4
4
Magnesium Production and
Processing
128
3
3
2
2
3
3
ch4
11
4
11
13
15
13
16
Petrochemical Production
9
3
10
12
13
12
15
Industrial Processes and Product Use 4-5
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Ferroalloy Production
1
+
1
1
+
+
+
Carbide Production and
Consumption
1
+
+
+
+
+
+
Iron and Steel Production &
Metallurgical Coke Production
1
1
+
+
+
+
+
Iron and Steel Production
1
1
+
+
+
+
+
Metallurgical Coke Production
NO
NO
NO
NO
NO
NO
NO
n2o
112
84
76
87
71
79
74
Nitric Acid Production
41
38
31
32
34
31
30
Adipic Acid Production
51
24
25
35
18
28
25
N20 from Product Uses
14
14
14
14
14
14
14
Caprolactam, Glyoxal, and
Glyoxylic Acid Production
6
7
5
5
5
4
5
Electronics Industry
+
+
1
1
1
1
1
HFCs
M
M
M
M
M
M
M
Substitution of Ozone Depleting
Substances3
M
M
M
M
M
M
M
HCFC-22 Production
3
1
+
+
+
+
+
Electronics Industry
NO
NO
+
+
+
+
+
Magnesium Production and
Processing
+
+
+
+
+
+
+
PFCs
M
M
M
M
M
M
M
Electronics Industry
+
+
+
+
+
+
+
Aluminum Production
M
M
M
M
M
M
M
Substitution of Ozone Depleting
Substances
NO
+
+
+
+
+
+
Electrical Transmission and
Distribution
NO
+
+
NO
+
+
+
sf6
1
1
+
+
+
+
+
Electrical Transmission and
Distribution
1
1
+
+
+
+
+
Magnesium Production and
Processing
+
+
+
+
+
+
+
Electronics Industry
+
+
+
+
+
+
+
nf3
+
+
+
+
+
+
+
Electronics Industry
+
+
+
+
+
+
+
+ Does not exceed 0.5 kt.
M (Mixture of gases)
NO (Not Occurring)
a Small amounts of PFC emissions also result from this source.
Note: Totals by gas may not sum due to independent rounding.
This chapter presents emission estimates calculated in accordance with the 2006IPCC Guidelines for National
Greenhouse Gas Inventories (2006 IPCC Guidelines) and its refinements. For additional detail on IPPU sources that
are not included in this Inventory report, please review Annex 5, Assessment of the Sources and Sinks of
Greenhouse Gas Emissions Not Included. These sources are not included due to various national circumstances,
such as emissions from a source may not currently occur in the United States, data are not currently available for
those emission sources (e.g., ceramics, non-metallurgical magnesium production, glyoxal and glyoxylic acid
production, Cm from direct reduced iron production), emissions are included elsewhere within the Inventory
report, or data suggest that emissions are not significant (e.g., other various fluorinated gas emissions from other
product uses). In terms of geographic scope, emissions reported in the IPPU chapter include those from all 50
states, including Hawaii and Alaska, as well as from District of Columbia and U.S. Territories to the extent to which
industries are occurring. While most IPPU sources do not occur in U.S. Territories (e.g., electronics manufacturing
does not occur in U.S. Territories), they are estimated and accounted for where they are known to occur (e.g.,
cement production, lime production, and electrical transmission and distribution). EPA will review this on an
4-6 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
ongoing basis to ensure emission sources are included across all geographic areas if they occur. Information on
planned improvements for specific IPPU source categories can be found in the Planned Improvements section of
the individual source category.
In addition, as mentioned in the Energy chapter of this report (Box 3-5), fossil fuels consumed for non-energy uses
for primary purposes other than combustion for energy (including lubricants, paraffin waxes, bitumen asphalt, and
solvents) are reported in the Energy chapter. According to the 2006IPCC Guidelines, these non-energy uses of
fossil fuels are to be reported under the IPPU, rather than the Energy sector; however, due to national
circumstances regarding the allocation of energy statistics and carbon balance data, the United States reports
these non-energy uses in the Energy chapter of this Inventory. Although emissions from these non-energy uses are
reported in the Energy chapter, the methodologies used to determine emissions are compatible with the 2006
IPCC Guidelines and are well documented and scientifically based. The methodologies used are described in
Section 3.2, Carbon Emitted from Non-Energy Uses of Fossil Fuels and Annex 2.3, Methodology for Estimating
Carbon Emitted from Non-Energy Uses of Fossil Fuels. The emissions are reported under the Energy chapter to
improve transparency, report a more complete carbon balance, and avoid double counting. For example, only the
emissions from the first use of lubricants and waxes are to be reported under the IPPU sector, and emissions from
use of lubricants in 2-stroke engines and emissions from secondary use of lubricants and waxes in waste
incineration with energy recovery are to be reported under the Energy sector. Reporting non-energy use emissions
from only first use of lubricants and waxes under IPPU would involve making artificial adjustments to the non-
energy use carbon balance and could potentially result in double counting of emissions. These artificial
adjustments would also be required for asphalt and road oil and solvents (which are captured as part of
petrochemical feedstock emissions) and could also potentially result in double counting of emissions. For more
information, see the Methodology discussion in Section 3.1, CO2 from Fossil Fuel Combustion, Section 3.2, Carbon
Emitted from Non-Energy Uses of Fossil Fuels and Annex 2.3, Methodology for Estimating Carbon Emitted from
Non-Energy Uses of Fossil Fuels.
Finally, as stated in the Energy chapter, portions of the fuel consumption data for seven fuel categories—coking
coal, distillate fuel, industrial other coal, petroleum coke, natural gas, residual fuel oil, and other oil—are
reallocated to the IPPU chapter, as they are consumed during non-energy related industrial process activity.
Emissions from uses of fossil fuels as feedstocks or reducing agents (e.g., petrochemical production, aluminum
production, titanium dioxide, zinc production) are reported in the IPPU chapter, unless otherwise noted due to
specific national circumstances. This approach is compatible with the 2006 IPCC Guidelines and is well documented
and scientifically based. The emissions from these feedstocks and reducing agents are reported under the IPPU
chapter to improve transparency and to avoid double counting of emissions under both the Energy and IPPU
sectors. More information on the methodology to adjust for these emissions within the Energy chapter is
described in the Methodology section of CO2 from Fossil Fuel Combustion (3.1 Fossil Fuel Combustion [CRF Source
Category 1A]) and Annex 2.1, Methodology for Estimating Emissions of CO2 from Fossil Fuel Combustion.
Additional information is listed within each IPPU emission source in which this approach applies.
Box 4-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals
In following the United Nations Framework Convention on Climate Change (UNFCCC) requirement under Article
4.1 to develop and submit national greenhouse gas emission inventories, the emissions and removals presented
in this report and this chapter are organized by source and sink categories and calculated using internationally
accepted methods provided by the Intergovernmental Panel on Climate Change (IPCC) in the 2006 IPCC
Guidelines for National Greenhouse Gas Inventories (2006 IPCC Guidelines) and its supplements and
refinements. Additionally, the calculated emissions and removals in a given year for the United States are
presented in a common format in line with the UNFCCC reporting guidelines for the reporting of inventories
under this international agreement. The use of consistent methods to calculate emissions and removals by all
nations providing their inventories to the UNFCCC ensures that these reports are comparable. The presentation
of emissions and removals provided in the IPPU chapter do not preclude alternative examinations, but rather,
this chapter presents emissions and removals in a common format consistent with how countries are to report
Inventories under the UNFCCC. The report itself, and this chapter, follows this standardized format, and
Industrial Processes and Product Use 4-7
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
provides an explanation of the application of methods used to calculate emissions and removals from industrial
processes and from the use of greenhouse gases in products.
QA/QC and Verification Procedures
For IPPU sources, a detailed QA/QC plan was developed and implemented for specific categories. This plan is
consistent with the U.S. Inventory QA/QC plan outlined in Annex 8 but tailored to include specific procedures
recommended for these sources. The IPPU QA/QC Plan does not replace the Inventory QA/QC Plan, but rather
provides more context for the IPPU sector. The IPPU QA/QC Plan provides the completed QA/QC forms for each
inventory reports, as well as, for certain source categories (e.g., key categories), more detailed documentation of
quality control checks and recalculations due to methodological changes.
Two types of checks were performed using this plan: (1) general (Tier 1) procedures consistent with Volume 1,
Chapter 6 of the 2006IPCC Guidelines that focus on annual procedures and checks to be used when gathering,
maintaining, handling, documenting, checking, and archiving the data, supporting documents, and files; and (2)
source category-specific (Tier 2) procedures that focus on checks and comparisons of the emission factors, activity
data, and methodologies used for estimating emissions from the relevant industrial process and product use
sources. Examples of these procedures include: checks to ensure that activity data and emission estimates are
consistent with historical trends to identify significant changes; that, where possible, consistent and reputable data
sources are used and specified across sources; that interpolation or extrapolation techniques are consistent across
sources; and that common datasets, units, and conversion factors are used where applicable. The IPPU QA/QC
plan also checked for transcription errors in data inputs required for emission calculations, including activity data
and emission factors; and confirmed that estimates were calculated and reported for all applicable and able
portions of the source categories for all years.
For sources that use data from EPA's Greenhouse Gas Reporting Program (GHGRP), EPA verifies annual facility-
level reports through a multi-step process (e.g., including a combination of pre-and post-submittal electronic
checks and manual reviews by staff) to identify potential errors and ensure that data submitted to EPA are
accurate, complete, and consistent.3 Based on the results of the verification process, EPA follows up with facilities
to resolve mistakes that may have occurred. The post-submittals checks are consistent with a number of general
and category-specific QC procedures, including: range checks, statistical checks, algorithm checks, and year-to-year
checks of reported data and emissions. See Box 4-2 below for more information on use of GHGRP data in this
chapter.
General QA/QC procedures (Tier 1) and calculation-related QC (category-specific, Tier 2) have been performed for
all IPPU sources. Consistent with the 2006 IPCC Guidelines, additional category-specific QC procedures were
performed for more significant emission categories (such as the comparison of reported consumption with
modeled consumption using EPA's Greenhouse Gas Reporting Program (GHGRP) data within Substitution of Ozone
Depleting Substances) or sources where significant methodological and data updates have taken place. The QA/QC
implementation did not reveal any significant inaccuracies, and all errors identified were documented and
corrected. Application of these procedures, specifically category-specific QC procedures and
updates/improvements as a result of QA processes (expert, public, and UNFCCC technical expert reviews), are
described further within respective source categories, in the Recalculations Discussion and Planned Improvement
sections.
For most IPPU categories, activity data are obtained via aggregation of facility-level data from EPA's GHGRP (see
Box 4-2 below and Annex 9), national commodity surveys conducted by U.S. Geological Survey (USGS) National
Minerals Information Center, U.S. Department of Energy (DOE), U.S. Census Bureau, and industry associations such
as Air-Conditioning, Heating, and Refrigeration Institute (AHRI), American Chemistry Council (ACC), and American
3 See https://www.epa.gov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
4-8 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Iron and Steel Institute (AISI) (specified within each source category). The emission factors used include those
2 derived from the EPA's GHGRP and application of IPCC default factors. Descriptions of uncertainties and
3 assumptions for activity data and emission factors are included within the uncertainty discussion sections for each
4 IPPU source category.
Box 4-2: Industrial Process and Product Use Data from EPA's Greenhouse Gas Reporting Program
EPA collects greenhouse gas emissions data from individual facilities and suppliers of certain fossil fuels and
industrial gases through its Greenhouse Gas Reporting Program (GHGRP). The GHGRP applies to direct
greenhouse gas emitters, fossil fuel suppliers, industrial gas suppliers, and facilities that inject CO2 underground
for sequestration or other reasons and requires reporting by sources or suppliers in 41 industrial categories.
Annual reporting is at the facility level, except for certain suppliers of fossil fuels and industrial greenhouse
gases.
In general, the threshold for reporting is 25,000 metric tons or more of CO2 Eq. per year, but reporting is
required for all facilities in some industries. Calendar year 2010 was the first year for which data were collected
for facilities subject to 40 CFR Part 98, though some source categories first collected data for calendar year
2011. For more information, see Annex 9, Use of EPA Greenhouse Gas Reporting Program in Inventory.
EPA uses annual GHGRP data in a number of categories to improve the national estimates presented in this
Inventory, consistent with IPCC guidelines (e.g., minerals, chemicals, product uses). Methodologies used in
EPA's GHGRP are consistent with IPCC guidelines, including higher tier methods; however, it should be noted
that the coverage and definitions for source categories (e.g., allocation of energy and IPPU emissions) in EPA's
GHGRP may differ from those used in this Inventory in meeting the UNFCCC reporting guidelines (IPCC 2011)
and is an important consideration when incorporating GHGRP data in the Inventory. In line with the UNFCCC
reporting guidelines, the Inventory is a comprehensive accounting of all emissions from source categories
identified in the 2006 IPCC Guidelines. EPA has paid particular attention to ensuring both completeness and
time-series consistency for major recalculations that have occurred from the incorporation of GHGRP data into
these categories, consistent with 2006 IPCC Guidelines and IPCC Technical Bulletin on Use of Facility-Specific
Data in National GHG Inventories,4
For certain source categories in this Inventory (e.g., nitric acid production, lime production, cement production,
petrochemical production, carbon dioxide consumption, ammonia production, and urea consumption for non-
agricultural purposes), EPA has integrated data values that have been calculated by aggregating GHGRP data
that are considered confidential business information (CBI) at the facility level. EPA, with industry engagement,
has put forth criteria to confirm that a given data aggregation shields underlying CBI from public disclosure. EPA
is only publishing data values that meet these aggregation criteria.5 Specific uses of aggregated facility-level
data are described in the respective methodological sections (e.g., including other sources using GHGRP data
that is not aggregated CBI, such as aluminum, electronics industry, electrical transmission and distribution,
HCFC-22 production, and magnesium production and processing). For other source categories in this chapter, as
indicated in the respective planned improvements sections,6 EPA is continuing to analyze how facility-level
GHGRP data may be used to improve the national estimates presented in this Inventory, giving particular
consideration to ensuring time-series consistency and completeness.
Additionally, EPA's GHGRP has and will continue to enhance QA/QC procedures and assessment of uncertainties
within the IPPU categories (see those categories for specific QA/QC details regarding the use of GHGRP data).
6
4 See http://www.ipcc-negip.iges.or.jp/public/tb/TFI Technical Bulletin l.pdf.
5 U.S. EPA Greenhouse Gas Reporting Program. Developments on Publication of Aggregated Greenhouse Gas Data, November
25, 2014. See http://www.epa.gov/ghgreporting/confidential-business-information-ghg-reporting.
6 Ammonia Production, Glass Production, Lead Production, and Other Fluorinated Gas Production.
Industrial Processes and Product Use 4-9
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
4.1 Cement Production (CRF Source
Category 2A1)
Cement production is an energy- and raw material-intensive process that results in the generation of carbon
dioxide (CO2) 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 chapter.
During the clinker production process, the key reaction occurs when calcium carbonate (CaCOs), in the form of
limestone or similar rocks or in the form of cement kiln dust (CKD), is heated in a cement kiln at a temperature
range of about 700 to 1,000 degrees Celsius (1,300 to 1,800 degrees Fahrenheit) to form lime (i.e., calcium oxide,
or CaO) and CO2 in a process known as calcination or calcining. The quantity of CO2 emitted during clinker
production is directly proportional to the lime content of the clinker. During calcination, each mole of CaCC>3
heated in the clinker kiln forms one mole of CaO and one mole of CO2. The CO2 is vented to the atmosphere as part
of the kiln exhaust:
CaC03 + heat -» CaO + C02
Next, over a temperature range of 1000 to 1450 degrees Celsius, the CaO combines with alumina, iron oxide and
silica that are also present in the clinker raw material mix to form hydraulically reactive compounds within white-
hot semifused (sintered) nodules of clinker. These "sintering" reactions are highly exothermic and produce few CO2
process emissions. The clinker is then rapidly cooled to maintain quality and then very finely ground with a small
amount of gypsum and potentially other materials (e.g., ground granulated blast furnace slag, etc.) to make
Portland and similar cements.
Masonry cement consists of plasticizers (e.g., ground limestone, lime, etc.) and portland cement, and the amount
of portland cement used accounts for approximately 3 percent of total clinker production (USGS 2022a). No
additional emissions are associated with the production of masonry cement. Carbon dioxide emissions that result
from the production of lime used to produce portland and masonry cement are included in Section 4.2 Lime
Production (CRF Source Category 2A2).
Carbon dioxide emitted from the chemical process of cement production is the second largest source of industrial
CO2 emissions in the United States. Cement is produced in 34 states and Puerto Rico. Texas, Missouri, California,
and Florida were the leading cement-producing states in 2021 and accounted for almost 44 percent of total U.S.
production (USGS 2022b). In 2021, shipments of cement were estimated to have slightly increased from 2020, and
net imports increased by about 20 percent compared to 2020 (USGS 2022b). Clinker production in 2021 increased
by 1.5 percent, compared to 2020 (EPA 2022; USGS 2022b). In 2021, U.S. clinker production totaled 79,400 kilotons
(EPA 2022). The resulting CO2 emissions were estimated to be 41.3 MMT CO2 Eq. (41,312 kt) (see Table 4-3). The
total construction value and cement shipments increased during the first nine months of 2021 compared to the
same time period in 2020. This increase was attributed to economic recovery from the COVID-19 pandemic.
Despite the slight increases, growth was constrained by increased costs, labor shortages, logistical issues, and
supply chain disruptions (USGS 2022b).
Table 4-3: CO2 Emissions from Cement Production (MMT CO2 Eq. and kt)
Year
1990
2005
2017
2018
2019
2020
2021
MMTCOz Eq.
33.5
46.2
40.3
39.0
40.9
40.7
41.3
kt
33,484
46,194
40,324
38,971
40,896
40,688
41,312
Greenhouse gas emissions from cement production, which are primarily driven by production levels, increased
every year from 1991 through 2006 but decreased in the following years until 2009. Since 1990, emissions have
increased by 23 percent. Emissions from cement production were at their lowest levels in 2009 (2009 emissions
4-10 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
are approximately 28 percent lower than 2008 emissions and 12 percent lower than 1990) due to the economic
recession and the associated decrease in demand for construction materials. Since 2009, emissions have increased
by nearly 40 percent due to increasing demand for cement. Cement continues to be a critical component of the
construction industry; therefore, the availability of public and private construction funding, as well as overall
economic conditions, have considerable impact on the level of cement production.
Methodology and Time-Series Consistency
Carbon dioxide emissions from cement production were estimated using the Tier 2 methodology from the 2006
IPCC Guidelines as this is a key category. The Tier 2 methodology was used because detailed and complete data
(including weights and composition) for carbonate(s) consumed in clinker production are not available,7 and thus a
rigorous Tier 3 approach is impractical. Tier 2 specifies the use of aggregated plant or national clinker production
data and an emission factor, which is the product of the average lime mass fraction for clinker of 65 percent and a
constant reflecting the mass of CO2 released per unit of lime. The U.S. Geological Survey (USGS) mineral
commodity expert for cement has confirmed that this is a reasonable assumption for the United States (Van Oss
2013a). This calculation yields an emission factor of 0.510 tons of CO2 per ton of clinker produced, which was
determined as follows:
Equation 4-1: 2006IPCCGuide/inesTier 1 Emission Factor for Clinker (precursor to Equation
2.4)
EFciinker = 0.650 CaO x [(44.01 g/mole CO2) -h (56.08 g/mole CaO)] = 0.510 tons C02/ton clinker
During clinker production, some of the raw materials, partially reacted raw materials, and clinker enters 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 which are then calcined, there are associated CChemissions.
At some plants, essentially all CKD is directly returned to the kiln, becoming part of the raw material feed, or is
likewise returned to the kiln after first being removed from the exhaust. In either case, the returned CKD becomes
a raw material, thus forming clinker, and the associated CO2 emissions are a component of those calculated for the
clinker overall. At some plants, however, the CKD cannot be returned to the kiln because it is chemically unsuitable
as a raw material or chemical issues limit the amount of CKD that can be so reused. Any clinker that cannot be
returned to the kiln is either used for other (non-clinker) purposes or is landfilled. The CO2 emissions attributable
to the non-returned calcinated portion of the CKD are not accounted for by the clinker emission factor and thus a
CKD correction factor should be applied to account for those emissions. The USGS reports the amount of CKD used
to produce clinker, but no information is currently available on the total amount of CKD produced annually.8
Because data are not currently available to derive a country-specific CKD correction factor, a default correction
factor of 1.02 (2 percent) was used to account for CKD CO2 emissions, as recommended by the IPCC (IPCC 2006).9
Total cement production emissions were calculated by adding the emissions from clinker production and the
emissions assigned to CKD.
7 As discussed further under "Planned Improvements," most cement-producing facilities that report their emissions to the
GHGRP use CEMS to monitor combined process and fuel combustion emissions for kilns, making it difficult to quantify the
process emissions on a facility-specific basis. In 2021, the percentage of facilities not using CEMS was 4 percent.
8 The USGS Minerals Yearbook: Cement notes that CKD values used for clinker production are likely underreported.
9 As stated on p. 2.12 of the 2006 IPCC Guidelines, Vol. 3, Chapter 2: "...As data on the amount of CKD produced may be scarce
(except possibly for plant-level reporting), estimating emissions from lost CKD based on a default value can be considered good
practice. The amount of C02 from lost CKD can vary, but ranges typically from about 1.5 percent (additional C02 relative to that
calculated for clinker) for a modern plant to about 20 percent for a plant losing a lot of highly calcinated CKD (van Oss 2005). In
the absence of data, the default CKD correction factor (CFckd) is 1.02 (i.e., add 2 percent to the CO2 calculated for clinker). If no
calcined CKD is believed to be lost to the system, the CKD correction factor will be 1.00 (van Oss 2005)..."
Industrial Processes and Product Use 4-11
-------
1 Small amounts of impurities (i.e., not calcium carbonate) may exist in the raw limestone used to produce clinker.
2 The proportion of these impurities is generally minimal, although a small amount (1 to 2 percent) of magnesium
3 oxide (MgO) may be desirable as a flux. Per the IPCC Tier 2 methodology, a correction for MgO is not used, since
4 the amount of MgO from carbonate is likely very small and the assumption of a 100 percent carbonate source of
5 CaO already yields an overestimation of emissions (IPCC 2006).
6 The 1990 through 2012 activity data for clinker production were obtained from USGS (Van Oss 2013a, Van Oss
7 2013b). Clinker production data for 2013 were also obtained from USGS (USGS 2014). USGS compiled the data (to
8 the nearest ton) through questionnaires sent to domestic clinker and cement manufacturing plants, including
9 facilities in Puerto Rico. Clinker production values in the current Inventory report utilize GHGRP data for the years
10 2014 through 2021 (EPA 2022). Clinker production data are summarized in Table 4-4. Details on how this GHGRP
11 data compares to USGS reported data can be found in the section on QA/QC and Verification.
12 Table 4-4: Clinker Production (kt)
Year
1990
2005
2017
2018
2019
2020
2021
Clinker
64,355
88,783
77,500
74,900
78,600
78,200
79,400
13 Notes: Clinker production from 1990 through 2021 includes Puerto Rico (relevant U.S. Territories).
14 Methodological approaches were applied to the entire time series to ensure time-series consistency from 1990
15 through 2021. The methodology for cement production spliced activity data from two different sources: USGS for
16 1990 through 2013 and GHGRP starting in 2014. Consistent with the 2006 IPCC Guidelines, the overlap technique
17 was applied to compare the two data sets for years where there was overlap, with findings that the data sets were
18 consistent and adjustments were not needed.
19 Uncertainty - TO BE UPDATED FOR FINAL INVENTORY REPORT
20 The uncertainties contained in these estimates are primarily due to uncertainties in the lime content of clinker and
21 in the percentage of CKD recycled inside the cement kiln. Uncertainty is also associated with the assumption that
22 all calcium-containing raw materials are CaCC>3, when a small percentage likely consists of other carbonate and
23 non-carbonate raw materials. The lime content of clinker varies from 60 to 67 percent; 65 percent is used as a
24 representative value (Van Oss 2013a). This contributes to the uncertainty surrounding the emission factor for
25 clinker which has an uncertainty range of ±3 percent with uniform densities (Van Oss 2013b). The amount of CO2
26 from CKD loss can range from 1.5 to 8 percent depending upon plant specifications, and uncertainty was estimated
27 at ±5 percent with uniform densities (Van Oss 2013b). Additionally, some amount of CO2 is reabsorbed when the
28 cement is used for construction. As cement reacts with water, alkaline substances such as calcium hydroxide are
29 formed. During this curing process, these compounds may react with CO2 in the atmosphere to create calcium
30 carbonate. This reaction only occurs in roughly the outer 0.2 inches of the total thickness. Because the amount of
31 CO2 reabsorbed is thought to be minimal, it was not estimated. EPA assigned default uncertainty bounds of ±3
32 percent for clinker production, based on expert judgment (Van Oss 2013b).
33 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-5. Based on the
34 uncertainties associated with total U.S. clinker production, the CO2 emission factor for clinker production, and the
35 emission factor for additional CO2 emissions from CKD, 2021 CO2 emissions from cement production were
36 estimated to be between 38.3 and 43.1 MMT CO2 Eq. at the 95 percent confidence level. This confidence level
37 indicates a range of approximately 6 percent below and 6 percent above the emission estimate of 40.7 MMT CO2
38 Eq.
39 Table 4-5: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Cement
40 Production (MMT CO2 Eq. and Percent)
Source Gas
2021 Emission Estimate
Uncertainty Range Relative to Emission Estimate a
(MMT C02 Eq.)
(MMT C02 Eq.) (%)
Lower Upper Lower Upper
4-12 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Bound Bound
Bound
Bound
Cement Production
C02
40.7
38.3 43.1
-6%
+6%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
1 QA/QC and Verification
2 General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S. Inventory
3 QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006IPCC Guidelines as described in the
4 introduction of the IPPU chapter (see Annex 8 for more details).
5 EPA relied upon the latest guidance from the IPCC on the use of facility-level data in national inventories and
6 applied a category-specific QC process to compare activity data from EPA's GHGRP with existing data from USGS
7 surveys. This was to ensure time-series consistency of the emission estimates presented in the Inventory. Total
8 U.S. clinker production is assumed to have low uncertainty because facilities routinely measure this for economic
9 reasons and because both USGS and GHGRP take multiple steps to ensure that reported totals are accurate. EPA
10 verifies annual facility-level GHGRP reports through a multi-step process that is tailored to the reporting industry
11 (e.g., combination of electronic checks including range checks, statistical checks, algorithm checks, year-to-year
12 comparison checks, along with manual reviews involving outside data checks) to identify potential errors and
13 ensure that data submitted to EPA are accurate, complete, and consistent (EPA 2015). Based on the results of the
14 verification process, EPA follows up with facilities to resolve mistakes that may have occurred.10 Facilities are also
15 required to monitor and maintain records of monthly clinker production per section 98.84 of the GHGRP regulation
16 (40 CFR 98.84).
17 EPA's GHGRP requires all facilities producing Portland cement to report greenhouse gas emissions, including CO2
18 process emissions from each kiln, CO2 combustion emissions from each kiln, CH4 and N2O combustion emissions
19 from each kiln, and CO2, CH4, and N2O emissions from each stationary combustion unit other than kilns (40 CFR
20 Part 98 Subpart H). Source-specific quality control measures for the Cement Production category are included in
21 section 98.84, Monitoring and QA/QC Requirements.
22 As mentioned above, EPA compares GHGRP clinker production data to the USGS clinker production data. For the
23 year 2014 and 2020, USGS and GHGRP clinker production data showed a difference of approximately 1 percent. In
24 2018 the difference was approximately 3 percent. In 2015, 2016, 2017, 2019, and 2021, that difference was less
25 than 1 percent between the two sets of activity data. This difference resulted in a difference in emissions
26 compared to USGS data of about 0.1 MMT CO2 Eq. in 2015, 2016, 2017, 2019, and 2021. The information collected
27 by the USGS National Minerals Information Center surveys continue to be an important data source.
28 Recalculations Discussion
29 No recalculations were performed for the 1990 through 2020 portion of the time series.
30 Planned Improvements
31 EPA is continuing to evaluate and analyze data reported under EPA's GHGRP that would be useful to improve the
32 emission estimates for the Cement Production source category. Most cement production facilities reporting under
33 EPA's GHGRP use Continuous Emission Monitoring Systems (CEMS) to monitor and report CO2 emissions, thus
34 reporting combined process and combustion emissions from kilns. In implementing further improvements and
35 integration of data from EPA's GHGRP, the latest guidance from the IPCC on the use of facility-level data in national
36 inventories will be relied upon, in addition to category-specific QC methods recommended by the 2006 IPCC
10 See GHGRP Verification Fact Sheet https://www.epa.gov/sites/production/files/2Q15-
07/documents/ghgrp verification factsheet.pdf.
Industrial Processes and Product Use 4-13
-------
1 Guidelines.11 EPA's long-term improvement plan includes continued assessment of the feasibility of using
2 additional GHGRP information beyond aggregation of reported facility-level clinker data, in particular
3 disaggregating the combined process and combustion emissions reported using CEMS, to separately present
4 national process and combustion emissions streams consistent with IPCC and UNFCCC guidelines. This long-term
5 planned analysis is still in development and has not been applied for this current Inventory.
6 In response to feedback from Portland Cement Association (PCA) during the Public Review comment period of a
7 previous Inventory, EPA plans to work with PCA to discuss additional long-term improvements to review methods
8 and data used to estimate CO2 emissions from cement production to account for organic material in the raw
9 material and to discuss the carbonation that occurs across the duration of the cement product. Work includes
10 identifying data and studies on the average carbon content for organic materials in kiln feed in the United States
11 and CO2 reabsorption rates via carbonation for various cement products. This information is not reported by
12 facilities subject to GHGRP reporting.
13
14 4.2 Lime Production (CRF Source Category
15 2A2)
16 Lime is an important manufactured product with many industrial, chemical, and environmental applications. Lime
17 production involves three main processes: stone preparation, calcination, and hydration. Carbon dioxide (CO2) is
18 generated during the calcination stage, when limestone—consisting of calcium carbonate (CaCOs) and/or
19 magnesium carbonate (MgCOs)—is roasted at high temperatures in a kiln to produce calcium oxide (CaO) and CO2.
20 The CO2 is given off as a gas and is normally emitted to the atmosphere.
21 CaCO3 —>CaO + C02
22 Some facilities, however, recover CO2 generated during the production process for use in sugar refining and
23 precipitated calcium carbonate (PCC) production.12 PCC is used as a filler or coating in the paper, food, and plastic
24 industries and is derived from reacting hydrated high-calcium quicklime with CO2, a production process that does
25 not result in net emissions of CO2 to the atmosphere. Emissions from fuels consumed for energy purposes during
26 the production of lime are included in the Energy chapter.
27 For U.S. operations, the term "lime" actually refers to a variety of chemical compounds. These include CaO, or
28 high-calcium quicklime; calcium hydroxide (Ca(OH)2), or hydrated lime; dolomitic quicklime ([CaOMgO]); and
29 dolomitic hydrate ([Ca(OH)2*MgO] or [Ca(OH)2*Mg(OH)2]).
30 The current lime market is approximately distributed across six end-use categories, as follows: metallurgical uses,
31 35 percent; environmental uses, 29 percent; chemical and industrial uses, 21 percent; construction uses, 10
32 percent; miscellaneous uses, 3 percent; and refractory dolomite, 1 percent (USGS 2021c). The major uses are in
33 steel making, chemical and industrial applications (such as the manufacture of fertilizer, glass, paper and pulp, and
34 precipitated calcium carbonate, and in sugar refining), flue gas desulfurization (FGD) systems at coal-fired electric
35 power plants, construction, and water treatment, as well as uses in mining, pulp and paper and precipitated
36 calcium carbonate manufacturing (USGS 2022a). Lime is also used as a CO2 scrubber, and there has been
37 experimentation on the use of lime to capture CO2 from electric power plants. Both lime (CaO) and limestone
11 See IPCC Technical Bulletin on Use of Facility-Specific Data in National Greenhouse Gas Inventories http://www.ipcc-
nBBip.iBes.or.jp/public/tb/TFI Technical Bulletin l.pdf.
12 The amount of C02 captured for sugar refining and PCC production is reported within the CRF tables under CRF Source
Category 2H3, but within this report, they are included in this chapter.
4-14 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 (CaCC>3) can be used as a sorbent for FGD systems. Emissions from limestone consumption for FGD systems are
2 reported under Section 4.4 Other Process Uses of Carbonate Production (CRF Source Category 2A4).
3 Emissions from lime production have fluctuated over the time series depending on lime end-use markets -
4 primarily the steel making industry and FGD systems for utility and industrial plants - and also energy costs. One
5 significant change to lime end-use since 1990 has been the increase in demand for lime for FGD at coal-fired
6 electric power plants, which can be attributed to compliance with sulfur dioxide (SO2) emission regulations of the
7 Clean Air Act Amendments of 1990. Phase I went into effect on January 1,1995, followed by Phase II on January 1,
8 2000. To supply lime for the FGD market, the lime industry installed more than 1.8 million tons per year of new
9 capacity by the end of 1995 (USGS 1996). The need for air pollution controls continued to drive the FGD lime
10 market, which had doubled between 1990 and 2019 (USGS 1991 and 2020d).
11 The U.S. lime industry temporarily shut down some individual gas-fired kilns and, in some case, entire lime plants
12 during 2000 and 2001, due to significant increases in the price of natural gas. Lime production continued to
13 decrease in 2001 and 2002, a result of lower demand from the steel making industry, lime's largest end-use
14 market, when domestic steel producers were affected by low priced imports and slowing demand (USGS 2002).
15 Emissions from lime production increased and then peaked in 2006 at approximately 30.3 percent above 1990
16 levels, due to strong demand from the steel and construction markets (road and highway construction projects),
17 before dropping to its lowest level in 2009 at approximately 2.5 percent below 1990 emissions, driven by the
18 economic recession and downturn in major markets including construction, mining, and steel (USGS 2007, 2008,
19 2010). In 2010, the lime industry began to recover as the steel, FGD, and construction markets also recovered
20 (USGS 2011 and 2012). Fluctuation in lime production since 2015 has been driven largely by demand from the steel
21 making industry (USGS 2018b, 2019, 2020b, 2021c). In 2020, a decline in lime production was a result of plants
22 temporarily closing as a result of the global COVID-19 pandemic (USGS 2022a).
23 Lime production in the United States—including Puerto Rico—was reported to be 16,774 kilotons in 2021, an
24 increase of about 5.7 percent compared to 2020 levels (USGS 2022b). Compared to 1990, lime production
25 increased by about 5.9 percent. At year-end 2021, 73 primary lime plants were operating in the United States,
26 including Puerto Rico according to the USGS MCS (USGS 2022a).13 Principal lime producing states were, in
27 alphabetical order, Alabama, Kentucky, Missouri, Ohio, and Texas (USGS 2022a).
28 U.S. lime production resulted in estimated net CO2 emissions of 11.9 MMT CO2 Eq. (11,870 kt) (see Table 4-6 and
29 Table 4-7). Carbon dioxide emissions from lime production increased by about 5.1 percent compared to 2020
30 levels. Compared to 1990, CO2 emissions have increased by about 1.5 percent. The trends in CO2 emissions from
31 lime production are directly proportional to trends in production, which are described above.
32 Table 4-6: CO2 Emissions from Lime Production (MMT CO2 Eq. and kt)
Year 1990 2005 2017 2018 2019 2020 2021
MMT C02 Eq. 11.7 14.6 12.9 13.1 12.1 11.3 11.9
kt 11,700 14,552 12,882 13,106 12,112 11,299 11,870
33
34 Table 4-7: Gross, Recovered, and Net CO2 Emissions from Lime Production (kt)
~Year 1990 2005 2017 2018 2019 2020 2021
Gross 11,959 15,074 13,283 13,609 12,676 11,875 12,586
Recovered3 259 522 401 503 564 576 716
Net Emissions 11,700 14,552 12,882 13,106 12,112 11,299 11,870
Note: Totals may not sum due to independent rounding.
a For sugar refining and PCC production.
13 In 2021, 68 operating primary lime facilities in the United States reported to the EPA Greenhouse Gas Reporting Program.
Industrial Processes and Product Use 4-15
-------
1 Methodology and Time-Series Consistency
2 To calculate emissions, the amounts of high-calcium and dolomitic lime produced were multiplied by their
3 respective emission factors using the Tier 2 approach from the 2006IPCC Guidelines. The emission factor is the
4 product of the stoichiometric ratio between CO2 and CaO, and the average CaO and MgO content for lime. The
5 CaO and MgO content for lime is assumed to be 95 percent for both high-calcium and dolomitic lime (IPCC 2006).
6 The emission factors were calculated as follows:
7 Equation 4-2: 2006IPCCGuide/inesTier 2 Emission Factor for Lime Production, High-
8 Calcium Lime (Equation 2.9)
9 EFHigh-Calcium Lime = [(44.01 g/mole C02) 4- (56.08 g/mole CaO)] x (0.9500 CaO/lime) = 0.7455 g C02/g lime
10 Equation 4-3: 2006IPCC Guide/inesTier 2 Emission Factor for Lime Production, Dolomitic
11 Lime (Equation 2.9)
12 E FDolomitic Lime — [(88.02 g/mole C02) 4 (96.39 g/mole CaO*MgO)] x (0.9500 CaO*MgO /lime) = 0.8675 g
13 C02/g lime
14 Production was adjusted to remove the mass of chemically combined water found in hydrated lime, determined
15 according to the molecular weight ratios of H2O to (Ca(OH)2 and [Ca(OH)2*Mg(OH)2]) (IPCC 2006). These factors set
16 the chemically combined water content to 27 percent for high-calcium hydrated lime, and 30 percent for dolomitic
17 hydrated lime.
18 The 2006 IPCC Guidelines (Tier 2 method) also recommends accounting for emissions from lime kiln dust (LKD)
19 through application of a correction factor. LKD is a byproduct of the lime manufacturing process typically not
20 recycled back to kilns. LKD is a very fine-grained material and is especially useful for applications requiring very
21 small particle size. Most common LKD applications include soil reclamation and agriculture. Emissions from the
22 application of lime for agricultural purposes are reported in the Agriculture chapter under 5.5 Liming (CRF Source
23 Category 3G). Currently, data on annual LKD production is not readily available to develop a country-specific
24 correction factor. Lime emission estimates were multiplied by a factor of 1.02 to account for emissions from LKD
25 (IPCC 2006). See the Planned Improvements section associated with efforts to improve uncertainty analysis and
26 emission estimates associated with LKD.
27 Lime emission estimates were further adjusted to account for the amount of CO2 captured for use in on-site
28 processes. All the domestic lime facilities are required to report these data to EPA under its GHGRP. The total
29 national-level annual amount of CO2 captured for on-site process use was obtained from EPA's GHGRP (EPA 2022)
30 based on reported facility-level data for years 2010 through 2021. The amount of CO2 captured/recovered for non-
31 marketed on-site process use is deducted from the total gross emissions (i.e., from lime production and LKD). The
32 net lime emissions are presented in Table 4-6 and Table 4-7. GHGRP data on CO2 removals (i.e., CO2
33 captured/recovered) was available only for 2010 through 2021. Since GHGRP data are not available for 1990
34 through 2009, IPCC "splicing" techniques were used as per the 2006 IPCC Guidelines on time-series consistency
35 (IPCC 2006, Volume 1, Chapter 5).
36 Lime production data (i.e., lime sold and non-marketed lime used by the producer) by type (i.e., high-calcium and
37 dolomitic quicklime, high-calcium and dolomitic hydrated lime, and dead-burned dolomite) for 1990 through 2021
38 (see Table 4-8) were obtained from U.S. Geological Survey (USGS) Minerals Yearbook (USGS 1992 through 2022b)
39 and are compiled by USGS to the nearest ton. Dead-burned dolomite data are additionally rounded by USGS to no
40 more than one significant digit to avoid disclosing company proprietary data. Production data for the individual
41 quicklime (i.e., high-calcium and dolomitic) and hydrated lime (i.e., high-calcium and dolomitic) types were not
42 provided prior to 1997. These were calculated based on total quicklime and hydrated lime production data from
43 1990 through 1996 and the three-year average ratio of the individual lime types from 1997 to 1999. Natural
44 hydraulic lime, which is produced from CaO and hydraulic calcium silicates, is not manufactured in the United
45 States (USGS 2018a). Total lime production was adjusted to account for the water content of hydrated lime by
46 converting hydrate to oxide equivalent based on recommendations from the IPCC and using the water content
4-16 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 values for high-calcium hydrated lime and dolomitic hydrated lime mentioned above, and is presented in Table 4-9
2 (IPCC 2006). The CaO and CaOMgO contents of lime, both 95 percent, were obtained from the IPCC (IPCC 2006).
3 Table 4-8: High-Calcium- and Dolomitic-Quicklime, High-Calcium- and Dolomitic-Hydrated,
4 and Dead-Burned-Dolomite Lime Production (kt)
Year
1990
2005
2017
2018
2019
2020
2021
High-Calcium Quicklime
11,166
14,100
12,200
12,400
11,300
10,700
11,200
Dolomitic Quicklime
2,234
2,990
2,650
2,810
2,700
2,390
2,700
High-Calcium Hydrated
1,781
2,220
2,360
2,430
2,430
2,320
2,430
Dolomitic Hydrated
319
474
276
265
267
252
244
Dead-Burned Dolomite
342
200
200
200
200
200
200
5
6 Table 4-9: Adjusted Lime Production (kt)
Year 1990
2005
2017 2018 2019 2020 2021
High-Calcium 12,466
Dolomitic 2,800
15,721
3,522
13,923 14,174 13,074 12,394 12,974
3,043 3,196 3,087 2,766 3,071
Note: Minus water content of hydrated lime.
7 Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
8 through 2021.
9 Uncertainty - TO BE UPDATED FOR FINAL INVENTORY REPORT
10 The uncertainties contained in these estimates can be attributed to slight differences in the chemical composition
11 of lime products and CO2 recovery rates for on-site process use over the time series. Although the methodology
12 accounts for various formulations of lime, it does not account for the trace impurities found in lime, such as iron
13 oxide, alumina, and silica. Due to differences in the limestone used as a raw material, a rigid specification of lime
14 material is impossible. As a result, few plants produce lime with exactly the same properties.
15 In addition, a portion of the CO2 emitted during lime production will actually be reabsorbed when the lime is
16 consumed, especially at captive lime production facilities. As noted above, lime has many different chemical,
17 industrial, environmental, and construction applications. In many processes, CO2 reacts with the lime to create
18 calcium carbonate (e.g., water softening). Carbon dioxide reabsorption rates vary, however, depending on the
19 application. For example, 100 percent of the lime used to produce precipitated calcium carbonate reacts with CO2,
20 whereas most of the lime used in steel making reacts with impurities such as silica, sulfur, and aluminum
21 compounds. Quantifying the amount of CO2 that is reabsorbed would require a detailed accounting of lime use in
22 the United States and additional information about the associated processes where both the lime and byproduct
23 CO2 are "reused." Research conducted thus far has not yielded the necessary information to quantify CO2
24 reabsorption rates.14 Some additional information on the amount of CO2 consumed on site at lime facilities,
25 however, has been obtained from EPA's GHGRP.
26 In some cases, lime is generated from calcium carbonate byproducts at pulp mills and water treatment plants.15
27 The lime generated by these processes is included in the USGS data for commercial lime consumption. In the
14 Representatives of the National Lime Association estimate that C02 reabsorption that occurs from the use of lime may offset
as much as a quarter of the C02 emissions from calcination (Males 2003).
15 Some carbide producers may also regenerate lime from their calcium hydroxide byproducts, which does not result in
emissions of C02. In making calcium carbide, quicklime is mixed with coke and heated in electric furnaces. The regeneration of
lime in this process is done using a waste calcium hydroxide (hydrated lime) [CaC2 + 2H20 -> C2H2 + Ca(OH) 2], not calcium
carbonate [CaCOs]. Thus, the calcium hydroxide is heated in the kiln to simply expel the water [Ca(OH)2 + heat -> CaO + H20],
and no C02 is released.
Industrial Processes and Product Use 4-17
-------
1 pulping industry, mostly using the Kraft (sulfate) pulping process, lime is consumed in order to causticize a process
2 liquor (green liquor) composed of sodium carbonate and sodium sulfide. The green liquor results from the dilution
3 of the smelt created by combustion of the black liquor where biogenic carbon (C) is present from the wood. Kraft
4 mills recover the calcium carbonate "mud" after the causticizing operation and calcine it back into lime—thereby
5 generating CO —for reuse in the pulping process. Although this re-generation of lime could be considered a lime
6 manufacturing process, the CO emitted during this process is mostly biogenic in origin and therefore is not
7 included in the industrial processes totals (Miner and Upton 2002). In accordance with IPCC methodological
8 guidelines, any such emissions are calculated by accounting for net C fluxes from changes in biogenic C reservoirs
9 in wooded or crop lands (see the Land Use, Land-Use Change, and Forestry chapter).
10 In the case of water treatment plants, lime is used in the softening process. Some large water treatment plants
11 may recover their waste calcium carbonate and calcine it into quicklime for reuse in the softening process. Further
12 research is necessary to determine the degree to which lime recycling is practiced by water treatment plants in the
13 United States.
14 Another uncertainty is the assumption that calcination emissions for LKD are around 2 percent. The National Lime
15 Association (NLA) has commented that the estimates of emissions from LKD in the United States could be closer to
16 6 percent. They also note that additional emissions (approximately 2 percent) may also be generated through
17 production of other byproducts/wastes (off-spec lime that is not recycled, scrubber sludge) at lime plants (Seeger
18 2013). Publicly available data on LKD generation rates, total quantities not used in cement production, and types of
19 other byproducts/wastes produced at lime facilities are limited. NLA compiled and shared historical emissions
20 information and quantities for some waste products reported by member facilities associated with generation of
21 total calcined byproducts and LKD, as well as methodology and calculation worksheets that member facilities
22 complete when reporting. There is uncertainty regarding the availability of data across the time series needed to
23 generate a representative country-specific LKD factor. Uncertainty of the activity data is also a function of the
24 reliability and completeness of voluntarily reported plant-level production data. Further research, including
25 outreach and discussion with NLA, and data is needed to improve understanding of additional calcination
26 emissions to consider revising the current assumptions that are based on IPCC guidelines. More information can be
27 found in the Planned Improvements section below.
28 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-10. Lime CO emissions
29 for 2021 were estimated to be between 11.1 and 11.5 MMT CO Eq. at the 95 percent confidence level. This
30 confidence level indicates a range of approximately 2 percent below and 2 percent above the emission estimate of
31 11.9 MMT CO Eq.
32 Table 4-10: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Lime
33 Production (MMT CO2 Eq. and Percent)
Source
Gas
2021 Emission Estimate
(MMTCO. Eq.)
Uncertainty Range Relative to Emission Estimate-1
(MMTCO' Eq.) (%)
Lower Upper
Bound Bound
Lower
Bound
Upper
Bound
Lime Production
C02
11.9
11.1 11.5
+2%
Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
34 QA/QC and Verification
35 General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S. Inventory
36 QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006 IPCC Guidelines as noted in the introduction
37 of the IPPU chapter (see Annex 8 for more details).
38 More details on the greenhouse gas calculation, monitoring and QA/QC methods associated with reporting on CO2
39 captured for onsite use applicable to lime manufacturing facilities can be found under Subpart S (Lime
4-18 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Manufacturing) of the GHGRP regulation (40 CFR Part 98).16 EPA verifies annual facility-level GHGRP reports
through a multi-step process (e.g., combination of electronic checks and manual reviews) to identify potential
errors and ensure that data submitted to EPA are accurate, complete, and consistent (EPA 2022).17 Based on the
results of the verification process, EPA follows up with facilities to resolve mistakes that may have occurred. The
post-submittals checks are consistent with a number of general and category-specific QC procedures, including:
range checks, statistical checks, algorithm checks, and year-to-year checks of reported data and emissions.
Recalculations Discussion
No recalculations were performed for the 1990 through 2020 portion of the time series.
Planned Improvements
EPA plans to review GHGRP emissions and activity data reported to EPA under Subpart S of the GHGRP regulation
(40 CFR Part 98), and aggregated activity data on lime production by type in particular. In addition, initial review of
data has identified that several facilities use CEMS to report emissions. Under Subpart S, if a facility is using a
CEMS, they are required to report combined combustion emissions and process emissions. EPA continues to
review how best to incorporate GHGRP and notes that particular attention will be made to also ensuring time-
series consistency of the emissions estimates presented in future Inventory reports, consistent with IPCC and
UNFCCC guidelines. This is required because the facility-level reporting data from EPA's GHGRP, with the program's
initial requirements for reporting of emissions in calendar year 2010, are not available for all inventory years (i.e.,
1990 through 2009) as required for this Inventory. In implementing improvements and integration of data from
EPA's GHGRP, the latest guidance from the IPCC on the use of facility-level data in national inventories will be
relied upon.18
Future improvements involve improving and/or confirming the representativeness of current assumptions
associated with emissions from production of LKD and other byproducts/wastes as discussed in the Uncertainty
section, per comments from the NLA provided during a prior Public Review comment period for a previous
Inventory (i.e., 1990 through 2018). EPA met with NLA in summer of 2020 for clarification on data needs and
available data and to discuss planned research into GHGRP data. Previously, EPA met with NLA in spring of 2015 to
outline specific information required to apply IPCC methods to develop a country-specific correction factor to
more accurately estimate emissions from production of LKD. In 2016, NLA compiled and shared historical
emissions information reported by member facilities on an annual basis under voluntary reporting initiatives from
2002 through 2011 associated with generation of total calcined byproducts and LKD. Reporting of LKD was only
differentiated for the years 2010 and 2011. This emissions information was reported on a voluntary basis
consistent with NLA's facility-level reporting protocol, which was also provided to EPA. To reflect information
provided by NLA, EPA updated the qualitative description of uncertainty. At the time of this Inventory, this planned
improvement is in process and has not been incorporated into this current Inventory report.
16 See http://www.ecfr.gov/cgi-bin/text-idx7tpN/ecfrbrowse/Title40/40cfr98 main Q2.tpl.
17 See https://www.epa.gov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
18 See http://www.ipcc-nggip.iges.or.ip/public/tb/TFI Technical Bulletin l.pdf.
Industrial Processes and Product Use 4-19
-------
1
2
4.3 Glass Production (CRF Source Category
2A3)
3 Glass production is an energy and raw-material intensive process that results in the generation of carbon dioxide
4 (CO2) from both the energy consumed in making glass and the glass production process itself. Emissions from fuels
5 consumed for energy purposes during the production of glass are included in the Energy sector.
6 Glass production employs a variety of raw materials in a glass-batch. These include formers, fluxes, stabilizers, and
7 sometimes colorants. The major raw materials (i.e., fluxes and stabilizers) that emit process-related CO2 emissions
8 during the glass melting process are limestone, dolomite, and soda ash. The main former in all types of glass is
9 silica (SiCh). Other major formers in glass include feldspar and boric acid (i.e., borax). Fluxes are added to lower the
10 temperature at which the batch melts. Most commonly used flux materials are soda ash (sodium carbonate,
11 Na2CC>3) and potash (potassium carbonate, K2O). Stabilizers make glass more chemically stable and keep the
12 finished glass from dissolving and/or falling apart. Commonly used stabilizing agents in glass production are
13 limestone (CaCCh), dolomite (CaCOsMgCOs), alumina (AI2O3), magnesia (MgO), barium carbonate (BaCOs),
14 strontium carbonate (SrCOs), lithium carbonate (IJ2CO3), and zirconia (ZrCh) (DOE 2002). Glass makers also use a
15 certain amount of recycled scrap glass (cullet), which comes from in-house return of glassware broken in the
16 production process or other glass spillage or retention, such as recycling or from cullet broker services.
17 The raw materials (primarily soda ash, limestone, and dolomite) release CO2 emissions in a complex high-
18 temperature chemical reaction during the glass melting process. This process is not directly comparable to the
19 calcination process used in lime manufacturing, cement manufacturing, and process uses of carbonates (i.e.,
20 limestone/dolomite use) but has the same net effect in terms of generating process CO2 emissions (IPCC 2006).
21 The U.S. glass industry can be divided into four main categories: containers, flat (window) glass, fiber glass, and
22 specialty glass. The majority of commercial glass produced is container and flat glass (EPA 2009). The United States
23 is one of the major global exporters of glass. Domestically, demand comes mainly from the construction, auto,
24 bottling, and container industries. There are more than 1,700 facilities that manufacture glass in the United States,
25 with the largest companies being Corning, Guardian Industries, Owens-Illinois, and PPG Industries.19
26 The glass container sector is one of the leading soda ash consuming sectors in the United States. In 2021, glass
27 production accounted for 48 percent of total domestic soda ash consumption (USGS 2022). Emissions from soda
28 ash production are reported in 4.12 Soda Ash Production (CRF Source Category 2B7).
29 In 2021, 2,280 kilotons of soda ash, 1,397 kilotons of limestone, 893 kilotons of dolomite, and 2 kilotons of other
30 carbonates were consumed for glass production (USGS 2022; EPA 2022). Use of soda ash, limestone, dolomite, and
31 other carbonates in glass production resulted in aggregate CO2 emissions of 2.0 MMT CO2 Eq. (1,969 kt) (see Table
32 4-11). Overall, emissions have decreased by 13 percent compared to 1990. Emissions increased by 6 percent
33 compared to 2020 levels.
34 Emissions from glass production have remained relatively consistent over the time series with some fluctuations
35 since 1990. In general, these fluctuations were related to the behavior of the export market and the U.S. economy.
36 Specifically, the extended downturn in residential and commercial construction and automotive industries
37 between 2008 and 2010 resulted in reduced consumption of glass products, causing a drop in global demand for
38 limestone/dolomite and soda ash and resulting in lower emissions. Some commercial food and beverage package
39 manufacturers are shifting from glass containers towards lighter and more cost-effective polyethylene
40 terephthalate (PET) based containers, putting downward pressure on domestic consumption of soda ash (USGS
19 Excerpt from Glass & Glass Product Manufacturing Industry Profile, First Research. Available online at:
http://www.firstresearch.com/lndustry-Research/Glass-and-Glass-Product-Manufacturing.html.
4-20 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 1995 through 2015b). Glass production in 2021 was steady, changing by no more than 5 percent over the course of
2 the year (Federal Reserve 2022).
3 Table 4-11: CO2 Emissions from Glass Production (MMT CO2 Eq. and kt)
Year
1990
2005
2017
2018
2019
2020
2021
MMT CO? Eq.
2.3
2.4
2.0
2.0
1.9
1.9
2.0
kt
2,262
2,401
1,984
1,989
1,940
1,858
1,969
4
5 Methodology and Time-Series Consistency
6 Carbon dioxide emissions were calculated based on the 2006IPCC Guidelines Tier 3 method by multiplying the
7 quantity of input carbonates (limestone, dolomite, soda ash, and other carbonates) by the carbonate-based
8 emission factor (in metric tons CC>2/metric ton carbonate) and the average carbonate-based mineral mass fraction.
9 2010 through 2021
10 For this Inventory, the methodology for estimating CO2 emissions from glass production for years 2010 through
11 2021 has added new activity data reported to the U.S. EPA Greenhouse Gas Reporting Program (GHGRP) on the
12 quantities of a group of other carbonates (i.e., barium carbonate, potassium carbonate, lithium carbonate, and
13 strontium carbonate) used for glass production (EPA 2022). The methodology continues to use the quantities of
14 limestone and dolomite used for glass production obtained from GHGRP (EPA 2022). USGS data on the quantity of
15 soda ash used for glass production continues to be used because it was obtained directly from the soda ash
16 producers and includes use by smaller artisanal glass operations, which are excluded in the GHGRP data.
17 GHGRP collects data from glass production facilities with greenhouse gas emissions greater than 25,000 metric
18 tons CO2 Eq. The reporting threshold is used to exclude artisanal glass operations that are expected to have much
19 lower greenhouse gas emissions than the threshold. These smaller facilities have not been accounted for yet for
20 this portion of the time series for limestone, dolomite, or other carbonates due to limited data. Facilities report the
21 total quantity of each type of carbonate used in glass production each year to GHGRP, with data collection starting
22 in 2010 (EPA 2022).
23 Using the total quantities of each carbonate, EPA calculated the metric tons of emissions resulting from glass
24 production by multiplying the quantity of input carbonates (i.e., limestone, dolomite, soda ash, and other
25 carbonates) by carbonate-based emission factors in metric tons CCh/metric ton carbonate (limestone, 0.43971;
26 dolomite, 0.47732; soda ash, 0.41492; and other carbonates, 0.262), and by the average carbonate-based mineral
27 mass fraction for each year. IPCC default values were used for limestone, dolomite, and soda ash emission factors,
28 and the emission factor for other carbonates is based on expert judgment (Icenhour 2022). The average carbonate-
29 based mineral mass fractions from the GHGRP, averaged across 2010 through 2015, indicate that soda ash
30 contained 98.7 percent sodium carbonate (Na2COs). This averaged value is used to estimate emissions for 1990
31 through 2009, described below. The previous methodology assumed that soda ash contained 100 percent sodium
32 carbonate (Na2COs).
33 1990 through 2009
34 Data from GHGRP on the quantity of limestone, dolomite, and other carbonates used in glass production are not
35 available for 1990 through 2009. Additionally, USGS does not collect data on the quantity of other carbonates used
36 for glass production.
37 To address time-series consistency, total emissions from 1990 to 2009 were calculated using the Federal Reserve
38 Industrial Production Index for glass production in the United States as a surrogate for the total quantity of
39 carbonates used in glass production. The production index measures real output expressed as a percentage of real
40 output in a base year, which is currently 2017 (Federal Reserve 2021). Since January 1971, the Federal Reserve has
41 released the monthly glass production index for NAICS code 3272 (Glass and Glass Product Manufacturing) as part
Industrial Processes and Product Use 4-21
-------
1 of release G.17, "Industrial Production and Capacity Utilization" (Federal Reserve 2022). The monthly index values
2 for each year were averaged to calculate an average annual glass production index value. Total annual process
3 emissions were calculated by taking a ratio of the average annual glass production index for each year to the
4 average annual glass production index for base year 2017, and multiplying by the calculated 2017 emissions
5 (process-related) based on GHGRP data.
6 Emissions from limestone, dolomite, and other carbonate consumption were disaggregated from total annual
7 emissions, using the average percent contribution of each to annual emissions from these three carbonates for
8 2010 through 2015 based on GHGRP data: 64.3 percent limestone, 35.6 percent dolomite, and 0.1 percent other
9 carbonates.
10 The methodology for estimating CO2 emissions from the use of soda ash for glass production and data sources for
11 the amount of soda ash used in glass production are consistent with the methodology used for 2010 through 2021.
12 Because data on the average mineral mass fraction for soda ash is only available starting in 2010, the values for
13 2010 through 2015 are averaged, as described above, and used to calculate emissions for 1990 to 2009.
14 Data on soda ash used for glass production for 1990 through 2021 were obtained from the U.S. Bureau of Mines
15 (1991 and 1993a), the USGS Minerals Yearbook: Soda Ash (USGS 1995 through 2015b), and USGS Mineral Industry
16 Surveys for Soda Ash (USGS 2017 through 2021). Data on limestone, dolomite, and other carbonates used for glass
17 production and on average carbonate-based mineral mass fraction for 2010 through 2021 were obtained from
18 GHGRP (EPA 2022). The quantities of limestone, dolomite, and other carbonates were calculated for 1990 through
19 2009 using the Federal Reserve Industrial Production Index (Federal Reserve 2022).
20 The amount of limestone, dolomite, soda ash, and other carbonates used in glass production each year and the
21 annual average Federal Reserve production indices for glass production are shown in Table 4-12.
22 Table 4-12: Limestone, Dolomite, Soda Ash, and Other Carbonates Used in Glass Production
23 (kt) and Average Annual Production Index for Glass and Glass Product Manufacturing
Activity
1990
2005
2017
2018
2019
2020
2021
Limestone
1,405
1,686
1,488
1,442
1,370
1,334
1,397
Dolomite
718
861
806
871
883
824
893
Soda Ash
3,177
3,050
2,360
2,280
2,220
2,130
2,280
Other Carbonates
2
3
2
2
2
2
2
Total
5,302
5,599
4,656
4,596
4,475
4,289
4,572
Production Index3
94.3
113.1
100
102.5
99.8
93.2
93.7
a Average Annual Production Index uses 2017 as the base year.
Note: Totals may not sum due to independent rounding.
24 As discussed above, methodological approaches were applied to the entire time series to ensure consistency in
25 emissions from 1990 through 2021. Consistent with the 2006IPCC Guidelines, the overlap technique was applied
26 to compare USGS and GHGRP data sets for 2010 through 2021. To address the inconsistencies, adjustments were
27 made as described above.
28 Uncertainty - TO BE UPDATED FOR FINAL INVENTORY REPORT
29 The methodology and activity data used in this Inventory reduced uncertainty for glass production, compared to
30 the previous Inventory. Uncertainty levels presented in this section in previous Inventories arose in part due to
31 variations in the chemical composition of limestone used in glass production. For example in addition to calcium
32 carbonate, limestone may contain smaller amounts of magnesia, silica, and sulfur, among other minerals (e.g.,
33 potassium carbonate, strontium carbonate and barium carbonate, and dead burned dolomite). The methodology
34 in this Inventory report uses GHGRP data on the average mass fraction of each mineral in the limestone and
35 dolomite used in glass production for each year from 2010-2020.
36 The data and methodology used in this Inventory report also reduce uncertainty associated with activity data. The
37 methodology uses the amount of limestone and dolomite used in glass manufacturing which is reported directly by
4-22 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 the glass manufacturers for years 2010 through 2020 and the amount of soda ash used in glass manufacturing
2 which is reported by soda ash producers for the full time series. The emissions from other carbonates reported to
3 GHGRP-barium carbonate (BaCOs), potassium carbonate (K2CO3), lithium carbonate (IJ2CO3), and strontium
4 carbonate (SrCOs)-are not included in these estimates.
5 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-13. In 2020, glass
6 production CO2 emissions were estimated to be between 1.8 and 1.9 MMT CO2 Eq. at the 95 percent confidence
7 level. This indicates a range of approximately 2 percent below and 2 percent above the emission estimate of 1.9
8 MMT CO2 Eq.
9 Table 4-13: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Glass
10 Production (MMT CO2 Eq. and Percent)
Source Gas
2021 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.)
(MMT C02 Eq.)
(%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Glass Production C02
1.9
1.8
1.9
-2%
+2%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence
interval.
11 QA/QC and Verification
12 General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S. Inventory
13 QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006IPCC Guidelines as described in the
14 introduction of the IPPU chapter (see Annex 8 for more details). For the GHGRP data, EPA verifies annual facility-
15 level reports through a multi-step process (e.g., including a combination of pre-and post-submittal electronic
16 checks and manual reviews by staff) to identify potential errors and ensure that data submitted to EPA are
17 accurate, complete, and consistent (EPA 2015).20 Based on the results of the verification process, EPA follows up
18 with facilities to resolve mistakes that may have occurred. The post-submittals checks are consistent with a
19 number of general and category-specific QC procedures, including: range checks, statistical checks, algorithm
20 checks, and year-to-year checks of reported data and emissions.
21 Recalculations Discussion
22 For the current Inventory, refinements to the methodology were implemented, using more complete activity data
23 from GHGRP for 2010 through 2021 and the industrial production index for glass and glass product manufacturing
24 from the Federal Reserve for 1990 through 2009 to address time-series consistency. These refinements are
25 described under the Methodology and Time-Series Consistency section. The revised values for 1990 through 2020
26 resulted in decreased emissions estimates prior to 2018 and slight increases for 2019 and 2020. Across the time
27 series, emissions decreased by an average of 1.0 percent compared to the previous Inventory. Annual emission
28 changes during the time series ranged from a 0.1 percent increase in 2019 and 2020 (1 kt CO2) to a 1.4 percent
29 decrease in 1999 (27 kt CO2).
30 Planned Improvements
31 EPA plans to evaluate updates to uncertainty levels for the activity data and mineral mass fraction values from
32 EPA's GHGRP. This is a near-term planned improvement.
20 GHGRP Report Verification Factsheet. See https://www.epa.gov/sites/production/files/2015-
07/documents/ghgrp verification factsheet.pdf.
Industrial Processes and Product Use 4-23
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Some glass producing facilities in the United States do not report to EPA's GHGRP because they fall below the
reporting threshold for this industry. EPA will continue ongoing research on the availability of data to better assess
the completeness of emission estimates from glass production and how to refine the methodology to ensure
complete national coverage of this category. When reporting began in 2010, EPA received data from more facilities
that were above the reporting threshold than expected, and total emissions were higher than expected for all glass
production facilities in the United States (EPA 2009). Research will include reassessing previous assessments of
GHGRP industry coverage using the reporting threshold of 25,000 metric tons CO2 Eq. This is a medium-term
planned improvement.
4.4 Other Process Uses of Carbonates (CRF
Source Category 2A4)
Limestone (CaCOs), dolomite (CaCOsMgCOs),21 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, glass
production, and environmental pollution control. This section addresses only limestone, dolomite, and soda ash use.
For industrial applications, carbonates such as limestone and dolomite are heated sufficiently enough to calcine the
material and generate CO2 as a byproduct.
CaCO3 —> CaO + C02
MgC03 —> MgO + C02
Examples of such applications include limestone used as a flux or purifier in metallurgical furnaces, as a sorbent in
flue gas desulfurization (FGD) systems for utility and industrial plants, and as a raw material for the production of
glass, lime, and cement. Emissions from limestone and dolomite used in the production of cement, lime, glass, and
iron and steel are excluded from the Other Process Uses of Carbonates category and reported under their respective
source categories (e.g., Section 4.3, Glass Production). Emissions from soda ash production are reported under
Section 4.12, Soda Ash Production (CRF Source Category 2B7). Emissions from soda ash consumption associated
with glass manufacturing are reported under Section 4.3, Glass Production (CRF Source Category 2A3). Emissions
from the use of limestone and dolomite in liming of agricultural soils are included in the Agriculture chapter under
Section 5.5, Liming (CRF Source Category 3G). Emissions from fuels consumed for energy purposes during these
processes are accounted for in the Energy chapter under Section 3.1, Fossil Fuel Combustion (CRF Source Category
1A). Both lime (CaO) and limestone (CaCOs) can be used as a sorbent for FGD systems. Emissions from lime
consumption for FGD systems and from sugar refining are reported under Section 4.3 Lime Production (CRF Source
Category 2A2). Emissions from the use of dolomite in primary magnesium metal production are reported under
Section 4.20, Magnesium Production and Processing (CRF Source Category 2C4).
Limestone and dolomite are widely distributed throughout the world in deposits of varying sizes and degrees of
purity. Large deposits of limestone occur in nearly every state in the United States, and significant quantities are
extracted for industrial applications. In 2018, the leading limestone producing states were Texas, Florida, Ohio,
Missouri, and Pennsylvania, which contributed 46 percent of the total U.S. output (USGS 2022a). Dolomite deposits
are found in the United States, Canada, Mexico, Europe, Africa, and Brazil. In the United States, the leading dolomite
producing states are Pennsylvania, New York, and Utah which currently contribute more than a third of the total
U.S. output (USGS 2022a). Internationally, two types of soda ash are produced: natural and synthetic. In 2019, 93
percent of the global soda ash production came from China, the United States, Russia, Germany, India, Turkey,
Poland, and France. The United States only produces natural soda ash and only in two states: Wyoming and
California (USGS 2021c).
21 Limestone and dolomite are collectively referred to as limestone by the industry, and intermediate varieties are seldom
distinguished.
4-24 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 In 2021,12,789 kilotons (kt) of limestone, 2,826 kt of dolomite, and 2,360 kt of soda ash were consumed for these
2 emissive applications, which excludes consumption for the production of cement, lime, glass, and iron and steel
3 (Willett 2022, USGS 2022b). Usage of limestone, dolomite and soda ash resulted in aggregate CO2 emissions of 8.0
4 MMT CO2 Eq. (7,968 kt) (see Table 4-14 and Table 4-15). The 2021 emissions decreased 5 percent compared to
5 2020, primarily as a result of decreased limestone consumption attributed to flux stone. Growth in the public and
6 private construction markets contributed to an increase in consumption of crushed stone in 2021. Overall
7 emissions have increased 29 percent from 1990 through 2021.
8 Table 4-14: CO2 Emissions from Other Process Uses of Carbonates (MMT CO2 Eq.)
Year
1990
2005
2017
2018
2019
2020
2021
Flux Stone
2.6
2.6
2.4
2.8
2.9
3.4
2.8
FGD
1.4
3.0
5.6
2.2
3.2
3.0
3.1
Soda Ash Consumption3
1.4
1.3
1.1
1.1
1.0
1.0
1.0
Other Miscellaneous Usesb
0.8
0.5
0.8
1.3
1.2
1.0
1.0
Total
6.2
7.5
9.9
7.4
8.4
8.4
8.0
a Soda ash consumption not associated with glass manufacturing.
b "Other miscellaneous uses" include chemical stone, mine dusting or acid water treatment,
and acid neutralization.
Note: Totals may not sum due to independent rounding.
9
10 Table 4-15: CO2 Emissions from Other Process Uses of Carbonates (kt)
Year
1990
2005
2017
2018
2019
2020
2021
Flux Stone
2,592
2,649
2,441
2,795
2,936
3,450
2,799
FGD
1,432
2,973
5,598
2,229
3,202
2,997
3,135
Soda Ash
Consumption3
1,390
1,305
1,058
1,069
1,036
958
979
Other
Miscellaneous
Usesb
819
533
771
1,259
1,248
994
1,038
Total
6,233
7,459
9,869
7,351
8,422
8,399
7,951
a Soda ash consumption not associated with glass manufacturing.
b "Other miscellaneous uses" include chemical stone, mine dusting or acid water treatment, and acid
neutralization.
Note: Totals may not sum due to independent rounding.
11 Methodology and Time-Series Consistency
12 Carbon dioxide emissions were calculated based on the 2006IPCC Guidelines Tier 2 method by multiplying the
13 quantity of limestone or dolomite consumed by the emission factor for limestone or dolomite calcination,
14 respectively: 0.43971 metric ton CCh/metric ton carbonate for limestone and 0.47732 metric ton CCh/metric ton
15 carbonate for dolomite.22 This methodology was used for flux stone, flue gas desulfurization systems, chemical
16 stone, mine dusting or acid water treatment, and acid neutralization. Flux stone used during the production of iron
17 and steel was deducted from the Other Process Uses of Carbonates source category estimate and attributed to the
18 Iron and Steel Production source category estimate. Similarly, limestone and dolomite consumption for glass
19 manufacturing, cement, and lime manufacturing are excluded from this category and attributed to their respective
20 categories.
22 2006 IPCC Guidelines, Volume 3: Chapter 2, Table 2.1.
Industrial Processes and Product Use 4-25
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Consumption data for 1990 through 2021 of limestone and dolomite used for flux stone, flue gas desulfurization
systems, chemical stone, mine dusting or acid water treatment, and acid neutralization (see Table 4-16) were
obtained from the U.S. Geological Survey (USGS) Minerals Yearbook: Crushed Stone Annual Report (1995a through
2022), preliminary data for 2021 from USGS Crushed Stone Commodity Expert (Willett 2022), American Iron and
Steel Institute limestone and dolomite consumption data (AISI 2018 through 2021), and the U.S. Bureau of Mines
(1991 and 1993a), which are reported to the nearest ton. In addition, the estimated values for limestone and
dolomite consumption for flux stone used during the production of iron and steel were adjusted using emissions
data from the EPA's Greenhouse Gas Reporting Program (GHGRP) subpart Qfor the iron and steel sector to
account for the impacts of the COVID-19 pandemic in 2020 and 2021. Iron and steel GHGRP process emissions data
increased by approximately 12 percent from 2020 to 2021 (EPA 2022). This adjustment method is consistent with
the method used in Section 4.17 Iron and Steel Production (CRF Source Category 2C1) and Metallurgical Coke
Production.
During 1990 and 1992, the USGS did not conduct a detailed survey of limestone and dolomite consumption by
end-use; therefore, data on consumption by end use for 1990 was estimated by applying the 1991 ratios of total
limestone and dolomite consumption by end use 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 end uses to the 1992 total values.
In 1991, the U.S. Bureau of Mines, now known as the USGS, began compiling production and end use information
through surveys of crushed stone manufacturers. Manufacturers provided different levels of detail in survey
responses, so information was divided into three categories: (1) production by end-use, as reported by
manufacturers (i.e., "specified" production); (2) production reported by manufacturers without end-uses specified
(i.e., "unspecified-reported" production); and (3) estimated additional production by manufacturers who did not
respond to the survey (i.e., "unspecified-estimated" production). Additionally, each year the USGS withholds data
on certain limestone and dolomite end-uses due to confidentiality agreements regarding company proprietary
data. For the purposes of this analysis, emissive end-uses that contained withheld data were estimated using one
of the following techniques: (1) the value for all the withheld data points for limestone or dolomite use was
distributed evenly to all withheld end-uses; (2) the average percent of total limestone or dolomite for the withheld
end-use in the preceding and succeeding years; or (3) the average fraction of total limestone or dolomite for the
end-use over the entire time period.
A large quantity of crushed stone was reported to the USGS under the category "unspecified uses." A portion of
this consumption is believed to be limestone or dolomite used for emissive end uses. The quantity listed for
"unspecified uses" was, therefore, allocated to all other reported end-uses according to each end-use's fraction of
total consumption in that year.23
Table 4-16: Limestone and Dolomite Consumption (kt)
Activity
1990
2005
2017
2018
2019
2020
2021
Flux Stone
5,842
5,745
5,447
6,242
6,551
7,592
6,124
Limestone
5,237
2,492
4,216
4,891
5,088
4,361
3,299
Dolomite
605
3,254
1,230
1,351
1,463
2,961
2,826
FGD
3,258
6,761
12,732
5,068
7,282
6,817
7,129
Other Miscellaneous Uses
1,835
1,212
1,754
2,862
2,834
2,260
2,361
Total
10,935
13,719
19,932
14,172
16,667
16,669
15,615
Note: Totals may not sum due to independent rounding.
Excluding glass manufacturing which is reported under Section 4.3 Glass Production (CRF Source Category 2A3),
most soda ash is consumed in chemical production, with minor amounts used in soap production, pulp and paper,
flue gas desulfurization, and water treatment. As soda ash is consumed for these purposes, CO2 is usually emitted.
In these applications, it is assumed that one mole of carbon is released for every mole of soda ash used. Thus,
23 This approach was recommended by USGS, the data collection agency.
4-26 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 approximately 0.113 metric tons of carbon (or 0.415 metric tons of CO2) are released for every metric ton of soda
2 ash consumed. The activity data for soda ash consumption for 1990 to 2021 (see Table 4-17) were obtained from
3 the U.S. Geological Survey (USGS) Minerals Yearbook for Soda Ash (1994 through 2015b) and USGS Mineral
4 Industry Surveys for Soda Ash (USGS 2017a, 2018, 2019, 2020b, 2021d, 2022b). Soda ash consumption data were
5 collected by the USGS from voluntary surveys of the U.S. soda ash industry.
6 Table 4-17: Soda Ash Consumption Not Associated with Glass Manufacturing (kt)
Activity
1990
2005 ¦
2017
2018
2019
2020
2021
Soda Asha
3,351
3,1441
2,550
2,576
2,497
2,310
2,360
a Soda ash consumption is sales reported by producers which exclude imports. Historically, imported soda ash is less
than 1 percent of the total U.S. consumption (Kostick 2012).
7 Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
8 through 2021.
9 Uncertainty - TO BE UPDATED FOR FINAL INVENTORY REPORT
10 The uncertainty levels presented in this section account for uncertainty associated with activity data. Data on
11 limestone and dolomite consumption are collected by USGS through voluntary national surveys. USGS contacts the
12 mines (i.e., producers of various types of crushed stone) for annual sales data. Data on other carbonate
13 consumption are not readily available. The producers report the annual quantity sold to various end-users and
14 industry types. USGS estimates the historical response rate for the crushed stone survey to be approximately 70
15 percent, and the rest is estimated by USGS. Large fluctuations in reported consumption exist, reflecting year-to-
16 year changes in the number of survey responders. The uncertainty resulting from a shifting survey population is
17 exacerbated by the gaps in the time series of reports. The accuracy of distribution by end use is also uncertain
18 because this value is reported by the producer/mines and not the end user. Additionally, there is significant
19 inherent uncertainty associated with estimating withheld data points for specific end uses of limestone and
20 dolomite. Lastly, much of the limestone consumed in the United States is reported as "other unspecified uses;"
21 therefore, it is difficult to accurately allocate this unspecified quantity to the correct end-uses. EPA contacted the
22 USGS National Minerals Information Center Crushed Stone commodity expert to assess the current uncertainty
23 ranges associated with the limestone and dolomite consumption data compiled and published by USGS. During
24 this discussion, the expert confirmed that EPA's range of uncertainty was still reasonable (Willett 2017).
25 Uncertainty in the estimates also arises in part due to variations in the chemical composition of limestone. In
26 addition to calcium carbonate, limestone may contain smaller amounts of magnesia, silica, and sulfur, among
27 other minerals. The exact specifications for limestone or dolomite used as flux stone vary with the
28 pyrometallurgical process and the kind of ore processed.
29 For emissions from soda ash consumption, the primary source of uncertainty results from the fact that these
30 emissions are dependent upon the type of processing employed by each end-use. Specific emission factors for
31 each end-use are not available, so a Tier 1 default emission factor is used for all end-uses. Therefore, there is
32 uncertainty surrounding the emission factors from the consumption of soda ash. Additional uncertainty comes
33 from the reported consumption and allocation of consumption within sectors that is collected on a quarterly basis
34 by the USGS. Efforts have been made to categorize company sales within the correct end-use sector.
35 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-18. Carbon dioxide
36 emissions from other process uses of carbonates in 2021 were estimated to be between 8.2 and 12.9 MMT CO2 Eq.
37 at the 95 percent confidence level. This indicates a range of approximately 19 percent below and 28 percent above
38 the emission estimate of 8.0 MMT CO2 Eq.
Industrial Processes and Product Use 4-27
-------
1 Table 4-18: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Other
2 Process Uses of Carbonates (MMT CO2 Eq. and Percent)
2021 Emission
Source Gas Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.)
(MMT C02 Eq.) (%)
Lower Upper Lower Upper
Bound Bound Bound Bound
Other Process Uses
t C02 8.0
of Carbonates
8.2 12.9 -19% +28%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence
interval.
3 QA/QC and Verification
4 General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S. Inventory
5 QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006IPCC Guidelines as described in the
6 introduction of the IPPU chapter (see Annex 8 for more details).
7 Recalculations Discussion
8 For the current Inventory, updated USGS data on limestone and dolomite consumption was available for 2019 and
9 2020, resulting in updated emissions estimates for those years. Compared to the previous Inventory, emissions for
10 2019 decreased by 14.7 percent (1,449 kt CO2 Eq.) and emissions for 2020 decreased by 18.8 percent (1,843 kt CO2
11 Eq.).
12 Planned Improvements
13 In response to comments received during previous Inventory reports from the UNFCCC, EPA has inquired to the
14 availability of ceramics and non-metallurgical magnesia data. EPA is assessing potential activity data from USGS
15 that spans the full time series for ceramics production . Data on non-metallurgical magnesia is not currently
16 reported by survey respondents to USGS, and EPA continues to conduct outreach with other entities. This
17 improvement remains ongoing, and EPA plans to continue to update this Planned Improvements section in future
18 reports as more information becomes available.
19 EPA also plans to review the uncertainty ranges assigned to activity data. This planned improvement is currently
20 planned as a medium-term improvement.
21 4.5 Ammonia Production (CRF Source
22 Category 2B1)
23 Emissions of carbon dioxide (CO2) occur during the production of synthetic ammonia (NH3), primarily through the
24 use of natural gas, petroleum coke, or naphtha as a feedstock. The natural gas-, naphtha-, and petroleum coke-
25 based processes produce CO2 and hydrogen (H2), the latter of which is used in the production of ammonia. The
26 brine electrolysis process for production of ammonia does not lead to process-based CO2 emissions. Due to
27 national circumstances, emissions from fuels consumed for energy purposes during the production of ammonia
28 are accounted for in the Energy chapter. More information on this approach can be found in the Methodology
29 section below.
4-28 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
Ammonia production requires a source of nitrogen (N) and hydrogen (H). Nitrogen is obtained from air through
liquid air distillation or an oxidative process where air is burnt and the residual nitrogen is recovered. In the United
States, the majority of ammonia is produced using a natural gas feedstock as the hydrogen source. One synthetic
ammonia production plant located in Kansas is producing ammonia from petroleum coke feedstock. In some U.S.
plants, some of the CO2 produced by the process is captured and used to produce urea rather than being emitted
to the atmosphere. In 2021,16 companies operated 35 ammonia producing facilities in 16 states. Approximately
60 percent of domestic ammonia production capacity is concentrated in Louisiana, Oklahoma, and Texas (USGS
2022).
Synthetic ammonia production from natural gas feedstock consists of five principal process steps. The primary
reforming step converts methane (CH4) to CO2, carbon monoxide (CO), and hydrogen (H2) in the presence of a
catalyst. Only 30 to 40 percent of the CH4 feedstock to the primary reformer is converted to CO and CO2 in this
step of the process. The secondary reforming step converts the remaining CH4 feedstock to CO and CO2. In the shift
conversion step, the CO in the process gas from the secondary reforming step (representing approximately 15
percent of the process gas) is converted to CO2 in the presence of a catalyst, water, and air. Carbon dioxide is
removed from the process gas by the shift conversion process, and the H2 is combined with the nitrogen (N2) gas in
the process gas during the ammonia synthesis step to produce ammonia. The CO2 is included in a waste gas stream
with other process impurities and is absorbed by a scrubber solution. In regenerating the scrubber solution, CO2 is
released from the solution.
The conversion process for conventional steam reforming of CH4, including the primary and secondary reforming
and the shift conversion processes, is approximately as follows:
0.88C7/4 +1.26Air +1.24H20 0.88C02 + N2 +3H2
N2 +3 H2 ->2 NH3
To produce synthetic ammonia from petroleum coke, the petroleum coke is gasified and converted to CO2 and H2.
These gases are separated, and the H2 is used as a feedstock to the ammonia production process, where it is
reacted with N2 to form ammonia.
Not all of the CO2 produced during the production of ammonia is emitted directly to the atmosphere. Some of the
ammonia and some of the CO2 produced by the synthetic ammonia process are used as raw materials in the
production of urea [COfNFhh], which has a variety of agricultural and industrial applications.
The chemical reaction that produces urea is:
2nh3+ co2 -> NH2COONH4 -> CO(NH2)2 +h2o
Only the CO2 emitted directly to the atmosphere from the synthetic ammonia production process is accounted for
in determining emissions from ammonia production. The CO2 that is captured during the ammonia production
process and used to produce urea does not contribute to the CO2 emission estimates for ammonia production
presented in this section. Instead, CO2 emissions resulting from the consumption of urea are attributed to the urea
consumption or urea application source category (under the assumption that the carbon stored in the urea during
its manufacture is released into the environment during its consumption or application). Emissions of CO2 resulting
from agricultural applications of urea are accounted for in Section 5.6 Urea Fertilization (CRF Source Category 3H)
of the Agriculture chapter. Emissions of CO2 resulting from non-agricultural applications of urea (e.g., use as a
feedstock in chemical production processes) are accounted for in Section 4.6 Urea Consumption for Non-
Agricultural Purposes of this chapter.
Emissions from fuel used for energy at ammonia plants are accounted for in the Energy chapter. The consumption
of natural gas and petroleum coke as fossil fuel feedstocks for NH3 production are adjusted for within the Energy
chapter as these fuels were consumed during non-energy related activities. More information on this methodology
is described in Annex 2.1, Methodology for Estimating Emissions of C02from Fossil Fuel Combustion.
Total emissions of CO2 from ammonia production in 2021 were 12.2 MMT CO2 Eq. (12,207 kt) and are summarized
in Table 4-19 and Table 4-20. Ammonia production relies on natural gas as both a feedstock and a fuel, and as
such, market fluctuations and volatility in natural gas prices affect the production of ammonia. Since 1990,
Industrial Processes and Product Use 4-29
-------
1 emissions from ammonia production have decreased by about 15 percent. Emissions in 2021 decreased by about 6
2 percent from the 2020 levels. One facility in Kansas produces ammonia from petroleum coke and began operations
3 in 2000. All other facilities use natural gas as feedstock.
4 Emissions from ammonia production increased steadily from 2015 to 2018, due to the addition of new ammonia
5 production facilities and new production units at existing facilities in 2016, 2017, and 2018. Agriculture continues
6 to drive demand for nitrogen fertilizers, accounting for approximately 88 percent of domestic ammonia
7 consumption.
8 Table 4-19: CO2 Emissions from Ammonia Production (MMT CO2 Eq.)
Gas 1990 2005 2017 2018 2019 2020 2021
CO; 14.4 10.2 A 12.5 12.7 12.4 13.0 12.2
9 Table 4-20: CO2 Emissions from Ammonia Production (kt)
Gas 1990 2005 2017 2018 2019 2020 2021
CO^ 14,404 10,234 12,481 12,669 12,401 13,006 12,207
10 Methodology and Time-Series Consistency
11 For this Inventory, the methodology for estimating CO2 emissions from the production of synthetic ammonia is a
12 country-specific approach consistent with the 2006IPCC Guidelines (IPCC 2006) and is based on Tier 3 methods.,
13 This Inventory report includes methodological refinements for 2010 to 2021 that directly use the process CO2
14 emissions reported to subpart G of the U.S. EPA Greenhouse Gas Reporting Program (GHGRP) (EPA 2022) and for
15 1990 to 2009 based on reported and calculated data on natural gas and petroleum coke feedstock used for
16 ammonia production.
17 Emissions from fuel used for energy at ammonia plants are accounted for in the Energy chapter. This differs from
18 the 2006 IPCC Guidance for ammonia which indicates that "in the case of ammonia production no distinction is
19 made between fuel and feedstock emissions with all emissions accounted for in the IPPU Sector;" however,
20 accurate data on fuel use for ammonia production is not known at this time. Data on total fuel use (including fuel
21 used for ammonia feedstock and fuel used for energy) for ammonia production are not known in the United
22 States. The Energy Information Administration (EIA), where energy use data is obtained for the Inventory (see the
23 Energy chapter), does not provide data broken out by industrial category; data is only available at the broad
24 industry sector level. Furthermore, the GHGRP data used in the analysis is based on feedstock use and not fuel use.
25 4.5.1.1 Natural Gas Feedstock
26 In 2017, facilities started reporting data to GHGRP on the quantity of natural gas feedstock used for ammonia
27 production and the carbon content of the natural gas feedstock (EPA 2022). Using these data and reported process
28 CO2 emissions, the average molecular weight of the feedstock and the average carbon content were derived for
29 years 2017 through 2021. The quantity of natural gas feedstock for 2010 to 2016 was then calculated using GHGRP
30 CO2 emissions for 2010 through 2016, average molecular weight of the feedstock for 2017 through 2021, and
31 average carbon content for 2017 through 2021.
32 To estimate natural gas feedstock use for 1990 to 2009, the ratio of natural gas feedstock quantity to ammonia
33 production quantity was calculated for each year and averaged over the years from 2010 to 2014, using the
34 calculated quantity of natural gas feedstock and total ammonia production for 2010 through 2014 (ACC 2021). The
35 years 2010 to 2014 were used to determine the average ratio of natural gas feedstock quantity to ammonia
36 production because that period was deemed to better represent historic ammonia production from 1990 to 2009.
4-30 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 This 2010 to 2014 average ratio was multiplied by total ammonia production for each year from 1990 to 2009 to
2 determine natural gas feedstock use.
3 CO2 emissions from the production of synthetic ammonia from natural gas feedstock for 1990 to 2009 were
4 estimated using the natural gas feedstock quantity as determined from above and the Inventory CO2 emissions
5 factor and heating content value for natural gas, consistent with values used in the Energy chapter. In terms of
6 reporting under GHGRP, 22 facilities reported from 2010 to 2012; 23 from 2013 to 2015; 26 in 2016; 28 in 2017
7 and 29 from 2018 to 2021, therefore, earlier years exclude the newer facilities that might not represent historic
8 information.
9 4.5.1.2 Petroleum Coke Feedstock
10 CO2 emissions from the production of synthetic ammonia from petroleum coke feedstock for 2000 to 2009 were
11 estimated by multiplying the following: quantity of petroleum coke feedstock reported by the facility (Coffeyville
12 2005, 2006, 2007a, 2007b, 2009, 2010, 2011, and 2012; CVR 2012 through 2021); the Inventory heating content
13 value for petroleum coke which is consistent with values used in the Energy chapter; and a stoichiometric CO2/C
14 factor of 44/12.
15 4.5.1.3 Urea Production Adjustments
16 Emissions of CO2 from ammonia production from both feedstocks and for all years from 1990 to 2021 were
17 adjusted to account for the use of some of the CO2 emissions from ammonia production as a raw material in the
18 production of urea. The CO2 emissions reported for ammonia production are reduced by a factor of 0.733, which
19 corresponds to a stoichiometric CCh/urea factor of 44/60, assuming complete conversion of ammonia (NH3) and
20 CO2 to urea (IPCC 2006; EFMA 2000), and multiplied by total annual domestic urea production.
21 All synthetic ammonia production and subsequent urea production are assumed to be from the same process—
22 conventional catalytic reforming of natural gas feedstock, with the exception of ammonia production from
23 petroleum coke feedstock at the one plant located in Kansas.
24 Data on facility-level process emissions for 2010 through 2021 on natural gas feedstock used and carbon content
25 of the natural gas feedstock starting in 2017 were obtained from GHGRP (EPA 2022). Total ammonia production
26 data for 2011 through 2021 were obtained from American Chemistry Council (ACC 2021). For years before 2011,
27 ammonia production data were obtained from the Census Bureau of the U.S. Department of Commerce (U.S.
28 Census Bureau 1991 through 1994,1998 through 2011) as reported in Current Industrial Reports Fertilizer
29 Materials and Related Products annual and quarterly reports. Natural gas and petroleum coke heating values come
30 from national-level data (EIA 2022), and natural gas and petroleum coke carbon contents are the same as used in
31 the Energy chapter calculations.
32 Data on urea production for 2010 through 2021 were obtained from GHGRP (EPA 2022). Urea production data for
33 2009 through 2010 were obtained from the U.S. Census Bureau (U.S. Census Bureau 2010 and 2011). Urea
34 production data for 1990 through 2008 were obtained from the USGS Minerals Yearbook: Nitrogen (USGS 1994-
35 2009). The U.S. Census Bureau ceased collection of urea production statistics in 2011. Total ammonia production,
36 total urea production, and recovered CO2 consumed for urea production are shown in Table 4-21.
37 Table 4-21: Total Ammonia Production, Total Urea Production, and RecoveredCOz Consumed
38 for Urea Production (kt)
Year
1990
2005
2017
2018
2019
2020
2021
Total Ammonia Production
Total Urea Production
15,425
7,450
10,143
5,270
14,070
9,030
16,010
10,700
16,410
11,400
17,020
11,500
15,420
10,500
Recovered C02 Consumed for
Urea Production
5,463
3,865
6,622
7,847
8,360
8,433
7,700
Industrial Processes and Product Use 4-31
-------
1 Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
2 through 2021. The methodology for ammonia production spliced activity data from different sources: U. S. Census
3 Bureau data for 1990 through 2010, ACC data beginning in 2011, and GHGRP data beginning in 2010 and
4 2017. Consistent with the 2006IPCC Guidelines, the overlap technique was applied to compare the two data sets
5 for years where there was overlap, with findings that the data sets were consistent and adjustments were not
6 needed.
7 Uncertainty - TO BE UPDATED FOR FINAL INVENTORY REPORT
8 The uncertainties presented in this section are primarily due to how accurately the emission factor used represents
9 an average across all ammonia plants using natural gas feedstock. Uncertainties are also associated with ammonia
10 production estimates and the assumption that all ammonia production and subsequent urea production was from
11 the same process—conventional catalytic reforming of natural gas feedstock, with the exception of one ammonia
12 production plant located in Kansas that is manufacturing ammonia from petroleum coke feedstock. Uncertainty is
13 also associated with the representativeness of the emission factor used for the petroleum coke-based ammonia
14 process. It is also assumed that ammonia and urea are produced at co-located plants from the same natural gas
15 raw material. The uncertainty of the total urea production activity data, based on USGS Minerals Yearbook:
16 Nitrogen data, is a function of the reliability of reported production data and is influenced by the completeness of
17 the survey responses. EPA assigned a default uncertainty range of ±5 percent for both ammonia production and
18 the emission factor used for the petroleum coke-based ammonia process, consistent with the ranges in Section
19 3.2.3.2 of the 2006 IPCC Guidelines, and ±10 percent for urea production, based on expert judgment.
20 Recovery of CO2 from ammonia production plants for purposes other than urea production (e.g., commercial sale,
21 etc.) has not been considered in estimating the CO2 emissions from ammonia production, as data concerning the
22 disposition of recovered CO2 are not available. Such recovery may or may not affect the overall estimate of CO2
23 emissions depending upon the end use to which the recovered CO2 is applied. Further research is required to
24 determine whether byproduct CO2 is being recovered from other ammonia production plants for application to
25 end uses that are not accounted for elsewhere; however, for reporting purposes, CO2 consumption for urea
26 production is provided in this chapter.
27 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-22. Carbon dioxide
28 emissions from ammonia production in 2021 were estimated to be between 11.4 and 14.1 MMT CO2 Eq. at the 95
29 percent confidence level. This indicates a range of approximately 10 percent below and 11 percent above the
30 emission estimate of 12.0 MMT CO2 Eq.
31 Table 4-22: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from
32 Ammonia Production (MMT CO2 Eq. and Percent)
2021 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
Source Gas
(MMT C02 Eq.)
(MMT CO
zEq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Ammonia Production C02
12.0
11.4
14.1
-10%
+11%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
33 QA/QC and Verification
34 General quality assurance/quality control (QA/QC) procedures were applied to ammonia production emission
35 estimates consistent with the U.S. Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006
36 IPCC Guidelines as described in the introduction of the IPPU chapter (see Annex 8 for more details). More details
37 on the greenhouse gas calculation, monitoring and QA/QC methods applicable to ammonia facilities can be found
4-32 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 under Subpart G (Ammonia Production) of the regulation (40 CFR Part 98).24 EPA verifies annual facility-level
2 GHGRP reports through a multi-step process (e.g., combination of electronic checks and manual reviews) to
3 identify potential errors and ensure that data submitted to EPA are accurate, complete, and consistent.25 Based on
4 the results of the verification process, EPA follows up with facilities to resolve mistakes that may have occurred.
5 The post-submittals checks are consistent with a number of general and category-specific QC procedures, including
6 range checks, statistical checks, algorithm checks, and year-to-year checks of reported data and emissions.
7 More details on the greenhouse gas calculation, monitoring, and QA/QC methods applicable to reporting of urea
8 produced at ammonia production facilities can be found under Section 4.6 Urea Consumption for Non-Agricultural
9 Purposes.
10 Recalculations
11 Based on the updated methodology, recalculations were performed for emissions from ammonia for years 1990
12 through 2020. Compared to the previous Inventory, total CO2 emissions from ammonia production (from natural
13 gas and petroleum coke feedstocks) increased by an average of 8.7 percent (961 kt CO2) per year, ranging from a
14 decrease of 4.8 percent (507 kt CO2) in 2015 to an increase of 13.3 percent (1,203 kt CO2) in 2007.
15 Planned Improvements
16 Currently the Inventory does not separately track fuel energy use for ammonia production. To be more consistent
17 with 2006IPCC Guidelines, EPA is considering whether to include natural gas fuel use as part of ammonia
18 production emissions as a future improvement. The data are still being evaluated as part of EPA's efforts to
19 disaggregate other industrial sector categories' energy use in the Energy chapter of the Inventory. If possible, this
20 will be incorporated in future Inventory reports. If incorporated, the fuel energy use and emissions will be
21 removed from current reporting under Energy to avoid double counting.
22 4.6 Urea Consumption for Non-Agricultural
23 Purposes
24 Urea is produced using ammonia (NH3) and carbon dioxide (CO2) as raw materials. All urea produced in the United
25 States is assumed to be produced at ammonia production facilities where both ammonia and CO2 are generated.
26 There were 35 plants producing ammonia in the United States in 2021, with two additional plants sitting idle for
27 the entire year (USGS 2022b).
28 The chemical reaction that produces urea is:
29 2NH3+ C02 -> NH2COONH4 -> CO(NH2)2 + H20
30 This section accounts for CO2 emissions associated with urea consumed exclusively for non-agricultural purposes.
31 Emissions of CO2 resulting from agricultural applications of urea are accounted for in Section 5.6 Urea Fertilization
32 (CRF Source Category 3H) of the Agriculture chapter.
33 The industrial applications of urea include its use in adhesives, binders, sealants, resins, fillers, analytical reagents,
34 catalysts, intermediates, solvents, dyestuffs, fragrances, deodorizers, flavoring agents, humectants and
35 dehydrating agents, formulation components, monomers, paint and coating additives, photosensitive agents, and
24 See http://www.ecfr.gov/cgi-bin/text-idx7tpN/ecfrbrowse/Title40/40cfr98 main Q2.tpl.
25 See https://www.epa.eov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
Industrial Processes and Product Use 4-33
-------
1 surface treatments agents. In addition, urea is used for abating nitrogen oxide (NOx) emissions from coal-fired
2 power plants and diesel transportation motors.
3 Emissions of CO2 from urea consumed for non-agricultural purposes in 2021 were estimated to be 5.0 MMT CO2
4 Eq. (4,989 kt) and are summarized in Table 4-23 and Table 4-24. Net CO2 emissions from urea consumption for
5 non-agricultural purposes have increased by approximately 32 percent from 1990 to 2021 and decreased by
6 approximately 14.0 percent from 2020 to 2021.
7 Table 4-23: CO2 Emissions from Urea Consumption for Non-Agricultural Purposes (MMT CO2
8 Eq.)
Source 1990 2005 2017 2018 2019 2020 2021
Urea Consumption 3.8 3.7 5.2 6.1 6.2 5.8 5.0
9 Table 4-24: CO2 Emissions from Urea Consumption for Non-Agricultural Purposes (kt)
Source 1990 2005 2017 2018 2019 2020 2021
Urea Consumption 3,784 3,653 5,161 6,111 6,154 5,814 4,989
10 Methodology and Time-Series Consistency
11 Emissions of CO2 resulting from urea consumption for non-agricultural purposes are estimated by multiplying the
12 amount of urea consumed in the United States for non-agricultural purposes by a factor representing the amount
13 of CO2 used as a raw material to produce the urea. This method is based on the assumption that all of the carbon
14 in urea is released into the environment as CO2 during use, consistent with the Tier 1 method used to estimate
15 emissions from ammonia production in the 2006IPCC Guidelines (IPCC 2006) which states that the "CO2 recovered
16 [from ammonia production] for downstream use can be estimated from the quantity of urea produced where CO2
17 is estimated by multiplying urea production by 44/60, the stoichiometric ratio of CO2 to urea."
18 The amount of urea consumed for non-agricultural purposes in the United States is estimated by deducting the
19 quantity of urea fertilizer applied to agricultural lands, which is obtained directly from the Agriculture chapter (see
20 Table 5-25), from the total domestic supply of urea as reported in Table 4-25. The domestic supply of urea is
21 estimated based on the amount of urea produced plus urea imports and minus urea exports. A factor of 0.733 tons
22 of CO2 per ton of urea consumed is then applied to the resulting supply of urea for non-agricultural purposes to
23 estimate CO2 emissions from the amount of urea consumed for non-agricultural purposes. The 0.733 tons of CO2
24 per ton of urea emission factor is based on the stoichiometry of carbon in urea. This corresponds to a
25 stoichiometric ratio of CO2 to urea of 44/60, assuming complete conversion of carbon in urea to CChflPCC 2006;
26 EFMA2000).
27 Urea production data for 1990 through 2008 were obtained from the U.S. Geological Survey (USGS) Minerals
28 Yearbook: Nitrogen (USGS 1994 through 2009a). Urea production data for 2009 through 2010 were obtained from
29 the U.S. Census Bureau (2011). The U.S. Census Bureau ceased collection of urea production statistics in 2011.
30 Urea production data for 2011 through 2021 were obtained from GHGRP(EPA 2018; EPA 2022a; EPA 2022b).
31 Urea import data for 2021 were not available at the time of publication and were estimated using 2020 values.
32 Urea import data for 2013 to 2020 were obtained from the USGS Minerals Yearbook: Nitrogen (USGS 2021a). Urea
33 import data for 2011 and 2012 were taken from U.S. Fertilizer Import/Exports from the United States Department
34 of Agriculture (USDA) Economic Research Service Data Sets (U.S. Department of Agriculture 2012). USDA
35 suspended updates to this data after 2012. Urea import data for the previous years were obtained from the U.S.
36 Census Bureau Current Industrial Reports Fertilizer Materials and Related Products annual and quarterly reports for
37 1997 through 2010 (U.S. Census Bureau 2001 through 2011), The Fertilizer Institute (TFI 2002) for 1993 through
38 1996, and the United States International Trade Commission Interactive Tariff and Trade DataWeb (U.S. ITC 2002)
39 for 1990 through 1992 (see Table 4-25).
4-34 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Urea export data for 2021 were not available at the time of publication and were estimated using 2020 values.
2 Urea export data for 2013 to 2020 were obtained from the USGS Minerals Yearbook: Nitrogen (USGS 2021a). Urea
3 export data for 1990 through 2012 were taken from U.S. Fertilizer Import/Exports from USDA Economic Research
4 Service Data Sets (U.S. Department of Agriculture 2012). USDA suspended updates to this data after 2012.
5 Table 4-25: Urea Production, Urea Applied as Fertilizer, Urea Imports, and Urea Exports (kt)
Year
1990
2005
2017
2018
2019
2020
2021
Urea Production
7,450
5,270
9,030
10,700
11,400
11,500
10,500
Urea Applied as Fertilizer
3,296
4,779
6,630
6,734
6,859
6,984
7,109
Urea Imports
1,860
5,026
5,510
5,110
4,410
4,190
4,190
Urea Exports
854
536
872
743
559
111
111
Urea Consumed for Non-
Agricultural Purposes
5,160
4,981
7,038
8,333
8,392
7,929
6,804
6 Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
7 through 2021. The methodology for urea consumption for non-agricultural purposes spliced activity data from
8 different sources: USGS data for 1990 through 2008, U. S. Census Bureau data for 2009 and 2010, and GHGRP data
9 beginning in 2011. Consistent with the 2006IPCC Guidelines, the overlap technique was applied to compare the
10 data sets for years where there was overlap, with findings that the data sets were consistent and adjustments
11 were not needed.
12 Uncertainty - TO BE UPDATED FOR FINAL INVENTORY REPORT
13 There is limited publicly available data on the quantities of urea produced and consumed for non-agricultural
14 purposes. Therefore, the amount of urea used for non-agricultural purposes is estimated based on a balance that
15 relies on estimates of urea production, urea imports, urea exports, and the amount of urea used as fertilizer. The
16 primary uncertainties associated with this source category are associated with the accuracy of these estimates as
17 well as the fact that each estimate is obtained from a different data source. Because urea production estimates are
18 no longer available from the USGS, there is additional uncertainty associated with urea produced beginning in
19 2011. There is also uncertainty associated with the assumption that all of the carbon in urea is released into the
20 environment as CO2 during use.
21 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-26. Carbon dioxide
22 emissions associated with urea consumption for non-agricultural purposes during 2021 were estimated to be
23 between 5.1 and 6.8 MMT CO2 Eq. at the 95 percent confidence level. This indicates a range of approximately 14
24 percent below and 14 percent above the emission estimate of 5.0 MMT CO2 Eq.
25 Table 4-26: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Urea
26 Consumption for Non-Agricultural Purposes (MMT CO2 Eq. and Percent)
2021 Emission Estimate
Gas
(MMT C02 Eq.)
Uncertainty Range Relative to Emission Estimate3
Source
(MMT C02 Eq.) (%)
Lower Upper Lower Upper
Bound Bound Bound Bound
Urea Consumption
for Non-Agricultural
C02 5.0
5.1 6.8 -14% +14%
Purposes
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
Industrial Processes and Product Use 4-35
-------
1 QA/QC and Verification
2 General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S. Inventory
3 QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006IPCC Guidelines as described in the
4 introduction of the IPPU chapter (see Annex 8 for more details).
5 More details on the greenhouse gas calculation, monitoring and QA/QC methods applicable to reporting of urea
6 production occurring at ammonia facilities can be found under Subpart G (Ammonia Manufacturing) of the
7 regulation (40 CFR Part 98).26 EPA verifies annual facility-level GHGRP reports through a multi-step process (e.g.,
8 combination of electronic checks and manual reviews) to identify potential errors and ensure that data submitted
9 to EPA are accurate, complete, and consistent.27 Based on the results of the verification process, EPA follows up
10 with facilities to resolve mistakes that may have occurred. The post-submittals checks are consistent with a
11 number of general and category-specific QC procedures, including range checks, statistical checks, algorithm
12 checks, and year-to-year checks of reported data and emissions. EPA also conducts QA checks of GHGRP reported
13 urea production data against external datasets including the USGS Minerals Yearbook data. The comparison shows
14 consistent trends in urea production over time.
is Recalculations Discussion
16 Based on updated quantities of urea applied for agricultural uses for 2015 to 2020, updated urea imports from
17 USGS for 2020, and updated urea exports from USGS for 2020, recalculations were performed for 2015 through
18 2020. Compared to the previous Inventory, CO2 emissions from urea consumption for non-agricultural purposes
19 decreased by less than 1 percent (25 kt CO2) for 2015, less than 1 percent (41 kt CO2) for 2016, and less than 1
20 percent (21 kt CO2) for 2017; increased by 1.33 percent (80 kt CO2) for 2018 and by 1.82 percent (110 kt CO2) for
21 2019; and decreased by 2.81 percent (168 kt CO2) for 2020.
22 Planned Improvements
23 At this time, there are no specific planned improvements for estimating CO2 emissions from urea consumption for
24 non-agricultural purposes.
25 4.7 Nitric Acid Production (CRF Source
26 Category 2B2)
27 Nitrous oxide (N2O) is emitted during the production of nitric acid (HNO3), an inorganic compound used primarily
28 to make synthetic commercial fertilizers. Nitric acid is also a major component in the production of adipic acid—a
29 feedstock for nylon—and explosives. Virtually all of the nitric acid produced in the United States is manufactured
30 by the high-temperature catalytic oxidation of ammonia (EPA 1998). There are two different nitric acid production
31 methods: weak nitric acid and high-strength nitric acid. The first method utilizes oxidation, condensation, and
32 absorption to produce nitric acid at concentrations between 30 and 70 percent nitric acid. High-strength acid (90
33 percent or greater nitric acid) can be produced from dehydrating, bleaching, condensing, and absorption of the
34 weak nitric acid. Most U.S. plants were built between 1960 and 2000. As of 2021, there were 31 active nitric acid
35 production plants, including one high-strength nitric acid production plant in the United States (EPA 2010; EPA
36 2022).
26 See http://www.ecfr.gov/cgi-bin/text-idx7tpN/ecfrbrowse/Title40/40cfr98 main Q2.tpl.
27 See https://www.epa.eov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
4-36 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 The basic process technology for producing nitric acid has not changed significantly over time. During this process,
2 N2O is formed as a byproduct and released from reactor vents into the atmosphere. Emissions from fuels
3 consumed for energy purposes during the production of nitric acid are included in the Energy chapter.
4 Nitric acid is made from the reaction of ammonia (NH3) with oxygen (O2) in two stages. The overall reaction is:
5 4NH3 + 802 -> 4HNO:i +4H2
6 Currently, the nitric acid industry in the United States controls emissions of NO and NO2 (i.e., NOx), using a
7 combination of non-selective catalytic reduction (NSCR) and selective catalytic reduction (SCR) technologies. In the
8 process of destroying NOx, NSCR systems are also very effective at destroying N2O. Five nitric acid plants had NSCR
9 systems installed between 1964 and 1977, over half due to the finalization of the Nitric Acid Plant New Source
10 Performance Standards (NSPS) which went into effect in 1971. Four additional nitric acid plants had NSCR systems
11 installed between 2016 and 2018, as a result of EPA Consent Decrees to control NOx emissions more effectively.
12 NSCR systems are used in approximately one-third of the weak acid production plants. For N2O abatement, U.S.
13 facilities are using both tertiary (i.e., NSCR and SCR) and secondary controls (i.e., catalysts added to the ammonia
14 reactor to lessen potential N2O production).
15 Emissions from the production of nitric acid are generally directly proportional to the annual amount of nitric acid
16 produced because emissions are calculated as the product of the total annual production and plant-specific
17 emission factors. There are a few instances, however, where that relationship has not been directly proportional.
18 For example in 2015 and 2019, nitric acid production decreased and emissions increased, compared to the
19 respective preceding years. N2O emissions for those years are calculated based on data from the GHGRP as
20 discussed in the Methodology section below. According to data from plants reporting to GHGRP, plant-specific
21 operations can affect the emission factor used, including: (1) site-specific fluctuations in ambient temperature and
22 humidity, (2) catalyst age and condition, (3) process changes, such as fluctuations in process pressure or
23 temperature and replacing the ammonia catalyst, (4) the addition, removal, maintenance, and utilization of
24 abatement technologies, and (5) the number of nitric acid trains, which are reaction vessels where ammonia is
25 oxidized to form nitric acid. Changes in those operating conditions for the years in question (2015 and 2019)
26 caused changes in emission factors, which resulted in emissions changing disproportionally to production in those
27 years.
28 Nitrous oxide emissions from this source were estimated to be 7.9 MMT CO2 Eq. (30 kt of N2O) in 2021 (see Table
29 4-27). Emissions from nitric acid production have decreased by 27 percent since 1990, while production has
30 increased by 8 percent over the same time period (see Table 4-27). Emissions have decreased by 39 percent since
31 1997, the highest year of production in the time series. From 2020 to 2021, nitric acid production decreased by 2.1
32 percent, leading to an overall decrease in emissions from nitric acid production of 4.8 percent from 2020 to 2021.
33 Table 4-27: N2O Emissions from Nitric Acid Production (MMT CO2 Eq. and kt N2O)
Year
1990
2005
2017
2018
2019
2020
2021
MMTCOz Eq.
10.8
10.1
CO
00
8.5
8.9
8.3
7.9
kt N20
41
38
31
32
34
31
30
34 Methodology and Time-Series Consistency
35 Emissions of N2O were calculated using the estimation methods provided by the 2006IPCC Guidelines and a
36 country-specific method utilizing EPA's GHGRP. The 2006 IPCC Guidelines Tier 2 method was used to estimate
37 emissions from nitric acid production for 1990 through 2009, and a country-specific approach similar to the IPCC
38 Tier 3 method was used to estimate N2O emissions for 2010 through 2021.
39 2010 through 2021
40 Process N2O emissions and nitric acid production data were obtained directly from EPA's GHGRP for 2010 through
41 2021 by aggregating reported facility-level data (EPA 2022).
Industrial Processes and Product Use 4-37
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Since 2010, in the United States, all nitric acid facilities that produce weak nitric acid (30 to 70 percent) have been
required to report annual greenhouse gas emissions data to EPA as per the requirements of the GHGRP (Subpart
V). Beginning with 2018, the rule was changed to include facilities that produce nitric acid of any strength. The only
facility that produces high-strength nitric acid also produces weak nitric acid. All greenhouse gas emissions from
nitric acid production originate from the production of weak nitric acid.
Process emissions and nitric acid production reported to the GHGRP provide complete estimates of greenhouse
gas emissions for the United States because there are no reporting thresholds. While facilities are allowed to stop
reporting to the GHGRP if the total reported emissions from nitric acid production are less than 25,000 metric tons
CO2 Eq. per year for five consecutive years or less than 15,000 metric tons CO2 Eq. per year for three consecutive
years, no facilities have stopped reporting as a result of these provisions.28 All nitric acid facilities are required to
either calculate process N2O emissions using a site-specific emission factor that is the average of the emission
factor determined through annual performance tests for each nitric acid train under typical operating conditions or
directly measure process N2O emissions using monitoring equipment.29
Emissions from facilities vary from year to year, depending on the amount of nitric acid produced with and without
abatement technologies and other conditions affecting the site-specific emission factor. To maintain consistency
across the time series and with the rounding approaches taken by other data sets, GHGRP nitric acid data are
rounded and are shown in Table 4-28
1990 through 2009
Using GHGRP data for 2010,30 country-specific N2O emission factors were calculated for nitric acid production with
abatement and without abatement (i.e., controlled and uncontrolled emission factors). The following 2010
emission factors were derived for production with abatement and without abatement: 3.3 kg INhO/metric ton
HNO3 produced at plants using abatement technologies (e.g., tertiary systems such as NSCR systems) and 5.99 kg
INhO/metric ton HNO3 produced at plants not equipped with abatement technology. Country-specific weighted
emission factors were derived by weighting these emission factors by percent production with abatement and
without abatement over time periods 1990 through 2008 and 2009. These weighted emission factors were used to
estimate N2O emissions from nitric acid production for years prior to the availability of GHGRP data (i.e., 1990
through 2008 and 2009). A separate weighted emission factor is included for 2009 due to data availability for that
year.
EPA verified the installation dates of N2O abatement technologies for all facilities based on GHGRP facility-level
information and confirmed that all abatement technologies were accounted for in the derived emission factors
(EPA 2021). No changes to N2O abatement levels from 1990 through 2008 or for 2009 were made due to the
review of GHGRP-reported N2O abatement installation dates. Due to the lack of information on abatement
equipment utilization, it is assumed that once abatement technology was installed in facilities, the equipment was
consistently operational for the duration of the time series considered in this report (especially NSCRs).
The country-specific weighted N2O emission factors were used in conjunction with annual production to estimate
N2O emissions for 1990 through 2009, using the following equations:
28 See 40 CFR 98.2(i)(l) and 40 CFR 98.2(i)(2) for more information about these provisions.
29 Facilities must use standard methods - either EPA Method 320 or ASTM D6348-03 for annual performance tests - and must
follow associated QA/QC procedures consistent with category-specific QC of direct emission measurements during these
performance tests.
30 National N20 process emissions, national production, and national share of nitric acid production with abatement and
without abatement technology was aggregated from the GHGRP facility-level data for 2010 to 2017 (i.e., percent production
with and without abatement).
4-38 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Equation 4-4: 2006IPCCGuidelinesTier 3: N2O Emissions From Nitric Acid Production
(Equation 3.6)
E, = P.- X EF.
weighted, i
[(%Pc,i X EFc) + (%PunCii x EFunc)J
where,
Annual N2O Emissions for year i (kg/yr)
Annual nitric acid production for year i (metric tons HNO3)
Weighted N2O emission factor for year i (kg INhO/metric ton HNO3)
Percent national production of HNO3 with N2O abatement technology (%)
N2O emission factor, with abatement technology (kg INhO/metric ton HNO3)
Percent national production of HNO3 without N2O abatement technology (%)
N2O emission factor, without abatement technology (kg INhO/metric ton HNO3)
year from 1990 through 2009
• For 2009: Weighted N2O emission factor = 5.46 kg INhO/metric ton HNO3.
• For 1990 through 2008: Weighted N2O emission factor = 5.66 kg INhO/metric ton HNO3.
Nitric acid production data for the United States for 1990 through 2009 were obtained from the U.S. Census
Bureau (U.S. Census Bureau 2008, 2009, 2010a, 2010b) (see Table 4-28). EPA used GHGRP facility-level information
to verify that all reported N2O abatement equipment were incorporated into the estimation of N2O emissions from
nitric acid production over the full time series (EPA 2021).
Table 4-28: Nitric Acid Production (kt)
Year
1990
2005
2017
2018
2019
2020
2021
Production (kt)
7,200
6,710
7,780
8,210
8,080
7,970
7,800
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2021. The methodology for nitric acid production spliced activity data from two different sources: U. S.
Census Bureau production data for 1990 through 2009 and GHGRP production data starting in 2010. Consistent
with the 2006 IPCC Guidelines, the overlap technique was applied to compare the two data sets for years where
there was overlap, with findings that the data sets were consistent and adjustments were not needed.
Uncertainty-TO BE UPDATED FOR FINAL INVENTORY REPORT
Uncertainty associated with the parameters used to estimate N2O emissions includes the share of U.S. nitric acid
production attributable to each emission abatement technology (i.e., utilization) over the time series (especially
prior to 2010), and the associated emission factors applied to each abatement technology type. While some
information has been obtained through outreach with industry associations, limited information is available over
the time series (especially prior to 2010) for a variety of facility level variables, including plant-specific production
levels, plant production technology (e.g., low or high pressure, etc.), and abatement technology destruction and
removal efficiency rates. Production data prior to 2010 were obtained from National Census Bureau, which does
not provide uncertainty estimates with their data. Facilities reporting to EPA's GHGRP must measure production
using equipment and practices used for accounting purposes. While emissions are often directly proportional to
production, the emission factor for individual facilities can vary significantly from year to year due to site-specific
fluctuations in ambient temperature and humidity, catalyst age and condition, nitric acid production process
changes, the addition or removal of abatement technologies, and the number of nitric acid trains at the facility. At
this time, EPA does not estimate uncertainty of the aggregated facility-level information. As noted in the QA/QC
and verification section below, EPA verifies annual facility-level reports through a multi-step process (e.g.,
combination of electronic checks and manual reviews by staff) to identify potential errors and ensure that data
Industrial Processes and Product Use 4-39
-------
1 submitted to EPA are accurate, complete, and consistent. The annual production reported by each nitric acid
2 facility under EPA's GHGRP and then aggregated to estimate national N2O emissions is assumed to have low
3 uncertainty. EPA assigned an uncertainty range of ±5 percent for facility-reported N2O emissions, consistent with
4 section 3.4.3.1 of the 2006IPCC Guidelines, and ±2 percent for nitric acid production, consistent with section
5 3.3.3.2 of the 2006 IPCC Guidelines.
6 The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 4-29. Nitrous oxide
7 emissions from nitric acid production were estimated to be between 8.8 and 9.8 MMT CO2 Eq. at the 95 percent
8 confidence level. This indicates a range of approximately 5 percent below to 5 percent above the 2021 emissions
9 estimate of 8.9 MMT CO2 Eq.
10 Table 4-29: Approach 2 Quantitative Uncertainty Estimates for N2O Emissions from Nitric
11 Acid Production (MMT CO2 Eq. and Percent)
2021 Emission Estimate
Source Gas
(MMT C02 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMTCO2 Eq.)
(%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Nitric Acid Production N20 8.9
8.8 9.8
-5% +5%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
12 QA/QC and Verification
13 General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S. Inventory
14 QA/QC plan, which is in accordance with Volume 1, Chapter 6 of the 2006 IPCC Guidelines as described in the
15 introduction of the IPPU chapter (see Annex 8 for more details). More details on the greenhouse gas calculation,
16 monitoring and QA/QC methods applicable to nitric acid facilities can be found under Subpart V: Nitric Acid
17 Production of the GHGRP regulation (40 CFR Part 98).31
18 The main QA/QC activities are related to annual performance testing, which must follow either EPA Method 320 or
19 ASTM D6348-03. EPA verifies annual facility-level GHGRP reports through a multi-step process that is tailored to
20 the Subpart (e.g., combination of electronic checks including range checks, statistical checks, algorithm checks,
21 year-to-year comparison checks, along with manual reviews) to identify potential errors and ensure that data
22 submitted to EPA are accurate, complete, and consistent. Based on the results of the verification process, EPA
23 follows up with facilities to resolve mistakes that may have occurred (EPA 2015).32 EPA's review of observed
24 trends noted that while emissions have generally mirrored production, in 2015 and 2019 nitric acid production
25 decreased compared to the previous year and emissions increased. While review is ongoing, based on feedback
26 from the verification process to date, these changes are due to facility-specific changes (e.g., in the nitric
27 production process and management of abatement equipment).
28 Recalculations Discussion
29 For the current Inventory, CC>2-equivalent estimates of total N2O emissions from nitric acid production have been
30 revised to reflect the 100-year global warming potentials (GWPs) provided in the IPCC Fifth Assessment Report
31 (AR5) (IPCC 2013). AR5 GWP values differ slightly from those presented in the IPCC Fourth Assessment Report
32 (AR4) (IPCC 2007) (used in the previous inventories). The AR5 GWPs have been applied across the entire time
33 series for consistency. The GWP of N2O has decreased from 298 to 265, leading to an overall decrease in estimates
31 See Subpart V monitoring and reporting regulation http://www.ecfr.gov/cgi-bin/text-
idx?tpl=/ecfrbrowse/Title40/40cfr98 main Q2.tpl.
32 See GHGRP Verification Factsheet https://www.epa.gov/sites/production/files/2015-
07/documents/ghgrp verification factsheet.pdf.
4-40 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 of CCh-equivalent N2O emissions. Compared to the previous Inventory which applied 100-year GWP values from
2 AR4, N2O emissions decreased by 11 percent for each year of the time series, ranging from a decrease of 1.0 MMT
3 CO2 Eq. in 2020 to 1.6 MMT CO2 Eq. in 1997. Further discussion on this update and the overall impacts of updating
4 the inventory GWPs to reflect the IPCC Fifth Assessment Report can be found in Chapter 9, Recalculations and
5 Improvements.
6 Planned Improvements
7 Pending resources, EPA is considering a near-term improvement to both review and refine quantitative uncertainty
8 estimates and the associated qualitative discussion.
9 4.8 Adipic Acid Production (CRF Source
10 Category 2B3)
11 Adipic acid is produced through a two-stage process during which nitrous oxide (N2O) is generated in the second
12 stage. Emissions from fuels consumed for energy purposes during the production of adipic acid are accounted for
13 in the Energy chapter. The first stage of manufacturing usually involves the oxidation of cyclohexane to form a
14 cyclohexanone/cyclohexanol mixture. The second stage involves oxidizing this mixture with nitric acid to produce
15 adipic acid. Nitrous oxide is generated as a byproduct of the nitric acid oxidation stage and is emitted in the waste
16 gas stream (Thiemens and Trogler 1991). The second stage is represented by the following chemical reaction:
17 ('CH2)5CO(cyclohexanone) + (CH2)zCHOH (cyclohexanol) + wHN03
18 -» HOOC(CH2)4COOH(adipic acid) + xN20 + yH20
19 Process emissions from the production of adipic acid vary with the types of technologies and level of emission
20 controls employed by a facility. In 1990, two major adipic acid-producing plants had N2O abatement technologies
21 in place and, as of 1998, three major adipic acid production facilities had control systems in place (Reimer et al.
22 1999). In 2021, thermal reduction was applied as an N2O abatement measure at one adipic acid facility (EPA 2022).
23 Worldwide, only a few adipic acid plants exist. The United States, Europe, and China are the major producers, with
24 the United States accounting for the largest share of global adipic acid production capacity in recent years. In 2021,
25 the United States had two companies with a total of two adipic acid production facilities (one in Texas and one in
26 Florida), following the ceased operations of a third major production facility at the end of 2015 (EPA 2022).
27 Adipic acid is a white crystalline solid used in the manufacture of synthetic fibers, plastics, coatings, urethane
28 foams, elastomers, and synthetic lubricants. Commercially, it is the most important of the aliphatic dicarboxylic
29 acids, which are used to manufacture polyesters. Eighty-four percent of all adipic acid produced in the United
30 States is used in the production of nylon 6,6; 9 percent is used in the production of polyester polyols; 4 percent is
31 used in the production of plasticizers; and the remaining 4 percent is accounted for by other uses, including
32 unsaturated polyester resins and food applications (ICIS 2007). Food grade adipic acid is used to provide some
33 foods with a "tangy" flavor (Thiemens and Trogler 1991).
34 Compared to 1990, national adipic acid production in 2021 has increased by less than 1 percent to approximately
35 760,000 metric tons (ACC 2022). Nitrous oxide emissions from adipic acid production were estimated to be 6.6
36 MMT CO2 Eq. (25 kt N2O) in 2021 (see Table 4-30). Over the period 1990 through 2021, facilities have reduced
37 emissions by 51 percent due to the widespread installation of pollution control measures in the late 1990s. The
38 COVID-19 pandemic may have partially influenced the 11 percent decrease in N2O emissions from adipic acid
39 production between 2020 and 2021.
40 Significant changes in the amount of time that the N2O abatement device at one facility was in operation has been
41 the main cause of fluctuating emissions in recent years. These fluctuations are most evident for years where trends
42 in emissions and adipic acid production were not directly proportional: (1) between 2016 and 2017, (2) between
Industrial Processes and Product Use 4-41
-------
1 2017 and 2018, and (3) between 2019 and 2020. As noted above, changes in control measures and abatement
2 technologies at adipic acid production facilities, including maintenance of equipment, can result in annual emission
3 fluctuations. Little additional information is available on drivers of trends, and the amount of adipic acid produced
4 is not reported under EPA's GHGRP.
5 Table 4-30: N2O Emissions from Adipic Acid Production (MMT CO2 Eq. and kt N2O)
Year
1990
2005
2017
2018
2019
2020
2021
MMTCOz Eq.
13.5
6.3
6.6
9.3
4.7
7.4
6.6
kt N20
51
24
25
35
18
28
25
e Methodology and Time-Series Consistency
7 Emissions are estimated using both Tier 2 and Tier 3 methods consistent with the 2006IPCC Guidelines. Due to
8 confidential business information (CBI), plant names are not provided in this section; therefore, the four adipic
9 acid-producing facilities that have operated over the time series will be referred to as Plants 1 through 4. As noted
10 above, one currently operating facility uses thermal reduction as an N2O abatement technology.
11 2010 through 2021
12 All emission estimates for 2010 through 2021 were obtained through analysis of GHGRP data (EPA 2010 through
13 2022), which is consistent with the 2006 IPCC Guidelines Tier 3 method. Facility-level greenhouse gas emissions
14 data were obtained from EPA's GHGRP for the years 2010 through 2021 (EPA 2010 through 2022) and aggregated
15 to national N2O emissions. Consistent with IPCC Tier 3 methods, all adipic acid production facilities are required to
16 either calculate N2O emissions using a facility-specific emission factor developed through annual performance
17 testing under typical operating conditions or directly measure N2O emissions using monitoring equipment.33
is 1990 through 2009
19 For years 1990 through 2009, which were prior to EPA's GHGRP reporting, for both Plants 1 and 2, emission
20 estimates were obtained directly from the plant engineers and account for reductions due to control systems in
21 place at these plants during the time series. These prior estimates are considered CBI and hence are not published
22 (Desai 2010, 2011). These estimates were based on continuous process monitoring equipment installed at the two
23 facilities.
24 For Plant 4,1990 through 2009 N2O emissions were estimated using the following Tier 2 equation from the 2006
25 IPCC Guidelines:
26 Equation 4-5: 2006IPCCGuide/inesTier 2: N2O Emissions From Adipic Acid Production
27 (Equation 3.8)
28
29 where,
30 Eaa
31 Qaa
32 EFaa
33 DF
34 UF
Eaa = Qaa x EFaa X (1 - [DF X UF])
N2O emissions from adipic acid production, metric tons
Quantity of adipic acid produced, metric tons
Emission factor, metric ton INhO/metric ton adipic acid produced
N2O destruction factor
Abatement system utility factor
33 Facilities must use standard methods, either EPA Method 320 or ASTM D6348-03 for annual performance testing, and must
follow associated QA/QC procedures during these performance tests consistent with category-specific QC of direct emission
measurements.
4-42 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 The adipic acid production is multiplied by an emission factor (i.e., N2O emitted per unit of adipic acid produced),
2 which has been estimated to be approximately 0.3 metric tons of N2O per metric ton of product (IPCC 2006). The
3 "N2O destruction factor" in the equation represents the percentage of N2O emissions that are destroyed by the
4 installed abatement technology. The "abatement system utility factor" represents the percentage of time that the
5 abatement equipment operates during the annual production period. Plant-specific production data for Plant 4
6 were obtained across the time series through personal communications (Desai 2010, 2011). The plant-specific
7 production data were then used for calculating emissions as described above.
8 For Plant 3, 2005 through 2009 emissions were obtained directly from the plant (Desai 2010, 2011). For 1990
9 through 2004, emissions were estimated using plant-specific production data and the IPCC factors as described
10 above for Plant 4. Plant-level adipic acid production for 1990 through 2003 was estimated by allocating national
11 adipic acid production data to the plant level using the ratio of known plant capacity to total national capacity for
12 all U.S. plants (ACC 2022; CMR 2001,1998; CW 1999; C&EN 1992 through 1995). For 2004, actual plant production
13 data were obtained and used for emission calculations (CW 2005).
14 Plant capacities for 1990 through 1994 were obtained from Chemical & Engineering News, "Facts and Figures" and
15 "Production of Top 50 Chemicals" (C&EN 1992 through 1995). Plant capacities for 1995 and 1996 were kept the
16 same as 1994 data. The 1997 plant capacities were taken from Chemical Market Reporter, "Chemical Profile: Adipic
17 Acid" (CMR 1998). The 1998 plant capacities for all four plants and 1999 plant capacities for three of the plants
18 were obtained from Chemical Week, Product Focus: Adipic Acid/Adiponitrile (CW 1999). Plant capacities for the
19 year 2000 for three of the plants were updated using Chemical Market Reporter, "Chemical Profile: Adipic Acid"
20 (CMR 2001). For 2001 through 2003, the plant capacities for three plants were held constant at year 2000
21 capacities. Plant capacity for 1999 to 2003 for the one remaining plant was kept the same as 1998.
22 National adipic acid production data (see Table 4-31) from 1990 through 2021 were obtained from the American
23 Chemistry Council (ACC 2022).
24 Table 4-31: Adipic Acid Production (kt)
Year
1990 I
2005 I
2017
2018
2019
2020
2021
Production (kt)
755 1
865 |
830
825
810
710
760
25 Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
26 through 2021. The methodology for adipic acid production spliced activity data from multiple sources: plant-
27 specific emissions data and publicly available plant capacity data for 1990 through 2009 and GHGRP emission data
28 starting in 2010. Consistent with the 2006 IPCC Guidelines, the overlap technique was applied to compare the two
29 data sets for years where there was overlap, with findings that the data sets were consistent and adjustments
30 were not needed.
31 Uncertainty - TO BE UPDATED FOR FINAL INVENTORY REPORT
32 Uncertainty associated with N2O emission estimates includes the methods used by companies to monitor and
33 estimate emissions. While some information has been obtained through outreach with facilities, limited
34 information is available over the time series on these methods, abatement technology destruction and removal
35 efficiency rates, and plant-specific production levels. EPA assigned an uncertainty range of ±5 percent for facility-
36 reported N2O emissions, consistent with section 3.4.3.2 of the 2006 IPCC Guidelines.
37 The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 4-32. Nitrous oxide
38 emissions from adipic acid production for 2021 were estimated to be between 7.9 and 8.7 MMT CO2 Eq. at the 95
39 percent confidence level. These values indicate a range of approximately 5 percent below to 5 percent above the
40 2021 emission estimate of 7.4 MMT CO2 Eq.
41 Table 4-32: Approach 2 Quantitative Uncertainty Estimates for N2O Emissions from Adipic
42 Acid Production (MMT CO2 Eq. and Percent)
Source Gas 2021 Emission Estimate Uncertainty Range Relative to Emission Estimate3
Industrial Processes and Product Use 4-43
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
(MMT CO? Eq.)
(MMT CO? Eq.)
(%)
Lower Upper
Lower
Upper
Bound Bound
Bound
Bound
Adipic Acid Production
N20
7.4
7.9 8.7
-5%
+5%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S. Inventory
QA/QC plan, which is in accordance with Volume 1, Chapter 6 of the 2006IPCC Guidelines as described in the
introduction of the IPPU chapter (see Annex 8 for more details).
More details on the greenhouse gas calculation, monitoring and QA/QC methods applicable to adipic acid facilities
can be found under Subpart E (Adipic Acid Production) of the GHGRP regulation (40 CFR Part 98).34 The main
QA/QC activities are related to annual performance testing, which must follow either EPA Method 320 or ASTM
D6348-03. EPA verifies annual facility-level GHGRP reports through a multi-step process (e.g., combination of
electronic checks and manual reviews) to identify potential errors and ensure that data submitted to EPA are
accurate, complete, and consistent (EPA 2015).35 Based on the results of the verification process, EPA follows up
with facilities to resolve mistakes that may have occurred. The post-submittals checks are consistent with a
number of general and category-specific QC procedures, including range checks, statistical checks, algorithm
checks, and year-to-year comparisons of reported data.
Recalculations Discussion
For the current Inventory, CCh-equivalent estimates of total N2O emissions from adipic acid production have been
revised to reflect the 100-year global warming potentials (GWPs) provided in the IPCC Fifth Assessment Report
(AR5) (IPCC 2013). AR5 GWP values differ slightly from those presented in the IPCC Fourth Assessment Report
(AR4) (IPCC 2007) (used in the previous inventories). The AR5 GWPs have been applied across the entire time
series for consistency. The GWP of N2O has decreased from 298 to 265, leading to an overall decrease in estimates
of CCh-equivalent N2O emissions. Compared to the previous Inventory which applied 100-year GWP values from
AR4, N2O emissions decreased by 11.1 percent for each year of the time series, ranging from a decrease of 0.3
MMT CO2 Eq. in 2008 to 1.9 MMT CO2 Eq. in 1995. Further discussion on this update and the overall impacts of
updating the inventory GWPs to reflect the IPCC Fifth Assessment Report can be found in Chapter 9, Recalculations
and Improvements.
Planned Improvements
EPA plans to review GHGRP facility reported information on the date of abatement technology installation in order
to better reflect trends and changes in emissions abatement within the industry across the time series. To date,
the facility using the facility-specific emission factor developed through annual performance testing has reported
no installation and no utilization of N2O abatement technology. The facility using direct measurement of N2O
emissions has reported the use of thermal reduction as an N2O abatement technology but is not required to report
the date of installation.
34 See http://www.ecfr.eov/cei-bin/text-idx7tph/ecfrbrowse/Title40/40cfr98 main Q2.tpl.
35 See https://www.epa.eov/sites/production/files/2015-07/documents/eherp verification factsheet.pdf.
4-44 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
4.9 Caprolactam, Glyoxal and Glyoxylic
Acid Production (CRF Source Category 2B4)
Caprolactam
Caprolactam (CsHnNO) is a colorless monomer produced for nylon-6 fibers and plastics. A substantial proportion
of the fiber is used in carpet manufacturing. Most commercial processes used for the manufacture of caprolactam
begin with benzene, but toluene can also be used. The production of caprolactam can give rise to significant
emissions of nitrous oxide (N2O).
During the production of caprolactam, emissions of N2O can occur from the ammonia oxidation step, emissions of
carbon dioxide (CO2) from the ammonium carbonate step, emissions of sulfur dioxide (SO2) from the ammonium
bisulfite step, and emissions of non-methane volatile organic compounds (NMVOCs). Emissions of CO2, SO2 and
NMVOCs from the conventional process are unlikely to be significant in well-managed plants. Modified
caprolactam production processes are primarily concerned with elimination of the high volumes of ammonium
sulfate that are produced as a byproduct of the conventional process (IPCC 2006).
In the most commonly used process where caprolactam is produced from benzene, benzene is hydrogenated to
cyclohexane which is then oxidized to produce cyclohexanone (CsHioO). The classical route (Raschig process) and
basic reaction equations for production of caprolactam from cyclohexanone are (IPCC 2006):
Oxidation of NH3 to NO/N02
I
NH3 reacted with C02/H20 to yield ammonium carbonate (NH4)2C03
I
(NH4)2C03 reacted with NO/N02 (from NH3 oxidation) to yield ammonium nitrite (NH4N02)
I
NH3 reacted with S02/H20 to yield ammonium bisulphite (NH4HS03)
I
NH4N02 and (NH4HS03) reacted to yield hydroxylamine disulphonate (NOH(S03NH4)2)
I
(NOH(S03NH4)2) hydrolised to yield hydroxylamine sulphate ((NH2OH)2.H2S04) and
ammonium sulphate ((NH4)2S04)
I
Cylohexanone reaction-.
1
C6H10O + ~(NH20H)2.H2S04(+NH3 and H2S04) -> C6H10NOH + (NH4)2S04 + H20
I
Beckmann rearrangement:
C6H10NOH (+H2S04 and S02) -> C^H^NO.H^O,^ (+4NH3 and H20) -> C^NO + 2(NH4)2S04
In 2004, three facilities produced caprolactam in the United States (ICIS 2004). Another facility, Evergreen
Recycling, was in operation from 2000 to 2001 (ICIS 2004; Textile World 2000) and from 2007 through 2015 (DOE
2011; Shaw 2015). Caprolactam production at Fibrant LLC (formerly DSM Chemicals) in Georgia ceased in 2018
Industrial Processes and Product Use 4-45
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
(Cline 2019). As of 2021, two companies in the United States produced caprolactam at two facilities: AdvanSix
(formerly Honeywell) in Virginia (AdvanSix 2022) and BASF in Texas (BASF 2022).
Nitrous oxide emissions from caprolactam production in the United States were estimated to be 1.2 MMT CO2 Eq,
(5 kt N2O) in 2021 (see Table 4-33). National emissions from caprolactam production decreased by approximately
17 percent over the period of 1990 through 2021. Emissions in 2021 increased by approximately 6 percent from
the 2020 levels. This annual increase returned caprolactam production to levels consistent with 2019 before the
COVID-19 pandemic.
Table 4-33: N2O Emissions from Caprolactam Production (MMT CO2 Eq. and kt N2O)
Year 1990 2005 2017 2018 2019 2020 2021
MMT C02 Eq. 1.5 1.9 1.3 1.3 1.2 1.2 1.2
kt N20 6 7 5 5 5 4 5
Glyoxal
Glyoxal is mainly used as a crosslinking agent for vinyl acetate/acrylic resins, disinfectant, gelatin hardening agent,
textile finishing agent (permanent-press cotton, rayon fabrics), and wet-resistance additive (paper coatings) (IPCC
2006). It is also used for enhanced oil-recovery. It is produced from oxidation of acetaldehyde with concentrated
nitric acid, or from the catalytic oxidation of ethylene glycol, and N2O is emitted in the process of oxidation of
acetaldehyde.
Glyoxal (ethanedial) (C2H2O2) is produced from oxidation of acetaldehyde (ethanal) (C2H4O) with concentrated
nitric acid (HNO3). Glyoxal can also be produced from catalytic oxidation of ethylene glycol (ethanediol)
(CH2OHCH2OH).
Glyoxylic Acid
Glyoxylic acid is produced by nitric acid oxidation of glyoxal. Glyoxylic acid is used for the production of synthetic
aromas, agrochemicals, and pharmaceutical intermediates (IPCC 2006).
EPA does not currently estimate the emissions associated with the production of Glyoxal and Glyoxylic Acid due to
a lack of publicly available information on the industry in the United States. See Annex 5 for additional information.
Methodology and Time-Series Consistency
Emissions of N2O from the production of caprolactam were calculated using the estimation methods provided by
the 2006 IPCC Guidelines. The 2006 IPCC Guidelines Tier 1 method was used to estimate emissions from
caprolactam production for 1990 through 2021, as shown in this formula:
Equation 4-6: 2006IPCCGuide/inesTier 1: N2O Emissions From Caprolactam Production
(Equation 3.9)
en2o = EF x CP
where,
Enzo = Annual N2O Emissions (kg)
EF = N2O emission factor (default) (kg N20/metric ton caprolactam produced)
CP = Caprolactam production (metric tons)
During the caprolactam production process, N2O is generated as a byproduct of the high temperature catalytic
oxidation of ammonia (NH3), which is the first reaction in the series of reactions to produce caprolactam. The
amount of N2O emissions can be estimated based on the chemical reaction shown above. Based on this formula,
which is consistent with an IPCC Tier 1 approach, approximately 111.1 metric tons of caprolactam are required to
4-46 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 generate one metric ton of N2O, resulting in an emission factor of 9.0 kg N2O per metric ton of caprolactam (IPCC
2 2006). When applying the Tier 1 method, the 2006 IPCC Guidelines state that it is good practice to assume that
3 there is no abatement of N2O emissions and to use the highest default emission factor available in the guidelines.
4 In addition, EPA did not find support for the use of secondary catalysts to reduce N2O emissions, such as those
5 employed at nitric acid plants.
6 The activity data for caprolactam production (see Table 4-34) from 1990 to 2021 were obtained from the American
7 Chemistry Council's Guide to the Business of Chemistry (ACC 2022). EPA will continue to analyze and assess
8 alternative sources of production data as a quality control measure.
9 Table 4-34: Caprolactam Production (kt)
Year
1990 I
2005 I
2017
2018
2019
2020
2021
Production (kt)
626 |
795 |
545
530
515
490
520
10
11 Carbon dioxide and methane (CH4) emissions may also occur from the production of caprolactam, but currently the
12 IPCC does not have methodologies for calculating these emissions associated with caprolactam production.
13 Methodological approaches, consistent with IPCC guidelines, have been applied to the entire time series to ensure
14 consistency in emissions from 1990 through 2021.
15 Uncertainty - TO BE UPDATED FOR FINAL INVENTORY REPORT
16 Estimation of emissions of N2O from caprolactam production can be treated as analogous to estimation of
17 emissions of N2O from nitric acid production. Both production processes involve an initial step of NH3 oxidation,
18 which is the source of N2O formation and emissions (IPCC 2006). Therefore, uncertainties for the default emission
19 factor values in the 2006 IPCC Guidelines are an estimate based on default values for nitric acid plants. In general,
20 default emission factors for gaseous substances have higher uncertainties because mass values for gaseous
21 substances are influenced by temperature and pressure variations and gases are more easily lost through process
22 leaks. The default values for caprolactam production have a relatively high level of uncertainty due to the limited
23 information available (IPCC 2006). EPA assigned an uncertainty range of ±40 percent for facility-reported N2O
24 emissions, consistent with Section 3.5.2.1 of the 2006 IPCC Guidelines.
25 The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 4-35. Nitrous oxide
26 emissions from Caprolactam, Glyoxal and Glyoxylic Acid Production for 2021 were estimated to be between 0.8
27 and 1.6 MMT CO2 Eq. at the 95 percent confidence level. These values indicate a range of approximately 31
28 percent below to 32 percent above the 2021 emission estimate of 1.2 MMT CO2 Eq.
29 Table 4-35: Approach 2 Quantitative Uncertainty Estimates for N2O Emissions from
30 Caprolactam, Glyoxal and Glyoxylic Acid Production (MMT CO2 Eq. and Percent)
2021 Emission Estimate
Source Gas
(MMT C02 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.)
(%)
Lower Upper
Lower Upper
Bound Bound
Bound Bound
Caprolactam Production N20 1.2
0.8 1.6
-31% +32%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
31 QA/QC and Verification
32 General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S. Inventory
33 QA/QC plan, which is in accordance with Volume 1, Chapter 6 of the 2006 IPCC Guidelines as described in the
34 introduction of the IPPU chapter (see Annex 8 for more details).
Industrial Processes and Product Use 4-47
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Recalculations Discussion
Recalculations were performed for 2020 to reflect updated caprolactam production data from the American
Chemistry Council's Guide to the Business of Chemistry (ACC 2022). In addition, for the current Inventory, CO2-
equivalent total emission estimates of N2O from caprolactam production have been revised to reflect the 100-year
global warming potentials (GWPs) provided in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). AR5 GWP
values differ slightly from those presented in the IPCC Fourth Assessment Report (AR4) (IPCC 2007) (used in the
previous inventories). The AR5 GWPs have been applied across the entire time series for consistency. The GWP of
N2O decreased from 298 to 265, leading to an overall decrease in estimates of calculated CC>2-equivalent N2O
emissions. Compared to the previous Inventory, which applied 100-year GWP values from AR4, annual N2O
emissions decreased by 11 percent each year, ranging from a decrease of 0.15 MMT CO2 Eq. in 2020 to 0.25 MMT
CO2 Eq. in 2010 and 2011. Further discussion on this update and the overall impacts of updating the Inventory
GWP values to reflect the IPCC Fifth Assessment Report can be found in Chapter 9, Recalculations and
Improvements.
Planned Improvements
Pending resources, EPA will research other available datasets for caprolactam production and industry trends,
including facility-level data. EPA continues to research the production process and emissions associated with the
production of glyoxal and glyoxylic acid. Preliminary data suggests that glyoxal and glyoxylic acid may no longer be
produced domestically and are largely imported to the United States. EPA is working to identify historical data to
understand if any production of these chemicals has occurred since 1990. During the Expert Review period for the
current Inventory report, EPA continues to seek expert solicitation on data available for these emission source
categories. This planned improvement is subject to data availability and will be implemented in the medium- to
long-term.
4.10 Carbide Production and Consumption
(CRF Source Category 2B5)
Carbon dioxide (CO2) and methane (CH4) are emitted from the production of silicon carbide (SiC), a material used
for industrial abrasive applications as well as metallurgical and other non-abrasive applications in the United
States. Emissions from fuels consumed for energy purposes during the production of silicon carbide are accounted
for in the Energy chapter. Additionally, some metallurgical and non-abrasive applications of SiC are emissive, and
while emissions should be accounted for where they occur based on 2006 IPCC Guidelines, emissions from SiC
consumption are accounted for here until additional data on SiC consumption by end-use are available.
To produce SiC, silica sand or quartz (SiCh) is reacted with carbon (C) in the form of petroleum coke. A portion
(about 35 percent) of the carbon contained in the petroleum coke is retained in the SiC. The remaining C is emitted
as CO2, Cm, or carbon monoxide (CO). The overall reaction is shown below, but in practice, it does not proceed
according to stoichiometry:
Si02 + 3C —> SiC + 2CO (+ 02 —> 2COz)
Carbon dioxide and CH4 are also emitted during the production of calcium carbide, a chemical used to produce
acetylene. Carbon dioxide is implicitly accounted for in the storage factor calculation for the non-energy use of
petroleum coke in the Energy chapter. As noted in Annex 5 to this report, CH4 emissions from calcium carbide
production are not estimated because data are not available. EPA is continuing to investigate the inclusion of these
emissions in future Inventory reports.
Markets for manufactured abrasives, including SiC, are heavily influenced by activity in the U.S. manufacturing
sector, especially in the aerospace, automotive, furniture, housing, and steel manufacturing sectors. Specific
4-48 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
applications of abrasive-grade SiC in 2017 included antislip abrasives, blasting abrasives, bonded abrasives, coated
abrasives, polishing and buffing compounds, tumbling media, and wire-sawing abrasives. Approximately 50
percent of SiC is used in metallurgical applications, which include primarily iron and steel production, and other
non-abrasive applications, which include use in advanced or technical ceramics and refractories (USGS 1991a
through 2021; Washington Mills 2021).
As a result of the economic downturn in 2008 and 2009, demand for SiC decreased in those years. Low-cost
imports, particularly from China, combined with high relative operating costs for domestic producers, continue to
put downward pressure on the production of SiC in the United States. Consumption of SiC in the United States has
recovered somewhat from its low in 2009 to 2020; 2021 consumption data was withheld to avoid disclosing
company proprietary data (USGS 1991b through 2021b).
Silicon carbide was manufactured by two facilities in the United States, one of which produced primarily non-
abrasive SiC (USGS 2021). USGS production values for the United States consists of SiC used for abrasives and for
metallurgical and other non-abrasive applications (USGS 2021). During the COVID-19 pandemic in 2020, the U.S.
Department of Homeland Security considered abrasives manufacturing part of the critical manufacturing sector,
and as a result, pandemic "stay-at-home" orders issued in March 2020 did not affect the abrasives manufacturing
industry. These plants remained at full operation (USGS 2021a). Consumption of SiC decreased by approximately
25 percent in 2020 due to the pandemic and a sharp decline in imports and rebounded with an increase of
approximately 30 percent from 2020 to 2021, remaining below pre-pandemic levels (U.S. Census Bureau 2005
through 2021).
Carbon dioxide emissions from SiC production and consumption in 2021 were 0.2 MMT CO2 Eq. (172 kt CO2), which
are about 29 percent lower than emissions in 1990 (243 kt) (see Table 4-36and Table 4-37). Approximately 53
percent of these emissions resulted from SiC production, while the remainder resulted from SiC consumption.
Methane emissions from SiC production in 2021 were 0.01 MMT CO2 Eq. (0.4 kt CH4) (see Table 4-36 andTable
4-37). Emissions have not fluctuated greatly in recent years.
Table 4-36: CO2 and ChU Emissions from Silicon Carbide Production and Consumption (MMT
COz Eq.)
Year 1990 2005 2017 2018 2019 2020 2021
Production
C02 0.2 0.1 0.1 0.1 0.1 0.1 0.1
CH4 + + + + + + +
Consumption
C02 Ol Ol Ol Ol Ol Ol 0.1
Total 012 012 02 02 02 02 0.2
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Table 4-37: CO2 and ChU Emissions from Silicon Carbide Production and Consumption (kt)
Year
1990
2005
2017
2018
2019
2020
2021
Production
C02
170
92
92
92
92
92
92
ch4
1
+
+
+
+
+
+
Consumption
C02
73
121
90
93
84
62
80
Total
243
213
181
184
175
154
172
+ Does not exceed 0.5 kt
Note: Totals may not sum due to independent rounding.
Industrial Processes and Product Use 4-49
-------
1 Methodology and Time-Series Consistency
2 Emissions of CO2 and CH4 from the production of SiC were calculated using the Tier 1 method provided by the 2006
3 IPCC Guidelines. Annual estimates of SiC production were multiplied by the default emission factors, as shown
4 below:
5 Equation 4-7: 2006IPCCGuide/inesTier 1: Emissions from Carbide Production (Equation
6 3.11)
1 Esc,C02 = EFsc,co2 X Qsc
/I metric ton\
8 Esc,CH4 = EFsc CH4 x Qsc x [ 100Qkg )
9 where,
10 Esc,co2 = CO2 emissions from production of SiC, metric tons
11 EFsc,co2 = Emission factor for production of SiC, metric ton CCh/metric ton SiC
12 Qsc = Quantity of SiC produced, metric tons
13 Esc,ch4 = Cm emissions from production of SiC, metric tons
14 EFsc,ch4 = Emission factor for production of SiC, kilogram CH4/metric ton SiC
15 Emission factors were taken from the 2006 IPCC Guidelines:
16 • 2.62 metric tons C02/metric ton SiC
17 • 11.6 kg Cl-U/metric ton SiC
18 Production data includes silicon carbide manufactured for abrasive applications as well as for metallurgical and
19 other non-abrasive applications (USGS 2021).
20 Silicon carbide industrial abrasives production data for 1990 through 2021 were obtained from the U.S. Geological
21 Survey (USGS) Minerals Yearbook: Manufactured Abrasives (USGS 1991a through 2021). Silicon carbide production
22 data published by USGS have been rounded to the nearest 5,000 metric tons to avoid disclosing company
23 proprietary data. For the period 1990 through 2001, reported USGS production data include production from two
24 facilities located in Canada that ceased operations in 1995 and 2001. Using SiC production data from Canada (ECCC
25 2022), U.S. SiC production for 1990 through 2001 was adjusted to reflect only U.S. production.
26 SiC consumption for the entire time series is estimated using USGS consumption data (USGS 1991b through 2021b)
27 and data from the U.S. International Trade Commission (USITC) database on net imports and exports of SiC (U.S.
28 Census Bureau 2005 through 2021) (see Table 4-38). Total annual SiC consumption (utilization) was estimated by
29 subtracting annual exports of SiC from the annual total of national SiC production and annual imports.
30 Emissions of CO2 from SiC consumption for metallurgical uses were calculated by multiplying the annual utilization
31 of SiC for metallurgical uses (reported annually in the USGS Minerals Yearbook: Silicon) by the carbon content of
32 SiC (30.0 percent), which was determined according to the molecular weight ratio of SiC. Because USGS withheld
33 consumption data for metallurgical uses from publication for 2017, 2018, and 2021 due to concerns of disclosing
34 company-specific sensitive information, SiC consumption for 2017 and 2018 were estimated using 2016 values,
35 and SiC consumption for 2021 was estimated using the 2020 value.
36 Emissions of C02from SiC consumption for other non-abrasive uses were calculated by multiplying the annual SiC
37 consumption for non-abrasive uses by the carbon content of SiC (30 percent). The annual SiC consumption for non-
38 abrasive uses was calculated by multiplying the annual SiC consumption (production plus net imports) by the
39 percentage used in metallurgical and other non-abrasive uses (50 percent) (USGS 1991a through 2021) and then
40 subtracting the SiC consumption for metallurgical use.
41 The petroleum coke portion of the total CO2 process emissions from silicon carbide production is adjusted for
42 within the Energy chapter, as these fuels were consumed during non-energy related activities. Additional
43 information on the adjustments made within the Energy sector for Non-Energy Use of Fuels is described in both
4-50 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
the Methodology section of CO2 from Fossil Fuel Combustion (Section 3.1) and Annex 2.1, Methodology for
Estimating Emissions of CO2 from Fossil Fuel Combustion.
Table 4-38: Production and Consumption of Silicon Carbide (Metric Tons)
Year 1990
2005
2017 2018 2019 2020 2021
Production 65,000
Consumption 132,465
35,000
220,149
35,000 35,000 35,000 35,000 35,000
163,492 168,526 152,412 113,756 146,312
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2021.
Uncertainty-TO BE UPDATED FOR FINAL INVENTORY REPORT
Silicon carbide production data published by the USGS is rounded to the nearest 5,000 tons and has been
consistently reported at 35,000 tons since 2003 to avoid disclosure of company proprietary data. This translates to
an uncertainty range of ±7 percent for SiC production (USGS 2021). There is uncertainty associated with the
emission factors used because they are based on stoichiometry as opposed to monitoring of actual SiC production
plants. An alternative is to calculate emissions based on the quantity of petroleum coke used during the
production process rather than on the amount of silicon carbide produced; however, these data were not
available. For CFU, there is also uncertainty associated with the hydrogen-containing volatile compounds in the
petroleum coke (IPCC 2006). Consistent with the range in Section 3.6.3.1 of the 2006IPCC Guidelines, EPA assigned
an uncertainty of ±10 percent for the Tier 1 CO2 and CFU emission factors for the SiC production processes. There is
also uncertainty associated with the use or destruction of CH4 generated from the process, in addition to
uncertainty associated with levels of production, net imports, consumption levels, and the percent of total
consumption that is attributed to metallurgical and other non-abrasive uses. Consistent with the range in Section
3.6.3.2 of the 2006 IPCC Guidelines, EPA assigned an uncertainty range of ±5 percent for the primary data inputs
for consumption (i.e., crude imports, ground and refined imports, crude exports, ground and refined exports,
utilization [metallurgical applications]) to calculate overall uncertainty from SiC production.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-39. Silicon carbide
production and consumption CO2 emissions from 2021 were estimated to be between 9 percent below and 9
percent above the emission estimate of 0.17 MMT CO2 Eq. at the 95 percent confidence level. Silicon carbide
production CFU emissions were estimated to be between 7 percent below and 7 percent above the emission
estimate of 0.01 MMT CO2 Eq. at the 95 percent confidence level.
Table 4-39: Approach 2 Quantitative Uncertainty Estimates for ChU and CO2 Emissions from
Silicon Carbide Production and Consumption (MMT CO2 Eq. and Percent)
Source Gas
2021 Emission Estimate Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (MMT C02 Eq.) (%)
Lower Upper Lower Upper
Bound Bound Bound Bound
C02 0.17 0.14 0.17 -9% +9%
Silicon Carbide Production
and Consumption
Silicon Carbide Production CH4 + + + -7% +7%
+ Does not exceed 0.05 MMT C02 Eq.
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S. Inventory
QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006 IPCC Guidelines as described in the
introduction of the IPPU chapter (see Annex 8 for more details).
Industrial Processes and Product Use 4-51
-------
i Recalculations Discussion
2 Recalculations were performed for 1990 through 2001 to account for updated data on SiC production from
3 Canada, which is used to revise production data to reflect only U.S. production. Compared to the previous
4 Inventory, estimates of CO2 emissions in 1997 increased by 3 kt CO2, and estimates of CH4 emissions increased by
5 11 metric tons CH4.
6 Updated USITC data on 2019 SiC exports and 2020 SiC imports resulted in updated SiC consumption estimates for
7 those years. Compared to the previous Inventory, SiC consumption values for 2019 and 2020 increased by less
8 than 2 metric tons and 20 metric tons, respectively. These minimal increases did not impact emissions estimates,
9 compared to the previous Inventory.
10 In addition, for the current Inventory, CC>2-equivalent estimates of total CH4 emissions from carbide production
11 have been revised to reflect the 100-year global warming potentials (GWPs) provided in the IPCC Fifth Assessment
12 Report (AR5) (IPCC 2013). AR5 GWP values differ slightly from those presented in the IPCC Fourth Assessment
13 Report (AR4) (IPCC 2007) (used in the previous inventories). The AR5 GWPs have been applied across the entire
14 time series for consistency. The GWP of CH4 increased from 25 to 28, leading to an overall increase in estimates for
15 CC>2-equivalent CH4 emissions. Compared to the previous Inventory, which applied 100-year GWP values from AR4,
16 annual CC>2-equivalent CFU emissions increased by 12 percent each year, ranging from an increase of 1.0 kt CO2 Eq.
17 in 2002 to 2.3 kt CO2 Eq. in 1990. The net impact on the entire category from these updates was an average annual
18 0.7 percent increase in emissions for the time series. Further discussion on this update and the overall impacts of
19 updating the Inventory GWP values to reflect the IPCC Fifth Assessment Report can be found in Chapter 9,
20 Recalculations and Improvements.
22 EPA is initiating research for data on SiC consumption by end-use for consideration in updating emissions
23 estimates from SiC consumption and to account for emissions where they occur. This planned improvement is
24 subject to data availability and will be implemented in the medium- to long-term given significance of emissions.
25 EPA has not integrated aggregated facility-level GHGRP information to inform estimates of CO2 and Cl-Ufrom SiC
26 production and consumption. The aggregated information (e.g., activity data and emissions) associated with silicon
27 carbide did not meet criteria to shield underlying confidential business information (CBI) from public disclosure.
28 EPA plans to examine the use of GHGRP silicon carbide emissions data for possible use in emission estimates
29 consistent with both Volume 1, Chapter 6 of the 2006 IPCC Guidelines and the latest IPCC guidance on the use of
30 facility-level data in national inventories. This planned improvement is ongoing and has not been incorporated into
31 this Inventory report. This is a long-term planned improvement.
34 Titanium dioxide (TiCh) is manufactured using one of two processes: the chloride process and the sulfate process.
35 The chloride process uses petroleum coke and chlorine as raw materials and emits process-related carbon dioxide
36 (CO2). Emissions from fuels consumed for energy purposes during the production of titanium dioxide are
37 accounted for in the Energy chapter. The sulfate process does not use petroleum coke or other forms of carbon as
38 a raw material and does not emit CO2. The chloride process is based on the following chemical reactions and does
39 emit CO2:
21 Planned Improvements
33
32
4.11 Titanium Dioxide Production (CRF
Source Category 2B6)
41
40
2FeTi03 + 7CZ2 ~F 3C —> 2TiCl^ + 2FeCl^ + 3C02
2TiCl4 + 2O2 ~* 2Ti02 ~l~
4-52 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
The carbon in the first chemical reaction is provided by petroleum coke, which is oxidized in the presence of the
chlorine and FeTiCh (rutile ore) to form CO2. Since 2004, all TiCh produced in the United States has been produced
using the chloride process, and a special grade of "calcined" petroleum coke is manufactured specifically for this
purpose.
The principal use of TiCh is as a white pigment in paint, lacquers, and varnishes. It is also used as a pigment in the
manufacture of plastics, paper, and other products. In 2021, U.S. TiC>2 production totaled 1,100,000 metric tons
(USGS 2022). Five plants produced TiCh in the United States in 2021.
Emissions of CO2 from titanium dioxide production in 2021 were estimated to be 1.5 MMT CO2 Eq. (1,474 kt CO2),
which represents an increase of 12 percent since 1990 (see Table 4-40). Compared to 2020, emissions from
titanium dioxide production increased by 24 percent in 2021, due to a 24 percent increase in production. The
annual production increase in 2021 represents a return to production levels seen in 2019 before the COVID-19
pandemic.
Table 4-40: CO2 Emissions from Titanium Dioxide (MMT CO2 Eq. and kt)
Year
1990
2005
2017
2018
2019
2020
2021
MMTCOz Eq.
1.2
1.8
1.7
1.5
1.5
1.2
1.5
kt
1,195
1,755
1,688
1,541
1,474
1,193
1,474
Methodology and Time-Series Consistency
Emissions of CO2 from TiC>2 production were calculated by multiplying annual national TiC>2 production by chloride
process-specific emission factors using a Tier 1 approach provided in 2006IPCC Guidelines. The Tier 1 equation is
as follows:
Equation 4-8: 2006IPCCGuide/inesTier 1: CO2 Emissions from Titanium Production
(Equation 3.12)
Etd = EFtd X Qtd
where,
Etd = CO2 emissions from Ti02 production, metric tons
EFtd = Emission factor (chloride process), metric ton CCh/metric ton TiC>2
Qtd = Quantity of Ti02 produced, metric tons
The petroleum coke portion of the total CO2 process emissions from Ti02 production is adjusted for within the
Energy chapter as these fuels were consumed during non-energy related activities. Additional information on the
adjustments made within the Energy sector for Non-Energy Use of Fuels is described in both the Methodology
section of CO2 from Fossil Fuel Combustion (Section 3.1 Fossil Fuel Combustion) and Annex 2.1, Methodology for
Estimating Emissions of CO2 from Fossil Fuel Combustion.
Data were obtained for the total amount of Ti02 produced each year. For years prior to 2004, it was assumed that
Ti02 was produced using the chloride process and the sulfate process in the same ratio as the ratio of the total U.S.
production capacity for each process. As of 2004, the last remaining sulfate process plant in the United States
closed; therefore, 100 percent of production since 2004 used the chloride process (USGS 2005). An emission factor
of 1.34 metric tons C02/metric ton Ti02 was applied to the estimated chloride-process production (IPCC 2006). It
was assumed that all Ti02 produced using the chloride process was produced using petroleum coke, although
some Ti02 may have been produced with graphite or other carbon inputs.
The emission factor for the Ti02 chloride process was taken from the 2006 IPCC Guidelines. Titanium dioxide
production data and the percentage of total Ti02 production capacity that used the chloride process for 1990
through 2018 (see Table 4-41) were obtained through the U.S. Geological Survey (USGS) Minerals Yearbook:
Titanium (USGS 1991 through 2022). Production data for 2019 were obtained from the USGS Minerals Yearbook:
Titanium, advanced data release of the 2019 tables (USGS 2021). Production data for 2020 and 2021 were
Industrial Processes and Product Use 4-53
-------
1 obtained from the Minerals Commodity Summaries: Titanium and Titanium Dioxide (USGS 2022).36 Data on the
2 percentage of total TiCh production capacity that used the chloride process were not available for 1990 through
3 1993, so data from the 1994 USGS Minerals Yearbook were used for these years. Because a sulfate process plant
4 closed in September 2001, the chloride process percentage for 2001 was estimated (Gambogi 2002). By 2002, only
5 one sulfate process plant remained online in the United States, and this plant closed in 2004 (USGS 2005).
6 Table 4-41: Titanium Dioxide Production (kt)
Year
1990 |
2005 |
2017
2018
2019
2020
2021
Production
979 |
1,310 |
1,260
1,150
1,100
890
1,100
7 Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
8 through 2021.
9 Uncertainty - TO BE UPDATED FOR FINAL INVENTORY REPORT
10 Each year, the USGS collects titanium industry data for titanium mineral and pigment production operations. If
11 TiC>2 pigment plants do not respond, production from the operations is estimated based on prior year production
12 levels and industry trends. Variability in response rates fluctuates from 67 to 100 percent of TiCh pigment plants
13 over the time series. EPA currently uses an uncertainty range of ±5 percent for the primary data inputs (i.e., TiC>2
14 production and chloride process capacity values) to calculate overall uncertainty from TiCh production, consistent
15 with the range in Section 3.7.3.2 of the 2006IPCC Guidelines.
16 Although some TiCh may be produced using graphite or other carbon inputs, information and data regarding these
17 practices were not available. Titanium dioxide produced using graphite inputs, for example, may generate differing
18 amounts of CChper unit of TiCh produced as compared to that generated using petroleum coke in production.
19 While the most accurate method to estimate emissions would be to base calculations on the amount of reducing
20 agent used in each process rather than on the amount of TiC>2 produced, sufficient data were not available to do
21 so.
22 As of 2004, the last remaining sulfate-process plant in the United States closed. Since annual TiCh production was
23 not reported by USGS by the type of production process used (chloride or sulfate) prior to 2004 and only the
24 percentage of total production capacity by process was reported, the percent of total TiC>2 production capacity that
25 was attributed to the chloride process was multiplied by total TiC>2 production to estimate the amount of TiC>2
26 produced using the chloride process. Finally, the emission factor was applied uniformly to all chloride-process
27 production, and no data were available to account for differences in production efficiency among chloride-process
28 plants. In calculating the amount of petroleum coke consumed in chloride-process TiC>2 production, literature data
29 were used for petroleum coke composition. Certain grades of petroleum coke are manufactured specifically for
30 use in the TiC>2 chloride process; however, this composition information was not available. Consistent with the
31 range in Table 3.9 of the 2006 IPCC Guidelines, EPA assigned an uncertainty range of ±15 percent for the Tier 1 CO2
32 emission factor for the titanium dioxide (chloride route) production process.
33 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-42. Titanium dioxide
34 consumption CO2 emissions from 2021 were estimated to be between 1.2 and 1.5 MMT CO2 Eq. at the 95 percent
35 confidence level. This indicates a range of approximately 13 percent below and 13 percent above the emission
36 estimate of 1.3 MMT CO2 Eq.
36 EPA has not integrated aggregated facility-level GHGRP information for Titanium Dioxide production facilities (40 CFR Part 98
Subpart EE). The relevant aggregated information (activity data, emission factor) from these facilities did not meet criteria to
shield underlying CBI from public disclosure.
4-54 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Table 4-42: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Titanium
2 Dioxide Production (MMT CO2 Eq. and Percent)
2021 Emission Estimate
Source Gas
(MMT C02 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.)
1
[%)
Lower Upper
Bound Bound
Lower
Bound
Upper
Bound
Titanium Dioxide Production C02 1.3
1.2 1.5
-13%
+13%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
3 QA/QC and Verification
4 General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S. Inventory
5 QA/QC plan, which is in accordance with Volume 1, Chapter 6 of the 2006IPCC Guidelines as described in the
6 introduction of the IPPU chapter (see Annex 8 for more details).
7 Recalculations Discussion
8 Updated USGS data on TiC>2 production was available for 2020, resulting in updated emissions estimates for that
9 year. Compared to the previous Inventory, emissions for 2020 decreased by 12 percent (110 kt CO2 Eq.).
10 Planned Improvements
11 EPA plans to examine the use of GHGRP titanium dioxide emissions and other data for possible use in emission
12 estimates consistent with both Volume 1, Chapter 6 of the 2006 IPCC Guidelines and the latest IPCC guidance on
13 the use of facility-level data in national inventories.37 This planned improvement is ongoing and has not been
14 incorporated into this Inventory report. This is a long-term planned improvement given significance of these
15 emissions.
16 4.12 Soda Ash Production (CRF Source
17 Category 2B7)
18 Carbon dioxide (CO2) is generated as a byproduct of calcining trona ore to produce soda ash and is eventually
19 emitted into the atmosphere. In addition, CO2 may also be released when soda ash is consumed. Emissions from
20 soda ash consumption not associated with glass production are reported under Section 4.4 Other Process Uses of
21 Carbonates (CRF Category 2A4), and emissions from fuels consumed for energy purposes during the production
22 and consumption of soda ash are accounted for in the Energy chapter.
23 Calcining involves placing crushed trona ore into a kiln to convert sodium bicarbonate into crude sodium carbonate
24 that will later be filtered into pure soda ash. The emission of CO2 during trona-based production is based on the
25 following reaction:
26 2Na2C03 ¦ NaHC03 ¦ 2H20(Trona) -» 3Na2C03(Soda Ash) + 5H20 +C02
TJ Soda ash (sodium carbonate, Na2CC>3) is a white crystalline solid that is readily soluble in water and strongly
28 alkaline. Commercial soda ash is used as a raw material in a variety of industrial processes and in many familiar
29 consumer products such as glass, soap and detergents, paper, textiles, and food. The largest use of soda ash is for
37 See http://www.ipcc-nggip.iges.or.jp/public/tb/TFI Technical Bulletin l.pdf.
Industrial Processes and Product Use 4-55
-------
1 glass manufacturing. Emissions from soda ash used in glass production are reported under Section 0, Glass
2 Production (CRF Source Category 2A3). In addition, soda ash is used primarily to manufacture many sodium-based
3 inorganic chemicals, including sodium bicarbonate, sodium chromates, sodium phosphates, and sodium silicates
4 (USGS 2018b). Internationally, two types of soda ash are produced: natural and synthetic. The United States
5 produces only natural soda ash and is second only to China in total soda ash production. Trona is the principal ore
6 from which natural soda ash is made.
7 The United States represents about one-fifth of total world soda ash output (USGS 2021a). Only two states
8 produce natural soda ash: Wyoming and California. Of these two states, net emissions of CO2 from soda ash
9 production were only calculated for Wyoming, due to specifics regarding the production processes employed in
10 the state.38 Based on 2021 reported data, the estimated distribution of soda ash by end-use in 2021 (excluding
11 glass production) was chemical production, 53 percent; other uses, 16 percent; wholesale distributors (e.g., for use
12 in agriculture, water treatment, and grocery wholesale), 11 percent; soap and detergent manufacturing, 10
13 percent; flue gas desulfurization, 7 percent; water treatment, 2 percent; and pulp and paper production, 2 percent
14 (USGS 2022b).39
15 U.S. natural soda ash is competitive in world markets because it is generally considered a better-quality raw
16 material than synthetically produced soda ash, and most of the world's soda ash is synthetic. Although the United
17 States continues to be a major supplier of soda ash, China surpassed the United States in soda ash production in
18 2003, becoming the world's leading producer.
19 In 2021, CO2 emissions from the production of soda ash from trona ore were 1.7 MMT CO2 Eq. (1,714 kt CO2) (see
20 Table 4-43). Total emissions from soda ash production in 2021 increased by approximately 17 percent compared to
21 emissions in 2020, as soda ash production returned to 2018 levels observed before the COVID-19 pandemic.
22 Emissions have increased by approximately 20 percent from 1990 levels.
23 Trends in emissions have remained relatively constant over the time series with some fluctuations since 1990. In
24 general, these fluctuations were related to the behavior of the export market and the U.S. economy. The U.S. soda
25 ash industry saw a decline in domestic and export sales caused by adverse global economic conditions in 2009,
26 followed by a steady increase in production through 2019 before a significant decrease in 2020 due to the COVID-
27 19 pandemic.
28 Table 4-43: CO2 Emissions from Soda Ash Production (MMT CO2 Eq. and kt CO2)
Year
1990
2005
2017
2018
2019
2020
2021
MMT CO? Eq.
1.4
1.7
1.8
1.7
1.8
1.5
1.7
kt C02
1,431
1,655
1,753
1,714
1,792
1,461
1,714
29 Methodology and Time-Series Consistency
30 During the soda ash production process, trona ore is calcined in a rotary kiln and chemically transformed into a
31 crude soda ash that requires further processing. Carbon dioxide and water are generated as byproducts of the
32 calcination process. Carbon dioxide emissions from the calcination of trona ore can be estimated based on the
38 In California, soda ash is manufactured using sodium carbonate-bearing brines instead of trona ore. To extract the sodium
carbonate, the complex brines are first treated with C02 in carbonation towers to convert the sodium carbonate into sodium
bicarbonate, which then precipitates from the brine solution. The precipitated sodium bicarbonate is then calcined back into
sodium carbonate. Although C02 is generated as a byproduct, the C02 is recovered and recycled for use in the carbonation stage
and is not emitted. A facility in a third state, Colorado, produced soda ash until the plant was idled in 2004. The lone producer
of sodium bicarbonate no longer mines trona ore in the state. For a brief time, sodium bicarbonate was produced using soda
ash feedstocks mined in Wyoming and shipped to Colorado. Prior to 2004, because the trona ore was mined in Wyoming, the
production numbers given by the USGS included the feedstocks mined in Wyoming and shipped to Colorado. In this way, the
sodium bicarbonate production that took place in Colorado was accounted for in the Wyoming numbers.
39 Percentages may not add up to 100 percent due to independent rounding.
4-56 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 chemical reaction shown above. Based on this formula and the IPCC default emission factor of 0.0974 metric tons
2 CO2 per metric ton of trona ore, both of which are consistent with an IPCC Tier 1 approach, one metric ton of CO2
3 is emitted when approximately 10.27 metric tons of trona ore are processed (IPCC 2006). Thus, the 17.6 million
4 metric tons of trona ore mined in 2021 for soda ash production (USGS 2022b) resulted in CO2 emissions of
5 approximately 1.7 MMT CO2 Eq. (1,714 kt).
6 Once produced, most soda ash is consumed in chemical production, with minor amounts used in soap production,
7 pulp and paper, flue gas desulfurization, and water treatment (excluding soda ash consumption for glass
8 manufacturing). As soda ash is consumed for these purposes, additional CO2 is usually emitted. Consistent with the
9 2006 IPCC Guidelines for National Greenhouse Gas Inventories, emissions from soda ash consumption in chemical
10 production processes are reported under Section 4.4 Other Process Uses of Carbonates (CRF Category 2A4).
11 Data is not currently available for the quantity of trona used in soda ash production. Because trona ore produced is
12 used primarily for soda ash production, EPA assumes that all trona produced was used in soda ash production. The
13 activity data for trona ore production (see Table 4-44) for 1990 through 2021 were obtained from the U.S.
14 Geological Survey (USGS) Minerals Yearbook for Soda Ash (1994 through 2015b) and USGS Mineral Industry
15 Surveys for Soda Ash (USGS 2016 through 2017, 2018a, 2019, 2020, 2021, 2022b). Soda ash production40 data
16 were collected by the USGS from voluntary surveys of the U.S. soda ash industry. EPA will continue to analyze and
17 assess opportunities to use facility-level data from EPA's GHGRP to improve the emission estimates for the Soda
18 Ash Production source category consistent with IPCC41 and UNFCCC guidelines.
19 Table 4-44: Trona Ore Used in Soda Ash Production (kt)
Year
1990
2005
2017
2018
2019
2020
2021
Trona Ore Use3
14,700
17,000
18,000
17,600
18,400
15,000
17,600
a Trona ore use is assumed to be equal to trona ore production.
20 Methodological approaches were applied to the entire time series to ensure consistency in emissions estimates
21 from 1990 through 2021.
22 Uncertainty - TO BE UPDATED FOR FINAL INVENTORY REPORT
23 Emission estimates from soda ash production have relatively low associated uncertainty levels because reliable
24 and accurate data sources are available for the emission factor and activity data for trona-based soda ash
25 production. One source of uncertainty is the purity of the trona ore used for manufacturing soda ash. The emission
26 factor used for this estimate assumes the ore is 100 percent pure and likely overestimates the emissions from soda
27 ash manufacture. The average water-soluble sodium carbonate-bicarbonate content for ore mined in Wyoming
28 ranges from 85.5 to 93.8 percent (USGS 1995c).
29 EPA is aware of one facility producing soda ash from a liquid alkaline feedstock process, based on EPA's GHGRP.
30 Soda ash production data was collected by the USGS from voluntary surveys. A survey request was sent to each of
31 the five soda ash producers, all of which responded, representing 100 percent of the total production data (USGS
32 2022b). EPA assigned a default uncertainty range of ±5 percent for trona production, consistent with the ranges in
33 Section 3.8.2.2 of the 2006 IPCC Guidelines, and -15 percent to 0 percent range for the trona emission factor,
34 based on expert judgment on the purity of mined trona.
35 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-45. Soda ash production
36 CO2 emissions for 2021 were estimated to be between 1.3 and 1.5 MMT CO2 Eq. at the 95 percent confidence
40 EPA has assessed the feasibility of using emissions information (including activity data) from EPA's GHGRP program. At this
time, the aggregated information associated with production of soda ash did not meet criteria to shield underlying confidential
business information (CBI) from public disclosure.
41 See http://www.ipcc-nggip.iges.or.jp/public/tb/TFI Technical Bulletin l.pdf.
Industrial Processes and Product Use 4-57
-------
1 level. This indicates a range of approximately 9 percent below and 8 percent above the emission estimate of 1.7
2 MMT CO2 Eq.
3 Table 4-45: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Soda Ash
4 Production (MMT CO2 Eq. and Percent)
Source
Gas
2021 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.)
(MMT C02 Eq.)
(%)
Lower Upper
Lower
Upper
Bound Bound
Bound
Bound
Soda Ash Production
C02
1.7
1.3 1.5
-9%
+8%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
5 QA/QC and Verification
6 General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S. Inventory
7 QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006IPCC Guidelines as described in the
8 introduction of the IPPU chapter (see Annex 8 for more details).
9 Recalculations Discussion
10 No recalculations were performed for the 1990 through 2020 portion of the time series.
11 Planned Improvements
12 EPA is assessing planned improvements for future reports, but at this time has no specific planned improvements
13 for estimating CO2 emissions from soda ash production.
14 4.13 Petrochemical Production (CRF Source
15 Category 2B8)
16 The production of some petrochemicals results in carbon dioxide (CO2) and methane (CH4) emissions.
17 Petrochemicals are chemicals isolated or derived from petroleum or natural gas. Carbon dioxide emissions from
18 the production of acrylonitrile, carbon black, ethylene, ethylene dichloride, ethylene oxide, and methanol, and CH4
19 emissions from the production of methanol and acrylonitrile are presented here and reported under IPCC Source
20 Category 2B8. The petrochemical industry uses primary fossil fuels (i.e., natural gas, coal, petroleum, etc.) for non-
21 fuel purposes in the production of carbon black and other petrochemicals. Emissions from fuels and feedstocks
22 transferred out of the system for use in energy purposes (e.g., indirect or direct process heat or steam production)
23 are currently accounted for in the Energy sector. The allocation and reporting of emissions from feedstocks
24 transferred out of the system for use in energy purposes to the Energy chapter is consistent with the 2006 IPCC
25 Guidelines.
26 Worldwide, more than 90 percent of acrylonitrile (vinyl cyanide, C3H3N) is made by way of direct ammoxidation of
27 propylene with ammonia (NH3) and oxygen over a catalyst. This process is referred to as the SOHIO process,
28 named after the Standard Oil Company of Ohio (SOHIO) (IPCC 2006). The primary use of acrylonitrile is as the raw
29 material for the manufacture of acrylic and modacrylic fibers. Other major uses include the production of plastics
30 (acrylonitrile-butadiene-styrene [ABS] and styrene-acrylonitrile [SAN]), nitrile rubbers, nitrile barrier resins,
31 adiponitrile, and acrylamide. All U.S. acrylonitrile facilities use the SOHIO process (AN 2014). The SOHIO process
4-58 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
involves a fluidized bed reaction of chemical-grade propylene, ammonia, and oxygen over a catalyst. The process
produces acrylonitrile as its primary product, and the process yield depends on the type of catalyst used and the
process configuration. The ammoxidation process produces byproduct CO2, carbon monoxide (CO), and water
from the direct oxidation of the propylene feedstock and produces other hydrocarbons from side reactions.
Carbon black is a black powder generated by the incomplete combustion of an aromatic petroleum- or coal-based
feedstock at a high temperature. Most carbon black produced in the United States is added to rubber to impart
strength and abrasion resistance, and the tire industry is by far the largest consumer. The other major use of
carbon black is as a pigment. The predominant process used in the United States to produce carbon black is the
furnace black (or oil furnace) process. In the furnace black process, carbon black oil (a heavy aromatic liquid) is
continuously injected into the combustion zone of a natural gas-fired furnace. Furnace heat is provided by the
natural gas and a portion of the carbon black feedstock; the remaining portion of the carbon black feedstock is
pyrolyzed to carbon black. The resultant CO2 and uncombusted CH4 are released from thermal incinerators used as
control devices, process dryers, and equipment leaks. Three facilities in the United States use other types of
carbon black processes. Specifically, one facility produces carbon black by the thermal cracking of acetylene-
containing feedstocks (i.e., acetylene black process), a second facility produces carbon black by the thermal
cracking of other hydrocarbons (i.e., thermal black process), and a third facility produces carbon black by the open
burning of carbon black feedstock (i.e., lamp black process) (EPA 2000).
Ethylene (C2H4) is consumed in the production processes of the plastics industry including polymers such as high,
low, and linear low density polyethylene (HDPE, LDPE, LLDPE); polyvinyl chloride (PVC); ethylene dichloride;
ethylene oxide; and ethylbenzene. Virtually all ethylene is produced from steam cracking of ethane, propane,
butane, naphtha, gas oil, and other feedstocks. The representative chemical equation for steam cracking of ethane
to ethylene is shown below:
C2H6 -» C2H4 + H2
Small amounts of Cm are also generated from the steam cracking process. In addition, CO2 and Cm emissions
result from combustion units.
Ethylene dichloride (C2H4CI2) is used to produce vinyl chloride monomer, which is the precursor to polyvinyl
chloride (PVC). Ethylene dichloride was also used as a fuel additive until 1996 when leaded gasoline was phased
out. Ethylene dichloride is produced from ethylene by either direct chlorination, oxychlorination, or a combination
of the two processes (i.e., the "balanced process"); most U.S. facilities use the balanced process. The direct
chlorination and oxychlorination reactions are shown below:
C2H4 + Cl2 -» C2H4Cl2 (direct chlorination)
C2H4 + i02 + 2HCI -» C2H4Cl2 + 2H20 (oxychlorination)
C2H4 + 302 -» 2C02 + 2H20 (direct oxidation of ethylene during oxychlorination)
In addition to the byproduct CO2 produced from the direction oxidation of the ethylene feedstock, CO2 and Cm
emissions are also generated from combustion units.
Ethylene oxide (C2H4O) is used in the manufacture of glycols, glycol ethers, alcohols, and amines. Approximately 70
percent of ethylene oxide produced worldwide is used in the manufacture of glycols, including monoethylene
glycol. Ethylene oxide is produced by reacting ethylene with oxygen over a catalyst. The oxygen may be supplied to
the process through either an air (air process) or a pure oxygen stream (oxygen process). The byproduct CO2 from
the direct oxidation of the ethylene feedstock is removed from the process vent stream using a recycled carbonate
solution, and the recovered CO2 may be vented to the atmosphere or recovered for further utilization in other
sectors, such as food production (IPCC 2006). The combined ethylene oxide reaction and byproduct CO2 reaction is
exothermic and generates heat, which is recovered to produce steam for the process. The ethylene oxide process
also produces other liquid and off-gas byproducts (e.g., ethane that may be burned for energy recovery within the
process. Almost all facilities, except one in Texas, use the oxygen process to manufacture ethylene oxide (EPA
2008).
Industrial Processes and Product Use 4-59
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Methanol (CH3OH) is a chemical feedstock most often converted into formaldehyde, acetic acid and olefins. It is
also an alternative transportation fuel, as well as an additive used by municipal wastewater treatment facilities in
the denitrification of wastewater. Methanol is most commonly synthesized from a synthesis gas (i.e., "syngas" - a
mixture containing H2, CO, and CO2) using a heterogeneous catalyst. There are a number of process techniques
that can be used to produce syngas. Worldwide, steam reforming of natural gas is the most common method;
most methanol producers in the United States also use steam reforming of natural gas to produce syngas. Other
syngas production processes in the United States include partial oxidation of natural gas and coal gasification.
Emissions of CO2 and CH4 from petrochemical production in 2021 were 33.2 MMT CO2 Eq. (33,170 kt CO2) and 0.4
MMT CO2 Eq. (15 kt CH4), respectively (see Table 4-46 and Table 4-47). Carbon dioxide emissions from
petrochemical production are driven primarily from ethylene production, while CH4 emissions are almost entirely
from methanol production. Since 1990, total CO2 emissions from petrochemical production increased by 53
percent, and CH4 emissions increased by 65 percent. Emissions of CO2 and Cl-Uwere higher in 2021 than in any
preceding year. Compared to 2020, CO2 emissions increased 11 percent in 2021, and Cm emissions increased 21
percent. The increases are due primarily to increased ethylene and methanol production, which have been driven
by the increased natural gas production in the United States over the past decade, and to recovery from a strong
hurricane season that temporarily shut down many facilities in Texas and Louisiana in 2020. Emissions from carbon
black also increased significantly in 2021 as the industry began to recover from the lower production in 2020 as a
result of the COVID-19 pandemic.
Table 4-46: CO2 and ChU Emissions from Petrochemical Production (MMT CO2 Eq.)
Year
1990
2005
2017
2018
2019
2020
2021
CO?
21.6
27.4
28.9
29.3
30.7
29.8
33.2
Carbon Black
3.4
4.3
3.3
3.4
3.3
2.6
3.0
Ethylene
13.1
19.0
20.0
19.4
20.7
20.7
22.8
Ethylene Dichloride
0.3
0.5
0.4
0.4
0.5
0.5
0.4
Ethylene Oxide
1.1
1.5
1.3
1.3
1.4
1.7
1.9
Acrylonitrile
1.2
1.3
1.0
1.3
1.0
0.9
0.9
Methanol
2.5
0.8
2.9
3.5
3.8
3.5
4.2
ch4
0.2
0.1
0.3
0.3
0.3
0.3
0.4
Acrylonitrile
+
+
+
+
+
+
+
Methanol
0.2
0.1
0.3
0.3
0.4
0.3
0.4
Total
21.9
27.5
29.2
29.7
31.1
30.1
33.6
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals by gas may not sum due to independent rounding.
Table 4-47: CO2 and ChU Emissions from Petrochemical Production (kt)
Year
1990
2005
2017
2018
2019
2020
2021
CO?
21,611
27,383
28,890
29,314
30,702
29,780
33,170
Carbon Black
3,381
4,269
3,310
3,440
3,300
2,610
3,000
Ethylene
13,126
19,024
20,000
19,400
20,700
20,700
22,800
Ethylene Dichloride
254
455
412
440
503
456
376
Ethylene Oxide
1,123
1,489
1,250
1,300
1,370
1,680
1,930
Acrylonitrile
1,214
1,325
1,040
1,250
990
850
850
Methanol
2,513
821
2,878
3,484
3,839
3,484
4,214
ch4
9
3
10
12
13
12
15
Acrylonitrile
+
+
+
+
+
+
+
Methanol
9
3
10
12
13
12
14
+ Does not exceed 0.5 kt CH4.
Note: Totals by gas may not sum due to independent rounding.
4-60 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Methodology and Time-Series Consistency
Emissions of CO2 and Cm were calculated using the estimation methods provided by the 2006IPCC Guidelines and
country-specific methods from EPA's GHGRP. The 2006 IPCC Guidelines Tier 1 method was used to estimate CO2
and Cm emissions from production of acrylonitrile and methanol,42 and a country-specific approach similar to the
IPCC Tier 2 method was used to estimate CO2 emissions from production of carbon black, ethylene oxide, ethylene,
and ethylene dichloride. The Tier 2 method for petrochemicals is a total feedstock carbon (C) mass balance
method used to estimate total CO2 emissions, but it is not applicable for estimating CH4 emissions.
As noted in the 2006 IPCC Guidelines, the total feedstock C mass balance method (Tier 2) is based on the
assumption that all of the C input to the process is converted either into primary and secondary products or into
CO2. Further, the guideline states that while the total C mass balance method estimates total C emissions from the
process, it does not directly provide an estimate of the amount of the total C emissions emitted as CO2, CH4, or
non-CI-U volatile organic compounds (NMVOCs). This method accounts for all the C as CO2, including CH4.
Note, a small subset of facilities reporting under EPA's GHGRP use Continuous Emission Monitoring Systems
(CEMS) to monitor CO2 emissions from process vents and/or stacks from stationary combustion units. These
facilities are required to also report CO2, CH4 and N2O emissions from combustion of process off-gas in flares. The
CO2 emissions from flares are included in aggregated CO2 results. Preliminary analysis of aggregated annual reports
shows that flared CH4 and N2O emissions are less than 500 kt CO2 Eq./year. EPA's GHGRP team is still reviewing
these data across reported years, and EPA plans to address this more completely in future reports.
Carbon Black, Ethylene, Ethylene Dichloride, and Ethylene Oxide
2010 through 2021
Carbon dioxide emissions and national production were aggregated directly from EPA's GHGRP dataset for 2010
through 2021 (EPA 2022). In 2021, data reported to the GHGRP included CO2 emissions of 3,000,000 metric tons
from carbon black production; 22,800,000 metric tons of CChfrom ethylene production; 376,000 metric tons of
CO2 from ethylene dichloride production; and 1,930,000 metric tons of CO2 from ethylene oxide production. These
emissions reflect application of a country-specific approach similar to the IPCC Tier 2 method and were used to
estimate CO2 emissions from the production of carbon black, ethylene, ethylene dichloride, and ethylene oxide.
Since 2010, EPA's GHGRP, under Subpart X, requires all domestic producers of petrochemicals to report annual
emissions and supplemental emissions information (e.g., production data, etc.) to facilitate verification of reported
emissions. Under EPA's GHGRP, most petrochemical production facilities are required to use either a mass balance
approach or CEMS to measure and report emissions for each petrochemical process unit to estimate facility-level
process CO2 emissions; ethylene production facilities also have a third option. The mass balance method is used by
most facilities43 and assumes that all the carbon input is converted into primary and secondary products,
byproducts, or is emitted to the atmosphere as CO2. To apply the mass balance, facilities must measure the volume
or mass of each gaseous and liquid feedstock and product, mass rate of each solid feedstock and product, and
carbon content of each feedstock and product for each process unit and sum for their facility. To apply the
optional combustion methodology, ethylene production facilities must measure the quantity, carbon content, and
molecular weight of the fuel to a stationary combustion unit when that fuel includes any ethylene process off-gas.
These data are used to calculate the total CO2 emissions from the combustion unit. The facility must also estimate
the fraction of the emissions that is attributable to burning the ethylene process off-gas portion of the fuel. This
42 EPA has not integrated aggregated facility-level GHGRP information for acrylonitrile and methanol production. The
aggregated information associated with production of these petrochemicals did not meet criteria to shield underlying CBI from
public disclosure.
43 A few facilities producing ethylene dichloride, ethylene, and methanol used C02 CEMS; those C02 emissions have been
included in the aggregated GHGRP emissions presented here.
Industrial Processes and Product Use 4-61
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
fraction is multiplied by the total emissions to estimate the emissions from ethylene production. The QA/QC and
Verification section below has a discussion of non-CC>2 emissions from ethylene production facilities.
All non-energy uses of residual fuel and some non-energy uses of "other oil" are assumed to be used in the
production of carbon black; therefore, consumption of these fuels is adjusted for within the Energy chapter to
avoid double-counting of emissions from fuel used in the carbon black production presented here within IPPU
sector. Additional information on the adjustments made within the Energy sector for Non-Energy Use of Fuels is
described in both the Methodology section of CChfrom Fossil Fuel Combustion (3.1 Fossil Fuel Combustion (IPCC
Source Category 1A)) and Annex 2.1, Methodology for Estimating Emissions of CO2 from Fossil Fuel Combustion.
1990 through 2009
Prior to 2010, for each of these 4 types of petrochemical processes, an average national CO2 emission factor was
calculated based on the GHGRP data and applied to production for earlier years in the time series (i.e., 1990
through 2009) to estimate CO2 emissions from carbon black, ethylene, ethylene dichloride, and ethylene oxide
production. For carbon black, ethylene, ethylene dichloride, and ethylene oxide carbon dioxide emission factors
were derived from EPA's GHGRP data by dividing annual CO2 emissions for petrochemical type "\" with annual
production for petrochemical type "i" and then averaging the derived emission factors obtained for each calendar
year 2010 through 2013 (EPA 2019). The years 2010 through 2013 were used in the development of carbon dioxide
emission factors as these years are more representative of operations in 1990 through 2009 for these facilities.
The average emission factors for each petrochemical type were applied across all prior years because
petrochemical production processes in the United States have not changed significantly since 1990, though some
operational efficiencies have been implemented at facilities over the time series.
The average country-specific CO2 emission factors that were calculated from the GHGRP data are as follows:
• 2.59 metric tons CCh/metric ton carbon black produced
• 0.79 metric tons CCh/metric ton ethylene produced
• 0.040 metric tons CCh/metric ton ethylene dichloride produced
• 0.46 metric tons CCh/metric ton ethylene oxide produced
Annual production data for carbon black for 1990 through 2009 were obtained from the International Carbon
Black Association (Johnson 2003 and 2005 through 2010). Annual production data for ethylene, ethylene
dichloride, and ethylene oxide for 1990 through 2009 were obtained from the American Chemistry Council's
(ACC's) Business of Chemistry (ACC 2022a).
Acrylonitrile
Carbon dioxide and methane emissions from acrylonitrile production were estimated using the Tier 1 method in
the 2006 IPCC Guidelines. Annual acrylonitrile production data were used with IPCC default Tier 1CO2 and CH4
emission factors to estimate emissions for 1990 through 2021. Emission factors used to estimate acrylonitrile
production emissions are as follows:
• 0.18 kg CHVmetric ton acrylonitrile produced
• 1.00 metric tons CCh/metric ton acrylonitrile produced
Annual acrylonitrile production data for 1990 through 2021 were obtained from ACC's Business of Chemistry (ACC
2022a). EPA is not able to apply the aggregated facility-level GHGRP information for acrylonitrile production
needed for a Tier 2 approach. The aggregated information associated with production of these petrochemicals did
not meet criteria to shield underlying CBI from public disclosure.
Methanol
Carbon dioxide and methane emissions from methanol production were estimated using the Tier 1 method in the
2006 IPCC Guidelines. Annual methanol production data were used with IPCC default Tier 1CO2 and CH4 emission
4-62 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
factors to estimate emissions for 1990 through 2021. Emission factors used to estimate methanol production
emissions are as follows:
• 2.3 kg Cm/metric ton methanol produced
• 0.67 metric tons CCh/metric ton methanol produced
Annual methanol production data for 1990 through 2021 were obtained from the ACC's Business of Chemistry (ACC
2022a, ACC 2022b). EPA is not able to apply the aggregated facility-level GHGRP information for methanol
production needed for a Tier 2 approach. The aggregated information associated with production of these
petrochemicals did not meet criteria to shield underlying CBI from public disclosure.
Table 4-48: Production of Selected Petrochemicals (kt)
Chemical
1990
2005
2017
2018
2019
2020
2021
Carbon Black
1,307
1,651
1,240
1,280
1,210
990
1,140
Ethylene
16,542
23,975
27,800
30,500
32,400
33,500
34,700
Ethylene Dichloride
6,283
11,260
12,400
12,500
12,600
11,900
11,500
Ethylene Oxide
2,429
3,220
3,350
3,310
3,800
4,680
4,860
Acrylonitrile
1,214
1,325
1,040
1,250
990
850
850
Methanol
3,750
1,225
4,295
5,200
5,730
5,200
6,290
As noted earlier in the introduction section of the Petrochemical Production section, the allocation and reporting
of emissions from both fuels and feedstocks transferred out of the system for use in energy purposes to the Energy
chapter differs slightly from the 2006IPCC Guidelines. According to the 2006IPCC Guidelines, emissions from fuel
combustion from petrochemical production should be allocated to this source category within the IPPU chapter.
Due to national circumstances, EIA data on primary fuel for feedstock use within the energy balance are presented
by commodity only, with no resolution on data by industry sector (i.e., petrochemical production). In addition,
under EPA's GHGRP, reporting facilities began reporting in 2014 on annual feedstock quantities for mass balance
and CEMS methodologies (79 FR 63794), as well as the annual average carbon content of each feedstock (and
molecular weight for gaseous feedstocks) for the mass balance methodology beginning in reporting year 2017 (81
FR 89260).44 The United States is currently unable to report non-energy fuel use from petrochemical production
under the IPPU chapter due to CBI issues. Therefore, consistent with 2006 IPCC Guidelines, fuel consumption data
reported by EIA are modified to account for these overlaps to avoid double-counting. More information on the
non-energy use of fossil fuel feedstocks for petrochemical production can be found in Annex 2.3.
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2021. The methodology for ethylene production, ethylene dichloride production, and ethylene oxide
production spliced activity data from two different sources: ACC for 1990 through 2009 and GHGRP for 2010
through 2021. Consistent with the 2006 IPCC Guidelines, the overlap technique was applied to compare the two
data sets for years where there was overlap. For ethylene production, the data sets were determined to be
consistent, and adjustments were not needed. For ethylene dichloride production and ethylene oxide production,
the data sets were determined to be inconsistent. The GHGRP data includes production of ethylene dichloride and
ethylene oxide as intermediates while it is unclear if the ACC data does; therefore, no adjustments were made to
the ethylene dichloride and ethylene oxide activity data for 1990 through 2009 because the 2006 IPCC Guidelines
indicate that it is not good practice to use the overlap technique when the data sets are inconsistent. The
methodology for carbon black production also spliced activity data from two different sources: ICBA for 1990
through 2009 and GHGRP for 2010 through 2021. The overlap technique was applied to these data for 2010 and
2011. The data sets were determined to be consistent, and adjustments were not needed.
44 See https://www.epa.gov/ghgreporting/historical-rulemakings.
Industrial Processes and Product Use 4-63
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Uncertainty-TO BE UPDATED FOR FINAL INVENTORY REPORT
The CO2 and CH4 emission factors used for methanol and acrylonitrile production are based on a limited number of
studies. Using plant-specific factors instead of default or average factors could increase the accuracy of the
emission estimates; however, such data were not available for the current Inventory report. For methanol, EPA
assigned an uncertainty range of ±30 percent for the CO2 emission factor and -80 percent to +30 percent for the
Cm emission factor, consistent with the ranges in Table 3.27 of the 2006IPCC Guidelines. For acrylonitrile, EPA
assigned an uncertainty range of ±60 percent for the CO2 emission factor and ±10 percent for the CH4 emission
factor, consistent with the ranges in Table 3.27 of the 2006 IPCC Guidelines. The results of the quantitative
uncertainty analysis for the CO2 emissions from carbon black production, ethylene, ethylene dichloride, and
ethylene oxide are based on reported GHGRP data. Refer to the Methodology section for more details on how
these emissions were calculated and reported to EPA's GHGRP. EPA assigned CO2 emissions from carbon black,
ethylene, ethylene dichloride, and ethylene oxide production an uncertainty range of ±5 percent, consistent with
the ranges in Table 3.27 of the 2006 IPCC Guidelines. In the absence of other data, these values have been
assessed as reasonable. There is some uncertainty in the applicability of the average emission factors for each
petrochemical type across all prior years. While petrochemical production processes in the United States have not
changed significantly since 1990, some operational efficiencies have been implemented at facilities over the time
series.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-49. Petrochemical
production CO2 emissions from 2020 were estimated to be between 28.4 and 31.7 MMT CO2 Eq. at the 95 percent
confidence level. This indicates a range of approximately 5 percent below to 6 percent above the emission
estimate of 30.0 MMT CO2 Eq. Petrochemical production CH4 emissions from 2020 were estimated to be between
0.11 and 0.39 MMT CO2 Eq. at the 95 percent confidence level. This indicates a range of approximately 57 percent
below to 47 percent above the emission estimate of 0.3 MMT CO2 Eq.
Table 4-49: Approach 2 Quantitative Uncertainty Estimates for ChU Emissions from
Petrochemical Production and CO2 Emissions from Petrochemical Production (MMT CO2 Eq.
and Percent)
Source
Gas
2021 Emission Estimate
(MMT C02 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Petrochemical
Production
Petrochemical
Production
C02
30.0
28.4
31.7
-5%
+6%
ch4
0.3
0.11
0.39
-57%
+47%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
QA/QC and Verification
For Petrochemical Production, QA/QC activities were conducted consistent with the U.S. Inventory QA/QC plan, as
described in the QA/QC and Verification Procedures section of the IPPU chapter and Annex 8. Source-specific
quality control measures for this category included the QA/QC requirements and verification procedures of EPA's
GHGRP. More details on the greenhouse gas calculation, monitoring and QA/QC methods applicable to
petrochemical facilities can be found under Subpart X (Petrochemical Production) of the regulation (40 CFR Part
98).45 EPA verifies annual facility-level GHGRP reports through a multi-step process (e.g., combination of electronic
checks and manual reviews) to identify potential errors and ensure that data submitted to EPA are accurate,
45 See http://www.ecfr.gov/cgi-bin/text-idx7tph/ecfrbrowse/Title40/40cfr98 main Q2.tpl.
4-64 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
complete, and consistent (EPA 2015).46 Based on the results of the verification process, EPA follows up with
facilities to resolve mistakes that may have occurred. The post-submittals checks are consistent with a number of
general and category-specific QC procedures, including range checks, statistical checks, algorithm checks, and year-
to-year checks of reported data and emissions. EPA also conducts QA checks of GHGRP reported production data
by petrochemical type against external datasets.
For ethylene, ethylene dichloride, and ethylene oxide, it is possible to compare CO2 emissions calculated using the
GHGRP data to the CO2 emissions that would have been calculated using the Tier 1 approach if GHGRP data were
not available. For ethylene, the GHGRP emissions were within 5 percent of the emissions calculated using the Tier
1 approach prior to 2017; in 2017 through 2021, the GHGRP emissions have been between 7 percent and 18
percent lower than what would be calculated using the Tier 1 approach. For ethylene dichloride, the GHGRP
emissions are typically higher than the Tier 1 emissions by up to 25 percent, but in 2021, GHGRP emissions were a
few percentage points lower than the Tier 1 emissions. For ethylene oxide, GHGRP emissions typically vary from
the Tier 1 emissions by up to ±20 percent, but in 2021, the GHGRP emissions are significantly higher than the Tier 1
emissions. This is likely due to GHGRP data capturing the production of ethylene oxide as an intermediate in the
onsite production of ethylene glycol.
EPA's GHGRP mandates that all petrochemical production facilities report their annual emissions of CO2, CH4, and
N2O from each of their petrochemical production processes. Source-specific quality control measures for the
Petrochemical Production category included the QA/QC requirements and verification procedures of EPA's GHGRP.
The QA/QC requirements differ depending on the calculation methodology used.
As part of a planned improvement effort, EPA has assessed the potential of using GHGRP data to estimate CH4
emissions from ethylene production. As discussed in the Methodology section above, CO2 emissions from ethylene
production in this chapter are based on data reported under the GHGRP, and these emissions are calculated using
a Tier 2 approach that assumes all of the carbon in the fuel (i.e., ethylene process off-gas) is converted to CO2.
Ethylene production facilities also calculate and report CH4 emissions under the GHGRP when they use the optional
combustion methodology. The facilities calculate CH4 emissions from each combustion unit that burns off-gas from
an ethylene production process unit using a Tier 1 approach based on the total quantity of fuel burned, a default
higher heating value, and a default emission factor. Because multiple other types of fuel in addition to the ethylene
process unit off-gas may be burned in these combustion units, the facilities also report an estimate of the fraction
of emissions that is due to burning the ethylene process off-gas component of the total fuel. Multiplying the total
emissions by the estimated fraction provides an estimate of the CH4 emissions from the ethylene production
process unit. These ethylene production facilities also calculate CH4 emissions from flares that burn process vent
emissions from ethylene processes. The emissions are calculated using either a Tier 2 approach based on
measured gas volumes and measured carbon content or higher heating value, or a Tier 1 approach based on the
measured gas flow and a default emission factor. Nearly all ethylene production facilities use the optional
combustion methodology under the GHGRP, and the sum of reported CH4 emissions from combustion in stationary
combustion units and flares at all of these facilities is on the same order of magnitude as the combined CH4
emissions presented in this chapter from methanol and acrylonitrile production. The CH4 emissions from ethylene
production under the GHGRP have not been included in this chapter because this approach double counts carbon
(i.e., all of the carbon in the CH4 emissions is also included in the CO2 emissions from the ethylene process units).
EPA continues to assess the GHGRP data for ways to better disaggregate the data and incorporate it into the
inventory.
These facilities are also required to report emissions of N2O from combustion of ethylene process off-gas in both
stationary combustion units and flares. Facilities using CEMS (consistent with a Tier 3 approach) are also required
to report emissions of CH4 and N2O from combustion of petrochemical process-off gases in flares. Preliminary
analysis of the aggregated reported CH4 and N2O emissions from facilities using CEMS and N2O emissions from
facilities using the optional combustion methodology suggests that these annual emissions are less than 0.4
percent of total petrochemical emissions, which is not significant enough to prioritize for inclusion in the report at
46 See https://www.epa.gov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
Industrial Processes and Product Use 4-65
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
this time. Pending resources and significance, EPA may include these N2O emissions in future reports to enhance
completeness.
Future QC efforts to validate the use of Tier 1 default emission factors and report on the comparison of Tier 1
emission estimates and GHGRP data are described below in the Planned Improvements section.
Recalculations Discussion
The acrylonitrile and methanol production quantities for 2020 were updated with the revised values in ACC's
Business of Chemistry (ACC 2022a, ACC 2022b). These changes resulted in a 0.8 percent (240 kt) decrease in total
petrochemical CO2 Eq. emissions for 2020, compared to the previous Inventory.
In addition, for the current Inventory, CC>2-equivalent estimates of total CH4 emissions from acrylonitrile and
methanol production have been revised to reflect the 100-year global warming potentials (GWPs) provided in the
IPCC Fifth Assessment Report (AR5) (IPCC 2013). AR5 GWP values differ slightly from those presented in the IPCC
Fourth Assessment Report (AR4) (IPCC 2007) (used in the previous inventories). The AR5 GWPs have been applied
across the entire time series for consistency. The GWP of CH4 increased from 25 to 28, leading to an overall
increase in estimates for CC>2-equivalent CH4 emissions. Compared to the previous Inventory, which applied 100-
year GWP values from AR4, annual Cm emissions increased by 12 percent each year, ranging from an increase of
5.4 kt CO2 Eq. in 2011 to 42.1 kt CO2 Eq. in 1997. The net impact on the entire category from these updates was an
average annual 0.1 percent increase in emissions for the time series. Further discussion on this update and the
overall impacts of updating the Inventory GWP values to reflect the IPCC Fifth Assessment Report can be found in
Chapter 9, Recalculations and Improvements.
Planned Improvements
Improvements include completing category-specific QC of activity data and emission factors, along with further
assessment of CH4 and N2O emissions to enhance completeness in reporting of emissions from U.S. petrochemical
production, pending resources, significance and time-series consistency considerations. For example, EPA is
planning additional assessment of ways to use CH4 data from the GHGRP in the Inventory. One possible approach
EPA is assessing would be to adjust the CO2 emissions from the GHGRP downward by subtracting the carbon that is
also included in the reported CH4 emissions, per the discussion in the Petrochemical Production QA/QC and
Verification section, above. As of this current report, timing and resources have not allowed EPA to complete this
analysis of activity data, emissions, and emission factors and remains a priority improvement within the IPPU
chapter.
Pending resources, a secondary potential improvement for this source category would focus on continuing to
analyze the fuel and feedstock data from EPA's GHGRP to better disaggregate energy-related emissions and
allocate them more accurately between the Energy and IPPU sectors of the Inventory. It is important to ensure no
double counting of emissions between fuel combustion, non-energy use of fuels, and industrial process emissions.
For petrochemical feedstock production, EPA review of the categories suggests this is not a significant issue since
the non-energy use industrial release data includes different categories of sources and sectors than those included
in the IPPU emissions category for petrochemicals. As noted previously in the methodology section, data
integration is not available at this time because feedstock data from the EIA used to estimate non-energy uses of
fuels are aggregated by fuel type, rather than disaggregated by both fuel type and particular industries. Also,
GHGRP-reported data on quantities of fuel consumed as feedstocks by petrochemical producers are unable to be
used due to the data failing GHGRP CBI aggregation criteria. EPA will continue to look for ways to incorporate this
data into future Inventories that will allow for easier data integration between the non-energy uses of fuels
category and the petrochemicals category presented in this chapter. This planned improvement is still under
development and has not been completed to report on progress in this current Inventory.
4-66 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
4.14 HCFC-22 Production (CRF Source
Category 2B9a)
Trifluoromethane (HFC-23 or CHF3) is generated as a byproduct during the manufacture of chlorodifluoromethane
(HCFC-22), which is primarily employed in refrigeration and air conditioning systems and as a chemical feedstock
for manufacturing synthetic polymers. Between 1990 and 2000, U.S. production of HCFC-22 increased significantly
as HCFC-22 replaced chlorofluorocarbons (CFCs) in many applications. Between 2000 and 2007, U.S. production
fluctuated but generally remained above 1990 levels. In 2008 and 2009, U.S. production declined markedly and has
remained near 2009 levels since. Because HCFC-22 depletes stratospheric ozone, its production for non-feedstock
uses was phased out in 2020 under the U.S. Clean Air Act.47 Feedstock production, however, is permitted to
continue indefinitely.
HCFC-22 is produced by the reaction of chloroform (CHCb) and hydrogen fluoride (HF) in the presence of a catalyst,
SbCls. The reaction of the catalyst and HF produces SbClxFy, (where x + y = 5), which reacts with chlorinated
hydrocarbons to replace chlorine atoms with fluorine. The HF and chloroform are introduced by submerged piping
into a continuous-flow reactor that contains the catalyst in a hydrocarbon mixture of chloroform and partially
fluorinated intermediates. The vapors leaving the reactor contain HCFC-21 (CHCbF), HCFC-22 (CHCIF2), HFC-23
(CHF3), HCI, chloroform, and HF. The under-fluorinated intermediates (HCFC-21) and chloroform are then
condensed and returned to the reactor, along with residual catalyst, to undergo further fluorination. The final
vapors leaving the condenser are primarily HCFC-22, HFC-23, HCI and residual HF. The HCI is recovered as a useful
byproduct, and the HF is removed. Once separated from HCFC-22, the HFC-23 may be released to the atmosphere,
recaptured for use in a limited number of applications, or destroyed.
Two facilities produced HCFC-22 in the United States in 2021. Emissions of HFC-23 from this activity in 2021 were
estimated to be 2.2 MMT CO2 Eq. (0.1 kt) (see Table 4-50). This quantity represents a 27 percent increase from
2020 emissions and a 94 percent decrease from 1990 emissions. The decrease from 1990 emissions was caused
primarily by changes in the HFC-23 emission rate (kg HFC-23 emitted/kg HCFC-22 produced). The increase from
2020 emissions was caused by both an increase in the HFC-23 emission rate at one plant and an increase in the
total quantity of HCFC-22 produced. The long-term decrease in the emission rate is primarily attributable to six
factors: (a) five plants that did not capture and destroy the HFC-23 generated have ceased production of HCFC-22
since 1990; (b) one plant that captures and destroys the HFC-23 generated began to produce HCFC-22; (c) one
plant implemented and documented a process change that reduced the amount of HFC-23 generated; (d) the
same plant began recovering HFC-23, primarily for destruction and secondarily for sale; (e) another plant began
destroying HFC-23; and (f) the same plant, whose emission rate was higher than that of the other two plants,
ceased production of HCFC-22 in 2013.
Table 4-50: HFC-23 Emissions from HCFC-22 Production (MMT COz Eq. and kt HFC-23)
Year
1990
2005
2017
2018
2019
2020
2021
MMTCOz Eq.
38.6
16.8
4.3
2.7
3.1
1.8
2.2
kt HFC-23
3
1
0.3
0.2
0.3
0.1
0.2
Methodology and Time-Series Consistency
To estimate HFC-23 emissions for five of the eight HCFC-22 plants that have operated in the United States since
1990, methods comparable to the Tier 3 methods in the 2006IPCC Guidelines (IPCC 2006) were used throughout
the time series. Emissions for 2010 through 2021 were obtained through reports submitted by U.S. HCFC-22
47 As construed, interpreted, and applied in the terms and conditions of the Montreal Protocol on Substances that Deplete the
Ozone Layer [42 U.S.C. §7671m(b), CAA §614],
Industrial Processes and Product Use 4-67
-------
1 production facilities to EPA's Greenhouse Gas Reporting Program (GHGRP). EPA's GHGRP mandates that all HCFC-
2 22 production facilities report their annual emissions of HFC-23 from HCFC-22 production processes and HFC-23
3 destruction processes. Previously, data were obtained by EPA through collaboration with an industry association
4 that received voluntarily reported HCFC-22 production and HFC-23 emissions annually from all U.S. HCFC-22
5 producers from 1990 through 2009. These emissions were aggregated and reported to EPA on an annual basis.
6 For the other three plants, the last of which closed in 1993, methods comparable to the Tier 1 method in the 2006
1 IPCC Guidelines were used. Emissions from these three plants have been calculated using the recommended
8 emission factor for unoptimized plants operating before 1995 (0.04 kg HCFC-23/kg HCFC-22 produced).
9 The five plants that have operated since 1994 measure (or, for the plants that have since closed, measured)
10 concentrations of HFC-23 as well as mass flow rates of process streams to estimate their generation of HFC-23.
11 Plants using thermal oxidation to abate their HFC-23 emissions monitor the performance of their oxidizers to verify
12 that the HFC-23 is almost completely destroyed. One plant that releases a small fraction of its byproduct HFC-23
13 periodically measures HFC-23 concentrations at process vents using gas chromatography. This information is
14 combined with information on quantities of products (e.g., HCFC-22) to estimate HFC-23 emissions.
15 To estimate 1990 through 2009 emissions, reports from an industry association were used that aggregated HCFC-
16 22 production and HFC-23 emissions from all U.S. HCFC-22 producers and reported them to EPA (ARAP 1997,1999,
17 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, and 2010). To estimate 2010 through 2021
18 emissions, facility-level data (including both HCFC-22 production and HFC-23 emissions) reported through EPA's
19 GHGRP were analyzed. In 1997 and 2008, comprehensive reviews of plant-level estimates of HFC-23 emissions and
20 HCFC-22 production were performed (RTI1997; RTI 2008). The 1997 and 2008 reviews enabled U.S. totals to be
21 reviewed, updated, and where necessary, corrected, and also for plant-level uncertainty analyses (Monte-Carlo
22 simulations) to be performed for 1990,1995, 2000, 2005, and 2006. Estimates of annual U.S. HCFC-22 production
23 are presented inTable 4-51.
24 Table 4-51: HCFC-22 Production (kt)
Year
1990
2005
2012
2017
2018
2019
2020
2021
Production
139
156
96
C
C
C
C
C
C(CBI)
Note: HCFC-22 production in 2013 through 2020 is considered Confidential Business
Information (CBI) as there were only two producers of HCFC-22 in those years.
25 Uncertainty
26 The uncertainty analysis presented in this section was based on a plant-level Monte Carlo Stochastic Simulation for
27 2006. The Monte Carlo analysis used estimates of the uncertainties in the individual variables in each plant's
28 estimating procedure. This analysis was based on the generation of 10,000 random samples of model inputs from
29 the probability density functions for each input. A normal probability density function was assumed for all
30 measurements and biases except the equipment leak estimates for one plant; a log-normal probability density
31 function was used for this plant's equipment leak estimates. The simulation for 2006 yielded a 95-percent
32 confidence interval for U.S. emissions of 6.8 percent below to 9.6 percent above the reported total.
33 The relative errors yielded by the Monte Carlo Stochastic Simulation for 2006 were applied to the U.S. emission
34 estimate for 2021. The resulting estimates of absolute uncertainty are likely to be reasonably accurate because (1)
35 the methods used by the two remaining plants to estimate their emissions are not believed to have changed
36 significantly since 2006, and (2) although the distribution of emissions among the plants has changed between
37 2006 and 2021 (because one plant has closed), the plant that currently accounts for most emissions had a relative
38 uncertainty in its 2006 (as well as 2005) emissions estimate that was similar to the relative uncertainty for total
39 U.S. emissions. Thus, the closure of one plant is not likely to have a large impact on the uncertainty of the national
40 emission estimate.
41 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-52. HFC-23 emissions
42 from HCFC-22 production were estimated to be between 2.1 and 2.5 MMT CO2 Eq. at the 95 percent confidence
4-68 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 level. This indicates a range of approximately 7 percent below and 10 percent above the emission estimate of 2.2
2 MMT CO2 Eq.
3 Table 4-52: Approach 2 Quantitative Uncertainty Estimates for HFC-23 Emissions from
4 HCFC-22 Production (MMT CO2 Eq. and Percent)
Source
2021 Emission Estimate
Gas
(MMT CO? Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
HCFC-22 Production
HFC-23 2.2
2.1 2.5
-7% +10%
a Range of emissions reflects a 95 percent confidence interval.
5 QA/QC and Verification
6 General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S. Inventory
7 QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006IPCC Guidelines as described in the
8 introduction of the IPPU chapter (see Annex 8 for more details). Under the GHGRP, EPA verifies annual facility-level
9 reports through a multi-step process (e.g., including a combination of pre-and post-submittal electronic checks and
10 manual reviews by staff) to identify potential errors and ensure that data submitted to EPA are accurate,
11 complete, and consistent (EPA 20 15).48 Based on the results of the verification process, EPA follows up with
12 facilities to resolve mistakes that may have occurred. The post-submittals checks are consistent with a number of
13 general and category-specific QC procedures, including: range checks, statistical checks, algorithm checks, and
14 year-to-year checks of reported data and emissions.
15 The GHGRP also requires source-specific quality control measures for the HCFC-22 Production category. Under
16 EPA's GHGRP, HCFC-22 producers are required to (1) measure concentrations of HFC-23 and HCFC-22 in the
17 product stream at least weekly using equipment and methods (e.g., gas chromatography) with an accuracy and
18 precision of 5 percent or better at the concentrations of the process samples, (2) measure mass flows of HFC-23
19 and HCFC-22 at least weekly using measurement devices (e.g., flowmeters) with an accuracy and precision of 1
20 percent of full scale or better, (3) calibrate mass measurement devices at the frequency recommended by the
21 manufacturer using traceable standards and suitable methods published by a consensus standards organization,
22 (4) calibrate gas chromatographs at least monthly through analysis of certified standards, and (5) document these
23 calibrations.
24 Recalculations
25 For the current Inventory, the C02-equivalent estimates of total HFC-23 emissions from HCFC-22 production have
26 been revised to reflect the 100-year global warming potential (GWP) for HFC-23 provided in the IPCC Fifth
27 Assessment Report (AR5) (IPCC 2013). AR5 GWP values differ slightly from those presented in the IPCC Fourth
28 Assessment Report (AR4) (IPCC 2007) (used in the previous inventories). The AR5 GWP has been applied across the
29 entire time series for consistency. With this change, the GWP of HFC-23 has decreased from 14,800 to 12,400,
30 leading to a decrease of 16 percent in C02-equivalent HFC-23 emissions in every year compared to the previous
31 inventory. Further discussion on this update and the overall impacts of updating the inventory GWPs to reflect the
32 IPCC Fifth Assessment Report can be found in Chapter 9, Recalculations and Improvements.
48 EPA (2015). Greenhouse Gas Reporting Program Report Verification. Available online at:
https://www.epa.gov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
Industrial Processes and Product Use 4-69
-------
1 4.15 Carbon Dioxide Consumption (CRF
2 Source Category 2BIO)
3 Carbon dioxide (CO2) is used for a variety of commercial applications, including food processing, chemical
4 production, carbonated beverage production, and refrigeration, and is also used in petroleum production for
5 enhanced oil recovery (EOR). CO2 used for EOR is injected underground to enable additional petroleum to be
6 produced. For the purposes of this analysis, CO2 used in food and beverage applications is assumed to be emitted
7 to the atmosphere. A further discussion of CO2 used in EOR is described in the Energy chapter in Box 3-6 titled
8 "Carbon Dioxide Transport, Injection, and Geological Storage" and is not included in this section.
9 Carbon dioxide is produced from naturally-occurring CO2 reservoirs, as a byproduct from the energy and industrial
10 production processes (e.g., ammonia production, fossil fuel combustion, ethanol production), and as a byproduct
11 from the production of crude oil and natural gas, which contain naturally occurring CO2 as a component.
12 In 2021, the amount of CO2 produced and captured for commercial applications and subsequently emitted to the
13 atmosphere was 5.0 MMT CO2 Eq. (4,990 kt) (see Table 4-53). This is less than a 1 percent increase (20 kt) from
14 2020 levels and is an increase of approximately 239 percent since 1990.
15 Table 4-53: CO2 Emissions from CO2 Consumption (MMT CO2 Eq. and kt)
Year
1990
2005
2017
2018
2019
2020
2021
MMTCOz Eq.
1.5
1.4
4.6
4.1
4.9
5.0
5.0
kt
1,472
1,375
4,580
4,130
4,870
4,970
4,990
is Methodology and Time-Series Consistency
17 Carbon dioxide emission estimates for 1990 through 2021 were based on the quantity of CO2 extracted and
18 transferred for industrial applications (i.e., non-EOR end-uses). Some of the CO2 produced by these facilities is used
19 for EOR, and some is used in other commercial applications (e.g., chemical manufacturing, food and beverage).
20 2010 through 2021
21 For 2010 through 2021, data from EPA's GHGRP (Subpart PP) were aggregated from facility-level reports to
22 develop a national-level estimate for use in the Inventory (EPA 2022). Facilities report CO2 extracted or produced
23 from natural reservoirs and industrial sites, and CO2 captured from energy and industrial processes and transferred
24 to various end-use applications to EPA's GHGRP. This analysis includes only reported CO2 transferred to food and
25 beverage end-uses. EPA is continuing to analyze and assess integration of CO2 transferred to other end-uses to
26 enhance the completeness of estimates under this source category. Other end-uses include industrial applications,
27 such as metal fabrication. EPA is analyzing the information reported to ensure that other end-use data excludes
28 non-emissive applications and publication will not reveal CBI. Additionally, a small amount of CO2 is used as a
29 refrigerant; use and emissions from this application are reported under Section 4.24 Substitution of Ozone
30 Depleting Substances (CRF Source Category 2F). Reporters subject to EPA's GHGRP Subpart PP are also required to
31 report the quantity of CO2 that is imported and/or exported. Currently, these data are not publicly available
32 through the GHGRP due to data confidentiality reasons and hence are excluded from this analysis.
33 Facilities subject to Subpart PP of EPA's GHGRP are required to measure CO2 extracted or produced. More details
34 on the calculation and monitoring methods applicable to extraction and production facilities can be found under
35 Subpart PP: Suppliers of Carbon Dioxide of the regulation, Part 98.49 The number of facilities that reported data to
49 See http://www.ecfr.gov/cgi-bin/text-idx7tpk/ecfrbrowse/Title40/40cfr98 main Q2.tpl.
4-70 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 EPA's GHGRP Subpart PP (Suppliers of Carbon Dioxide) for 2010 through 2021 is much higher (ranging from 44 to
2 53) than the number of facilities included in the Inventory for the 1990 to 2009 time period prior to the availability
3 of GHGRP data (4 facilities). The difference is largely due to the fact the 1990 to 2009 data includes only CO2
4 transferred to end-use applications from naturally occurring CO2 reservoirs and excludes industrial sites.
5 1990 through 2009
6 For 1990 through 2009, data from EPA's GHGRP are not available. For this time period, CO2 production data from
7 four naturally-occurring CO2 reservoirs were used to estimate annual CO2 emissions. These facilities were Jackson
8 Dome in Mississippi, Bravo and West Bravo Domes in New Mexico, and McCallum Dome in Colorado. The facilities
9 in Mississippi and New Mexico produced CO2 for use in both EOR and in other commercial applications (e.g.,
10 chemical manufacturing, food production). The fourth facility in Colorado (McCallum Dome) produced CO2 for
11 commercial applications only (New Mexico Bureau of Geology and Mineral Resources 2006).
12 Carbon dioxide production data and the percentage of production that was used for non-EOR applications for the
13 Jackson Dome, Mississippi facility were obtained from Advanced Resources International (ARI 2006, 2007) for 1990
14 to 2000, and from the Annual Reports of Denbury Resources (Denbury Resources 2002 through 2010) for 2001 to
15 2009 (see Table 4-54). Denbury Resources reported the average CO2 production in units of MMCF CO2 per day for
16 2001 through 2009 and reported the percentage of the total average annual production that was used for EOR.
17 Production from 1990 to 1999 was set equal to 2000 production, due to lack of publicly available production data
18 for 1990 through 1999. Carbon dioxide production data for the Bravo Dome and West Bravo Dome were obtained
19 from ARI for 1990 through 2009 (ARI 1990 to 2010). Data for the West Bravo Dome facility were only available for
20 2009. The percentage of total production that was used for non-EOR applications for the Bravo Dome and West
21 Bravo Dome facilities for 1990 through 2009 were obtained from New Mexico Bureau of Geology and Mineral
22 Resources (Broadhead 2003; New Mexico Bureau of Geology and Mineral Resources 2006). Production data for the
23 McCallum Dome (Jackson County), Colorado facility were obtained from the Colorado Oil and Gas Conservation
24 Commission (COGCC) for 1999 through 2009 (COGCC 2014). Production data for 1990 to 1998 and percentage of
25 production used for EOR were assumed to be the same as for 1999, due to lack of publicly available data.
26 Table 4-54: CO2 Production (kt CO2) and the Percent Used for Non-EOR Applications
Total C02
Jackson Dome,
Bravo Dome,
West Bravo Dome,
McCallum Dome,
Production
MS
NM
NM
CO
from Extraction
%
C02 Production
CO2 Production
C02 Production
C02 Production
and Capture
Non-
Year
(kt) (% Non-EOR)
(kt) (% Non-EOR)
(kt) (% Non-EOR)
(kt) (% Non-EOR)
Facilities (kt)
EOR3
1990
1,344 (100%)
63 (1%)
+
65 (100%)
NE
NE
2005
1,254 (27%)
58 (1%)
+
63 (100%)
NE
NE
2017
IE
IE
IE
IE
59,900b
8%
2018
IE
IE
IE
IE
58,400b
7%
2019
IE
IE
IE
IE
61,300b
8%
2020
IE
IE
IE
IE
44,700b
11%
2021
IE
IE
IE
IE
43,980b
11%
+ Does not exceed 0.5 percent.
NE (Not Estimated)
IE (Included Elsewhere)
a Includes only food and beverage applications.
b For 2010 through 2021, the publicly available GHGRP data were aggregated at the national level based on GHGRP CBI
criteria. The Dome-specific C02 production values are accounted for (i.e. included elsewhere) in the Total C02 Production
from Extraction and Capture Facilities values starting in 2010 and are not able to be disaggregated.
27 Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
28 through 2021. The methodology for CO2 consumption spliced activity data from two different sources: Industry
Industrial Processes and Product Use 4-71
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
data for 1990 through 2009 and GHGRP data starting in 2010. Consistent with the 2006IPCC Guidelines, the
overlap technique was applied to compare the two data dets for years where there was overlap. The data sets
were determined to be inconsistent; the GHGRP data includes CO2 from industrial sources while the industry data
does not. No adjustments were made to the activity data for 1990 through 2009 because the 2006 IPCC Guidelines
indicate that it is not good practice to use the overlap technique when the data sets are inconsistent.
Uncertainty-TO BE UPDATED FOR FINAL INVENTORY REPORT
There is uncertainty associated with the data reported through EPA's GHGRP. Specifically, there is uncertainty
associated with the amount of CO2 consumed for food and beverage applications, given the GHGRP does have
provisions that Subpart PP reporters are not required to report to the GHGRP if their emissions fall below certain
thresholds, in addition to the exclusion of the amount of CO2 transferred to all other end-use categories. This latter
category might include CO2 quantities that are being used for non-EOR industrial applications such as firefighting.
Second, uncertainty is associated with the exclusion of imports/exports data for CO2 suppliers. Currently these
data are not publicly available through EPA's GHGRP and hence are excluded from this analysis. EPA verifies annual
facility-level reports through a multi-step process (e.g., combination of electronic checks and manual reviews by
staff) to identify potential errors and ensure that data submitted to EPA are accurate, complete, and consistent.
Based on the results of the verification process, EPA follows up with facilities to resolve mistakes that may have
occurred.50
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-55. Carbon dioxide
consumption CO2 emissions for 2021 were estimated to be between 4.7 and 5.2 MMT CO2 Eq. at the 95 percent
confidence level. This indicates a range of approximately 5 percent below to 5 percent above the emission
estimate of 5.0 MMT CO2 Eq.
Table 4-55: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from CO2
Consumption (MMT CO2 Eq. and Percent)
Source Gas
2021 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.)
(MMT C02
Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
C02 Consumption C02
5.0
4.7
5.2
-5%
+5%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S. Inventory
QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006 IPCC Guidelines as described in the
introduction of the IPPU chapter (see Annex 8 for more details). More details on the greenhouse gas calculation,
monitoring and QA/QC methods applicable to CO2 Consumption can be found under Subpart PP (Suppliers of
Carbon Dioxide) of the regulation (40 CFR Part 98).51 EPA verifies annual facility-level GHGRP reports through a
multi-step process (e.g., combination of electronic checks and manual reviews) to identify potential errors and
ensure that data submitted to EPA are accurate, complete, and consistent (EPA 20 15).52 Based on the results of the
verification process, EPA follows up with facilities to resolve mistakes that may have occurred. The post-submittals
50 See https://www.epa.eov/sites/production/files/2015-07/documents/eherp verification factsheet.pdf.
51 See http://www.ecfr.eov/cei-bin/text-idx7tph/ecfrbrowse/Title40/40cfr98 main Q2.tpl.
52 See https://www.epa.eov/sites/production/files/2015-07/documents/eherp verification factsheet.pdf.
4-72 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 checks are consistent with a number of general and category-specific QC procedures, including range checks,
2 statistical checks, algorithm checks, and year-to-year checks of reported data and emissions.
3 Recalculations Discussion
4 No recalculations were performed for the 1990 through 2020 portion of the time series.
5 Planned Improvements
6 EPA will continue to evaluate the potential to include additional GHGRP data on other emissive end-uses to
7 improve the accuracy and completeness of estimates for this source category. Particular attention will be made to
8 ensuring time-series consistency of the emissions estimates presented in future Inventory reports, consistent with
9 IPCC and UNFCCC guidelines. This is required as the facility-level reporting data from EPA's GHGRP, with the
10 program's initial requirements for reporting of emissions in calendar year 2010, are not available for all inventory
11 years (i.e., 1990 through 2009) as required for this Inventory. In implementing improvements and integration of
12 data from EPA's GHGRP, the latest guidance from the IPCC on the use of facility-level data in national inventories
13 will be relied upon.53
14 These improvements are still in process and will be incorporated into future Inventory reports. These are near-to
15 medium-term improvements.
it 4.16 Phosphoric Acid Production (CRF
I? Source Category 2B10)
18 Phosphoric acid (H3PO4) is a basic raw material used in the production of phosphate-based fertilizers. Phosphoric
19 acid production from natural phosphate rock is a source of carbon dioxide (CO2) emissions, due to the chemical
20 reaction of the inorganic carbon (calcium carbonate) component of the phosphate rock.
21 Phosphate rock is mined in Florida and North Carolina, which account for more than 75 percent of total domestic
22 output, and in Idaho and Utah (USGS 2022). It is used primarily as a raw material for wet-process phosphoric acid
23 production. The composition of natural phosphate rock varies, depending on the location where it is mined.
24 Natural phosphate rock mined in the United States generally contains inorganic carbon in the form of calcium
25 carbonate (limestone) and may also contain organic carbon.
26 The phosphoric acid production process involves chemical reaction of the calcium phosphate (Ca3(PC>4)2)
27 component of the phosphate rock with sulfuric acid (H2SO4) and recirculated phosphoric acid (H3PO4) (EFMA 2000).
28 The generation of CO2, however, is due to the associated limestone-sulfuric acid reaction, as shown below:
29 CaC03 + H2S04 + H20 -> CaS04 -2H20 + C02
30 Total U.S. phosphate rock production in 2021 was an estimated 23.0 million metric tons (USGS 2022). Total imports
31 of phosphate rock to the United States in 2021 were 2.4 million metric tons (USGS 2022). Between 2017 and 2020,
32 most of the imported phosphate rock (87 percent) came from Peru, with 13 percent from Morocco (USGS 2022).
33 All phosphate rock mining companies in the United States are vertically integrated with fertilizer plants that
34 produce phosphoric acid located near the mines. The phosphoric acid production facilities that use imported
35 phosphate rock are located in Louisiana.
36 Between 1990 and 2021, domestic phosphate rock production decreased by nearly 54 percent. Total CO2
37 emissions from phosphoric acid production were 0.9 MMT CO2 Eq. (909 kt CO2) in 2021 (see Table 4-56). Domestic
53 See http://www.ipcc-nggiD.iges.or.jp/public/tb/TFl Technical Bulletin l.pdf.
Industrial Processes and Product Use 4-73
-------
1 consumption of phosphate rock in 2021 was estimated to have decreased 1 percent relative to 2020 levels. The
2 COVID-19 pandemic did not impact the domestic phosphate rock market as both the fertilizer industry and related
3 agricultural businesses were considered essential industries and were unaffected by pandemic "stay-at-home"
4 orders issued in March 2020 (USGS 2021a).
5 Table 4-56: CO2 Emissions from Phosphoric Acid Production (MMT CO2 Eq. and kt)
Year
1990
2005
2017
2018
2019
2020
2021
MMTCOz Eq.
1.5
1.3
1.0
0.9
0.9
0.9
0.9
kt
1,529
1,342
1,025
937
909
901
909
e Methodology and Time-Series Consistency
7 The United States uses a country-specific methodology consistent with an IPCC Tier 1 approach to calculate
8 emissions from production of phosphoric acid from phosphate rock.54 Carbon dioxide emissions from production
9 of phosphoric acid from phosphate rock are estimated by multiplying the average amount of inorganic carbon
10 (expressed as CO2) contained in the natural phosphate rock as calcium carbonate by the amount of phosphate rock
11 that is used annually to produce phosphoric acid, accounting for domestic production and net imports for
12 consumption. The estimation methodology is as follows:
13 Equation 4-9: CO2 Emissions from Phosphoric Acid Production
14 f = r x o
'-"pa °pr Vpr
15 where,
Epa = CO2 emissions from phosphoric acid production, metric tons
Cpr = Average amount of carbon (expressed as CO2) in natural phosphate rock, metric ton
CO2/ metric ton phosphate rock
QPr = Quantity of phosphate rock used to produce phosphoric acid
16
17 The CO2 emissions calculation methodology assumes that all of the inorganic C (calcium carbonate) content of the
18 phosphate rock reacts to produce CO2 in the phosphoric acid production process and is emitted with the stack gas.
19 The methodology also assumes that none of the organic C content of the phosphate rock is converted to CO2 and
20 that all of the organic C content remains in the phosphoric acid product.
21 From 1993 to 2004, the U.S. Geological Survey (USGS) Mineral Yearbook: Phosphate Rock disaggregated phosphate
22 rock mined annually in Florida and North Carolina from phosphate rock mined annually in Idaho and Utah, and
23 reported the annual amounts of phosphate rock exported and imported for consumption (see Table 4-57). For the
24 years 1990 through 1992, and 2005 through 2021, only nationally aggregated mining data was reported by USGS.
25 For the years 1990,1991, and 1992, the breakdown of phosphate rock mined in Florida and North Carolina and the
26 amount mined in Idaho and Utah are approximated using data reported by USGS for the average share of U.S.
27 production in those states from 1993 to 2004. For the years 2005 through 2021, the same approximation method
28 is used, but the share of U.S. production based on production capacity in those states were obtained from the
29 USGS commodity specialist for phosphate rock (USGS 2012; USGS 2021b). For 1990 through 2021, data on U.S.
30 domestic consumption of phosphate rock, consisting of domestic reported sales and use of phosphate rock,
31 exports of phosphate rock (primarily from Florida and North Carolina), and imports of phosphate rock for
32 consumption, were obtained from USGS Minerals Yearbook: Phosphate Rock (USGS 1994 through 2015b) and from
33 USGS Minerals Commodity Summaries: Phosphate Rock (USGS 2016 through 2021a, 2022). From 2004 through
34 2021, the USGS reported no exports of phosphate rock from U.S. producers (USGS 2022).
35 The carbonate content of phosphate rock varies depending upon where the material is mined. Composition data
36 for domestically mined and imported phosphate rock were provided by the Florida Institute of Phosphate
54 The 2006 IPCC Guidelines do not provide a method for estimating process emissions (C02) from Phosphoric Acid Production.
4-74 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Research, now known as the Florida Industrial and Phosphate Research Institute (FIPR 2003a). Phosphate rock
2 mined in Florida contains approximately 1 percent inorganic C, and phosphate rock imported from Morocco
3 contains approximately 1.46 percent inorganic C. Calcined phosphate rock mined in North Carolina and Idaho
4 contains approximately 0.41 percent and 0.27 percent inorganic C, respectively (see Table 4-57). Similar to the
5 phosphate rock mined in Morocco, phosphate rock mined in Peru contains approximately 5 percent CO2 (Golder
6 Associates and M3 Engineering 2016).
7 Carbonate content data for phosphate rock mined in Florida are used to calculate the CO2 emissions from
8 consumption of phosphate rock mined in Florida and North Carolina (more than 75 percent of domestic
9 production), and carbonate content data for phosphate rock mined in Morocco and Peru are used to calculate CO2
10 emissions from consumption of imported phosphate rock. The CO2 emissions calculation assumes that all of the
11 domestic production of phosphate rock is used in uncalcined form. As of 2006, the USGS noted that one phosphate
12 rock producer in Idaho produces calcined phosphate rock; however, no production data were available for this
13 single producer (USGS 2006). The USGS confirmed that no significant quantity of domestic production of
14 phosphate rock is in the calcined form (USGS 2012).
15 Table 4-57: Phosphate Rock Domestic Consumption, Exports, and Imports (kt)
Location/Year 1990
2005
2017 2018 2019 2020 2021
U.S. Domestic Consumption3 49,800
FL and NC 42,494
ID and UT 7,306
Exports—FL and NC 6,240
Imports 451
35,200
28,160
7,040
0
2,630
26,300 23,300 23,400 22,600 23,000
20,510 18,170 18,250 17,630 17,940
5,790 5,130 5,150 4,970 5,060
0 0 0 0 0
2,470 2,770 2,140 2,520 2,400
Total U.S. Consumption 44,011
37,830
28,770 26,070 25,540 25,120 25,400
a U.S. domestic consumption values are based on reported phosphate rock sold or used by producers.
Note: Totals may not sum due to independent rounding.
16 Table 4-58: Chemical Composition of Phosphate Rock (Percent by Weight)
North
Central
North
Carolina
Idaho
Composition
Florida
Florida
(calcined)
(calcined)
Morocco
Peru
Total Carbon (as C)
1.60
1.76
0.76
0.60
1.56
NA
Inorganic Carbon (as C)
1.00
0.93
0.41
0.27
1.46
NA
Organic Carbon (as C)
0.60
0.83
0.35
0.00
0.10
NA
Inorganic Carbon (as C02)
3.67
3.43
1.50
1.00
5.00
5.00
NA (Not Available)
Sources: FIPR (2003a), Golder Associates and M3 Engineering (2016)
17 Methodological approaches were applied to the entire time series to ensure consistency in emissions estimates
18 from 1990 through 2020.
19 Uncertainty - TO BE UPDATED FOR FINAL INVENTORY REPORT
20 Phosphate rock production data used in the emission calculations were developed by the USGS through monthly
21 and semiannual voluntary surveys of the active phosphate rock mines during 2021. Prior to 2006, USGS provided
22 the data disaggregated regionally; however, beginning in 2006, only total U.S. phosphate rock production was
23 reported. Regional production for 2021 was estimated based on regional production data from 2017 to 2020 and
24 multiplied by regionally-specific emission factors. There is uncertainty associated with the degree to which the
25 estimated 2021 regional production data represents actual production in those regions. Total U.S. phosphate rock
26 production data are not considered to be a significant source of uncertainty because all the domestic phosphate
27 rock producers report their annual production to the USGS. Data for exports of phosphate rock used in the
28 emission calculations are reported to the USGS by phosphate rock producers and are not considered to be a
29 significant source of uncertainty. Data for imports for consumption are based on international trade data collected
Industrial Processes and Product Use 4-75
-------
1 by the U.S. Census Bureau. These U.S. government economic data are not considered to be a significant source of
2 uncertainty. Based on expert judgement of the USGS, EPA assigned a default uncertainty range of ±5 percent to
3 the percentage of phosphate rock produced from Florida and North Carolina, and ±5 percent to phosphoric acid
4 production and imports.
5 An additional source of uncertainty in the calculation of CO2 emissions from phosphoric acid production is the
6 carbonate composition of phosphate rock, as the composition of phosphate rock varies depending upon where the
7 material is mined and may also vary over time. The Inventory relies on one study (FIPR 2003a) of chemical
8 composition of the phosphate rock; limited data are available beyond this study. Another source of uncertainty is
9 the disposition of the organic carbon content of the phosphate rock. A representative of FIPR indicated that in the
10 phosphoric acid production process, the organic C content of the mined phosphate rock generally remains in the
11 phosphoric acid product, which is what produces the color of the phosphoric acid product (FIPR 2003b). Organic
12 carbon is therefore not included in the calculation of CO2 emissions from phosphoric acid production.
13 A third source of uncertainty is the assumption that all domestically-produced phosphate rock is used in
14 phosphoric acid production and used without first being calcined. Calcination of the phosphate rock would result
15 in conversion of some of the organic C in the phosphate rock into CO2; however, according to air permit
16 information available to the public, at least one facility has calcining units permitted for operation (NCDENR 2013).
17 Finally, USGS indicated that in 2021 less than 5 percent of domestically-produced phosphate rock was used to
18 manufacture elemental phosphorus and other phosphorus-based chemicals, rather than phosphoric acid (USGS
19 2022). According to USGS, there is only one domestic producer of elemental phosphorus, in Idaho, and no data
20 were available concerning the annual production of this single producer. Elemental phosphorus is produced by
21 reducing phosphate rock with coal coke, and it is therefore assumed that 100 percent of the carbonate content of
22 the phosphate rock will be converted to CO2 in the elemental phosphorus production process. The calculation for
23 CO2 emissions assumes that phosphate rock consumption, for purposes other than phosphoric acid production,
24 results in CChemissions from 100 percent of the inorganic carbon content in phosphate rock, but none from the
25 organic carbon content.
26 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-59. 2021 phosphoric acid
27 production CO2 emissions were estimated to be between 0.8 and 1.2 MMT CO2 Eq. at the 95 percent confidence
28 level. This indicates a range of approximately 18 percent below and 20 percent above the emission estimate of 0.9
29 MMT CO2 Eq.
30 Table 4-59: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from
31 Phosphoric Acid Production (MMT CO2 Eq. and Percent)
2021 Emission Estimate
Source Gas
(MMT C02 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.)
(%)
Lower Upper
Lower Upper
Bound Bound
Bound Bound
Phosphoric Acid Production C02 0.9
IN
1
00
O
-18% +20%
3 Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
32 QA/QC and Verification
33 For more information on the general QA/QC process applied to this source category, consistent with the U.S.
34 Inventory QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006IPCC Guidelines as described in the
35 introduction of the IPPU chapter (see Annex 8 for more details).
36 Recalculations Discussion
37 Recalculations were performed for 2020 to reflect an updated value for the total U.S. production of phosphate
38 rock based on updated USGS data. This update resulted in a decrease of 37 kt CO2 in 2020.
4-76 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Planned Improvements
EPA continues to evaluate potential improvements to the Inventory estimates for this source category, which
include direct integration of EPA's GHGRP data for 2010 through 2021 along with assessing applicability of
reported GHGRP data to update the inorganic C content of phosphate rock for prior years to ensure time-series
consistency. Specifically, EPA would need to assess that averaged inorganic C content data (by region or other
approaches) meets GHGRP confidential business information (CBI) screening criteria. EPA would then need to
assess the applicability of GHGRP data for the averaged inorganic C content (by region or other approaches) from
2010 through 2021, along with other information to inform estimates in prior years in the required time series
(1990 through 2009) based on the sources of phosphate rock used in production of phosphoric acid over time. In
implementing improvements and integration of data from EPA's GHGRP, the latest guidance from the IPCC on the
use of facility-1 eve I data in national inventories will be relied upon.55 These long-term planned improvements are
still in development by EPA and have not been implemented into the current Inventory report.
4.17 Iron and Steel Production (CRF Source
Category 2C1) and Metallurgical Coke
Production
Iron and steel production is a multi-step process that generates process-related emissions of carbon dioxide (CO2)
and methane (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 during the production of iron and steel
are accounted for in the Energy chapter.
Iron and steel production includes seven distinct production processes: metallurgical coke production, sinter
production, direct reduced iron (DRI) production, pellet production, pig iron56 production, electric arc furnace
(EAF) steel production, and basic oxygen furnace (BOF) steel production. The number of production processes at a
particular plant is dependent upon the specific plant configuration. Most process CO2 generated from the iron and
steel industry is a result of the production of crude iron.
In addition to the production processes mentioned above, CO2 is also generated at iron and steel mills through the
consumption of process byproducts (e.g., blast furnace gas, coke oven gas) used for various purposes including
heating, annealing, and electricity generation. Process byproducts sold off-site for use as synthetic natural gas are
also accounted for in these calculations. In general, CO2 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 2021, approximately eleven integrated iron and steel steelmaking facilities utilized BOFs to refine and produce
steel from iron, and raw steel was produced at 101 facilities across the United States. Approximately 29 percent of
steel production was attributed to BOFs and 71 percent to EAFs (USGS 2022). The trend in the United States for
integrated facilities has been a shift towards fewer BOFs and more EAFs. EAFs use scrap steel as their main input
and use significantly less energy than BOFs. There are also 14 cokemaking facilities, of which 3 facilities are co-
55 See http://www.ipcc-nggip.iges.or.jp/public/tb/TFl Technical Bulletin l.pdf.
56 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 over time as industry trends have shifted. Throughout this report, pig iron will be used
interchangeably with crude iron, but it should be noted that in other data sets or reports pig iron and crude iron may not be
used interchangeably and may provide different values.
Industrial Processes and Product Use 4-77
-------
1 located with integrated iron and steel facilities (ACCCI 2021). In the United States, 6 states account for roughly 52
2 percent of total raw steel production: Indiana, Alabama, Tennessee, Kentucky, Mississippi, and Arkansas (AISI
3 2022).
4 Total annual production of crude steel in the United States was fairly constant between 2000 and 2008 and ranged
5 from a low of 99,320,000 tons to a high of 109,880,000 tons (2001 and 2004, respectively). Due to the decrease in
6 demand caused by the global economic downturn (particularly from the automotive industry), crude steel
7 production in the United States sharply decreased to 65,459,000 tons in 2009. Crude steel production was fairly
8 constant from 2011 through 2014, and after a dip in production from 2014 to 2015, crude steel production has
9 slowly and steadily increased for the past few years. Crude steel production dipped again in 2020 due to the
10 COVID-19 pandemic and increased close to pre-pandemic levels in 2021. The United States was the fourth largest
11 producer of raw steel in the world, behind China, India and Japan, accounting for approximately 4.4 percent of
12 world production in 2021 (AISI 2004 through 2022).
13 The majority of CO2 emissions from the iron and steel production process come from the use of metallurgical coke
14 in the production of pig iron and from the consumption of other process byproducts, with lesser amounts emitted
15 from the use of carbon-containing flux and from the removal of carbon from pig iron used to produce steel.
16 According to the 2006IPCC Guidelines, the production of metallurgical coke from coking coal is considered to be an
17 energy use of fossil fuel, and the use of coke in iron and steel production is considered to be an industrial process
18 source. The 2006 IPCC Guidelines suggest that emissions from the production of metallurgical coke should be
19 reported separately in the Energy sector, while emissions from coke consumption in iron and steel production
20 should be reported in the Industrial Processes and Product Use sector. The approaches and emission estimates for
21 both metallurgical coke production and iron and steel production, however, are presented here because much of
22 the relevant activity data is used to estimate emissions from both metallurgical coke production and iron and steel
23 production. For example, some byproducts (e.g., coke oven gas) of the metallurgical coke production process are
24 consumed during iron and steel production, and some byproducts of the iron and steel production process (e.g.,
25 blast furnace gas) are consumed during metallurgical coke production. Emissions associated with the consumption
26 of these byproducts are attributed at the point of consumption. Emissions associated with the use of conventional
27 fuels (e.g., natural gas, fuel oil) for electricity generation, heating and annealing, or other miscellaneous purposes
28 downstream of the iron and steelmaking furnaces are reported in the Energy chapter.
29 Metallurgical Coke Production
30 Emissions of CO2 from metallurgical coke production in 2021 were 3.2 MMT CO2 Eq. (3,224 kt CO2) (see Table 4-60
31 and Table 4-61). Emissions increased by 39 percent from 2020 to 2021 and have decreased by 43 percent since
32 1990. Coke production in 2021 was about 21 percent higher than in 2020 and 55 percent below 1990 (EIA 2022,
33 AISI 2022).
34 Significant activity data for 2021 and 2020 were not available in time for publication of this report due to industry
35 consolidation that impacts the publication of data without revealing confidential business information. Activity
36 data for these years were estimated using 2019 values adjusted based on GHGRP emissions data, as described in
37 the Methodology and Time-Series Consistency section below.
38 Table 4-60: CO2 Emissions from Metallurgical Coke Production (MMT CO2 Eq.)
Gas 1990 2005 2017 2018 2019 2020 2021
CO; 5.6 A 3.9 A Z0 13 ^0 23 3,2_
39 Table 4-61: CO2 Emissions from Metallurgical Coke Production (kt)
Gas 1990 2005 2017 2018 2019 2020 2021
C02 5,608 3,921 1,978 1,282 3,006 2,325 3,224
4-78 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Iron and Steel Production
2 Emissions of CO2 and CH4 from iron and steel production in 2021 were 38.8 MMT CO2 Eq. (38,817 kt) and 0.0082
3 MMT CO2 Eq. (0.3 kt Cm), respectively (see Table 4-62 through Table 4-65). Emissions from iron and steel
4 production increased by 10 percent from 2020 to 2021 and have decreased by 61 percent since 1990, due to
5 restructuring of the industry, technological improvements, and increased scrap steel utilization. Carbon dioxide
6 emission estimates include emissions from the consumption of carbonaceous materials in the blast furnace, EAF,
7 and BOF, as well as blast furnace gas and coke oven gas consumption for other activities at the steel mill.
8 Significant activity data for 2021 and 2020 were not available in time for publication of this report due to industry
9 consolidation that impacts the publication of data without revealing confidential business information. Activity
10 data for these years were estimated using 2019 values adjusted based on GHGRP emissions data, as described in
11 the Methodology and Time-Series Consistency section below.
12 In 2021, domestic production of pig iron increased by 21 percent from 2020 levels. Overall, domestic pig iron
13 production has declined since the 1990s; pig iron production in 2021 was 54 percent lower than in 2000 and 55
14 percent below 1990. Carbon dioxide emissions from iron production have decreased by 80 percent (36.6 MMT CO2
15 Eq.) since 1990. Carbon dioxide emissions from steel production have decreased by 26 percent (2.1 MMT CO2 Eq.)
16 since 1990, while overall CO2 emissions from iron and steel production have declined by 61 percent (60.3 MMT
17 CO2 Eq.) from 1990 to 2021.
18 Table 4-62: CO2 Emissions from Iron and Steel Production (MMT CO2 Eq.)
Source/Activity Data
1990
2005
2017
2018
2019
2020
2021
Sinter Production
2.4
1.7
0.9
0.9
0.9
0.7
0.8
Iron Production
45.7
17.7
8.2
9.6
9.4
8.4
9.1
Pellet Production
1.8
1.5
0.9
0.9
0.9
0.8
0.8
Steel Production
8.0
9.4
6.2
5.8
5.8
5.6
5.9
Other Activities3
41.2
35.9
22.4
24.1
23.2
19.8
22.1
Total
99.1
66.2
38.8
41.6
40.1
35.4
38.8
a Includes emissions from blast furnace gas and coke oven gas combustion for activities at the steel mill other
than consumption in blast furnace, EAFs, or BOFs.
Note: Totals may not sum due to independent rounding.
19 Table 4-63: CO2 Emissions from Iron and Steel Production (kt)
Source/Activity Data
1990
2005
2017
2018
2019
2020
2021
Sinter Production
2,448
1,663
869
937
876
749
836
Iron Production
45,706
17,661
8,237
9,581
9,360
8,409
9,121
Pellet Production
1,817
1,503
867
924
878
751
838
Steel Production
7,964
9,395
6,218
5,754
5,812
5,657
5,902
Other Activities3
41,194
35,934
22,396
24,149
23,158
19,820
22,119
Total
99,129
66,156
38,832
41,576
40,084
35,387
38,817
a Includes emissions from blast furnace gas and coke oven gas combustion for activities at the steel mill
other than consumption in blast furnace, EAFs, or BOFs.
Note: Totals may not sum due to independent rounding.
20 Table 4-64: Cm Emissions from Iron and Steel Production (MMT CO2 Eq.)
Source/Activity Data
1990
2005
2017
2018
2019
2020
2021
Sinter Production
+
+
+
+
+
+
+
+ Does not exceed 0.05 MMT C02 Eq.
Industrial Processes and Product Use 4-79
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Table 4-65: ChU Emissions from Iron and Steel Production (kt)
Source/Activity Data
1990
2005
2017
2018
2019
2020
2021
Sinter Production
+
+
+
+
+
+
+
+ Does not exceed 0.05 MMT C02 Eq.
Methodology and Time-Series Consistency
Emission estimates presented in this chapter utilize a country-specific approach based on Tier 2 methodologies
provided by the 2006IPCC Guidelines. These Tier 2 methodologies call for a mass balance accounting of the
carbonaceous inputs and outputs during the iron and steel production process and the metallurgical coke
production process. Tier 1 methods are used for certain iron and steel production processes (i.e., sinter
production, pellet production and DRI production) for which available data are insufficient to apply a Tier 2
method (e.g., country-specific carbon contents of inputs and outputs are not known). The majority of emissions
are captured with higher tier methods, as sinter production, pellet production, and DRI production only account
for roughly 8 percent of total iron and steel production emissions.
The Tier 2 methodology equation is as follows:
Equation 4-10: CO2 Emissions from Coke, Pig Iron, EAF Steel, and BOF Steel Production,
based on 2006 IPCC Guidelines Tier 2 Methodologies
Eco2
^(<2a X Ca) - ^(<2fc X Cb)
44
X
12
where,
Eco2 = Emissions from coke, pig iron, EAF steel, or BOF steel production, metric tons
a = Input material a
b = Output material b
Qa = Quantity of input material a, metric tons
Ca = Carbon content of input material a, metric tons C/metric ton material
Qb = Quantity of output material b, metric tons
Cb = Carbon content of output material b, metric tons C/metric ton material
44/12 = Stoichiometric ratio of CO2 to C
The Tier 1 methodology equations are as follows:
Equation 4-11: 2006IPCCGuide/inesTier 1: Emissions from Sinter, Direct Reduced Iron, and
Pellet Production (Equations 4.6,4.7, and 4.8)
ES,P = QSX EFS,P
Ed,C02 = Qd X EFd,c02
Ep,co2 = Qp x EFp,C02
Es,p = Emissions from sinter production process for pollutant p (CO2 or CH4), metric ton
Qs = Quantity of sinter produced, metric tons
EFs,p = Emission factor for pollutant p (CO2 or CH4), metric ton p/metric ton sinter
Ed,co2 = Emissions from DRI production process for CO2, metric ton
Qd = Quantity of DRI produced, metric tons
EFd,co2 = Emission factor for CO2, metric ton C02/metric ton DRI
EP,co2 = Emissions from pellet production process for CO2, metric ton
QP = Quantity of pellets produced, metric tons
EFP,co2 = Emission factor for CO2, metric ton C02/metric ton pellets produced
where,
4-80 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
A significant number of activity data that serve as inputs to emissions calculations were unavailable for 2021 and
2020 at the time of publication and were estimated using 2019 values. In addition, to account for the impacts of
the COVID-19 pandemic in 2020, the EPA used process emissions data from the EPA's Greenhouse Gas Reporting
Program (GHGRP) subpart Q for the iron and steel sector to adjust the estimated values for 2021 and 2020. GHGRP
process emissions data decreased by approximately 14 percent from 2019 to 2020 and increased by approximately
12% from 2020 to 2021 (EPA 2022). These percentage changes were applied to 2019 activity data values to
produce an estimate for 2021 and 2020 data.
Metallurgical Coke Production
Coking coal is used to manufacture metallurgical coke which is used primarily as a reducing agent in the production
of iron and steel but is also used in the production of other metals including zinc and lead (see Zinc Production and
Lead Production sections of this chapter). Emissions associated with producing metallurgical coke from coking coal
are estimated and reported separately from emissions that result from the iron and steel production process. To
estimate emissions from metallurgical coke production, a Tier 2 method provided by the 2006IPCC Guidelines was
utilized. The amount of carbon contained in materials produced during the metallurgical coke production process
(i.e., coke, coke breeze and coke oven gas) is deducted from the amount of carbon contained in materials
consumed during the metallurgical coke production process (i.e., natural gas, blast furnace gas, and coking coal).
For calculations, activity data for these inputs, including natural gas, blast furnace gas, and coking coke consumed
for metallurgical coke production, are in units consistent with the carbon content values. Light oil, which is
produced during the metallurgical coke production process, is excluded from the deductions due to data
limitations. The amount of carbon contained in these materials is calculated by multiplying the material-specific
carbon content by the amount of material consumed or produced (see Table 4-66). The amount of coal tar
produced was approximated using a production factor of 0.03 tons of coal tar per ton of coking coal consumed.
The amount of coke breeze produced was approximated using a production factor of 0.075 tons of coke breeze per
ton of coking coal consumed (Steiner 2008; DOE 2000). Data on the consumption of carbonaceous materials (other
than coking coal) as well as coke oven gas production were available for integrated steel mills only (i.e., steel mills
with co-located coke plants); therefore, carbonaceous material (other than coking coal) consumption and coke
oven gas production were excluded from emission estimates for merchant coke plants. Carbon contained in coke
oven gas used for coke-oven underfiring was not included in the deductions to avoid double-counting.
Table 4-66: Material Carbon Contents for Metallurgical Coke Production
Material
kg C/kg
Coal Tara
0.62
Cokea
0.83
Coke Breeze3
0.83
Coking Coalb
0.75
Material
kg C/GJ
Coke Oven Gasc
12.1
Blast Furnace Gasc
70.8
a Source: IPCC (2006), Vol. 3 Chapter 4, Table 4.3
b Source: EIA (2017b)
c Source: IPCC (2006), Vol. 2 Chapter 1, Table 1.3
Although the 2006 IPCC Guidelines provide a Tier 1 Cm emission factor for metallurgical coke production (i.e., 0.1 g
Cm per metric ton of coke production), it is not appropriate to use because CO2 emissions were estimated using
the Tier 2 mass balance methodology. The mass balance methodology makes a basic assumption that all carbon
that enters the metallurgical coke production process either exits the process as part of a carbon-containing
output or as CO2 emissions. This is consistent with a preliminary assessment of aggregated facility-level
greenhouse gas CH4 emissions reported by coke production facilities under EPA's GHGRP. The assessment indicates
that CH4 emissions from coke production are insignificant and below 500 kt or 0.05 percent of total national
Industrial Processes and Product Use 4-81
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
23
24
25
26
27
28
29
30
31
32
33
34
35
emissions. Pending resources and significance, EPA continues to assess the possibility of including these emissions
in future Inventories to enhance completeness but has not incorporated these emissions into this report.
Data relating to the mass of coking coal consumed at metallurgical coke plants and the mass of metallurgical coke
produced at coke plants were taken from the Energy Information Administration (EIA) Quarterly Coal Report:
October through December (EIA 1998 through 2019) and EIA Quarterly Coal Report: January through March (EIA
2021 through 2022) (see Table 4-67). Data on the volume of natural gas consumption, blast furnace gas
consumption, and coke oven gas production for metallurgical coke production at integrated steel mills were
obtained from the American Iron and Steel Institute (AISI) Annual Statistical Report (AISI 2004 through 2022) and
through personal communications with AISI (Steiner 2008) (see
Table 4-68). These data from the AISI Annual Statistical Report were withheld for 2021 and 2020, so the 2019
values were used as estimated data for the missing 2021 and 2020 values and adjusted using GHGRP emissions
data, as described earlier in this Methodology and Time-Series Consistency section.
The factor for the quantity of coal tar produced per ton of coking coal consumed was provided by AISI (Steiner
2008). The factor for the quantity of coke breeze produced per ton of coking coal consumed was obtained through
Table 2-1 of the report Energy and Environmental Profile of the U.S. Iron and Steel Industry (DOE 2000). Currently,
data on natural gas consumption and coke oven gas production at merchant coke plants were not available and
were excluded from the emission estimate. Carbon contents for metallurgical coke, coal tar, coke oven gas, and
blast furnace gas were provided by the 2006IPCC Guidelines. The carbon content for coke breeze was assumed to
equal the carbon content of coke. Carbon contents for coking coal was from EIA.
Table 4-67: Production and Consumption Data for the Calculation of CO2 Emissions from
Metallurgical Coke Production (Thousand Metric Tons)
Source/Activity Data
1990
2005
2017
2018
2019
2020
id2i
Metallurgical Coke Production
Coking Coal Consumption at Coke Plants
35,269
21,259
15,910
16,635
16,261
13,076
15,957
Coke Production at Coke Plants
25,054
15,167
11,746
12,525
11,676
9,392
11,381
Coke Breeze Production
2,645
1,594
1,193
1,248
1,220
981
1,197
Coal Tar Production
1,058
638
477
499
488
392
479
Table 4-68: Production and Consumption Data for the Calculation of CO2 Emissions fror
Metallurgical Coke Production (Million ft3)
Source/Activity Data
1990
2005
2017
2018
2019
2020
2021
Metallurgical Coke Production
Coke Oven Gas Production
250,767
114,213
74,997
80,750
77,692
66,492
74,206
Natural Gas Consumption
599
2,996
2,103
2,275
2,189
1,873
2,091
Blast Furnace Gas Consumption
24,602
4,460
3,683
4,022
3,914
3,350
3,738
Iron and Steel Production
To estimate emissions from pig iron production in the blast furnace, the amount of carbon contained in the
produced pig iron and blast furnace gas were deducted from the amount of carbon contained in inputs (i.e.,
metallurgical coke, sinter, natural ore, pellets, natural gas, fuel oil, coke oven gas, carbonate fluxes or slagging
materials, and direct coal injection). For calculations, activity data for these inputs, including coke consumed for
pig iron production, are in units consistent with the carbon content values. The carbon contained in the pig iron,
blast furnace gas, and blast furnace inputs was estimated by multiplying the material-specific carbon content by
each material type (see Table 4-69). In the absence of a default carbon content value from the 2006 IPCC
Guidelines for pellet, sinter, or natural ore consumed for pig iron production, a country-specific approach based on
Tier 2 methodology is used. Pellet, sinter, and natural ore used as an input for pig iron production is assumed to
have the same carbon content as direct reduced iron (2 percent). Carbon in blast furnace gas used to pre-heat the
4-82 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
blast furnace air is combusted to form CO2 during this process. Carbon contained in blast furnace gas used as a
blast furnace input was not included in the deductions to avoid double-counting.
Emissions from steel production in EAFs were estimated by deducting the carbon contained in the steel produced
from the carbon contained in the EAF anode, charge carbon, and scrap steel added to the EAF. Small amounts of
carbon from DRI and pig iron to the EAFs were also included in the EAF calculation. For BOFs, estimates of carbon
contained in BOF steel were deducted from carbon contained in inputs such as natural gas, coke oven gas, fluxes
(i.e., limestone and dolomite), and pig iron. In each case, the carbon was calculated by multiplying material-specific
carbon contents by each material type (see Table 4-69). For EAFs, the amount of EAF anode consumed was
approximated by multiplying total EAF steel production by the amount of EAF anode consumed per metric ton of
steel produced (0.002 metric tons EAF anode per metric ton steel produced [Steiner 2008]). The amount of carbon-
containing flux (i.e., limestone and dolomite) used in EAF and BOF steel production was deducted from the "Other
Process Uses of Carbonates" source category (CRF Source Category 2A4) to avoid double-counting.
Carbon dioxide emissions from the consumption of blast furnace gas and coke oven gas for other activities
occurring at the steel mill were estimated by multiplying the amount of these materials consumed for these
purposes by the material-specific carbon content (see Table 4-69).
Table 4-69: Material Carbon Contents for Iron and Steel Production
Material
kg C/kg
Coke
0.83
Direct Reduced Iron
0.02
Dolomite
0.13
EAF Carbon Electrodes
0.82
EAF Charge Carbon
0.83
Limestone
0.12
Pig Iron
0.04
Steel
0.01
Material
kg C/GJ
Coke Oven Gas
12.1
Blast Furnace Gas
70.8
Source: IPCC (2006), Table 4.3. Coke Oven Gas and
Blast Furnace Gas, Table 1.3.
Carbon dioxide emissions associated with sinter production, direct reduced iron production, pellet production, pig
iron production, steel production, and other steel mill activities were summed to calculate the total CO2 emissions
from iron and steel production (see Table 4-62 and Table 4-63).
The sinter production process results in fugitive emissions of CH4, which are emitted via leaks in the production
equipment, rather than through the emission stacks or vents of the production plants. The fugitive emissions were
calculated by applying Tier 1 emission factors taken from the 2006 IPCC Guidelines for sinter production (see Table
4-70). Although the 2006 IPCC Guidelines also provide a Tier 1 methodology for CH4 emissions from pig iron
production, it is not appropriate to use because CO2 emissions for pig iron production are estimated using the Tier
2 mass balance methodology. The mass balance methodology makes a basic assumption that all carbon that enters
the pig iron production process either exits the process as part of a carbon-containing output or as CO2 emissions;
the estimation of CH4 emissions is precluded. Annual analysis of facility-level emissions reported during iron
production further supports this assumption and indicates that CFU emissions are below 500 kt CO2 Eq. and well
below 0.05 percent of total national emissions. The production of direct reduced iron could also result in emissions
of Cm through the consumption of fossil fuels (e.g., natural gas, etc.); however, these emission estimates are
excluded due to data limitations. Pending further analysis and resources, EPA may include these emissions in
future reports to enhance completeness. EPA is still assessing the possibility of including these emissions in future
reports and have not included this data in the current report.
Industrial Processes and Product Use 4-83
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Table 4-70: ChU Emission Factors for Sinter and Pig Iron Production
Material Produced
Factor
Unit
Sinter
0.07
kg CH4/metric ton
Source: IPCC (2006), Table 4.2.
Emissions of CChfrom sinter production, direct reduced iron production, and pellet production were estimated by
multiplying total national sinter production, total national direct reduced iron production, and total national pellet
production by Tier 1 CO2 emission factors (see Table 4-71). Because estimates of sinter production, direct reduced
iron production, and pellet production were not available, production was assumed to equal consumption.
Table 4-71: CO2 Emission Factors for Sinter Production, Direct Reduced Iron Production, and
Pellet Production
Metric Ton C02/Metric
Material Produced
Ton
Sinter
0.2
Direct Reduced Iron
0.7
Pellet Production
0.03
Source: IPCC (2006), Table 4.1.
The consumption of coking coal, natural gas, distillate fuel, and coal used in iron and steel production are adjusted
for within the Energy chapter to avoid double-counting of emissions reported within the IPPU chapter as these
fuels were consumed during non-energy related activities. More information on this methodology and examples of
adjustments made between the IPPU and Energy chapters are described in Annex 2.1, Methodology for Estimating
Emissions of CO2 from Fossil Fuel Combustion.
Sinter consumption and pellet consumption data for 1990 through 2020 were obtained from AISI's Annual
Statistical Report (AISI 2004 through 2022) and through personal communications with AISI (Steiner 2008) (see
Table 4-72). These data from the AISI Annual Statistical Report were withheld for 2021 and 2020, so the 2019
values were used as estimated data for the missing 2021 and 2020 values and adjusted using GHGRP emissions
data, as described earlier in this Methodology and Time-Series Consistency section.
In general, direct reduced iron (DRI) consumption data were obtained from the U.S. Geological Survey (USGS)
Minerals Yearbook - Iron and Steel Scrap (USGS 1991 through 2020) and personal communication with the USGS
Iron and Steel Commodity Specialist (Tuck 2020). Data for DRI consumed in EAFs were not available for the years
1990 and 1991. EAF DRI consumption in 1990 and 1991 was calculated by multiplying the total DRI consumption
for all furnaces by the EAF share of total DRI consumption in 1992. Additionally, data for DRI consumed in EAFs
were not available for 2021 at the time of publication, so 2020 values were used as estimated data for the missing
2021 values and adjusted using GHGRP emissions data, as described earlier in this Methodology and Time-Series
Consistency section. Data for DRI consumed in BOFs were not available for the years 1990 through 1993. BOF DRI
consumption in 1990 through 1993 was calculated by multiplying the total DRI consumption for all furnaces
(excluding EAFs and cupola) by the BOF share of total DRI consumption (excluding EAFs and cupola) in 1994.
The Tier 1 CO2 emission factors for sinter production, direct reduced iron production and pellet production were
obtained through the 2006 IPCC Guidelines (IPCC 2006). Time-series data for pig iron production, coke, natural gas,
fuel oil, sinter, and pellets consumed in the blast furnace; pig iron production; and blast furnace gas produced at
the iron and steel mill and used in the metallurgical coke ovens and other steel mill activities were obtained from
AISI's Annual Statistical Report (AISI 2004 through 2021) and through personal communications with AISI (Steiner
2008) (see Table 4-72 and Table 4-73). Data including blast furnace gas, coke oven gas, natural gas, limestone,
sinter, and natural ore consumption for blast furnaces, coke production, and steelmaking furnaces (EAFs and BOFs)
from the AISI Annual Statistical Report were withheld for 2021 and 2020, so the 2019 values were used as
estimated data for the missing 2021 and 2020 values and adjusted using GHGRP emissions data, as described
earlier in this Methodology and Time-Series Consistency section. Similarly, the percent of total steel production for
4-84 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
EAF and BOF steelmaking processes were withheld for 2021, so the 2020 values were used as estimated data for
the missing 2021 values and adjusted using GHGRP emissions data, as described earlier in this Methodology and
Time-Series Consistency section.
Data for EAF steel production, carbon-containing flux, EAF charge carbon, and natural gas consumption were
obtained from AISI's Annual Statistical Report (AISI 2004 through 2022) and through personal communications
with AISI (AISI 2006 through 2016, Steiner 2008). The factor for the quantity of EAF anode consumed per ton of
EAF steel produced was provided by AISI (Steiner 2008). Data for BOF steel production, carbon-containing flux,
natural gas, natural ore, pellet, sinter consumption as well as BOF steel production were obtained from AISI's
Annual Statistical Report (AISI 2004 through 2022) and through personal communications with AISI (Steiner 2008).
Data for EAF consumption of natural gas and BOF consumption of coke oven gas, limestone, and natural ore from
the AISI Annual Statistical Report were not available for 2021, so 2020 values were used as estimated data for the
missing 2021 values and adjusted using GHGRP emissions data, as described earlier in this Methodology and Time-
Series Consistency section. Data for EAF and BOF scrap steel, pig iron, and DRI consumption were obtained from
the USGS Minerals Yearbook - Iron and Steel Scrap (USGS 1991 through 2020). These data were not available for
2021 at the time of publication, so the 2020 values were used as estimated data for the missing 2021 values and
adjusted using GHGRP emissions data, as described earlier in this Methodology and Time-Series Consistency
section. Data on coke oven gas and blast furnace gas consumed at the iron and steel mill (other than in the EAF,
BOF, or blast furnace) were obtained from AISI's Annual Statistical Report (AISI 2004 through 2021) and through
personal communications with AISI (Steiner 2008). These data were not available for 2021, so 2020 values were
used as estimated data for the missing 2021 values and adjusted using GHGRP emissions data, as described earlier
in this Methodology and Time-Series Consistency section. Some data from the AISI Annual Statistical Report on
natural gas consumption were withheld for 2020, so the 2019 values were used as estimated data for the missing
2020 values and adjusted using GHGRP emissions data, as described earlier in this Methodology and Time-Series
Consistency section.
Data on blast furnace gas and coke oven gas sold for use as synthetic natural gas were obtained from ElA's Natural
Gas Annual (EIA 2020). Carbon contents for direct reduced iron, EAF carbon electrodes, EAF charge carbon,
limestone, dolomite, pig iron, and steel were provided by the 2006IPCC Guidelines. The carbon contents for
natural gas, fuel oil, and direct injection coal were obtained from EIA (EIA 2017b) and EPA (EPA 2010). Heat
contents for fuel oil and direct injection coal were obtained from EIA (EIA 1992, 2011); natural gas heat content
was obtained from Table 37 of AISI's Annual Statistical Report (AISI 2004 through 2021). Heat contents for coke
oven gas and blast furnace gas were provided in Table 37 of AISI's Annual Statistical Report (AISI 2004 through
2021) and confirmed by AISI staff (Carroll 2016).
Table 4-72: Production and Consumption Data for the Calculation of CO2 and ChU Emissions
from Iron and Steel Production (Thousand Metric Tons)
Source/Activity Data
1990
2005
2017
2018
2019
2020
2021
Sinter Production
12,239
8,315
4,347
4,687
4,378
3,747
4,182
Direct Reduced Iron Production
517
1,303
C
C
C
C
C
Pellet Production
60,563
50,096
28,916
30,793
29,262
25,044
27,949
Pig Iron Production
Coke Consumption
24,946
13,832
7,101
7,618
7,291
6,240
6,964
Pig Iron Production
49,669
37,222
22,395
24,058
22,302
18,320
22,246
Direct Injection Coal
Consumption
1,485
2,573
2,125
2,569
2,465
2,110
2,354
EAF Steel Production
EAF Anode and Charge Carbon
Consumption
67
1,127
1,127
1,133
1,137
1,118
1,130
Scrap Steel Consumption
42,691
46,600
C
C
C
C
C
Flux Consumption
319
695
998
998
998
998
998
EAF Steel Production
33,511
52,194
55,825
58,904
61,172
51,349
57,307
BOF Steel Production
Pig Iron Consumption
47,307
34,400
C
C
C
C
C
Industrial Processes and Product Use 4-85
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Scrap Steel Consumption
14,713
11,400
C
C
C
C
C
Flux Consumption
576
582
408
408
363
311
347
BOF Steel Production
43,973
42,705
25,788
27,704
26,591
21,384
23,865
C (Confidential)
Table 4-73: Production and Consumption Data for the Calculation of CO2 Emissions from
Iron and Steel Production (Million ft3 unless otherwise specified)
Source/Activity Data
1990
2005
2017
2018
2019
2020
2021
Pig Iron Production
Natural Gas Consumption
56,273
59,844
38,142
40,204
37,934
32,465
36,232
Fuel Oil Consumption
(thousand gallons)
163,397
16,170
4,352
3,365
2,321
1,986
2,217
Coke Oven Gas
Consumption
22,033
16,557
12,459
13,337
12,926
11,063
12,346
Blast Furnace Gas
Production
1,439,380
1,299,980
808,499
871,860
836,033
715,509
798,522
EAF Steel Production
Natural Gas Consumption
15,905
19,985
8,105
8,556
9,115
7,801
8,706
BOF Steel Production
Coke Oven Gas
Consumption
3,851
524
374
405
389
333
372
Other Activities
Coke Oven Gas
Consumption
224,883
97,132
62,164
67,008
64,377
55,096
61,489
Blast Furnace Gas
Consumption
1,414,778
1,295,520
804,816
867,838
832,119
712,159
794,783
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2021.
Uncertainty-TO BE UPDATED FOR FINAL INVENTORY REPORT
The estimates of CO2 emissions from metallurgical coke production are based on assessing uncertainties in
material production and consumption data and average carbon contents. Uncertainty is associated with the total
U.S. coking coal consumption, total U.S. coke production, and materials consumed during this process. Data for
coking coal consumption and metallurgical coke production are from different data sources (EIA) than data for
other carbonaceous materials consumed at coke plants (AISI), which does not include data for merchant coke
plants. There is uncertainty associated with the fact that coal tar and coke breeze production were estimated
based on coke production because coal tar and coke breeze production data were not available. Since merchant
coke plant data is not included in the estimate of other carbonaceous materials consumed at coke plants, the mass
balance equation for CO2 from metallurgical coke production cannot be reasonably completed; therefore, for the
purpose of this analysis, uncertainty parameters are applied to primary data inputs to the calculation (i.e., coking
coal consumption and metallurgical coke production) only.
The estimates of CO2 emissions from iron and steel production are based on material production and consumption
data and average carbon contents. There is uncertainty associated with the assumption that pellet production,
direct reduced iron and sinter consumption are equal to production. There is uncertainty with the
representativeness of the associated IPCC default emission factors. There is uncertainty associated with the
assumption that all coal used for purposes other than coking coal is for direct injection coal. There is also
uncertainty associated with the carbon contents for pellets, sinter, and natural ore, which are assumed to equal
the carbon contents of direct reduced iron, when consumed in the blast furnace. There is uncertainty associated
with the consumption of natural ore under current industry practices. For EAF steel production, there is
4-86 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
uncertainty associated with the amount of EAF anode and charge carbon consumed due to inconsistent data
throughout the time series. Also for EAF steel production, there is uncertainty associated with the assumption that
100 percent of the natural gas attributed to "steelmaking furnaces" by AISI is process-related and nothing is
combusted for energy purposes. Uncertainty is also associated with the use of process gases such as blast furnace
gas and coke oven gas. Data are not available to differentiate between the use of these gases for processes at the
steel mill versus for energy generation (i.e., electricity and steam generation); therefore, all consumption is
attributed to iron and steel production. These data and carbon contents produce a relatively accurate estimate of
CO2 emissions; however, there are uncertainties associated with each.
For calculating the emissions estimates from iron and steel and metallurgical coke production, EPA utilizes a
number of data points taken from the AISI Annual Statistical Report (ASR). This report serves as a benchmark for
information on steel companies in United States, regardless if they are a member of AISI, which represents
integrated producers (i.e., blast furnace and EAF). During the compilation of the 1990 through 2016 Inventory
report EPA initiated conversation with AISI to better understand and update the qualitative and quantitative
uncertainty metrics associated with AISI data elements. AISI estimates their data collection response rate to range
from 75 to 90 percent, with certain sectors of the iron and steel industry not being covered by the ASR; therefore,
there is some inherent uncertainty in the values provided in the AISI ASR, including material production and
consumption data. There is also some uncertainty to which materials produced are exported to Canada. As
indicated in the introduction to this section, the trend for integrated facilities has moved to more use of EAFs and
fewer BOFs. This trend may not be completely captured in the current data which also increases uncertainty. EPA
currently uses an uncertainty range of ±10 percent for the primary data inputs (e.g., consumption and production
values for each production process, heat and carbon content values) to calculate overall uncertainty from iron and
steel production, consistent with the ranges in Table 4.4 of the 2006IPCC Guidelines. During EPA's discussion with
AISI, AISI noted that an uncertainty range of ±5 percent would be a more appropriate approximation to reflect
their coverage of integrated steel producers in the United States. EPA will continue to assess the best range of
uncertainty for these values. Consistent with the ranges in Table 4.4 of the 2006 IPCC Guidelines, EPA assigned an
uncertainty range of ±25 percent for the Tier 1CO2 emission factors for the sinter, direct reduced iron, and pellet
production processes.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-74 for metallurgical coke
production and iron and steel production. Total CO2 emissions from metallurgical coke production and iron and
steel production for 2020 were estimated to be between 31.4 and 44.2 MMT CO2 Eq. at the 95 percent confidence
level. This indicates a range of approximately 17 percent below and 17 percent above the emission estimate of
35.4 MMT CO2 Eq. Total CH4 emissions from metallurgical coke production and iron and steel production for 2020
were estimated to be between 0.005 and 0.008 MMT CO2 Eq. at the 95 percent confidence level. This indicates a
range of approximately 21 percent below and 23 percent above the emission estimate of 0.007 MMT CO2 Eq.
Table 4-74: Approach 2 Quantitative Uncertainty Estimates for CO2 and ChU Emissions from
Iron and Steel Production and Metallurgical Coke Production (MMT CO2 Eq. and Percent)
Source
Gas
2021 Emission Estimate
(MMT C02 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Metallurgical Coke & Iron
and Steel Production
Metallurgical Coke & Iron
and Steel Production
C02
35.4
31.4
44.2
-17%
+17%
ch4
+
+
+
-21%
+23%
+ Does not exceed 0.05 MMT C02 Eq.
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
Industrial Processes and Product Use 4-87
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
QA/QC and Verification
For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
Chapter 6 of the 2006IPCC Guidelines, see the QA/QC and Verification Procedures section in the introduction of
the IPPU chapter.
Recalculations Discussion
Recalculations were performed for the year 2020 with updated values for DRI, pig iron, and scrap steel
consumption for both BOF and EAF steel production. Compared to the previous Inventory, CO2 emissions from
steel production increased by less than 1 percent (7 kt CO2).
In addition, for the current Inventory, C02-equivalent estimates of CH4 emissions from sinter production have been
revised to reflect the 100-year global warming potentials (GWPs) provided in the IPCC Fifth Assessment Report
(AR5) (IPCC 2013). AR5 GWP values differ slightly from those presented in the IPCC Fourth Assessment Report
(AR4) (IPCC 2007) (used in the previous Inventories). The AR5 GWPs have been applied across the entire time
series for consistency. The GWP of C02-equivalent CH4 increased from 25 to 28 between the AR4 and AR5 reports,
leading to an overall increase in calculated C02-equivalent Cm emissions. Compared to the previous Inventory,
which applied 100-year GWP values from AR4, annual CFU emissions from sinter production increased by 12
percent each year, ranging from 0.78 kt CO2 Eq. in 2009 to 2.6 kt CO2 Eq. in 1993. The net impact on the entire
category from these updates was an annual 0.002 percent increase in emissions for each year of the time series,
reflecting the relative low contribution of CH4 emissions to the overall category. Further discussion on this update
and the overall impacts of updating the Inventory GWP values to reflect the IPCC Fifth Assessment Report can be
found in Chapter 9, Recalculations and Improvements.
Planned Improvements
Significant activity data for 2021 and 2020 were not available for this report and were estimated using 2019 values
and adjusted using GHGRP emissions data. EPA will continue to explore sources of 2021 and 2020 data and other
estimation approaches. EPA will evaluate and analyze data reported under EPA's GHGRP to improve the emission
estimates for Iron and Steel Production process categories. Particular attention will be made to ensure time-series
consistency of the emissions estimates presented in future Inventory reports, consistent with IPCC and UNFCCC
guidelines. This is required as the facility-level reporting data from EPA's GHGRP, with the program's initial
requirements for reporting of emissions in calendar year 2010, are not available for all inventory years (i.e., 1990
through 2009) as required for this Inventory. In implementing improvements and integration of data from EPA's
GHGRP, the latest guidance from the IPCC on the use of facility-level data in national inventories will be relied
upon.57 This is a near to medium-term improvement, and per preliminary work, EPA estimates that the earliest
this improvement could be incorporated is the 2024 Inventory submission.
Additional improvements include accounting for emission estimates for the production of metallurgical coke in the
Energy chapter as well as identifying the amount of carbonaceous materials, other than coking coal, consumed at
merchant coke plants. Other potential improvements include identifying the amount of coal used for direct
injection and the amount of coke breeze, coal tar, and light oil produced during coke production. Efforts will also
be made to identify information to better characterize emissions from the use of process gases and fuels within
the Energy and IPPU chapters. Additional efforts will be made to improve the reporting between the IPPU and
Energy chapters, particularly the inclusion of a quantitative summary of the carbon balance in the United States.
This planned improvement is a long-term improvement and is still in development. It is not included in this current
Inventory report. EPA estimates that the earliest this improvement could be incorporated is the 2024 Inventory
submission.
57 See http://www.ipcc-nggiD.iges.or.jp/public/tb/TFl Technical Bulletin l.pdf.
4-88 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 4.18 Ferroalloy Production (CRF Source
2 Category 2C2)
3 Carbon dioxide (CO2) and methane (CH4) are emitted from the production of several ferroalloys. Ferroalloys are
4 composites of iron (Fe) and other elements such as silicon (Si), manganese (Mn), and chromium (Cr). Emissions
5 from fuels consumed for energy purposes during the production of ferroalloys are accounted for in the Energy
6 chapter. Emissions from the production of two types of ferrosilicon (25 to 55 percent and 56 to 95 percent silicon),
7 silicon metal (96 to 99 percent silicon), and miscellaneous alloys (32 to 65 percent silicon) have been calculated.
8 Emissions from the production of ferrochromium and ferromanganese are not included because of the small
9 number of manufacturers of these materials in the United States. Government information disclosure rules
10 prevent the publication of production data for these production facilities. Additionally, production of
11 ferrochromium in the United States ceased in 2009 (USGS 2013a).
12 Similar to emissions from the production of iron and steel, CO2 is emitted when metallurgical coke is oxidized
13 during a high-temperature reaction with iron and the selected alloying element. Due to the strong reducing
14 environment, CO is initially produced and eventually oxidized to CO2. A representative reaction equation for the
15 production of 50 percent ferrosilicon (FeSi) is given below:
16 Fe203 + 2Si02 + 7C —> 2FeSi + 7C0
17 While most of the carbon contained in the process materials is released to the atmosphere as CO2, a percentage is
18 also released as CFU and other volatiles. The amount of CH4 that is released is dependent on furnace efficiency,
19 operation technique, and control technology.
20 Ferroalloys are used to alter the material properties of the steel. Ferroalloys are produced in conjunction with the
21 iron and steel industry, often at co-located facilities, and production trends closely follow that of the iron and steel
22 industry. As of 2018,11 facilities in the United States produce ferroalloys (USGS 2022b).
23 Emissions of CO2 from ferroalloy production in 2021 were 1.6 MMT CO2 Eq. (1,567 kt CO2) (see Table 4-75 and
24 Table 4-76), which is a 14 percent increase since 2020 and a 27 percent reduction since 1990. Emissions of CFU
25 from ferroalloy production in 2021 were 0.01 MMT CO2 Eq. (0.4 kt CH4), which is a 14 percent increase since 2020
26 and a 35 percent decrease since 1990. The decrease in emissions since 1990 can largely be attributed to the
27 closure of two facilities in 2018. The increase in emissions from 2020 can be attributed to one facility reopening its
28 ferrosilicon production facility after shutting down in 2020 due to decreased demand during the COVID-19
29 pandemic (USGS 2022a).
30 Table 4-75: CO2 and ChU Emissions from Ferroalloy Production (MMT CO2 Eq.)
Gas 1990 2005 2017 2018 2019 2020 2021
C02 22 1A Z0 2A L6 1A HF
CH4 + + + + + + +
Total 2J2 1A 2.0 2.1 1.6 1.4 1.6
+ Does not exceed 0.05 MMT C02 Eq.
31 Table 4-76: CO2 and ChU Emissions from Ferroalloy Production (kt)
Gas 1990 2005 2017 2018 2019 2020 2021
C02 2,152 1,392 1,975 2,063 1,598 1,377 1,567
CH4 1 + 1 1 + + +
+ Does not exceed 0.5 kt
Industrial Processes and Product Use 4-89
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Methodology and Time-Series Consistency
Emissions of CO2 and CH4 from ferroalloy production were calculated58 using a Tier 1 method from the 2006IPCC
Guidelines by multiplying annual ferroalloy production by material-specific default emission factors provided by
IPCC (IPCC 2006). The Tier 1 equations for CO2 and CFU emissions are as follows:
Equation 4-12: 2006 IPCC Guidelines Tier 1: CO2 Emissions for Ferroalloy Production
(Equation 4.15)
ECo2 = ^(MPi X EFO
i
where,
Eco2 = CO2 emissions, metric tons
MP, = Production of ferroalloy type /', metric tons
EFi = Generic emission factor for ferroalloy type /', metric tons CCh/metric ton specific
ferroalloy product
Equation 4-13: 2006IPCC Guidelines Tier 1: ChU Emissions for Ferroalloy Production
(Equation 4.18)
ECHi = Y^MPi X EF0
i
where,
Ech4 = Cm emissions, kg
MP, = Production of ferroalloy type /', metric tons
EFi = Generic emission factor for ferroalloy type /', kg Cl-U/metric ton specific ferroalloy product
Default emission factors were used because country-specific emission factors are not currently available. The
following emission factors were used to develop annual CO2 and CH4 estimates:
• Ferrosilicon, 25 to 55 percent Si and Miscellaneous Alloys, 32 to 65 percent Si: 2.5 metric tons CCh/metric
ton of alloy produced, 1.0 kg Cl-U/metric ton of alloy produced.
• Ferrosilicon, 56 to 95 percent Si: 4.0 metric tons CCh/metric ton alloy produced, 1.0 kg Cl-U/metric ton of
alloy produced.
• Silicon Metal: 5.0 metric tons CCh/metric ton metal produced, 1.2 kg Cl-U/metric ton metal produced.
It was assumed that 100 percent of the ferroalloy production was produced using petroleum coke in an electric arc
furnace process (IPCC 2006), although some ferroalloys may have been produced with coking coal, wood, other
biomass, or graphite carbon inputs. The amount of petroleum coke consumed in ferroalloy production was
calculated assuming that the petroleum coke used is 90 percent carbon (C) and 10 percent inert material (Onder
and Bagdoyan 1993).
The use of petroleum coke for ferroalloy production is adjusted for within the Energy chapter as this fuel was
consumed during non-energy related activities. Additional information on the adjustments made within the Energy
sector for Non-Energy Use of Fuels is described in both the Methodology section of CO2 from Fossil Fuel
Combustion (3.1 Fossil Fuel Combustion [CRF Source Category 1A]) and Annex 2.1, Methodology for Estimating
Emissions of CO2 from Fossil Fuel Combustion.
58 EPA has not integrated aggregated facility-level GHGRP information to inform these estimates. The aggregated information
(e.g., activity data and emissions) associated with production of ferroalloys did not meet criteria to shield underlying
confidential business information (CBI) from public disclosure.
4-90 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Ferroalloy production data for 1990 through 2021 (see Table 4-77) were obtained from the U.S. Geological Survey
(USGS) through the Minerals Yearbook: Silicon (USGS 1996 through 2022). The following data were available from
the USGS publications for the time series:
• Ferrosilicon, 25 to 55 percent Si: Annual production data were available from 1990 through 2010.
• Ferrosilicon, 56 to 95 percent Si: Annual production data were available from 1990 through 2010.
• Silicon Metal: Annual production data were available from 1990 through 2005. Production data for 2005
were used as estimates for 2006 through 2010 because data for these years were not available due to
government information disclosure rules.
• Miscellaneous Alloys, 32 to 65 percent Si: Annual production data were available from 1990 through
1998. Starting 1999, USGS reported miscellaneous alloys and ferrosilicon containing 25 to 55 percent
silicon as a single category.
Starting with the 2011 publication, USGS ceased publication of production quantity by ferroalloy product and
began reporting all the ferroalloy production data as a single category (i.e., Total Silicon Materials Production). This
is due to the small number of ferroalloy manufacturers in the United States and government information
disclosure rules. Ferroalloy product shares developed from the 2010 production data (i.e., ferroalloy product
production divided by total ferroalloy production) were used with the total silicon materials production quantity to
estimate the production quantity by ferroalloy product type for 2011 through 2021 (USGS 2017 through 2022).
Table 4-77: Production of Ferroalloys (Metric Tons)
Year 1990
2005
2017 2018 2019 2020 2021
Ferrosilicon 25%-55% 321,385
Ferrosilicon 56%-95% 109,566
Silicon Metal 145,744
Misc. Alloys 32-65% 72,442
123,000
86,100
148,000
NA
181,775 189,846 147,034 126,681 144,227
160,390 167,511 129,736 111,778 127,259
175,835 183,642 142,229 122,541 139,514
NA NA NA NA NA
NA (Not Available) for product type, aggregated along with ferrosilicon (25-55% Si)
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2021.
Uncertainty-TO BE UPDATED FOR FINAL INVENTORY REPORT
Annual ferroalloy production was reported by the USGS in three broad categories until the 2010 publication:
ferroalloys containing 25 to 55 percent silicon (including miscellaneous alloys), ferroalloys containing 56 to 95
percent silicon, and silicon metal (through 2005 only, 2005 value used as an estimate for 2006 through 2010).
Starting with the 2011 Minerals Yearbook: Silicon, USGS started reporting all the ferroalloy production under a
single category: total silicon materials production. The total silicon materials quantity was allocated across the
three categories, based on the 2010 production shares for the three categories. Refer to the Methodology section
for further details. Additionally, production data for silvery pig iron (alloys containing less than 25 percent silicon)
are not reported by the USGS to avoid disclosing proprietary company data. Emissions from this production
category, therefore, were not estimated.
Some ferroalloys may be produced using wood or other biomass as a primary or secondary carbon source
(carbonaceous reductants); however, information and data regarding these practices were not available. Emissions
from ferroalloys produced with wood or other biomass would not be counted under this source because wood-
based carbon is of biogenic origin.59 Even though emissions from ferroalloys produced with coking coal or graphite
inputs would be counted in national trends, they may be generated with varying amounts of CO2 per unit of
ferroalloy produced. The most accurate method for these estimates would be to base calculations on the amount
59 Emissions and sinks of biogenic carbon are accounted for in the Land Use, Land-Use Change, and Forestry chapter.
Industrial Processes and Product Use 4-91
-------
1 of reducing agent used in the process, rather than the amount of ferroalloys produced. These data, however, were
2 not available, and are also often considered confidential business information.
3 Emissions of Cm from ferroalloy production will vary depending on furnace specifics, such as type, operation
4 technique, and control technology. Higher heating temperatures and techniques such as sprinkle charging would
5 reduce CFU emissions; however, specific furnace information was not available or included in the Cm emission
6 estimates.
7 Consistent with the ranges for the Tier 1 calculation methodology in Table 4.9 of Section 4.3.3.2 of the 2006IPCC
8 Guidelines, EPA assigned a default uncertainty range of ±25 percent for the primary emission factors (i.e.,
9 ferrosilicon 25-55% Si, ferrosilicon 56-95% Si, and silicon metal), and an uncertainty range of ±5 percent for the
10 2010 production values for ferrosilicon 25-55% Si, ferrosilicon 56-95% Si, and silicon metal production and the
11 2021 total silicon materials production value used to calculate emissions from the overall 2021 ferroalloy
12 production.
13 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-78. Ferroalloy
14 production CO2 emissions from 2021 were estimated to be between 1.2 and 1.6 MMT CO2 Eq. at the 95 percent
15 confidence level. This indicates a range of approximately 13 percent below and 13 percent above the emission
16 estimate of 1.6 MMT CO2 Eq. Ferroalloy production CH4 emissions were estimated to be between a range of
17 approximately 12 percent below and 13 percent above the emission estimate of 0.01 MMT CO2 Eq.
18 Table 4-78: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from
19 Ferroalloy Production (MMT CO2 Eq. and Percent)
2021 Emission Estimate
Source Gas
(MMT C02 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Ferroalloy Production C02 1.6
1.2
1.6
-13%
+13%
Ferroalloy Production CH4 +
+
+
-12%
+13%
+ Does not exceed 0.05 MMT C02 Eq.
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
20 QA/QC and Verification
21 For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
22 Chapter 6 of the 2006 IPCC Guidelines, see the QA/QC and Verification Procedures section in the introduction of
23 the IPPU chapter and Annex 8.
24 Recalculations Discussion
25 Recalculations were completed for 2014 based on revised total silicon materials production data from USGS.
26 Compared to the previous Inventory, estimates of CO2 emissions from ferroalloy production in 2014 increased by
27 4.8 percent (92 kt CO2), and estimates of CFU emissions increased by 4.9 percent (0.026 kt CH4).
28 In addition, for the current Inventory, CC>2-equivalent estimates of total CFU emissions from ferroalloy production
29 have been revised to reflect the 100-year global warming potentials (GWPs) provided in the IPCC Fifth Assessment
30 Report (AR5) (IPCC 2013). AR5 GWP values differ slightly from those presented in the IPCC Fourth Assessment
31 Report (AR4) (IPCC 2007) (used in the previous inventories). The AR5 GWPs have been applied across the entire
32 time series for consistency. The GWP of CH4 increased from 25 to 28 between the AR4 and AR5 reports, leading to
33 an overall increase in CC>2-equivalent estimates for CFU emissions. Compared to the previous Inventory, which
34 applied 100-year GWP values from AR4, annual CFU emissions increased by 12 percent each year, ranging from 1.1
35 kt CO2 Eq. in 2003 to 2.0 kt CO2 Eq. in 1990. The net impact on the entire category from these updates was an
36 average annual 0.09 percent increase in emissions for each year of the time series. Further discussion on this
4-92 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 update and the overall impacts of updating the Inventory GWP values to reflect the IPCC Fifth Assessment Report
2 can be found in Chapter 9, Recalculations and Improvements.
3
4 Planned Improvements
5 Pending available resources and prioritization of improvements for more significant sources, EPA will continue to
6 evaluate and analyze data reported under EPA's GHGRP that would be useful to improve the emission estimates
7 and category-specific QC procedures for the Ferroalloy Production source category. Given the small number of
8 facilities and reporting thresholds, particular attention will be made to ensure completeness and time-series
9 consistency of the emissions estimates presented in future Inventory reports, consistent with IPCC and UNFCCC
10 guidelines. This is required as the facility-level reporting data from EPA's GHGRP, with the program's initial
11 requirements for reporting of emissions in calendar year 2010, are not available for all inventory years (i.e., 1990
12 through 2009) as required for this Inventory. In implementing improvements and integration of data from EPA's
13 GHGRP, the latest guidance from the IPCC on the use of facility-level data in national inventories will be relied
14 upon.60 This is a long-term planned improvement, and EPA is still assessing the possibility of incorporating this
15 improvement into the Inventory. This improvement has not been included in the current Inventory report.
16 4.19 Aluminum Production (CRF Source
I? Category 2C3)
18 Aluminum is a lightweight, malleable, and corrosion-resistant metal that is used in many manufactured products,
19 including aircraft, automobiles, bicycles, and kitchen utensils. As of recent reporting, the United States was the
20 ninth61 largest producer of primary aluminum, tied with Iceland with an aluminum production of 880 thousand
21 metric tons, with approximately 1.3 percent of the world total production (USGS 2021). The United States was also
22 a major importer of primary aluminum. The production of primary aluminum—in addition to consuming large
23 quantities of electricity—results in process-related emissions of carbon dioxide (CO2) and two perfluorocarbons
24 (PFCs): perfluoromethane (CF4) and perfluoroethane (C2F6).
25 Carbon dioxide is emitted during the aluminum smelting process when alumina (aluminum oxide, AI2O3) is reduced
26 to aluminum using the Hall-Heroult reduction process. The reduction of the alumina occurs through electrolysis in
27 a molten bath of natural or synthetic cryolite (NasAIFs). The reduction cells contain a carbon (C) lining that serves
28 as the cathode. Carbon is also contained in the anode, which can be a C mass of paste, coke briquettes, or
29 prebaked C blocks from petroleum coke. During reduction, most of this C is oxidized and released to the
30 atmosphere as CO2.
31 Process emissions of CO2 from aluminum production were estimated to be 1.5 MMT CO2 Eq. (1,541 kt) in 2021 (see
32 Table 4-79). The C anodes consumed during aluminum production consist of petroleum coke and, to a minor
33 extent, coal tar pitch. The petroleum coke portion of the total CO2 process emissions from aluminum production is
34 considered to be a non-energy use of petroleum coke and is accounted for here and not under the CO2 from Fossil
35 Fuel Combustion source category of the Energy sector. Similarly, the coal tar pitch portion of these CO2 process
36 emissions is accounted for here.
60 See http://www.ipcc-nggiD.iges.or.jp/public/tb/TFl Technical Bulletin l.pdf.
61 Based on the U.S. USGS (2021) Aluminum factsheet, assuming all countries grouped under the "other countries" categories
all have lower production than the U.S. Available at: https://pubs.usgs.gov/periodicals/mcs2022/mcs2022-aluminum.pdf
Industrial Processes and Product Use 4-93
-------
l Table 4-79: CO2 Emissions from Aluminum Production (MMT CO2 Eq. and kt)
Year
1990
2005
2017
2018
2019
2020
2021
MMTCOz Eq.
6.8
5 4.1
1.2
1.5
1.9
1.7
1.5
kt C02
6,831
4,142
1,205
1,455
1,880
1,748
1,541
2 In addition to CO2 emissions, the aluminum production industry is also a source of PFC emissions. During the
3 smelting process, when the alumina ore content of the electrolytic bath falls below critical levels required for
4 electrolysis, rapid voltage increases occur, which are termed High Voltage Anode Effects (HVAEs) HVAEs cause C
5 from the anode and fluorine from the dissociated molten cryolite bath to combine, thereby producing fugitive
6 emissions of CF4 and C2F6. In general, the magnitude of emissions for a given smelter and level of production
7 depends on the frequency and duration of these anode effects. As the frequency and duration of the anode effects
8 increase, emissions increase. Another type of anode effect, Low Voltage Anode Effects (LVAEs), became a concern
9 in the early 2010s as the aluminum industry increasingly began to use cell technologies with higher amperage and
10 additional anodes (IPCC 2019). LVAEs emit CF4 and are included in PFC emission totals from 2006 forward.
11 Since 1990, emissions of CF4 and C2F6 have both declined by 95 and 97 percent respectively, to 0.82 MMT CO2 Eq.
12 of CF4 (0.1 kt) and 0.10 MMT CO2 Eq. of C2F6 (0.01 kt) in 2021, respectively, as shown in Table 4-80 and Table 4-81.
13 This decline is due both to reductions in domestic aluminum production and to actions taken by aluminum
14 smelting companies to reduce the frequency and duration of anode effects. These actions include technology and
15 operational changes such as employee training, use of computer monitoring, and changes in alumina feeding
16 techniques. Since 1990, aluminum production has declined by 78 percent, while the combined CF4 and C2F6
17 emission rate (per metric ton of aluminum produced) has been reduced by 78 percent. PFC emissions decreased by
18 approximately 36 percent between 2020 and 2021. Aluminum production also decreased in 2021, down 13
19 percent from 2020.
20 Table 4-80: PFC Emissions from Aluminum Production (MMT CO2 Eq.)
Gas
1990
2005
2017
2018
2019
2020
2021
cf4
16.1
2.6
0.7
1.0
1.1
1.2
0.8
c2f6
0 -y
0.5
0.3
0.3
0.2
0.2
0.1
Total
19.3
3.1
1.0
1.4
1.4
1.4
0.9
Note: Totals may not sum due to independent rounding.
21
22 Table 4-81: PFC Emissions from Aluminum Production (kt)
Gas
1990
2005
2017
2018
2019
2020
2021
cf4
2.4
0.4
0.1
0.2
0.2
0.2
0.1
c2f6
0.29
0.05
0.03
0.03
0.03
0.02
0.01
23 In 2021, U.S. primary aluminum production totaled approximately 0.88 million metric tons, a 13 percent decrease
24 from 2020 production levels (USGS 2022). In 2021, three companies managed production at six operational
25 primary aluminum smelters in five states. Two smelters operated at full capacity during 2021, while four smelters
26 operated at reduced capacity (USGS 2022). Domestic smelters were operating at about 55 percent of capacity of
27 1.64 million tons per year at year end 2021 (USGS 2022).
28 Methodology and Time-Series Consistency
29 Process CO2 and PFC (i.e., CF4 and C2F6) emission estimates from primary aluminum production for 2010 through
30 2021 are available from EPA's GHGRP Subpart F (Aluminum Production) (EPA 2022). Under EPA's GHGRP, facilities
31 began reporting primary aluminum production process emissions (for 2010) in 2011; as a result, GHGRP data (for
32 2010 through 2021) are available to be incorporated into the Inventory. EPA's GHGRP mandates that all facilities
33 that contain an aluminum production process must report: CF4 and C2F6 emissions from anode effects in all
4-94 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 prebake and S0derberg electrolysis cells, CO2 emissions from anode consumption during electrolysis in all prebake
2 and S0derberg cells, and all CO2 emissions from onsite anode baking. To estimate the process emissions, EPA's
3 GHGRP uses the process-specific equations detailed in Subpart F (aluminum production).62 These equations are
4 based on the Tier 2/Tier 3 IPCC (2006) methods for primary aluminum production, and Tier 1 methods when
5 estimating missing data elements. It should be noted that the same methods (i.e., 2006 IPCC Guidelines) were used
6 for estimating the emissions prior to the availability of the reported GHGRP data in the Inventory. Prior to 2010,
7 aluminum production data were provided through EPA's Voluntary Aluminum Industrial Partnership (VAIP).
8 As previously noted, the use of petroleum coke for aluminum production is adjusted for within the Energy chapter
9 as this fuel was consumed during non-energy related activities. Additional information on the adjustments made
10 within the Energy sector for Non-Energy Use of Fuels is described in both the Methodology section of CO2 from
11 Fossil Fuel Combustion (3.1 Fossil Fuel Combustion [CRF Source Category 1A]) and Annex 2.1, Methodology for
12 Estimating Emissions of CO2 from Fossil Fuel Combustion.
13 Process CO2 Emissions from Anode Consumption and Anode Baking
14 Carbon dioxide emission estimates for the years prior to the introduction of EPA's GHGRP in 2010 were estimated
15 using 2006 IPCC Guidelines methods, but individual facility reported data were combined with process-specific
16 emissions modeling. These estimates were based on information previously gathered from EPA's Voluntary
17 Aluminum Industrial Partnership (VAIP) program, U.S. Geological Survey (USGS) Mineral Commodity reviews, and
18 The Aluminum Association (USAA) statistics, among other sources. Since pre- and post-GHGRP estimates use the
19 same methodology, emission estimates are comparable across the time series.
20 Most of the CO2 emissions released during aluminum production occur during the electrolysis reaction of the C
21 anode, as described by the following reaction:
22 2AI2O3 + 3C -> 4A1 + 3C02
23 For prebake smelter technologies, CO2 is also emitted during the anode baking process. These emissions can
24 account for approximately 10 percent of total process CO2 emissions from prebake smelters.
25 Depending on the availability of smelter-specific data, the CO2 emitted from electrolysis at each smelter was
26 estimated from: (1) the smelter's annual anode consumption, (2) the smelter's annual aluminum production and
27 rate of anode consumption (per ton of aluminum produced) for previous and/or following years, or (3) the
28 smelter's annual aluminum production and IPCC default CO2 emission factors. The first approach tracks the
29 consumption and carbon content of the anode, assuming that all C in the anode is converted to CO2. Sulfur, ash,
30 and other impurities in the anode are subtracted from the anode consumption to arrive at a C consumption figure.
31 This approach corresponds to either the IPCC Tier 2 or Tier 3 method, depending on whether smelter-specific data
32 on anode impurities are used. The second approach interpolates smelter-specific anode consumption rates to
33 estimate emissions during years for which anode consumption data are not available. This approach avoids
34 substantial errors and discontinuities that could be introduced by reverting to Tier 1 methods for those years. The
35 last approach corresponds to the IPCC Tier 1 method (IPCC 2006) and is used in the absence of present or historic
36 anode consumption data.
37 The equations used to estimate CO2 emissions in the Tier 2 and 3 methods vary depending on smelter type (IPCC
38 2006). For Prebake cells, the process formula accounts for various parameters, including net anode consumption,
39 and the sulfur, ash, and impurity content of the baked anode. For anode baking emissions, the formula accounts
40 for packing coke consumption, the sulfur and ash content of the packing coke, as well as the pitch content and
41 weight of baked anodes produced. For S0derberg cells, the process formula accounts for the weight of paste
62 Code of Federal Regulations, Title 40: Protection of Environment, Part 98: Mandatory Greenhouse Gas Reporting, Subpart
F—Aluminum Production. See https://www.ecfr.gov/cgi-bin/text-
idx?SID=24a41781dfe4218b339e914de03e8727&mc=true&node=pt40.23.98&rgn=div5#sp40.23.98.f.
Industrial Processes and Product Use 4-95
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
consumed per metric ton of aluminum produced, and pitch properties, including sulfur, hydrogen, and ash
content.
Through the VAIP, anode consumption (and some anode impurity) data have been reported for 1990, 2000, 2003,
2004, 2005, 2006, 2007, 2008, and 2009. Where available, smelter-specific process data reported under the VAIP
were used; however, if the data were incomplete or unavailable, information was supplemented using industry
average values recommended by IPCC (2006). Smelter-specific CO2 process data were provided by 18 of the 23
operating smelters in 1990 and 2000, by 14 out of 16 operating smelters in 2003 and 2004,14 out of 15 operating
smelters in 2005,13 out of 14 operating smelters in 2006, 5 out of 14 operating smelters in 2007 and 2008, and 3
out of 13 operating smelters in 2009. For years where CO2 emissions data or CO2 process data were not reported
by these companies, estimates were developed through linear interpolation, and/or assuming representative (e.g.,
previously reported or industry default) values.
In the absence of any previous historical smelter-specific process data (i.e., 1 out of 13 smelters in 2009; 1 out of
14 smelters in 2006, 2007, and 2008; 1 out of 15 smelters in 2005; and 5 out of 23 smelters between 1990 and
2003), CO2 emission estimates were estimated using Tier 1 S0derberg and/or Prebake emission factors (metric ton
of CO2 per metric ton of aluminum produced) from IPCC (2006).
Process PFC Emissions from Anode Effects
High Voltage Anode Effects
Smelter-specific PFC emissions from aluminum production for 2010 through 2021 were reported to EPA under its
GHGRP. To estimate their PFC emissions from HVAEs and report them under EPA's GHGRP, smelters use an
approach identical to the Tier 3 approach in the 2006 IPCC Guidelines (IPCC 2006). Specifically, they use a smelter-
specific slope coefficient as well as smelter-specific operating data to estimate an emission factor using the
following equation:
PFC = S xAE
AE = F xD
where,
PFC
= CF4 or C2F6, kg/MT aluminum
S
= Slope coefficient, PFC/AE
AE
= Anode effect, minutes/cell-day
F
= Anode effect frequency per cell-day
D
= Anode effect duration, minutes
They then multiply this emission factor by aluminum production to estimate PFC emissions from HVAEs. All U.S.
aluminum smelters are required to report their emissions under EPA's GHGRP.
Perfluorocarbon emissions for the years prior to 2010 were estimated using the same equation, but the slope-
factor used for some smelters was technology-specific rather than smelter-specific, making the method a Tier 2
rather than a Tier 3 approach for those smelters. Emissions and background data were reported to EPA under the
VAIP. For 1990 through 2009, smelter-specific slope coefficients were available and were used for smelters
representing between 30 and 94 percent of U.S. primary aluminum production. The percentage changed from year
to year as some smelters closed or changed hands and as the production at remaining smelters fluctuated. For
smelters that did not report smelter-specific slope coefficients, IPCC technology-specific slope coefficients were
applied (IPCC 2006). The slope coefficients were combined with smelter-specific anode effect data collected by
aluminum companies and reported under the VAIP to estimate emission factors over time. For 1990 through 2009,
smelter-specific anode effect data were available for smelters representing between 80 and 100 percent of U.S.
primary aluminum production. Where smelter-specific anode effect data were not available, representative values
(e.g., previously reported or industry averages) were used.
4-96 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
For all smelters, emission factors were multiplied by annual production to estimate annual emissions at the
smelter level. For 1990 through 2009, smelter-specific production data were available for smelters representing
between 30 and 100 percent of U.S. primary aluminum production. (For the years after 2000, this percentage was
near the high end of the range.) Production at non-reporting smelters was estimated by calculating the difference
between the production reported under VAIP and the total U.S. production supplied by USGS, and then allocating
this difference to non-reporting smelters in proportion to their production capacity. Emissions were then
aggregated across smelters to estimate national emissions.
Table 4-82: Summary of HVAE Emissions
Year
1990
2005
2017
2018
2019
2020
2021
MMTCOz Eq.
19.3
3.1
0.9
1.4
1.4
1.4
0.9
Low Voltage Anode Effects
LVAE emissions of CF4 were estimated for 2006 through 2021 based on the Tier 1 (technology-specific, production-
based) method in the 2019 Refinement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories (IPCC
2019). Prior to 2006, LVAE emissions are believed to have been negligible.63 The Tier 1 method is used in the LVAE
emissions calculations from aluminum production in the absence of smelter-specific data available to quantify the
LVAE-specific process emissions. National aluminum production estimates (allocated to smelters as described
below) and the technology used in individual smelters were the best available data to perform the emissions
calculations, as smelter-specific production data is not publicly available.
The following equation was used to estimate LVAE PFC emissions:
Equation 4-14: CF4 Emissions Resulting from Low Voltage Anode Effects
In the LVAE emissions calculations, the Metal Production (MP) factor is calculated differently for the years 2006
through 2009 than for 2010 and beyond. For years prior to GHGRP reporting (2006 through 2009), the MP factor is
calculated by dividing the annual production reported by USGS with the total U.S. capacity reported for this
specific year, based on the USGS yearbook and applying this national utilization factor to each facility's production
capacity to obtain an estimated facility production value. For GHGRP reporting years (2010+), the methodology to
calculate the MP value was changed to allocate the total annual production reported by USAA, based on the
distribution of CO2 emissions amongst the operating smelters in a specific year. The latter improves the accuracy of
the LVAE emissions estimates over assuming capacity utilization is the same at all smelters. The main drawback of
using this methodology to calculate the MP factor is that, in some instances, it led to production estimates that are
slightly larger (less than six percent) than the production capacity reported that year. In practice, this is most likely
explained by the differences in process efficiencies at each facility and to a lesser extent, differences in
measurements and methods used by each facility to obtain their CO2 estimates and the degree of uncertainty in
the USGS annual production reporting.
63 The 2019 Refinement states, "Since 2006, the global aluminum industry has undergone changes in technology and operating
conditions that make LVAE emissions much more prevalent12; these changes have occurred not only through uptake of newer
technologies (e.g., PFPBL to PFPBM) but also during upgrades within the same technology in order to maximize productivity and
reduce energy use" (IPCC 2019). Footnote #12 uses the example of PFPBL, which is prevalent in the United States, as an older
technology that has been upgraded.
LVAE ECF4 = LVAEEFCF4 X MP
where,
LVAE Ecf4
LVAE EFcf4
MP
LVAE emissions of CF4 from aluminum production, kg CF4
LVAE emission factor for CF4 (default by cell technology type)
metal production by cell technology type, tons Al.
Industrial Processes and Product Use 4-97
-------
1 Once LVAE emissions were estimated, they were then combined with HVAE emissions estimates to calculate total
2 PFC emissions from aluminum production.
3 Table 4-83: Summary of LVAE Emissions
Year
2006
2017
2018
2019
2020
2021
MMT CO? Eq.
0.13
0.05
0.05
0.07
0.06
0.05
4 Production Data
5 Between 1990 and 2009, production data were provided under the VAIP by 21 of the 23 U.S. smelters that
6 operated during at least part of that period. For the non-reporting smelters, production was estimated based on
7 the difference between reporting smelters and national aluminum production levels as reported to USGS, with
8 allocation to specific smelters based on reported production capacities (USGS 1990 through 2009).
9 National primary aluminum production data for 2010 through 2021 were compiled using USGS Mineral Industry
10 Surveys, and the USGS Mineral Commodity Summaries.
11 Table 4-84: Production of Primary Aluminum (kt)
Year
1990
2005
2017
2018
2019
2020
2021
Production (kt)
4,048
2,481
741
891
1,093
1,012
880
12 Methodological approaches were applied to the entire time-series to ensure time-series consistency from 1990
13 through 2020.
14 Uncertainty
15 Uncertainty was estimated for the CO2, CF4, and C2F6 emission values reported by each individual facility to EPA's
16 GHGRP, taking into consideration the uncertainties associated with aluminum production, anode effect minutes,
17 and slope factors. The uncertainty bounds used for these parameters were established based on information
18 collected under the VAIP and held constant through 2021. Uncertainty surrounding the reported CO2, CF4, and C2F6
19 emission values were determined to have a normal distribution with uncertainty ranges of approximately 6
20 percent below to 6 percent above, 16 percent below to 16 percent above, and 20 percent below to 20 percent
21 above their 2021 emission estimates, respectively.
22 For LVAE, since emission values were not reported through EPA's GHGRP but estimated instead through a Tier 1
23 methodology, the uncertainty analysis examined uncertainty associated with primary capacity data as well as
24 technology-specific emission factors. Uncertainty for each facility's primary capacity, reported in the USGS
25 Yearbook, was estimated to have a Pert Beta distribution with an uncertainty range of 7 percent below to 7
26 percent above the capacity estimates based on the uncertainty of reported capacity data, the number of years
27 since the facility reported new capacity data, and uncertainty in capacity utilization. Uncertainty was applied to
28 LVAE emission factors according to technology using the uncertainty ranges provided in the 2019 Refinement to
29 the 2006IPCC Guidelines. An uncertainty range for Horizontal Stud S0derberg (HSS) technology was not provided
30 in the 2019 Refinement to the 2006 IPCC Guidelines due to insufficient data, so a normal distribution and
31 uncertainty range of ±99 percent was applied for that technology based on expert judgment. A Monte Carlo
32 analysis was applied to estimate the overall uncertainty of the CO2, CF4, and C2F6 emission estimates for the U.S.
33 aluminum industry as a whole, and the results are provided below.
34 The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 4-85. Aluminum
35 production-related CO2 emissions were estimated to be between 1.50 and 1.58 MMT CO2 Eq. at the 95 percent
36 confidence level. This indicates a range of approximately 2 percent below to 3 percent above the emission
37 estimate of 1.54 MMT CO2 Eq. Also, production-related CF4 emissions were estimated to be between 0.75 and 0.89
38 MMT CO2 Eq. at the 95 percent confidence level. This indicates a range of approximately 9 percent below to 9
39 percent above the emission estimate of 0.82 MMT CO2 Eq. Aluminum production-related C2F6 emissions were
4-98 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
estimated to be between 0.09 and 0.11 MMT CO2 Eq. at the 95 percent confidence level. This indicates a range of
approximately 11 percent below to 11 percent above the emission estimate of 0.10 MMT CO2 Eq. Finally,
Aluminum production-related aggregated PFCs emissions were estimated to be between 0.85 and 0.99 MMT CO2
Eq. at the 95 percent confidence level. This indicates a range of approximately 8 percent below to 8 percent above
the emission estimate of 0.922 MMT CO2 Eq.
Table 4-85: Approach 2 Quantitative Uncertainty Estimates for CO2 and PFC Emissions from
Aluminum Production (MMT CO2 Eq. and Percent)
Source
Gas
2021 Emission Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Aluminum Production
C02
1.54
1.50
1.58
-2%
3%
Aluminum Production
cf4
0.82
0.75
0.89
-9%
9%
Aluminum Production
c2f6
0.10
0.09
0.11
-11%
11%
Aluminum Production
PFCs
0.92
0.85
0.99
-8%
8%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S. Inventory
QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006IPCC Guidelines as described in the
introduction of the IPPU chapter (see Annex 8 for more details). For the GHGRP data, EPA verifies annual facility-
level reports through a multi-step process (e.g., including a combination of pre-and post-submittal electronic
checks and manual reviews by staff) to identify potential errors and ensure that data submitted to EPA are
accurate, complete, and consistent (EPA 20 15).64 Based on the results of the verification process, EPA follows up
with facilities to resolve mistakes that may have occurred. The post-submittals checks are consistent with a
number of general and category-specific QC procedures, including: range checks, statistical checks, algorithm
checks, and year-to-year checks of reported data and emissions.
Recalculations Discussion
The total primary aluminum production estimates were updated to reflect data reported to the USGS (as detailed
in Production Data section above) for all years 1990 to 2021. Previously, production estimates from the U.S.
Aluminum Association and other external resources were used for some years. The data from USGS are compiled
from the U.S. Geological Survey monthly surveys sent to the primary aluminum smelters owned by the companies
operating in the United States. In recent years, all companies who were sent the surveys responded, making USGS
data the most accurate available. These data source modifications did lead to minor differences in the greenhouse
gas emissions calculations for some years between 2000 and 2009. No historical or current production estimates
publicly available were found to be broken down into smelter specific production estimates. In addition, for the
current Inventory, C02-equivalent emissions totals of CF4 and C2F6 from Aluminum production have been revised to
reflect the 100-year global warming potentials (GWPs) provided in the IPCC Fifth Assessment Report (AR5) (IPCC
2013). AR5 GWP values differ slightly from those presented in the IPCC Fourth Assessment Report (AR4) (IPCC
2007) (used in the previous inventories). The AR5 GWPs have been applied across the entire time series for
consistency. The GWPs of CF4 and C2F6 have decreased, leading to an overall decrease in calculated C02-equivalent
emissions from Aluminum production. Compared to the previous Inventory which applied 100-year GWP values
from AR4, the average annual change in CF4 emissions was a 10 percent decrease and the average annual change
in CQ2-equivalent C2F6 emissions was a 9 percent decrease for the time series. The net impact from these updates
64 GHGRP Report Verification Factsheet. See https://www.epa.gov/sites/production/files/2Q15-
07/documents/ghgrp verification factsheet.pdf.
Industrial Processes and Product Use 4-99
-------
1 was an average annual 10 percent decrease in CCh-equivalent total PFC emissions for the time series. Further
2 discussion on this update and the overall impacts of updating the Inventory GWP values to reflect the IPCC Fifth
3 Assessment Report can be found in Chapter 9, Recalculations and Improvements.
4 4.20 Magnesium Production and Processing
s (CRF Source Category 2C4)
6 The magnesium metal production and casting industry uses sulfur hexafluoride (SFs) as a cover gas to prevent the
7 rapid oxidation of molten magnesium in the presence of air. Sulfur hexafluoride has been used in this application
8 around the world for more than thirty years. A dilute gaseous mixture of SF6 with dry air and/or carbon dioxide
9 (CO2) is blown over molten magnesium metal to induce and stabilize the formation of a protective crust. A small
10 portion of the SF6 reacts with the magnesium to form a thin molecular film of mostly magnesium oxide and
11 magnesium fluoride. The amount of SF6 reacting in magnesium production and processing is considered to be
12 negligible and thus all SF6 used is assumed to be emitted into the atmosphere. Alternative cover gases, such as
13 AM-cover™ (containing HFC-134a), Novec™ 612 (FK-5-1-12) and dilute sulfur dioxide (SO2) systems can and are
14 being used by some facilities in the United States. However, many facilities in the United States are still using
15 traditional SF6 cover gas systems. Carbon dioxide is also released during primary magnesium production if
16 carbonate based raw materials, such as dolomite, are used. During the processing of these raw materials to
17 produce magnesium, calcination occurs which results in a release of CO2 emissions.
18 The magnesium industry emitted 1.1 MMT CO2 Eq. (0.05 kt) of SF6, 0.04 MMT CO2 Eq. (0.03 kt) of HFC-134a, and
19 0.003 MMT CO2 Eq. (2.9 kt) of CO2 in 2021. This represents an increase of approximately 24 percent from total
20 2020 emissions (see Table 4-86 and Table 4-87) and an increase in SF6 emissions by 26 percent. In 2021, total HFC-
21 134a emissions decreased from 0.052 MMT CO2 Eq. to 0.040 MMT CO2 Eq., or a 24 percent decrease as compared
22 to 2020 emissions. FK 5-1-12 emissions in 2021 were consistent with 2020. The emissions of the carrier gas, CO2,
23 decreased from 2.97 kt in 2020 to 2.92 kt in 2021, or 2 percent.
24 Table 4-86: SF6, HFC-134a, FK 5-1-12 and CO2 Emissions from Magnesium Production and
25 Processing (MMT CO2 Eq.)
Year
1990
2005
2017
2018
2019
2020
2021
sf6
5.4
2.9
1.0
1.1
0.9
0.9
1.1
HFC-134a
0.0
0.0
0.1
0.1
0.1
0.1
+
C02
0.1
+
+
+
+
+
+
FK 5-1-12°
0.0
0.0
+
+
+
+
+
Total
5.5
2.9
1.1
1.1
1.0
0.9
1.2
+ Does not exceed 0.05 MMT C02 Eq.
a Emissions of FK 5-1-12 are not included in totals.
Note: Totals may not sum due to independent rounding.
26 Table 4-87: SF6, HFC-134a, FK 5-1-12 and CO2 Emissions from Magnesium Production and
27 Processing (kt)
Year
1990
2005
2017
2018
2019
2020
2021
sf6
0.2
0.1
+
+
+
+
+
HFC-134a
0.0
0.0
0.1
0.1
+
+
+
C02
128.4
3.3
3.3
1.6
2.4
3.0
2.9
FK 5-1-12°
0.0
o.o
+
+
+
+
+
+ Does not exceed 0.05 kt
a Emissions of FK 5-1-12 are not included in totals.
4-100 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Methodology and Time-Series Consistency
Emission estimates for the magnesium industry incorporate information provided by industry participants in EPA's
SFs Emission Reduction Partnership for the Magnesium Industry as well as emissions data reported through
Subpart T (Magnesium Production and Processing) of EPA's GHGRP. The Partnership started in 1999 and, in 2010,
participating companies represented 100 percent of U.S. primary and secondary production and 16 percent of the
casting sector production (i.e., die, sand, permanent mold, wrought, and anode casting). SF6 emissions for 1999
through 2010 from primary production, secondary production (i.e., recycling), and die casting were generally
reported by Partnership participants. Partners reported their SF6 consumption, which is assumed to be equivalent
to emissions. Along with SF6, some Partners reported their HFC-134a and FK 5-1-12 consumed, which is also
assumed to be equal to emissions. The last reporting year under the Partnership was 2010. Emissions data for
2011 through 2020 are obtained through EPA's GHGRP. Under the program, owners or operators of facilities that
have a magnesium production or casting process must report emissions from use of cover or carrier gases, which
include SF6, HFC-134a, FK 5-1-12 and CO2. Consequently, cover and carrier gas emissions from magnesium
production and processing were estimated for three time periods, depending on the source of the emissions data:
1990 through 1998 (pre-EPA Partnership), 1999 through 2010 (EPA Partnership), and 2011 through 2021 (EPA
GHGRP). The methodologies described below also make use of magnesium production data published by the U.S.
Geological Survey (USGS) as available.
1990 through 1998
To estimate emissions for 1990 through 1998, industry SF6 emission factors were multiplied by the corresponding
metal production and consumption (casting) statistics from USGS. For this period, it was assumed that there was
no use of HFC-134a or FK 5-1-12 cover gases, and hence emissions were not estimated for these alternatives.
Sulfur hexafluoride emission factors from 1990 through 1998 were based on a number of sources and
assumptions. Emission factors for primary production were available from U.S. primary producers for 1994 and
1995. The primary production emission factors were 1.2 kg SF6 per metric ton for 1990 through 1993, and 1.1 kg
SFs per metric ton for 1994 through 1997. The emission factor for secondary production from 1990 through 1998
was assumed to be constant at the 1999 average Partner value. An emission factor for die casting of 4.1 kg SF6 per
metric ton, which was available for the mid-1990s from an international survey (Gjestland and Magers 1996), was
used for years 1990 through 1996. For 1996 through 1998, the emission factor for die casting was assumed to
decline linearly to the level estimated based on Partner reports in 1999. This assumption is consistent with the
trend in SF6 sales to the magnesium sector that was reported in the RAND survey of major SF6 manufacturers,
which showed a decline of 70 percent from 1996 to 1999 (RAND 2002). Sand casting emission factors for 1990
through 2001 were assumed to be the same as the 2002 emission factor. The emission factors for the other
processes (i.e., permanent mold, wrought, and anode casting), about which less is known, were assumed to remain
constant at levels defined in Table 4-86. The emission factors for the other processes (i.e., permanent mold,
wrought, and anode casting) were based on discussions with industry representatives.
The quantities of CO2 carrier gas used for each production type have been estimated using the 1999 estimated CO2
emissions data and the annual calculated rate of change of SF6 use in the 1990 through 1999 time period. For each
year and production type, the rate of change of SF6 use between the current year and the subsequent year was
first estimated. This rate of change was then applied to the CO2 emissions of the subsequent year to determine the
CO2 emission of the current year.
Carbon dioxide emissions from the calcination of dolomite in the primary production of magnesium were
calculated based on the 2006IPCC Guidelines Tier 2 method by multiplying the estimated primary production of
magnesium by an emissions factor of 3.62 kilogram of CO2 per kilogram of magnesium produced.65 For 1990
through 1998, production was estimated to be equal to the production capacity of the facility.
65 See https://www.ipcc-nggip.iges.or.jP/public/2CX36gl/pdf/3 Volume3/V3 4 Ch4 Metal lndustry.pdf.
Industrial Processes and Product Use 4-101
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
1999 through 2010
The 1999 through 2010 emissions from primary and secondary production were based on information provided by
EPA's industry Partners. In some instances, there were years of missing Partner data, including SF6 consumption
and metal processed. For these situations, emissions were estimated through interpolation where possible, or by
holding company-reported emissions (as well as production) constant from the previous year. For alternative cover
gases, including HFC-134a and FK 5-1-12, mainly reported data was relied upon. That is, unless a Partner reported
using an alternative cover gas, it was not assumed it was used. Emissions of alternate gases were also estimated
through linear interpolation where possible.
The die casting emission estimates for 1999 through 2010 were also based on information supplied by industry
Partners. When a Partner was determined to be no longer in production, its metal production and usage rates
were set to zero. Missing data on emissions or metal input was either interpolated or held constant at the last
available reported value. In 1999 through 2010, Partners were assumed to account for all die casting tracked by
USGS. For 1999, die casters who were not Partners were assumed to be similar to Partners who cast small parts.
Due to process requirements, these casters consume larger quantities of SF6 per metric ton of processed
magnesium than casters that process large parts. Consequently, emission estimates from this group of die casters
were developed using an average emission factor of 5.2 kg SF6 per metric ton of magnesium. This emission factor
was developed using magnesium production and SF6 usage data for the year 1999. In 2008, the derived emission
factor for die casting began to increase after many years of largely decreasing emission factors. As determined
through an analysis of activity data reported from the USGS, this increase is due to a temporary decrease in
production at many facilities between 2008 and 2010, which reflects the change in production that occurred
during the recession.
The emissions from other casting operations were estimated by multiplying emission factors (kg SF6 per metric ton
of metal produced or processed) by the amount of metal produced or consumed from USGS, with the exception of
some years for which Partner sand casting emissions data are available. The emission factors for sand casting
activities were acquired through the data reported by the Partnership for 2002 to 2006. For 1999 through 2001,
the sandcasting emission factor was held constant at the 2002 Partner-reported level. For 2007 through 2010, the
sandcasting Partner did not report and the reported emission factor from 2005 was applied to the Partner and to
all other sand casters. Activity data for 2005 was obtained from USGS (USGS 2005b).
The emission factors for primary production, secondary production and sand casting for the 1999 to 2010 are not
published to protect company-specific production information. However, the emission factor for primary
production has not risen above the average 1995 Partner value of 1.1 kg SF6 per metric ton. The emission factors
for the other industry sectors (i.e., permanent mold, wrought, and anode casting) were based on discussions with
industry representatives. The emission factors for casting activities are provided below in Table 4-88.
The emissions of HFC-134a and FK-5-1-12 were included in the estimates for only instances where Partners
reported that information to the Partnership. Emissions of these alternative cover gases were not estimated for
instances where emissions were not reported.
Carbon dioxide carrier gas emissions were estimated using the emission factors developed based on GHGRP-
reported carrier gas and cover gas data, by production type. It was assumed that the use of carrier gas, by
production type, is proportional to the use of cover gases. Therefore, an emission factor, in kg CO2 per kg cover gas
and weighted by the cover gases used, was developed for each of the production types. GHGRP data, on which
these emissions factors are based, was available for primary, secondary, die casting and sand casting. The emission
factors were applied to the quantity of all cover gases used (SF6, HFC-134a, and FK-5-1-12) by production type in
this time period for producers that reported CO2 emissions from 2011-2020 through the GHGP. Carrier gas
emissions for the 1999 through 2010 time period were only estimated for those Partner companies that reported
using CO2 as a carrier gas through the GHGRP. Using this approach helped ensure time-series consistency.
Emissions of carrier gases for permanent mold, wrought, and anode processes were estimated using the ratio of
total CO2 emissions to total cover gas emissions for primary, secondary, die and sand in a given year and the total
SFs emissions from each permanent mold, wrought, and anodes processes respectively in that same year. CO2
emissions from the calcination of dolomite were estimated using the same approach as described above. At the
4-102 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 end of 2001, the sole magnesium production plant operating in the United States that produced magnesium metal
2 using a dolomitic process that resulted in the release of CO2 emissions ceased its operations (USGS 1995b through
3 2020).
4 Table 4-88: SF6 Emission Factors (kg SF6 per metric ton of magnesium)
Year
Die Casting3
Permanent Mold
Wrought
Anodes
1999
1.75b
2
1
1
2000
0.72
2
1
1
2001
0.72
2
1
1
2002
0.71
2
1
1
2003
0.81
2
1
1
2004
0.79
2
1
1
2005
0.77
2
1
1
2006
0.88
2
1
1
2007
0.64
2
1
1
2008
0.97
2
1
1
2009
1.41
2
1
1
2010
1.43
2
1
1
a Weighted average includes all die casters, Partners and non-Partners. For
the majority of the time series (2000 through 2010), Partners made up
100 percent of die casters in the United States.
b Weighted average that includes an estimated emission factor of 5.2 kg
SF6 per metric ton of magnesium for die casters that do not participate in
the Partnership.
5 2011 through 2021
6 For 2011 through 2021, for the primary and secondary producers, GHGRP-reported cover and carrier gases
7 emissions data were used. For sand and die casting, some emissions data was obtained through EPA's GHGRP.
8 Additionally, in 2018 a new GHGRP reporter began reporting permanent mold emissions. The balance of the
9 emissions for this industry segment was estimated based on previous Partner reporting (i.e., for Partners that did
10 not report emissions through EPA's GHGRP) or were estimated by multiplying emission factors by the amount of
11 metal produced or consumed. Partners who did not report through EPA's GHGRP were assumed to have continued
12 to emit SF6 at the last reported level, which was from 2010 in most cases, unless publicly available sources
13 indicated that these facilities have closed or otherwise eliminated SF6 emissions from magnesium production (ARB
14 2015). Many Partners that did report through the GHGRP showed increases in SF6 emissions driven by increased
15 production related to a continued economic recovery after the 2008 recession. One Partner in particular reported
16 an anonymously large increase in SF6 emissions from 2010 to 2011, further driving increases in emissions between
17 the two time periods of inventory estimates. All Partners were assumed to have continued to consume magnesium
18 at the last reported level. Where the total metal consumption estimated for the Partners fell below the U.S. total
19 reported by USGS, the difference was multiplied by the emission factors discussed in the section above, i.e., non-
20 partner emission factors. For the other types of production and processing (i.e., permanent mold, wrought, and
21 anode casting), emissions were estimated by multiplying the industry emission factors with the metal production
22 or consumption statistics obtained from USGS (USGS 2022). USGS data for 2021 were not yet available at the time
23 of the analysis, so the 2020 values were held constant through 2021 as an estimate.
24 Emissions of carrier gases for permanent mold, wrought, and anode processes were estimated using an approach
25 consistent with the 1999 through 2010 time series.
26 Methodological approaches were applied to the entire time series to ensure time-series consistency from 1990
27 through 2021. 2006IPCC Guidance methodologies were used throughout the timeseries, mainly either a Tier 2 or
28 Tier 3 approach depending on available data.
Industrial Processes and Product Use 4-103
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Uncertainty
Uncertainty surrounding the total estimated emissions in 2021 is attributed to the uncertainties around SF6, HFC-
134a, and CO2 emission estimates. To estimate the uncertainty surrounding the estimated 2021SF6 emissions from
magnesium production and processing, the uncertainties associated with three variables were estimated: (1)
emissions reported by magnesium producers and processors for 2021 through EPA's GHGRP, (2) emissions
estimated for magnesium producers and processors that reported via the Partnership in prior years but did not
report 2021 emissions through EPA's GHGRP, and (3) emissions estimated for magnesium producers and
processors that did not participate in the Partnership or report through EPA's GHGRP. An uncertainty of 5 percent
was assigned to the emissions (usage) data reported by each GHGRP reporter for all the cover and carrier gases
(per the 2006IPCC Guidelines). If facilities did not report emissions data during the current reporting year through
EPA's GHGRP, SFs emissions data were held constant at the most recent available value reported through the
Partnership. The uncertainty associated with these values was estimated to be 30 percent for each year of
extrapolation (per the 2006 IPCC Guidelines). The uncertainty of the total inventory estimate remained relatively
constant between 2020 and 2021.
Alternate cover gas and carrier gases data was set equal to zero if the facilities did not report via the GHGRP. For
those industry processes that are not represented in the Partnership, such as permanent mold and wrought
casting, SF6 emissions were estimated using production and consumption statistics reported by USGS and
estimated process-specific emission factors (see Table 4-89). The uncertainties associated with the emission
factors and USGS-reported statistics were assumed to be 75 percent and 25 percent, respectively. Emissions
associated with die casting and sand casting activities utilized emission factors based on Partner reported data
with an uncertainty of 75 percent. In general, where precise quantitative information was not available on the
uncertainty of a parameter, a conservative (upper-bound) value was used.
Additional uncertainties exist in these estimates that are not addressed in this methodology, such as the basic
assumption that SF6 neither reacts nor decomposes during use. The melt surface reactions and high temperatures
associated with molten magnesium could potentially cause some gas degradation. Previous measurement studies
have identified SF6 cover gas degradation in die casting applications on the order of 20 percent (Bartos et al. 2007).
Sulfur hexafluoride may also be used as a cover gas for the casting of molten aluminum with high magnesium
content; however, the extent to which this technique is used in the United States is unknown.
The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 4-89. Total emissions
associated with magnesium production and processing were estimated to be between 1.05 and 1.21 MMT CO2 Eq.
at the 95 percent confidence level. This indicates a range of approximately 6.7 percent below to 7.0 percent above
the 2021 emission estimate of 1.13 MMT CO2 Eq. The uncertainty estimates for 2021 are slightly lower to the
uncertainty reported for 2020 in the previous Inventory. This decrease in uncertainty is attributed to the increased
proportion of SF6 emissions that were calculated using data from GHGRP reporting facilities, which are more
accurate than emissions calculated using proxy or estimation methods for non-reporters.
Table 4-89: Approach 2 Quantitative Uncertainty Estimates for SFe, HFC-134a and CO2
Emissions from Magnesium Production and Processing (MMT CO2 Eq. and Percent)
Source
Gas
2021 Emission Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower
Upper
Lower Upper
Bound
Bound
Bound Bound
Magnesium
Production
SF6, HFC-
134a, C02
1.16
1.08
1.24
-6.7% 7.0%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence
interval.
4-104 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S. Inventory
QA/QC plan, which is in accordance with Volume 1, Chapter 6 of 2006IPCC Guidelines as described in the
introduction of the IPPU chapter (see Annex 8 for more details). For the GHGRP data, EPA verifies annual facility-
level reports through a multi-step process (e.g., including a combination of pre-and post-submittal electronic
checks and manual reviews by staff) to identify potential errors and ensure that data submitted to EPA are
accurate, complete, and consistent (EPA 2015).66 Based on the results of the verification process, EPA follows up
with facilities to resolve mistakes that may have occurred. The post-submittals checks are consistent with a
number of general and category-specific QC procedures, including: range checks, statistical checks, algorithm
checks, and year-to-year checks of reported data and emissions.
Recalculations Discussion
GHGRP-reported emissions for CO2 and SF6 were updated for a die casting and a permanent mold facility for their
2020 reported emissions data resulting in resulting in decreased 2020 CO2 and SF6 emissions. Another die casting
facility that was a late reporters to the GHGRP have had emissions back casted to 2001, increasing SF6 emissions in
those years (Kramer 2000). CO2 emissions from one facility which was previously interpolated for 2014 has
emissions data available on the FLIGHT tool and has been updated accordingly, resulting in a decrease in 2014 CO2
emissions.
One facility's Fluorinated Ketone and CO2 emissions from 2016 were updated as an interpolation between
reported 2015 and 2017 emissions, in alignment with previous updates to that facility's SF6 emissions, leading to
increased CO2 emissions and decreased fluorinated ketone emissions. HFC-134a emissions from one facility which
were not previously accounted in the estimate summary have been accounted for, leading to an increase in 2019
HFC-134a emissions. CO2 emissions from one facility were previously held constant from their 2018 emissions,
further research indicated that holding emissions from their 2017 emissions was more reflective of current
conditions and was updated, resulting in increased 2019 and 2020 CO2 emissions from that facility.
In addition, for the current Inventory, C02-equivalent estimates of total emissions of SF6, HFC-134a, CO2, and
Fluorinated Ketone have been revised to reflect the 100-year global warming potentials (GWPs) provided in the
IPCC Fifth Assessment Report (AR5) (IPCC 2013). AR5 GWP values differ slightly from those presented in the IPCC
Fourth Assessment Report (AR4) (IPCC 2007) (used in the previous inventories). The AR5 GWPs have been applied
across the entire time series for consistency. The GWP value for SF6 increased from 22,800 to 23,500 leading to an
increase in calculated C02-equivalent emissions. The GWP value for HFC-13a decreased from 1,430 to 1,300
leading to a decrease in calculated C02-equivalent emissions. Compared to the previous Inventory which applied
100-year GWP values from AR4, the average annual change in SF6 emissions was a 3.1 percent increase and the
average annual change in HFC-134a emissions was 4.5 percent decrease for the time series. While the GWP value
CO2 remained the same, calculations of CO2 emissions from Permanent Mold, Wrought, and Anode Emissions tied
to emissions of SF6 led to 0.02 percent increase in CO2 emissions. Overall, emissions from magnesium production
and processing increased over the time series. The net impact from these updates was an average annual 2.8
percent increase in emissions for the time series. Further discussion on this update and the overall impacts of
updating the Inventory GWP values to reflect the IPCC Fifth Assessment Report can be found in Chapter 9,
Recalculations and Improvements.
66 GHGRP Report Verification Factsheet. See https://www.epa.gov/sites/production/files/2Q15-
07/documents/ghgrp verification factsheet.pdf.
Industrial Processes and Product Use 4-105
-------
1 Planned Improvements
2 Cover gas research conducted over the last decade has found that SF6 used for magnesium melt protection can
3 have degradation rates on the order of 20 percent in die casting applications (Bartos et al. 2007). Current emission
4 estimates assume (per the 2006IPCC Guidelines) that all SF6 utilized is emitted to the atmosphere. Additional
5 research may lead to a revision of the 2006 IPCC Guidelines to reflect this phenomenon and until such time,
6 developments in this sector will be monitored for possible application to the Inventory methodology.
7 Additional emissions are generated as byproducts from the use of alternate cover gases, which are not currently
8 accounted for. Research on this topic is developing, and as reliable emission factors become available, these
9 emissions will be incorporated into the Inventory.
10 An additional die casting facility that was a late reporter to the GHGRP will have emissions back cast based on
11 further outreach to determine what years they started die casting. This value will be taken out of the non-reported
12 emissions from die casters for the years affected.
13 4.21 Lead Production (CRF Source Category
14 2C5)
15 In 2021, lead was produced in the United States using only secondary production processes. Until 2014, lead
16 production in the United States involved both primary and secondary processes—both of which emit carbon
17 dioxide (CO2) (Sjardin 2003). Emissions from fuels consumed for energy purposes during the production of lead are
18 accounted for in the Energy chapter.
19 Primary production of lead through the direct smelting of lead concentrate produces CO2 emissions as the lead
20 concentrates are reduced in a furnace using metallurgical coke (Sjardin 2003). Primary lead production, in the form
21 of direct smelting, previously occurred at a single smelter in Missouri. This primary lead smelter was closed at the
22 end of 2013, and a small amount of residual lead was processed during demolition of the facility in 2014 (USGS
23 2015). Beginning in 2015, primary lead production no longer occurred in the United States.
24 Similar to primary lead production, CO2 emissions from secondary lead production result when a reducing agent,
25 usually metallurgical coke, is added to the smelter to aid in the reduction process. Carbon dioxide emissions from
26 secondary production also occur through the treatment of secondary raw materials (Sjardin 2003). Secondary
27 production primarily involves the recycling of lead acid batteries and post-consumer scrap at secondary smelters.
28 Secondary lead production in the United States has fluctuated over the past 20 years, reaching a high of 1,180,000
29 metric tons in 2007, and declined for three successive years between 2019 and 2021. In 2021, secondary lead
30 production accounted for 100 percent of total U.S. lead production. The lead-acid battery industry accounted for
31 about 92 percent of the reported U.S. lead consumption in 2021 (USGS 2022b).
32 In 2021, secondary lead production in the United States decreased by approximately 4 percent compared to 2020,
33 due to the closure of a secondary lead smelter in South Carolina (Battery Industry 2021) and reduced production
34 from several other secondary lead smelters (USGS 2022b). Secondary lead production in 2021 is 7 percent higher
35 than in 1990 (USGS 1994 and 2022b). The United States has become more reliant on imported refined lead, owing
36 to the closure of the last primary lead smelter in 2013. Exports of spent starting-lighting-ignition (SLI) batteries
37 decreased between 2014 and 2017, and subsequently recovered beginning in 2018. Exports were 14 percent
38 higher in the first 9 months of 2021 compared to the same time period in 2014 (USGS 2015 through 2022b). In the
39 first 9 months of 2021, 25.5 million spent SLI lead-acid batteries were exported, 29 percent more than that in the
40 same time period in 2020 (USGS 2022b).
41 Emissions of CO2 from lead production in 2021 were 0.4 MMT CO2 Eq. (446 kt), which is a 4 percent decrease
42 compared to 2020 and a 14 percent decrease compared to 1990 (see Table 4-90).
4-106 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
The United States was the third largest mine producer of lead in the world, behind China and Australia, and
accounted for approximately 7 percent of world production in 2021 (USGS 2022b).
Table 4-90: CO2 Emissions from Lead Production (MMT CO2 Eq. and kt)
Year
1990
2005
2017
2018
2019
2020
2021
MMTCOz Eq.
0.5
0.6
0.5
0.5
0.5
0.5
0.4
kt
516
553
513
527
531
464
446
Methodology and Time-Series Consistency
The methods used to estimate emissions for lead production67 are based on Sjardin's work (Sjardin 2003) for lead
production emissions and Tier 1 methods from the 2006IPCC Guidelines. The Tier 1 equation is as follows:
Equation 4-15: 2006IPCC Guidelines Tier 1: CO2 Emissions From Lead Production (Equation
4.32)
C02 Emissions = (DS x EFDS) + (5 x EFS)
where,
DS = Lead produced by direct smelting, metric ton
S = Lead produced from secondary materials
EFds = Emission factor for direct smelting, metric tons CCh/metric ton lead product
EFs = Emission factor for secondary materials, metric tons CCh/metric ton lead product
For primary lead production using direct smelting, Sjardin (2003) and the 2006 IPCC Guidelines provide an emission
factor of 0.25 metric tons CCh/metric ton lead. For secondary lead production, Sjardin (2003) and the 2006 IPCC
Guidelines provide an emission factor of 0.25 metric tons CCh/metric ton lead for direct smelting, as well as an
emission factor of 0.2 metric tons CCh/metric ton lead produced for the treatment of secondary raw materials (i.e.,
pretreatment of lead acid batteries). Since the secondary production of lead involves both the use of the direct
smelting process and the treatment of secondary raw materials, Sjardin recommends an additive emission factor
to be used in conjunction with the secondary lead production quantity. The direct smelting factor (0.25) and the
sum of the direct smelting and pretreatment emission factors (0.45) are multiplied by total U.S. primary and
secondary lead production, respectively, to estimate CO2 emissions.
The production and use of coking coal for lead production is adjusted for within the Energy chapter as this fuel was
consumed during non-energy related activities. Additional information on the adjustments made within the Energy
sector for Non-Energy Use of Fuels is described in both the Methodology section of CO2 from Fossil Fuel
Combustion (Section 3.1 Fossil Fuel Combustion (CRF Source Category 1A)) and Annex 2.1, Methodology for
Estimating Emissions of CO2 from Fossil Fuel Combustion.
The 1990 through 2021 activity data for primary and secondary lead production (see Table 4-91) were obtained
from the U.S. Geological Survey (USGS 1995 through 2022b).
Table 4-91: Lead Production (Metric Tons)
Year 1990 2005 2017 2018 2019 2020 2021
Primary 404,000 143,000 0 0 0 0 0
Secondary 922,000 1,150,000 1,140,000 1,170,000 1,180,000 1,030,000 990,000
67 EPA has not integrated aggregated facility-level Greenhouse Gas Reporting Program (GHGRP) information to inform these
estimates. The aggregated information (e.g., activity data and emissions) associated with Lead Production did not meet criteria
to shield underlying confidential business information (CBI) from public disclosure.
Industrial Processes and Product Use 4-107
-------
1 Methodological approaches discussed below were applied to applicable years to ensure time-series consistency in
2 emissions from 1990 through 2021.
3 Uncertainty - TO BE UPDATED FOR FINAL INVENTORY REPORT
4 Uncertainty associated with lead production relates to the emission factors and activity data used. The direct
5 smelting emission factor used in primary production is taken from Sjardin (2003) who averaged the values
6 provided by three other studies (Dutrizac et al. 2000; Morris et al. 1983; Ullman 1997). For secondary production,
7 Sjardin (2003) added a CO2 emission factor associated with battery treatment. The applicability of these emission
8 factors to plants in the United States is uncertain. Consistent with the ranges in Table 4.23 of the 2006IPCC
9 Guidelines for a Tier 1 emission factor by process type, EPA assigned an uncertainty range of ±20 percent for these
10 emission factors.
11 There is also a smaller level of uncertainty associated with the accuracy of primary and secondary production data
12 provided by the USGS which is collected via voluntary surveys; the uncertainty of the activity data is a function of
13 the reliability of reported plant-level production data and the completeness of the survey response. EPA currently
14 uses an uncertainty range of ±10% for these activity data elements, consistent with the ranges in Table 4.23 of the
15 2006 IPCC Guidelines for Tier 1 national production data.
16 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-92. Lead production CO2
17 emissions in 2021 were estimated to be between 0.4 and 0.6 MMT CO2 Eq. at the 95 percent confidence level. This
18 indicates a range of approximately 15 percent below and 16 percent above the emission estimate of 0.5 MMT CO2
19 Eq.
20 Table 4-92: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Lead
21 Production (MMT CO2 Eq. and Percent)
Source Gas
2021 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.)
(MMT C02 Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Lead Production C02
0.4
0.4
0.6
-15%
+16%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
22 QA/QC and Verification
23 For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
24 Chapter 6 of the 2006 IPCC Guidelines, see the QA/QC and Verification Procedures section in the introduction of
25 the IPPU chapter.
26 Initial review of activity data show that EPA's GHGRP Subpart R lead production data and resulting emissions are
27 fairly consistent with those reported by USGS. EPA is still reviewing available GHGRP data, reviewing QC analysis to
28 understand differences in data reporting (i.e., threshold implications), and assessing the possibility of including this
29 planned improvement in future Inventory reports (see Planned Improvements section below). Currently, GHGRP
30 data are used for QA purposes only.
31 Recalculations Discussion
32 Recalculations were implemented for 2014, 2018, 2019, and 2020, based on revised USGS data for secondary lead
33 production. Compared to the previous Inventory, emissions increased by 4 percent (18 kt CO2) for 2014, 3 percent
34 (14 kt CO2) for 2018, and less than 1 percent (4 kt CO2) for 2019. Emissions decreased by 6 percent (31 kt CO2) for
35 2020.
4-108 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Planned Improvements
2 Pending resources and prioritization of improvements for more significant sources, EPA will continue to evaluate
3 and analyze data reported under EPA's GHGRP that would be useful to improve the emission estimates and
4 category-specific QC for the Lead Production source category, in particular considering completeness of reported
5 lead production given the reporting threshold. Particular attention will be made to ensuring time-series
6 consistency of the emissions estimates presented in future Inventory reports, consistent with IPCC and UNFCCC
7 guidelines. This is required as the facility-level reporting data from EPA's GHGRP, with the program's initial
8 requirements for reporting of emissions in calendar year 2010, are not available for all inventory years (i.e., 1990
9 through 2009) as required for this Inventory. In implementing improvements and integration of data from EPA's
10 GHGRP, the latest guidance from the IPCC on the use of facility-level data in national inventories will be relied
11 upon.68
12 4.22 Zinc Production (CRF Source Category
13 2C6)
14 Zinc production in the United States consists of both primary and secondary processes. Of the primary and
15 secondary processes currently used in the United States, only the electrothermic and Waelz kiln secondary
16 processes result in non-energy carbon dioxide (CO2) emissions (Viklund-White 2000). Emissions from fuels
17 consumed for energy purposes during the production of zinc are accounted for in the Energy chapter.
18 The majority of zinc produced in the United States is used for galvanizing. Galvanizing is a process where zinc
19 coating is applied to steel in order to prevent corrosion. Zinc is used extensively for galvanizing operations in the
20 automotive and construction industry. Zinc is also used in the production of zinc alloys and brass and bronze alloys
21 (e.g., brass mills, copper foundries, and copper ingot manufacturing). Zinc compounds and dust are also used, to a
22 lesser extent, by the agriculture, chemicals, paint, and rubber industries.
23 Production of zinc can be conducted with a range of pyrometallurgical (e.g., electrothermic furnace, Waelz kiln,
24 flame reactor, batch retorts, Pinto process, and PIZO process) and hydrometallurgical (e.g., hydrometallurgical
25 recovery, solvent recovery, solvent extraction-electrowinning, and electrolytic) processes. Hydrometallurgical
26 production processes are assumed to be non-emissive since no carbon is used in these processes (Sjardin 2003).
27 Primary production in the United States is conducted through the electrolytic process, while secondary techniques
28 include the electrothermic and Waelz kiln processes, as well as a range of other processes. Worldwide primary zinc
29 production also employs a pyrometallurgical process using an Imperial Smelting Furnace; however, this process is
30 not used in the United States (Sjardin 2003).
31 In the electrothermic process, roasted zinc concentrate and secondary zinc products enter a sinter feed where
32 they are burned to remove impurities before entering an electric retort furnace. Metallurgical coke is added to the
33 electric retort furnace as a carbon-containing reductant. This concentration step, using metallurgical coke and high
34 temperatures, reduces the zinc oxides and produces vaporized zinc, which is then captured in a vacuum
35 condenser. This reduction process also generates non-energy CO2 emissions.
36 ZnO + C -»Zn(gas) + C02 (Reaction 1)
37 ZnO + CO -» Zn(gas) + C02 (Reaction 2)
38 In the Waelz kiln process, electric arc furnace (EAF) dust, which is captured during the recycling of galvanized steel,
39 enters a kiln along with a reducing agent (typically carbon-containing metallurgical coke). When kiln temperatures
40 reach approximately 1,100 to 1,200 degrees Celsius, zinc fumes are produced, which are combusted with air
68 See http://www.ipcc-nggiD.iges.or.jp/public/tb/TFl Technical Bulletin l.pdf.
Industrial Processes and Product Use 4-109
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
entering the kiln. This combustion forms zinc oxide, which is collected in a baghouse or electrostatic precipitator,
and is then leached to remove chloride and fluoride. The use of carbon-containing metallurgical coke in a high-
temperature fuming process results in non-energy CO2 emissions. Through this process, approximately 0.33 metric
tons of zinc is produced for every metric ton of EAF dust treated (Viklund-White 2000).
In the flame reactor process, a waste feed stream, which can include EAF dust, is processed in a high-temperature
environment (greater than 2,000 °C) created by the combustion of natural gas or coal and oxygen-enriched air.
Volatile metals, including zinc, are forced into the gas phase and drawn into a combustion chamber, where air is
introduced and oxidation occurs. The metal oxide product is then collected in a dust collection system (EPA 1992).
In 2021, the only companies in the United States that used emissive technology to produce secondary zinc
products were Befesa Holding US Inc (Befesa) and Steel Dust Recycling (SDR). The secondary zinc facilities operated
by Befesa were acquired from American Zinc Recycling (AZR) (formerly "Horsehead Corporation") in 2021. PIZO
Operating Company, LLC (PIZO) operated a secondary zinc production facility that processed EAF dust in
Blytheville, AR from 2009 to 2012.
For Befesa, EAF dust is recycled in Waelz kilns at their Calumet, IL; Palmerton, PA; Rockwood, TN; and Barnwell, SC
facilities. The former AZR facility in Beaumont, TX processed EAF dust via flame reactor from 1993 through 2009
(AZR 2021, Horsehead 2014). These Waelz kiln and flame reactor facilities produce intermediate zinc products
(crude zinc oxide or calcine). Prior to 2014, most of output from these facilities were transported to their Monaca,
PA facility where the products were smelted into refined zinc using electrothermic technology. In April 2014, the
Monaca smelter was permanently closed and replaced by a new facility in Mooresboro, NC in 2014.
The Mooresboro facility uses a hydrometallurgical process (i.e., solvent extraction with electrowinning technology)
to produce zinc products, which is assumed to be non-emissive as described above. Production at the Mooresboro
facility was idled in April 2016 and re-started in March 2020, with plans to be at full capacity by 2021 (Recycling
Today 2020). Direct consumption of coal, coke, and natural gas were replaced with electricity consumption
(Horsehead 2012b). The Mooresboro facility uses leaching and solvent extraction (SX) technology combined with
electrowinning, melting, and casting technology. In this process, Waelz Oxide (WOX) is first washed in water to
remove soluble elements such as chlorine, potassium, and sodium, and then is leached in a sulfuric acid solution to
dissolve the contained zinc creating a pregnant liquor solution (PLS). The PLS is then processed in a solvent
extraction step in which zinc is selectively extracted from the PLS using an organic solvent creating a purified zinc-
loaded electrolyte solution. The loaded electrolyte solution is then fed into the electrowinning process in which
electrical energy is applied across a series of anodes and cathodes submerged in the electrolyte solution causing
the zinc to deposit on the surfaces of the cathodes. As the zinc metal builds upon these surfaces, the cathodes are
periodically harvested in order to strip the zinc from their surfaces (Horsehead 2015).
SDR recycles EAF dust into intermediate zinc products using Waelz kilns and sells the intermediate products to
companies who smelt it into refined products.
Emissions of CO2 from zinc production in 2021 were estimated to be 1.0 MMT CO2 Eq. (969 kt CO2) (see Table
4-93). All 2021CO2 emissions resulted from secondary zinc production processes. Emissions from zinc production
in the United States have increased overall since 1990 due to a gradual shift from non-emissive primary production
to emissive secondary production. In 2021, emissions were estimated to be 53 percent higher than they were in
1990. Emissions decreased 1 percent from 2020 levels.
In 2021, global zinc mine production, or primary production, recovered from the reduced output experienced in
2020 due largely to the COVID-19 pandemic. U.S. primary zinc production mirrored this global trend. While total
refined zinc production increased in 2020 due to the reopening of an idled secondary zinc refinery, consumption of
refined zinc decreased in association with a decline in the U.S. steel industry as a result of the COVID-19 pandemic.
Refined zinc production increased in 2021, along with zinc consumption (USGS 2022).
Table 4-93: CO2 Emissions from Zinc Production (MMT CO2 Eq. and kt)
Year
1990
2005
2017
2018
2019
2020
2021
MMTCOz Eq.
0.6
1.0
0.9
1.0
1.0
1.0
1.0
kt
632
1,030
900
999
1,026
977
969
4-110 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 In 2021, United States primary and secondary refined zinc production were estimated to total 220,000 metric tons
2 (USGS 2022) (see Table 4-94). Domestic zinc mine production increased in 2021 compared to 2020 owing partially
3 to a decrease in production at the Red Dog Mine in Alaska and the closure of the Pend Oreille Mine in Washington
4 State in July 2019. Primary zinc production (primary slab zinc) in 2018 is used as an estimate for 2019 through 2021
5 due to the lack of available data. Secondary zinc production in 2020 increased by 250 percent compared to 2019
6 and was largely influenced by the reopening of the idled AZR secondary zinc refinery in Mooresboro, NC in March
7 2020 (USGS 2021; AZP 2021). From 2020 to 2021, secondary zinc production increased by 51 percent. Secondary
8 zinc production from the reopened facility was estimated by subtracting estimated primary zinc production from
9 the total zinc production value obtained from the USGS Minerals Yearbook: Zinc. Production of secondary zinc
10 reached its lowest point in the time series in 2019, following the closure of the Monaca, PA smelter in 2014 and
11 technical and environmental issues with the Mooresboro, NC facility which reopened in 2020, as noted above.
12 Table 4-94: Zinc Production (Metric Tons)
Year
1990
2005
2017
2018
2019
2020
2021
Primary
Secondary
262,704
95,708
191,120
156,000
117,000
15,000
101,000
15,000
101,000
14,000
101,000
79,000
101,000
119,000
Total
358,412
347,120
132,000
116,000
115,000
180,000
220,000
13 Methodology and Time-Series Consistency
14 The methods used to estimate non-energy CO2 emissions from zinc production69 using the electrothermic primary
15 production and Waelz kiln secondary production processes are based on Tier 1 methods from the 2006IPCC
16 Guidelines (IPCC 2006). The Tier 1 equation used to estimate emissions from zinc production is as follows:
17 Equation 4-16: 2006IPCCGuide/inesTier 1: CO2 Emissions From Zinc Production (Equation
18 4.33)
^002 ~ Zn x EFdefault
20 where,
21 Eco2 = CO2 emissions from zinc production, metric tons
22 Zn = Quantity of zinc produced, metric tons
23 EFdefauit = Default emission factor, metric tons CCh/metric ton zinc produced
24 The Tier 1 emission factors provided by IPCC for Waelz kiln-based secondary production were derived from
25 metallurgical coke consumption factors and other data presented in Vikland-White (2000). These coke
26 consumption factors as well as other inputs used to develop the Waelz kiln emission factors are shown below. IPCC
27 does not provide an emission factor for electrothermic processes due to limited information; therefore, the Waelz
28 kiln-specific emission factors were also applied to zinc produced from electrothermic processes. Starting in 2014,
29 refined zinc produced in the United States used hydrometallurgical processes and is assumed to be non-emissive.
30 For Waelz kiln-based production, IPCC recommends the use of emission factors based on EAF dust consumption, if
31 possible, rather than the amount of zinc produced since the amount of reduction materials used is more directly
32 dependent on the amount of EAF dust consumed. Since only a portion of emissive zinc production facilities
33 consume EAF dust, the emission factor based on zinc production is applied to the non-EAF dust consuming
34 facilities, while the emission factor based on EAF dust consumption is applied to EAF dust consuming facilities.
69 EPA has not integrated aggregated facility-level Greenhouse Gas Reporting Program (GHGRP) information to inform these
estimates. The aggregated information (e.g., activity data and emissions) associated with Zinc Production did not meet criteria
to shield underlying confidential business information (CBI) from public disclosure.
Industrial Processes and Product Use 4-111
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
The Waelz kiln emission factor based on the amount of zinc produced was developed based on the amount of
metallurgical coke consumed for non-energy purposes per ton of zinc produced (i.e., 1.19 metric tons coke/metric
ton zinc produced) (Viklund-White 2000), and the following equation:
Equation 4-17: Waelz Kiln CO2 Emission Factor for Zinc Produced
1.19 metric tons coke 0.85 metric tons C 3.67 metric tons C02 3.70 metric tons C02
EFwaelz Kiln : '¦ : '¦ * : '¦ : * * '
metric tons zinc metric tons coke metric tons C metric tons zinc
Refined zinc production levels for AZR's Monaca, PA facility (utilizing electrothermic technology) were available
from the company for years 2005 through 2013 (Horsehead 2008, 2011, 2012, 2013, and 2014). The Monaca
facility was permanently shut down in April 2014 and replaced by AZR's new facility in Mooresboro, NC. The new
facility uses hydrometallurgical process to produce refined zinc products. Hydrometallurgical production processes
are assumed to be non-emissive since no carbon is used in these processes (Sjardin 2003).
Metallurgical coke consumption for non-EAF dust consuming facilities for 1990 through 2004 were extrapolated
using the percentage change in annual refined zinc production at secondary smelters in the United States, as
provided by the U.S. Geological Survey (USGS) Minerals Yearbook: Zinc (USGS 1995 through 2006). Metallurgical
coke consumption for 2005 through 2013 were based on the secondary zinc production values obtained from the
Horsehead Corporation Annual Report Form 10-K: 2005 through 2008 from the 2008 10-K (Horsehead Corp 2009);
2009 and 2010 from the 2010 10-K (Horsehead Corp. 2011); and 2011 through 2013 from the associated 10-K
(Horsehead Corp. 2012a, 2013, 2014). Metallurgical coke consumption levels for 2014 and later were zero due to
the closure of the AZR (formerly "Horsehead Corporation") electrothermic furnace facility in Monaca, PA. The
secondary zinc produced values for each year were then multiplied by the 3.70 metric tons CCh/metric ton zinc
produced emission factor to develop CO2 emission estimates for the AZR electrothermic furnace facility.
The Waelz kiln emission factor based on the amount of EAF dust consumed was developed based on the amount
of metallurgical coke consumed per ton of EAF dust consumed (i.e., 0.4 metric tons coke/metric ton EAF dust
consumed) (Viklund-White 2000), and the following equation:
Equation 4-18: Waelz Kiln CO2 Emission Factor for EAF Dust Consumed
OA metric tons coke 0.85 metric tons C 3.67 metric tons C02 1.24 metric tons C02
E T7 A T7 X X
ivci nr tons EAF Dust metric tons coke metric tons C metric tons EAF Dust
Metallurgical coke consumption for EAF dust consuming facilities for 1990 through 2021 were calculated based on
the values of EAF dust consumed. The values of EAF dust consumed for Befesa, SDR, and PIZO are explained below.
The total amount of EAF dust consumed by the Waelz kilns currently operated by Befesa was available from AZR
(formerly "Horsehead Corporation") in financial reports for years 2006 through 2015 (Horsehead 2007, 2008,
2010a, 2011, 2012a, 2013, 2014, 2015, and 2016), from correspondence with AZR for 2016through 2019 (AZR
2020), and from correspondence with Befesa for 2020 and 2021 (Befesa 2022). The EAF dust consumption values
for each year were then multiplied by the 1.24 metric tons CCh/metric ton EAF dust consumed emission factor to
develop CO2 emission estimates for Befesa's Waelz kiln facilities.
The amount of EAF dust consumed by SDR and their total production capacity were obtained from SDR's facility in
Alabama for the years 2011 through 2021 (SDR 2012, 2014, 2015, 2017, 2018, 2021, 2022). The SDR facility has
been operational since 2008, underwent expansion in 2011 to include a second unit (operational since early- to
mid-2012), and expanded its capacity again in 2017 (SDR 2018). Annual consumption data for SDR was not publicly
available for the years 2008, 2009, and 2010. These data were estimated using data for AZR's Waelz kilns for 2008
through 2010 (Horsehead 2007, 2008, 2010a, 2010b, 2011). Annual capacity utilization ratios were calculated using
AZR's annual consumption and total capacity for the years 2008 through 2010. AZR's annual capacity utilization
ratios were multiplied with SDR's total capacity to estimate SDR's consumption for each of the years, 2008 through
2010 (SDR 2013). The 1.24 metric tons CCh/metric ton EAF dust consumed emission factor was then applied to
SDR's estimated EAF dust consumption to develop CO2 emission estimates for those Waelz kiln facilities.
4-112 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
PIZO's facility in Arkansas was operational from 2009 to 2012 (PIZO 2021). The amount of EAF dust consumed by
PIZO's facility for 2009 through 2012 was not publicly available. EAF dust consumption for PIZO's facility for 2009
and 2010 were estimated by calculating annual capacity utilization of AZR's Waelz kilns and multiplying this
utilization ratio by PIZO's total capacity (PIZO 2012). EAF dust consumption for PIZO's facility for 2011 through
2012 were estimated by applying the average annual capacity utilization rates for AZR and SDR (Grupo PROMAX)
to PIZO's annual capacity (Horsehead 2012; SDR 2012; PIZO 2012). The 1.24 metric tons C02/metric ton EAF dust
consumed emission factor was then applied to PIZO's estimated EAF dust consumption to develop CO2 emission
estimates for those Waelz kiln facilities.
The production and use of coking coal for zinc production is adjusted for within the Energy chapter as this fuel was
consumed during non-energy related activities. Additional information on the adjustments made within the Energy
sector for Non-Energy Use of Fuels is described in both the Methodology section of CO2 from Fossil Fuel
Combustion (3.1 Fossil Fuel Combustion (CRF Source Category 1A)) and Annex 2.1, Methodology for Estimating
Emissions of CO2 from Fossil Fuel Combustion.
Beginning with the 2017 USGS Minerals Commodity Summary: Zinc, United States primary and secondary refined
zinc production were reported as one value, total refined zinc production. Prior to this publication, primary and
secondary refined zinc production statistics were reported separately. For years 2016 through 2021, only one
facility produced primary zinc. Primary zinc produced from this facility was subtracted from the USGS 2016 to 2020
total zinc production statistic to estimate secondary zinc production for these years.
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2021.
Uncertainty-TO BE UPDATED FOR FINAL INVENTORY REPORT
The uncertainty associated with these estimates is two-fold, relating to activity data and emission factors used.
First, there is uncertainty associated with the amount of EAF dust consumed in the United States to produce
secondary zinc using emission-intensive Waelz kilns. The estimate for the total amount of EAF dust consumed in
Waelz kilns is based on combining the totals for (1) the EAF dust consumption value obtained for the kilns currently
operated by Befesa (and formerly operated by AZR or Horsehead Corporation) and (2) an EAF dust consumption
value obtained from the Waelz kiln facility operated by SDR. For the 1990 through 2015 estimates, EAF dust
consumption values for the kilns currently operated by Befesa were obtained from annual financial reports to the
Securities and Exchange Commission (SEC) by AZR. In 2016, AZR reorganized as a private company and ceased
providing annual reports to the SEC (Recycling Today 2017). EAF dust consumption values for subsequent years
from the Befesa kilns and SDR have been obtained from personal communication with facility representatives.
Since actual EAF dust consumption information is not available for PIZO's facility (2009 through 2010) and SDR's
facility (2008 through 2010), the amount is estimated by multiplying the EAF dust recycling capacity of the facility
(available from the company's website) by the capacity utilization factor for AZR (which was available from
Horsehead Corporation financial reports).The EAF dust consumption for PIZO's facility for 2011 through 2012 was
estimated by multiplying the average capacity utilization factor developed from AZR and SDR's annual capacity
utilization rates by PIZO's EAF dust recycling capacity. Therefore, there is uncertainty associated with the
assumption used to estimate PIZO's annual EAF dust consumption values for 2009 through 2012 and SDR's annual
EAF dust consumption values for 2008 through 2010. EPA uses an uncertainty range of ±5 percent for these EAF
dust consumption data inputs, based upon expert elicitation from the USGS commodity specialist.
Second, there is uncertainty associated with the emission factors used to estimate CO2 emissions from secondary
zinc production processes. The Waelz kiln emission factors are based on materials balances for metallurgical coke
and EAF dust consumed as provided by Viklund-White (2000). Therefore, the accuracy of these emission factors
depend upon the accuracy of these materials balances. Data limitations prevented the development of emission
factors for the electrothermic process. Therefore, emission factors for the Waelz kiln process were applied to both
electrothermic and Waelz kiln production processes. Consistent with the ranges in Table 4.25 of the 2006IPCC
Guidelines, EPA assigned an uncertainty range of ±20 percent for the Tier 1 Waelz kiln emission factors, which are
provided by Viklund-White in the form of metric tons of coke per metric ton of EAF dust consumed and metric tons
Industrial Processes and Product Use 4-113
-------
1 of coke per metric ton of zinc produced. In order to convert coke consumption rates to CO2 emission rates, values
2 for the heat and carbon content of coke were obtained from Table 4.2 - Tier 2 of the 2006IPCC Guidelines, An
3 uncertainty range of ±10 percent was assigned to these coke data elements based upon Table 4.25, Tier 2 -
4 National Reducing Agent & Process Materials Data.
5 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-95. Zinc production CO2
6 emissions from 2021 were estimated to be between 0.8 and 1.2 MMT CO2 Eq. at the 95 percent confidence level.
7 This indicates a range of approximately 19 percent below and 20 percent above the emission estimate of 1.0 MMT
8 CO2 Eq.
9 Table 4-95: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Zinc
10 Production (MMT CO2 Eq. and Percent)
Source Gas 2021 Emission Estimate Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (MMT C02 Eq.) (%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Zinc Production C02 1.0
0.8
1.2
-19%
+20%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
11 QA/QC and Verification
12 For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
13 Chapter 6 of the 2006 IPCC Guidelines, see the QA/QC and Verification Procedures section in the introduction of
14 the IPPU chapter.
15 Recalculations Discussion
16 Recalculations were performed for the year 2020 based on updated EAF dust consumption data. Compared to the
17 previous Inventory, emissions from zinc production decreased by 3 percent (31 kt CO2).
is Planned Improvements
19 Pending resources and prioritization of improvements for more significant sources, EPA will continue to evaluate
20 and analyze data reported under EPA's GHGRP that would be useful to improve the emission estimates and
21 category-specific QC for the Zinc Production source category, in particular considering completeness of reported
22 zinc production given the reporting threshold. Given the small number of facilities in the United States, particular
23 attention will be made to risks for disclosing CBI and ensuring time-series consistency of the emissions estimates
24 presented in future Inventory reports, consistent with IPCC and UNFCCC guidelines. This is required as the facility-
25 level reporting data from EPA's GHGRP, with the program's initial requirements for reporting of emissions in
26 calendar year 2010, are not available for all inventory years (i.e., 1990 through 2009) as required for this Inventory.
27 In implementing improvements and integration of data from EPA's GHGRP, the latest guidance from the IPCC on
28 the use of facility-level data in national inventories will be relied upon.70 This is a long-term planned improvement,
29 and EPA is still assessing the possibility of including this improvement in future Inventory reports.
30
70 See http://www.ipcc-nggip.iges.or.jp/public/tb/TFI Technical Bulletin l.pdf.
4-114 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 4.23 Electronics Industry (CRF Source
2 Category 2E)
3 The electronics industry uses multiple greenhouse gases in its manufacturing processes. In semiconductor
4 manufacturing, these include long-lived fluorinated greenhouse gases used for plasma etching and chamber
5 cleaning (CRF Source Category 2E1), fluorinated heat transfer fluids used for temperature control and other
6 applications (CRF Source Category 2E4), and nitrous oxide (N2O) used to produce thin films through chemical vapor
7 deposition and in other applications (reported under CRF Source Category 2H3). Similar to semiconductor
8 manufacturing, the manufacturing of micro-electro-mechanical systems (MEMS) devices (reported under CRF
9 Source Category 2E5 Other) and photovoltaic (PV) cells (CRF Source Category 2E3) requires the use of multiple
10 long-lived fluorinated greenhouse gases for various processes.
11 The gases most commonly employed in the electronics industry are trifluoromethane (hydrofluorocarbon (HFC)-23
12 or CHF3), perfluoromethane (CF4), perfluoroethane (C2F6), nitrogen trifluoride (NF3), and sulfur hexafluoride (SFs),
13 although other fluorinated compounds such as perfluoropropane (C3F8) and perfluorocyclobutane (c-C4Fs) are also
14 used. The exact combination of compounds is specific to the process employed.
15 In addition to emission estimates for these seven commonly used fluorinated gases, this Inventory contains
16 emissions estimates for N2O and other HFCs and unsaturated, low-GWP PFCs including CsFs, C4F6, HFC-32, HFC-41,
17 and HFC-134a. These additional HFCs and PFCs are emitted from etching and chamber cleaning processes in much
18 smaller amounts, accounting for 0.02 percent of emissions (in CO2 Eq.) from these processes.
19 For semiconductors, a single 300 mm silicon wafer that yields between 400 to 600 semiconductor products
20 (devices or chips) may require more than 100 distinct fluorinated-gas-using process steps, principally to deposit
21 and pattern dielectric films. Plasma etching (or patterning) of dielectric films, such as silicon dioxide and silicon
22 nitride, is performed to provide pathways for conducting material to connect individual circuit components in each
23 device. The patterning process uses plasma-generated fluorine atoms, which chemically react with exposed
24 dielectric film to selectively remove the desired portions of the film. The material removed as well as undissociated
25 fluorinated gases flow into waste streams and, unless emission abatement systems are employed, into the
26 atmosphere. Plasma enhanced chemical vapor deposition (PECVD) chambers, used for depositing dielectric films,
27 are cleaned periodically using fluorinated and other gases. During the cleaning cycle the gas is converted to
28 fluorine atoms in plasma, which etches away residual material from chamber walls, electrodes, and chamber
29 hardware. Undissociated fluorinated gases and other products pass from the chamber to waste streams and,
30 unless abatement systems are employed, into the atmosphere.
31 In addition to emissions of unreacted gases, some fluorinated compounds can also be transformed in the plasma
32 processes into different fluorinated compounds which are then exhausted, unless abated, into the atmosphere.
33 For example, when C2F6 is used in cleaning or etching, CF4 is typically generated and emitted as a process
34 byproduct. In some cases, emissions of the byproduct gas can rival or even exceed emissions of the input gas, as is
35 the case for NF3 used in remote plasma chamber cleaning, which often generates CF4 as a byproduct.
36 Besides dielectric film etching and PECVD chamber cleaning, much smaller quantities of fluorinated gases are used
37 to etch polysilicon films and refractory metal films like tungsten.
38 Nitrous oxide is used in manufacturing semiconductor devices to produce thin films by CVD and nitridation
39 processes as well as for N-doping of compound semiconductors and reaction chamber conditioning (Doering
40 2000).
41 Liquid perfluorinated compounds are also used as heat transfer fluids (F-HTFs) for temperature control, device
42 testing, cleaning substrate surfaces and other parts, and soldering in certain types of semiconductor
43 manufacturing production processes. Leakage and evaporation of these fluids during use is a source of fluorinated
44 gas emissions (EPA 2006). Unweighted F-HTF emissions consist primarily of perfluorinated amines,
45 hydrofluoroethers, perfluoropolyethers (specifically, PFPMIEs), and perfluoroalkylmorpholines. Three percent or
Industrial Processes and Product Use 4-115
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
less consist of HFCs, PFCs, and SF6 (where PFCs are defined as compounds including only carbon and fluorine). With
the exceptions of the hydrofluoroethers and most of the HFCs, all of these compounds are very long-lived in the
atmosphere and have global warming potentials (GWPs) near 10,000.71
MEMS and photovoltaic cell manufacturing require thin film deposition and etching of material with a thickness of
one micron or more, so the process is less intricate and complex than semiconductor manufacturing. The
manufacturing process is different than semiconductors, but generally employs similar techniques. Like
semiconductors, MEMS and photovoltaic cell manufacturers use fluorinated compounds for etching, cleaning
reactor chambers, and temperature control. CF4, SF6, and the Bosch process (which consists of alternating steps of
SFs and C4F8) are used to manufacture MEMS (EPA 2010). Photovoltaic cell manufacturing predominately uses CF4,
to etch crystalline silicon wafers, and C2F6 or NF3 during chamber cleaning after deposition of SiNx films (IPCC
2006), although other F-GHGs may be used. Similar to semiconductor manufacturing, both MEMS and photovoltaic
cell manufacturing use N2O in depositing films and other manufacturing processes. MEMS and photovoltaic
manufacturing may also employ HTFs for cooling process equipment (EPA 2010).
Emissions from all fluorinated greenhouse gases (including F-HTFs) and N2O for semiconductors, MEMS and
photovoltaic cells manducating are presented in Table 4-96 below for the years 1990, 2005, and the period 2017 to
2021. The rapid growth of the electronics industry and the increasing complexity (growing number of layers and
functions)72 of electronic products led to an increase in emissions of 152 percent between 1990 and 1999, when
emissions peaked at 8.4 MMT CO2 Eq. Emissions began to decline after 1999, reaching a low point in 2009 before
rebounding to 2006 emission levels and more or less plateauing at the current level, which represents a 43 percent
decline from 1999 to 2021. Together, industrial growth, adoption of emissions reduction technologies (including
but not limited to abatement technologies) and shifts in gas usages resulted in a net increase in emissions of
approximately 45 percent between 1990 and 2021. Total emissions from semiconductor manufacture in 2021
were higher than 2020 emissions, increasing by 10 percent, largely due to a large increase in SF6 and CF4 emissions.
The increases in SF6 are seen in facilities that manufacture 200 mm wafer size that do not have abatement systems
installed. Increases in CF4 can be attributed to facilities that manufacture 300 mm wafer sizes that do have
abatement systems installed.
For U.S. semiconductor manufacturing in 2021, total CC>2-equivalent emissions of all fluorinated greenhouse gases
and N2O from deposition, etching, and chamber cleaning processes were estimated to be 4.8 MMT CO2 Eq. This is a
decrease in emissions from 1999 of 43 percent, and an increase in emissions from 1990 of 45 percent. These
trends are driven by the above stated reasons.
Photovoltaic cell and MEMS manufacturing emissions of all fluorinated greenhouse gases are in Table 4-96. While
EPA has developed a simple methodology to estimate emissions from non-reporters and to back-cast emissions
from these sources for the entire time series, there is very high uncertainty associated with these emission
estimates.
The emissions reported by facilities manufacturing MEMS included emissions of C2F6, C3F8, C-C4F8, CF4, HFC-23, NF3,
N2O and SFs,73 and were equivalent to only 0.110 percent to 0.249 percent of the total reported emissions from
71 The GWP of PFPMIE, a perfluoropolyether used as an F-HTF, is included in the IPCC Fourth Assessment Report with a value of
10,300. The GWPs of the perfluorinated amines and perfluoroalkylmorpholines that are used as F-HTFs have not been
evaluated in the peer-reviewed literature. However, evaluations by the manufacturer indicate that their GWPs are near 10,000
(78 FR 20632), which is expected given that these compounds are both saturated and fully fluorinated. EPA assigns a default
GWP of 10,000 to compounds that are both saturated and fully fluorinated and that do not have chemical-specific GWPs in
either the Fourth or the Fifth Assessment Reports.
72 Complexity is a term denoting the circuit required to connect the active circuit elements (transistors) on a chip. Increasing
miniaturization, for the same chip size, leads to increasing transistor density, which, in turn, requires more complex
interconnections between those transistors. This increasing complexity is manifested by increasing the levels (i.e., layers) of
wiring, with each wiring layer requiring fluorinated gas usage for its manufacture.
73 Gases not reported by MEMS manufacturers to the GHGRP are currently listed as "NE" in the CRF. Since no facilities report
using these gases, emissions of these gases are not estimated for this sub-sector. However, there is insufficient data to
definitively conclude that they are not used by non-reporting facilities.
4-116 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
electronics manufacturing in 2011 to 2021. F-GHG emissions, the primary type of emissions for MEMS, ranged
from 0.0003 to 0.012 MMT CO2 Eq. from 1991 to 2021. Based upon information in the World Fab Forecast (WFF), it
appears that some GHGRP reporters that manufacture both semiconductors and MEMS are reporting their
emissions as only from semiconductor manufacturing (GHGRP reporters must choose a single classification per
fab). Emissions from non-reporters have not been estimated.
Total CCh-equivalent emissions from manufacturing of photovoltaic cells were estimated to range from 0.0003
MMT CO2 Eq. to 0.0330 MMT CO2 Eq. from 1998 to 2021 and were equivalent to between 0.003 percent to 0.60
percent of the total reported emissions from electronics manufacturing. F-GHG emissions, the primary type of
emissions for photovoltaic cells, ranged from 0.0003 to 0.032 MMT CO2 Eq. from 1998 to 2021. Emissions from
manufacturing of photovoltaic cells were estimated using an emission factor developed from reported data from a
single manufacturer between 2015 and 2016. This emission factor was then applied to production capacity
estimates from non-reporting facilities. Reported emissions from photovoltaic cell manufacturing consisted of CF4,
C2F6, c-C4F8, CHFs, NFb, and N2O.74
Emissions of F-HTFs, grouped by HFCs, PFCs or SF6 are presented in Table 4-96. Emissions of F-HTFs that are not
HFCs, PFCs or SF6 are not included in inventory totals and are included for informational purposes only.
Since reporting of F-HTF emissions began under EPA's GHGRP in 2011, total F-HTF emissions (reported and
estimated non-reported) have fluctuated between 0.6 MMT CO2 Eq. and 0.9 MMT CO2 Eq., with an overall
declining trend between 2011 to 2021. An analysis of the data reported to EPA's GHGRP indicates that F-HTF
emissions account for anywhere between 11 percent and 17 percent of total annual emissions (F-GHG, N2O and F-
HTFs) from semiconductor manufacturing.75 Table 4-98 shows F-HTF emissions in tons by compound group based
on reporting to EPA's GHGRP during years 2014 through 2020.76
Table 4-96: PFC, HFC, SFe, NF3, and N2O Emissions from Electronics Industry (MMT CO2 Eq.)
Year
1990
2005
2017
2018
2019
2020
2021
cf4
0.8
1.0
1.5
1.6
1.5
1.5
1.6
c2f6
1.8
1.8
1.1
1.0
0.9
0.8
0.8
CsFs
+
0.1
0.1
0.1
0.1
0.1
0.1
C4Fs
0.0
0.1
0.1
0.1
0.1
0.1
0.1
HFC-23
0.2
0.2
0.3
0.3
0.3
0.3
0.4
sf6
0.5
0.8
0.7
0.8
0.8
0.8
0.9
nf3
+
0.4
0.5
0.5
0.5
0.6
0.6
C4F6
+
+
+
+
+
+
+
CsFs
+
+
+
+
+
+
+
ch2f2
+
+
+
+
+
+
+
ch3f
+
+
+
+
+
+
+
CH2FCF3
+
+
+
+
+
+
+
Total Semiconductors
3.3
4.3
4.3
4.4
4.1
4.1
4.5
cf4
0.0
+
+
+
+
+
+
c2f6
0.0
+
+
+
+
+
+
CsFs
0.0
+
0.0
0.0
0.0
0.0
0.0
C4Fs
0.0
+
+
+
+
+
+
74 Gases not reported by PV manufacturers to the GHGRP are currently listed as "NE" in the CRF. Since no facilities report using
these gases, emissions of these gases are not estimated for this sub-sector. However, there is insufficient data to definitively
conclude that they are not used by non-reporting facilities.
75 Emissions data for HTFs (in tons of gas) from the semiconductor industry from 2011 through 2020 were obtained from the
EPA GHGRP annual facility emissions reports.
76 Many fluorinated heat transfer fluids consist of perfluoropolymethylisopropyl ethers (PFPMIEs) of different molecular
weights and boiling points that are distilled from a mixture. "BP 200 °C" (and similar terms below) indicate the boiling point of
the fluid in degrees Celsius. For more information, see https://www.regulations.gov/document?D=EPA-HQ-QAR-2009-0927-
0276.
Industrial Processes and Product Use 4-117
-------
HFC-23
0.0
+
+
+
+
+
+
sf6
0.0
+
+
+
+
+
+
nf3
0.0
0.0
+
+
+
+
+
Total MEMS
0.0
+
+
+
+
+
+
cf4
0.0
+
+
+
+
+
+
c2f6
0.0
+
+
+
+
+
+
C4Fs
0.0
+
+
+
+
+
+
HFC-23
0.0
+
+
+
+
+
+
sf6
0.0
0.0
0.0
0.0
0.0
0.0
0.0
nf3
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Total PV
0.0
+
+
+
+
+
+
N20 (Semiconductors)
+
0.1
0.2
0.2
0.2
0.3
0.3
N20 (MEMS)
0.0
+
+
+
+
+
+
N20 (PV)
0.0
+
+
+
+
+
+
Total N20
+
0.1
0.2
0.2
0.2
0.3
0.3
HFC, PFC and SF6 F-HTFs
0.0
+
+
+
+
+
+
Total Electronics Industry
3.3
4.5
4.6
4.7
4.3
4.4
4.8
+ Does not exceed 0.05 MMT C02 Eq.
1 Table 4-97: PFC, HFC, SFe, NF3, and N2O Emissions from Semiconductor Manufacture (Metric
2 Tons)
Year
1990
2005
2017
2018
2019
2020
2021
cf4
114.8
145.3
219.8
234.7
219.1
224.8
236.7
c2f6
160.0
163.4
97.7
92.9
79.1
70.4
75.8
CsFs
0.4
7.3
11.7
12.1
10.1
9.0
10.6
C4Fs
0.0
10.9
5.8
6.0
5.7
5.7
6.3
HFC-23
14.6
14.1
25.7
26.5
25.5
26.6
30.3
sf6
21.7
33.4
30.1
33.2
32.3
31.9
38.4
nf3
2.8
26.2
32.8
34.0
33.1
36.0
39.5
c4f6
0.7
0.9
0.9
0.8
0.9
0.8
1.0
C5Fs
0.5
0.6
0.8
0.5
1.2
0.4
0.4
ch2f2
0.6
0.8
1.1
0.9
1.0
1.1
1.0
ch3f
1.4
1.8
2.3
2.4
2.5
2.8
2.9
ch2fcf3
+
+
+
+
+
+
+
n2o
135.9
463.3
912.9
853.8
781.6
993.9
1,062.1
+ Does not exceed 0.05 MT.
3 Table 4-98: F-HTF Emissions from Electronics Manufacture by Compound Group (kt CO2 Eq.)
Year
1990
2005
2017
2018
2019
2020
2021
HFCs
0.0
1.0
3.6
2.7
1.1
0.9
1.1
PFCs
0.0
3.8
9.1
10.0
8.4
7.8
5.4
sf6
0.0
5.4
16.7
13.2
6.0
12.8
9.0
HFEs
0.0
41.2
2.9
4.6
1.3
5.3
3.8
PFPMIEs
0.0
109.8
148.5
183.0
171.7
150.2
148.3
Perfluoalkylromorpholines
0.0
65.9
52.3
58.6
56.5
61.0
53.5
Perfluorotrialkylamines
0.0
208.6
384.1
410.7
363.6
380.4
359.8
Total F-HTFs
0.0
435.8
617.2
682.8
608.6
618.3
580.9
Note: Emissions of F-HTFs that are not HFCs, PFCs or SF6 are not included in inventory totals and are included for
informational purposes only. Emissions presented for informational purposes include HFEs, PFPMIEs,
perfluoroalkylmorpholines, and perfluorotrialkylamines.
4-118 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Methodology and Time-Series Consistency
Emissions are based on data reported through Subpart I, Electronics Manufacture, of EPA's GHGRP, semiconductor
manufacturing Partner-reported emissions data received through EPA's PFC77 Reduction/Climate Partnership,
EPA's PFC Emissions Vintage Model (PEVM)—a model that estimates industry emissions from etching and chamber
cleaning processes in the absence of emission control strategies (Burton and Beizaie 2001)78—and estimates of
industry activity (i.e., total manufactured layer area and manufacturing capacity). The availability and applicability
of reported emissions data from the EPA Partnership and EPA's GHGRP and activity data differ across the 1990
through 2021 time series. Consequently, fluorinated greenhouse gas (F-GHG) emissions from etching and chamber
cleaning processes for semiconductors were estimated using seven distinct methods, one each for the periods
1990 through 1994,1995 through 1999, 2000 through 2006, 2007 through 2010, 2011 and 2012, 2013 and 2014,
and 2015 through 2021. Nitrous oxide emissions were estimated using five distinct methods, one each for the
period 1990 through 1994,1995 through 2010, 2011 and 2012, 2013 and 2014, and 2015 through 2021. The
methodology discussion below for these time periods focuses on semiconductor emissions from etching, chamber
cleaning, and uses of N2O. Other emissions for MEMS, photovoltaic cells, and HTFs were estimated using the
approaches described immediately below.
MEMS
GHGRP-reported emissions (F-GHG and N2O) from the manufacturing of MEMS are available for the years 2011 to
2021. Emissions from manufacturing of MEMS for years prior to 2011 were calculated by linearly interpolating
emissions between 1990 (at zero MMT CO2 Eq.) and 2011, the first year where emissions from manufacturing of
MEMS was reported to the GHGRP. Based upon information in the World Fab Forecast (WFF), it appears that some
GHGRP reporters that manufacture both semiconductors and MEMS are reporting their emissions as only from
semiconductor manufacturing; however, emissions from MEMS manufacturing are likely being included in
semiconductor totals. Emissions were not estimated for non-reporters.
Photovoltaic Cells
GHGRP-reported emissions (F-GHG and N2O) from the manufacturing of photovoltaic cells are available for 2011,
2012, 2015, and 2016 from two manufacturers. EPA estimates the emissions from manufacturing of PVs from non-
reporting facilities by multiplying the estimated capacity of non-reporters by a calculated F-GHG emission factor
and N2O emission factor based on GHGRP reported emissions from the manufacturer (in MMT CO2 Eq. per
megawatt) that reported emissions in 2015 and 2016. This manufacture's emissions are expected to be more
representative of emissions from the sector, as their emissions were consistent with consuming only CF4for
etching processes and are a large-scale manufacturer, representing 28 percent of the U.S. production capacity in
2016. The second photovoltaic manufacturer only produced a small fraction of U.S. production (<4 percent). They
also reported the use of NF3 in remote plasma cleaning processes, which does not have an emission factor in Part
98 for PV manufacturing, requiring them to report emissions equal to consumption. The total F-GHG emissions
from non-reporters are then disaggregated into individual gases using the gas distribution from the 2015 to 2016
manufacturer. Manufacturing capacities in megawatts were drawn from DisplaySearch, a 2015 Congressional
Research Service Report on U.S. Solar Photovoltaic Manufacturing, and self-reported capacity by GHGRP reporters.
EPA estimated that during the 2015 to 2016 period, 28 percent of manufacturing capacity in the United States was
represented through reported GHGRP emissions. Capacities are estimated for the full time series by linearly scaling
the total U.S. capacity between zero in 1997 to the total capacity reported of crystalline silicon (c-Si) PV
manufacturing in 2000 in DisplaySearch and then linearly scaling between the total capacity of c-Si PV
77 In the context of the EPA Partnership and PEVM, PFC refers to perfluorocompounds, not perfluorocarbons.
78 A Partner refers to a participant in the U.S. EPA PFC Reduction/Climate Partnership for the Semiconductor Industry. Through
a Memorandum of Understanding (MoU) with the EPA, Partners voluntarily reported their PFC emissions to the EPA by way of a
third party, which aggregated the emissions through 2010.
Industrial Processes and Product Use 4-119
-------
1 manufacturing in DisplaySearch in 2009 to the total capacity of c-Si PV manufacturing reported in the
2 Congressional Research Service report in 2012. Capacities were held constant for non-reporters for 2012 to 2019.
3 In 2020, non-reporter capacity declined due to the closure of several PV manufacturing plants. This capacity was
4 held constant for 2021. Average emissions per MW from the GHGRP reporter in 2015 and 2016 were then applied
5 to the total capacity prior to 2015. Emissions for 2014 from the GHGRP reporter that reported in 2015 and 2016
6 were scaled to the number of months open in 2014. For 1998 through 2021, emissions per MW (capacity) from the
7 GHGRP reporter were applied to the non-reporters. For 2017 through 2021, there are no reported PV emissions.
8 Therefore, emissions were estimated using the EPA-derived emission factor and estimated manufacturing capacity
9 from non-reporters only.
10 HTFs
11 Facility emissions of F-HTFs from semiconductor manufacturing are reported to EPA under its GHGRP and are
12 available for the years 2011 through 2021. EPA estimates the emissions of F-HTFs from non-reporting
13 semiconductor facilities by calculating the ratio of GHGRP-reported fluorinated HTF emissions to GHGRP reported
14 F-GHG emissions from etching and chamber cleaning processes, and then multiplying this ratio by the F-GHG
15 emissions from etching and chamber cleaning processes estimated for non-reporting facilities. Fluorinated HTF use
16 in semiconductor manufacturing is assumed to have begun in the early 2000s and to have gradually displaced
17 other HTFs (e.g., de-ionized water and glycol) in semiconductor manufacturing (EPA 2006). For time-series
18 consistency, EPA interpolated the share of F-HTF emissions to F-GHG emissions between 2000 (at 0 percent) and
19 2011 (at 17 percent) and applied these shares to the unadjusted F-GHG emissions during those years to estimate
20 the fluorinated HTF emissions.
21 Semiconductors
22 1990 through 1994
23 From 1990 through 1994, Partnership data were unavailable, and emissions were modeled using PEVM (Burton
24 and Beizaie 2001).79 The 1990 to 1994 emissions are assumed to be uncontrolled, since reduction strategies such
25 as chemical substitution and abatement were yet to be developed.
26 PEVM is based on the recognition that fluorinated greenhouse gas emissions from semiconductor manufacturing
27 vary with: (1) the number of layers that comprise different kinds of semiconductor devices, including both silicon
28 wafer and metal interconnect layers, and (2) silicon consumption (i.e., the area of semiconductors produced) for
29 each kind of device. The product of these two quantities, Total Manufactured Layer Area (TMLA), constitutes the
30 activity data for semiconductor manufacturing. PEVM also incorporates an emission factor that expresses
31 emissions per unit of manufactured layer-area. Emissions are estimated by multiplying TMLA by this emission
32 factor.
33 PEVM incorporates information on the two attributes of semiconductor devices that affect the number of layers:
34 (1) linewidth technology (the smallest manufactured feature size),80 and (2) product type (discrete, memory or
79 Various versions of the PEVM exist to reflect changing industrial practices. From 1990 to 1994 emissions estimates are from
PEVM vl.0, completed in September 1998. The emission factor used to estimate 1990 to 1994 emissions is an average of the
1995 and 1996 emissions factors, which were derived from Partner reported data for those years.
80 By decreasing features of Integrated Circuit components, more components can be manufactured per device, which
increases its functionality. However, as those individual components shrink it requires more layers to interconnect them to
achieve the functionality. For example, a microprocessor manufactured with 65 nm feature sizes might contain as many as 1
billion transistors and require as many as 11 layers of component interconnects to achieve functionality, while a device
manufactured with 130 nm feature size might contain a few hundred million transistors and require 8 layers of component
interconnects (ITRS 2007).
4-120 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 logic).81 For each linewidth technology, a weighted average number of layers is estimated using VLSI product-
2 specific worldwide silicon demand data in conjunction with complexity factors (i.e., the number of layers per
3 Integrated Circuit (IC) specific to product type (Burton and Beizaie 2001; ITRS 2007). PEVM derives historical
4 consumption of silicon (i.e., square inches) by linewidth technology from published data on annual wafer starts
5 and average wafer size (VLSI Research, Inc. 2012).
6 The emission factor in PEVM is the average of four historical emission factors, each derived by dividing the total
7 annual emissions reported by the Partners for each of the four years between 1996 and 1999 by the total TMLA
8 estimated for the Partners in each of those years. Over this period, the emission factors varied relatively little (i.e.,
9 the relative standard deviation for the average was 5 percent). Since Partners are believed not to have applied
10 significant emission reduction measures before 2000, the resulting average emission factor reflects uncontrolled
11 emissions and hence may be use here to estimate 1990 through 1994 emissions. The emission factor is used to
12 estimate U.S. uncontrolled emissions using publicly available data on world (including U.S.) silicon consumption.
13 As it was assumed for this time period that there was no consequential adoption of fluorinated-gas-reducing
14 measures, a fixed distribution of fluorinated-gas use was assumed to apply to the entire U.S. industry to estimate
15 gas-specific emissions. This distribution was based upon the average fluorinated-gas purchases made by
16 semiconductor manufacturers during this period and the application of IPCC default emission factors for each gas
17 (Burton and Beizaie 2001).
18 PEVM only addressed the seven main F-GHGs (CF4, C2F6, C3F8, C-C4F8, HFC-23, SF6, and NF3) used in semiconductor
19 manufacturing. Through reporting under Subpart I of EPA's GHGRP, data on other F-GHGs (C4F6, CsFs, HFC-32, HFC-
20 41, HFC-134a) used in semiconductor manufacturing became available and EPA was therefore able to extrapolate
21 this data across the entire 1990 to 2021 timeseries. To estimate emissions for these "other F-GHGs", emissions
22 data from Subpart I between 2014 to 2016 were used to estimate the average share or percentage contribution of
23 these gases as compared to total F-GHG emissions. Subpart I emission factors were updated for 2014 by EPA as a
24 result of a larger set of emission factor data becoming available, so reported data from 2011 through 2013 was not
25 utilized for the average. To estimate non-reporter emissions from 2011-2021, the average emissions data from
26 Subpart I of 2011 to 2021 was used.
27 To estimate N2O emissions, it was assumed the proportion of N2O emissions estimated for 1995 (discussed below)
28 remained constant for the period of 1990 through 1994.
29 1995 through 1999
30 For 1995 through 1999, total U.S. emissions were extrapolated from the total annual emissions reported by the
31 Partners (1995 through 1999). Partner-reported emissions are considered more representative (e.g., in terms of
32 capacity utilization in a given year) than PEVM-estimated emissions and are used to generate total U.S. emissions
33 when applicable. The emissions reported by the Partners were divided by the ratio of the total capacity of the
34 plants operated by the Partners and the total capacity of all of the semiconductor plants in the United States; this
35 ratio represents the share of capacity attributable to the Partnership. This method assumes that Partners and non-
36 Partners have identical capacity utilizations and distributions of manufacturing technologies. Plant capacity data is
37 contained in the World Fab Forecast (WFF) database and its predecessors, which is updated quarterly. Gas-specific
38 emissions were estimated using the same method as for 1990 through 1994.
39 For this time period emissions of other F-GHGs (C4F6, CsFs, HFC-32, HFC-41, HFC-134a) were estimated using the
40 method described above for 1990 to 1994.
41 For this time period, the N2O emissions were estimated using an emission factor that was applied to the annual,
42 total U.S. TMLA manufactured. The emission factor was developed using a regression-through-the-origin (RTO)
81 Memory devices manufactured with the same feature sizes as microprocessors (a logic device) require approximately one-
half the number of interconnect layers, whereas discrete devices require only a silicon base layer and no interconnect layers
(ITRS 2007). Since discrete devices did not start using PFCs appreciably until 2004, they are only accounted for in the PEVM
emissions estimates from 2004 onwards.
Industrial Processes and Product Use 4-121
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
model: GHGRP reported N2O emissions were regressed against the corresponding TMLA of facilities that reported
no use of abatement systems. Details on EPA's GHGRP reported emissions and development of emission factor
using the RTO model are presented in the 2011 through 2012 section. The total U.S. TMLA for 1995 through 1999
was estimated using PEVM.
2000 through 2006
Emissions for the years 2000 through 2006—the period during which Partners began the consequential application
of fluorinated greenhouse gas-reduction measures—were estimated using a combination of Partner-reported
emissions and adjusted PEVM modeled emissions. The emissions reported by Partners for each year were
accepted as the quantity emitted from the share of the industry represented by those Partners. Remaining
emissions, those from non-Partners, were estimated using PEVM, with one change. To ensure time-series
consistency and to reflect the increasing use of remote clean technology (which increases the efficiency of the
production process while lowering emissions of fluorinated greenhouse gases), the average non-Partner emission
factor (PEVM emission factor) was assumed to begin declining gradually during this period. Specifically, the non-
Partner emission factor for each year was determined by linear interpolation, using the end points of 1999 (the
original PEVM emission factor) and 2011 (a new emission factor determined for the non-Partner population based
on GHGRP-reported data, described below).
The portion of the U.S. total emissions attributed to non-Partners is obtained by multiplying PEVM's total U.S.
emissions figure by the non-Partner share of U.S. total silicon capacity for each year as described above.82 Gas-
specific emissions from non-Partners were estimated using linear interpolation between the gas-specific emissions
distributions of 1999 (assumed to be the same as that of the total U.S. Industry in 1994) and 2011 (calculated from
a subset of non-Partners that reported through the GHGRP as a result of emitting more than 25,000 MT CO2 Eq.
per year). Annual updates to PEVM reflect published figures for actual silicon consumption from VLSI Research,
Inc., revisions and additions to the world population of semiconductor manufacturing plants, and changes in IC
fabrication practices within the semiconductor industry (see ITRS 2008 and Semiconductor Equipment and
Materials Industry 2011).83,84,85
For this time period emissions of other F-GHGs (C4F6, CsFs, HFC-32, HFC-41, HFC-134a) were estimated using the
method described above for 1990 to 1994.
82 This approach assumes that the distribution of linewidth technologies is the same between Partners and non-Partners. As
discussed in the description of the method used to estimate 2007 emissions, this is not always the case.
83 Special attention was given to the manufacturing capacity of plants that use wafers with 300 mm diameters because the
actual capacity of these plants is ramped up to design capacity, typically over a 2 to 3 year period. To prevent overstating
estimates of partner-capacity shares from plants using 300 mm wafers, design capacities contained in WFF were replaced with
estimates of actual installed capacities for 2004 published by Citigroup Smith Barney (2005). Without this correction, the
partner share of capacity would be overstated, by approximately 5 percent. For perspective, approximately 95 percent of all
new capacity additions in 2004 used 300 mm wafers, and by year-end those plants, on average, could operate at approximately
70 percent of the design capacity. For 2005, actual installed capacities were estimated using an entry in the World Fab Watch
database (April 2006 Edition) called "wafers/month, 8-inch equivalent," which denoted the actual installed capacity instead of
the fully-ramped capacity. For 2006, actual installed capacities of new fabs were estimated using an average monthly ramp rate
of 1100 wafer starts per month (wspm) derived from various sources such as semiconductor fabtech, industry analysts, and
articles in the trade press. The monthly ramp rate was applied from the first-quarter of silicon volume (FQSV) to determine the
average design capacity over the 2006 period.
84 In 2006, the industry trend in co-ownership of manufacturing facilities continued. Several manufacturers, who are Partners,
now operate fabs with other manufacturers, who in some cases are also Partners and in other cases are not Partners. Special
attention was given to this occurrence when estimating the Partner and non-Partner shares of U.S. manufacturing capacity.
85 Two versions of PEVM are used to model non-Partner emissions during this period. For the years 2000 to 2003 PEVM
v3.2.0506.0507 was used to estimate non-Partner emissions. During this time, discrete devices did not use PFCs during
manufacturing and therefore only memory and logic devices were modeled in the PEVM v3.2.0506.0507. From 2004 onwards,
discrete device fabrication started to use PFCs, hence PEVM V4.0.0701.0701, the first version of PEVM to account for PFC
emissions from discrete devices, was used to estimate non-Partner emissions for this time period.
4-122 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Nitrous oxide emissions were estimated using the same methodology as the 1995 through 1999 methodology.
2007 through 2010
For the years 2007 through 2010, emissions were also estimated using a combination of Partner reported
emissions and adjusted PEVM modeled emissions to provide estimates for non-Partners; however, two
improvements were made to the estimation method employed for the previous years in the time series. First, the
2007 through 2010 emission estimates account for the fact that Partners and non-Partners employ different
distributions of manufacturing technologies, with the Partners using manufacturing technologies with greater
transistor densities and therefore greater numbers of layers.86 Second, the scope of the 2007 through 2010
estimates was expanded relative to the estimates for the years 2000 through 2006 to include emissions from
research and development (R&D) fabs. This additional enhancement was feasible through the use of more detailed
data published in the WFF. PEVM databases were updated annually as described above. The published world
average capacity utilization for 2007 through 2010 was used for production fabs, while for R&D fabs a 20 percent
figure was assumed (SIA 2009).
In addition, publicly available utilization data was used to account for differences in fab utilization for
manufacturers of discrete and IC products for 2010 emissions for non-Partners. The Semiconductor Capacity
Utilization (SICAS) Reports from SIA provides the global semiconductor industry capacity and utilization,
differentiated by discrete and IC products (SIA 2009 through 2011). PEVM estimates were adjusted using
technology-weighted capacity shares that reflect the relative influence of different utilization. Gas-specific
emissions for non-Partners were estimated using the same method as for 2000 through 2006.
For this time period emissions of other F-GHGs (CsFs, CH2F2, CH3F, CH2FCF3, C2H2F4) were estimated using the
method described above for 1990 to 1994.
Nitrous oxide emissions were estimated using the same methodology as the 1995 through 1999 methodology.
2011 through 2012
The fifth method for estimating emissions from semiconductor manufacturing covers the period 2011 through
2012. This methodology differs from previous years because the EPA's Partnership with the semiconductor
industry ended (in 2010) and reporting under EPA's GHGRP began. Manufacturers whose estimated uncontrolled
emissions equal or exceed 25,000 MT CO2 Eq. per year (based on default F-GHG-specific emission factors and total
capacity in terms of substrate area) are required to report their emissions to EPA. This population of reporters to
EPA's GHGRP included both historical Partners of EPA's PFC Reduction/Climate Partnership as well as non-Partners
some of which use gallium arsenide (GaAs) technology in addition to Si technology.87 Emissions from the
population of manufacturers that were below the reporting threshold were also estimated for this time period
using EPA-developed emission factors and estimates of facility-specific production obtained from WFF. Inventory
totals reflect the emissions from both reporting and non-reporting populations.
Under EPA's GHGRP, semiconductor manufacturing facilities report emissions of F-GHGs (for all types of F-GHGs)
used in etch and clean processes as well as emissions of fluorinated heat transfer fluids. (Fluorinated heat transfer
fluids are used to control process temperatures, thermally test devices, and clean substrate surfaces, among other
applications.) They also report N2O emissions from CVD and other processes. The F-GHGs and N2O were
aggregated, by gas, across all semiconductor manufacturing GHGRP reporters to calculate gas-specific emissions
86 EPA considered applying this change to years before 2007 but found that it would be difficult due to the large amount of
data (i.e., technology-specific global and non-Partner TMLA) that would have to be examined and manipulated for each year.
This effort did not appear to be justified given the relatively small impact of the improvement on the total estimate for 2007
and the fact that the impact of the improvement would likely be lower for earlier years because the estimated share of
emissions accounted for by non-Partners is growing as Partners continue to implement emission-reduction efforts.
87 GaAs and Si technologies refer to the wafer on which devices are manufactured, which use the same PFCs but in different
ways.
Industrial Processes and Product Use 4-123
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
for the GHGRP-reporting segment of the U.S. industry. At this time, emissions that result from heat transfer fluid
use that are HFC, PFC and SF6 are included in the total emission estimates from semiconductor manufacturing, and
these GHGRP-reported emissions have been compiled and presented in Table 4-96. F-HTF emissions resulting from
other types of gases (e.g., HFEs) are not presented in semiconductor manufacturing totals in Table 4-96 and Table
4-97 but are shown in Table 4-98 for informational purposes.
Changes to the default emission factors and default destruction or removal efficiencies (DREs) used for GHGRP
reporting affected the emissions trend between 2013 and 2014. These changes did not reflect actual emission rate
changes but data improvements. Therefore, for the current Inventory, EPA adjusted the time series of GHGRP-
reported data for 2011 through 2013 to ensure time-series consistency using a series of calculations that took into
account the characteristics of a facility (e.g., wafer size and abatement use). To adjust emissions for facilities that
did not report abatement in 2011 through 2013, EPA simply applied the revised emission factors to each facility's
estimated gas consumption by gas, process type and wafer size. In 2014, EPA also started collecting information on
fab-wide DREs and the gases abated by process type, which were used in calculations for adjusting emissions from
facilities that abated F-GHGs in 2011 through 2013.
• To adjust emissions for facilities that abated emissions in 2011 through 2013, EPA first calculated the
quantity of gas abated in 2014 using reported F-GHG emissions, the revised default DREs (or the
estimated site-specific DRE,88 if a site-specific DRE was indicated), and the fab-wide DREs reported in
2014.89 To adjust emissions for facilities that abated emissions in 2011 through 2013, EPA first estimated
the percentage of gas passing through abatement systems for remote plasma clean in 2014 using the ratio
of emissions reported for CF4 and NF3.
• EPA then estimated the quantity of NF3 abated for remote plasma clean in 2014 using the ratio of
emissions reported for CF4 (which is not abated) and NF3. This abated quantity was then subtracted from
the total abated quantity calculated as described in the bullet above.
• To account for the resulting remaining abated quantity, EPA assumed that the percentage of gas passing
through abatement systems was the same across all remaining gas and process type combinations where
abatement was reported for 2014.
• The percentage of gas abated was then assumed to be the same in 2011 through 2013 (if the facility
claimed abatement that year) as in 2014 for each gas abated in 2014.
The revised emission factors and DREs were then applied to the estimated gas consumption for each facility by gas,
process type and wafer size.90
For the segment of the semiconductor industry that is below EPA's GHGRP reporting threshold, and for R&D
facilities, which are not covered by EPA's GHGRP, emission estimates are based on EPA-developed emission factors
for the F-GHGs and N2O and estimates of manufacturing activity. The new emission factors (in units of mass of CO2
Eq./TMLA [million square inches (MSI)]) are based on the emissions reported under EPA's GHGRP by facilities
without abatement and on the TMLA estimates for these facilities based on the WFF (SEMI 2012, 2013).91 In a
refinement of the method used to estimate emissions for the non-Partner population for prior years, different
88 EPA generally assumed site-specific DREs were as follows: CF4, Etch (90 percent); all other gases, Etch (98 percent); NF3,
Clean (95 percent); CF4, Clean (80 percent), and all other gases, Clean (80 percent). There were a few exceptions where a higher
DRE was assumed to ensure the calculations operated correctly when there was 100 percent abatement.
89 If abatement information was not available for 2014 or the reported incorrectly in 2014, data from 2015 or 2016 was
substituted.
90 Since facilities did not report by fab before 2014, fab-wide DREs were averaged if a facility had more than one fab. For
facilities that reported more than one wafer size per facility, the percentages of a facility's emissions per wafer size were
estimated in 2014 and applied to earlier years, if possible. If the percentage of emissions per wafer size were unknown, a 50/50
split was used.
91 EPA does not have information on fab-wide DREs for this time period, so it is not possible to estimate uncontrolled emissions
from fabs that reported point-of-use abatement. These fabs were therefore excluded from the regression analysis. (They are
still included in the national totals.)
4-124 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 emission factors were developed for different subpopulations of fabs, disaggregated by wafer size (200 mm and
2 300 mm). For each of these groups, a subpopulation-specific emission factor was obtained using a regression-
3 through-the-origin (RTO) model: facility-reported aggregate emissions of seven F-GHGs (CF4, C2F6, C3F8, C-C4F8,
4 CHF3, SFs and NF3)92 were regressed against the corresponding TMLA to estimate an aggregate F-GHG emissions
5 factor (CO2 Eq./MSI TMLA), and facility-reported N2O emissions were regressed against the corresponding TMLA to
6 estimate a N2O emissions factor (CO2 Eq./MSI TMLA). For each subpopulation, the slope of the RTO model is the
7 emission factor for that subpopulation. Information on the use of point-of-use abatement by non-reporting fabs
8 was not available; thus, EPA conservatively assumed that non-reporting facilities did not use point-of-use
9 abatement.
10 For 2011 and 2012, estimates of TMLA relied on the capacity utilization of the fabs published by the U.S. Census
11 Bureau's Historical Data Quarterly Survey of Plant Capacity Utilization (USCB 2011, 2012). Similar to the
12 assumption for 2007 through 2010, facilities with only R&D activities were assumed to utilize only 20 percent of
13 their manufacturing capacity. All other facilities in the United States are assumed to utilize the average percent of
14 the manufacturing capacity without distinguishing whether fabs produce discrete products or logic products.
15 Non-reporting fabs were then broken out into subpopulations by wafer size (200 mm and 300 mm), using
16 information available through the WFF. The appropriate emission factor was applied to the total TMLA of each
17 subpopulation of non-reporting facilities to estimate the CC>2-equivalent emissions of that subpopulation.
18 Gas-specific, CC>2-equivalent emissions for each subpopulation of non-reporting facilities were estimated using the
19 corresponding reported distribution of gas-specific, CC>2-equivalent emissions from which the aggregate emission
20 factors, based on GHGRP-reported data, were developed. Estimated in this manner, the non-reporting population
21 accounted for 4.9 and 5.0 percent of U.S. emissions in 2011 and 2012, respectively. The GHGRP-reported emissions
22 and the calculated non-reporting population emissions are summed to estimate the total emissions from
23 semiconductor manufacturing.
24 2013 and 2014
25 For 2013 and 2014, as for 2011 and 2012, F-GHG and N2O emissions data received through EPA's GHGRP were
26 aggregated, by gas, across all semiconductor-manufacturing GHGRP reporters to calculate gas-specific emissions
27 for the GHGRP-reporting segment of the U.S. industry. However, for these years WFF data was not available.
28 Therefore, an updated methodology that does not depend on the WFF derived activity data was used to estimate
29 emissions for the segment of the industry that are not covered by EPA's GHGRP. For the facilities that did not
30 report to the GHGRP (i.e., which are below EPA's GHGRP reporting threshold or are R&D facilities), emissions were
31 estimated based on the proportion of total U.S. emissions attributed to non-reporters for 2011 and 2012. EPA used
32 a simple averaging method by first estimating this proportion for both F-GHGs and N2O for 2011, 2012, and 2015
33 and 2016, resulting in one set of proportions for F-GHGs and one set for N2O, and then applied the average of each
34 set to the 2013 and 2014 GHGRP reported emissions to estimate the non-reporters' emissions. Fluorinated gas-
35 specific, CC>2-equivalent emissions for non-reporters were estimated using the corresponding reported distribution
36 of gas-specific, CC>2-equivalent emissions reported through EPA's GHGRP for 2013 and 2014.
37 GHGRP-reported emissions in 2013 were adjusted to capture changes to the default emission factors and default
38 destruction or removal efficiencies used for GHGRP reporting, affecting the emissions trend between 2013 and
39 2014. EPA used the same method to make these adjustments as described above for 2011 and 2012 GHGRP data.
40 2015 through 2021
41 Similar to the methods described above for 2011 and 2012, and 2013 and 2014, EPA relied upon emissions data
42 reported directly through the GHGRP. For 2015 through 2021, EPA took an approach similar to the one used for
43 2011 and 2012 to estimate emissions for the segment of the semiconductor industry that is below EPA's GHGRP
92 Only seven gases were aggregated because inclusion of F-GHGs that are not reported in the Inventory results in
overestimation of emission factor that is applied to the various non-reporting subpopulations.
Industrial Processes and Product Use 4-125
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
reporting threshold, and for R&D facilities, which are not covered by EPA's GHGRP. However, in a change from
previous years, EPA was able to develop new annual emission factors for 2015 through 2021 using TMLA from WFF
and a more comprehensive set of emissions, i.e., fabs with as well as without abatement control, as new
information about the use of abatement in GHGRP fabs and fab-wide were available. Fab-wide DREs represent
total fab CO2 Eq.-weighted controlled F-GHG and N2O emissions (emissions after the use of abatement) divided by
total fab CO2 Eq.-weighted uncontrolled F-GHG and N2O emissions (emission prior to the use of abatement).
Using information about reported emissions and the use of abatement and fab-wide DREs, EPA was able to
calculate uncontrolled emissions (each total F-GHG and N2O) for every GHGRP reporting fab. Using this, coupled
with TMLA estimated using methods described above (see 2011 through 2012), EPA derived emission factors by
year, gas type (F-GHG or N2O), and wafer size (200 mm and less or 300 mm) by dividing the total annual emissions
reported by GHGRP reporters by the total TMLA estimated for those reporters. These emission factors were
multiplied by estimates of non-reporter TMLA to arrive at estimates of total F-GHG and N2O emissions for non-
reporters for each year. For each wafer size, the total F-GHG emissions were disaggregated into individual gases
using the shares of total emissions represented by those gases in the emissions reported to the GHGRP by
unabated fabs producing that wafer size.
Data Sources
GHGRP reporters, which consist of former EPA Partners and non-Partners, estimated their emissions using a
default emission factor method established by EPA. Like the Tier 2c Method in the 2019 Refinement to the 2006
IPCC Guidelines, this method uses different emission and byproduct generation factors for different F-GHGs and
process types and uses factors for different wafer sizes (i.e., 300mm vs. 150 and 200mm) and CVD clean subtypes
(in situ thermal, in situ plasma, and remote plasma). Starting with 2014 reported emissions, EPA's GHGRP required
semiconductor manufacturers to apply updated emission factors to estimate their F-GHG emissions. For the years
2011 through 2013 reported emissions, semiconductor manufacturers used older emission factors to estimate
their F-GHG emissions (Federal Register / Vol. 75, No. 230 /December 1, 2010, 74829). Subpart I emission factors
were updated for 2014 by EPA as a result of a larger set of emission factor data becoming available as part of the
Subpart I petition process, which took place from 2011 through 2013. In addition to semiconductor manufacturing,
GHGRP also includes reported emissions from MEMS and PV producers.
Historically, semiconductor industry partners estimated and reported their emissions using a range of methods
and uneven documentation. It is assumed that most Partners used a method at least as accurate as the IPCC's Tier
2a Methodology, recommended in the 2006 IPCC Guidelines. Partners are estimated to have accounted for
between 56 and 79 percent of F-GHG emissions from U.S. semiconductor manufacturing between 1995 and 2010,
with the percentage declining in recent years as Partners increasingly implemented abatement measures.
Estimates of operating plant capacities and characteristics for Partners and non-Partners were derived from the
Semiconductor Equipment and Materials Industry (SEMI) WFF (formerly World Fab Watch) database (1996 through
2012, 2013, 2016, 2018, and 2021) (e.g., Semiconductor Materials and Equipment Industry 2021). Actual
worldwide capacity utilizations for 2008 through 2010 were obtained from Semiconductor International Capacity
Statistics (SICAS) (SIA 2009 through 2011). Estimates of the number of layers for each linewidth was obtained from
International Technology Roadmap for Semiconductors: 2013 Edition (Burton and Beizaie 2001; ITRS 2007; ITRS
2008; ITRS 2011; ITRS 2013). PEVM utilized the WFF, SICAS, and ITRS, as well as historical silicon consumption
estimates published by VLSI. Actual quarterly U.S. capacity utilizations for 2011, 2012, 2014 to 2021 were obtained
from the U.S. Census Bureau's Historical Data Quarterly Survey of Plant Capacity Utilization (USCB 2011, 2012,
2015, 2016, 2017, 2018, 2019, 2020, 2021).
Estimates of PV manufacturing capacity, which are used to calculate emissions from non-reporting facilities, are
based on data from two sources. A historical market analysis from DisplaySearch provided estimates of U.S.
manufacturing capacity from 2000-2009 (DisplaySearch 2010). Domestic PV cell production for 2012 was obtained
from a Congressional Research Service report titled U.S. Solar Photovoltaic Manufacturing: Industry Trends, Global
Competition, Federal Support (Platzer 2015).
4-126 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Uncertainty
2 A quantitative uncertainty analysis of this source category was performed using the IPCC-recommended Approach
3 2 uncertainty estimation methodology, the Monte Carlo Stochastic Simulation technique. The Monte Carlo
4 Stochastic Simulation was performed on the total emissions estimate from the Electronics Industry, represented in
5 equation form as:
6 Equation 4-19: Total Emissions from Electronics Industry
7 Total Emissions (Et) = Semiconductors F-GHG and N2O Emissions (Esemi)+ MEMS F-GHG and N2O Emissions
8 (Emems) + PV F-GHG and N2O Emissions (Epv) + HFC, PFC and SF6 F-HTFs Emissions (Ehtf)
9 The uncertainty in the total emissions for the Electronics Industry, presented in Table 4-99 below, results from the
10 convolution of four distributions of emissions, namely from semiconductors manufacturing, MEMS manufacturing,
11 PV Manufacturing and emissions of Heat Transfer Fluids. The approaches for estimating uncertainty in each of the
12 sources are described below:
13 Semiconductors Manufacture Emission Uncertainty
14 The Monte Carlo Stochastic Simulation was performed on the emissions estimate from semiconductor
15 manufacturing, represented in equation form as:
16 Equation 4-20: Total Emissions from Semiconductor Manufacturing
17 Semiconductors F-GHG and N2O Emissions (Esemi) = GHGRP Reported F-GHG Emissions (ER,F-GHG,semi) + Non-
18 Reporters' Estimated F-GHG Emissions (ENR,F-GHG,semi) + GHGRP Reported N2O Emissions (ER.Nzo.semi) + Non-
19 Reporters' Estimated N2O Emissions (ENR,N20,semi)
20 The uncertainty in Esemi results from the convolution of four distributions of emissions, Er,f-ghg,semi ER,N2o,semi Enr,f-
21 GHG,semi and ENR,N2o,semi. The approaches for estimating each distribution and combining them to arrive at the
22 reported 95 percent confidence interval (CI) for Esemi are described in the remainder of this section.
23 The uncertainty estimate of Er, f-ghg,semi, or GHGRP-reported F-GHG emissions, is developed based on gas-specific
24 uncertainty estimates of emissions for two industry segments, one processing 200 mm or less wafers and one
25 processing 300 mm wafers. Uncertainties in emissions for each gas and industry segment are based on an
26 uncertainty analysis conducted during the assessment of emission estimation methods for the Subpart I
27 rulemaking in 2012 (see Technical Support for Modifications to the Fluorinated Greenhouse Gas Emission
28 Estimation Method Option for Semiconductor Facilities under Subpart I, docket EPA-HQ-OAR-2011-0028).93 This
29 assessment relied on facility-specific gas information by gas and wafer size, and incorporated uncertainty
30 associated with both emission factors and gas consumption quantities. The 2012 analysis did not consider the use
31 of abatement.
32 For the industry segment that manufactured 200 mm wafers, estimates of uncertainty at a 95 percent CI ranged
33 from ±29 percent for C3F8 to ±10 percent for CF4. For the corresponding 300 mm industry segment, estimates of
34 uncertainty at the 95 percent CI ranged from ±36 percent for C4F8 to ±16 percent for CF4. For gases for which
93 On November 13, 2013, EPA published a final rule revising Subpart I (Electronics Manufacturing) of the GHGRP (78 FR
68162). The revised rule includes updated default emission factors and updated default destruction and removal efficiencies
that are slightly different from those that semiconductor manufacturers were required to use to report their 2012 emissions.
The uncertainty analyses that were performed during the development of the revised rule focused on these updated defaults
but are expected to be reasonably representative of the uncertainties associated with the older defaults, particularly for
estimates at the country level. (They may somewhat underestimate the uncertainties associated with the older defaults at the
facility level.) For simplicity, the 2012 estimates are assumed to be unbiased although in some cases, the updated (and
therefore more representative) defaults are higher or lower than the older defaults. Multiple models and sensitivity scenarios
were run for the Subpart I analysis. The uncertainty analysis presented here made use of the Input gas and wafer size model
(Model 1) under the following conditions: Year = 2010, f = 20, n = SIA3.
Industrial Processes and Product Use 4-127
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
uncertainty was not analyzed in the 2012 assessment (e.g., CH2F2), EPA applied the 95 percent CI range equivalent
to the range for the gas and industry segment with the highest uncertainty from the 2012 assessment. These gas
and wafer-specific uncertainty estimates were developed to represent uncertainty at a facility-level, but they are
applied to the total emissions across all the facilities that did not abate emissions as reported under EPA's GHGRP
at a national-level. Hence, it is noted that the uncertainty estimates used may be overestimating the uncertainties
at a national-level.
For those facilities reporting abatement of emissions under EPA's GHGRP, estimates of uncertainties for the no
abatement industry segments are modified to reflect the use of full abatement (abatement of all gases from all
cleaning and etching equipment) and partial abatement. These assumptions used to develop uncertainties for the
partial and full abatement facilities are identical for 200 mm and 300 mm wafer processing facilities. For all
facilities reporting gas abatement, a triangular distribution of destruction or removal efficiency is assumed for each
gas. The triangular distributions range from an asymmetric and highly uncertain distribution of zero percent
minimum to 90 percent maximum with 70 percent most likely value for CF4 to a symmetric and less uncertain
distribution of 85 percent minimum to 95 percent maximum with 90 percent most likely value for C4F8, NF3, and
SFs. For facilities reporting partial abatement, the distribution of fraction of the gas fed through the abatement
device, for each gas, is assumed to be triangularly distributed as well. It is assumed that no more than 50 percent
of the gases are abated (i.e., the maximum value) and that 50 percent is the most likely value, and the minimum is
zero percent. Consideration of abatement then resulted in four additional industry segments, two 200-mm wafer-
processing segments (one fully and one partially abating each gas) and two 300-mm wafer-processing segment
(one fully and the other partially abating each gas). Gas-specific emission uncertainties were estimated by
convolving the distributions of unabated emissions with the appropriate distribution of abatement efficiency for
fully and partially abated facilities using a Monte Carlo simulation.
The uncertainty in ER,F-GHG,semi is obtained by allocating the estimates of uncertainties to the total GHGRP-reported
emissions from each of the six industry segments, and then running a Monte Carlo simulation which results in the
95 percent CI for emissions from GHGRP-reporting facilities (ER,F-GHG,semi).
The uncertainty in ER,N2o,semi is obtained by assuming that the uncertainty in the emissions reported by each of the
GHGRP reporting facilities results from the uncertainty in quantity of N2O consumed and the N2O emission factor
(or utilization). Similar to analyses completed for Subpart I (see Technical Support for Modifications to the
Fluorinated Greenhouse Gas Emission Estimation Method Option for Semiconductor Facilities under Subpart I,
docket EPA-HQ-OAR-2011-0028), the uncertainty of N2O consumed was assumed to be 20 percent. Consumption
of N2O for GHGRP reporting facilities was estimated by back-calculating from emissions reported and assuming no
abatement. The quantity of N2O utilized (the complement of the emission factor) was assumed to have a triangular
distribution with a minimum value of zero percent, mode of 20 percent and maximum value of 84 percent. The
minimum was selected based on physical limitations, the mode was set equivalent to the Subpart I default N2O
utilization rate for chemical vapor deposition, and the maximum was set equal to the maximum utilization rate
found in ISMI Analysis of Nitrous Oxide Survey Data (ISMI 2009). The inputs were used to simulate emissions for
each of the GHGRP reporting, INhO-emitting facilities. The uncertainty for the total reported N2O emissions was
then estimated by combining the uncertainties of each facilities' reported emissions using Monte Carlo simulation.
The estimate of uncertainty in Enr, F-GHG.semi and Enr, N2o,semi entailed developing estimates of uncertainties for the
emissions factors and the corresponding estimates of TMLA.
The uncertainty in TMLA depends on the uncertainty of two variables—an estimate of the uncertainty in the
average annual capacity utilization for each level of production of fabs (e.g., full scale or R&D production) and a
corresponding estimate of the uncertainty in the number of layers manufactured. For both variables, the
distributions of capacity utilizations and number of manufactured layers are assumed triangular for all categories
of non-reporting fabs. The most probable utilization is assumed to be 82 percent, with the highest and lowest
utilization assumed to be 89 percent, and 70 percent, respectively. For the triangular distributions that govern the
number of possible layers manufactured, it is assumed the most probable value is one layer less than reported in
the ITRS; the smallest number varied by technology generation between one and two layers less than given in the
ITRS and largest number of layers corresponded to the figure given in the ITRS.
The uncertainty bounds for the average capacity utilization and the number of layers manufactured are used as
4-128 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 inputs in a separate Monte Carlo simulation to estimate the uncertainty around the TMLA of both individual
2 facilities as well as the total non-reporting TMLA of each sub-population.
3 The uncertainty around the emission factors for non-reporting facilities is dependent on the uncertainty of the
4 total emissions (MMT CO2 Eq. units) and the TMLA of each reporting facility in that category. For each wafer size
5 for reporting facilities, total emissions were regressed on TMLA (with an intercept forced to zero) for 10,000
6 emission and 10,000 TMLA values in a Monte Carlo simulation, which results in 10,000 total regression coefficients
7 (emission factors). The 2.5th and the 97.5th percentile of these emission factors are determined, and the bounds
8 are assigned as the percent difference from the estimated emission factor.
9 The next step in estimating the uncertainty in emissions of reporting and non-reporting facilities in semiconductor
10 manufacture is convolving the distribution of reported emissions, emission factors, and TMLA using Monte Carlo
11 simulation. For this Monte Carlo simulation, the distributions of the reported F-GHG gas- and wafer size-specific
12 emissions are assumed to be normally distributed, and the uncertainty bounds are assigned at 1.96 standard
13 deviations around the estimated mean. The were some instances, though, where departures from normality were
14 observed for variables, including for the distributions of the gas- and wafer size-specific N2O emissions, TMLA, and
15 non-reporter emission factors, both for F-GHGs and N2O. As a result, the distributions for these parameters were
16 assumed to follow a pert beta distribution.
17 MEMS Manufacture Emission Uncertainty
18 The Monte Carlo Stochastic Simulation was performed on the emissions estimate from MEMS manufacturing,
19 represented in equation form as:
20 Equation 4-21: Total Emissions from MEMS Manufacturing
21 MEMS F-GHG and N2O Emissions (Emems) = GHGRP Reported F-GHG Emissions (Er,f-ghg,mems) + GHGRP
22 Reported N2O Emissions (Er.nzo, mems)
23 Emissions from MEMS manufacturing are only quantified for GHGRP reporters. MEMS manufacturers that report
24 to the GHGRP all report the use of 200 mm wafers. Some MEMS manufacturers report using abatement
25 equipment. Therefore, the estimates of uncertainty at the 95 percent CI for each gas emitted by MEMS
26 manufacturers are set equal to the gas-specific uncertainties for manufacture of 200mm semiconductor wafers
27 with partial abatement. The same assumption is applied for uncertainty levels for GHGRP reported MEMS N2O
28 emissions (Er,n2o,mems).
29 PV Manufacture Emission Uncertainty
30 The Monte Carlo Stochastic Simulation was performed on the emissions estimate from PV manufacturing,
31 represented in equation form as:
32 Equation 4-22: Total Emissions from PV Manufacturing
33 PV F-GHG and N2O Emissions (Epv) = Non-Reporters' Estimated F-GHG Emissions (Enr,f-ghg,pv) + Non-
34 Reporters' Estimated N2O Emissions (Enr.nzo.pv)
35 Emissions from PV manufacturing are only estimated for non-GHGRP reporters. There were no reported emissions
36 from PV manufacturing in GHGRP in 2021. The "Non-Reporters' Estimated F-GHG Emissions" term in Equation 4-22
37 was estimated using an emission factor developed using emissions from reported data in 2015 and 2016 and total
38 non-reporters' capacity. Due to a lack of information and data and because they represent similar physical and
39 chemical processes, the uncertainty at the 95 percent CI level for non-reporter PV capacity is assumed to be the
40 same as the uncertainty in non-reporter TMLA for semiconductor manufacturing. Similarly, the uncertainty for the
41 PV manufacture emission factors are assumed to be the same as the uncertainties in emission factors used for
42 non-reporters in semiconductor manufacture.
43 Heat Transfer Fluids Emission Uncertainty
44 There is a lack of data related to the uncertainty of emission estimates of heat transfer fluids used for electronics
45 manufacture. Therefore, per the 2006IPCC Guidelines (IPCC 2006, Volume 3, Chapter 6), uncertainty bounds of 20
Industrial Processes and Product Use 4-129
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
percent were applied to estimate uncertainty associated with the various types of heat transfer fluids, including
PFCs, HFC, and SF6, at the national level.
The results of the Approach 2 quantitative uncertainty analysis for electronics manufacturing are summarized in
Table 4-99. These results were obtained by convolving—using Monte Carlo simulation—the distributions of
emissions for each reporting and non-reporting facility that manufactures semiconductors, MEMS, or PVs and use
heat transfer fluids. The emissions estimate for total U.S. F-GHG, N2O, and HTF emissions from electronics
manufacturing were estimated to be between 4.88 and 5.50 MMT CO2 Eq. at a 95 percent CI level. This range
represents 6 percent below to 6 percent above the 2021 emission estimate of 5.19 MMT CO2 Eq. for all emissions
from electronics manufacture. This range and the associated percentages apply to the estimate of total emissions
rather than those of individual gases. Uncertainties associated with individual gases will be somewhat higher than
the aggregate but were not explicitly modeled.
Table 4-99: Approach 2 Quantitative Uncertainty Estimates for HFC, PFC, SFe, NF3 and N2O
Emissions from Electronics Manufacture (MMT CO2 Eq. and Percent)
2021 Emission
Estimate
Uncertainty Range Relative to Emission Estimate3
Source
Gas
(MMT CO? Eq.)
(MMT CO? Eq.)
(%)
Lower
Upper
Lower Upper
Boundb
Boundb
Bound Bound
Electronics
Industry
HFC, PFC, SF6,
NFs, and N20
4.79
4.51
5.08
-6% 6%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
b Absolute lower and upper bounds were calculated using the corresponding lower and upper bounds in percentages.
QA/QC and Verification
For its GHGRP data, EPA verifies annual facility-level reports through a multi-step process (e.g., including a
combination of pre-and post-submittal electronic checks and manual reviews by staff) to identify potential errors
and ensure that data submitted to EPA are accurate, complete, and consistent (EPA 2015).94 Based on the results
of the verification process, EPA follows up with facilities to resolve mistakes that may have occurred. The post-
submittals checks are consistent with a number of general and category-specific QC procedures including range
checks, statistical checks, algorithm checks, and year-to-year checks of reported data and emissions.
For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
Chapter 6 of the 2006IPCC Guidelines, see the QA/QC and Verification Procedures section in the introduction of
the IPPU chapter and Annex 8 for more details.
Recalculations Discussion
Any resubmitted emissions data reported to EPA's GHGRP from all prior years were updated in this Inventory.
Additionally, EPA made the following changes:
• To estimate non-reporter F-GHG and N2O emissions, EPA relies on data reported through Subpart I and
the World Fab Forecast. This process requires EPA to map facilities that report through Subpart I and
which are also represented in the World Fab Forecast. For this Inventory update, EPA identified and
made corrections to a few instances of this mapping based on new information and additional reviews of
the data. This had minimal effects on emission estimates.
94 GHGRP Report Verification Factsheet. See https://www.epa.gov/sites/production/files/2Q15-
07/documents/ghgrp verification factsheet.pdf.
4-130 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
• EPA re-ran regression analyses for years 2010 to 2020 to reflect updates to Subpart I and the World Fab
Forecast. These changes had minor effects on the emission factors, standard error, and R2values for all
years. This resulted in the recalculation of non-reporter's F-GHG and N2O estimates for all years.
• To estimate emissions for "other F-GHGs" in the years prior to 2011, emissions data from Subpart I were
used to estimate the average share or percentage contribution of these gases as compared to total F-
GHG emissions. Previously, the emissions data between 2011-2020 was used to calculate this average.
However, the average in this Inventory was updated to only include 2014-2016. This change was made to
make a more realistic estimate of the distribution of other F-GHGs pre-2011. This will also hold the pre-
2011 other F-GHGs emissions constant in future inventories. Emissions data from 2011-2013 was not
used as the 2011-2013 data did not reflect the updated emissions factors in Subpart I.
• To estimate emissions of HFCs, PFCs, and SF6 from F-HTFs between 2001 and 2010, emissions data from
Subpart I were used to estimate the average share or percentage contribution of these gases as
compared to total F-HTFs emissions. Previously, this average was calculated using Subpart I data from
2011 to 2021. However, to estimate the distribution of these gases between 2001 and 2010 more
realistically, emissions data from 2011 to 2013 was averaged instead. This will hold the pre-2011
emissions constant in future inventories.
• Previously, F-GHG emissions from a PV manufacturer not-reporting through the GHGRP were held
constant from 2013 through the most recent Inventory year. EPA determined that this manufacturer
ceased operations in 2019, so their reported emissions were changed to zero for 2020 and beyond.
• To improve the uncertainty analysis for this source category other F-GHGs from semiconductor
manufacturing, HFC, PFC, and SF6 emissions from the use of heat transfer fluids and emissions resulting
from the manufacturing of PVs and MEMS were included in total uncertainty estimates.
Overall, the impact of these recalculations led to an average decrease of 0.004 MMT CO2 Eq. (0.083 percent) across
the time series (1990 through 2020).
For the current Inventory, estimates of CC>2-equivalent F-GHGs, N2O, and F-HTF emissions from the electronics
inventory have been revised to reflect the 100-year GWPs provided in the IPCC Fifth Assessment Report (AR5)
(IPCC 2013). AR5 GWP values differ slightly from those presented in the IPCC Fourth Assessment Report (AR4) (IPCC
2007) used in the previous inventories. The AR5 GWPs have been applied across the entire time series for
consistency. The GWPs of CF4, C2F6, and NF3, three of the most significant contributors to total emissions in this
source category, have decreased, leading to a decrease in calculated CC>2-equivalent emissions from those F-GHGs.
In contrast, the GWP of SF6, another significant contributor to total emissions in this source category, has
increased, leading to an increase in calculated CC>2-equivalent emissions for this F-GHG. Compared to the previous
Inventory which applied 100-year GWP values from AR4, the average annual change in CC>2-equivalent emissions
across the time series 1990-2020 for CF4, C2F6, NF3, and SF6 were 11 percent decrease, 9 percent decrease, 9
percent decrease, and 8 percent increase, respectively. The net impact from these updates and the additional
updates noted above was an average annual 7.5 percent decrease in CC>2-equivalent emissions for the time series.
Further discussion on this update and the overall impacts of updating the Inventory GWP values to reflect the IPCC
Fifth Assessment Report can be found in Chapter 9, Recalculations and Improvements.
Planned Improvements
The Inventory methodology uses data reported through the EPA Partnership (for earlier years) and EPA's GHGRP
(for later years) to extrapolate the emissions of the non-reporting population. While these techniques are well
developed, the understanding of the relationship between the reporting and non-reporting populations is limited.
Further analysis of the reporting and non-reporting populations could aid in the accuracy of the non-reporting
population extrapolation in future years. In addition, the accuracy of the emissions estimates for the non-reporting
population could be further increased through EPA's further investigation of and improvement upon the accuracy
of estimated activity in the form of TMLA.
Industrial Processes and Product Use 4-131
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
The Inventory uses utilization from two different sources for various time periods-SEMI to develop PEVM and to
estimate non-Partner emissions for the period 1995 to 2010 and U.S. Census Bureau for 2011 through 2021. SEMI
reported global capacity utilization for manufacturers through 2011. U.S. Census Bureau capacity utilization
include U.S. semiconductor manufacturers as well as assemblers. Further analysis on the impacts of using a new
and different source of utilization data could prove to be useful in better understanding of industry trends and
impacts of utilization data sources on historical emission estimates.
Estimates of semiconductor non-reporter and non-Partner emissions are based on EPA-developed emission factors
for the time periods pre-2010, 2011 through 2012, and 2015 through 2021. Based on the data available for these
time periods, the methods used to develop emission factors for non-reporters and non-Partners are slightly
inconsistent for semiconductors (e.g., how data representing emissions and TMLA from the manufacture of various
wafer sizes are aggregated or disaggregated for purposes of calculating emission factors). Further analyses to
support potentially adjusting the methods for developing these emission factors could be done to better ensure
consistency across the time series.
The methodology for estimating semiconductor emissions from non-reporters uses data from the International
Technology Roadmap for Semiconductors (ITRS) on the number of layers associated with various technology node
sizes. The ITRS has now been replaced by the International Roadmap for Devices and Systems (IRDS), which has
published updated data on the number of layers used in each device type and node size (in nanometers).
Incorporating this updated dataset will improve the accuracy of emissions estimates from non-reporting
semiconductor fabs.
4.24 Substitution of Ozone Depleting
Substances (CRF Source Category 2F)
Hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), and carbon dioxide (CO2) are used as alternatives to several
classes of ozone-depleting substances (ODSs) that are being phased out under the terms of the Montreal Protocol
and the Clean Air Act Amendments of 1990.95 Ozone-depleting substances—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 extinguishing, and aerosols. Although HFCs and PFCs are not harmful to the stratospheric ozone
layer, they are potent greenhouse gases. On December 27, 2020, the American Innovation and Manufacturing
(AIM) Act was enacted by Congress and directs EPA to address HFCs by phasing down production and consumption
(i.e., production plus import minus export), maximizing reclamation and minimizing releases from equipment, and
facilitating the transition to next-generation technologies through sector-based restrictions. Emission estimates for
HFCs, PFCs, and CO2 used as substitutes for ODSs are provided in Table 4-100 and Table 4-101.96
Table 4-100: Emissions of HFCs, PFCs, and CO2 from ODS Substitutes (MMT CO2 Eq.)
Gas
1990
2005
2017
2018
2019
2020
2021
HFC-23
0.0
+
+
+
+
+
+
HFC-32
0.0
0.3
5.3
6.1
6.8
7.7
9.4
HFC-125
+
8.2
45.4
48.6
52.9
57.5
65.9
HFC-134a
+
72.8
58.8
56.4
55.3
54.1
50.0
95 [42 U.S.C § 7671, CAA Title VI],
96 Emissions of ODSs are not included here consistent with UNFCCC reporting guidelines for national inventories noted in Box
4-1. See Annex 6.2 for more details on emissions of ODSs. Emissions from C02 used in the food and beverage industry are
separately reported in Chapter 4.15 Carbon Dioxide Consumption but does not include C02 in ODS substitute use sectors as a
refrigerant, foam blowing agent, or fire extinguishing agent.
4-132 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
HFC-143a
+
10.0
30.1
29.7
29.9
29.9
30.0
HFC-236fa
0.0
0.9
1.0
0.9
0.9
0.9
0.8
cf4
0.0
+
+
+
+
+
+
C02
+
+
+
+
+
+
+
Others3
0.3
7.1
15.5
16.0
16.1
15.9
16.3
Total
0.3
99.4
156.1
157.8
162.0
166.1
172.5
+ Does not exceed 0.05 MMT C02 Eq.
a Others represent an unspecified mix of HFCs and PFCs, which includes HFC-152a, HFC-227ea, HFC-245fa,
HFC-365mfc, HFC-43-10mee, HCFO-1233zd(E), HFO-1234yf, HFO-1234ze(E), HFO-1336mzz(Z), C4Fi0, and
PFC/PFPEs, the latter being a proxy for a diverse collection of PFCs and perfluoropolyethers (PFPEs)
employed for solvent applications. For estimating purposes, the GWP value used for PFC/PFPEs was based
upon n-C6Fi4.
Note: Totals may not sum due to independent rounding.
Table 4-101: Emissions of HFCs, PFCs, and CO2 from ODS Substitution (Metric Tons)
Gas
1990
2005
2017
2018
2019
2020
2021
HFC-23
0
1
2
2
2
2
2
HFC-32
0
397
7,832
8,937
10,077
11,374
13,846
HFC-125
+
2,580
14,308
15,335
16,682
18,153
20,803
HFC-134a
+
56,029
45,264
43,419
42,558
41,590
38,447
HFC-143a
+
2,093
6,264
6,188
6,230
6,234
6,240
HFC-236fa
0
118
124
118
112
108
104
cf4
0
2
6
7
7
7
8
C02
14
1,325
2,879
3,093
3,303
3,516
3,734
Others3
M
M
M
M
M
M
M
+ Does not exceed 0.5 MT.
M (Mixture of Gases).
a Others represent an unspecified mix of HFCs and PFCs, which includes HFC-152a, HFC-227ea, HFC-245fa, HFC-365mfc,
HFC-43-10mee, HCFO-1233zd(E), HFO-1234yf, HFO-1234ze(E), HFO-1336mzz(Z), C4Fi0, and PFC/PFPEs, the latter being a
proxy for a diverse collection of PFCs and perfluoropolyethers (PFPEs) employed for solvent applications.
In 1990 and 1991, the only significant emissions of HFCs and PFCs as substitutes to ODSs were relatively small
amounts of HFC-152a—used as an aerosol propellant and also a component of the refrigerant blend R-500 used in
chillers. Beginning in 1992, HFC-134a was used in growing amounts as a refrigerant in motor vehicle air-
conditioners and in refrigerant blends such as R-404A.97 In 1993, the use of HFCs in foam production began, and in
1994 ODS substitutes for halons entered widespread use in the United States as halon production was phased out.
In 1995, these compounds also found applications as solvents. Non-fluorinated ODS substitutes, such as CO2, have
been used in place of ODS in certain foam production and fire extinguishing uses since the 1990s.
The use and subsequent emissions of HFCs, PFCs, and CO2 as ODS substitutes has been increasing from small
amounts in 1990 to 172.5 MMT CO2 Eq. emitted in 2021. This increase was in large part the result of efforts to
phase out CFCs, HCFCs, and other ODSs in the United States. Use and emissions of HFCs are expected to start
decreasing in the next few years and continue downward as production and consumption of HFCs are phased
down to 15 percent of their baseline levels by 2036 through an allowance allocation and trading program
established by EPA. Improvements in recovery practices and the use of alternative gases and technologies, through
voluntary actions and in response to potential future regulations under the AIM Act, will also contribute to a
reduction in HFC use and emissions.
Table 4-102 presents emissions of HFCs, PFCs, and CO2 as ODS substitutes by end-use sector for 1990 through
2021. The refrigeration and air-conditioning sector is further broken down by sub-sector. The end-use sectors that
contributed the most toward emissions of HFCs, PFCs, and CO2 as ODS substitutes in 2021 include refrigeration and
97 R-4Q4A contains HFC-125, HFC-143a, and HFC-134a.
Industrial Processes and Product Use 4-133
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
air-conditioning (139.1 MMT CO2 Eq., or approximately 81 percent), aerosols (17.7 MMT CO2 Eq., or approximately
10 percent), and foams (10.8 MMT CO2 Eq., or approximately 6 percent). Within the refrigeration and air-
conditioning end-use sector residential unitary AC, part of the Residential Stationary Air-conditioning subsector
shown below, was the highest emitting end-use (38.5 MMT CO2 Eq.), followed by large retail food, which is part of
the Commercial Refrigeration subsector. Each of the end-use sectors is described in more detail below.
Table 4-102: Emissions of HFCs, PFCs, and CO2 from ODS Substitutes (MMT CO2 Eq.) by
Sector
Sector
1990
2005
2017
2018
2019
2020
2021
Refrigeration/Air Conditioning
+
83.0
120.2
122.4
126.2
130.3
139.1
Commercial Refrigeration
+
14.9
40.8
39.6
40.2
40.6
41.0
Domestic Refrigeration
+
0.2
1.2
1.2
1.2
1.2
1.1
Industrial Process
Refrigeration
+
1.8
12.6
13.8
15.0
16.2
17.4
Transport Refrigeration
+
1.6
6.4
6.9
7.4
7.9
8.4
Mobile Air Conditioning
+
61.5
30.7
28.7
26.6
24.6
22.9
Residential Stationary Air
Conditioning
+
1.2
22.8
26.0
29.1
32.9
41.1
Commercial Stationary Air
Conditioning
+
1.7
5.7
6.2
6.6
6.9
7.3
Aerosols
0.2
10.2
17.7
16.7
17.0
17.3
17.7
Foams
+
3.5
13.8
14.2
14.1
13.7
10.8
Solvents
+
1.6
1.9
2.0
2.0
2.0
2.1
Fire Protection
+
1.1
2.4
2.6
2.7
2.7
2.8
Total
0.3
99.4
156.1
157.8
162.0
166.1
172.5
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Refrigeration/Air Conditioning
The refrigeration and air-conditioning sector includes a wide variety of equipment types that have historically used
CFCs or HCFCs. End-uses within this sector include motor vehicle air-conditioning, retail food refrigeration,
refrigerated transport (e.g., ship holds, truck trailers, railway freight cars), household refrigeration, residential and
small commercial air-conditioning and heat pumps, chillers (large comfort cooling), cold storage facilities, and
industrial process refrigeration (e.g., systems used in food processing, chemical, petrochemical, pharmaceutical, oil
and gas, and metallurgical industries). As the ODS phaseout has taken effect, most equipment has been retrofitted
or replaced to use HFC-based substitutes. Common HFCs in use today in refrigeration/air-conditioning equipment
are HFC-134a, R-410A,98 R-404A, and R-507A." Lower-GWP options such as hydrofluoroolefin (HFO)-1234yf in
motor vehicle air-conditioning, R-717 (ammonia) in cold storage and industrial applications, and R-744 (carbon
dioxide) and HFC/HFO blends in retail food refrigeration, are also being used. Manufacturers of residential and
commercial air conditioning have announced their plans to use HFC-32 and R-454B100 in the future, and at least
one manufacturer has announced the availability of chillers operating on HFC-32 as of 2023 (Carrier, 2023). These
refrigerants are emitted to the atmosphere during equipment operation (as a result of component failure, leaks,
and purges), as well as at manufacturing (if charged at the factory), installation, servicing, and disposal events.
98 R-410A contains HFC-32 and HFC-125.
99 R-507A, also called R-507, contains HFC-125 and HFC-143a.
100 R_454B contains HFC-32 and HFO-1234yf.
4-134 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Aerosols
Aerosol propellants are used in metered dose inhalers (MDIs) and a variety of personal care products and
technical/specialty products (e.g., duster sprays and safety horns). Pharmaceutical companies that produce MDIs—
a type of inhaled therapy used to treat asthma and chronic obstructive pulmonary disease—have replaced the use
of CFCs with HFC-propellant alternatives. The earliest ozone-friendly MDIs were produced with HFC-134a, but the
industry is using HFC-227ea as well. Conversely, since the use of CFC propellants in other types of aerosols was
banned in 1978, most non-medical consumer aerosol products have not transitioned to HFCs, but to "not-in-kind"
technologies, such as solid or roll-on deodorants and finger-pump sprays. The transition away from ODSs in
specialty aerosol products has also led to the introduction of non-fluorocarbon alternatives (e.g., hydrocarbon
propellants) in certain applications, in addition to HFC-134a or HFC-152a. Other low-GWP options such as HFO-
1234ze(E) are being used as well. These propellants are released into the atmosphere as the aerosol products are
used.
Foams
Chlorofluorocarbons and HCFCs have traditionally been used as foam blowing agents to produce polyurethane
(PU), polystyrene, polyolefin, and phenolic foams, which are used in a wide variety of products and applications.
Since the Montreal Protocol, flexible PU foams as well as other types of foam, such as polystyrene sheet,
polyolefin, and phenolic foam, have transitioned almost completely away from fluorocompounds into alternatives
such as CO2 and hydrocarbons. The majority of rigid PU foams have transitioned to HFCs—primarily HFC-134a and
HFC-245fa. Today, these HFCs are used to produce PU appliance, PU commercial refrigeration, PU spray, and PU
panel foams—used in refrigerators, vending machines, roofing, wall insulation, garage doors, and cold storage
applications. In addition, HFC-152a, HFC-134a, and CO2 are used to produce polystyrene sheet/board foam, which
is used in food packaging and building insulation. Low-GWP fluorinated foam blowing agents in use include HFO-
1234ze(E) and HCFO-1233zd(E). Emissions of blowing agents occur when the foam is manufactured as well as
during the foam lifetime and at foam disposal, depending on the particular foam type.
Solvents
Chlorofluorocarbons, methyl chloroform (1,1,1-trichloroethane or TCA), and to a lesser extent carbon tetrachloride
(CCU) were historically used as solvents in a wide range of cleaning applications, including precision, electronics,
and metal cleaning. Since their phaseout, metal cleaning end-use applications have primarily transitioned to non-
fluorocarbon solvents and not-in-kind processes. The precision and electronics cleaning end-uses have transitioned
in part to high-GWP gases, due to their high reliability, excellent compatibility, good stability, low toxicity, and
selective solvency. These applications rely on HFC-43-10mee, HFC-365mfc, HFC-245fa, and to a lesser extent, PFCs.
Electronics cleaning involves removing flux residue that remains after a soldering operation for printed circuit
boards and other contamination-sensitive electronics applications. Precision cleaning may apply to either
electronic components or to metal surfaces, and is characterized by products, such as disk drives, gyroscopes, and
optical components, that require a high level of cleanliness and generally have complex shapes, small clearances,
and other cleaning challenges. The use of these solvents yields fugitive emissions of these HFCs and PFCs.
Fire Protection
Fire protection applications include portable fire extinguishers ("streaming" applications) that originally used halon
1211, and total flooding applications that originally used halon 1301, as well as some halon 2402. Since the
production and import of virgin halons were banned in the United States in 1994, the halon replacement agent of
choice in the streaming sector has been dry chemical, although HFC-236fa is also used to a limited extent. In the
total flooding sector, HFC-227ea has emerged as the primary replacement for halon 1301 in applications that
require clean agents. Other HFCs, such as HFC-23 and HFC-125, are used in smaller amounts. The majority of HFC-
227ea in total flooding systems is used to protect essential electronics, as well as in civil aviation, military mobile
weapons systems, oil/gas/other process industries, and merchant shipping. Fluoroketone FK-5-1-12 is also used as
Industrial Processes and Product Use 4-135
-------
1 a low-GWP option and 2-BTP is being considered. As fire protection equipment is tested or deployed, emissions of
2 these fire protection agents occur.
3 Methodology and Time-Series Consistency
4 A detailed Vintaging Model of ODS-containing equipment and products was used to estimate the actual—versus
5 potential—emissions of various ODS substitutes, including HFCs, PFCs, and CO2. The name of the model refers to
6 the fact that it tracks the use and emissions of various compounds for the annual "vintages" of new equipment
7 that enter service in each end-use. The Vintaging Model predicts ODS and ODS substitute use in the United States
8 based on modeled estimates of the quantity of equipment or products sold each year containing these chemicals
9 and the amount of the chemical required to manufacture and/or maintain equipment and products over time.
10 Emissions for each end-use were estimated by applying annual leak rates and release profiles, which account for
11 the lag in emissions from equipment as they leak over time. By aggregating the data for 78 different end-uses, the
12 model produces estimates of annual use and emissions of each compound. Further information on the Vintaging
13 Model is contained in Annex 3.9.
14 Methodological approaches were applied to the entire time series to ensure time-series consistency from 1990
15 through 2021.
is Uncertainty
17 Given that emissions of ODS substitutes occur from thousands of different kinds of equipment and from millions of
18 point and mobile sources throughout the United States, emission estimates must be made using analytical tools
19 such as the Vintaging Model or the methods outlined in IPCC (2006). Though the model is more comprehensive
20 than the IPCC default methodology, significant uncertainties still exist with regard to the levels of equipment sales,
21 equipment characteristics, and end-use emissions profiles that were used to estimate annual emissions for the
22 various compounds.
23 The uncertainty analysis quantifies the level of uncertainty associated with the aggregate emissions across the 78
24 end-uses in the Vintaging Model. In order to calculate uncertainty, functional forms were developed to simplify
25 some of the complex "vintaging" aspects of some end-use sectors, especially with respect to refrigeration and air-
26 conditioning, and to a lesser degree, fire extinguishing. These sectors calculate emissions based on the entire
27 lifetime of equipment, not just equipment put into commission in the current year, thereby necessitating
28 simplifying equations. The functional forms used variables that included growth rates, emission factors, transition
29 from ODSs, change in charge size as a result of the transition, disposal quantities, disposal emission rates, and
30 either stock (e.g., number of air conditioning units in operation) for the current year or ODS consumption before
31 transition to alternatives began (e.g., in 1985 for most end-uses). Uncertainty was estimated around each variable
32 within the functional forms based on expert judgment, and a Monte Carlo analysis was performed.
33 The most significant sources of uncertainty for the ODS Substitutes source category include the total stock of
34 refrigerant installed in industrial process refrigeration and cold storage equipment, as well as the charge size for
35 technical aerosols using HFC-134a.
36 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-103. Substitution of
37 ozone depleting substances HFC and PFC emissions were estimated to be between 165.2 and 197.8 MMT CO2 Eq.
38 at the 95 percent confidence level. This indicates a range of approximately 4.2 percent below to 14.7 percent
39 above the emission estimate of 172.5 MMT CO2 Eq.
4-136 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Table 4-103: Approach 2 Quantitative Uncertainty Estimates for HFC and PFC Emissions
2 from ODS Substitutes (MMT CO2 Eq. and Percent)
2021 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
Source
Gases
(MMT CO? Eq.)
(MMT CO? Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Substitution of Ozone
Depleting Substances
HFCs and
PFCs
172.5
165.2
197.8
-4.2%
+14.7%
3 a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
4
5 QA/QC and Verification
6 For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
1 Chapter 6 of the 2006IPCC Guidelines, see the QA/QC and Verification Procedures section in the introduction of
8 the IPPU chapter. Category specific QA/QC findings are described below.
9 The QA and verification process for individual gases and sources in the Vintaging Model includes review against up-
10 to-date market information, including equipment stock estimates, leak rates, and sector transitions to new
11 chemicals and technologies. In addition, comparisons against published emission and consumption sources by gas
12 and by source are performed when available as described further below. Independent peer reviews of the
13 Vintaging Model are periodically performed, including one conducted in 2017 (EPA 2018), to confirm Vintaging
14 Model estimates and identify updates. The HFCs and PFCs within the unspecified mix of HFCs and PFCs are
15 modelled and verified individually in the same process as all other gases and sources in the Vintaging Model. For
16 the purposes of reporting emissions to protect Confidential Business Information (CBI), some HFCs and PFCs are
17 grouped into an unspecified mix. In addition, comparisons against published emission and consumption sources by
18 gas and by source are performed when available as described further below.
19 Comparison of Reported Consumption to Modeled Consumption of HFCs
20 Data from EPA's Greenhouse Gas Reporting Program (GHGRP)101 was also used to perform quality assurance as a
21 reference scenario check on the modeled net supply of HFCs, from which the modeled emissions from this source
22 category are derived as specified in 2006 IPCC Guidelines for National Greenhouse Gas Inventories.
23 Reported Net Supply (GHGRP Top-Down Estimate). Consumption patterns demonstrated through data reported
24 under GHGRP Subpart OO (Suppliers of Industrial Greenhouse Gases) and Subpart QQ (Importers and Exporters of
25 Fluorinated Greenhouse Gases Contained in Pre-Charged Equipment or Closed-Cell Foams) were compared to the
26 modeled demand for new saturated HFCs used as ODS substitutes from the Vintaging Model. The collection of
27 data from suppliers of HFCs enables EPA to calculate the reporters' aggregated net supply-the sum of the
28 quantities of chemical produced or imported into the United States less the sum of the quantities of chemical
29 transformed (used as a feedstock in the production of other chemicals), destroyed, or exported from the United
30 States.102This allows for an overall quality assurance check on the modeled demand for new chemical in the
31 Vintaging Model as a proxy for total amount supplied, which is similar to net supply, as an input to the emission
32 calculations in the model. Under EPA's GHGRP, suppliers (i.e., producers, importers, and exporters) of HFCs under
101 For the GHGRP data, EPA verifies annual facility-level and company-level reports through a multi-step process (e.g.,
including a combination of pre-and post-submittal electronic checks and manual reviews by staff) to identify potential errors
and ensure that data submitted to EPA are accurate, complete, and consistent (EPA 2015). Based on the results of the
verification process, EPA follows up with facilities to resolve mistakes that may have occurred. The post-submittals checks are
consistent with a number of general and category-specific QC procedures, including range checks, statistical checks, algorithm
checks, and year-to-year checks of reported data.
102 Chemical that is exported, transformed, or destroyed—unless otherwise imported back to the United States—will never be
emitted in the United States.
Industrial Processes and Product Use 4-137
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Subpart OO103 began annually reporting their production, transformation, destruction, imports, and exports to EPA
in 2011 (for supply that occurred in 2010) and suppliers of HFCs under Subpart QQ began annually reporting their
imports and exports to EPA in 2012 (for supply that occurred in 2011).
Note, GHGRP data reported under subparts QQ and OO are not used directly to estimate emissions of ODS
Substitutes because they do not include complete information on the sectors or end-uses in which that chemical
will be used. Therefore, it does not provide the data that would be needed to calculate the source or time that a
chemical is emitted. Reports to the GHGRP on production and bulk import (Subpart 00) do not currently include
any information on expected end-uses. Published data on fluorinated gases contained in pre-charged equipment
and closed-cell foams (Subpart QQ) does not provide information on the type of product imported or exported.
Furthermore, the information from both subparts would not capture the entire market in the United States.
Modeled Consumption (Vintaging Model Bottom-Up Estimate). The Vintaging Model, used to estimate emissions
from this source category, calculates chemical demand based on the quantity of equipment and products sold,
serviced and retired each year, and the amount of the chemical required to manufacture and/or maintain the
equipment and products on an end-use basis.104 It is assumed that the total demand equals the amount supplied
by either new production, chemical import, or quantities recovered (often reclaimed) and placed back on the
market. In the Vintaging Model, demand for new chemical, as a proxy for consumption, is calculated as any
chemical demand (either for new equipment or for servicing existing equipment) that cannot be met through
recycled or recovered material.105 No distinction is made in the Vintaging Model between whether that need is
met through domestic production or imports. To calculate emissions, the Vintaging Model estimates the quantity
released from equipment over time, which varies by product type as detailed in Annex 3.9.1. Thus, verifying the
Vintaging Model's calculated consumption against GHGRP reported data, which does not provide details on the
end-uses where the chemical is used, is not an exact comparison of the Vintaging Model's emission estimates, but
is believed to provide an overall check of the underlying data.
Overall, the Vintaging Model estimates for consumption are lower than the GHGRP data by an average of 9.8
percent across the time series (i.e., 2012 through 2020). The difference is greatest during the last three years (2018
through 2020). A summary of findings from this comparison, potential causes for differences, and related planned
improvements are discussed below. Annex 3.9.2 provides additional information on the comparison of the data
from the GHGRP and Vintaging Model, and a more detailed discussion of the results.
Comparison of Emissions Derived from Atmospheric Measurements to Modeled Emissions
Emissions of some fluorinated greenhouse gases are estimated for the contiguous United States from the National
Oceanic and Atmospheric Administration (NOAA) and were used to perform additional quality control by
comparing the emission estimates derived from atmospheric measurements by NOAA to the bottom-up emission
estimates from the Vintaging Model. The 2019 Refinement to the 2006IPCC Guidelines for National Greenhouse
Gas Inventories (IPCC 2019) Volume 1: General Guidance and Reporting, Chapter 6: Quality Assurance, Quality
Control and Verification notes that atmospheric concentration measurements can provide independent data sets
as a basis for comparison with inventory estimates. Further, it identified fluorinated gases as one of most suitable
103 Among other provisions, the AIM Act of 2020 directed EPA to develop a U.S. production baseline and a U.S. consumption
baseline and to phase down HFC production and consumption relative to those baselines. Data reported to the GHGRP under
Subpart 00 are relevant to the production and consumption baselines. The data below include aggregated Subpart 00 data for
AIM-listed HFCs for reporting years 2012 through 2021 from all companies that reported AIM-listed HFCs, though not all species
were reported in each reporting year.
104 The model builds an inventory of the in-use stock of equipment and products and ODSs and HFCs in each of the sub-
applications. Emissions are subsequently estimated by applying annual and disposal emission rates to each population of
equipment and products. See Annex 3.9.1. for further details on the model.
105 The Vintaging Model does not calculate "consumption" as defined under the Montreal Protocol and the AIM Act, because
the model includes chemical supplied to pre-charge equipment made overseas and sent to the domestic market and does not
include chemical produced or imported in the United States but placed in products shipped to foreign markets.
4-138 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
greenhouse gases for such comparisons. The 2019 Refinement makes this conclusion on fluorinated gases based
on the lack of natural sources, the potential uncertainties in bottom-up inventory methods for some sources, the
long life of many of these gases, and the well-known loss mechanisms. Unlike the more abundant gases in the
Inventory, since there are no known natural sources of HFCs, the HFC emission sources included in this Inventory
account for the majority of total emissions detectable in the atmosphere, and the estimates derived from
atmospheric measurements are driven solely by anthropogenic emissions.
The 2019 Refinement provides guidance on conducting such comparisons (as summarized in Table 6.2 of IPCC 2019
Volume 1, Chapter 6) and provides guidance on using such comparisons to identify areas of improvement in
national inventories (as summarized in Box 6.5 of IPCC 2019 Volume 1, Chapter 6).
Emission estimates for four key HFCs (HFC-134a, HFC-125, HFC-143a, and HFC-32) from Hu et al. (2017) for 2008
through 2014 were examined in the 2022 Inventory (EPA 2022b). Recently updated estimates from 2008 through
2020 provided from Hu et al. (2022) were used here for an updated comparison over a longer time series. This
provides a quality check on the modeled emissions reported above. Hu et al. (2017) provided similar comparisons;
here additional emissions estimates from Hu et al. (2022) are incorporated and the EPA data used in Hu et al.
(2017) was updated to reflect the current Inventory estimates and extended to the whole time series. Annex 3.9.2
provides additional details on the data from NOAA as compared to the Vintaging Model and a more detailed
discussion of the results. Potential Inventory updates identified due to the current comparison with atmospheric
data are noted in the Planned Improvements section below.
Summary of Comparisons
Comparing the Vintaging Model's estimates to GHGRP-reported estimates of supply and emissions estimates
derived from atmospheric measurements, particularly for more widely used chemicals, can help validate the
model. These comparisons show that Vintaging Model consumption estimates are well within the same order of
magnitude as the actual consumption data as reported to EPA's GHGRP although the differences in reported net
supply and modeled demand are still significant, in particular for more recent years. Using a Tier 2 bottom-up
modeling methodology to estimate emissions requires assumptions and expert judgment so it is expected that the
model will have limitations. The differences (i.e., higher net supply seen in GHGRP compared to the modeled
supply) are likely due to temporal discrepancies, including 1) the top-down data are reported at the time of actual
production or import, and the bottom-up data are calculated at the time of actual placement on the market and 2)
stockpiling of chemicals by suppliers and distributors to produce or import additional quantities of HFCs for various
reasons such as expectations that prices may increase, or supplies may decrease, in the future (e.g., in response to
regulations under the AIM Act). Based on information collected by the EPA during previous ODS phasedowns at
the time, such stockpiling behavior was seen, and it is concluded that such behavior similarly exists amongst HFC
suppliers in anticipation of current and recently promulgated controls on HFCs. Any such activity would increase
the GHGRP data as compared to the modeled data. This effect is likely the major reason why there is a divergence
in the comparison above, with the GHGRP data in 2017 through 2020 (i.e., the years following agreement of the
Kigali Amendment to the Montreal Protocol) significantly higher than the modeled data. Improvements of the
model methodology to incorporate a temporal factor could be investigated. Additional discussion on potential
reasons for differences are discussed in Annex 3.9.2.
The comparisons of modeled emissions for four key HFCs show reasonable agreement with atmospheric
measurement derivations of emissions from Hu et. al (2017, 2022), though certain chemicals and during certain
years differences can be significant, most notably modeled emissions of HFC-134a were more than two standard
deviations (2 s.d.) higher than those seen through atmospheric measurements for the years 2008, 2009, and 2011
through 2013, and more than 2 s.d. below the atmospheric measurements for the years 2017 to 2020. Hence,
areas for further research that may improve the modeling are highlighted in planned improvements.
Considering the strengths and weaknesses of three independent approaches for estimating consumption and
emissions of these HFCs, in most instances the estimates provide added confidence in EPA's understanding of total
U.S. emissions for these chemicals and how they've change over time and, furthermore, the comparisons have
helped identify areas for potential improvement in the future. Annex 3.9.2 provides a more detailed discussion of
the results.
Industrial Processes and Product Use 4-139
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
Recalculations Discussion
For the current Inventory, updates to the Vintaging Model included updating 2021 growth rates for residential and
commercial unitary air-conditioning to align with annual sales estimates published by AHRI. Projected growth rates
were updated for residential unitary air-conditioning to align with projected residential housing available from the
Energy Information Administration (EIA) and commercial unitary air-conditioning growth rates were updated
based on new commercial floorspace growth projections from EIA (EPA 2022c).
Refrigerant transitions for road transport and modern rail transport were updated to reflect manufacturer
announcements regarding the use of R-452A in place of R-404A (EPA 2022d). Manufacturing emissions for
domestic refrigerator foam were adjusted to only include equipment manufactured within the United States,
including those that are produced for export, and excluding those that are imported with foam.
The current Inventory also began reporting CO2 emissions from ODS substitute use as a refrigerant, foam blowing
agent, and fire extinguishing agent. The impact of this addition has very little effect to total emissions across the
timer series; for example, CO2 emissions represent 0.002 percent of CCh-equivalent total emissions in 2021.
In addition, for the current Inventory, CC>2-equivalent emissions totals of HFCs and PFCs from ODS substitutes have
been revised to reflect the 100-year GWPs provided in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). AR5
GWP values differ slightly from those presented in AR4 (IPCC 2007) used in the previous inventories. The AR5
GWPs have been applied across the entire time series for consistency. The GWPs of HFC-134a and HFC-125, the
two most significant contributors to total emissions in this source category, have decreased, from 1,430 to 1,300
and from 3,500 to 3,170, respectively, leading to a decrease in calculated CC>2-equivalent emissions for those HFCs.
In contrast, the GWPs of HFC-32 and HFC-143a, the third and fourth most significant contributors to total
emissions in this source category, have increased, from 675 to 677 and from 4,470 to 4,800, respectively, leading
to an increase in calculated CC>2-equivalent emissions for those HFCs. Compared to the previous Inventory which
applied 100-year GWP values from AR4, the average annual changes in CC>2-equivalent emissions across the time
series 1990-2020 for the four most prevalent HFCs were a 9 percent decrease for HFC-134a, 9 percent decrease for
HFC-125, 0.3 percent increase for HFC-32, and 7 percent increase for HFC-143a. The net impact from these
updates and the additional updates noted above was an average annual 5.6 percent decrease in total emissions for
the time series. Further discussion on this update and the overall impacts of updating the GWP values to reflect
the IPCC Fifth Assessment Report can be found in Chapter 9, Recalculations and Improvements.
Planned Improvements
Future improvements to the Vintaging Model are planned for the Refrigeration and Air-conditioning, Fire
Suppression, and Aerosols sectors. Specifically, refrigerated storage space estimates published biannually from the
United States Department of Agriculture (USDA) are being compared to cold storage warehouse space currently
estimated in the Vintaging Model. EPA is also reviewing the addition of an end-use representing multi-split air-
conditioning units. Streaming agent fire suppression lifetimes, market size, and growth rates and flooding agent
fire suppression market transitions are under review to align more closely with real world activities. In addition,
further refinement of HFC consumption in MDIs is expected from review of data collected on HFC use for MDI
production, imports, and exports in response to requests for application-specific allowances for MDIs. EPA expects
these revisions to be prepared for the 2024 or 2025 Inventory submission.
As discussed above, future reporting under the AIM Act may provide useful information for verification purposes
and possible improvements to the Vintaging Model, such as information on HFC stockpiling behaviors. EPA expects
this reporting by early 2023 and incorporation into the 2024 or 2025 report. Should the data suggest structural
changes to the model, such as the handling of stockpiles before use, EPA expects to introduce the revised model
for the 2025 or 2026 Inventory submission.
Several potential improvements to the Inventory were identified in the 2022 Inventory submission based on the
comparisons discussed above—net supply values from the GHGRP and emission estimates derived from
atmospheric measurements—and remain valid. To estimated HFC emissions for just the contiguous United States,
matching the coverage by the atmospheric measurements, EPA will investigate the availability of data from Alaska,
4-140 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Hawaii, and U.S. territories. This is planned by the 2025 Inventory submission. To improve estimates of HFC-125
and HFC-143a, further research into the refrigeration market can be made. Research in this industry on the shift
away from blends such as R-404A or success in lowering emission rates could be used to improve the Inventory
estimate. This is planned for the 2024 Inventory cycle. That said, for the years where both the atmospheric
measurements and the model display a roughly constant emission of HFC-143a at similar levels, the new results
suggest robust estimates for the refrigeration market. Uncertainty estimates by species would aid in comparisons
to atmospheric data. EPA will explore the possibility of revising the Monte Carlo analysis to differentiate between
species, starting with the higher-emitted HFCs identified above, in a future (i.e., 2024 or 2025) Inventory
submission. Reclamation reports and, when available, information gathered under the AIM Act, could be used to
improve the understanding of how chemical moves through the economy and could resolve some of the temporal
effects discussed in Annex 3.9.2. This would likely require revisions to the basic model structure and could be
introduced for the 2025 or 2026 Inventory submission. The additional data from the atmospheric measurements
suggests additional items to investigate. The faster uptick in HFC-32 and HFC-125 emissions suggests additional
emissions of R-410A compared to the model's estimation. Further investigation into the emission rate, whether
that varies over time, stocks, lifetimes, and other factors will be investigated for the 2025 Inventory submission.
4.25 Electrical Transmission and Distribution
(CRF Source Category 2G1)
The largest use of sulfur hexafluoride (SFs), both in the United States and internationally, is as an electrical
insulator and interrupter in equipment that transmits and distributes electricity (RAND 2004). The gas has been
employed by the electric power industry in the United States since the 1950s because of its dielectric strength and
arc-quenching characteristics. It is used in gas-insulated substations, circuit breakers, and other switchgear. SF6 has
replaced flammable insulating oils in many applications and allows for more compact substations in dense urban
areas. Another greenhouse gas emitted in much smaller amounts by the electric power industry is
tetrafluoromethane (CF4), which is mixed with SF6 to avoid liquefaction at low temperatures (Middleton 2000).
While mixed gas circuit breakers are more common in extremely cold climates in geographies outside of the
United States, some U.S. manufacturers of electrical equipment are emitting CF4 during the manufacturing of
equipment designed to hold the SF6/CF4 gas mixture. However, no electrical transmission and distribution facilities
in the United States have reported emissions of or equipment using CF4. SF6 emissions exceed PFC emissions from
electric power systems on both a GWP-unweighted and CC>2-equivalent basis.
Fugitive emissions of SF6 and CF4 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. Emissions of SF6 and CF4 from equipment manufacturing and from electrical transmission
and distribution systems were estimated to be 5.98 MMT CO2 Eq. (0.3 kt) in 2021. This quantity represents a 76
percent decrease from the estimate for 1990 (see Table 4-104 and Table 4-105). There are a few potential causes
for this decrease: a sharp increase in the price of SF6 during the 1990s and a growing awareness of the
environmental impact of SF6 emissions through programs such as EPA's voluntary SF6 Emission Reduction
Partnership for Electric Power Systems (Partnership) and EPA's GHGRP, regulatory drivers at the state and local
levels, and research and development of alternative gases to SF6 that can be used in gas-insulated substations.
Utilities participating in the Partnership have lowered their emission factor from 13 percent in 1999 (kg SF6 emitted
per kg of nameplate capacity) to 1 percent in 2021. SF6 emissions reported by electric power systems to EPA's
GHGRP have decreased by 42 percent from 2011 to 2021,106 with much of the reduction seen from utilities that
106 Analysis of emission trends from facilities reporting to EPA's GHGRP is imperfect due to an inconsistent group of reporters
year to year. A facility that has reported total non-biogenic greenhouse gas emissions below 15,000 metric tons of carbon
Industrial Processes and Product Use 4-141
-------
1 are not participants in the Partnership. These utilities may be making relatively large reductions in emissions as
2 they take advantage of relatively large and/or inexpensive emission reduction opportunities (i.e., "low hanging
3 fruit," such as replacing major leaking circuit breakers) that Partners have already taken advantage of under the
4 voluntary program (Ottinger et al. 2014). However, total emissions from electrical transmission and distribution in
5 2021 were higher than 2020 emissions, increasing by 2.17 percent, largely due to a large increase in transmission
6 miles.
7 Table 4-104: SF6 and CF4 Emissions from Electric Power Systems and Electrical Equipment
8 Manufacturers (MMT CO2 Eq.)
Electrical
Electric Power Equipment
Year Systems Manufacturers Total
1990 243 03 24.7
2005 11.2 0.7 11.8
2017 5.2 0.3 5.5
2018 4.9 0.3 5.2
2019 5.7 0.4 6.1
2020 5.3 0.5 5.9
2021 5.6 0.4 6.0
Note: Totals may not sum due to independent rounding.
9 Table 4-105: SF6 and CF4 Emissions from Electric Power Systems and Electrical Equipment
10 Manufacturers (kt)
Year SF6 Emissions CF4 Emissions
1990 1.0 NO
2005 0.5 0.00031
2017 0.2 +
2018 0.2 NO
2019 0.3 0.00006
2020 0.2 0.00002
202 1 03 0.00016
+ Does not exceed 0.000005 kt.
NO (Not Occurring)
11 Methodology and Time-Series Consistency
12 The estimates of emissions from Electrical Transmission and Distribution are comprised of emissions from electric
13 power systems and emissions from the manufacture of electrical equipment. The methodologies for estimating
14 both sets of emissions are described below.
dioxide equivalent (MT C02 Eq.) for three consecutive years or below 25,000 MT C02 Eq. for five consecutive years to EPA's
GHGRP can discontinue reporting for all direct emitter subparts. For this sector, most of the variability in the group of reporters
is due to facilities exiting the GHGRP due to being below one of these thresholds; however, facilities must re-enter the program
if their emissions at a later date are above 25,000 MT C02 Eq., which may occur for a variety of reasons, including changes in
facility size and changes in emission rates.
4-142 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1990 through 1998 Emissions from Electric Power Systems
Emissions from electric power systems from 1990 through 1998 were estimated based on (1) the emissions
estimated for this source category in 1999, which, as discussed in the next section, were based on the emissions
reported during the first year of EPA's SF6 Emission Reduction Partnership for Electric Power Systems (Partnership),
and (2) the RAND survey of global SF6 emissions. Because most utilities participating in the Partnership reported
emissions only for 1999 through 2011, modeling was used to estimate SF6 emissions from electric power systems
for the years 1990 through 1998. To perform this modeling, U.S. emissions were assumed to follow the same
trajectory as global emissions from this source during the 1990 through 1999 period. To estimate global emissions,
the RAND survey of global SF6 sales was used, together with the following equation for estimating emissions, which
is derived from the mass-balance equation for chemical emissions (Volume 3, Equation 7.3) in the 2006IPCC
Guidelines.107 (Although Equation 7.3 of the 2006 IPCC Guidelines appears in the discussion of substitutes for
ozone-depleting substances, it is applicable to emissions from any long-lived pressurized equipment that is
periodically serviced during its lifetime.)
Equation 4-23: Estimation for SF6 Emissions from Electric Power Systems
Emissions (kilograms SFs) = SF6 purchased to refill existing equipment (kilograms) + nameplate capacity of retiring
equipment (kilograms)108
Note that the above equation holds whether the gas from retiring equipment is released or recaptured; if the gas
is recaptured, it is used to refill existing equipment, thereby lowering the amount of SF6 purchased by utilities for
this purpose.
Gas purchases by utilities and equipment manufacturers from 1961 through 2003 are available from the RAND
(2004) survey. To estimate the quantity of SF6 released or recovered from retiring equipment, the nameplate
capacity of retiring equipment in a given year was assumed to equal 81.2 percent of the amount of gas purchased
by electrical equipment manufacturers 40 years previous (e.g., in 2000, the nameplate capacity of retiring
equipment was assumed to equal 81.2 percent of the gas purchased in 1960). The remaining 18.8 percent was
assumed to have been emitted at the time of manufacture. The 18.8 percent emission factor is an average of IPCC
default SFs emission rates for Europe and Japan for 1995 (IPCC 2006). The 40-year lifetime for electrical equipment
is also based on IPCC (2006). The results of the two components of the above equation were then summed to yield
estimates of global SF6 emissions from 1990 through 1999.
U.S. emissions between 1990 and 1999 are assumed to follow the same trajectory as global emissions during this
period. To estimate U.S. emissions, global emissions for each year from 1990 through 1998 were divided by the
estimated global emissions from 1999. The result was a time series of factors that express each year's global
emissions as a multiple of 1999 global emissions. Historical U.S. emissions were estimated by multiplying the factor
for each respective year by the estimated U.S. emissions of SF6 from electric power systems in 1999 (estimated to
be 14.0 MMT C02 Eq.).
Two factors may affect the relationship between the RAND sales trends and actual global emission trends. One is
utilities' inventories of SF6 in storage containers. When SF6 prices rise, utilities are likely to deplete internal
inventories before purchasing new SF6 at the higher price, in which case SF6 sales will fall more quickly than
emissions. On the other hand, when SF6 prices fall, utilities are likely to purchase more SF6 to rebuild inventories, in
which case sales will rise more quickly than emissions. This effect was accounted for by applying 3-year smoothing
to utility SFs sales data. The other factor that may affect the relationship between the RAND sales trends and
actual global emissions is the level of imports from and exports to Russia and China. SF6 production in these
countries is not included in the RAND survey and is not accounted for in any another manner by RAND. However,
107 Ideally, sales to utilities in the United States between 1990 and 1999 would be used as a model. However, this information
was not available. There were only two U.S. manufacturers of SF6 during this time period, so it would not have been possible to
conceal sensitive sales information by aggregation.
108 Nameplate capacity is defined as the amount of SF6 within fully charged electrical equipment.
Industrial Processes and Product Use 4-143
-------
1 atmospheric studies confirm that the downward trend in estimated global emissions between 1995 and 1998 was
2 real (see the Uncertainty discussion below).
3 1999 through 2021 Emissions from Electric Power Systems
4 Emissions from electric power systems from 1999 to 2021 were estimated based on: (1) reporting from utilities
5 participating in EPA's SF6 Emission Reduction Partnership for Electric Power Systems (Partners), which began in
6 1999; (2) reporting from utilities covered by EPA's GHGRP, which began in 2012 for emissions occurring in 2011
7 (GHGRP-Only Reporters); and (3) the relationship between utilities' reported emissions and their transmission
8 miles as reported in the 2001, 2004, 2007, 2010, 2013, and 2016 Utility Data Institute (UDI) Directories of Electric
9 Power Producers and Distributors (UDI 2001, 2004, 2007, 2010, 2013, and 2017), and 2019, 2020, and 2021
10 Homeland Infrastructure Foundation-Level Data (HIFLD) (HIFLD 2019, 2020, and 2021), which was applied to the
11 electric power systems that do not report to EPA (Non-Reporters). Total U.S. transmission mileage was
12 interpolated between 2016 and 2019 to estimate transmission mileage of electric power systems in 2017 and
13 2018. (Transmission miles are defined as the miles of lines carrying voltages above 34.5 kV).
14 Partners
15 Over the period from 1999 to 2021, Partner utilities, which for inventory purposes are defined as utilities that
16 either currently are or previously have been part of the Partnership,109 represented 49 percent, on average, of
17 total U.S. transmission miles. Partner utilities estimated their emissions using a Tier 3 utility-level mass balance
18 approach (IPCC 2006). If a Partner utility did not provide data for a particular year, emissions were interpolated
19 between years for which data were available or extrapolated based on Partner-specific transmission mile growth
20 rates. In 2012, many Partners began reporting their emissions (for 2011 and later years) through EPA's GHGRP
21 (discussed further below) rather than through the Partnership. In 2021, less than 1 percent of the total emissions
22 attributed to Partner utilities were reported through Partnership reports. Approximately 99.7 percent of the total
23 emissions attributed to Partner utilities were reported and verified through EPA's GHGRP.110 Overall, the emission
24 rates reported by Partners have decreased significantly throughout the time series.
25 Non-Partners
26 Non-Partners consist of two groups: Utilities that have reported to the GHGRP beginning in 2012 (reporting 2011
27 emissions) or later years (GHGRP-only Reporters) and utilities that have never reported to the GHGRP (Non-
28 Reporters). EPA's GHGRP requires users of SF6 in electric power systems to report emissions if the facility has a
29 total SFs nameplate capacity that exceeds 17,820 pounds. (This quantity is the nameplate capacity that would
30 result in annual SF6 emissions equal to 25,000 metric tons of C02 equivalent at the historical emission rate reported
31 under the Partnership.) As under the Partnership, electric power systems that report their SF6 emissions under
32 EPA's GHGRP are required to use the Tier 3 utility-level mass-balance approach. GHGRP-Only Reporters accounted
33 for 16 percent of U.S. transmission miles and 13 percent of estimated U.S. emissions from electric power system in
34 2021.111
109 Starting in the 1990 to 2015 Inventory, partners who had reported three years or less of data prior to 2006 were removed.
Most of these Partners had been removed from the list of current Partners but remained in the Inventory due to the
extrapolation methodology for non-reporting partners.
110 Only data reported as of August 12, 2022 are used in the emission estimates for the prior year of reporting. Emissions for
Partners that did not report to the Partnership or GHGRP are extrapolated for three years using a utility-specific transmission
mile growth rate. After four consecutive years of non-reporting they are included in the 'non-reporting Partners' category. It
should be noted that data reported through EPA's GHGRP must go through a verification process. For electric power systems,
verification involved a series of electronic range, completeness, and algorithm checks for each report submitted.
111 GHGRP-reported and Partner transmission miles from a number of facilities were equal to zero with non-zero emissions.
These facilities emissions were added to the emissions totals for their respective parent companies when identifiable and not
included in the regression equation when not identifiable or applicable. Other facilities reported non-zero transmission miles
with zero emissions, or zero transmission miles and zero emissions. These facilities were not included in the development of the
4-144 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 From 1999 through 2010, emissions from both GHGRP-only Reporters and Non-Reporters were estimated in the
2 same way. From 1999 through 2008, emissions were estimated using the results of a regression analysis that
3 correlated the 1999 emissions from Partner utilities with their 1999 transmission miles.112 The 1999 regression
4 coefficient (emission factor) was held constant through 2008 and multiplied by the transmission miles estimated
5 for the non-Partners for each year.
6 The 1999 regression equation for Non-Partners was developed based on the emissions reported by a subset of
7 Partner utilities who reported non-zero emissions and non-zero transmission miles (representing approximately 50
8 percent of total U.S. transmission miles). The regression equation for 1999 is displayed in the equation below.
9 Equation 4-24: Regression Equation for Estimating SF6 Emissions of Non-Reporting Facilities
10 in 1999
11
12 Emissions (kg) = 0.771 x Transmission Miles
13
14 For reasons discussed further below in the Recalculations section, the emission factor for the non-Partners was
15 assumed to decrease beginning in 2009, trending toward the regression coefficient (emission factor) calculated for
16 the GHGRP-only reporters based on their reported 2011 emissions and transmission miles. Emission factors for
17 2009 and 2010 were linearly interpolated between the 1999 and 2011 emission factors. For 2009, the emissions of
18 non-Partners were estimated by multiplying their transmission miles by the interpolated 2009 emission factor
19 (0.65 kg/transmission mile).
20 The 2011 regression equation was developed based on the emissions reported by GHGRP-Only Reporters who
21 reported non-zero emissions and non-zero transmission miles (representing approximately 23 percent of total U.S.
22 transmission miles). The regression equation for 2011 is displayed below.
23 Equation 4-25: Regression Equation for Estimating SF6 Emissions of GHGRP-Only Reporters
24 in 2011
25
26 Emissions (kg) = 0.397 x Transmission Miles
27
28 For 2011 and later years, the emissions of GHGRP-only reporters were generally equated to their reported
29 emissions, unless they did not report. The emissions of GHGRP-only reporters that have years of non-reporting
30 between reporting years are gap filled by interpolating between reported values.
31 For 2010 and later years, the emissions of non-Reporters were estimated by multiplying their transmission miles by
32 the estimated 2010 emission factor (0.52 kg/transmission mile), which was held constant from 2010 through 2021.
33 Off-ramping GHGRP Facilities
34 The GHGRP program has an "off-ramp" provision (40 CFR Part 98.2(i)) that exempts facilities from reporting under
35 certain conditions. If reported total greenhouse gas emissions are below 15,000 metric tons of carbon dioxide
36 equivalent (MT CO2 Eq.) for three consecutive years or below 25,000 MT CO2 Eq. for five consecutive years, the
37 facility may elect to discontinue reporting. Emissions of GHGRP reporters that have off-ramped are extrapolated
38 for three years of non-reporting using a utility-specific transmission mile growth rate. After three consecutive years
39 of non-reporting, emissions for facilities that off-ramped from GHGRP were estimated using an emissions rate
40 derived from the reported emissions and transmission miles of GHGRP-only reporters in the respective year.
regression equations (discussed further below). These emissions are already implicitly accounted for in the relationship
between transmission miles and emissions.
112 In the United States, SF6 is contained primarily in transmission equipment rated above 34.5 kV.
Industrial Processes and Product Use 4-145
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Table 4-106: GHGRP-only Average Emission Rate (kg per mile)
2011
2017
2018
2019
2020
2021
Average emission rate
0.40
0.24
0.22
0.28
0.26
0.25
Table 4-107: Categorization of Utilities and Timeseries for Application of Corresponding
Categorization of Utilities
Timeseries
Partners
1999-2021
Non-Partners (GHGRP-Only)
2011-2021
Non-Partners (Remaining Non-
Reporting Utilities)
1999-2021
Off-ramping GHGRP Facilities
2017-2021
Total Industry Emissions
As a final step, total electric power system emissions from 1999 through 2021 were determined for each year by
summing the Partner reported and estimated emissions (reported data was available through the EPA's SF6
Emission Reduction Partnership for Electric Power Systems), the GHGRP-only reported emissions, off-ramping
GHGRP Facilities (non-reporters), non-reporters who eventually report to GHGRP, and the non-reporting utilities'
emissions.
Non-Partner Transmission Miles
Data on transmission miles for each Non-Reporter for the years 2000, 2003, 2006, and 2009, 2012, and 2016 were
obtained from the 2001, 2004, 2007, 2010, 2013, and 2017 UDI Directories of Electric Power Producers and
Distributors, respectively (UDI 2001, 2004, 2007, 2010, 2013, and 2017). For 2019, 2020, and 2021 non-reporter
transmission mileage was derived by subtracting reported transmission mileage data from the total U.S.
transmission mileage from 2019, 2020, and 2021 HIFLD Data (HIFLD 2019, 2020, and 2021). The following trends in
transmission miles have been observed over the time series:
• The U.S. transmission system grew by over 22,000 miles between 2000 and 2003 yet declined by almost
4,000 miles between 2003 and 2006. Given these fluctuations, periodic increases are assumed to occur
gradually. Therefore, transmission mileage was assumed to increase at an annual rate of 1.2 percent
between 2000 and 2003 and decrease by 0.20 percent between 2003 and 2006.
• The U.S. transmission system's annual growth rate grew to 1.7 percent from 2006 to 2009 as transmission
miles increased by more than 33,000 miles.
• The annual growth rate for 2009 through 2012 was calculated to be 1.5 percent as transmission miles
grew yet again by over 30,000 miles during this time period.
• The annual transmission mile growth rate for 2012 through 2016 was calculated to be 0.4 percent, as
transmission miles increased by approximately 10,250 miles.
• The annual transmission mile growth rate for 2016 through 2020 was calculated to be 0.7 percent, as
transmission miles increased by approximately 20,300 miles.
• The annual transmission mile growth rate for 2020 through 2021 was calculated to be 2.2 percent, as
transmission miles increased by approximately 16,152 miles.
Transmission miles for each year for non-reporters were calculated by interpolating between UDI reported values
obtained from the 2001, 2004, 2007, 2010, 2013 and 2017 UDI directories and 2019 HIFLD data. In cases where a
non-reporter previously reported the GHGRP or the Partnership, transmission miles were interpolated between
the most recently reported value and the next available UDI value.
4-146 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
1990 through 2021 Emissions from Manufacture of Electrical Equipment
Three different methods were used to estimate 1990 to 2021 emissions from original electrical equipment
manufacturers (OEMs).
• OEM SFs emissions from 1990 through 2000 were derived by assuming that manufacturing emissions
equaled 10 percent of the quantity of SF6 provided with new equipment. The 10 percent emission rate is
the average of the "ideal" and "realistic" manufacturing emission rates (4 percent and 17 percent,
respectively) identified in a paper prepared under the auspices of the International Council on Large
Electric Systems (CIGRE) in February 2002 (O'Connell et al. 2002). The quantity of SF6 provided with new
equipment was estimated based on statistics compiled by the National Electrical Manufacturers
Association (NEMA). These statistics were provided for 1990 to 2000.
• OEM SFs emissions from 2000 through 2010 were estimated by (1) interpolating between the emission
rate estimated for 2000 (10 percent) and an emission rate estimated for 2011 based on reporting by
OEMs through the GHGRP (5.7 percent), and (2) estimating the quantities of SF6 provided with new
equipment for 2001 to 2010. The quantities of SF6 provided with new equipment were estimated using
Partner reported data and the total industry SF6 nameplate capacity estimate (156.5 MMT CO2 Eq. in
2010). Specifically, the ratio of new nameplate capacity to total nameplate capacity of a subset of
Partners for which new nameplate capacity data was available from 1999 to 2010 was calculated. These
ratios were then multiplied by the total industry nameplate capacity estimate for each year to derive the
amount of SF6 provided with new equipment for the entire industry. Additionally, to obtain the 2011
emission rate (necessary for estimating 2001 through 2010 emissions), the estimated 2011 emissions
(estimated using the third methodology listed below) were divided by the estimated total quantity of SF6
provided with new equipment in 2011. The 2011 quantity of SF6 provided with new equipment was
estimated in the same way as the 2001 through 2010 quantities.
• OEM CF4 emissions from 1991 through 2010 were estimated by using an average ratio of reported SF6 and
CF4 emissions from 2011 through 2013. This ratio was applied to the estimated SF6 emissions for 1991
through 2010 to arrive at CF4 emissions. CF4 emissions are estimated starting in 1991 and assumed zero
prior to 1991 based on the entry of the CF4/SF6 gas mixture into the market (Middleton 2000).
• OEM emissions from 2011 through 2021 were estimated using the SF6 and CF4 emissions from OEMs
reporting to the GHGRP, and an assumption that these reported emissions account for a conservatively
low estimate of 50 percent of the total emissions from all U.S. OEMs.
• OEM SFs emissions from facilities off-ramping from the GHGRP were determined by extrapolation. First,
emission growth rates were calculated for each reporting year for each OEM reporting facility as well as
an average emissions growth rate (2011 to present). Averages of reported emissions from last three
consecutive reporting years were multiplied by the average growth rate for each off-ramping OEM to
estimate emissions for the non-reporting year(s).
Methodological approaches were applied to the entire time series to ensure time-series consistency from 1990
through 2021.
Uncertainty
To estimate the uncertainty associated with emissions of SF6 and CF4 from Electrical Transmission and Distribution,
uncertainties associated with four quantities were estimated: (1) emissions from Partners, (2) emissions from
GHGRP-Only Reporters, (3) emissions from Non-Reporters, and (4) emissions from manufacturers of electrical
equipment. A Monte Carlo analysis was then applied to estimate the overall uncertainty of the emissions estimate.
Total emissions from the SF6 Emission Reduction Partnership include emissions from both reporting (through the
Partnership or EPA's GHGRP) and non-reporting Partners. For reporting Partners, individual Partner-reported SF6
data was assumed to have an uncertainty of 10 percent. Based on a Monte Carlo analysis, the cumulative
Industrial Processes and Product Use 4-147
-------
1 uncertainty of all Partner-reported data was estimated to be 6.3 percent. The uncertainty associated with
2 extrapolated or interpolated emissions from non-reporting Partners was assumed to be 20 percent.
3 For GHGRP-Only Reporters, reported SF6 data was assumed to have an uncertainty of 10 percent. Based on a
4 Monte Carlo analysis, the cumulative uncertainty of all GHGRP-Only reported data was estimated to be 8.3
5 percent.
6 As discussed below, EPA has substantially revised its method for estimating emissions from non-Reporters,
7 assuming that the average emission rate of non-Reporters has declined much more slowly than the average
8 emission rate of reporting facilities rather than declining at the same rate. This assumption brings the U.S. SF6
9 emissions estimated in this Inventory into better agreement with the U.S. SF6 emissions inferred from atmospheric
10 observations. However, it must be emphasized that the actual emission rates of non-Reporters remain unknown. It
11 is possible that they are lower or even higher than estimated here. One possibility is that SF6 sources other than
12 electric power systems are contributing to the emissions inferred from atmospheric observations, implying that
13 the emissions from non-Reporters are lower than estimated here. Another is that the emissions inferred from
14 atmospheric measurements are over- (or under-) estimated, implying that emissions from no-Reporters could be
15 either lower or higher than estimated here. These uncertainties are difficult to quantify and are not reflected in the
16 estimated uncertainty below. The estimated uncertainty below accounts only for the two sources of uncertainty
17 associated with the regression equations used to estimate emissions in 2019 from Non-Reporters: (1) uncertainty
18 in the coefficients (as defined by the regression standard error estimate), and (2) the uncertainty in total
19 transmission miles for Non-Reporters. Uncertainties were also estimated regarding (1) estimates of SF6 and CF4
20 emissions from OEMs reporting to EPA's GHGRP, and (2) the assumption on the percent share of OEM emissions
21 from OEMs reporting to EPA's GHGRP.
22 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-108. Electrical
23 Transmission and Distribution SF6 and CF4 emissions were estimated to be between 4.5 and 7.3 MMT CO2 Eq. at
24 the 95 percent confidence level. This indicates a range of approximately 23 percent below and 25 percent above
25 the emission estimate of 5.8 MMT CO2 Eq.
26 Table 4-108: Approach 2 Quantitative Uncertainty Estimates for SF6 and CF4 Emissions from
27 Electrical Transmission and Distribution (MMT CO2 Eq. and Percent)
2021 Emission
Estimate
Uncertainty Range Relative to 2018 Emission Estimate3
Source
Gas
(MMT CO? Eq.)
(MMT CO? Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
sf6
Electrical Transmission
and
6.0
4.6
7.5
-23%
+25%
and Distribution
cf4
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
28 In addition to the uncertainty quantified above for the 2021 estimate, there is uncertainty associated with the
29 emission rates of GHGRP-only facilities before 2011 and of non-Reporters throughout the time series. As noted
30 above in the discussion of the uncertainty of non-Reporters for 2021, these uncertainties are difficult to quantify.
31 There is also uncertainty associated with using global SF6 sales data to estimate U.S. emission trends from 1990
32 through 1999. However, the trend in global emissions implied by sales of SF6 appears to reflect the trend in global
33 emissions implied by changing SF6 concentrations in the atmosphere. That is, emissions based on global sales
34 declined by 29 percent between 1995 and 1998 (RAND 2004), and emissions based on atmospheric measurements
35 declined by 17 percent over the same period (Levin et al. 2010).
36 Several pieces of evidence indicate that U.S. SF6 emissions were reduced as global emissions were reduced. First,
37 the decreases in sales and emissions coincided with a sharp increase in the price of SF6 that occurred in the mid-
38 1990s and that affected the United States as well as the rest of the world. A representative from DILO, a major
39 manufacturer of SF6 recycling equipment, stated that most U.S. utilities began recycling rather than venting SF6
4-148 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
within two years of the price rise. Finally, the emissions reported by the one U.S. utility that reported its emissions
for all the years from 1990 through 1999 under the Partnership showed a downward trend beginning in the mid-
1990s.
QA/QC and Verification
For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
Chapter 6 of the 2006IPCC Guidelines, see the QA/QC and Verification Procedures section in the introduction of
the IPPU chapter and Annex 8 for more details. Category specific QC findings are described below.
For the GHGRP data, EPA verifies annual facility-level reports through a multi-step process (e.g., including a
combination of pre-and post-submittal electronic checks and manual reviews by staff) to identify potential errors
and ensure that data submitted to EPA are accurate, complete, and consistent (EPA 2015).113 Based on the results
of the verification process, EPA follows up with facilities to resolve mistakes that may have occurred. The post-
submittals checks are consistent with a number of general and category-specific QC procedures including: range
checks, statistical checks, algorithm checks, and year-to-year checks of reported data and emissions.
Comparison of Emissions Derived from Atmospheric Measurements to
Emissions from Bottom-up Estimates
Emissions of SF6 have been estimated for the contiguous United States by the National Oceanic and Atmospheric
Administration (NOAA) based on atmospheric measurements. To provide additional quality control for the SF6
emissions estimates presented in this Inventory, U.S. EPA and NOAA compared the 2007-2018 emission estimates
derived from atmospheric measurements by NOAA to the emission estimates for SF6-emitting source categories in
this Inventory, of which electrical transmission and distribution is by far the largest.114 The 2019 Refinement to the
2006 IPCC Guidelines for National Greenhouse Gas Inventories (IPCC 2019) Volume 1: General Guidance and
Reporting, Chapter 6: Quality Assurance, Quality Control and Verification notes that atmospheric concentration
measurements can provide independent data sets as a basis for comparison with inventory estimates. Further, it
identifies fluorinated gases as particularly suited for such comparisons. The 2019 Refinement makes this conclusion
for fluorinated gases based on their lack of significant natural sources,115 their generally long atmospheric
lifetimes, their well-known loss mechanisms, and the potential uncertainties in bottom-up inventory methods for
some of their sources. Unlike non-fluorinated greenhouse gases (CO2, Cm, and N2O), SF6 has no significant natural
sources; therefore, the SF6 estimates derived from atmospheric measurements are driven overwhelmingly by
anthropogenic emissions. The 2019 Refinement provides guidance on conducting such comparisons (as
summarized in Table 6.2 of IPCC [2019] Volume 1, Chapter 6) and provides guidance on using such comparisons to
identify areas of improvement in national inventories (as summarized in Box 6.5 of IPCC 2019 Volume 1, Chapter
6). Emission estimates for SF6 from Hu et al. (2022) were used in this comparison.
As shown in Figure 4-3, a significant gap existed between the atmosphere-derived emissions for 2007-2018
available in Hu et al., and the inventory estimates for the same years in the 1990 through 2020 Inventory,
particularly in 2010 and earlier years, before reporting through the GHGRP began. With the revisions in
methodology described above and below in the Recalculations Discussion section, the gap between the
atmosphere-derived emissions and the estimates in this Inventory is smaller. Nevertheless, differences remain
between the atmosphere-derived emissions and the Inventory estimates, especially before 2011. EPA is continuing
to research potential contributors to this difference. One potential contributor to the difference before 2011 is an
SFs production plant that operated in Metropolis, Illinois, through 2010, and which is currently unaccounted for in
113 GHGRP Report Verification Factsheet. See https://www.epa.gov/sites/production/files/2Q15-
07/documents/ehgrp verification factsheet.pdf.
114 Other SF6-emitting source categories included in this Inventory include Magnesium Production and Processing and
Electronics Manufacturing.
115 See Harnisch and Eisenhauer (1998).
Industrial Processes and Product Use 4-149
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
the Inventory. While EPA never received reported emissions from this plant, based on production capacity data
from 2006 and the broad range of emission factors observed for production of SF6 and other fluorinated gases, the
plant's SFs emissions would likely have ranged between 30 and 300 metric tons yr1 (Hu et al. 2022). Emissions at
the upper end of this range would explain most of the gap in 2007 and 2008, and a tapering down of emissions
through 2010 might have been expected as the plant reduced production on its way to shutting down. EPA plans
to include estimates of emissions from this plant in a future submission of the Inventory. See Planned
Improvements section below.
Figure 4-3: U.S. Emissions of SF6 Comparison3
1,200
1,000
800
¦a 600
400
200
0
EPA 1990-2020
EPA 1990-2021
NOAA
2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021
aSources: NOAA data from Hu et al. (2023); EPA 1990-2020 Inventory estimates from EPA (2022).
Recalculations Discussion
The historical emissions estimated for this source category have undergone major revisions for the period 1990
through 2021, namely for non-Partners based on the comparison with atmospheric data. Other, relatively smaller
recalculations include an adjustment to OEM SF6 emissions to address GHGRP off-ramping facilities and a
correction to earlier year data for two facilities:
• To determine emissions from OEM facilities that have ceased reporting to the GHGRP as a result of the
off-ramping provision, emissions were estimated by multiplying the average of reported emissions from
the prior three consecutive years by the average growth rate of SF6 emissions for all reporting years.
• Significant incongruities were identified and corrected in the reported data for two historical nameplate
capacities of reporter facilities with one instance in 2011 and the other instance in 2013. In each instance,
corrections were made by calculating the expected nameplate capacity using data reported by the facility
in the prior year.
Updates were also made to reporter emissions where facilities had resubmitted data.
Recalculations of Non-Partner Emissions
As discussed above, results of research conducted by the National Oceanic Atmospheric Administration (Hu et al.
2022) reveal that total U.S. emissions of SF6 were likely significantly higher than previously estimated in the
inventory, particularly for the years before 2012, when reporting of emissions from electric power systems began
4-150 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
under the GHGRP. In addition, the research indicates that U.S. emissions of SF6 trended strongly downward from
2008 to 2009, and the downward trend continued through 2012.
In evaluating possible drivers for the difference and the trend, EPA identified non-Partner utilities as a potentially
significant contributor. As discussed above, non-Partner utilities consist of two groups: (1) utilities that were
required to report to the GHGRP for the first time in 2012 (GHGRP-only reporters) and (2) utilities that have never
been required to the GHGRP because they fall under the reporting threshold (non-reporters). The emission rates of
the GHGRP-only facilities before 2011 are not known, and the emission rates of non-reporters are not known for
any year of the time series. A simple assumption would be that the emission rates of the non-Partners have been
the same as those of the Partners. However, this assumption is uncertain because the Partners and non-Partners
are distinct populations whose emission rates may have varied in magnitude, trend, or both. For example, both the
Partners and the GHGRP-only reporters have reduced their emission rates over time. The extent to which non-
Partners and, for more recent years, non-reporters have also reduced emission rates depends on how much the
observed reductions are due to industry-wide trends (such as improved electrical equipment design and materials
and greater availability of SF6 recycling equipment) versus emission reduction efforts that result directly from
tracking and reporting emissions (such as improved SF6 handling practices and equipment refurbishment or
replacement campaigns). In general, non-reporting facilities would be expected to show reductions related to
industry-wide trends, but not reductions related to tracking and reporting emissions.
EPA has previously revised assumptions regarding the emission rates of non-Partner utilities based on ongoing
review and statistical analysis of data from the Partnership and the GHGRP. In U.S. Inventories submitted in 2012
and earlier years, non-Partners were assumed to have the same emission rate per transmission mile as the
Partners (except certain outliers) had in 1999, when the Partnership began. Because Partners significantly
decreased their emission rates as the Partnership continued, the assumption that non-Partners continued to emit
at the Partners' 1999 rate caused the estimated emission rates for Partners and non-Partners to diverge over time.
In 2012, the submittal of the first set of reports (for 2011) by GHGRP-only utilities provided some insight into the
emission rates of non-Partner utilities. When the emission rates of Partners and GHGRP-only facilities were
compared in 2012, no statistically significant difference was found. Thus, in the U.S. Inventories submitted in 2013
through 2022, EPA assumed that the emission rates per transmission mile of non-reporting utilities (and of GHGRP-
only utilities before 2011) were similar to those of Partners (before 2011) and then of GHGRP reporters (in and
after 2011). Specifically, non-reporter emissions for 2011 and later years were estimated by multiplying non-
reporter transmission miles by regression coefficients derived for reporting facilities for the same year. Non-
reporter and GHGRP-only emissions for 1999 through 2006 were estimated by linearly interpolating between the
1999 regression coefficient (based on 1999 Partner data) and 2006 regression coefficient. Non-reporter and
GHGRP-only emissions for 2007 through 2010 were estimated by linearly interpolating between the 2006
regression coefficient and the 2011 regression coefficient.
The results of the comparison with the atmosphere-derived emissions suggest that, rather than decreasing in
tandem with the emission rates of the Partners from 1999 onward, the emission rates of the non-Partners may
have remained high until 2008, decreasing sharply thereafter. In 2008, EPA began to develop the GHGRP, and the
final rule establishing the GHGRP scope and reporting requirements for electric power systems was published in
2010. Thus, the trend is consistent with the hypothesis that non-Partner utilities, faced with the possibility of being
required to calculate and report their SF6 emissions, began to take action to understand and reduce those
emissions in 2009. Resources for tracking, and to some extent, reducing, emissions were available on EPA's
website for the Partnership and elsewhere. The importance of tracking and reporting emissions to emission
reduction efforts is supported by analysis of the emissions reported by both Partner and GHGRP-only utilities. Both
sets of data show that emissions declined most rapidly during the first three years of reporting (1999-2001 for the
Partners; 2011-2013 for the GHGRP-only utilities). In addition, while there was no statistically significant difference
(at the 95 percent confidence level) between the Partner and GHGRP-only facility emission rates in 2011,
subsequent analysis of the data shows that the emission rates of the GHGRP-only facilities were, on average,
higher than those of the Partners, but that the difference was rapidly narrowed in subsequent years. This is
consistent with Partners having already made cost-effective reductions in earlier years that the GHGRP-only
facilities implemented as they began reporting.
Industrial Processes and Product Use 4-151
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Given these atmospheric findings, the trends in emission reductions upon initial reporting, and because emissions
from non-reporting electric power systems are a significant source of uncertainty in the current U.S. SF6 inventory,
EPA revised the methodology used to estimate non-reporter emissions. To recalculate non-Partner emissions from
1999 through 2010, an updated regression coefficient (emissions as a function of transmission miles) that includes
outliers for 1999 was calculated to estimate non-reporter emissions for 1999. In addition, a new regression
coefficient was calculated for 2011 that includes GHGRP-only Reporters. New emissions rates (SF6 emissions/
Transmission Miles) were calculated for 1999 and 2011. The 1999 emissions rate was held constant to estimate
non-Partner emissions from 2000-2008. Emissions from 2009-2010 were based on the interpolated emission rate
between 2008 (still held at the 1999 emission rate) and the 2011 emission rate from the GHGRP-only reporters, as
discussed above. The interpolated 2010 emission rate was used for estimating non-reporter emissions from 2010-
2021. As a result of the revision to the methodology used to estimate non-reporter emissions in this Inventory,
non-reporter SF6 emissions estimates increased by 94 percent at an average, for years 1999 through 2020, in
comparison to the 1999 through 2020 Inventory emission estimates. Non-reporting facilities were assumed to
have significantly lowered their emissions rates in anticipation of the GHGRP, but not to have made additional
substantial improvements after determining that they were not subject to the rule. Of note, even though the
emissions per transmission mile are being held constant for non-reporters, the implied emission rate in terms of
emissions per nameplate capacity is still decreasing, although at a slower rate than for reporters, as the average
nameplate capacity per transmission mile continues to increase.
As a result of the recalculations, SF6 emissions from electrical transmission and distribution increased by 50
percent for 2020 relative to the previous report. On average, SF6 emission estimates for 1999 through 2020
increased by approximately 23 percent per year.
Revision of Global Warming Potentials (GWPs)
For the current Inventory, calculated C02-equivalent estimates of total SF6 and CF4 emissions from electrical
transmission and distribution have been revised to reflect the 100-year global warming potentials (GWPs) provided
in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). AR5 GWP values differ slightly from those presented in the
IPCC Fourth Assessment Report (AR4) (IPCC 2007) (used in the previous Inventories). The AR5 GWPs have been
applied across the entire time series for consistency. The GWP of SF6 has increased, leading to an overall increase
in emissions from C02-equivalent SF6 emissions. The GWP of CF4 has decreased, leading to a decrease in CO2-
equivalent CF4 emissions. Compared to the previous Inventory which applied 100-year GWP values from AR4, the
average annual change in SF6 C02-equivalent emissions was a 3.1 percent increase and the average annual change
in CF4 C02-equivalent emissions was a 10.3 percent decrease for the time series. Further discussion on this update
and the overall impacts of updating the Inventory GWP values to reflect the IPCC Fifth Assessment Report can be
found in Chapter 9, Recalculations and Improvements.
Planned Improvements
EPA plans to revisit the methodology for determining emissions from the manufacture of electrical equipment, in
particular, the assumption that emissions reported by OEMs account for a conservatively low estimate of 50
percent of the total emissions from all U.S. OEMs. Additional market research will be required to confirm or modify
the assumptions regarding the portion of industry not reporting to the GHGRP program. EPA also plans to review
available data to reflect the emissions from the missing SF6 production facility, and allocate and report those
emissions under the appropriate category (i.e., fluorochemical production category) in future Inventories.
4-152 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
4.26 Nitrous Oxide from Product Uses (CRF
Source Category 2G3)
Nitrous oxide (N2O) is a clear, colorless, oxidizing liquefied gas with a slightly sweet odor which is used in a wide
variety of specialized product uses and applications. The amount of N2O that is actually emitted depends upon the
specific product use or application.
There are a total of three N2O production facilities currently operating in the United States (Ottinger 2021). Nitrous
oxide 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.
Small 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, which is used to inflate airbags;
• Fuel oxidant in auto racing; and
• Oxidizing agent in blowtorches used by jewelers and others (Heydorn 1997).
Production of N2O in 2021 was approximately 15 kt (see Table 4-109).
Table 4-109: N2O Production (kt)
Year
1990
2005
2017
2018
2019
2020
2021
Production (kt)
16
15
15
15
15
15
15
Nitrous oxide emissions were 3.8 MMT CO2 Eq. (14 kt N2O) in 2021 (see Table 4-110). Production of N2O stabilized
during the 1990s because medical markets had found other substitutes for anesthetics, and more medical
procedures were being performed on an outpatient basis using local anesthetics that do not require N2O. The use
of N2O as a propellant for whipped cream has also stabilized due to the increased popularity of cream products
packaged in reusable plastic tubs (Heydorn 1997).
Table 4-110: N2O Emissions from N2O Product Usage (MMT CO2 Eq. and kt)
Year 1990 2005 2017 2018 2019 2020 2021
MMT C02 Eq. 3.8 3.8 3.8 3.8 3.8 3.8 3.8
kt 14 14 14 14 14 14 14
Methodology ana rime-Series Consistency
Emissions from N2O product uses were estimated using the following equation:
Equation 4-26: N2O Emissions from Product Use
Epu = X a X ERa*)
where,
Epu = N2O emissions from product uses, metric tons
P = Total U.S. production of N2O, metric tons
a = specific application
Sa = Share of N2O usage by application a
Industrial Processes and Product Use 4-153
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
ERa = Emission rate for application a, percent
The share of total quantity of N2O usage by end-use represents the share of national N2O produced that is used by
the specific subcategory (e.g., anesthesia, food processing). In 2020, the medical/dental industry used an
estimated 89.5 percent of total N2O produced, followed by food processing propellants at 6.5 percent. All other
subcategories, including semiconductor manufacturing, atomic absorption spectrometry, sodium azide production,
auto racing, and blowtorches, used the remainder of the N2O produced. This subcategory breakdown changed
slightly in the mid-1990s. For instance, the small share of N2O usage in the production of sodium azide declined
significantly during the 1990s. Due to the lack of information on the specific time period of the phase-out in this
market subcategory, most of the N2O usage for sodium azide production is assumed to have ceased after 1996,
with the majority of its small share of the market assigned to the larger medical/dental consumption subcategory
(Heydorn 1997). For 1990 through 1996, N2O usage was allocated across the following subcategories: medical
applications, food processing propellant, and sodium azide production. A usage emissions rate was then applied
for each subcategory to estimate the amount of N2O emitted.
Only the medical/dental and food propellant subcategories were assumed to release emissions into the
atmosphere that are not captured under another source category, and therefore these subcategories were the
only usage subcategories with emission rates. Emissions of N2O from semiconductor manufacturing are described
in Section 4.23 Electronics Industry (CRF Source Category 2E) and reported under CRF Source Category 2H3. For
the medical/dental subcategory, due to the poor solubility of N2O in blood and other tissues, none of the N2O is
assumed to be metabolized during anesthesia and quickly leaves the body in exhaled breath. Therefore, an
emission factor of 100 percent was used for this subcategory (IPCC 2006). For N2O used as a propellant in
pressurized and aerosol food products, none of the N2O is reacted during the process and all of the N2O is emitted
to the atmosphere, resulting in an emission factor of 100 percent for this subcategory (IPCC 2006). For the
remaining subcategories, all of the N2O is consumed or reacted during the process, and therefore the emission rate
was considered to be zero percent (Tupman 2002).
The 1990 through 1992 N2O production data were obtained from SRI Consulting's Nitrous Oxide, North America
(Heydorn 1997). Nitrous oxide production data for 1993 through 1995 were not available. Production data for
1996 was specified as a range in two data sources (Heydorn 1997; Tupman 2002). In particular, for 1996, Heydorn
(1997) estimates N2O production to range between 13.6 and 18.1 thousand metric tons. Tupman (2002) provided a
narrower range (15.9 to 18.1 thousand metric tons) for 1996 that falls within the production bounds described by
Heydorn (1997). Tupman (2002) data are considered more industry-specific and current; therefore, the midpoint
of the narrower production range was used to estimate N2O emissions for years 1993 through 2001 (Tupman
2002). The 2002 and 2003 N2O production data were obtained from the Compressed Gas Association Nitrous
Oxide Fact Sheet and Nitrous Oxide Abuse Hotline (CGA 2002, 2003). These data were also provided as a range. For
example, in 2003, CGA (2003) estimates N2O production to range between 13.6 and 15.9 thousand metric tons.
Due to the lack of publicly available data, production estimates for years 2004 through 2021 were held constant at
the 2003 value.
The 1996 share of the total quantity of N2O used by each subcategory was obtained from SRI Consulting's Nitrous
Oxide, North America (Heydorn 1997). The 1990 through 1995 share of total quantity of N2O used by each
subcategory was kept the same as the 1996 number provided by SRI Consulting. The 1997 through 2001 share of
total quantity of N2O usage by sector was obtained from communication with a N2O industry expert (Tupman
2002). The 2002 and 2003 share of total quantity of N2O usage by sector was obtained from CGA (2002, 2003). Due
to the lack of publicly available data, the share of total quantity of N2O usage data for years 2004 through 2021
was assumed to equal the 2003 value. The emission factor for the food processing propellant industry was
obtained from SRI Consulting's Nitrous Oxide, North America (Heydorn 1997) and confirmed by a N2O industry
expert (Tupman 2002). The emission factor for all other subcategories was obtained from communication with a
N2O industry expert (Tupman 2002). The emission factor for the medical/dental subcategory was obtained from
the 2006 IPCC Guidelines.
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2021.
4-154 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Uncertainty-TO BE UPDATED FOR FINAL INVENTORY REPORT
2 The overall uncertainty associated with the 2021 N2O emission estimate from N2O product usage was calculated
3 using the 2006IPCC Guidelines (2006) Approach 2 methodology. Uncertainty associated with the parameters used
4 to estimate N2O emissions include production data, total market share of each end use, and the emission factors
5 applied to each end use, respectively. The uncertainty associated with N2O production data is ±25 percent, based
6 on expert judgment. The uncertainty associated with the market share for the medical/dental subcategory is ±0.56
7 percent, and uncertainty for the market share of food propellant subcategory is ±25 percent, both based on expert
8 judgment. Uncertainty for emission factors was assumed to be zero, consistent with the 2006 IPCC Guidelines.
9 The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 4-111. Nitrous oxide
10 emissions from N2O product usage were estimated to be between 3.2 and 5.2 MMT CO2 Eq. at the 95 percent
11 confidence level. This indicates a range of approximately 24 percent below to 24 percent above the emission
12 estimate of 3.8 MMT CO2 Eq.
13 Table 4-111: Approach 2 Quantitative Uncertainty Estimates for N2O Emissions from N2O
14 Product Usage (MMT CO2 Eq. and Percent)
Source
Gas
2021 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.)
(MMT C02 Eq.)
(%]
1
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
N20 from Product Uses
N20
3.8
3.2
5.2
-24%
+24%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
15 QA/QC and Verification
16 For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
17 Chapter 6 of the 2006 IPCC Guidelines, see the QA/QC and Verification Procedures section in the introduction of
18 the IPPU chapter.
19 Recalculations Discussion
20 For the current Inventory, CC>2-equivalent estimates of total N2O emissions from N2O product uses have been
21 revised to reflect the 100-year global warming potentials (GWPs) provided in the IPCC Fifth Assessment Report
22 (AR5) (IPCC 2013). AR5 GWP values differ slightly from those presented in the IPCC Fourth Assessment Report
23 (AR4) (IPCC 2007) (used in the previous inventories). The AR5 GWPs have been applied across the entire time
24 series for consistency. The GWP of N2O decreased from 298 to 265, leading to an overall decrease in estimates for
25 calculated CC>2-equivalent N2O emissions. Compared to the previous Inventory, which applied 100-year GWP
26 values from AR4, annual calculated CC>2-equivalent N2O emissions decreased by 11 percent each year, ranging from
27 a decrease of 430 kt CO2 Eq. in 1992 to 519 kt CO2 Eq. for 1997 through 2001. Further discussion on this update
28 and the overall impacts of updating the Inventory GWP values to reflect the IPCC Fifth Assessment Report can be
29 found in Chapter 9, Recalculations and Improvements.
30 Planned Improvements
31 EPA recently initiated an evaluation of alternative production statistics for cross-verification and updating time-
32 series activity data, emission factors, assumptions, etc., and a reassessment of N2O product use subcategories that
33 accurately represent trends. This evaluation includes conducting a literature review of publications and research
34 that may provide additional details on the industry. This work remains ongoing, and thus far no additional sources
35 of data have been found to update this category.
Industrial Processes and Product Use 4-155
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Pending additional resources and planned improvement prioritization, EPA may also evaluate production and use
cycles, and the potential need to incorporate a time lag between production and ultimate product use and
resulting release of N2O. Additionally, planned improvements include considering imports and exports of N2O for
product uses.
Finally, for future Inventories, EPA will examine data from EPA's GHGRP to improve the emission estimates for the
N2O 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 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.
4.27 Industrial Processes and Product Use
Sources of Precursor Gases—TO BE
UPDATED FOR FINAL INVENTORY REPORT
In addition to the main greenhouse gases addressed above, many industrial processes can result in emissions of
various greenhouse gas precursors. The reporting requirements of the UNFCCC116 request that information be
provided on precursor emissions, which include carbon monoxide (CO), nitrogen oxides (NOx), non-methane
volatile organic compounds (NMVOCs), and sulfur dioxide (SO2). These gases are not direct greenhouse gases, but
indirectly impact Earth's radiative balance by altering the concentrations of greenhouse gases (e.g., ozone) and
atmospheric aerosol (e.g., particulate sulfate). Combustion byproducts such as CO and NOx are emitted from
industrial applications that employ thermal incineration as a control technology. NMVOCs, commonly referred to
as "hydrocarbons," are the primary gases emitted from most processes employing organic or petroleum-based
products, and can also result from the product storage and handling.
Accidental releases of precursors associated with product use and handling can constitute major emissions in this
category. In the United States, emissions from product use are primarily the result of solvent evaporation,
whereby the lighter hydrocarbon molecules in the solvents escape into the atmosphere. The major categories of
product uses include: degreasing, graphic arts, surface coating, other industrial uses of solvents (e.g., electronics),
dry cleaning, and non-industrial uses (e.g., uses of paint thinner). Product usage in the United States also results in
the emission of small amounts of hydrofluorocarbons (HFCs) and hydrofluoroethers (HFEs), which are included
under Substitution of Ozone Depleting Substances in this chapter.
Total emissions of NOx, CO, NMVOCs, and SO2 from non-energy industrial processes and product use from 1990 to
2021 are reported in Table 4-112.
Table 4-112: NOx, CO, NMVOC, and SO2 Emissions from Industrial Processes and Product
Use (kt)
Gas/Source 1990 2005 2017
NOx 580 557 387
Mineral Industry 246 329 220
Other Industrial Processes3 93 109 70
Metal Industry 88 60 60
Chemical Industry 152 55 37
Product Usesb 13 1
2018
2019
2020
2021
393
369
369
369
227
214
214
214
72
67
67
67
57
54
54
54
38
34
34
34
+
1
1
1
116 See http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
4-156 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
CO 4,129 1,557 1,007 1,028 978 978 978
Metal Industry 2,395 752 425 443 434 434 434
Other Industrial Processes3 608 420 311 309 280 280 280
Mineral Industry 49 194 163 164 159 159 159
Chemical Industry 1,073 189 107 112 104 104 104
Product Usesb 5 2 1 + 2 2 2
NMVOCs 7,638 5,850 3,767 3,726 3,531 3,531 3,531
Product Usesb 5,216 3,851 2,696 2,627 2,446 2,446 2,446
Other Industrial Processes3 1,720 1,709 959 980 970 970 970
Chemical Industry 575 213 68 71 68 68 68
Mineral Industry 16 32 24 26 26 26 26
Metal Industry 111 45 20 22 21 21 21
S02 1,307 828 508 486 392 392 392
Other Industrial Processes3 129 227 243 232 165 165 165
Chemical Industry 269 228 101 97 87 87 87
Mineral Industry 250 215 87 88 81 81 81
Metal Industry 659 158 77 68 58 58 58
Product Usesb NO + + + + + j_
+ Does not exceed 0.5 kt.
3 Other Industrial Processes includes storage and transport, other industrial processes (manufacturing of
agriculture, food, and kindred products; wood, pulp, paper, and publishing products; rubber and
miscellaneous plastic products; machinery products; construction; transportation equipment; and textiles,
leather, and apparel products), and miscellaneous sources (catastrophic/accidental release, other
combustion (structural fires), health services, repair shops, and fugitive dust). It does not include agricultural
fires or slash/prescribed burning, which are accounted for under the Field Burning of Agricultural Residues
source.
b Product Uses includes the following categories: solvent utilization (degreasing, graphic arts, dry cleaning,
surface coating, other industrial, and nonindustrial).
Note: Totals by gas may not sum due to independent rounding.
Methodology and Time-Series Consistency
Emission estimates for 1990 through 2021 were obtained from data published on the National Emissions Inventory
(NEI) Air Pollutant Emissions Trends Data website (EPA 2022a). For Table 4-112, NEI reported emissions of CO, NOx,
SO2, and NMVOCs and recategorized from NEI Tier 1/Tier 2 source categories to those more closely aligned with
IPCC categories, based on EPA (2022).117 NEI Tier 1 emission categories related to the IPPU sector categories in this
report include: chemical and allied product manufacturing, metals processing, storage and transport, solvent
utilization, other industrial processes, and miscellaneous sources. As described in detail in the NEI Technical
Support Documentation (TSD) (EPA 2021), NEI emissions are estimated through a combination of emissions data
submitted directly to the EPA by state, local, and tribal air agencies, as well as additional information added by the
Agency from EPA emissions programs, such as the emission trading program, Toxics Release Inventory (TRI), and
data collected during rule development or compliance testing.
Methodological approaches were applied to the entire time series to ensure time-series consistency from 1990
through 2021, which are described in detail in the NEI'sTSD and on EPA's Air Pollutant Emission Trends web site
(EPA 2021a; EPA 2021b). Updates to historical activity data are documented in NEI's TSD (EPA 2021). A quantitative
uncertainty analysis was not performed.
117 The NEI estimates and reports emissions from six criteria air pollutants (CAPs) and 187 hazardous air pollutants (HAPs) in
support of National Ambient Air Quality Standards. Reported NEI emission estimates are grouped into 60 sectors and 15 Tier 1
source categories, which broadly cover similar source categories to those presented in this chapter. For this report, EPA has
mapped and regrouped emissions of greenhouse gas precursors (CO, NOx, S02, and NMVOCs) from NEI Tier 1/Tier 2 categories
to better align with IPCC source categories, and to ensure consistency and completeness to the extent possible. See Annex 6.6
for more information on this mapping.
Industrial Processes and Product Use 4-157
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
5. Agriculture
Agricultural activities contribute directly to emissions of greenhouse gases through a variety of processes. This
chapter provides an assessment of methane (Cm) from enteric fermentation, livestock manure management, rice
cultivation and field burning of agricultural residues and nitrous oxide (N2O) emissions from agricultural soil
management, livestock manure management, and field burning of agricultural residues; as well as carbon dioxide
(CO2) emissions from liming and urea fertilization (see Figure 5-1). Additional CO2, CH4 and N2O fluxes from
agriculture-related land-use and land-use conversion activities, such as cultivation of cropland, management on
grasslands, grassland fires, aquaculture, and conversion of forest land to cropland, are presented in the Land Use,
Land-Use Change, and Forestry (LULUCF) chapter. Carbon dioxide emissions from stationary and mobile on-farm
energy use and CH4 and N2O emissions from stationary on-farm energy use are reported in the Energy chapter
under the Industrial sector emissions. Methane and N2O emissions from mobile on-farm energy use are reported
in the Energy chapter under mobile fossil fuel combustion emissions.
Figure 5-1: 2021 Agriculture Sector Greenhouse Gas Emission Sources
Management
Enteric Fermentation
Manure Management
Rice Cultivation
Urea Fertilization
Liming
Field Burning of Agricultural Residues
60 80 100 120 140 160 180 200 220
MMT CO2 Eq.
In 2021, the Agriculture sector was responsible for emissions of 589.3 MMT CO2 Eq.,1 or 9.3 percent of total U.S.
greenhouse gas emissions. Emissions of N2O by agricultural soil management through activities such as fertilizer
application and other agricultural practices that increased nitrogen availability in the soil was the largest source of
U.S. N2O emissions, accounting for 74.1 percent, and the largest source of emissions from the Agriculture sector,
accounting for 48.4 percent of total sector emissions. Methane emissions from enteric fermentation and manure
management represent 26.8 percent and 9.1 percent of total CFU emissions from anthropogenic activities,
1 Following the current reporting requirements under the United Nations Framework Convention on Climate Change (UNFCCC),
this Inventory report presents C02 equivalent values based on the IPCC Fifth Assessment Report (AR5) GWP values. See the
Introduction chapter as well as Chapter 9 for more information.
Agriculture 5-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
respectively, and 33.1 and 14.2 percent of Agriculture sector emissions, respectively. Of all domestic animal types,
beef and dairy cattle were the largest emitters of Cm. Rice cultivation and field burning of agricultural residues
were minor sources of Cm. Manure management and field burning of agricultural residues were also small sources
of N2O emissions. Urea fertilization and liming accounted for 0.1 percent and 0.06 percent of total CChemissions
from anthropogenic activities, respectively.
Table 5-1 and Table 5-2 present emission estimates for the Agriculture sector. Between 1990 and 2021, CChand
Cm emissions from agricultural activities increased by 16.2 percent and 15.7 percent, respectively, while N2O
emissions from agricultural activities fluctuated from year to year but increased by 4.1 percent overall. Trends in
sources of agricultural emissions over the 1990 to 2021 time series are shown in Figure 5-2.
Figure 5-2: Trends in Agriculture Sector Greenhouse Gas Emission Sources
700
650
600
550
500
450
S 400
o
£ 350
s 300
250
200
150
100
50
0
Oi-irNn^-LnvDr^oocnoi-Hrsjro^-LnvDrv.ooo^Oi-irNjfOTrLovDrv.coo^Oi-H
G^cr>cr>cr\cr>oooooooooO'-ii-iT-iT-i'—iT-i'-i'-i'-iT-irMrM
cncncr>cricr>av>cr>cr>cr>crioooooooooooooooooooooo
HHHHHHHHHHfNjfNJNfNJfNlfMNfMfNfVJlNfNJfNfVJlNfMNlNfNJfMfNJOJ
Each year, some emission estimates in the Agriculture sector of the Inventory are recalculated and revised with
improved methods and/or data. In general, recalculations are made to the U.S. greenhouse gas emission estimates
either to incorporate new methodologies or, most commonly, to update recent historical data. These
improvements are implemented consistently across the previous Inventory's time series (i.e., 1990 through 2020)
to ensure that the trend is accurate. This year's notable updates include: Agricultural Soil Management: a)
incorporating the most recently released cropping and land use history data from the National Resources
Inventory (NRI), b) incorporating remote sensing data regarding tillage practices collected through OpTIS, c)
incorporating updated cropland management data from the U.S. Department of Agriculture Conservation Effects
and Assessment Project (USDA-CEAP2) into the DayCent model, d) modifying the statistical imputation method for
the management activity data associated with tillage practices, mineral fertilization, manure amendments, cover
crop management, planting and harvest dates using gradient boosting instead of an artificial neural network, e)
constraining synthetic N fertilization and manure N applications in the Tier 3 method at the state scale rather than
the national scale, and f) re-calibrating the soil C module in the DayCent model using Bayesian method. In total,
the methodological and historic data improvements made to the Agriculture sector in this Inventory increased
greenhouse gas emissions by an average of 0.2 MMT CO2 Eq. (less than 0.1 percent) across the time series. For
Field Burning of Agricultural Residues
Urea Fertilization
I Liming
Rice Cultivation
I Manure Management
I Enteric Fermentation
I Agricultural Soil Management
_ in ">
cxi vo — iS co
LO ^ LO ^ gj B
5-2 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
more information on specific methodological updates, please see the Recalculations discussions within the
respective source category sections of this chapter. In addition, for the current Inventory, CCh-equivalent
emissions totals of Cm and N2O have been revised to reflect the 100-year global warming potentials (GWPs)
provided in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). Further discussion on this update and the overall
impacts of updating the Inventory GWP values to reflect the IPCC Fifth Assessment Report can be found in Chapter
9, Recalculations and Improvements.
Emissions reported in the Agriculture chapter include those from all states; however, for Hawaii and Alaska some
agricultural practices that can increase nitrogen availability in the soil, and thus cause N2O emissions, are not
included (see chapter sections on "Uncertainty and Time-Series Consistency" and "Planned Improvements" for
more details). Emissions from the Agriculture sector occurring in U.S. Territories and the District of Columbia are
not estimated due to incomplete data, with the exception of urea fertilization in Puerto Rico. EPA continues to
identify and review available data on an ongoing basis to include agriculture emissions from U.S. Territories, to the
extent they are occurring, in future Inventories. Other minor outlying U.S. Territories in the Pacific Islands have no
permanent populations (e.g., Baker Island) and therefore EPA assumes no agricultural activities are occurring. See
Annex 5 for more information on EPA's assessment of the sources not included in this Inventory.
Table 5-1: Emissions from Agriculture (MMT CO2 Eq.)
Gas/Source
1990
2005
2017
2018
2019
2020
2021
CO?
7.1
7.9
7.9
7.2
7.2
8.0
8.3
Urea Fertilization
2.4
3.5
4.9
4.9
5.0
5.1
5.2
Liming
4.7
4.4
3.1
2.2
2.2
2.9
3.0
ch4
240.4
263.7
277.5
281.2
280.4
281.0
278.2
Enteric Fermentation
183.1
188.2
195.9
196.8
197.3
196.2
194.9
Manure Management
39.0
54.9
64.4
66.5
65.7
66.7
66.0
Rice Cultivation
17.9
20.2
16.7
17.4
16.9
17.6
16.8
Field Burning of Agricultural Residues
0.4
0.5
0.5
0.5
0.5
0.5
0.5
n2o
290.9
295.4
315.7
329.4
315.7
297.0
302.8
Agricultural Soil Management
278.4
280.8
298.7
312.1
298.2
279.3
285.2
Manure Management
12.4
14.5
16.9
17.2
17.4
17.5
17.4
Field Burning of Agricultural Residues
0.1
0.2
0.2
0.2
0.2
0.2
0.2
Total
538.4
567.0
601.2
617.8
603.3
586.0
589.3
Note: Totals may not sum due to independent rounding.
ible 5-2: Emissions from Agriculture (kt)
Gas/Source
1990
2005
2017
2018
2019
2020
2021
CO?
7,106
7,856
7,931
7,178
7,234
8,037
8,260
Urea Fertilization
2,417
3,504
4,862
4,939
5,030
5,122
5,214
Liming
4,690
4,351
3,069
2,240
2,203
2,915
3,047
ch4
8,587
9,419
9,911
10,043
10,013
10,036
9,937
Enteric Fermentation
6,539
6,722
6,998
7,028
7,046
7,007
6,962
Manure Management
1,394
1,960
2,300
2,375
2,348
2,383
2,358
Rice Cultivation
640
720
596
623
602
630
600
Field Burning of Agricultural Residues
15
17
17
17
17
17
17
N20
1,098
1,115
1,191
1,243
1,191
1,121
1,143
Agricultural Soil Management
1,050
1,060
1,127
1,178
1,125
1,054
1,076
Manure Management
47
55
64
65
65
66
66
Field Burning of Agricultural Residues
1
1
1
1
1
1
1
Note: Totals by gas may not sum due to independent rounding.
Agriculture 5-3
-------
1
In following the United Nations Framework Convention on Climate Change (UNFCCC) requirement under Article
4.1 to develop and submit national greenhouse gas emission inventories, the emissions and removals presented
in this report and this chapter are organized by source and sink categories and calculated using internationally-
accepted methods provided by the Intergovernmental Panel on Climate Change (IPCC) in the 2006IPCC
Guidelines for National Greenhouse Gas Inventories (2006 IPCC Guidelines). Additionally, the calculated
emissions and removals in a given year for the United States are presented in a common format in line with the
UNFCCC reporting guidelines for the reporting of inventories under this international agreement. The use of
consistent methods to calculate emissions and removals by all nations providing their inventories to the
UNFCCC ensures that these reports are comparable. The presentation of emissions provided in the Agriculture
chapter do not preclude alternative examinations, but rather, this chapter presents emissions in a common
format consistent with how countries are to report Inventories under the UNFCCC. The report itself, and this
chapter, follow this standardized format and provide an explanation of the application of methods used to
calculate emissions from agricultural activities.
5 Methane is produced as part of normal digestive processes in animals. During digestion, microbes resident in an
6 animal's digestive system ferment food consumed by the animal. This microbial fermentation process, referred to
7 as enteric fermentation, produces Cm as a byproduct, which can be exhaled or eructated by the animal. The
8 amount of Cm produced and emitted by an individual animal depends primarily upon the animal's digestive
9 system, and the amount and type of feed it consumes.2
10 Ruminant animals (e.g., cattle, buffalo, sheep, goats, and camels) are the major emitters of Cm because of their
11 unique digestive system. Ruminants possess a rumen, or large "fore-stomach," in which microbial fermentation
12 breaks down the feed they consume into products that can be absorbed and metabolized. The microbial
13 fermentation that occurs in the rumen enables them to digest coarse plant material that non-ruminant animals
14 cannot. Ruminant animals, consequently, have the highest Cm emissions per unit of body mass among all animal
15 types.
16 Non-ruminant animals (e.g., swine, horses, and mules and asses) also produce Cm emissions through enteric
17 fermentation, although this microbial fermentation occurs in the large intestine. These non-ruminants emit
18 significantly less CFU on a per-animal-mass basis than ruminants because the capacity of the large intestine to
19 produce CFU is lower.
20 In addition to the type of digestive system, an animal's feed quality and feed intake also affect Cm emissions. In
21 general, lower feed quality and/or higher feed intake leads to higher CFU emissions. Feed intake is positively
22 correlated to animal size, growth rate, level of activity and production (e.g., milk production, wool growth,
23 pregnancy, or work). Therefore, feed intake varies among animal types as well as among different management
24 practices for individual animal types (e.g., animals in feedlots or grazing on pasture).
2 C02 emissions from livestock are not estimated because annual net C02 emissions are assumed to be zero - the C02
photosynthesized by plants is returned to the atmosphere as respired C02 (IPCC 2006).
5-4 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
2
4
3
5.1 Enteric Fermentation (CRF Source
Category 3A)
-------
1 Methane emission estimates from enteric fermentation are provided in Table 5-3 and Table 5-4. Total livestock CFU
2 emissions in 2021 were 194.9 MMT CO2 Eq. (6,962 kt). Beef cattle remain the largest contributor of CH4 emissions
3 from enteric fermentation, accounting for 71 percent in 2021. Emissions from dairy cattle in 2021 accounted for 25
4 percent, and the remaining emissions were from swine, horses, sheep, goats, American bison, mules and asses.3
5 Table 5-3: ChU Emissions from Enteric Fermentation (MMT CO2 Eq.)
Livestock Type
1990
2005
2017
2018
2019
2020
2021
Beef Cattle
132.8
139.6
140.9
141.2
141.7
140.4
139.1
Dairy Cattle
43.3
41.3
48.0
48.6
48.5
48.8
49.1
Swine
2.3
2.6
3.0
3.1
3.2
3.2
3.1
Horses
1.1
2.0
1.4
1.4
1.3
1.2
1.1
Sheep
2.9
1.5
1.3
1.3
1.3
1.3
1.3
Goats
0.6
0.7
0.7
0.7
0.7
0.7
0.7
American Bison
0.1
0.5
0.4
0.4
0.4
0.5
0.5
Mules and Asses
+
0.1
0.1
0.1
0.1
0.1
0.1
Total
183.1
188.2
195.9
196.8
197.3
196.2
194.9
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
6 Table 5-4: ChU Emissions from Enteric Fermentation (kt)
Livestock Type
1990
2005
2017
2018
2019
2020
2021
Beef Cattle
4,742
4,986
5,033
5,042
5,062
5,013
4,967
Dairy Cattle
1,547
1,473
1,715
1,737
1,732
1,744
1,754
Swine
81
92
108
110
115
116
111
Horses
40
70
51
48
46
43
40
Sheep
102
55
47
47
47
47
47
Goats
23
26
24
24
25
25
23
American Bison
4
17
15
15
16
16
17
Mules and Asses
1
2
3
3
3
3
3
Total 6,539 6,722 6,998 7,028 7,046 7,007 6,962
Note: Totals may not sum due to independent rounding.
7 From 1990 to 2021, emissions from enteric fermentation have increased by 6.5 percent. From 2020 to 2021,
8 emissions decreased by 0.6 percent, largely driven by a decrease in beef cattle populations. While emissions
9 generally follow trends in cattle populations, over the long term there are exceptions. For example, while dairy
10 cattle emissions increased 13.4 percent over the entire time series, the population has declined by 3.5 percent,
11 and milk production increased 62 percent (USDA 2021a, USDA 2022). These trends indicate that while emissions
12 per head are increasing, emissions per unit of product (i.e., meat, milk) are decreasing.
13 Generally, from 1990 to 1995 emissions from beef cattle increased and then decreased from 1996 to 2004. These
14 trends were mainly due to fluctuations in beef cattle populations and increased digestibility of feed for feedlot
15 cattle. Beef cattle emissions generally increased from 2004 to 2007, as beef cattle populations increased, and an
3 Enteric fermentation emissions from poultry are not estimated because no IPCC method has been developed for determining
enteric fermentation CH4 emissions from poultry; at this time, developing a country-specific method would require a
disproportionate amount of resources given the small magnitude of this source category. Enteric fermentation emissions from
camels are not estimated because there is no significant population of camels in the United States. Given the insignificance of
estimated camel emissions in terms of the overall level and trend in national emissions, there are no immediate improvement
plans to include this emissions category in the Inventory. See Annex 5 for more information on significance of estimated camel
emissions.
Agriculture 5-5
-------
1 extensive literature review indicated a trend toward a decrease in feed digestibility for those years. Beef cattle
2 emissions decreased again from 2007 to 2014, as populations again decreased, but increased from 2015 to 2019,
3 consistent with another increase in population over those same years. Emissions and populations slightly declined
4 from 2019 to 2021.
5 Emissions from dairy cattle generally trended downward from 1990 to 2004, along with an overall dairy cattle
6 population decline during the same period. Similar to beef cattle, dairy cattle emissions rose from 2004 to 2007
7 due to population increases and a decrease in feed digestibility (based on an analysis of more than 350 dairy cow
8 diets used by producers across the United States). Dairy cattle emissions continued to trend upward from 2007 to
9 2019, generally in line with dairy cattle population changes.
10 Regarding trends in other animals, populations of sheep have steadily declined, with an overall decrease of 54
11 percent since 1990. Horse populations are 1 percent greater than they were in 1990, but their numbers have been
12 declining by an average of 4 percent annually since 2007. Goat populations increased by about 20 percent through
13 2007 followed by a steady decrease through 2012. After a steady increase of 1 percent annually through 2020,
14 goat populations dropped by 5 percent in 2021. Swine populations have trended upward through most of the time
15 series, increasing 43 percent from 1990 to 2020. However, swine populations decreased by around 4 percent from
16 2020 to 2021. The population of American bison more than quadrupled over the 1990 to 2020 time period, while
17 the population of mules and asses increased by a factor of five.
19 Livestock enteric fermentation emission estimate methodologies fall into two categories: cattle and other
20 domesticated animals. Cattle, due to their large population, large size, and particular digestive characteristics,
21 account for the majority of enteric fermentation CH4 emissions from livestock in the United States. A more detailed
22 methodology (i.e., IPCC Tier 2) was therefore applied to estimate emissions for all cattle. Emission estimates for
23 other domesticated animals (horses, sheep, swine, goats, American bison, and mules and asses) were estimated
24 using the IPCC Tier 1 approach, as suggested by the 2006 IPCC Guidelines (see the Planned Improvements section).
25 While the large diversity of animal management practices cannot be precisely characterized and evaluated,
26 significant scientific literature exists that provides the necessary data to estimate cattle emissions using the IPCC
27 Tier 2 approach. The Cattle Enteric Fermentation Model (CEFM), developed by EPA and used to estimate cattle CH4
28 emissions from enteric fermentation using IPCC's Tier 2 method, incorporates this information and other analyses
29 of livestock population, feeding practices, and production characteristics. For the current Inventory, CEFM results
30 for 1990 through 2020 were carried over from the 1990 to 2020 Inventory (i.e., 2022 Inventory submission) to
31 focus resources on CEFM improvements, and a simplified approach was used to estimate 2021 enteric emissions
32 from cattle.
33 See Annex 3.10 for more detailed information on the methodology and data used to calculate CH4 emissions from
34 enteric fermentation. In addition, variables and the resulting emissions are also available at the state level in Annex
18
Methodology and Time-Series Consistency
35 3.10.
36 1990-2020 Inventory Methodology for Cattle
37 National cattle population statistics were disaggregated into the following cattle sub-populations:
38
• Dairy Cattle
39
40
41
o Calves
o Heifer Replacements
o Cows
42
• Beef Cattle
43
o Calves
5-6 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
o Heifer Replacements
o Heifer and Steer Stockers
o Animals in Feedlots (Heifers and Steer)
o Cows
o Bulls
Calf birth rates, end-of-year population statistics, detailed feedlot placement information, and slaughter weight
data were used to create a transition matrix that models cohorts of individual animal types and their specific
emission profiles. The key variables tracked for each of the cattle population categories are described in Annex
3.10. These variables include performance factors such as pregnancy and lactation as well as average weights and
weight gain. Annual cattle population data were obtained from the U.S. Department of Agriculture's (USDA)
National Agricultural Statistics Service (NASS) QuickStats database (USDA 2021a).
Diet characteristics were estimated by region for dairy, grazing beef, and feedlot beef cattle. These diet
characteristics were used to calculate digestible energy (DE) values (expressed as the percent of gross energy
intake digested by the animal) and CH4 conversion rates (Ym) (expressed as the fraction of gross energy converted
to CH4) for each regional population category. The IPCC recommends Ym ranges of 3.0±1.0 percent for feedlot
cattle and 6.5±1.0 percent for other well-fed cattle consuming temperate-climate feed types (IPCC 2006). Given
the availability of detailed diet information for different regions and animal types in the United States, DE and Ym
values unique to the United States were developed. The diet characterizations and estimation of DE and Ym values
were based on information from state agricultural extension specialists, a review of published forage quality
studies and scientific literature, expert opinion, and modeling of animal physiology.
The diet characteristics for dairy cattle were based on Donovan (1999) and an extensive review of nearly 20 years
of literature from 1990 through 2009. Estimates of DE were national averages based on the feed components of
the diets observed in the literature for the following year groupings: 1990 through 1993,1994 through 1998,1999
through 2003, 2004 through 2006, 2007, and 2008 onward.4 Base year Ym values by region were estimated using
Donovan (1999). As described in ERG (2016), a ruminant digestion model (COWPOLL, as selected in Kebreab et al.
2008) was used to evaluate Ym for each diet evaluated from the literature, and a function was developed to adjust
regional values over time based on the national trend. Dairy replacement heifer diet assumptions were based on
the observed relationship in the literature between dairy cow and dairy heifer diet characteristics.
For feedlot animals, the DE and Ym values used for 1990 were recommended by Johnson (1999). Values for DE and
Ym for 1991 through 1999 were linearly extrapolated based on the 1990 and 2000 data. DE and Ym values for 2000
onwards were based on survey data in Galyean and Gleghorn (2001) and Vasconcelos and Galyean (2007).
For grazing beef cattle, Ym values were based on Johnson (2002), DE values for 1990 through 2006 were based on
specific diet components estimated from Donovan (1999), and DE values from 2007 onwards were developed from
an analysis by Archibeque (2011), based on diet information in Preston (2010) and USDA-APHIS:VS (2010). Weight
and weight gains for cattle were estimated from Holstein (2010), Doren et al. (1989), Enns (2008), Lippke et al.
(2000), Pinchack et al. (2004), Platter et al. (2003), Skogerboe et al. (2000), and expert opinion. See Annex 3.10 for
more details on the method used to characterize cattle diets and weights in the United States.
Calves younger than 4 months are not included in emission estimates because calves consume mainly milk and the
IPCC recommends the use of a Ym of zero for all juveniles consuming only milk. Diets for calves aged 4 to 6 months
are assumed to go through a gradual weaning from milk decreasing to 75 percent at 4 months, 50 percent at age 5
months, and 25 percent at age 6 months. The portion of the diet made up with milk still results in zero emissions.
For the remainder of the diet, beef calf DE and Ym are set equivalent to those of beef replacement heifers, while
dairy calf DE is set equal to that of dairy replacement heifers and dairy calf Ym is provided at 4 and 7 months of age
by Soliva (2006). Estimates of Ym for 5- and 6-month-old dairy calves are linearly interpolated from the values
provided for 4 and 7 months.
4 Due to inconsistencies in the 2003 literature values, the 2002 values were used for 2003 as well.
Agriculture 5-7
-------
1 To estimate CFU emissions, the population was divided into state, age, sub-type (i.e., dairy cows and replacements,
2 beef cows and replacements, heifer and steer stockers, heifers and steers in feedlots, bulls, beef calves 4 to 6
3 months, and dairy calves 4 to 6 months), and production (i.e., pregnant, lactating) groupings to more fully capture
4 differences in Cm emissions from these animal types. The transition matrix was used to simulate the age and
5 weight structure of each sub-type on a monthly basis in order to more accurately reflect the fluctuations that
6 occur throughout the year. Cattle diet characteristics were then used in conjunction with Tier 2 equations from
7 IPCC (2006) to produce Cm emission factors for the following cattle types: dairy cows, beef cows, dairy
8 replacements, beef replacements, steer stockers, heifer stockers, steer feedlot animals, heifer feedlot animals,
9 bulls, and calves. To estimate emissions from cattle, monthly population data from the transition matrix were
10 multiplied by the calculated emission factor for each cattle type. More details are provided in Annex 3.10.
11 2021 Inventory Methodology for Cattle
12 As noted above, a simplified approach for cattle enteric emissions was used in lieu of the CEFM for 2021 to focus
13 resources on CEFM improvements. First, 2021 populations for each of the CEFM cattle subpopulations were
14 estimated, then these populations were multiplied by the corresponding 2020 implied emission factors developed
15 from the CEFM for the 1990 to 2020 Inventory. Dairy cow, beef cow, and bull populations for 2021 were based on
16 data directly from the USDA-NASS QuickStats database (USDA 2021a, USDA 2022). Because the remaining CEFM
17 cattle sub-population categories do not correspond exactly to the remaining QuickStats cattle categories, 2021
18 populations for these categories were estimated by extrapolating the 2020 populations based on percent changes
19 from 2020 to 2021 in similar QuickStats categories, consistent with Volume 1, Chapter 5 of the 2006 IPCC
20 Guidelines on time-series consistency. Table 5-5 lists the QuickStats categories used to estimate the percent
21 change in population for each of the CEFM categories.
22 Table 5-5: Cattle Sub-Population Categories for 2021 Population Estimates
23 Non-Cattle Livestock
24 Emission estimates for other animal types were based on average emission factors (Tier 1 default IPCC emission
25 factors) representative of entire populations of each animal type. Methane emissions from these animals
26 accounted for a minor portion of total CFU emissions from livestock in the United States from 1990 through 2021.
27 Additionally, the variability in emission factors for each of these other animal types (e.g., variability by age,
28 production system, and feeding practice within each animal type) is less than that for cattle.
29 Annual livestock population data for 1990 to 2021 for sheep; swine; goats; horses; mules and asses; and American
30 bison were obtained for available years from USDA-NASS (USDA 2022; USDA 2019). Horse, goat, and mule and ass
31 population data were available for 1987,1992,1997, 2002, 2007, 2012, and 2017 (USDA 2019); the remaining
CEFM Cattle Category
USDA-NASS QuickStats Cattle Category
Dairy Calves
Dairy Cows
Dairy Replacements 7-11 months
Dairy Replacements 12-23 months
Bulls
Beef Calves
Beef Cows
Beef Replacements 7-11 months
Beef Replacements 12-23 months
Steer Stockers
Heifer Stockers
Steer Feedlot
Heifer Feedlot
Cattle, Calves
Cattle, Cows, Milk
Cattle, Heifers, GE 500 lbs, Milk Replacement
Cattle, Heifers, GE 500 lbs, Milk Replacement
Cattle, Bulls, GE 500 lbs
Cattle, Calves
Cattle, Cows, Beef
Cattle, Heifers, GE 500 lbs, Beef Replacement
Cattle, Heifers, GE 500 lbs, Beef Replacement
Cattle, Steers, GE 500 lbs
Cattle, Heifers, GE 500 lbs, (Excl. Replacement)
Cattle, On Feed
Cattle, On Feed
5-8 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
years between 1990 and 2021 were interpolated and extrapolated from the available estimates (with the
exception of goat populations being held constant between 1990 and 1992). American bison population estimates
were available from USDA for 2002, 2007, 2012, and 2017 (USDA 2019) and from the National Bison Association
(1999) for 1990 through 1999. Additional years were based on observed trends from the National Bison
Association (1999), interpolation between known data points, and extrapolation beyond 2012, as described in
more detail in Annex 3.10.
Methane emissions from sheep, goats, swine, horses, American bison, and mules and asses were estimated by
using emission factors utilized in Crutzen et al. (1986, cited in IPCC 2006; IPCC 2019). These emission factors are
representative of typical animal sizes, feed intakes, and feed characteristics in developed countries. For American
bison, the emission factor for buffalo was used and adjusted based on the ratio of live weights to the 0.75 power.
The methodology is the same as that recommended by IPCC (2006).
Uncertainty
A quantitative uncertainty analysis for this source category was performed using the IPCC-recommended Approach
2 uncertainty estimation methodology based on a Monte Carlo Stochastic Simulation technique as described in ICF
(2003). These uncertainty estimates were developed for the 1990 through 2001 Inventory (i.e., 2003 submission to
the UNFCCC). While there are plans to update the uncertainty to reflect recent methodological updates and
forthcoming changes (see Planned Improvements, below), at this time the uncertainty estimates were directly
applied to the 2021 emission estimates in this Inventory.
A total of 185 primary input variables (177 for cattle and 8 for non-cattle) were identified as key input variables for
the uncertainty analysis. A normal distribution was assumed for almost all activity- and emission factor-related
input variables. Triangular distributions were assigned to three input variables (specifically, cow-birth ratios for the
three most recent years included in the 2001 model run) to ensure only positive values would be simulated. For
some key input variables, the uncertainty ranges around their estimates (used for inventory estimation) were
collected from published documents and other public sources; others were based on expert opinion and best
estimates. In addition, both endogenous and exogenous correlations between selected primary input variables
were modeled. The exogenous correlation coefficients between the probability distributions of selected activity-
related variables were developed through expert judgment.
Among the individual cattle sub-source categories, beef cattle account for the largest amount of Cm emissions, as
well as the largest degree of uncertainty in the emission estimates—due mainly to the difficulty in estimating the
diet characteristics for grazing members of this animal group. Among non-cattle, horses represent the largest
percent of uncertainty in the previous uncertainty analysis because the Food and Agricultural Organization (FAO)
of the United Nations population estimates used for horses at that time had a higher degree of uncertainty than
for the USDA population estimates used for swine, goats, and sheep. The horse populations are drawn from the
same USDA source as the other animal types, and therefore the uncertainty range around horses is likely
overestimated. Cattle calves, American bison, mules and asses were excluded from the initial uncertainty estimate
because they were not included in emission estimates at that time.
The uncertainty ranges associated with the activity data-related input variables were plus or minus 10 percent or
lower. However, for many emission factor-related input variables, the lower- and/or the upper-bound uncertainty
estimates were over 20 percent. The results of the quantitative uncertainty analysis are summarized in Table 5-6.
Based on this analysis, enteric fermentation Cm emissions in 2021 were estimated to be between 173.5 and 230.0
MMT CO2 Eq. at a 95 percent confidence level, which indicates a range of 11 percent below to 18 percent above
the 2021 emission estimate of 194.9 MMT CO2 Eq.
Agriculture 5-9
-------
1 Table 5-6: Approach 2 Quantitative Uncertainty Estimates for ChU Emissions from Enteric
2 Fermentation (MMT CO2 Eq. and Percent)
2021 Emission
Source
Gas
Estimate
Uncertainty Range Relative to Emission Estimate3'b'c
(MMT CO? Eq.)
(MMT CO? Eq.)
(%)
Lower Upper
Lower Upper
Bound Bound
Bound Bound
Enteric Fermentation
ch4
194.9
173.5 230.0
-11% +18%
a Range of emissions estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence
interval.
b Note that the relative uncertainty range was estimated with respect to the 2001 emission estimates from the
2003 submission and applied to the 2021 estimates.
c The overall uncertainty calculated in 2003, and applied to the 2021 emission estimate, did not include uncertainty
estimates for calves, American bison, and mules and asses. Additionally, for bulls the emissions estimate was
based on the Tier 1 methodology. Since bull emissions are now estimated using the Tier 2 method, the uncertainty
surrounding their estimates is likely lower than indicated by the previous uncertainty analysis.
3 QA/QC and Verification
4 In order to ensure the quality of the emission estimates from enteric fermentation, the General (IPCC Tier 1) and
5 category-specific (Tier 2) Quality Assurance/Quality Control (QA/QC) procedures were implemented consistent
6 with the U.S. Inventory QA/QC plan outlined in Annex 8. Category-specific or Tier 2 QA procedures included
7 independent review of emission estimate methodologies from previous inventories.
8 As part of the quality assurance process, average implied emissions factors for U.S. dairy and beef cattle were
9 developed based on CEFM output and compared to emission factors for other countries provided by IPCC (2006).
10 This comparison is discussed in further detail in Annex 3.10.
11 Over the past few years, particular importance has been placed on harmonizing the data exchange between the
12 enteric fermentation and manure management source categories. The current Inventory now utilizes the transition
13 matrix from the CEFM for estimating cattle populations and weights for both source categories, and the CEFM is
14 used to output volatile solids and nitrogen excretion estimates using the diet assumptions in the model in
15 conjunction with the energy balance equations from the IPCC (2006). This approach facilitates the QA/QC process
16 for both of these source categories. As noted in the Methodology discussion above, a simplified approach for cattle
17 enteric emissions was used in lieu of the CEFM for 2021.
is Recalculations Discussion
19 EPA updated the global warming potential (GWP) for calculating C02-equivalent emissions of CFU (from 25 to 28)
20 to reflect the 100-year GWPs provided in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). The previous
21 Inventory used 100-year GWPs provided in the IPCC Fourth Assessment Report (AR4). The AR5 GWPs have been
22 applied across the entire time series for consistency. This update resulted in an average annual increase of 12
23 percent for C02-equivalent Cm emissions for the time series from 1990 to 2020 compared to the previous
24 Inventory. Further discussion on this update and the overall impacts of updating the Inventory GWP values to
25 reflect the AR5 can be found in Chapter 9, Recalculations and Improvements.
26 Planned Improvements
27 Regular annual data reviews and updates are necessary to maintain an emissions inventory that reflects the
28 current base of knowledge. In addition to the documented approaches currently used to address data availability,
5-10 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 EPA conducts the following annual assessments to identify and determine the applicability of newer data when
2 updating the estimates to extend time series each year:
3 • Further research to improve the estimation of dry matter intake (as gross energy intake) using data from
4 appropriate production systems;
5 • Updating input variables that are from older data sources, such as beef births by month, beef and dairy
6 annual calving rates, and beef cow lactation rates;
7 • Investigating the availability of data for dairy births by month, to replace the current assumption that
8 births are evenly distributed throughout the year;
9 • Investigating the availability of annual data for the DE, Ym, and crude protein values of specific diet and
10 feed components for grazing and feedlot animals;
11 • Further investigation on additional sources or methodologies for estimating DE for dairy cattle, given the
12 many challenges in characterizing dairy cattle diets;
13 • Further evaluation of the assumptions about weights and weight gains for beef cows, such that trends
14 beyond 2007 are updated, rather than held constant; and
15 • Further evaluation of the estimated weight for dairy cows (i.e., 1,500 lbs) that is based solely on Holstein
16 cows as mature dairy cow weight is likely slightly overestimated, based on knowledge of the breeds of
17 dairy cows in the United States.
18 Depending upon the outcome of ongoing investigations, future improvement efforts for enteric fermentation
19 could include some of the following options which are additional to the regular updates, and may or may not have
20 implications for regular updates once addressed:
21 • Potentially updating to a Tier 2 methodology for other animal types (i.e., sheep, swine, goats, horses);
22 efforts to move to Tier 2 will consider the emissions significance of livestock types;
23 • Investigation of methodologies and emission factors for including enteric fermentation emission
24 estimates from poultry;
25 • Comparison of the current CEFM with other models that estimate enteric fermentation emissions for
26 quality assurance and verification;
27 • Investigation of recent research implications suggesting that certain parameters in enteric models may be
28 simplified without significantly diminishing model accuracy; and
29 • Recent changes that have been implemented to the CEFM warrant an assessment of the current
30 uncertainty analysis; therefore, a revision of the quantitative uncertainty surrounding emission estimates
31 from this source category will be initiated. EPA plans to perform this uncertainty analysis following the
32 completed updates to the CEFM.
33 EPA is continuously investigating these recommendations and potential improvements and working with USDA and
34 other experts to utilize the best available data and methods for estimating emissions. Many of these
35 improvements are major updates and may take multiple years to implement in full.
Agriculture 5-11
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
5.2 Manure Management (CRF Source
Category 3B)
The treatment, storage, and transportation of livestock manure can produce anthropogenic Cm and N2O
emissions.5 Methane is produced by the anaerobic decomposition of manure and nitrous oxide is produced from
direct and indirect pathways through the processes of nitrification and denitrification; in addition, there are many
underlying factors that can affect these resulting emissions from manure management, as described below.
When livestock manure is stored or treated in systems that promote anaerobic conditions (e.g., as a liquid/slurry in
lagoons, ponds, tanks, or pits), the decomposition of the volatile solids component in the manure tends to produce
Cm. When manure is handled as a solid (e.g., in stacks or drylots) or deposited on pasture, range, or paddock
lands, it tends to decompose aerobically and produce CO2 and little or no CH4. Ambient temperature, moisture,
and manure storage or residency time affect the amount of CH4 produced because they influence the growth of
the bacteria responsible for CH4 formation. For non-liquid-based manure systems, moist conditions (which are a
function of rainfall and humidity) can promote CH4 production. Manure composition, which varies by animal diet,
growth rate, and animal type (particularly the different animal digestive systems), also affects the amount of CH4
produced. In general, the greater the energy content of the feed, the greater the potential for CH4 emissions.
However, some higher-energy feeds also are more digestible than lower quality forages, which can result in less
overall waste excreted from the animal.
As previously stated, N2O emissions are produced through both direct and indirect pathways. Direct N2O emissions
are produced as part of the nitrogen (N) cycle through the nitrification and denitrification of the N in livestock dung
and urine.6 There are two pathways for indirect N2O emissions. The first is the result of the volatilization of N in
manure (as NH3 and NOx) and the subsequent deposition of these gases and their products (NhV and NO3") onto
soils and the surface of lakes and other waters. The second pathway is the runoff and leaching of N from manure
into the groundwater below, into riparian zones receiving drain or runoff water, or into the ditches, streams,
rivers, and estuaries into which the land drainage water eventually flows.
The production of direct N2O emissions from livestock manure depends on 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. For direct N2O emissions to occur, the manure must first be handled aerobically
where organic N is mineralized or decomposed to NH4 which is then nitrified to NO3 (producing some N2O as a
byproduct) (nitrification). Next, the manure must be handled anaerobically where the nitrate is then denitrified to
N2O and N2 (denitrification). NOx can also be produced during denitrification (Groffman et al. 2000; Robertson and
Groffman 2015). These emissions are most likely to occur in dry manure handling systems that have aerobic
conditions, but that also contain pockets of anaerobic conditions due to saturation. Avery small portion of the
total N excreted is expected to convert to N2O in the waste management system (WMS).
Indirect N2O emissions are produced when nitrogen is lost from the system through volatilization (as NH3 or NOx)
or through runoff and leaching. The vast majority of volatilization losses from these operations are NH3. Although
there are also some small losses of NOx, there are no quantified estimates available for use, so losses due to
volatilization are only based on NH3 loss factors. Runoff losses would be expected from operations that house
animals or store manure in a manner that is exposed to weather. Runoff losses are also specific to the type of
5 C02 emissions from livestock are not estimated because annual net C02 emissions are assumed to be zero - the C02
photosynthesized by plants is returned to the atmosphere as respired C02 (IPCC 2006).
6 Direct and indirect N20 emissions from dung and urine spread onto fields either directly as daily spread or after it is removed
from manure management systems (i.e., lagoon, pit, etc.) and from livestock dung and urine deposited on pasture, range, or
paddock lands are accounted for and discussed in the Agricultural Soil Management source category within the Agriculture
sector.
5-12 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
animal housed on the operation due to differences in manure characteristics. Little information is known about
leaching from manure management systems as most research focuses on leaching from land application systems.
However, storage systems are often designed to minimize leaching (e.g., clay soil or synthetic liners in lagoons).
Since leaching losses are expected to be minimal, leaching losses are coupled with runoff losses and the
runoff/leaching estimate provided in this chapter does not account for any leaching losses.
Estimates of Cm emissions from manure management in 2021 were 66.0 MMT CO2 Eq. (2,358 kt); in 1990,
emissions were 39.0 MMT CO2 Eq. (1,394 kt). This represents a 69 percent increase in emissions from 1990.
Emissions increased on average by 0.8 MMT CO2 Eq. (2 percent) annually over this period. The majority of this
increase is due to swine and dairy cow manure, where emissions increased 38 and 124 percent, respectively. From
2020 to 2021, there was a 1 percent decrease in total CH4 emissions from manure management, mainly due to a
decrease in swine and poultry populations.
Although a large quantity of managed manure in the United States is handled as a solid, producing little CH4, the
general trend in manure management, particularly for dairy cattle and swine (which are both shifting towards
larger facilities), is one of increasing use of liquid systems. Also, new regulations controlling the application of
manure nutrients to land have shifted manure management practices at smaller dairies from daily spread systems
to storage and management of the manure on site. In many cases, manure management systems with the most
substantial methane emissions are those associated with confined animal management operations where manure
is handled in liquid-based systems. Nitrous oxide emissions from manure management vary significantly between
the types of management system used and can also result in indirect emissions due to other forms of nitrogen loss
from the system (IPCC 2006).
While national dairy animal populations have decreased since 1990, some states have seen increases in their dairy
cattle populations as the industry becomes more concentrated in certain areas of the country and the number of
animals contained on each facility increases. These areas of concentration, such as California, New Mexico, and
Idaho, tend to utilize more liquid-based systems to manage (flush or scrape) and store manure. Thus, the shift
toward larger dairy cattle and swine facilities since 1990 has translated into an increasing use of liquid manure
management systems, which have higher potential CH4 emissions than dry systems. This significant shift in both
the dairy cattle and swine industries was accounted for by incorporating state and WMS-specific CH4 conversion
factor (MCF) values in combination with the 1992,1997, 2002, 2007, 2012, and 2017 farm-size distribution data
reported in the U.S. Department of Agriculture (USDA) Census of Agriculture (USDA 2019d).
In 2021, total N2O emissions from manure management were estimated to be 17.4 MMT CO2 Eq. (66 kt); in 1990,
emissions were 12.4 MMT CO2 Eq. (47 kt). These values include both direct and indirect N2O emissions from
manure management. Nitrous oxide emissions have increased since 1990. Multiple drivers increase N2O emissions,
such as increasing nitrogen excretion rates for some animal types (see Annex, Table A-163) and increasing
numbers of animals on feedlots versus other dry systems (e.g., pasture). Across the entire time series, the overall
net effect is that N2O emissions showed a 40 percent increase from 1990 to 2021, but recent declines in a few
animal populations (e.g., swine and calves) resulted in a 0.5 percent decrease from 2020 to 2021.
Table 5-7 and Table 5-8 provide estimates of CH4 and N2O emissions from manure management by animal
category.7
Table 5-7: ChU and N2O Emissions from Manure Management (MMT CO2 Eq.)
Gas/Animal Type 1990 2005 2017 2018 2019 2020 2021
CH4a 39.0 54.9 64.4 66.5 65.7 66.7 66.0
Dairy Cattle 16.0 26.4 35.0 35.8 34.6 35.5 35.9
7 Manure management emissions from camels are not estimated because there is no significant population of camels in the
United States. Given the insignificance of estimated camel emissions in terms of the overall level and trend in national
emissions, there are no immediate improvement plans to include this emissions category in the Inventory. See Annex 5 for
more information on significance of estimated camel emissions.
Agriculture 5-13
-------
Swine
17.4
23.5
23.5
24.7
25.0
25.1
24.0
Poultry
3.7
3.6
3.8
3.9
4.0
4.0
3.9
Beef Cattle
1.8
1.9
2.0
2.0
2.0
2.0
2.0
Horses
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Sheep
0.1
0.1
0.06
0.06
0.05
0.05
0.05
Goats
+
+
+
+
+
+
+
American Bison
+
+
+
+
+
+
+
Mules and Asses
+
+
+
+
+
+
+
N2Ob
12.4
14.5
16.9
17.2
17.4
17.5
17.4
Beef Cattle
5.2
6.4
7.9
8.1
8.2
8.3
8.3
Dairy Cattle
4.6
4.8
5.4
5.4
5.4
5.5
5.5
Swine
1.1
1.4
1.7
1.8
1.9
1.9
1.8
Poultry
1.2
1.4
1.5
1.5
1.5
1.5
1.5
Sheep
0.1
0.3
0.3
0.3
0.3
0.3
0.3
Horses
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Goats
+
+
+
+
+
+
+
Mules and Asses
+
+
+
+
+
+
+
American Bisonc
NA
NA
NA
NA
NA
NA
NA
Total
51.4
69.4
81.3
83.7
83.1
84.2
83.4
+ Does not exceed 0.05 MMT C02 Eq.
NA (Not Available)
a Accounts for CH4 reductions due to capture and destruction of CH4 at facilities using anaerobic
digesters.
b Includes both direct and indirect N20 emissions.
cThere are no American bison N20 emissions from managed systems; American bison are
maintained entirely on pasture, range, and paddock.
Notes: N20 emissions from manure deposited on pasture, range and paddock are included in the
Agricultural Soils Management sector. Totals may not sum due to independent rounding.
l Table 5-8: ChU and N2O Emissions from Manure Management (kt)
Gas/Animal Type
1990
2005
2017
2018
2019
2020
2021
CH4a
1,394
1,960
2,300
2,375
2,348
2,383
2,358
Dairy Cattle
572
943
1,248
1,278
1,237
1,269
1,283
Swine
621
812
840
882
891
895
858
Poultry
131
130
136
139
144
142
141
Beef Cattle
63
67
70
70
71
71
71
Horses
4
5
3
3
3
3
3
Sheep
3
2
2
2
2
2
2
Goats
+
+
+
+
+
+
+
American Bison
+
+
+
+
+
+
+
Mules and Asses
+
+
+
+
+
+
+
N2Ob
47
55
64
65
65
66
66
Beef Cattle
20
24
30
30
31
31
31
Dairy Cattle
17
18
20
21
20
21
21
Swine
4
5
7
7
7
7
7
Poultry
5
5
5
6
6
6
6
Sheep
+
1
1
1
1
1
1
Horses
+
+
+
+
+
+
+
Goats
+
+
+
+
+
+
+
Mules and Asses
+
+
+
+
+
+
+
American Bisonc
NA
NA
NA
NA
NA
NA
NA
5-14 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
+ Does not exceed 0.5 kt.
NA (Not Available)
a Accounts for CH4 reductions due to capture and destruction of CH4 at facilities using anaerobic
digesters.
b Includes both direct and indirect N20 emissions.
cThere are no American bison N20 emissions from managed systems; American bison are
maintained entirely on pasture, range, and paddock.
Notes: N20 emissions from manure deposited on pasture, range and paddock are included in the
Agricultural Soils Management sector. Totals by gas may not sum due to independent rounding.
Methodology and Time-Series Consistency
The methodologies presented in IPCC (2006) form the basis of the Cm and N2O emission estimates for each animal
type, including Tier 1, Tier 2, and use of the CEFM previously described for Enteric Fermentation. These
methodologies use:
• IPCC (2006; 2019) Tier 1 default N2O emission factors and MCFs for dry systems
• U.S. specific MCFs for liquid systems (ERG 2001)
• U.S. specific values for volatile solids (VS) production rate and nitrogen excretion rate for some animal
types, including cattle values from the CEFM
This combination of Tier 1 and Tier 2 methods was applied to all livestock animal types. This section presents a
summary of the methodologies used to estimate CFU and N2O emissions from manure management. For the
current Inventory, time-series results were carried over from the 1990 to 2020 Inventory (i.e., 2022 submission)
and a simplified approach was used to estimate manure management emissions for 2021.
See Annex 3.11 for more detailed information on the methodologies (including detailed formulas and emission
factors), data used to calculate CH4 and N2O emissions, and emission results (including input variables and results
at the state-level) from manure management.
Methane Calculation Methods
The following inputs were used in the calculation of manure management CFU emissions for 1990 through 2020:
• Animal population data (by animal type and state);
• Typical animal mass (TAM) data (by animal type);
• Portion of manure managed in each WMS, by state and animal type;
• VS production rate (by animal type and state or United States);
• Methane producing potential (Bo) of the volatile solids (by animal type); and
• Methane conversion factors (MCF), the extent to which the CFU producing potential is realized for each
type of WMS (by state and manure management system, including the impacts of any biogas collection
efforts).
Methane emissions were estimated by first determining activity data, including animal population, TAM, WMS
usage, and waste characteristics. The activity data sources are described below:
• Annual animal population data for 1990 through 2020 for all livestock types, except goats, horses, mules
and asses, and American bison were obtained from the USDA-NASS. For cattle, the USDA populations
were utilized in conjunction with birth rates, detailed feedlot placement information, and slaughter
weight data to create the transition matrix in the Cattle Enteric Fermentation Model (CEFM) that models
cohorts of individual animal types and their specific emission profiles. The key variables tracked for each
of the cattle population categories are described in Section 5.1 and in more detail in Annex 3.10. Goat
population data for 1992,1997, 2002, 2007, 2012, and 2017; horse and mule and ass population data for
1987,1992,1997, 2002, 2007, 2012, and 2017; and American bison population for 2002, 2007, 2012, and
Agriculture 5-15
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
2017 were obtained from the Census of Agriculture (USDA 2019d). American bison population data for
1990 through 1999 were obtained from the National Bison Association (1999).
• The TAM is an annual average weight that was obtained for animal types other than cattle from
information in USDA's Agricultural Waste Management Field Handbook (USDA 1996), the American
Society of Agricultural Engineers, Standard D384.1 (ASAE 1998) and others (Meagher 1986; EPA 1992;
Safley 2000; ERG 2003b; IPCC 2006; ERG 2010a). For a description of the TAM data used for cattle, see
Annex 3.10.
• WMS usage was estimated for swine and dairy cattle for different farm size categories using state and
regional data from USDA (USDA APHIS 1996; Bush 1998; Ott 2000; USDA 2016c) and EPA (ERG 2000a; EPA
2002a and 2002b; ERG 2018, ERG 2019). For beef cattle and poultry, manure management system usage
data were not tied to farm size but were based on other data sources (ERG 2000a; USDA APHIS 2000; UEP
1999). For other animal types, manure management system usage was based on previous estimates (EPA
1992). American bison WMS usage was assumed to be the same as not on feed (NOF) cattle, while mules
and asses were assumed to be the same as horses.
• VS production rates for all cattle except for calves were calculated by head for each state and animal type
in the CEFM. VS production rates by animal mass for all other animals were determined using data from
USDA's Agricultural Waste Management Field Handbook (USDA 1996 and 2008; ERG 2010b and 2010c)
and data that was not available in the most recent Handbook were obtained from the American Society of
Agricultural Engineers, Standard D384.1 (ASAE 1998) or the 2006 IPCC Guidelines (IPCC 2006). American
bison VS production was assumed to be the same as NOF bulls.
• Bo was determined for each animal type based on literature values (Morris 1976; Bryant et al. 1976;
Hashimoto 1981; Hashimoto 1984; EPA 1992; Hill 1982; Hill 1984).
• MCFs for dry systems were set equal to default IPCC factors based on state climate for each year (IPCC
2006; IPCC 2019). MCFs for liquid/slurry, anaerobic lagoon, and deep pit systems were calculated based
on the forecast performance of biological systems relative to temperature changes as predicted in the
van't Hoff-Arrhenius equation which is consistent with IPCC (2006) Tier 2 methodology.
• Data from anaerobic digestion systems with CH4 capture and combustion were obtained from the EPA
AgSTAR Program, including information available in the AgSTAR project database (EPA 2021). Anaerobic
digester emissions were calculated based on estimated methane production and collection and
destruction efficiency assumptions (ERG 2008).
• For all cattle except for calves, the estimated amount of VS (kg per animal-year) managed in each WMS
for each animal type, state, and year were taken from the CEFM, assuming American bison VS production
to be the same as NOF bulls. For animals other than cattle, the annual amount of VS (kg per year) from
manure excreted in each WMS was calculated for each animal type, state, and year. This calculation
multiplied the animal population (head) by the VS excretion rate (kg VS per 1,000 kg animal mass per
day), the TAM (kg animal mass per head) divided by 1,000, the WMS distribution (percent), and the
number of days per year (365.25).
The estimated amount of VS managed in each WMS was used to estimate the CH4 emissions (kg CH4 per year) from
each WMS. The amount of VS (kg per year) was multiplied by the Bo (m3 CH4 per kg VS), the MCF for that WMS
(percent), and the density of CH4 (kg CH4 per m3 CH4). The CH4 emissions for each WMS, state, and animal type
were summed to determine the total U.S. CH4 emissions. See details in Step 5 of Annex 3.11.
The following approach was used in the calculation of manure management CH4 emissions for 2021:
• Obtain 2021 national-level animal population data: Sheep, poultry, and swine data were downloaded
from USDA-NASS Quickstats (USDA 2022). Cattle populations were obtained from the CEFM (see NIR
Section 5.land Annex 3.10). Data for goats, horses, bison, mules, and asses were extrapolated based on
the 2011 through 2020 population values to reflect recent trends in animal populations.
5-16 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
• Multiply the national populations by the animal-specific 2020 implied emission factors8 for CFU to
calculate national-level 2021 CFU emissions estimates by animal type. These methods were utilized in
order to maintain time-series consistency as referenced in Volume 1, Chapter 5 of the 2006IPCC
Guidelines.
Nitrous Oxide Calculation Methods
The following inputs were used in the calculation of direct and indirect manure management N2O emissions for
1990 through 2020:
• Animal population data (by animal type and state);
• TAM data (by animal type);
• Portion of manure managed in each WMS (by state and animal type);
• Total Kjeldahl N excretion rate (Nex);
• Direct N2O emission factor (EFwms);
• Indirect N2O emission factor for volatilization (EFvoiatiiization);
• Indirect N2O emission factor for runoff and leaching (EFrunoff/ieach);
• Fraction of N loss from volatilization of NH3 and NOx (Fracgas); and
• Fraction of N loss from runoff and leaching (Fracmnoff/ieach).
Nitrous oxide emissions were estimated by first determining activity data, including animal population, TAM, WMS
usage, and waste characteristics. The activity data sources (except for population, TAM, and WMS, which were
described above) are described below:
• Nex for all cattle except for calves were calculated by head for each state and animal type in the CEFM.
Nex rates by animal mass for all other animals were determined using data from USDA's Agricultural
Waste Management Field Handbook (USDA 1996 and 2008; ERG 2010b and 2010c) and data from the
American Society of Agricultural Engineers, Standard D384.1 (ASAE 1998) and IPCC (2006). American bison
Nex were assumed to be the same as NOF bulls.9
• All N2O emission factors (direct and indirect) were taken from IPCC (2006).
• Country-specific estimates for the fraction of N loss from volatilization (Fracgas) and runoff and leaching
(FraCrunoff/ieach) were developed. Fracgas values were based on WMS-specific volatilization values as
estimated from EPA's National Emission Inventory - Ammonia Emissions from Animal Agriculture
Operations (EPA 2005). Fracmnoff/ieaching values were based on regional cattle runoff data from EPA's Office
of Water (EPA 2002b; see Annex 3.11).
To estimate N2O emissions for cattle (except for calves), the estimated amount of N excreted (kg per animal-year)
that is managed in each WMS for each animal type, state, and year were taken from the CEFM. For calves and
other animals, the amount of N excreted (kg per year) in manure in each WMS for each animal type, state, and
year was calculated. The population (head) for each state and animal was multiplied by TAM (kg animal mass per
8 An implied emission factor is defined as emissions divided by the relevant measure of activity; the implied emission factor is
equal to emissions per activity data unit. For source/sink categories that are composed of several subcategories, the emissions
and activity data are summed up across all subcategories. Hence, the implied emission factors are generally not equivalent to
the emission factors used to calculate emission estimates, but are average values that could be used, with caution, in data
comparisons (UNFCCC 2017).
9 Nex of American bison on grazing lands are accounted for and discussed in the Agricultural Soil Management source category
and included under pasture, range and paddock (PRP) emissions. Because American bison are maintained entirely on
unmanaged WMS and N20 emissions from unmanaged WMS are not included in the Manure Management source category,
there are no N20 emissions from American bison included in the Manure Management source category.
Agriculture 5-17
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
head) divided by 1,000, the nitrogen excretion rate (Nex, in kg N per 1,000 kg animal mass per day), WMS
distribution (percent), and the number of days per year.
Direct N2O emissions were calculated by multiplying the amount of N excreted (kg per year) in each WMS by the
N2O direct emission factor for that WMS (EFwms, in kg N2O-N per kg N) and the conversion factor of N2O-N to N2O.
These emissions were summed over state, animal, and WMS to determine the total direct N2O emissions (kg of
N2O per year). See details in Step 6 of Annex 3.11.
Indirect N2O emissions from volatilization (kg N2O per year) were then calculated by multiplying the amount of N
excreted (kg per year) in each WMS by the fraction of N lost through volatilization (Fracgas) divided by 100, the
emission factor for volatilization (EFvoiatiiization, in kg N2O per kg N), and the conversion factor of N2O-N to N2O.
Indirect N2O emissions from runoff and leaching (kg N2O per year) were then calculated by multiplying the amount
of N excreted (kg per year) in each WMS by the fraction of N lost through runoff and leaching (Fracmnoff/ieach)
divided by 100, the emission factor for runoff and leaching (EFrunoff/ieach, in kg N2O per kg N), and the conversion
factor of N2O-N to N2O. The indirect N2O emissions from volatilization and runoff and leaching were summed to
determine the total indirect N2O emissions. See details in Step 6 of Annex 3.11.
Following these steps, direct and indirect N2O emissions were summed to determine total N2O emissions (kg N2O
per year) for the years 1990 to 2020.
Methodological approaches, changes to historic data, and other parameters were applied to the entire time series
to ensure consistency in emissions estimates from 1990 through 2020. In some cases, the activity data source
changed over the time series. For example, updated WMS distribution data were applied to 2016 for dairy cows
and 2009 for swine. While previous WMS distribution data were from another data source, EPA integrated the
more recent data source to reflect the best available current WMS distribution data for these animals. EPA
assumed a linear interpolation distribution for years between the two data sources. Refer to Annex 3.11 for more
details on data sources and methodology.
The following approach was used in the calculation of manure management N2O emissions for 2021:
• Obtain 2021 national-level animal population data: Sheep, poultry, and swine data were downloaded
from USDA-NASS Quickstats (USDA 2022). Cattle populations were obtained from the CEFM, see Section
5.1 and Annex 3.10 (Enteric Fermentation). Data for goats, horses, bison, mules, and asses were
extrapolated based on the 2011 through 2020 population values to reflect recent trends in animal
populations.
• The national populations were multiplied by the animal-specific 2020 implied emission factors for N2O
(which combines both direct and indirect N2O) to calculate national-level 2021 N2O emissions estimates
by animal type. These methods were utilized in order to maintain time-series consistency as referenced in
Volume 1, Chapter 5 of the 2006IPCC Guidelines.
Uncertainty
An analysis (ERG 2003a) was conducted for the manure management emission estimates presented in the 1990
through 2001 Inventory (i.e., 2003 submission to the UNFCCC) to determine the uncertainty associated with
estimating CFU and N2O emissions from livestock manure management. The quantitative uncertainty analysis for
this source category was performed in 2002 through the IPCC-recommended Approach 2 uncertainty estimation
methodology, the Monte Carlo Stochastic Simulation technique. The uncertainty analysis was developed based on
the methods used to estimate CFU and N2O emissions from manure management systems. A normal probability
distribution was assumed for each source data category. The series of equations used were condensed into a single
equation for each animal type and state. The equations for each animal group contained four to five variables
around which the uncertainty analysis was performed for each state. While there are plans to update the
uncertainty to reflect recent manure management updates and forthcoming changes (see Planned Improvements,
below), at this time the uncertainty estimates were directly applied to the 2021 emission estimates.
5-18 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 5-9. Manure management
Cm emissions in 2021 were estimated to be between 54.1 and 79.2 MMT CO2 Eq. at a 95 percent confidence level,
which indicates a range of 18 percent below to 20 percent above the actual 2021 emission estimate of 66.0 MMT
CO2 Eq. At the 95 percent confidence level, N2O emissions were estimated to be between 14.6 and 21.6 MMT CO2
Eq. (or approximately 16 percent below and 24 percent above the actual 2021 emission estimate of 17.4 MMT CO2
Eq.).
Table 5-9: Approach 2 Quantitative Uncertainty Estimates for CH4 and N2O (Direct and
Indirect) Emissions from Manure Management (MMT CO2 Eq. and Percent)
2021 Emission
Source
Gas
Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT CO? Eq.)
(MMT CO?
Eq.)
(%)
Lower
Upper
Lower Upper
Bound
Bound
Bound Bound
Manure Management
ch4
66.0
54.1
79.2
-18% +20%
Manure Management
n2o
17.4
14.6
21.6
-16% +24%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence
interval.
QA/QC and Verification
General (Tier 1) and category-specific (Tier 2) QA/QC activities were conducted consistent with the U.S. Inventory
QA/QC plan outlined in Annex 8. Tier 2 activities focused on comparing estimates for the previous and current
Inventories for N2O emissions from managed systems and CH4 emissions from livestock manure. All errors
identified were corrected. Order of magnitude checks were also conducted, and corrections made where needed.
In addition, manure N data were checked by comparing state-level data with bottom-up estimates derived at the
county level and summed to the state level. Similarly, a comparison was made by animal and WMS type for the full
time series, between national level estimates for N excreted, both for pasture and managed systems, and the sum
of county estimates for the full time series. This was done to ensure consistency between excreted N within the
manure management sector and those data provided to the managed soils sector. All errors identified were
corrected.
Time-series data, including population, are validated by experts to ensure they are representative of the best
available U.S.-specific data. The U.S.-specific values for TAM, Nex, VS, Bo, and MCF were also compared to the IPCC
default values and validated by experts. Although significant differences exist in some instances, these differences
are due to the use of U.S.-specific data and the differences in U.S. agriculture as compared to other countries. The
U.S. manure management emission estimates use the most reliable country-specific data, which are more
representative of U.S. animals and systems than the IPCC (2006) default values.
For additional verification of the 1990 to 2020 estimates, the implied CH4 emission factors for manure
management (kg of CH4 per head per year) were compared against the default IPCC (2006) values. Table 5-10
presents the implied emission factors of kg of CH4 per head per year used for the manure management emission
estimates as well as the IPCC (2006) default emission factors. The U.S. implied emission factors fall within the
range of the IPCC (2006) default values, except in the case of sheep, goats, and some years for horses and dairy
cattle. The U.S. implied emission factors are greater than the IPCC (2006) default value for those animals due to
the use of U.S.-specific data for typical animal mass and VS excretion. There is an increase in implied emission
factors for dairy cattle and swine across the time series. This increase reflects the dairy cattle and swine industry
trend towards larger farm sizes; large farms are more likely to manage manure as a liquid and therefore produce
more CH4 emissions.
Agriculture 5-19
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Table 5-10: IPCC (2006) Implied Emission Factor Default Values Compared with Calculated
Values for ChU from Manure Management (kg/head/year)
IPCC Default
Animal Type
CH4 Emission
Factors
Implied CH4 Emission Factors (kg/head/year)
(ke/head/vear)a
1990
2005
2017
2018
2019
2020
2021
Dairy Cattle
48-112
29.3
53.0
66.0
67.3
65.6
67.5
67.5
Beef Cattle
1-2
0.8
0.8
0.9
0.9
0.9
0.9
0.9
Swine
10-45
11.5
13.3
11.6
12.0
11.6
11.6
11.6
Sheep
0.19-0.37
0.3
0.4
0.4
0.4
0.4
0.4
0.4
Goats
0.13-0.26
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Poultry
0.02-1.4
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Horses
1.56-3.13
1.9
1.4
1.2
1.2
1.2
1.2
1.2
American Bison
NA
0.8
0.9
0.9
0.9
0.9
0.9
0.9
Mules and Asses
0.76-1.14
0.4
0.4
0.4
0.4
0.4
0.4
0.4
Note: CH4 implied emission factors were not calculated for 2021 due to the simplified emissions estimation
approach used to estimate emissions for that year. 2020 values were used for 2021.
NA (Not Applicable)
a Ranges reflect 2006 IPCC Guidelines (Volume 4, Table 10.14) default emission factors for North America across
different climate zones.
In addition, default IPCC (2006) emission factors for N2O were compared to the U.S. Inventory implied N2O
emission factors. Default N2O emission factors from the 2006 IPCC Guidelines were used to estimate N2O emission
from each WMS in conjunction with U.S.-specific Nex values. The implied emission factors differed from the U.S.
Inventory values due to the use of U.S.-specific Nex values and differences in populations present in each WMS
throughout the time series.
Recalculations Discussion
EPA updated global warming potentials (GWP) for calculating CC>2-equivalent emissions of CFU and N2O to reflect
the 100-year GWPs provided in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). The previous Inventory used
100-year GWPs provided in the IPCC Fourth Assessment Report (AR4). The AR5 GWPs have been applied across the
entire time series for consistency. The GWP of CFU has increased from 25 to 28, leading to an increase in the
calculated CC>2-equivalent emissions of CFU, while the GWP of N2O has decreased from 298 to 265, leading to a
decrease in the calculated CC>2-equivalent emissions of N2O. The cumulative effect of these recalculations had a
low impact on the overall manure management emission estimates.
On average, CC>2-equivalent total emissions increased by 5.7 percent for each year of the time series compared to
the previous Inventory. Further discussion on this update and the overall impacts of updating the Inventory GWP
values to reflect the AR5 can be found in Chapter 9, Recalculations and Improvements.
Planned Improvements
Regular annual data reviews and updates are necessary to maintain an emissions inventory that reflects the
current base of knowledge. In addition to the documented approaches currently used to address data availability,
EPA conducts data assessments and is actively pursuing the following investigations for the 2024 Inventory
submission:
• Continuing to investigate new sources of WMS data. EPA is working with the USDA Natural Resources
Conservation Service to collect data for potential improvements to the Inventory.
• Determining appropriate updates to other default N2O emission factors to reflect IPCC (2019).
Many of the improvements identified below are major updates and may take multiple years to fully
implement. Potential improvements (long-term improvements) for future Inventory years include:
5-20 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 • Revising the anaerobic digestion estimates to estimate CFU emissions reductions due to the use of
2 anaerobic digesters (the Inventory currently estimates only emissions from anaerobic digestion systems).
3 • Investigating the updated IPCC 2019 Refinement default N2O emissions factor for anaerobic digesters.
4 Historically, EPA has not estimated N2O emissions from digesters as the default guidance was no
5 emissions. Incorporating AgSTAR data for N2O emissions, like CFU emissions, is a longer-term goal for EPA.
6 • Investigating updates to the current anaerobic digester MCFs based on IPCC (2019).
7 • Investigating the typical animal masses used in each the Enteric Fermentation and Manure Management
8 inventories and confirm they align.
9 EPA is aware of the following potential updates or improvements but notes that implementation will be based on
10 available resources and data availability:
11 • Updating the Bo data used in the Inventory, as data become available. EPA is conducting outreach with
12 counterparts from USDA as to available data and research on Bo.
13 • Comparing CH4 and N2O emission estimates with estimates from other models and more recent studies
14 and compare the results to the Inventory.
15 • Comparing manure management emission estimates with on-farm measurement data to identify
16 opportunities for improved estimates.
17 • Comparing VS and Nex data to literature data to identify opportunities for improved estimates.
18 • Determining if there are revisions to the U.S.-specific method for calculating liquid systems for MCFs
19 based on updated guidance from the IPCC 2019 Refinement.
20 • Investigating improved emissions estimate methodologies for swine pit systems with less than one month
21 of storage (the recently updated swine WMS data included this WMS category).
22 • Improving the linkages with the Enteric Fermentation source category estimates. For future Inventories, it
23 may be beneficial to have the CEFM and Manure Management calculations in the same model, as they
24 rely on much of the same activity data and on each other's outputs to properly calculate emissions.
25 • Revising the uncertainty analysis to address changes that have been implemented to the CH4 and N2O
26 estimates. The plan is to align the timing of the updated Manure Management uncertainty analysis with
27 the uncertainty analysis for Enteric Fermentation.
28 5.3 Rice Cultivation (CRF Source Category
- 3C)
30 Most of the world's rice is grown on flooded fields (Baicich 2013) that create anaerobic conditions leading to CH4
31 production through a process known as methanogenesis. Approximately 60 to 90 percent of the CH4 produced by
32 methanogenic bacteria in flooded rice fields is oxidized in the soil and converted to CO2 by methanotrophic
33 bacteria. The remainder is emitted to the atmosphere (Holzapfel-Pschorn et al. 1985; Sass et al. 1990) or
34 transported as dissolved CH4 into groundwater and waterways (Neue et al. 1997). Methane is transported to the
35 atmosphere primarily through the rice plants, but some CH4 also escapes via ebullition (i.e., bubbling through the
36 water) and to a much lesser extent by diffusion through the water (van Bodegom et al. 2001).
37 Water management is arguably the most important factor affecting CFU emissions in rice cultivation, and improved
38 water management has the largest potential to mitigate emissions (Yan et al. 2009). Upland rice fields are not
39 flooded, and therefore do not produce CH4, but large amounts of CFUcan be emitted in continuously irrigated
Agriculture 5-21
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
fields, which is the most common practice in the United States (USDA 2012). Single or multiple aeration events
with drainage of a field during the growing season can significantly reduce these emissions (Wassmann et al.
2000a), but drainage may also increase N2O emissions. Deepwater rice fields (i.e., fields with flooding depths
greater than one meter, such as natural wetlands) tend to have fewer living stems reaching the soil, thus reducing
the amount of CH4 transport to the atmosphere through the plant compared to shallow-flooded systems (Sass
2001).
Other management practices also influence CH4 emissions from flooded rice fields including rice residue straw
management and application of organic amendments, in addition to cultivar selection due to differences in the
amount of root exudates10 among rice varieties (Neue et al. 1997). These practices influence the amount of
organic matter available for methanogenesis, and some practices, such as mulching rice straw or composting
organic amendments, can reduce the amount of labile carbon and limit CH4 emissions (Wassmann et al. 2000b).
Fertilization practices also influence Cm emissions, particularly the use of fertilizers with sulfate, which can reduce
Cm emissions (Wassmann et al. 2000b; Linquist et al. 2012). Other environmental variables also impact the
methanogenesis process such as soil temperature and soil type. Soil temperature regulates the activity of
methanogenic bacteria, which in turn affects the rate of CH4 production. Soil texture influences decomposition of
soil organic matter, but is also thought to have an impact on oxidation of CH4 in the soil (Sass et al. 1994).
Rice is currently cultivated in thirteen states, including Arkansas, California, Florida, Illinois, Kentucky, Louisiana,
Minnesota, Mississippi, Missouri, New York, South Carolina, Tennessee and Texas. Soil types, rice varieties, and
cultivation practices vary across the United States, but most farmers apply fertilizers and do not harvest crop
residues. In addition, a second, ratoon rice crop is sometimes grown in the Southeastern region of the country.
Ratoon crops are produced from regrowth of the stubble remaining after the harvest of the first rice crop.
Methane emissions from ratoon crops are higher than those from the primary crops due to the increased amount
of labile organic matter available for anaerobic decomposition in the form of relatively fresh crop residue straw.
Emissions tend to be higher in rice fields if the residues have been in the field for less than 30 days before planting
the next rice crop (Lindau and Bollich 1993; IPCC 2006; Wang et al. 2013).
A combination of Tier 1 and 3 methods are used to estimate CH4 emissions from rice cultivation across most of the
time series, while a surrogate data method has been applied to estimate national emissions for 2016 to 2021 in
this Inventory due to lack of data in the later years of the time series. National emission estimates based on
surrogate data will be recalculated in a future Inventory with the Tier 1 and 3 methods as data becomes available.
Overall, rice cultivation is a minor source of CH4 emissions in the United States relative to other source categories
(see Table 5-11, Table 5-12, and Figure 5-3). Most emissions occur in Arkansas, California, Louisiana, Mississippi,
Missouri and Texas. In 2021, CH4 emissions from rice cultivation were 16.8 MMT CO2 Eq. (600 kt). Annual emissions
fluctuate between 1990 and 2021, which is largely due to differences in the amount of rice harvested areas over
time, which has been decreasing over the past two decades. Consequently, emissions in 2021 are 6 percent lower
than emissions in 1990.
Table 5-11: ChU Emissions from Rice Cultivation (MMT CO2 Eq.)
State
1990
2005
2017
2018
2019
2020
2021
Arkansas
6.0
00
00
NE
NE
NE
NE
NE
California
3.7
3.8
NE
NE
NE
NE
NE
Florida
+
+
NE
NE
NE
NE
NE
Illinois
+
+
NE
NE
NE
NE
NE
Kentucky
+
+
NE
NE
NE
NE
NE
Louisiana
2.9
3.2
NE
NE
NE
NE
NE
Minnesota
+
0.1
NE
NE
NE
NE
NE
10 The roots of rice plants add organic material to the soil through a process called "root exudation." Root exudation is thought
to enhance decomposition of the soil organic matter and release nutrients that the plant can absorb for production. The
amount of root exudate produced by a rice plant over a growing season varies among rice varieties.
5-22 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Mississippi
1.3
1.5
NE
NE
NE
NE
NE
Missouri
0.6
1.3
NE
NE
NE
NE
NE
New York
+
+
NE
NE
NE
NE
NE
South Carolina
+
+
NE
NE
NE
NE
NE
Tennessee
+
+
NE
NE
NE
NE
NE
Texas
3.4
1.5
NE
NE
NE
NE
NE
Total
17.9
20.2
16.7
17.4
16.9
17.6
16.8
+ Does not exceed 0.05 MMT C02 Eq.
NE (Not Estimated). State-level emissions are not estimated for 2016 through 2021 in this Inventory. A
surrogate data method is used to estimate emissions for these years and are produced only at the
national scale.
Note: Totals may not sum due to independent rounding.
l Table 5-12: ChU Emissions from Rice Cultivation (kt)
State
1990
2005
2017
2018
2019
2020
2021
Arkansas
216
315
NE
NE
NE
NE
NE
California
131
134
NE
NE
NE
NE
NE
Florida
+
1
NE
NE
NE
NE
NE
Illinois
+
+
NE
NE
NE
NE
NE
Kentucky
+
+
NE
NE
NE
NE
NE
Louisiana
103
113
NE
NE
NE
NE
NE
Minnesota
1
2
NE
NE
NE
NE
NE
Mississippi
45
55
NE
NE
NE
NE
NE
Missouri
22
45
NE
NE
NE
NE
NE
New York
+
+
NE
NE
NE
NE
NE
South Carolina
+
+
NE
NE
NE
NE
NE
Tennessee
+
+
NE
NE
NE
NE
NE
Texas
122
54
NE
NE
NE
NE
NE
Total
640
720
596
623
602
630
600
+ Does not exceed 0.5 kt.
NE (Not Estimated). State-level emissions are not estimated for 2016 through 2021 in this Inventory. A
surrogate data method is used to estimate emissions for these years and are produced only at the
national scale.
Note: Totals may not sum due to independent rounding.
Agriculture 5-23
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Figure 5-3: Annual ChU Emissions from Rice Cultivation, 2015
¦ >20
Note: Only national-scale emissions are estimated for 2016 through 2021 in this Inventory using the surrogate data method
described in the Methodology section; therefore, the fine-scale emission patterns in this map are based on the estimates for
2015.
Methodology and Time-Series Consistency
The methodology used to estimate Cm emissions from rice cultivation is based on a combination of IPCC Tier 1 and
3 approaches. The Tier 3 method utilizes the DayCent process-based model to estimate CHU emissions from rice
cultivation (Cheng et al. 2013), and has been tested in the United States (see Annex 3.12) and Asia (Cheng et al.
2013, 2014). The model simulates hydrological conditions and thermal regimes, organic matter decomposition,
root exudation, rice plant growth and its influence on oxidation of ChU, as well as CHj transport through the plant
and via ebullition (Cheng et al. 2013). The 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 ChU emissions, DayCent simulates soil C stock changes and N2O emissions (Parton et
al. 1987 and 1998; Del Grosso et al. 2010), and allows for a seamless set of simulations for crop rotations that
include both rice and non-rice crops.
The 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 percent by volume). The Tier 3 method using DayCent
has not been fully tested for estimating emissions associated with these conditions.
The Tier 1 method for estimating CH4 emissions from rice production utilizes a default base emission rate and
scaling factors (IPCC 2006). The base emission rate represents emissions for continuously flooded fields with no
5-24 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
organic amendments. Scaling factors are used to adjust the base emission rate for water management and organic
amendments that differ from continuous flooding with no organic amendments. The method accounts for pre-
season and growing season flooding; types and amounts of organic amendments; and the number of rice
production seasons within a single year (i.e., single cropping, ratooning, etc.). The Tier 1 analysis is implemented in
the Agriculture and Land Use National Greenhouse Gas Inventory (ALU) software (Ogle et al. 2016).11
Rice cultivation areas are based on crop and land use histories recorded in the USDA National Resources Inventory
(NRI) survey (USDA-NRCS 2018). The NRI is a statistically-based sample of all non-federal land, and includes
489,178 survey locations in agricultural land for the conterminous United States and Hawaii of which 1,960 include
one or more years of rice cultivation. The Tier 3 method is used to estimate Cm emissions from 1,655 of the NRI
survey locations, and the remaining 305 survey locations are estimated with the Tier 1 method. Each NRI survey
location is associated with an "expansion factor" that allows scaling of Cm emission to the entire land base with
rice cultivation (i.e., each expansion factor represents the amount of area with the same land-use/management
history as the survey location). Land-use and some management information in the NRI (e.g., crop type, soil
attributes, and irrigation) were collected on a 5-year cycle beginning in 1982, along with cropping rotation data in
4 out of 5 years for each 5-year time period (i.e., 1979 to 1982,1984 to 1987,1989 to 1992, and 1994 to 1997).
The NRI program began collecting annual data in 1998, with data through 2015 (USDA-NRCS 2018). The current
Inventory only uses NRI data through 2015, and the harvested rice areas in each state are presented in Table 5-13.
Table 5-13: Rice Area Harvested (1,000 Hectares)
State/Crop
1990
2005
2017
2018
2019
2020
2021
Arkansas
600
784
NE
NE
NE
NE
NE
California
249
236
NE
NE
NE
NE
NE
Florida
0
4
NE
NE
NE
NE
NE
Illinois
0
0
NE
NE
NE
NE
NE
Kentucky
0
0
NE
NE
NE
NE
NE
Louisiana
381
402
NE
NE
NE
NE
NE
Minnesota
4
9
NE
NE
NE
NE
NE
Mississippi
123
138
NE
NE
NE
NE
NE
Missouri
48
94
NE
NE
NE
NE
NE
New York
1
0
NE
NE
NE
NE
NE
South Carolina
0
0
NE
NE
NE
NE
NE
Tennessee
0
1
NE
NE
NE
NE
NE
Texas
302
118
NE
NE
NE
NE
NE
Total
1,707
1,788
NE
NE
NE
NE
NE
NE (Not Estimated). Area data will be updated in the next inventory.
Note: Totals may not sum due to independent rounding.
The Southeastern states have sufficient growing periods for a ratoon crop in some years (Table 5-14). For example,
the growing season length is occasionally sufficient for ratoon crops to be grown on about 1 percent of the rice
fields in Arkansas. No data are available about ratoon crops in Missouri or Mississippi, and the average amount of
ratooning in Arkansas was assigned to these states. Ratoon cropping occurs much more frequently in Louisiana
(LSU 2015 for years 2000 through 2013, 2015) and Texas (TAMU 2015 for years 1993 through 2015), averaging 32
percent and 45 percent of rice acres planted, respectively. Florida also has a large fraction of area with a ratoon
crop (49 percent). Ratoon rice crops are not grown in California.
11 See http://www.nrel.colostate.edu/proiects/ALUsoftware/.
Agriculture 5-25
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Table 5-14: Average Ratooned Area as Percent of Primary Growth Area (Percent)
State
1990-2015
Arkansas3
California
Florida15
Louisiana0
Mississippi3
Missouri3
Texasd
1%
0%
49%
32%
1%
1%
45%
3 Arkansas: 1990-2000 (Slaton 1999 through 2001); 2001-2011 (Wilson 2002 through 2007, 2009 through 2012); 2012-2013
(Hardke 2013, 2014). Estimates of ratooning for Missouri and Mississippi are based on the data from Arkansas.
b Florida - Ratoon: 1990-2000 (Schueneman 1997,1999 through 2001); 2001 (Deren 2002); 2002-2003 (Kirstein 2003
through 2004, 2006); 2004 (Cantens 2004 through 2005); 2005-2013 (Gonzalez 2007 through 2014).
c Louisiana: 1990-2013 (Linscombe 1999, 2001 through 2014).
dTexas: 1990-2002 (Klosterboer 1997,1999 through 2003); 2003-2004 (Stansel 2004 through 2005); 2005 (Texas Agricultural
Experiment Station 2006); 2006-2013 (Texas Agricultural Experiment Station 2007 through 2014).
While rice crop production in the United States includes a minor amount of land with mid-season drainage or
alternate wet-dry periods, the majority of rice growers use continuously flooded water management systems
(Hardke 2015; UCCE 2015; Hollier 1999; Way et al. 2014). Therefore, continuous flooding was assumed in the
DayCent simulations and the Tier 1 method. Variation in flooding can be incorporated in future Inventories if water
management data are collected.
Winter flooding is another key practice associated with water management in rice fields, and the impact of winter
flooding on Cm emissions is addressed in the Tier 3 and Tier 1 analyses. Flooding is used to prepare fields for the
next growing season, and to create waterfowl habitat (Young 2013; Miller et al. 2010; Fleskes et al. 2005).
Fitzgerald et al. (2000) suggests that as much as 50 percent of the annual emissions may occur during winter
flooding. Winter flooding is a common practice with an average of 34 percent of fields managed with winter
flooding in California (Miller et al. 2010; Fleskes et al. 2005), and approximately 21 percent of the fields managed
with winter flooding in Arkansas (Wilson and Branson 2005 and 2006; Wilson and Runsick 2007 and 2008; Wilson
et al. 2009 and 2010; Hardke and Wilson 2013 and 2014; Hardke 2015). No data are available on winter flooding
forTexas, Louisiana, Florida, Missouri, or Mississippi. Forthese states, the average amount of flooding is assumed
to be similar to Arkansas. In addition, the amount of flooding is assumed to be relatively constant over the
Inventory time series.
A surrogate data method is used to estimate emissions from 2016 to 2021 associated with the rice CH4 emissions
for Tier 1 and 3 methods. Specifically, a linear regression model with autoregressive moving-average (ARMA)
errors was used to estimate the relationship between the surrogate data and emissions data from 1990 through
2015, which were derived using the Tier 1 and 3 methods (Brockwell and Davis 2016). Surrogate data are based on
rice commodity statistics from USDA-NASS.12 See Box 5-2 for more information about the surrogate data method.
Box 5-2: Surrogate Data Method
An approach to extend the time series is needed to estimate emissions from Rice Cultivation because there are
gaps in activity data at the end of the time series. This is mainly due to the fact that the National Resources
Inventory (NRI) does not release data every year, and the NRI is a key data source for estimating greenhouse gas
emissions.
A surrogate data method has been selected to impute missing emissions at the end of the time series. A linear
regression model with autoregressive moving-average (ARMA) errors (Brockwell and Davis 2016) is used to
estimate the relationship between the surrogate data and the observed 1990 to 2015 emissions data that has
12 See https://quickstats.nass.usda.gov/.
5-26 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
been compiled using the inventory methods described in this section. The model to extend the time series is
given by
Y=xp+ £,
where Y is the response variable (e.g., CH emissions), xp is the surrogate data that is used to predict the
missing emissions data, and e is the remaining unexplained error. Models with a variety of surrogate data were
tested, including commodity statistics, weather data, or other relevant information. Parameters are estimated
from the observed data for 1990 to 2015 using standard statistical techniques, and these estimates are used to
predict the missing emissions data for 2016 to 2021.
A critical issue in using splicing methods is to adequately account for the additional uncertainty introduced by
predicting emissions with related information without compiling the full inventory. For example, predicting CH
emissions will increase the total variation in the emission estimates for these specific years, compared to those
years in which the full inventory is compiled. This added uncertainty is quantified within the model framework
using a Monte Carlo approach. The approach requires estimating parameters for results in each Monte Carlo
simulation for the full inventory (i.e., the surrogate data model is refit with the emissions estimated in each
Monte Carlo iteration from the full inventory analysis with data from 1990 to 2015).
1
2 In order to ensure time-series consistency, the same methods are applied from 1990 to 2015, and a surrogate data
3 method is used to approximate emissions for the remainder of the 2016 to 2021 time series based on the
4 emissions data from 1990 to 2015. This surrogate data method is consistent with data splicing methods in IPCC
5 (2006).
6 Uncertainty
7 Sources of uncertainty in the Tier 3 method include management practices, uncertainties in model structure (i.e.,
8 algorithms and parameterization), and variance associated with the NRI sample. Sources of uncertainty in the IPCC
9 (2006) Tier 1 method include the emission factors, management practices, and variance associated with the NRI
10 sample. A Monte Carlo analysis was used to propagate uncertainties in the Tier 1 and 3 methods. For 2016 to 2021,
11 there is additional uncertainty propagated through the Monte Carlo analysis associated with the surrogate data
12 method (See Box 5-2 for information about propagating uncertainty with the surrogate data method). The
13 uncertainties from the Tier 1 and 3 approaches are combined to produce the final Cm emissions estimate using
14 simple error propagation (IPCC 2006). Additional details on the uncertainty methods are provided in Annex 3.12.
15 Rice cultivation CFU emissions in 2021 were estimated to be between 4.2 and 29.4 MMT CO2 Eq. at a 95 percent
16 confidence level, which indicates a range of 75 percent below to 75 percent above the 2021 emission estimate of
17 16.8 MMT CO2 Eq. (see Table 5-15).
18 Table 5-15: Approach 2 Quantitative Uncertainty Estimates for ChU Emissions from Rice
19 Cultivation (MMT CO2 Eq. and Percent)
Source
Inventory
Method
Gas
2021 Emission
Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT CO? Eq.)
(MMT CO?
Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Rice Cultivation
Tier 3
ch4
14.0
1.4
26.6
-90%
+90%
Rice Cultivation
Tier 1
ch4
2.8
1.5
4.1
-48%
+48%
Rice Cultivation
Total
ch4
16.8
4.2
29.4
-75%
+75%
a Range of emission estimates is the 95 percent confidence interval.
Agriculture 5-27
-------
1 QA/QC and Verification
2 General (Tier 1) and category-specific (Tier 2) QA/QC activities were conducted consistent with the U.S. Inventory
3 QA/QC plan outlined in Annex 8. Quality control measures include checking input data, model scripts, and results
4 to ensure data are properly handled throughout the inventory process. Inventory reporting forms and text are
5 reviewed and revised as needed to correct transcription errors.
6 Model results are compared to field measurements to verify if results adequately represent Cm emissions. The
7 comparisons included over 17 long-term experiments, representing about 238 combinations of management
8 treatments across all the sites. A statistical relationship was developed to assess uncertainties in the model
9 structure, adjusting the estimates for model bias and assessing precision in the resulting estimates (methods are
10 described in Ogle et al. 2007). See Annex 3.12 for more information.
11 Recalculations Discussion
12 EPA updated global warming potential (GWP) for calculating C02-equivalent emissions of CFU (from 25 to 28) to
13 reflect the 100-year GWPs provided in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). The previous Inventory
14 used 100-year GWPs provided in the IPCC Fourth Assessment Report (AR4). This update was applied across the
15 entire time series for consistency. As a result of this change, C02-equivalent emissions increased by an annual
16 average of 1.9 MMT CO2 Eq., or 12 percent, over the time series from 1990 to 2020 compared to the previous
17 Inventory. Further discussion on this update and the overall impacts of updating the Inventory GWP values to
18 reflect the AR5 can be found in Chapter 9, Recalculations and Improvements.
19 Planned Improvements
20 A key planned improvement for rice cultivation is to fill several gaps in the management activity including
21 compiling new data on water management, organic amendments and ratooning practices in rice cultivation
22 systems. This improvement is expected to be completed for the next Inventory, but may not be prioritized
23 depending on the needs for other inventory improvements in the Agriculture sector.
24 5.4 Agricultural Soil Management (CRF
25 Source Category 3D)
26 Nitrous oxide is naturally produced in soils through the microbial processes of nitrification and denitrification that
27 is driven by the availability of mineral nitrogen (N) (Firestone and Davidson 1989).13 Mineral N is made available in
28 soils through decomposition of soil organic matter and plant litter, as well as asymbiotic fixation of N from the
29 atmosphere.14 Several agricultural activities increase mineral N availability in soils that lead to direct N2O
30 emissions at the site of a management activity (see Figure 5-4) (Mosier et al. 1998). These activities include
31 synthetic N fertilization; application of managed livestock manure; application of other organic materials such as
32 biosolids (i.e., treated sewage sludge); deposition of manure on soils by domesticated animals in pastures, range,
33 and paddocks (PRP) (i.e., unmanaged manure); retention of crop residues (N-fixing legumes and non-legume crops
13 Nitrification and denitrification are driven by the activity of microorganisms in soils. Nitrification is the aerobic microbial
oxidation of ammonium (NH4+) to nitrate (NOs ), and denitrification is the anaerobic microbial reduction of nitrate to N2. Nitrous
oxide is a gaseous intermediate product in the reaction sequence of nitrification and denitrification.
14 Asymbiotic N fixation is the fixation of atmospheric N2 by bacteria living in soils that do not have a direct relationship with
plants.
5-28 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 and forages); and drainage of organic soils15 (i.e., Histosols) (IPCC 2006). Additionally, agricultural soil management
2 activities, including irrigation, drainage, tillage practices, cover crops, and fallowing of land, can influence N
3 mineralization from soil organic matter and levels of asymbiotic N fixation. Indirect emissions of N2O occur when N
4 is transported from a site and is subsequently converted to N2O; there are two pathways for indirect emissions: (1)
5 volatilization and subsequent atmospheric deposition of applied/mineralized N, and (2) surface runoff and leaching
6 of applied/mineralized N into groundwater and surface water.16 Direct and indirect emissions from agricultural
7 lands are included in this section (i.e., cropland and grassland as defined in Section 6.1 Representation of the U.S.
8 Land Base). Nitrous oxide emissions from Forest Land and Settlements soils are found in Sections 6.2 and 6.10,
9 respectively.
15 Drainage of organic soils in former wetlands enhances mineralization of N-rich organic matter, thereby increasing N20
emissions from these soils.
16 These processes entail volatilization of applied or mineralized N as NH3 and NOx, transformation of these gases in the
atmosphere (or upon deposition), and deposition of the N primarily in the form of particulate NH4+, nitric acid (HNO3), and NOx.
In addition, hydrological processes lead to leaching and runoff of NO3" that is converted to N20 in aquatic systems, e.g.,
wetlands, rivers, streams and lakes. Note: N20 emissions are not estimated for aquatic systems associated with N inputs from
terrestrial systems in order to avoid double-counting.
Agriculture 5-29
-------
1 Figure 5-4: Sources and Pathways of N that Result in N2O Emissions from Agricultural Soil
2 Management
Sources and Pathways of N that Result in N2O Emissions from Agricultural Soil Management
- ft " ~
I TfiTtuzER Synthetic N Fertilizers
Synthetic N fertilizer applied to Soil
Organic
Amendments
Includes both commerciiJ and
norvco/n mercisl fertilizers (i.e.,
animal manure compost
sewage sludge tankage «c)
Urine and Dung fr om
Grazing Animals
Manure deposited on pasture range
and paddock ij 1.
Crop Residues
Includes above- and belowground
residues for al I crops (non-N and N-
fixing (and from perennial forage
crops and pastures following renewal
Mineralization of
Soil Organic Matter
Includes N converted to mineral form
upon decomposition of soil organic
matter
Asymbiotic Fixation
Fixation of atmospheric N2 by bacteria
living m soilsthat do not have a direct
relationship with plants
This graphic illustrates the sources and pathways of nitrogen that result
in direct and Indirect N,0 emissions from soils using the methodologies
described in this Inventory. Emission pathways are shown with arrows.
On the lower right-hand side is a cut-away view of a representative
section of a managed soil; histosol cultivation is represented here.
N Flows:
0
N Inputs to
Managed Soils
Direct N*0
Emissions
N Volatilization
and Deposition
0
Indirect N2O
Emissions
Histosol
Cultivation
Agricultural soils produce the majority of N2O emissions in the United States. Estimated emissions in 2021 are
285.2 MMT CO2 Eq. (1,076 kt) (see Table 5-16 and Table 5-17). Annual N2O emissions from agricultural soils are 2.5
percent greater in 2021 compared to 1990, but emissions fluctuated between 1990 and 2021 due to inter-annual
variability largely associated with weather patterns, synthetic fertilizer use, and crop production. From 1990 to
2021, cropland accounted for 69 percent of total direct emissions on average from agricultural soil management,
while grassland accounted for 31 percent. On average, 78 percent of indirect emissions are from croplands and 22
5-30 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 percent from grasslands. Estimated direct and indirect N2O emissions by sub-source category are shown in Table
2 5-18 and Table 5-19.
3 Table 5-16: N2O Emissions from Agricultural Soils (MMT CO2 Eq.)
Activity
1990
2005
2017
2018
2019
2020
2021
Direct
252.6
255.9
270.4
282.0
268.4
253.6
257.7
Cropland
172.7
175.9
188.3
195.6
184.7
177.3
178.4
Grassland
79.9
80.1
82.1
86.4
83.7
76.4
79.3
Indirect
25.8
24.8
28.2
30.1
29.8
25.6
27.5
Cropland
19.9
19.1
22.4
23.7
23.5
20.0
21.9
Grassland
5.9
5.7
5.9
6.4
6.3
5.6
5.7
Total
278.4
280.8
298.7
312.1
298.2
279.3
285.2
Notes: Estimates for 2021 are based on a data splicing method, except for other organic N
amendments that are based on a data splicing method for 2018 to 2021 (See Methodology
section). Totals may not sum due to independent rounding. Quality control procedures uncovered
minor errors in the estimates that will be corrected in the final version of this Inventory.
4 Table 5-17: N2O Emissions from Agricultural Soils (kt)
Activity
1990
2005
2017
2018
2019
2020
2021
Direct
953
966
1,021
1,064
1,013
957
972
Cropland
651.5
663.6
710.7
738.2
697.1
668.9
673.2
Grassland
301.5
302.2
309.8
325.9
315.7
288.3
299.1
Indirect
97
94
107
114
112
97
104
Cropland
75.0
72.1
84.5
89.4
88.5
75.6
82.6
Grassland
22.3
21.7
22.1
24.3
23.8
21.2
21.4
Total
1,050
1,060
1,127
1,178
1,125
1,054
1,076
Notes: Estimates for 2021 are based on a data splicing method, except for other organic N that are based
on a data splicing method for 2018 to 2021 (See Methodology section). Totals may not sum due to
independent rounding. Quality control procedures uncovered minor errors in the estimates that will be
corrected in the final version of this Inventory.
5 Table 5-18: Direct N2O Emissions from Agricultural Soils by Land Use Type and N Input Type
6 (MMT CO2 Eq.)
Activity
1990
2005
2017
2018
2019
2020
2021
Cropland
172.6
175.8
188.1
195.6
184.7
177.3
178.3
Mineral Soils
169.2
172.5
185.1
192.6
181.8
174.3
175.4
Synthetic Fertilizer
58.0
61.8
65.6
64.9
62.0
60.6
59.5
Organic Amendment3
11.1
12.0
13.6
13.6
13.5
13.7
13.9
Residue Nb
26.4
26.2
26.2
28.5
25.3
28.8
24.6
Mineralization and
Asymbiotic Fixation
73.6
72.5
79.7
85.6
81.0
71.3
77.4
Drained Organic Soils
3.4
3.2
3.0
3.0
2.9
2.9
2.9
Grassland
80.0
80.2
82.4
86.4
83.7
76.4
79.4
Mineral Soils
77.7
77.9
80.1
84.2
81.4
74.1
77.1
Synthetic Fertilizer
+
+
+
+
+
+
+
PRP Manure
12.3
10.8
10.2
10.8
10.1
9.8
10.3
Managed Manurec
+
+
+
+
+
+
+
Biosolids (i.e., treated
Sewage Sludge)
0.2
0.4
0.6
0.6
0.6
0.6
0.6
Residue Nd
21.4
22.4
22.7
22.3
22.5
22.7
20.8
Mineralization and
Asymbiotic Fixation
43.8
44.3
46.6
50.5
48.3
41.0
45.3
Agriculture 5-31
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Drained Organic Soils 23 22 23 22 22 23 2.3
Total 252.6 255.9 270.4 282.0 268.4 253.6 257.7
+ Does not exceed 0.05 MMT C02 Eq.
a Organic amendment inputs include managed manure, daily spread manure, and commercial organic
fertilizers (i.e., dried blood, dried manure, tankage, compost, and other).
b Cropland residue N inputs include N in unharvested cover crops as well as harvested crops.
c Managed manure inputs include managed manure and daily spread manure amendments that are applied to
grassland soils.
d Grassland residue N inputs include residual biomass, both legumes and grasses, that is ungrazed and
becomes dead organic matter.
Notes: Estimates for 2021 are based on a data splicing method, except for other organic N amendments that
are based on a data splicing method for 2018 to 2021 (See Methodology section). Totals may not sum due to
independent rounding. Quality control procedures uncovered minor errors in the estimates that will be
corrected in the final version of this Inventory.
Table 5-19: Indirect N2O Emissions from Agricultural Soils (MMT CO2 Eq.)
Activity
1990
2005
2017
2018
2019
2020
2021
Cropland
19.9
19.1
22.4
23.7
23.5
20.0
21.9
Volatilization & Atm.
Deposition
6.3
6.6
7.1
7.6
6.7
7.2
7.1
Surface Leaching & Run-Off
13.6
12.5
15.3
16.1
16.7
12.9
14.8
Grassland
5.9
5.7
5.9
6.4
6.3
5.6
5.7
Volatilization & Atm.
Deposition
3.5
3.5
3.4
3.5
3.3
3.1
3.2
Surface Leaching & Run-Off
2.4
2.2
2.4
3.0
3.0
2.6
2.5
Total
25.8
24.8
28.2
30.1
29.8
25.6
27.5
Notes: Estimates for 2021 are based on a data splicing method, except for other organic N amendments
that are based on a data splicing method for 2018 to 2021 (See Methodology section). Totals may not sum
due to independent rounding. Quality control procedures uncovered minor errors in the estimates that
will be corrected in the final version of this Inventory.
Figure 5-5 and Figure 5-6 show regional patterns for direct N2O emissions. Figure 5-7 and Figure 5-8 show indirect
N2O emissions from volatilization, and Figure 5-9 and Figure 5-10 show the indirect N2O emissions from leaching
and runoff in croplands and grasslands, respectively.
Direct N2O emissions from croplands occur throughout all of the cropland regions but tend to be high in the
Midwestern Corn Belt Region (particularly, Illinois, Iowa, Kansas, Minnesota, Nebraska), where a large portion of
the land is used for growing highly fertilized corn and N-fixing soybean crops (see Figure 5-5). There are high
emissions from the Southeastern region, and portions of the Great Plains, such as North Dakota and Montana.
Emissions are also high in the Lower Mississippi River Basin from Missouri to Louisiana, and highly productive
irrigated areas, such as Platte River, which flows from Colorado and Wyoming through Nebraska, Snake River
Valley in Idaho, and the Central Valley in California. Direct emissions are low in mountainous regions of the Eastern
United States because only a small portion of land is cultivated, and in much of the Western United States where
rainfall and access to irrigation water are limited, in addition to mountainous, which are generally not suitable for
crop production.
Direct N2O emissions from grasslands are more evenly distributed throughout the United States compared to
emissions from cropland due to suitable areas for grazing in most regions (see Figure 5-6). Total emissions tend be
highest in the Great Plains and western United States where a large proportion of the land is dominated by
grasslands with cattle and sheep grazing (particularly Texas, Montana, New Mexico, Oklahoma, and South Dakota).
5-32 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
Figure 5-5: Croplands, 2020 Annual Direct N2O Emissions Estimated Using the Tier 3
DayCent Model
4 Note: Only national-scale emissions are estimated for 2021 using a splicing method, and therefore the fine-scale
5 emission patterns in this map are based on Inventory data from 2020.
6 Figure 5-6: Grasslands, 2020 Annual Direct N2O Emissions Estimated Using the Tier 3
7 DayCent Model
9 Note: Only national-scale emissions are estimated for 2021 using a splicing method, and therefore the fine-scale
10 emission patterns in this map are based on Inventory data from 2020.
Agriculture 5-33
-------
1 Indirect N2O emissions from volatilization in croplands have a similar pattern as the direct N?.0 emissions with
2 higher emissions in the Midwestern Corn Belt, Lower Mississippi River Basin, Southeastern region, and parts of the
3 Great Plains and irrigated areas of the Western United States. Indirect N2O emissions from volatilization in
4 grasslands are higher in the Eastern and Central United States, along with relatively small areas scattered around
5 the Western United States. The higher emissions are partly due to large additions of PRP manure N, which in turn,
6 stimulates NH3 volatilization.
7 Indirect N2O emissions from surface runoff and leaching of applied/mineralized N in croplands is highest in the
8 Midwestern Corn Belt. There are also relatively high emissions associated with N management in the Lower
9 Mississippi River Basin, Piedmont region of the Southeastern United States and the Mid-Atlantic states. In addition,
10 areas of high emissions occur in portions of the Great Plains that have relatively large areas of irrigated croplands
11 with high leaching rates of applied/mineralized N. Indirect N2O emissions from surface runoff and leaching of
12 applied/mineralized N in grasslands are higher in the eastern United States and coastal Northwest region. These
13 regions have greater precipitation and higher levels of leaching and runoff compared to arid to semi-arid regions in
14 the Western United States.
15 Figure 5-7; Croplands, 2020 Annual Indirect N2O Emissions from Volatilization Using the
16 Tier 3 DayCent Model
18 Note: Only national-scale emissions are estimated for 2021 using a splicing method, and therefore the fine-scale
19 emission patterns in this map are based on Inventory data from 2020.
5-34 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Figure 5-8: Grasslands, 2020 Annual Indirect N2O Emissions from Volatilization Using the
2 Tier 3 DayCent Model
4 Note: Only national-scale emissions are estimated for 2021 using a splicing method, and therefore the fine-scale
5 emission patterns in this map are based on Inventory data from 2020.
6 Figure 5-9: Croplands, 2020 Annual Indirect N2O Emissions from Leaching and Runoff Using
7 the Tier 3 DayCent Model
9 Note: Only national-scale emissions are estimated for 2021 using a splicing method, and therefore the fine-scale
10 emission patterns in this map are based on Inventory data from 2020.
Agriculture 5-35
-------
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Figure 5-10: Grasslands, 2020 Annual Indirect N2O Emissions from Leaching and Runoff
Using the Tier 3 DayCent Model
Note: Only national-scale emissions are estimated for 2021 using a splicing method, and therefore the fine-scale
emission patterns in this map are based on Inventory data from 2020.
Methodology and Time-Series Consistency
The 2006IPCC Guidelines (IPCC 2006) divide emissions from the agricultural soil management source category into
five components, including (1) direct emissions from N additions to cropland and grassland mineral soils from
synthetic fertilizers, biosolids (i.e., treated sewage sludge), crop residues (legume N-fixing and non-legume crops),
and organic amendments; (2) direct emissions from soil organic matter mineralization due to land use and
management change; (3) direct emissions from drainage of organic soils in croplands and grasslands; (4) direct
emissions from soils due to manure deposited by livestock on PRP grasslands; and (5) indirect emissions from soils
and water from N additions and manure deposition to soils that lead to volatilization, leaching, or runoff of N and
subsequent conversion to N2O.
In this source category, the United States reports on all croplands, as well as all managed grasslands, whereby
anthropogenic greenhouse gas emissions are estimated in a manner consistent with the managed land concept
(IPCC 2006), including direct and indirect N2O emissions from asymbiotic fixation17 and mineralization of N
associated with decomposition of soil organic matter and residues. One recommendation from IPCC (2006) that
has not been completely adopted is the estimation of emissions from grassland pasture renewal, which involves
occasional plowing to improve forage production in pastures. Currently no data are available to address pasture
renewal.
In addition, estimates of N2O emissions from managed croplands and grasslands are not available for Alaska and
Hawaii except for managed manure and PRP N, and biosolid additions for Alaska, and managed manure and PRP N,
17 W inputs from asymbiotic N fixation are not directly addressed in 2006 IPCC Guidelines, but are a component of the N inputs
and total emissions from managed lands and are included in the Tier 3 approach developed for this source.
5-36 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
biosolids additions, and crop residue for Hawaii. There is a planned improvement to include the additional sources
of emissions in a future inventory.
Direct N2O Emissions
The methodology used to estimate direct N2O emissions from agricultural soil management in the United States is
based on a combination of IPCCTier 1 and 3 approaches, along with application of a splicing method for latter
years in the Inventory time series (IPCC 2006; Del Grosso et al. 2010). A Tier 3 process-based model (DayCent) is
used to estimate direct emissions from a variety of crops that are grown on mineral (i.e., non-organic) soils, as well
as the direct emissions from non-federal grasslands except for applications of biosolids (i.e., treated sewage
sludge) (Del Grosso et al. 2010). The Tier 3 approach has been specifically designed and tested to estimate N2O
emissions in the United States, accounting for more of the environmental and management influences on soil N2O
emissions than the IPCC Tier 1 method (see Box 5-3 for further elaboration). Moreover, the Tier 3 approach
addresses direct N2O emissions and soil C stock changes from mineral cropland soils in a single analysis. Carbon
and N dynamics are linked in plant-soil systems through biogeochemical processes of microbial decomposition and
plant production (McGill and Cole 1981). Coupling the two source categories (i.e., agricultural soil C and N2O) in a
single inventory analysis ensures that there is consistent activity data and treatment of the processes, and
interactions are considered between C and N cycling in soils.
Crop and land use histories are based on the USDA National Resources Inventory (NRI) (USDA-NRCS 2020), and
extended through 2020 using the USDA-NASS Crop Data Layer Product (USDA-NASS 2021, Johnson and Mueller
2010). The areas have been modified in the original NRI survey through a process in which the Forest Inventory
and Analysis (FIA) survey data and the National Land Cover Dataset (Yang et al. 2018) are harmonized with the NRI
data. This process ensures that the land use areas are consistent across all land use categories (See Section 6.1,
Representation of the U.S. Land Base for more information).
The NRI is a statistically-based sample of all non-federal land,18 and includes 364,334 survey locations on
agricultural land for the conterminous United States that are included in the Tier 3 method. The Tier 1 approach is
used to estimate the emissions from 161,161 locations in the NRI survey across the time series, which are
designated as cropland or grassland (discussed later in this section). Each survey location is associated with an
"expansion factor" that allows scaling of N2O emissions from NRI survey locations to the entire country (i.e., each
expansion factor represents the amount of area with the same land-use/management history as the survey
location). Each NRI survey location was sampled on a 5-year cycle from 1982 until 1997. For cropland, data were
collected in 4 out of 5 years in the cycle (i.e., 1979 through 1982,1984 through 1987,1989 through 1992, and 1994
through 1997). In 1998, the NRI program began collecting annual data, which are currently available through 2017
(USDA-NRCS 2020). For 2018-2020, the time series is extended with the crop data provided in USDA-NASS CDL
(USDA-NASS 2021). CDL data have a 30 to 58 m spatial resolution, depending on the year. NRI survey locations are
overlaid on the CDL in a geographic information system, and the crop types are extracted to extend the cropping
histories for the inventory analysis.
Box 5-3: Tier 1 vs. Tier 3 Approach for Estimating N2O Emissions
The IPCC (2006) Tier 1 approach is based on multiplying activity data on different N inputs (i.e., synthetic
fertilizer, manure, N fixation, etc.) by the appropriate default IPCC emission factors to estimate N2O emissions
on an input-by-input basis. The Tier 1 approach requires a minimal amount of activity data, readily available in
most countries (e.g., total N applied to crops); calculations are simple; and the methodology is highly
transparent. In contrast, the Tier 3 approach developed for this Inventory is based on application of a process-
based model (i.e., DayCent) that represents the interaction of N inputs, land use and management, as well as
environmental conditions at specific locations, such as freeze-thaw effects that generate pulses of N2O
18 The NRI survey does include sample points on federal lands, but the program does not collect data from those sample
locations.
Agriculture 5-37
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
emissions (Wagner-Riddle et al. 2017, Del Grosso et al. 2022). Consequently, the Tier 3 approach accounts for
land-use and management impacts and their interaction with environmental factors, such as weather patterns
and soil characteristics, in a more comprehensive manner, which will enhance or dampen anthropogenic
influences. However, the Tier 3 approach requires more detailed activity data (e.g., crop-specific N fertilization
rates), additional data inputs (e.g., daily weather, soil types), and considerable computational resources and
programming expertise. The Tier 3 methodology is less transparent, and thus it is critical to evaluate the output
of Tier 3 methods against measured data in order to demonstrate that the method is an improvement over
lower tier methods for estimating emissions (IPCC 2006). Another important difference between the Tier 1 and
Tier 3 approaches relates to assumptions regarding N cycling. Tier 1 assumes that N added to a system is subject
to N2O emissions only during that year and cannot be stored in soils and contribute to N2O emissions in
subsequent years. This is a simplifying assumption that may create bias in estimated N2O emissions for a specific
year. In contrast, the process-based model in the Tier 3 approach includes the legacy effect of N added to soils
in previous years that is re-mineralized from soil organic matter and emitted as N2O during subsequent years.
DayCent is used to estimate N2O emissions associated with production of alfalfa hay, barley, corn, cotton, dry
beans, grass hay, grass-clover hay, lentils, oats, onions, peanuts, peas, potatoes, rice, sorghum, soybeans, sugar
beets, sunflowers, sweet potatoes, tobacco, tomatoes, and wheat, but is not applied to estimate N2O emissions
from other crops or rotations with other crops,19 such as sugarcane, some vegetables, and perennial/horticultural
crops. Areas that are converted between agriculture (i.e., cropland and grassland) and other land uses, such as
forest land, wetland and settlements, are not simulated with DayCent. DayCent is also not used to estimate
emissions from land areas with very gravelly, cobbly, or shaley soils in the topsoil (greater than 35 percent by
volume in the top 30 cm of the soil profile), or to estimate emissions from drained organic soils (Histosols). The Tier
3 method has not been fully tested for estimating N2O emissions associated with these crops and rotations, land
uses, as well as organic soils or cobbly, gravelly, and shaley mineral soils. In addition, federal grassland areas are
not simulated with DayCent due to limited activity data on land use histories. For areas that are not included in the
DayCent simulations, Tier 1 methods are used to estimate emissions, including (1) direct emissions from N inputs
for crops on mineral soils that are not simulated by DayCent; (2) direct emissions from PRP N additions on federal
grasslands; (3) direct emissions for land application of biosolids (i.e., treated sewage sludge) to soils; and (4) direct
emissions from drained organic soils in croplands and grasslands.
A splicing method is used to estimate soil N2O emissions for 2021 at the national scale because new NRI activity
data have not been incorporated into the analysis for those years. Specifically, linear regression models with
autoregressive moving-average (ARMA) errors (Brockwell and Davis 2016) are used to estimate the relationship
between surrogate data and the 1990 to 2020 emissions that are derived using the Tier 3 method. Surrogate data
for these regression models includes corn and soybean yields from USDA-NASS statistics,20 and weather data from
the PRISM Climate Group (PRISM 2022). For the Tier 1 method, a linear-time series model is used to estimate
emissions for 2021 without surrogate data. In addition, the linear time series model is used to estimate emissions
data for 2018 to 2021 for other organic N amendments (i.e., commercial organic fertilizer) due to a gap in the
activity data during the latter part of the time series (TVA 1991 through 1994; AAPFCO 1995 through 2022). See
Box 5-4 for more information about the splicing method. Emission estimates for years with imputed data will be
recalculated in future Inventory reports when new NRI data and other organic amendment N data are available.
19 A small proportion of the major commodity crop production, such as corn and wheat, is included in the Tier 1 analysis
because these crops are rotated with other crops or land uses (e.g., forest lands) that are not simulated by DayCent.
20 See https://quickstats.nass.usda.gov/.
5-38 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
Box 5-4: Data Splicing Method
An approach to extend the time series is needed for Agricultural Soil Management because there are typically
activity data gaps at the end of the time series. This is mainly because the NRI survey program, which provides
critical information for estimating greenhouse gas emissions and removals, does not release data every year.
Splicing methods have been used to impute missing data at the end of the emission time series for both the Tier
1 and 3 methods. Specifically, a linear regression model with autoregressive moving-average (ARMA) errors
(Brockwell and Davis 2016) is used to estimate emissions based on the emissions data that has been compiled
using the inventory methods described in this section. The model to extend the time series is given by the
equation:
Y = xp + £,
where Y is the response variable (e.g., soil nitrous oxide), xp for the Tier 3 method contains specific surrogate
data depending on the response variable, and £ is the remaining unexplained error. Models with a variety of
surrogate data were tested, including commodity statistics, weather data, or other relevant information. The
term xp for the Tier 1 method only contains year as a predictor of emission patterns over the time series
(change in emissions per year), and therefore, is a linear time series model with no surrogate data. Parameters
are estimated using standard statistical techniques, and used in the model described above to predict the
missing emissions data.
A critical issue with splicing methods is to account for the additional uncertainty introduced by predicting
emissions without compiling the full inventory. Specifically, uncertainty will increase for years with imputed
estimates based on the splicing methods, compared to those years in which the full inventory is compiled. This
additional uncertainty is quantified within the model framework using a Monte Carlo approach. Consequently,
the uncertainty from the original inventory data is combined with the uncertainty in the data splicing model.
The approach requires estimating parameters in the data splicing models in each Monte Carlo simulation for the
full inventory (i.e., the surrogate data model is refit with the draws of parameters values that are selected in
each Monte Carlo iteration, and used to produce estimates with inventory data). Therefore, the data splicing
method generates emissions estimates from each surrogate data model in the Monte Carlo analysis, which are
used to derive confidence intervals in the estimates for the missing emissions data. Furthermore, the 95 percent
confidence intervals are estimated using the 3 sigma rules assuming a unimodal density (Pukelsheim 1994).
2
3 Tier 3 Approach for Mineral Cropland Soils
4 The DayCent biogeochemical model (Parton et al. 1998; Del Grosso et al. 2001 and 2011) is used to estimate direct
5 N2O emissions from mineral cropland soils that are managed for production of a wide variety of crops (see list in
6 previous section) based on the crop histories in the 2017 NRI (USDA-NRCS 2020), and extended through 2020 using
7 CDL (USDA-NASS 2021). Crops simulated by DayCent are grown on approximately 85 percent of total cropland area
8 in the United States. The model simulates net primary productivity (NPP) using the NASA-CASA production
9 algorithm MODIS Enhanced Vegetation Index (EVI) products, MOD13Q1 and MYD13Q121 (Potter et al. 1993, 2007).
10 The model simulates soil temperature and water dynamics, using daily weather data from a 4-kilometer gridded
11 product developed by the PRISM Climate Group (2022), and soil attributes from the Soil Survey Geographic
12 Database (SSURGO) (Soil Survey Staff 2020). DayCent is used to estimate direct N2O emissions due to mineral N
13 available from the following sources: (1) application of synthetic fertilizers; (2) application of livestock manure; (3)
21 NPP is estimated with the NASA-CASA algorithm for most of the cropland that is used to produce major commodity crops in
the central United States from 2000 to 2020. Other regions and years prior to 2000 are simulated with a method that
incorporates water, temperature, and moisture stress on crop production (see Metherell et al. 1993), but does not incorporate
the additional information about crop condition provided with remote sensing data.
Agriculture 5-39
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
retention of crop residues in the field for N-fixing legumes and non-legume crops and subsequent mineralization of
N during microbial decomposition (i.e., leaving residues in the field after harvest instead of burning or collecting
residues); (4) mineralization of N from decomposition of soil organic matter; and (5) asymbiotic fixation.
Management activity data from several sources supplement the activity data from the NRI. The USDA-NRCS
Conservation Effects and Assessment Project (CEAP) provides data on a variety of cropland management activities,
and is used to inform the inventory analysis about tillage practices, mineral fertilization, manure amendments,
cover crop management, as well as planting and harvest dates (USDA-NRCS 2022; USDA-NRCS 2018; USDA-NRCS
2012). CEAP data are collected at a subset of NRI survey locations, and currently provide management information
from approximately 2002 to 2006 and 2013 to 2016. These data are combined with other datasets in an
imputation analysis. This imputation analysis is comprised of three steps: a) determine the trends in management
activity across the time series by combining information from several datasets (discussed below); b) use Gradient
Boosting (Friedman 2001) to determine the likely management practice at a given NRI survey location; and c)
assign management practices from the CEAP survey to the specific NRI locations using a predictive mean matching
method for certain variables that are adapted to reflect the trending information (Little 1988, van Buuren 2012).
Gradient boosting is a machine learning technique used in regression and classification tasks, among others. It
combines predictions from multiple weak prediction models and outperforms many complicated machine learning
algorithms. It makes the best predictions at specific NRI survey locations or at state or region level models. The
predictive mean matching method identifies the most similar management activity recorded in the CEAP surveys
that match the prediction from the gradient boosting algorithm. The matching ensures that imputed management
activities are realistic for each NRI survey location, and not odd or physically unrealizable results that could be
generated by the gradient boosting. There are six complete imputations of the management activity data using
these methods.
To determine trends in mineral fertilization and manure amendments, CEAP data are combined with information
on fertilizer use and rates by crop type for different regions of the United States from the USDA Economic
Research Service. The data collection program was known as the Cropping Practices Surveys through 1995 (USDA-
ERS 1997), and is now part of data collection known as the Agricultural Resource Management Surveys (ARMS)
(USDA-ERS 2020). Additional data on fertilization practices are compiled through other sources particularly the
National Agricultural Statistics Service (USDA-NASS 1992,1999, 2004). To determine the trends in tillage
management, CEAP data are combined with Conservation Technology Information Center data between 1989 and
2004 (CTIC 2004) and OpTIS Data Product22 for 2008 to 2020 (Hagen et al. 2020). The CTIC data are adjusted for
long-term adoption of no-till agriculture (Towery 2001). For cover crops, CEAP data are combined with information
from USDA Census of Agriculture (USDA-NASS 2012, 2017) and the OpTIS data product23 (Hagen et al. 2020). It is
assumed that cover crop management was minimal prior to 1990 and the rates increased linearly over the decade
to the levels of cover crop management in the CEAP survey.
The IPCC method considers crop residue N and N mineralized from soil organic matter as activity data. However,
they are not treated as activity data in DayCent simulations because residue production, symbiotic N fixation (e.g.,
legumes), mineralization of N from soil organic matter, and asymbiotic N fixation are internally generated by the
model as part of the simulation. In other words, DayCent accounts for the influence of symbiotic N fixation,
mineralization of N from soil organic matter and crop residue retained in the field, and asymbiotic N fixation on
N2O emissions, but these are not model inputs.
The N2O emissions from crop residues are reduced by approximately 3 percent (the assumed average burned
portion for crop residues in the United States) to avoid double counting associated with non-CC>2 greenhouse gas
emissions from agricultural residue burning. Estimated levels of residue burning are based on state inventory data
(ILENR 1993; Oregon Department of Energy 1995; Noller 1996; Wisconsin Department of Natural Resources 1993;
Cibrowski 1996).
22 OpTIS data on tillage practices provided by Regrow Agriculture, Inc.
23 OpTIS data on cover crop management provided by Regrow Agriculture, Inc.
5-40 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
Uncertainty in the emission estimates from DayCent is associated with input uncertainty due to missing
management data in the NRI survey that is imputed from other sources; model uncertainty due to incomplete
specification of C and N dynamics in the DayCent model parameters and algorithms; and sampling uncertainty
associated with the statistical design of the NRI survey. To assess input uncertainty, C and N dynamics at each NRI
survey location are simulated six times using the imputation product and other model driver data. Uncertainty in
parameterization and model algorithms are determined using a structural uncertainty estimator derived from
fitting a linear mixed-effect model (Ogle et al. 2007; Del Grosso et al. 2010). Sampling uncertainty is assessed using
NRI replicate sampling weights. These data are combined in a Monte Carlo stochastic simulation with 1,000
iterations for 1990 through 2020. For each iteration, there is a random selection of management data from the
imputation product (select one of the six imputations), random selection of parameter values and random effects
for the linear mixed-effect model (i.e., structural uncertainty estimator), and random selection of a set of survey
weights from the replicates associated with the NRI survey design.
In order to ensure time-series consistency, the DayCent model is applied from 1990 to 2020, and a linear
extrapolation method is used to approximate emissions for 2021 based on the pattern in emissions data from 1990
to 2020 (See Box 5-4). The pattern is determined using a linear regression model with moving-average (ARMA)
errors. Linear extrapolation is a standard data splicing method for approximating missing values at the end of an
inventory time series (IPCC 2006). The time series will be updated with the Tier 3 method in the future as new
activity data are incorporated into the analysis.
Nitrous oxide emissions from managed agricultural lands are the result of interactions among anthropogenic
activities (e.g., N fertilization, manure application, tillage) and other driving variables, such as weather and soil
characteristics. These factors influence key processes associated with N dynamics in the soil profile, including
immobilization of N by soil microbial organisms, decomposition of organic matter, plant uptake, leaching, runoff,
and volatilization, as well as the processes leading to N2O production (nitrification and denitrification). It is not
possible to partition N2O emissions into each anthropogenic activity directly from model outputs due to the
complexity of the interactions (e.g., N2O emissions from synthetic fertilizer applications cannot be distinguished
from those resulting from manure applications). To approximate emissions by activity, the amount of synthetic N
fertilizer added to the soil, or mineral N made available through decomposition of soil organic matter and plant
litter, as well as asymbiotic fixation of N from the atmosphere, is determined for each N source and then divided
by the total amount of mineral N in the soil according to the DayCent model simulation. For 2021, the contribution
of each N source is based on the average of values that are estimated for 2018 to 2020. The percentages are then
multiplied by the total of direct N2O emissions in order to approximate the portion attributed to N management
practices. This approach is only an approximation because it assumes that all N made available in soil has an equal
probability of being released as N2O, regardless of its source, which is unlikely to be the case (Delgado et al. 2009).
However, this approach allows for further disaggregation of emissions by source of N, which is valuable for
reporting purposes and is analogous to the reporting associated with the IPCC (2006) Tier 1 method, in that it
associates portions of the total soil N2O emissions with individual sources of N.
Tier 1 Approach for Mineral Cropland Soils
The IPCC (2006) Tier 1 methodology is used to estimate direct N2O emissions for mineral cropland soils that are not
simulated by DayCent (e.g., DayCent has not been parametrized to simulate all crop types and some soil types such
as Histosols). For the Tier 1 method, estimates of direct N2O emissions from N applications are based on mineral
soil N that is made available from the following practices: (1) the application of synthetic commercial fertilizers; (2)
application of managed manure and non-manure commercial organic fertilizers; and (3) decomposition and
mineralization of nitrogen from above- and below-ground crop residues in agricultural fields (i.e., crop biomass
that is not harvested). Non-manure commercial organic amendments are only included in the Tier 1 analysis
because these data are not available at the county-level, which is necessary for the DayCent simulations.
Consequently, all commercial organic fertilizer, as well as manure that is not added to crops in the DayCent
simulations, are included in the Tier 1 analysis. The following sources are used to derive activity data:
Agriculture 5-41
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
• A process-of-elimination approach is used to estimate synthetic N fertilizer additions for crop areas that are
not simulated by DayCent. The total amount of fertilizer used on farms has been estimated at the county-level
by the USGS using sales records from 1990 to 2012 (Brakebill and Gronberg 2017). For 2013 through 2017,
fertilizer sales data from AAPFCO (AAPFCO 2013 through 2022)24 after adjusting for the proportion of on-farm
application to determine the amount applied to crops. The amount of fertilizer applied after 2017 is estimated
using the data splicing method described in Box 5-4 for the linear time series model. Then the portion of
fertilizer applied to crops and grasslands simulated by DayCent is subtracted from the on-farm sales data (see
Tier 3 Approach for Mineral Cropland Soils and Direct N2O Emissions from Grassland Soils sections for
information on data sources), and the remainder of the total fertilizer used on farms is assumed to be applied
to crops that are not simulated by DayCent. At a minimum, 3 percent of state-level on-farm fertilizer sales are
assumed to be applied to cropland in the Tier 1 method.
• Similarly, a process-of-elimination approach is used to estimate manure N additions for crops that are not
simulated by DayCent. The total amount of manure available for land application to soils has been estimated
with methods described in the Manure Management section (Section 5.2) and annex (Annex 3.11). The
amount of manure N applied in the Tier 3 approach to crops and grasslands is subtracted from total annual
manure N available for land application (see Tier 3 Approach for Mineral Cropland Soils and Direct N2O
Emissions from Grassland Soils sections for information on data sources). This difference is assumed to be
applied to crops that are not simulated by DayCent.
• Commercial organic fertilizer additions are based on organic fertilizer consumption statistics through 20 1725,
which are converted from mass of fertilizer to units of N using average organic fertilizer N content, ranging
between 2.3 to 4.2 percent across the time series (TVA 1991 through 1994; AAPFCO 1995 through 2022).
Commercial fertilizers include dried manure and biosolids (i.e., treated sewage sludge), but the amounts are
removed from the commercial fertilizer data to avoid double counting26 with the manure N dataset described
above and the biosolids (i.e., treated sewage sludge) amendment data discussed later in this section.
• Crop residue N is derived by combining amounts of above- and below-ground biomass, which are determined
based on NRI crop area data (USDA-NRCS 2020), as extended using the CDL data (USDA-NASS 2021), crop
production yield statistics (USDA-NASS 2022), dry matter fractions (IPCC 2006), linear equations to estimate
above-ground biomass given dry matter crop yields from harvest (IPCC 2006), ratios of below-to-above-ground
biomass (IPCC 2006), and N contents of the residues (IPCC 2006). N inputs from residue were reduced by 3
percent to account for average residue burning portions in the United States.
The total amounts of soil mineral N from applied synthetic and organic fertilizers, manure N additions and crop
residues are multiplied by the IPCC (2006) default emission factor to derive an estimate of direct N2O emissions
using the Tier 1 method. Further elaboration on the methodology and data used to estimate N2O emissions from
mineral soils are described in Annex 3.12.
In order to ensure time-series consistency, the Tier 1 methods are applied from 1990 to 2020, and a linear
extrapolation method27 is used to approximate emissions for 2021 based on the emission patterns between 1990
and 2020 (See Box 5-4). The exceptions include crop residue N which is estimating using the Tier 1 method for
24 The fertilizer consumption data in AAPFCO are recorded in "fertilizer year" totals, (i.e., July to June), but are converted to
calendar year totals. This is done by assuming that approximately 35 percent of fertilizer usage occurred from July to December
and 65 percent from January to June (TVA 1992b).
25 Soil N20 emissions are imputed using data splicing methods for commercial fertilizers, i.e., other organic fertilizers, after
2017 because the activity data are not available.
26 Commercial organic fertilizers include dried blood, tankage, compost, and other, but the dried manure and biosolids (i.e.,
treated sewage sludge) are also included in other datasets in this Inventory. Consequently, the proportions of dried manure and
biosolids, which are provided in the reports (TVA 1991 through 1994; AAPFCO 1995 through 2022), are used to estimate the N
amounts in dried manure and biosolids. To avoid double counting, the resulting N amounts for dried manure and biosolids are
subtracted from the total N in commercial organic fertilizers before estimating emissions using the Tier 1 method.
5-42 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
1990 to 2021 with no linear extrapolation, and for other organic N fertilizers (i.e., commercial fertilizers), which are
estimated with linear time series model for 2018 to 2021 due to a gap in the activity data during the latter part of
the time series (TVA 1991 through 1994; AAPFCO 1995 through 2022). For the extrapolation, the emission pattern
is determined using a linear regression model with moving-average (ARMA) errors. Linear extrapolation is a
standard data splicing method for approximating missing values at the end of an inventory time series (IPCC 2006).
As with the Tier 3 method, the time series that is based on the splicing methods will be recalculated in a future
Inventory report with updated activity data.
Tier 1 and 3 Approaches from Mineral Grassland Soils
As with N2O emissions from croplands, the Tier 3 process-based approach with application of the DayCent model
and Tier 1 method described in IPCC (2006) are combined to estimate emissions from non-federal grasslands and
PRP manure N additions for federal grasslands, respectively. Grassland includes pasture and rangeland that
produce grass or mixed grass/legume forage primarily for livestock grazing. Rangelands are extensive areas of
native grassland that are not intensively managed, while pastures are seeded grassland (possibly following tree
removal) that may also have additional management, such as irrigation, fertilization, or inter-seeding legumes.
DayCent is used to simulate N2O emissions from NRI survey locations (USDA-NRCS 2020) on non-federal grasslands
resulting from manure deposited by livestock directly onto pastures and rangelands (i.e., PRP manure), N fixation
from legume seeding, managed manure amendments (i.e., manure other than PRP manure such as daily spread or
manure collected from other animal waste management systems such as lagoons and digesters), and synthetic
fertilizer application. Other N inputs are simulated within the DayCent framework, including N input from
mineralization due to decomposition of soil organic matter and N inputs from senesced grass litter, as well as
asymbiotic fixation of N from the atmosphere. The simulations used the same weather, soil, and synthetic N
fertilizer data as discussed under the Tier 3 Approach in the Mineral Cropland Soils section. Synthetic N fertilization
rates are based on data from the Carbon Sequestration Rural Appraisals (CSRA) conducted by the USDA-NRCS
(USDA-NRCS, unpublished data). The CSRA was a solicitation of expert knowledge from USDA-NRCS staff
throughout the United States to support the Inventory. Biological N fixation is simulated within DayCent, and
therefore is not an input to the model.
Manure N deposition from grazing animals in PRP systems (i.e., PRP manure N) is a key input of N to grasslands.
The amounts of PRP manure N applied on non-federal grasslands for each NRI survey location are based on the
amount of N excreted by livestock in PRP systems that is estimated in the Manure Management section (See
Section 5.2 and Annex 3.11). The total amount of N excreted in each county is divided by the grassland area to
estimate the N input rate associated with PRP manure. The resulting rates are a direct input into the DayCent
simulations. The N input is subdivided between urine and dung based on a 50:50 split. DayCent simulations of non-
federal grasslands accounted for approximately 71 percent of total PRP manure N in aggregate across the
country.28 The remainder of the PRP manure N in each state is assumed to be excreted on federal grasslands, and
the N2O emissions are estimated using the IPCC (2006) Tier 1 method.
Biosolids (i.e., treated sewage sludge) are assumed to be applied on grasslands29. Application of biosolids is
estimated from data compiled by EPA (1993,1999, 2003), McFarland (2001), and NEBRA (2007) (see Section 7.2
Wastewater Treatment for a detailed discussion of the methodology for estimating treated sewage sludge
available for land application application). Biosolids data are only available at the national scale, and it is not
possible to associate application with specific soil conditions and weather at NRI survey locations. Therefore,
DayCent could not be used to simulate the influence of biosolids on N2O emissions from grassland soils, and
consequently, emissions from biosolids are estimated using the IPCC (2006) Tier 1 method.
28 A small amount of PRP N (less than 1 percent) is deposited in grazed pasture that is in rotation with annual crops, and is
reported in the grassland N20 emissions.
29 A portion of biosolids may be applied to croplands, but there is no national dataset to disaggregate the amounts between
cropland and grassland.
Agriculture 5-43
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Soil N2O emission estimates from DayCent are adjusted using a structural uncertainty estimator accounting for
uncertainty in model algorithms and parameter values (Del Grosso et al. 2010). There is also sampling uncertainty
for the NRI survey that is propagated with replicate sampling weights associated with the survey. N2O emissions
for the PRP manure N deposited on federal grasslands and applied biosolids N are estimated using the Tier 1
method by multiplying the N input by the default emission factor. Emissions from manure N are estimated at the
state level and aggregated to the entire country, but emissions from biosolids N are calculated exclusively at the
national scale. Further elaboration on the methodology and data used to estimate N2O emissions from mineral
soils are described in Annex 3.12.
Soil N2O emissions and 95 percent confidence intervals are estimated for each year between 1990 and 2020 based
on the Tier 1 and 3 methods, except for biosolids (discussed below). In order to ensure time-series consistency,
emissions from 2021 are estimated using a splicing method as described in Box 5-4, with a linear extrapolation
based on the emission patterns in the 1990 to 2020 data. Linear extrapolation is a standard data splicing method
for approximating emissions at the end of a time series (IPCC 2006). As with croplands, estimates for 2021 will be
recalculated in a future Inventory when the activity data are updated. Biosolids application data are compiled
through 2021 in this Inventory, and therefore soil N2O emissions and confidence intervals are estimated using the
Tier 1 method for all years without application of the splicing method.
Tier 1 Approach for Drainage of Organic Soils in Croplands and Grasslands
The IPCC (2006) Tier 1 method is used to estimate direct N2O emissions due to drainage of organic soils in
croplands and grasslands at a state scale. State-scale estimates of the total area of drained organic soils are
obtained from the 2017 NRI (USDA-NRCS 2020) using soils data from the Soil Survey Geographic Database
(SSURGO) (Soil Survey Staff 2020). Temperature data from the PRISM Climate Group (PRISM 2022) are used to
subdivide areas into temperate and tropical climates according to the climate classification from IPCC (2006). To
estimate annual emissions, the total temperate area is multiplied by the IPCC default emission factor for
temperate regions, and the total tropical area is multiplied by the IPCC default emission factor for tropical regions
(IPCC 2006).
Total Direct N2O Emissions from Cropland and Grassland Soils
Annual direct emissions from the Tier 1 and 3 approaches for mineral and drained organic soils occurring in both
croplands and grasslands are summed to obtain the total direct N2O emissions from agricultural soil management
(see Table 5-16 and Table 5-17). Further elaboration on the methodology and data used to estimate soil N2O
emissions are described in Annex 3.12.
Indirect N2O Emissions Associated with Nitrogen Management in Cropland and
Grasslands
Indirect N2O emissions occur when synthetic N applied or made available through anthropogenic activity is
transported from the soil either in gaseous or aqueous forms and later converted into N2O. There are two
pathways leading to indirect emissions. The first pathway results from volatilization of N as NOx and NH3 following
application of synthetic fertilizer, organic amendments (e.g., manure, biosolids), and deposition of PRP manure.
Nitrogen made available from mineralization of soil organic matter and residue, including N incorporated into
crops and forage from symbiotic N fixation, and input of N from asymbiotic fixation also contributes to volatilized
N emissions. Volatilized N can be returned to soils through atmospheric deposition, and a portion of the deposited
N is emitted to the atmosphere as N2O. The second pathway occurs via leaching and runoff of soil N (primarily in
the form of NO3") that is made available through anthropogenic activity on managed lands, including organic and
synthetic fertilization, organic amendments, mineralization of soil organic matter and residue, and inputs of N into
the soil from asymbiotic fixation. The NO3" is subject to denitrification in water bodies, which leads to N2O
emissions. Regardless of the eventual location of the indirect N2O emissions, the emissions are assigned to the
original source of the N for reporting purposes, which here includes croplands and grasslands.
5-44 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Tier 1 and 3 Approaches for Indirect N2O Emissions from Atmospheric Deposition of Volatilized N
2 The Tier 3 DayCent model and IPCC (2006) Tier 1 methods are combined to estimate the amount of N that is
3 volatilized and eventually emitted as N2O. DayCent is used to estimate N volatilization for land areas whose direct
4 emissions are simulated with DayCent (i.e., most commodity and some specialty crops and most grasslands). The N
5 inputs included are the same as described for direct N2O emissions in the Tier 3 Approach for Mineral Cropland
6 and Grassland Soils sections. Nitrogen volatilization from all other areas is estimated using the Tier 1 method with
7 default IPCC fractions for N subject to volatilization (i.e., synthetic and manure N on croplands not simulated by
8 DayCent, other organic N inputs (i.e., commercial fertilizers), PRP manure N excreted on federal grasslands, and
9 biosolids [i.e., treated sewage sludge] application on grasslands).
10 The IPCC (2006) default emission factor is multiplied by the amount of volatilized N generated from both DayCent
11 and Tier 1 methods to estimate indirect N2O emissions occurring with re-deposition of the volatilized N from 1990-
12 2020 (see Table 5-19). A linear extrapolation data splicing method, described in Box 5-4, is applied to estimate
13 emissions from 2021 based on the emission patterns from 1990 to 2020. Linear extrapolation is a standard data
14 splicing method for estimating emissions at the end of a time series (IPCC 2006). Further elaboration on the
15 methodology and data used to estimate indirect N2O emissions are described in Annex 3.12.
16 Tier 1 and 3 Approaches for Indirect N2O Emissions from Leaching/Runoff
17 As with the calculations of indirect emissions from volatilized N, the Tier 3 DayCent model and IPCC (2006) Tier 1
18 method are combined to estimate the amount of N that is subject to leaching and surface runoff into water bodies,
19 and eventually emitted as N2O. DayCent is used to simulate the amount of N transported from lands in the Tier 3
20 Approach. Nitrogen transport from all other areas is estimated using the Tier 1 method and the IPCC (2006) default
21 factor for the proportion of N subject to leaching and runoff associated with N applications on croplands that are
22 not simulated by DayCent, applications of biosolids on grasslands, other organic N fertilizer applications, crop
23 residue N inputs, and PRP manure N excreted on federal grasslands.
24 For both the DayCent Tier 3 and IPCC (2006) Tier 1 methods, nitrate leaching is assumed to be an insignificant
25 source of indirect N2O in cropland and grassland systems in arid regions, as discussed in IPCC (2006). In the United
26 States, the threshold for significant nitrate leaching is based on the potential evapotranspiration (PET) and rainfall
27 amount, similar to IPCC (2006), and is assumed to be negligible in regions where the amount of precipitation does
28 not exceed 80 percent of PET (Note: All irrigated systems are assumed to have significant amounts of leaching of N
29 even in drier climates).
30 For leaching and runoff data estimated by the Tier 3 and Tier 1 approaches, the IPCC (2006) default emission factor
31 is used to estimate indirect N2O emissions that occur in groundwater and waterways (see Table 5-19). Further
32 elaboration on the methodology and data used to estimate indirect N2O emissions are described in Annex 3.12.
33 In order to ensure time-series consistency, indirect soil N2O emissions are estimated using the Tier 1 and 3
34 approaches from 1990 to 2020 and then a linear extrapolation data splicing method, described in Box 5-4, is
35 applied to estimate emissions from 2021 based on the emission patterns from 1990 to 2020. Linear extrapolation
36 is a standard data splicing method for estimating emissions at the end of a time series (IPCC 2006). As with the
37 direct N2O emissions, the time series will be recalculated in a future Inventory when new activity data are
38 incorporated into the analysis.
39 Uncertainty
40 Uncertainty is estimated for each of the following five components of N2O emissions from agricultural soil
41 management: (1) direct emissions simulated by DayCent; (2) the components of indirect emissions (N volatilized
42 and leached or runoff) simulated by DayCent; (3) direct emissions estimated with the IPCC (2006) Tier 1 method;
43 (4) the components of indirect emissions (N volatilized and leached or runoff) estimated with the IPCC (2006) Tier
44 1 method; and (5) indirect emissions estimated with the IPCC (2006) Tier 1 method. Uncertainty in direct emissions
45 as well as the components of indirect emissions that are estimated from DayCent are derived from a Monte Carlo
Agriculture 5-45
-------
1 Analysis (consistent with IPCC Approach 2), addressing uncertainties in model inputs and structure (i.e., algorithms
2 and parameterization) (Del Grosso et al. 2010). For 2021 (and 2018 to 2021 for other organic N fertilizers)here is
3 additional uncertainty propagated through the Monte Carlo Analysis associated with the splicing method (See Box
4 5-4) except for the Tier 1 method for biosolids and crop residue N inputs, which do not use the data splicing
5 method for 2021.
6 Simple error propagation methods (IPCC 2006) are used to derive confidence intervals for direct emissions
7 estimated with the IPCC (2006) Tier 1 method, the proportion of volatilization and leaching or runoff estimated
8 with the IPCC (2006) Tier 1 method, and indirect N2O emissions. Uncertainty in the splicing method is also included
9 in the error propagation for 2021 (see Box 5-4). Additional details on the uncertainty methods are provided in
10 Annex 3.12.
11 Table 5-20 shows the combined uncertainty for soil N2O emissions. The estimated direct soil N2O emissions range
12 from 35 percent below to 74 percent above the 2021 emission estimate of 257.7 MMT CO2 Eq. The combined
13 uncertainty for indirect soil N2O emissions ranges from 64 percent below to 146 percent above the 2021 estimate
14 of 27.5 MMTCO2 Eq.
15 Table 5-20: Quantitative Uncertainty Estimates of N2O Emissions from Agricultural Soil
16 Management in 2021 (MMT CO2 Eq. and Percent)
2021 Emission
Source
Gas
Estimate
Uncertainty Range Relative to Emission Estimate
(MMT C02 Eq.)
(MMT CO
2 Eq.) (%)
Lower
Upper Lower Upper
Bound
Bound Bound Bound
Direct Soil N20 Emissions
N20
257.7
168.0
447.7 -35% 74%
Indirect Soil N20 Emissions
n2o
27.5
9.9
67.8 -64% 146%
Note: Due to lack of data, uncertainties in PRP manure N production, other organic fertilizer amendments, and
biosolids (i.e., treated sewage sludge) amendments to soils are currently treated as certain. These sources of
uncertainty will be included in a future Inventory (IPCC 2006). Quality control procedures uncovered minor errors in
the estimates that will be corrected in the final version of this Inventory.
17 Additional uncertainty is associated with an incomplete estimation of N2O emissions from managed croplands and
18 grasslands in Hawaii and Alaska. The Inventory currently includes the N2O emissions from managed manure and
19 PRP N, and biosolid additions for Alaska and managed manure and PRP N, biosolid additions, and crop residue for
20 Hawaii. Land areas used for agriculture in Alaska and Hawaii are small relative to major crop commodity states in
21 the conterminous United States, so the emissions are likely to be minor for the other sources of N (e.g., synthetic
22 fertilizer and crop residue inputs. Regardless, there is a planned improvement to include the additional sources of
23 emissions in a future Inventory.
24 QA/QC and Verification
25 General (Tier 1) and category-specific (Tier 2) QA/QC activities were conducted consistent with the U.S. Inventory
26 QA/QC plan outlined in Annex 8. DayCent results for N2O emissions and NO3" leaching are compared with field data
27 representing various cropland and grassland systems, soil types, and climate patterns (Del Grosso et al. 2005; Del
28 Grosso et al. 2008), and further evaluated by comparing the model results to emission estimates produced using
29 the IPCC (2006) Tier 1 method for the same sites. Nitrous oxide measurement data for cropland are available for
30 79 sites with 829 observations of management practice effects, and measurement data for grassland are available
31 for 11 sites with 17 observations of management practice effects. Nitrate leaching data are available for 9 sites,
32 representing 230 observations of management practice effects. In general, DayCent predicted N2O emission and
33 nitrate leaching for these sites reasonably well. See Annex 3.12 for more detailed information about the
34 comparisons.
5-46 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Databases containing input data and probability distribution functions required for DayCent simulations of
croplands and grasslands and unit conversion factors have been checked, in addition to the program scripts that
are used to run the Monte Carlo uncertainty analysis. Major errors were found in the synthetic N application rates
for the Tier 3 method, with overapplication based on comparisons to the synthetic fertilizer sales data. Other
errors were identified in the application of the structural uncertainty estimator for direct and indirect soil N2O
emissions. All of these errors were corrected. Databases containing input data, emission factors, and calculations
required for the Tier 1 method have been checked and updated as needed. Links between spreadsheets have also
been checked, updated, and corrected as needed.
Recalculations Discussion
Several improvements have been implemented in this Inventory leading to the need for recalculations. These
improvements included a) incorporating new USDA-NRCS NRI data through 2017; b) extending the time series for
crop histories through 2020 using USDA-NASS CDL data; c) incorporating USDA-NRCS CEAP survey data for 2013 to
2016; d) incorporating cover crop and tillage management information from the OpTIS remote-sensing data
product from 2008 to 2020; e) modifying the statistical imputation method for the management activity data
associated with about tillage practices, mineral fertilization, manure amendments, cover crop management,
planting and harvest dates using gradient boosting instead of an artificial neural network; f) updating time series of
synthetic N fertilizer sales data, PRP N and manure N available for application to soils; g) constraining synthetic N
fertilization and manure N applications in the Tier 3 method at the state scale rather than the national scale; h) re-
calibrating the soil C module in the DayCent model using Bayesian methods; and i) application of global warming
potential (GWP) values from the IPCC Fifth Assessment Report (AR5) (IPCC 2013). The updated GWP for calculating
CCh-equivalent emissions N2O (updated from 298 to 265) reflects the 100-year GWPs provided in the IPCC AR5.
The previous Inventory used 100-year GWPs provided in the IPCC Fourth Assessment Report (AR4). This update
was applied across the entire time series. Further discussion on this update and the overall impacts of updating the
Inventory GWP values to reflect the AR5 can be found in Chapter 9, Recalculations and Improvements.
These combined impact from these changes resulted in an average annual decrease in emissions of 35.3 MMT CO2
Eq., or 11 percent, from 1990 to 2020 relative to the previous Inventory.
Planned Improvements
Several planned improvements are underway associated with improving the DayCent biogeochemical model.
These improvements include a better representation of plant phenology, particularly senescence events following
grain filling in crops. In addition, crop parameters associated with temperature and water stress effects on plant
production will be further improved in DayCent with additional model calibration. In addition, there is an
improvement underway to calibrate the N submodule in order to more accurately predict N-gas losses and nitrate
leaching rates. Experimental study sites will continue to be added for quantifying model structural uncertainty with
priority given to studies that have continuous (daily) measurements of N2O (e.g., Scheer et al. 2013). In addition,
improvements are underway to simulate crop residue burning in the DayCent model based on the amount of crop
residues burned according to the data that is used in the Field Burning of Agricultural Residues source category
(see Section 5.7).
For Tier 1, there is a planned improvement to include all sources of N for Alaska and Hawaii in the Inventory for
agricultural soil management, which currently only addresses managed manure N and PRP N, and biosolids
additions for grasslands in both states, in addition to crop residue N inputs for Hawaii. There is also an
improvement to incorporate the Tier 1 emission factor for N2O emissions from drained organic soils by using the
revised factors in the 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories:
Wetlands (IPCC 2014). There is a planned improvement for the Tier 1 method associated with estimating soil N2O
emissions from N mineralization due to soil organic matter decomposition that is accelerated with land use
conversions to cropland and grassland. Lastly, a review of available data on biosolids (i.e., treated sewage sludge)
Agriculture 5-47
-------
1 application will also be undertaken to improve the distribution of biosolids application on croplands, grasslands
2 and settlements.
3 Other suggested improvements identified through public review are being evaluated for future Inventory
4 submissions. Improvements are expected to be completed for the next Inventory (i.e., 2024 submission to the
5 UNFCCC, 1990 through 2022 Inventory). However, the timeline may be extended if there are insufficient resources
6 to fund all or part of these planned improvements.
7 5.5 Liming (CRF Source Category 3G)
8 Crushed limestone (CaCOs) and dolomite (CaMgfCOsh) are added to soils by land managers to increase soil pH
9 (i.e., to reduce acidification). Carbon dioxide emissions occur as these compounds react with hydrogen ions in
10 soils. The rate of degradation of applied limestone and dolomite depends on the soil conditions, soil type, climate
11 regime, and whether limestone or dolomite is applied. Emissions from limestone and dolomite that are used in
12 industrial processes (e.g., cement production, glass production, etc.) are reported in the IPPU chapter. Emissions
13 from liming of soils have fluctuated between 1990 and 2021 in the United States, ranging from 2.2 MMT CO2 Eq. to
14 6.0 MMT CO2 Eq. across the entire time series. In 2021, liming of soils in the United States resulted in emissions of
15 3.0 MMT CO2 Eq. (0.8 MMT C), representing a 35 percent decrease in emissions since 1990 (see Table 5-21 and
16 Table 5-22). The trend is driven by variation in the amount of limestone and dolomite applied to soils over the time
17 period.
18 Table 5-21: Emissions from Liming (MMT CO2 Eq.)
Source 1990
2005
2017 2018 2019 2020 2021
Limestone 4.1
Dolomite 0.6
3.9
0.4
2.9 2.0 1.9 2.5 2.6
0.2 0.2 0.3 0.4 0.4
Total 4.7
4.4
3.1 2.2 2.2 2.9 3.0
Note: Totals may not sum due to independent rounding.
19 Table 5-22: Emissions from Liming (MMT C)
Source 1990
2005
2017 2018 2019 2020 2021
Limestone 1.1
Dolomite 0.2
1.1
0.1
0.8 0.6 0.5 0.7 0.7
+ 0.1 0.1 0.1 0.1
Total 1.3
1.2
0.8 0.6 0.6 0.8 0.8
+ Does not exceed 0.05 MMT C
Note: Totals may not sum due to independent rounding.
20 Methodology and Time-Series Consistency
21 Carbon dioxide emissions from application of limestone and dolomite to soils were estimated using a Tier 2
22 methodology consistent with IPCC (2006). The annual amounts of limestone and dolomite, which are applied to
23 soils (see Table 5-23), were multiplied by CO2 emission factors from West and McBride (2005). These country-
24 specific emission factors (0.059 metric ton C/metric ton limestone, 0.064 metric ton C/metric ton dolomite) are
25 lower than the IPCC default emission factors because they account for the portion of carbonates that are
26 transported from soils through hydrological processes and eventually deposited in ocean basins (West and
27 McBride 2005). This analysis of lime dissolution is based on studies in the Mississippi River basin, where the vast
28 majority of lime application occurs in the United States (West 2008). Moreover, much of the remaining lime
29 application is occurring under similar precipitation regimes, and so the emission factors are considered a
30 reasonable approximation for all lime application in the United States (West 2008) (See Box 5-5).
5-48 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
The annual application rates of limestone and dolomite were derived from estimates and industry statistics
provided in the U.S. Geological Survey (USGS) Minerals Yearbook (Tepordei 1993 through 2006; Willett 2007a,
2007b, 2009, 2010, 2011a, 2011b, 2013a, 2014, 2015, 2016, 2017, 2020a, 2022a, 2022b, 2022c), as well as
preliminary data that will eventually be published in the Minerals Yearbook for the latter part of the time series
(Willett 2022d). Data for the final year of the inventory is based on the Mineral Industry Surveys, as discussed
below (USGS 2022). The U.S. Geological Survey (USGS; U.S. Bureau of Mines prior to 1997) compiled production
and use information through surveys of crushed stone manufacturers. However, manufacturers provided different
levels of detail in survey responses so the estimates of total crushed limestone and dolomite production and use
were divided into three components: (1) production by end-use, as reported by manufacturers (i.e., "specified"
production); (2) production reported by manufacturers without end-uses specified (i.e., "unspecified" production);
and (3) estimated additional production by manufacturers who did not respond to the survey (i.e., "estimated"
production).
Box 5-5: Comparison of the Tier 2 U.S. Inventory Approach and IPCC (2006) Default Approach
Emissions from liming of soils were estimated using a Tier 2 methodology based on emission factors specific to
the United States that are lower than the IPCC (2006) default emission factors. Most lime application in the
United States occurs in the Mississippi River basin, or in areas that have similar soil and rainfall regimes as the
Mississippi River basin. Under these conditions, a significant portion of dissolved agricultural lime leaches
through the soil into groundwater. Groundwater moves into channels and is transported to larger rives and
eventually the ocean where CaCC>3 precipitates to the ocean floor (West and McBride 2005). The U.S.-specific
emission factors (0.059 metric ton C/metric ton limestone and 0.064 metric ton C/metric ton dolomite) are
about half of the IPCC (2006) emission factors (0.12 metric ton C/metric ton limestone and 0.13 metric ton
C/metric ton dolomite). For comparison, the 2021 U.S. emission estimate from liming of soils is 3.0 MMT CO2
Eq. using the country-specific factors. In contrast, emissions would be estimated at 6.2 MMT CO2 Eq. using the
IPCC (2006) default emission factors.
Data on "specified" limestone and dolomite amounts were used directly in the emission calculation because the
end use is provided by the manufacturers and can be used to directly determine the amount applied to soils.
However, it is not possible to determine directly how much of the limestone and dolomite is applied to soils for
manufacturer surveys in the "unspecified" and "estimated" categories. For these categories, the amounts of
crushed limestone and dolomite applied to soils were determined by multiplying the percentage of total
"specified" limestone and dolomite production that is applied to soils, by the total amounts of "unspecified" and
"estimated" limestone and dolomite production. In other words, the proportion of total "unspecified" and
"estimated" crushed limestone and dolomite that was applied to soils is proportional to the amount of total
"specified" crushed limestone and dolomite that was applied to soils.
In addition, data were not available for 1990,1992, and 2021 on the fractions of total crushed stone production
that were limestone and dolomite, and on the fractions of limestone and dolomite production that were applied to
soils. To estimate the 1990 and 1992 data, a set of average fractions were calculated using the 1991 and 1993
data. These average fractions were applied to the quantity of "total crushed stone produced or used" reported for
1990 and 1992 in the 1994 Minerals Yearbook (Tepordei 1996). To estimate 2021 data, 2020 fractions were applied
to the 2021 estimates of total crushed stone. The basis for these estimates is from the USGS Mineral Industry
Surveys: Crushed Stone and Sand and Gravel in the First Quarter of 2022 (USGS 2022).
The primary source for limestone and dolomite activity data is the Minerals Yearbook, published by the Bureau of
Mines through 1996 and by the USGS from 1997 to the present. In 1994, the "Crushed Stone" chapter in the
Minerals Yearbook began rounding (to the nearest thousand metric tons) quantities for total crushed stone
produced or used. It then reported revised (rounded) quantities for each of the years from 1990 to 1993. In order
to minimize the inconsistencies in the activity data, these revised production numbers have been used in all of the
subsequent calculations.
Agriculture 5-49
-------
1
Table 5-23: Applied Minerals (MMT)
Mineral
1990
2005
2017
2018
2019
2020
2021
Limestone
19.0
18.1
13.4
9.4
8.9
11.7
12.2
Dolomite
2.4
1.9
0.7
0.9
1.2
1.6
1.7
2 Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
3 through 2021. In addition, the same methods are applied throughout the time series, and the activity data are
4 extended in the last two years of the time series based on proportions of specified, unspecified and estimated
5 agricultural limestone and dolomite so that estimates are consistent with the previous year's data. These years will
6 be recalculated when additional data are available on the amounts of limestone and dolomite that are used for
7 agricultural purposes.
8 Uncertainty
9 Uncertainty regarding the amount of limestone and dolomite applied to soils was estimated at ±15 percent with
10 normal densities (Tepordei 2003; Willett 2013b). Analysis of the uncertainty associated with the emission factors
11 included the fraction of lime dissolved by nitric acid versus the fraction that reacts with carbonic acid, and the
12 portion of bicarbonate that leaches through the soil and is transported to the ocean. Uncertainty regarding the
13 time associated with leaching and transport was not addressed in this analysis, but is assumed to be a relatively
14 small contributor to the overall uncertainty (West 2005). The probability distribution functions for the fraction of
15 lime dissolved by nitric acid and the portion of bicarbonate that leaches through the soil were represented as
16 triangular distributions between ranges of zero and 100 percent of the estimates. The uncertainty surrounding
17 these two components largely drives the overall uncertainty. The emission factor distributions were truncated at 0
18 so that emissions were not less than 0.
19 A Monte Carlo (Approach 2) uncertainty analysis was applied to estimate the uncertainty in CO2 emissions from
20 liming. The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 5-24. Carbon
21 dioxide emissions from carbonate lime application to soils in 2021 were estimated to be between 0.46 and 5.88
22 MMT CO2 Eq. at the 95 percent confidence level. This confidence interval represents a range of 85 percent below
23 to 94 percent above the 2021 emission estimate of 3.0 MMT CO2 Eq. All of the carbon in the carbonate lime
24 applied to agricultural soils is not emitted to the atmosphere due to the dominance of the carbonate lime
25 dissolving in carbonic acid rather than nitric acid (West and McBride 2005).
26 Table 5-24: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Liming
27 (MMT CO2 Eq. and Percent)
Source
Gas
2021 Emission Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Liming
C02
3.04
0.46 5.88
-85% 94%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
28 QA/QC and Verification
29 A source-specific QA/QC plan for liming has been developed and implemented, consistent with the U.S. Inventory
30 QA/QC plan outlined in Annex 8. The quality control effort focused on the Tier 1 procedures for this Inventory.
31 Quality control uncovered small errors in the national data estimates of total stone sold or used for most years in
32 the inventory time series. These errors were due to changes in the estimates from the original values, which were
33 recalculated and published by USGS in subsequent reports. No other errors were found.
5-50 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
Recalculations Discussion
2 Limestone and dolomite application data for 2018, 2019, 2020 were updated with the recently acquired data from
3 Willett, J.C. (2022a, 2022b, 2022c), rather than approximated by a ratio method, which was used in the previous
4 Inventory. There were also corrections to the national data estimates of total stone sold or used (both limestone
5 and dolomite) based on quality control. With these revisions, the emissions decreased by an average of 0.5
6 percent for inventory time series from 1990 to 2020 relative to the previous Inventory.
7 5.6 Urea Fertilization (CRF Source Category
8 3H)
9 The use of urea (CO(NH2)2) as a fertilizer leads to greenhouse gas emissions through the release of CChthat was
10 fixed during the production of urea. In the presence of water and urease enzymes, urea that is applied to soils as
11 fertilizer is converted into ammonium (NhV), hydroxyl ion (OH), and bicarbonate (HCO3 ). The bicarbonate then
12 evolves into CO2 and water. Emissions from urea fertilization in the United States were 5.2 MMT CO2 Eq. (1.4 MMT
13 C) in 2021 (Table 5-25 and Table 5-26). Carbon dioxide emissions have increased by 116 percent between 1990 and
14 2021 due to an increasing amount of urea that is applied to soils. The variation in emissions across the time series
15 is driven by differences in the amounts of fertilizer applied to soils each year. Carbon dioxide emissions associated
16 with urea that is used for non-agricultural purposes are reported in the IPPU chapter (Section 4.6).
17 Table 5-25: CO2 Emissions from Urea Fertilization (MMT CO2 Eq.)
Source 1990 2005 2017 2018 2019 2020 2021
Urea Fertilization 2.4 J 3.5 4.9 4.9 5.0 5.1 5.2
18 Table 5-26: CO2 Emissions from Urea Fertilization (MMT C)
Source 1990 2005 2017 2018 2019 2020 2021
Urea Fertilization 0.7 1.0 1.3 1.3 1.4 1.4 1.4
19 Methodology and Time-Series Consistency
20 Carbon dioxide emissions from the application of urea to agricultural soils were estimated using the IPCC (2006)
21 Tier 1 methodology. The method assumes that C in the urea is released after application to soils and converted to
22 CO2. The annual amounts of urea applied to croplands (see Table 5-27) were derived from the state-level fertilizer
23 sales data provided in Commercial Fertilizer reports (TVA 1991,1992,1993,1994; AAPFCO 1995 through 2022).30
24 These amounts were multiplied by the default IPCC (2006) emission factor (0.20 metric tons of C per metric ton of
25 urea), which is equal to the C content of urea on an atomic weight basis. National estimates from Urea Fertilization
26 also include emissions from Puerto Rico.
27 Fertilizer sales data are reported in fertilizer years (July previous year through June current year), so a calculation
28 was performed to convert the data to calendar years (January through December). According to monthly fertilizer
30 The amount of urea consumed for non-agricultural purposes in the United States is reported in the Industrial Processes and
Product Use chapter, Section 4.6 Urea Consumption for Non-Agricultural Purposes.
Agriculture 5-51
-------
1 use data (TVA 1992b), 35 percent of total fertilizer used in any fertilizer year is applied between July and December
2 of the previous calendar year, and 65 percent is applied between January and June of the current calendar year.
3 Fertilizer sales data for the 2018 through 2021 fertilizer years were not available for this Inventory. Therefore, urea
4 application in the 2018 through 2021 fertilizer years were estimated using a linear, least squares trend of
5 consumption over the data from the previous five years (2013 through 2017) at the state scale. A trend of five
6 years was chosen as opposed to a longer trend as it best captures the current inter-annual variability in
7 consumption. State-level estimates of CO2 emissions from the application of urea to agricultural soils were
8 summed to estimate total emissions for the entire United States. The fertilizer year data is then converted into
9 calendar year (Table 5-27) data using the method described above.
10 Table 5-27: Applied Urea (MMT)
1990
2005
2017
2018
2019
2020
2021
Urea Fertilizer3
3.3
CO
'sT
6.6
6.7
6.9
7.0
7.1
a These numbers represent amounts applied to all agricultural land, including Cropland Remaining
Cropland, Land Converted to Cropland, Grassland Remaining Grassland, Land Converted to
Grassland, Settlements Remaining Settlements, Land Converted to Settlements, Forest Land
Remaining Forest Land and Land Converted to Forest Land, as it is not currently possible to
apportion the data by land-use/conversion category.
11 Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
12 through 2021. In addition, the same methods are applied in all years and the activity data are extended using a
13 data splicing method with a linear extrapolation based on the last four years of urea fertilization data to ensure
14 consistency in the time series. These years will be recalculated when additional data are available on urea
15 fertilization.
is Uncertainty
17 An Approach 2 Monte Carlo analysis is conducted as described by the IPCC (2006). The largest source of
18 uncertainty is the default emission factor, which assumes that 100 percent of the C in CO(NH2)2 applied to soils is
19 emitted as CO2. The uncertainty surrounding this factor incorporates the possibility that some of the C may not be
20 emitted to the atmosphere, and therefore the uncertainty range is set from 50 percent emissions to the maximum
21 emission value of 100 percent using a triangular distribution. In addition, urea consumption data have uncertainty
22 that is represented as a normal density. Due to the highly skewed distribution of the resulting emissions from the
23 Monte Carlo uncertainty analysis, the estimated emissions are based on the analytical solution to the equation,
24 and the confidence interval is approximated based on the values at 2.5 and 97.5 percentiles.
25 Carbon dioxide emissions from urea fertilization of agricultural soils in 2021 are estimated to be between 2.99 and
26 5.39 MMT CO2 Eq. at the 95 percent confidence level. This indicates a range of 43 percent below to 3 percent
27 above the 2021 emission estimate of 5.2 MMT CO2 Eq. (Table 5-28).
28 Table 5-28: Quantitative Uncertainty Estimates for CO2 Emissions from Urea Fertilization
29 (MMT CO2 Eq. and Percent)
30
Uncertainty Range Relative to Emission
Source Gas
2021 Emission Estimate
Estimate3
(MMT CO? Eq.)
(MMT CO? Eq.)
(%)
Lower Upper
Lower Upper
Bound Bound
Bound Bound
Urea Fertilization C02
5.2
2.99 5.39
-43% +3%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence
interval.
5-52 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 There are additional uncertainties that are not quantified in this analysis. There is uncertainty surrounding the
2 assumptions underlying conversion of fertilizer years to calendar years. These uncertainties are negligible over
3 multiple years because an over- or under-estimated value in one calendar year is addressed with a corresponding
4 increase or decrease in the value for the subsequent year. In addition, there is uncertainty regarding the fate of C
5 in urea that is incorporated into solutions of urea ammonium nitrate (UAN) fertilizer. Emissions of CO2 from UAN
6 applications to soils are not estimated in the current Inventory (see Planned Improvements).
7 QA/QC and Verification
8 A source-specific QA/QC plan for Urea Fertilization has been developed and implemented, consistent with the U.S.
9 Inventory QA/QC plan.
10 Recalculations Discussion
11 The new AAPFCO report on urea consumption (2022) provided revisions to previous estimates of urea fertilization
12 for Idaho and Oklahoma in addition to data for all states in 2017. With the new year of data, data splicing methods
13 were used to adjust the fertilization values for 2018 to 2020 based on the most recent 5 years of data (2013-2017).
14 These modifications resulted in an average reduction in emissions of 1 percent for 2015 to 2020.
15 Planned Improvements
16 A key planned improvement is to incorporate Urea Ammonium Nitrate (UAN) in the estimation of Urea CO2
17 emissions. Activity data for UAN have been identified, but additional information is needed to fully incorporate this
18 type of fertilizer into the analysis, which will be completed in a future Inventory.
19 5.7 Field Burning of Agricultural Residues
20 (CRF Source Category 3F)
21 Crop production creates large quantities of agricultural crop residues, which farmers manage in a variety of ways.
22 For example, crop residues can be left in the field and possibly incorporated into the soil with tillage; collected and
23 used as fuel, animal bedding material, supplemental animal feed, or construction material; composted and applied
24 to soils; transported to landfills; or burned in the field. The 2006IPCC Guidelines does not consider field burning of
25 crop residues to be a net source of CO2 emissions because it is assumed the C released to the atmosphere as CO2
26 during burning is reabsorbed during the next growing season by the crop (IPCC 2006). However, crop residue
27 burning is a net source of CFU, N2O, CO, and NOx, which are released during combustion.
28 In the United States, field burning of agricultural residues occurs in southeastern states, the Great Plains, and the
29 Pacific Northwest (McCarty 2011). The primary crops that are managed with residue burning include corn, cotton,
30 lentils, rice, soybeans, sugarcane and wheat (McCarty 2009). In 2021, CFU and N2O emissions from field burning of
31 agricultural residues were 0.5 MMT CO2 Eq. (17 kt) and 0.2 MMT CO2 Eq. (1 kt), respectively (Table 5-29 and Table
32 5-30). Annual emissions of CFU and N2O have increased from 1990 to 2021 by 14 percent and 16 percent,
33 respectively. The increase in emissions over time is partly due to higher yielding crop varieties with larger amounts
34 of residue production and fuel loads, but also linked with an increase in the area burned for some of the crop
35 types.
Agriculture 5-53
-------
1 Table 5-29: ChU and N2O Emissions from Field Burning of Agricultural Residues (MMT CO2
2 Eq.)
Gas/Crop Type
1990
2005
2017
2018
2019
2020
2021
ch4
0.4
0.5
0.5
0.5
0.5
0.5
0.5
Maize
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Rice
0.1 I
0.1
0.1
0.1
0.1
0.1
0.1
Wheat
0.2
0.2
0.1
0.1
0.1
0.1
0.1
Barley
+
+
+
+
+
+
+
Oats
+
+
+
+
+
+
+
Other Small Grains
+
+
+
+
+
+
+
Sorghum
+
+
+
+
+
+
+
Cotton
+
+
+
+
+
+
+
Grass Hay
+
+
+
+
+
+
+
Legume Hay
+
+
+
+
+
+
+
Peas
+
+
+
+
+
+
+
Sunflower
+
+
+
+
+
+
+
Tobacco
+
+
+
+
+
+
+
Vegetables
+
+
+
+
+
+
+
Chickpeas
+
+
+
+
+
+
+
Dry Beans
+
+
+
+
+
+
+
Lentils
+
+
+
+
+
+
+
Peanuts
+
+
+
+
+
+
+
Soybeans
+
+
+
+
+
+
+
Potatoes
+
+
+
+
+
+
+
Sugarbeets
+
+
+
+
+
+
+
n2o
0.1
0.2
0.2
0.2
0.2
0.2
0.2
Maize
+
+
+
+
+
+
+
Rice
+
+
+
+
+
+
+
Wheat
0.1
0.1
+
+
+
+
+
Barley
+
+
+
+
+
+
+
Oats
+
+
+
+
+
+
+
Other Small Grains
+
+
+
+
+
+
+
Sorghum
+
+
+
+
+
+
+
Cotton
+
+
+
+
+
+
+
Grass Hay
+
+
+
+
+
+
+
Legume Hay
+
+
+
+
+
+
+
Peas
+
+
+
+
+
+
+
Sunflower
+
+
+
+
+
+
+
Tobacco
+
+
+
+
+
+
+
Vegetables
+
+
+
+
+
+
+
Chickpeas
+
+
+
+
+
+
+
Dry Beans
+
+
+
+
+
+
+
Lentils
+
+
+
+
+
+
+
Peanuts
+
+
+
+
+
+
+
Soybeans
+
+
+
+
+
+
+
Potatoes
+
+
+
+
+
+
+
Sugarbeets
+
+
+
+
+
+
+
Total
0.6
0.7
0.7
0.6
0.6
0.6
0.6
+ Does not exceed 0.05 MMT C02 Eq.
5-54 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Note: Totals may not sum due to independent rounding.
1
2 Table 5-30: ChU, N2O, CO, and NOx Emissions from Field Burning of Agricultural Residues
3 (kt)
Gas/Crop Type
1990
2005
2017
2018
2019
2020
2021
ch4
15
17
17
17
17
17
17
Maize
2
4
5
5
5
5
5
Rice
3
3
3
2
3
2
3
Wheat
6
6
5
5
5
5
5
Barley
+
+
+
+
+
+
+
Oats
+
+
+
+
+
+
+
Other Small Grains
+
+
+
+
+
+
+
Sorghum
+
+
+
+
+
+
+
Cotton
1
2
1
1
1
1
1
Grass Hay
+
+
+
+
+
+
+
Legume Hay
+
+
+
+
+
+
+
Peas
+
+
+
+
+
+
+
Sunflower
+
+
+
+
+
+
+
Tobacco
+
+
+
+
+
+
+
Vegetables
+
+
+
+
+
+
+
Chickpeas
+
+
+
+
+
+
+
Dry Beans
+
+
+
+
+
+
+
Lentils
+
+
+
+
+
+
+
Peanuts
+
+
+
+
+
+
+
Soybeans
1
2
2
2
2
2
2
Potatoes
+
+
+
+
+
+
+
Sugarbeets
+
+
+
+
+
+
+
n2o
1
1
1
1
1
1
1
Maize
+
+
+
+
+
+
+
Rice
+
+
+
+
+
+
+
Wheat
+
+
+
+
+
+
+
Barley
+
+
+
+
+
+
+
Oats
+
+
+
+
+
+
+
Other Small Grains
+
+
+
+
+
+
+
Sorghum
+
+
+
+
+
+
+
Cotton
+
+
+
+
+
+
+
Grass Hay
+
+
+
+
+
+
+
Legume Hay
+
+
+
+
+
+
+
Peas
+
+
+
+
+
+
+
Sunflower
+
+
+
+
+
+
+
Tobacco
+
+
+
+
+
+
+
Vegetables
+
+
+
+
+
+
+
Chickpeas
+
+
+
+
+
+
+
Dry Beans
+
+
+
+
+
+
+
Lentils
+
+
+
+
+
+
+
Peanuts
+
+
+
+
+
+
+
Soybeans
+
+
+
+
+
+
+
Agriculture 5-55
-------
Potatoes
+
+
+
+
+
+
+
Sugarbeets
+
+
+
+
+
+
+
CO
315
363
339
338
337
336
336
NOx
13
15
14
14
14
14
14
+ Does not exceed 0.5 kt.
Note: Totals by gas may not sum due to independent rounding.
1 Methodology and Time-Series Consistency
2 A country-specific Tier 2 method is used to estimate greenhouse gas emissions from field burning of agricultural
3 residues from 1990 to 2014 (for more details comparing the country-specific approach to the IPCC (2006) default
4 approach, see Box 5-6), and a data splicing method with a linear extrapolation is applied to complete the emissions
5 time series from 2015 to 2021. The following equation is used to estimate the amounts of C and N released (R/,
6 where / is C or N) from burning.
7 Equation 5-1: Elemental C or N Released through Oxidation of Crop Residues
8
9
10
11
12 where,
13 Crop Production (CP)
14 Residue: Crop Ratio (RCR)
15
16 Dry Matter Fraction (DMF)
17
18 Fraction C or N (Fj)
19
20 Fraction Burned (FB)
21 Combustion Efficiency (CE)
22 Area Burned (AB)
23 Crop Area Harvested (CAH)
24
25 Crop production data are available by state and year from USDA (2019) for twenty-one crops that are burned in
26 the conterminous United States, including maize, rice, wheat, barley, oats, other small grains, sorghum, cotton,
27 grass hay, legume hay, peas, sunflower, tobacco, vegetables, chickpeas, dry beans, lentils, peanuts, soybeans,
28 potatoes, and sugarbeets.31 Crop area data are based on the 2015 National Resources Inventory (NRI) (USDA-NRCS
29 2018). In order to estimate total crop production, the crop yield data from USDA Quick Stats crop yields is
30 multiplied by the NRI crop areas. The production data for the crop types are presented in Table 5-31. Alaska and
31 Hawaii are not included in the current analysis, but there is a planned improvement to estimate residue burning
32 emissions for these two states in a future Inventory.
R, = CPx RCR x DMF x F x FB x CE
AB
FB =
CAH
= Annual production of crop, by state, kt crop production
= Amount of residue produced per unit of crop production, kt residue/kt crop
production
= Amount of dry matter per unit of residue biomass for a crop, kt residue dry
matter/ kt residue biomass
= Fraction of C or N per unit of dry matter for a crop, kt C or N /kt residue dry
matter
= Proportion of residue biomass consumed, unitless
= Proportion of residue actually burned, unitless
= Total area of crop burned, by state, ha
= Total area of crop harvested, by state, ha
31 Sugarcane and Kentucky bluegrass (produced on farms for turf grass installations) may have small areas of burning that are
not captured in the sample of locations that were used in the remote sensing analysis (see Planned Improvements).
5-56 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
7
8
9
10
11
The amount of elemental C or N released through oxidation of the crop residues is used in the following equation
to estimate the amount of Cm, CO, N2O, and NOx emissions (Eg, where g is the specific gas, i.e., CFU, CO, N2O, and
NOx) from the field burning of agricultural residues:
Equation 5-2: Emissions from Crop Residue Burning
Eg = X EFg x CF
where,
Emission ratio [EFg)
Conversion Factor (CF)
= emission ratio by gas, g CH4-C or CO-C/g C released, or g N2O-N or NOx
N/g N released
= conversion by molecular weight ratio of CH4-C to C (16/12), CO-C to C
(28/12), N2O-N to N (44/28), or NOx-N to N (30/14)
12
Box 5-6: Comparison of Tier 2 U.S. Inventory Approach and IPCC (2006) Default Approach
Emissions from Field Burning of Agricultural Residues are calculated using a Tier 2 methodology that is based on
the method developed by the IPCC/UNEP/OECD/IEA (1997). The rationale for using the IPCC/UNEP/OECD/IEA
(1997) approach rather than the method provided in the 2006 IPCC Guidelines is as follows: (1) the equations
from both guidelines rely on the same underlying variables (though the formats differ); (2) the IPCC (2006)
equation was developed to be broadly applicable to all types of biomass burning, and, thus, is not specific to
agricultural residues; (3) the IPCC (2006) method provides emission factors based on the dry matter content
rather than emission rates related to the amount of C and N in the residues; and (4) the IPCC (2006) default
factors are provided only for four crops (corn, rice, sugarcane, and wheat) while this Inventory includes
emissions from twenty-one crops.
A comparison of the methods in the current Inventory and the default IPCC (2006) approach was undertaken for
2014 to determine the difference in estimates between the two approaches. To estimate greenhouse gas
emissions from field burning of agricultural residues using the IPCC (2006) methodology, the following
equation—cf. IPCC (2006) Equation 2.27—was used with default factors and country-specific values for mass of
fuel.
Equation 5-3: Estimation of Greenhouse Gas Emissions from Fire
Emissions (kt) =AB x Mb x Cf x Get x 10~6
where,
Area Burned (AB) = Total area of crop burned (ha)
Mass of Fuel (Mb) = U.S.- Specific Values using NASS Statistics32 (metric tons dry matter)
Combustion Factor (Cf) = IPCC (2006) default combustion factor with fuel biomass consumption
(metric tons dry matter ha"1)
Emission Factor (Gef) = IPCC (2006) emission factor (g kg1 dry matter burnt)
The IPCC (2006) Tier 1 method approach resulted in 21 percent lower emissions of CFU and 44 percent lower
emissions of N2O compared to this Inventory. In summary, the IPCC/UNEP/OECD/IEA (1997) method is
considered more appropriate for U.S. conditions because it is more flexible for incorporating country-specific
data. Emissions are estimated based on specific C and N content of the fuel, which is converted into CH4, CO,
32 NASS yields are used to derive mass of fuel values because IPCC (2006) only provides default values for 4 of the 21 crops
included in the Inventory.
Agriculture 5-57
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
N2O and NOx, compared to IPCC (2006) approach that is based on dry matter rather than elemental
composition.
Table 5-31: Agricultural Crop Production (kt of Product)
Crop
1990
2005
2011
2012
2013
2014
Maize
296,065
371,256
399,531
349,739
436,565
453,524
Rice
9,543
11,751
9,890
10,445
10,894
12,380
Wheat
79,805
68,077
61,082
69,388
67,388
62,602
Barley
9,281
5,161
3,891
5,382
4,931
5,020
Oats
5,969
2,646
1,661
1,743
1,806
2,042
Other Small Grains
2,651
2,051
1,259
1,657
1,902
2,492
Sorghum
23,687
14,382
9,196
11,288
18,680
18,436
Cotton
4,605
6,106
5,200
5,357
3,982
4,396
Grass Hay
44,150
49,880
44,670
40,821
45,588
46,852
Legume Hay
90,360
91,819
82,440
71,435
79,669
82,844
Peas
51
660
206
488
599
447
Sunflower
1,015
1,448
820
1,274
987
907
Tobacco
1,154
337
286
466
481
542
Vegetables
0
1,187
1,201
1,973
1,844
2,107
Chickpeas
0
5
+
1
+
+
Dry Beans
467
1,143
1,024
1,260
1,110
1,087
Lentils
0
101
46
95
72
76
Peanuts
1,856
2,176
1,982
2,854
2,072
2,735
Soybeans
56,612
86,980
87,556
85,843
94,756
110,560
Potatoes
18,924
20,026
19,800
19,776
20,234
19,175
Sugarbeets
24,951
25,635
27,345
32,791
31,890
31,737
+ Absolute value does not exceed 0.05 MMT C02 Eq.
Note: The amount of crop production has not been compiled for 2015 to 2021 so a data splicing method
is used to estimate emissions for this portion of the time series.
The area burned is determined based on an analysis of remote sensing products (McCarty et al. 2009, 2010, 2011).
The presence of fires has been analyzed at 3,600 survey locations in the NRI from 1990 to 2002 with LANDFIRE
data products developed from 30 m Landsat imagery (LANDFIRE 2008), and from 2003 through 2014 using 1 km
Moderate Resolution Imaging Spectroradiometer imagery (MODIS) Global Fire Location Product (MCD14ML) using
combined observations from Terra and Aqua satellites (Giglio et al. 2006). A sample of states are included in the
analysis with high, medium and low burning rates for agricultural residues, including Arkansas, California, Florida,
Indiana, Iowa and Washington. The area burned is determined directly from the analysis for these states.
For other states within the conterminous United States, the area burned for the 1990 through 2014 portion of the
time series is estimated from a logistical regression model that has been developed from the data collected from
the remote sensing products for the six states. The logistical regression model is used to predict occurrence of fire
events. Several variables are tested in the logistical regression including a) the historical level of burning in each
state (high, medium or low levels of burning) based on an analysis by McCarty et al. (2011), b) year that state laws
limit burning of fields, in addition to c) mean annual precipitation and mean annual temperature from a 4-
kilometer gridded product from the PRISM Climate Group (2015). A K-fold model fitting procedure is used due to
low frequency of burning and likelihood that outliers could influence the model fit. Specifically, the model is
trained with a random selection of sample locations and evaluated with the remaining sample. This process is
repeated ten times to select a model that is most common among the set of ten, and avoid models that appear to
5-58 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
be influenced by outliers due to the random draw of survey locations for training the model. In order to address
uncertainty, a Monte Carlo analysis is used to sample the parameter estimates for the logistical regression model
and produce one thousand estimates of burning for each crop in the remaining forty-two states included in this
Inventory. State-level area burned data are divided by state-level crop area data to estimate the percent of crop
area burned by crop type for each state. Table 5-32 shows the resulting percentage of crop residue burned at the
national scale by crop type. State-level estimates are also available upon request.
Table 5-32: U.S. Average Percent Crop Area Burned by Crop (Percent)
Crop
1990
2005
2011
2012
2013
2014
Maize
+
+
+
+
+
+
Rice
8%
8%
4%
5%
4%
6%
Wheat
1%
2%
2%
2%
2%
1%
Barley
1%
+
1%
1%
1%
1%
Oats
1%
1%
1%
1%
2%
1%
Other Small Grains
1%
1%
1%
1%
1%
1%
Sorghum
1%
1%
1%
1%
1%
1%
Cotton
1%
1%
1%
1%
1%
1%
Grass Hay
+
+
+
+
+
+
Legume Hay
+
+
+
+
+
+
Peas
+
+
1%
+
+
+
Sunflower
+
+
+
+
+
+
Tobacco
2%
2%
2%
2%
3%
3%
Vegetables
+
+
+
+
+
+
Chickpeas
+
1%
+
+
0%
0%
Dry Beans
1%
1%
1%
1%
+
+
Lentils
+
+
1%
+
+
+
Peanuts
3%
3%
3%
3%
3%
3%
Soybeans
+
+
+
1%
1%
1%
Potatoes
+
+
+
+
+
+
Sugarbeets
+
+
+
+
+
+
+ Does not exceed 0.5 percent
Additional parameters are needed to estimate the amount of burning, including residue: crop ratios, dry matter
fractions, carbon fractions, nitrogen fractions and combustion efficiency. Residue: crop product mass ratios,
residue dry matter fractions, and the residue N contents are obtained from several sources (IPCC 2006 and sources
at bottom of Table 5-33). The residue C contents for all crops are based on IPCC (2006) default value for
herbaceous biomass. The combustion efficiency is assumed to be 90 percent for all crop types
(IPCC/UNEP/OECD/IEA 1997). See Table 5-33 for a summary of the crop-specific conversion factors. Emission ratios
and mole ratio conversion factors for all gases are based on the Revised 1996 IPCC Guidelines
(IPCC/UNEP/OECD/IEA 1997) (see Table 5-34).
Table 5-33: Parameters for Estimating Emissions from Field Burning of Agricultural Residues
Crop
Residue/Crop
Ratio
Dry
Matter
Fraction
Carbon
Fraction
Nitrogen
Fraction
Combustion
Efficiency
(Fraction)
Maize
0.707
0.56
0.47
0.01
0.90
Rice
1.340
0.89
0.47
0.01
0.90
Wheat
1.725
0.89
0.47
0.01
0.90
Barley
1.181
0.89
0.47
0.01
0.90
Agriculture 5-59
-------
Oats
1.374
0.89
0.47
0.01
0.90
Other Small Grains
1.777
0.88
0.47
0.01
0.90
Sorghum
0.780
0.60
0.47
0.01
0.90
Cotton
7.443
0.93
0.47
0.01
0.90
Grass Hay
0.208
0.90
0.47
0.02
0.90
Legume Hay
0.290
0.67
0.47
0.01
0.90
Peas
1.677
0.91
0.47
0.01
0.90
Sunflower
1.765
0.88
0.47
0.01
0.90
Tobacco
0.300
0.87
0.47
0.01
0.90
Vegetables
0.708
0.08
0.47
0.01
0.90
Chickpeas
1.588
0.91
0.47
0.01
0.90
Dry Beans
0.771
0.90
0.47
0.01
0.90
Lentils
1.837
0.91
0.47
0.02
0.90
Peanuts
1.600
0.94
0.47
0.02
0.90
Soybeans
1.500
0.91
0.47
0.01
0.90
Potatoes
0.379
0.25
0.47
0.02
0.90
Sugarbeets
0.196
0.22
0.47
0.02
0.90
Notes: Chickpeas: IPCC (2006), Table 11.2; values are for Beans & pulses.
Cotton: Combined sources (Heitholt et al. 1992; Halevy 1976; Wells and Meredith 1984; Sadras and
Wilson 1997; Pettigrew and Meredith 1997; Torbert and Reeves 1994; Gerik et al. 1996; Brouder
and Cassmen 1990; Fritschi et al. 2003; Pettigrew et al. 2005; Bouquet and Breitenbeck 2000;
Mahroni and Aharonov 1964; Bange and Milroy 2004; Hollifield et al. 2000; Mondino et al. 2004;
Wallach etal. 1978).
Lentils: IPCC (2006), Table 11.2; Beans & pulses.
Peas: IPCC (2006), Table 11.2; values are for Beans & pulses.
Peanuts: IPCC (2006); Table 11.2; Root ratio and belowground N content values are for Root crops,
other.
Sugarbeets: IPCC (2006); Table 11.2; values are forTubers.
Sunflower: IPCC (2006), Table 11.2; values are for Grains.
Sugarcane: combined sources (Wiedenfels 2000, Dua and Sharma 1976; Singels & Bezuidenhout
2002; Stirling et al. 1999; Sitompul et al. 2000).
Tobacco: combined sources (Beyaert 1996; Moustakas and Ntzanis 2005; Crafts-Brandner et al. 1994;
Hopkinson 1967; Crafts-Brandner et al. 1987).
Vegetables (Combination of carrots, lettuce/cabbage, melons, onions, peppers and tomatoes):
Carrots: McPharlin et al. (1992); Gibberd et al. (2003); Reid and English (2000); Peach et al. (2000);
see IPCC Tubers for R:S and N fraction.
Lettuce, cabbage: combined sources (Huett and Dettman 1991; De Pinheiro Henriques & Marcelis
2000; Huett and Dettman 1989; Peach et al. 2000; Kage et al. 2003; Tan et al. 1999; Kumar et al.
1994; MacLeod et al. 1971; Jacobs et al. 2004; Jacobs et al. 2001; Jacobs et al. 2002); values from
IPCC Grains used for N fraction.
Melons: Valantin et al. (1999); squash for R:S; IPCC Grains for N fraction.
Onion: Peach et al. (2000), Halvorson et al. (2002); IPCC (2006) Tubers for N fraction.
Peppers: combined sources (Costa and Gianquinto 2002; Marcussi et al. 2004; Tadesse et al. 1999;
Diaz-Perez et al. 2008); IPCC Grains for N fraction.
Tomatoes: Scholberg et al. (2000a,b); Akintoye et al. (2005); values for AGR-N and BGR-N are from
Grains.
l Table 5-34: Greenhouse Gas Emission Ratios and Conversion Factors
Gas
Emission Ratio
Conversion Factor
CH4:C
0.005a
16/12
CO:C
0.060a
28/12
N20:N
0.007b
44/28
NOx:N
0.121b
30/14
5-60 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
a Mass of C compound released (units of C) relative to
mass of total C released from burning (units of C).
b Mass of N compound released (units of N) relative to
mass of total N released from burning (units of N).
1 For this Inventory, new activity data on the burned areas have not been analyzed for 2015 to 2021. To complete
2 the emissions time series, a linear extrapolation of the trend is applied to estimate the emissions in the last seven
3 years of the inventory. Specifically, a linear regression model with autoregressive moving-average (ARMA) errors is
4 used to estimate the trend in emissions over time from 1990 through 2014, and the trend is used to approximate
5 the Cm, N2O, CO and NOx from 2015 to 2021 (Brockwell and Davis 2016). The Tier 2 method described previously
6 will be applied to recalculate the emissions for the last seven years in the time series (2015 to 2021) in a future
7 Inventory.
8 In order to ensure time-series consistency, the same method is applied from 1990 to 2014, and a linear
9 extrapolation method is used to approximate emissions for the remainder of the time series based on the
10 emissions data from 1990 to 2014. This extrapolation method is consistent with data splicing methods in IPCC
11 (2006).
12 Uncertainty
13 Emissions are estimated using a linear regression model with autoregressive moving-average (ARMA) errors for
14 2021. The linear regression ARMA model produced estimates of the upper and lower bounds to quantify
15 uncertainty (Table 5-35), and the results are summarized in Table 5-35. Methane emissions from field burning of
16 agricultural residues in 2021 are between 0.4 and 0.6 MMT CO2 Eq. at a 95 percent confidence level. This indicates
17 a range of 16 percent below and 16 percent above the 2021 emission estimate of 0. MMT CO2 Eq. Nitrous oxide
18 emissions are between 0.1 and 0.2 MMT CO2 Eq., or approximately 19 percent below and 19 percent above the
19 2021 emission estimate of 0.2 MMT CO2 Eq.
20 Table 5-35: Approach 2 Quantitative Uncertainty Estimates for ChU and N2O Emissions from
21 Field Burning of Agricultural Residues (MMT CO2 Eq. and Percent)
2021 Emission
Source
Gas
Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT CO? Eq.)
(MMT CO?
Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Field Burning of Agricultural
Residues
ch4
0.5
0.4
0.6
-16%
16%
Field Burning of Agricultural
Residues
n2o
0.2
0.1
0.2
-19%
19%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
22 Due to data limitations, there are additional uncertainties in agricultural residue burning, particularly the potential
23 omission of burning associated with Kentucky bluegrass (produced on farms for turf grass installation) and
24 sugarcane (see Annex 5 on sugarcane).
25 QA/QC and Verification
26 A source-specific QA/QC plan for field burning of agricultural residues is implemented with Tier 1 analyses,
27 consistent with the U.S. Inventory QA/QC plan outlined in Annex 8. Quality control measures included checking
28 input data, model scripts, and results to ensure data are properly handled throughout the inventory process.
29 Inventory reporting forms and text are reviewed and revised as needed to correct transcription errors. An error
30 was identified in the calculation of the emissions using the IPCC (2006) equation, which was corrected in Box 5.6.
Agriculture 5-61
-------
1 Recalculations Discussion
2 EPA updated the global warming potentials (GWPs) for calculating CC>2-equivalent emissions of CFU (from 25 to 28)
3 and N2O (from 298 to 265) to reflect the 100-year GWPs provided in the IPCC Fifth Assessment Report (AR5) (IPCC
4 2013). The previous Inventory used 100-year GWPs provided in the IPCC Fourth Assessment Report (AR4). This
5 update was applied across the entire time series to ensure consistency.
6 As a result of this change, CC>2-equivalent CFU emissions increased by an annual average of 0.05 MMT CO2 Eq., or
7 12 percent, over the time series from 1990 to 2020 compared to the previous Inventory. In contrast, N2O
8 emissions decreased by an annual average of 0.02 MMT CO2 Eq., or 11 percent, over the time series from 1990 to
9 2020 compared to the previous Inventory. Further discussion on this update and the overall impacts of updating
10 the Inventory GWP values to reflect the AR5 can be found in Chapter 9, Recalculations and Improvements.
11 Planned Improvements
12 A key planned improvement is to estimate the emissions associated with field burning of agricultural residues in
13 the states of Alaska and Hawaii. In addition, a new method is in development that will directly link agricultural
14 residue burning with the Tier 3 methods that are used in several other source categories, including Agricultural Soil
15 Management, Cropland Remaining Cropland, and Land Converted to Cropland chapters of the Inventory. The
16 method is based on simulating burning events directly within the DayCent process-based model framework using
17 information derived from remote sensing fire products as described in the Methodology section. This
18 improvement will lead to greater consistency in the methods for across sources, ensuring mass balance of C and N
19 in the Inventory analysis.
20 As previously noted in this chapter, remote sensing data were used in combination with a resource survey to
21 estimate non-CC>2 emissions and these data did not allow identification of burning of sugarcane (see Annex 5). EPA
22 has received feedback on this category/crop type, which includes average estimates of emissions of sugarcane
23 burning found in academic literature. EPA plans to incorporate the burning of sugarcane into the analysis during a
24 future Inventory when an updated analysis is conducted (see Annex 5).
5-62 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Forestry
This chapter provides an assessment of the greenhouse gas fluxes resulting from land use and land-use change in
the United States.1 The Intergovernmental Panel on Climate Change's 2006IPCC Guidelines for National
Greenhouse Gas Inventories (IPCC 2006) recommends reporting fluxes according to changes within and
conversions between all land use types including: Forest Land, Cropland, Grassland, Wetlands, and Settlements (as
well as Other Land).
The greenhouse gas flux from Forest Land Remaining Forest Land is reported for all forest ecosystem carbon (C)
pools (i.e., aboveground biomass, belowground biomass, dead wood, litter, and mineral and organic soils),
harvested wood pools, and non-carbon dioxide (non-CCh) emissions from forest fires, the application of synthetic
nitrogen fertilizers to forest soils, and the draining of organic soils. Fluxes from Land Converted to Forest Land are
included for aboveground biomass, belowground biomass, dead wood, litter, and C stock changes from mineral
soils, while C stock changes from drained organic soils and all non-CC>2 emissions from Land Converted to Forest
Land are included in the fluxes from Forest Land Remaining Forest Land as it is not currently possible to separate
these fluxes by conversion category.
Fluxes are reported for four agricultural land use/land-use change categories: Cropland Remaining Cropland, Land
Converted to Cropland, Grassland Remaining Grassland, and Land Converted to Grassland. The reported
greenhouse gas fluxes from these agricultural lands include changes in soil organic C stocks in mineral and organic
soils due to land use and management, and for the subcategories of Forest Land Converted to Cropland and Forest
Land Converted to Grassland, the changes in aboveground biomass, belowground biomass, dead wood, and litter C
stocks are also reported. The greenhouse gas flux from Grassland Remaining Grassland also includes estimates of
non-CC>2 emissions from grassland fires occurring on both Grassland Remaining Grassland and Land Converted to
Grassland.
Fluxes from Wetlands Remaining Wetlands include changes in C stocks and methane (Cm) and nitrous oxide (N2O)
emissions from managed peatlands, aboveground and belowground biomass, dead organic matter, soil C stock
changes and CFU emissions from coastal wetlands, as well as N2O emissions from aquaculture. In addition, CH4
emissions from reservoirs and other constructed waterbodies are included for the subcategory Flooded Land
Remaining Flooded Land. Estimates for Land Converted to Wetlands include aboveground and belowground
biomass, dead organic matter and soil C stock changes, and CFU emissions from land converted to vegetated
coastal wetlands. Carbon dioxide (CO2) and CFU emissions are included for reservoirs and other constructed
waterbodies under the subcategory Land Converted to Flooded Land.
1 The term "flux" is used to describe the exchange of C02 to and from the atmosphere, with net flux of C02 being either positive
or negative depending on the overall balance. Removal and long-term storage of C02 from the atmosphere is also referred to as
"carbon sequestration."
Land Use, Land-Use Change, and Forestry 6-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Fluxes from Settlements Remaining Settlements include changes in C stocks from organic soils, N2O emissions from
nitrogen fertilizer additions to soils, and CChfluxes from settlement trees and landfilled yard trimmings and food
scraps. The reported greenhouse gas flux from Land Converted to Settlements includes changes in C stocks in
mineral and organic soils due to land use and management for all land use conversions to settlements, and the C
stock changes in aboveground biomass, belowground biomass, dead wood, and litter are also included for the
subcategory Forest Land Converted to Settlements.
In 2021, the land use, land-use change, and forestry (LULUCF) sector resulted in a net increase in C stocks (i.e., net
CO2 removals) of 832.0 MMT CO2 Eq. This represents an offset of approximately 13.1 percent of total (i.e., gross)
greenhouse gas emissions in 2021. Emissions of CH4 and N2O from LULUCF activities in 2021 were 66.0 and 11.8
MMT CO2 Eq., respectively, and combined represent 1.2 percent of total greenhouse gas emissions.3 In 2021, the
overall net flux from LULUCF resulted in a removal of 754.2 MMT CO2 Eq. Emissions, removals and net greenhouse
gas flux from LULUCF are summarized in Figure 6-1 and Table 6-1 by land use and category, and Table 6-2 and
Table 6-3 by gas in MMT CO2 Eq. and kt, respectively. Trends in LULUCF sources and sinks over the 1990 to 2021
time series are shown in Figure 6-2.
Flooded Land Remaining Flooded Land was the largest source of non-CC>2emissions from LULUCF in 2021,
accounting for 58.4 percent of the LULUCF sector emissions. Non-CC>2 emissions from forest fires are the second
largest source of LULUCF sector emissions; these emissions have increased 341.4 percent since 1990 and account
for 31.4 percent of LULUCF emissions in 2021. Coastal Wetlands Remaining Coastal Wetlands and Settlements
Remaining Settlements soils accounted for 5.7 and 2.6 percent of non-CC>2emissions from LULUCF in 2021,
respectively, and the remaining sources account for less than one percent each.
2 LULUCF Carbon Stock Change is the net C stock change from the following categories: Forest Land Remaining Forest Land,
Land Converted to Forest Land, Cropland Remaining Cropland, Land Converted to Cropland, Grassland Remaining Grassland,
Land Converted to Grassland, Wetlands Remaining Wetlands, Land Converted to Wetlands, Settlements Remaining
Settlements, and Land Converted to Settlements.
3 LULUCF emissions include the CH4 and N20 emissions reported for Peatlands Remaining Peatlands, forest fires, drained
organic soils, grassland fires, and Coastal Wetlands Remaining Coastal Wetlands; CH4 emissions from Land Converted to Coastal
Wetlands, Flooded Land Remaining Flooded Land, and Land Converted to Flooded Land; and N20 emissions from forest soils
and settlement soils.
6-2 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Figure 6-1: 2021 LULUCF Chapter Greenhouse Gas Sources and Sinks
Forest Land Remaining Forest Land
Settlements Remaining Settlements
Land Converted to Forest Land
Land Converted to Grassland
Cropland Remaining Cropland
Wetlands Remaining Wetlands
Non-CCh Emissions from Peatlands Remaining Peatlands
Non-CCh Emissions from Drained Organic Soils
CH4 Emissions from Land Converted to Flooded Land
Cm Emissions from Land Converted to Coastal Wetlands
Land Converted to Wetlands
N2O Emissions from Forest Soils
Non-CCh Emissions from Grassland Fires
N2O Emissions from Settlement Soils
Non-CCh Emissions from Coastal Wetlands Remaining Coastal Wetlands
Grassland Remaining Grassland
Non-CC>2 Emissions from Forest Fires
Non-CCh Emissions from Flooded Land Remaining Flooded Land
Land Converted to Cropland
Land Converted to Settlements
Note: Parentheses in horizontal axis indicate net sequestration.
(695.4)
Carbon Stock Change
I Non-CCh Emissions
l< 0.5|
l< 0.5|
|< 0.51
l< 0.51
|< 0.51
|< 0.5|
(250) (200) (150) (100) (50)
MMT CCh Eq.
50 100
Figure 6-2; Trends in Emissions and Removals (Net CO2 Flux) from Land Use, Land-Use
Change, and Forestry
400
300
Wetlands Remaining Wetlands
Land Converted to Wetlands
I Land Converted to Settlements
Land Converted to Grassland
200 a) ™
in
co in in
^ CO ^ (M ^
*T" I"***
-i ld <0 in
Land Converted to Forest Land
I Land Converted to Cropland
Grassland Remaining Grassland
Cropland Remaining Cropland
u") m r--. tt- ^ o CN ko
' ' H d ^ ^ ^
>0 00 r-v
I Settlements Remaining Settlements
I Forest Land Remaining Forest Land
¦ Net Emissions (Sources and Sinks)
100
0
-100
-200
-300
-400
-500
-600
-700
-800
-900
-1,000
Ci Ol Ci
G*i C> C> Q*i
f\l (N f"\l (N fN fN
(N N N (N (N
(N fN fN fM fN
Land Use, Land-Use Change, and Forestry 6-3
-------
1 Table 6-1: Emissions and Removals (Net Flux) from Land Use, Land-Use Change, and
2 Forestry (MMT CO2 Eq.)
Land-Use Category
1990
2005
2017
2018
2019
2020
2021
Forest Land Remaining Forest Land
(815.8)
(695.4)
(695.2)
(692.9)
(638.1)
(684.0)
(670.5)
Changes in Forest Carbon Stocks3
(821.4)
(714.2)
(710.7)
(704.4)
(649.3)
(707.4)
(695.4)
Non-C02 Emissions from Forest Firesb
5.5
18.3
15.0
11.0
10.8
23.0
24.4
N20 Emissions from Forest Soilsc
0.1
0.4
0.4
0.4
0.4
0.4
0.4
Non-C02 Emissions from Drained Organic
Soilsd
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Land Converted to Forest Land
(98.5)
(98.4)
(98.3)
(98.3)
(98.3)
(98.3)
(98.3)
Changes in Forest Carbon Stockse
(98.5)
(98.4)
(98.3)
(98.3)
(98.3)
(98.3)
(98.3)
Cropland Remaining Cropland
(23.2)
(29.0)
(22.3)
(16.6)
(14.5)
(23.3)
(18.9)
Changes in Mineral and Organic Soil
Carbon Stocks
(23.2)
(29.0)
(22.3)
(16.6)
(14.5)
(23.3)
(18.9)
Land Converted to Cropland
54.8
54.7
56.6
56.3
56.3
56.7
56.5
Changes in all Ecosystem Carbon Stocks'
54.8
54.7
56.6
56.3
56.3
56.7
56.5
Grassland Remaining Grassland
8.8
11.7
11.6
11.9
14.6
6.7
10.6
Changes in Mineral and Organic Soil
Carbon Stocks
8.7
11.0
10.9
11.3
14.0
6.0
10.0
Non-C02 Emissions from Grassland Fires5
0.2
0.7
0.6
0.6
0.6
0.6
0.6
Land Converted to Grassland
(6.7)
(40.1)
(24.5)
(24.2)
(23.3)
(25.9)
(24.7)
Changes in all Ecosystem Carbon Stocks'
(6.7)
(40.1)
(24.5)
(24.2)
(23.3)
(25.9)
(24.7)
Wetlands Remaining Wetlands
41.5
43.1
41.8
41.8
41.8
41.8
41.8
Changes in Organic Soil Carbon Stocks in
Peatlands
1.1
1.1
0.8
0.8
0.8
0.7
0.7
Non-C02 Emissions from Peatlands
Remaining Peatlands
+
+
+
+
+
+
+
Changes in Biomass, DOM, and Soil
Carbon Stocks in Coastal Wetlands
(8.4)
(7.7)
(8.8)
(8.8)
(8.8)
(8.8)
(8.8)
CH4 Emissions from Coastal Wetlands
Remaining Coastal Wetlands
4.2
4.2
4.3
4.3
4.3
4.3
4.3
N20 Emissions from Coastal Wetlands
Remaining Coastal Wetlands
0.1
0.2
0.1
0.1
0.1
0.1
0.1
CH4 Emissions from Flooded Land
Remaining Flooded Land
44.6
45.3
45.4
45.4
45.4
45.4
45.4
Land Converted to Wetlands
3.3
1.4
0.8
0.8
0.8
0.6
0.6
Changes in Biomass, DOM, and Soil
Carbon Stocks in Land Converted to
Coastal Wetlands
0.5
0.5
(+)
(+)
(+)
(+)
(+)
CH4 Emissions from Land Converted to
Coastal Wetlands
0.3
0.3
0.2
0.2
0.2
0.2
0.2
Changes in Land Converted to Flooded
Land
1.4
0.4
0.4
0.4
0.4
0.3
0.3
CH4 Emissions from Land Converted to
Flooded Land
1.1
0.3
0.3
0.3
0.3
0.2
0.2
Settlements Remaining Settlements
(107.8)
(113.9)
(125.6)
(125.0)
(124.5)
(131.6)
(132.5)
Changes in Organic Soil Carbon Stocks
11.3
12.2
16.0
15.9
15.9
15.9
15.9
Changes in Settlement Tree Carbon
Stocks
(96.4)
(117.4)
(129.6)
(129.5)
(129.3)
(136.7)
(137.8)
N20 Emissions from Settlement Soilsh
1.8
2.8
1.9
2.0
2.0
2.0
2.1
Changes in Yard Trimming and Food
Scrap Carbon Stocks in Landfills
(24.5)
(11.4)
(13.8)
(13.4)
(13.1)
(12.8)
(12.6)
Land Converted to Settlements
62.5
85.0
80.9
81.0
81.1
81.0
81.0
Changes in all Ecosystem Carbon Stocks'
62.5
85.0
80.9
81.0
81.1
81.0
81.0
LULUCF Emissions'
57.9
72.4
68.3
64.4
64.2
76.4
77.8
ch4
53.5
61.3
60.1
57.3
56.9
65.4
66.0
6-4 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
N20 4A 11.1 8.3 7.0 7.3 11.0 11.8
LULUCF Carbon Stock Change1'
(938.9)
(853.5)
(842.5)
(829.5)
(768.2)
(852.5)
(832.0)
LULUCF Sector Net Totalk
(881.0)
(781.1)
(774.2)
(765.1)
(704.0)
(776.2)
(754.2)
+ Absolute value does not exceed 0.05 MMT C02 Eq.
a Includes the net changes to carbon stocks stored in all forest ecosystem pools (estimates include C stock changes from
drained organic soils from both Forest Land Remaining Forest Land and Land Converted to Forest Land) and harvested
wood products.
b Estimates include CH4 and N20 emissions from fires on both Forest Land Remaining Forest Land and Land Converted to
Forest Land.
c Estimates include N20 emissions from N fertilizer additions on both Forest Land Remaining Forest Land and Land
Converted to Forest Land.
d Estimates include CH4 and N20 emissions from drained organic soils on both Forest Land Remaining Forest Land and
Land Converted to Forest Land. Carbon stock changes from drained organic soils are included with the Forest Land
Remaining Forest Land forest ecosystem pools.
e Includes the net changes to carbon stocks stored in all forest ecosystem pools.
f Includes changes in mineral and organic soil carbon stocks for all land-use conversions to cropland, grassland, and
settlements. Also includes aboveground/belowground biomass, dead wood, and litter carbon stock changes for
conversion of forest land to cropland, grassland, and settlements.
s Estimates include CH4 and N20 emissions from fires on both Grassland Remaining Grassland and Land Converted to
Grassland.
h Estimates include N20 emissions from N fertilizer additions on both Settlements Remaining Settlements and Land
Converted to Settlements because it is not possible to separate the activity data at this time.
' LULUCF emissions include the CH4 and N20 emissions reported for Peatlands Remaining Peatlands, forest fires, drained
organic soils, grassland fires, and Coastal Wetlands Remaining Coastal Wetlands; CH4 emissions from Land Converted to
Coastal Wetlands, Flooded Land Remaining Flooded Land, and Land Converted to Flooded Land; and N20 emissions from
forest soils and settlement soils.
' LULUCF Carbon Stock Change includes any C stock gains and losses from all land use and land-use conversion categories.
k The LULUCF Sector Net Total is the net sum of all LULUCF CH4 and N20 emissions to the atmosphere plus LULUCF net
carbon stock changes in units of MMT C02 Eq.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
The C stock changes and emissions of CH4 and N2O from LULUCF are summarized in Table 6-2 (MMT CO2 Eq.) and
Table 6-3 (kt). Total net C sequestration in the LULUCF sector decreased by approximately 11.4 percent between
1990 and 2021. This decrease was primarily due to a decline in the rate of net C accumulation in Forest Land, as
well as an increase in emissions from Land Converted to Settlements.4 Specifically, there was a net C accumulation
in Settlements Remaining Settlements, which increased from 1990 to 2021, while the net C accumulation in Forest
Land Remaining Forest Land and Land Converted to Wetlands slowed over this period. Net C accumulation
remained steady from 1990 to 2021 in Land Converted to Forest Land, Cropland Remaining Cropland, Land
Converted to Cropland, and Wetlands Remaining Wetlands, while net C accumulation fluctuated in Grassland
Remaining Grassland.
Flooded Land Remaining Flooded Land was the largest source of CH4 emissions from LULUCF in 2021, totaling 45.4
MMT CO2 Eq. (1,623 kt of CH4). Forest fires resulted in CH4 emissions of 15.5 MMT CO2 Eq. 554 kt of CH4). Coastal
Wetlands Remaining Coastal Wetlands resulted in CH4 emissions of 4.3 MMT CO2 Eq. (154 kt of CH4). Grassland
fires resulted in CH4 emissions of 0.3 MMT CO2 Eq. (12 kt of CH4). Land Converted to Flooded Land and Land
Converted to Wetlands each resulted in CH4 emissions of 0.2 MMT CO2 Eq. (6 kt of CH4). Drained organic soils on
forest lands and Peatlands Remaining Peatlands resulted in CH4 emissions of less than 0.05 MMT CO2 Eq. each.
For N2O emissions, forest fires were the largest source from LULUCF in 2021, totaling 8.9 MMT CO2 Eq. (34 kt of
N2O). Nitrous oxide emissions from fertilizer application to settlement soils in 2021 totaled to 2.1 MMT CO2 Eq. (8
kt of N2O). This represents an increase of 14.9percent since 1990. Additionally, the application of synthetic
4 Carbon sequestration estimates are net figures. The C stock in a given pool fluctuates due to both gains and losses. When
losses exceed gains, the C stock decreases, and the pool acts as a source. When gains exceed losses, the C stock increases, and
the pool acts as a sink; also referred to as net C sequestration or removal.
Land Use, Land-Use Change, and Forestry 6-5
-------
1 fertilizers to forest soils in 2021 resulted in N2O emissions of 0.4 MMT CO2 Eq. (2 kt of N2O). Nitrous oxide
2 emissions from fertilizer application to forest soils have increased by 455.1 percent since 1990, but still account for
3 a relatively small portion of overall emissions. Grassland fires resulted in N2O emissions of 0.3 MMT CO2 Eq. (1 kt of
4 N2O). Coastal Wetlands Remaining Coastal Wetlands resulted in N2O emissions of 0.1 MMT CO2 Eq. (1 kt of N2O).
5 Drained organic soils on forest lands resulted in N2O emissions of 0.1 MMT CO2 Eq. (less than 0.05 kt of N2O), and
6 Peatlands Remaining Peatlands resulted in N2O emissions of less than 0.05 MMT CO2 Eq.
7 Table 6-2: Emissions and Removals from Land Use, Land-Use Change, and Forestry by Gas
8 (MMT COz Eq.)
Gas/Land-Use Category
1990
2005
2017
2018
2019
2020
2021
Carbon Stock Change (C02)a
(938.9)
(853.5)
(842.5)
(829.5)
(768.2)
(852.5)
(832.0)
Forest Land Remaining Forest Land
(821.4)
(714.2)
(710.7)
(704.4)
(649.3)
(707.4)
(695.4)
Land Converted to Forest Land
(98.5)
(98.4)
(98.3)
(98.3)
(98.3)
(98.3)
(98.3)
Cropland Remaining Cropland
(23.2)
(29.0)
(22.3)
(16.6)
(14.5)
(23.3)
(18.9)
Land Converted to Cropland
54.8
54.7
56.6
56.3
56.3
56.7
56.5
Grassland Remaining Grassland
8.7
11.0
10.9
11.3
14.0
6.0
10.0
Land Converted to Grassland
(6.7)
(40.1)
(24.5)
(24.2)
(23.3)
(25.9)
(24.7)
Wetlands Remaining Wetlands
(7.4)
(6.60)
(7.95)
(7.99)
(8.03)
(8.06)
(8.09)
Land Converted to Wetlands
1.9
0.8
0.3
0.3
0.3
0.3
0.3
Settlements Remaining Settlements
(109.6)
(116.6)
(127.5)
(127.0)
(126.5)
(133.6)
(134.5)
Land Converted to Settlements
62.5
85.0
80.9
81.0
81.1
81.0
81.0
ch4
53.5
61.3
60.1
57.3
56.9
65.4
66.0
Forest Land Remaining Forest Land:
Forest Firesb
3.2
10.9
9.6
6.9
6.4
15.0
15.5
Forest Land Remaining Forest Land:
Drained Organic Soilsd
+
+
+
+
+
+
+
Grassland Remaining Grassland:
Grassland Firesc
0.1
0.4
0.3
0.3
0.3
0.3
0.3
Wetlands Remaining Wetlands: Flooded
Land Remaining Flooded Land
44.6
45.3
45.4
45.4
45.4
45.4
45.4
Wetlands Remaining Wetlands: Coastal
Wetlands Remaining Coastal Wetlands
4.2
4.2
4.3
4.3
4.3
4.3
4.3
Wetlands Remaining Wetlands:
Peatlands Remaining Peatlands
+
+
+
+
+
+
+
Land Converted to Wetlands: Land
Converted to Flooded Land
1.1
0.3
0.3
0.3
0.3
0.2
0.2
Land Converted to Wetlands: Land
Converted to Coastal Wetlands
0.3
0.3
0.2
0.2
0.2
0.2
0.2
n2o
4.4
11.1
8.3
7.0
7.3
11.0
11.8
Forest Land Remaining Forest Land:
Forest Firesb
2.3
7.4
5.4
4.2
4.4
8.0
8.9
Forest Land Remaining Forest Land:
Forest Soils'
0.1
0.4
0.4
0.4
0.4
0.4
0.4
Forest Land Remaining Forest Land:
Drained Organic Soilsd
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Grassland Remaining Grassland:
Grassland Firesc
0.1
0.3
0.3
0.3
0.3
0.3
0.3
Wetlands Remaining Wetlands: Coastal
Wetlands Remaining Coastal Wetlands
0.1
0.2
0.1
0.1
0.1
0.1
0.1
Wetlands Remaining Wetlands:
Peatlands Remaining Peatlands
+
+
+
+
+
+
+
Settlements Remaining Settlements:
Settlement Soilse
1.8
2.8
1.9
2.0
2.0
2.0
2.1
LULUCF Carbon Stock Change3
(938.8)
(853.5)
(842.5)
(829.5)
(768.2)
(852.5)
(832.0)
LULUCF Emissions8
57.9
72.4
68.3
64.4
64.2
76.4
77.8
LULUCF Sector Net Totalh
(881.0)
(781.1)
(774.2)
(765.1)
(704.0)
(776.2)
(754.2)
6-6 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
+ Absolute value does not exceed 0.05 MMT C02 Eq.
a LULUCF Carbon Stock Change is the net C stock change from the following categories: Forest Land Remaining Forest
Land, Land Converted to Forest Land, Cropland Remaining Cropland, Land Converted to Cropland, Grassland Remaining
Grassland, Land Converted to Grassland, Wetlands Remaining Wetlands, Land Converted to Wetlands, Settlements
Remaining Settlements, and Land Converted to Settlements.
b Estimates include CH4 and N20 emissions from fires on both Forest Land Remaining Forest Land and Land Converted to
Forest Land.
c Estimates include CH4 and N20 emissions from drained organic soils on both Forest Land Remaining Forest Land and
Land Converted to Forest Land.
d Estimates include CH4 and N20 emissions from fires on both Grassland Remaining Grassland and Land Converted to
Grassland.
e Estimates include N20 emissions from N fertilizer additions on both Forest Land Remaining Forest Land and Land
Converted to Forest Land.
f Estimates include N20 emissions from N fertilizer additions on both Settlements Remaining Settlements and Land
Converted to Settlements.
s LULUCF emissions include the CH4 and N20 emissions reported for Peatlands Remaining Peatlands, forest fires, drained
organic soils, grassland fires, and Coastal Wetlands Remaining Coastal Wetlands; CH4 emissions from Flooded Land
Remaining Flooded Land, Land Converted to Flooded Land, and Land Converted to Coastal Wetlands; and N20 emissions
from forest soils and settlement soils.
h The LULUCF Sector Net Total is the net sum of all LULUCF CH4 and N20 emissions to the atmosphere plus LULUCF net
carbon stock changes in units of MMT C02 Eq.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
1 Table 6-3: Emissions and Removals from Land Use, Land-Use Change, and Forestry by Gas
2 (kt)
Gas/Land-Use Category
1990
2005
2017
2018
2019
2020
2021
Carbon Stock Change (C02)a
(938,856)
(853,529)
(842,516)
(829,501)
(768,224)
(852,534)
(832,039)
Forest Land Remaining Forest Land
(821,444)
(714,232)
(710,697)
(704,446)
(649,336)
(707,426)
(695,354)
Land Converted to Forest Land
(98,452)
(98,429)
(98,322)
(98,263)
(98,253)
(98,254)
(98,254)
Cropland Remaining Cropland
(23,176)
(29,001)
(22,293)
(16,597)
(14,544)
(23,335)
(18,940)
Land Converted to Cropland
54,792
54,651
56,597
56,327
56,280
56,725
56,511
Grassland Remaining Grassland
8,694
11,040
10,928
11,266
13,997
6,046
10,005
Land Converted to Grassland
(6,684)
(40,098)
(24,467)
(24,205)
(23,304)
(25,921)
(24,669)
Wetlands Remaining Wetlands
(7,372)
(6,601)
(7,953)
(7,990)
(8,031)
(8,059)
(8,095)
Land Converted to Wetlands
1884
820
339
341
349
250
256
Settlements Remaining Settlements
(109,567)
(116,642)
(127,510)
(126,961)
(126,469)
(133,610)
(134,514)
Land Converted to Settlements
62,469
84,965
80,860
81,026
81,087
81,050
81,014
ch4
1,911
2,190
2,145
2,048
2,032
2,336
2,356
Forest Land Remaining Forest Land:
Forest Firesb
116
390
342
245
228
534
554
Forest Land Remaining Forest Land:
Drained Organic Soilsd
1
1
1
1
1
1
1
Grassland Remaining Grassland:
Grassland Firesc
3
13
12
12
12
12
12
Wetlands Remaining Wetlands:
Flooded Land Remaining Flooded
Land
1,592.8
1,617.0
1,620.7
1,620.8
1,620.9
1,622.7
1,622.8
Wetlands Remaining Wetlands:
Coastal Wetlands Remaining Coastal
Wetlands
149
151
153
153
153
154
154
Wetlands Remaining Wetlands:
Peatlands Remaining Peatlands
+
+
+
+
+
+
+
Land Converted to Wetlands: Land
Converted to Flooded Land
39
9
9
9
9
6
6
Land Converted to Wetlands: Land
Converted to Coastal Wetlands
10
10
8
7
7
7
6
Land Use, Land-Use Change, and Forestry 6-7
-------
n2o
17
42
31
27
27
41
45
Forest Land Remaining Forest Land:
Forest Firesb
9
28
21
16
17
30
34
Forest Land Remaining Forest Land:
Forest Soils'
+
2
2
2
2
2
2
Forest Land Remaining Forest Land:
Drained Organic Soilsd
+
+
+
+
+
+
+
Grassland Remaining Grassland:
Grassland Firesc
+
1
1
1
1
1
1
Wetlands Remaining Wetlands:
Coastal Wetlands Remaining Coastal
Wetlands
+
1
+
1
1
1
1
Wetlands Remaining Wetlands:
Peatlands Remaining Peatlands
+
+
+
+
+
+
+
Settlements Remaining Settlements:
Settlement Soilse
7
10
7
7
8
8
8
+ Absolute value does not exceed 0.5 kt.
a LULUCF Carbon Stock Change is the net C stock change from the following categories: Forest Land Remaining Forest Land, Land
Converted to Forest Land, Cropland Remaining Cropland, Land Converted to Cropland, Grassland Remaining Grassland, Land
Converted to Grassland, Wetlands Remaining Wetlands, Land Converted to Wetlands, Settlements Remaining Settlements, and
Land Converted to Settlements.
b Estimates include CH4 and N20 emissions from fires on both Forest Land Remaining Forest Land and Land Converted to Forest
Land.
c Estimates include CH4 and N20 emissions from drained organic soils on both Forest Land Remaining Forest Land and Land
Converted to Forest Land.
d Estimates include CH4 and N20 emissions from fires on both Grassland Remaining Grassland and Land Converted to Grassland.
e Estimates include N20 emissions from N fertilizer additions on both Forest Land Remaining Forest Land and Land Converted to
Forest Land.
f Estimates include N20 emissions from N fertilizer additions on both Settlements Remaining Settlements and Land Converted to
Settlements.
Notes: Totals by gas may not sum due to independent rounding. Parentheses indicate net sequestration.
1 Each year, some emission and sink estimates in the LULUCF sector of the Inventory are recalculated and revised
2 with improved methods and/or data. In general, recalculations are made to the U.S. greenhouse gas emissions and
3 sinks estimates either to incorporate new methodologies or, most commonly, to update recent historical data.
4 These improvements are implemented consistently across the previous Inventory's time series (i.e., 1990 to 2020)
5 to ensure that the trend is accurate. Of the updates implemented for this Inventory, the most significant include
6 (1) Flooded Land Remaining Flooded Land and Land Converted to Flooded Land: the National Wetland Inventory
7 (NWI) is now used as the primary data source for flooded land surface area rather than the National Hydrography
8 Data (NHD as the primary geospatial data source, (2) Forest Lands: use of new data from the National Forest
9 Inventory (NFI) as well as updated fire data and harvested wood products' (HWP) data, and using plot-level soil
10 orders based on the more refined gridded National Soil Survey Geographic Database (gNATSGO) dataset rather
11 than the Digital General Soil Map of the United States (STATSG02) dataset which had been used in previous
12 Inventories; and (3) Coastal Wetlands: an update was made to the activity data to remove any estuarine forested
13 wetland areas that were located outside of states classified as subtropical since those wetlands fall under Forest
14 Land Remaining Forest Land and to remove any estuarine forested wetland areas that were located outside of
15 states classified as subtropical since, states classified as wet temperate, cold temperate and Mediterranean
16 climate zones fall under the category of Land Converted to Forest Land. Together, these updates for 2020
17 decreased total C sequestration by 40.4 MMT CO2 Eq. (5.0 percent) and increased total non-CC>2 emissions by 23.4
18 MMT CO2 Eq. (52.6 percent), compared to the previous Inventory (i.e., 1990 to 2020). In addition, for the current
19 Inventory, CC>2-equivalent emissions totals of CFU and N2O have been revised to reflect the 100-year global
20 warming potentials (GWPs) provided in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). Further discussion on
21 this update and the overall impacts of updating the Inventory GWP values to reflect the IPCC Fifth Assessment
22 Report can be found in Chapter 9, Recalculations and Improvements.
6-8 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
For more information on specific methodological updates, please see the Recalculations discussion within the
respective source category section of this chapter.
Emissions and removals reported in the LULUCF chapter include those from all states; however, for Hawaii and
Alaska some emissions and removals from land use and land-use change are not included (see chapter sections on
Uncertainty and Planned Improvements for more details). In addition, U.S. Territories are not included for most
categories. EPA continues to review available data on an ongoing basis to include emissions and removals from
U.S. Territories in future inventories to the extent they are occurring (e.g., see Box 6-2). See Annex 5 for more
information on EPA's assessment of the emissions and removals not included in this Inventory.
Box 6-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals
In following the United Nations Framework Convention on Climate Change (UNFCCC) requirement under Article
4.1 to develop and submit national greenhouse gas emission inventories, the gross emissions total presented in
this report for the United States excludes emissions and removals from LULUCF. The LULUCF Sector Net Total
presented in this report for the United States includes emissions and removals from LULUCF. All emissions and
removals estimates are calculated using internationally accepted methods provided by the IPCC in the 2006
IPCC Guidelines for National Greenhouse Gas Inventories (2006 IPCC Guidelines), 2013 Supplement to the 2006
IPCC Guidelines for National GHG Inventories: Wetlands, and the 2019 Refinement to the 2006 IPCC Guidelines
for National GHG Inventories. Additionally, the calculated emissions and removals in a given year for the United
States are presented in a common manner in line with the UNFCCC reporting guidelines for the reporting of
inventories under this international agreement.5 The use of consistent methods to calculate emissions and
removals by all nations providing their inventories to the UNFCCC ensures that these reports are comparable.
The presentation of emissions and removals provided in the Land Use Land-Use Change and Forestry chapter
does not preclude alternative examinations, but rather, this chapter presents emissions and removals in a
common format consistent with how countries are to report Inventories under the UNFCCC. The report itself,
and this chapter, follow this standardized format, and provides an explanation of the application of methods
used to calculate emissions and removals.
6.1 Representation of the U.S. Land Base
A national land use representation system that is consistent and complete, both temporally and spatially, is
needed in order to assess land use and land-use change status and the associated greenhouse gas fluxes over the
Inventory time series. This system should be consistent with IPCC (2006), such that all countries reporting on
national greenhouse gas fluxes to the UNFCCC should: (1) describe the methods and definitions used to determine
areas of managed and unmanaged lands in the country (Table 6-4), (2) describe and apply a consistent set of
definitions for land-use categories over the entire national land base and time series (i.e., such that increases in
the land areas within particular land-use categories are balanced by decreases in the land areas of other categories
unless the national land base is changing) (Table 6-5), and (3) account for greenhouse gas fluxes on all managed
lands. The IPCC (2006, Vol. IV, Chapter 1) considers all anthropogenic greenhouse gas emissions and removals
associated with land use and management to occur on managed land, and all emissions and removals on managed
land should be reported based on this guidance (See IPCC (2010), Ogle et al. (2018) for further discussion).
Consequently, managed land serves as a proxy for anthropogenic emissions and removals. This proxy is intended
to provide a practical framework for conducting an inventory, even though some of the greenhouse gas emissions
and removals on managed land are influenced by natural processes that may or may not be interacting with the
^ See http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
Land Use, Land-Use Change, and Forestry 6-9
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
anthropogenic drivers. This section of the Inventory has been developed in order to comply with this guidance.
While the 2019 Refinement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories provide guidance
for factoring out natural emissions and removals, the United States does not apply this guidance and estimates all
emissions/removals on managed land regardless of whether the driver was natural.
Three databases are used to track land management in the United States and are used as the basis to classify
United States land area into the thirty-six IPCC land use and land-use change categories (Table 6-5) (IPCC 2006).
The three primary databases are the U.S. Department of Agriculture (USDA) National Resources Inventory (NRI),6
the USDA Forest Service (USFS) Forest Inventory and Analysis (FIA)7 Database, and the Multi-Resolution Land
Characteristics Consortium (MRLC) National Land Cover Dataset (NLCD).8
The total land area included in the United States Inventory is 936 million hectares across the 50 states.9
Approximately 886 million hectares of this land base is considered managed and 50 million hectares is
unmanaged, a distribution that has remained stable over the time series of the Inventory (Table 6-5). In 2021, the
United States had a total of 280 million hectares of managed forest land (0.71 percent decrease compared to
1990). There are 160 million hectares of cropland (8.3 percent decrease compared to 1990), 339 million hectares
of managed Grassland (0.4 percent increase compared to 1990), 39 million hectares of managed Wetlands (4.6
percent increase compared to 1990), 47 million hectares of Settlements (41 percent increase compared to 1990),
and 21 million hectares of managed Other Land (1.0 percent decrease compared to 1990) (Table 6-5).
Wetlands are not differentiated between managed and unmanaged with the exception of remote areas in Alaska,
and so are reported mostly as managed.10 In addition, C stock changes are not currently estimated for the entire
managed land base, which leads to discrepancies between the managed land area data presented here and in the
subsequent sections of the Inventory (e.g., Grassland Remaining Grassland within interior Alaska).1112 Planned
improvements are under development to estimate C stock changes and greenhouse gas emissions on all managed
land and to ensure consistency between the total area of managed land in the land-representation description and
the remainder of the Inventory.
Dominant land uses vary by region, largely due to climate patterns, soil types, geology, proximity to coastal
regions, and historical settlement patterns (Figure 6-3). Forest land tends to be more common in the eastern
United States, mountainous regions of the western United States, and Alaska. Cropland is concentrated in the mid-
continent region of the United States, and Grassland is more common in the western United States and Alaska.
Wetlands are fairly ubiquitous throughout the United States, though they are more common in the upper Midwest
6 NRI data are available at https://www.nrcs.usda.Eov/wps/portal/nrcs/main/national/techriical/nra/nri/.
7 FIA data are available at https://www.fia.fs.usda.gov/tools-data/index.php.
8 NLCD data are available at http://www.mrlc.gov/ and MRLC is a consortium of several U.S. government agencies.
9 The current land representation does not include areas from U.S. Territories, but there are planned improvements to include
these regions in future Inventories. U.S. Territories represent approximately 0.1 percent of the total land base for the United
States. See Box 6-2.
10 According to the IPCC (2006), wetlands are considered managed if they are created through human activity, such as dam
construction, or the water level is artificially altered by human activity. Distinguishing between managed and unmanaged
wetlands in the conterminous United States and Alaska is difficult due to limited data availability. Wetlands are not
characterized within the NRI with information regarding water table management. As a result, all Wetlands in the conterminous
United States and Hawaii are reported as managed in the Land Representation, but emission/removal estimates only developed
for those wetlands that are included under the Flooded Lands, Coastal Wetlands or Peat Extraction categories. See the Planned
Improvements section of the Inventory for future refinements to the Wetland area estimates.
11 Other discrepancies occur because the coastal wetlands analysis is based on another land use product (NOAA C-CAP) that is
not currently incorporated into the land representation analysis for this section, which relies on the NRI and NLCD for wetland
areas. EPA anticipates addressing these discrepancies in future Inventories.
12 These "managed area" discrepancies also occur in the Common Reporting Format (CRF) tables submitted to the UNFCCC.
6-10 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 and eastern portions of the country, as well as coastal regions. Settlements are more concentrated along the
2 coastal margins and in the eastern states.
3 Table 6-4: Managed and Unmanaged Land Area by Land-Use Categories for All 50 States
4 (Thousands of Hectares)
Land Use Categories
1990
2005
2017
2018
2019
2020
2021
Managed Lands
886,533
886,530
886,531
886,531
886,531
886,531
886,531
Forest
282,357
281,755
281,057
280,870
280,686
280,519
280,363
Croplands
174,496
165,622
161,922
161,394
160,693
160,111
160,077
Grasslands
337,639
339,694
338,053
338,264
338,722
339,138
338,989
Settlements
33,427
40,210
45,595
45,972
46,306
46,654
46,970
Wetlands
37,704
38,661
39,108
39,251
39,380
39,382
39,438
Other
20,910
20,588
20,796
20,779
20,743
20,727
20,693
Unmanaged Lands
49,708
49,711
49,710
49,710
49,710
49,710
49,710
Forest
10,260
10,260
10,264
10,264
10,264
10,264
10,269
Croplands
0
0
0
0
0
0
0
Grasslands
24,666
24,686
24,696
24,696
24,696
24,696
24,691
Settlements
0
0
0
0
0
0
0
Wetlands
4,048
4,047
4,058
4,058
4,058
4,058
4,058
Other
10,734
10,718
10,692
10,692
10,692
10,692
10,692
Total Land Areas
936,241
936,241
936,241
936,241
936,241
936,241
936,241
Forest
292,617
292,016
291,321
291,134
290,951
290,782
290,632
Croplands
174,496
165,622
161,922
161,394
160,693
160,111
160,077
Grasslands
362,305
364,380
362,749
362,960
363,417
363,834
363,680
Settlements
33,427
40,210
45,595
45,972
46,307
46,654
46,971
Wetlands
41,752
42,708
43,167
43,310
43,439
43,441
43,496
Other
31,644
31,306
31,488
31,471
31,435
31,419
31,385
5
6 Table 6-5: Land Use and Land-Use Change for the U.S. Managed Land Base for All 50 States
7 (Thousands of Hectares)
Land Use & Land-Use
Change Categories3
1990
2005
2017
2018
2019
2020
2021
Total Forest Land
282,357
281,755
281,057
280,870
280,686
280,519
280,363
FF
281,232
280,457
279,841
279,778
279,616
279,446
279,298
CF
216
154
110
101
87
83
82
GF
805
1,028
959
855
862
867
869
WF
13
23
19
19
16
15
14
SF
11
18
19
19
19
19
19
OF
79
77
108
99
86
89
81
Total Cropland
174,496
165,622
161,922
161,394
160,693
160,111
160,077
CC
162,265
150,400
148,327
149,721
149,504
149,817
150,586
FC
178
83
64
63
63
63
66
GC
11,673
14,623
13,121
11,231
10,758
9,914
9,132
WC
119
178
102
99
98
86
81
SC
75
102
122
107
105
101
97
OC
186
235
186
173
166
129
115
Total Grassland
337,639
339,694
338,053
338,264
338,722
339,138
338,989
GG
328,320
316,625
318,704
321,748
322,632
323,883
325,096
FG
591
642
722
733
746
726
704
CG
8,177
17,746
16,075
13,594
13,491
13,205
12,200
WG
168
466
199
181
172
159
143
SG
43
525
283
230
190
139
100
OG
341
3,692
2,070
1,778
1,491
1,026
746
Total Wetlands
37,704
38,661
39,108
39,251
39,380
39,382
39,438
Land Use, Land-Use Change, and Forestry 6-11
-------
WW
37,148
36,636
37,727
38,020
38,283
38,426
38,613
FW
38
73
71
69
57
57
51
CW
145
637
403
362
310
261
221
GW
326
1,169
662
564
501
415
342
SW
0
38
21
17
14
10
2
OW
47
107
225
220
216
212
210
Total Settlements
33,427
40,210
45,595
45,972
46,306
46,654
46,970
SS
30,561
31,445
39,875
40,771
41,617
42,467
43,189
FS
301
503
483
467
449
460
456
CS
1,231
3,604
2,110
1,917
1,726
1,528
1,366
GS
1,276
4,371
2,919
2,630
2,349
2,062
1,830
WS
4
59
39
30
25
18
14
OS
54
229
169
157
141
120
115
Total Other Land
20,910
20,588
20,796
20,779
20,743
20,727
20,693
00
20,175
17,019
17,874
18,059
18,305
18,563
18,817
FO
53
81
97
96
98
100
106
CO
287
603
670
629
582
540
489
GO
371
2,764
1,929
1,772
1,541
1,309
1,068
WO
22
100
208
206
206
205
204
SO
2
21
18
17
11
10
10
Grand Total
886,533
886,530
886,531
886,531
886,531
886,531
886,531
6-12 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Figure 6-3: Percent of Total Land Area for Each State in the General Land Use Categories for
2 2021
Forest Lands
Croplands
Other Lands
Land Use, Land-Use Change, and Forestry 6-13
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Methodology and Time-Series Consistency
IPCC (2006) describes three approaches for representing land areas. Approach 1 provides data on the total area for
each individual land use category, but does not provide detailed information on transfer of land area between
categories following land-use change and is not spatially explicit other than at the national or regional level. With
Approach 1, total net conversions between categories can be detected, but not the individual changes (i.e.,
additions and/or losses) between the land-use categories that led to those net changes. Approach 2 introduces
tracking of individual land-use changes between the categories (e.g., Forest Land to Cropland, Cropland to Forest
Land, and Grassland to Cropland), using survey samples or other forms of data, but does not provide spatially-
explicit location data. Approach 3 extends Approach 2 by providing spatially-explicit location data, such as surveys
with spatially identified sample locations and maps obtained from remote sensing products. The three approaches
are not presented as hierarchical tiers and are not mutually exclusive.
According to IPCC (2006), the approach or mix of approaches selected by an inventory agency should reflect
calculation needs and national circumstances. For this analysis, the NRI, FIA, and the NLCD have been combined to
provide a complete representation of land use for managed lands. These data sources are described in more detail
later in this section. NRI, FIA and NLCD are Approach 3 data sources that provide spatially-explicit representations
of land use and land-use conversions. Lands are treated as remaining in the same category (e.g., Cropland
Remaining Cropland) if a land-use change has not occurred in the last 20 years. Otherwise, the land is classified in a
land-use change category based on the current use and most recent use before conversion to the current use (e.g.,
Cropland Converted to Forest Land).
Definitions of Land Use in the United States
Managed and Unmanaged Land
The United States definition of managed land is similar to the general definition of managed land provided by the
IPCC (2006), but with some additional elaboration to reflect national circumstances. Based on the following
definitions, most lands in the United States are classified as managed:
• Managed Land: Land is considered managed if direct human intervention has influenced its condition.
Direct intervention occurs mostly in areas accessible to human activity and includes altering or
maintaining the condition of the land to produce commercial or non-commercial products or services; to
serve as transportation corridors or locations for buildings, landfills, or other developed areas for
commercial or non-commercial purposes; to extract resources or facilitate acquisition of resources; or to
provide social functions for personal, community, or societal objectives where these areas are readily
accessible to society.13
• Unmanaged Land: All other land is considered unmanaged. Unmanaged land is largely comprised of areas
inaccessible to society due to the remoteness of the locations. Though these lands may be influenced
13 Wetlands are an exception to this general definition, because these lands, as specified by IPCC (2006), are only considered
managed if they are created through human activity, such as dam construction, or the water level is artificially altered by
human activity. Distinguishing between managed and unmanaged wetlands in the United States is difficult due to limited data
availability. Wetlands are not characterized within the NRI with information regarding water table management or origin (i.e.,
constructed rather than natural origin). Therefore, unless wetlands are converted into cropland or grassland, it is not possible
to know if they are artificially created or if the water table is managed based on the use of NRI data. As a result, most wetlands
are reported as managed with the exception of wetlands in remote areas of Alaska, but emissions from managed wetlands are
only reported for coastal regions, flooded lands (e.g., reservoirs) and peatlands where peat extraction occurs due to insufficient
activity data to estimate emissions and limited resources to improve the Inventory. See the Planned Improvements section of
the Inventory for future refinements to the wetland area estimates.
6-14 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
indirectly by human actions such as atmospheric deposition of chemical species produced in industry or
CO2 fertilization, they are not influenced by a direct human intervention.14
In addition, land that is previously managed remains in the managed land base for 20 years before re-classifying
the land as unmanaged in order to account for legacy effects of management on C stocks.15 Unmanaged land is
also re-classified as managed over time if anthropogenic activity is introduced into the area based on the definition
of managed land.
Land-Use Categories
As with the definition of managed lands, IPCC (2006) provides general non-prescriptive definitions for the six main
land-use categories: Forest Land, Cropland, Grassland, Wetlands, Settlements and Other Land. In order to reflect
national circumstances, country-specific definitions have been developed, based predominantly on criteria used in
the land use surveys for the United States. Specifically, the definition of Forest Land is based on the FIA definition
of forest,16 while definitions of Cropland, Grassland, and Settlements are based on the NRI.17The definitions for
Other Land and Wetlands are based on the IPCC (2006) definitions for these categories.
• Forest Land: A land-use category that includes areas at least 120 feet (36.6 meters) wide and at least one
acre (0.4 hectare) in size with at least 10 percent cover (or equivalent stocking) by live trees including land
that formerly had such tree cover and that will be naturally or artificially regenerated. Trees are woody
plants having a more or less erect perennial stem(s) capable of achieving at least 3 inches (7.6 cm) in
diameter at breast height, or 5 inches (12.7 cm) diameter at root collar, and a height of 16.4 feet (5 m) at
maturity in situ. Forest Land includes all areas recently having such conditions and currently regenerating
or capable of attaining such condition in the near future. Forest Land also includes transition zones, such
as areas between forest and non-forest lands that have at least 10 percent cover (or equivalent stocking)
with live trees and forest areas adjacent to urban and built-up lands. Unimproved roads and trails,
streams, and clearings in forest areas are classified as forest if they are less than 120 feet (36.6 m) wide or
an acre (0.4 ha) in size. However, land is not classified as Forest Land if completely surrounded by urban
or developed lands, even if the criteria are consistent with the tree area and cover requirements for
Forest Land. These areas are classified as Settlements. In addition, Forest Land does not include land that
is predominantly under an agricultural land use (Nelson et al. 2020).
• Cropland: A land-use category that includes areas used for the production of adapted crops for harvest;
this category includes both cultivated and non-cultivated lands. Cultivated crops include row crops or
close-grown crops and also pasture in rotation with cultivated crops. Non-cultivated cropland includes
continuous hay, perennial crops (e.g., orchards) and horticultural cropland. Cropland also includes land
with agroforestry, such as alley cropping and windbreaks,18 if the dominant use is crop production,
assuming the stand or woodlot does not meet the criteria for Forest Land. Lands in temporary fallow or
14 There are some areas, such as Forest Land and Grassland in Alaska that are classified as unmanaged land due to the
remoteness of their location.
15 There are examples of managed land transitioning to unmanaged land in the U.S. For example, in 2018,100 hectares of
managed grassland converted to unmanaged because data indicated that no further grazing occurred. Livestock data are
collected annually by the Department of Agriculture, and no livestock had occurred in the area since the mid-1970s, and
therefore there was no longer active management through livestock grazing. The area is also remote, at least 10 miles from
roads and settlements, and therefore the land was no longer managed based on the implementation criteria.
16 See https://www.fia.fs.usda.gov/library/field-guides-methods-proc/docs/2022/core_ver9-2_9_2022_SW_HW%20table.pdf,
page 23.
17 See https://www.nrcs.usda.gov/wps/portal/nrcs/main/national/technical/nra/nri/.
18 Currently, there is no data source to account for biomass C stock change associated with woody plant growth and losses in
alley cropping systems and windbreaks in cropping systems, although these areas are included in the Cropland land base.
Land Use, Land-Use Change, and Forestry 6-15
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
enrolled in conservation reserve programs (i.e., set-asides19) are also classified as Cropland, as long as
these areas do not meet the Forest Land criteria. Roads through Cropland, including interstate highways,
state highways, other paved roads, gravel roads, dirt roads, and railroads are excluded from Cropland
area estimates and are, instead, classified as Settlements.
• Grassland: A land-use category on which the plant cover is composed principally of grasses, grass-like
plants (i.e., sedges and rushes), forbs, or shrubs suitable for grazing and browsing, and includes both
pastures and native rangelands. This includes areas where practices such as clearing, burning, chaining,
and/or chemicals are applied to maintain the grass vegetation. Land is also categorized as Grassland if
there have been three or fewer years of continuous hay production.20 Savannas, deserts, and tundra are
considered Grassland.21 Drained wetlands are considered Grassland if the dominant vegetation meets the
plant cover criteria for Grassland. Woody plant communities of low forbs, shrubs and woodlands, such as
sagebrush, mesquite, chaparral, mountain shrubland, and pinyon-juniper, are also classified as Grassland
if they do not meet the criteria for Forest Land. Grassland includes land managed with agroforestry
practices, such as silvopasture and windbreaks, if the land is principally grass, grass-like plants, forbs, and
shrubs suitable for grazing and browsing, and assuming the stand or woodlot does not meet the criteria
for Forest Land. Roads through Grassland, including interstate highways, state highways, other paved
roads, gravel roads, dirt roads, and railroads are excluded from Grassland and are, instead, classified as
Settlements.
• Wetlands: A land-use category that includes land covered or saturated by water for all or part of the year,
in addition to lakes, reservoirs, and rivers. In addition, all coastal wetlands are considered managed
regardless of whether the water level is changed or if they were created by human activity. Certain areas
that fall under the managed Wetlands definition are included in other land uses based on the IPCC
guidance and national circumstances, including lands that are flooded for most or just part of the year in
Croplands (e.g., rice cultivation and cranberry production), Grasslands (e.g., wet meadows dominated by
grass cover) and Forest Lands (e.g., Riparian Forests near waterways). See Section 6.8 Wetlands
Remaining Wetlands for more information.
• Settlements: A land-use category representing developed areas consisting of units equal to or greater
than 0.25 acres (0.1 ha) that includes residential, industrial, commercial, and institutional land;
construction sites; public administrative sites; railroad yards; cemeteries; airports; golf courses; sanitary
landfills; sewage treatment plants; water control structures and spillways; parks within urban and built-up
areas; and highways, railroads, and other transportation facilities. Also included are all tracts that may
meet the definition of Forest Land, and tracts of less than 10 acres (4.05 ha) that may meet the definitions
for Cropland, Grassland, or Other Land but are completely surrounded by urban or built-up land, and so
are included in the Settlements category. Rural transportation corridors located within other land uses
(e.g., Forest Land, Cropland, and Grassland) are also included in Settlements.
• Other Land: 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. Following the guidance provided by the IPCC (2006), C stock
changes and non-CC>2 emissions are not estimated for Other Lands because these areas are largely devoid
of biomass, litter and soil C pools. However, C stock changes and non-C02 emissions should be estimated
for Land Converted to Other Land during the first 20 years following conversion to account for legacy
effects.
19 A set-aside is cropland that has been taken out of active cropping and converted to some type of vegetative cover, including,
for example, native grasses or trees, but is still classified as cropland based on national circumstances.
20 Areas with four or more years of continuous hay production are Cropland because the land is typically more intensively
managed with cultivation, greater amounts of inputs, and other practices. Occasional harvest of hay from grasslands typically
does not involve cultivation or other intensive management practices.
21 2006 IPCC Guidelines do not include provisions to separate desert and tundra as land-use categories.
6-16 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Land Use Data Sources: Description and Application to U.S.
Land Area Classification
U.S. Land Use Data Sources
The three main sources for land use data in the United States are the NRI, FIA, and the NLCD (Table 6-6). These
data sources are combined to account for land use in all 50 states. FIA and NRI data are used when available for an
area because these surveys contain additional information on management, site conditions, crop types, biometric
measurements, and other data that are needed to estimate C stock changes, N2O, and CH4 emissions on those
lands. If NRI and FIA data are not available for an area, however, then the NLCD product is used to represent the
land use.
Table 6-6: Data Sources Used to Determine Land Use and Land Area for the Conterminous
——V """ *
NRI
FIA
NLCD
Forest Land
Conterminous
United States
Non-Federal
Federal
Hawaii
Non-Federal
Federal
Alaska
Non-Federal
Federal
Croplands, Grasslands, Other Lands, Settlements, and Wetlands
Conterminous
United States
Non-Federal •
Federal
Hawaii
Non-Federal •
Federal
Alaska
Non-Federal
Federal
National Resources Inventory
For the Inventory, the NRI is the official source of data for land use and land-use change on non-federal lands in
the conterminous United States and Hawaii, and is also used to determine the total land base for the
conterminous United States and Hawaii. The NRI is a statistically-based survey conducted by the USDA Natural
Resources Conservation Service and is designed to assess soil, water, and related environmental resources on non-
federal lands. The NRI has a stratified multi-stage sampling design, where primary sample units are stratified on
the basis of county and township boundaries defined by the United States Public Land Survey (Nusser and Goebel
1997). Within a primary sample unit (typically a 160 acre [64.75 ha] square quarter-section), three sample points
are selected according to a restricted randomization procedure. Each point in the survey is assigned an area weight
(expansion factor) based on other known areas and land use information (Nusser and Goebel 1997). The NRI
survey utilizes data obtained from remote sensing imagery and site visits in order to provide detailed information
on land use and management, particularly for Croplands and Grasslands (i.e., agricultural lands), and is used as the
basis to account for C stock changes in agricultural lands (except federal Grasslands). The NRI survey was
conducted every 5 years between 1982 and 1997, but shifted to annualized data collection in 1998. The land use
between five-year periods from 1982 and 1997 are assumed to be the same for a five-year time period if the land
Land Use, Land-Use Change, and Forestry 6-17
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
use is the same at the beginning and end of the five-year period (Note: most of the data has the same land use at
the beginning and end of the five-year periods). If the land use had changed during a five-year period, then the
change is assigned at random to one of the five years. For crop histories, years with missing data are estimated
based on the sequence of crops grown during years preceding and succeeding a missing year in the NRI history.
This gap-filling approach allows for development of a full time series of land use data for non-federal lands in the
conterminous United States and Hawaii. This Inventory incorporates data through 2017 from the NRI. The land use
patterns are assumed to remain the same from 2018 through 2021 for this Inventory, but the time series will be
updated when new data are integrated into the land representation analysis.
Forest Inventory and Analysis
The FIA program, conducted by the USFS, is the official source of data on forest land area and management data
for the Inventory and is another statistically-based survey for the United States. FIA engages in a hierarchical
system of sampling, with sampling categorized as Phases 1 through 3, in which sample points for phases are
subsets of the previous phase. Phase 1 refers to collection of remotely-sensed data (either aerial photographs or
satellite imagery) primarily to classify land into forest or non-forest and to identify landscape patterns like
fragmentation and urbanization. Phase 2 is the collection of field data on a network of ground plots that enable
classification and summarization of area, tree, and other attributes associated with forest land uses. Phase 3 plots
are a subset of Phase 2 plots where data on indicators of forest health are measured. Data from all three phases
are also used to estimate C stock changes for forest land. Historically, FIA inventory surveys have been conducted
periodically, with all plots in a state being measured at a frequency of every five to 10 years. A new national plot
design and annual sampling design was introduced by the FIA program in 1998 and is now used in all states.
Annualized sampling means that a portion of plots throughout each state is sampled each year, with the goal of
measuring all plots once every five to seven years in the eastern United States and once every ten years in the
western United States. See Annex 3.13 to see the specific survey data available by state. The most recent year of
available data varies state by state (range of most recent data is from 2018 through 2021; see Table A-202 in
Annex 3.13).
National Land Cover Dataset
As noted above, while the NRI survey sample covers the conterminous United States and Hawaii, land use data are
only collected on non-federal lands. Gaps exist in the land representation when the NRI and FIA datasets are
combined, such as federal grasslands operated by Bureau of Land Management (BLM), USDA, and National Park
Service, as well as Alaska.22 The NLCD is used to account for land use on federal lands in the conterminous United
States and Hawaii, in addition to federal and non-federal lands in Alaska with the exception of forest lands in
Alaska.
NLCD products provide land-cover for 1992, 2001, 2004, 2006, 2008, 2011, 2013, 2016, and 2019 in the
conterminous United States (Yang et al. 2018; Fry et al. 2011; Homer et al. 2007, 2015), and also for Alaska in 2001,
2011, and 2016 and Hawaii in 2001. A NLCD change product is not available for Hawaii because data are only
available for one year, i.e., 2001. The NLCD products are based primarily on Landsat Thematic Mapper imagery at a
30-meter resolution, and the land cover categories have been aggregated into the 36 IPCC land-use categories for
the conterminous United States and Alaska, and into the six IPCC land-use categories for Hawaii. The land use
patterns are assumed to remain the same after the last year of data in the time series, which is 2001 for Hawaii,
2019 for the conterminous United States and 2016 for Alaska, but the time series will be updated when new data
are released.
For the conterminous United States, the aggregated maps of IPCC land-use categories obtained from the NLCD
products were used in combination with the NRI database to represent land use and land-use change for federal
22 The NRI survey program does not include U.S. Territories with the exception of non-federal lands in Puerto Rico. The FIA
program recently began implementing surveys of forest land in U.S. Territories and those data will be used in the years ahead.
Furthermore, NLCD does not include coverage for all U.S. Territories.
6-18 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
lands, with the exception of forest lands, which are based on FIA. Specifically, NRI survey locations designated as
federal lands were assigned a land use/land-use change category based on the NLCD maps that had been
aggregated into the IPCC categories. This analysis addressed shifts in land ownership across years between federal
or non-federal classes as represented in the NRI survey (i.e., the ownership is classified for each survey location in
the NRI). The sources of these additional data are discussed in subsequent sections of the report.
Managed Land Designation
Lands are designated as managed in the United States based on the definition provided earlier in this section. The
following criteria are used in order to apply the definition in an analysis of managed land:
• All croplands and settlements are designated as managed so only grassland, forest land, wetlands or other
lands may be designated as unmanaged land;23
• All forest lands with active fire protection are considered managed;
• All forest lands designated for timber harvests are considered managed;
• All grasslands are considered managed at a county scale if there are grazing livestock in the county;
• Other areas are considered managed if accessible based on the proximity to roads and other
transportation corridors, and/or infrastructure;
• Protected lands maintained for recreational and conservation purposes are considered managed (i.e.,
managed by public and/or private organizations);
• Lands with active and/or past resource extraction are considered managed; and
• Lands that were previously managed but subsequently classified as unmanaged, remain in the managed
land base for 20 years following the conversion to account for legacy effects of management on C stocks.
The analysis of managed lands, based on the criteria listed above, is conducted using a geographic information
system (Ogle et al. 2018). Lands that are used for crop production or settlements are determined from the NLCD
(Fry et al. 2011; Homer et al. 2007; Homer et al. 2015). Forest lands with active fire management are determined
from maps of federal and state management plans from the National Atlas (U.S. Department of Interior 2005) and
Alaska Interagency Fire Management Council (1998). It is noteworthy that all forest lands in the conterminous
United States have active fire protection, and are therefore designated as managed regardless of accessibility or
other criteria. In addition, forest lands with timber harvests are designated as managed based on county-level
estimates of timber products in the U.S. Forest Service Timber Products Output Reports (U.S. Department of
Agriculture 2012). Timber harvest data lead to additional designation of managed forest land in Alaska. The
designation of grasslands as managed is based on grazing livestock population data at the county scale from the
USDA National Agricultural Statistics Service (U.S. Department of Agriculture 2015). Accessibility is evaluated based
on a 10-km buffer surrounding road and train transportation networks using the ESRI Data and Maps product (ESRI
2008), and a 10-km buffer surrounding settlements using NLCD.
Lands maintained for recreational purposes are determined from analysis of the Protected Areas Database (U.S.
Geological Survey 2012). The Protected Areas Database includes lands protected from conversion of natural
habitats to anthropogenic uses and describes the protection status of these lands. Lands are considered managed
that are protected from development if the regulations allow for extractive or recreational uses or suppression of
natural disturbance (e.g., forest lands with active fire protection). Lands that are protected from development and
not accessible to human intervention, including no suppression of disturbances or extraction of resources, are not
included in the managed land base.
Multiple data sources are used to determine lands with active resource extraction: Alaska Oil and Gas Information
System (Alaska Oil and Gas Conservation Commission 2009), Alaska Resource Data File (U.S. Geological Survey
23 All wetlands are considered managed in this Inventory with the exception of remote areas in Alaska. Distinguishing between
managed and unmanaged wetlands in the conterminous United States and Hawaii is difficult due to limited data availability.
Wetlands are not characterized within the NRI with information regarding water table management. Regardless, a planned
improvement is underway to subdivide managed and unmanaged wetlands.
Land Use, Land-Use Change, and Forestry 6-19
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
2012), Active Mines and Mineral Processing Plants (U.S. Geological Survey 2005), and Coal Production and
Preparation Report (U.S. Energy Information Administration 2011). A buffer of 3,300 and 4,000 meters is
established around petroleum extraction and mine locations, respectively, to account for the footprint of
operation and impacts of activities on the surrounding landscape. The buffer size is based on visual analysis of
disturbance to the landscape for approximately 130 petroleum extraction sites and 223 mines. After applying the
criteria identified above, the resulting managed land area is overlaid on the NLCD to estimate the area of managed
land by land use for both federal and non-federal lands in Alaska. The remaining land represents the unmanaged
land base. The resulting spatial product is also used to identify NRI survey locations that are considered managed
and unmanaged for the conterminous United States and Hawaii.24
Approach for Combining Data Sources
The managed land base in the United States has been classified into the 36 IPCC land use/land-use conversion
categories (Table 6-5) using definitions developed to meet national circumstances, while adhering to IPCC
guidelines (2006).25 In practice, the land was initially classified into land use subcategories within the NRI, FIA, and
NLCD datasets, and then aggregated into the 36 broad land use and land-use change categories identified in IPCC
(2006).
All three datasets provide information on forest land areas in the conterminous United States, but the area data
from FIA serve as the official dataset for forest land. Therefore, another step in the analysis is to address the
inconsistencies in the representation of the forest land among the three databases. NRI and FIA have different
criteria for classifying forest land in addition to different sampling designs, leading to discrepancies in the resulting
estimates of forest land area on non-federal land in the conterminous United States. Similarly, there are
discrepancies between the NLCD and FIA data for defining and classifying forest land on federal lands. Any change
in forest land area in the NRI and NLCD also requires a corresponding change in other land use areas because of
the dependence between the forest land area and the amount of land designated as other land uses, such as the
amount of grassland, cropland, and wetlands (i.e., areas for the individual land uses must sum to the total
managed land area of the country).
FIA is the main database for forest statistics, and consequently, the NRI and NLCD are adjusted to achieve
consistency with FIA estimates of forest land in the conterminous United States. Adjustments are made in the
Forest Land Remaining Forest Land, Land Converted to Forest Land, and Forest Land converted to other uses (i.e.,
Grassland, Cropland, Settlements, Other Lands, and Wetlands). All adjustments are made at the state scale to
address the discrepancies in areas associated with forest land and conversions to and from Forest Land. There are
three steps in this process. The first step involves adjustments to Land Converted to Forest Land (Grassland,
Cropland, Settlements, Other Lands, and Wetlands), followed by a second step in which there are adjustments in
Forest Land converted to another land use (i.e., Grassland, Cropland, Settlements, Other Lands, and Wetlands),
and the last step is to adjust Forest Land Remaining Forest Land.
In the first step, Land Converted to Forest Land in the NRI and NLCD are adjusted to match the state-level
estimates in the FIA data for non-federal and federal Land Converted to Forest Land, respectively. FIA data have
not provided specific land-use categories that are converted to forest land in the past, but rather a sum of all land
converted to forest land.26 The NRI and NLCD provide information on specific land-use conversions, such as
Grassland Converted to Forest Land. Therefore, adjustments at the state level to NRI and NLCD are made
proportional to the amount of specific land-use conversions into forest land for the state, prior to any further
adjustments. For example, if 50 percent of the land-use change to forest land is associated with Grassland
24 The exception is cropland and settlement areas in the NRI, which are classified as managed, regardless of the managed land
base obtained from the spatial analysis described in this section.
25 Definitions are provided in the previous section.
26 The FIA program has started to collect data on the specific land uses that are converted to Forest Land, which will be further
investigated and incorporated into a future Inventory.
6-20 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Converted to Forest Land in a state according to NRI or NLCD, then half of the discrepancy with FIA data in the area
of Land Converted to Forest Land is addressed by increasing or decreasing the area in Grassland Converted to
Forest Land. Moreover, any increase or decrease in Grassland Converted to Forest Land in NRI or NLCD is
addressed by a corresponding change in the area of Grassland Remaining Grassland, so that the total amount of
managed area is not changed within an individual state.
In the second step, state-level areas are adjusted in the NRI and NLCD to address discrepancies with FIA data for
forest land converted to other uses. Similar to Land Converted to Forest Land, FIA have not provided information
on the specific land-use changes in the past,27 so areas associated with forest land conversion to other land uses in
NRI and NLCD are adjusted proportional to the amount of area in each conversion class in these datasets.
In the final step, the area of Forest Land Remaining Forest Land in each state according to the NRI and NLCD is
adjusted to match the FIA estimates for non-federal and federal land, respectively. It is assumed that the majority
of the discrepancy in Forest Land Remaining Forest Land is associated with less-precise estimates of Grassland
Remaining Grassland and Wetlands Remaining Wetlands in the NRI and NLCD. This step also assumes that there
are no changes in the land-use conversion categories. Therefore, corresponding adjustments are made in the area
estimates of Grassland Remaining Grassland and Wetlands Remaining Wetlands from the NRI and NLCD. This
adjustment balances the change in Forest Land Remaining Forest Land area, which ensures no change in the
overall amount of managed land within an individual state. The adjustments are based on the proportion of land
within each of these land-use categories at the state level according to NRI and NLCD (i.e., a higher proportion of
Grassland led to a larger adjustment in Grassland area).
The modified NRI data are then aggregated to provide the land use and land-use change data for non-federal lands
in the conterminous United States, and the modified NLCD data are aggregated to provide the land use and land-
use change data for federal lands. Data for all land uses in Hawaii are based on NRI for non-federal lands and on
NLCD for federal lands. Land use data in Alaska are based on the NLCD data after adjusting this dataset to be
consistent with forest land areas in the FIA (Table 6-6). The result is land use and land-use change data for the
conterminous United States, Hawaii, and Alaska.
A summary of the details on the approach used to combine data sources for each land use are described below.
• Forest Land: Land representation for both non-federal and federal forest lands in the conterminous
United States and Alaska are based on the FIA. FIA is used as the basis for both forest land area data as
well as to estimate C stocks and fluxes on forest land in the conterminous United States and Alaska. FIA
does have survey plots in Alaska that are used to determine the C stock changes, and the associated area
data for this region are harmonized with NLCD using the methods described above. NRI is used in the
current report to provide forest land areas on non-federal lands in Hawaii, and NLCD is used for federal
lands. FIA data is being collected in Hawaii and U.S. Territories, however there is insufficient data to make
population estimates for this Inventory.
• Cropland: Cropland is classified using the NRI, which covers all non-federal lands within 49 states
(excluding Alaska), including state and local government-owned land as well as tribal lands. NRI is used as
the basis for both cropland area data as well as to estimate soil C stocks and fluxes on cropland. NLCD is
used to determine cropland area and soil C stock changes on federal lands in the conterminous United
States and Hawaii. NLCD is also used to determine croplands in Alaska, but C stock changes are not
estimated for this region in the current Inventory.
• Grassland: Grassland on non-federal lands is classified using the NRI within 49 states (excluding Alaska),
including state and local government-owned land as well as tribal lands. NRI is used as the basis for both
grassland area data as well as to estimate soil C stocks and non-CC>2 greenhouse emissions on grassland.
Grassland area and soil C stock changes are determined using the classification provided in the NLCD for
27 The FIA program has started to collect data on specific land uses following conversion from Forest Land, which will be further
investigated and incorporated into a future Inventory.
Land Use, Land-Use Change, and Forestry 6-21
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
federal land within the conterminous United States. NLCD is also used to estimate the areas of federal and
non-federal grasslands in Alaska, and the federal grasslands in Hawaii, but the current Inventory does not
include C stock changes in these areas.
• Wetlands: NRI captures wetlands on non-federal lands within 49 states (excluding Alaska), while the land
representation data for federal wetlands and wetlands in Alaska are based on the NLCD.28
• Settlements: NRI captures non-federal settlement area in 49 states (excluding Alaska). If areas of forest
land or grassland under 10 acres (4.05 ha) are contained within settlements or urban areas, they are
classified as settlements (urban) in the NRI database. If these parcels exceed the 10-acre (4.05 ha)
threshold and are grassland, they are classified as grassland by NRI. Regardless of size, a forested area is
classified as non-forest by FIA if it is located within an urban area. Land representation for settlements on
federal lands and Alaska is based on the NLCD.
• Other Land: Any land that is not classified into one of the previous five land-use categories is categorized
as other land using the NRI for non-federal areas in the conterminous United States and Hawaii and using
the NLCD for the federal lands in all regions of the United States and for non-federal lands in Alaska.
Some lands can be classified into one or more categories due to multiple uses that meet the criteria of more than
one definition. However, a ranking has been developed for assignment priority in these cases. The ranking process
is from highest to lowest priority based on the following order:
Settlements > Cropland > Forest Land > Grassland > Wetlands > Other Land
Settlements are given the highest assignment priority because they are extremely heterogeneous with a mosaic of
patches that include buildings, infrastructure, and travel corridors, but also open grass areas, forest patches,
riparian areas, and gardens. The latter examples could be classified as grassland, forest land, wetlands, and
cropland, respectively, but when located in close proximity to settlement areas, they tend to be managed in a
unique manner compared to non-settlement areas. Consequently, these areas are assigned to the Settlements
land-use category. Cropland is given the second assignment priority, because cropping practices tend to dominate
management activities on areas used to produce food, forage, or fiber. The consequence of this ranking is that
crops in rotation with pasture are classified as cropland, and land with woody plant cover that is used to produce
crops (e.g., orchards) is classified as cropland, even though these areas may also meet the definitions of grassland
or forest land, respectively. Similarly, wetlands are considered croplands if they are used for crop production, such
as rice or cranberries. Forest land occurs next in the priority assignment because traditional forestry practices tend
to be the focus of the management activity in areas with woody plant cover that are not croplands (e.g., orchards)
or settlements (e.g., housing subdivisions with significant tree cover). Grassland occurs next in the ranking, while
wetlands and then other land complete the list.
The assignment priority does not reflect the level of importance for reporting greenhouse gas emissions and
removals on managed land, but is intended to classify all areas into a discrete land-use category. Currently, the
IPCC does not make provisions in the guidelines for assigning land to multiple uses. For example, a wetland is
classified as forest land if the area has sufficient tree cover to meet the stocking and stand size requirements.
Similarly, wetlands are classified as cropland if they are used for crop production, such as rice, or as grassland if
they are composed principally of grasses, grass-like plants (i.e., sedges and rushes), forbs, or shrubs suitable for
grazing and browsing. Regardless of the classification, emissions and removals from these areas should be included
in the Inventory if the land is considered managed, and therefore impacted by anthropogenic activity in
accordance with the guidance provided by the IPCC (2006).
28 This analysis does not distinguish between managed and unmanaged wetlands except for remote areas in Alaska, but there
is a planned improvement to subdivide managed and unmanaged wetlands for the entire land base.
6-22 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
QA/QC and Verification
The land base obtained from the NRI, FIA, and NLCD was compared to the Topological^ Integrated Geographic
Encoding and Referencing (TIGER) survey (U.S. Census Bureau 2010). The United States Census Bureau gathers
data on the population and economy and has a database of land areas for the country. The area estimates of land-
use categories, based on NRI, FIA, and NLCD, are obtained from remote sensing data instead of the land survey
approach used by the United States Census Survey. The Census does not provide a time series of land-use change
data or land management information, which is needed for estimating greenhouse gas emissions from land use
and land-use change. Regardless, the Census does provide sufficient information to provide a quality assurance
check on the Inventory data. There are 46 million more hectares of land in the United States according to the
Census, compared to the total area estimate of 936 million hectares obtained from the combined NRI, FIA, and
NLCD data. Much of this difference is associated with open waters in coastal regions and the Great Lakes, which is
included in the TIGER Survey of the Census, but not included in the land representation using the N Rl, FIA and
NLCD. There is only a 0.4 percent difference when open water in coastal regions is removed from the TIGER data.
General QC procedures for data gathering and data documentation also were applied consistent with the QA/QC
and Verification Procedures described in Annex 8.
Recalculations Discussion
Major updates were made in this Inventory associated with the release of new land use and land cover data. The
land representation data were recalculated from the previous Inventory with the following datasets: a) updated
FIA data from 1990 to 2021 for the conterminous United States and Alaska, b) updated NRI data from 1990 to 2017
for the conterminous United States and Hawaii, and c) updated NLCD data for the conterminous United States
from 2001 through 2019 and Alaska from 2001 through 2016. With these recalculations, managed forest land
essentially remained the same as the previous Inventory across the time series from 1990 to 2021 according to the
new FIA data. According to the new NRI and NLCD data, as well as harmonization of these data with the new FIA
data (See section "Approach for Combining Data Sources"), grassland and settlements remained essentially
unchanged from the previous Inventory and cropland, wetlands, and other land decreased by an average of 0.1
percent, 0.9 percent, and 5.8 percent, respectively.
Planned Improvements
Research is underway to harmonize NRI and FIA sampling frames to improve consistency and facilitate estimation
using multi-frame sampling. This includes development of a common land use classification schema between the
two land inventories that can be used in the harmonization process. These steps will allow for population
estimation exclusive of auxiliary information (e.g., NLCD). The multi-frame sample will also serve as reference data
for the development of spatially explicit and spatially continuous map products for each year in the Inventory time
series. Another key planned improvement for the Inventory is to fully incorporate area data by land use type for
U.S. Territories. Fortunately, most of the managed land in the United States is included in the current land use
data, but a complete reporting of all lands in the United States is a key goal for the near future. Preliminary land
use area data for U.S. Territories by land-use category are provided in Box 6-2.
Box 6-2: Preliminary Estimates of Land Use in U.S. Territories
Several programs have developed land cover maps for U.S. Territories using remote sensing imagery, including
the Gap Analysis Program, Caribbean Land Cover project, National Land Cover Dataset (NLCD), USFS Pacific
Islands Imagery Project, and the National Oceanic and Atmospheric Administration (NOAA) Coastal Change
Analysis Program (C-CAP). Land-cover data can be used to inform a land use classification if there is a time series
to evaluate the dominate practices. For example, land that is principally used for timber production with tree
cover over most of the time series is classified as forest land even if there are a few years of grass dominance
following timber harvest. These products were reviewed and evaluated for use in the national Inventory as a
step towards implementing a planned improvement to include U.S. Territories in the land representation for the
Land Use, Land-Use Change, and Forestry 6-23
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Inventory. Recommendations are to use the NOAA C-CAP Regional Land Cover Database for the smaller island
Territories (U.S. Virgin Islands, Guam, Northern Marianas Islands, and American Samoa) because this program is
ongoing and therefore will be continually updated. The C-CAP product does not cover the entire territory of
Puerto Rico, so the NLCD was used for this area. The final selection of land-cover products for these territories is
still under discussion. Results are presented below (in hectares). The total land area of all U.S. Territories is 1.05
million hectares, representing 0.1 percent of the total land base for the United States (see Table 6-7).
Table 6-7: Total Land Area (Hectares) by Land Use Category for U.S. Territories
Puerto Rico
U.S. Virgin
Islands
Guam
Northern
Marianas
Islands
American
Samoa
Total
Cropland
19,712
138
236
289
389
20,764
Forest Land
404,004
13,107
24,650
25,761
15,440
482,962
Grasslands
299,714
12,148
15,449
13,636
1,830
342,777
Other Land
5,502
1,006
1,141
5,186
298
13,133
Settlements
130,330
7,650
11,146
3,637
1,734
154,496
Wetlands
24,525
4,748
1,633
260
87
31,252
Total
883,788
38,796
54,255
48,769
19,777
1,045,385
Note: Totals may not sum due to independent rounding.
Methods in the 2013 Supplement to the 2006 Guidelines for National Greenhouse Gas Inventories: Wetlands (IPCC
2014) have been applied to estimate emissions and removals from coastal wetlands. Specifically, greenhouse gas
emissions from coastal wetlands have been developed for the Inventory using the NOAA C-CAP land cover product.
The NOAA C-CAP product is not used directly in the land representation analysis, however, so a planned
improvement for future Inventories is to reconcile the coastal wetlands data from the C-CAP product with the
wetlands area data provided in the NRI, FIA and NLCD. Estimates from flooded lands are also included in this
Inventory, but data are not directly used in the land representation analysis at this time; this is a planned
improvement to includes for future inventories. In addition, the current Inventory does not include a classification
of managed and unmanaged wetlands, except for remote areas in Alaska. Consequently, there is a planned
improvement to classify managed and unmanaged wetlands for the conterminous United States and Hawaii, and
more detailed wetlands datasets will be evaluated and integrated into the analysis to meet this objective.
6.2 Forest Land Remaining Forest Land
(CRF Category 4A1)
Changes in Forest Carbon Stocks (CRF Category 4A1)
Delineation of Carbon Pools
For estimating carbon (C) stocks or stock change (flux), C in forest ecosystems can be divided into the following five
storage pools (IPCC 2006):
• Aboveground biomass, which includes all living biomass above the soil including stem, stump, branches,
bark, seeds, and foliage. This category includes live understory.
• Belowground biomass, which includes all living biomass of coarse living roots greater than 2 millimeters
(mm) diameter.
6-24 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
• Dead wood, which includes all non-living woody biomass either standing, lying on the ground (but not
including litter), or in the soil.
3
4
• Litter, which includes all duff, humus, and fine woody debris above the mineral soil as well as woody
fragments with diameters of up to 7.5 cm.
5
6
• Soil organic C (SOC), including all organic material in soil to a depth of 1 meter but excluding the coarse
roots of the belowground pools.
7 In addition, there are two harvested wood pools included when estimating C flux:
8 • Harvested wood products (HWP) in use.
9 • HWP in solid waste disposal sites (SWDS).
10 Forest Carbon Cycle
11 Carbon is continuously cycled among the previously defined C storage pools and the atmosphere as a result of
12 biogeochemical processes in forests (e.g., photosynthesis, respiration, decomposition, and disturbances such as
13 fires or pest outbreaks) and anthropogenic activities (e.g., harvesting, thinning, and replanting). As trees
14 photosynthesize and grow, C is removed from the atmosphere and stored in living tree biomass. As trees die and
15 otherwise deposit litter and debris on the forest floor, C is released to the atmosphere and is also transferred to
16 the litter, dead wood, and soil pools by organisms that facilitate decomposition.
17 The net change in forest C is not equivalent to the net flux between forests and the atmosphere because timber
18 harvests do not cause an immediate flux of all harvested biomass C to the atmosphere. Instead, harvesting
19 transfers a portion of the C stored in wood to a "product pool." Once in a product pool, the C is emitted over time
20 as CO2 in the case of decomposition and as CO2, CH4, N2O, CO, and NOxwhen the wood product combusts. The rate
21 of emission varies considerably among different product pools. For example, if timber is harvested to produce
22 energy, combustion releases C immediately, and these emissions are reported for information purposes in the
23 Energy sector while the harvest (i.e., the associated reduction in forest C stocks) and subsequent combustion are
24 implicitly estimated in the Land Use, Land-Use Change, and Forestry (LULUCF) sector (i.e., the portion of harvested
25 timber combusted to produce energy does not enter the HWP pools). Conversely, if timber is harvested and used
26 as lumber in a house, it may be many decades or even centuries before the lumber decays and C is released to the
27 atmosphere. If wood products are disposed of in SWDS, the C contained in the wood may be released many years
28 or decades later or may be stored almost permanently in the SWDS. These latter fluxes, with the exception of CH4
29 from wood in SWDS, which is included in the Waste sector, are also estimated in the LULUCF sector.
30 Net Change in Carbon Stocks within Forest Land of the United States
31 This section describes the general method for quantifying the net changes in C stocks in the five C storage pools
32 and two harvested wood pools (a more detailed description of the methods and data is provided in Annex 3.13).
33 The underlying methodology for determining C stock and stock change relies on data from the national forest
34 inventory (NFI) conducted by the Forest Inventory and Analysis (FIA) program within the USDA Forest Service. The
35 annual NFI is implemented across all U.S. forest lands within the conterminous 48 states and Alaska and
36 inventories have been initiated in Hawaii and some of the U.S. Territories. The methods for estimation and
37 monitoring are continuously improved and these improvements are reflected in the C estimates (Domke et al.
38 2022). First, the total C stocks are estimated for each C storage pool at the individual NFI plot, next the annual net
39 changes in C stocks for each pool at the population are estimated, and then the changes in stocks are summed for
40 all pools to estimate total net flux at the population level (e.g., U.S. state). Changes in C stocks from disturbances,
41 such natural disturbances (e.g., wildfires, insects/disease, wind) or harvesting, are included in the net changes (See
42 Box 6-3 for more information). For instance, an inventory conducted after a fire implicitly includes only the C
43 stocks remaining on the NFI plot. The IPCC (2006) recommends estimating changes in C stocks from forest lands
44 according to several land-use types and conversions, specifically Forest Land Remaining Forest Land and Land
Land Use, Land-Use Change, and Forestry 6-25
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Converted to Forest Land, with the former being lands that have been forest lands for 20 years or longer and the
latter being lands (i.e., croplands, grassland, wetlands, settlements and other lands) that have been converted to
forest lands for less than 20 years. The methods and data used to delineate forest C stock changes by these two
categories continue to improve and in order to facilitate this delineation, a combination of modeling approaches
for C estimation were used in this Inventory.
Forest Area in the United States
Approximately 32 percent of the U.S. land area is estimated to be forested based on the U.S. definition of forest
land as provided in Section 6.1 Representation of the U.S. Land Base. All annual NFI plots included in the public FIA
database as of August 2022 (which includes data collected through 2021- note that the ongoing COVID 19
pandemic has resulted in delays in data collection in many states) were used in this Inventory. The NFIs from the
conterminous United States (USDA Forest Service 2022a, 2022b) and Alaska comprise an estimated 280 million
hectares of forest land that are considered managed and are included in the current Inventory. Some differences
also exist in forest land area estimates from the latest update to the Resources Planning Act (RPA) Assessment
(Oswalt et al. 2019) and the forest land area estimates included in this report, which are based on the annual NFI
data through 2021 for all states (USDA Forest Service 2022b; Nelson et al. 2020). Sufficient annual NFI data are not
yet available for Hawaii and the U.S. Territories to include them in this section of the Inventory but estimates of
these areas are included in Oswalt et al. (2019). While Hawaii and U.S. Territories have relatively small areas of
forest land and thus may not substantially influence the overall C budget for forest land, these regions will be
added to the forest C estimates as sufficient data become available. Since Hawaii was not included in this section
of the current Inventory, this results in small differences in the area estimates reported in this section and those
reported in Section 6.1 Representation of the U.S. Land Base. Also, it is not possible to separate Forest Land
Remaining Forest Land from Land Converted to Forest Land in Wyoming because of the split annual cycle method
used for population estimation, this prevents harmonization of forest land in Wyoming with the NRI/NLCD method
used in section 6.1 Representation of the U.S. Land Base (CRF Category 4.1).29 Agroforestry systems that meet the
definition of forest land are also not currently included in the current Inventory since they are not explicitly
inventoried (i.e., classified as an agroforestry system) by either the FIA program or the Natural Resources Inventory
(NRI)30 of the USDA Natural Resources Conservation Service (Perry et al. 2005).
An estimated 67 percent (208 million hectares) of U.S. forests in Alaska, Hawaii and the conterminous United
States are classified as timberland, meaning they meet minimum levels of productivity and have not been removed
from production. Approximately ten percent of Alaska forest land and 73 percent of forest land in the
conterminous United States are classified as timberland. Of the remaining non-timberland, nearly 33 million
hectares are reserved forest lands (withdrawn by law from management for production of wood products) and 102
million hectares are lower productivity forest lands (Oswalt et al. 2019). Historically, the timberlands in the
conterminous 48 states have been more frequently or intensively surveyed than the forest lands removed from
production because they do not meet the minimum level of productivity.
Since the late 1980s, gross forest land area in Alaska, Hawaii, and the conterminous United States has increased by
about 13 million hectares (Oswalt et al. 2019). The southern region of the United States contains the most forest
land (Figure 6-4). A substantial portion of this accrued forest land is from the conversion of abandoned croplands
to forest (e.g., Woodall et al. 2015b). Estimated forest land area in the conterminous United States and Alaska
represented in this Inventory is stable, but there are substantial conversions as described in Section 6.1
Representation of the U.S. Land Base and each of the land conversion sections for each land-use category (e.g.,
Land Converted to Cropland, Land Converted to Grassland). The major influences on the net C flux from forest land
29 See Annex 3.13, Table A-213 for annual differences between the forest area reported in Section 6.1 Representation of the
U.S. Land Base and Section 6.2 Forest Land Remaining Forest Land.
30 The Natural Resources Inventory of the USDA Natural Resources Conservation Service is described in Section 6.1
Representation of the U.S. Land Base.
6-26 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
across the 1990 to 2021 time series are management activities, natural disturbance, particularly wildfire, and the
ongoing impacts of current and previous land-use conversions. These activities affect the net flux of C by altering
the amount of C stored in forest ecosystems and also the area converted to forest land. For example, intensified
management of forests that leads to an increased rate of growth of aboveground biomass (and possible changes to
the other C storage pools) may increase the eventual biomass density of the forest, thereby increasing the uptake
and storage of C in the aboveground biomass pool.31 Though harvesting forests removes much of the C in
aboveground biomass (and possibly changes C density in other pools), on average, the estimated volume of annual
net growth in aboveground tree biomass in the conterminous United States is essentially twice the volume of
annual removals on timberlands (Oswalt et al. 2019). The net effects of forest management and changes in Forest
Land Remaining Forest Land are captured in the estimates of C stocks and fluxes presented in this section.
Figure 6-4: Changes in Forest Area by Region for Forest Land Remaining Forest Land in the
conterminous United States and Alaska (1990-2021)
100-i
Si 90 H
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Forest Carbon Stocks and Stock Change
In the Forest Land Remaining Forest Land category, forest management practices, the regeneration of forest areas
cleared more than 20 years prior to the reporting year, and timber harvesting have resulted in net removal (i.e.,
net sequestration or accumulation) of C each year from 1990 through 2021. The rate of forest clearing in the 17th
century following European settlement had slowed by the late 19th century. Through the later part of the 20th
century, many areas of previously forested land in the United States were allowed to revert to forests or were
actively reforested. The impacts of these land-use changes still influence C fluxes from these forest lands. More
recently, the 1970s and 1980s saw a resurgence of federally sponsored forest management programs (e.g., the
Forestry Incentive Program) and soil conservation programs (e.g., the Conservation Reserve Program), which have
focused on tree planting, improving timber management activities, combating soil erosion, and converting
marginal cropland to forests. In addition to forest regeneration and management, forest harvests and natural
disturbance have also affected net C fluxes. Because most of the timber harvested from U.S. forest land is used in
wood products, and many discarded wood products are disposed of in SWDS rather than by incineration,
significant quantities of C in harvested wood are transferred to these long-term storage pools rather than being
released rapidly to the atmosphere (Skog 2008). By maintaining current harvesting practices and regeneration
activities on forested lands, along with continued input of harvested products into the HWP pool, C stocks in the
Forest Land Remaining Forest Land category are likely to continue to increase in the near term, though possibly at
a lower rate. Changes in C stocks in the forest ecosystem and harvested wood pools associated with Forest Land
Remaining Forest Land were estimated to result in net removal of 695.4 MMT CO2 Eq. (189.6 MMT C) in 2021
(Table 6-8, Table 6-9, Table A-210, Table A-211 and state-level estimates in Table A-214). The estimated net uptake
of C in the Forest Ecosystem was 592.5 MMT CO2 Eq. (161.6 MMT C) in 2021 (Table 6-8 and Table 6-9). The
majority of this uptake in 2021, 409.1 MMT CO2 Eq. (111.6 MMT C), was from aboveground biomass. Overall,
estimates of average C density in forest ecosystems (including all pools) increased consistently over the time series
with an average of approximately 192 MT C ha 1 from 1990 to 2021. This was calculated by dividing the Forest Land
area estimates by Forest Ecosystem C Stock estimates for every year (see Table 6-10 and Table A-212) and then
calculating the mean across the entire time series, i.e., 1990 through 2021. The increasing forest ecosystem C
density, when combined with relatively stable forest area, results in net C accumulation over time. Aboveground
live biomass is responsible for the majority of net C uptake among all forest ecosystem pools (Figure 6-5). These
increases may be influenced in some regions by reductions in C density or forest land area due to natural
disturbances (e.g., wildfire, weather, insects/disease), particularly in Alaska. The inclusion of all managed forest
land in Alaska has increased the interannual variability in carbon stock change estimates over the time series, and
much of this variability can be attributed to severe fire years (e.g., 2019). The distribution of carbon in forest
ecosystems in Alaska is substantially different from forests in the conterminous United States. In Alaska, more than
11 percent of forest ecosystem C is stored in the litter carbon pool whereas in the conterminous United States,
only 7 percent of the total ecosystem C stocks are in the litter pool. Much of the litter material in forest
ecosystems is combusted during fire (IPCC 2006) leading to substantial C losses in this pool during severe fire years
(Figure 6-5, Table A-217).
The estimated net uptake of C in HWP was 102.8 MMT C02 Eq. (28.0 MMT C) in 2021 (Table 6-8, Table 6-9, Table
A-210, and Table A-211). The majority of this uptake, 65.1 MMT CO2 Eq. (17.7 MMT C), was from wood and paper
in SWDS. Products in use accounted for an estimated 37.8 MMT CO2 Eq. (10.3 MMT C) in 2021.
Table 6-8: Net CO2 Flux from Forest Ecosystem Pools in Forest Land Remaining Forest Land
and Harvested Wood Pools (MMT CO2 Eq.)
Carbon Pool
1990
2005
2017
2018
2019
2020
2021
Forest Ecosystem
(697.7)
(608.2)
(610.4)
(610.5)
(559.8)
(610.8)
(592.5)
Aboveground Biomass
(499.1)
(443.8)
(425.9)
(428.0)
(410.8)
(419.0)
(409.1)
Belowground Biomass
(101.8)
(89.8)
(84.5)
(85.1)
(81.6)
(83.1)
(81.1)
Dead Wood
(100.8)
(97.9)
(100.0)
(102.7)
(98.2)
(102.3)
(101.1)
Litter
0.9
22.5
(2.0)
1.6
30.4
(1.9)
1.9
Soil (Mineral)
3.2
0.5
(0.1)
0.6
0.7
(5.4)
(4.0)
6-28 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Soil (Organic)
(0.8)
(0.4)
1.4
2.3
(1.1)
0.1
0.1
Drained Organic Soil3
0.8
0.8
0.8
0.8
0.8
0.8
0.8
Harvested Wood
(123.8)
(106.0)
(100.3)
(94.0)
(89.6)
(96.6)
(102.8)
Products in Use
(54.8)
(42.6)
(34.9)
(28.9)
(25.1)
(32.0)
(37.8)
SWDS
(69.0)
(63.4)
(65.3)
(65.1)
(64.5)
(64.6)
(65.1)
Total Net Flux
(821.4)
(714.2)
(710.7)
(704.4)
(649.3)
(707.4)
(695.4)
a These estimates include C stock changes from drained organic soils from both Forest Land Remaining
Forest Land and Land Converted to Forest Land. See the section below on C02, CH4, and N20 Emissions
from Drained Organic Soils for the methodology used to estimate the C02 emissions from drained organic
soils. Also, Table 6-20 and 6-21 for non-C02 emissions from drainage of organic soils from both Forest Land
Remaining Forest Land and Land Converted to Forest Land.
Notes: Forest ecosystem C stock changes do not include forest stocks in U.S. Territories because managed
forest land for U.S. Territories is not currently included in Section 6.1 Representation of the U.S. Land Base.
The forest ecosystem C stock changes do not include Hawaii because there is not sufficient NFI data to
support inclusion at this time. However, managed forest land area for Hawaii is included in Section 6.1
Representation of the U.S. Land Base, so there are small differences in the forest land area estimates in this
Section and Section 6.1. Also, it is not possible to separate Forest Land Remaining Forest Land from Land
Converted to Forest Land in Wyoming because of the split annual cycle method used for population
estimation, this prevents harmonization of forest land in Wyoming with the NRI/NLCD method used in
section 6.1 Representation of the U.S. Land Base (CRF Category 4.1). See Annex 3.13, Table A-213 for
annual differences between the forest area reported in Section 6.1 Representation of the U.S. Land Base
and Section 6.2 Forest Land Remaining Forest Land. The forest ecosystem C stock changes do not include
trees on non-forest land (e.g., agroforestry systems and settlement areas—see Section 6.10Settlements
Remaining Settlements for estimates of C stock change from settlement trees). Forest ecosystem C stocks
on managed forest land in Alaska were compiled using the gain-loss method as described in Annex 3.13.
Parentheses indicate net C uptake (i.e., a net removal of C from the atmosphere). Total net flux is an
estimate of the actual net flux between the total forest C pool and the atmosphere. Harvested wood
estimates are based on results from annual surveys and models. Totals may not sum due to independent
rounding.
1 Table 6-9: Net C Flux from Forest Ecosystem Pools in Forest Land Remaining Forest Land
2 and Harvested Wood Pools (MMT C)
Carbon Pool
1990
2005
2017
2018
2019
2020
2021
Forest Ecosystem
(190.3)
(165.9)
(166.5)
(166.5)
(152.7)
(166.6)
(161.6)
Aboveground Biomass
(136.1)
(121.0)
(116.1)
(116.7)
(112.0)
(114.3)
(111.6)
Belowground Biomass
(27.8)
(24.5)
(23.0)
(23.2)
(22.3)
(22.7)
(22.1)
Dead Wood
(27.5)
(26.7)
(27.3)
(28.0)
(26.8)
(27.9)
(27.6)
Litter
0.2
6.1
(0.6)
0.4
8.3
(0.5)
0.5
Soil (Mineral)
0.9
0.1
(0.0)
0.2
0.2
(1.5)
(1.1)
Soil (Organic)
(0.2)
(0.1)
0.4
0.6
(0.3)
0.0
0.0
Drained Organic Soil3
0.21
0.2
0.2
0.2
0.2
0.2
0.2
Harvested Wood
(33.8)
(28.9)
(27.3)
(25.6)
(24.4)
(26.3)
(28.0)
Products in Use
(14.9)
(11.6)
(9.5)
(7.9)
(6.8)
(8.7)
(10.3)
SWDS
(18.8)
(17.3)
(17.8)
(17.8)
(17.6)
(17.6)
(17.7)
Total Net Flux
(224.0)
(194.8)
(193.8)
(192.1)
(177.1)
(192.9)
(189.6)
a These estimates include carbon stock changes from drained organic soils from both Forest Land Remaining
Forest Land and Land Converted to Forest Land. See the section below on C02, CH4, and N20 Emissions from
Drained Organic Soils for the methodology used to estimate the C flux from drained organic soils. Also, see
Table 6-20 and 6-21 for greenhouse gas emissions from non-C02 gases changes from drainage of organic soils
from Forest Land Remaining Forest Land and Land Converted to Forest Land.
Notes: Forest ecosystem C stock changes do not include forest stocks in U.S. Territories because managed
forest land for U.S. Territories is not currently included in Section 6.1 Representation of the U.S. Land Base.
The forest ecosystem C stock changes do not include Hawaii because there is not sufficient NFI data to support
inclusion at this time. However, managed forest land area for Hawaii is included in 6.1 Representation of the
U.S. Land Base so there are small differences in the forest land area estimates in this Section and Section 6.1.
Land Use, Land-Use Change, and Forestry 6-29
-------
Also, it is not possible to separate Forest Land Remaining Forest Land from Land Converted to Forest Land in
Wyoming because of the split annual cycle method used for population estimation, this prevents
harmonization of forest land in Wyoming with the NRI/NLCD method used in section 6.1 Representation of the
U.S. Land Base (CRF Category 4.1). See Annex 3.13, Table A-213 for annual differences between the forest area
reported in Section 6.1 Representation of the U.S. Land Base and Section 6.2 Forest Land Remaining Forest
Land. The forest ecosystem C stock changes do not include trees on non-forest land (e.g., agroforestry systems
and settlement areas—see Section 6.10Settlements Remaining Settlements for estimates of C stock change
from settlement trees). Forest ecosystem C stocks on managed forest land in Alaska were compiled using the
gain-loss method as described in Annex 3.13. Parentheses indicate net C uptake (i.e., a net removal of C from
the atmosphere). Total net flux is an estimate of the actual net flux between the total forest C pool and the
atmosphere. Harvested wood estimates are based on results from annual surveys and models. Totals may not
sum due to independent rounding.
1 Stock estimates for forest ecosystem and harvested wood C storage pools are presented in Table 6-10. Together,
2 the estimated aboveground biomass and soil C pools account for a large proportion of total forest ecosystem C
3 stocks. Forest land area estimates are also provided in Table 6-10, but these do not precisely match those in
4 Section 6.1 Representation of the U.S. Land Base for Forest Land Remaining Forest Land. This is because the forest
5 land area estimates in Table 6-10 only include managed forest land in the conterminous U.S. and Alaska while the
6 area estimates in Section 6.1 also include all managed forest land in Hawaii. Differences also exist because forest
7 land area estimates are based on the latest NFI data through 2021, and woodland areas previously included as
8 forest land have been separated and included in the Grassland categories in this Inventory.32
9 Table 6-10: Forest Area (1,000 ha) and C Stocks in Forest Land Remaining Forest Land and
10 Harvested Wood Pools (MMT C)
1990
2005
2018
2019
2020
2021
2022
Forest Area (1,000 ha)
282,150
281,096
280,467
280,299
280,120
279,962
279,800
Carbon Pools (MMT C)
Forest Ecosystem
51,354
54,098
56,303
56,470
56,623
56,790
56,951
Aboveground Biomass
11,899
13,849
15,406
15,523
15,635
15,749
15,861
Belowground Biomass
2,344
2,740
3,052
3,076
3,098
3,121
3,143
Dead Wood
1,948
2,359
2,717
2,745
2,771
2,799
2,827
Litter
3,929
3,922
3,896
3,896
3,888
3,888
3,888
Soil (Mineral)
25,920
25,911
25,914
25,914
25,914
25,915
25,916
Soil (Organic)
5,315
5,318
5,318
5,317
5,317
5,317
5,317
Harvested Wood
1,895
2,353
2,645
2,671
2,695
2,721
2,749
Products in Use
1,249
1,447
1,516
1,523
1,530
1,539
1,549
SWDS
646
906
1,129
1,147
1,165
1,182
1,200
Total C Stock
53,249
56,451
58,948
59,141
59,318
59,511
59,701
Notes: Forest area and C stock estimates include all Forest Land Remaining Forest Land in the conterminous 48 states and
Alaska. Forest ecosystem C stocks do not include forest stocks in U.S. Territories because managed forest land for U.S.
Territories is not currently included in Section 6.1 Representation of the U.S. Land Base. The forest ecosystem C stocks do
not include Hawaii because there is not sufficient NFI data to support inclusion at this time. However, managed forest land
area for Hawaii is included in Section 6.1 Representation of the U.S. Land Base so there are small differences in the forest
land area estimates in this Section and Section 6.1. Also, it is not possible to separate Forest Land Remaining Forest Land
from Land Converted to Forest Land in Wyoming because of the split annual cycle method used for population estimation,
this prevents harmonization of forest land in Wyoming with the NRI/NLCD method used in section 6.1 Representation of
the U.S. Land Base (CRF Category 4.1). See Annex 3.13, Table A-213 for annual differences between the forest area
reported in Section 6.1 Representation of the U.S. Land Base and Section 6.2 Forest Land Remaining Forest Land. The
forest ecosystem C stocks do not include trees on non-forest land (e.g., agroforestry systems and settlement areas—see
Section 6.10 Settlements Remaining Settlements for estimates of C stock change from settlement trees). Forest ecosystem
C stocks on managed forest land in Alaska were compiled using the gain-loss method as described in Annex 3.13.
32 See Annex 3.13, Table A-213 for annual differences between the forest area reported in Section 6.1 Representation of the
U.S. Land Base and Section 6.2 Forest Land Remaining Forest Land.
6-30 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Harvested wood product stocks include exports, even if the logs are processed in other countries, and exclude imports.
Harvested wood estimates are based on results from annual surveys and models. Totals may not sum due to independent
rounding. Population estimates compiled using FIA data are assumed to represent stocks as of January 1 of the inventory
year. Flux is the net annual change in stock. Thus, an estimate of flux for 2021 requires estimates of C stocks for 2021 and
2022.
Figure 6-5: Estimated Net Annual Changes in C Stocks for All C Pools in Forest Land
Remaining Forest Land in the Conterminous United States and Alaska (1990-2021)
20-1
| 8
£ >•
So
& i—
§ S
•e ^
.2 jjj?
§5
S> "
¦§1
iS V)
d
£ s
0~«
-20-
-40-
-60-
-80-1
-100-
-120-
-140-
-160-
-180-
-200-
-220-
-240-J
| i i i i | i i i l | i
1990 1995 2000
All forest ecosystem pools
Aboveground biomass
Belowground biomass
Dead wood
Litter
Soil (mineral)
i | i i
2005
Year
1 I 1
2010
1 I 1
2015
' I 1
2020
¦ Soil (organic)
Drained Organic Soil
Harvested Wood Products (HWP)
Products in use
Solid waste disposal sites
Total net change
(forest ecosystem + HWP)
Box 6-3: CO2 Emissions from Forest Fires
As stated previously, the forest inventory approach implicitly includes ail C losses due to disturbances such as
forest fires, because only C remaining in the forest is estimated. Net C stock change is estimated by subtracting
consecutive C stock estimates. A forest fire disturbance removes C from the forest. The inventory data from the
NFI on which net C stock estimates are based already reflect this C loss. Therefore, estimates of net annual
changes in C stocks for U.S. forest land already includes CO2 emissions from forest fires occurring in the
conterminous states as well as the portion of managed forest lands in Alaska. Because it is of interest to
quantify the magnitude of CO2 emissions from fire disturbance, these separate estimates are highlighted here.
Note that these CO2 estimates are based on the same methodology as applied for the non-CCh greenhouse gas
Land Use, Land-Use Change, and Forestry 6-31
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
emissions from forest fires that are also quantified in a separate section below as required by IPCC Guidance
and UNFCCC reporting requirements.
Emissions estimates are developed using IPCC (2006) methodology and based on U.S.-specific data and models
to quantify the primary fire-specific components: area burned; availability and combustibility of fuel; fire
severity (or consumption); and CO2 and non-CC>2 emissions. Estimated CO2 emissions for fires on forest lands in
the conterminous U.S. and in Alaska for 2021 are 203 MMT CO2 per year (Table 6-11). This estimate is an
embedded component of the net annual forest C stock change estimates provided previously (i.e., Table 6-9),
but this separate approach to estimating CO2 emissions is necessary in order to associate these emissions with
fire. See the discussion in Annex 3.13 for more details on this methodology. Note that in Alaska, a portion of the
forest lands are considered unmanaged, therefore the estimates for Alaska provided in Table 6-11 include only
managed forest land within the state, which is consistent with C stock change estimates provided above.
Table 6-11: Estimates of CO2 (MMT per Year) Emissions3 from Forest Fires in the
Conterminous 48 States and Alaska
C02 emitted from fires on forest
C02 emitted from
land in the Conterminous 48
fires on forest land in
Total C02 emitted
Year
States (MMT yr1)
Alaska (MM Tyrx)
(MMTyr1)
1990
13.6
38.6
52.2
2005
31.1
137.4
168.4
2017
119.0
4.5
123.5
2018
87.4
7.6
95.0
2019
22.3
77.9
100.2
2020
181.2
1.6
182.8
2021
196.6
5.9
202.6
a These emissions have already been included in the estimates of net annual changes in C stocks, which include the
amount sequestered minus any emissions, including the assumption that combusted wood may continue to
decay through time.
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
The methodology described herein is consistent with the 2006 IPCC Guidelines. Forest ecosystem C stocks and net
annual C stock change were determined according to the stock-difference method for the conterminous United
States, which involved applying C estimation factors to annual forest inventories across time to obtain C stocks and
then subtracting between the years to obtain the stock change. The gain-loss method was used to estimate C
stocks and net annual C stock changes in Alaska. The approaches for estimating carbon stocks and stock changes
on Forest Land Remaining Forest Land are described in Annex 3.13. All annual NFI plots available in the public FIA
database (USDA Forest Service 2022b) were used in the current Inventory. Additionally, NFI plots established and
measured in 2014 as part of a pilot inventory in interior Alaska were also included in this Inventory as were plots
established and measured since 2015 as part of the operational NFI in interior Alaska. Some of the data from the
pilot and operational NFI in interior Alaska are not yet available in the public FIA database. Only plots which meet
the definition of forest land (see Section 6.1 Representation of the U.S. Land Base) are measured in the NFI; as part
of the pre-field process in the FIA program, all plots or portions of plots (i.e., conditions) are classified into a land-
use category. This land use information on each forest and non-forest plot was used to estimate forest land area
and land converted to and from forest land over the time series. The estimates in this section of the report are
based on land use information from the NFI and they may differ from the other land-use categories where area
estimates reported in the Land Representation were not updated (see Section 6.1 Representation of the U.S. Land
Base). Further, Hawaii was not included in this section of the current Inventory, which also contributes to small
6-32 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
differences in the area estimates reported in this section and those reported in Section 6.1 Representation of the
U.S. Land Base (See Annex 3.13 for details on differences).
To implement the stock-difference approach, forest land conditions in the conterminous United States were
observed on NFI plots at time to and at a subsequent time ti=to+s, where s is the time step (time measured in
years) and is indexed by discrete (e.g., 5 year) forest age classes. The inventory from to to ti was then projected to
2021. This projection approach requires simulating changes in the age-class distribution resulting from forest aging
and disturbance events and then applying C density estimates for each age class to obtain population estimates for
the nation. In cases where there are ti estimates in the last year (e.g., 2021) of the NFI no projections are
necessary for those plots. To implement the gain-loss approach in Alaska, forest land conditions in Alaska were
observed on NFI plots from 2004 to 2021. Plot-level data from the NFI were harmonized with auxiliary data
describing climate, forest structure, disturbance, and other site-specific conditions to develop non-parametric
models to predict carbon stocks by forest ecosystem carbon pool as well as fluxes over the entire inventory period,
1990 to 2021. First, carbon stocks for each forest ecosystem carbon pool were predicted for the year 2016 for all
base intensity NFI plot locations (each plot representing approximately 2,403 ha) in coastal southeast and
southcentral Alaska and for 1/5 intensity plots in interior Alaska (each plot representing 12,015 ha). Next, the
chronosequence of sampled NFI plots and auxiliary information (e.g., climate, forest structure, disturbance, and
other site-specific data) were used to predict annual gains and losses for each forest ecosystem carbon pool. The
annual gains and losses were then combined with the stock estimates and disturbance information to compile
plot- and population-level carbon stocks and fluxes for each year from 1990 to 2021. To estimate C stock changes
in harvested wood, estimates were based on factors such as the allocation of wood to various primary and end-use
products as well as half-life (the time at which half of the amount placed in use will have been discarded from use)
and expected disposition (e.g., product pool, SWDS, combustion). An overview of the different methodologies and
data sources used to estimate the C in forest ecosystems within the conterminous states and Alaska and harvested
wood products for all of the United States is provided below. See Annex 3.13 for details and additional information
related to the methods and data.
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2021. Details on the emission/removal trends and methodologies through time are described in more
detail in the Introduction and Methodology sections.
Forest Ecosystem Carbon from Forest Inventory
The United States applied the compilation approach described in Woodall et al. (2015a) for the current Inventory
which removes the older periodic inventory data, which may be inconsistent with annual inventory data, from the
estimation procedures. This approach enables the delineation of forest C accumulation by forest growth, land-use
change, and natural disturbances such as fire. Development will continue on a system that attributes changes in
forest C to disturbances and delineates Land Converted to Forest Land from Forest Land Remaining Forest Land. As
part of this development, C pool science will continue and will be expanded to improve the estimates of C stock
transfers from forest land to other land uses and include techniques to better identify land-use change (see the
Planned Improvements section below).
Unfortunately, the annual FIA inventory system does not extend into the 1970s, necessitating the adoption of a
system to estimate carbon stocks prior to the establishment of the annual forest inventory. The estimation of
carbon stocks prior to the annual national forest inventory consisted of a modeling framework comprised of a
forest dynamics module (age transition matrices) and a land use dynamics module (land area transition matrices).
The forest dynamics module assesses forest uptake, forest aging, and disturbance effects (e.g., disturbances such
as wind, fire, and floods identified by foresters on inventory plots). The land use dynamics module assesses C stock
transfers associated with afforestation and deforestation (Woodall et al. 2015b). Both modules are developed
from land use area statistics and C stock change or C stock transfer by age class. The required inputs are estimated
from more than 625,000 forest and non-forest observations recorded in the FIA national database (U.S. Forest
Service 2022a, b, c). Model predictions prior to the annual inventory period are constructed from the estimation
system using the annual estimates. The estimation system is driven by the annual forest inventory system
Land Use, Land-Use Change, and Forestry 6-33
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
conducted by the FIA program (Frayerand Furnival 1999; Bechtold and Patterson 2005; USDA Forest Service
2022d, 2022a). The FIA program relies on a rotating panel statistical design with a sampling intensity of one 674.5
m2 ground plot per 2,403 ha of land and water area. A five or seven-panel design, with 20 percent or 14.3 percent
of the field plots typically measured each year within a state, is used in the eastern United States and a ten-panel
design, with typically 10 percent of the field plots measured each year within a state, is used in the western United
States. The interpenetrating hexagonal design across the U.S. landscape enables the sampling of plots at various
intensities in a spatially and temporally unbiased manner. Typically, tree and site attributes are measured with
higher sample intensity while other ecosystem attributes such as downed dead wood are sampled during summer
months at lower intensities. The first step in incorporating FIA data into the estimation system is to identify annual
inventory datasets by state. Inventories include data collected on permanent inventory plots on forest lands and
were organized as separate datasets, each representing a complete inventory, or survey, of an individual state at a
specified time. Many of the annual inventories reported for states are represented as "moving window" averages,
which mean that a portion—but not all—of the previous year's inventory is updated each year (USDA Forest
Service 2022d). Forest C estimates are organized according to these state surveys, and the frequency of surveys
varies by state.
Using this FIA data, separate estimates were prepared for the five C storage pools identified by IPCC (2006) and
described above. All estimates were based on data collected from the extensive array of permanent, annual forest
inventory plots and associated models (e.g., live tree belowground biomass) in the United States (USDA Forest
Service 2022b, 2022c). Carbon conversion factors were applied at the disaggregated level of each inventory plot
and then appropriately expanded to population estimates.
Carbon in Biomass
Live tree C pools include aboveground and belowground (coarse root) biomass of live trees with diameter at breast
height (dbh) of at least 2.54 cm at 1.37 m above the litter. Separate estimates were made for above- and
belowground biomass components. If inventory plots included data on individual trees, aboveground and
belowground (coarse roots) tree C was based on Woodall et al. (2011a), which is also known as the component
ratio method (CRM), and is a function of tree volume, species, and diameter. An additional component of foliage,
which was not explicitly included in Woodall et al. (2011a), was added to each tree following the same CRM
method.
Understory vegetation is a minor component of biomass, which is defined in the FIA program as all biomass of
undergrowth plants in a forest, including woody shrubs and trees less than 2.54 cm dbh. For this Inventory, it was
assumed that 10 percent of total understory C mass is belowground (Smith et al. 2006). Estimates of C density
were based on information in Birdsey (1996) and biomass estimates from Jenkins et al. (2003). Understory biomass
represented over 1 percent of C in biomass, but its contribution rarely exceeded 2 percent of the total carbon
stocks or stock changes across all forest ecosystem C pools each year.
Carbon in Dead Organic Matter
Dead organic matter is calculated as three separate pools—standing dead trees, downed dead wood, and litter—
with C stocks estimated from sample data or from models as described below. The standing dead tree C pool
includes aboveground and belowground (coarse root) biomass for trees of at least 12.7 cm dbh. Calculations
followed the basic method applied to live trees (Woodall et al. 2011a) with additional modifications to account for
decay and structural loss (Domke et al. 2011; Harmon et al. 2011). Downed dead wood estimates are based on
measurement of a subset of FIA plots for downed dead wood (Domke et al. 2013; Woodall and Monleon 2008;
Woodall et al. 2013). Downed dead wood is defined as pieces of dead wood greater than 7.5 cm diameter, at
transect intersection, that are not attached to live or standing dead trees. This includes stumps and roots of
harvested trees. To facilitate the downscaling of downed dead wood C estimates from the state-wide population
estimates to individual plots, downed dead wood models specific to regions and forest types within each region
are used. Litter C is the pool of organic C (also known as duff, humus, and fine woody debris) above the mineral
soil and includes woody fragments with diameters of up to 7.5 cm. A subset of FIA plots are measured for litter C.
6-34 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
A modeling approach, using litter C measurements from FIA plots (Domke et al. 2016), was used to estimate litter
C for every FIA plot used in the estimation framework.
Carbon in Forest Soil
Soil carbon is the largest terrestrial C sink with much of that C in forest ecosystems. The FIA program has been
consistently measuring soil attributes as part of the annual inventory since 2001 and has amassed an extensive
inventory of soil measurement data on forest land in the conterminous U.S. and coastal Alaska (O'Neill et al. 2005).
Observations of mineral and organic soil C on forest land from the FIA program and the International Soil Carbon
Monitoring Network were used to develop and implement a modeling approach that enabled the prediction of
mineral and organic (i.e., undrained organic soils) soil C to a depth of 100 cm from empirical measurements to a
depth of 20 cm and included site-, stand-, and climate-specific variables that yield predictions of soil C stocks
specific to forest land in the United States (Domke et al. 2017). This new approach allowed for separation of
mineral and organic soils, the latter also referred to as Histosols, in the Forest Land Remaining Forest Land
category. Note that mineral and organic (i.e., undrained organic soils) soil C stock changes are reported to a depth
of 100 cm for Forest Land Remaining Forest Land to remain consistent with past reporting in this category,
however for consistency across land-use categories, mineral (e.g., cropland, grassland, settlements) soil C is
reported to a depth of 30 cm in Section 6.3 Land Converted to Forest Land. Estimates of C stock changes from
organic soils shown in Table 6-8 and Table 6-9 include the emissions from drained organic forest soils, and the
methods used to develop these estimates can be found in the Drained Organic Soils section below.
Harvested Wood Carbon
Estimates of the HWP contribution to forest C sinks and emissions (hereafter called "HWP contribution") were
based on methods described in Skog (2008) using the WOODCARB II model. These methods are based on IPCC
(2006) guidance for estimating the HWP contribution. IPCC (2006) provides methods that allow for reporting of the
HWP contribution using one of several different methodological approaches: Production, stock change and
atmospheric flow, as well as a default method that assumes there is no change in HWP C stocks (see Annex 3.13
for more details about each approach). The United States uses the production approach to report HWP
contribution. Under the production approach, C in exported wood was estimated as if it remains in the United
States, and C in imported wood was not included in the estimates. Though reported U.S. HWP estimates are based
on the production approach, estimates resulting from use of the two alternative approaches, the stock change and
atmospheric flow approaches, are also presented for comparison (see Annex 3.13). Annual estimates of change
were calculated by tracking the annual estimated additions to and removals from the pool of products held in end
uses (i.e., products in use such as housing or publications) and the pool of products held in SWDS. The C loss from
harvest is reported in the Forest Ecosystem component of the Forest Land Remaining Forest Land and Land
Converted to Forest Land sections and for informational purposes in the Energy sector, but the non-C02 emissions
associated with biomass energy are included in the Energy sector emissions (see Chapter 3). EPA includes HWP
within the forest chapter because forests are the source of wood that goes into the HWP estimates.
Solidwood products include lumber and panels. End-use categories for solidwood include single and multifamily
housing, alteration and repair of housing, and other end uses. There is one product category and one end-use
category for paper. Additions to and removals from pools were tracked beginning in 1900, with the exception of
additions of softwood lumber to housing, which began in 1800. Solidwood and paper product production and
trade data were taken from USDA Forest Service and other sources (Hair and Ulrich 1963; Hair 1958; USDC Bureau
of Census 1976; Ulrich 1985,1989; Steer 1948; AF&PA 2006a, 2006b; Howard 2003, 2007; Howard and Jones 2016;
Howard and Liang 2019). Estimates for disposal of products reflects the change over time in the fraction of
products discarded to SWDS (as opposed to burning or recycling) and the fraction of SWDS that were in sanitary
landfills versus dumps.
There are five annual HWP variables that were used in varying combinations to estimate HWP contribution using
any one of the three main approaches listed above. These are:
(1A) annual change of C in wood and paper products in use in the United States,
Land Use, Land-Use Change, and Forestry 6-35
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
(IB) annual change of C in wood and paper products in SWDS in the United States,
(2A) annual change of C in wood and paper products in use in the United States and other countries where the
wood came from trees harvested in the United States,
(2B) annual change of C in wood and paper products in SWDS in the United States and other countries where
the wood came from trees harvested in the United States,
(3) C in imports of wood, pulp, and paper to the United States,
(4) C in exports of wood, pulp and paper from the United States, and
(5) C in annual harvest of wood from forests in the United States.
The sum of variables 2A and 2B yielded the estimate for HWP contribution under the production estimation
approach. A key assumption for estimating these variables that adds uncertainty in the estimates was that
products exported from the United States and held in pools in other countries have the same half-lives for
products in use, the same percentage of discarded products going to SWDS, and the same decay rates in SWDS as
they would in the United States.
Uncertainty
A quantitative uncertainty analysis placed bounds on the flux estimates for forest ecosystems through a
combination of sample-based and model-based approaches to uncertainty estimation for forest ecosystem CO2
flux using IPCC Approach 1 (Table 6-12 and Table A-214 for state-level uncertainties). A Monte Carlo Stochastic
Simulation of the methods described above, and probabilistic sampling of C conversion factors, were used to
determine the HWP uncertainty using IPCC Approach 2. See Annex 3.13 for additional information. The 2021 net
annual change for forest C stocks was estimated to be between -773.6 and -618.1 MMT CO2 Eq. around a central
estimate of-695.4 MMT CO2 Eq. at a 95 percent confidence level. This includes a range of-665.6 to -519.5 MMT
CO2 Eq. around a central estimate of-592.5 MMT CO2 Eq. for forest ecosystems and -130.9 to -77.8 MMT CO2 Eq.
around a central estimate of-102.8 MMT CO2 Eq. for HWP.
Table 6-12: Quantitative Uncertainty Estimates for Net CO2 Flux from Forest Land
Remaining Forest Land: Changes in Forest C Stocks (MMT CO2 Eq. and Percent)
2021 Flux Estimate Uncertainty Range Relative to Flux Estimate
(MMT C02 Eq.) (MMT CP2 Eq.) (%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Forest Ecosystem C Pools3
C02
(592.5)
(665.6)
(519.5)
-12.3%
12.3%
Harvested Wood Products'5
C02
(102.8)
(130.9)
(77.8)
-27.3%
24.3%
Total Forest
C02
(695.4)
(773.6)
(618.1)
-11.3%
11.1%
a Range of flux estimates predicted through a combination of sample-based and model-based uncertainty for a 95 percent
confidence interval, IPCC Approach 1.
b Range of flux estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval, IPCC
Approach 2.
Notes: Parentheses indicate negative values or net uptake. Totals may not sum due to independent rounding.
QA/QC and Verification
The FIA program has conducted consistent forest surveys based on extensive statistically-based sampling of most
of the forest land in the conterminous U.S., dating back to 1952. The FIA program includes numerous quality
assurance and quality control (QA/QC) procedures, including calibration among field crews, duplicate surveys of
some plots, and systematic checking of recorded data. Because of the statistically-based sampling, the large
number of survey plots, and the quality of the data, the survey databases developed by the FIA program form a
strong foundation for C stock estimates. Field sampling protocols, summary data, and detailed inventory databases
6-36 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
are archived and are publicly available (USDA Forest Service 2022d).
General quality control procedures were used in performing calculations to estimate C stocks based on survey
data. For example, the C datasets, which include inventory variables such as areas and volumes, were compared to
standard inventory summaries such as the forest resource statistics of Oswalt et al. (2019) or selected population
estimates generated from the FIA database, which are available at an FIA internet site (USDA Forest Service
2022b). Agreement between the C datasets and the original inventories is important to verify accuracy of the data
used.
Estimates of the HWP variables and the HWP contribution under the production estimation approach use data
from U.S. Census and USDA Forest Service surveys of production and trade and other sources (Hair and Ulrich
1963; Hair 1958; USDC Bureau of Census 1976; Ulrich 1985,1989; Steer 1948; AF&PA 2006a, 2006b; Howard 2003,
2007; Howard and Jones 2016; Howard and Liang 2019; AF&PA 2021; FAO 2021). Factors to convert wood and
paper to units of C are based on estimates by industry and U.S. Forest Service published sources (see Annex 3.13).
The WOODCARB II model uses estimation methods suggested by IPCC (2006). Estimates of annual C change in
solidwood and paper products in use were calibrated to meet two independent criteria. The first criterion is that
the WOODCARB II model estimate of C in houses standing in 2001 needs to match an independent estimate of C in
housing based on U.S. Census and USDA Forest Service survey data. Meeting the first criterion resulted in an
estimated half-life of about 80 years for single family housing built in the 1920s, which is confirmed by other U.S.
Census data on housing. The second criterion is that the WOODCARB II model estimate of wood and paper being
discarded to SWDS needs to match EPA estimates of discards used in the Waste sector each year over the period
1990 to 2000 (EPA 2006). These criteria help reduce uncertainty in estimates of annual change in C in products in
use in the United States and, to a lesser degree, reduce uncertainty in estimates of annual change in C in products
made from wood harvested in the United States. In addition, WOODCARB II landfill decay rates have been
validated by ensuring that estimates of CH4 emissions from landfills based on EPA (2006) data are reasonable in
comparison to CH4 estimates based on WOODCARB II landfill decay rates.
Recalculations Discussion
The methods used in the current Inventory to compile estimates for forest ecosystem carbon stocks and stock
changes and HWPs from 1990 through 2021 are consistent with those used in the previous (1990 through 2020)
Inventory. Population estimates of carbon stocks and stock changes were compiled using NFI data from each U.S.
state and national estimates were compiled by summing over all states. New NFI data in most states were
incorporated in the latest Inventory which contributed to decreases in forest land area estimates and carbon
stocks, particularly in Alaska where new data from 2018 to 2021, particularly litter and soil data, were included
(Table 6-13). Fire data sources were also updated for Alaska through 2021 and this combined with the new NFI
data for the years 2018 through 2021 resulted in substantial changes in carbon stocks and stock changes. Soil
(organic) carbon stocks decreased in the latest Inventory relative to the previous Inventory and mineral soil carbon
stocks increased slightly in this Inventory relative to the previous Inventory. These changes can be attributed to
obtaining plot-level soil orders using the more refined gridded National Soil Survey Geographic Database
(gNATSGO) dataset (Soil Survey Staff 2020a, 2020b), rather than the Digital General Soil Map of the United States
(STATSG02) dataset which had been used in previous Inventories (Table 6-13). This resulted in a structural change
in the soil carbon estimates for mineral and organic soils across the entire time series, particularly in Alaska where
new data on forest area was included for the years 2018 through 2021 (Table 6-8). Finally, recent land-use change
in Alaska (since 2015) also contributed to variability in soil carbon stocks and stock changes in recent years in the
time series, which led to differences in estimates in the previous Inventory and the current Inventory. New data
included in the HWP time-series result in a minor decrease (< 1 percent) in carbon stocks in the HWP pools but a
substantial increase (60 percent) in the carbon stock change estimates for Products in Use and to a lesser extent (2
percent) in SWDS between the previous Inventory and the current Inventory. With the easing of the global
pandemic and the return of consumers to the marketplace, there was a rebound in the purchase and accumulation
of both paper and solid wood products. This rebound is expected to continue in 2022.
Land Use, Land-Use Change, and Forestry 6-37
-------
1 Table 6-13: Recalculations of Forest Area (1,000 ha) and C Stocks in Forest Land Remaining
2 Forest Land and Harvested Wood Pools (MMT C)
2021 Estimate, 2021 Estimate, 2022 Estimate,
Previous Inventory Current Inventory Current Inventory
Forest Area (1000 ha)
281,951
279,962
279,800
Carbon Pools (MMT C)
Forest
58,316
56,790
56,951
Aboveground Biomass
15,688
15,749
15,861
Belowground Biomass
3,106
3,121
3,143
Dead Wood
2,896
2,799
2,827
Litter
3,810
3,888
3,888
Soil (Mineral)
25,459
25,915
25,916
Soil (Organic)
7,357
5,317
5,317
Harvested Wood
2,718
2,721
2,749
Products in Use
1,536
1,539
1,549
SWDS
1,182
1,182
1,200
Total Stock
61,034
59,511
59,701
Note: Totals may not sum due to independent rounding.
3 Table 6-14: Recalculations of Net C Flux from Forest Ecosystem Pools in Forest Land
4 Remaining Forest Land and Harvested Wood Pools (MMT C)
Carbon Pool (MMT C)
2020 Estimate,
Previous Inventory
2020 Estimate,
Current Inventory
2021 Estimate,
Current Inventory
Forest
(159.4)
(166.6)
(161.6)
Aboveground Biomass
(108.7)
(114.3)
(111.6)
Belowground Biomass
(21.6)
(22.7)
(22.1)
Dead Wood
(27.7)
(27.9)
(27.6)
Litter
(0.5)
(0.5)
0.5
Soil (Mineral)
(1.1)
(1.5)
(1.1)
Soil (Organic)
0.1
0.0
0.0
Drained organic soil
0.2
0.2
0.2
Harvested Wood
(22.8)
(26.3)
(28.0)
Products in Use
(5.5)
(8.7)
(10.3)
SWDS
(17.3)
(17.6)
(17.7)
Total Net Flux
(182.2)
(192.9)
(189.6)
Note: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
5 Planned Improvements
6 Reliable estimates of forest C stocks and changes across the diverse ecosystems of the United States require a high
7 level of investment in both annual monitoring and associated analytical techniques. Development of improved
8 monitoring/reporting techniques is a continuous process that occurs simultaneously with annual Inventory
9 submissions. Planned improvements can be broadly assigned to the following categories: development of a robust
10 estimation and reporting system, individual C pool estimation, coordination with other land-use categories, and
11 annual inventory data incorporation.
12 While this Inventory submission includes C change by Forest Land Remaining Forest Land and Land Converted to
13 Forest Land and C stock changes for all IPCC pools in these two categories, there are many improvements that are
14 still necessary. The estimation approach used for the conterminous United States in the current Inventory for the
15 forest land category operates at the state scale, whereas previously the western United States and southeast and
16 southcentral coastal Alaska operated at a regional scale. While this is an improvement over previous Inventories
17 and led to improved estimation and separation of land-use categories in the current Inventory, research is
6-38 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
underway to leverage all FIA data and auxiliary information (i.e., remotely sensed information) to operate at finer
spatial and temporal scales. As in past submissions, emissions and removals associated with natural (e.g., wildfire,
insects, and disease) and human (e.g., harvesting) disturbances are implicitly included in the report given the
design of the annual NFI, but not explicitly estimated. In addition to integrating auxiliary information into the
estimation framework and leveraging all NFI plot measurements, alternative estimators are also being evaluated
which will eliminate latency in population estimates from the NFI, improve annual estimation and characterization
of interannual variability, facilitate attribution of fluxes to particular activities, and allow for streamlined
harmonization of NFI data with auxiliary data products. This will also facilitate separation of prescribed and wildfire
emissions in future reports. The transparency and repeatability of estimation and reporting systems will be
improved through the dissemination of open-source code (e.g., R programming language) in concert with the
public availability of the annual NFI (USDA Forest Service 2022b). Also, several FIA database processes are being
institutionalized to increase efficiency and QA/QC in reporting and further improve transparency, completeness,
consistency, accuracy, and availability of data used in reporting. Finally, a combination of approaches was used to
estimate uncertainty associated with C stock changes in the Forest Land Remaining Forest Land category in this
report. There is research underway investigating more robust approaches to estimate total uncertainty (Clough et
al. 2016), which will be considered in future Inventory reports.
The modeling framework used to estimate downed dead wood within the dead wood C pool (Smith et al. 2022)
will be updated similar to the litter (Domke et al. 2016) and soil C pools (Domke et al. 2017). Finally, components of
other pools, such as C in belowground biomass (Russell et al. 2015) and understory vegetation (Russell et al. 2014;
Johnson et al. 2017), are being explored but may require additional investment in field inventories before
improvements can be realized in the Inventory report.
The foundation of forest C estimation and reporting is the annual NFI. The ongoing annual surveys by the FIA
program are expected to improve the accuracy and precision of forest C estimates as new state surveys become
available (USDA Forest Service 2022b). With the exception of Wyoming (which will have sufficient remeasurements
in the years ahead), all other states in the conterminous United States now have sufficient annual NFI data to
consistently estimate C stocks and stock changes for the future using the state-level compilation system. The FIA
program continues to install permanent plots in Alaska as part of the operational NFI, and as more plots are added
to the NFI, they will be used to improve estimates for all managed forest land in Alaska. The methods used to
include all managed forest land in the conterminous United States will be used in future Inventories for Hawaii and
U.S. Territories as forest C data become available (only a small number of plots from Hawaii are currently available
from the annualized sampling design). To that end, research is underway to incorporate all NFI information (both
annual and periodic data) and the dense time series of remotely sensed data in multiple inferential frameworks for
estimating greenhouse gas emissions and removals as well as change (i.e., disturbance or land-use changes)
detection and attribution across the entire reporting period and all managed forest land in the United States.
Leveraging this auxiliary information will aid the efforts to improve estimates for interior Alaska as well as the
entire inventory system. In addition to fully inventorying all managed forest land in the United States, the more
intensive sampling (i.e., more samples) of fine woody debris, litter, and SOC on a subset of FIA plots continues and
will substantially improve spatial and temporal resolution of C pools (Westfall et al. 2013) as this information
becomes available (Woodall et al. 2011b). Increased sample intensity of some C pools and using annualized
sampling data as it becomes available for those states currently not reporting are planned for future submissions.
There will also be improved methods and models to characterize standing live and dead tree carbon in the next
Inventory. The NFI sampling frame extends beyond the forest land-use category (e.g., woodlands, which fall into
the grasslands land-use category, and urban areas, which fall into the settlements land-use category) with
inventory-relevant information for trees outside of forest land. These data will be utilized as they become available
in the NFI.
Non-C02 Emissions from Forest Fires
Emissions of non-CC>2 gases from forest fires were estimated using U.S.-specific data and models for annual area of
forest burned, fuel, consumption, and emission consistent with IPCC (2006). In 2021, emissions from this source
Land Use, Land-Use Change, and Forestry 6-39
-------
1 were estimated to be 15.5 MMT CO2 Eq. of CFU and 8.9 MMT CO2 Eq. of N2O (Table 6-15; kt units provided in Table
2 6-16). The estimates of non-CC>2 emissions from forest fires include the conterminous 48 states plus managed
3 forest land in Alaska (Ogle et al. 2018).
4 Table 6-15: N011-CO2 Emissions from Forest Fires (MMT CO2 Eq.)a
Gas
1990
2005
2017
2018
2019
2020
2021
ch4
3.2
10.9
9.6
6.9
6.4
15.0
15.5
n2o
2.3
7.4
5.4
4.2
4.4
8.0
8.9
Total
5.5
18.3
15.0
11.0
10.8
23.0
24.4
a These estimates include Non-C02 emissions from forest fires on Forest Land Remaining
Forest Land and Land Converted to Forest Land.
Note: Totals may not sum due to independent rounding
5 Table 6-16: N011-CO2 Emissions from Forest Fires (kt)a
Gas
1990
2005
2017
2018
2019
2020
2021
ch4
116
39.
342
245
228
534
554
n2o
9
28
21
16
17
30
34
CO
2985
10,039
7,298
5,347
5,885
11,080
11,798
NOx
48
145
122
100
89
171
201
a These estimates include Non-C02 emissions from forest fires on Forest Land Remaining
Forest Land and Land Converted to Forest Land.
6 Methodology and Time-Series Consistency
7 Non-CC>2 emissions from forest fires—primarily CFU and N2O emissions—were calculated consistent with IPCC
8 (2006) methodology, which included U.S.-specific data and models on area burned, fuel, consumption, and
9 emission. The annual estimates were calculated by the Wildland Fire Emissions Inventory System (WFEIS, French et
10 al. 2011, 2014) with area burned based on Monitoring Trends in Burn Severity (MTBS, Eidenshink et al. 2007) or
11 MODIS burned area mapping (MODIS MCD64A1, Giglio et al. 2018) data. The MTBS data available for this report
12 (MTBS 2022) included fires through 2020, and the MODIS-based records include 2001 through 2021. Emissions
13 reported here are calculated from MTBS data for the 1990 to 2020 interval, and the 2001 through 2021 emissions
14 are also based on MODIS burned areas. Where both the MTBS and MODIS sources are available, the predictions
15 are averaged. Note that N2O emissions are not included in WFEIS calculations; the emissions provided here are
16 based on the average N2O to CO2 ratio of 0.000166 following Larkin et al. (2014). See Emissions from Forest Fires in
17 Annex 3.13 for further details on all fire-related emissions calculations for forests. Consistent use of available data
18 sources, data processing, and calculation methods were applied to the entire time series to ensure time-series
19 consistency from 1990 through 2021.
20 Uncertainty
21 Uncertainty estimates for non-C02 emissions from forest fires are based on a Monte Carlo (IPCC Approach 2)
22 approach to propagate variability among the alternate WFEIS annual estimates per state. Uncertainty in parts of
23 the WFEIS system are not currently quantified. Among potential sources for future analysis are burned areas from
24 MTBS or MODIS, the fuels models or the Consume model (Prichard et al. 2014). See Annex 3.13 for the quantities
25 and assumptions employed to define and propagate uncertainty. The results of the Approach 2 quantitative
26 uncertainty analysis are summarized in Table 6-17.
6-40 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Table 6-17: Quantitative Uncertainty Estimates of Non-C02 Emissions from Forest Fires
2 (MMT CO2 Eq. and Percent)3
Source
Gas
2021 Emission Estimate
(MMT C02 Eq.)
Uncertainty Range Relative to Emission Estimateb
(MMT C02 Eq.) (%)
Lower
Upper
Lower Upper
Bound
Bound
Bound Bound
Non-C02 Emissions from
Forest Fires
ch4
15.5
10.5
20.5
-32% 32%
Non-C02 Emissions from
Forest Fires
n2o
8.9
2.6
15.3
-71% 72%
a These estimates include Non-C02 emissions from forest fires on Forest Land Remaining Forest Land and Land
Converted to Forest Land.
b Range of flux estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
3 QA/QC and Verification
4 Tier 1 and Tier 2 QA/QC activities were conducted consistent with the U.S. QA/QC plan. Source-specific quality
5 control measures for estimating non-C02 emissions from forest fires included checking input data, documentation,
6 and calculations to ensure data were properly handled through the inventory process and results were consistent
7 with values expected from those calculations. The QA/QC procedures did not reveal any inaccuracies or incorrect
8 input values.
9 Recalculations Discussion
10 The methods used in the current (1990 through 2021) Inventory to compile estimates of non-C02 emissions from
11 forest fires represent a slight change relative to the previous (1990 through 2020) Inventory. The basic
12 components of calculating forest fire emissions (IPCC 2006) remain unchanged, but the WFEIS-based estimates
13 now include both MTBS and MODIS based burns and two alternate fuel models where available. An additional
14 source of change leading to recalculations are recent and ongoing updates to the MTBS fire records (i.e., including
15 both most-recent as well as possible updates to past years' fires).
16 The EPA also updated global warming potentials (GWP) for calculating C02-equivalent emissions of CH4 (from 25 to
17 28) and N20 (from 298 to 265) to reflect the 100-year GWP values provided in the IPCC Fifth Assessment Report
18 (AR5) (IPCC 2013). The previous Inventory used 100-year GWPs provided in the IPCC Fourth Assessment Report
19 (AR4). This update was applied across the entire time series.
20 The net result of implementing AR5 GWP values and other improvements listed above was an average annual
21 increase of 0.2 MMT CO2 Eq., or 1 percent, in total non-C02 emissions from forest fires across the entire time
22 series. Further discussion on this update and the overall impacts of updating the Inventory GWP values to reflect
23 the AR5 can be found in Chapter 9, Recalculations and Improvements.
24 Planned Improvements
25 Continuing improvements are planned for developing better fire and site-specific estimates for forest fires. The
26 focus will be on addressing three aspects of reporting: best use of WFEIS, better resolution of uncertainty as
27 discussed above, and identification of burned areas that are not captured by MTBS records.
28 N20 Emissions from N Additions to Forest Soils
29 Of the synthetic nitrogen (N) fertilizers applied to soils in the United States, no more than one percent is applied to
30 forest soils. Application rates are similar to those occurring on cropland soils, but in any given year, only a small
31 proportion of total forested land receives N fertilizer. This is because forests are typically fertilized only twice
Land Use, Land-Use Change, and Forestry 6-41
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
during their approximately 40-year growth cycle (once at planting and once midway through their life cycle). While
the rate of N fertilizer application for the area of forests that receives N fertilizer in any given year is relatively high,
the annual application rate is quite low over the entire area of forest land.
N additions to soils result in direct and indirect N2O emissions. Direct emissions occur on-site due to the N
additions. Indirect emissions result from fertilizer N that is transformed and transported to another location
through volatilization in the form of ammonia [NH3] and nitrogen oxide [NOx], in addition to leaching and runoff of
nitrates [NO3], and later converted into N2O at off-site locations from the original N application. The indirect
emissions are assigned to forest land because the management activity leading to the emissions occurred in forest
land.
Direct soil N2O emissions from Forest Land Remaining Forest Land and Land Converted to Forest Land33 in 2021
were 0.3 MMT CO2 Eq. (1.2 kt), and the indirect emissions were 0.1 MMT CO2 Eq. (0.4 kt). Total emissions for 2021
were 0.4 MMT CO2 Eq. (1.5 kt) and have increased by 455 percent from 1990 to 2021. Total forest soil N2O
emissions are summarized in Table 6-18.
Table 6-18: N2O Fluxes from Soils in Forest Land Remaining Forest Land and Land Converted
to Forest Land (MMT CO2 Eq. and kt N2O)
1990
2005
2017
2018
2019
2020
2021
Direct N20 Fluxes from Soils
MMTCO2 Eq.
0.1
0.3
0.3
0.3
0.3
0.3
0.3
kt N20
0.2
1.2
1.2
1.2
1.2
1.2
1.2
Indirect N20 Fluxes from Soils
MMTC02 Eq.
+
0.1
0.1
0.1
0.1
0.1
0.1
kt N20
0.1
0.4
0.4
0.4
0.4
0.4
0.4
Total
MMT C02 Eq.
0.1
0.4
0.4
0.4
0.4
0.4
0.4
kt N20
0.3
1.5
1.5
1.5
1.5
1.5
1.5
+ Does not exceed 0.05 MMT C02 Eq. or 0.05 kt.
Notes: Totals may not sum due to independent rounding. The N20 emissions from Land Converted
to Forest Land are included with Forest Land Remaining Forest Land because it is not currently
possible to separate the activity data by land-use conversion category.
Methodology and Time-Series Consistency
The IPCC Tier 1 approach is used to estimate N2O from soils within Forest Land Remaining Forest Land and Land
Converted to Forest Land. According to U.S. Forest Service statistics for 1996 (USDA Forest Service 2001),
approximately 75 percent of trees planted are for timber, and about 60 percent of national total harvested forest
area is in the southeastern United States. Although southeastern pine plantations represent the majority of
fertilized forests in the United States, this Inventory also incorporated N fertilizer application to commercial
Douglas-fir stands in western Oregon and Washington. For the Southeast, estimates of direct N2O emissions from
fertilizer applications to forests are based on the area of pine plantations receiving fertilizer in the southeastern
United States and estimated application rates (Albaugh et al. 2007; Fox et al. 2007). Fertilizer application is rare for
hardwoods and therefore not included in the inventory (Binkley et al. 1995). For each year, the area of pine
receiving N fertilizer is multiplied by the weighted average of the reported range of N fertilization rates (121 lbs. N
per acre). Area data for pine plantations receiving fertilizer in the Southeast are not available for 2005 through
2021, so data from 2004 are used for these years. For commercial forests in Oregon and Washington, only fertilizer
applied to Douglas-fir is addressed in the inventory because the vast majority (approximately 95 percent) of the
total fertilizer applied to forests in this region is applied to Douglas-fir (Briggs 2007). Estimates of total Douglas-fir
33 The N20 emissions from Land Converted to Forest Land are included with Forest Land Remaining Forest Land because it is
not currently possible to separate the activity data by land-use conversion category.
6-42 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 area and the portion of fertilized area are multiplied to obtain annual area estimates of fertilized Douglas-fir
2 stands. Similar to the Southeast, data are not available for 2005 through 2021, so data from 2004 are used for
3 these years. The annual area estimates are multiplied by the typical rate used in this region (200 lbs. N per acre) to
4 estimate total N applied (Briggs 2007), and the total N applied to forests is multiplied by the IPCC (2006) default
5 emission factor of one percent to estimate direct N2O emissions.
6 For indirect emissions, the volatilization and leaching/runoff N fractions for forest land are calculated using the
7 IPCC default factors of 10 percent and 30 percent, respectively. The amount of N volatilized is multiplied by the
8 IPCC default factor of one percent for the portion of volatilized N that is converted to N2O off-site. The amount of
9 N leached/runoff is multiplied by the IPCC default factor of 0.075 percent for the portion of leached/runoff N that
10 is converted to N2O off-site. The resulting estimates are summed to obtain total indirect emissions.
11 The same method is applied in all years of this Inventory to ensure time-series consistency from 1990 through
12 2021.
13 Uncertainty
14 The amount of N2O emitted from forests depends not only on N inputs and fertilized area, but also on a large
15 number of variables, including organic C availability, oxygen gas partial pressure, soil moisture content, pH,
16 temperature, and tree planting/harvesting cycles. The effect of the combined interaction of these variables on N2O
17 flux is complex and highly uncertain. IPCC (2006) does not incorporate any of these variables into the default
18 methodology, except variation in estimated fertilizer application rates and estimated areas of forested land
19 receiving N fertilizer. All forest soils are treated equivalently under this methodology. Furthermore, only
20 applications of synthetic N fertilizers to forest are captured in this Inventory, so applications of organic N fertilizers
21 are not estimated. However, the total quantity of organic N inputs to soils in the United States is included in the
22 inventory for Agricultural Soil Management (Section 5.4) and Settlements Remaining Settlements (Section 6.10).
23 Uncertainties exist in the fertilization rates, annual area of forest lands receiving fertilizer, and the emission
24 factors. Fertilization rates are assigned a default level34 of uncertainty at ±50 percent, and area receiving fertilizer
25 is assigned a ±20 percent according to expert knowledge (Binkley 2004). The uncertainty ranges around the 2004
26 activity data and emission factor input variables are directly applied to the 2021 emission estimates. IPCC (2006)
27 provided estimates for the uncertainty associated with direct and indirect N2O emission factor for synthetic N
28 fertilizer application to soils.
29 Uncertainty is quantified using simple error propagation methods (IPCC 2006). The results of the quantitative
30 uncertainty analysis are summarized in Table 6-19. Direct N2O fluxes from soils in 2021 are estimated to be
31 between 0.1 and 1.0 MMT CO2 Eq. at a 95 percent confidence level. This indicates a range of 59 percent below and
32 211 percent above the emission estimate of 0.3 MMT CO2 Eq. for 2021. Indirect N2O emissions in 2021 are 0.1
33 MMT CO2 Eq. and have a range are between 0.01 and 0.3 MMT CO2 Eq., which is 86 percent below to 238 percent
34 above the emission estimate for 2021.
35 Table 6-19: Quantitative Uncertainty Estimates of N2O Fluxes from Soils in Forest Land
36 Remaining Forest Land and Land Converted to Forest Land (MMT CO2 Eq. and Percent)
Source
Gas
2021 Emission Estimate
(MMT C02 Eq.)
Uncertainty Range Relative to Emission Estimate
(MMTC02Eq.) (%)
Forest Land Remaining Forest
Lower
Upper
Lower
Upper
Land
Bound
Bound
Bound
Bound
Direct N20 Fluxes from Soils
N20
0.3
0.1
1.0
-59%
+211%
Indirect N20 Fluxes from Soils
n2o
0.1
+
0.3
-86%
+238%
+ Does not exceed 0.05 MMT C02 Eq.
34 Uncertainty is unknown for the fertilization rates so a conservative value of ±50 percent is used in the analysis.
Land Use, Land-Use Change, and Forestry 6-43
-------
1 QA/QC and Verification
2 The spreadsheet containing fertilizer applied to forests and calculations for N2O and uncertainty ranges are
3 checked and verified based on the sources of these data.
4 Recalculations Discussion
5 EPA updated global warming potential (GWP) for calculating CC>2-equivalent emissions of N2O (from 298 to 265) to
6 reflect the 100-year GWP values provided in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). The previous
7 Inventory used 100-year GWP values provided in the IPCC Fourth Assessment Report (AR4). This update was
8 applied across the entire time series.
9 As a result of this change, calculated CC>2-equivalent emissions decreased by an annual average of 0.04 MMT CO2
10 Eq., or 11 percent, over the time series from 1990 to 2020 compared to the previous Inventory.
11 Further discussion on this update and the overall impacts of updating the Inventory GWP values to reflect the AR5
12 can be found in Chapter 9, Recalculations and Improvements.
13 C02, CH4, and N20 Emissions from Drained Organic Soils35
14 Drained organic soils on forest land are identified separately from other forest soils largely because mineralization
15 of the exposed or partially dried organic material results in continuous CO2 and N2O emissions (IPCC 2006). In
16 addition, the 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands
17 (IPCC 2014) calls for estimating CH4 emissions from these drained organic soils and the ditch networks used to
18 drain them.
19 Organic soils are identified on the basis of thickness of organic horizon and percent organic matter content. All
20 organic soils are assumed to have originally been wet, and drained organic soils are further characterized by
21 drainage or the process of artificially lowering the soil water table, which exposes the organic material to drying
22 and the associated emissions described in this section. The land base considered here is drained inland organic
23 soils that are coincident with forest area as identified by the NFI of the USDA Forest Service (USDA Forest Service
24 2022b).
25 The estimated area of drained organic soils on forest land is 70,849 ha and did not change over the time series
26 based on the data used to compile the estimates in the current Inventory. These estimates are based on
27 permanent plot locations of the NFI (USDA Forest Service 2022b) coincident with mapped organic soil locations
28 (STATSG02 2016), which identifies forest land on organic soils. Forest sites that are drained are not explicitly
29 identified in the data, but for this estimate, planted forest stands on sites identified as mesic or xeric (which are
30 identified in USDA Forest Service 2022c, d) are labeled "drained organic soil" sites.
31 Land use, region, and climate are broad determinants of emissions as are more site-specific factors such as
32 nutrient status, drainage level, exposure, or disturbance. Current data are limited in spatial precision and thus lack
33 site specific details. At the same time, corresponding emissions factor data specific to U.S. forests are similarly
34 lacking. Tier 1 estimates are provided here following IPCC (2014). Total annual non-CC>2 emissions on forest land
35 with drained organic soils in 2021 are estimated as 0.8 MMT CO2 Eq. per year (Table 6-20; kt units provided in
36 6-21).
37 The Tier 1 methodology provides methods to estimate emissions of CO2 from three pathways: direct emissions
38 primarily from mineralization; indirect, or off-site, emissions associated with dissolved organic carbon releasing
39 CO2 from drainage waters; and emissions from (peat) fires on organic soils. Data about forest fires specifically
35 Estimates of C02 emissions from drained organic soils are described in this section but reported in Table 6-8 and Table 6-9 for
both Forest Land Remaining Forest Land and Land Converted to Forest Land in order to allow for reporting of all C stock changes
on forest lands in a complete and comprehensive manner.
6-44 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 located on drained organic soils are not currently available; as a result, no corresponding estimate is provided
2 here. Non-CCh emissions provided here include Cm and N2O. Methane emissions generally associated with anoxic
3 conditions do occur from the drained land surface, but the majority of these emissions originate from ditches
4 constructed to facilitate drainage at these sites. Emission of N2O can be significant from these drained organic soils
5 in contrast to the very low emissions from wet organic soils.
6 Table 6-20: N011-CO2 Emissions from Drained Organic Forest Soilsa'b (MMT CO2 Eq.)
Source
1990
2005
2017
2018
2019
2020
2021
ch4
+
+
+
+
+
+
+
n2o
0.1
0.1
0.1
0.1
0. 1
0.1
0.1
Total
0.1
0.1
0.1
0.1
0.1
0.1
0.1
+ Does not exceed 0.05 MMT C02 Eq.
a This table includes estimates from Forest Land Remaining Forest Land and Land Converted to
Forest Land.
b Estimates of C02 emissions from drained organic soils are described in this section but reported
in Table 6-8 and Table 6-9 for both Forest Land Remaining Forest Land and Land Converted to
Forest Land in order to allow for reporting of all C stock changes on forest lands in a complete
and comprehensive manner.
Note: Totals may not sum due to independent rounding.
7 Table 6-21: Non-C02 Emissions from Drained Organic Forest Soilsa'b (kt)
Source
1990
2005
2017
2018
2019
2020
2021
ch4
1
1
1
1
1
1
1
n2o
+
+
+
+
+
+
+
+ Does not exceed 0.5 kt.
a This table includes estimates from Forest Land Remaining Forest Land and Land Converted to
Forest Land.
b Estimates of C02 emissions from drained organic soils are described in this section but reported
in Table 6-8 and Table 6-9 for both Forest Land Remaining Forest Land and Land Converted to
Forest Land in order to allow for reporting of all C stock changes on forest lands in a complete
and comprehensive manner.
8 Methodology and Time-Series Consistency
9 The Tier 1 methods for estimating CO2, Cm and N2O emissions from drained inland organic soils on forest lands
10 follow IPCC (2006), with extensive updates and additional material presented in the 2013 Supplement to the 2006
11 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands (IPCC 2014). With the exception of quantifying
12 area of forest on drained organic soils, which is user-supplied, all quantities necessary for Tier 1 estimates are
13 provided in Chapter 2, Drained Inland Organic Soils of IPCC (2014).
14 Estimated area of drained organic soils on forest land is 70,849 ha based on analysis of the permanent NFI of the
15 USDA Forest Service and did not change over the time series. The most recent plot data per state within the
16 inventories were used in a spatial overlay with the STATSG02 (2016) soils data, and forest plots coincident with the
17 soil order histosol were selected as having organic soils. Information specific to identifying "drained organic" are
18 not in the inventory data so an indirect approach was employed here. Specifically, artificially regenerated forest
19 stands (inventory field STDORGCD=l) on mesic or xeric sites (inventory field 11
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Table 6-22: States identified as having Drained Organic Soils, Area of Forest on Drained
Organic Soils, and Sampling Error
State
Forest on Drained Organic Soil
(1,000 ha)
Sampling Error (68.3% as ±
Percentage of Estimate)
Florida
2.4
79
Georgia
3.7
71
Michigan
18.7
34
Minnesota
30.2
19
North Carolina
1.3
99
Virginia
2.3
102
Washington
2.1
101
Wisconsin
10.1
30
Total
70.8
14
Note: Totals may not sum due to independent rounding.
The Tier 1 methodology provides methods to estimate emissions for three pathways of C emission as CO2. Note
that subsequent mention of equations and tables in the remainder of this section refer to Chapter 2 of IPCC (2014).
The first pathway-direct CO2 emissions-is calculated according to Equation 2.3 and Table 2.1 as the product of
forest area and emission factor for temperate drained forest land. The second pathway—indirect, or off-site,
emissions—is associated with dissolved organic carbon (DOC) releasing CO2 from drainage waters according to
Equation 2.4 and Table 2.2, which represent a default composite of the three pathways for this flux: (1) the flux of
DOC from natural (undrained) organic soil; (2) the proportional increase in DOC flux from drained organic soils
relative to undrained sites; and (3) the conversion factor for the part of DOC converted to CO2 after export from a
site. The third pathway—emissions from (peat) fires on organic soils—assumes that the drained organic soils burn
in a fire, but not any wet organic soils. However, this Inventory currently does not include emissions for this
pathway because data on the combined fire and drained organic soils information are not available at this time;
this may become available in the future with additional analysis.
Non-C02 emissions, according to the Tier 1 method, include methane (CH4), nitrous oxide (N2O), and carbon
monoxide (CO). Emissions associated with peat fires include factors for Cm and CO in addition to CO2, but fire
estimates are assumed to be zero for the current Inventory, as discussed above. Methane emissions generally
associated with anoxic conditions do occur from the drained land surface, but the majority of these emissions
originate from ditches constructed to facilitate drainage at these sites. From this, two separate emission factors
are used, one for emissions from the area of drained soils and a second for emissions from drainage ditch
waterways. Calculations are conducted according to Equation 2.6 and Tables 2.3 and 2.4, which includes the
default fraction of the total area of drained organic soil which is occupied by ditches. Emissions of N2O can be
significant from these drained soils in contrast to the very low emissions from wet organic soils. Calculations are
conducted according to Equation 2.7 and Table 2.5, which provide the estimate as kg N per year.
Methodological calculations were applied to the entire set of estimates for 1990 through 2021. Year-specific data
are not available. Estimates are based on a single year and applied as the annual estimates over the interval.
Uncertainty
Uncertainties are based on the sampling error associated with forest area of drained organic soils and the
uncertainties provided in the Chapter 2 (IPCC 2014) emissions factors (Table 6-23). The estimates and resulting
quantities representing uncertainty are based on the IPCC Approach 1-error propagation. However, probabilistic
sampling of the distributions defined for each emission factor produced a histogram result that contained a mean
and 95 percent confidence interval. The primary reason for this approach was to develop a numerical
representation of uncertainty with the potential for combining with other forest components. The methods and
parameters applied here are identical to previous inventories, but input values were resampled for this Inventory,
which results in minor changes in the number of significant digits in the resulting estimates, relative to past values.
The total non-C02 emissions in 2021 from drained organic soils on Forest Land Remaining Forest Land and Land
6-46 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Converted to Forest Land were estimated to be between 0 and 0.150 MMT CO2 Eq. around a central estimate of
2 0.068 MMT CO2 Eq. at a 95 percent confidence level.
3 Table 6-23: Quantitative Uncertainty Estimates for Non-C02 Emissions on Drained Organic
4 Forest Soils (MMT CO2 Eq. and Percent)3
2021 Emission
Source Estimate Uncertainty Range Relative to Emission Estimate
(MMT C02 Eq.) (MMT CP2 Eq.) (%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
ch4
+
+
+
-69%
+82%
n2o
0.1
+
0.1
-118%
+132%
Total
0.1
+
0.2
-107%
+120%
+ Does not exceed 0.05 MMT C02 Eq.
a Range of flux estimates predicted through a combination of sample-based and IPCC defaults for a 95
percent confidence interval, IPCC Approach 1.
Note: Totals may not sum due to independent rounding.
5 QA/QC and Verification
6 IPCC (2014) guidance cautions of a possibility of double counting some of these emissions. Specifically, the off-site
7 emissions of dissolved organic C from drainage waters may be double counted if soil C stock and change is based
8 on sampling and this C is captured in that sampling. Double counting in this case is unlikely since plots identified as
9 drained were treated separately in this chapter. Additionally, some of the non-CC>2 emissions may be included in
10 either the Wetlands or sections on N2O emissions from managed soils. These paths to double counting emissions
11 are unlikely here because these issues are taken into consideration when developing the estimates and this
12 chapter is the only section directly including such emissions on forest land.
13 Recalculations Discussion
14 The EPA updated global warming potentials (GWP) for calculating CC>2-equivalent emissions of CH4 (from 25 to 28)
15 and N20 (from 298 to 265) to reflect the 100-year GWPs provided in the IPCC Fifth Assessment Report (AR5) (IPCC
16 2013). The previous Inventory used 100-year GWPs provided in the IPCC Fourth Assessment Report (AR4). This
17 update was applied across the entire time series. As a result of this change, there was a minimal decrease in
18 average annual calculated CC>2-equivalent total emissions from drained organic forest soils from 1990 through
19 2020 compared to the previous Inventory. Further discussion on this update and the overall impacts of updating
20 the Inventory GWP values to reflect the AR5 can be found in Chapter 9, Recalculations and Improvements.
21 Planned Improvements
22 Additional data will be compiled to update estimates of forest areas on drained organic soils as new reports and
23 geospatial products become available.
24
Land Use, Land-Use Change, and Forestry 6-47
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
6.3 Land Converted to Forest Land (CRF
Source Category 4A2)
The C stock change estimates for Land Converted to Forest Land that are provided in this Inventory include all
forest land in an inventory year that had been in another land use(s) during the previous 20 years.36 For example,
cropland or grassland converted to forest land during the past 20 years would be reported in this category.
Converted lands are in this category for 20 years as recommended in the 2006IPCC Guidelines (IPCC 2006), after
which they are classified as Forest Land Remaining Forest Land. Estimates of C stock changes from all pools (i.e.,
aboveground and belowground biomass, dead wood, litter and soils), as recommended by IPCC (2006), are
included in the Land Converted to Forest Land category of this Inventory.
Area of Land Converted to Forest in the United States37
Land conversion to and from forests has occurred regularly throughout U.S. history. The 1970s and 1980s saw a
resurgence of federally sponsored forest management programs (e.g., the Forestry Incentive Program) and soil
conservation programs (e.g., the Conservation Reserve Program), which have focused on tree planting, improving
timber management activities, combating soil erosion, and converting marginal cropland to forests. Recent
analyses suggest that net accumulation of forest area continues in areas of the United States, in particular the
northeastern United States (Woodall et al. 2015b). Specifically, the annual conversion of land from other land-use
categories (i.e., Cropland, Grassland, Wetlands, Settlements, and Other Lands) to Forest Land resulted in a fairly
continuous net annual accretion of Forest Land area from over the time series at an average rate of 1.0 million ha
year"1.
Over the 20-year conversion period used in the Land Converted to Forest Land category, the conversion of
cropland to forest land resulted in the largest source of C transfer and uptake, accounting for approximately 39
percent of the uptake annually. Estimated C uptake has remained relatively stable over the time series across all
conversion categories (see Table 6-24). The net flux of C from all forest pool stock changes in 2021 was -98.3 MMT
C02 Eq. (-26.8 MMT C) (Table 6-24 and Table 6-25).
Mineral soil C stocks increased slightly over the time series for Land Converted to Forest Land. The small gains are
associated with Cropland Converted to Forest Land, Settlements Converted to Forest Land, and Other Land
Converted to Forest Land. Much of this conversion is from soils that are more intensively used under annual crop
production or settlement management, or are conversions from other land, which has little to no soil C. In
contrast, Grassland Converted to Forest Land leads to a loss of soil C across the time series, which negates some of
the gain in soil C with the other land-use conversions. Managed Pasture to Forest Land is the most common
conversion. This conversion leads to a loss of soil C because pastures are mostly improved in the United States with
fertilization and/or irrigation, which enhances C input to soils relative to typical forest management activities.
36 The annual NFI data used to compile estimates of carbon transfer and uptake in this section are based on 5- to 10-yr
remeasurements so the exact conversion period was limited to the remeasured data over the time series.
37 The estimates reported in this section only include the 48 conterminous states in the United States. Land use conversions to
forest land in Alaska are currently included in the Forest Land Remaining Forest Land section because currently there is
insufficient data to separate the changes and estimates for Hawaii were not included because there is insufficient NFI data to
support inclusion at this time. Also, it is not possible to separate Forest Land Remaining Forest Land from Land Converted to
Forest Land in Wyoming because of the split annual cycle method used for population estimation, this prevents harmonization
of forest land in Wyoming with the NRI/NLCD method used in section 6.1 Representation of the U.S. Land Base (CRF Category
4.1). See Annex 3.13, Table A-213 for annual differences between the forest area reported in Section 6.1 Representation of the
U.S. Land Base and Section 6.3 Land Converted to Forest Land.
6-48 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Table 6-24: Net CO2 Flux from Forest C Pools in Land Converted to Forest Land by Land Use
2 Change Category (MMT CO2 Eq.)
Land Use/Carbon Pool
1990
2005
2017
2018
2019
2020
2021
Cropland Converted to Forest Land
(38.5)
(38.1)
(37.9)
(37.8)
(37.8)
(37.8)
(37.8)
Aboveground Biomass
(22.2)
(22.0)
(21.9)
(21.9)
(21.9)
(21.9)
(21.9)
Belowground Biomass
(4.3)
(4.3)
(4.2)
(4.2)
(4.2)
(4.2)
(4.2)
Dead Wood
(4.8)
(4.8)
(4.8)
(4.8)
(4.8)
(4.8)
(4.8)
Litter
(6.9)
(6.8)
(6.8)
(6.8)
(6.8)
(6.8)
(6.8)
Mineral Soil
(0.3)
(0.3)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Grassland Converted to Forest Land
(12.2)
(12.2)
(12.3)
(12.3)
(12.3)
(12.3)
(12.3)
Aboveground Biomass
(6.1)
(6.2)
(6.2)
(6.2)
(6.2)
(6.2)
(6.2)
Belowground Biomass
(1.0)
(1.0)
(1.0)
(1.0)
(1.0)
(1.0)
(1.0)
Dead Wood
(1.2)
(1.2)
(1.2)
(1.2)
(1.2)
(1.2)
(1.2)
Litter
(4.1)
(4.1)
(4.1)
(4.1)
(4.1)
(4.1)
(4.1)
Mineral Soil
0.2
0.3
0.3
0.3
0.3
0.3
0.3
Other Land Converted to Forest Land
(9.9)
(10.5)
(10.7)
(10.7)
(10.7)
(10.7)
(10.7)
Aboveground Biomass
(4.7)
(4.7)
(4.8)
(4.8)
(4.8)
(4.8)
(4.8)
Belowground Biomass
(0.8)
(0.8)
(0.8)
(0.8)
(0.8)
(0.8)
(0.8)
Dead Wood
(1.3)
(1.3)
(1.3)
(1.3)
(1.3)
(1.3)
(1.3)
Litter
(2.5)
(2.5)
(2.5)
(2.5)
(2.5)
(2.5)
(2.5)
Mineral Soil
(0.6)
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
Settlements Converted to Forest Land
(34.4)
(34.2)
(34.0)
(34.0)
(34.0)
(34.0)
(34.0)
Aboveground Biomass
(21.0)
(20.9)
(20.7)
(20.7)
(20.7)
(20.7)
(20.7)
Belowground Biomass
(4.0)
(4.0)
(3.9)
(3.9)
(3.9)
(3.9)
(3.9)
Dead Wood
(4.0)
(4.0)
(3.9)
(3.9)
(3.9)
(3.9)
(3.9)
Litter
(5.4)
(5.4)
(5.3)
(5.3)
(5.3)
(5.3)
(5.3)
Mineral Soil
(0.1)
(0.04)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Wetlands Converted to Forest Land
(3.4)
(3.4)
(3.4)
(3.4)
(3.4)
(3.4)
(3.4)
Aboveground Biomass
(1.5)
(1.5)
(1.5)
(1.5)
(1.5)
(1.5)
(1.5)
Belowground Biomass
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
Dead Wood
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
Litter
(1.3)
(1.3)
(1.3)
(1.3)
(1.3)
(1.3)
(1.3)
Mineral Soil
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Total Aboveground Biomass Flux
(55.5)
(55.3)
(55.2)
(55.1)
(55.1)
(55.1)
(55.1)
Total Belowground Biomass Flux
(10.4)
(10.3)
(10.3)
(10.3)
(10.3)
(10.3)
(10.3)
Total Dead Wood Flux
(11.6)
(11.6)
(11.6)
(11.6)
(11.6)
(11.6)
(11.6)
Total Litter Flux
(20.1)
(20.1)
(20.1)
(20.1)
(20.1)
(20.1)
(20.1)
Total Mineral Soil Flux
(0.8)
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
Total Flux
(98.5)
(98.4)
(98.3)
(98.3)
(98.3)
(98.3)
(98.3)
Land Use, Land-Use Change, and Forestry 6-49
-------
Notes: Totals may not sum due to independent rounding. Parentheses indicate net uptake. Forest ecosystem C stock
changes from land conversion in Alaska are currently included in the Forest Land Remaining Forest Land section because
there is insufficient data to separate the changes at this time. Forest ecosystem C stock changes from land conversion do
not include U.S. Territories because managed forest land in U.S. Territories is not currently included in Section 6.1
Representation of the U.S. Land Base. The forest ecosystem C stock changes from land conversion do not include Hawaii
because there is insufficient NFI data to support inclusion at this time. Also, it is not possible to separate Forest Land
Remaining Forest Land from Land Converted to Forest Land in Wyoming because of the split annual cycle method used for
population estimation, this prevents harmonization of forest land in Wyoming with the NRI/NLCD method used in section
6.1 Representation of the U.S. Land Base (CRF Category 4.1). See Annex 3.13, Table A-217 for annual differences between
the forest area reported in Section 6.1 Representation of the U.S. Land Base and Section 6.3 Land Converted to Forest Land.
The forest ecosystem C stock changes from land conversion do not include trees on non-forest land (e.g., agroforestry
systems and settlement areas—see Section 6.10 Settlements Remaining Settlements for estimates of C stock change from
settlement trees). It is not possible to separate emissions from drained organic soils between Forest Land Remaining Forest
Land and Land Converted to Forest Land so estimates for all organic soils are included in Table 6-8 and Table 6-9 of the
Forest Land Remaining Forest Land section of the Inventory.
1 Table 6-25: Net C Flux from Forest C Pools in Land Converted to Forest Land by Land Use
2 Change Category (MMT C)
Land Use/Carbon Pool
1990
2005
2017
2018
2019
2020
2021
Cropland Converted to Forest
Land
(10.8)
(10.8)
(10.3)
(10.3)
(10.3)
(10.3)
(10.3)
Aboveground Biomass
(6.3)
(6.3)
(6.0)
(6.0)
(6.0)
(6.0)
(6.0)
Belowground Biomass
(1.2)
(1.2)
(1.2)
(1.2)
(1.2)
(1.2)
(1.2)
Dead Wood
(1.4)
(1.4)
(1.3)
(1.3)
(1.3)
(1.3)
(1.3)
Litter
(1.9)
(1.9)
(1.8)
(1.8)
(1.8)
(1.8)
(1.8)
Mineral Soil
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Grassland Converted to Forest
Land
(3.1)
(3.2)
(3.4)
(3.4)
(3.4)
(3.4)
(3.4)
Aboveground Biomass
(1.6)
(1.6)
(1.7)
(1.7)
(1.7)
(1.7)
(1.7)
Belowground Biomass
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
Dead Wood
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
Litter
(1.0)
(1.0)
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
Mineral Soil
0.0
0.1
0.1
0.1
0.1
0.1
0.1
Other Land Converted to Forest
Land
(2.7)
(2.9)
(2.9)
(2.9)
(2.9)
(2.9)
(2.9)
Aboveground Biomass
(1.3)
(1.3)
(1.3)
(1.3)
(1.3)
(1.3)
(1.3)
Belowground Biomass
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Dead Wood
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
Litter
(0.7)
(0.7)
(0.7)
(0.7)
(0.7)
(0.7)
(0.7)
Mineral Soil
(0.2)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
Settlements Converted to Forest
Land
(9.3)
(9.3)
(9.3)
(9.3)
(9.3)
(9.3)
(9.3)
Aboveground Biomass
(5.7)
(5.7)
(5.7)
(5.7)
(5.7)
(5.7)
(5.7)
Belowground Biomass
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
Dead Wood
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
Litter
(1.5)
(1.5)
(1.5)
(1.5)
(1.5)
(1.5)
(1.5)
Mineral Soil
+
+
+
+
+
+
+
Wetlands Converted to Forest
Land
(0.9)
(0.9)
(0.9)
(0.9)
(0.9)
(0.9)
(0.9)
Aboveground Biomass
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
Belowground Biomass
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Dead Wood
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Litter
(0.3)
(0.3)
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
Mineral Soil
+
+
+
+
+
+
+
Total Aboveground Biomass Flux
(15.2)
(15.3)
(15.0)
(15.0)
(15.0)
(15.0)
(15.0)
Total Belowground Biomass Flux
(2.9)
(2.9)
(2.8)
(2.8)
(2.8)
(2.8)
(2.8)
6-50 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Total Dead Wood Flux
(3.2)
(3.2)
(3.2)
(3.2)
(3.2)
(3.2)
(3.2)
Total Litter Flux
(5.4)
(5.4)
(5.5)
(5.5)
(5.5)
(5.5)
(5.5)
Total Mineral Soil Flux
(0.2)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
Total Flux
(26.9)
(27.0)
(26.8)
(26.8)
(26.8)
(26.8)
(26.8)
+ Absolute value does not exceed 0.05 MMT C.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net uptake. Forest ecosystem C
stock changes from land conversion in Alaska are currently included in the Forest Land Remaining Forest Land
section because there is not sufficient data to separate the changes at this time. Forest ecosystem C stock
changes from land conversion do not include U.S. Territories because managed forest land in U.S. Territories is
not currently included in Section 6.1 Representation of the U.S. Land Base. The forest ecosystem C stock changes
from land conversion do not include Hawaii because there is not sufficient NFI data to support inclusion at this
time. Also, it is not possible to separate Forest Land Remaining Forest Land from Land Converted to Forest Land in
Wyoming because of the split annual cycle method used for population estimation, this prevents harmonization
of forest land in Wyoming with the NRI/NLCD method used in section 6.1 Representation of the U.S. Land Base
(CRF Category 4.1). See Annex 3.13, Table A-217 for annual differences between the forest area reported in
Section 6.1 Representation of the U.S. Land Base and Section 6.3 Land Converted to Forest Land. The forest
ecosystem C stock changes from land conversion do not include trees on non-forest land (e.g., agroforestry
systems and settlement areas—see Section 6.10 Settlements Remaining Settlements for estimates of C stock
change from settlement trees). It is not possible to separate emissions from drained organic soils between Forest
Land Remaining Forest Land and Land Converted to Forest Land so estimates for organic soils are included in
Table 6-8 and Table 6-9 of the Forest Land Remaining Forest Land section of the Inventory.
Methodology and Time-Series Consistency
The following section includes a description of the methodology used to estimate stock changes in all forest C
pools for Land Converted to Forest Land. National Forest Inventory data and IPCC (2006) defaults for reference C
stocks were used to compile separate estimates for the five C storage pools. Estimates for Aboveground and
Belowground Biomass, Dead Wood and Litter were based on data collected from the extensive array of
permanent, annual NFI plots and associated models (e.g., live tree belowground biomass estimates) in the United
States (USDA Forest Service 2022b, 2022c). Carbon conversion factors were applied at the individual plot and then
appropriately expanded to state population estimates, which are summed to provide the national estimate. To
ensure consistency in the Land Converted to Forest Land category where C stock transfers occur between land-use
categories, all soil estimates are based on methods from Ogle et al. (2003, 2006) and IPCC (2006).
The methods used for estimating carbon stocks and stock changes in the Land Converted to Forest Land are
consistent with those used for Forest Land Remaining Forest Land. For land-use conversion, IPCC (2006) default
biomass C stock values were applied in the year of conversion on individual plots to estimate the carbon stocks
removed due to land-use conversion from Croplands and Grasslands. There is no biomass loss data or IPCC (2006)
defaults to include transfers, losses, or gains of carbon in the year of the conversion for other land use (i.e., Other
Lands, Settlements, Wetlands) conversions to Forest Land so these were incorporated for these conversion
categories. All annual NFI plots included in the public FIA database as of August 2022 were used in this Inventory.
Forest Land conditions were observed on NFI plots at time to and at a subsequent time ti=to+s, where s is the time
step (time measured in years) and is indexed by discrete (e.g., 5 year) forest age classes. The inventory from to was
then projected from ti to 2021. This projection approach requires simulating changes in the age-class distribution
resulting from forest aging and disturbance events and then applying C density estimates for each age class to
obtain population estimates for the nation.
Carbon in Biomass
Live tree C pools include aboveground and belowground (coarse root) biomass of live trees with diameter at breast
height (dbh) of at least 2.54 cm at 1.37 m above the forest floor. Separate estimates were made for above and
belowground biomass components. If inventory plots included data on individual trees, above- and belowground
tree C was based on Woodall et al. (2011a), which is also known as the component ratio method (CRM), and is a
Land Use, Land-Use Change, and Forestry 6-51
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
function of volume, species, and diameter. An additional component of foliage, which was not explicitly included in
Woodall et al. (2011a), was added to each tree following the same CRM method.
Understory vegetation is a minor component of biomass and is defined as all biomass of undergrowth plants in a
forest, including woody shrubs and trees less than 2.54 cm dbh. For the current Inventory, it was assumed that 10
percent of total understory C mass is belowground (Smith et al. 2006). Estimates of C density were based on
information in Birdsey (1996) and biomass estimates from Jenkins et al. (2003). Understory biomass represented
over one percent of C in biomass, but its contribution rarely exceeded 2 percent of the total.
Biomass losses associated with conversion from Grassland and Cropland to Forest Land were assumed to occur in
the year of conversion. To account for these losses, IPCC (2006) defaults for aboveground and belowground
biomass on Grasslands and aboveground biomass on Croplands were subtracted from sequestration in the year of
the conversion. As previously discussed, for all other land use (i.e., Other Lands, Settlements, Wetlands)
conversions to Forest Land no biomass loss data were available, and no IPCC (2006) defaults currently exist to
include transfers, losses, or gains of carbon in the year of the conversion, so none were incorporated for these
conversion categories. As defaults or country-specific data become available for these conversion categories, they
will be incorporated.
Carbon in Dead Organic Matter
Dead organic matter was initially calculated as three separate pools—standing dead trees, downed dead wood,
and litter—with C stocks estimated from sample data or from models. The standing dead tree C pool includes
aboveground and belowground (coarse root) biomass for trees of at least 12.7 cm dbh. Calculations followed the
basic method applied to live trees (Woodall et al. 2011a) with additional modifications to account for decay and
structural loss (Domke et al. 2011; Harmon et al. 2011). Downed dead wood estimates are based on measurement
of a subset of FIA plots for downed dead wood (Domke et al. 2013; Woodall and Monleon 2008; Woodall et al.
2013). Downed dead wood is defined as pieces of dead wood greater than 7.5 cm diameter, at transect
intersection, that are not attached to live or standing dead trees. This includes stumps and roots of harvested
trees. To facilitate the downscaling of downed dead wood C estimates from the state-wide population estimates to
individual plots, downed dead wood models specific to regions and forest types within each region are used. Litter
C is the pool of organic C (also known as duff, humus, and fine woody debris) above the mineral soil and includes
woody fragments with diameters of up to 7.5 cm. A subset of FIA plots are measured for litter C. A modeling
approach, using litter C measurements from FIA plots (Domke et al. 2016) was used to estimate litter C for every
FIA plot used in the estimation framework. Dead organic matter C stock estimates are included for all land-use
conversions to Forest Land.
Mineral Soil Carbon Stock Changes
A Tier 2 method is applied to estimate mineral soil C stock changes for Land Converted to Forest Land (Ogle et al.
2003, 2006; IPCC 2006). For this method, land is stratified by climate, soil types, land use, and land management
activity, and then assigned reference carbon levels and factors for the forest land and the previous land use. The
difference between the stocks is reported as the stock change under the assumption that the change occurs over
20 years. Reference C stocks have been estimated from data in the National Soil Survey Characterization Database
(USDA-NRCS 1997), and U.S.-specific stock change factors have been derived from published literature (Ogle et al.
2003, 2006). Land use and land-use change patterns are determined from a combination of the Forest Inventory
and Analysis Dataset (FIA), the 2015 National Resources Inventory (NRI) (USDA-NRCS 2018), and National Land
Cover Dataset (NLCD) (Yang et al. 2018). See Annex 3.12 (Methodology for Estimating N2O Emissions, CFU
Emissions and Soil Organic C Stock Changes from Agricultural Soil Management) for more information about this
method. Note that soil C in this Inventory is reported to a depth of 100 cm in the Forest Land Remaining Forest
Land category (Domke et al. 2017) while other land-use categories report soil C to a depth of 30 cm. However, to
ensure consistency in the Land Converted to Forest Land category where C stock transfers occur between land-use
categories, soil C estimates were based on a 30 cm depth using methods from Ogle et al. (2003, 2006) and IPCC
(2006), as described in Annex 3.12.
6-52 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
In order to ensure time-series consistency, the same methods are applied from 1990 to 2015 so that changes
reflect anthropogenic activity and not methodological adjustments. Mineral soil organic C stock changes from 2016
to 2021 are estimated using a linear extrapolation method described in Box 6-4 of the Methodology section in
Cropland Remaining Cropland. The extrapolation is based on a linear regression model with moving-average
(ARMA) errors using the 1990 to 2015 emissions data and is a standard data splicing method for estimating
emissions at the end of a time series if activity data are not available (IPCC 2006). The Tier 2 method described
previously will be applied to recalculate the 2016 to 2021 emissions in a future Inventory.
Uncertainty
A quantitative uncertainty analysis placed bounds on the flux estimates for Land Converted to Forest Land through
a combination of sample-based and model-based approaches to uncertainty for forest ecosystem CO2 Eq. flux
(IPCC Approach 1). Uncertainty estimates for forest pool C stock changes were developed using the same
methodologies as described in the Forest Land Remaining Forest Land section for aboveground and belowground
biomass, dead wood, and litter. The exception was when IPCC default estimates were used for reference C stocks
in certain conversion categories (i.e., Cropland Converted to Forest Land and Grassland Converted to Forest Land).
In those cases, the uncertainties associated with the IPCC (2006) defaults were included in the uncertainty
calculations. IPCC Approach 2 was used for mineral soils and is described in the Cropland Remaining Cropland
section.
Uncertainty estimates are presented in Table 6-26 for each land conversion category and C pool. Uncertainty
estimates were obtained using a combination of sample-based and model-based approaches for all non-soil C
pools (IPCC Approach 1) and a Monte Carlo approach (IPCC Approach 2) was used for mineral soil. Uncertainty
estimates were combined using the error propagation model (IPCC Approach 1). The combined uncertainty for all
C stocks in Land Converted to Forest Land ranged from 11 percent below to 11 percent above the 2021C stock
change estimate of-98.3 MMT CO2 Eq.
Table 6-26: Quantitative Uncertainty Estimates for Forest C Pool Stock Changes (MMT CO2
Eq. per Year) in 2021 from Land Converted to Forest Land by Land Use Change
zuzi MUX Uncertainty Range Relative to Flux Range3
Land Use/Carbon Pool Estimate
(MMT C02 Eq.) (MMT CP2 Eq.) (%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Cropland Converted to Forest Land
(37.8)
(46.5)
(29.2)
-23%
23%
Aboveground Biomass
(21.9)
(30.3)
(13.5)
-38%
38%
Belowground Biomass
(4.2)
(5.3)
(3.2)
-25%
25%
Dead Wood
(4.8)
(6.0)
(3.5)
-26%
26%
Litter
(6.8)
(7.8)
(5.7)
-16%
16%
Mineral Soils
(0.2)
(0.5)
0.1
-135%
135%
Grassland Converted to Forest Land
(12.3)
(14.8)
(9.9)
-20%
20%
Aboveground Biomass
(6.2)
(7.6)
(4.9)
-22%
22%
Belowground Biomass
(1.0)
(1.3)
(0.7)
-28%
28%
Dead Wood
(1.2)
(1.4)
(1.1)
-12%
12%
Litter
(4.1)
(4.7)
(3.6)
-13%
13%
Mineral Soils
0.3
(0.1)
0.6
-137%
137%
Other Lands Converted to Forest Land
(10.7)
(13.0)
(8.3)
-22%
22%
Aboveground Biomass
(4.8)
(6.9)
(2.7)
-44%
44%
Belowground Biomass
(0.8)
(1.3)
(0.4)
-51%
51%
Dead Wood
(1.3)
(1.9)
(0.8)
-42%
42%
Litter
(2.5)
(3.2)
(1.9)
-25%
25%
Mineral Soils
(1.1)
(1.9)
(0.4)
-68%
68%
Settlements Converted to Forest Land
(34.0)
(40.5)
(27.5)
-19%
19%
Aboveground Biomass
(20.7)
(26.9)
(14.5)
-30%
30%
Land Use, Land-Use Change, and Forestry 6-53
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Belowground Biomass
(3.9)
(5.3)
(2.6)
-33%
33%
Dead Wood
(3.9)
(5.1)
(2.8)
-29%
29%
Litter
(5.3)
(6.2)
(4.4)
-17%
17%
Mineral Soil
(0.1)
(0.1)
(0.0)
-47%
47%
Wetlands Converted to Forest Land
(3.4)
(3.6)
(3.3)
-5%
5%
Aboveground Biomass
(1.5)
(1.7)
(1.4)
-9%
9%
Belowground Biomass
(0.3)
(0.3)
(0.3)
-11%
11%
Dead Wood
(0.4)
(0.4)
(0.3)
-12%
12%
Litter
(1.3)
(1.3)
(1.2)
-5%
5%
Mineral Soils
0.0
0.0
0.0
NA
NA
Total: Aboveground Biomass
(55.1)
(65.9)
(44.4)
-19%
19%
Total: Belowground Biomass
(10.3)
(12.0)
(8.5)
-17%
17%
Total: Dead Wood
(11.6)
(13.4)
(9.8)
-15%
15%
Total: Litter
(20.1)
(21.7)
(18.5)
-8%
8%
Total: Mineral Soils
(1.1)
(1.7)
(0.6)
-51%
51%
Total: Lands Converted to Forest Lands
(98.3)
(109.4)
(87.1)
-11%
11%
+ Absolute value does not exceed 0.05 MMT C02 Eq.
NA (Not Applicable)
a Range of flux estimate for 95 percent confidence interval.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net uptake. It is not possible to separate
emissions from drained organic soils between Forest Land Remaining Forest Land and Land Converted to Forest Land so
estimates for organic soils are included in Table 6-8 and Table 6-9 of the Forest Land Remaining Forest Land section of the
Inventory.
QA/QC and Verification
See QA/QC and Verification sections under Forest Land Remaining Forest Land and for mineral soil estimates
Cropland Remaining Cropland.
Recalculations Discussion
The approach for estimating carbon stock changes in Land Converted to Forest Land is consistent with the
methods used for Forest Land Remaining Forest Land and is described in Annex 3.13. The Land Converted to Forest
Land estimates in this Inventory are based on the land-use change information in the annual NFI. All conversions
are based on empirical estimates compiled using plot remeasurements from the NFI, IPCC (2006) default biomass C
stocks removed from Croplands and Grasslands in the year of conversion on individual plots and the Tier 2 method
for estimating mineral soil C stock changes (Ogle et al. 2003, 2006; IPCC 2006). All annual NFI plots included in the
public FIA database as of August 2022 were used in this Inventory. This is the fourth year that remeasurement data
from the annual NFI were available throughout the conterminous United States (with the exception of Wyoming)
to estimate land-use conversion. The availability of remeasurement data from the annual NFI allowed for
consistent plot-level estimation of C stocks and stock changes for Forest Land Remaining Forest Land and the Land
Converted to Forest Land categories. Estimates in the previous Inventory were based on state-level carbon density
estimates and a combination of NRI data and NFI data in the eastern United States. The refined analysis in this
Inventory resulted in changes in the Land Converted to Forest Land categories. Overall, the Land Converted to
Forest Land C stock changes decreased by approximately 1 percent in 2020 between the previous Inventory and
the current Inventory (Table 6-27). This decrease is directly attributed to the incorporation of annual NFI data into
the compilation system.
6-54 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Table 6-27: Recalculations of the Net C Flux from Forest C Pools in Land Converted to Forest
Land by Land Use Change Category (MMT C)
Conversion category
2020 Estimate,
2020 Estimate,
2021 Estimate,
and Carbon pool (MMT C)
Previous Inventory
Current Inventory
Current Inventory
Cropland Converted to Forest Land
(10.8)
(10.3)
(10.3)
Aboveground Biomass
(6.3)
(6.0)
(6.0)
Belowground Biomass
(1.2)
(1.2)
(1.2)
Dead Wood
(1.4)
(1.3)
(1.3)
Litter
(1.9)
(1.8)
(1.8)
Mineral soil
(0.1)
(0.1)
(0.1)
Grassland Converted to Forest Land
(3.2)
(3.4)
(3.4)
Aboveground Biomass
(1.7)
(1.7)
(1.7)
Belowground Biomass
(0.3)
(0.3)
(0.3)
Dead Wood
(0.3)
(0.3)
(0.3)
Litter
(1.1)
(1.1)
(1.1)
Mineral soil
0.1
0.1
0.1
Other Land Converted to Forest Land
(3.0)
(2.9)
(2.9)
Aboveground Biomass
(1.3)
(1.3)
(1.3)
Belowground Biomass
(0.2)
(0.2)
(0.2)
Dead Wood
(0.4)
(0.4)
(0.4)
Litter
(0.7)
(0.7)
(0.7)
Mineral soil
(0.3)
(0.3)
(0.3)
Settlements Converted to Forest Land
(9.3)
(9.3)
(9.3)
Aboveground Biomass
(5.7)
(5.7)
(5.7)
Belowground Biomass
(1.1)
(1.1)
(1.1)
Dead Wood
(1.1)
(1.1)
(1.1)
Litter
(1.5)
(1.5)
(1.5)
Mineral soil
(0.0)
(0.0)
(0.0)
Wetlands Converted to Forest Land
(0.9)
(0.9)
(0.9)
Aboveground Biomass
(0.4)
(0.4)
(0.4)
Belowground Biomass
(0.1)
(0.1)
(0.1)
Dead Wood
(0.1)
(0.1)
(0.1)
Litter
(0.3)
(0.4)
(0.4)
Mineral soil
0.0
0.0
0.0
Total Aboveground Biomass Flux
(15.3)
(15.0)
(15.0)
Total Belowground Biomass Flux
(2.9)
(2.8)
(2.8)
Total Dead Wood Flux
(3.2)
(3.2)
(3.2)
Total Litter Flux
(5.4)
(5.5)
(5.5)
Total SOC (mineral) Flux
(0.3)
(0.3)
(0.3)
Total Flux
(27.1)
(26.8)
(26.8)
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Planned Improvements
There are many improvements necessary to improve the estimation of carbon stock changes associated with land-
use conversion to forest land over the entire time series. First, soil C has historically been reported to a depth of
100 cm in the Forest Land Remaining Forest Land category (Domke et al. 2017) while other land-use categories
(e.g., Grasslands and Croplands) report soil carbon to a depth of 30 cm. To ensure greater consistency in the Land
Converted to Forest Land category where C stock transfers occur between land-use categories, all mineral soil
estimates in the Land Converted to Forest Land category in this Inventory are based on methods from Ogle et al.
(2003, 2006) and IPCC (2006). Methods have recently been developed (Domke et al. 2017) to estimate soil C to
depths of 20, 30, and 100 cm in the Forest Land category using in situ measurements from the Forest Inventory
and Analysis program within the USDA Forest Service and the International Soil Carbon Network. In subsequent
Inventories, a common reporting depth will be defined for all land-use conversion categories and Domke et al.
Land Use, Land-Use Change, and Forestry 6-55
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
(2017) will be used in the Forest Land Remaining Forest Land and Land Converted to Forest Land categories to
ensure consistent reporting across all forest land. Second, there will be improved methods and models to
characterize standing live and dead tree carbon in the next Inventory. Third, due to the 5 to 10-year
remeasurement periods within the FIA program and limited land-use change information available over the entire
time series, estimates presented in this section may not reflect the entire 20-year conversion history. Work is
underway to integrate the dense time series of remotely sensed data into a new estimation system, which will
facilitate land conversion estimation over the entire time series.
6.4 Cropland Remaining Cropland (CRF
Category 4B1)
Carbon (C) in cropland ecosystems occurs in biomass, dead organic matter, and soils. However, C storage in
cropland biomass and dead organic matter is relatively ephemeral and does not need to be reported according to
the IPCC (2006), with the exception of C 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,
C is found in organic and inorganic forms of C, but soil organic C is the main source and sink for atmospheric CO2 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.38
Well-drained mineral soils typically contain from 1 to 6 percent organic C by weight, whereas mineral soils with
high water tables for substantial periods of a year may contain significantly more C (NRCS 1999). Conversion of
mineral soils from their native state to agricultural land uses can cause up to half of the soil organic C to be lost to
the atmosphere due to enhanced microbial decomposition. The rate and ultimate magnitude of C loss depends on
subsequent management practices, climate and soil type (Ogle et al. 2005). Agricultural practices, such as clearing,
drainage, tillage, planting, grazing, crop residue management, fertilization, application of biosolids (i.e., treated
sewage sludge) and flooding, can modify both organic matter inputs and decomposition, and thereby result in a
net C stock change (Paustian et al. 1997a; Lai 1998; Conant et al. 2001; Ogle et al. 2005; Griscom et al. 2017; Ogle
et al. 2019). Eventually, the soil can reach a new equilibrium that reflects a balance between C inputs (e.g.,
decayed plant matter, roots, and organic amendments such as manure and crop residues) and C loss through
microbial decomposition of organic matter (Paustian et al. 1997b).
Organic soils, also referred to as histosols, include all soils with more than 12 to 20 percent organic C by weight,
depending on clay content (NRCS 1999; Brady and Weil 1999). The organic layer of these soils can be very deep
(i.e., several meters), and form under inundated conditions that results in minimal decomposition of plant
residues. When organic soils are prepared for crop production, they are drained and tilled, leading to aeration of
the soil that accelerates both the decomposition rate and CO2 emissions.39 Due to the depth and richness of the
organic layers, C loss from drained organic soils can continue over long periods of time, which varies depending on
climate and composition (i.e., decomposability) of the organic matter (Armentano and Menges 1986). Due to
deeper drainage and more intensive management practices, the use of organic soils for annual crop production
leads to higher C loss rates than drainage of organic soils in grassland or forests (IPCC 2006).
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 United States Department of Agriculture
38 Carbon dioxide emissions associated with liming and urea application are also estimated but are included in the Liming and
Urea Fertilization sections of the Agriculture chapter of the Inventory.
39 N20 emissions from drained organic soils are included in the Agricultural Soil Management section of the Agriculture chapter
of the Inventory.
6-56 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
(USDA) National Resources Inventory (NRI) for non-federal lands (USDA-NRCS 2018a) and the National Land Cover
Dataset for federal lands (Yang et al. 2018; Homer et al. 2007; Fry et al. 2011; Homer et al. 2015). 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), in addition to cropland that has been enrolled in the Conservation Reserve Program (CRP)40 (i.e.,
considered set-aside cropland).
There are several discrepancies between the current land representation (See Section 6.1) and the area data that
have been used in the inventory for Cropland Remaining Cropland. First, the current land representation is based
on the latest NRI dataset, which includes data through 2017, but these data have not been incorporated into the
Cropland Remaining Cropland Inventory. Second, cropland in Alaska is not included in the Inventory, and third,
some miscellaneous croplands are also not included in the Inventory due to limited understanding of greenhouse
gas emissions from these management systems (e.g., aquaculture). These differences lead to discrepancies
between the managed area in Cropland Remaining Cropland and the cropland area included in the Inventory
analysis (Table 6-31). Improvements are underway to incorporate the latest NRI dataset, croplands in Alaska and
miscellaneous croplands as part of future C inventories (See Planned Improvements Section).
Land use and land management of mineral soils are the largest contributor to total net C stock change, especially
in the early part of the time series (see Table 6-28 and Table 6-29). In 2021, mineral soils are estimated to
sequester 51.8 MMT CO2 Eq. from the atmosphere (14.1 MMT C). This rate of C storage in mineral soils represents
about a 11 percent decrease in the rate since the initial reporting year of 1990. Carbon dioxide emissions from
organic soils are 32.9 MMT CO2 Eq. (9.0 MMT C) in 2021, which is a 6 percent decrease compared to 1990. In total,
United States agricultural soils in Cropland Remaining Cropland sequestered approximately 18.9 MMT CO2 Eq. (5.2
MMT C) in 2021.
Table 6-28: Net CO2 Flux from Soil C Stock Changes in Cropland Remaining Cropland (MMT
COz Eq.)
Soil Type 1990 2005 2017 2018 2019 2020 2021
Mineral Soils (58.2) (62.4) (55.1) (49.4) (47.4) (56.2) (51.8)
Organic Soils 35;0 3^4 32.8 32.8 32.9 32.9 32.9
Total Net Flux (23.2) (29.0) (22.3) (16.6) (14.5) (23.3) (18.9)
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Table 6-29: Net CO2 Flux from Soil C Stock Changes in Cropland Remaining Cropland (MMT
C)
Soil Type
1990
2005
2017
2018
2019
2020
2021
Mineral Soils
(15.9)
(17.0)
(15.0)
(13.5)
(12.9)
(15.3)
(14.1)
Organic Soils
9.5
9.1
8.9
8.9
9.0
9.0
9.0
Total Net Flux
(6.3)
(7.9)
(6.1)
(4.5)
(4.0)
(6.4)
(5.2)
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Soil organic C stocks increase in Cropland Remaining Cropland largely due to conservation tillage (i.e., reduced- and
no-till practices), land set-aside from production in the Conservation Reserve Program, annual crop production
with hay or pasture in rotations, and manure amendments. However, there is a decline in the net amount of C
sequestration (i.e., 2021 is 18 percent less than 1990 for mineral and organic soils), and this decline is due to lower
sequestration rates in set-aside lands, less impact of manure amendments and annual crop production with hay
and pasture in rotation. Soil organic C losses from drainage of organic soils are relatively stable across the time
40 The Conservation Reserve Program (CRP) is a land conservation program administered by the Farm Service Agency (FSA). In
exchange for a yearly rental payment, farmers enrolled in the program agree to remove environmentally sensitive land from
agricultural production and plant species that will improve environmental health and quality. Contracts for land enrolled in CRP
are 10 to 15 years in length. The long-term goal of the program is to re-establish valuable land cover to help improve water
quality, prevent soil erosion, and reduce loss of wildlife habitat.
Land Use, Land-Use Change, and Forestry 6-57
-------
1 series with a small decline associated with the land base declining for Cropland Remaining Cropland on organic
2 soils since 1990.
3 The spatial variability in the 2015 annual soil organic C stock changes41 are displayed in Figure 6-6 and Figure 6-7
4 for mineral and organic soils, respectively. Isolated areas with high rates of C accumulation occur throughout the
5 agricultural land base in the United States, but there are more concentrated areas. In particular, higher rates of net
6 C accumulation in mineral soils occur in the Corn Belt region, which is the region with the largest amounts of
7 conservation tillage, along with moderate rates of CRP enrollment. The regions with the highest rates of emissions
8 from drainage of organic soils occur in the Southeastern Coastal Region (particularly Florida), upper Midwest and
9 Northeast surrounding the Great Lakes, and isolated areas along the Pacific Coast (particularly California), which
10 coincides with the largest concentrations of organic soils in the United States that are used for agricultural
11 production.
12 Figure 6-6; Total Net Annual Soil C Stock Changes for Mineral Soils under Agricultural
13 Management within States, 2015, Cropland Remaining Cropland
~ -1 to 1
14
15 Note: Only national-scale soil organic C stock changes are estimated for 2016 to 2021 in the current Inventory
16 using a surrogate data method, and therefore the fine-scale emission patterns in this map are based on
17 inventory data from 2015. Negative values represent a net increase in soil organic C stocks, and positive values
18 represent a net decrease in soil organic C stocks.
41 Only national-scale emissions are estimated for 2016 to 2021 in this Inventory using the surrogate data method, and
therefore the fine-scale emission patterns in this map are based on inventory data from 2015.
6-58 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Figure 6-7: Total Net Annual Soil C Stock Changes for Organic Soils under Agricultural
2 Management within States, 2015, Cropland Remaining Cropland
¦ >40
3
4 Note: Only national-scale soil organic C stock changes are estimated for 2016 to 2021 in the current Inventory
5 using a surrogate data method, and therefore the fine-scale emission patterns in this map are based on
6 inventory data from 2015.
7 Methodology and Time-Series Consistency
8 The following section includes a description of the methodology used to estimate changes in soil organic C stocks
9 for Cropland Remaining Cropland, including (1) agricultural land use and management activities on mineral soils;
10 and (2) agricultural land use and management activities on organic soils. Carbon dioxide emissions and removals42
11 due to changes in mineral soil organic C stocks are estimated using a Tier 3 method for the majority of annual
12 crops (Ogle et al. 2010). A Tier 2 IPCC method is used for the remaining crops not included in the Tier 3 method
13 (see list of crops in the Mineral Soil Carbon Stock Changes section below) (Ogle et al. 2003, 2006). In addition, a
14 Tier 2 method is used for very gravelly, cobbly, or shaley soils (i.e., classified as soils that have greater than 35
15 percent of soil volume comprised of gravel, cobbles, or shale, regardless of crop). Emissions from organic soils are
16 estimated using a Tier 2 IPCC method, While a combination of Tier 2 and 3 methods are used to estimate C stock
17 changes across most of the time series, a surrogate data method has been applied to estimate stock changes in the
18 last few years of the Inventory. Stock change estimates based on surrogate data will be recalculated in a future
19 Inventory report using the Tier 2 and 3 methods when data become available.
20 Soil organic C stock changes on non-federal lands are estimated for Cropland Remaining Cropland (as well as
21 agricultural land falling into the IPCC categories Land Converted to Cropland, Grassland Remaining Grassland, and
22 Land Converted to Grassland) according to land use histories recorded in the USDA NRI survey (USDA-NRCS 2018a).
23 The NRI is a statistically-based sample of all non-federal land, and includes approximately 489,178 survey locations
24 in agricultural land for the conterminous United States and Hawaii. Each survey location is associated with an
42 Removals occur through uptake of C02 into crop and forage biomass that is later incorporated into soil C pools.
Land Use, Land-Use Change, and Forestry 6-59
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
"expansion factor" that allows scaling of C stock changes from NRI survey locations to the entire country (i.e., each
expansion factor represents the amount of area that is expected to have the same land use/management history
as the sample point). Land use and some management information (e.g., crop type, soil attributes, and irrigation)
are collected for each NRI point on a 5-year cycle beginning from 1982 through 1997. For cropland, data has been
collected for 4 out of 5 years during each survey cycle (i.e., 1979 through 1982,1984 through 1987,1989 through
1992, and 1994 through 1997). In 1998, the NRI program began collecting annual data, and the annual data are
currently available through 2017, however this Inventory uses the previous NRI with annual data through 2015
(USDA-NRCS 2018a). NRI survey locations are classified as Cropland Remaining Cropland in a given year between
1990 and 2015 if the land use has been cropland for a continuous time period of at least 20 years. NRI survey
locations are classified according to land use histories starting in 1979, and consequently the classifications are
based on less than 20 years from 1990 to 1998. This may have led to an overestimation of Cropland Remaining
Cropland in the early part of the time series to the extent that some areas are converted to cropland between
1971 and 1978.
Mineral Soil Carbon Stock Changes
An IPCC Tier 3 model-based approach (Ogle et al. 2010) is applied to estimate organic C stock changes for mineral
soils on the majority of land that is used to produce annual crops and forage crops that are harvested and used as
feed (e.g., hay and silage) in the United States. These crops include alfalfa hay, barley, corn, cotton, grass hay,
grass-clover hay, oats, peanuts, potatoes, rice, sorghum, soybeans, sugar beets, sunflowers, tobacco and wheat,
but is not applied to estimate organic C stock changes from other crops or rotations with other crops. The model-
based approach uses the DayCent biogeochemical model (Parton et al. 1998; Del Grosso et al. 2001, 2011) to
estimate soil organic C stock changes, soil nitrous oxide (N2O) emissions from agricultural soil management, and
methane (CH4) emissions from rice cultivation. Carbon and N dynamics are linked in plant-soil systems through the
biogeochemical processes of microbial decomposition and plant production (McGill and Cole 1981). Coupling the
two source categories (i.e., agricultural soil C and N2O) in a single inventory analysis ensures that there is a
consistent treatment of the processes and interactions between C and N cycling in soils.
The remaining crops on mineral soils are estimated using an IPCC Tier 2 method (Ogle et al. 2003), including some
vegetables, perennial/horticultural crops, and crops that are rotated with these crops. The Tier 2 method is also
used for very gravelly, cobbly, or shaley soils (greater than 35 percent by volume), and soil organic C stock changes
on federal croplands. Mineral soil organic C stocks are estimated using a Tier 2 method for these areas because the
DayCent model, which is used for the Tier 3 method, has not been fully tested for estimating C stock changes
associated with these crops and rotations, as well as cobbly, gravelly, or shaley soils. In addition, there is
insufficient information to simulate croplands on federal lands using DayCent.
A surrogate data method is used to estimate soil organic C stock changes from 2016 to 2021 at the national scale
for land areas included in the Tier 2 and Tier 3 methods. Specifically, linear regression models with autoregressive
moving-average (ARMA) errors (Brockwell and Davis 2016) are used to estimate the relationship between
surrogate data and the 1990 to 2015 stock change data that are derived using the Tier 2 and 3 methods. Surrogate
data for these regression models include corn and soybean yields from USDA-NASS statistics,43 and weather data
from the PRISM Climate Group (PRISM 2018). See Box 6-4 for more information about the surrogate data method.
Stock change estimates for 2016 to 2021 will be recalculated in future Inventories with an updated time series of
activity data.
43 See https://quickstats.nass.usda.gov/.
6-60 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Box 6-4: Surrogate Data Method
Time series extension is needed because there are typically gaps at the end of the time series. This is mainly
because the NRI, which provides critical data for estimating greenhouse gas emissions and removals, does not
release new activity data every year.
A surrogate data method has been used to impute missing emissions at the end of the time series for soil
organic C stock changes in Cropland Remaining Cropland, Land Converted to Cropland, Grassland Remaining
Grassland, and Land Converted to Grassland. A linear regression model with autoregressive moving-average
(ARMA) errors (Brockwell and Davis 2016) is used to estimate the relationship between the surrogate data and
the modeled 1990 to 2015 emissions data that has been compiled using the inventory methods described in this
section. The model to extend the time series is given by
Y = xp + £,
where Y is the response variable (e.g., soil organic carbon), xp contains specific surrogate data depending on the
response variable, and £ is the remaining unexplained error. Models with a variety of surrogate data were
tested, including commodity statistics, weather data, or other relevant information. Parameters are estimated
from the emissions data for 1990 to 2015 using standard statistical techniques, and these estimates are used to
predict the missing emissions data for 2016 to 2021.
A critical issue with application of splicing methods is to adequately account for the additional uncertainty
introduced by predicting emissions rather than compiling the full inventory. Consequently, uncertainty will
increase for years with imputed estimates based on the splicing methods, compared to those years in which the
full inventory is compiled. This added uncertainty is quantified within the model framework using a Monte Carlo
approach. The approach requires estimating parameters for results in each iteration of the Monte Carlo analysis
for the full inventory (i.e., the surrogate data model is refit with the emissions estimated in each Monte Carlo
iteration from the full inventory analysis with data from 1990 to 2015), estimating emissions from each model
and deriving confidence intervals combining uncertainty across all iterations. This approach propagates
uncertainties through the calculations from the original inventory and the surrogate data method. Furthermore,
the 95 percent confidence intervals are estimated using the 3 sigma rules assuming a unimodal density
(Pukelsheim 1994).
Tier 3 Approach. Mineral soil organic C stocks and stock changes are estimated to a 30 cm depth using the
DayCent biogeochemical44 model (Parton et al. 1998; Del Grosso et al. 2001, 2011), which simulates cycling of C, N,
and other nutrients in cropland, grassland, forest, and savanna ecosystems. The DayCent model utilizes the soil C
modeling framework developed in the Century model (Parton et al. 1987,1988,1994; Metherell et al. 1993), but
has been refined to simulate dynamics at a daily time-step. Input data on land use and management are specified
at a daily resolution and include land-use type, crop/forage type, and management activities (e.g., planting,
harvesting, fertilization, manure amendments, tillage, irrigation, cover crops, and grazing; more information is
provided below). The model simulates net primary productivity (NPP) using the NASA-CASA production algorithm
MODIS Enhanced Vegetation Index (EVI) products, MOD13Q1 and MYD13Q1, for most croplands45 (Potter et al.
1993, 2007). The model simulates soil temperature and water dynamics, using daily weather data from a 4-
kilometer gridded product developed by the PRISM Climate Group (2018), and soil attributes from the Soil Survey
44 Biogeochemical cycles are the flow of chemical elements and compounds between living organisms and the physical
environment.
45 NPP is estimated with the NASA-CASA algorithm for most of the cropland that is used to produce major commodity crops in
the central United States from 2000 to 2015. Other regions and years prior to 2000 are simulated with a method that
incorporates water, temperature and moisture stress on crop production (see Metherell et al. 1993), but does not incorporate
the additional information about crop condition provided with remote sensing data.
Land Use, Land-Use Change, and Forestry 6-61
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Geographic Database (SSURGO) (Soil Survey Staff 2019). This method is more accurate than the Tier 1 and 2
approaches provided by the IPCC (2006) because the simulation model treats changes as continuous over time as
opposed to the simplified discrete changes represented in the default method (see Box 6-5 for additional
information).
Box 6-5: Tier 3 Approach for Soil C Stocks Compared to Tier 1 or 2 Approaches
A Tier 3 model-based approach is used to estimate soil organic C stock changes for the majority of agricultural
land with mineral soils. This approach results in a more complete and accurate estimation of soil organic C stock
changes and entails several fundamental differences from the IPCC Tier 1 or 2 methods, as described below.
1) The IPCC Tier 1 and 2 methods are simplified approaches for estimating soil organic C stock changes
and classify land areas into discrete categories based on highly aggregated information about climate
(six regions), soil (seven types), and management (eleven management systems) in the United States.
In contrast, the Tier 3 model incorporates the same variables (i.e., climate, soils, and management
systems) with considerably more detail both temporally and spatially, and captures multi-dimensional
interactions through the more complex model structure.
2) The IPCC Tier 1 and 2 methods have a coarser spatial resolution in which data are aggregated to soil
types in climate regions, of which there about 30 combinations in the United States. In contrast, the
Tier 3 model simulates soil C dynamics at about 350,000 individual NRI survey locations in crop fields
and grazing lands.
The IPCC Tier 1 and 2 methods use a simplified approach for estimating changes in C stocks that assumes a step-
change from one equilibrium level of the C stock to another equilibrium level. In contrast, the Tier 3 approach
simulates a continuum of C stock changes that may reach a new equilibrium over an extended period of time
depending on the environmental conditions (i.e., a new equilibrium often requires hundreds to thousands of
years to reach). More specifically, the DayCent model, which is used in the United States Inventory, simulates
soil C dynamics (and CO2 emissions and uptake) on a daily time step based on C emissions and removals from
plant production and decomposition processes. These changes in soil organic C stocks are influenced by
multiple factors that affect primary production and decomposition, including changes in land use and
management, weather variability and secondary feedbacks between management activities, climate, and soils.
Historical land-use patterns and irrigation histories are simulated with DayCent based on the 2015 USDA NRI
survey (USDA-NRCS 2018a). Additional sources of activity data are used to supplement the activity data from the
NRI. The USDA-NRCS Conservation Effects and Assessment Project (CEAP) provides data on a variety of cropland
management activities, and is used to inform the inventory analysis about tillage practices, mineral fertilization,
manure amendments, cover cropping management, as well as planting and harvest dates (USDA-NRCS 2018b;
USDA-NRCS 2012). CEAP data are collected at a subset of NRI survey locations, and currently provide management
information from approximately 2002 to 2006. These data are combined with other datasets in an imputation
analysis that extend the time series from 1990 to 2015. This imputation analysis is comprised of three steps: a)
determine the trends in management activity across the time series by combining information across several
datasets (discussed below), b) use an artificial neural network to determine the likely management practice at a
given NRI survey location (Cheng and Titterington 1994), and c) assign management practices from the CEAP
survey to the specific NRI locations using predictive mean matching methods that is adapted to reflect the trending
information (Little 1988, van Buuren 2012). The artificial neural network is a machine learning method that
approximates nonlinear functions of inputs and searches through a very large class of models to impute an initial
value for management practices at specific NRI survey locations. The predictive mean matching method identifies
the most similar management activity recorded in the CEAP survey that matches the prediction from the artificial
neural network. Predictive mean matching ensures that imputed management activities are realistic for each NRI
survey location, and not odd or physically unrealizable results that could be generated by the artificial neural
network. There are six complete imputations of the management activity data using these methods.
6-62 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
To determine trends in mineral fertilization and manure amendments from 1979 to 2015, CEAP data are combined
with information on fertilizer use and rates by crop type for different regions of the United States from the USDA
Economic Research Service. The data collection program was known as the Cropping Practices Surveys through
1995 (USDA-ERS 1997), and is now part of a data collection program known as the Agricultural Resource
Management Surveys (ARMS) (USDA-ERS 2018). Additional data on fertilization practices are compiled through
other sources particularly the National Agricultural Statistics Service (USDA-NASS 1992,1999, 2004). The donor
survey data from CEAP contain both mineral fertilizer rates and manure amendment rates, so that the selection of
a donor via predictive mean matching yields the joint imputation of both rates. This approach captures the
relationship between mineral fertilization and manure amendment practices for U.S. croplands based directly on
the observed patterns in the CEAP survey data.
To determine the trends in tillage management from 1979 to 2015, CEAP data are combined with Conservation
Technology Information Center data between 1989 and 2004 (CTIC 2004) and USDA-ERS Agriculture Resource
Management Surveys (ARMS) data from 2002 to 2015 (Claasen et al. 2018). CTIC data are adjusted for long-term
adoption of no-till agriculture (Towery 2001). It is assumed that the majority of agricultural lands are managed
with full tillage prior to 1985. For cover crops, CEAP data are combined with information from 2011 to 2016 in the
USDA Census of Agriculture (USDA-NASS 2012, 2017). It is assumed that cover cropping was minimal prior to 1990
and the rates increased linearly over the decade to the levels of cover crop management derived from the CEAP
survey.
Uncertainty in the C stock estimates from DayCent associated with management activity includes input uncertainty
due to missing management data in the NRI survey, which is imputed from other sources as discussed above;
model uncertainty due to incomplete specification of C and N dynamics in the DayCent model algorithms and
associated parameterization; and sampling uncertainty associated with the statistical design of the NRI survey. To
assess input uncertainty, the C and N dynamics at each NRI survey location are simulated six times using the
imputation product and other model driver data. Uncertainty in parameterization and model algorithms are
determined using a structural uncertainty estimator as described in Ogle et al. (2007, 2010). Sampling uncertainty
is assessed using the NRI replicate sampling weights.
Carbon stocks and 95 percent confidence intervals are estimated for each year between 1990 and 2015 using the
DayCent model. However, note that the areas have been modified in the original NRI survey through the process in
which the Forest Inventory and Analysis (FIA) survey data and the National Land Cover Dataset (Homer et al. 2007;
Fry et al. 2011; Homer et al. 2015) are harmonized with the NRI data. This process ensures that the areas of Forest
Land Remaining Forest Land and Land Converted to Forest Land are consistent with other land-use categories
while maintaining a consistent time series for the total land area of the United States. For example, if the FIA
estimate less Cropland Converted to Forest Land than the NRI, then the amount of area for this land-use
conversion is reduced in the NRI dataset and re-classified as Cropland Remaining Cropland (See Section 6.1,
Representation of the U.S. Land Base for more information). Further elaboration on the methodology and data
used to estimate stock changes from mineral soils are described in Annex 3.12 of EPA (2022).
In order to ensure time-series consistency, the Tier 3 method is applied from 1990 to 2015 so that changes reflect
anthropogenic activity and not methodological adjustments. In addition, soil organic C stock changes from 2016 to
2021 are approximated with a linear extrapolation of emission patterns from 1990 to 2015. The extrapolation is
based on a linear regression model with moving-average (ARMA) errors (See Box 6-4). Linear extrapolation is a
standard data splicing method for approximating emissions at the end of a time series (IPCC 2006). Time series of
activity data will be updated in a future inventory, and emissions from 2016 to 2021 will be recalculated.
Tier 2 Approach. In the IPCC Tier 2 method, data on climate, soil types, land use, and land management activity are
used to classify land area and apply appropriate factors to estimate soil organic C stock changes to a 30 cm depth
(Ogle et al. 2003, 2006). The primary source of activity data for land use, crop and irrigation histories is the 2015
NRI survey (USDA-NRCS 2018a). Each NRI survey location is classified by soil type, climate region, and management
condition using data from other sources. Survey locations on federal lands are included in the NRI, but land use
and cropping history are not compiled for these locations in the survey program (i.e., NRI is restricted to data
collection on non-federal lands). Therefore, land-use patterns for the NRI survey locations on federal lands are
Land Use, Land-Use Change, and Forestry 6-63
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
based on the National Land Cover Database (NLCD) (Yang et al. 2018; Fry et al. 2011; Homer et al. 2007; Homer et
al. 2015).
Additional management activities needed for the Tier 2 method are based on the imputation product described for
the Tier 3 approach, including tillage practices, mineral fertilization, and manure amendments that are assigned to
NRI survey locations. The one exception are activity data on wetland restoration of Conservation Reserve Program
land that are obtained from Euliss and Gleason (2002). Climate zones in the United States are classified using mean
precipitation and temperature (1950 to 2000) variables from the WorldClim data set (Hijmans et al. 2005) and
potential evapotranspiration data from the Consortium for Spatial Information (CGIAR-CSI) (Zomer et al. 2008,
2007) (Figure A-9). IPCC climate zones are then assigned to NRI survey locations.
Reference C stocks are estimated using the National Soil Survey Characterization Database (NRCS 1997) with
cultivated cropland as the reference condition, rather than native vegetation as used in IPCC (2006). Soil
measurements under agricultural management are much more common and easily identified in the National Soil
Survey Characterization Database (NRCS 1997) than are soils under a native condition, and therefore cultivated
cropland provides a more robust sample for estimating the reference condition. Country-specific C stock change
factors are derived from published literature to determine the impact of management practices on soil organic C
storage (Ogle et al. 2003, 2006). The factors represent changes in tillage, cropping rotations, intensification, and
land-use change between cultivated and uncultivated conditions. However, country-specific factors associated
with organic matter amendments are not estimated due to an insufficient number of studies in the United States
to analyze the impacts. Instead, factors from IPCC (2006) are used to estimate the effect of those activities.
Changes in soil organic C stocks for mineral soils are estimated 1,000 times for 1990 through 2015, using a Monte
Carlo stochastic simulation approach and probability distribution functions for the country-specific stock change
factors, reference C stocks, and land use activity data (Ogle et al. 2003; Ogle et al. 2006). Further elaboration on
the methodology and data used to estimate stock changes from mineral soils are described in Annex 3.12 of EPA
(2022).
In order to ensure time-series consistency, the Tier 2 method is applied from 1990 to 2015 so that changes reflect
anthropogenic activity and not methodological adjustments. In addition, soil organic C stock changes for the
remainder of the time series are approximated with a linear extrapolation of emission patterns from 1990 to 2015.
The extrapolation is based on a linear regression model with moving-average (ARMA) errors (See Box 6-4). Linear
extrapolation is a standard data splicing method for approximating emissions at the end of a time series (IPCC
2006). As with the Tier 3 method, time series of activity data will be updated in a future inventory, and emissions
from 2016 to 2021 will be recalculated (see Planned Improvements section).
Organic Soil Carbon Stock Changes
Annual C emissions from drained organic soils in Cropland Remaining Cropland are estimated using the Tier 2
method provided in IPCC (2006), with country-specific C loss rates (Ogle et al. 2003) rather than default IPCC rates.
The final estimates include a measure of uncertainty as determined from a Monte Carlo Simulation with 1,000
iterations. Emissions are based on the land area data for drained organic soils from 1990 to 2015 for Cropland
Remaining Cropland in the 2015 NRI (USDA-NRCS 2018a). Further elaboration on the methodology and data used
to estimate stock changes from organic soils are described in Annex 3.12 of EPA (2022).
In order to ensure time-series consistency, the same Tier 2 method is applied from 1990 to 2015 so that changes
reflect anthropogenic activity and not methodological adjustments. In addition, soil organic C stock changes for the
remainder of the time series are approximated with a linear extrapolation of emission patterns from 1990 to 2015.
The extrapolation is based on a linear regression model with moving-average (ARMA) errors (See Box 6-4). Linear
extrapolation is a standard data splicing method for approximating emissions at the end of a time series (IPCC
2006). Estimates for 2016 to 2021 will be recalculated in a future inventory when new activity data are
incorporated into the analysis.
6-64 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Uncertainty
Uncertainty is quantified for changes in soil organic C stocks associated with Cropland Remaining Cropland
(including both mineral and organic soils). Uncertainty estimates are presented in Table 6-30 for each subsource
(mineral and organic soil C stocks) and the methods that are used in the Inventory analyses (i.e., Tier 2 and Tier 3).
Uncertainty for the Tier 2 and 3 approaches is derived using a Monte Carlo approach (see Annex 3.12 of EPA 2022
for further discussion). For 2016 to 2021, additional uncertainty is propagated through the Monte Carlo Analysis
that is associated with the surrogate data method. Soil organic C stock changes from the Tier 2 and 3 approaches
are combined using the simple error propagation method provided by the IPCC (2006). The combined uncertainty
is calculated by taking the square root of the sum of the squares of the standard deviations of the uncertain
quantities.
The combined uncertainty for soil organic C stocks in Cropland Remaining Cropland ranges from 406 percent below
to 406 percent above the 2021 stock change estimate of -18.9 MMT CO2 Eq. The large relative uncertainty around
the 2021 stock change estimate is mostly due to variation in soil organic C stock changes that is not explained by
the surrogate data method, leading to high prediction error.
Table 6-30: Approach 2 Quantitative Uncertainty Estimates for Soil C Stock Changes
occurring within Cropland Remaining Cropland (MMT CO2 Eq. and Percent)
Source
2021 Flux Estimate
Uncertainty Range Relative to Flux Estimate3
(MMT CO? Eq.)
(MMT CO?
Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Mineral Soil C Stocks: Cropland Remaining
Cropland, Tier 3 Inventory Methodology
(46.6)
(120.8)
27.6
-159%
159%
Mineral Soil C Stocks: Cropland Remaining
Cropland, Tier 2 Inventory Methodology
(5.2)
(12.3)
1.8
-134%
134%
Organic Soil C Stocks: Cropland Remaining
Cropland, Tier 2 Inventory Methodology
32.9
13.9
51.9
-58%
58%
Combined Uncertainty for Flux associated
with Agricultural Soil Carbon Stock Change in
(18.9)
(95.9)
58.0
-406%
406%
Cropland Remaining Cropland
a Range of C stock change estimates predicted by Monte Carlo Stochastic Simulation with a 95 percent confidence interval.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Uncertainty is also associated with lack of reporting of agricultural woody biomass and dead organic matter C stock
changes. However, woody biomass C stock changes are likely minor in perennial crops, such as orchards and nut
plantations. There will be removal and replanting of tree crops each year, but the net effect on biomass C stock
changes is probably minor because the overall area and tree density is relatively constant across time series. In
contrast, agroforestry practices, such as shelterbelts, riparian forests and intercropping with trees, may have more
significant changes over the Inventory time series, compared to perennial woody crops, at least in some regions of
the United States, but there are currently no datasets to evaluate the trends. Changes in litter C stocks are also
assumed to be negligible in croplands over annual time frames, although there are certainly significant changes at
sub-annual time scales across seasons. This trend may change in the future, particularly if crop residue becomes a
viable feedstock for bioenergy production.
QA/QC and Verification
Quality control measures included checking input data, model scripts, and results to ensure data are properly
handled throughout the inventory process. Inventory reporting forms and text are reviewed and revised as needed
to correct transcription errors. Results from the DayCent model are compared to field measurements and soil
monitoring sites associated with the NRI (Spencer et al. 2011), and a statistical relationship has been developed to
Land Use, Land-Use Change, and Forestry 6-65
-------
1 assess uncertainties in the predictive capability of the model (Ogle et al. 2007). The comparisons include 72 long-
2 term experiment sites and 142 NRI soil monitoring network sites, with 948 observations across all of the sites (see
3 Annex 3.12 of EPA 2022 for more information).
7 There are two key improvements planned for the inventory, including a) incorporating the latest land use data
8 from the USDA National Resources Inventory, and b) conducting an analysis of C stock changes in Alaska for
9 cropland. This latter improvement will be conducted using the Tier 2 method for mineral and organic soils that is
10 described earlier in this section. The analysis will initially focus on land-use change, which typically has a larger
11 impact on soil organic C stock changes than management practices, but will be further refined over time to
12 incorporate management data. These two improvements will resolve most of the differences between the
13 managed land base for Cropland Remaining Cropland and amount of area currently included in Cropland
14 Remaining Cropland Inventory (See Table 6-31).
15 Table 6-31: Comparison of Managed Land Area in Cropland Remaining Cropland and Area in
16 the Current Cropland Remaining Cropland Inventory (Thousand Hectares)
4 Recalculations Discussion
5 There are no recalculations in the time series from the previous Inventory.
6 Planned Improvements
Area (Thousand Hectares)
Year
Managed Land Inventory
Difference
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
162,265 162,134
161,834 161,692
161,336 161,223
159,567 159,420
157,880 157,703
157,269 157,025
156,630 156,380
156,010 155,738
152,330 151,987
151,429 151,105
151,246 150,952
150,725 150,442
150,417 150,146
151,043 150,814
150,769 150,616
150,400 150,275
149,893 149,762
150,100 150,003
149,706 149,694
149,646 149,714
149,215 149,314
148,619 148,815
148,290 148,495
148,653 148,989
149,136 149,463
131
142
113
147
178
244
250
271
343
324
294
283
271
229
153
126
131
97
11
-68
-100
-195
-205
-336
-327
6-66 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
2015 148,520 148,851 -331
2016
148,432 * *
2017
148,327 * *
2018
149,721 * *
2019
149,504 * *
2020
149,817 * *
2021
150,586 * *
NRI data have not been incorporated into the inventory after 2015,
designated with asterisks (*).
1 There are several other planned improvements underway related to the plant production module in DayCent. A
2 key improvement for a future Inventory will be to incorporate additional management activity data from the
3 USDA-NRCS Conservation Effects Assessment Project survey. The CEAP survey has compiled new data in recent
4 years. Crop parameters associated with temperature effects on plant production will be further improved in
5 DayCent with additional model calibration. Senescence events following grain filling in crops, such as wheat, are
6 being modified based on recent model algorithm development, and will be incorporated. There will also be further
7 testing and parameterization of the DayCent model to reduce the bias in model predictions for grasslands, which
8 was discovered through model evaluation by comparing output to measurement data from 72 experimental sites
9 and 142 NRI soil monitoring network sites (See QA/QC and Verification section).
10 Improvements are underway to simulate crop residue burning in the DayCent model based on the amount of crop
11 residues burned according to the data that are used in the Field Burning of Agricultural Residues source category
12 (see Section 5.7). This improvement will more accurately represent the C inputs to the soil that are associated with
13 residue burning. In addition, a review of available data on biosolids (i.e., treated sewage sludge) application will be
14 undertaken to improve the distribution of biosolids application on croplands, grasslands and settlements.
15 Many of these improvements are expected to be completed for the 1990 through 2022 Inventory (i.e., 2024
16 submission to the UNFCCC). However, the timeline may be extended if there are insufficient resources to fund all
17 or part of these planned improvements.
18
19 6.5 Land Converted to Cropland (CRF
20 Category 4B2)
21 Land Converted to Cropland includes all cropland in an inventory year that had been in another land use(s) during
22 the previous 20 years (USDA-NRCS 2018), and used to produce food or fiber, or forage that is harvested and used
23 as feed (e.g., hay and silage). For example, Grassland or Forest Land Converted to Cropland during the past 20
24 years would be reported in this category. Recently converted lands are retained in this category for 20 years as
25 recommended by IPCC (2006).
26 Land-use change can lead to large losses of C to the atmosphere, particularly conversions from forest land
27 (Houghton et al. 1983; Houghton and Nassikas 2017). Moreover, conversion of forest to another land use (i.e.,
28 deforestation) is one of the largest anthropogenic sources of emissions to the atmosphere globally, although this
29 source may be declining according to a recent assessment (Tubiello et al. 2015).
30 The 2006 IPCC Guidelines recommend reporting changes in biomass, dead organic matter and soil organic C stocks
31 with land-use change. All soil organic C stock changes are estimated and reported for Land Converted to Cropland,
32 but reporting of C stock changes for aboveground and belowground biomass, dead wood, and litter pools is limited
Land Use, Land-Use Change, and Forestry 6-67
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
to Forest Land Converted to Cropland and Grassland Converted Cropland for woodland conversions (i.e., woodland
conversion to cropland).46
There are several discrepancies between the current land representation (See Section 6.1) and the area data that
have been used in the inventory for Land Converted to Cropland. First, the current land representation is based on
the latest NRI dataset, which includes data through 2017, but these data have not been incorporated into the Land
Converted to Cropland Inventory. Second, cropland in Alaska is not included in the Inventory, but is a relatively
small amount of U.S. cropland area (approximately 28,700 hectares). Third, some miscellaneous croplands are also
not included in the Inventory due to limited understanding of greenhouse gas emissions from these management
systems (e.g., aquaculture). These differences lead to small discrepancies between the managed area in Land
Converted to Cropland and the cropland area included in the Land Converted to Cropland Inventory analysis (Table
6-35). Improvements are underway to incorporate the latest NRI dataset, croplands in Alaska and miscellaneous
croplands as part of future C inventories (See Planned Improvements section).
Forest Land Converted to Cropland is the largest source of emissions from 1990 to 2021, accounting for
approximately 86 percent of the average total loss of C among all of the land-use conversions in Land Converted to
Cropland. The pattern is due to the large losses of biomass and dead organic matter C for Forest Land Converted to
Cropland. The next largest source of emissions is Grassland Converted to Cropland accounting for approximately
17 percent of the total emissions (Table 6-32 and Table 6-33). The net change in total C stocks for 2021 led to CO2
emissions to the atmosphere of 56.5 MMT CO2 Eq. (15.4 MMT C), including 29.8 MMT CO2 Eq. (8.1 MMT C) from
aboveground biomass C losses, 5.8 MMT CO2 Eq. (1.6 MMT C) from belowground biomass C losses, 5.8 MMT CO2
Eq. (1.6 MMT C) from dead wood C losses, 8.2 MMT C02 Eq. (2.2 MMT C) from litter C losses, 3.2 MMT C02 Eq. (0.9
MMT C) from mineral soils and 3.8 MMT CO2 Eq. (1.0 MMT C) from drainage and cultivation of organic soils.
Emissions in 2021 are 3 percent higher than emissions in the initial reporting year, i.e., 1990.
Table 6-32: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes in
Land Converted to Cropland by Land Use Change Category (MMT CO2 Eq.)
1990
2005
2017
2018
2019
2020
2021
Grassland Converted to Cropland
8.0
8.6
9.8
9.6
9.6
9.9
9.8
Aboveground Live Biomass
0.6
0.6
0.6
0.6
0.6
0.6
0.6
Belowground Live Biomass
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Dead Wood
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Litter
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Mineral Soils
4.1
4.0
5.4
5.1
5.1
5.5
5.3
Organic Soils
2.7
3.5
3.3
3.3
3.3
3.3
3.3
Forest Land Converted to Cropland
48.2
48.1
48.5
48.5
48.5
48.5
48.5
Aboveground Live Biomass
28.8
28.9
29.2
29.2
29.2
29.2
29.2
Belowground Live Biomass
5.6
5.6
5.7
5.7
5.7
5.7
5.7
Dead Wood
5.5
5.5
5.5
5.5
5.5
5.5
5.5
Litter
7.8
7.8
8.0
8.0
8.0
8.0
8.0
Mineral Soils
0.4
0.2
0.1
0.1
0.1
0.2
0.1
Organic Soils
0.1
0.1
+
+
+
+
+
Other Lands Converted to Cropland
(2.2)
(2.9)
(2.2)
(2.2)
(2.3)
(2.3)
(2.3)
Mineral Soils
(2.3)
(2.9)
(2.2)
(2.2)
(2.3)
(2.3)
(2.3)
Organic Soils
0.2
0.1
+
+
+
+
+
Settlements Converted to Cropland
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Mineral Soils
(0.1)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Organic Soils
+
+
+
+
+
+
+
46 Changes in biomass C stocks are estimated for Forest Land Converted to Cropland and Grassland Converted to Cropland for
woodland conversions. There is a planned improvement to include the effect of other land-use conversions, in addition to
herbaceous grassland conversions to cropland in a future Inventory. Note: changes in dead organic matter are assumed
negligible for other land-use conversions to cropland, except Forest Land and woodland conversions.
6-68 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Wetlands Converted to Cropland
0.8
0.9
0.6
0.6
0.6
0.6
0.7
Mineral Soils
0.3
0.3
0.2
0.2
0.2
0.2
0.2
Organic Soils
0.6
0.6
0.3
0.4
0.4
0.4
0.4
Aboveground Live Biomass
29.4
29.5
29.8
29.8
29.8
29.8
29.8
Belowground Live Biomass
5.7
5.7
5.8
5.8
5.8
5.8
5.8
Dead Wood
5.7
5.7
5.8
5.8
5.8
5.8
5.8
Litter
8.0
8.1
8.2
8.2
8.2
8.2
8.2
Total Mineral Soil Flux
2.3
1.3
3.4
3.1
3.0
3.5
3.2
Total Organic Soil Flux
3.7
4.3
3.7
3.7
3.7
3.8
3.8
Total Net Flux
54.8
54.7
56.6
56.3
56.3
56.7
56.5
+ Does not exceed 0.05 MMT C02 Eq.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
1 Table 6-33: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes in
2 Land Converted to Cropland (MMT C)
1990
2005
2017
2018
2019
2020
2021
Grassland Converted to Cropland
2.2
2.4
2.7
2.6
2.6
2.7
2.7
Aboveground Live Biomass
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Belowground Live Biomass
+
+
+
+
+
+
+
Dead Wood
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Litter
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Mineral Soils
1.1
1.1
1.5
1.4
1.4
1.5
1.5
Organic Soils
0.7
1.0
0.9
0.9
0.9
0.9
0.9
Forest Land Converted to Cropland
13.1
13.1
13.2
13.2
13.2
13.2
13.2
Aboveground Live Biomass
7.9
7.9
8.0
8.0
8.0
8.0
8.0
Belowground Live Biomass
1.5
1.5
1.5
1.5
1.5
1.5
1.5
Dead Wood
1.5
1.5
1.5
1.5
1.5
1.5
1.5
Litter
2.1
2.1
2.2
2.2
2.2
2.2
2.2
Mineral Soils
0.1
+
+
+
+
+
+
Organic Soils
+
+
+
+
+
+
+
Other Lands Converted to Cropland
(0.6)
(0.8)
(0.6)
(0.6)
(0.6)
(0.6)
(0.6)
Mineral Soils
(0.6)
(0.8)
(0.6)
(0.6)
(0.6)
(0.6)
(0.6)
Organic Soils
+
+
+
+
+
+
+
Settlements Converted to Cropland
+
+
+
+
+
+
+
Mineral Soils
+
+
+
+
+
+
+
Organic Soils
+
+
+
+
+
+
+
Wetlands Converted to Cropland
0.2
0.3
0.2
0.2
0.2
0.2
0.2
Mineral Soils
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Organic Soils
0.2
0.2
0.1
0.1
0.1
0.1
0.1
Aboveground Live Biomass
8.0
8.1
8.1
8.1
8.1
8.1
8.1
Belowground Live Biomass
1.6
1.6
1.6
1.6
1.6
1.6
1.6
Dead Wood
1.6
1.6
1.6
1.6
1.6
1.6
1.6
Litter
2.2
2.2
2.2
2.2
2.2
2.2
2.2
Total Mineral Soil Flux
0.6
0.4
0.9
0.8
0.8
0.9
0.9
Total Organic Soil Flux
1.0
1.2
1.0
1.0
1.0
1.0
1.0
Total Net Flux
14.9
14.9
15.4
15.4
15.3
15.5
15.4
+ Does not exceed 0.05 MMT C.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
3 Methodology and Time-Series Consistency
4 The following section includes a description of the methodology used to estimate C stock changes for Land
5 Converted to Cropland, including (1) loss of aboveground and belowground biomass, dead wood and litter C with
6 conversion of forest lands to croplands, as well as (2) the impact from all land-use conversions to cropland on
7 mineral and soil organic C stocks.
Land Use, Land-Use Change, and Forestry 6-69
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Biomass, Dead Wood and Litter Carbon Stock Changes
A Tier 2 method is applied to estimate biomass, dead wood, and litter C stock changes for Forest Land Converted
to Cropland and Grassland Converted to Cropland for woodland conversions. Estimates are calculated in the same
way as those in the Forest Land Remaining Forest Land category using data from the USDA Forest Service, Forest
Inventory and Analysis (FIA) program (USDA Forest Service 2022). However, there are no country-specific data for
cropland biomass, so default biomass values (IPCC 2006) were used to estimate the carbon stocks for the new
cropland (litter and dead wood carbon stocks were assumed to be zero since no reference C density estimates
exist for croplands). The difference between the stocks is reported as the stock change under the assumption that
the change occurred in the year of the conversion. If FIA plots include data on individual trees, aboveground and
belowground C density estimates are based on Woodall et al. (2011). Aboveground and belowground biomass
estimates also include live understory which is a minor component of biomass defined as all biomass of
undergrowth plants in a forest, including woody shrubs and trees less than 2.54 cm dbh. For this Inventory, it was
assumed that 10 percent of total understory C mass is belowground (Smith et al. 2006). Estimates of C density are
based on information in Birdsey (1996) and biomass estimates from Jenkins et al. (2003).
For dead organic matter, if FIA plots include data on standing dead trees, standing dead tree C density is estimated
following the basic method applied to live trees (Woodall et al. 2011) with additional modifications to account for
decay and structural loss (Domke et al. 2011; Harmon et al. 2011). If FIA plots include data on downed dead wood,
downed dead wood C density is estimated based on measurements of a subset of FIA plots for downed dead wood
(Domke et al. 2013; Woodall and Monleon 2008). Downed dead wood is defined as pieces of dead wood greater
than 7.5 cm diameter, at transect intersection, that are not attached to live or standing dead trees. This includes
stumps and roots of harvested trees. To facilitate the downscaling of downed dead wood C estimates from the
state-wide population estimates to individual plots, downed dead wood models specific to regions and forest types
within each region are used. Litter C is the pool of organic C (also known as duff, humus, and fine woody debris)
above the mineral soil and includes woody fragments with diameters of up to 7.5 cm. A subset of FIA plots are
measured for litter C. If FIA plots include litter material, a modeling approach using litter C measurements from FIA
plots is used to estimate litter C density (Domke et al. 2016). In order to ensure time-series consistency, the same
methods are applied from 1990 to 2021 so that changes reflect anthropogenic activity and not methodological
adjustments. See Annex 3.13 for more information about reference C density estimates for forest land and the
compilation system used to estimate carbon stock changes from forest land. See the Grassland Remaining
Grassland section for more information about estimation of biomass, deadwood and litter C stock changes for
woodlands.
Soil Carbon Stock Changes
Soil organic stock changes are estimated for Land Converted to Cropland according to land use histories recorded
in the 2015 USDA NRI survey for non-federal lands (USDA-NRCS 2018). Land use and some management
information (e.g., crop type, soil attributes, and irrigation) had been collected for each NRI point on a 5-year cycle
beginning in 1982. In 1998, the NRI program began collecting annual data, which are currently available through
2017, however this Inventory uses the previous NRI with annual data available through 2015 (USDA-NRCS 2018).
NRI survey locations are classified as Land Converted to Cropland in a given year between 1990 and 2015 if the
land use is cropland but had been another use during the previous 20 years. NRI survey locations are classified
according to land use histories starting in 1979, and consequently the classifications are based on less than 20
years from 1990 to 1998, which may have led to an underestimation of Land Converted to Cropland in the early
part of the time series to the extent that some areas are converted to cropland from 1971 to 1978. For federal
lands, the land use history is derived from land cover changes in the National Land Cover Dataset (Yang et al. 2018;
Homer et al. 2007; Fry et al. 2011; Homer et al. 2015).
6-70 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Mineral Soil Carbon Stock Changes
An IPCC Tier 3 model-based approach (Ogle et al. 2010) is applied to estimate C stock changes from 1990 to 2015
for mineral soils on the majority of land that is used to produce annual crops and forage crops that are harvested
and used as feed (e.g., hay and silage) in the United States. These crops include alfalfa hay, barley, corn, cotton,
grass hay, grass-clover hay, oats, peanuts, potatoes, rice, sorghum, soybeans, sugar beets, sunflowers, tobacco,
and wheat. Soil organic C stock changes on the remaining mineral soils are estimated with the IPCC Tier 2 method
(Ogle et al. 2003), including land used to produce some vegetables and perennial/horticultural crops and crops
rotated with these crops; land on very gravelly, cobbly, or shaley soils (greater than 35 percent by volume); and
land converted from another land use or federal ownership.47
For the years 2016 to 2021, a surrogate data method is used to estimate soil organic C stock changes at the
national scale for land areas included in the Tier 2 and Tier 3 methods. Specifically, linear regression models with
autoregressive moving-average (ARMA) errors (Brockwell and Davis 2016) are used to estimate the relationship
between surrogate data and the 1990 to 2015 stock change data from the Tier 2 and 3 methods. Surrogate data
for these regression models include corn and soybean yields from USDA-NASS statistics,48 and weather data from
the PRISM Climate Group (PRISM 2018). See Box 6-4 in the Methodology section of Cropland Remaining Cropland
for more information about the surrogate data method. Stock change estimates for 2016 to 2021 will be
recalculated in future Inventories when the time series of activity data are updated.
Tier 3 Approach. For the Tier 3 method, mineral soil organic C stocks and stock changes are estimated using the
DayCent biogeochemical model (Parton et al. 1998; Del Grosso et al. 2001, 2011). The DayCent model utilizes the
soil C modeling framework developed in the Century model (Parton et al. 1987,1988,1994; Metherell et al. 1993),
but has been refined to simulate dynamics at a daily time-step. National estimates are obtained by using the
model to simulate historical land-use change patterns as recorded in the USDA NRI survey (USDA-NRCS 2018).
Carbon stocks and 95 percent confidence intervals are estimated for each year between 1990 and 2015. See the
Cropland Remaining Cropland section and Annex 3.12 of EPA (2022) for additional discussion of the Tier 3
methodology for mineral soils.
In order to ensure time-series consistency, the Tier 3 method is applied from 1990 to 2015 so that changes reflect
anthropogenic activity and not methodological adjustments. Soil organic C stock changes from 2016 to 2021 are
approximated using a linear extrapolation of emission patterns from 1990 to 2015. The extrapolation is based on a
linear regression model with moving-average (ARMA) errors (described in Box 6-4 of the Methodology section in
Cropland Remaining Cropland). Linear extrapolation is a standard data splicing method for estimating emissions at
the end of a time series (IPCC 2006). Time series of activity data will be updated in a future Inventory, and
emissions from 2016 to 2021 will be recalculated.
Tier 2 Approach. For the mineral soils not included in the Tier 3 analysis, soil organic C stock changes are estimated
using a Tier 2 Approach, as described in the Tier 2 Approach for mineral soils in Cropland Remaining Cropland. In
order to ensure time-series consistency, the Tier 2 method is applied from 1990 to 2015 so that changes reflect
anthropogenic activity and not methodological adjustments. In addition, soil organic C stock changes are
approximated for the remainder of the 2016 to 2021 time series with a linear extrapolation of emission patterns
from 1990 to 2015. The extrapolation is based on a linear regression model with moving-average (ARMA) (See Box
6-4 of the Methodology section in Cropland Remaining Cropland). Linear extrapolation is a standard data splicing
method for estimating emissions at the end of a time series (IPCC 2006). As with the Tier 3 method, time series of
activity data will be updated in a future Inventory, and emissions from 2016 to 2021 will be recalculated.
47 Federal land is not a land use, but rather an ownership designation that is treated as grassland for purposes of these
calculations. The specific land use on federal lands is not identified in the NRI survey (USDA-NRCS 2018).
48 See https://quickstats.nass.usda.gov/.
Land Use, Land-Use Change, and Forestry 6-71
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Organic Soil Carbon Stock Changes
Annual C emissions from drained organic soils in Land Converted to Cropland are estimated using the Tier 2
method provided in IPCC (2006), with country-specific C loss rates (Ogle et al. 2003) as described in the Cropland
Remaining Cropland section for organic soils. Further elaboration on the methodology is also provided in Annex
3.12 of EPA (2022).
In order to ensure time-series consistency, the Tier 2 methods are applied from 1990 to 2015 so that changes
reflect anthropogenic activity and not methodological adjustments. In addition, soil organic C stock changes for the
remainder of the time series (i.e., 2016 to 2021) are approximated with a linear extrapolation of emission patterns
from 1990 to 2015. The extrapolation is based on a linear regression model with moving-average (ARMA) errors
(See Box 6-4 of the Methodology section in Cropland Remaining Cropland). Linear extrapolation is a standard data
splicing method for approximating emissions at the end of a time series (IPCC 2006). Estimates for 2016 to 2021
will be recalculated in a future inventory when new activity data are incorporated into the analysis.
Uncertainty
The uncertainty analyses for biomass, dead wood and litter C losses with Forest Land Converted to Cropland and
Grassland Converted to Cropland for woodland conversions are conducted in the same way as the uncertainty
assessment for forest ecosystem C flux associated with Forest Land Remaining Forest Land. Sample and model-
based error are combined using simple error propagation methods provided by the IPCC (2006) by taking the
square root of the sum of the squares of the standard deviations of the uncertain quantities. For additional details,
see the Uncertainty Analysis in Annex 3.13.
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 described in Cropland Remaining Cropland. The uncertainty for annual C
emission estimates from drained organic soils in Land Converted to Cropland is estimated using a Monte Carlo
approach, which is also described in the Cropland Remaining Cropland section. For 2016 to 2021, there is
additional uncertainty propagated through the Monte Carlo Analysis associated with the surrogate data method,
which is also described in Cropland Remaining Cropland.
Uncertainty estimates are presented in Table 6-34 for each subsource (i.e., biomass C stocks, dead wood C stocks,
litter C stocks, soil organic C stocks for mineral and organic soils) and the method applied in the Inventory analysis
(i.e., Tier 2 and Tier 3). Uncertainty estimates for the total C stock changes for biomass, dead organic matter and
soils are combined using the simple error propagation methods provided by the IPCC (2006), as discussed in the
previous paragraph. The combined uncertainty for total C stocks in Land Converted to Cropland ranged from 94
percent below to 94 percent above the 2021 stock change estimate of 56.5 MMT CO2 Eq. The large relative
uncertainty in the 2021 estimate is mostly due to variation in soil organic C stock changes that is not explained by
the surrogate data method, leading to high prediction error with this splicing method.
Table 6-34: Approach 2 Quantitative Uncertainty Estimates for Soil, Dead Organic Matter
and Biomass C Stock Changes occurring within Land Converted to Cropland (MMT CO2 Eq.
and Percent)
2021 Flux Estimate Uncertainty Range Relative to Flux Estimate3
(MMT C02 Eq.) (MMT CP2 Eq.) (%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Grassland Converted to Cropland
9.8
(25.6)
45.1
-362%
362%
Aboveground Live Biomass
0.6
(0.1)
1.3
-125%
125%
Belowground Live Biomass
0.1
+
0.2
-137%
120%
Dead Wood
0.2
(0.1)
0.5
-134%
123%
Litter
0.2
(0.1)
0.5
-134%
119%
Mineral Soil C Stocks: Tier 3
1.0
(34.1)
36.1
-3,546%
3,546%
6-72 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Mineral Soil C Stocks: Tier 2
4.3
1.2
7.4
-71%
71%
Organic Soil C Stocks: Tier 2
3.3
0.8
5.8
-75%
75%
Forest Land Converted to Cropland
48.5
8.7
88.3
-82%
82%
Aboveground Live Biomass
29.2
(7.9)
66.4
-127%
127%
Belowground Live Biomass
5.7
(1.5)
12.9
-127%
127%
Dead Wood
5.5
(1.5)
12.6
-127%
127%
Litter
8.0
(2.2)
18.1
-127%
127%
Mineral Soil C Stocks: Tier 2
0.1
(0.1)
0.4
-145%
145%
Organic Soil C Stocks: Tier 2
+
(0.1)
0.2
-2,595%
2,595%
Other Lands Converted to Cropland
(2.3)
(3.8)
(0.8)
-66%
66%
Mineral Soil C Stocks: Tier 2
(2.3)
(3.8)
(0.8)
-66%
66%
Organic Soil C Stocks: Tier 2
+
+
+
0%
0%
Settlements Converted to Cropland
(0.1)
(0.3)
+
-116%
116%
Mineral Soil C Stocks: Tier 2
(0.2)
(0.3)
+
-90%
90%
Organic Soil C Stocks: Tier 2
+
+
0.1
-85%
85%
Wetlands Converted to Croplands
0.7
+
1.3
-98%
98%
Mineral Soil C Stocks: Tier 2
0.2
+
0.5
-110%
110%
Organic Soil C Stocks: Tier 2
0.4
(0.2)
1.0
-142%
142%
Total: Land Converted to Cropland
56.5
3.2
109.8
-94%
94%
Aboveground Live Biomass
29.8
(7.3)
67.0
-125%
125%
Belowground Live Biomass
5.8
(1.4)
13.0
-125%
125%
Dead Wood
5.8
(1.3)
12.8
-123%
122%
Litter
8.2
(1.9)
18.3
-124%
124%
Mineral Soil C Stocks: Tier 3
1.0
(34.1)
36.1
-3,546%
3,546%
Mineral Soil C Stocks: Tier 2
2.2
(1.2)
5.7
-155%
155%
Organic Soil C Stocks: Tier 2
3.8
1.2
6.4
-68%
68%
+ Does not exceed 0.05 MMT C02 Eq.
a Range of C stock change estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
1 Uncertainty is also associated with lack of reporting of agricultural biomass and dead organic matter C stock
2 changes. Biomass C stock changes are likely minor in perennial crops, such as orchards and nut plantations, given
3 the small amount of change in land that is used to produce these commodities in the United States. In contrast,
4 agroforestry practices, such as shelterbelts, riparian forests and intercropping with trees, may have led to larger
5 changes in biomass C stocks at least in some regions of the United States. However, there are currently no datasets
6 to evaluate the trends. Changes in dead organic matter C stocks are assumed to be negligible with conversion of
7 land to croplands with the exception of forest lands, which are included in this analysis. This assumption will be
8 further explored in a future Inventory.
9 QA/QC and Verification
10 See the QA/QC and Verification section in Cropland Remaining Cropland for information on QA/QC steps.
11 Recalculations Discussion
12 Recalculations are associated with new FIA data from 1990 to 2021 on biomass, dead wood and litter C stocks in
13 Grassland Converted to Cropland (i.e., woodland conversion to cropland), updated FIA data from 1990 to 2021 on
14 biomass, dead wood and litter C stocks in Forest Land Converted to Cropland, and updated estimates for mineral
15 soils from 2016 to 2021 using the linear extrapolation method. As a result, Land Converted to Cropland has an
16 estimated larger C loss of 2.6 MMT CO2 Eq. on average over the time series. This represents a 4.9 percent increase
17 in C stock changes for Land Converted to Grassland compared to the previous Inventory.
Land Use, Land-Use Change, and Forestry 6-73
-------
1
2
3
4
5
6
7
8
9
10
11
12
Planned Improvements
There are two key improvements planned for the Inventory, including a) incorporating the latest land use data
from the USDA National Resources Inventory, and b) conducting an analysis of C stock changes in Alaska for
cropland. These two improvements will resolve most of the discrepancies between the managed land base for
Land Converted to Cropland and amount of area currently included in Land Converted to Cropland Inventory (See
Table 6-35). Another planned improvement is to estimate the biomass C stock changes for other land-use changes
besides Forest Land Converted to Cropland and Grassland Converted to Cropland for woodland conversion.
Additional planned improvements are discussed in the Planned Improvements section of Cropland Remaining
Cropland.
Table 6-35: Comparison of Managed Land Area in Land Converted to Cropland and the Area
in the current Land Converted to Cropland Inventory (Thousand Hectares)
Area (Thousand Hectares)
Year
Managed Land
Inventory
Difference
1990
12,230
12,308
-77
1991
12,561
12,654
-94
1992
12,858
12,943
-85
1993
14,093
14,218
-125
1994
15,266
15,400
-134
1995
15,439
15,581
-143
1996
15,740
15,888
-148
1997
15,919
16,073
-154
1998
17,263
17,440
-177
1999
17,659
17,819
-160
2000
17,518
17,693
-175
2001
17,441
17,600
-158
2002
17,311
17,487
-177
2003
16,064
16,257
-194
2004
15,136
15,317
-182
2005
15,221
15,424
-202
2006
15,149
15,410
-262
2007
14,734
14,923
-189
2008
14,248
14,399
-150
2009
13,762
13,814
-52
2010
13,888
13,905
-17
2011
14,209
14,186
22
2012
14,450
14,429
21
2013
13,991
13,752
239
2014
13,464
13,050
414
2015
13,561
13,049
512
2016
13,519
*
*
2017
13,594
*
*
2018
11,673
*
*
2019
11,189
*
*
2020
10,293
*
*
2021
9,491
*
*
NRI data have not been incorporated into the inventory after 2015, designated with asterisks (*).
6-74 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
6.6 Grassland Remaining Grassland (CRF
Category 4C1)
Carbon (C) in grassland ecosystems occurs in biomass, dead organic matter, and soils. Soils are the largest pool of C
in grasslands, and have the greatest potential for longer-term storage or release of C. Biomass and dead organic
matter C pools are relatively ephemeral compared to the soil C pool, with the exception of C stored in tree and
shrub biomass that occurs in grasslands. The 2006IPCC 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 grasslands49),
and may be extended to include agroforestry management associated with grasslands in the future. For soil
organic C, 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.50
Grassland Remaining Grassland includes all grassland in an Inventory year that had been grassland for a continuous
time period of at least 20 years (USDA-NRCS 2018). 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 interseeding of legumes. Woodlands are also considered
grassland and are areas of continuous tree cover that do not meet the definition of forest land (See Land
Representation section for more information about the criteria for forest land).
There are two discrepancies between the current land representation (See Section 6.1) and the area data that
have been used in the inventory for Grassland Remaining Grassland. First, the current land representation is based
on the latest NRI dataset, which includes data through 2017, but these data have not yet been incorporated into
the Grassland Remaining Grassland Inventory. Second, grassland in Alaska is not included in the Inventory, and is
approximately 50 million hectares. These differences lead to discrepancies between the managed area in
Grassland Remaining Grassland and the grassland area included in the Grassland Remaining Grassland Inventory
analysis (Table 6-39). Improvements are underway to incorporate the latest NRI dataset, and grasslands in Alaska
as part of future C inventories (See Planned Improvements Section).
For Grassland Remaining Grassland, there has been considerable variation in C stocks between 1990 and 2021.
These changes are driven by variability in weather patterns and associated interaction with land management
activity. Moreover, changes are small on a per hectare rate basis across the time series even in the years with a
larger total change in stocks. The net change in total C stocks for 2021 led to net CO2 emissions to the atmosphere
of 10.0 MMT CO2 Eq. (2.7 MMT C), including 2.1 MMT CO2 Eq. (0.6 MMT C) from net losses of aboveground
biomass C, 0.3 MMT CO2 Eq. (0.1 MMT C) from net losses in belowground biomass C, 3.0 MMT CO2 Eq. (0.8 MMT
C) from net losses in dead wood C, less than 0.05 MMT CO2 Eq. (less than 0.05 MMT C) from net gains in litter C,
0.8 MMT CO2 Eq. (0.2 MMT C) from net gains in mineral soil organic C, and 5.4 MMT CO2 Eq. (1.5 MMT C) from
losses of C due to drainage and cultivation of organic soils (Table 6-36 and Table 6-37). Losses of carbon are 15
percent higher in 2021 compared to 1990, but as noted previously, stock changes are highly variable from 1990 to
2021, with an average annual change of 9.4 MMT CO2 Eq. (2.6 MMT C).
49 Woodlands are considered grasslands in the U.S. Land Representation because they do not meet the definition of Forest
Land.
50 C02 emissions associated with liming and urea fertilization are also estimated but included in the Agriculture chapter of the
report.
Land Use, Land-Use Change, and Forestry 6-75
-------
1 Table 6-36: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes in
2 Grassland Remaining Grassland (MMT CO2 Eq.)
1990
2005
2017
2018
2019
2020
2021
Aboveground Live Biomass
1.4
1.7
2.1
2.1
2.1
2.1
2.1
Belowground Live Biomass
0.2
0.3
0.3
0.3
0.3
0.3
0.3
Dead Wood
3.2
3.2
3.0
3.0
3.0
3.0
3.0
Litter
(0.3)
(0.1)
+
+
+
+
+
Mineral Soils
(2.2)
0.8
0.1
0.4
3.2
(4.8)
(0.8)
Organic Soils
6.3
5.2
5.4
5.4
5.4
5.4
5.4
Total Net Flux
8.7
11.0
10.9
11.3
14.0
6.0
10.0
+ Does not exceed 0.05 MMT C02 Eq.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net
sequestration.
3 Table 6-37: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes in
4 Grassland Remaining Grassland (MMT C)
1990
2005
2017
2018
2019
2020
2021
Aboveground Live Biomass
0.4
0.5
0.6
0.6
0.6
0.6
0.6
Belowground Live Biomass
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Dead Wood
0.9
0.9
0.8
0.8
0.8
0.8
0.8
Litter
(0.1)
+
+
+
+
+
+
Mineral Soils
(0.6)
0.2
+
0.1
0.9
(1.3)
(0.2)
Organic Soils
1.7
1.4
1.5
1.5
1.5
1.5
1.5
Total Net Flux
2.4
3.0
3.0
3.1
3.8
1.6
2.7
+ Does not exceed 0.05 MMT C
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net
sequestration.
5 The spatial variability in soil organic C stock changes for 201551 is displayed in Figure 6-8 for mineral soils and in
6 Figure 6-9 for organic soils. Although relatively small on a per-hectare basis, grassland soils gained C in isolated
7 areas that mostly occurred in pastures of the eastern United States. For organic soils, the regions with the highest
8 rates of emissions coincide with the largest concentrations of organic soils used for managed grassland, including
9 the Southeastern Coastal Region (particularly Florida), upper Midwest and Northeast, and a few isolated areas
10 along the Pacific Coast.
51 Only national-scale emissions are estimated for 2016 to 2021 in the current Inventory using the surrogate data method, and
therefore the fine-scale emission patterns in this map are based on inventory data from 2015.
6-76 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Figure 6-8: Total Net Annual Soil C Stock Changes for Mineral Soils under Agricultural
2 Management within States, 2015, Grassland Remaining Grassland
3
4 Note: Only national-scale soil organic C stock changes are estimated for 2016 to 2021 in the current Inventory using a
5 surrogate data method, and therefore the fine-scale emission patterns in this map are based on inventory data from
6 2015. Negative values represent a net increase in soil organic C stocks, and positive values represent a net decrease in
7 soil organic C stocks.
Land Use, Land-Use Change, and Forestry 6-77
-------
1
2
Figure 6-9: Total Net Annual Soil C Stock Changes for Organic Soils under Agricultural
Management within States, 2015, Grassland Remaining Grassland
¦ > 40
3
4 Note: Only national-scale soil organic carbon stock changes are estimated for 2016 to 2021 in the current Inventory
5 using a surrogate data method, and therefore the fine-scale emission patterns in this map are based on inventory
6 data from 2015.
7 Methodology and Time-Series Consistency
8 The following section includes a description of the methodology used to estimate C stock changes for Grassland
9 Remaining Grassland, including (1) aboveground and belowground biomass, dead wood and litter C for woodlands,
10 as well as (2) soil organic C stocks for mineral and organic soils.
11 Biomass, Dead Wood and Litter Carbon Stock Changes
12 Woodlands are lands that do not meet the definition of forest land or agroforestry (see Section 6.1 Representation
13 of the U.S. Land Base), but include woody vegetation with C storage in aboveground and belowground biomass,
14 dead wood and litter C (IPCC 2006) as described in the Forest Land Remaining Forest Land section. Carbon stocks
15 and net annual C stock change were determined according to the stock-difference method for the conterminous
16 United States, which involved applying C estimation factors to annual forest inventories across time to obtain C
17 stocks and then subtracting the values between years to estimate the stock changes. The methods for estimating
18 carbon stocks and stock changes for woodlands in Grassland Remaining Grassland are consistent with those in the
19 Forest Land Remaining Forest Land section and are described in Annex 3.13. All annual National Forest Inventory
20 (NFI) plots available in the public FIA database (USDA Forest Service 2022) were used in the current Inventory.
21 While the NFI is an all-lands inventory, only those plots that meet the definition of forest land are typically
22 measured. However, in some cases, particularly in the Central Plains and Southwest United States, woodlands have
23 been measured as part of the survey. This analysis is limited to those plots and is not considered a comprehensive
24 assessment of trees outside of forest land that meet the definition of grassland. The same methods are applied
25 from 1990 to 2021 in order to ensure time-series consistency. This methodology is consistent with IPCC (2006).
6-78 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Soil Carbon Stock Changes
The following section includes a brief description of the methodology used to estimate changes in soil organic C
stocks for Grassland Remaining Grassland, including: (1) agricultural land use and management activities on
mineral soils; and (2) agricultural land use and management activities on organic soils. Further elaboration on the
methodologies and data used to estimate stock changes from mineral and organic soils are provided in the
Cropland Remaining Cropland section and Annex 3.12 of EPA (2022).
Soil organic C stock changes are estimated for Grassland Remaining Grassland on non-federal lands according to
land use histories recorded in the 2015 USDA NRI survey (USDA-NRCS 2018). Land use and some management
information (e.g., grass type, soil attributes, and irrigation) were originally collected for each NRI survey location
on a 5-year cycle beginning in 1982. In 1998, the NRI program began collecting annual data, and the annual data
are currently available through 2017, however this Inventory uses the previous NRI with annual data through 2015
(USDA-NRCS 2015). NRI survey locations are classified as Grassland Remaining Grassland in a given year between
1990 and 2015 if the land use had been grassland for 20 years. NRI survey locations are classified according to land
use histories starting in 1979, and consequently the classifications are based on less than 20 years from 1990 to
1998. This may have led to an overestimation of Grassland Remaining Grassland in the early part of the time series
to the extent that some areas are converted to grassland between 1971 and 1978. For federal lands, the land use
history is derived from land cover changes in the National Land Cover Dataset (Yang et al. 2018; Homer et al. 2007;
Fry et al. 2011; Homer et al. 2015).
Mineral Soil Carbon Stock Changes
An IPCC Tier 3 model-based approach (Ogle et al. 2010) is applied to estimate C stock changes from 1990 to 2015
for most mineral soils in Grassland Remaining Grassland. The C stock changes for the remaining soils are estimated
with an IPCC Tier 2 method (Ogle et al. 2003), including gravelly, cobbly, or shaley soils (greater than 35 percent by
volume), the additional stock changes associated with biosolids (i.e., treated sewage sludge) amendments, and
federal land.52
A surrogate data method is used to estimate soil organic C stock changes from 2016 to 2021 at the national scale
for land areas included in the Tier 2 and Tier 3 methods. Specifically, linear regression models with autoregressive
moving-average (ARMA) errors (Brockwell and Davis 2016) are used to estimate the relationship between
surrogate data and the 1990 to 2015 emissions data from the Tier 2 and 3 methods. Surrogate data for these
regression models are based on weather data from the PRISM Climate Group (PRISM Climate Group 2018). See
Box 6-4 in the Methodology section of Cropland Remaining Cropland for more information about the surrogate
data method.
Tier 3 Approach. Mineral soil organic C stocks and stock changes for Grassland Remaining Grassland are estimated
using the DayCent biogeochemical53 model (Parton et al. 1998; Del Grosso et al. 2001, 2011), as described in
Cropland Remaining Cropland. The DayCent model utilizes the soil C modeling framework developed in the
Century model (Parton et al. 1987,1988,1994; Metherell et al. 1993), but has been refined to simulate dynamics
at a daily time-step. Historical land-use patterns and irrigation histories are simulated with DayCent based on the
2015 USDA NRI survey (USDA-NRCS 2018).
The amount of manure produced by each livestock type is calculated for managed and unmanaged waste
management systems based on methods described in Section 5.2 Manure Management and Annex 3.11. Manure N
deposition from grazing animals (i.e., pasture/range/paddock (PRP) manure) is an input to the DayCent model to
estimate the influence of PRP manure on C stock changes for lands included in the Tier 3 method. Carbon stocks
52 Federal land is not a land use, but rather an ownership designation that is treated as grassland for purposes of these
calculations. The specific land use on federal lands is not identified in the NRI survey (USDA-NRCS 2018).
53 Biogeochemical cycles are the flow of chemical elements and compounds between living organisms and the physical
environment.
Land Use, Land-Use Change, and Forestry 6-79
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
and 95 percent confidence intervals are estimated for each year between 1990 and 2015 using the NRI survey
data. Further elaboration on the Tier 3 methodology and data used to estimate C stock changes from mineral soils
are described in Annex 3.12 of EPA (2022).
In order to ensure time-series consistency, the same methods are applied from 1990 to 2015 so that changes
reflect anthropogenic activity and not methodological adjustments. In addition, soil organic C stock changes from
2016 to 2021 are approximated using a linear extrapolation of emission patterns from 1990 to 2015. The
extrapolation is based on a linear regression model with moving-average (ARMA) errors, described in Box 6-4 of
the Methodology section in Cropland Remaining Cropland. Linear extrapolation is a standard data splicing method
for estimating emissions at the end of a time series (IPCC 2006). Stock change estimates for 2016 to 2021 will be
recalculated in future Inventories with an updated time series of activity data (see the Planned Improvements
section in Cropland Remaining Cropland).
Tier 2 Approach. The Tier 2 approach is based on the same methods described in the Tier 2 portion of Cropland
Remaining Cropland section for mineral soils, with the exception of the manure N deposition from grazing animals
(i.e., PRP manure), and the land use and management data that are used in the Inventory for federal grasslands.
First, the PRP N manure is included in the Tier 2 method that is not deposited on lands included in the Tier 3
method. Second, the NRI (USDA-NRCS 2018) provides land use and management histories for all non-federal lands,
and is the basis for the Tier 2 analysis for these areas. However, NRI does not provide land use information on
federal lands. The land use data for federal lands is based on the National Land Cover Database (NLCD) (Yang et al.
2018; Fry et al. 2011; Homer et al. 2007; Homer et al. 2015). In addition, the Bureau of Land Management (BLM)
manages some of the federal grasslands, and compiles information on grassland condition through the BLM
Rangeland Inventory (BLM 2014). To estimate soil organic C stock changes from federal grasslands, rangeland
conditions in the BLM data are aligned with IPCC grassland management categories of nominal, moderately
degraded, and severely degraded in order to apply the appropriate emission factors. Further elaboration on the
Tier 2 methodology and data used to estimate C stock changes from mineral soils are described in Annex 3.12 of
EPA (2022).
In order to ensure time-series consistency, the Tier 2 method is applied from 1990 to 2015 so that changes reflect
anthropogenic activity and not methodological adjustments. In addition, soil organic C stock changes are
approximated for the remainder of the time series with a linear extrapolation of emission patterns from 1990 to
2015. The extrapolation is based on a linear regression model with moving-average (ARMA) (See Box 6-4 of the
Methodology section in Cropland Remaining Cropland). Linear extrapolation is a standard data splicing method for
estimating emissions at the end of a time series (IPCC 2006). As with the Tier 3 method, time series of activity data
will be updated in a future Inventory, and emissions from 2016 to 2021 will be recalculated.
Additional Mineral C Stock Change Calculations
A Tier 2 method is used to adjust annual C stock change estimates for mineral soils between 1990 and 2021 to
account for additional C stock changes associated with biosolids (i.e., treated sewage sludge) amendments.
Estimates of the amounts of biosolids N applied to agricultural land are derived from national data on biosolids
generation, disposition, and N content (see Section 7.2, Wastewater Treatment for a detailed discussion of the
methodology for estimating treated sewage sludge available for land application application). Although biosolids
can be added to land managed for other land uses, it is assumed that agricultural amendments only occur in
Grassland Remaining Grassland. Total biosolids generation data for 1988,1996, and 1998, in dry mass units, are
obtained from EPA (1999) and estimates for 2004 are obtained from an independent national biosolids survey
(NEBRA 2007). These values are linearly interpolated to estimate values for the intervening years, and linearly
extrapolated to estimate values for years since 2004. Nitrogen application rates from Kellogg et al. (2000) are used
to determine the amount of area receiving biosolids amendments. The soil organic C storage rate is estimated at
0.38 metric tons C per hectare per year for biosolids amendments to grassland as described above. The stock
change rate is based on country-specific factors and the IPCC default method (see Annex 3.12 of EPA (2022) for
further discussion).
6-80 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Organic Soil Carbon Stock Changes
2 Annual C emissions from drained organic soils in Grassland Remaining Grassland are estimated using the Tier 2
3 method in IPCC (2006), which utilizes country-specific C loss rates (Ogle et al. 2003) rather than default IPCC rates.
4 For more information, see the Cropland Remaining Cropland section for organic soils and Annex 3.12 of EPA
5 (2022).
6 In order to ensure time-series consistency, the Tier 2 methods are applied from 1990 to 2015 so that changes
7 reflect anthropogenic activity and not methodological adjustments. In addition, soil organic C stock changes for the
8 remainder of the time series (i.e., 2016 to 2021) are approximated with a linear extrapolation of emission patterns
9 from 1990 to 2015. The extrapolation is based on a linear regression model with moving-average (ARMA) errors
10 (See Box 6-4 of the Methodology section in Cropland Remaining Cropland). Linear extrapolation is a standard data
11 splicing method for approximating emissions at the end of a time series (IPCC 2006). Estimates for 2016 to 2021
12 will be recalculated in future Inventories with an updated time series of activity data.
is Uncertainty
14 The uncertainty analysis for biomass, dead wood and litter C losses with woodlands is conducted in the same way
15 as the uncertainty assessment for forest ecosystem C flux associated with Forest Land Remaining Forest Land.
16 Sample and model-based error are combined using simple error propagation methods provided by the IPCC (2006)
17 by taking the square root of the sum of the squares of the standard deviations of the uncertain quantities. For
18 additional details, see the Uncertainty Analysis in Annex 3.13.
19 Uncertainty analysis for mineral soil organic C stock changes using the Tier 3 and Tier 2 methodologies are based
20 on a Monte Carlo approach that is described in the Cropland Remaining Cropland section and Annex 3.12 of EPA
21 (2022). The uncertainty for annual C emission estimates from drained organic soils in Grassland Remaining
22 Grassland is estimated using a Monte Carlo approach, which is also described in the Cropland Remaining Cropland
23 section. For 2016 to 2021, there is additional uncertainty propagated through the Monte Carlo Analysis associated
24 with the surrogate data method.
25 Uncertainty estimates are presented in Table 6-38 for each subcategory (i.e., soil organic C stocks for mineral and
26 organic soils) and the method applied in the Inventory analysis (i.e., Tier 2 and Tier 3). Uncertainty estimates from
27 the Tier 2 and 3 approaches are combined using the simple error propagation methods provided by the IPCC
28 (2006), i.e., by taking the square root of the sum of the squares of the standard deviations of the uncertain
29 quantities.
30 The combined uncertainty for soil organic C stocks in Grassland Remaining Grassland ranges from more than 1417
31 percent below and above the 2021 stock change estimate of 10.0 MMT CO2 Eq. The large relative uncertainty is
32 mostly due to high levels of uncertainty in the Tier 3 method and variation in soil organic C stock changes that is
33 not explained by the surrogate data method.
34 Table 6-38: Approach 2 Quantitative Uncertainty Estimates for C Stock Changes Occurring
35 Within Grassland Remaining Grassland (MMT CO2 Eq. and Percent)
Source
2021 Flux Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Flux Estimate3
(MMT C02 Eq.) (%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Woodland Biomass:
Aboveground live biomass 2.1 1.8 2.3 -12% 11%
Land Use, Land-Use Change, and Forestry 6-81
-------
Belowground live biomass
Dead wood
Litter
Mineral Soil C Stocks Grassland Remaining
Grassland, Tier 3 Methodology
Mineral Soil C Stocks: Grassland Remaining
Grassland, Tier 2 Methodology
Mineral Soil C Stocks: Grassland Remaining
Grassland, Tier 2 Methodology (Change in Soil
C due to Biosolids [i.e., Treated Sewage
Sludge] Amendments)
Organic Soil C Stocks: Grassland Remaining
Grassland, Tier 2 Methodology
Combined Uncertainty for Flux Associated
with Carbon Stock Changes Occurring in
Grassland Remaining Grassland 10.0 (131.7) 151.7 -1417% 1417%
+ Does not exceed 0.05 MMT C02 Eq.
a Range of C stock change estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net
sequestration.
1 Uncertainty is also associated with a lack of reporting on biomass, dead wood and litter C stock changes for
2 agroforestry systems. Changes in biomass and dead organic matter C stocks are assumed to be negligible in other
3 grasslands, largely comprised of herbaceous biomass, although there are certainly significant changes at sub-
4 annual time scales across seasons.
5 QA/QC and Verification
6 See the QA/QC and Verification section in Cropland Remaining Cropland.
7 Recalculations Discussion
8 Recalculations are associated with updated FIA data from 1990 to 2021 on biomass, dead wood and litter C stocks
9 in woodlands for Grassland Remaining Grassland, and updated estimates for mineral soils from 2016 to 2021 using
10 the linear extrapolation method. As a result of these new data, Grassland Remaining Grassland has a larger loss of
11 at 2.2 MMT CO2 Eq. compared to the previous Inventory, or 28 percent on average over the time series for
12 Grassland Remaining Grassland compared to the previous Inventory.
13 Planned Improvements
14 There are two key improvements planned for the Inventory, including a) incorporating the latest land use data
15 from the USDA National Resources Inventory, and b) conducting an analysis of C stock changes in Alaska for
16 grassland. While both improvements are needed, the latter improvement is a significant development that will
17 resolve the majority of the discrepancy between the managed land base for Grassland Remaining Grassland and
18 amount of area currently included in Grassland Remaining Grassland Inventory (see Table 6-39).
0.3 0.3 0.3 -4% 4%
3.0 2.6 3.4 -13% 14%
+ + 0.1 -20% 20%
1.8 (139.6) 143.2 -7961% 7961%
(0.9) (10.0) 8.1 -960% 960%
(1.7) (2.5) (0.8) -50% 50%
5.4 1.2 9.6 -77% 77%
6-82 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Table 6-39: Comparison of Managed Land Area in Grassland Remaining Grassland and the
2 Area in the current Grassland Remaining Grassland Inventory (Thousand Hectares)
Area (Thousand Hectares)
Year
Managed Land
Inventory
Difference
1990
328,320
277,406
50,914
1991
327,812
276,918
50,894
1992
327,355
276,422
50,933
1993
325,620
274,484
51,137
1994
324,006
272,813
51,194
1995
323,134
271,975
51,159
1996
322,284
271,123
51,160
1997
321,526
270,259
51,268
1998
319,596
268,174
51,422
1999
318,701
267,301
51,400
2000
317,690
266,202
51,488
2001
316,849
265,649
51,200
2002
316,455
265,192
51,263
2003
316,780
265,403
51,377
2004
316,810
265,421
51,389
2005
316,625
265,123
51,502
2006
316,344
264,804
51,540
2007
316,326
264,749
51,577
2008
316,496
264,878
51,618
2009
316,792
265,099
51,693
2010
316,652
264,942
51,711
2011
316,403
264,627
51,776
2012
316,294
264,413
51,881
2013
317,153
265,239
51,914
2014
318,024
266,180
51,844
2015
318,146
266,234
51,912
2016
318,513
*
*
2017
318,704
*
*
2018
321,748
*
*
2019
322,632
*
*
2020
323,883
*
*
2021
325,096
*
*
3 NRI data have not been incorporated into the inventory after 2015, designated with asterisks (*).
4 Additionally, a review of available data on biosolids (i.e., treated sewage sludge) application will be undertaken to
5 improve the distribution of biosolids application on croplands, grasslands and settlements. For information about
6 other improvements, see the Planned Improvements section in Cropland Remaining Cropland.
7 Non-C02 Emissions from Grassland Fires (CRF Source Category
a 4C1)
9 Fires are common in grasslands, and are thought to have been a key feature shaping the evolution of the grassland
10 vegetation in North America (Daubenmire 1968; Anderson 2004). Fires can occur naturally through lightning
11 strikes, but are also an important management practice to remove standing dead vegetation and improve forage
Land Use, Land-Use Change, and Forestry 6-83
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
for grazing livestock. Woody and herbaceous biomass will be oxidized in a fire, although in this section the current
focus is primarily on herbaceous biomass.54 Biomass burning emits a variety of trace gases including non-CC>2
greenhouse gases such as Cm and N2O, as well as CO and NOx that can become greenhouse gases when they react
with other gases in the atmosphere (Andreae and Merlet 2001). IPCC (2006) recommends reporting non-CC>2
greenhouse gas emissions from all wildfires and prescribed burning occurring in managed grasslands.
Biomass burning in grassland of the United States (Including burning emissions in Grassland Remaining Grassland
and Land Converted to Grassland) is a relatively small source of emissions, but it has increased by over 300 percent
since 1990. In 2021, Cm and N2O emissions from biomass burning in grasslands were 0.3 MMT CO2 Eq. (12 kt) and
0.3 MMT CO2 Eq. (1 kt), respectively. Annual emissions from 1990 to 2021 have averaged approximately 0.3 MMT
CO2 Eq. (12 kt) of CH4 and 0.3 MMT C02 Eq. (1 kt) of l\l20 (see Table 6-40 and Table 6-41).
Table 6-40: ChU and N2O Emissions from Biomass Burning in Grassland (MMT CO2 Eq.)
1990
2005
2017
2018
2019
2020
2021
ch4
0.1
0.4
0.3
0.3
0.3
0.3
0.3
n2o
0.1
0.3
0.3
0.3
0.3
0.3
0.3
Total Net Flux
0.2
0.7
0.6
0.6
0.6
0.6
0.6
Table 6-41: ChU, N2O, CO, and NOx Emissions from Biomass Burning in Grassland (kt)
1990
2005
2017
2018
2019
2020
2021
ch4
3
13
12
12
12
12
12
n2o
+
1
1
1
1
1
1
CO
84
358
345
331
341
334
339
NOx
5
21
21
20
20
20
20
+ Does not exceed 0.5 kt.
Methodology and Time-Series Consistency
The following section includes a description of the methodology used to estimate non-CC>2 greenhouse gas
emissions from biomass burning in grassland, including (1) determination of the land base that is classified as
managed grassland; (2) assessment of managed grassland area that is burned each year, and (3) estimation of
emissions resulting from the fires. For this Inventory, the IPCC Tier 1 method is applied to estimate non-CC>2
greenhouse gas emissions from biomass burning in grassland from 1990 to 2014 (IPCC 2006). A data splicing
method is used to estimate the emissions from 2015 to 2021, which is discussed later in this section.
The land area designated as managed grassland is based primarily on the National Resources Inventory (NRI)
(Nusser and Goebel 1997; USDA-NRCS 2015). NRI has survey locations across the entire United States, but does not
classify land use on federally-owned areas, and so survey locations on federal lands are designated as grassland
using land cover data from the National Land Cover Dataset (NLCD) (Fry et al. 2011; Homer et al. 2007; Homer et
al. 2015) (see Section 6.1 Representation of the U.S. Land Base).
The area of biomass burning in grasslands (Grassland Remaining Grassland and Land Converted to Grassland) is
determined using 30-m fire data from the Monitoring Trends in Burn Severity (MTBS) program for 1990 through
2014.55 NRI survey locations on grasslands are designated as burned in a year if there is a fire within 500 m of the
survey point according to the MTBS fire data. The area of biomass burning is estimated from the NRI spatial
weights and aggregated to the country (Table 6-42).
54 A planned improvement is underway to incorporate woodland tree biomass into the Inventory for non-C02 emissions from
grassland fires.
55 See http://www.mtbs.gov.
6-84 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
Table 6-42: Thousands of Grassland Hectares Burned Annually
Year
Thousand Hectares
1990
317
2005
1,343
2017
2018
2019
2020
2021
NE
NE
NE
NE
NE
Notes: Burned area was not estimated (NE) for 2015 to 2021 but will be updated
in a future Inventory. Burned area for the year 2014 is estimated to be 1,659
thousand hectares.
2 For 1990 to 2014, the total area of grassland burned is multiplied by the IPCC default factor for grassland biomass
3 (4.1 tonnes dry matter per ha) (IPCC 2006) to estimate the amount of combusted biomass. A combustion factor of
4 1 is assumed in this Inventory, and the resulting biomass estimate is multiplied by the IPCC default grassland
5 emission factors for CFU (2.3 g Cm per kg dry matter), N2O (0.21 g N2O per kg dry matter), CO (65 g CO per kg dry
6 matter) and NOx (3.9 g NOx per kg dry matter) (IPCC 2006). The Tier 1 analysis is implemented in the Agriculture
7 and Land Use National Greenhouse Gas Inventory (ALU) software (Ogle et al. 2016).56
8 A linear extrapolation of the trend in the time series is applied to estimate emissions for 2015 to 2021. Specifically,
9 a linear regression model with autoregressive moving-average (ARMA) errors (Brockwell and Davis 2016) is used to
10 derive the trend in emissions over time from 1990 to 2014, and the trend is used to approximate the 2015 to 2021
11 emissions. The Tier 1 method described previously will be applied to recalculate the 2015 to 2021 emissions in a
12 future Inventory.
13 The same methods are applied from 1990 to 2014, and a data splicing method is used to extend the time series
14 from 2015 to 2021 ensuring a consistent time series of emissions data. The trend extrapolation is a standard data
15 splicing method for estimating emissions at the end of a time series if activity data are not available (IPCC 2006).
17 Emissions are estimated using a linear regression model with ARMA errors for 2015 to 2021. The model produces
18 estimates for the upper and lower bounds of the emission estimate and the results are summarized in Table 6-43.
19 Methane emissions from Biomass Burning in Grassland for 2021 are estimated to be between approximately 0.0
20 and 0.8 MMT CO2 Eq. at a 95 percent confidence level. This indicates a range of 100 percent below and 138
21 percent above the 2021 emission estimate of 0.3 MMT CO2 Eq. Nitrous oxide emissions are estimated to be
22 between approximately 0.0 and 0.7 MMT CO2 Eq., or 100 percent below and 143 percent above the 2021 emission
23 estimate of 0.3 MMT CO2 Eq.
i6 Uncertainty
56 See http://www.nrel.colostate.edu/proiects/ALUsoftware/.
Land Use, Land-Use Change, and Forestry 6-85
-------
1 Table 6-43: Uncertainty Estimates for Non-C02 Greenhouse Gas Emissions from Biomass
2 Burning in Grassland (MMT CO2 Eq. and Percent)
3
Source
Gas
2021 Emission Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Grassland Burning
Grassland Burning
ch4
n2o
0.3
0.3
+ 0.8
+ 0.7
-100% +145%
-100% +145%
+ Does not exceed 0.05 MMT C02 Eq.
a Range of emission estimates predicted by linear regression time-series model for a 95 percent confidence interval.
4 Uncertainty is also associated with lack of reporting of emissions from biomass burning in grassland of Alaska.
5 Grassland burning emissions could be relatively large in this region of the United States, and therefore extending
6 this analysis to include Alaska is a planned improvement for the Inventory. There is also uncertainty due to lack of
7 reporting combustion of woody biomass, and this is another planned improvement.
8 QA/QC and Verification
9 Quality control measures included checking input data, model scripts, and results to ensure data are properly
10 handled throughout the inventory process. Inventory reporting forms and text are reviewed and revised as needed
11 to correct transcription errors. Quality control identified problems with input data for common reporting format
12 tables in the spreadsheets, which have been corrected.
13 Recalculations Discussion
14 EPA updated global warming potentials (GWP) for calculating the C02-equivalent emissions of CH4 (from 25 to 28)
15 and N20 (from 298 to 265) to reflect the 100-year GWPs provided in the IPCC Fifth Assessment Report (AR5) (IPCC
16 2013). The previous Inventory used 100-year GWPs provided in the IPCC Fourth Assessment Report (AR4). This
17 update was applied across the entire time series. As a result of this change, there was a net decrease in calculated
18 C02-equivalent emissions by an annual average of less than 0.05 MMT CO2 Eq., or 0.03 percent, over the time
19 series from 1990 to 2020 compared to the previous Inventory. Further discussion on this update and the overall
20 impacts of updating the inventory GWP values to reflect the AR5 can be found in Chapter 9, Recalculations and
21 Improvements.
22 Planned Improvements
23 A data splicing method is applied to estimate emissions in the latter part of the time series, which introduces
24 additional uncertainty in the emissions data. Therefore, a key improvement for the next Inventory will be to
25 update the time series with new activity data from the Monitoring Trends in Burn Severity program and recalculate
26 the emissions. Two other planned improvements have been identified for this source category, including a)
27 incorporation of country-specific grassland biomass factors, and b) extending the analysis to include Alaska. In the
28 current Inventory, biomass factors are based on a global default for grasslands that is provided by the IPCC (2006).
29 There is considerable variation in grassland biomass, however, which would affect the amount of fuel available for
30 combustion in a fire. Alaska has an extensive area of grassland and includes tundra vegetation, although some of
31 the areas are not managed. There has been an increase in fire frequency in boreal forest of the region (Chapin et
32 al. 2008), and this may have led to an increase in burning of neighboring grassland areas. There is also an effort
33 under development to incorporate grassland fires into DayCent model simulations. Lastly, a future Inventory will
34 incorporate non-C02 greenhouse emissions from burning woodland tree biomass in grasslands. These
35 improvements are expected to reduce uncertainty and produce more accurate estimates of non-C02 greenhouse
36 gas emissions from grassland burning.
6-86 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
6.7 Land Converted to Grassland (CRF
Category 4C2)
Land Converted to Grassland includes all grassland in an Inventory year that had been in another land use(s) during
the previous 20 years (USDA-NRCS 2018).57 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). Grassland includes pasture and rangeland that are used primarily but not
exclusively 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 interseeding of legumes.
Land use change can lead to large losses of C to the atmosphere, particularly conversions from forest land
(Houghton et al. 1983, Houghton and Nassikas 2017). Moreover, conversion of forest 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 according to a recent assessment (Tubiello et al. 2015).
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 Grassland, but there
is limited reporting of other pools in this Inventory. Losses of aboveground and belowground biomass, dead wood
and litter C from Forest Land Converted to Grassland are reported, as well as gains and losses associated with
conversions to woodlands58 from other land uses, including Croplands Converted to Grasslands, Settlements
Converted to Grasslands and Other Lands Converted to Grasslands. However, the current Inventory does not
include the gains and losses in aboveground and belowground biomass, dead wood and litter C for other land-use
conversions to grassland that are not woodlands.59
There are two discrepancies between the current land representation (See Section 6.1) and the area data that
have been used in the inventory for Land Converted to Grassland. First, the current land representation is based on
the latest NRI dataset, which includes data through 2017, but these data have not yet been incorporated into the
Land Converted to Grassland Inventory. Second, grassland in Alaska is not included in the Inventory. These
differences lead to discrepancies between the managed area in Land Converted to Grassland and the grassland
area included in the Land Converted to Grassland Inventory analysis (Table 6-47). Improvements are underway to
incorporate the latest NRI dataset, and grasslands in Alaska as part of future C inventories (See Planned
Improvements Section).
The largest C losses with Land Converted to Grassland are associated with aboveground biomass, belowground
biomass, and litter C losses from Forest Land Converted to Grassland (see Table 6-44 and Table 6-45). These three
pools led to net emissions in 2021 of 12.6, 2.2, and 4.8 MMT CO2 Eq. (3.4, 0.6, and 1.3 MMT C), respectively. In
contrast, land use and management of mineral soils in Land Converted to Grassland led to an increase in soil
organic C stocks, estimated at 42.6 MMT CO2 Eq. (11.6 MMT C) in 2021. The gains are primarily associated with
Other Land Converted to Grassland, and also due to Cropland Converted to Grassland, which leads to less intensive
57 NRI survey locations are classified according to land use histories starting in 1979, and consequently the classifications are
based on less than 20 years from 1990 to 2001. This may have led to an underestimation of Land Converted to Grassland in the
early part of the time series to the extent that some areas are converted to grassland between 1971 and 1978.
58 Woodlands are considered grasslands in the U.S. Land Representation because they do not meet the definition of Forest
Land.
59 Changes in biomass C stocks are not currently reported for other conversions to grassland (other than forest land conversion
to grassland and other land -use conversions to woodlands), but this is a planned improvement for a future Inventory. Note:
changes in dead organic matter are assumed negligible for other land-use conversions (i.e., other than forest land) to grassland
based on the Tier 1 method in IPCC (2006).
Land Use, Land-Use Change, and Forestry 6-87
-------
1 management of the soil. Drainage of organic soils for grassland management led to CO2 emissions to the
2 atmosphere of 1.8 MMT CO2 Eq. (0.5 MMT C). The total net C stock change in 2021 for Land Converted to
3 Grassland is estimated as a gain of 24.7 MMT CO2 Eq. (6.7 MMT C), which represents an increase in C stock change
4 of 269 percent compared to the initial reporting year of 1990.
5 Table 6-44: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for
6 Land Converted to Grassland (MMT CO2 Eq.)
1990
2005
2017
2018
2019
2020
2021
Cropland Converted to Grassland
(19.1)
(24.2)
(18.6)
(18.5)
(18.0)
(20.3)
(19.3)
Aboveground Live Biomass
(0.4)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
Belowground Live Biomass
(0.1)
(0.1)
+
+
+
+
+
Dead Wood
(0.2)
(0.2)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Litter
(0.2)
(0.2)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Mineral Soils
(18.9)
(25.0)
(19.4)
(19.3)
(18.7)
(21.0)
(20.1)
Organic Soils
0.6
1.5
1.4
1.3
1.3
1.3
1.3
Forest Land Converted to Grassland
20.1
20.2
19.7
19.7
19.6
19.6
19.6
Aboveground Live Biomass
13.3
13.1
12.6
12.6
12.6
12.6
12.6
Belowground Live Biomass
2.3
2.3
2.2
2.2
2.2
2.2
2.2
Dead Wood
(0.3)
(0.2)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Litter
4.8
4.9
4.8
4.8
4.8
4.8
4.8
Mineral Soils
(0.1)
(0.1)
+
+
(0.1)
+
+
Organic Soils
+
0.2
0.2
0.2
0.2
0.2
0.2
Other Lands Converted to Grassland
(7.2)
(34.5)
(24.6)
(24.4)
(24.0)
(24.3)
(24.0)
Aboveground Live Biomass
(1.6)
(1.5)
(1.3)
(1.3)
(1.3)
(1.3)
(1.3)
Belowground Live Biomass
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Dead Wood
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
Litter
(0.8)
(0.8)
(0.7)
(0.7)
(0.7)
(0.7)
(0.7)
Mineral Soils
(4.2)
(31.7)
(22.2)
(21.9)
(21.6)
(21.9)
(21.6)
Organic Soils
+
+
0.1
0.1
0.1
0.1
0.1
Settlements Converted to Grassland
(0.6)
(1.7)
(1.3)
(1.2)
(1.2)
(1.3)
(1.2)
Aboveground Live Biomass
(0.2)
(0.2)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Belowground Live Biomass
+
+
+
+
+
+
+
Dead Wood
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Litter
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Mineral Soils
(0.2)
(1.4)
(1.0)
(0.9)
(0.9)
(1.0)
(0.9)
Organic Soils
+
+
+
+
+
+
+
Wetlands Converted to Grassland
0.1
0.2
0.3
0.3
0.3
0.2
0.2
Mineral Soils
+
+
+
+
+
+
+
Organic Soils
0.1
0.2
0.2
0.2
0.2
0.2
0.2
Aboveground Live Biomass
11.1
11.1
11.0
11.0
10.9
10.9
10.9
Belowground Live Biomass
2.1
2.0
2.0
2.0
2.0
2.0
2.0
Dead Wood
(0.9)
(0.8)
(0.7)
(0.7)
(0.7)
(0.7)
(0.7)
Litter
3.7
3.8
3.9
3.9
3.9
3.9
3.9
Total Mineral Soil Flux
(23.4)
(58.2)
(42.5)
(42.2)
(41.3)
(43.9)
(42.6)
Total Organic Soil Flux
0.8
1.9
1.9
1.9
1.8
1.8
1.8
Total Net Flux
(6.7)
(40.1)
(24.5)
(24.2)
(23.3)
(25.9)
(24.7)
7 + Does not exceed 0.05 MMT C02 Eq.
8 Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
9 Table 6-45: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for
10 Land Converted to Grassland (MMT C)
1990
2005
2017
2018
2019
2020
2021
Cropland Converted to Grassland
(5.2)
(6.6)
(5.1)
(5.1)
(4.9)
(5.5)
(5.3)
Aboveground Live Biomass
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Belowground Live Biomass
+
+
+
+
+
+
+
6-88 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Dead Wood
+
+
+
+
+
+
+
Litter
(0.1)
+
+
+
+
+
+
Mineral Soils
(5.2)
(6.8)
(5.3)
(5.3)
(5.1)
(5.7)
(5.5)
Organic Soils
0.2
0.4
0.4
0.4
0.4
0.4
0.4
Forest Land Converted to Grassland
5.5
5.5
5.4
5.4
5.4
5.4
5.4
Aboveground Live Biomass
3.6
3.6
3.4
3.4
3.4
3.4
3.4
Belowground Live Biomass
0.6
0.6
0.6
0.6
0.6
0.6
0.6
Dead Wood
(0.1)
(0.1)
+
+
+
+
+
Litter
1.3
1.3
1.3
1.3
1.3
1.3
1.3
Mineral Soils
+
+
+
+
+
+
+
Organic Soils
+
+
0.1
0.1
0.1
0.1
0.1
Other Lands Converted to Grassland
(2.0)
(9.4)
(6.7)
(6.6)
(6.5)
(6.6)
(6.5)
Aboveground Live Biomass
(0.4)
(0.4)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
Belowground Live Biomass
(0.1)
+
+
+
+
+
+
Dead Wood
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Litter
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Mineral Soils
(1.2)
(8.6)
(6.1)
(6.0)
(5.9)
(6.0)
(5.9)
Organic Soils
+
+
+
+
+
+
+
Settlements Converted to Grassland
(0.2)
(0.5)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
Aboveground Live Biomass
+
+
+
+
+
+
+
Belowground Live Biomass
+
+
+
+
+
+
+
Dead Wood
+
+
+
+
+
+
+
Litter
+
+
+
+
+
+
+
Mineral Soils
+
(0.4)
(0.3)
(0.3)
(0.2)
(0.3)
(0.3)
Organic Soils
+
+
+
+
+
+
+
Wetlands Converted to Grassland
+
0.1
0.1
0.1
0.1
0.1
0.1
Mineral Soils
+
+
+
+
+
+
+
Organic Soils
+
0.1
0.1
0.1
0.1
0.1
0.1
Aboveground Live Biomass
3.0
3.0
3.0
3.0
3.0
3.0
3.0
Belowground Live Biomass
0.6
0.6
0.5
0.5
0.5
0.5
0.5
Dead Wood
(0.3)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Litter
1.0
1.0
1.1
1.1
1.1
1.1
1.1
Total Mineral Soil Flux
(6.4)
(15.9)
(11.6)
(11.5)
(11.3)
(12.0)
(11.6)
Total Organic Soil Flux
0.2
0.5
0.5
0.5
0.5
0.5
0.5
Total Net Flux
(1.8)
(10.9)
(6.7)
(6.6)
(6.4)
(7.1)
(6.7)
1 + Does not exceed 0.05 MMT C.
2 Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
3 Methodology and Time-Series Consistency
4 The following section includes a description of the methodology used to estimate C stock changes for Land
5 Converted to Grassland, including (1) loss of aboveground and belowground biomass, dead wood and litter C with
6 Forest Land Converted to Grassland and other land-use conversions to woodlands, as well as (2) the impact from
7 all land-use conversions to grassland on mineral and organic soil organic C stocks.
8 Biomass, Dead Wood, and Litter Carbon Stock Changes
9 A Tier 3 method is applied to estimate biomass, dead wood and litter C stock changes for Forest Land Converted to
10 Grassland and other land-use conversions to woodlands (i.e., Croplands Converted to Grasslands, Settlements
11 Converted to Grasslands and Other Lands Converted to Grasslands). Estimates are calculated in the same way as
12 those in the Forest Land Remaining Forest Land category using data from the USDA Forest Service, Forest
13 Inventory and Analysis (FIA) program (USDA Forest Service 2022). There are limited data on the herbaceous
14 grassland C stocks following conversion so default biomass estimates (IPCC 2006) for grasslands are used to
15 estimate C stock changes (Note: litter and dead wood C stocks are assumed to be zero following conversion
Land Use, Land-Use Change, and Forestry 6-89
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
because no reference C density estimates exist for grasslands). The difference between the stocks is reported as
the stock change under the assumption that the change occurred in the year of the conversion.
The amount of biomass C that is lost abruptly with Forest Land Converted to Grasslands is estimated based on the
amount of C before conversion and the amount of C following conversion according to remeasurements in the FIA
program. This approach is consistent with IPCC (2006) that assumes there is an abrupt change during the first year,
but does not necessarily capture the slower change over the years following conversion until a new steady state is
reached. It was determined that using an IPCC Tier I approach that assumes all C is lost in the year of conversion
for Forest Land Converted to Grasslands in the West and Great Plains states does not accurately characterize the
transfer of C in woody biomass during abrupt or gradual land-use change. To estimate this transfer of C in woody
biomass, state-specific C densities for woody biomass remaining on these former forest lands following conversion
to grasslands were developed and included in the estimation of C stock changes from Forest Land Converted to
Grasslands in the West and Great Plains states. A review of the literature in grassland and rangeland ecosystems
(Asner et al. 2003; Huang et al. 2009; Tarhouni et al. 2016), as well as an analysis of FIA data, suggests that a
conservative estimate of 50 percent of the woody biomass C density was lost during conversion from Forest Land
to Grasslands. This estimate was used to develop state-specific C density estimates for biomass, dead wood, and
litter for Grasslands in the West and Great Plains states and these state-specific C densities were applied in the
compilation system to estimate the C losses associated with conversion from forest land to grassland in the West
and Great Plains states. Further, losses from forest land to what are characterized as woodlands are included in
this category using FIA plot re-measurements and the methods and models described hereafter.
If FIA plots include data on individual trees, aboveground and belowground C density estimates are based on
Woodall et al. (2011). Aboveground and belowground biomass estimates also include live understory which is a
minor component of biomass defined as all biomass of undergrowth plants in a forest, including woody shrubs and
trees less than 2.54 cm dbh. For this Inventory, it was assumed that 10 percent of total understory C mass is
belowground (Smith et al. 2006). Estimates of C density are based on information in Birdsey (1996) and biomass
estimates from Jenkins et al. (2003).
If FIA plots include data on standing dead trees, standing dead tree C density is estimated following the basic
method applied to live trees (Woodall et al. 2011) with additional modifications to account for decay and structural
loss (Domke et al. 2011; Harmon et al. 2011). If FIA plots include data on downed dead wood, downed dead wood
C density is estimated based on measurements of a subset of FIA plots for downed dead wood (Domke et al. 2013;
Woodall and Monleon 2008). Downed dead wood is defined as pieces of dead wood greater than 7.5 cm diameter
that are not attached to live or standing dead trees at transect intersection. This includes stumps and roots of
harvested trees. To facilitate the downscaling of downed dead wood C estimates from the state-wide population
estimates to individual plots, downed dead wood models specific to regions and forest types within each region
are used. Litter C is the pool of organic C (also known as duff, humus, and fine woody debris) above the mineral
soil and includes woody fragments with diameters of up to 7.5 cm. A subset of FIA plots is measured for litter C. If
FIA plots include litter material, a modeling approach using litter C measurements from FIA plots is used to
estimate litter C density (Domke et al. 2016). The same methods are applied from 1990 to 2021 in order to ensure
time-series consistency. See Annex 3.13 for more information about reference C density estimates for forest land.
See the Grassland Remaining Grassland section for more information about estimation of biomass, deadwood and
litter C stock changes for woodlands.
Soil Carbon Stock Changes
Soil organic C stock changes are estimated for Land Converted to Grassland according to land use histories
recorded in the 2015 USDA NRI survey for non-federal lands (USDA-NRCS 2018). Land use and some management
information (e.g., crop type, soil attributes, and irrigation) were originally collected for each NRI survey locations
on a 5-year cycle beginning in 1982. In 1998, the NRI Program began collecting annual data, and the annual data
are currently available through 2017, however this Inventory uses the previous NRI with annual data through 2015
(USDA-NRCS 2018). NRI survey locations are classified as Land Converted to Grassland in a given year between
1990 and 2015 if the land use is grassland but had been classified as another use during the previous 20 years. NRI
6-90 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
survey locations are classified according to land use histories starting in 1979, and consequently the classifications
are based on less than 20 years from 1990 to 1998. This may have led to an underestimation of Land Converted to
Grassland in the early part of the time series to the extent that some areas are converted to grassland between
1971 and 1978. For federal lands, the land use history is derived from land cover changes in the National Land
Cover Dataset (Yang et al. 2018; Homer et al. 2007; Fry et al. 2011; Homer et al. 2015).
Mineral Soil Carbon Stock Changes
An IPCC Tier 3 model-based approach (Ogle et al. 2010) is applied to estimate C stock changes in mineral soils for
most of the area in Land Converted to Grassland. C stock changes on the remaining area are estimated with an
IPCC Tier 2 approach (Ogle et al. 2003), including prior cropland used to produce vegetables, tobacco, and
perennial/horticultural crops; land areas with very gravelly, cobbly, or shaley soils (greater than 35 percent by
volume); and land converted to grassland from another land use other than cropland.
A surrogate data method is used to estimate soil organic C stock changes from 2016 to 2021 at the national scale
for land areas included in the Tier 2 and Tier 3 methods. Specifically, linear regression models with autoregressive
moving-average (ARMA) errors (Brockwell and Davis 2016) are used to estimate the relationship between
surrogate data and the 1990 to 2015 emissions data that are derived using the Tier 2 and 3 methods. Surrogate
data for these regression models includes weather data from the PRISM Climate Group (PRISM Climate Group
2018). See Box 6-4 in the Methodology section of Cropland Remaining Cropland for more information about the
surrogate data method.
Tier 3 Approach. Mineral soil organic C stocks and stock changes are estimated using the DayCent
biogeochemical60 model (Parton et al. 1998; Del Grosso et al. 2001, 2011). The DayCent model utilizes the soil C
modeling framework developed in the Century model (Parton et al. 1987,1988,1994; Metherell et al. 1993), but
has been refined to simulate dynamics at a daily time-step. Historical land use patterns and irrigation histories are
simulated with DayCent based on the 2015 USDA NRI survey (USDA-NRCS 2018). Carbon stocks and 95 percent
confidence intervals are estimated for each year between 1990 and 2015. See the Cropland Remaining Cropland
section and Annex 3.12 for additional discussion of the Tier 3 methodology for mineral soils.
In order to ensure time-series consistency, the same methods are applied from 1990 to 2015 so that changes
reflect anthropogenic activity and not methodological adjustments. In addition, soil organic C stock changes from
2016 to 2021 are approximated using a linear extrapolation of emission patterns from 1990 to 2015. The
extrapolation is based on a linear regression model with moving-average (ARMA) errors, described in 6.4 of the
Methodology section in Cropland Remaining Cropland. Linear extrapolation is a standard data splicing method for
estimating emissions at the end of a time series (IPCC 2006). Stock change estimates for 2016 to 2021 will be
recalculated in future Inventories with an updated time series of activity data (see the Planned Improvements
section in Cropland Remaining Cropland).
Tier 2 Approach. For the mineral soils not included in the Tier 3 analysis, soil organic C stock changes are estimated
using a Tier 2 Approach, as described in the Tier 2 Approach for mineral soils in Grassland Remaining Grassland and
Annex 3.12. In order to ensure time-series consistency, the Tier 2 method is applied from 1990 to 2015 so that
changes reflect anthropogenic activity and not methodological adjustments. In addition, soil organic C stock
changes are approximated for the remainder of the time series with a linear extrapolation of emission patterns
from 1990 to 2015. The extrapolation is based on a linear regression model with moving-average (ARMA) (See Box
6-4 of the Methodology section in Cropland Remaining Cropland). Linear extrapolation is a standard data splicing
method for estimating emissions at the end of a time series (IPCC 2006). As with the Tier 3 method, stock change
estimates for 2016 to 2021 will be recalculated in future Inventories with an updated time series of activity data.
60 Biogeochemical cycles are the flow of chemical elements and compounds between living organisms and the physical
environment.
Land Use, Land-Use Change, and Forestry 6-91
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Organic Soil Carbon Stock Changes
Annual C emissions from drained organic soils in Land Converted to Grassland are estimated using the Tier 2
method provided in IPCC (2006), with country-specific C loss rates (Ogle et al. 2003) as described in the Cropland
Remaining Cropland section. Further elaboration on the methodology is also provided in Annex 3.12 for organic
soils.
In order to ensure time-series consistency, the Tier 2 method is applied from 1990 to 2015 so that changes reflect
anthropogenic activity and not methodological adjustments. In addition, soil organic C stock changes are
approximated for the remainder of the time series with a linear extrapolation of emission patterns from 1990 to
2015. The extrapolation is based on a linear regression model with moving-average (ARMA) (See Box 6-4 of the
Methodology section in Cropland Remaining Cropland). Linear extrapolation is a standard data splicing method for
estimating emissions at the end of a time series (IPCC 2006). Annual C emissions from drained organic soils from
2016 to 2021 will be recalculated in future Inventories with an updated time series of activity data.
Uncertainty
The uncertainty analyses for biomass, dead wood and litter C losses with Forest Land Converted to Grassland and
other land-use conversions to woodlands are conducted in the same way as the uncertainty assessment for forest
ecosystem C flux in the Forest Land Remaining Forest Land category. Sample and model-based error are combined
using simple error propagation methods provided by the IPCC (2006), by taking the square root of the sum of the
squares of the standard deviations of the uncertain quantities. For additional details see the Uncertainty Analysis
in Annex 3.13.
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 described in the Cropland Remaining Cropland section and Annex 3.12.
The uncertainty for annual C emission estimates from drained organic soils in Land Converted to Grassland is
estimated using a Monte Carlo approach, which is also described in the Cropland Remaining Cropland section. For
2016 to 2021, there is additional uncertainty propagated through the Monte Carlo Analysis associated with a
surrogate data method, which is also described in Cropland Remaining Cropland.
Uncertainty estimates are presented in Table 6-46 for each subsource (i.e., biomass C stocks, mineral and organic C
stocks in soils) and the method applied in the inventory analysis (i.e., Tier 2 and Tier 3). Uncertainty estimates from
the Tier 2 and 3 approaches are combined using the simple error propagation methods provided by the IPCC
(2006), as discussed in the previous paragraph. The combined uncertainty for total C stocks in Land Converted to
Grassland ranges from 149 percent below to 149 percent above the 2021 stock change estimate of 24.7 MMT CO2
Eq. The large relative uncertainty around the 2021 stock change estimate is partly due to large uncertainties in
biomass and dead organic matter C losses with Forest Land Conversion to Grassland, in addition to variation in soil
organic C stock changes that is not explained by the surrogate data method.
Table 6-46: Approach 2 Quantitative Uncertainty Estimates for Soil, Dead Organic Matter
and Biomass C Stock Changes occurring within Land Converted to Grassland (MMT CO2 Eq.
and Percent)
2021 Flux Estimate3 Uncertainty Range Relative to Flux Estimate3
(MMT C02 Eq.) (MMT CP2 Eq.) (%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Cropland Converted to Grassland
(19.3)
(48.9)
10.3
-153%
153%
Aboveground Live Biomass
(0.3)
(0.6)
0.1
-129%
129%
Belowground Live Biomass
+
(0.1)
+
-167%
100%
Dead Wood
(0.1)
(0.3)
+
-133%
129%
Litter
(0.1)
(0.3)
+
-114%
127%
Mineral Soil C Stocks: Tier 3
(16.2)
(45.6)
13.1
-181%
181%
6-92 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Mineral Soil C Stocks: Tier 2
(3.8)
(7.2)
(0.5)
-88%
88%
Organic Soil C Stocks: Tier 2
1.3
(0.1)
2.7
-105%
105%
Forest Land Converted to Grassland
19.6
5.4
33.9
-73%
73%
Aboveground Live Biomass
12.6
(0.6)
25.7
-104%
104%
Belowground Live Biomass
2.2
(0.1)
4.5
-105%
105%
Dead Wood
(0.1)
+
+
-100%
117%
Litter
4.8
(0.2)
9.9
-105%
104%
Mineral Soil C Stocks: Tier 2
+
(0.2)
0.1
-324%
324%
Organic Soil C Stocks: Tier 2
0.2
+
0.4
-119%
119%
Other Lands Converted to Grassland
(24.0)
(40.5)
(7.5)
-69%
69%
Aboveground Live Biomass
(1.3)
(2.1)
(0.5)
-63%
62%
Belowground Live Biomass
(0.2)
(0.3)
(0.1)
-68%
52%
Dead Wood
(0.4)
(0.6)
(0.1)
-66%
61%
Litter
(0.7)
(1.1)
(0.3)
-62%
63%
Mineral Soil C Stocks: Tier 2
(21.6)
(38.0)
(5.1)
-76%
76%
Organic Soil C Stocks: Tier 2
0.1
+
0.2
-163%
163%
Settlements Converted to Grassland
(1.2)
(2.0)
(0.5)
-61%
62%
Aboveground Live Biomass
(0.1)
(0.2)
+
-61%
73%
Belowground Live Biomass
+
+
+
-108%
100%
Dead Wood
(0.1)
(0.1)
+
-42%
29%
Litter
(0.1)
(0.1)
+
-46%
63%
Mineral Soil C Stocks: Tier 2
(0.9)
(1.7)
(0.2)
-80%
80%
Organic Soil C Stocks: Tier 2
+
+
+
-289%
289%
Wetlands Converted to Grasslands
0.2
+
0.5
-120%
120%
Mineral Soil C Stocks: Tier 2
+
(0.1)
0.1
-933%
933%
Organic Soil C Stocks: Tier 2
0.2
+
0.5
-119%
119%
Total: Land Converted to Grassland
(24.7)
(61.4)
12.1
-149%
149%
Aboveground Live Biomass
10.9
(2.2)
24.1
-120%
120%
Belowground Live Biomass
2.0
(0.3)
4.3
-116%
116%
Dead Wood
(0.7)
(1.0)
(0.4)
-48%
47%
Litter
3.9
(1.2)
9.0
-130%
130%
Mineral Soil C Stocks: Tier 3
(16.2)
(45.6)
13.1
-181%
181%
Mineral Soil C Stocks: Tier 2
(26.4)
(43.2)
(9.6)
-64%
64%
Organic Soil C Stocks: Tier 2
1.8
0.4
3.2
-79%
79%
+ Absolute value does not exceed 0.05 MMT C02 Eq.
a Range of C stock change estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
1 Uncertainty is also associated with a lack of reporting on biomass, dead wood and litter C stock changes for
2 conversions to agroforestry systems and herbaceous grasslands. The influence of agroforestry is difficult to address
3 because there are currently no datasets to evaluate the trends in the area and associated C stocks in agroforestry
4 systems. The influence of land-use change to herbaceous grasslands and agroforestry will be further explored in a
5 future Inventory.
6 QA/QC and Verification
7 See the QA/QC and Verification section in Cropland Remaining Cropland and Grassland Remaining Grassland for
8 information on QA/QC steps.
9 Recalculations Discussion
10 Recalculations are associated with new FIA data from 1990 to 2021 on biomass, dead wood and litter C stocks
11 associated with conversions to woodlands from Cropland Converted to Grassland, Other Land Converted to
12 Grassland, and Settlements Converted to Grassland; updated FIA data from 1990 to 2021 on biomass, dead wood
13 and litter C stocks from Forest Land Converted to Grassland; and updated estimates for mineral soils from 2016 to
Land Use, Land-Use Change, and Forestry 6-93
-------
1 2021 using the linear extrapolation method. As a result, Land Converted to Grassland has an estimated increase in
2 C stock changes of 2.9 MMT CO2 Eq. on average over the time series, representing a 23 percent increase in C
3 sequestration compared to the previous Inventory.
5 There are two key improvements planned for the inventory, including a) incorporating the latest land use data
6 from the USDA National Resources Inventory, and b) conducting an analysis of C stock changes in Alaska for
7 cropland. These two improvements will resolve the majority of the discrepancy between the managed land base
8 for Land Converted to Grassland and amount of area currently included in Land Converted to Grassland Inventory
9 (See Table 6.47).
10 Table 6-47: Comparison of Managed Land Area in Land Converted to Grassland and Area in
11 the current Land Converted to Grassland Inventory (Thousand Hectares)
4
Planned Improvements
Area (Thousand Hectares)
Year
Managed Land
Inventory
Difference
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
9,319
9,514
9,733
11,641
13,391
14,060
14,749
15,431
19,309
20,164
21,295
22,387
22,863
22,495
23,164
23,070
23,409
23,144
23,448
23,339
23,415
23,557
23,383
22,196
20,856
20,811
20,083
19,349
16,517
16,090
15,254
13,892
9,394
9,485
9,691
11,566
13,378
13,994
14,622
15.162
19,052
19,931
20,859
21,968
22,392
22,008
22,547
22,447
22,702
22,428
22,661
22,581
22,634
22,750
22,596
21,439
20.163
20,210
-75
29
43
75
14
66
127
269
258
234
436
418
471
487
617
622
707
716
787
758
780
806
787
757
693
601
6-94 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
NRI data have not been incorporated into the inventory after 2015,
designated with asterisks (*).
1 In addition, the amount of biomass C that is lost abruptly or the slower changes that continue to occur over a
2 decade or longer with Forest Land Converted to Grasslands will be further refined in a future Inventory. The
3 current values are estimated based on the amount of C before conversion and an estimated level of C left after
4 conversion based on limited plot data from the FIA and published literature for the Western United States and
5 Great Plains Regions. The amount of C left after conversion will be further investigated with additional data
6 collection, particularly in the Western United States and Great Plains, including tree biomass, understory biomass,
7 dead wood and litter C pools. In addition, biomass C stock changes will be estimated for conversions from other
8 land uses to herbaceous grasslands. For information about other improvements, see the Planned Improvements
9 section in Cropland Remaining Cropland.
10
u 6.8 Wetlands Remaining Wetlands (CRF
12 Category 4D1)
13 Wetlands Remaining Wetlands includes all wetlands in an Inventory year that have been classified as a wetland for
14 the previous 20 years, and in this Inventory, the flux estimates include Peatlands, Coastal Wetlands, and Flooded
15 Land.
is Peatlands Remaining Peatlands
17 Emissions from Managed Peatlands
18 Managed peatlands are peatlands that have been cleared and drained for the production of peat. The production
19 cycle of a managed peatland has three phases: land conversion in preparation for peat extraction (e.g., clearing
20 surface biomass, draining), extraction (which results in the emissions reported under Peatlands Remaining
21 Peatlands), and abandonment, restoration, rewetting, or conversion of the land to another use.
22 Carbon dioxide emissions from the removal of biomass and the decay of drained peat constitute the major
23 greenhouse gas flux from managed peatlands. Managed peatlands may also emit Cm and N2O. The natural
24 production of Cm is largely reduced but not entirely eliminated when peatlands are drained in preparation for
25 peat extraction (Strack et al. 2004 as cited in the 2006IPCC Guidelines). Drained land surface and ditch networks
26 contribute to the Cm flux in peatlands managed for peat extraction. Methane emissions were considered
27 insignificant under the IPCC Tier 1 methodology (IPCC 2006), but are included in the emissions estimates for
28 Peatlands Remaining Peatlands consistent with the 2013 Supplement to the 2006 IPCC Guidelines for National
29 Greenhouse Gas Inventories: Wetlands (IPCC 2013). Nitrous oxide emissions from managed peatlands depend on
30 site fertility. In addition, abandoned and restored peatlands continue to release greenhouse gas emissions.
31 Although methodologies are provided to estimate emissions and removals from rewetted organic soils (which
32 includes rewetted/restored peatlands) in IPCC (2013) guidelines, information on the areal extent of
33 rewetted/restored peatlands in the United States is currently unavailable. The current Inventory estimates CO2,
34 Cm and N2O emissions from peatlands managed for peat extraction in accordance with IPCC (2006 and 2013)
35 guidelines.
Land Use, Land-Use Change, and Forestry 6-95
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
C02, N2O, and CH4 Emissions from Peatlands Remaining Peatlands
IPCC (2013) recommends reporting CO2, N2O, and Cm emissions from lands undergoing active peat extraction (i.e.,
Peatlands Remaining Peatlands) as part of the estimate for emissions from managed wetlands. Peatlands occur
where plant biomass has sunk to the bottom of water bodies and water-logged areas and exhausted the oxygen
supply below the water surface during the course of decay. Due to these anaerobic conditions, much of the plant
matter does not decompose but instead forms layers of peat over decades and centuries. In the United States,
peat is extracted for horticulture and landscaping growing media, and for a wide variety of industrial, personal
care, and other products. It has not been used for fuel in the United States for many decades. Peat is harvested
from two types of peat deposits in the United States: Sphagnum bogs in northern states (e.g., Minnesota) and
wetlands in states further south (e.g., Florida). The peat from Sphagnum bogs in northern states, which is nutrient-
poor, is generally corrected for acidity and mixed with fertilizer. Production from more southerly states is relatively
coarse (i.e., fibrous) but nutrient-rich.
IPCC (2006 and 2013) recommend considering both on-site and off-site emissions when estimating CO2 emissions
from Peatlands Remaining Peatlands using the Tier 1 approach. Current IPCC methodologies estimate only on-site
N2O and Cm emissions. This is because off-site N2O estimates are complicated by the risk of double-counting
emissions from nitrogen fertilizers added to horticultural peat where subsequent runoff or leaching into
waterbodies can result in indirect N2O emissions that are already included within the Agricultural Soil Management
category.
On-site emissions from managed peatlands occur as the land is drained and cleared of vegetation, and the
underlying peat is exposed to sun, weather and oxygen. As this occurs, some peat deposit is lost and CO2 is emitted
from the oxidation of the peat. Since N2O emissions from saturated ecosystems tend to be low unless there is an
exogenous source of nitrogen, N2O emissions from drained peatlands are dependent on nitrogen mineralization
and therefore on soil fertility. Peatlands located on highly fertile/nutrient-rich soils, mostly made up of southern
peatlands in Florida, contain significant amounts of organic nitrogen in inert/microbially inaccessible forms.
Draining land in preparation for peat extraction allows bacteria to convert the organic nitrogen into nitrates
through nitrogen mineralization which leach to the surface where they are reduced to N2O during nitrification.
Nitrate availability also contributes to the activity of methanogens and methanotrophs that result in CH4 emissions
(Blodau 2002; Treat et al. 2007 as cited in IPCC 2013). Drainage ditches, which are constructed to drain the land in
preparation for peat extraction, also contribute to the flux of CH4 through in situ production and lateral transfer of
CH4 from the organic soil matrix (IPCC 2013).
Off-site CO2 emissions from managed peatlands occur from waterborne dissolved organic carbon losses and the
horticultural and landscaping use of peat. Dissolved organic carbon from water drained off peatlands reacts within
aquatic ecosystems and is converted to CO2, which is then emitted to the atmosphere (Billet et al. 2004 as cited in
IPCC 2013). During the horticultural and landscaping use of peat, nutrient-poor (but fertilizer-enriched) peat tends
to be used in bedding plants and in greenhouse and plant nursery production, whereas nutrient-rich (but relatively
coarse) peat is used directly in landscaping, athletic fields, golf courses, and plant nurseries. Most (nearly 94
percent) of the CO2 emissions from peat occur off-site, as the peat is processed and sold to firms which, in the
United States, use it predominantly for the aforementioned horticultural and landscaping purposes.
Total emissions from Peatlands Remaining Peatlands are estimated to be 0.7 MMT CO2 Eq. in 2021 (see Table 6-48
and Table 6-49) comprising 0.7 MMT C02 Eq. (700 kt) of C02, 0.004 MMT C02 Eq. (0.15 kt) of CH4 and 0.0005 MMT
CO2 Eq. (0.002 kt) of N2O. Total emissions in 2021 are 4.5 percent less than total emissions in 2020.
Total emissions from Peatlands Remaining Peatlands have fluctuated between 0.7 and 1.3 MMT CO2 Eq. across the
time series with a decreasing trend from 1990 until 1993, followed by an increasing trend until reaching peak
emissions in 2000. After 2000, emissions generally decreased until 2006 and then increased until 2009. The trend
reversed in 2009 and total emissions have generally decreased between 2009 and 2021. Carbon dioxide emissions
from Peatlands Remaining Peatlands have fluctuated between 0.7 and 1.3 MMT CO2 across the time series, and
these emissions drive the trends in total emissions. Methane and N2O emissions remained close to zero across the
time series.
6-96 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Table 6-48: Emissions from Peatlands Remaining Peatlands (MMT CO2 Eq.)
Gas
1990
2005
2017
2018
2019
2020
2021
CO?
1.1
1.1
0.8
0.8
0.8
0.7
0.7
Off-site
1.0
1.0
0.8
0.7
0.7
0.7
0.7
On-site
0.1
0.1
0.1
0.1
0.1
+
+
CH4 (On-site)
+
+
+
+
+
+
+
N20 (On-site)
+
+
+
+
+
+
+
Total
1.1
1.1
0.8
0.8
0.8
0.7
0.7
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Table 6-49: Emissions from Peatlands Remaining Peatlands (kt)
Gas
1990
2005
2017
2018
2019
2020
2021
C02
1,055
1,101
829
795
757
733
700
Off-site
985
1,030
774
744
707
683
653
On-site
70
71
55
51
50
50
48
CH4 (On-site)
+
+
+
+
+
+
+
N20 (On-site)
+
+
+
+
+
+
+
+ Does not exceed 0.5 kt
Note: Totals by gas may not sum due to independent rounding.
Methodology and Time-Series Consistency
Off-Site CO2 Emissions
Carbon dioxide emissions from domestic peat production were estimated using a Tier 1 methodology consistent
with IPCC (2006). Off-site CO2 emissions from Peatlands Remaining Peatlands were calculated by apportioning the
annual weight of peat produced in the United States (Table 6-50) into peat extracted from nutrient-rich deposits
and peat extracted from nutrient-poor deposits using annual percentage-by-weight figures. These nutrient-rich
and nutrient-poor production values were then multiplied by the appropriate default C fraction conversion factor
taken from IPCC (2006) in order to obtain off-site emission estimates. For the conterminous 48 states, both annual
percentages of peat type by weight and domestic peat production data were sourced from estimates and industry
statistics provided in the Minerals Yearbook and Mineral Commodity Summaries from the U.S. Geological Survey
(USGS 1995 through 2018; USGS 2022a; USGS 2022b; USGS 2022c). Hawaii is assumed to have no peat production
due to its absence from these sources. To develop these data, the U.S. Geological Survey (USGS; U.S. Bureau of
Mines prior to 1997) obtained production and use information by surveying domestic peat producers. On average,
about 75 percent of the peat operations respond to the survey; USGS estimates data for non-respondents on the
basis of prior-year production levels (Apodaca 2011).
The estimates for Alaska rely on reported peat production from the annual Alaska's Mineral Industry reports
(DGGS 1993 through 2015). Similar to the U.S. Geological Survey, the Alaska Department of Natural Resources,
Division of Geological & Geophysical Surveys (DGGS) solicits voluntary reporting of peat production from producers
for the Alaska's Mineral Industry report. However, the report does not estimate production for the non-reporting
producers, resulting in larger inter-annual variation in reported peat production from Alaska depending on the
number of producers who report in a given year (Szumigala 2011). In addition, in both the conterminous 48 states
and Alaska, large variations in peat production can also result from variation in precipitation and the subsequent
changes in moisture conditions, since unusually wet years can hamper peat production. The methodology
estimates emissions from Alaska separately from the conterminous 48 states because Alaska previously conducted
its own mineral surveys and reported peat production by volume, rather than by weight (Table 6-51). However,
volume production data were used to calculate off-site CO2 emissions from Alaska applying the same methodology
Land Use, Land-Use Change, and Forestry 6-97
-------
1 but with volume-specific C fraction conversion factors from IPCC (2006).61 Peat production was not reported for
2 2015 in Alaska's Mineral Industry 2014 report (DGGS 2015), and reliable data are not available beyond 2012, so
3 Alaska's peat production in 2013 through 2021 (reported in cubic yards) was assumed to be equal to the 2012
4 value.
5 Consistent with IPCC (2013) guidelines, off-site CO2 emissions from dissolved organic carbon were estimated based
6 on the total area of peatlands managed for peat extraction, which is calculated from production data using the
7 methodology described in the On-Site CO2 Emissions section below. Carbon dioxide emissions from dissolved
8 organic C were estimated by multiplying the area of managed peatlands by the default emission factor for
9 dissolved organic C provided in IPCC (2013).
10 The United States has largely imported peat from Canada for horticultural purposes; in 2021, imports of Sphagnum
11 moss (nutrient-poor) peat from Canada represented 96 percent of total U.S. peat imports and 80 percent of U.S.
12 domestic consumption (USGS 2022c). Most peat produced in the United States is reed-sedge peat, generally from
13 southern states, which is classified as nutrient-rich by IPCC (2006). To be consistent with the Tier 1 method, only
14 domestic peat production is accounted for when estimating off-site emissions. Higher-tier calculations of CO2
15 emissions from apparent consumption would involve consideration of the percentages of peat types stockpiled
16 (nutrient-rich versus nutrient-poor) as well as the percentages of peat types imported and exported.
17 Table 6-50: Peat Production of Conterminous 48 States (kt)
Type of Deposit
1990
2005
2017
2018
2019
2020
2021
Nutrient-Rich
595.1
657.6
423.3
416.7
410.4
430.7
378.0
Nutrient-Poor
55.4
27.4
74.7
62.3
45.6
13.3
42.0
Total Production
692.0
685.0
498.0
479.0
456.0
444.0
420.0
Sources: United States Geological Survey (USGS) (1991-2017) Minerals Yearbook: Peat (1994-2016); United
States Geological Survey (USGS) (2018) Minerals Yearbook: Peat - Tables-only release (2018); United States
Geological Survey (USGS) (2021) Mineral Commodity Summaries: Peat (2021).
18 Table 6-51: Peat Production of Alaska (Thousand Cubic Meters)
1990
2005
2017
2018
2019
2020
2021
Total Production
49.7
47.8
93.1
93.1
93.1
93.1
93.1
Sources: Division of Geological & Geophysical Surveys (DGGS), Alaska Department of Natural Resources
(1997-2015) Alaska's Mineral Industry Report (1997-2014).
19 On-site CO2 Emissions
20 IPCC (2006) suggests basing the calculation of on-site emission estimates on the area of peatlands managed for
21 peat extraction differentiated by the nutrient type of the deposit (rich versus poor). Information on the area of
22 land managed for peat extraction is currently not available for the United States, but consistent with IPCC (2006),
23 an average production rate for the industry was applied to derive a land area estimate. In a mature industrialized
24 peat industry, such as exists in the United States and Canada, the vacuum method can extract up to 100 metric
25 tons per hectare per year (Cleary et al. 2005 as cited in IPCC 2006).62 The area of land managed for peat extraction
26 in the conterminous 48 states of the United States was estimated using both nutrient-rich and nutrient-poor
27 production data and the assumption that 100 metric tons of peat are extracted from a single hectare in a single
61 Peat produced from Alaska was assumed to be nutrient poor; as is the case in Canada, "where deposits of high-quality [but
nutrient poor] Sphagnum moss are extensive" (USGS 2008).
62 The vacuum method is one type of extraction that annually "mills" or breaks up the surface of the peat into particles, which
then dry during the summer months. The air-dried peat particles are then collected by vacuum harvesters and transported from
the area to stockpiles (IPCC 2006).
6-98 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 year, see Table 6-52. The annual land area estimates were then multiplied by the IPCC (2013) default emission
2 factor in order to calculate on-site CO2 emission estimates.
3 Production data are not available by weight for Alaska. In order to calculate on-site emissions resulting from
4 Peatlands Remaining Peatlands in Alaska, the production data by volume were converted to weight using annual
5 average bulk peat density values, and then converted to land area estimates using the assumption that a single
6 hectare yields 100 metric tons, see Table 6-53. The IPCC (2006) on-site emissions equation also includes a term
7 that accounts for emissions resulting from the change in C stocks that occurs during the clearing of vegetation
8 prior to peat extraction. Area data on land undergoing conversion to peatlands for peat extraction is also
9 unavailable for the United States. However, USGS records show that the number of active operations in the United
10 States has been declining since 1990; therefore, it seems reasonable to assume that no new areas are being
11 cleared of vegetation for managed peat extraction. Other changes in C stocks in living biomass on managed
12 peatlands are also assumed to be zero under the Tier 1 methodology (IPCC 2006 and 2013).
13 Table 6-52: Peat Production Area of Conterminous 48 States (Hectares)
1990a
2005
2017
2018
2019
2020
2021
Nutrient-Rich
5,951
6,576
4,233
4,167
4,104
4,307
3,780
Nutrient-Poor
554
274
747
623
456
133
420
Total Production
6,920
6,850
4,980
4,790
4,560
4,440
4,200
a A portion of the production in 1990 is of unknown nutrient type, resulting in a total production value greater than the
sum of nutrient-rich and nutrient-poor.
14 Table 6-53: Peat Production Area of Alaska (Hectares)
1990
2005
2017
2018
2019
2020
2021
Nutrient-Rich
0
0
0
0
0
0
0
Nutrient-Poor
286
104
333
212
329
428
428
Total Production
286
104
333
212
329
428
428
15 On-site N2O Emissions
16 IPCC (2006) indicates the calculation of on-site N2O emission estimates using Tier 1 methodology only considers
17 nutrient-rich peatlands managed for peat extraction. These area data are not available directly for the United
18 States, but the on-site CO2 emissions methodology above details the calculation of nutrient-rich area data from
19 production data. In order to estimate N2O emissions, the land area estimate of nutrient-rich Peatlands Remaining
20 Peatlands was multiplied by the appropriate default emission factor taken from IPCC (2013). See Planned
21 Improvements section for additional information on the basis of land area estimates.
22 On-site CH4 Emissions
23 IPCC (2013) also suggests basing the calculation of on-site Cm emission estimates on the total area of peatlands
24 managed for peat extraction. Area data is derived using the calculation from production data described in the On-
25 site CO2 Emissions section above. In order to estimate CH4 emissions from drained land surface, the land area
26 estimate of Peatlands Remaining Peatlands was multiplied by the emission factor for direct CH4 emissions taken
27 from IPCC (2013). In order to estimate CH4 emissions from drainage ditches, the total area of peatland was
28 multiplied by the default fraction of peatland area that contains drainage ditches, and the appropriate emission
29 factor taken from IPCC (2013). See Table 6-54 for the calculated area of ditches and drained land.
30 Table 6-54: Peat Production (Hectares)
1990
2005
2017
2018
2019
2020
2021
Conterminous 48 States
Area of Drained Land
6,574
6,508
4,731
4,551
4,332
4,218
3,990
Land Use, Land-Use Change, and Forestry 6-99
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Area of Ditches
346
343
249
240
228
222
210
Total Production
6,920
6,850
4,980
4,790
4,560
4,440
4,200
Alaska
Area of Drained Land
272
99
317
202
312
407
407
Area of Ditches
14
5
17
11
16
21
21
Total Production
286
104
333
212
329
428
212
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2021. The same data sources were used throughout the time series, when available. When data were
unavailable or the available data were outliers, missing values were estimated based on the past available data.
Uncertainty
A Monte Carlo (Approach 2) uncertainty analysis was applied to estimate the uncertainty of CO2, Cm, and N2O
emissions from Peatlands Remaining Peatlands for 2021, using the following assumptions:
• The uncertainty associated with peat production data was estimated to be ± 25 percent (Apodaca 2008)
and assumed to be normally distributed.
• The uncertainty associated with peat production data stems from the fact that the USGS receives data
from the smaller peat producers but estimates production from some larger peat distributors. The peat
type production percentages were assumed to have the same uncertainty values and distribution as the
peat production data (i.e., ± 25 percent with a normal distribution).
• The uncertainty associated with the reported production data for Alaska was assumed to be the same as
for the conterminous 48 states, or ± 25 percent with a normal distribution. It should be noted that the
DGGS estimates that around half of producers do not respond to their survey with peat production data;
therefore, the production numbers reported are likely to underestimate Alaska peat production
(Szumigala 2008).
• The uncertainty associated with the average bulk density values was estimated to be ± 25 percent with a
normal distribution (Apodaca 2008).
• IPCC (2006 and 2013) gives uncertainty values for the emissions factors for the area of peat deposits
managed for peat extraction based on the range of underlying data used to determine the emission
factors. The uncertainty associated with the emission factors was assumed to be triangularly distributed.
• The uncertainty values surrounding the C fractions were based on IPCC (2006) and the uncertainty was
assumed to be uniformly distributed.
• The uncertainty values associated with the fraction of peatland covered by ditches was assumed to be ±
100 percent with a normal distribution based on the assumption that greater than 10 percent coverage,
the upper uncertainty bound, is not typical of drained organic soils outside of The Netherlands (IPCC
2013).
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 6-55. Carbon dioxide
emissions from Peatlands Remaining Peatlands in 2021 were estimated to be between 0.6 and 0.8 MMT CO2 Eq. at
the 95 percent confidence level. This indicates a range of 16 percent below to 16 percent above the 2021 emission
estimate of 0.7 MMT CO2 Eq. Methane emissions from Peatlands Remaining Peatlands in 2021 were estimated to
be between 0.002 and 0.007 MMT CO2 Eq. This indicates a range of 58 percent below to 80 percent above the
2021 emission estimate of 0.004 MMT CO2 Eq. Nitrous oxide emissions from Peatlands Remaining Peatlands in
2021 were estimated to be between 0.0003 and 0.0008 MMT CO2 Eq. at the 95 percent confidence level. This
indicates a range of 52 percent below to 53 percent above the 2021 emission estimate of 0.0005 MMT CO2 Eq.
6-100 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Table 6-55: Approach 2 Quantitative Uncertainty Estimates for CO2, Cm, and N2O Emissions
2 from Peatlands Remaining Peatlands (MMT CO2 Eq. and Percent)
2021 Emission
Source
Gas
Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT CO? Eq.)
(MMT CO?
Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Peatlands Remaining Peatlands
C02
0.7
0.6
0.8
-16%
16%
Peatlands Remaining Peatlands
ch4
+
+
+
-58%
80%
Peatlands Remaining Peatlands
n2o
+
+
+
-52%
53%
+ Does not exceed 0.05 MMT C02 Eq.
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
3 QA/QC and Verification
4 A QA/QC analysis was performed to review input data and calculations, and no issues were identified. In addition,
5 the emission trends were analyzed to ensure they reflected activity data trends.
6 Recalculations Discussion
7 The conterminous 48 states peat production estimates for Peatlands Remaining Peatlands were updated using the
8 Peat section of the Mineral Commodity Summaries 2022. The 2022 edition updated 2018, 2019, and 2020 peat
9 production data and provided peat type production estimates for 2021. The updated data increased previously
10 estimated emissions for 2018 by 0.4 percent, 2019 by 0.2 percent, and 2020 by 3.5 percent versus estimated
11 emissions for 2018, 2019, and 2020 in the previous (i.e., 1990 through 2020) Inventory for Peatlands Remaining
12 Peatlands.
13 Although Alaska peat production data for 2015 through 2021 were unavailable, 2014 data are available in the
14 Alaska's Mineral Industry 2014 report. However, the reported values represented an apparent 98 percent
15 decrease in production since 2012. Due to the uncertainty of the most recent data, 2013, 2014, 2015, 2016, 2017,
16 2018, 2019, and 2020 values were assumed to be equal to the 2012 value, seen in the Alaska's Mineral Industry
17 2013 report. If updated Alaska data are available for the next Inventory cycle, this will result in a recalculation in
18 the next (i.e., 1990 through 2021) Inventory report.
19 EPA updated global warming potentials (GWP) for calculating C02-equivalent emissions of CFU (from 25 to 28) and
20 N2O (from 298 to 265) to reflect the 100-year GWPs provided in the IPCC Fifth Assessment Report (AR5) (IPCC
21 2013). The previous Inventory used 100-year GWPs provided in the IPCC Fourth Assessment Report (AR4). This
22 update was applied across the entire time series. This change resulted in an 11 percent reduction in CO2 Eq.
23 emissions for N2O across the time series, as well as a 12 percent increase in CO2 Eq. emissions for CFU across the
24 time series. Further discussion on this update and the overall impacts of updating the Inventory GWP values to
25 reflect the AR5 can be found in Chapter 9, Recalculations and Improvements.
26 The cumulative effect of all of these changes was an average increase of 0.2 percent across the time series, with
27 the smallest increase of 0.05 percent (0.0005 MMT CO2 Eq.) in 1996 to the largest increase of 3.6 percent (0.03
28 MMTCO2 Eq.) in 2020.
29 Planned Improvements
30 EPA notes the following improvements may be implemented or investigated within the next two or three
31 inventory cycles pending time and resource constraints:
32 • The implied emission factors will be calculated and included in this chapter for future Inventories.
33 Currently, the N2O emissions calculation uses different land areas than the CO2 and CFU emission
34 calculations (see Methodology and Time Series Consistency in this chapter), so estimating the implied
Land Use, Land-Use Change, and Forestry 6-101
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
emission factor per total land area is not appropriate. The inclusion of implied emission factors in this
chapter will provide another method of QA/QC and verification for Inventory data.
EPA notes the following improvements will continue to be investigated as time and resources allow, but there are
no immediate plans to implement until data are available or identified:
• In order to further improve estimates of CO2, N2O, and Cm emissions from Peatlands Remaining
Peatlands, future efforts will investigate if improved data sources exist for determining the quantity of
peat harvested per hectare and the total area of land undergoing peat extraction.
• EPA plans to identify a new source for Alaska peat production. The current source has not been reliably
updated since 2012 and Alaska Department of Natural Resources indicated future publication of data has
been discontinued.
• Edits to the trends and methodology sections are planned based on expert review comments.
Coastal Wetlands Remaining Coastal Wetlands
Consistent with ecological definitions of wetlands,63 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 Land
Representation. Guidance on quantifying greenhouse gas emissions and removals on Coastal Wetlands is provided
in the 2013 Supplement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands (Wetlands
Supplement), which recognizes the particular importance of vascular plants in sequestering CO2 from the
atmosphere within biomass, dead organic material (DOM; including litter and dead wood stocks) and soils. Thus,
the Wetlands Supplement provides specific guidance on quantifying emissions and removals on organic and
mineral soils that are covered or saturated for part of the year by tidal fresh, brackish or saline water and are
vegetated by vascular plants and may extend seaward to the maximum depth of vascular plant vegetation. The
United States calculates emissions and removals based upon the stock change method for soil carbon (C) and the
gain-loss method for biomass and DOM. Presently, this Inventory does not calculate the lateral flux of C to or from
any land use. Lateral transfer of organic C to coastal wetlands and to marine sediments within U.S. waters is the
subject of ongoing scientific investigation; there is currently no IPCC methodological guidance for lateral fluxes of
C.
The United States recognizes both Vegetated Wetlands and Unvegetated Open Water as Coastal Wetlands. Per
guidance provided by the Wetlands Supplement, sequestration of C into biomass, DOM and soil C pools is
recognized only in Vegetated Coastal Wetlands and does not occur in Unvegetated Open Water Coastal Wetlands.
The United States takes the additional step of recognizing that C stock losses occur when Vegetated Coastal
Wetlands are converted to Unvegetated Open Water Coastal Wetlands.
This Inventory includes all privately- and publicly-owned coastal wetlands (i.e., mangroves and tidal marsh) along
the oceanic shores of the conterminous United States, but does not include Coastal Wetlands Remaining Coastal
Wetlands in Alaska, Hawaii, or any of the United States Territories. Seagrasses are not currently included within
the Inventory due to insufficient data on distribution, change through time and C stocks or C stock changes as a
result of anthropogenic influence (see Planned Improvements).
Under the Coastal Wetlands Remaining Coastal Wetlands category, the following emissions and removals are
quantified in this chapter:
1) Carbon stock changes and Cm emissions on Vegetated Coastal Wetlands Remaining Vegetated Coastal
Wetlands,
63 See https://water.usBS.gov/nwsum/WSP2425/definitions.html; accessed August 2021.
6-102 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
2) Carbon stock changes on Vegetated Coastal Wetlands Converted to Unvegetated Open Water Coastal
Wetlands,
3) Carbon stock changes on Unvegetated Open Water Coastal Wetlands Converted to Vegetated Coastal
Wetlands, and
4) Nitrous Oxide Emissions from Aquaculture in Coastal Wetlands.
Vegetated coastal wetlands hold C in all five C pools (i.e., aboveground biomass, belowground biomass, dead
organic matter [DOM; dead wood and litter], and soil), though typically soil C and, to a lesser extent, aboveground
and belowground biomass are the dominant pools, depending on wetland type (i.e., forested vs. marsh).
Vegetated Coastal Wetlands are net accumulators of C over centuries to millennia as soils accumulate C under
anaerobic soil conditions and C accumulates in plant biomass. Large emissions from soil C and biomass stocks
occur when Vegetated Coastal Wetlands are converted to Unvegetated Open Water Coastal Wetlands (e.g., when
Vegetated Coastal Wetlands are lost due to subsidence, channel cutting through Vegetated Coastal Wetlands), but
are still recognized as Coastal Wetlands in this Inventory. These C stock losses resulting from conversion to
Unvegetated Open Water Coastal Wetlands can cause the release of decades to centuries of accumulated soil C, as
well as the standing stock of biomass C. Conversion of Unvegetated Open Water Coastal Wetlands to Vegetated
Coastal Wetlands, either through restoration efforts or naturally, initiates the building of C stocks within soils and
biomass. In applying the Wetlands Supplement methodologies for estimating CFU emissions, coastal wetlands in
salinity conditions greater than 18 parts per thousand have little to no CFU emissions compared to those
experiencing lower salinity brackish and freshwater conditions. Therefore, conversion of Vegetated Coastal
Wetlands to or from Unvegetated Open Water Coastal Wetlands are conservatively assumed to not result in a
change in salinity condition and are assumed to have no impact on Cm emissions. The Wetlands Supplement
provides methodologies to estimate N2O emissions from coastal wetlands that occur due to aquaculture. The N2O
emissions from aquaculture result from the N derived from consumption of the applied food stock that is then
excreted as N load available for conversion to N2O. While N2O emissions can also occur due to anthropogenic N
loading from the watershed and atmospheric deposition, these emissions are not reported here to avoid double-
counting of indirect N2O emissions with the Agricultural Soils Management, Forest Land and Settlements
categories.
The Wetlands Supplement provides methodologies for estimating C stock changes and CH4 emissions from
mangroves, tidal marshes and seagrasses. Depending upon their height and area, C stock changes from mangroves
may be reported under the Forest Land category or under Coastal Wetlands. If mangrove stature is 5 m or greater
or if there is evidence that trees can obtain that height, mangroves are reported under the Forest Land category
because they meet the definition of Forest Land. Mangrove forests that are less than 5 m are reported under
Coastal Wetlands because they meet the definition of Wetlands. All other non-drained, intact coastal marshes are
reported under Coastal Wetlands.
Because of human activities and level of regulatory oversight, all coastal wetlands within the conterminous United
States are included within the managed land area described in Section 6.1, and as such, estimates of C stock
changes, emissions of CH4, and emissions of N2O from aquaculture from all coastal wetlands are included in this
Inventory. At the present stage of inventory development, Coastal Wetlands are not explicitly shown in the Land
Representation analysis while work continues to harmonize data from NOAA's Coastal Change Analysis Program
(C-CAP)64 with NRI, FIA and NLDC data used to compile the Land Representation. However, a check was
undertaken to confirm that Coastal Wetlands recognized by C-CAP represented a subset of Wetlands recognized by
the NRI for marine coastal states.
The greenhouse gas fluxes for all four wetland categories described above are summarized in Table 6-56. Coastal
Wetlands Remaining Coastal Wetlands are generally a net C sink, with the fluxes ranging from -3.3 to -4.4 MMT
CO2 Eq. across the majority of the time series; however, between 2006 and 2010, they were a net source of
64 See https://coast.noaa.gov/digitalcoast/tools/lca.html; accessed August 2021.
Land Use, Land-Use Change, and Forestry 6-103
-------
1 emissions (ranging from 5.6 to 5.9 MMT CO2 Eq.), resulting from a large loss of vegetated coastal wetlands to open
2 water due to hurricanes (Table 6-56). Recognizing removals of CChto soil of 10.2 MMT CO2 Eq. and Cm emissions
3 of 4.3 MMT CO2 Eq. in 2021, Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands are a net sink of
4 5.9MMT CO2 Eq. Loss of coastal wetlands, primarily in the Mississippi Delta as a result of hurricane impacts and
5 sediment diversion and other human impacts, recognized as Vegetated Coastal Wetlands Converted to
6 Unvegetated Coastal Wetlands, drive an emission of 1.5 MMT CO2 Eq. since 2011, primarily from soils. Building of
7 new wetlands from open water, recognized as Unvegetated Coastal Wetlands Converted to Vegetated Coastal,
8 results each year in removal of 0.1 MMT CO2 Eq. Aquaculture is a minor industry in the United States, resulting in
9 an emission of N2O across the time series of between 0.1 to 0.2 MMT CO2 Eq. In total, Coastal Wetlands are a net
10 sink of 4.4 MMT C02 Eq. in 2021.
11 Table 6-56: Emissions and Removals from Coastal Wetlands Remaining Coastal Wetlands
12 (MMT COz Eq.)
Land Use/Carbon Pool
1990
2005
2017
2018
2019
2020
2021
Vegetated Coastal Wetlands
Remaining Vegetated Coastal
Wetlands
(6.0)
(6.0)
(5.9)
(5.9)
(5.9)
(5.9)
(5.9)
Biomass C Flux
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Soil C Flux
(10.1)
(10.2)
(10.2)
(10.2)
(10.2)
(10.2)
(10.2)
Net CH4 Flux
4.2
4.2
4.3
4.3
4.3
4.3
4.3
Vegetated Coastal Wetlands
Converted to Unvegetated Open
Water Coastal Wetlands
1.8
2.6
1.5
1.5
1.5
1.5
1.5
Biomass C Flux
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Dead Organic Matter C Flux
+
+
+
+
+
+
+
Soil C Flux
1.7
2.5
1.5
1.5
1.5
1.5
1.5
Unvegetated Open Water Coastal
Wetlands Converted to Vegetated
Coastal Wetlands
(+)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Biomass C Flux
(+)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Dead Organic Matter C Flux
(+)
(+)
+
+
+
+
+
Soil C Flux
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Net N20 Flux from Aquaculture in
Coastal Wetlands
0.1
0.2
0.1
0.1
0.1
0.1
0.1
Total Biomass C Flux
+
+
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Total Dead Organic Matter C Flux
(+)
(+)
+
+
+
+
+
Total Soil C Flux
(8.4)
(7.7)
(8.7)
(8.7)
(8.7)
(8.7)
(8.7)
Total CH4 Flux
4.2
4.2
4.3
4.3
4.3
4.3
4.3
Total N20 Flux
0.1
0.2
0.1
0.1
0.1
0.1
0.1
Total Flux
(4.1)
(3.3)
(4.4)
(4.4)
(4.4)
(4.4)
(4.4)
+ Absolute value does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
13 Emissions and Removals from Vegetated Coastal Wetlands
14 Remaining Vegetated Coastal Wetlands
15 The conterminous United States currently has 2.98 million hectares of intertidal Vegetated Coastal Wetlands
16 Remaining Vegetated Coastal Wetlands comprised of tidally influenced palustrine emergent marsh (661,731 ha),
17 palustrine scrub shrub (133,365 ha) and estuarine emergent marsh (1,893,276 ha), estuarine scrub shrub (94,667
18 ha) and estuarine forested wetlands (195,221 ha). Mangroves fall under both estuarine forest and estuarine scrub
19 shrub categories depending upon height. Dwarf mangroves, found in subtropical states along the Gulf of Mexico,
20 do not attain the height status to be recognized as Forest Land, and are therefore always classified within
21 Vegetated Coastal Wetlands. Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands are found in
6-104 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
cold temperate (53,970 ha), warm temperate (896,287 ha), subtropical (1,965,242 ha) and Mediterranean (62,761
ha) climate zones.
Soils are the largest C pool in Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands, reflecting long-
term removal of atmospheric CO2 by vegetation and transfer into the soil pool in the form of both autochthonous
and allochthonous decaying organic matter. Soil C emissions are not assumed to occur in coastal wetlands that
remain vegetated. This Inventory includes changes in C stocks in both biomass and soils. Changes in DOM C stocks
are not included. Methane emissions from decomposition of organic matter in anaerobic conditions are present at
salinity less than half that of sea water. Mineral and organic soils are not differentiated in terms of C stock changes
or Cm emissions.
Table 6-57 through Table 6-59 summarize nationally aggregated biomass and soil C stock changes and CH4
emissions on Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands. Intact Vegetated Coastal
Wetlands Remaining Vegetated Coastal Wetlands hold a total biomass C stock of 35.95 MMT C. Removals from
biomass C stocks in 2021 were 0.05 MMT CO2 Eq. (0.01 MMT C), which has increased over the time series (Table
6-57 and Table 6-58). Carbon dioxide emissions from biomass in Vegetated Coastal Wetlands Remaining Vegetated
Coastal Wetlands between 2002 and 2011, with very low sequestration between 2002 and 2006 and emissions of
0.21 MMT CO2 Eq. between 2007 and 2011, are not inherently typical and are a result of coastal wetland loss over
time. Most of the coastal wetland loss has occurred in palustrine and estuarine emergent wetlands. Vegetated
coastal wetlands maintain a large C stock within the top 1 meter of soil (estimated to be 804 MMT C) to which C
accumulated at a rate of 10.2 MMT CO2 Eq. (2.8 MMT C) in 2021, a value that has remained relatively constant
across the reporting period. For Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands, methane
emissions of 4.3 of MMT CO2 Eq. (154 kt CH4) in 2021 (Table 6-59) offset C removals resulting in a net removal of
5.9 MMT CO2 Eq. in 2021; this rate has been relatively consistent across the reporting period. Dead organic matter
stock changes are not calculated in Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands since this
stock is considered to be in a steady state when using Tier 1 methods (IPCC 2014). Due to federal regulatory
protection, loss of Vegetated Coastal Wetlands through human activities slowed considerably in the 1970s and the
current annual rates of C stock change and CH4 emissions are relatively constant over time.
Table 6-57: Net CO2 Flux from C Stock Changes in Vegetated Coastal Wetlands Remaining
Vegetated Coastal Wetlands (MMT CO2 Eq.)
Year
1990
2005
2017
2018
2019
2020
2021
Biomass Flux
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Soil Flux
(10.1)
(10.2)
(10.2)
(10.2)
(10.2)
(10.2)
(10.2)
Total C Stock Change
(10.2)
(10.2)
(10.2)
(10.2)
(10.2)
(10.2)
(10.2)
+ Absolute value does not exceed 0.05 MMT C02 Eq.
Note: Parentheses indicate net sequestration. Totals may not sum due to independent rounding.
Table 6-58: Net CO2 Flux from C Stock Changes in Vegetated Coastal Wetlands Remaining
Vegetated Coastal Wetlands (MMT C)
Year
1990
2005
2017
2018
2018
2019
2020
Biomass Flux
(+)
+
(+)
(+)
(+)
(+)
(+)
Soil Flux
(2.8)
(2.8)
(2.8)
(2.8)
(2.8)
(2.8)
(2.8)
Total C Stock Change
(2.8)
(2.8)
(2.8)
(2.8)
(2.8)
(2.8)
(2.8)
+ Absolute value does not exceed 0.05 MMT C.
Note: Parentheses indicate net sequestration. Totals may not sum due to independent rounding.
Table 6-59: CH4 Emissions from Vegetated Coastal Wetlands Remaining Vegetated Coastal
Wetlands (MMT COz Eq. and kt CH4)
Year
1990
2005
2017
2018
2019
2020
2021
Methane Emissions (MMT C02 Eq.)
4.2
4.2
4.3
4.3
4.3
4.3
4.3
Methane Emissions (kt CH4)
149
151
153
153
153
154
154
Land Use, Land-Use Change, and Forestry 6-105
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Methodology and Time-Series Consistency
The following section includes a description of the methodology used to estimate changes in biomass C stocks, soil
C stocks and emissions of CFU for Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands. Dead
organic matter is not calculated for Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands since it is
assumed to be in steady state (IPCC 2014).
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2021.
Biomass Carbon Stock Changes
Above- and belowground biomass C stocks for palustrine (freshwater) and estuarine (saline) marshes are
estimated for Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands on land below the elevation of
high tides (taken to be mean high water spring tide elevation) and as far seawards as the extent of intertidal
vascular plants according to the national LiDAR dataset, the national network of tide gauges and land use histories
recorded in the 1996, 2001, 2006, 2010, and 2016 NOAA C-CAP surveys (NOAA OCM 2020). C-CAP areas are
calculated at the state/territory level and summed according to climate zone to national values. Federal and non-
federal lands are represented. Trends in land cover change are extrapolated to 1990 and 2021 from these datasets.
Based upon NOAA C-CAP, coastal wetlands are subdivided into palustrine and estuarine classes and further
subdivided into emergent marsh, scrub shrub and forest classes (Table 6-60). Biomass is not sensitive to soil
organic matter content but is differentiated based on climate zone. Aboveground biomass C stocks for non-
forested wetlands data are derived from a national assessment combining field plot data and aboveground
biomass mapping by remote sensing (Byrd et al. 2017; Byrd et al. 2018; Byrd et al. 2020). The aboveground
biomass C stock for subtropical estuarine forested wetlands (dwarf mangroves that are not classified as forests due
to their stature) is derived from a meta-analysis by Lu and Megonigal (2017). Root to shoot ratios from the
Wetlands Supplement (Table 6-62; IPCC 2014) were used to account for belowground biomass, which were
multiplied by the aboveground C stock. Above- and belowground values were summed to obtain total biomass C
stocks. Biomass C stock changes per year for Wetlands Remaining Wetlands were determined by calculating the
difference in area between that year and the previous year to calculate gain/loss of area for each climate type,
which was multiplied by the mean biomass for that climate type.
Table 6-60: Area of Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands,
Vegetated Coastal Wetlands Converted to Unvegetated Open Water Coastal Wetlands, and
Unvegetated Open Water Coastal Wetlands Converted to Vegetated Coastal Wetlands (ha)
Year
1990
2005
2017
2018
2019
2020
2021
Vegetated Coastal Wetlands
Remaining Vegetated Coastal
Wetlands
2,975,477
2,985,783
2,973,256
2,974,523
2,975,789
2,977,055
2,978,322
Vegetated Coastal Wetlands
Converted to Unvegetated
Open Water Coastal Wetlands
1,720
2,515
1,488
1,488
1,488
1,488
1,488
Unvegetated Open Water
Coastal Wetlands Converted to
Vegetated Coastal Wetlands
952
1,769
2,406
2,406
2,406
2,406
2,406
Table 6-61: Aboveground Biomass Carbon Stocks for Vegetated Coastal Wetlands (t C ha1)
Climate Zone
Wetland Type Cold Temperate Warm Temperate Subtropical Mediterranean
Palustrine Scrub/Shrub Wetland 3^25 3X7 Z24 4~69
Palustrine Emergent Wetland 3.25 3.17 2.24 4.69
6-106 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Estuarine Forested Wetland N/A N/A 17.83 N/A
Estuarine Scrub/Shrub Wetland 3.05 3.05 2.43 3.44
Estuarine Emergent Wetland 3.05 3.10 2.43 3.44
Source: All data from Byrd et al. (2017, 2018 and 2020) except for subtropical estuarine forested wetlands,
which is from Lu and Megonigal (2017); N/A means there are currently no estuarine forested wetlands that
are less than 5 meters tall; these forested wetlands meet the definition of forest land and are included in
the Forest Land chapter.
l Table 6-62: Root to Shoot Ratios for Vegetated Coastal Wetlands
Climate Zone
Wetland Type
Cold Temperate
Warm Temperate
Subtropical
Mediterranean
Palustrine Scrub/Shrub Wetland
1.15
1.15
3.65
3.63
Palustrine Emergent Wetland
1.15
1.15
3.65
3.63
Estuarine Forested Wetland
N/A
N/A
0.96
N/A
Estuarine Scrub/Shrub Wetland
2.11
2.11
3.65
3.63
Estuarine Emergent Wetland
2.11
2.11
3.65
3.63
Source: All values from IPCC (2014); N/A means there are currently no estuarine forested wetlands that are
less than 5 meters tall; these forested wetlands meet the definition of forest land and are included in the
Forest Land chapter.
2 Soil Carbon Stock Changes
3 Soil C stock changes are estimated for Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands for
4 both mineral and organic soils. Soil C stock changes, stratified by climate zones and wetland classes, are derived
5 from a synthesis of peer-reviewed literature (Table 6-63; Lynch 1989; Orson et al. 1990; Kearny & Stevenson 1991;
6 Roman et al. 1997; Craft et al. 1998; Orson et al. 1998; Merrill 1999; Hussein et al. 2004; Church et al. 2006; Koster
7 et al. 2007; Callaway et al. 2012a&b; Bianchi et al. 2013; Crooks et al. 2014; Weston et al. 2014; Villa & Mitsch
8 2015; Marchio et al. 2016; Noe et al. 2016).
9 Tier 2 estimates of soil C removals associated with annual soil C accumulation on managed Vegetated Coastal
10 Wetlands Remaining Vegetated Coastal Wetlands were developed with country-specific soil C removal factors
11 multiplied by activity data of land area for Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands.
12 The methodology follows Eq. 4.7, Chapter 4 of the Wetlands Supplement, and is applied to the area of Vegetated
13 Coastal Wetlands Remaining Vegetated Coastal Wetlands on an annual basis. To estimate soil C stock changes, no
14 differentiation is made between organic and mineral soils since currently no statistical evidence supports
15 disaggregation (Holmquist et al. 2018).
16 Table 6-63: Annual Soil Carbon Accumulation Rates for Vegetated Coastal Wetlands (t C ha"1
17 yr1)
Climate Zone
Cold Temperate
Warm Temperate
Subtropical
Mediterranean
Palustrine Scrub/Shrub Wetland
1.01
1.54
0.45
0.85
Palustrine Emergent Wetland
1.01
1.54
0.45
0.85
Estuarine Forested Wetland
N/A
N/A
0.87
N/A
Estuarine Scrub/Shrub Wetland
1.01
0.82
1.09
0.85
Estuarine Emergent Wetland
2.17
0.82
1.09
0.85
Source: All data from Lu and Megonigal (2017)65; N/A means there are currently no estuarine forested
wetlands that are less than 5 meters tall; these forested wetlands meet the definition of forest land and are
included in the Forest Land chapter.
65 See https://github.com/Smithsonian/Coastal-Wetland-NGGI-Data-Public; accessed August 2022.
Land Use, Land-Use Change, and Forestry 6-107
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
So/7 Methane Emissions
Tier 1 estimates of Cm emissions for Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands are
derived from the same wetland map used in the analysis of wetland soil C fluxes, produced from C-CAP, LiDAR and
tidal data, in combination with default Cm emission factors provided in Table 4.14 of the Wetlands Supplement.
The methodology follows Equation 4.9, Chapter 4 of the Wetlands Supplement; Tier 1 emissions factors are
multiplied by the area of freshwater (palustrine) coastal wetlands. The Cl-U fluxes applied are determined based on
salinity; only palustrine wetlands are assumed to emit Cm. Estuarine coastal wetlands in the C-CAP classification
include wetlands with salinity less than 18 ppt, a threshold at which methanogenesis begins to occur (Poffenbarger
et al. 2011), but the dataset currently does not differentiate estuarine wetlands based on their salinities and, as a
result, Cm emissions from estuarine wetlands are not included at this time.
Uncertainty
Underlying uncertainties in the estimates of soil and biomass C stock changes and CH4 emissions 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 Cl-U 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. Because mean soil and biomass C stocks for each available community class
are in a fairly narrow range, the same overall uncertainty was assigned to each, respectively (i.e., applying
approach for asymmetrical errors, the largest uncertainty for any soil C stock value should be applied in the
calculation of error propagation; IPCC 2000). Uncertainty for root to shoot ratios, which are used for quantifying
belowground biomass, are derived from the 2013 Wetlands Supplement. Uncertainties for Cl-U flux are the Tier 1
default values reported in the 2013 IPCC Wetlands Supplement. Overall uncertainty of the NOAA C-CAP remote
sensing product is 15 percent. This is in the range of remote sensing methods (±10 to 15 percent; IPCC 2003).
However, there is significant uncertainty in salinity ranges for tidal and non-tidal estuarine wetlands and activity
data used to apply Cl-U flux emission factors (delineation of an 18 ppt boundary) that will need significant
improvement to reduce uncertainties. Details on the emission/removal trends and methodologies through time
are described in more detail in the introduction and the Methodology section. The combined uncertainty was
calculated using the IPCC Approach 1 method of summing the squared uncertainty for each individual source (C-
CAP, soil, biomass and CH4) and taking the square root of that total.
Uncertainty estimates are presented in Table 6-64 for each subcategory (i.e., soil C, biomass C and Cl-U emissions).
The combined uncertainty across all subcategory is 37.0 percent below and above the estimate of -6.4 MMT CO2
Eq, which is primarily driven by the uncertainty in the CH4 estimates because there is high variability in CH4
emissions when the salinity is less than 18 ppt. In 2021, the total flux was -6.4 MMT CO2 Eq., with lower and upper
estimates of-8.7 and -4.0 MMT CO2 Eq.
Table 6-64: IPCC Approach 1 Quantitative Uncertainty Estimates for C Stock Changes and
CH4 Emissions occurring within Vegetated Coastal Wetlands Remaining Vegetated Coastal
Wetlands in 2021 (MMT CO2 Eq. and Percent)
Source/Sink
Gas
2021 Estimate
Uncertainty Range Relative to Estimate
(MMT CO? Eq.)
(MMT CO? Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Biomass C Stock Change
C02
(0.05)
(0.06)
(0.03)
-24.1%
24.1%
Soil C Stock Change
C02
(10.2)
(12.0)
(8.4)
-18.7%
18.7%
CH4 emissions
ch4
4.3
3.0
5.6
-29.9%
29.9%
Total Flux
(5.9)
(8.1)
(3.8)
-37.0%
37.0%
Note: Parentheses indicate net sequestration. Totals may not sum due to independent rounding.
6-108 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 QA/QC and Verification
2 NOAA provided the National LiDAR Dataset, tide data, and C-CAP land cover and land cover change mapping, all of
3 which are subject to agency internal QA/QC assessment. Acceptance of final datasets into archive and
4 dissemination are contingent upon the product compilation being compliant with mandatory QA/QC requirements
5 (McCombs et al. 2016). QA/QC and verification of soil C stock datasets have been provided by the Smithsonian
6 Environmental Research Center and Coastal Wetland Inventory team leads who reviewed summary tables against
7 reviewed sources. Biomass C stocks are derived from peer-review literature and reviewed by the U.S. Geological
8 Survey prior to publishing, by the peer-review process during publishing, and by the Coastal Wetland Inventory
9 team leads before inclusion in this Inventory. A team of two evaluated and verified there were no computational
10 errors within the calculation worksheets. Soil and biomass C stock change data are based upon peer-reviewed
11 literature and CFU emission factors derived from the Wetlands Supplement.
12 Recalculations Discussion
13 An update was made to the activity data to remove any estuarine forested wetland areas that were located
14 outside of states classified as subtropical since those wetlands fall under Forest Land Remaining Forest Land. The
15 resulting changes in emissions and removals were minimal and did not affect source or sink status, but resulted in
16 a slight decrease in removals between 1990 and 2001 (0.03 MMT CO2 Eq.) and 2012 to 2020 (0.001 MMT CO2 Eq.)
17 and a slight increase in emissions between 2002 and 2006 (0.04-0.06 MMT CO2 Eq.) and 2007 to 2011 (0.001 MMT
18 CO2 Eq.). The change did not affect CFU emissions because no emission factor currently is applied to estuarine
19 wetlands.
20 In addition, the EPA updated the global warming potential (GWP) for calculating C02-equivalent emissions of CFU
21 (from 25 to 28) to reflect the 100-year GWP provided in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). The
22 previous Inventory used the 100-year GWP provided in the IPCC Fourth Assessment Report (AR4). This update was
23 applied across the entire time series. This change resulted in an average annual increase of 0.46 MMT CO2 Eq., or
24 12 percent, in calculated C02-equivalent CFU emissions from Vegetated Coastal Wetlands Remaining Vegetated
25 Coastal Wetlands from 1990 through 2020 compared to the previous Inventory. Further discussion on this update
26 and the overall impacts of updating the inventory GWP values to reflect the AR5 can be found in Chapter 9,
27 Recalculations and Improvements.
28 Planned Improvements
29 Harmonization across all spatial datasets used to calculate activity data is underway. Once completed, a better
30 representation of forested tidal wetlands, palustrine tidal wetlands, and forest land near the tidal boundary will be
31 obtained.
32 Administered by the Smithsonian Environmental Research Center, the Coastal Wetland Carbon Research
33 Coordination Network has established a U.S. country-specific database of soil C stock and biomass estimates for
34 coastal wetlands.66 This dataset is currently in review and may be update in coming months. Refined error analysis
35 combining land cover change and C stock estimates will be provided as new data are incorporated. Through this
36 work, a model is in development to represent updated changes in soil C stocks for estuarine emergent wetlands.
37 Work is currently underway to examine the feasibility of incorporating seagrass soil and biomass C stocks into the
38 Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands estimates. Additionally, investigation into
39 quantifying the distribution, area, and emissions resulting from impounded waters (i.e., coastal wetlands where
40 tidal connection to the ocean has been restricted or eliminated completely) is underway.
41
66 See https://serc.si.edu/coastalcarbon; accessed August 2021.
Land Use, Land-Use Change, and Forestry 6-109
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Emissions from Vegetated Coastal Wetlands Converted to
Unvegetated Open Water Coastal Wetlands
Vegetated Coastal Wetlands Converted to Unvegetated Open Water Coastal Wetlands is a source of emissions
from soil, biomass, and DOM C stocks. An estimated 1,488 ha of Vegetated Coastal Wetlands were converted to
Unvegetated Open Water Coastal Wetlands in 2021, which largely occurred within estuarine and palustrine
emergent wetlands. Prior to 2006, annual conversion to unvegetated open water coastal wetlands was higher than
current rates: 1,720 between 1990 and 2000 and 2,515 ha between 2001 and 2005. The Mississippi Delta
represents more than 40 percent of the total coastal wetland of the United States, and over 90 percent of the area
of Vegetated Coastal Wetlands Converted to Unvegetated Open Water Coastal Wetlands. The drivers of coastal
wetlands loss include legacy human impacts on sediment supply through rerouting river flow, direct impacts of
channel cutting on hydrology, salinity and sediment delivery, and accelerated subsidence from aquifer extraction.
Each of these drivers directly contributes to wetland erosion and subsidence, while also reducing the resilience of
the wetland to build with sea-level rise or recover from hurricane disturbance. Over recent decades, the rate of
Mississippi Delta wetland loss has slowed, though episodic mobilization of sediment occurs during hurricane
events (Couvillion et al. 2011; Couvillion et al. 2016). The land cover analysis between the 2006 and 2011C-CAP
surveys coincides with two such events, hurricanes Katrina and Rita (both making landfall in the late summer of
2005), that occurred between these C-CAP survey dates. The subsequent 2016 C-CAP survey determined that
erosion rates had slowed.
Shallow nearshore open water within the U.S. Land Representation is recognized as falling under the Coastal
Wetlands category within this Inventory. While high resolution mapping of coastal wetlands provides data to
support IPCC Approach 2 methods for tracking land cover change, the depth in the soil profile to which sediment is
lost is less clear. This Inventory adopts the Tier 1 methodological guidance from the Wetlands Supplement for
estimating emissions following the methodology for excavation (see Methodology section, below) when Vegetated
Coastal Wetlands are converted to Unvegetated Open Water Coastal Wetlands, assuming aim depth of disturbed
soil. This 1 m depth of disturbance is consistent with estimates of wetland C loss provided in the literature and the
Wetlands Supplement (Crooks et al. 2009; Couvillion et al. 2011; Delaune and White 2012; IPCC 2014). The same
assumption on depth of soils impacted by erosion has been applied here. It is a reasonable Tier 1 assumption,
based on experience, but estimates of emissions are sensitive to the depth to which the assumed disturbances
have occurred (Holmquist et al. 2018). A Tier 1 assumption is also adopted in that all mobilized C is immediately
returned to the atmosphere (as assumed for terrestrial land-use categories), rather than redeposited in long-term
C storage. The science is currently under evaluation to adopt more refined emissions factors for mobilized coastal
wetland C based upon the geomorphic setting of the depositional environment.
In 2021, there were 1,488 ha of Vegetated Coastal Wetlands Converted to Unvegetated Open Water Coastal
Wetlands (Table 6-60) across all wetland types and climates, which resulted in 1.5 MMT CO2 Eq. (0.4 MMT C) and
0.06 MMT CO2 Eq. (0.02 MMT C) lost through soil and biomass, respectively, with minimal DOM C stock loss (Table
6-65, and Table 6-66). Across the reporting period, the area of vegetated coastal wetlands converted to
unvegetated open water coastal wetlands was greatest between the 2006 to 2011 C-CAP reporting period (11,373
ha) and has decreased since then to current levels (Table 6-60).
Table 6-65: Net CO2 Flux from C Stock Changes in Vegetated Coastal Wetlands Converted to
Unvegetated Open Water Coastal Wetlands (MMT CO2 Eq.)
Year
1990
2005
2017
2018
2019
2020
2021
Biomass Flux
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Dead Organic Matter Flux
+
+
+
+
+
+
+
Soil Flux
1.7
2.5
1.5
1.5
1.5
1.5
1.5
Total C Stock Change
1.8
2.6
1.5
1.5
1.5
1.5
1.5
+ Absolute value does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
6-110 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Table 6-66: Net CO2 Flux from C Stock Changes in Vegetated Coastal Wetlands Converted to
2 Unvegetated Open Water Coastal Wetlands (MMT C)
Year
1990
2005
2017
2018
2019
2020
2021
Biomass Flux
+
+
+
+
+
+
+
Dead Organic Matter Flux
+
+
+
+
+
+
+
Soil Flux
0.5
0.7
0.4
0.4
0.4
0.4
0.4
Total C Stock Change
0.5
0.7
0.4
0.4
0.4
0.4
0.4
+ Absolute value does not exceed 0.05 MMT C.
Note: Totals may not sum due to independent rounding.
3 Methodology and Time-Series Consistency
4 The following section includes a brief description of the methodology used to estimate changes in soil, biomass
5 and DOM C stocks for Vegetated Coastal Wetlands Converted to Unvegetated Open Water Coastal Wetlands.
6 Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
7 through 2021.
8 Biomass Carbon Stock Changes
9 Biomass C stock changes for palustrine and estuarine marshes are estimated for Vegetated Coastal Wetlands
10 Converted to Unvegetated Open Water Coastal Wetlands on lands below the elevation of high tides (taken to be
11 mean high water spring tide elevation) within the U.S. Land Representation according to the national LiDAR
12 dataset, the national network of tide gauges and land use histories recorded in the 1996, 2001, 2006, 2010, and
13 2016 NOAA C-CAP surveys. C-CAP areas are calculated at the state/territory level and summed according to
14 climate zone to national values. Publicly-owned and privately-owned lands are represented. Trends in land cover
15 change are extrapolated to 1990 and 2021 from these datasets. The C-CAP database provides peer reviewed
16 country-specific mapping to support IPCC Approach 3 quantification of coastal wetland distribution, including
17 conversion to and from open water. Biomass C stocks are not sensitive to soil organic content but are
18 differentiated based on climate zone. Non-forested aboveground biomass C stock data are derived from a national
19 assessment combining field plot data and aboveground biomass mapping by remote sensing (Byrd et al. 2017; Byrd
20 et al. 2018; Byrd et al. 2020). The aboveground biomass C stock for estuarine forested wetlands (dwarf mangroves
21 that are not classified as forests due to their stature) is derived from a meta-analysis by Lu and Megonigal (201767;
22 Table 6-61). Aboveground biomass C stock data for all subcategories are not available and thus assumptions were
23 applied using expert judgment about the most appropriate assignment of a C stock to a disaggregation of a
24 community class. Root to shoot ratios from the Wetlands Supplement were used to account for belowground
25 biomass, which were multiplied by the aboveground C stock (Table 6-62; IPCC 2014). Above- and belowground
26 values were summed to obtain total biomass C stocks. Conversion to open water results in emissions of all biomass
27 C stocks during the year of conversion; therefore, emissions are calculated by multiplying the C-CAP derived area
28 of vegetated coastal wetlands lost that year in each climate zone by its mean biomass.
29 Dead Organic Matter
30 Dead organic matter (DOM) C stocks, which include litter and dead wood stocks for subtropical estuarine forested
31 wetlands, are an emission from Vegetated Coastal Wetlands Converted to Unvegetated Open Water Coastal
32 Wetlands across all years in the time series. Data on DOM C stocks are not currently available for either palustrine
33 or estuarine scrub/shrub wetlands for any climate zone. Data for estuarine forested wetlands in other climate
34 zones are not included since there is no estimated loss of these forests to unvegetated open water coastal
35 wetlands across any year based on C-CAP data. For subtropical estuarine forested wetlands, Tier 1 estimates of
67 See https://github.com/Smithsonian/Coastal-Wetland-NGGI-Data-Public; accessed October 2022.
Land Use, Land-Use Change, and Forestry 6-111
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
mangrove DOM were used (IPCC 2014). Trends in land cover change are derived from the NOAA C-CAP dataset and
extrapolated to cover the entire 1990 through 2021 time series. Conversion to open water results in emissions of
all DOM C stocks during the year of conversion; therefore, emissions are calculated by multiplying the C-CAP
derived area of vegetated coastal wetlands lost that year by its Tier 1 DOM C stock.
Soil Carbon Stock Changes
Soil C stock changes are estimated for Vegetated Coastal Wetlands Converted to Unvegetated Open Water Coastal
Wetlands. Country-specific soil C stocks were updated in 2018 based upon analysis of an assembled dataset of
1,959 cores from across the conterminous United States (Holmquist et al. 2018). This analysis demonstrated that it
was not justified to stratify C stocks based upon mineral or organic soil classification, climate zone, or wetland
classes; therefore, a single soil C stock of 2701 C ha 1 was applied to all classes. Following the Tier 1 approach for
estimating CO2 emissions with extraction provided within the Wetlands Supplement, soil C loss with conversion of
Vegetated Coastal Wetlands to Unvegetated Open Water Coastal Wetlands is assumed to affect soil C stock to
one-meter depth (Holmquist et al. 2018) with all emissions occurring in the year of wetland conversion, and
multiplied by activity data of vegetated coastal wetland area converted to unvegetated open water wetlands. The
methodology follows Eq. 4.6 in the Wetlands Supplement.
Soil Methane Emissions
A Tier 1 assumption has been applied that salinity conditions are unchanged and hence CH4 emissions are assumed
to be zero with conversion of Vegetated Coastal Wetlands to Unvegetated Open Water Coastal Wetlands.
Uncertainty
Underlying uncertainties in 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 DOM stocks. Assumptions that underlie the methodological approaches
applied and uncertainties linked to interpretation of remote sensing data are also included in this uncertainty
assessment. The IPCC default assumption of 1 m of soil erosion with anthropogenic activities was adopted to
provide standardization in U.S. tidal C accounting (Holmquist et al. 2018). This depth of potentially erodible tidal
wetland soil has not been comprehensively addressed since most soil cores analyzed were shallow (e.g., less than
50 cm) and do not necessarily reflect the depth to non-wetland soil or bedrock (Holmquist et al. 2018). Uncertainty
specific to coastal wetlands include differentiation of palustrine and estuarine community classes, which
determines the soil C stock applied. Because mean soil and biomass C stocks for each available community class
are in a fairly narrow range, the same overall uncertainty was assigned to each (i.e., applying approach for
asymmetrical errors, the largest uncertainty for any soil C stock value should be applied in the calculation of error
propagation; IPCC 2000). For aboveground biomass C stocks, the mean standard error was very low and largely
influenced by the uncertainty associated with the estimated map area (Byrd et al. 2018). Uncertainty for root to
shoot ratios, which are used for quantifying belowground biomass, are derived from the Wetlands Supplement.
Uncertainty for subtropical estuarine forested wetland DOM stocks was derived from those listed for the Tier 1
estimates (IPCC 2014). Overall uncertainty of the NOAA C-CAP remote sensing product is 15 percent. This is in the
range of remote sensing methods (+/-10 to 15 percent; IPCC 2003). The combined uncertainty was calculated by
summing the squared uncertainty for each individual source (C-CAP, soil, biomass, and DOM) and taking the
square root of that total.
Uncertainty estimates are presented in Table 6-67 for each subcategory (i.e., soil C, biomass C, and DOM
emissions). The combined uncertainty across all subcategory is 32.0 percent above and below the estimate of 1.5
6-112 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
MMT CO2 Eq, which is driven by the uncertainty in the soil C estimates. In 2021, the total C flux was 1.5 MMT CO2
Eq., with lower and upper estimates of 1.0 and 2.0 MMT CO2 Eq.
Table 6-67: Approach 1 Quantitative Uncertainty Estimates for CO2 Flux Occurring within
Vegetated Coastal Wetlands Converted to Unvegetated Open Water Coastal Wetlands in
2020 (MMT CO2 Eq. and Percent)
Source
2021 Flux
Estimate
(MMT C02 Eq.)
Uncertainty Range Relative to Flux Estimate
(MMT C02 Eq.) (%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Biomass C Stock
0.06
0.05
0.08
-24.1%
24.1%
Dead Organic Matter C Stock
0.0005
0.000
0.001
-25.8%
25.8%
Soil C Stock
1.5
1.3
1.7
-15.0%
15.0%
Total Flux
1.5
1.0
2.0
-32.0%
32.0%
QA/QC and Verification
Data provided by NOAA (i.e., National LiDAR Dataset, NOS Tide Data, and C-CAP land cover and land cover change
mapping) undergo internal agency QA/QC procedures. Acceptance of final datasets into archive and dissemination
are contingent upon assurance that the data product is compliant with mandatory NOAA QA/QC requirements
(McCombs et al. 2016). QA/QC and Verification of the soil C stock dataset have been provided by the Smithsonian
Environmental Research Center and by the Coastal Wetlands project team leads who reviewed the estimates
against primary scientific literature. Biomass C stocks are derived from peer-review literature and reviewed by the
U.S. Geological Survey prior to publishing, by the peer-review process during publishing, and by the Coastal
Wetland Inventory team leads before inclusion in the Inventory. For subtropical estuarine forested wetlands, Tier 1
estimates of mangrove DOM were used (IPCC 2014). Land cover estimates were assessed to ensure that the total
land area did not change over the time series in which the inventory was developed, and were verified by a second
QA team. A team of two evaluated and verified there were no computational errors within the calculation
worksheets.
Recalculations Discussion
An update was made to the activity data to remove any estuarine forested wetland areas that were located
outside of states classified as subtropical since those wetlands fall under Forest Land Remaining Forest Land. The
resulting change in emissions and removals was negligible (±0.0001 MMT CO2 Eq.) and did not affect whether a
given year was a source or sink.
Planned Improvements
The depth of soil C affected by conversion of vegetated coastal wetlands converted to unvegetated coastal
wetlands will be updated from the IPCC default assumption of 1 m of soil erosion when mapping and modeling
advancements can quantitatively improve accuracy and precision. Improvements are underway to address this,
first conducting a review of literature publications. Until the time where these more detailed and spatially
distributed data are available, the IPCC default assumption that the top 1 m of soil is disturbed by anthropogenic
activity will be applied. This is a longer-term improvement.
More detailed research is in development that provides a longer-term assessment and more highly refined rates of
wetlands loss across the Mississippi Delta (e.g., Couvillion et al. 2016). The Mississippi Delta is the largest extent of
coastal wetlands in the United States. Higher resolution imagery analysis would improve quantification of
conversation to open water, which occurs not only at the edge of the marsh but also within the interior. Improved
Land Use, Land-Use Change, and Forestry 6-113
-------
1 mapping could provide a more refined regional Approach 2-3 land representation to support the national-scale
2 assessment provided by C-CAP.
3 An approach for calculating the fraction of remobilized coastal wetland soil C returned to the atmosphere as CO2 is
4 currently under review and may be included in future reports.
5 Research by USGS is investigating higher resolution mapping approaches to quantify conversion of coastal
6 wetlands is also underway. Such approaches may form the basis for a full Approach 3 land representation
7 assessment in future years. C-CAP data harmonization with the National Land Cover Dataset (NLCD) will be
8 incorporated into a future iteration of the Inventory.
9 Stock Changes from Unvegetated Open Water Coastal
10 Wetlands Converted to Vegetated Coastal Wetlands
11 Open water within the U.S. land base, as described in Section 6.1 Representation of the U.S. Land Base, is
12 recognized as Coastal Wetlands within this Inventory. The appearance of vegetated tidal wetlands on lands
13 previously recognized as open water reflects either the building of new vegetated marsh through sediment
14 accumulation or the transition from other lands uses through an intermediary open water stage as flooding
15 intolerant plants are displaced and then replaced by wetland plants. Biomass, DOM and soil C accumulation on
16 Unvegetated Open Water Coastal Wetlands Converted to Vegetated Coastal Wetlands begins with vegetation
17 establishment.
18 Within the United States, conversion of Unvegetated Open Water Coastal Wetlands to Vegetated Coastal
19 Wetlands is predominantly due to engineered activities, which include active restoration of wetlands (e.g.,
20 wetlands restoration in San Francisco Bay), dam removals or other means to reconnect sediment supply to the
21 nearshore (e.g., Atchafalaya Delta, Louisiana, Couvillion et al. 2011). Wetland restoration projects have been
22 ongoing in the United States since the 1970s. Early projects were small, a few hectares in size. By the 1990s,
23 restoration projects, each hundreds of hectares in size, were becoming common in major estuaries. In several
24 coastal areas e.g., San Francisco Bay, Puget Sound, Mississippi Delta and south Florida, restoration activities are in
25 planning and implementation phases, each with the goal of recovering tens of thousands of hectares of wetlands.
26 In 2021, 2,406 ha of unvegetated open water coastal wetlands were converted to vegetated coastal wetlands
27 across all wetland types and climates, which has steadily increased over the reporting period (Table 6-59). This
28 resulted in 0.007 MMT C02 Eq. (0.002 MMT C) and 0.1 MMT C02 Eq. (0.03 MMT C) sequestered in soil and
29 biomass, respectively (Table 6-68 and Table 6-69). The soil C stock has increased during the Inventory reporting
30 period, likely due to increasing vegetated coastal wetland restoration over time. While DOM C stock increases are
31 present, they are minimal in the early part of the time series and zero in the later because there are no
32 conversions from unvegetated open water coastal wetlands to subtropical estuarine forested wetlands between
33 2011 and 2016 (and by proxy through 2021), and that is the only coastal wetland type where DOM data is currently
34 available.
35 Throughout the reporting period, the amount of Open Water Coastal Wetlands Converted to Vegetated Coastal
36 Wetlands has increased over time, reflecting the increase in engineered restoration activities mentioned above.
37 Table 6-68: CO2 Flux from C Stock Changes from Unvegetated Open Water Coastal Wetlands
38 Converted to Vegetated Coastal Wetlands (MMT CO2 Eq.)
Year
1990
2005
2017
2018
2019
2020
2021
Biomass C Flux
(+)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Dead Organic Matter C Flux
(+)
(+)
0
0
0
0
0
Soil C Flux
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Total C Stock Change
(+)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
6-114 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
+ Absolute value does not exceed 0.05 MMT C02 Eq.
Notes: Parentheses indicate net sequestration. Totals may not sum due to independent rounding.
1 Table 6-69: CO2 Flux from C Stock Changes from Unvegetated Open Water Coastal Wetlands
2 Converted to Vegetated Coastal Wetlands (MMT C)
Year
1990
2005
2017
2018
2019
2020
2021
Biomass C Flux
(0.01)
(0.02)
(0.03)
(0.03)
(0.03)
(0.03)
(0.03)
Dead Organic Matter C Flux
(+)
(+)
0
0
0
0
0
Soil C Flux
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Total C Stock Change
(0.01)
(0.02)
(0.03)
(0.03)
(0.03)
(0.03)
(0.03)
+ Absolute value does not exceed 0.005 MMT C.
Notes: Parentheses indicate net sequestration. Totals may not sum due to independent rounding.
3 Methodology and Time-Series Consistency
4 The following section includes a brief description of the methodology used to estimate changes in soil, biomass
5 and DOM C stocks, and Cm emissions for Unvegetated Open Water Coastal Wetlands Converted to Vegetated
6 Coastal Wetlands.
7 Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
8 through 2021.
9 Biomass Carbon Stock Changes
10 Quantification of regional coastal wetland biomass C stock changes for palustrine and estuarine marsh vegetation
11 are presented for Unvegetated Open Water Coastal Wetlands Converted to Vegetated Coastal Wetlands on lands
12 below the elevation of high tides (taken to be mean high water spring tide elevation) according to the national
13 LiDAR dataset, the national network of tide gauges and land use histories recorded in the 1996, 2001, 2005, 2011,
14 and 2016 NOAA C-CAP surveys. C-CAP areas are calculated at the state/territory level and summed according to
15 climate zone to national values. Privately-owned and publicly-owned lands are represented. Trends in land cover
16 change are extrapolated to 1990 and 2021 from these datasets (Table 6-58). C-CAP provides peer reviewed high
17 resolution -level mapping of coastal wetland distribution, including conversion to and from open water. Biomass C
18 stock is not sensitive to soil organic content but differentiated based on climate zone. Data for non-forested
19 wetlands are derived from a national assessment combining field plot data and aboveground biomass mapping by
20 remote sensing (Table 6-61; Byrd et al. 2017; Byrd et al. 2018; Byrd et al. 2020). The aboveground biomass C stock
21 for subtropical estuarine forested wetlands (dwarf mangroves that are not classified as forests due to their stature)
22 is derived from a meta-analysis by Lu and Megonigal (20 1768). Aboveground biomass C stock data for all
23 subcategories are not available and thus assumptions were applied using expert judgment about the most
24 appropriate assignment of a C stock to a disaggregation of a community class. Root to shoot ratios from the
25 Wetlands Supplement were used to account for belowground biomass, which were multiplied by the aboveground
26 C stock (Table 6-62; IPCC 2014). Above- and belowground values were summed to obtain total biomass C stocks.
27 Conversion of open water to Vegetated Coastal Wetlands results in the establishment of a standing biomass C
28 stock; therefore, stock changes that occur are calculated by multiplying the C-CAP derived area gained that year in
29 each climate zone by its mean biomass. While the process of revegetation of unvegetated open water wetlands
68 See https://github.com/Smithsonian/Coastal-Wetland-NGGI-Data-Public: accessed September 2022.
Land Use, Land-Use Change, and Forestry 6-115
-------
1 can take many years to occur, it is assumed in the calculations that the total biomass is reached in the year of
2 conversion.
3 Dead Organic Matter
4 Dead organic matter (DOM) C stocks, which include litter and dead wood stocks, are included for subtropical
5 estuarine forested wetlands for Vegetated Coastal Wetlands Converted to Unvegetated Open Water Coastal
6 Wetlands across all years. Tier 1 default or country-specific data on DOM are not currently available for either
7 palustrine or estuarine scrub/shrub wetlands for any climate zone. Data for estuarine forested wetlands in other
8 climate zones are not included since there is no estimated loss of these forests to unvegetated open water coastal
9 wetlands across any year based on C-CAP data. Tier 1 estimates of subtropical estuarine forested wetland DOM
10 were used (IPCC 2014). Trends in land cover change are derived from the NOAA C-CAP dataset and extrapolated to
11 cover the entire 1990 through 2021 time series. Dead organic matter removals are calculated by multiplying the C-
12 CAP derived area gained that year by its Tier 1 DOM C stock. Similar to biomass C stock gains, gains in DOM can
13 take many years to occur, but for this analysis, the total DOM stock is assumed to accumulate during the first year
14 of conversion.
15 Soil Carbon Stock Change
16 Soil C stock changes are estimated for Unvegetated Open Water Coastal Wetlands Converted to Vegetated Coastal
17 Wetlands. Country-specific soil C removal factors associated with soil C accretion, stratified by climate zones and
18 wetland classes, are derived from a synthesis of peer-reviewed literature and updated this year based upon
19 refined review of the dataset (Lynch 1989; Orson et al. 1990; Kearny & Stevenson 1991; Roman et al. 1997; Craft et
20 al. 1998; Orson et al. 1998; Merrill 1999; Hussein et al. 2004; Church et al. 2006; Koster et al. 2007; Callaway et al.
21 2012 a & b; Bianchi et al. 2013; Crooks et al. 2014; Weston et al. 2014; Villa & Mitsch 2015; Marchio et al. 2016;
22 Noe et al. 2016). Soil C stock changes are stratified based upon wetland class (Estuarine, Palustrine) and subclass
23 (Emergent Marsh, Scrub Shrub). For soil C stock change, no differentiation is made for soil type (i.e., mineral,
24 organic). Soil C removal factors were developed from literature references that provided soil C removal factors
25 disaggregated by climate region and vegetation type by salinity range (estuarine or palustrine) as identified using
26 NOAA C-CAP as described above (see Table 6-63 for values).
27 Tier 2 level estimates of C stock changes associated with annual soil C accumulation in Vegetated Coastal Wetlands
28 were developed using country-specific soil C removal factors multiplied by activity data on Unvegetated Coastal
29 Wetlands converted to Vegetated Coastal Wetlands. The methodology follows Eq. 4.7, Chapter 4 of the Wetlands
30 Supplement, and is applied to the area of Unvegetated Coastal Wetlands converted to Vegetated Coastal Wetlands
31 on an annual basis.
32 Soil Methane Emissions
33 A Tier 1 assumption has been applied that salinity conditions are unchanged and hence Cm emissions are assumed
34 to be zero with conversion of Vegetated Open Water Coastal Wetlands to Vegetated Coastal Wetlands.
35 Uncertainty
36 Underlying uncertainties in estimates of soil and biomass C stock changes include uncertainties associated with
37 country-specific (Tier 2) literature values of these C stocks, assumptions that underlie the methodological
38 approaches applied and uncertainties linked to interpretation of remote sensing data. Uncertainty specific to
39 coastal wetlands include differentiation of palustrine and estuarine community classes that determines the soil C
40 stock applied. Because mean soil and biomass C stocks for each available community class are in a fairly narrow
41 range, the same overall uncertainty was applied to each, respectively (i.e., applying approach for asymmetrical
42 errors, the largest uncertainty for any soil C stock value should be applied in the calculation of error propagation;
43 IPCC 2000). For aboveground biomass C stocks, the mean standard error was very low and largely influenced by
44 error in estimated map area (Byrd et al. 2018). Uncertainty for root to shoot ratios, which are used for quantifying
6-116 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
belowground biomass (Table 6-62), are derived from the Wetlands Supplement. Uncertainty for subtropical
estuarine forested wetland DOM stocks were derived from those listed for the Tier 1 estimates (IPCC 2014).
Overall uncertainty of the NOAA C-CAP remote sensing product is 15 percent. This is in the range of remote
sensing methods (±10 to 15 percent; IPCC 2003). The combined uncertainty was calculated by summing the
squared uncertainty for each individual source (C-CAP, soil, biomass, and DOM) and taking the square root of that
total.
Uncertainty estimates are presented in Table 6-70 for each subcategory (i.e., soil C, biomass C and DOM
emissions). The combined uncertainty across all subsources is 33.4 percent above and below the estimate of-0.1
MMT CO2 Eq. In 2021, the total C flux was -0.1 MMT CO2 Eq., with lower and upper estimates of-0.1 and -0.07
MMTCO2 Eq.
Table 6-70: Approach 1 Quantitative Uncertainty Estimates for C Stock Changes Occurring
within Unvegetated Open Water Coastal Wetlands Converted to Vegetated Coastal Wetlands
in 2021 (MMT CO2 Eq. and Percent)
Source
2021 Flux Estimate
Uncertainty Range
Relative to Flux Estimate
(MMT CO? Eq.)
(MMT C02
Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Biomass C Stock Flux
(0.1)
(0.12)
(0.08)
-20.0%
20.0%
Dead Organic Matter C Stock Flux
0
0
0
-25.8%
25.8%
Soil C Stock Flux
(0.007)
(0.008)
(0.005)
-18.78%
18.1%
Total Flux
(0.1)
(0.14)
(0.07)
-33.8%
33.8%
Notes: Parentheses indicate net sequestration. Totals may not sum due to independent rounding.
QA/QC and Verification
NOAA provided data (i.e., National LiDAR Dataset, NOS Tide Data, and C-CAP land cover and land cover change
mapping), which undergo internal agency QA/QC assessment procedures. Acceptance of final datasets into the
archive for dissemination are contingent upon assurance that the product is compliant with mandatory NOAA
QA/QC requirements (McCombs et al. 2016). QA/QC and Verification of soil C stock dataset has been provided by
the Smithsonian Environmental Research Center and Coastal Wetlands project team leads who reviewed the
summary tables against primary scientific literature. Aboveground biomass C reference stocks are derived from an
analysis by the Blue Carbon Monitoring project and reviewed by U.S. Geological Survey prior to publishing, the
peer-review process during publishing, and the Coastal Wetland Inventory team leads before inclusion in the
inventory. Root to shoot ratios and DOM data are derived from peer-reviewed literature and undergo review as
per IPCC methodology. Land cover estimates were assessed to ensure that the total land area did not change over
the time series in which the inventory was developed and verified by a second QA team. A team of two evaluated
and verified there were no computational errors within calculation worksheets. Two biogeochemists at the USGS,
also members of the NASA Carbon Monitoring System Science Team, corroborated the simplifying assumption that
where salinities are unchanged CH4 emissions are constant with conversion of Unvegetated Open Water Coastal
Wetlands to Vegetated Coastal Wetlands.
Recalculations Discussion
An update was made to the activity data to remove any estuarine forested wetland areas that were located
outside of states classified as subtropical since those wetlands fall under Forests Remaining Forests. The resulting
change in emissions and removals was negligible (±0.0001 MMT CO2 Eq.) and did not affect whether a given year
was a source or sink.
Land Use, Land-Use Change, and Forestry 6-117
-------
1 Planned Improvements
2 Administered by the Smithsonian Environmental Research Center, the Coastal Wetland Carbon Research
3 Coordination Network has established a U.S. country-specific database of published data quantifying soil C stock
4 and biomass in coastal wetlands. Reference values for soil and biomass C stocks will be updated as new data
5 emerge. Refined error analysis combining land cover change, soil and biomass C stock estimates will be updated at
6 those times.
7 The USGS is investigating higher resolution mapping approaches to quantify conversion of coastal wetlands. Such
8 approaches may form the basis for a full Approach 3 land representation assessment in future years. C-CAP data
9 harmonization with the National Land Cover Dataset (NLCD) will be incorporated into a future iteration of the
10 inventory.
11 N20 Emissions from Aquaculture in Coastal Wetlands
12 Shrimp and fish cultivation in coastal areas increases nitrogen loads resulting in direct emissions of N2O. Nitrous
13 oxide is generated and emitted as a byproduct of the conversion of ammonia (contained in fish urea) to nitrate
14 through nitrification and nitrate to N2 gas through denitrification (Hu et al. 2012). Nitrous oxide emissions can be
15 readily estimated from data on fish production (IPCC 2014).
16 Aquaculture production in the United States has fluctuated slightly from year to year, with resulting N2O emissions
17 between 0.1 and 0.2 MMT CO2 Eq. between 1990 and 2021 (Table 6-71). Aquaculture production data were
18 updated through 2019; data through 2021 are not yet available and in this analysis are held constant with 2019
19 emissions of 0.2 MMT C02 Eq. (0.5 Kt N2O).
20 Table 6-71: N2O Emissions from Aquaculture in Coastal Wetlands (MMT CO2 Eq. and kt N2O)
Year
1990
2005
2017
2018
2019
2020
2021
Emissions (MMT C02 Eq.)
0.1
0.2
0.1
0.1
0.1
0.1
0.1
Emissions (kt N20)
0.4
0.6
0.5
0.5
0.5
0.5
0.5
21 Methodology and Time-Series Consistency
22 The methodology to estimate N2O emissions from Aquaculture in Coastal Wetlands follows the Tier 1 guidance in
23 the Wetlands Supplement by applying country-specific fisheries production data and the IPCC Tier 1 default
24 emission factor.
25 Each year NOAA Fisheries document the status of U.S. marine fisheries in the annual report of Fisheries of the
26 United States (National Marine Fisheries Service 2022), from which activity data for this analysis is derived.69 The
27 fisheries report has been produced in various forms for more than 100 years, primarily at the national level, on
28 U.S. recreational catch and commercial fisheries landings and values. In addition, data are reported on U.S.
29 aquaculture production, the U.S. seafood processing industry, imports and exports offish-related products, and
30 domestic supply and per capita consumption of fisheries products. Within the aquaculture chapter, the mass of
31 production for catfish, striped bass, tilapia, trout, crawfish, salmon and shrimp are reported. While some of these
32 fisheries are produced on land and some in open water cages within coastal wetlands, all have data on the
33 quantity of food stock produced, which is the activity data that is applied to the IPCC Tier 1 default emissions
34 factor to estimate emissions of N2O from aquaculture. It is not apparent from the data as to the amount of
35 aquaculture occurring above the extent of high tides on river floodplains. While some aquaculture occurs on
36 coastal lowland floodplains, this is likely a minor component of tidal aquaculture production because of the need
37 for a regular source of water for pond flushing. The estimation of N2O emissions from aquaculture is not sensitive
69 See https://www.fisheries.noaa.gov/resource/document/fisheries-united-states-2019-report; accessed August 2021.
6-118 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 to salinity using IPCC approaches, and as such, the location of aquaculture ponds within the boundaries of coastal
2 wetlands does not influence the calculations.
3 Other open water shellfisheries for which no food stock is provided, and thus no additional N inputs, are not
4 applicable for estimating N2O emissions (e.g., clams, mussels, and oysters) and have not been included in the
5 analysis. The IPCC Tier 1 default emissions factor of 0.00169 kg N2O-N per kg offish/shellfish produced is applied to
6 the activity data to calculate total N2O emissions.
7 Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
8 through 2021.
9 Uncertainty
10 Uncertainty estimates are based upon the Tier 1 default 95 percent confidence interval provided in Table 4.15,
11 chapter 4 of the Wetlands Supplement for N2O emissions and on expert judgment of the NOAA Fisheries of the
12 United States fisheries production data. Given the overestimate of fisheries production from coastal wetland areas
13 due to the inclusion offish production in non-coastal wetland areas, this is a reasonable initial first approximation
14 for an uncertainty range.
15 Uncertainty estimates for N2O emissions from aquaculture production are presented in Table 6-72 for N2O
16 emissions. The combined uncertainty is 116 percent above and below the estimate of 0.13 MMT CO2 Eq. In 2021,
17 the total flux was 0.13 MMT CO2 Eq., with lower and upper estimates of 0.00 and 0.29 MMT CO2 Eq.
18 Table 6-72: Approach 1 Quantitative Uncertainty Estimates for N2O Emissions from
19 Aquaculture Production in Coastal Wetlands in 2021 (MMT CO2 Eq. and Percent)
2021 Emissions
Estimate
Uncertainty Range Relative to Emissions Estimate3
Source
(MMT CO? Eq.)
(MMT CO?
Eq.)
(%)
Lower
Upper
Lower Upper
Bound
Bound
Bound Bound
Combined Uncertainty for N20 Emissions
for Aquaculture Production in Coastal
0.13
0.00
0.29
-116% 116%
Wetlands
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
20 QA/QC and Verification
21 NOAA provided internal QA/QC review of reported fisheries data. The Coastal Wetlands Inventory team consulted
22 with the Coordinating Lead Authors of the Coastal Wetlands chapter of the Wetlands Supplement to assess which
23 fisheries production data to include in estimating emissions from aquaculture. It was concluded that N2O emissions
24 estimates should be applied to any fish production to which food supplement is supplied be they pond or coastal
25 open water and that salinity conditions were not a determining factor in production of N2O emissions.
26 Recalculations Discussion
27 A NOAA report was released in 2022 that contains updated fisheries data through 2019 and the 2019 production
28 estimate was revised from 308,550 to 298,336 MT, although it did not affect the resulting emissions (National
29 Marine Fisheries Service 2022). The updated production value was applied for 2019, and the 2019 value was
30 applied in 2020 and 2021. This resulted in a slight reduction of N2O emissions by 0.01 MMT CO2 Eq. (0.02 kt N2O), a
31 3.3 percent decrease, for 2018 and 2019 compared to the previous Inventory.
32 In addition, the EPA updated the global warming potential (GWP) for calculating C02-equivalent emissions of N2O
33 (from 298 to 265) to reflect the 100-year GWP provided in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). The
34 previous Inventory used the 100-year GWP provided in the IPCC Fourth Assessment Report (AR4). This update was
Land Use, Land-Use Change, and Forestry 6-119
-------
1 applied across the entire time series. The net result of this change was an average annual decrease of 0.02 MMT
2 CO2 Eq.in N2O emissions from aquaculture for the 1990-2020 period. Further discussion on this update and the
3 overall impacts of updating the inventory GWP values to reflect the AR5 can be found in Chapter 9, Recalculations
4 and Improvements.
5 Together, the combined net effect of implementing these two recalculations was an average annual decrease in
6 N2O emissions from aquaculture of 13.4 percent from 1990 through 2020 compared to the previous Inventory.
7 Flooded Land Remaining Flooded Land
8 Flooded lands are defined as water bodies where human activities have 1) caused changes in the amount of
9 surface area covered by water, typically through water level regulation (e.g., constructing a dam), 2) waterbodies
10 where human activities have changed the hydrology of existing natural waterbodies thereby altering water
11 residence times and/or sedimentation rates, in turn causing changes to the natural emission of greenhouse gases,
12 and 3) waterbodies that have been created by excavation, such as canals, ditches and ponds (IPCC 2019). Flooded
13 lands include waterbodies with seasonally variable degrees of inundation, but these waterbodies would be
14 expected to retain some inundated area throughout the year under normal conditions.
15 Flooded lands are broadly classified as "reservoirs" or "other constructed waterbodies" (IPCC 2019). Other
16 constructed waterbodies include canals/ditches and ponds (flooded land <8 ha surface area). Reservoirs are
17 defined as flooded land greater than 8 ha. IPCC guidance (IPCC 2019) provides default emission factors for
18 reservoirs, ponds, and canals/ditches.
19 Land that has been flooded for greater than 20 years is defined as Flooded Land Remaining Flooded Land and land
20 flooded for 20 years or less is defined as Land Converted to Flooded Land. The distinction is based on literature
21 reports that Cm and CO2 emissions are high immediately following flooding, but decline to a steady background
22 level approximately 20 years after flooding (Abril et al. 2005, Barros et al. 2011, Teodoru et al. 2012). Emissions of
23 Cm are estimated for Flooded Land Remaining Flooded Land, but CO2 emissions are not included as they are
24 primarily the result of decomposition of organic matter entering the waterbody from the catchment or contained
25 in inundated soils and are captured in Chapter 6, Land Use, Land-Use Change, and Forestry.
26 Nitrous oxide emissions from flooded lands are largely related to input of organic or inorganic nitrogen from the
27 watershed. These inputs from runoff/leaching/deposition are largely driven by anthropogenic activities such as
28 land-use change, wastewater disposal or fertilizer application in the watershed or application of fertilizer or feed in
29 aquaculture. These emissions are not included here to avoid double-counting of N2O emissions which are captured
30 in other source categories, such as indirect N2O emissions from managed soils (Section 5.4, Agricultural Soil
31 Management)) and wastewater management (Section 7.2, Wastewater Treatment and Discharge).
32 Emissions from Flooded Land Remaining Flooded Land-
33 Reservoirs
34 Reservoirs are designed to store water for a wide range of purposes including hydropower, flood control, drinking
35 water, and irrigation. The permanently wetted portion of reservoirs are typically surrounded by periodically
36 inundated land referred to as a "drawdown zone" or "inundation area." Greenhouse gas emissions from
37 inundation areas are considered significant and similar per unit area to the emissions from the water surface and
38 are therefore included in the total reservoir surface area when estimating greenhouse gas emissions from flooded
39 land. Lakes converted into reservoirs without substantial changes in water surface area or water residence times
40 are not considered to be managed flooded land (see Area Estimates below) (IPCC 2019).
41 In 2021, the United States and Puerto Rico hosted 9.7 million hectares of reservoir surface area in the Flooded
42 Land Remaining Flooded Land category (see Methodology and Time-Series Consistency below for calculation
43 details). These reservoirs are distributed across all six of the aggregated climate zones used to define flooded land
44 emission factors (Figure 6-) (IPCC 2019).
6-120 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Figure 6-10: U.S. Reservoirs (black polygons) in the Flooded Land Remaining Flooded Land
Category in 2021.
Climate Zone
[~~1 boreal
0 cool temperate
tropical dry/montane
| tropical moist/wet
1 I warm temperate dry
O warm temperate moist
Alaska
100 mi
500 mi
Note: Colors represent climate zone used to derive IPCC default emission factors.
Methane is produced in reservoirs through the microbial breakdown of organic matter. Per unit area, ChU emission
rates tend to scale positively with temperature and system productivity (i.e., abundance of algae), but negatively
with system size (i.e., depth, surface area). Methane produced in reservoirs can be emitted from the reservoir
surface or exported from the reservoir when ChUrich water passes through the dam. This exported CHa can be
released to the atmosphere as the water passes through hydropower turbines or the downstream river channel.
Methane emitted to the atmosphere via this pathway is referred to as "downstream emissions."
Table 6-73 and Table 6-74 below summarize nationally aggregated CH4 emissions from reservoirs. The increase in
CH4 emissions through the time series is attributable to reservoirs matriculating from the Land Converted to
Flooded Land category into the Flooded Land Remaining Flooded Land category.
Table 6-73: CH4 Emissions from Flooded Land Remaining Flooded Land—Reservoirs (MMT
CO2 Eq.)
Source 1990
2005
2017 2018 2019 2020 2021
Reservoirs
Surface Emission 25.9
Downstream Emission 2.3
26.4
2,4
26.5 26.5 26.5 26.5 26.5
2,4 2.4 2.4 2.4 2,4
Total 28.2
28.8
28.9 28.9 28.9 28.9 28.9
Note: Totals may riot sum to due independent rounding.
Table 6-74: CH4 Emissions from Flooded Land Remaining Flooded Land—Reservoirs (kt ChU)
Source 1990
2005
2017 2018 2019 2020 2021
Reservoirs
Surface Emission 924
Downstream Emission 83
943
85
946 946 946 948 948
85 85 85 85 85
Total 1,007
1,028
1,032 1,032 1,032 1,033 1,033
Note: Totals may not sum to due independent rounding.
Land Use, Land-Use Change, and Forestry 6-121
-------
1
2 Methane emissions from reservoirs in Texas, Florida, and Louisiana (Figure 6-11, Table 6-75) compose 33 percent
3 of national Cm emissions from reservoirs in 2021. Emissions from these states are particularly high due to 1) the
4 large expanse of reservoirs in these states (Table 6-78) and 2) the high CFU emission factor for the tropical
5 dry/montane and topical moist climate zones which encompass a majority of the flooded land area in these states
6 (Figure 6-, Table 6-76).
7 Methane emissions from reservoirs in Flooded Land Remaining Flooded Land increased 2.5 percent from 1990 to
8 2021 due to the matriculation of reservoirs in Land Converted to Flooded Land to Flooded Land Remaining Flooded
9 Land.
10 Figure 6-11: Total ChU Emissions (Downstream + Surface) from Reservoirs in Flooded Land
11 Remaining Flooded Land in 2021 (kt ChU)
100 mi 500 mi
12
13
14 Table 6-75: Surface and Downstream ChU Emissions from Reservoirs in Flooded Land
15 Remaining Flooded Land in 2021 (kt ChU)
State
Surface
Downstream
Total
Alabama
24
2
26
Alaska
1
+
1
Arizona
15
1
16
Arkansas
26
2
28
California
39
4
43
Colorado
6
1
7
Connecticut
3
+
3
Delaware
3
+
3
District of Columbia
+
+
+
Florida
126
11
137
Georgia
35
3
38
Hawaii
1
+
1
Idaho
10
1
10
Illinois
17
2
19
6-122 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
Indiana
6
1
6
Iowa
6
1
6
Kansas
9
1
9
Kentucky
13
1
14
Louisiana
59
5
65
Maine
13
1
15
Maryland
13
1
14
Massachusetts
5
+
5
Michigan
9
1
9
Minnesota
17
2
18
Mississippi
19
2
20
Missouri
17
2
19
Montana
14
1
15
Nebraska
11
1
12
Nevada
17
2
18
New Hampshire
3
+
3
New Jersey
11
1
12
New Mexico
5
+
6
New York
12
1
14
North Carolina
32
3
35
North Dakota
14
1
15
Ohio
6
1
7
Oklahoma
24
2
26
Oregon
16
1
17
Pennsylvania
6
1
6
Puerto Rico
+
+
+
Rhode Island
1
+
1
South Carolina
37
3
40
South Dakota
13
1
14
Tennessee
18
2
20
Texas
128
11
139
Utah
22
2
24
Vermont
2
+
2
Virginia
24
2
26
Washington
25
2
27
West Virginia
2
+
2
Wisconsin
10
1
11
Wyoming
5
+
5
+ Indicates values less than 0.5 kt
Methodology and Time-Series Consistency
Estimates of CFU emission for reservoirs in Flooded Land Remaining Flooded Land follow the Tier 1 methodology in
the 2019 Refinement to the 2006IPCC Guidelines (IPCC 2019). Methane emissions from the surface of these
flooded lands are calculated as the product of flooded land surface area and a climate-specific emission factor
(Table 6-76). Downstream emissions are calculated as 9 percent of the surface emission (Tier 1 default). Total CFU
emissions from reservoirs are calculated as the sum of surface and downstream emissions. National emissions are
calculated as the sum of state emissions.
The IPCC default surface emission factors used in the Tier 1 methodology are derived from model-predicted (G-res
model, Prairie et al. 2017) emission rates for all reservoirs in the Global Reservoir and Dam (GRanD) database
(Lehner et al. 2011). Predicted emission rates were aggregated by the 11 IPCC climate zones (IPCC 2019, Table
7A.2) which were collapsed into six climate zones using a regression tree approach. All six aggregated climate zone
are present in the United States.
Land Use, Land-Use Change, and Forestry 6-123
-------
1 Table 6-76: IPCC (2019) Default ChU Emission Factors for Surface Emission from Reservoirs
2 in Flooded Land Remaining Flooded Land
Climate
Surface emission factor
(MT CH4 ha1 y1)
Boreal
0.0136
Cool Temperate
0.054
Warm Temperate Dry
0.1509
Warm Temperate Moist
0.0803
Tropical Dry/Montane
0.2837
Tropical Moist/Wet
0.1411
Note: downstream CH4 emissions are calculated as 9 percent of surface
emissions. Downstream emissions are not calculated for C02.
3 Area estimates
4 U.S. reservoirs were identified from the NHDWaterbody layer in the National Hydrography Dataset Plus V2
5 (NHD)70, the National Inventory of Dams (NID)71, the National Wetlands Inventory (NWI)72, and the Navigable
6 Waterways (NW) network73. The NHD only covers the conterminous U.S., whereas the NID, NW and NWI also
7 include Alaska, Hawaii, and Puerto Rico.
8 Waterbodies in the NHDWaterbody layer that were greater than or equal to 8 ha in surface area, not identified as
9 canal/ditch in NHD, and met any of the following criteria were considered reservoirs: 1) the waterbody was
10 classified as "Reservoir" in the NHDWaterbody layer, 2) the waterbody name in the NHDWaterbody layer included
11 "Reservoir", 3) the waterbody in the NHDWaterbody layer was located in close proximity (up to 100 m) to a dam in
12 the NID, 4) the NHDWaterbody GNIS name was similar to a nearby NID feature (between 100 m to 1000 m).
13 EPA assumes that all features included in the NW network are subject to water-level management to maintain
14 minimum water depths required for navigation and are therefore managed flooded lands. Navigable Waterway
15 features greater than 8 ha in surface area are defined as reservoirs.
16 NWI features were considered "managed" if they had a Special Modifier value indicating the presence of
17 management activities (Figure 6-12). To be included in the flooded lands inventory, the managed flooded land had
18 to be wet or saturated for at least one season per year (see 'Water Regime' in Figure 6-12). NWI features that met
19 these criteria, were greater than 8 ha in surface area, and were not a canal/ditch (see Emissions from Land
20 Converted to Flooded Land - Other Constructed Waterbodies) were defined as reservoirs.
21 Surface areas for identified flooded lands were taken from the NHD, NWI or NW. If features from the NHD, NWI, or
22 NW datasets overlapped, duplicated areas were erased. The first step was to take the final NWI Flooded Lands
23 features and use it to identify overlapping NHD features. If the NHD feature had its center in a NWI feature, it was
24 removed from analysis. Next, remaining NHD features were erased from any remaining overlapping NWI features.
25 Final selections of NHD and NWI features were used to erase any overlapping NW waterbodies.
26 Reservoir age was determined by assuming the waterbody was created the same year as a nearby (up to 100 m)
27 NID feature. If no nearby NID feature was identified, it was assumed the waterbody was greater than 20-years old
28 throughout the time series.
29
70 See https://www.usgs.gov/core-science-systems/ngp/national-hydrography
71 See https://nid.sec.usace.armv.mil.
72 See https://www.fws.gov/program/national-wetlands-inventorv/data-download
73 See https://hifld-geoplatform.opendata.arcgis.com/maps/aaa3767c7d2b41f69e7528f99cf2fb76 Q/about
6-124 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Figure 6-12: Selected Features from NWI that Meet Flooded Lands Criteria
MODIFIERS
In order to more adequately descnbe the wetland and deepwater habitats, one each of the water regime, water chemistry, soil, or
special modifiers may be applied at the class or lower level in the hierarchy
Water Regime
Special Modifiers
Water Chemistry
Soil
Nontidal
A Temporarily Flooded
B Seasonally Saturated
Saltwater Tidal
L Subtidal
Freshwater Tidal
Q Regularly Flooded-Fresh Tidal
R Seasonally Flooded-Fresh Tidal
b Beaver
Halinity/Salinity pH Modifiers for
Fresh Water
1 Hyperhaline / Hypersaline a Acid
2 Euhaline / Eusaline t Circumneutral
3 Mixohaline / M ixosaline (Brackish) i Alkaline
4 Polyhaline
5 Mesohaline
6 Oligohaline
0 Fresh
g Organic
n Mineral
M Irreqularlv Exposed
d Partly Drained/Ditched
f Farmed
m Managed
h Diked/Impounded
r Artificial Substrate
s SdoII
x Excavated |
C Seasonally Flooded
N Regularly Flooded
P Irregularly Flooded
S Temporarily Flooded- Fresh Tidal
D Continuously Saturated
T Semipermanently Flooded-Fresh Tidal
V Permanently Flooded-Fresh Tidal
E Seasonally Flooded /
Saturated
F Semipermanently Flooded
G Intermittently Exposed
H Permanently Flooded
J Intermittently Flooded
K Artificially Flooded |
Must also meet one selected special modifier (red box) to be included in the flooded lands inventory
I ~| Included in the flooded lands inventory if it meets water regime qualifier (gold box)
2 Source (modified): https://www.fws.gov/sites/default/files/documents/wetlands-and-deepwater-map-code-diagram.pdf
3 IPCC (2019) allows for the exclusion of managed waterbodies from the inventory if the water surface area or
4 residence time was not substantially changed by the construction of the dam. The guidance does not quantify
5 what constitutes a "substantial" change, but here EPA excludes the U.S. Great Lakes from the inventory based on
6 expert judgment that neither the surface area nor water residence time was substantially altered by their
7 associated dams.
8 Reservoirs were disaggregated by state (using boundaries from the 2016 U.S. Census Bureau74) and climate zone.
9 Downstream and surface emissions for cross-state reservoirs were allocated to states based on the surface area
10 that the reservoir occupied in each state. Only the U.S. portion of reservoirs that cross country borders were
11 included in the inventory.
12 The surface area of reservoirs in Flooded Land Remaining Flooded Land increased by approximately 4 percent from
13 1990 to 2021 (Table 6-77) due to reservoirs matriculating into Flooded Land Remaining Flooded Land when they
14 reached 20 years of age.
15 Table 6-77: National Totals of Reservoir Surface Area in Flooded Land Remaining Flooded
16 Land (millions of ha)
Surface Area (millions of ha)
1990
2005
2017
2018
2019
2020
2021
Reservoir
9.40
9.61
9.64
9.65
9.65
9.67
9.67
17
18 Table 6-78: State Breakdown of Reservoir Surface Area in Flooded Land Remaining Flooded
19 Land (millions of ha)
State
1990
2005
2017
2018
2019
2020
2021
Alabama
0.24
0.24
0.24
0.24
0.24
0.24
0.24
Alaska
0.02
0.02
0.02
0.02
0.02
0.02
0.02
Arizona
0.06
0.06
0,06
0,06
0,06
0,06
0,06
Arkansas
0.29
0.30
0.30
0.30
0.30
0.30
0.30
California
0.35
0.36
0.36
0.36
0.36
0.36
0.36
Colorado
0.03
0.08
0.08
0.08
0.08
0.08
0.08
Connecticut
0.03
0.03
0.03
0.03
0.03
0.03
0.03
Delaware
0.03
0.03
0.03
0.03
0.03
0.03
0.03
District of Columbia
+
+
+
+
+
+
+
Florida
0.88
0.89
0,89
0,89
0,89
0,89
0,89
74 See https://www.census.gov/geographies/mapping-files/time-series/geo/carto-boundarv-file.html.
Land Use, Land-Use Change, and Forestry 6-125
-------
Georgia
0.29
0.30
0.30
0.30
0.30
0.30
0.30
Hawaii
+
+
+
+
+
+
+
Idaho
0.13
0.15
0.15
0.15
0.15
0.15
0.15
Illinois
0.21
0.22
0.22
0.22
0.22
0.23
0.23
Indiana
0.06
0.07
0.07
0.07
0.07
0.07
0.07
Iowa
0.07
0.08
0.08
0.08
0.08
0.08
0.08
Kansas
0.07
0.09
0.09
0.09
0.09
0.09
0.09
Kentucky
0.15
0.16
0.16
0.16
0.16
0.16
0.16
Louisiana
0.41
0.42
0.42
0.42
0.42
0.42
0.42
Maine
0.23
0.24
0.25
0.25
0.25
0.25
0.25
Maryland
0.16
0.16
0.16
0.16
0.16
0.16
0.16
Massachusetts
0.07
0.07
0.07
0.07
0.07
0.07
0.07
Michigan
0.15
0.15
0.15
0.15
0.15
0.15
0.15
Minnesota
0.30
0.31
0.31
0.31
0.31
0.31
0.31
Mississippi
0.18
0.18
0.18
0.18
0.18
0.18
0.18
Missouri
0.20
0.20
0.20
0.20
0.20
0.21
0.21
Montana
0.24
0.26
0.26
0.26
0.26
0.26
0.26
Nebraska
0.13
0.13
0.13
0.13
0.13
0.13
0.13
Nevada
0.09
0.09
0.09
0.09
0.09
0.09
0.09
New Hampshire
0.06
0.06
0.06
0.06
0.06
0.06
0.06
New Jersey
0.13
0.13
0.13
0.13
0.13
0.13
0.13
New Mexico
0.05
0.05
0.05
0.05
0.05
0.05
0.05
New York
0.21
0.21
0.21
0.21
0.21
0.21
0.21
North Carolina
0.40
0.40
0.40
0.40
0.40
0.40
0.40
North Dakota
0.25
0.25
0.25
0.25
0.25
0.25
0.25
Ohio
0.08
0.08
0.08
0.08
0.08
0.08
0.08
Oklahoma
0.27
0.27
0.27
0.27
0.27
0.27
0.27
Oregon
0.21
0.21
0.21
0.21
0.21
0.21
0.21
Pennsylvania
0.08
0.08
0.08
0.08
0.08
0.08
0.08
Puerto Rico
+
+
+
+
+
+
+
Rhode Island
0.02
0.02
0.02
0.02
0.02
0.02
0.02
South Carolina
0.31
0.32
0.33
0.33
0.33
0.33
0.33
South Dakota
0.24
0.24
0.24
0.24
0.24
0.24
0.24
Tennessee
0.22
0.23
0.23
0.23
0.23
0.23
0.23
Texas
0.66
0.67
0.67
0.67
0.67
0.67
0.67
Utah
0.18
0.19
0.19
0.19
0.19
0.19
0.19
Vermont
0.04
0.04
0.04
0.04
0.04
0.04
0.04
Virginia
0.30
0.30
0.30
0.30
0.30
0.30
0.30
Washington
0.26
0.26
0.26
0.26
0.26
0.26
0.26
West Virginia
0.03
0.03
0.03
0.03
0.03
0.03
0.03
Wisconsin
0.18
0.18
0.18
0.18
0.18
0.18
0.18
Wyoming
0.09
0.09
0.09
0.09
0.09
0.09
0.09
Total
9.40
9.61
9.64
9.65
9.65
9.67
9.67
+ Indicates values less than 0.005 million Ha
Note: Totals may not sum due to independent rounding.
1 Uncertainty
2 Uncertainty in estimates of Cm emissions from reservoirs in Flooded Land Remaining Flooded Land (Table 6-79)
3 are developed using the IPCC Approach 2 and include uncertainty in the default emission factors and land areas.
4 Uncertainty ranges for the emission factors are provided in the 2019 Refinement to the 2006 IPCC Guidelines (IPCC
5 2019). Uncertainties in the spatial data include 1) uncertainty in area estimates from the NHD, NWI, and NW, and
6 2) uncertainty in the location of dams in the NID. Overall uncertainties in these spatial datasets are unknown, but
7 uncertainty for remote sensing products is assumed to be ± 10 -15 percent based on IPCC guidance (IPCC 2003).
8 An uncertainty range of ± 15 percent for the reservoir area estimates is assumed and is based on expert judgment.
6-126 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Table 6-79: Approach 2 Quantitative Uncertainty Estimates for ChU Emissions from
2 Reservoirs in Flooded Land Remaining Flooded Land
Source
Gas
2021 Emission Estimate
Uncertainty Range Relative to Emission
Estimate3
(MMT CO? Eq.)
(MMT CO:
i Eq-)
(%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Reservoir
Surface
Downstream
ch4
ch4
26.5
2.39
26.2
2.32
26.8
2.7
-1.2%
-3%
1.1%
13%
Total
ch4
28.9
28.6
29.4
-1%
1.7%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
3 QA/QC and Verification
4 The National Hydrography Data (NHD) is managed by the USGS in collaboration with many other federal, state, and
5 local entities. Extensive QA/QC procedures are incorporated into the curation of the NHD. The National Inventory
6 of Dams (NID) is maintained by the U.S. Army Corps of Engineers (USACE) in collaboration with the Federal
7 Emergency Management Agency (FEMA) and state regulatory offices. USACE resolves duplicative and conflicting
8 data from 68 data sources, which helps obtain the more complete, accurate, and updated NID. The Navigable
9 Waterways (NW) dataset is part of the U.S. Department of Transportation (USDOT)/Bureau of Transportation
10 Statistics (BTS) National Transportation Atlas Database (NTAD). The NW is a comprehensive network database of
11 the nation's navigable waterways updated on a continuing basis. U.S. Fish and Wildlife Service is the principal
12 agency in charge of wetland mapping including the National Wetlands Inventory (NWI). Quality and consistency of
13 the Wetlands Layer is supported by federal wetlands mapping and classification standards, which were developed
14 under the oversight of the Federal Geographic Data Committee (FGDC) with input by the FGDC Wetlands
15 Subcommittee. This dataset is part of the FGDC Water-Inland Theme, which is co-chaired by the FWS and the U.S.
16 Geological Survey.
17 General QA/QC procedures were applied to activity data, documentation, and emission calculations consistent
18 with the U.S. Inventory QA/QC plan, which is in accordance with Vol. 1 Chapter 6 of the 2006IPCC Guidelines (see
19 Annex 8 for more details). All calculations were executed independently in Excel and R. Ten percent of state and
20 national totals were randomly selected for comparison between the two approaches to ensure there were no
21 computational errors.
22 Recalculations Discussion
23 The 1990 through 2021 Inventory uses the National Wetland Inventory (NWI) as the primary data source for
24 flooded land surface area, whereas the 1990 through 2020 Inventory report used the National Hydrography Data
25 (NHD) as the primary geospatial data source. The NWI is far more detailed than the NHD, resulting in increased
26 emission estimates across the time series. The NWI also includes Alaska, Hawaii, and Puerto Rico which were not
27 estimated in the 1990 through 2020 Inventory.
28 Emissions from reservoirs in Flooded Land Remaining Flooded Land were further increased by correcting the
29 creation date of several large reservoirs in South Dakota, North Dakota, Alabama, Arkansas, Georgia, and South
30 Carolina. These reservoirs were incorrectly classified as Land Converted to Flooded Land for a portion of the 1990-
31 2020 time series, but are classified as Flooded Land Remaining Flooded Land throughout the 1990 through 2021
32 Inventory time series.
33 The 1990 through 2020 Inventory distinguished between reservoirs and inundation areas. Inundation areas were
34 defined as periodically flooded lands that bordered a permanently flooded reservoir. The NWI includes both
35 permanently and periodically flooded lands, but doesn't consistently discriminate between them, therefore
36 inundation areas and reservoirs are consolidated into reservoirs for the 1990 through 2021 Inventory.
Land Use, Land-Use Change, and Forestry 6-127
-------
1 In addition, the EPA updated the global warming potential (GWP) for Cm (from 25 to 28) to reflect the 100-year
2 GWP provided in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). The previous Inventory used the 100-year
3 GWP provided in the IPCC Fourth Assessment Report (AR4). This update was applied across the entire time series.
4 Further discussion on this update and the overall impacts of updating the inventory GWP values to reflect the AR5
5 can be found in Chapter 9, Recalculations and Improvements.
6 The net effect of these recalculations was an average annual increase in Cm emission estimates from reservoirs of
7 10.3 MMT CO2 Eq., or 56 percent, over the time series from 1990 to 2020 compared to the previous Inventory.
8 Planned Improvements
9 The EPA is currently measuring greenhouse gas emissions from 108 reservoirs in the conterminous United States.
10 The survey will be complete in September 2023 and the data will be used to develop country-specific emission
11 factors for U.S. reservoirs. At the earliest, these emission factors will be used in the 2025 Inventory submission.
12 Emissions from Flooded Land Remaining Flooded Land-Other
13 Constructed Waterbodies
14 The IPCC (IPCC 2019) provides emission factors for several types of "other constructed waterbodies" including
15 freshwater ponds and canals/ditches. IPCC (2019) describes ponds as waterbodies that are "...constructed by
16 excavation and/or construction of walls to hold water in the landscape for a range of uses, including agricultural
17 water storage, access to water for livestock, recreation, and aquaculture." Furthermore, the IPCC "Decision tree
18 for types of Flooded Land" (IPCC 2019, Fig. 7.2) defines a size threshold of 8 ha to distinguish reservoirs from
19 "other constructed waterbodies." For this Inventory, ponds are defined as managed flooded land that are 1) less
20 than 8 ha in surface area, and 2) not categorized as canals/ditches. IPCC (2019) further distinguishes saline versus
21 brackish ponds, with the former supporting lower CFU emissions than the latter. Activity data on pond salinity are
22 not uniformly available for the conterminous United States and all ponds in the inventory are assumed to be
23 freshwater. Ponds often receive high organic matter and nutrient loadings, may have low oxygen levels, and are
24 often sites of substantial CFU emissions from anaerobic sediments.
25 Canals and ditches (terms are used interchangeably) are linear water features constructed to transport water (i.e.,
26 stormwater drainage, aqueduct), to irrigate or drain land, to connect two or more bodies of water, or to serve as a
27 waterway for watercraft. The geometry and construction of canals and ditches varies widely and includes narrow
28 earthen channels (<1 m wide) and concrete lined aqueducts in excess of 50 m wide. Canals and ditches can be
29 extensive in many agricultural, forest and settlement areas, and may also be significant sources of emissions in
30 some circumstances.
31 Methane emissions from freshwater ponds in Flooded Land Remaining Flooded Land increased by less than 1
32 percent from 1990 to 2021. Methane emissions from canals and ditches have remained constant throughout the
33 time series because age data are not available for canals and ditches, thus they are assumed to be greater than 20-
34 years old in 1990 and are included in Flooded Land Remaining Flooded Land throughout the time series. Overall,
35 Cm emissions from other constructed waterbodies have remained fairly constant since 1990 (Table 6-80 and Table
36 6-81).
37 Table 6-80: ChU Emissions from Other Constructed Waterbodies in Flooded Land Remaining
38 Flooded Land (MMT COz Eq.)
Source
1990
2005
2017
2018
2019
2020
2021
Other Constructed Waterbodies
Canals and Ditches
2.3
2.3
2.3
2.3
2.3
2.3
2.3
Freshwater Ponds
14.1
14.2
14.2
14.2
14.2
14.2
14.2
Total
16.4
16.5
16.5
16.5
16.5
16.5
16.5
Note: Totals may not sum due to independent rounding.
6-128 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Table 6-81: ChU Emissions from Other Constructed Waterbodies in Flooded Land Remaining
Flooded Land (kt ChU)
Source
1990
2005
2017
2018
2019
2020
2021
Other Constructed Waterbodies
Canals and Ditches
80.9
80.9
80.9
80.9
80.9
80.9
80.9
Freshwater Ponds
505.2
507.8
508.3
508.3
508.4
508.4
508.5
Total
586.0
588.7
589.2
589.2
589.3
589.3
589.3
Note: Totals may not sum due to independent rounding.
Florida and Louisiana have the greatest methane emissions from canals and ditches in the United States (Figure
6-13, Table 6-82). Presumably, most of these canals serve to drain the extensive wetland complexes in these states
(Davis, 1973). California has the third greatest methane emissions from canals and ditches. Canals and ditches in
California primarily serve to convey water from the mountains to urban and agricultural areas. Michigan and
Minnesota have the fourth and fifth largest methane emissions from canals and ditches. These systems serve to
drain historic wetlands to facilitate row-crop agriculture. Florida, Texas, and Georgia have the greatest methane
emissions from freshwater ponds, although states throughout the eastern United States make significant
contributions to the national total. These patterns of emissions are in accordance with the distribution of other
constructed waterbodies in the United States.
Table 6-82: ChU Emissions from Other Constructed Waterbodies in Flooded Land Remaining
Flooded Land in 2021 (kt ChU)
State
Canals and Ditches
Freshwater Ponds
Total
Alabama
+
12.4
12.5
Alaska
+
+
+
Arizona
1.5
1.1
2.6
Arkansas
3.1
11.4
14.5
California
7.0
13.5
20.4
Colorado
2.9
5.7
8.6
Connecticut
+
2.4
2.4
Delaware
+
1.3
1.3
District of Columbia
+
+
+
Florida
15.6
47.8
63.4
Georgia
+
26.0
26.2
Hawaii
+
+
0.6
Idaho
1.7
3.8
5.5
Illinois
1.0
14.3
15.3
Indiana
1.7
11.7
13.4
Iowa
+
13.0
13.4
Kansas
+
16.3
16.4
Kentucky
+
8.3
8.5
Louisiana
9.4
8.9
18.3
Maine
+
5.6
5.6
Maryland
+
2.7
3.1
Massachusetts
+
3.2
3.2
Michigan
5.4
12.1
17.5
Minnesota
4.7
16.2
20.9
Mississippi
1.6
14.0
15.6
Missouri
2.4
23.1
25.4
Montana
2.0
12.0
14.0
Nebraska
2.0
19.6
21.6
Nevada
0.7
1.0
1.7
New Hampshire
+
1.6
1.6
New Jersey
+
4.7
5.1
New Mexico
0.8
2.4
3.2
Land Use, Land-Use Change, and Forestry 6-129
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
New York
+
10.9
11.3
North Carolina
2.6
16.8
19.4
North Dakota
0.8
13.1
13.9
Ohio
0.8
9.2
10.0
Oklahoma
+
21.8
21.9
Oregon
1.0
5.5
6.5
Pennsylvania
+
4.5
4.6
Puerto Rico
+
+
+
Rhode Island
+
+
+
South Carolina
1.3
14.5
15.8
South Dakota
+
18.4
18.7
Tennessee
+
8.7
8.9
Texas
4.6
38.6
43.2
Utah
0.8
3.3
4.1
Vermont
+
1.0
1.1
Virginia
0.5
9.6
10.1
Washington
+
3.1
3.6
West Virginia
+
1.6
1.6
Wisconsin
+
4.1
4.4
Wyoming
0.9
6.0
6.8
Total
80.9
508.5
589.3
+ Indicates values less than 0.5 kt
Note: Totals may not sum due to independent rounding.
Figure 6-13: 2021 ChU Emissions from A) Ditches and Canals and B) Freshwater Ponds in
Flooded Land Remaining Flooded Land (kt ChU)
A. CH4 Emissions from Ditches and Canals B. CH4 Emissions from Freshwater Ponds
Methodology and Time-Series Consistency
Estimates of CFU emissions for other constructed waterbodies in Flooded Land Remaining Flooded Land follow the
Tier 1 methodology in IPCC (2019). All calculations are performed at the state level and summed to obtain national
estimates. Based on IPCC guidance, methane emissions from the surface of these flooded lands are calculated as
the product of flooded land surface area and an emission factor (Table 6-83). Although literature data on
greenhouse gas emissions from canals and ditches is relatively sparse, they have the highest default emission
factor of all flooded land types (Table 6-83). Default emission factors for freshwater ponds are on the higher end of
those for reservoirs. There are insufficient data to support climate-specific emission factors for ponds or canals and
ditches. Downstream emissions are not inventoried for other constructed waterbodies because 1) many of these
systems are not associated with dams (e.g., excavated ponds and ditches), and 2) there are insufficient data to
derive downstream emission factors for other constructed waterbodies that are associated with dams (IPCC 2019).
6-130 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Table 6-83: IPCC (2019) Default ChU Emission Factors for Surface Emissions from Other
2 Constructed Waterbodies in Flooded Land Remaining Flooded Land
Surface emission factor
Other Constructed Waterbody
(MT CH4 ha1 y1)
Freshwater ponds
0.183
Canals and ditches
0.416
3 Area estimates
4 Other constructed waterbodies were identified from the NHDWaterbody layer in the National Hydrography
5 Dataset Plus V2 (NHD)75, the National Inventory of Dams (NID)76, the National Wetlands Inventory (NWI)77, and
6 the Navigable Waterways (NW) network.78 The NHD only covers the conterminous US, whereas the NID, NW and
7 NWI also include Alaska, Hawaii, District of Columbia, and Puerto Rico. The following paragraphs present the
8 criteria used to identify other constructed waterbodies in the NHD, NW, and NWI.
9 Waterbodies in the NHDWaterbody layer that were greater than 20-years old, less than 8 ha in surface area, not
10 identified as canal/ditch in NHD, and met any of the following criteria were considered freshwater ponds in
11 Flooded Land Remaining Flooded Land: 1) the waterbody was classified "Reservoir" in the NHDWaterbody layer, 2)
12 the waterbody name in the NHDWaterbody layer included "Reservoir", 3) the waterbody in the NHDWaterbody
13 layer was located in close proximity (up to 100 m) to a dam in the NID, 4) the NHDWaterbody GNIS name was
14 similar to nearby NID feature (between 100 m to 1000 m).
15 EPA assumes that all features included in the NW are subject to water-level management to maintain minimum
16 water depths required for navigation and are therefore managed flooded lands. NW features that were less than 8
17 ha in surface area and not identified as canals/ditch (see below) were considered freshwater ponds. Only 2.1
18 percent of NW features met these criteria, and they were primarily associated with larger navigable waterways,
19 such as lock chambers on impounded rivers.
20 NWI features were considered "managed" if they had a special modifier value indicating the presence of
21 management activities (Figure 6-12). To be included in the flooded lands inventory, the managed flooded land had
22 to be wet or saturated for at least one season per year (see "Water Regime' in Figure 6-12). NWI features that met
23 these criteria, were less than 8 ha in surface area, and were not a canal/ditch (see below) were defined as
24 freshwater ponds.
25 Canals and ditches, a subset of other constructed waterbodies, were identified in the NWI by their morphology.
26 Unlike a natural water body, canals and ditches are typically narrow, linear features with abrupt angular turns.
27 Figure 6-14 contrasts the unique shape of ditches/canals vs more natural water features.
28
75 See https://www.usgs.gov/core-science-svstems/ngp/national-hvdrography.
76 See https://nid.sec.usace.armv.mil.
77 See https://www.fws.gov/program/national-wetlands-inventorv/data-download
78 See https://hifld-geoplatform.opendata.arcgis.com/maps/aaa3767c7d2b41f69e7528f99cf2fb76_0/about
Land Use, Land-Use Change, and Forestry 6-131
-------
1 Figure 6-14: Left: NWI Features Identified as Carta Is/Ditches (pink) by Unique Narrow,
2 Linear/Angular Morphology. Right: Non-Canal/Ditches with More Natural Morphology (blue)
rf'm
3 This morphology was identified systematically using shape attributes in a decision tree model. A training set of 752
4 features were identified as either "ditch" or "not ditch" using expert judgment. The training set was used to train a
5 decision tree which was used to categorize millions of NWI features based on three shape attribute ratios (Figure
6 6-12).
7 Table 8-84: Predictors used in Decision Tree to Identify Canal/Ditches
Shape Length : # of Shape Vertices
Shape Area : Shape Length
Shape Area : # of Shape Vertices
8 The decision tree built a model using 80 percent of the 752 training features and used the 20 percent to validate
9 the model. The model was 93.1 percent accurate. Below are the validation results (Table 6-85).
10 Table 6-85: Validation Results for Ditch/Canal Classification Decision Tree
Truth
Prediction Ditch/Canal Not Ditch/Canal
Ditch/Canal 49 5
Not Ditch/Canal 8 27
11 The decision tree model was then applied to the entire NWI dataset using the following shape attribute ratios
12 (Figure 6-15).
13
6-132 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Figure 6-15: Structure of Decision Tree Used to Identify Canals/Ditches
NOT DITCH
0.71
100%
-fycV t Area_Length < 2.2 -fi»T
DITCH
0.14
27% J
Area Vertices >= 9.8
1
DITCH
0 03
23%
NOT DITCH
092
73%
Length_Vertlces >= 34
Area Vertices < 474
NOT DITCH
0 82
NOT DITCH
0 96
67%
Surface areas for other constructed waterbodies were taken from NHD, NWI or the NW. If features from the NHD,
NWI, or the NW datasets overlapped, these areas were erased. The first step was to take the final NWI Flooded
Lands features and use it to identify overlapping NHD features. If the NHD feature had its center in a NWI feature,
it was removed from analysis. Next, remaining NHD features were erased from any remaining overlapping NWI
features. Final selections of NHD and NWI features were used to erase any overlapping NW waterbodies.
The age of other constructed waterbody features was determined by assuming the waterbody was created the
same year as a nearby (up to 100 m) NID feature. If no nearby NID feature was identified, it was assumed the
waterbody was greater than 20-years old throughout the time series. No canal/ditch features were associated with
a nearby dam, therefore all canal/ditch features were assumed to be greater than 20-years old through the time
series.
For the year 2021, this Inventory contains 2,778,529 ha of freshwater ponds and 194,412 ha of canals and ditches
in Flooded Land Remaining Flooded Land (Table 6-86). The surface area of freshwater ponds increased by 18,069
Ha (0.6 percent) from 1990 to 2021 due to flooded lands matriculating from Land Converted to Flooded Land to
Flooded Land Remaining Flooded Land. All canals and ditches were assumed to be greater than 20-years old
throughout the time series, thus the surface area of these flooded lands is constant throughout the time series.
Table 6-86: National Surface Area Totals in Flooded Land Remaining Flooded Land - Other
Constructed Waterbodies (ha)
1990
2005
2017 2018 2019 2020 2021
Canals and ditches 194,412
Freshwater ponds 2,760,460
194,412
2,775,096
194,412 194,412 194,412 194,412 194,412
2,777,613 2,777,854 2,778,136 2,778,394 2,778,529
Total 2,954,871
2,969,508
2,972,024 2,972,266 2,972,548 2,972,805 2,972,941
Note: Totals may not sum due to independent rounding.
Canals and ditches in the conterminous United States are most abundant in the Gulf Coast states and California
(Figure 6-16A, Table 6-87). Florida contains 20 percent of all U.S. canal and ditch surface area, most of which were
constructed in the early 1900s for drainage, flood protection, and water storage purposes. Freshwater ponds are
Land Use, Land-Use Change, and Forestry 6-133
-------
1 more widely distributed across the United States (Figure 6-16B, Table 6-88). Florida also has the greatest surface
2 area of freshwater ponds, equivalent to 9 percent of all freshwater pond surface area in the United States.
3 Figure 6-16: 2021 Surface Area of A) Ditches and Canals and B) Freshwater Ponds in Flooded
4 Land Remaining Flooded Land (hectares)
A. Area of Ditches and Canals B. Area Freshwater Ponds
5
6 Table 6-87: State Totals of Surface Area in Flooded Land Remaining Flooded Land— Canals
and Ditches (ha)
State
1990
2005
2017
2018
2019
2020
2021
Alabama
228
228
228
228
228
228
228
Alaska
115
115
115
115
115
115
115
Arizona
3,536
3,536
3,536
3,536
3,536
3,536
3,536
Arkansas
7,349
7,349
7,349
7,349
7,349
7,349
7,349
California
16,725
16,725
16,725
16,725
16,725
16,725
16,725
Colorado
6,874
6,874
6,874
6,874
6,874
6,874
6,874
Connecticut
28
28
28
28
28
28
28
Delaware
130
130
130
130
130
130
130
District of Columbia
1
1
1
1
1
1
1
Florida
37,482
37,482
37,482
37,482
37,482
37,482
37,482
Georgia
352
352
352
352
352
352
352
Hawaii
538
538
538
538
538
538
538
Idaho
4,027
4,027
4,027
4,027
4,027
4,027
4,027
Illinois
2,489
2,489
2,489
2,489
2,489
2,489
2,489
Indiana
4,064
4,064
4,064
4,064
4,064
4,064
4,064
Iowa
867
867
867
867
867
867
867
Kansas
258
258
258
258
258
258
258
Kentucky
672
672
672
672
672
672
672
Louisiana
22,565
22,565
22,565
22,565
22,565
22,565
22,565
Maine
56
56
56
56
56
56
56
Maryland
967
967
967
967
967
967
967
Massachusetts
132
132
132
132
132
132
132
Michigan
12,897
12,897
12,897
12,897
12,897
12,897
12,897
Minnesota
11,235
11,235
11,235
11,235
11,235
11,235
11,235
Mississippi
3,936
3,936
3,936
3,936
3,936
3,936
3,936
Missouri
5,670
5,670
5,670
5,670
5,670
5,670
5,670
Montana
4,740
4,740
4,740
4,740
4,740
4,740
4,740
Nebraska
4,864
4,864
4,864
4,864
4,864
4,864
4,864
Nevada
1,587
1,587
1,587
1,587
1,587
1,587
1,587
New Hampshire
103
103
103
103
103
103
103
New Jersey
944
944
944
944
944
944
944
New Mexico
2,002
2,002
2,002
2,002
2,002
2,002
2,002
New York
925
925
925
925
925
925
925
6-134 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
North Carolina
6,321
6,321
6,321
6,321
6,321
6,321
6,321
North Dakota
1,819
1,819
1,819
1,819
1,819
1,819
1,819
Ohio
1,819
1,819
1,819
1,819
1,819
1,819
1,819
Oklahoma
278
278
278
278
278
278
278
Oregon
2,498
2,498
2,498
2,498
2,498
2,498
2,498
Pennsylvania
143
143
143
143
143
143
143
Puerto Rico
249
249
249
249
249
249
249
Rhode Island
1
1
1
1
1
1
1
South Carolina
3,226
3,226
3,226
3,226
3,226
3,226
3,226
South Dakota
703
703
703
703
703
703
703
Tennessee
442
442
442
442
442
442
442
Texas
11,152
11,152
11,152
11,152
11,152
11,152
11,152
Utah
1,875
1,875
1,875
1,875
1,875
1,875
1,875
Vermont
95
95
95
95
95
95
95
Virginia
1,306
1,306
1,306
1,306
1,306
1,306
1,306
Washington
1,125
1,125
1,125
1,125
1,125
1,125
1,125
West Virginia
28
28
28
28
28
28
28
Wisconsin
887
887
887
887
887
887
887
Wyoming
2,086
2,086
2,086
2,086
2,086
2,086
2,086
Total
194,412
194,412
194,412
194,412
194,412
194,412
194,412
1
2 Table 6-88: State Totals of Surface Area in Flooded Land Remaining Flooded Land—
3 Freshwater Ponds (ha)
State
1990
2005
2017
2018
2019
2020
2021
Alabama
67,304
67,639
67,655
67,658
67,658
67,658
67,658
Alaska
2,449
2,456
2,456
2,456
2,456
2,456
2,456
Arizona
6,153
6,199
6,208
6,211
6,211
6,215
6,215
Arkansas
62,194
62,510
62,510
62,510
62,510
62,510
62,510
California
73,388
73,589
73,647
73,647
73,653
73,659
73,660
Colorado
30,871
31,143
31,157
31,167
31,167
31,168
31,168
Connecticut
13,001
13,055
13,058
13,058
13,058
13,058
13,058
Delaware
7,006
7,010
7,010
7,010
7,010
7,010
7,010
District of Columbia
22
22
22
22
22
22
22
Florida
261,027
261,150
261,191
261,191
261,195
261,195
261,195
Georgia
140,246
142,014
142,090
142,090
142,093
142,099
142,099
Hawaii
2,229
2,236
2,238
2,238
2,238
2,238
2,238
Idaho
20,678
20,780
20,781
20,781
20,781
20,781
20,781
Illinois
77,370
77,913
77,985
78,001
78,006
78,016
78,016
Indiana
63,427
63,918
64,003
64,006
64,011
64,011
64,011
Iowa
67,833
69,748
70,668
70,749
70,911
71,023
71,096
Kansas
87,134
89,134
89,189
89,202
89,209
89,215
89,231
Kentucky
44,788
45,164
45,189
45,189
45,189
45,189
45,189
Louisiana
48,756
48,884
48,889
48,889
48,889
48,894
48,894
Maine
30,645
30,694
30,703
30,703
30,703
30,703
30,703
Maryland
14,739
14,890
14,942
14,942
14,942
14,944
14,945
Massachusetts
17,327
17,386
17,425
17,432
17,438
17,444
17,446
Michigan
66,159
66,310
66,342
66,347
66,347
66,355
66,355
Minnesota
88,283
88,509
88,585
88,592
88,599
88,622
88,634
Mississippi
76,062
76,212
76,230
76,230
76,235
76,240
76,241
Missouri
125,673
125,955
125,970
125,970
125,971
125,972
125,972
Montana
65,130
65,484
65,506
65,506
65,510
65,510
65,510
Nebraska
105,741
106,970
107,124
107,177
107,189
107,211
107,219
Nevada
5,641
5,644
5,680
5,690
5,690
5,694
5,694
New Hampshire
8,744
8,769
8,780
8,780
8,780
8,780
8,781
New Jersey
25,780
25,782
25,782
25,782
25,782
25,782
25,782
Land Use, Land-Use Change, and Forestry 6-135
-------
New Mexico
13,020
13,025
13,025
13,025
13,025
13,025
13,025
New York
59,452
59,707
59,811
59,811
59,813
59,813
59,816
North Carolina
91,555
91,608
91,613
91,613
91,613
91,613
91,613
North Dakota
71,758
71,763
71,784
71,784
71,784
71,784
71,784
Ohio
49,844
50,177
50,340
50,351
50,365
50,391
50,406
Oklahoma
119,199
119,310
119,310
119,310
119,312
119,313
119,313
Oregon
29,950
29,958
29,960
29,967
29,967
29,967
29,967
Pennsylvania
24,724
24,740
24,749
24,749
24,749
24,749
24,749
Puerto Rico
851
851
851
851
851
851
851
Rhode Island
2,521
2,529
2,536
2,536
2,536
2,536
2,536
South Carolina
78,075
78,748
78,960
78,961
78,972
78,976
78,976
South Dakota
100,444
100,661
100,713
100,714
100,732
100,733
100,736
Tennessee
46,824
47,525
47,546
47,555
47,560
47,567
47,567
Texas
210,149
210,711
210,721
210,721
210,721
210,721
210,721
Utah
17,817
17,871
17,882
17,882
17,882
17,884
17,884
Vermont
5,692
5,705
5,709
5,709
5,709
5,709
5,709
Virginia
52,327
52,327
52,327
52,327
52,327
52,327
52,327
Washington
17,013
17,058
17,081
17,081
17,081
17,081
17,081
West Virginia
8,902
8,932
8,938
8,938
8,938
8,938
8,938
Wisconsin
22,037
22,181
22,189
22,189
22,189
22,189
22,189
Wyoming
32,508
32,540
32,554
32,554
32,554
32,554
32,554
Total
2,760,460
2,775,096
2,777,613
2,777,854
2,778,136
2,778,394
2,778,529
1 Uncertainty
2 Uncertainty in estimates of Cm emissions from other constructed waterbodies (ponds, canals/ditches) in Flooded
3 Land Remaining Flooded Land (Table 6-89) are estimated using IPCC Approach 2 and include uncertainty in the
4 default emission factors and the flooded land area inventory. Uncertainty in default emission factors is provided in
5 the 2019 Refinement to the 2006 IPCC Guidelines (IPCC 2019). Uncertainties in the spatial data include 1)
6 uncertainty in area estimates from the NHD, NWI, and NW, and 2) uncertainty in the location of dams in the NID.
7 Overall uncertainties in these spatial datasets are unknown, but uncertainty for remote sensing products is
8 assumed to be ± 10 -15 percent based on IPCC guidance (IPCC 2003). An uncertainty range of ± 15 percent for the
9 flooded land area estimates is assumed and is based on expert judgment.
10 Table 6-89: Approach 2 Quantitative Uncertainty Estimates for ChU Emissions from Other
11 Constructed Waterbodies in Flooded Land Remaining Flooded Land
Source
Gas
2021 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
(MMTCOz Eq.)
(MMT CO
2 Eq.)
(%)
Lower Bound
Upper Bound
Lower Bound
Upper Bound
Canals and ditches
ch4
2.3
2.1
2.4
-5.3
7
Freshwater pond
ch4
14.2
14.2
14.2
-0.04
0.04
Total
ch4
16.5
16.4
16.7
-0.7
1
aRange of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
12 QA/QC and Verification
13 The National Hydrography Data (NHD) is managed by the USGS in collaboration many other federal, state, and
14 local entities. Extensive QA/QC procedures are incorporated into the curation of the NHD. The National Inventory
15 of Dams (NID) is maintained by the U.S. Army Corps of Engineers (USACE) in collaboration with the Federal
16 Emergency Management Agency (FEMA) and state regulatory offices. USACE resolves duplicative and conflicting
6-136 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 data from 68 data sources, which helps obtain the more complete, accurate, and updated NID.79 The Navigable
2 Waterways (NW) dataset is part of the U.S. Department of Transportation (USDOT)/Bureau of Transportation
3 Statistics (BTS) National Transportation Atlas Database (NTAD). The NW is a comprehensive network database of
4 the nation's navigable waterways updated on a continuing basis. U.S. Fish and Wildlife Service is the principal
5 agency in charge of wetland mapping including the National Wetlands Inventory (NWI). Quality and consistency of
6 the Wetlands Layer is supported by federal wetlands mapping and classification standards, which were developed
7 under the oversight of the Federal Geographic Data Committee (FGDC) with input by the FGDC Wetlands
8 Subcommittee. This dataset is part of the FGDC Water-Inland Theme, which is co-chaired by the FWS and the U.S.
9 Geological Survey.
10 General QA/QC procedures were applied to activity data, documentation, and emission calculations consistent
11 with the U.S. Inventory QA/QC plan, which is in accordance with Vol. 1 Chapter 6 of 2006IPCC Guidelines (see
12 Annex 8 for more details). All calculations were executed independently in Excel and R. Ten percent of state and
13 national totals were randomly selected for comparison between the two approaches to ensure there were no
14 computational errors.
15 Recalculations Discussion
16 The 1990 through 2021 Inventory uses the National Wetland Inventory (NWI) as the primary data source for
17 flooded land surface area, whereas the 1990 through 2020 Inventory used the National Hydrography Data (NHD as
18 the primary geospatial data source. The NWI is far more detailed than the NHD and also includes Alaska, Hawaii,
19 and Puerto Rico which were missing from 1990 through 2020 Inventory.
20 In addition, the EPA updated the global warming potential (GWP) for calculating C02-equivalent emissions of CH4
21 (from 25 to 28) to reflect the 100-year GWP provided in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). The
22 previous Inventory used the 100-year GWP provided in the IPCC Fourth Assessment Report (AR4). This update was
23 applied across the entire time series. Further discussion on this update and the overall impacts of updating the
24 inventory GWP values to reflect the AR5 can be found in Chapter 9, Recalculations and Improvements.
25 The net effect of these recalculations was an average annual increase in CH4 emission estimates from constructed
26 waterbodies of 15.4 MMT CO2 Eq., or a factor of 15.3, over the time series from 1990 to 2020 compared to the
27 previous Inventory.
28 Planned Improvements
29 Default emission factors for canals/ditches were derived from a global dataset that include few measurements
30 from U.S. systems. The EPA plans to conduct a literature survey to determine if sufficient data are available to
31 derive a country-specific emission factor.
32 Canal and ditch surface area included here may overlap with ditches and canals included in CH4 emission estimates
33 for ditches draining inland organic soils (IPCC 2013, section 2.2.2.1). EPA plans to reconcile ditch/canal surface
34 areas between the two managed land types (flooded land vs drained inland organic soils) in the next (i.e., 1990
35 through 2022) Inventory.
36 Features less than 8 ha in the NW that were not identified as Canal/Ditch were defined as freshwater ponds. Many
37 of these features are lock chambers connected to an upstream reservoir. These systems likely have emission rates
38 more similar to a reservoir than freshwater pond. In the 1990 through 2022 Inventory these systems will be
39 classified as reservoirs.
79 See https://www.epa.gov/national-aquatic-resource-surveys/national-lakes-assessment-2017-qualitv-assurance-proiect-
plan.
Land Use, Land-Use Change, and Forestry 6-137
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
6.9 Land Converted to Wetlands (CRF
Source Category 4D2)
Emissions and Removals from Land Converted to Vegetated
Coastal Wetlands
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 through
active restoration and creation of coastal wetlands through removal of hydrological barriers. Based upon NOAA C-
CAP, wetlands are subdivided into freshwater (Palustrine) and saline (Estuarine) classes and further subdivided
into emergent marsh, scrub shrub and forest classes All other land categories (i.e., Forest Land, Cropland,
Grassland, Settlements and Other Lands) are identified as having some area converting to Vegetated Coastal
Wetlands. This inventory does not include Land Converted to Unvegetated Open Water Coastal Wetlands (see
Planned Improvements section below). Between 1990 and 2021 the rate of annual transition for Land Converted
to Vegetated Coastal Wetlands ranged from 0 to 2,650 ha per year, depending on the type of land converted.80
Conversion rates from Forest Land were relatively consistent between 1990 and 2010 (ranging between 2,409 and
2,650 ha) and decreased to 625 ha starting in 2011; the majority of these conversions resulted in increases in the
area of palustrine wetlands, which also initiates Cm emissions when lands are inundated with fresh water.81 Little
to no conversion of Cropland, Grassland, Settlement, or Other Lands to vegetated coastal wetlands occurred
during the reporting period, with converted areas ranging from 0 to 25 ha per year.82
Conversion to coastal wetlands resulted in a biomass C stock loss of 0.1 MMT CO2 Eq. (0.03 MMT C) in 2021 (Table
6-90 and Table 6-91). Loss of forest biomass through conversion of Forest Lands to Vegetated Coastal Wetlands is
the primary driver behind biomass C stock change being a source rather than a sink across the time series.
Conversion of Cropland, Grassland, Settlement and Other Lands result in a net increase in biomass stocks.
Conversion of lands to vegetated coastal wetlands resulted in a DOM loss of 0.03 MMT CO2 Eq. (0.008 MMT C) in
2021 (Table 6-90 and Table 6-91), which is driven by the loss of DOM when Forest Land is converted to Vegetated
Coastal Wetlands. This is likely an overestimate of loss because wetlands inherently preserve dead organic
material. Conversion of Cropland, Grassland, Settlement and Other Land results in a net increase in DOM Across all
time periods, soil C accumulation resulting from Lands Converted to Vegetated Coastal Wetlands is a carbon sink
and has ranged between -0.15 and -0.3 MMT CO2 Eq. (-0.04 and -0.07 MMT C; Table 6-90 and Table 6-91).
Conversion of lands to coastal wetlands resulted in CH4 emissions of 0.18 MMT CO2 Eq. (6.4 kt CH4) in 2021 (Table
6-92). Methane emissions due to the conversion of Lands to Vegetated Coastal Wetlands are largely the result of
Forest Land converting to palustrine emergent and scrub shrub coastal wetlands in warm temperate climates.
Emissions were the highest between 1990 and 2001 (0.28 MMT CO2 Eq., 10.0 kt CH4) and have continually
80 Data from C-CAP; see https://coast.noaa.gov/dieitalcoast/tools/. Accessed September 2022.
81 Currently, the C-CAP dataset categorizes coastal wetlands as either palustrine (fresh water) or estuarine (presence of saline
water). This classification does not differentiate between estuarine wetlands with salinity < 18 ppt (when methanogenesis
begins to occur) and those that are >18 ppt (where negligible to no CH4 is produced); therefore, it is not possible at this time to
account for CH4 emissions from estuarine wetlands in the Inventory.
82 At the present stage of Inventory development, Coastal Wetlands are not explicitly shown in the Land Representation
analysis while work continues harmonizing data from NOAA's Coastal Change Analysis Program (C-CAP) with NRI, FIA and NLDC
data used to compile the Land Representation (NOAA OCM 2020).
6-138 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 decreased to current levels. This decrease was driven by a reduction in the rate of conversion of forest land to
2 palustrine scrub-shrubs and emergent wetlands.
3 Table 6-90: Net CO2 Flux from C Stock Changes in Land Converted to Vegetated Coastal
4 Wetlands (MMT CCh Eq.)
Land Use/Carbon Pool
1990
2005
2017
2018
2019
2020
2021
Cropland Converted to Vegetated Coastal
Wetlands
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Biomass C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Soil C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Forest Land Converted to Vegetated
Coastal Wetlands
0.49
0.50
(0.01)
+
0.01
0.02
0.03
Biomass C Stock
0.62
0.62
0.13
0.13
0.13
0.13
0.13
Dead Organic Matter C Flux
0.11
0.12
0.03
0.03
0.03
0.03
0.03
Soil C Stock
(0.23)
(0.24)
(0.17)
(0.16)
(0.15)
(0.14)
(0.13)
Grassland Converted to Vegetated Coastal
Wetlands
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Biomass C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Soil C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Other Land Converted to Vegetated
Coastal Wetlands
(0.03)
(0.03)
(0.02)
(0.02)
(0.02)
(0.02)
(0.02)
Biomass C Stock
(0.01)
(0.02)
(0.01)
(0.01)
(0.01)
(0.01)
(0.01)
Soil C Stock
(0.01)
(0.02)
(0.02)
(0.02)
(0.02)
(0.02)
(0.02)
Settlements Converted to Vegetated
Coastal Wetlands
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Biomass C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Soil C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Total Biomass Flux
0.60
0.60
0.12
0.12
0.12
0.12
0.12
Total Dead Organic Matter Flux
0.11
0.12
0.03
0.03
0.03
0.03
0.03
Total Soil C Flux
(0.25)
(0.25)
(0.18)
(0.18)
(0.17)
(0.16)
(0.15)
Total Flux
0.46
0.47
(0.03)
(0.02)
(0.01)
(+)
0.01
+ Absolute value does not exceed 0.005 MMT C02 Eq.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
5 Table 6-91: Net CO2 Flux from C Stock Changes in Land Converted to Vegetated Coastal
6 Wetlands (MMT C)
Land Use/Carbon Pool
1990
2005
2017
2018
2019
2020
2021
Cropland Converted to Vegetated Coastal
Wetlands
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Biomass C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Soil C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Forest Land Converted to Vegetated
Coastal Wetlands
0.13
0.14
(+)
+
+
0.006
0.01
Biomass C Stock
0.17
0.17
0.04
0.04
0.04
0.04
0.04
Dead Organic Matter C Flux
0.03
0.03
0.01
0.01
0.01
0.01
0.01
Soil C Stock
(0.06)
(0.06)
(0.05)
(0.04)
(0.04)
(0.04)
(0.04)
Grassland Converted to Vegetated Coastal
Wetlands
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Biomass C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Soil C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Other Land Converted to Vegetated
Coastal Wetlands
(0.01)
(0.01)
(0.01)
(0.01)
(0.01)
(0.01)
(0.01)
Biomass C Stock
(+)
(0.005)
(+)
(+)
(+)
(+)
(+)
Soil C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Settlements Converted to Vegetated
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Land Use, Land-Use Change, and Forestry 6-139
-------
Coastal Wetlands
Biomass C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Soil C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Total Biomass Flux
0.16
0.16
0.03
0.03
0.03
0.03
0.03
Total Dead Organic Matter Flux
0.03
0.03
0.01
0.01
0.01
0.01
0.01
Total Soil C Flux
(0.07)
(0.07)
(0.05)
(0.05)
(0.05)
(0.04)
(0.04)
Total Flux
0.13
0.13
(0.01)
(0.01)
(+)
(+)
+
+ Absolute value does not exceed 0.005 MMT C.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
1 Table 6-92: ChU Emissions from Land Converted to Vegetated Coastal Wetlands (MMT CO2
2 Eq. and kt CH4)
Land Use/Carbon Pool
1990
2005
2017
2018
2019
2020
2021
Cropland Converted to Vegetated Coastal
Wetlands
CH4 Emissions (MMT C02 Eq.)
+
+
+
+
+
+
+
CH4 Emissions (kt CH4)
+
0.01
0.04
0.04
0.04
0.05
0.05
Forest Land Converted to Vegetated
Coastal Wetlands
CH4 Emissions (MMT C02 Eq.)
0.28
0.27
0.20
0.19
0.18
0.17
0.16
CH4 Emissions (kt CH4)
9.88
9.74
7.22
6.85
6.48
6.10
5.76
Grassland Converted to Vegetated Coastal
Wetlands
CH4 Emissions (MMT C02 Eq.)
+
+
+
+
+
+
+
CH4 Emissions (kt CH4)
0.01
0.01
0.06
0.07
0.07
0.08
0.08
Other Land Converted to Vegetated Coastal
Wetlands
CH4 Emissions (MMT C02 Eq.)
+
+
0.01
0.01
0.01
0.01
0.01
CH4 Emissions (kt CH4)
0.08
0.14
0.40
0.43
0.47
0.50
0.52
Settlements Converted to Vegetated
Coastal Wetlands
CH4 Emissions (MMT C02 Eq.)
+
+
+
+
+
+
+
CH4 Emissions (kt CH4)
0.01
+
+
+
+
+
+
Total CH4 Emissions (MMT C02 Eq.)
0.28
0.28
0.22
0.21
0.20
0.19
0.18
Total CH4 Emissions (kt CH4)
9.98
9.91
7.72
7.39
7.06
6.73
6.41
+ Absolute value does not exceed 0.005 MMT C02 Eq. or 0.005 kt CH4.
Note: Totals may not sum due to independent rounding.
3 Methodology and Time-Series Consistency
4 The following section provides a description of the methodology used to estimate changes in biomass, dead
5 organic matter and soil C stocks and (Remissions for Land Converted to Vegetated Coastal Wetlands.
6 Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
7 through 2021.
8 Biomass Carbon Stock Changes
9 Biomass C stocks for Land Converted to Vegetated Coastal Wetlands are estimated for palustrine and estuarine
10 marshes for land below the elevation of high tides (taken to be mean high water spring tide elevation) and as far
11 seawards as the extent of intertidal vascular plants within the U.S. Land Representation according to the national
12 LiDAR dataset, the national network of tide gauges and land use histories recorded in the 1996, 2001, 2005, 2011,
13 and 2016 NOAA C-CAP surveys (NOAA OCM 2020). Both federal and non-federal lands are represented.
14 Delineating Vegetated Coastal Wetlands from ephemerally flooded upland Grasslands represents a particular
15 challenge in remote sensing. Moreover, at the boundary between wetlands and uplands, which may be gradual on
16 low lying coastlines, the presence of wetlands may be ephemeral depending upon weather and climate cycles and
6-140 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
as such, impacts on the emissions and removals will vary over these time frames. Trends in land cover change are
extrapolated to 1990 and 2021 from these datasets using the C-CAP change data closest in date to a given year.
Biomass is not sensitive to soil organic content. Aboveground biomass C stocks for non-forested coastal wetlands
are derived from a national assessment combining field plot data and aboveground biomass mapping by remote
sensing (Byrd et al. 2017; Byrd et al. 2018; Byrd et al. 2020). Aboveground biomass C removal data for all
subcategories are not available and thus assumptions were applied using expert judgment about the most
appropriate assignment to a disaggregation of a community class. The aboveground biomass C stock for estuarine
forested wetlands (dwarf mangroves that are not classified as forests due to their stature) is derived from a meta-
analysis by Lu and Megonigal (201783). Root to shoot ratios from the Wetlands Supplement were used to account
for belowground biomass, which were multiplied by the aboveground C stock (IPCC 2014) and summed with
aboveground biomass to obtain total biomass carbon stocks. Aboveground biomass C stocks for Forest Land,
Cropland, and Grassland that are lost with the conversion to Vegetated Coastal Wetlands were derived from Tier 1
default values (IPCC 2006; IPCC 2019). Biomass C stock changes are calculated by subtracting the biomass C stock
values of each land-use category (i.e., Forest Land, Cropland, and Grassland) from those of Vegetated Coastal
Wetlands in each climate zone and multiplying that value by the corresponding C-CAP derived area gained that
year in each climate zone. The difference between the stocks is reported as the stock change under the
assumption that the change occurred in the year of the conversion. The total coastal wetland biomass C stock
change is accounted for during the year of conversion; therefore, no interannual changes are calculated during the
remaining years it is in the category.
Dead Organic Matter
Dead organic matter (DOM) C stocks, which include litter and dead wood stocks, are accounted for in subtropical
estuarine forested wetlands for Lands Converted to Vegetated Coastal Wetlands across all years. Tier 1 estimates
of mangrove DOM C stocks were used for subtropical estuarine forested wetlands (IPCC 2014). Neither Tier 1 or 2
data on DOM are currently available for either palustrine or estuarine scrub/shrub wetlands for any climate zone
or estuarine forested wetlands in climates other than subtropical climates. Tier 1 DOM C stocks for Forest Land
converted to Vegetated Coastal Wetlands were derived from IPCC (2019) to account for the loss of DOM that
occurs with conversion. Changes in DOM are assumed to be negligible for other land-use conversions (i.e., other
than Forest Land) to coastal wetlands based on the Tier 1 method in IPCC (2006). Trends in land cover change are
derived from the NOAA C-CAP dataset and extrapolated to cover the entire 1990 through 2021 time series. Dead
organic matter removals are calculated by multiplying the C-CAP derived area gained that year by the difference
between Tier 1 DOM C stocks for Vegetated Coastal Wetlands and Forest Land. The difference between the stocks
is reported as the stock change under the assumption that the change occurred in the year of the conversion. The
coastal wetland DOM stock is assumed to be in steady state once established in the year of conversion; therefore,
no interannual changes are calculated.
Soil Carbon Stock Changes
Soil C removals are estimated for Land Converted to Vegetated Coastal Wetlands across all years. Soil C stock
changes, stratified by climate zones and wetland classes, are derived from a synthesis of peer-reviewed literature
(Lynch 1989; Orson et al. 1990; Kearny & Stevenson 1991; Roman et al. 1997; Craft et al. 1998; Orson et al. 1998;
Merrill 1999; Hussein et al. 2004; Church et al. 2006; Koster et al. 2007; Callaway et al. 2012 a & b; Bianchi et al.
2013; Crooks et al. 2014; Weston et al. 2014; Villa & Mitsch 2015; Marchio et al. 2016; Noe et al. 2016). To
estimate soil C stock changes, no differentiation is made for soil type (i.e., mineral, organic). Soil C removal data for
all subcategories are not available and thus assumptions were applied using expert judgment about the most
appropriate assignment to a disaggregation of a community class.
83 See https://github.com/Smithsonian/Coastal-Wetland-NGGl-Data-Public; accessed October 2021.
Land Use, Land-Use Change, and Forestry 6-141
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
As per IPCC (2014) guidance, Land Converted to Vegetated Coastal Wetlands is assumed to remain in this category
for up to 20 years before transitioning to Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands. Tier
2 level estimates of soil C stock changes associated with annual soil C accumulation from Land Converted to
Vegetated Coastal Wetlands were developed using country-specific soil C removal factors multiplied by activity
data of land area for Land Converted to Vegetated Coastal Wetlands for a given year in addition to the previous
19-year cumulative area. Guidance from the Wetlands Supplement allows for the rate of soil C accumulation to be
instantaneously equivalent to that in natural settings and that soil C accumulation is initiated when natural
vegetation becomes established; this is assumed to occur in the first year of conversion. No loss of soil C as a result
of land conversion to coastal wetlands is assumed to occur. Since the C-CAP coastal wetland area dataset begins in
1996, the area converted prior to 1996 is assumed to be the same as in 1996. Similarly, the coastal wetland area
data for 2017 through 2021 is assumed to be the same as in 2016. The methodology follows Eq. 4.7, Chapter 4 of
the IPCC Wetlands Supplement (IPCC 2014) and is applied to the area of Land Converted to Vegetated Coastal
Wetlands on an annual basis.
Soil Methane Emissions
Tier 1 estimates of Cm emissions for Land Converted to Vegetated Coastal Wetlands are derived from the same
wetland map used in the analysis of wetland soil C fluxes for palustrine wetlands, and are produced from C-CAP,
LiDAR and tidal data, in combination with default Cm emission factors provided in Table 4.14 of the IPCC Wetlands
Supplement. The methodology follows Eq. 4.9, Chapter 4 of the IPCC Wetlands Supplement. Because Land
Converted to Vegetated Coastal Wetlands is held in this category for up to 20 years before transitioning to
Vegetated Coastal Wetlands Remaining to Vegetated Coastal Wetlands, Cm emissions in a given year represent
the cumulative area held in this category for that year and the prior 19 years.
Uncertainty
Underlying uncertainties in estimates of soil C removal factors, biomass change, DOM, and Cm emissions include
error in uncertainties associated with Tier 2 literature values of soil C removal estimates, biomass stocks, DOM,
and IPCC default Cm emission factors, uncertainties linked to interpretation of remote sensing data, as well as
assumptions that underlie the methodological approaches applied.
Uncertainty specific to coastal wetlands include differentiation of palustrine and estuarine community classes,
which determines what flux is applied. Because mean soil and biomass C removal for each available community
class are in a fairly narrow range, the same overall uncertainty was assigned to each, respectively (i.e., applying
approach for asymmetrical errors, the largest uncertainty for any soil C stock value should be applied in the
calculation of error propagation; IPCC 2000). Uncertainties for Cm flux are the Tier 1 default values reported in the
Wetlands Supplement. Overall uncertainty of the NOAA C-CAP remote sensing product is 15 percent. This is in the
range of remote sensing methods (±10 to 15 percent; IPCC 2003). However, there is significant uncertainty in
salinity ranges for tidal and non-tidal estuarine wetlands and activity data used to estimate the Cm flux (e.g.,
delineation of an 18 ppt boundary), which will need significant improvement to reduce uncertainties. The
combined uncertainty was calculated by summing the squared uncertainty for each individual source (C-CAP, soil,
biomass, and DOM) and taking the square root of that total.
Uncertainty estimates are presented in Table 6-93 for each carbon pool and the Cm emissions. The combined
uncertainty is 42.6 percent above and below the estimate of 0.17 MMT CO2 Eq. In 2021, the total flux was 0.17
MMT CO2 Eq., with lower and upper estimates of 0.10 and 0.24 MMT CO2 Eq.
Table 6-93: Approach 1 Quantitative Uncertainty Estimates for C Stock Changes occurring
within Land Converted to Vegetated Coastal Wetlands in 2021 (MMT CO2 Eq. and Percent)
Source
2021 Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Estimate3
(MMT C02 Eq.) (%)
Lower Upper Lower Upper
6-142 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Bound
Bound
Bound
Bound
Biomass C Stock Flux
0.12
0.1
0.15
-20.0%
20.0%
Dead Organic Matter Flux
0.03
0.02
0.03
-25.8%
25.8%
Soil C Stock Flux
(0.15)
(0.2)
(0.1)
-18.7%
18.7%
Methane Emissions
0.18
0.13
0.18
-29.9%
29.9%
Total Uncertainty
0.18
0.11
0.26
-42.6%
42.6%
a Range of flux estimates based on error propagation at 95 percent confidence interval.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
QA/QC and Verification
NOAA provided National LiDAR Dataset, tide data, and C-CAP land cover and land cover change mapping, all of
which are subject to agency internal mandatory QA/QC assessment (McCombs et al. 2016). QA/QC and verification
of soil C stock dataset has been provided by the Smithsonian Environmental Research Center and Coastal Wetland
Inventory team leads. Biomass C stocks are derived from peer-review literature, reviewed by U.S. Geological
Survey prior to publishing, by the peer-review process during publishing, and by the Coastal Wetland Inventory
team leads prior to inclusion in the inventory and from IPCC reports. As a QC step, a check was undertaken
confirming that Coastal Wetlands recognized by C-CAP represent a subset of Wetlands recognized by the NRI for
marine coastal states. A team of two evaluated and verified there were no computational errors within the
calculation worksheets. Soil C stock, emissions/removals data are based upon peer-reviewed literature and Cm
emission factors are derived from the Wetlands Supplement.
Recalculations Discussion
An update was made to the activity data to remove any estuarine forested wetland areas that were located
outside of states classified as subtropical since, states classified as wet temperate, cold temperate and
mediterranean climate zones fall under the category of Land Converted to Forest Land.
In addition, EPA updated the global warming potential (GWP) for calculating C02-equivalent emissions of CH4 (from
25 to 28) to reflect the 100-year GWP values provided in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). The
previous Inventory used 100-year GWP values provided in the IPCC Fourth Assessment Report (AR4). This update
was applied across the entire time series. Further discussion on this update and the overall impacts of updating the
Inventory GWP values to reflect the AR5 can be found in Chapter 9, Recalculations and Improvements.
As a result of these changes, the recalculations resulted in net average increases to emissions totals ranging from
0.03 MMT CO2 Eq. to 0.02 MMT CO2 Eq. across the 1990 through 2020 time series compared to the previous
Inventory.
Planned Improvements
Administered by the Smithsonian Environmental Research Center, the Coastal Wetland Carbon Research
Coordination Network has established a U.S. country-specific database of soil C stocks and biomass for coastal
wetlands.84 This dataset will be updated periodically. Refined error analysis combining land cover change and C
stock estimates will be provided as new data are incorporated. Through this work, a model is in development to
represent changes in soil C stocks and will be incorporated into the next (i.e., 2024) Inventory submission.
Currently, the only coastal wetland conversion that is reported in the Inventory is Lands Converted to Vegetated
Coastal Wetlands. The next (2024) submission will include C stock change data for Lands Converted to
Unvegetated Open Water Coastal Wetlands.
84 See https://serc.si.edu/coastalcarbon; accessed August 2021.
Land Use, Land-Use Change, and Forestry 6-143
-------
1 Land Converted to Flooded Land
2 Flooded lands are defined as water bodies where human activities have 1) caused changes in the amount of
3 surface area covered by water, typically through water level regulation (e.g., constructing a dam), 2) waterbodies
4 where human activities have changed the hydrology of existing natural waterbodies thereby altering water
5 residence times and/or sedimentation rates, in turn causing changes to the natural production of greenhouse
6 gases, and 3) waterbodies that have been created by excavation, such as canals, ditches and ponds (IPCC 2019).
7 Flooded lands include waterbodies with seasonally variable degrees of inundation but would be expected to retain
8 some inundated area throughout the year under normal conditions.
9 Flooded lands are broadly classified as "reservoirs" or "other constructed waterbodies" (IPCC 2019). Reservoirs are
10 defined as flooded land greater than 8 ha and includes the seasonally flooded land on the perimeter of
11 permanently flooded land (i.e., inundation areas). IPCC guidance (IPCC 2019) provides default emission factors for
12 reservoirs and several types of "other constructed waterbodies" including freshwater ponds and canals/ditches.
13 Land that has been flooded for 20 years or greater is defined as Flooded Land Remaining Flooded Land and land
14 flooded for less than 20 years is defined as Land Converted to Flooded Land. The distinction is based on literature
15 reports that CO2 and CFU emissions are high immediately following flooding as labile organic matter is rapidly
16 degraded but decline to a steady background level approximately 20 years after flooding (Abril et al. 2005, Barros
17 et al. 2011, Teodoru et al. 2012). Both CO2 and CFU emissions are estimated for Land Converted to Flooded Land.
18 Nitrous oxide emissions from flooded lands are largely related to inputs of organic or inorganic nitrogen from the
19 watershed. These inputs from runoff/leaching/deposition are largely driven by anthropogenic activities such as
20 land-use change, wastewater disposal or fertilizer application in the watershed or application of fertilizer or feed in
21 aquaculture. These emissions are not included here to avoid double-counting N2O emissions which are captured in
22 other source categories, such as indirect N2O emissions from managed soils (Section 5.4, Agricultural Soil
23 Management) and wastewater management (Section 7.2, Wastewater Treatment and Discharge).
24 Emissions from Land Converted to Flooded Land-Reservoirs
25 Reservoirs are designed to store water for a wide range of purposes including hydropower, flood control, drinking
26 water, and irrigation. The permanently wetted portion of reservoirs are typically surrounded by periodically
27 inundated land referred to as a "drawdown zone" or "inundation area." Greenhouse gas emissions from
28 inundation areas are considered significant and similar per unit area to the emissions from the water surface and
29 are therefore included in the total reservoir surface area when estimating greenhouse gas emissions from flooded
30 land. Lakes converted into reservoirs without substantial changes in water surface area or water residence times
31 are not considered to be managed flooded land (see Area Estimates below) (IPCC 2019).
32 In 2021, the United States and Puerto Rico contained 63,804 hectares of reservoir surface area in Land Converted
33 to Flooded Land (see Methodology and Time-Series Consistency below for calculation details) distributed across all
34 six of the aggregated climate zones used to define flooded land emission factors (Figure 6-17) (IPCC 2019).
6-144 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Figure 6-17: U.S. Reservoirs (black polygons) in the Land Converted to Flooded Land
Category in 2021
Climate Zone
I I boreal
¦ cool temperate
¦ tropical dry/montane
¦ tropical moist/wet
¦ warm temperate dry
Owarm temperate moist
Alaska
"fc P'
- '¦
v
1000 mi
Note: Colors represent climate zone used to derive IPCC default emission factors. Reservoirs (indicated by black
polygons) are sparsely distributed across United States, but can be seen in IL, IN, and OH in this image.
Methane and CO2 are produced in reservoirs through the natural breakdown of organic matter. Per unit area
emission rates tend to scale positively with temperature and system productivity (i.e., abundance of algae).
Greenhouse gases produced in reservoirs can be emitted directly from the water surface and inundation areas or
as greenhouse gas-enriched water passes through the dam and the downstream river. Sufficient information exists
to estimate downstream CFU emissions using Tier 1 IPCC guidance (IPCC 2019), but no guidance is provided for
downstream CO2 emissions. Table 6-94 and Table 6-95 below summarize nationally aggregated CFU and CO2
emissions from reservoirs in Land Converted to Flooded Land. The decrease in CO2 and CFU emissions through the
time series is attributable to reservoirs matriculating from the Land Converted to Flooded Land category into the
Flooded Land Remaining Flooded Land Category. Emissions have been stable since 2005, reflecting the low rate of
new flooded land creation over the past 16 years.
Table 6-94: ChU Emissions from Land Converted to Flooded Land - Reservoirs (MMT CO2 Eq.)
Source
1990
2005
2017
2018
2019
2020
2021
Reservoirs
Surface Emissions
0.9
0.2
0.2
0.2
0.2
0.2
0.2
Downstream Emissions
0.1
+
+
+
+
+
+
Total
1.0
0.2
0.3
0.3
0.3
0.2
0.2
+lndicates values less than 0.05 MMT C02
Note: Totals may not sum due to independent rounding
Table 6-95: ChU Emissions from Land Converted to Flooded Land—Reservoirs (kt ChU)
Source 1990
2005
2017 2018 2019 2020 2021
Reservoirs
Surface Emissions 34
Downstream Emissions 3
OO T~i
8 8 8 6 6
11111
Total 37
9
9 9 9 6 6
Land Use, Land-Use Change, and Forestry 6-145
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
Table 6-96: CO2 Emissions from Land Converted to Flooded Land—Reservoirs (MMT CO2)
Source
1990
2005
2017
2018
2019
2020
2021
Reservoir
1.3
0.3
0.4
0.4
0.4
0.2
0.2
Table 6-97: CO2 Emissions from Land Converted to Flooded Land—Reservoirs (MMT C)
Source
1990
2005
2017
2018
2019
2020
2021
Reservoir
0.4
0.1
0.1
0.1
0.1
0.1
0.1
Methane and CO2 emissions from reservoirs in Minnesota were 8-fold greater than from any other state (Figure
6-18 and Table 6-98). This is attributed to ten reservoirs created in Minnesota after 2001 which impound 52,252 ha
of water, 99 percent of which is located in Mille Lacs Lake. North Dakota is the second largest source of CO2 and
Cm from reservoirs in Land Converted to Flooded Land. Ninety-five percent of Land Converted to Flooded Land
reservoir surface area in North Dakota is attributed to Devils Lake. Both Mille Lacs and Devils Lakes are natural
waterbodies provisioned with dams for water level management.
Figure 6-18: 2021 A) ChU and B) CO2 Emissions from U.S. Reservoirs in Land Converted to
Flooded Land
A. CH4 Emissions from Reservoirs
kt CH4 y"1
¦
4
100 mi
500 mi
B. C02 Emissions from Reservoirs
kt C02 y"
n200
150
100
100 mi
100 mi
Alaska
100 mi
500 mi
Alaska
Table 6-98: Methane and CO2 Emissions from Reservoirs in Land Converted to Flooded Land
in 2021 (kt ChU; kt CO2)
ou C02a
State Surface Downstream Total Surface
Alabama
0
0
0
0
Alaska
0
0
0
0
Arizona
0
0
0
0
Arkansas
+
+
+
6
California
+
+
+
+
Colorado
+
+
+
1
Connecticut
+
+
+
+
Delaware
0
0
0
0
District of Columbia
0
0
0
0
Florida
+
+
+
5
Georgia
+
+
+
+
Hawaii
0
0
0
0
Idaho
+
+
+
2
6-146 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
Illinois
+
+
+
+
Indiana
+
+
+
+
Iowa
+
+
+
1
Kansas
+
+
+
1
Kentucky
0
0
0
0
Louisiana
0
0
0
0
Maine
+
+
+
+
Maryland
+
+
+
+
Massachusetts
+
+
+
4
Michigan
+
+
+
+
Minnesota
4
+
5
195
Mississippi
0
0
0
0
Missouri
0
0
0
0
Montana
+
+
+
+
Nebraska
+
+
+
+
Nevada
+
+
+
+
New Hampshire
0
0
0
0
New Jersey
0
0
0
0
New Mexico
+
+
+
+
New York
+
+
+
+
North Carolina
0
0
0
0
North Dakota
1
+
1
23
Ohio
+
+
+
1
Oklahoma
+
+
+
2
Oregon
0
0
0
0
Pennsylvania
+
+
+
+
Puerto Rico
0
0
0
0
Rhode Island
0
0
0
0
South Carolina
0
0
0
0
South Dakota
+
+
+
+
Tennessee
+
+
+
1
Texas
+
+
+
+
Utah
+
+
+
1
Vermont
0
0
0
0
Virginia
0
0
0
0
Washington
+
+
+
+
West Virginia
0
0
0
0
Wisconsin
+
+
+
+
Wyoming
+
+
+
+
+ Indicates values greater than zero and less than 0.5 kt
aC02: Only surface C02 emissions are included in the Inventory
Methodology and Time-Series Consistency
Estimates of CFU and CO2 emissions for reservoirs in Land Converted to Flooded Land follow the Tier 1
methodology in the IPCC guidance (IPCC 2019). All calculations are performed at the state level and summed to
obtain national estimates. Emissions from the surface of these flooded lands are calculated as the product of
flooded land surface area and a climate-specific emission factor (Table 6-99). Downstream CFU emissions are
calculated as 9 percent of the surface CFU emission (Tier 1 default). The IPCC guidance (IPCC 2019) does not
address downstream CO2 emissions, presumably because there are insufficient data in the literature to estimate
this emission pathway.
The IPCC default surface emission factors are derived from model-predicted (G-res model, Prairie et al. 2017)
emission rates for all reservoirs in the Global Reservoir and Dam (GRanD) database (Lehner et al. 2011). Predicted
emission rates were aggregated by the 11 IPCC climate zones (IPCC 2019, Table 7A.2) which were collapsed into six
climate zones using a regression tree approach. All six aggregated climate zone are present in the United States.
Land Use, Land-Use Change, and Forestry 6-147
-------
1 Table 6-99: IPCC (2019) Default ChU and CO2 Emission Factors for Surface Emissions from
2 Reservoirs in Land Converted to Flooded Land
Surface emission factor
Climate
MT CH4 ha 1 y1
MT CO? ha 1 y1
Boreal
0.0277
3.45
Cool Temperate
0.0847
3.74
Warm Temperate Dry
0.1956
6.23
Warm Temperate Moist
0.1275
5.35
Tropical Dry/Montane
0.3923
10.82
Tropical Moist/Wet
0.2516
10.16
Note: downstream CH4 emissions are calculated as 9 percent of surface emissions.
Downstream emissions are not calculated for C02.
3 Area Estimates
4 U.S. reservoirs were identified from the NHDWaterbody layer in the National Hydrography Dataset Plus V2
5 (NHD),85 the National Inventory of Dams (NID),86the National Wetlands Inventory (NWI),87 and the Navigable
6 Waterways (NW) network.88 The NHD only covers the conterminous U.S., whereas the NID, NW and NWI also
7 include Alaska, Hawaii, and Puerto Rico. The following paragraphs present the criteria used to identify other
8 constructed waterbodies in the NHD, NW, and NWI.
9 Waterbodies in the NHDWaterbody layer that were less than or equal to 20-years old, greater than or equal to 8
10 ha in surface area, not identified as canal/ditch in NHD, and met any of the following criteria were considered
11 reservoirs in Land Converted to Flooded Land: 1) the waterbody was classified "Reservoir" in the NHDWaterbody
12 layer, 2) the waterbody name in the NHDWaterbody layer included "Reservoir", 3) the waterbody in the
13 NHDWaterbody layer was located in close proximity (up to 100 m) to a dam in the NID, 4) the NHDWaterbody GNIS
14 name was similar to nearby NID feature (between 100 m to 1000 m).
15 EPA assumes that all features included in the NW are subject to water-level management to maintain minimum
16 water depths required for navigation and are therefore managed flooded lands. NW features greater than 8 ha in
17 surface area are defined as reservoirs.
18 NWI features were considered "managed" if they had a special modifier value indicating the presence of
19 management activities (Figure 6-19). To be included in the flooded lands inventory, the managed flooded land had
20 to be wet or saturated for at least one season per year (see 'Water Regime' in Figure 6-19). NWI features that met
21 these criteria, were greater than 8 Ha in surface area, and were not a canal/ditch (see Emissions from Land
22 Converted to Flooded Land - Other Constructed Waterbodies) were defined as reservoirs.
23 Surface areas for identified flooded lands were taken from NHD, NWI or the NW. If features from the NHD, NWI, or
24 the NW datasets overlapped, duplicate areas were erased. The first step was to take the final NWI Flooded Lands
25 features and use it to identify overlapping NHD features. If the NHD feature had its center in a NWI feature, it was
26 removed from analysis. Next, remaining NHD features were erased from any remaining overlapping NWI features.
27 Final selections of NHD and NWI features were used to erase any overlapping NW waterbodies.
28 Reservoir age was determined by assuming they were created the same year as a nearby (up to 100 m) NID
29 feature. If no nearby NID feature was identified, it was assumed the feature was greater than 20-years old
85 See https://www.uses.gov/core-science-svstems/nep/national-hvdroeraphv.
86 See https://nid.sec.usace.armv.mil.
87 See https://www.fws.eov/program/national-wetlands-inventory/data-download.
88 See https://www.census.eov/geoeraphies/mapping-files/time-series/eeo/carto-boundary-file.html.
6-148 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 throughout the time series. Only reservoirs less than or equal to 20-years old are included in Land Converted to
2 Flooded Land.
3 Figure 6-19: Selected Features from NWI that meet Flooded Lands Criteria
MODIFIERS
In order to more adequately describe the wetland and deepwater habitats, one each of the water regime, water chemistry, soil, or
special modifiers may be applied at the class or lower level in the hierarchy
Water Regime
Special Modifiers
Water Chemistry
Soil
Nontidal
A Temporarily Flooded
B Seasonally Saturated
Saltwater Tidal
IL Subtidal
Freshwater Tidal
Q Regularly Flooded-Fresh Tidal
R Seasonally Flooded-Fresh Tidal
b Beaver
Halinity/Salinity pH Modifiers for
Fresh Water
1 Hyperhaline / Hypersaline a Acid
2 Euhaline / Eusaline t Circumneutral
3 Mixohaline / M ixosaline (Brackish) i Alkaline
4 Polyhaline
5 Mesohaline
6 Oligohaline
0 Fresh
g Organic
n Mineral
M Irreaularlv Exposed
d Partly Drained/Ditched
f Farmed
m Managed
h Diked/Impounded
r Artificial Substrate
s SDOil
x Excavated |
C Seasonally Flooded
N Regularly Flooded
P Irregularly Flooded
S Temporarily Flooded- Fresh Tidal
D Continuously Saturated
T Semipermanently Flooded-Fresh Tidal
V Permanently Flooded-Fresh Tidal
E Seasonally Flooded /
Saturated
F Semipermanently Flooded
G Intermittently Exposed
H Permanently Flooded
J Intermittently Flooded
K Artificially Flooded [
I 1 Must also meet one selected special modifier (red box) to be included in the flooded lands inventory
I ~| Included in the flooded lands inventory if it meets water regime qualifier (gold box)
^ Source (modified): https://www.fws.gov/sites/default/files/documents/wetlarids-and-deepwater-map-code-diagram.pdf
5 IPCC (2019) allows for the exclusion of managed waterbodies from the inventory if the water surface area or
6 residence time was not substantially changed by the construction of the dam. The guidance does not quantify
7 what constitutes a "substantial" change, but here EPA excludes the U.S. Great Lakes from the inventory based on
8 expert judgment that neither the surface area nor water residence time was substantially altered by their
9 associated dams.
10 Reservoirs were disaggregated by state (using boundaries from the 2016 U.S. Census Bureau89) and climate zone.
11 Downstream and surface emissions for cross-state reservoirs were allocated to states based on the surface area
12 that the reservoir occupied in each state. Only the U.S. portion of reservoirs that cross country borders were
13 included in the inventory.
14 The surface area of reservoirs in Land Converted to Flooded Land decreased by approximately 70 percent from
15 1990 to 2021 (Table 6-100). This is due to reservoirs that were less than 20-years old at the beginning of time
16 series entering the Flooded Land Remaining Flooded Land category when they exceeded 20 years of age. The rate
17 at which flooded land has aged out of the Land Converted to Flooded Land category has outpaced the rate of new
18 dam construction. New dam construction has slowed considerably during the time series with only four new dams
19 constructed in 2021,90 versus 538 in 1990 (Figure 6-20).
20 Table 6-100: National Totals of Reservoir Surface Area in Land Converted to Flooded Land
21 (thousands of ha)
22
Surface Area (thousands of ha)
1990
2005
2017
2018
2019
2020
2021
Reservoir
234
63
85
84
84
64
64
89 See https://www.cerisus.gov/geographies/mapping-files/time-series/eeo/carto-boundarv-file.html.
90 See https://nid.sec.usace.armv.mil.
Land Use, Land-Use Change, and Forestry 6-149
-------
1
Figure 6-20: Number of Dams Built per Year from 1990 through 2021
400-
w
E
03
T3
§
200-
0-
2
3 Table 6-101: State Breakdown of Reservoir Surface Area in Land Converted to Flooded Land
4 (thousands of ha)
State
1990
2005
2017
2018
2019
2020
2021
Alabama
5.4
0.0
0.0
0.0
0.0
0.0
0.0
Alaska
0.6
0.0
0.0
0.0
0.0
0.0
0.0
Arizona
0.0
0.1
0.0
0.0
0.0
0.0
0.0
Arkansas
9.6
0.9
1.2
1.2
1.2
1.2
1.2
California
16.2
1.0
0.1
0.1
0.1
0.1
0.1
Colorado
3.7
1.1
0.2
0.2
0.2
0.3
0.2
Connecticut
0.0
0.1
0.0
0.0
0.0
0.0
0.0
Delaware
0.0
0.0
0.0
0.0
0.0
0.0
0.0
District of Columbia
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Florida
14.1
2.1
1.1
0.8
0.8
0.8
0.5
Georgia
9.7
3.7
0.1
0.1
0.0
0.0
0.0
Hawaii
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Idaho
18.1
0.8
0.4
0.4
0.4
0.4
0.4
Illinois
8.8
10.5
9.5
9.5
9.5
0.1
0.1
Indiana
10.0
0.2
0.1
0.1
0.1
0.1
0.1
Iowa
6.6
2.0
0.4
0.1
0.2
0.2
0.2
Kansas
18.9
0.3
0.2
0.2
0.1
0.1
0.1
Kentucky
4.7
0.0
0.0
0.0
0.0
0.0
0.0
Louisiana
5.8
3.2
0.2
0.0
0.0
0.0
0.0
Maine
12.5
4.2
0.0
0.0
0.0
0.0
0.0
Maryland
0.5
0.0
0.1
0.1
0.1
0.1
0.1
Massachusetts
1.1
0.2
0.9
0.9
0.9
0.8
0.8
Michigan
8.5
0.9
0.1
0.1
0.1
0.1
0.1
Minnesota
6.1
4.5
52.4
52.4
52.4
52.3
52.3
Mississippi
2.2
0.0
0.0
0.0
0.0
0.0
0.0
Missouri
0.2
9.7
9.7
9.7
9.7
0.0
0.0
Montana
13.4
1.2
0.1
0.1
0.1
0.1
0.1
Nebraska
5.3
1.3
0.1
0.1
0.1
0.0
0.0
Nevada
1.3
0.9
0.1
0.0
0.0
0.0
0.0
New Hampshire
0.3
0.0
0.0
0.0
0.0
0.0
0.0
New Jersey
0.0
0.0
0.0
0.0
0.0
0.0
0.0
New Mexico
0.1
0.0
0.0
0.0
0.0
0.0
0.0
O O O O
(J) O CN
-------
New York
1.9
0.5
0.1
0.1
0.1
0.1
0.1
North Carolina
0.6
0.1
0.1
0.1
0.0
0.0
0.0
North Dakota
0.0
0.9
6.2
6.2
6.2
6.2
6.2
Ohio
6.4
0.4
0.2
0.2
0.2
0.2
0.1
Oklahoma
3.0
0.0
0.4
0.4
0.4
0.4
0.4
Oregon
1.5
0.0
0.0
0.0
0.0
0.0
0.0
Pennsylvania
1.2
0.0
0.0
0.0
0.0
0.0
0.0
Puerto Rico
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Rhode Island
0.1
0.0
0.0
0.0
0.0
0.0
0.0
South Carolina
14.0
6.2
0.0
0.0
0.0
0.0
0.0
South Dakota
0.4
3.3
0.8
0.8
0.8
0.0
0.0
Tennessee
3.0
0.0
0.1
0.1
0.1
0.1
0.1
Texas
10.1
0.0
0.0
0.0
0.0
0.0
0.0
Utah
1.6
0.0
0.2
0.2
0.2
0.2
0.2
Vermont
0.1
0.0
0.0
0.0
0.0
0.0
0.0
Virginia
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Washington
2.7
0.2
0.0
0.0
0.0
0.0
0.0
West Virginia
1.9
1.6
0.0
0.0
0.0
0.0
0.0
Wisconsin
1.7
0.3
0.0
0.0
0.1
0.1
0.1
Wyoming
0.2
0.2
0.1
0.1
0.1
0.1
0.1
Total
234.4
62.9
85.3
84.4
84.3
64.1
63.8
1 Uncertainty
2 Uncertainty in estimates of Cm and CO2 emissions from reservoirs on Land Converted to Flooded Land were
3 developed using IPCC Approach 2 and include uncertainty in the default emission factors and the flooded land area
4 inventory (Table 6-102). Uncertainty in emission factors is provided in the 2019 Refinement to the 2006 IPCC
5 Guidelines (IPCC 2019). Uncertainties in the spatial data include 1) uncertainty in area estimates from the NHD,
6 NWI, and NW, and 2) uncertainty in the location of dams in the NID. Overall uncertainties in these spatial datasets
7 are unknown, but uncertainty for remote sensing products is assumed to be ± 10 to 15 percent based on IPCC
8 guidance (IPCC 2003). An uncertainty range of ± 15 percent for the flooded land area estimates is assumed and is
9 based on expert judgment.
10 Table 6-102: Approach 2 Quantitative Uncertainty Estimates for ChU and CO2 Emissions from
11 Reservoirs in Land Converted to Flooded Land
2021 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
Source
Gas
(MMTCOz Eq.)
(MMT CO
2 Eq.)
(%)
Lower Bound
Upper Bound
Lower Bound
Upper Bound
Reservoir
Surface
ch4
0.16
0.14
0.18
-13.3%
13.4%
Surface
C02
0.25
0.21
0.28
-13.9%
15.0%
Downstream
ch4
+
+
0.05
-62.8%
221.0%
Total
0.42
0.36
0.49
-14.9%
16.8%
+ Indicates values less than 0.05 MMT C02 Eq.
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
12 QA/QC and Verification
13 The National Hydrography Data (NHD) is managed by the USGS in collaboration many other federal, state, and
14 local entities. Extensive QA/QC procedures are incorporated into the curation of the NHD. The National Inventory
15 of Dams (NID) is maintained by the U.S. Army Corps of Engineers (USACE) in collaboration with the Federal
16 Emergency Management Agency (FEMA) and state regulatory offices. USACE resolves duplicative and conflicting
Land Use, Land-Use Change, and Forestry 6-151
-------
1 data from 68 data sources, which helps obtain the more complete, accurate, and updated NID. The Navigable
2 Waterways (NW) dataset is part of the U.S. Department of Transportation (USDOT)/Bureau of Transportation
3 Statistics (BTS) National Transportation Atlas Database (NTAD). The NW is a comprehensive network database of
4 the nation's navigable waterways updated on a continuing basis. U.S. Fish and Wildlife Service is the principal
5 agency in charge of wetland mapping including the National Wetlands Inventory (NWI). Quality and consistency of
6 the Wetlands Layer is supported by federal wetlands mapping and classification standards, which were developed
7 under the oversight of the Federal Geographic Data Committee (FGDC) with input by the FGDC Wetlands
8 Subcommittee. This dataset is part of the FGDC Water-Inland Theme, which is co-chaired by the FWS and the U.S.
9 Geological Survey.
10 General QA/QC procedures were applied to activity data, documentation, and emission calculations consistent
11 with the U.S. Inventory QA/QC plan, which is in accordance with Vol. 1 Chapter 6 of the 2006IPCC Guidelines (see
12 Annex 8 for more details). All calculations were executed independently in Excel and R. Ten percent of state and
13 national totals were randomly selected for comparison between the two approaches to ensure there were no
14 computational errors.
15 Recalculations Discussion
16 The 1990 through 2021 Inventory uses the National Wetland Inventory (NWI) as the primary data source for
17 flooded land surface area, whereas the 1990 through 2020 Inventory report used the National Hydrography Data
18 (NHD) as the primary geospatial data source. The NWI includes Alaska, Hawaii, and Puerto Rico, which were
19 missing from 1990 through 2020 Inventory, but this had little effect on the emission estimates as Hawaii and
20 Puerto Rico had no reservoirs in Land Converted to Flooded Land. In 1990, Alaska had 637 ha of reservoirs in Land
21 Converted to Flooded Land, but all reservoirs in Alaska matriculated to Flooded Land Remaining Flooded Land by
22 2004.
23 The 1990 through 2020 Inventory distinguished between reservoirs and inundation areas. Inundation areas were
24 defined as periodically flooded lands that bordered a permanently flooded reservoir. The NWI includes both
25 permanently and periodically flooded lands, but does not consistently discriminate between them, therefore
26 inundation areas and reservoirs are lumped into reservoirs for the 1990 through 2021 Inventory.
27 The 1990 though 2021 Inventory includes corrections to the age of several large reservoirs in South Dakota, North
28 Dakota, Alabama, Arkansas, Georgia, and South Carolina. As result, these flooded lands are now included in
29 Flooded Land Remaining Flooded Land throughout the time series, whereas they were misclassified as Land
30 Converted to Flooded Land for a portion of the time series in the 1990 through 2020 Inventory. For the year 1990,
31 these corrections reduced the surface area, methane emissions, and carbon dioxide emissions of reservoirs in Land
32 Converted to Flooded Land by 138,375 ha, 18.8 kt CH4, and 0.7 MMT CO2, respectively.
33 Overall, the recalculations resulted in substantial reductions in methane and carbon dioxide emissions in the first
34 few years of the time series (e.g., decrease of 4.1 MMT CO2 Eq. in 1990), but the differences were minor by 2005
35 through 2020 (0.1 MMT CO2 Eq.).
36 In addition, the EPA updated the global warming potential (GWP) for CH4 (from 25 to 28) to reflect the 100-year
37 GWP provided in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). The previous Inventory used the 100-year
38 GWP provided in the IPCC Fourth Assessment Report (AR4). This update was applied across the entire time series.
39 Further discussion on this update and the overall impacts of updating the inventory GWP values to reflect the AR5
40 can be found in Chapter 9, Recalculations and Improvements.
41 The net effect of these recalculations for CH4 emissions from reservoirs was an average annual decrease of 0.3
42 MMT CO2 Eq., or 49 percent, over the time series from 1990 to 2020 compared to the previous Inventory.
6-152 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Planned Improvements
2 The EPA is currently measuring greenhouse gas emissions from 108 reservoirs in the conterminous United States.
3 The survey will be complete in September 2023 and the data will be used to develop country-specific emission
4 factors for U.S. reservoirs. At the earliest, these emission factors will be used in the 2025 Inventory submission.
5 Emissions from Land Converted to Flooded Land-Other
e Constructed Waterbodies
7 Freshwater ponds are the only type of flooded lands within the "other constructed waterbodies" subcategory of
8 Land Converted to Flooded Land that are included in this Inventory (see Methodology for details) because age data
9 are not available for canals and ditches. All canals and ditches are assumed to be greater than 20-years old
10 throughout the time series and are included in Flooded Land Remaining Flooded Land.
11 IPCC (2019) describes ponds as waterbodies that are "...constructed by excavation and/or construction of walls to
12 hold water in the landscape for a range of uses, including agricultural water storage, access to water for livestock,
13 recreation, and aquaculture." The IPCC "Decision tree for types of Flooded Land" (IPCC 2019, Fig. 7.2) elaborates
14 on this description by defining waterbodies less than 8 ha as a subset of "other constructed waterbodies." For this
15 Inventory, ponds are defined as managed flooded land not identified as "canal/ditch" (see Methods below) with
16 surface area less than 8 ha. IPCC (2019) further distinguishes saline versus brackish ponds, with the former
17 supporting lower CFU emission rates than the latter. Activity data on pond salinity is not uniformly available for the
18 United States and all ponds in Land Converted to Flooded Land are assumed to be freshwater. Ponds often receive
19 high organic matter and nutrient loadings, may have low oxygen levels, and are sites of substantial Cm and CO2
20 emissions from anaerobic sediments.
21 Methane and CO2 emissions from freshwater ponds decreased 95 percent from 1990 to 2021 due to flooded land
22 matriculating from Land Converted to Flooded Land to Flooded Land Remaining Flooded Land. In 2021, Nebraska,
23 Montana, and Iowa had the greatest CO2 and CFU emissions for freshwater ponds in Land Converted to Flooded
24 Land (Table 6-103 through Table 6-107, Figure 6-21).
25 Table 6-103: ChU Emissions from Other Constructed Waterbodies in Land Converted to
26 Flooded Land (MMT CCh Eq.)
Source
1990
2005
2017
2018
2019
2020
2021
Freshwater Ponds
0.1
+
+
+
+
+
+
+ Indicates values less than 0.05 MMT C02 Eq.
27 Table 6-104: ChU Emissions from Other Constructed Waterbodies in Land Converted to
28 Flooded Land (kt Cm)
Source
1990
2005
2017
2018
2019
2020
2021
Freshwater Ponds
3
1
+
+
+
+
+
+ Indicates values less than 0.5 kt
29 Table 6-105: CO2 Emissions from Other Constructed Waterbodies in Land Converted to
30 Flooded Land (MMT COz)
Source
1990
2005
2017
2018
2019
2020
2021
Freshwater Ponds
0.1
1 +
+
+
+
+
+
+ Indicates values less than 0.05 MMT C02 Eq.
Land Use, Land-Use Change, and Forestry 6-153
-------
1 Table 6-106: CO2 Emissions from Other Constructed Waterbodies in Land Converted to
2 Flooded Land (MMT C)
Source
1990
2005
2017
2018
2019
2020
2021
Freshwater Ponds
0.02
1 0.01
+
+
+
+
+
+ Indicates values less than 0.005 MMT C
3 Table 6-107: ChU and CO2 Emissions from Other Constructed Waterbodies in Land Converted
4 to Flooded Land in 2021 (MT CO2 Eq.)
Freshwater Ponds
State
ch4
CO?
Total
Alabama
0
0
0
Alaska
0
0
0
Arizona
0
0
0
Arkansas
1
1
3
California
151
162
313
Colorado
278
202
480
Connecticut
0
0
0
Delaware
0
0
1
District of Columbia
0
0
0
Florida
25
50
76
Georgia
134
234
368
Hawaii
0
0
0
Idaho
1
0
1
Illinois
130
121
251
Indiana
111
116
227
Iowa
425
393
818
Kansas
353
369
722
Kentucky
4
4
8
Louisiana
3
6
10
Maine
1
1
2
Maryland
100
104
204
Massachusetts
342
311
654
Michigan
37
27
64
Minnesota
330
241
570
Mississippi
65
127
191
Missouri
13
14
27
Montana
491
359
850
Nebraska
514
471
985
Nevada
113
93
206
New Hampshire
1
0
1
New Jersey
0
0
0
New Mexico
0
0
0
New York
121
96
217
North Carolina
6
6
11
North Dakota
47
34
82
Ohio
195
200
396
Oklahoma
0
0
0
Oregon
0
0
0
Pennsylvania
0
0
0
Puerto Rico
0
0
0
Rhode Island
0
0
0
South Carolina
46
48
94
South Dakota
378
276
655
Tennessee
13
13
26
Texas
0
0
0
6-154 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Utah 146 107 253
Vermont 0 0 0
Virginia 0 0 0
Washington 23 28 50
West Virginia 15 16 31
Wisconsin 34 25 59
Wyoming 29 21 51
TOTAL 4,677 4,277 8,954
Figure 6-21: 2021 A) ChU and B) CO2 Emissions from Other Constructed Waterbodies
(Freshwater Ponds) in Land Converted to Flooded Land (MT CO2 Eq.)
Methodology and Time-Series Consistency
Estimates of Cm and CO2 emissions for other constructed waterbodies in Land Converted to Flooded Land follow
the Tier 1 methodology in IPCC (2019). All calculations are performed at the state level and summed to obtain
national estimates. Greenhouse gas emissions from the surface of these flooded lands are calculated as the
product of flooded land surface area and an emission factor (Table 6-108). Due to a lack of empirical data on CO2
emissions from recently created ponds, IPCC (2019) states "For all types of ponds created by damming, the
methodology described above to estimate CO2 emissions from land converted to reservoirs may be used." This
Inventory uses IPCC default CO2 emission factors for land converted to reservoirs when estimating CO2 emissions
from land converted to freshwater ponds. IPCC guidance also states that "there is insufficient information available
to derive separate Cm emission factors for recently constructed ponds..." and allows for the use of IPCC default
Cm emission factors for land remaining flooded land. Downstream emissions are not inventoried for other
constructed waterbodies because 1) many of these systems are not associated with dams (e.g., excavated ponds
and ditches), and 2) there are insufficient data to derive downstream emission factors for other constructed
waterbodies that are associated with dams (IPCC 2019).
Table 6-108: IPCC Default Methane and CO2 Emission Factors for Other Constructed
Waterbodies in Land Converted to Flooded Land
Emission Factor
Other Constructed Waterbody
Climate Zone
MT CH4 ha 1 y1
MT C02 ha 1 y1
Freshwater ponds
Boreal
0.183
3.45
Freshwater ponds
Cool Temperate
0.183
3.74
Freshwater ponds
Warm Temperate Dry
0.183
6.23
Freshwater ponds
Warm Temperate Moist
0.183
5.35
Freshwater ponds
Tropical Dry/Montane
0.183
10.82
Freshwater ponds
Tropical Moist/Wet
0.183
10.16
Land Use, Land-Use Change, and Forestry 6-155
-------
1 Area estimates
2 Other constructed waterbodies were identified from the NHDWaterbody layer in the National Hydrography
3 Dataset Plus V2 (NHD)91, the National Inventory of Dams (NID)92, the National Wetlands Inventory (NWI)93, and
4 the Navigable Waterways (NW) network94. The NHD only covers the conterminous U.S., whereas the NID, NW and
5 NWI also include Alaska, Hawaii, and Puerto Rico..
6 Waterbodies in the NHDWaterbody layer that were less than or equal to 20-years old, less than 8 ha in surface
7 area, not identified as canal/ditch in NHD, and met any of the following criteria were considered freshwater ponds
8 in Land Converted to Flooded Land: 1) the waterbody was classified "Reservoir" in the NHDWaterbody layer, 2) the
9 waterbody name in the NHDWaterbody layer included "Reservoir", 3) the waterbody in the NHDWaterbody layer
10 was located in close proximity (up to 100 m) to a dam in the NID, 4) the NHDWaterbody GNIS name was similar to
11 nearby NID feature (between 100 m to 1000 m).
12 EPA assumes that all features included in the NW are subject to water-level management to maintain minimum
13 water depths required for navigation and are therefore managed flooded lands. NW features that were less than 8
14 ha in surface area and not identified as canals/ditch (see below) were considered freshwater ponds. Only 2.1
15 percent of NW features met these criteria, and they were primarily associated with larger navigable waterways,
16 such as lock chambers on impounded rivers.
17 NWI features were considered "managed" if they had a special modifier value indicating the presence of
18 management activities (Figure 6-19). To be included in the flooded lands inventory, the managed flooded land had
19 to be wet or saturated for at least one season per year (see 'Water Regime' in Figure 6-19). NWI features that met
20 these criteria, were less than 8 Ha in surface area, and were not a canal/ditch were defined as freshwater ponds.
21 Surface areas for other constructed waterbodies were taken from NHD, NWI or the NW. If features from the NHD,
22 NWI, or the NW datasets overlapped, duplicate areas were erased. The first step was to take the final NWI Flooded
23 Lands features and use it to identify overlapping NHD features. If the NHD feature had its center in a NWI feature,
24 it was removed from analysis. Next, remaining NHD features were erased from any remaining overlapping NWI
25 features. Final selections of NHD and NWI features were used to erase any overlapping NW waterbodies.
26 The age of other constructed waterbody features was determined by assuming the waterbody was created the
27 same year as a nearby (up to 100 m) NID feature. If no nearby NID feature was identified, it was assumed the
28 waterbody was greater than 20-years old throughout the time series. No canal/ditch features were associated with
29 a nearby dam, therefore all canal/ditch features were assumed to be greater than 20-years old through the time
30 series.
31 For the year 2021, this Inventory contains 913 ha of freshwater ponds in Land Converted to Flooded Land. The
32 surface area of freshwater ponds decreased by 94 percent from 1990 to 2021 due to flooded lands aging out of
33 Land Converted to Flooded Land more quickly than new flooded lands entered the category. The greatest
34 reduction in freshwater pond surface area occurred in Iowa, Kansas, and Georgia (Table 6-110). Freshwater ponds
35 in the 2021 inventory are most abundant in Nebraska, Montana, and Kansas (Figure 6-22).
36 Table 6-109: National Surface Area Totals of Other Constructed Waterbodies in Land
37 Converted to Flooded Land (ha)
Other Constructed Waterbody
1990
2005
2017
2018
2019
2020
2021
Freshwater Ponds
15,572
5 3800
1805
1574
1299
1041
913
91 See https://www.usgs.gov/core-science-svstems/ngp/national-hvdrography.
92 See https://nid.sec.usace.armv.mil.
93 See https://www.fws.gov/program/national-wetlands-inventorv/data-download.
94 https://hifld-geoplatform.opendata.arcgis.com/datasets/geoplatform::navigable-waterway-network-lines-l/about.
6-156 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
Figure 6-22: Surface Area of Other Constructed Waterbodies in Land Converted to Flooded
Land(ha)
hectares
_ 100
® 75
— 50
— 25
0 Alaska
1000 mi
3
4 Table 6-110: State Surface Area Totals of Other Constructed Waterbodies in Land Converted
5 to Flooded Land (ha)
State
1990
2005
2017
2018
2019
2020
2021
Alabama
344
19
3
0
0
0
0
Alaska
6
0
0
0
0
0
0
Arizona
46
16
7
4
4
0
0
Arkansas
316
0
0
0
0
0
0
California
241
86
43
42
37
30
29
Colorado
276
45
60
50
52
51
54
Connecticut
54
3
0
0
0
0
0
Delaware
4
0
0
0
0
0
0
District of Columbia
0
0
0
0
0
0
0
Florida
128
50
10
9
5
5
5
Georgia
1,804
87
35
35
32
26
26
Hawaii
7
2
0
0
0
0
0
Idaho
102
1
0
0
0
0
0
Illinois
556
115
56
41
36
26
25
Indiana
510
115
30
27
22
22
22
Iowa
2,227
1,403
511
430
268
156
83
Kansas
2,017
127
111
98
91
85
69
Kentucky
390
25
2
1
1
1
1
Louisiana
133
10
5
5
5
1
1
Maine
54
8
0
0
0
0
0
Maryland
177
57
17
22
22
21
19
Massachusetts
66
70
88
80
74
68
67
Michigan
158
45
19
15
15
7
7
Minnesota
263
133
110
103
96
73
64
Mississippi
160
34
23
23
18
13
13
Missouri
285
17
4
4
3
3
3
Montana
368
108
100
100
96
96
96
Land Use, Land-Use Change, and Forestry 6-157
-------
Nebraska
1,271
274
191
142
130
108
100
Nevada
13
57
36
26
25
22
22
New Hampshire
35
12
1
1
1
1
0
New Jersey
1
0
0
0
0
0
0
New Mexico
6
0
0
0
0
0
0
New York
287
120
29
29
27
27
24
North Carolina
53
7
1
1
1
1
1
North Dakota
11
21
9
9
9
9
9
Ohio
389
250
104
93
79
53
38
Oklahoma
111
3
3
3
0
0
0
Oregon
8
9
7
0
0
0
0
Pennsylvania
19
9
0
0
0
0
0
Puerto Rico
0
0
0
0
0
0
0
Rhode Island
9
7
0
0
0
0
0
South Carolina
819
228
25
24
13
9
9
South Dakota
232
94
97
95
78
77
74
Tennessee
712
42
23
14
9
3
2
Texas
565
9
0
0
0
0
0
Utah
55
20
30
30
30
29
29
Vermont
17
4
0
0
0
0
0
Virginia
0
0
0
0
0
0
0
Washington
54
23
0
0
4
4
4
West Virginia
31
6
3
3
3
3
3
Wisconsin
146
9
7
7
7
7
7
Wyoming
39
16
5
6
6
6
6
TOTAL
15,572
3,800
1,805
1,574
1,299
1,041
913
1 Uncertainty
2 Uncertainty in estimates of CO2 and Cmemissions from Land Converted to Flooded Land-Other Constructed
3 Water Bodies include uncertainty in the default emission factors and the flooded land area inventory. Uncertainty
4 in emission factors is provided in the 2019 Refinement to the 2006IPCC Guidelines (IPCC 2019). Uncertainties in the
5 spatial data include 1) uncertainty in area estimates from the NHD and NW, and 2) uncertainty in the location of
6 dams in the NID. Overall uncertainties in the NHD, NWI, NID, and NW are unknown, but uncertainty for remote
7 sensing products is ± 10 -15 percent (IPCC 2003). EPA assumes an uncertainty of ± 15 percent for the flooded land
8 area inventory based on expert judgment. These uncertainties do not include the underestimate of pond surface
9 area discussed above.
10 Table 6-111: Approach 2 Quantitative Uncertainty Estimates for ChU and CO2 Emissions from
11 Other Constructed Waterbodies in Land Converted to Flooded Land
Source Gas 2021 Emission Estimate Uncertainty Range Relative to Emission Estimate3
(kt C02 Eq.) (kt C02 Eq.) (%)
Lower Bound
Upper Bound
Lower Bound
Upper Bound
Freshwater ponds
ch4
4.70
4.60
4.80
-2.7
3.2
Freshwater ponds
C02
4.28
4.18
4.37
-2.2
2.2
Total
8.95
8.77
9.19
-2.1
2.6
aRange of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
Note: Totals may not sum due to independent rounding.
6-158 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 QA/QC and Verification
2 The National Hydrography Data (NHD) is managed by the USGS with collaboration from many other federal, state,
3 and local entities. Extensive QA/QC procedures are incorporated into the curation of the NHD. The National
4 Inventory of Dams (NID) is maintained by the U.S. Army Corps of Engineers (USACE) in collaboration with the
5 Federal Emergency Management Agency (FEMA) and state regulatory offices. USACE resolves duplicative and
6 conflicting data from 68 data sources, which helps obtain the more complete, accurate, and updated NID. The
7 Navigable Waterways (NW) dataset is part of the U.S. Department of Transportation (USDOT)/Bureau of
8 Transportation Statistics (BTS) National Transportation Atlas Database (NTAD). The NW is a comprehensive
9 network database of the nation's navigable waterways updated on a continuing basis. U.S. Fish and Wildlife Service
10 is the principal agency in charge of wetland mapping including the National Wetlands Inventory. Quality and
11 consistency of the Wetlands Layer is supported by federal wetlands mapping and classification standards, which
12 were developed under the oversight of the Federal Geographic Data Committee (FGDC) with input by the FGDC
13 Wetlands Subcommittee. This dataset is part of the FGDC Water-Inland Theme, which is co-chaired by the FWS and
14 the U.S. Geological Survey.
15 General QA/QC procedures were applied to activity data, documentation, and emission calculations consistent
16 with the U.S. Inventory QA/QC plan, which is in accordance with Vol. 1 Chapter 6 of the 2006IPCC Guidelines (see
17 Annex 8 for more details). All calculations were executed independently in Excel and R. Ten percent of state and
18 national totals were randomly selected for comparison between the two approaches to ensure there were no
19 computational errors.
20 Recalculations Discussion
21 Methane and carbon dioxide emissions from other constructed waterbodies in Land Converted to Flooded Land
22 were recalculated using updated geospatial data in the 1990 through 2021 Inventory. The updated geospatial data
23 is more detailed than what was used for the 1990 through 2020 Inventory, and includes Alaska, Hawaii, and Puerto
24 Rico, which were not included in the 1990 through 2020 Inventory. Despite these recalculations, CO2 emission
25 estimates agreed to within 0.005 MMT CO2 between the previous (i.e., 1990 through 2020) and current (i.e., 1990
26 through 2021) Inventories.
27 In addition, the EPA updated the global warming potential (GWP) for calculating C02-equivalent emissions of CH4
28 (from 25 to 28) to reflect the 100-year GWP provided in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). The
29 previous Inventory used the 100-year GWP provided in the IPCC Fourth Assessment Report (AR4). This update was
30 applied across the entire time series. Further discussion on this update and the overall impacts of updating the
31 Inventory GWP values to reflect the AR5 can be found in Chapter 9, Recalculations and Improvements.
32 The net effect of these recalculations for CH4 emissions from constructed waterbodies was an increase in
33 emissions amounting to an average annual 11 percent increase over the time series from 1990 to 2020 compared
34 to the previous Inventory.
35 Planned Improvements
36 Features < 8 ha in the NW that were not identified as Canal/Ditch were defined as freshwater ponds. Many of
37 these features are lock chambers connected to an upstream reservoir. These systems likely have emission rates
38 more similar to a reservoir than freshwater pond. In the next Inventory (i.e., 1990 through 2022) these systems will
39 be classified as reservoirs.
Land Use, Land-Use Change, and Forestry 6-159
-------
1 6.10 Settlements Remaining Settlements
2 (CRF Category 4E1)
3 Soil Carbon Stock Changes (CRF Category 4E1)
4 Soil organic C stock changes for Settlements Remaining Settlements occur in both mineral and organic soils.
5 However, the United States does not estimate changes in soil organic C stocks for mineral soils in Settlements
6 Remaining Settlements. This approach is consistent with the assumption of the Tier 1 method in the 2006IPCC
1 Guidelines (IPCC 2006) that inputs equal outputs, and therefore the soil organic C stocks do not change. This
8 assumption may be re-evaluated in the future if funding and resources are available to conduct an analysis of soil
9 organic C stock changes for mineral soils in Settlements Remaining Settlements.
10 Drainage of organic soils is common when wetland areas have been developed for settlements. Organic soils, also
11 referred to as Histosols, include all soils with more than 12 to 20 percent organic C by weight, depending on clay
12 content (NRCS 1999; Brady and Weil 1999). The organic layer of these soils can be very deep (i.e., several meters),
13 and form under inundated conditions that results in minimal decomposition of plant residues. Drainage of organic
14 soils leads to aeration of the soil that accelerates decomposition rate and CO2 emissions.95 Due to the depth and
15 richness of the organic layers, C loss from drained organic soils can continue over long periods of time, which
16 varies depending on climate and composition (i.e., decomposability) of the organic matter (Armentano and
17 Menges 1986).
18 Settlements Remaining Settlements includes all areas that have been settlements for a continuous time period of
19 at least 20 years according to the 2015 United States Department of Agriculture (USDA) National Resources
20 Inventory (NRI) (USDA-NRCS 2018)96 or according to the National Land Cover Dataset (NLCD) for federal lands
21 (Yang et al. 2018; Fry et al. 2011; Homer et al. 2007, 2015). There are discrepancies between the current land
22 representation (See Section 6.1) and the area data that have been used in the Inventory for Settlements Remaining
23 Settlements. First, the current land representation is based on the latest NRI dataset, which includes data through
24 2017, but these data have not been incorporated into the Settlements Remaining Settlements Inventory. Second,
25 Alaska and the small amount of settlements on federal lands are not included in this Inventory even though these
26 areas are part of the U.S. managed land base. These differences lead to discrepancies between the managed area
27 in Settlements Remaining Settlements and the settlement area included in the Inventory analysis (Table 6-113).
28 There is a planned improvement to include CO2 emissions from drainage of organic soils in settlements of Alaska
29 and federal lands as part of a future Inventory (See Planned Improvements Section).
30 CO2 emissions from drained organic soils in settlements are 15.9 MMT CO2 Eq. (4.3 MMT C) in 2021 (See Table
31 6-112 and Table 6-113). Although the flux is relatively small, the amount has increased by over 40 percent since
32 1990 due to an increase in area of drained organic soils in settlements.
33 Table 6-112: Net CO2 Flux from Soil C Stock Changes in Settlements Remaining Settlements
34 (MMT COz Eq.)
Soil Type 1990 2005 2017 2018 2019 2020 2021
Organic Soils 113 122 16.0 15.9 15.9 15.9 15.9
95 N20 emissions from soils are included in the N20 Emissions from Settlement Soils section.
96 NRI survey locations are classified according to land use histories starting in 1979, and consequently the classifications are
based on less than 20 years from 1990 to 1998. This may have led to an overestimation of Settlements Remaining Settlements
in the early part of the time series to the extent that some areas are converted to settlements between 1971 and 1978.
6-160 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Table 6-113: Net CO2 Flux from Soil C Stock Changes in Settlements Remaining Settlements
2 (MMT C)
Soil Type
1990
2005
2017
2018
2019
2020
2021
Organic Soils
3.1
3.3
4.4
4.3
4.3
4.3
4.3
3 Methodology and Time-Series Consistency
4 An IPCC Tier 2 method is used to estimate soil organic C stock changes for organic soils in Settlements Remaining
5 Settlements (IPCC 2006). Organic soils in Settlements Remaining Settlements are assumed to be losing C at a rate
6 similar to croplands due to deep drainage, and therefore emission rates are based on country-specific values for
7 cropland (Ogle et al. 2003).
8 The land area designated as settlements is based primarily on the 2018 NRI (USDA-NRCS 2018) with additional
9 information from the NLCD (Yang et al. 2018; Fry et al. 2011; Homer et al. 2007, 2015). It is assumed that all
10 settlement area on organic soils is drained, and those areas are provided in Table 6-114 (See Section 6.1,
11 Representation of the U.S. Land Base for more information). The area of drained organic soils is estimated from
12 the NRI spatial weights and aggregated to the country (Table 6-114). The area of land on organic soils in
13 Settlements Remaining Settlements has increased from 220 thousand hectares in 1990 to over 303 thousand
14 hectares in 2015. The area of land on organic soils have been incorporated into the inventory analysis for
15 Settlements Remaining Settlements through 2015.
16 Table 6-114: Thousands of Hectares of Drained Organic Soils in Settlements Remaining
17 Settlements
Area
Year
(Thousand Hectares)
1990
220
2005
235
2014
291
2015
303
2016
*
2017
*
2018
*
2019
*
2020
*
2021
*
NRI data have not been incorporated into the
inventory after 2015, designated with asterisks
18 To estimate CO2 emissions from drained organic soils across the time series from 1990 to 2015, the area of organic
19 soils by climate (i.e., cool temperate, warm temperate, subtropical) in Settlements Remaining Settlements is
20 multiplied by the appropriate country-specific emission factors for Cropland Remaining Cropland under the
21 assumption that there is deep drainage of the soils. The emission factors are 11.2 MT C per ha in cool temperate
22 regions, 14.0 MT C per ha in warm temperate regions, and 14.3 MT C per ha in subtropical regions (see Annex 3.12
23 for more information).
24 In order to ensure time-series consistency, the same methods are applied from 1990 to 2015, and a linear
25 extrapolation method is used to approximate emissions for the remainder of the 2016 to 2021 time series (See Box
26 6-4 in Cropland Remaining Cropland). The extrapolation is based on a linear regression model with moving-average
27 (ARMA) errors using the 1990 to 2015 emissions data, and is a standard data splicing method for imputing missing
Land Use, Land-Use Change, and Forestry 6-161
-------
1 emissions data in a time series (IPCC 2006). The Tier 2 method described previously will be applied in future
2 Inventories to recalculate the estimates beyond 2015 as new activity data are integrated into the analysis.
3 Uncertainty
4 Uncertainty for the Tier 2 approach is derived using a Monte Carlo approach, along with additional uncertainty
5 propagated through the Monte Carlo Analysis for 2016 to 2021 based on the linear time series model. The results
6 of the Approach 2 Monte Carlo uncertainty analysis are summarized in Table 6-115. Soil C losses from drained
7 organic soils in Settlements Remaining Settlements for 2021 are estimated to be between 7.3 and 24.4 MMT CO2
8 Eq. at a 95 percent confidence level. This indicates a range of 54 percent below and 54 percent above the 2021
9 emission estimate of 15.9 MMT CO2 Eq.
10 Table 6-115: Uncertainty Estimates for CO2 Emissions from Drained Organic Soils in
11 Settlements Remaining Settlements (MMT CO2 Eq. and Percent)
2021 Emission
Source
Gas
Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT CO? Eq.)
(MMT CO? Eq.)
(%)
Lower Upper
Lower Upper
Bound Bound
Bound Bound
Organic Soils
C02
15.9
7.3 24.4
-54% 54%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence
interval.
12 QA/QC and Verification
13 Quality control measures included checking input data, model scripts, and results to ensure data are properly
14 handled throughout the inventory process. Inventory reporting forms and text are reviewed and revised as needed
15 to correct transcription errors. No errors were found in this Inventory.
16 Recalculations Discussion
17 There were no recalculations to the 1990 through 2020 time series in this Inventory.
is Planned Improvements
19 There are two key improvements planned for the inventory, including a) incorporating the latest land use data
20 from the USDA National Resources Inventory, and b) estimating CO2 emissions from drainage of organic soils in
21 settlements of Alaska and federal lands in order to provide a complete inventory of emissions for this category.
22 These improvements will resolve most of the differences between the managed land base for Settlements
23 Remaining Settlements and amount of area currently included in Settlements Remaining Settlements Inventory
24 (See Table 6-116). These improvements will be made as funding and resources are available to expand the
25 inventory for this source category.
26 Table 6-116: Area of Managed Land in Settlements Remaining Settlements that is not
27 included in the current Inventory (Thousand Hectares)
Area (Thousand Hectares)
SRS Managed Land SRS Area Included
Year Area (Section 6.1) in Inventory Difference
1990 30,561 30,425 136
1991 30,559 30,430 129
1992 30,556 30,434 123
6-162 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
1993
30,483
30,346
138
1994
30,398
30,264
135
1995
30,336
30,206
130
1996
30,276
30,157
119
1997
30,207
30,105
101
1998
30,141
30,041
99
1999
30,087
29,992
95
2000
30,029
29,949
80
2001
29,976
29,889
87
2002
29,969
29,882
87
2003
30,493
30,378
115
2004
30,986
30,859
127
2005
31,445
31,370
75
2006
31,953
31,812
140
2007
32,410
32,317
93
2008
33,028
32,922
106
2009
33,604
33,494
111
2010
34,179
34,069
111
2011
34,744
34,662
82
2012
35,315
35,215
100
2013
36,238
36,156
81
2014
37,172
37,129
43
2015
38,040
38,058
-18
2016
38,952
*
*
2017
39,875
*
*
2018
40,771
*
*
2019
41,617
*
*
2020
42,467
*
*
2021
43,189
*
*
NRI data have not been incorporated into the inventory after 2015, designated with
asterisks (*).
Changes in Carbon Stocks in Settlement Trees (CRF Source
Category 4E1)
Settlements are land uses where human populations and activities are concentrated. In these areas, the
anthropogenic impacts on tree growth, stocking and mortality are particularly pronounced (Nowak 2012) in
comparison to forest lands where non-anthropogenic forces can have more significant impacts. Estimates included
in this section include net CO2 and C flux from trees on Settlements Remaining Settlements and Land Converted to
Settlements as it is not possible to report on these separately at this time.
Trees in settlement areas of the United States are estimated to account for an average annual net sequestration of
117.2 MMT CO2 Eq. (32.0 MMT C) over the period from 1990 through 2021. Net C sequestration from settlement
trees in 2021 is estimated to be 137.8 MMT CO2 Eq. (37.6 MMT C) (Table 6-117). Dominant factors affecting C flux
trends for settlement trees are changes in the amount of settlement area (increasing sequestration due to more
land and trees) and net changes in tree cover (e.g., tree losses vs 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 C flux estimates. Annual sequestration increased by 43 percent between
1990 and 2021 due to increases in settlement area and changes in total 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 are less than in forests and this tree cover varies significantly across the United States (e.g., Nowak and
Land Use, Land-Use Change, and Forestry 6-163
-------
1 Greenfield 2018a). To quantify the C stored in settlement trees, the methodology used here requires analysis per
2 unit area of tree cover, rather than per unit of total land area (as is done for Forest Lands).
3 Table 6-117: Net Flux from Trees in Settlements Remaining Settlements (MMT CO2 Eq. and
4 MMT C)a
Year
1990
2005
2017
2018
2019
2020
2021
MMTCO2 Eq.
(96.4)
(117.4)
(129.6)
(129.5)
(129.3)
(136.7)
(137.8)
MMT C
(26.3)
(32.0)
(35.4)
(35.3)
(35.3)
(37.3)
(37.6)
a These estimates include net C02 and C flux from trees on Settlements Remaining Settlements and Land
Converted to Settlements as it is not possible to report on these separately at this time.
Note: Parentheses indicate net sequestration.
5 Methodology and Time-Series Consistency
6 To estimate net carbon sequestration in settlement areas, three types of data are required for each state:
7 1. Settlement area
8 2. Percent tree cover in settlement areas
9 3. Carbon sequestration density per unit of tree cover
10 Settlement Area
11 Settlement area is defined in Section 6.1 Representation of the U.S. Land Base as a land-use category representing
12 developed areas. The data used to estimate settlement area within Section 6.1 comes from the latest NRI as
13 updated through 2017, with the extension of the time series through 2021 based on assuming the settlement area
14 is the same as 2017. NRI data is also harmonized with the FIA dataset, which are available through 2021, and the
15 NLCD dataset, which is available through 2019. This process of combining the datasets extends the time series to
16 ensure that there is a complete and consistent representation of land use data for all source categories in the
17 LULUCF sector. Annual estimates of CO2 flux (Table 6-117) were developed based on estimates of annual
18 settlement area and tree cover derived from NLCD developed lands. Developed land, which was used to estimate
19 tree cover in settlement areas, is about six percent higher than the area categorized as Settlements in the
20 Representation of the U.S. Land Base developed for this report.
21 Percent Tree Cover in Settlement Areas
22 Percent tree cover in settlement area by state is needed to convert settlement land area to settlement tree cover
23 area. Converting to tree cover area is essential as tree cover, and thus C estimates, can vary widely among states in
24 settlement areas due to variations in the amount of tree cover (e.g., Nowak and Greenfield 2018a). However, since
25 the specific geography of settlement area is unknown because they are based on NRI sampling methods, NLCD
26 developed land was used to estimate the percent tree cover to be used in settlement areas. NLCD developed land
27 cover classes 21-24 (developed, open space (21), low intensity (22), medium intensity (23), and high intensity (24))
28 were used to estimate percent tree cover in settlement area by state (U.S. Department of Interior 2018; MRLC
29 2013).
30 a)
31
32
33
34
35
36 b)
37
"Developed, Open Space - areas with a mixture of some constructed materials, but mostly vegetation in
the form of lawn grasses. Impervious surfaces account for less than 20 percent of total cover. These areas
most commonly include large-lot single-family housing units, parks, golf courses, and vegetation planted
in developed settings for recreation, erosion control, or aesthetic purposes." Plots designated as either
park, recreation, cemetery, open space, institutional or vacant land were classified as Developed Open
Space.
"Developed, Low Intensity - areas with a mixture of constructed materials and vegetation. Impervious
surfaces account for 20 to 49 percent of total cover. These areas most commonly include single-family
6-164 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
housing units." Plots designated as single family or low-density residential land were classified as
Developed, Low Intensity.
c) "Developed, Medium Intensity - areas with a mixture of constructed materials and vegetation.
Impervious surfaces account for 50 to 79 percent of the total cover. These areas most commonly include
single-family housing units." Plots designated as medium density residential, other urban or mixed urban
were classified as Developed, Medium Intensity.
d) "Developed High Intensity - highly developed areas where people reside or work in high numbers.
Examples include apartment complexes, row houses and commercial/industrial. Impervious surfaces
account for 80 to 100 percent of the total cover." Plots designated as either commercial, industrial, high
density residential, downtown, multi-family residential, shopping, transportation or utility were classified
as Developed, High Intensity.
As NLCD is known to underestimate tree cover (Nowak and Greenfield 2010), photo-interpretation of tree cover
within NLCD developed lands was conducted for the years of c. 2011 and 2016 using 1,000 random points to
determine an average adjustment factor for NLCD tree cover estimates in developed land and determine recent
tree cover changes. This photo-interpretation of change followed methods detailed in Nowak and Greenfield
(2018b). Percent tree cover (%TC) in settlement areas by state was estimated as:
%TC in state = state NLCD %TC x national photo-interpreted %TC / national NLCD %TC
Percent tree cover in settlement areas by year was set as follows:
• 1990 to 2011: used 2011 NLCD tree cover adjusted with 2011 photo-interpreted values
• 2012 to 2015: used 2011 NLCD tree cover adjusted with photo-interpreted values, which were
interpolated from values between 2011 and 2016
• 2016 to 2020: used 2011 NLCD tree cover adjusted with 2016 photo-interpreted values
Carbon Sequestration Density per Unit of Tree Cover
Methods for quantifying settlement tree biomass, C sequestration, and C emissions from tree mortality and
decomposition were taken directly from Nowak et al. (2013), Nowak and Crane (2002), and Nowak (1994). In
general, net C sequestration estimates followed three steps, each of which is explained further in the paragraphs
below. First, field data from cities and urban areas within entire states were used to estimate C in tree biomass
from field data on measured tree dimensions. Second, estimates of annual tree growth and biomass increment
were generated from published literature and adjusted for tree condition, crown competition, and growing season
to generate estimates of gross C sequestration in settlement trees for all 50 states and the District of Columbia.
Third, estimates of C emissions due to mortality and decomposition were subtracted from gross C sequestration
estimates to obtain estimates of net C sequestration. Carbon storage, gross and net sequestration estimates were
standardized per unit tree cover based on tree cover in the study area.
Settlement tree carbon estimates are based on published literature (Nowak et al. 2013; Nowak and Crane 2002;
Nowak 1994) as well as newer data from the i-Tree database97 and U.S. Forest Service urban forest inventory data
(e.g., Nowak et al. 2016, 2017) (Table 6-118). These data are based on collected field measurements in several U.S.
cities between 1989 and 2017. Carbon storage and sequestration in these cities were estimated using the U.S.
Forest Service's i-Tree Eco model (Nowak et al. 2008). This computer model uses standardized field data from
randomly located plots, along with local hourly air pollution and meteorological data, to quantify urban forest
structure, monetary values of the urban forest, and environmental effects, including total C stored and annual C
sequestration (Nowak et al. 2013).
97 See http://www.itreetools.org.
Land Use, Land-Use Change, and Forestry 6-165
-------
1 In each city, a random sample of plots were measured to assess tree stem diameter, tree height, crown height and
2 crown width, tree location, species, and canopy condition. The data for each tree were used to estimate total dry-
3 weight biomass using allometric models, a root-to-shoot ratio to convert aboveground biomass estimates to whole
4 tree biomass, and wood moisture content. Total dry weight biomass was converted to C by dividing by two (50
5 percent carbon content). An adjustment factor of 0.8 was used for open grown trees to account for settlement
6 trees having less aboveground biomass for a given stem diameter than predicted by allometric models based on
7 forest trees (Nowak 1994). Carbon storage estimates for deciduous trees include only C stored in wood. Estimated
8 C storage was divided by tree cover in the area to estimate carbon storage per square meter of tree cover.
9 Table 6-118: Carbon Storage (kg C/m2 tree cover), Gross and Net Sequestration (kg C/m2
10 tree cover/year) and Tree Cover (percent) among Sampled U.S. Cities (see Nowak et al.
11 2013)
Sequestration
City
Storage
SE
Gross
SE
Net
SE
Ratio3
Tree
Cover
SE
Adrian, Ml
12.17
1.88
0.34
0.04
0.13
0.07
0.36
22.1
2.3
Albuquerque, NM
5.61
0.97
0.24
0.03
0.20
0.03
0.82
13.3
1.5
Arlington, TX
6.37
0.73
0.29
0.03
0.26
0.03
0.91
22.5
0.3
Atlanta, GA
6.63
0.54
0.23
0.02
0.18
0.03
0.76
53.9
1.6
Austin, TX
3.57
0.25
0.17
0.01
0.13
0.01
0.73
30.8
1.1
Baltimore, MD
10.30
1.24
0.33
0.04
0.20
0.04
0.59
28.5
1.0
Boise, ID
7.33
2.16
0.26
0.04
0.16
0.06
0.64
7.8
0.2
Boston, MA
7.02
0.96
0.23
0.03
0.17
0.02
0.73
28.9
1.5
Camden, NJ
11.04
6.78
0.32
0.20
0.03
0.10
0.11
16.3
9.9
Casper, WY
6.97
1.50
0.22
0.04
0.12
0.04
0.54
8.9
1.0
Chester, PA
8.83
1.20
0.39
0.04
0.25
0.05
0.64
20.5
1.7
Chicago (region), IL
9.38
0.59
0.38
0.02
0.26
0.02
0.70
15.5
0.3
Chicago, IL
6.03
0.64
0.21
0.02
0.15
0.02
0.70
18.0
1.2
Corvallis, OR
10.68
1.80
0.22
0.03
0.20
0.03
0.91
32.6
4.1
El Paso, TX
3.93
0.86
0.32
0.05
0.23
0.05
0.72
5.9
1.0
Freehold, NJ
11.50
1.78
0.31
0.05
0.20
0.05
0.64
31.2
3.3
Gainesville, FL
6.33
0.99
0.22
0.03
0.16
0.03
0.73
50.6
3.1
Golden, CO
5.88
1.33
0.23
0.05
0.18
0.04
0.79
11.4
1.5
Grand Rapids, Ml
9.36
1.36
0.30
0.04
0.20
0.05
0.65
23.8
2.0
Hartford, CT
10.89
1.62
0.33
0.05
0.19
0.05
0.57
26.2
2.0
Houston, TX
4.55
0.48
0.31
0.03
0.25
0.03
0.83
18.4
1.0
Indiana15
8.80
2.68
0.29
0.08
0.27
0.07
0.92
20.1
3.2
Jersey City, NJ
4.37
0.88
0.18
0.03
0.13
0.04
0.72
11.5
1.7
Kansas'5
7.42
1.30
0.28
0.05
0.22
0.04
0.78
14.0
1.6
Kansas City (region),
MO/KS
7.79
0.85
0.39
0.04
0.26
0.04
0.67
20.2
1.7
Lake Forest Park, WA
12.76
2.63
0.49
0.07
0.42
0.07
0.87
42.4
0.8
Las Cruces, NM
3.01
0.95
0.31
0.14
0.26
0.14
0.86
2.9
1.0
Lincoln, NE
10.64
1.74
0.41
0.06
0.35
0.06
0.86
14.4
1.6
Los Angeles, CA
4.59
0.51
0.18
0.02
0.11
0.02
0.61
20.6
1.3
Milwaukee, Wl
7.26
1.18
0.26
0.03
0.18
0.03
0.68
21.6
1.6
Minneapolis, MN
4.41
0.74
0.16
0.02
0.08
0.05
0.52
34.1
1.6
Moorestown, NJ
9.95
0.93
0.32
0.03
0.24
0.03
0.75
28.0
1.6
Morgantown, WV
9.52
1.16
0.30
0.04
0.23
0.03
0.78
39.6
2.2
Nebraska15
6.67
1.86
0.27
0.07
0.23
0.06
0.84
15.0
3.6
New York, NY
6.32
0.75
0.33
0.03
0.25
0.03
0.76
20.9
1.3
North Dakota15
7.78
2.47
0.28
0.08
0.13
0.08
0.48
2.7
0.6
Oakland, CA
5.24
0.19
NA
NA
NA
NA
NA
21.0
0.2
Oconomowoc, Wl
10.34
4.53
0.25
0.10
0.16
0.06
0.65
25.0
7.9
Omaha, NE
14.14
2.29
0.51
0.08
0.40
0.07
0.78
14.8
1.6
Philadelphia, PA
8.65
1.46
0.33
0.05
0.29
0.05
0.86
20.8
1.8
6-166 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Phoenix, AZ
3.42
0.50
0.38
0.04
0.35
0.04
0.94
9.9
1.2
Roanoke, VA
9.20
1.33
0.40
0.06
0.27
0.05
0.67
31.7
3.3
Sacramento, CA
7.82
1.57
0.38
0.06
0.33
0.06
0.87
13.2
1.7
San Francisco, CA
9.18
2.25
0.24
0.05
0.22
0.05
0.92
16.0
2.6
Scranton, PA
9.24
1.28
0.40
0.05
0.30
0.04
0.74
22.0
1.9
Seattle, WA
9.59
0.98
0.67
0.06
0.55
0.05
0.82
27.1
0.4
South Dakotab
3.14
0.66
0.13
0.03
0.11
0.02
0.87
16.5
2.2
Syracuse, NY
9.48
1.08
0.30
0.03
0.22
0.04
0.72
26.9
1.3
Tennessee15
6.47
0.50
0.34
0.02
0.30
0.02
0.89
37.7
0.8
Washington, DC
8.52
1.04
0.26
0.03
0.21
0.03
0.79
35.0
2.0
Woodbridge, NJ
8.19
0.82
0.29
0.03
0.21
0.03
0.73
29.5
1.7
SE (Standard Error)
NA (Not Available)
a Ratio of net to gross sequestration
b Statewide assessment of urban areas
To determine gross sequestration rates, tree growth rates need to be estimated. Base growth rates were
standardized for open-grown trees in areas with 153 days of frost-free length based on measured data on tree
growth (Nowak et al. 2013). These growth rates were adjusted to local tree conditions based on length of frost-
free season, crown competition (as crown competition increased, growth rates decreased), and tree condition (as
tree condition decreased, growth rates decreased). Annual growth rates were applied to each sampled tree to
estimate gross annual sequestration-that is, the difference in C storage estimates between year 1 and year (x + 1)
represents the gross amount of C sequestered. These annual gross C sequestration rates for each tree were then
scaled up to city estimates using tree population information. Total C sequestration was divided by total tree cover
to estimate a gross carbon sequestration density (kg C/m2 of tree cover/year). The area of assessment for each city
or state was defined by its political boundaries; parks and other forested urban areas were thus included in
sequestration estimates.
Where gross C sequestration accounts for all C sequestered, net C sequestration for settlement trees considers C
emissions associated with tree death and removals. The third step in the methodology estimates net C emissions
from settlement trees based on estimates of annual mortality, tree condition, and assumptions about whether
dead trees were removed from the site. Estimates of annual mortality rates by diameter class and condition class
were obtained from a study of street-tree mortality (Nowak 1986). Different decomposition rates were applied to
dead trees left standing compared with those removed from the site. For removed trees, different rates were
applied to the removed/aboveground biomass in contrast to the belowground biomass (Nowak et al. 2002). The
estimated annual gross C emission rates for each plot were then scaled up to city estimates using tree population
information.
The full methodology development is described in the underlying literature, and key details and assumptions were
made as follows. The allometric models applied to the field data for the Nowak methodology for each tree were
taken from the scientific literature (see Nowak 1994, Nowak et al. 2002), but if no allometric model could be found
for the particular species, the average result for the genus or botanical relative was used. The adjustment (0.8) to
account for less live tree biomass in open-grown urban trees was based on information in Nowak (1994).
Measured tree growth rates for street (Frelich 1992; Fleming 1988; Nowak 1994), park (deVries 1987), and forest
(Smith and Shifley 1984) trees were standardized to an average length of growing season (153 frost free days) and
adjusted for site competition and tree condition. Standardized growth rates of trees of the same species or genus
were then compared to determine the average difference between standardized street tree growth and
standardized park and forest growth rates. Crown light exposure (CLE) measurements (number of sides and/or top
of tree exposed to sunlight) were used to represent forest, park, and open (street) tree growth conditions. Local
tree base growth rates were then calculated as the average standardized growth rate for open-grown trees
multiplied by the number of frost-free days divided by 153. Growth rates were then adjusted for CLE. The CLE-
adjusted growth rate was then adjusted based on tree condition to determine the final growth rate. Assumptions
for which dead trees would be removed versus left standing were developed specific to each land use and were
Land Use, Land-Use Change, and Forestry 6-167
-------
1 based on expert judgment of the authors. Decomposition rates were based on literature estimates (Nowak et al.
2 2013).
3 Estimates of gross and net sequestration rates for each of the 50 states and the District of Columbia (Table 6-119)
4 were compiled in units of C sequestration per unit area of tree canopy cover. These rates were used in conjunction
5 with estimates of state settlement area and developed land percent tree cover data to calculate each state's
6 annual net C sequestration by urban trees. This method was described in Nowak et al. (2013) and has been
7 modified here to incorporate developed land percent tree cover data.
8 Net annual C sequestration estimates were obtained for all 50 states and the District of Columbia by multiplying
9 the gross annual emission estimates by 0.73, the average ratio for net/gross sequestration (Table 6-119). However,
10 state specific ratios were used where available.
11 State Carbon Sequestration Estimates
12 The gross and net annual C sequestration values for each state were multiplied by each state's settlement area of
13 tree cover, which was the product of the state's settlement area and the state's tree cover percentage based on
14 NLCD developed land. The model used to calculate the total carbon sequestration amounts for each state, can be
15 written as follows:
16 Equation 6-1: Net State Annual Carbon Sequestration
17 Net state annual C sequestration (t C/yr) = Gross state sequestration rate (t C/ha/yr) x Net to Gross state
18 sequestration ratio x state settlement Area (ha) x % state tree cover in settlement area
19 The results for all 50 states and the District of Columbia are given in Table 6-119. This approach is consistent with
20 the default IPCC Gain-Loss methodology in IPCC (2006), although sufficient field data are not yet available to
21 separately determine interannual gains and losses in C stocks in the living biomass of settlement trees. Instead, the
22 methodology applied here uses estimates of net C sequestration based on modeled estimates of decomposition,
23 as given by Nowak et al. (2013).
24 Table 6-119: Estimated Annual C Sequestration, Tree Cover, and Annual C Sequestration per
25 Area of Tree Cover for settlement areas in the United States by State and the District of
26 Columbia (2021)
State
Gross Annual
Sequestration
(Metric Tons
C/Year)
Net Annual Tree
Sequestration Cover
(Metric Tons C/Year) (Percent)
Gross Annual
Sequestration
per Area of
Tree Cover
(kg C/m2/Year)
Net Annual
Sequestration
per Area of
Tree Cover
(kg C/m2/Year)
Net: Gross
Annual
Sequestration
Ratio
Alabama
2,237,744
1,630,587
53.2
0.376
0.274
0.73
Alaska
147,132
107,212
47.1
0.169
0.123
0.73
Arizona
165,651
120,706
4.5
0.388
0.283
0.73
Arkansas
1,311,140
955,394
48.6
0.362
0.264
0.73
California
2,015,600
1,468,717
16.8
0.426
0.311
0.73
Colorado
142,617
103,922
7.9
0.216
0.157
0.73
Connecticut
645,185
470,130
58.3
0.262
0.191
0.73
Delaware
101,454
73,927
24.3
0.366
0.267
0.73
DC
12,936
9,426
24.9
0.366
0.267
0.73
Florida
4,611,318
3,360,150
40.0
0.520
0.379
0.73
Georgia
3,855,749
2,809,586
56.0
0.387
0.282
0.73
Hawaii
302,417
220,363
41.4
0.637
0.464
0.73
Idaho
59,784
43,563
7.4
0.201
0.146
0.73
Illinois
670,100
488,285
15.4
0.310
0.226
0.73
Indiana
478,924
442,841
17.0
0.274
0.254
0.92
Iowa
177,970
129,682
8.5
0.263
0.191
0.73
Kansas
288,544
224,536
10.7
0.310
0.241
0.78
6-168 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Kentucky
983,018
716,300
36.5
0.313
0.228
0.73
Louisiana
1,579,396
1,150,865
46.7
0.435
0.317
0.73
Maine
441,832
321,952
55.2
0.242
0.176
0.73
Maryland
852,295
621,045
39.8
0.353
0.257
0.73
Massachusetts
1,087,795
792,648
56.9
0.278
0.203
0.73
Michigan
1,405,750
1,024,334
34.4
0.241
0.175
0.73
Minnesota
324,971
236,798
13.0
0.251
0.183
0.73
Mississippi
1,619,525
1,180,107
56.9
0.377
0.275
0.73
Missouri
876,489
638,675
23.0
0.313
0.228
0.73
Montana
45,227
32,956
4.8
0.201
0.147
0.73
Nebraska
97,883
82,600
7.3
0.261
0.220
0.84
Nevada
35,830
26,108
4.8
0.226
0.165
0.73
New Hampshire
389,857
284,079
58.9
0.238
0.174
0.73
New Jersey
958,420
698,376
40.5
0.321
0.234
0.73
New Mexico
189,487
138,075
10.1
0.288
0.210
0.73
New York
1,601,568
1,167,022
39.7
0.263
0.192
0.73
North Carolina
3,423,492
2,494,611
53.8
0.341
0.249
0.73
North Dakota
18,755
8,912
1.7
0.244
0.116
0.48
Ohio
1,275,219
929,220
28.1
0.271
0.198
0.73
Oklahoma
721,283
525,580
21.9
0.364
0.265
0.73
Oregon
674,215
491,283
39.6
0.265
0.193
0.73
Pennsylvania
1,896,783
1,382,137
39.9
0.267
0.195
0.73
Rhode Island
126,971
92,521
49.6
0.283
0.206
0.73
South Carolina
2,027,815
1,477,617
53.4
0.370
0.269
0.73
South Dakota
29,388
25,485
2.8
0.258
0.224
0.87
Tennessee
1,673,175
1,496,015
40.8
0.332
0.297
0.89
Texas
4,403,317
3,208,585
28.3
0.403
0.294
0.73
Utah
119,889
87,360
11.6
0.235
0.172
0.73
Vermont
186,736
136,070
50.2
0.234
0.170
0.73
Virginia
2,095,911
1,527,237
52.5
0.321
0.234
0.73
Washington
1,133,393
825,874
37.3
0.282
0.206
0.73
West Virginia
769,654
560,827
63.7
0.264
0.192
0.73
Wisconsin
711,367
518,355
25.7
0.246
0.180
0.73
Wyoming
29,597
21,566
4.7
0.199
0.145
0.73
Total
51,030,569
37,580,224
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 C sequestration for each of the 50 states and the District of Columbia. A 10
percent uncertainty was associated with settlement area estimates based on expert judgment. Uncertainty
associated with estimates of percent settlement tree coverage for each of the 50 states was based on standard
error associated with the photo-interpretation of national tree cover in developed lands. Uncertainty associated
with estimates of gross and net C sequestration for each of the 50 states and the District of Columbia was based on
standard error estimates for each of the state-level sequestration estimates (Table 6-120). These estimates are
based on field data collected in each of the 50 states and the District of Columbia, and uncertainty in these
estimates increases as they are scaled up to the national level.
Additional uncertainty is associated with the biomass models, conversion factors, and decomposition assumptions
used to calculate C 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 C estimates and the forest tree C
estimates (e.g., Nowak et al. 2013). Due to data limitations, settlement soil flux is not quantified as part of this
Land Use, Land-Use Change, and Forestry 6-169
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
analysis, while reconciliation of settlement tree and forest tree estimates will be addressed through the land-
representation effort described in the Planned Improvements section of this chapter.
A Monte Carlo (Approach 2) uncertainty analysis was applied to estimate the overall uncertainty of the
sequestration estimate in 2021. The results of this quantitative uncertainty analysis are summarized in Table
6-120. The change in C stocks in Settlement Trees in 2021 was estimated to be between -208.1 and -66.95 MMT
CO2 Eq. at a 95 percent confidence level. This analysis indicates a range of 51 percent more sequestration to 51
percent less sequestration than the 2021 flux estimate of-137.79 MMT CO2 Eq.
Table 6-120: Approach 2 Quantitative Uncertainty Estimates for Net CO2 Flux from Changes
in C Stocks in Settlement Trees (MMT CO2 Eq. and Percent)
Source Gas
2021 Flux Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Flux Estimate3
(MMT C02 Eq.) (%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Changes in C Stocks in
CO2
Settlement Trees
(137.8)
(208.1) (67.0)
-51% 51%
a Range of C stock change estimates predicted by Monte Carlo Stochastic Simulation with a 95 percent confidence
interval.
Note: Parentheses indicate negative values or net sequestration.
QA/QC and Verification
Tier 1 and Tier 2 QA/QC activities were conducted consistent with the U.S. QA/QC plan. Source-specific quality
control measures for settlement trees included checking input data, documentation, and calculations to ensure
data were properly handled through the inventory process. Errors that were found during this process were
corrected as necessary.
Recalculations Discussion
The compilation methods remained the same in the latest inventory relative to the previous Inventory. New data
from the NRI and NLCD resulted in an increase in the settlement area for 2020, leading to a 5 percent increase in
the net C sequestration (Table 6-121).
Table 6-121: Recalculations of the Settlement Tree Categories
Category
Previous Estimate
2020,
2022 Inventory
Current Estimate
2020,
2023 Inventory
Current Estimate
2021,
2023 Inventory
Settlement Area (km2)
447,973
466,511
469,705
Settlement Tree Coverage (km2)
143,019
150,541
151,694
Net C Flux (MMT C)
(35.4)
(37.3)
(37.6)
Net C02 Flux MMT C02 Eq.
(129.8)
(136.7)
(137.8)
Planned Improvements
A consistent representation of the managed land base in the United States is discussed in Section 6.1
Representation of the U.S. Land Base, and discusses a planned improvement by the USDA Forest Service to
reconcile the overlap between Settlement Trees and the forest land categories. Estimates for Settlement Trees are
based on tree cover in settlement areas. Work is needed to clarify how much of this settlement area tree cover
may also be accounted for in "forest" area assessments as some of these forests may be adjacent to settlement
areas. For example, "forest" as defined by the USDA Forest Service Forest Inventory and Analysis (FIA) program fall
within urban areas. Nowak et al. (2013) estimates that 1.5 percent of forest plots measured by the FIA program fall
within land designated as Census urban, suggesting that approximately 1.5 percent of the C reported in the Forest
6-170 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
source category might also be counted in the urban areas. The potential overlap with settlement areas is unknown
at this time but research is underway to develop spatially explicit and spatially continuous land representation
products which will eliminate the potential for double counting. Future research may also enable more complete
coverage of changes in the C stock of trees for all settlements land.
To provide more accurate emissions estimates in the future, the following actions will be taken:
a) Photo-interpret settlement tree cover in 2021 to update tree cover estimates and trends
b) Update photo-interpretation for settlement areas using 2016 NLCD developed land information
c) Develop spatially explicit and spatially continuous representations of land to eliminate the overlap
between forest and settlement areas, as well as allow for improved estimates in "settlement areas."
N-.0 Emissions from Settlement Soils (CRF Source Category
4E1)
Of the synthetic N fertilizers applied to soils in the United States, approximately 1 to 2 percent are currently
applied to lawns, golf courses, and other landscaping within settlement areas, and contributes to soil N2O
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 N 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 N2O.
N additions to soils result in direct and indirect N2O emissions. Direct emissions occur on-site due to the N
additions in the form of synthetic fertilizers and biosolids as well as enhanced mineralization of N in drained
organic soils. Indirect emissions result from fertilizer and biosolids N that is transformed and transported to
another location in a form other than N2O (i.e., ammonia [NH3] and nitrogen oxide [NOx] volatilization, nitrate
[NO3 ] leaching and runoff), and later converted into N2O at the off-site location. The indirect emissions are
assigned to settlements because the management activity leading to the emissions occurred in settlements.
Total N2O emissions from soils in Settlements Remaining Settlements98 are 2.1 MMT CO2 Eq. (8 kt of N2O) in 2021.
There is an overall increase of 15 percent from 1990 to 2021 due to an expanding settlement area leading to more
synthetic N fertilizer applications that peaked in the mid-2000s. Inter-annual variability in these emissions is
directly attributable to variability in total synthetic fertilizer consumption, area of drained organic soils, and
biosolids applications in the United States. Emissions from this source are summarized in Table 6-122.
Table 6-122: N2O Emissions from Soils in Settlements Remaining Settlements (MMT CO2 Eq.
and kt N2O)
1990
2005
2017
2018
2019
2020
2021
MMT CO? Eq.
Direct N20 Emissions from Soils
1.5
2.2
1.6
1.7
1.7
1.7
1.7
Synthetic Fertilizers
0.8
1.5
0.7
0.8
0.8
0.8
0.8
Biosolids
0.2
0.1
0.1
0.1
0.1
0.1
0.1
Drained Organic Soils
0.5
0.6
0.7
0.7
0.7
0.7
0.7
Indirect N20 Emissions from Soils
0.3
0.5
0.3
0.3
0.3
0.3
0.3
Total
1.8
2.8
1.9
2.0
2.0
2.0
2.1
kt N20
Direct N20 Emissions from Soils
6
8
6
6
6
6
6
Synthetic Fertilizers
3
5
3
3
3
3
3
98 Estimates of Soil N20 for Settlements Remaining Settlements include emissions from Land Converted to Settlements because
it was not possible to separate the activity data.
Land Use, Land-Use Change, and Forestry 6-171
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Biosolids 1 1 11111
Drained Organic Soils 2 2 3 3 3 3 3
Indirect N20 Emissions from Soils 1 2 11111
Total 7 10 7 7 8 8 8
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
For settlement soils, the IPCC Tier 1 approach is used to estimate soil N2O emissions from synthetic N fertilizer,
biosolids additions, and drained organic soils. Estimates of direct N2O emissions from soils in settlements are based
on the amount of N in synthetic commercial fertilizers applied to settlement soils, the amount of N in biosolids
applied to non-agricultural land and surface disposal (see Section 7.2—Wastewater Treatment and Discharge for a
detailed discussion of the methodology for estimating biosolids available for non-agricultural land application), and
the area of drained organic soils within settlements.
Nitrogen applications to settlement soils are estimated using data compiled by the USGS (Brakebill and Gronberg
2017). The USGS estimated on-farm and non-farm fertilizer use is based on sales records at the county level from
1987 through 2012 (Brakebill and Gronberg 2017). Non-farm N fertilizer is assumed to be applied to settlements
and forest lands; values for 2013 through 2017 are based on 2012 values adjusted for total annual total N fertilizer
sales in the United States (AAPFCO 2016 through 2022) because there are no activity data on non-farm application
after 2012. Settlement application is calculated by subtracting forest application from total non-farm fertilizer use.
The amount of synthetic fertilization from 2018 to 2021 is determined using a linear extrapolation method (See
Box 6-4 in Cropland Remaining Cropland). This method is based on a linear regression model with moving-average
(ARMA) errors using the 1990 to 2017 fertilization data, and linear extrapolation. The total amount of fertilizer N
applied to settlements is multiplied by the IPCC default emission factor (1 percent) to estimate direct N2O
emissions (IPCC 2006) for 1990 to 2021.
Biosolids applications are derived from national data on biosolids generation, disposition, and N content (see
Section 7.2, Wastewater Treatment for further detail). The total amount of N resulting from these sources is
multiplied by the IPCC default emission factor for applied N (one percent) to estimate direct N2O emissions (IPCC
2006) for 1990 to 2021.
The IPCC (2006) Tier 1 method is also used to estimate direct N2O emissions due to drainage of organic soils in
settlements at the national scale. Estimates of the total area of drained organic soils are obtained from the 2015
NRI (USDA-NRCS 2018) using soils data from the Soil Survey Geographic Database (SSURGO) (Soil Survey Staff
2011). To estimate annual emissions from 1990 to 2015, the total area is multiplied by the IPCC default emission
factor for temperate regions (IPCC 2006). This Inventory does not include soil N2O emissions from drainage of
organic soils in Alaska and federal lands, although this is a planned improvement for a future Inventory.
For indirect emissions, the total N applied from fertilizer and biosolids is multiplied by the IPCC default factors of
10 percent for volatilization and 30 percent for leaching/runoff to calculate the amount of N volatilized and the
amount of N leached/runoff. The amount of N volatilized is multiplied by the IPCC default factor of one percent for
the portion of volatilized N that is converted to N2O off-site and the amount of N leached/runoff is multiplied by
the IPCC default factor of 0.075 percent for the portion of leached/runoff N that is converted to N2O off-site. The
resulting estimates are summed to obtain total indirect emissions from 1990 to 2021 for biosolids and synthetic
fertilization.
In order to ensure time-series consistency, the same methods are applied from 1990 to 2021 for biosolids. For
synthetic fertilizer, a linear extrapolation method is used to approximate fertilizer application for the remainder of
the 2018 to 2021 time series and then used to estimate emissions. For drainage of organic soils, the methods
described above are applied for 1990 to 2015, and a linear extrapolation method is used to approximate emissions
for the remainder of the 2016 to 2021 time series (See Box 6-4 in Cropland Remaining Cropland). The extrapolation
is based on a linear regression model with moving-average (ARMA) errors using the 1990 to 2015 emissions data,
and, and is a standard data splicing method for imputing missing emissions data in a time series (IPCC 2006). The
6-172 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
time series will be recalculated in a future Inventory with the methods described previously for drainage of organic
soils.
Uncertainty
The amount of N2O emitted from settlement soils depends not only on N 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
C availability; rate, application method, and timing of N input; oxygen gas partial pressure; soil moisture content;
pH; temperature; and irrigation/watering practices. The effect of the combined interaction of these variables on
N2O emissions is complex and highly uncertain. The IPCC default methodology does not explicitly incorporate these
variables, except variation in the total amount of fertilizer N and biosolids application, which leads to uncertainty
in the results.
Uncertainties exist in both the fertilizer N and biosolids application rates in addition to the emission factors.
Uncertainty in fertilizer N application is assigned a default level of ±50 percent." Uncertainty in the area of
drained organic soils is based on the estimated variance from the NRI survey (USDA-NRCS 2018). There is also
additional uncertainty associated with the fit of the linear regression model for the data splicing methods that was
used to estimate emissions associated with drainage of organic soils.
Uncertainty is propagated through the calculations of N2O emissions from fertilizer N and drainage of organic soils
based on a Monte Carlo analysis. The results are combined with the uncertainty in N2O emissions from the
biosolids application using simple error propagation methods (IPCC 2006). The results are summarized in Table
6-123. Direct N2O emissions from soils in Settlements Remaining Settlements in 2021 are estimated to be between
0.7 and 3.1 MMT CO2 Eq. at a 95 percent confidence level. This indicates a range of 57 percent below to 85 percent
above the 2021 emission estimate of 1.7 MMT CO2 Eq. Indirect N2O emissions in 2021 are between 0.1 and 1.0
MMT CO2 Eq., ranging from 78 percent below to 223 percent above the estimate of 0.3 MMT CO2 Eq.
Table 6-123: Quantitative Uncertainty Estimates of N2O Emissions from Soils in Settlements
Remaining Settlements (MMT CO2 Eq. and Percent)
Source Gas
2021 Emissions
(MMT C02 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower
Bound
Upper
Bound
Lower Upper
Bound Bound
Settlements Remaining Settlements
Direct N20 Emissions from Soils N20
1.7
0.7
3.1
-57% 85%
Indirect N20 Emissions from Soils N20
0.3
0.1
1.0
-78% 223%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
Note: These estimates include direct and indirect N20 emissions from Settlements Remaining Settlements and Land
Converted to Settlements because it was not possible to separate the activity data.
QA/QC and Verification
The spreadsheet containing fertilizer, drainage of organic soils, and biosolids applied to settlements and
calculations for N2O and uncertainty ranges have been checked. An error was found in the uncertainty calculation
and also some links in the spreadsheets that were causing errors. These errors were corrected.
99 No uncertainty is provided with the USGS fertilizer consumption data (Brakebill and Gronberg 2017) so a conservative ±50
percent is used in the analysis. Biosolids data are also assumed to have an uncertainty of ±50 percent.
Land Use, Land-Use Change, and Forestry 6-173
-------
1 Recalculations Discussion
2 Recalculations are associated with updated estimates for total fertilizers sales in a new AAPFCO report (AAPFCO
3 2022), along with revisions to the estimates derived from the linear extrapolation method.
4 EPA also updated the global warming potential (GWP) for calculating CCh-equivalent emissions of N2O (from 298 to
5 265) to reflect the 100-year GWP provided in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). The previous
6 Inventory used the 100-year GWP provided in the IPCC Fourth Assessment Report (AR4). This update was applied
7 across the entire time series.
8 As a result, calculated CC>2-equivalent total N2O emissions from settlement soils have decreased by an average
9 value of 0.3 MMT CO2 Eq. across the time series. This represents a 12 percent decrease in emissions compared to
10 the previous Inventory.
11 Further discussion on this update and the overall impacts of updating the inventory GWP values to reflect the AR5
12 can be found in Chapter 9, Recalculations and Improvements.
13 Planned Improvements
14 This source will be extended to include soil N2O emissions from drainage of organic soils in settlements of Alaska
15 and federal lands in order to provide a complete inventory of emissions for this category. In addition, this
16 Inventory needs to be updated with the latest land use data from the USDA National Resources Inventory (See
17 Planned Improvements in Settlements Remaining Settlements). Data on fertilizer amounts from 2018 to 2021 and
18 latest area data on drained organic soils will be incorporated into a future Inventory and used to recalculate the
19 time series.
20 Changes in Yard Trimmings and Food Scrap Carbon Stocks in
21 Landfills (CRF Category 4E1)
22 In the United States, yard trimmings (i.e., grass clippings, leaves, and branches) and food scraps account for a
23 significant portion of the municipal waste stream, and a large fraction of the collected yard trimmings and food
24 scraps are put in landfills. A portion of the carbon (C) contained in landfilled yard trimmings and food scraps can be
25 stored for very long periods.
26 Carbon storage estimates within the Inventory are associated with particular land uses. For example, harvested
27 wood products are reported under Forest Land Remaining Forest Land because these wood products originated
28 from the forest ecosystem. Similarly, C stock changes in yard trimmings and food scraps are reported under
29 Settlements Remaining Settlements because the bulk of the C, which comes from yard trimmings, originates from
30 settlement areas. While the majority of food scraps originate from cropland and grassland, in this Inventory they
31 are reported with the yard trimmings in the Settlements Remaining Settlements section. Additionally, landfills are
32 considered part of the managed land base under settlements (see Section 6.1 Representation of the U.S. Land
33 Base), and reporting these C stock changes that occur entirely within landfills fits most appropriately within the
34 Settlements Remaining Settlements section. The CH4 emissions resulting from anaerobic decomposition of yard
35 trimmings and food scraps in landfills are reported in the Waste chapter, see Section 7.1—Landfills.
36 The estimated amount of yard trimmings collected annually has stagnated since 1990 and the fraction that is
37 landfilled has been declining since 1990. From 1970 to 1990, yard trimmings collected for disposal increased by
38 about 51 percent. In 1990, over 53 million metric tons (wet weight) of yard trimmings and food scraps are
39 estimated to have been generated (i.e., put at the curb for collection to be taken to disposal sites or to composting
40 facilities) (EPA 2020). Since then, programs banning or discouraging yard trimmings disposal to landfills have led to
41 an increase in backyard composting and the use of mulching mowers, and consequently a slowing of year-over-
42 year increases in the tonnage of yard trimmings generated. From 1990 to 2021, yard trimmings collected for
6-174 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 disposal are estimated to have increased 1.1. percent. At the same time, an increase in the number of municipal
2 composting facilities has reduced the proportion of collected yard trimmings that are discarded in landfills per
3 year—from 72 percent in 1990 to 30 percent in 2021. The net effect of the slight increase in generation and the
4 increase in composting is a 58 percent decrease in the quantity of yard trimmings disposed of in landfills since
5 1990. Composting trends and emissions estimations are presented in the Waste chapter, Section 7.3 Composting.
6 Food scrap generation has grown by an estimated 165 percent since 1990. Though the proportion of total food
7 scraps generated that are eventually discarded in landfills has decreased from an estimated 82 percent in 1990 to
8 55 percent in 2020, the tonnage disposed of in landfills has increased considerably (by an estimated 78 percent)
9 due to the increase in food scrap generation. Although the total tonnage of food scraps disposed of in landfills has
10 increased from 1990 to 2021, the difference in the amount of food scraps added from one year to the next
11 generally decreased, and consequently the annual carbon stock net changes from food scraps have generally
12 decreased as well (as shown in Table 6-124 and Table 6-125). Landfilled food scraps decompose over time,
13 producing CFU and CO2. Decomposition happens at a higher rate initially, then decreases. As decomposition
14 decreases, the carbon stock becomes more stable. Because the cumulative carbon stock left in the landfill from
15 previous years is (1) not decomposing as much as the carbon introduced from food scraps in a single more recent
16 year; and (2) is much larger than the carbon introduced from food scraps in a single more recent year, the total
17 carbon stock in the landfill is primarily driven by the more stable "older" carbon stock, thus resulting in decreasing
18 annual changes in later years.
19 Overall, the decrease in the landfill disposal rate of yard trimmings has more than compensated for the increase in
20 food scrap disposal in landfills, and the net result is a decrease in the annual net change in landfill C storage from
21 24.5 MMT C02 Eq. (6.7 MMT C) in 1990 to 12.6 MMT C02 Eq. (3.4 MMT C) in 2021 (Table 6-124 and Table 6-125), a
22 decrease of 51 percent over the time series.
23 Table 6-124: Net Changes in Yard Trimmings and Food Scrap Carbon Stocks in Landfills
24 (MMT COz Eq.)
Carbon Pool
1990
2005
2017
2018
2019
2020
2021
Yard Trimmings
(20.1)
(7.5)
(8.3)
(8.3)
(8.2)
(8.2)
(8.1)
Grass
(1.7)
(0.6)
(0.8)
(0.8)
(0.8)
(0.8)
(0.7)
Leaves
(8.7)
(3.4)
(3.8)
(3.8)
(3.8)
(3.8)
(3.7)
Branches
(9.8)
(3.4)
(3.7)
(3.7)
(3.7)
(3.7)
(3.6)
Food Scraps
(4.4)
(3.9)
(5.6)
(5.2)
(4.8)
(4.5)
(4.5)
Total Net Flux
(24.5)
(11.4)
(13.8)
(13.4)
(13.1)
(12.8)
(12.6)
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
25 Table 6-125: Net Changes in Yard Trimmings and Food Scrap Carbon Stocks in Landfills
26 (MMT C)
Carbon Pool
1990
2005
2017
2018
2019
2020
2021
Yard Trimmings
(5.5)
(2.0)
(2.3)
(2.3)
(2.2)
(2.2)
(2.2)
Grass
(0.5)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Leaves
(2.4)
(0.9)
(1.0)
(1.0)
(1.0)
(1.0)
(1.0)
Branches
(2.7)
(0.9)
(1.0)
(1.0)
(1.0)
(1.0)
(1.0)
Food Scraps
(1.2)
(1.1)
(1.5)
(1.4)
(1.3)
(1.2)
(1.2)
Total Net Flux
(6.7)
(3.1)
(3.8)
(3.7)
(3.6)
(3.5)
(3.4)
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
27 Methodology and Time-Series Consistency
28 When wastes of biogenic origin (such as yard trimmings and food scraps) are landfilled and do not completely
29 decompose, the C that remains is effectively removed from the C cycle. Empirical evidence indicates that yard
Land Use, Land-Use Change, and Forestry 6-175
-------
1 trimmings and food scraps do not completely decompose in landfills (Barlaz 1998, 2005, 2008; De la Cruz and
2 Barlaz 2010), and thus the stock of C in landfills can increase, with the net effect being a net atmospheric removal
3 of C. Estimates of net C flux resulting from landfilled yard trimmings and food scraps were developed by estimating
4 the change in landfilled C stocks between inventory years and uses a country-specific methodology based on the
5 methodology for estimating the amount of harvested wood products stored in solid waste disposal systems that is
6 provided in the Land Use, Land-Use Change, and Forestry sector in IPCC (2003) and the 2006IPCC Guidelines for
1 National Greenhouse Gas Inventories (IPCC 2006). Carbon stock estimates were calculated by determining the
8 mass of landfilled C resulting from yard trimmings and food scraps discarded in a given year; adding the
9 accumulated landfilled C from previous years; and subtracting the mass of C that was landfilled in previous years
10 and has since decomposed and been emitted as CO2 and CH4.
11 To determine the total landfilled C stocks for a given year, the following data and factors were assembled:
12 (1) The composition of the yard trimmings (i.e., the proportion of grass, leaves and branches);
13 (2) The mass of yard trimmings and food scraps discarded in landfills;
14 (3) The C storage factor of the landfilled yard trimmings and food scraps; and
15 (4) The rate of decomposition of the degradable C.
16 The composition of yard trimmings was assumed to be 30 percent grass clippings, 40 percent leaves, and 30
17 percent branches on a wet weight basis (Oshins and Block 2000). The yard trimmings were subdivided, because
18 each component has its own unique adjusted C storage factor (i.e., moisture content and C content) and rate of
19 decomposition. The mass of yard trimmings and food scraps disposed of in landfills was estimated by multiplying
20 the quantity of yard trimmings and food scraps discarded by the proportion of discards managed in landfills. Data
21 on discards (i.e., the amount generated minus the amount diverted to centralized composting facilities) for both
22 yard trimmings and food scraps were taken primarily from Advancing Sustainable Materials Management: Facts
23 and Figures 2018 (EPA 2020), which provides data for 1960,1970,1980,1990, 2000, 2005, 2010, 2015, 2017 and
24 2018. To provide data for some of the missing years, detailed backup data were obtained from the 2012, 2013, and
25 2014, 2015, and 2017 versions of the Advancing Sustainable Materials Management: Facts and Figures reports
26 (EPA 2019), as well as historical data tables that EPA developed for 1960 through 2012 (EPA 2016). Remaining
27 years in the time series for which data were not provided were estimated using linear interpolation. Since the
28 Advancing Sustainable Materials Management: Facts and Figures reports for 2019, 2020, and 2021 were
29 unavailable, landfilled material generation, recovery, and disposal data for 2019, 2020, and 2021 were proxied
30 equal to 2018 values.
31 The amount of C disposed of in landfills each year, starting in 1960, was estimated by converting the discarded
32 landfilled yard trimmings and food scraps from a wet weight to a dry weight basis, and then multiplying by the
33 initial (i.e., pre-decomposition) C content (as a fraction of dry weight). The dry weight of landfilled material was
34 calculated using dry weight to wet weight ratios (Tchobanoglous et al. 1993, cited by Barlaz 1998) and the initial C
35 contents and the C storage factors were determined by Barlaz (1998, 2005, 2008).
36 The amount of C remaining in the landfill for each subsequent year was tracked based on a simple model of C fate
37 based on a laboratory experiment simulating decomposition of landfilled biogenic materials by methanogenic
38 microbes (Barlaz 1998, 2005, 2008). Carbon remaining in landfilled materials is expressed as a proportion of initial
39 C content, shown in the row labeled "C Storage Factor, Proportion of Initial C Stored (%)" in Table 6-126.
40 The modeling approach applied to simulate U.S. landfill C flows builds on the findings of Barlaz (1998, 2005, 2008).
41 The proportion of C stored is assumed to persist in landfills. The remaining portion is assumed to degrade over
6-176 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
time, resulting in emissions of CFU and CO2.100 The degradable portion of the C is assumed to decay according to
first-order kinetics. The decay rates for each of the materials are shown in Table 6-126.
The first-order decay rates, k, for each waste component are derived from De la Cruz and Barlaz (2010):
• De la Cruz and Barlaz (2010) calculate first-order decay rates using laboratory data published in Eleazer et
al. (1997), and a correction factor,/, is calculated so that the weighted average decay rate for all
components is equal to the EPA AP-42 default decay rate (0.04) for mixed MSW for regions that receive
more than 25 inches of rain annually (EPA 1995). Because AP-42 values were developed using landfill data
from approximately 1990, De la Cruz and Barlaz used 1990 waste composition for the United States from
EPA's Characterization of Municipal Solid Waste in the United States: 1990 Update (EPA 1991) to calculate
/. De la Cruz and Barlaz multiplied this correction factor by the Eleazer et al. (1997) decay rates of each
waste component to develop field-scale first-order decay rates.
• De la Cruz and Barlaz (2010) also use other assumed initial decay rates for mixed MSW in place of the AP-
42 default value based on different types of environments in which landfills in the United States are
located, including dry conditions (less than 25 inches of rain annually, k=0.02) and bioreactor landfill
conditions (moisture is controlled for rapid decomposition, /c=0.12).
Similar to the methodology in the Landfills section of the Inventory (Section 7.1), which estimates CFU emissions,
the overall MSW decay rate is estimated by partitioning the U.S. landfill population into three categories based on
annual precipitation ranges of: (1) Less than 20 inches of rain per year, (2) 20 to 40 inches of rain per year, and (3)
greater than 40 inches of rain per year. These correspond to overall MSW decay rates of 0.020,0.038, and 0.057
year"1, respectively. De la Cruz and Barlaz (2010) calculate component-specific decay rates corresponding to the
first value (0.020 year"1), but not for the other two overall MSW decay rates.
To maintain consistency between landfill-related methodologies across the Inventory, EPA developed correction
factors (/) for decay rates of 0.038 and 0.057 year"1 through linear interpolation. A weighted national average
component-specific decay rate is calculated by assuming that waste generation is proportional to population (the
same assumption used in the landfill methane emission estimate), based on population data from the 2000 U.S.
Census. The percent of census population is calculated for each of the three categories of annual precipitation
(noted in the previous paragraph); the population data are used as a surrogate for the number of landfills in each
annual precipitation category. Precipitation range percentages weighted by population are updated over time as
new Census data are available, to remain consistent with percentages used in the Waste chapter, Section 7.1
Landfills. The component-specific decay rates are shown in Table 6-126.
De la Cruz and Barlaz (2010) also use other assumed initial decay rates for mixed MSW in place of the AP-42
default value based on different types of environments in which landfills in the United States are located, including
dry conditions (less than 25 inches of rain annually, k=0.02) and bioreactor landfill conditions (moisture is
controlled for rapid decomposition, /c=0.12).
For each of the four materials (grass, leaves, branches, food scraps), the stock of C in landfills for any given year is
calculated according to Equation 6-2:
Equation 6-2: Total C Stock for Yard Trimmings and Food Scraps in Landfills
f
LFG,t= 1 Win x (1 - MCI) x ICQx {[CSi~x ICQ + [(1 - (CS,x ICC)) x e-/rff" "J]}
n
where,
t = Year for which C stocks are being estimated (year),
100 The CH4 emissions resulting from anaerobic decomposition of yard trimmings and food scraps in landfills are reported in the
Waste chapter, Section 7.1 Landfills.
Land Use, Land-Use Change, and Forestry 6-177
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
/ =
Waste type for which C stocks are being estimated (grass, leaves, branches, food
scraps),
LFO,t =
Stock of C in landfills in year t, for waste / (metric tons),
Win
Mass of waste / disposed of in landfills in year n (metric tons, wet weight),
n =
Year in which the waste was disposed of (year, where 1960 < n
-------
1
Table 6-127: C Stocks in Yard Trimmings and Food Scraps in Landfills (MMT C)
Carbon Pool
1990
2005
2017
2018
2019
2020
2021
2022a
Yard Trimmings
156.0
203.1
229.4
231.6
233.9
236.1
238.4
240.6
Branches
14.6
18.1
20.5
20.7
20.9
21.1
21.3
21.5
Leaves
66.7
87.4
99.4
100.4
101.5
102.5
103.6
104.6
Grass
74.7
97.7
109.5
110.5
111.5
112.5
113.5
114.4
Food Scraps
17.9
33.2
45.4
46.9
48.3
49.6
50.9
52.1
Total Carbon Stocks
173.9
236.3
274.8
278.5
282.2
285.7
289.2
292.7
a 2022 C stock estimate was forecasted using 1990 to 2021 data.
Note: Totals may not sum due to independent rounding.
2 Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
3 through 2021. When available, the same data source was used across the entire time series for the analysis. When
4 data were unavailable, missing values were estimated using linear interpolation or forecasting, as noted above.
5 Uncertainty
6 The uncertainty analysis for landfilled yard trimmings and food scraps includes an evaluation of the effects of
7 uncertainty for the following data and factors: disposal in landfills per year (tons of C), initial C content, moisture
8 content, decay rate, and proportion of C stored. The C storage landfill estimates are also a function of the
9 composition of the yard trimmings (i.e., the proportions of grass, leaves and branches in the yard trimmings
10 mixture). There are respective uncertainties associated with each of these factors.
11 A Monte Carlo (Approach 2) uncertainty analysis was applied to estimate the overall uncertainty of the
12 sequestration estimate for 2021. The results of the Approach 2 quantitative uncertainty analysis are summarized in
13 Table 6-128. Total yard trimmings and food scraps CO2 flux in 2021 was estimated to be between -21.6 and -5.5
14 MMT CO2 Eq. at a 95 percent confidence level. This indicates a range of 72 percent below to 56 percent above the
15 2021 flux estimate of -12.6 MMT C02 Eq.
16 Table 6-128: Approach 2 Quantitative Uncertainty Estimates for CO2 Flux from Yard
17 Trimmings and Food Scraps in Landfills (MMT CO2 Eq. and Percent)
2021 Flux
Source Gas
Estimate
Uncertainty Range Relative to Flux Estimate3
(MMT CO? Eq.)
(MMT CO?
Eq.)
(%)
Lower
Upper
Lower Upper
Bound
Bound
Bound Bound
Yard Trimmings and Food
CO2
Scraps
(12.6)
(21.6)
(5.5)
-72% 56%
a Range of flux estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
Note: Parentheses indicate negative values or net C sequestration.
is QA/QC and Verification
19 Tier 1 and Tier 2 QA/QC activities were conducted consistent with the U.S. QA/QC plan. Source-specific quality
20 control measures for Landfilled Yard Trimmings and Food Scraps included checking that input data were properly
21 transposed within the spreadsheet, checking calculations were correct, and confirming that all activity data and
22 calculations documentation was complete and updated to ensure data were properly handled through the
23 inventory process.
24 Order of magnitude checks and checks of time-series consistency were performed to ensure data were updated
25 correctly and any changes in emissions estimates were reasonable and reflected changes in activity data. An
Land Use, Land-Use Change, and Forestry 6-179
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
annual change trend analysis was also conducted to ensure the validity of the emissions estimates. Errors that
were found during this process were corrected as necessary.
To ensure consistency across the LULUCF and Waste sectors, and the accuracy of emissions, EPA plans to perform
a comparison of the activity data used and carbon inputs between the Landfilled Yard Trimmings and Food Scraps,
and the Waste chapter, Section 7.1—Landfills categories.
Recalculations Discussion
No recalculations were performed for the 1990-2021 inventory, as the Advancing Sustainable Materials
Management: Facts and Figures report for 2019, 2020, and 2021 were not yet available.
Planned Improvements
EPA notes the following improvements may be implemented or investigated within the next two or three
inventory cycles pending time and resource constraints:
• MSW data more recent than 2018 have not been released through the Advancing Sustainable Materials
Management reports. EPA will monitor the release schedule for these data and evaluate data for
integration into the Inventory when released. Six new food waste management pathways were
introduced in the 2018 Advancing Sustainable Materials Management report. Time series data for all of
these pathways are not provided prior to 2018 but EPA plans to investigate potential data sources and/or
methods to address time-series consistency and apply these data to the time series.
• EPA has been made aware of inconsistencies in landfilled food scraps data reported to the EPA
Greenhouse Gas Reporting Program (GHGRP) and will evaluate changes to how landfilled and energy
recovery values for yard trimmings and food scraps are calculated.
EPA notes the following improvements will continued to be investigated as time and resources allow, but there are
no immediate plans to implement these improvements until data are available or identified:
• EPA also plans to continue to investigate updates to the decay rate estimates for food scraps, leaves,
grass, and branches, as well as evaluate using decay rates that vary over time based on Census population
and climate data changes over time. Currently the inventory calculations use 2010 U.S. Census data, but
2020 U.S. Census data may be available.
• Other improvements include investigation into yard waste composition to determine if changes need to
be made based on changes in residential practices. A review of available literature will be conducted to
determine if there are changes in the allocation of yard trimmings. For example, leaving grass clippings in
place is becoming a more common practice, thus reducing the percentage of grass clippings in yard
trimmings disposed in landfills. In addition, agronomists may be consulted for determining the mass of
grass per acre on residential lawns to provide an estimate of total grass generation for comparison with
Inventory estimates.
• EPA will continue to evaluate data from recent peer-reviewed literature that may modify the default C
storage factors, initial C contents, and decay rates for yard trimmings and food scraps in landfills -
particularly updates to population precipitation ranges used to calculate k values. Based upon this
evaluation, changes may be made to the default values.
• Finally, EPA plans to review available data to ensure all types of landfilled yard trimmings and food scraps
are being included in the Inventory estimates, such as debris from road construction and commercial food
waste not included in other Inventory estimates.
6-180 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
6.11 Land Converted to Settlements (CRF
Category 4E2)
Land Converted to Settlements includes all settlements in an Inventory year that had been in another land use(s)
during the previous 20 years (USDA-NRCS 2015).101 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).
Land use change can lead to large losses of carbon (C) to the atmosphere, particularly conversions from forest land
(Houghton et al. 1983). Moreover, conversion of forest to another land use (i.e., deforestation) is one of the largest
anthropogenic sources of emissions to the atmosphere globally (Schimel 1995), although this source may be
declining globally (Tubiello et al. 2015). 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 this Inventory. Loss of
aboveground and belowground biomass, dead wood and litter C are reported for Forest Land Converted to
Settlements and Woodlands associated with Grasslands Converted to Settlements, but not for other land-use
conversions to settlements.
There are discrepancies between the current land representation (See Section 6.1) and the area data that have
been used in the inventory for Land Converted to Settlements. First, the current land representation is based on
the latest NRI dataset, which includes data through 2017, but these data have not been incorporated into the Land
Converted to Settlements Inventory. Second, this Inventory includes all settlements in the conterminous United
States and Hawaii, but does not include settlements in Alaska. Areas of drained organic soils in settlements on
federal lands are also not included in this Inventory. These differences lead to discrepancies between the managed
area in Land Converted to Settlements and the settlement area included in the Inventory analysis (Table 6-128).
There is a planned improvement to include CO2 emissions from drainage of organic soils in settlements of Alaska
and federal lands as part of a future Inventory (See Planned Improvements Section).
Forest Land Converted to Settlements is the largest source of emissions from 1990 to 2021, accounting for
approximately 75 percent of the average total loss of C among all of the land-use conversions in Land Converted to
Settlements. Total losses of aboveground and belowground biomass, dead wood and litter C losses in 2021 for all
conversions are 38.9, 7.4, 6.6, and 9.7 MMT CO2 Eq., respectively (10.6, 2.0,1.8, and 2.6 MMT C). Mineral and
organic soils also lost 16.1 and 2.4 MMT CO2 Eq. in 2021 (4.4 and 0.6 MMT C). The total net flux is 81.0 MMT CO2
Eq. in 2021 (22.1 MMT C), which is a 30 percent increase in CO2 emissions compared to the emissions in the initial
reporting year of 1990 (Table 6-129 and
Table 6-130). The main driver of net emissions for this source category is the conversion of forest land to
settlements, with large losses of biomass, deadwood and litter C.
101 NRI survey locations are classified according to land use histories starting in 1979, and consequently the classifications are
based on less than 20 years from 1990 to 2001. This may have led to an underestimation of Land Converted to Settlements in
the early part of the time series to the extent that some areas are converted to settlements from 1971 to 1978.
Land Use, Land-Use Change, and Forestry 6-181
-------
1 Table 6-129: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for
2 Land Converted to Settlements (MMT CO2 Eq.)
1990
2005
2017
2018
2019
2020
2021
Cropland Converted to
Settlements
3.4
9.8
6.0
5.9
5.9
5.9
5.9
Mineral Soils
2.8
8.4
5.2
5.2
5.1
5.1
5.1
Organic Soils
0.6
1.3
0.8
0.8
0.8
0.8
0.8
Forest Land Converted to
Settlements
53.4
59.0
63.5
63.7
63.8
63.7
63.7
Aboveground Live Biomass
32.5
35.3
38.1
38.3
38.3
38.3
38.3
Belowground Live Biomass
6.2
6.8
7.3
7.3
7.3
7.3
7.3
Dead Wood
5.4
5.9
6.4
6.4
6.4
6.4
6.4
Litter
8.0
8.7
9.5
9.5
9.5
9.5
9.5
Mineral Soils
1.1
2.0
1.9
1.9
1.9
1.9
1.9
Organic Soils
0.2
0.3
0.3
0.3
0.3
0.3
0.3
Grassland Converted to
Settlements
6.0
17.1
12.3
12.2
12.2
12.2
12.2
Aboveground Live Biomass
0.4
0.5
0.5
0.5
0.5
0.5
0.5
Belowground Live Biomass
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Dead Wood
0.1
0.1
0.2
0.2
0.2
0.2
0.2
Litter
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Mineral Soils
4.6
14.9
10.4
10.4
10.4
10.3
10.3
Organic Soils
0.6
1.4
0.9
0.9
0.9
0.9
0.9
Other Lands Converted to
Settlements
(0.4)
(1.4)
(1.2)
(1.2)
(1.2)
(1.2)
(1.2)
Mineral Soils
(0.4)
(1.6)
(1.3)
(1.3)
(1.3)
(1.3)
(1.3)
Organic Soils
+
0.2
0.1
0.1
0.1
0.1
0.1
Wetlands Converted to
Settlements
+
0.5
0.4
0.4
0.4
0.3
0.3
Mineral Soils
+
0.1
0.1
0.1
0.1
0.1
0.1
Organic Soils
+
0.4
0.3
0.3
0.3
0.3
0.3
Total Aboveground Biomass Flux
32.9
35.8
38.7
38.8
38.9
38.9
38.9
Total Belowground Biomass Flux
6.3
6.8
7.4
7.4
7.4
7.4
7.4
Total Dead Wood Flux
5.5
6.0
6.5
6.5
6.6
6.6
6.6
Total Litter Flux
8.2
8.9
9.7
9.7
9.7
9.7
9.7
Total Mineral Soil Flux
8.1
23.8
16.2
16.2
16.2
16.2
16.1
Total Organic Soil Flux
1.4
3.6
2.4
2.4
2.4
2.4
2.4
Total Net Flux
62.5
85.0
80.9
81.0
81.1
81.0
81.0
+ Absolute value does not exceed 0.05 MMT C02 Eq.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
4 Table 6-130: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for
5 Land Converted to Settlements (MMT C)
1990 2005 2017 2018 2019 2020 2021
Cropland Converted to
Settlements
0.9
2.7
1.6
1.6
1.6
1.6
1.6
Mineral Soils
0.8
2.3
1.4
1.4
1.4
1.4
1.4
Organic Soils
0.2
0.4
0.2
0.2
0.2
0.2
0.2
Forest Land Converted to
Settlements
14.6
16.1
17.3
17.4
17.4
17.4
17.4
Aboveground Live Biomass
8.9
9.6
10.4
10.4
10.5
10.5
10.5
Belowground Live Biomass
1.7
1.8
2.0
2.0
2.0
2.0
2.0
Dead Wood
1.5
1.6
1.7
1.7
1.7
1.7
1.7
6-182 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Litter
2.2
2.4
2.6
2.6
2.6
2.6
2.6
Mineral Soils
0.3
0.5
0.5
0.5
0.5
0.5
0.5
Organic Soils
+
0.1
0.1
0.1
0.1
0.1
0.1
Grassland Converted to
Settlements
1.6
4.7
3.3
3.3
3.3
3.3
3.3
Aboveground Live Biomass
0.1
0.1
0.1
0.1
0.1
0.1
10.0
Belowground Live Biomass
+
+
+
+
+
+
+
Dead Wood
+
+
+
+
+
+
+
Litter
+
0.1
0.1
0.1
0.1
0.1
0.1
Mineral Soils
1.3
4.1
2.8
2.8
2.8
2.8
2.8
Organic Soils
0.2
0.4
0.2
0.2
0.2
0.2
0.2
Other Lands Converted to
Settlements
(0.1)
(0.4)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
Mineral Soils
(0.1)
(0.4)
(0.4)
(0.4)
(0.3)
(0.3)
(0.3)
Organic Soils
+
+
+
+
+
+
+
Wetlands Converted to
Settlements
+
0.1
0.1
0.1
0.1
0.1
0.1
Mineral Soils
+
+
+
+
+
+
+
Organic Soils
+
0.1
0.1
0.1
0.1
0.1
0.1
Total Aboveground Biomass Flux
9.0
9.8
10.5
10.6
10.6
10.6
10.6
Total Belowground Biomass Flux
1.7
1.9
2.0
2.0
2.0
2.0
2.0
Total Dead Wood Flux
1.5
1.6
1.8
1.8
1.8
1.8
1.8
Total Litter Flux
2.2
2.4
2.6
2.6
2.6
2.6
2.6
Total Mineral Soil Flux
2.2
6.5
4.4
4.4
4.4
4.4
4.4
Total Organic Soil Flux
0.4
1.0
0.7
0.6
0.6
0.6
0.6
Total Net Flux
17.0
23.2
22.1
22.1
22.1
22.1
22.1
+ Absolute value does not exceed 0.05 MMT C.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
1 Methodology and Time-Series Consistency
2 The following section includes a description of the methodology used to estimate C stock changes for Land
3 Converted to Settlements, including (1) loss of aboveground and belowground biomass, dead wood and litter C
4 with conversion to settlements from forest lands and woodlands designated in the grassland, as well as (2) the
5 impact from all land-use conversions to settlements on soil organic C stocks in mineral and organic soils.
6 Biomass, Dead Wood, and Litter Carbon Stock Changes
7 A Tier 2 method is applied to estimate biomass, dead wood, and litter C stock changes for Forest Land Converted
8 to Settlements and woodlands associated with Grassland Converted to Settlements. Estimates are calculated in the
9 same way as those in the Forest Land Remaining Forest Land category using data from the USDA Forest Service,
10 Forest Inventory and Analysis (FIA) program (USDA Forest Service 2022), however there is no country-specific data
11 for settlements so the biomass, litter, and dead wood carbon stocks on these converted lands were assumed to be
12 zero. The difference between the stocks is reported as the stock change under the assumption that the change
13 occurred in the year of the conversion.
14 If FIA plots include data on individual trees, aboveground and belowground C density estimates are based on
15 Woodall et al. (2011). Aboveground and belowground biomass estimates also include live understory, which is a
16 minor component of biomass defined as all biomass of undergrowth plants in a forest, including woody shrubs and
17 trees less than 2.54 cm dbh. For this Inventory, it was assumed that 10 percent of total understory C mass is
18 belowground (Smith et al. 2006). Estimates of C density are based on information in Birdsey (1996) and biomass
19 estimates from Jenkins et al. (2003).
20 This inventory also includes estimates of change in dead organic matter for standing dead, deadwood and litter. If
21 FIA plots include data on standing dead trees, standing dead tree C density is estimated following the basic method
Land Use, Land-Use Change, and Forestry 6-183
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
applied to live trees (Woodall et al. 2011) with additional modifications to account for decay and structural loss
(Domke et al. 2011; Harmon et al. 2011). If FIA plots include data on downed dead wood, downed dead wood C
density is estimated based on measurements of a subset of FIA plots for downed dead wood (Domke et al. 2013;
Woodall and Monleon 2008). Downed dead wood is defined as pieces of dead wood greater than 7.5 cm diameter,
at transect intersection, that are not attached to live or standing dead trees. This includes stumps and roots of
harvested trees. To facilitate the downscaling of downed dead wood C estimates from the state-wide population
estimates to individual plots, downed dead wood models specific to regions and forest types within each region
are used. Litter C is the pool of organic C (also known as duff, humus, and fine woody debris) above the mineral
soil and includes woody fragments with diameters of up to 7.5 cm. A subset of FIA plots is measured for litter C. If
FIA plots include litter material, a modeling approach using litter C measurements from FIA plots is used to
estimate litter C density (Domke et al. 2016).
In order to ensure time-series consistency, the same methods are applied from 1990 to 2021 so that changes
reflect anthropogenic activity and not methodological adjustments. See Annex 3.13 for more information about
reference C density estimates for forest land and the compilation system used to estimate carbon stock changes
from forest land.
Soil Carbon Stock Changes
Soil organic C stock changes are estimated for Land Converted to Settlements according to land use histories
recorded in the 2015 USDA NRI survey for non-federal lands (USDA-NRCS 2018). Land use and some management
information were originally collected for each NRI survey location on a 5-year cycle beginning in 1982. In 1998, the
NRI program began collecting annual data, and the annual data have been incorporated from the NRI into the
inventory analysis through 2015 (USDA-NRCS 2018).
NRI survey locations are classified as Land Converted to Settlements in a given year between 1990 and 2015 if the
land use is settlements but had been classified as another use during the previous 20 years. NRI survey locations
are classified according to land use histories starting in 1979, and consequently the classifications are based on less
than 20 years from 1990 to 1998. This may have led to an underestimation of Land Converted to Settlements in
the early part of the time series to the extent that some areas are converted to settlement between 1971 and
1978. For federal lands, the land use history is derived from land cover changes in the National Land Cover Dataset
(Yang et al. 2018; Fry et al. 2011; Homer et al. 2007, 2015).
Mineral Soil Carbon Stock Changes
An IPCCTier 2 method (Ogle et al. 2003) is applied to estimate C stock changes for Land Converted to Settlements
on mineral soils from 1990 to 2015. Data on climate, soil types, land use, and land management activity are used
to classify land area and apply appropriate stock change factors (Ogle et al. 2003, 2006). Reference C stocks are
estimated using the National Soil Survey Characterization Database (USDA-NRCS 1997) with cultivated cropland as
the reference condition, rather than native vegetation as used in IPCC (2006). Soil measurements under
agricultural management are much more common and easily identified in the National Soil Survey Characterization
Database (USDA-NRCS 1997) than are soils under a native condition, and therefore cultivated cropland provide a
more robust sample for estimating the reference condition. Country-specific C stock change factors are derived
from published literature to determine the impact of management practices on soil organic C storage (Ogle et al.
2003, Ogle et al. 2006). However, there are insufficient data to estimate a set of land use, management, and input
factors for settlements. Moreover, the 2015 NRI survey data (USDA-NRCS 2018) do not provide the information
needed to assign different land use subcategories to settlements, such as turf grass and impervious surfaces, which
is needed to apply the Tier 1 factors from the IPCC guidelines (2006). Therefore, the United States has adopted a
land use factor of 0.7 to represent a net loss of soil organic C with conversion to settlements under the assumption
that there are additional soil organic C losses with land clearing, excavation and other activities associated with
development. More specific factor values can be derived in future Inventories as data become available. See Annex
3.12 for additional discussion of the Tier 2 methodology for mineral soils.
6-184 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 In order to ensure time-series consistency, the same methods are applied from 1990 to 2015 so that changes
2 reflect anthropogenic activity and not methodological adjustments. Soil organic C stock changes from 2016 to 2021
3 are estimated using a linear extrapolation method described in Box 6-4 of the Methodology section in Cropland
4 Remaining Cropland. The extrapolation is based on a linear regression model with moving-average (ARMA) errors
5 using the 1990 to 2015 emissions data, and is a standard data splicing method for imputing missing emissions data
6 in a time series (IPCC 2006). The Tier 2 method described previously will be applied to recalculate the 2016 to 2021
7 emissions in a future Inventory.
8 Organic Soil Carbon Stock Changes
9 Annual C emissions from drained organic soils in Land Converted to Settlements are estimated using the Tier 2
10 method provided in IPCC (2006). The Tier 2 method assumes that organic soils are losing C at a rate similar to
11 croplands, and therefore uses the country-specific values for cropland (Ogle et al. 2003). To estimate CO2
12 emissions from 1990 to 2015, the area of organic soils in Land Converted to Settlements is multiplied by the Tier 2
13 emission factor, which is 11.2 MT C per ha in cool temperate regions, 14.0 MT C per ha in warm temperate regions
14 and 14.3 MT C per ha in subtropical regions (See Annex 3.12 for more information).
15 In order to ensure time-series consistency, the same methods are applied from 1990 to 2015, and a linear
16 extrapolation method is used to approximate emissions for the remainder of the 2016 to 2021 time series (See Box
17 6-4 of the Methodology section in Cropland Remaining Cropland. The extrapolation is based on a linear regression
18 model with moving-average (ARMA) errors using the 1990 to 2015 emissions data, and is a standard data splicing
19 method for imputing missing emissions data in a time series (IPCC 2006). Estimates will be recalculated in future
20 Inventories when new NRI data are incorporated into the inventory.
21 Uncertainty
22 The uncertainty analysis for C losses with Forest Land Converted to Settlements is conducted in the same way as
23 the uncertainty assessment for forest ecosystem C flux in the Forest Land Remaining Forest Land category. Sample
24 and model-based error are combined using simple error propagation methods provided by the IPCC (2006), i.e., by
25 taking the square root of the sum of the squares of the standard deviations of the uncertain quantities. For
26 additional details, see the Uncertainty Analysis in Annex 3.13. The uncertainty analysis for mineral soil organic C
27 stock changes and annual C emission estimates from drained organic soils in Land Converted to Settlements is
28 estimated using a Monte Carlo approach, which is described in the Cropland Remaining Cropland section.
29 Uncertainty estimates are presented in Table 6-131 for each subsource (i.e., biomass C, dead wood, litter, soil
30 organic C in mineral soils and organic soils) and the method applied in the inventory analysis (i.e., Tier 2 and Tier
31 3). Uncertainty estimates from the Tier 2 and 3 approaches are combined using the simple error propagation
32 methods provided by the IPCC (2006), i.e., as described in the previous paragraph. There are also additional
33 uncertainties propagated through the analysis associated with the data splicing methods applied to estimate soil
34 organic C stock changes from 2016 to 2021. The combined uncertainty for total C stocks in Land Converted to
35 Settlements ranges from 34 percent below to 34 percent above the 2021 stock change estimate of 81.0 MMT CO2
36 Eq.
37 Table 6-131: Approach 2 Quantitative Uncertainty Estimates for Soil, Dead Organic Matter
38 and Biomass C Stock Changes occurring within Land Converted to Settlements (MMT CO2 Eq.
39 and Percent)
2021 Flux Estimate
Uncertainty Range Relative to Flux Estimate3
Source
(MMT CO? Eq.)
(MMT CO?
Eq.)
(%)
Lower
Upper
Lower Upper
Bound
Bound
Bound Bound
Cropland Converted to Settlements
5.9
1.8
10.0
-69% 69%
Mineral Soil C Stocks
5.1
1.1
9.2
-79% 79%
Organic Soil C Stocks
0.8
0.1
1.5
-90% 90%
Land Use, Land-Use Change, and Forestry 6-185
-------
Forest Land Converted to Settlements
63.7
38.4
89.0
-40%
40%
Aboveground Biomass C Stocks
38.3
14.5
62.2
-62%
62%
Belowground Biomass C Stocks
7.3
2.8
11.9
-62%
62%
Dead Wood
6.4
2.4
10.4
-62%
62%
Litter
9.5
3.6
15.4
-62%
62%
Mineral Soil C Stocks
1.9
1.2
2.5
-35%
35%
Organic Soil C Stocks
0.3
0.1
0.5
-74%
74%
Grassland Converted to Settlements
11.2
5.6
16.8
-50%
50%
Aboveground Biomass C Stocks
0.5
0.2
0.8
-65%
63%
Belowground Biomass C Stocks
0.1
+
0.1
-49%
54%
Dead Wood
0.2
0.1
0.3
-53%
65%
Litter
0.2
0.1
0.3
-65%
56%
Mineral Soil C Stocks
10.3
4.8
15.9
-54%
54%
Organic Soil C Stocks
0.9
+
1.7
-95%
95%
Other Lands Converted to Settlements
-1.2
(2.0)
(0.3)
-73%
73%
Mineral Soil C Stocks
-1.3
(2.1)
(0.4)
-66%
66%
Organic Soil C Stocks
0.1
(0.1)
0.3
-175%
175%
Wetlands Converted to Settlements
0.3
(0.2)
0.9
-157%
157%
Mineral Soil C Stocks
0.1
+
0.1
-110%
110%
Organic Soil C Stocks
0.3
(0.3)
0.8
-191%
191%
Total: Land Converted to Settlements
81.0
53.4
108.6
-34%
34%
Aboveground Biomass C Stocks
38.9
14.5
62.2
-62%
62%
Belowground Biomass C Stocks
7.4
2.8
11.9
-62%
62%
Dead Wood
6.6
2.4
10.4
-62%
62%
Litter
9.7
3.6
15.4
-62%
62%
Mineral Soil C Stocks
16.1
9.2
23.1
-43%
43%
Organic Soil C Stocks
2.4
(6.3)
11.0
-366%
366%
+ Does not exceed 0.05 MMT C02 Eq.
a Range of C stock change estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
Note: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
1 QA/QC and Verification
2 Quality control measures included checking input data, model scripts, and results to ensure data are properly
3 handled throughout the inventory process. Inventory reporting forms and text are reviewed and revised as needed
4 to correct transcription errors. No errors were found in this Inventory.
5 Recalculations Discussion
6 Recalculations are associated with new FIA data from 1990 to 2021 on biomass, dead wood and litter C stocks in
7 Forest Land Converted to Settlements and woodland conversion associated with Grassland Converted to
8 Settlements, and updated estimates for mineral and organic soils from 2016 to 2021 using the linear extrapolation
9 method. As a result, Land Converted to Settlements has an estimated larger C loss of 2.3 MMT CO2 Eq. on average
10 over the time series. This represents a 2.9 percent increase in C stock changes for Land Converted to Settlements
11 compared to the previous Inventory.
12 Planned Improvements
13 There are two key improvements planned for the inventory, including a) incorporating the latest land use data
14 from the USDA National Resources Inventory, and b) develop an inventory of mineral soil organic C stock changes
15 in Alaska and losses of C from drained organic soils in federal lands. These improvements will resolve most of the
16 differences between the managed land base for Land Converted to Settlements and amount of area currently
17 included in Land Converted to Settlements Inventory (See Table 6-113).
6-186 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 There are plans to improve classification of trees in settlements and to include transfer of biomass from forest land
2 to those areas in this category. There are also plans to extend the Inventory to included C losses associated with
3 drained organic soils in settlements occurring on federal lands.
4 These improvements will be made as funding and resources are available to expand the inventory for this source
5 category.
6 Table 6-132: Area of Managed Land in Land Converted to Settlements that is not included in
7 the current Inventory (Thousand Hectares)
Area (Thousand Hectares)
LCS Area
LCS Area Not
LCS Managed Land
Included in
Included in
Year
Area (Section 6.1)
Inventory
Inventory
1990
2,865
2,861
5
1991
3,213
3,238
-25
1992
3,575
3,592
-17
1993
4,147
4,107
40
1994
4,712
4,630
82
1995
5,271
5,161
110
1996
5,844
5,658
186
1997
6,421
6,174
247
1998
6,938
6,650
288
1999
7,451
7,116
336
2000
7,981
7,568
413
2001
8,386
7,947
439
2002
8,722
8,284
437
2003
8,738
8,335
403
2004
8,755
8,345
410
2005
8,765
8,341
425
2006
8,740
8,352
387
2007
8,722
8,295
427
2008
8,546
8,111
434
2009
8,351
7,930
420
2010
8,157
7,725
432
2011
7,953
7,498
455
2012
7,744
7,298
446
2013
7,342
6,932
410
2014
6,952
6,586
366
2015
6,542
6,165
377
2016
6,122
*
*
2017
5,720
*
*
2018
5,201
*
*
2019
4,690
*
*
2020
4,188
*
*
2021
3,781
*
*
8 NRI data have not been incorporated into the inventory after 2015, designated with asterisks (*).
Land Use, Land-Use Change, and Forestry 6-187
-------
1 6.12 Other Land Remaining Other Land (CRF
2 Category 4F1)
3 Land use is constantly occurring, and areas under a number of differing land-use types remain in their respective
4 land-use type each year, just as other land can remain as other land. While the magnitude of Other Land
5 Remaining Other Land is known (see Table 6-4), research is ongoing to track C pools in this land use. Until such
6 time that reliable and comprehensive estimates of C for Other Land Remaining Other Land can be produced, it is
7 not possible to estimate CO2, Cm or N2O fluxes on Other Land Remaining Other Land at this time.
s 6.13 Land Converted to Other Land (CRF
9 Category 4F2)
10 Land-use change is constantly occurring, and areas under a number of differing land-use types are converted to
11 other land each year, just as other land is converted to other uses. While the magnitude of these area changes is
12 known (see Table 6-4), research is ongoing to track C across Other Land Remaining Other Land and Land Converted
13 to Other Land. Until such time that reliable and comprehensive estimates of C across these land-use and land-use
14 change categories can be produced, it is not possible to separate CO2, Cm or N2O fluxes on Land Converted to
15 Other Land from fluxes on Other Land Remaining Other Land at this time.
6-188 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
7. Waste
Waste management and treatment activities are sources of greenhouse gas emissions (see Figure 7-1 and Figure
7-2). Landfills accounted for approximately 16.9 percent of total U.S. anthropogenic methane (Cm) emissions in
2021, the third largest contribution of any Cm source in the United States. Additionally, wastewater treatment and
discharge, composting of organic waste, and anaerobic digestion at biogas facilities accounted for approximately
2.9 percent, 0.4 percent, and less than 0.1 percent of U.S. CFU emissions, respectively. Nitrous oxide (N2O)
emissions resulted from the discharge of wastewater treatment effluents into aquatic environments were
estimated, the wastewater treatment process itself, and composting. Together, these waste activities account for
5.9 percent of total U.S. N2O emissions. Nitrogen oxides (NOx), carbon monoxide (CO), and non-CFU volatile organic
compounds (NMVOCs) are emitted by waste activities and are addressed separately at the end of this chapter. A
summary of greenhouse gas emissions from the Waste chapter is presented in Table 7-1 and Table 7-2. Overall, in
2021, waste activities generated emissions of 169.2 MMT CO2 Eq., or 2.7 percent of total U.S. greenhouse gas
Emissions from landfills contributed 72.5 percent of waste sector emissions in 2021 and are primarily comprised of
Cm emissions from municipal solid waste landfills (see Figure 7-1). Landfill emissions decreased by 2.2 MMT CO2
Eq. (1.7 percent) since 2020. Emissions from wastewater treatment were the second largest source of waste-
related emissions in 2021, accounting for 24.8 percent of sector emissions. The remaining two sources of
emissions, composting and anaerobic digestion at biogas facilities, account for 2.6 percent and 0.1 percent of
waste sector emissions in 2021, respectively.
Figure 7-1: 2021 Waste Sector Greenhouse Gas Sources
emissions.
Wastewater Treatment
Anaerobic Digestion at
Biogas Facilities
Composting
Landfills
123
0 10 20 30 40 50 60 70 80 90 100 110 120 130
MMTCCh Eq.
Waste 7-1
-------
l Figure 7-2: Trends in Waste Sector Greenhouse Gas Sources
¦ Anaerobic Digestion at Biogas Facilities
o-rHrNjm'a-mvor^ooCT^o-r-irMmTa-mvor^coCT'io-.-irMfnTa-mvoiv.coerio,-!
(TiCTia,icricriCTicria^cricrioooooooooo->-t->-i-^-i->-t->-i->-t-^H->-H-^H-^-ir>sjrNi
cridCTicricriC^cricricric^oooooooooooooooooooooo
¦THT-H-^H-nHT-i-nH-^Hi-H-^H-nHrsjrMrNjrMrvjfNjrMrsjrMrNjrMfNjrMrMrvjrMrvjfMrMrMrMrNj
2
3 Table 7-1: Emissions from Waste (MMT CO2 Eq.)
Gas/Source
1990
2005
2017
2018
2019
2020
2021
ch4
220.9
172.5
148.3
150.8
152.9
148.8
146.4
Landfills
197.8
147.7
123.9
126.7
129.0
124.8
122.6
Wastewater Treatment
22.7
22.7
21.5
21.4
21.2
21.3
21.1
Composting
0.4
2.1
2.7
2.5
2.5
2.6
2.6
Anaerobic Digestion at Biogas
Facilities
+
+
0.2
0.2
0.2
0.2
0.2
n2o
15.1
19.5
22.6
22.9
23.1
22.7
22.7
Wastewater Treatment
14.8
18.1
20.6
21.2
21.3
20.9
20.9
Composting
0.3
1.5
1.9
1.8
1.8
1.8
1.8
Total
236.0
192.1
170.9
173.7
176.0
171.5
169.2
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
4 Table 7-2: Emissions from Waste (kt)
Gas/Source
1990
2005
2017
2018
2019
2020
2021
ch4
7,889
6,161
5,297
5,384
5,460
5,315
5,230
Landfills
7,063
5,275
4,424
4,525
4,607
4,456
4,379
Wastewater Treatment
811
809
770
763
755
761
753
Composting
15
75
98
90
91
92
92
Anaerobic Digestion at Biogas
Facilities
1
2
6
6
6
6
6
n2o
57
74
85
87
87
86
86
Wastewater Treatment
56
68
78
80
80
79
79
Composting
1
6
7
7
7
7
7
Note: Totals by gas may not sum due to independent rounding.
7-2 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Carbon dioxide (CO2), CH4, and N2O emissions from the incineration of waste are accounted for in the Energy
sector rather than in the Waste sector because almost all incineration of municipal solid waste (MSW) in the
United States occurs at waste-to-energy facilities where useful energy is recovered. Similarly, the Energy sector
also includes an estimate of emissions from burning waste tires and hazardous industrial waste, because virtually
all of the combustion occurs in industrial and utility boilers that recover energy. The incineration of waste in the
United States in 2021 resulted in 12.8 MMT CO2 Eq. emissions, more than half of which is attributable to the
combustion of plastics. For more details on emissions from the incineration of waste, see Section 7.5. Greenhouse
gas precursor emissions from the waste sector are presented in Section 7.6.
Each year, some emission and sink estimates in the Inventory are recalculated and revised with improved methods
and/or data. In general, recalculations are made to the U.S. greenhouse gas emission estimates either to
incorporate new methodologies or, most commonly, to update recent historical data. These improvements are
implemented consistently across the previous Inventory's time series (i.e., 1990 to 2020) to ensure that the trend
is accurate. For the current Inventory, minor improvements were implemented beyond routine activity data
updates, including revising the industrial food waste disposal factor for estimating emissions from industrial
landfills. In total, the methodological and historic data improvements made to the Waste sector in this Inventory
resulted in an average increase in greenhouse gas emissions across the time series by 0.7 MMT CO2 Eq. (0.4
percent). In addition, estimates of CC>2-equivalent emissions totals of CFU and N2O have been revised to reflect the
100-year global warming potentials (GWPs) provided in the IPCC Fifth Assessment Report (AR5) (IPCC 2013)1. AR5
GWP values differ slightly from those presented in the IPCC Fourth Assessment Report (AR4) (IPCC 2007) (used in
the previous Inventories). For more information on specific methodological updates, please see the Recalculations
Discussion for each category in this chapter.
Due to lack of data availability, EPA is not able to estimate emissions associated with sludge generated from the
treatment of industrial wastewater or the amount of CH4 flared at composting sites. Emissions reported in the
Waste chapter for landfills, wastewater treatment, and anaerobic digestion at biogas facilities include those from
all 50 states, including Hawaii and Alaska, the District of Columbia, and U.S. Territories. Emissions from landfills
include modern, managed sites in most U.S. Territories except for outlying Pacific Islands. Emissions from domestic
wastewater treatment include most U.S. Territories except for outlying Pacific Islands. Those emissions are likely
insignificant as those outlying Pacific Islands (e.g., Baker Island) have no permanent population. No industrial
wastewater treatment emissions are estimated for U.S. Territories, due to lack of data availability. However,
industrial wastewater treatment emissions are not expected for outlying Pacific Islands and assumed to be small
for other U.S. Territories. Emissions for composting include all 50 states, including Hawaii and Alaska, and Puerto
Rico, but not the remaining U.S. Territories. Composting emissions from U.S. Territories are assumed to be small.
Similarly, EPA is not aware of any anerobic digestion at biogas facilities in U.S. Territories but will review this on an
ongoing basis to include these emissions if they are occurring. See Annex 5 for more information on EPA's
assessment of the sources not included in this Inventory.
Box 7-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals, including
Relationship to Greenhouse Gas Reporting Data
In following the United Nations Framework Convention on Climate Change (UNFCCC) requirement under Article
4.1 to develop and submit national greenhouse gas emission inventories, the emissions and removals presented
in this report and this chapter are organized by source and sink categories and calculated using internationally-
accepted methods provided by the Intergovernmental Panel on Climate Change (IPCC) in the 2006 IPCC
Guidelines for National Greenhouse Gas Inventories (2006 IPCC Guidelines) and its supplements and
refinements. Additionally, the calculated emissions and removals in a given year for the United States are
presented in a common format in line with the UNFCCC reporting guidelines for the reporting of inventories
under this international agreement. The use of consistent methods to calculate emissions and removals by all
1 As specified in UNFCCC reporting guidelines, the GWPs used are those listed in table 8.A.1 in Annex 8.A of Chapter 8 of the
Fifth Assessment Report of the Intergovernmental Panel on Climate Change, excluding the value for fossil methane.
Waste 7-3
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
nations providing their inventories to the UNFCCC ensures that these reports are comparable. The presentation
of emissions and sinks provided in the Waste chapter do not preclude alternative examinations, but rather, this
chapter presents emissions and removals in a common format consistent with how countries are to report
Inventories under the UNFCCC. The report itself, and this chapter, follows this standardized format, and
provides an explanation of the application of methods used to calculate emissions and removals from waste
management and treatment activities.
EPA also collects greenhouse gas emissions data from individual facilities and suppliers of certain fossil fuels and
industrial gases through its Greenhouse Gas Reporting Program (GHGRP). The GHGRP applies to direct
greenhouse gas emitters, fossil fuel suppliers, industrial gas suppliers, and facilities that inject CO2 underground
for sequestration or other reasons and requires reporting by sources or suppliers in 41 industrial categories.
Annual reporting is at the facility level, except for certain suppliers of fossil fuels and industrial greenhouse
gases. In general, the threshold for reporting is 25,000 metric tons or more of CO2 Eq. per year. See Annex 9
"Use of EPA Greenhouse Gas Reporting Program in Inventory" for more information.
Waste Data from EPA's Greenhouse Gas Reporting Program
EPA uses annual GHGRP facility-level data in the Landfills category to compile the national estimate of emissions
from Municipal Solid Waste (MSW) landfills (see Section 7.1 of this chapter for more information). EPA uses
directly reported GHGRP data for net CH4 emissions from MSW landfills for the years 2010 to 2021 of the
Inventory. MSW landfills subject to the GHGRP began collecting data in 2010. These data are also used to
recalculate emissions from MSW landfills for the years 2005 to 2009 to ensure time-series consistency.
7.1 Landfills (CRF Source Category 5A1)
In the United States, solid waste is managed by landfilling, recovery through recycling or composting, and
combustion through waste-to-energy facilities. Disposing of solid waste in modern, managed landfills is the most
used waste management technique in the United States. More information on how solid waste data are collected
and managed in the United States is provided in Box 7-3. The municipal solid waste (MSW) and industrial waste
landfills referred to in this section are all modern landfills that must comply with a variety of regulations as
discussed in Box 7-2. Disposing of waste in illegal dumping sites is not considered to have occurred in years later
than 1980 and these sites are not considered to contribute to net emissions in this section for the timeframe of
1990 to the current Inventory year. MSW landfills, or sanitary landfills, are sites where MSW is managed to prevent
or minimize health, safety, and environmental impacts. Waste is deposited in different cells and covered daily with
soil; many have environmental monitoring systems to track performance, collect leachate, and collect landfill gas.
Industrial waste landfills are constructed in a similar way as MSW landfills, but are used to dispose of industrial
solid waste, such as RCRA Subtitle D wastes (e.g., non-hazardous industrial solid waste defined in Title 40 of the
Code of Federal Regulations [CFR] in section 257.2), commercial solid wastes, or conditionally exempt small-
quantity generator wastes (EPA 2016a).
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 methane (CH4)
producing anaerobic bacteria convert the fermentation products into stabilized organic materials and biogas
consisting of approximately 50 percent biogenic carbon dioxide (CO2) and 50 percent CH4, by volume. Landfill
biogas also contains trace amounts of non-methane organic compounds (NMOC) and volatile organic compounds
(VOC) that either result from decomposition byproducts or volatilization of biodegradable wastes (EPA 2008).
7-4 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Box 7-2: Description of a Modern, Managed Landfill in the United States
Modern, managed landfills are well-engineered facilities that are located, designed, operated, and monitored to
ensure compliance with federal, state, and tribal regulations. A modern, managed landfill is EPA's interpretation
of the IPCC's terminology of a managed solid waste disposal site. Municipal solid waste (MSW) landfills must be
designed to protect the environment from contaminants which may be present in the solid waste stream.
Additionally, many new landfills collect and destroy landfill gas through flares or landfill gas-to-energy projects.
Requirements for affected MSW landfills may include:
• Siting requirements to protect sensitive areas (e.g., airports, floodplains, wetlands, fault areas, seismic
impact zones, and unstable areas);
• Design requirements for new landfills to ensure that Maximum Contaminant Levels (MCLs) will not be
exceeded in the uppermost aquifer (e.g., composite liners and leachate collection systems);
• Leachate collection and removal systems;
• Operating practices (e.g., daily and intermediate cover, receipt of regulated hazardous wastes, use of
landfill cover material, access options to prevent illegal dumping, use of a collection system to prevent
stormwater run-on/run-off, record-keeping);
• Air monitoring requirements (explosive gases);
• Groundwater monitoring requirements;
• Closure and post-closure care requirements (e.g., final cover construction); and
• Corrective action provisions.
Specific federal regulations that affected MSW landfills must comply with include the 40 CFR Part 258 (Subtitle
D of RCRA), or equivalent state regulations and the NSPS 40 CFR Part 60 Subpart WWW and XXX.2 Additionally,
state and tribal requirements may exist.
Methane and CO2 are the primary constituents of landfill gas generation and emissions. Net carbon dioxide flux
from carbon stock changes of materials of biogenic origin in landfills are estimated and reported under the Land
Use, Land-Use Change, and Forestry (LULUCF) sector (see Chapter 6 of this Inventory). Nitrous oxide (N2O)
emissions from the disposal and application of sewage sludge on landfills are also not explicitly modeled as part of
greenhouse gas emissions from landfills. Nitrous oxide emissions from sewage sludge applied to landfills as a daily
cover or for disposal are expected to be relatively small because the microbial environment in an anaerobic landfill
is not very conducive to the nitrification and denitrification processes that result in N2O emissions. Furthermore,
the 2006IPCC Guidelines did not include a methodology for estimating N2O emissions from solid waste disposal
sites "because they are not significant." Therefore, only CH4 generation and emissions are estimated for landfills
under the Waste sector.
Methane generation and emissions from landfills are a function of several factors, including: (1) the total amount
and composition of waste-in-place, which is the total waste landfilled annually over the operational lifetime of a
landfill; (2) the characteristics of the landfill receiving waste (e.g., size, climate, cover material); (3) the amount of
Cm that is recovered and either flared or used for energy purposes; and (4) the amount of CH4 oxidized as the
landfill gas - that is not collected by a gas collection system - passes through the cover material into the
atmosphere. Each landfill has unique characteristics, but all managed landfills employ similar operating practices,
including the application of a daily and intermediate cover material over the waste being disposed of in the landfill
to prevent odor and reduce risks to public health. Based on recent literature, the specific type of cover material
used can affect the rate of oxidation of landfill gas (RTI 2011). The most used cover materials are soil, clay, and
sand. Some states also permit the use of green waste, tarps, waste derived materials, sewage sludge or biosolids,
2 For more information regarding federal MSW landfill regulations, see
http://www.epa.gov/osw/nonhaz/municipal/landfill/msw regs.htm.
Waste 7-5
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
and contaminated soil as a daily cover. Methane production typically begins within the first year after the waste is
disposed of in a landfill and will continue for 10 to 50 or more years as the degradable waste decomposes over
time.
In 2021, landfill Cm emissions were approximately 122.6 MMT CO2 Eq. (4,379 kt), representing the third largest
source of CH4 emissions in the United States, behind enteric fermentation and natural gas systems. Emissions from
MSW landfills accounted for approximately 85 percent of total landfill emissions (103.7 MMT CO2 Eq.), while
industrial waste landfills accounted for the remainder (18.9 MMT CO2 Eq.). Nationally, there are significantly less
industrial waste landfills (hundreds) compared to MSW landfills (thousands), which contributes to the lower
national estimate of Cm emissions for industrial waste landfills. Additionally, the average organic content of waste
streams disposed in industrial waste landfills is lower than MSW landfills. Estimates of operational MSW landfills in
the United States have ranged from 1,700 to 2,000 facilities (EPA 2022a; EPA 2022b; EPA 2020c; Waste Business
Journal [WBJ] 2016; WBJ 2010). The Environment Research & Education Foundation (EREF) conducted a
nationwide analysis of MSW management and counted 1,540 operational MSW landfills in 2013 (EREF 2016).
Conversely, there are approximately 3,200 MSW landfills in the United States that have been closed since 1980 (for
which a closure data is known, (EPA 2022b; WBJ 2010). While the number of active MSW landfills has decreased
significantly over the past 20 years, from approximately 6,326 in 1990 to as few as 1,540 in 2013, the average
landfill size has increased (EPA 2022a; EREF 2016; BioCycle 2010). Larger landfills may have deeper cells where a
greater amount of area will be anaerobic (more CH4 is generated in anaerobic versus aerobic areas) and larger
landfills tend to generate more Cm compared to a smaller landfill (assuming the same waste composition and age
of waste). Regarding industrial waste landfills, the WBJ database includes approximately 1,200 landfills accepting
industrial and/or construction and demolition debris for 2016 (WBJ 2016). Only 169 facilities with industrial waste
landfills met the reporting threshold under Subpart TT (Industrial Waste Landfills) in the first year (2011) of EPA's
Greenhouse Gas Reporting Program for this subpart (GHGRP codified in 40 CFR part 98), indicating that there may
be several hundred industrial waste landfills that are not required to report under EPA's GHGRP. Less industrial
waste landfills meet the GHGRP eligibility threshold because they typically accept waste streams with low to no
organic content, which will not decompose and generate CH4 when disposed.
The annual amount of MSW generated and subsequently disposed in MSW landfills varies annually and depends
on several factors (e.g., the economy, consumer patterns, recycling and composting programs, inclusion in a
garbage collection service). The estimated annual quantity of waste placed in MSW landfills increased 10 percent
from approximately 205 MMT in 1990 to 226 MMT in 2000, then decreased by 11 percent to 202 MMT in 2010,
and then increased by 7 percent to approximately 216 MMT in 2021 (see Annex 3.14, Table A-220). Emissions
decreased between 1990 to 2021 largely because of increased use of landfill gas collection and control systems,
closure of older landfills, better management practices, and increased diversion of organics through state and local
policy and regulations. The total amount of MSW generated is expected to increase as the U.S. population
continues to grow. The impacts of the coronavirus (COVID-19) pandemic with respect to landfilled waste cannot be
quantified as data sources such as the EPA's Advancing Sustainable Materials Management: Facts and Figures
report have not been published for 2019 through 2021. The quantities of waste landfilled for 2014 to 2021
(presented in Annex 3.14) are extrapolated based on population growth and the last national assessment of MSW
landfilled from 2013 (EREF 2016). Net CH4 emissions from MSW landfills have decreased since 1990 (see Table 7-3
and Table 7-4).
The estimated quantity of waste placed in industrial waste landfills (from the pulp and paper, and food processing
sectors) has remained relatively steady since 1990, ranging from 9.7 MMT in 1990 to 11.2 MMT in 2021 (see Annex
3.14, Table A-219). CH4 emissions from industrial waste landfills have also remained at similar levels recently,
ranging from 16.1 MMT CO2 Eq. in 2005 to 18.9 MMT CO2 Eq. in 2021 when accounting for both CH4 generation
and oxidation. The EPA has focused the industrial waste landfills source category on industrial sectors known to
generate and dispose of by-products that are organic and contribute to CH4 generation, which are the pulp and
paper and food processing sectors. Construction and demolition (C&D) landfills, another type of industrial waste
landfill, may accept waste that could degrade (e.g., treated wood), but these waste streams are unlikely to
generate significant amounts of CH4 and are therefore not as relevant to the purpose of national greenhouse gas
emissions estimate. There is also a general lack of data on annual quantities of waste disposed in industrial waste
7-6 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 landfills and the GHGRP Subpart TT (Industrial Waste Landfills) dataset has confirmed C&D landfills, for example,
2 are insignificant CH4 generators.
3 EPA's Landfill Methane Outreach Program (LMOP) collects information on landfill gas energy projects currently
4 operational or under construction throughout the United States. LMOP's Landfill and Landfill Gas Energy Database
5 contains certain information on the gas collection and control systems in place at landfills provided by
6 organizations that are a part of the program, which can include the amount of landfill gas collected and flared. In
7 2021, LMOP identified 7 new landfill gas-to-energy (LFGE) projects (EPA 2022b) that began operation.
8 Landfill gas collection and control is not accounted for at industrial waste landfills in this chapter (see the
9 Methodology discussion for more information).
10 Table 7-3: ChU Emissions from Landfills (MMT CO2 Eq.)
Activity
1990
2005
2017
2018
2019
2020
2021
MSW CH4 Generation3
230.0
303.7
327.0
332.6
341.4
342.2
334.8
Industrial CH4 Generation
13.6
17.9
20.4
20.6
20.7
20.9
21.0
MSW CH4 Recovered3
(23.8)
(148.4)
(192.9)
(195.2)
(201.4)
(206.3)
(201.5)
MSW CH4 Oxidized3
(20.6)
(23.6)
(28.6)
(29.2)
(29.6)
(29.9)
(29.6)
Industrial CH4 Oxidized
(1.4)
(1.8)
(2.0)
(2.1)
(2.1)
(2.1)
(2.1)
MSW net CH4 Emissions
185.5
131.6
105.5
108.2
110.4
106.0
103.7
Industrial CH4 Emissions'5
12.2
16.1
18.4
18.5
18.6
18.8
18.9
Total
197.8
147.7
123.9
126.7
129.0
124.8
122.6
a For years 1990 to 2004, the Inventory methodology for MSW landfills uses the first order decay methodology. A
methodological change occurs in year 2005. For years 2005 to 2021, directly reported net CH4 emissions from the
GHGRP data plus a scale-up factor are used to account for emissions from landfill facilities that are not subject to the
GHGRP. More details on the scale-up factor and how it was developed can be found in Annex 3.14. These data
incorporate CH4 recovered and oxidized for MSW landfills. As such, CH4 generation, CH4 oxidation, and CH4 recovery
are not calculated separately and totaled to net CH4 emissions. See the Methodology and Time-Series Consistency
section of this chapter for more information.
b Methane recovery is not calculated for industrial landfills because this is not a common practice in the United States.
Only 1 landfill of 167 that report to Subpart TT (Industrial Waste Landfills) of the GHGRP had an active gas collection
and control system during the year 2021 (EPA 2022a).
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values.
11 Table 7-4: ChU Emissions from Landfills (kt)
Activity
1990
2005
2017
2018
2019
2020
2021
MSW CH4 Generation3
8,214
10,845
11,680
11,878
12,193
12,222
11,958
Industrial CH4 Generation
484
638
729
734
739
745
750
MSW CH4 Recovered3
(851)
(5,301)
(6,891)
(6,970)
(7,193)
(7,367)
(7,195)
MSW CH4 Oxidized3
(736)
(843)
(1,021)
(1,044)
(1,058)
(1,069)
(1,059)
Industrial CH4 Oxidized
(48)
(64)
(73)
(73)
(74)
(75)
(75)
MSW net CH4 Emissions
6,627
4,701
3,768
3,864
3,942
3,786
3,704
Industrial net CH4 Emissions'5
436
575
656
661
665
671
675
Total
7,063
5,275
4,424
4,525
4,607
4,456
4,379
a For years 1990 to 2004, the Inventory methodology for MSW landfills uses the first order decay methodology. A
methodological change occurs in year 2005. For years 2005 to 2021, directly reported net CH4 emissions from the
GHGRP data plus a scale-up factor are used to account for emissions from landfill facilities that are not subject to
the GHGRP. More details on the scale-up factor and how it was developed can be found in Annex 3.14. These data
incorporate CH4 recovered and oxidized for MSW landfills. As such, CH4 generation, CH4 oxidation, and CH4 recovery
are not calculated separately and totaled to net CH4 emissions. See the Methodology and Time-Series Consistency
section of this chapter for more information.
Waste 7-7
-------
b Methane recovery is not calculated for industrial landfills because this is not a common practice in the United
States. Only 1 landfill of 167 that report to Subpart TT (Industrial Waste Landfills) of the GHGRP had an active gas
collection and control system during the year 2021 (EPA 2022a).
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values.
1 Methodology and Time-Series Consistency
2 Methodology Applied for MSW Landfills
3 A combination of IPCC Tier 2 and 3 approaches (IPCC 2006) are used over the reported timeseries to calculate
4 emissions from MSW Landfills, using two primary methods. The first method uses the first order decay (FOD)
5 model as described by the 2006 IPCC Guidelines to estimate Cm generation. The amount of Cm recovered and
6 combusted from MSW landfills is subtracted from the Cm generation and is then adjusted with an oxidation
7 factor. The oxidation factor represents the amount of Cm in a landfill that is oxidized to CO2 as it passes through
8 the landfill cover (e.g., soil, clay, geomembrane). This method is presented below and is similar to Equation HH-6 in
9 40 CFR Part 98.343 for MSW landfills, and Equation TT-6 in 40 CFR Part 98.463 for industrial waste landfills.
10 Equation 7-1: Landfill Methane Generation
11 CH4.MSW = (GCH4 - Yl-1 Rn) * (1 - OX)
12 where,
13 Cm,msw = Net Cm emissions from solid waste
14 Gch4,msw = Cm generation from MSW landfills, using emission factors for DOC, k, MCF, F from IPCC
15 (2006) and other peer-reviewed sources
16 R = Cm recovered and combusted
17 Ox = Cm oxidized from MSW landfills before release to the atmosphere, using Ox values from
18 IPCC (2006) and other peer-reviewed or scientifically validated literature (40 CFR Part 98)
19 The second method used to calculate Cm emissions from landfills, also called the back-calculation method, is
20 based on directly measured amounts of recovered Cm from the landfill gas and is expressed below and by
21 Equation HH-8 in 40 CFR Part 98.343. The two parts of the equation consider the portion of Cl-Uin the landfill gas
22 that is not collected by the landfill gas collection system, and the portion that is collected. First, the recovered CH4
23 is adjusted with the collection efficiency of the gas collection and control system and the fraction of hours the
24 recovery system operated in the calendar year. This quantity represents the amount of Cm in the landfill gas that is
25 not captured by the collection system; this amount is then adjusted for oxidation. The second portion of the
26 equation adjusts the portion of Cm in the collected landfill gas with the efficiency of the destruction device(s), and
27 the fraction of hours the destruction device(s) operated during the year.
28 The current Inventory uses both methods to estimate Cm emissions across the time series within EPA's Waste
29 Model, as summarized in Figure 7-3 below. This chapter provides a summary of the methods, activity data, and
30 parameters used. Additional step-wise explanations to generate the net emissions are provided in Annex 3.14.
31 Equation 7-2: Net Methane Emissions from MSW Landfills
32 cm,Solid Waste = [( ^ r) x(l - OX) + R x (l - CDE X fDest))]
\CE X /REC J
33 where,
34 cm ,solid waste — Net CH4 emissions from solid waste
35 R = Quantity of recovered CH4 from Equation HH-4 of EPA's GHGRP
36 CE = Collection efficiency estimated at the landfill, considering system coverage, operation, and
37 cover system materials from Table HH-3 of EPA's GHGRP. If area by soil cover type
38 information is not available, the default value of 0.75 should be used (percent)
39 fREc = fraction of hours the recovery system was operating (percent)
7-8 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
OX = oxidation factor (percent)
DE = destruction efficiency (percent)
foest = fraction of hours the destruction device was operating (fraction)
Figure 7-3; Methodologies Used Across the Time Series to Compile the U.S. Inventory of
Emission Estimates for MSW Landfills
1990 - 2004
2005 - 2009
2010 - 2016
2017 - Present
U.S.-specific first-order
decay (FOD) model
Back-casted EPA
GHGRP reported net
methane emissions
EPA GHGRP
reported net
methane emissions
EPA GHGRP
reported net
methane emissions
Annex Steps 1-3
Annex Step 4
Annex Step 5
Annex Step 6
IPCC 2006 Emission Factors:
• DOC = 0.20
• MCF = 1
• DOC, = 0.5
• OX = 0.10
• DE = 0.99
Activity Data:
• National waste generation
data multiplied by the
national disposal factor
Back-casted GHGRP
emissions plus a 9%
scale-up factor1'2
Recovery calculated from
four CH4 recovery
databases
• Back-calculated CH4
generation3
• Weighted average
oxidation factor based on
GHGRP data3
• Net GHGRP emissions
plus a 9% scale-up
factor2
• GHGRP CH4 recovery plus
a 9% scale-up factor
• Back-calculated CH4
generation3
• Weighted average
oxidation factor based on
GHGRP data 3
• Net GHGRP emissions
plus an 11% scale-up
factor2
• GHGRP CH4 recovery plus
an 11% scale-up factor
• Back-calculated CH4
generation3
• Weighted average
oxidation factor based on
GHGRP data 3
1 The intent of the scale-up factor is to estimate emissions from landfills that do not report to the GHGRP. More details on
the scale-up factor and how it was developed can be found in Annex 3.14. The back-casted emissions are calculated using
directly reported net methane emissions for GHGRP reporting years 2010 to 2016. The back-casted emissions are subject
to change in each Inventory based on new reporting year reports and resubmitted greenhouse gas reports for previous
years. This method is compatible with the 2006IPCC Guidelines because facilities reporting to the GHGRP either use the
FOD method, or directly measured methane recovery data with default emission factors either directly included in the
2006 IPCC Guidelines or scientifically validated through peer review.
2 Emission factors used by facilities reporting to GHGRP Subpart HH are facility-specific defaults derived from peer-reviewed
literature and the 2006 IPCC Guidelines.
3 Methane generation is back-calculated from the net MSW emissions, estimated methane recovery data, and the weighted
average oxidation factor based on GHGRP Subpart HH reported data of 0.18 between 2010 to 2016, and 0.21 between
2017 to 2020, and 0.22 in 2021.
The Waste Model is a spreadsheet developed by the IPCC for purposes of estimating methane emissions from solid
waste disposal sites, adapted to the United States by the inclusion and usage of U.S.-specific parameters. The
Waste Model contains activity and waste generation information from both the MSW and Industrial landfill sectors
and estimates the amount of CH4 emissions from each sector for each year of the time series, using both methods.
Prior to the 1990 through 2015 Inventory, only the FOD method was used. Methodological changes were made to
the 1990 through 2015 Inventory to incorporate higher tier data (i.e., CH4 emissions as directly reported to EPA's
GHGRP), which cannot be directly applied to earlier years in the time series without significant bias. The technique
used to merge the directly reported GHGRP data with the previous methodology is described as the overlap
technique in the Time-Series Consistency chapter of the 2006 IPCC Guidelines. Additional details on the technique
used is included in Annex 3.14, and a technical memorandum (RTI 2017).
A summary of the methodology used to generate the current 1990 to 2021 Inventory estimates for MSW landfills
is as follows and is also illustrated in Annex Figure A-19:
Waste 7-9
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
• 1940 to 1989: These years are included for historical waste disposal amounts. Estimates of the annual
quantity of waste landfilled for 1960 through 1988 were obtained from EPA's Anthropogenic Methane
Emissions in the United States, Estimates for 1990: Report to Congress (EPA 1993) and an extensive
landfill survey by the EPA's Office of Solid Waste in 1986 (EPA 1988). Although waste placed in landfills in
the 1940s and 1950s contributes very little to current Cm generation, estimates for those years were
included in the FOD model for completeness in accounting for Cm generation rates and are based on the
population in those years and the per capita rate for land disposal for the 1960s. For the Inventory
calculations, wastes landfilled prior to 1980 were broken into two groups: wastes disposed in managed,
anaerobic landfills (Methane Conversion Factor, MCF, of 1) and those disposed in uncategorized solid
waste disposal waste sites (MCF of 0.6) (IPCC 2006). Uncategorized sites represent those sites for which
limited information is known about the management practices. All calculations after 1980 assume waste
is disposed in managed, anaerobic landfills. The FOD method was applied to estimate annual Cm
generation. Methane recovery amounts were then subtracted, and the result was then adjusted with a 10
percent oxidation factor to derive the net emissions estimates. A detailed explanation of the methods
used are presented in Annex 3.14 Step 1.
• 1990 to 2004: The Inventory time series begins in 1990. The FOD method is exclusively used for this group
of years. The national total of waste generated (based on state-specific landfill waste generation data)
and a national average disposal factor for 1989 through 2004 were obtained from the State of Garbage
(SOG) survey every two years (i.e., 2002, 2004 as published in BioCycle 2006). In-between years were
interpolated based on population growth. For years 1989 to 2000, directly reported total MSW generation
data were used; for other years, the estimated MSW generation (excluding construction and demolition
waste and inerts) were presented in the reports and used in the Inventory. The FOD method was applied
to estimate annual CFU generation. Landfill-specific CFU recovery amounts (calculated from four CFU
recovery databases) were then subtracted from Cm generation and the result was adjusted with a 10
percent oxidation factor to derive the net emissions estimates. A detailed explanation of the methods
used are presented in Annex 3.14 Steps 1 through 3.
• 2005 to 2009: Emissions for these years are estimated using net CFU emissions that are reported by
landfill facilities under EPA's GHGRP. Because not all landfills in the United States are required to report to
EPA's GHGRP, a 9 percent scale-up factor is applied to the GHGRP emissions for completeness. The intent
of the scale-up factor is to account for emissions from landfills that do not report to the GHGRP.
Supporting information, including details on the technique used to estimate emissions for 2005 to 2009,
to develop the scale-up factor, and to ensure time-series consistency by incorporating the directly
reported GHGRP emissions is presented in Annex 3.14 Step 4 and in RTI 2018a. Separate estimates of CH4
generation, CH4 recovery, and oxidation are calculated from the net CH4 emissions. Landfill-specific CH4
recovery is calculated from four CH4 recovery databases. A single oxidation factor is not applied to the
annual CH4 generated as is done for 1990 to 2004 because the GHGRP emissions data are used, which
already take oxidation into account. The GHGRP allows facilities to use varying oxidation factors (i.e., 0,
10, 25, or 35 percent) depending on their facility-specific calculated CH4 flux rate. The effectively applied
average oxidation factor between 2005 to 2009 averages to 0.14. Methane generation is then back-
calculated using net CH4 emissions, CH4 recovery, and oxidation. A detailed explanation of the methods
used to develop the back-casted emissions and revised scale-up factor are presented in Annex 3.14 Step
4.
• 2010 to 2016: Net CH4 emissions as directly reported to the GHGRP are used with a 9 percent scale-up
factor to account for landfills that are not required to report to the GHGRP. A combination of the FOD
method and the back-calculated CH4 emissions were used by the facilities reporting to the GHGRP.
Landfills reporting to the GHGRP without gas collection and control apply the FOD method, while most
landfills with landfill gas collection and control apply the back-calculation method. Methane recovery is
calculated using reported GHGRP recovery data plus a 9 percent scale-up factor. Methane generation and
oxidation are back-calculated from the net GHGRP CH4 emissions applied and estimated CH4 recovery. The
7-10 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
average oxidation factor effectively applied is 0.18 percent. A detailed explanation of the methods used to
develop the revised scale-up factor are presented in Annex 3.14 Step 5.
• 2017 to 2021: The same methodology is applied as for 2010 through 2016 where a scale-up factor is
applied to account for landfills that are not required to report to the GHGRP. The scale-up factor was
revised for the 1990 to 2020 Inventory to change the methodology from total waste-in-place to only
considering waste disposed for non-reporting landfills in the past 50 years (i.e., since 1970). Additional
revisions made included incorporating facilities that have stopped reporting to the GHGRP, new additions
to the 2021 LMOP Database (EPA 2022b), corrections to the underlying database of non-reporting landfills
used to develop the 9 percent scale-up factor that were identified. For 2017 to 2021, a scale-up factor of
11 percent is applied annually to the GHGRP net reported CH4 emissions. Methane recovery is calculated
using reported GHGRP recovery data plus an 11 percent scale-up factor. Separate estimates of CH4
generation and oxidation are calculated from the net CH4 emissions applied. The average oxidation factor
effectively applied is 0.22 percent. A detailed explanation of the methods used to develop the revised
scale-up factor are presented in Annex 3.14 Step 6.
With regard to the time series and as stated in 2006IPCC Guidelines Volume 1: Chapter 5 Time-Series Consistency
(IPCC 2006), "the time series is a central component of the greenhouse gas inventory because it provides
information on historical emissions trends and tracks the effects of strategies to reduce emissions at the national
level. All emissions in a time series should be estimated consistently, which means that as far as possible, the time
series should be calculated using the same method and data sources in all years" (IPCC 2006). In some cases, it
may not be possible to use the same methods and consistent data sets for all years because of limited data
(activity data, emission factors, or other parameters) directly used in the calculation of emission estimates for
some historical years. In such cases, emissions or removals may need to be recalculated using alternative methods.
In this case, this chapter provides guidance on techniques to splice, or join methodologies together instead of
back-casting emissions back to 1990. One of those techniques is referred to as the overlap technique. The overlap
technique is recommended when new data becomes available for multiple years. This was the case with EPA's
GHGRP data for MSW landfills, where directly reported CH4 emissions data became available for more than 1,200
MSW landfills beginning in 2010. The GHGRP emissions data had to be merged with emissions from the FOD
method to avoid a drastic change in emissions in 2010, when the datasets were combined. EPA also had to
consider that according to IPCC's good practice, efforts should be made to reduce uncertainty in Inventory
calculations and that, when compared to the GHGRP data, the FOD method presents greater uncertainty.
In evaluating the best way to combine the two datasets, EPA considered either using the FOD method from 1990
to 2009, or using the FOD method for a portion of that time and back-casting the GHGRP emissions data to a year
where emissions from the two methodologies aligned. Plotting the back-casted GHGRP emissions against the
emissions estimates from the FOD method showed an alignment of the data in 2004 and later years which
facilitated the use of the overlap technique while also reducing uncertainty. A detailed explanation and a chart
showing the estimates across the time series considering the two method options is included in Annex 3.14. EPA
ultimately decided to back-cast the GHGRP emissions from 2009 to 2005 only, to merge the datasets and adhere to
the IPCC Good Practice Guidance for ensuring time-series consistency.
Supporting information, including details on the techniques used to ensure time-series consistency by
incorporating the directly-reported GHGRP emissions is presented in Annex 3.14.
Methodology Applied for Industrial Waste Landfills
Emissions from industrial waste landfills are estimated using a Tier 2 approach (IPCC 2006) and a tailored (country-
specific) IPCC waste model. Activity data used are industrial production data (ERG 2021) for two sectors (pulp and
paper manufacturing, and food and beverage manufacturing) to which country-specific default waste disposal
factors are applied (a separate disposal factor for each sector). The disposal factors, as described below, are based
on scientifically reviewed data, and are the same across the entire time series. The emission factors are based on
those recommended by the 2006 IPCC Guidelines and are the same across the entire time series.
Waste 7-11
-------
1 The FOD equation from IPCC 2006 is used via the waste model to estimate methane emissions:
2 Equation 7-3: Net Methane Emissions from Industrial Waste Landfills
3
CH4.IND = (GCH4 - * (1 - OX)
4 where,
5
6
7
8
9
10
R
OX
CH4 ,Solid Waste —
GcH4,lnd =
Net Cm emissions from solid waste
Cm generation from industrial waste landfills, using production data multiplied by a
disposal factor and emission factors for DOC, k, MCF, F (IPCC 2006)
Cm recovered and combusted (no recovery is assumed for industrial waste landfills)
Cm oxidized from industrial waste landfills before release to the atmosphere (using the
2006 IPCC Guidelines value for OX of 0.10)
11 The activity data used in the emission calculations are production data (e.g., the amount of meat, poultry,
12 vegetables processed; the amount of paper produced) versus disposal data. There are currently no facility-specific
13 data sources that track and report the amount and type of waste disposed of in the universe of industrial waste
14 landfills in the United States. EPA's GHGRP provides some insight into waste disposal in industrial waste landfills
15 but is not comprehensive. Data reported to the GHGRP on industrial waste landfills suggests that most of the
16 organic waste which would result in methane emissions is disposed at pulp and paper and food processing
17 facilities. Of the 168 facilities that reported to Subpart TT of the GHGRP in 2019, 92 (54 percent) are in the North
18 American Industrial Classification System (NAICS) for Pulp, Paper, and Wood Products (NAICS 321 and 322) and 12
19 (7 percent) are in Food Manufacturing (NAICS 311).
20 Based on this limited information, the Inventory methodology assumes most of the organic waste placed in
21 industrial waste landfills originates from the food processing (meat, vegetables, fruits) and pulp and paper sectors,
22 thus estimates of industrial landfill emissions focused on these two sectors. EPA validated this assumption through
23 an analysis of the Subpart TT of the GHGRP in the 2016 reporting year (RTI 2018b). The Subpart TT waste disposal
24 information for pulp and paper facilities correlates well with the activity data currently used to estimate Inventory
25 emissions; however, the waste disposal information in Subpart TT related to food and beverage facilities are
26 approximately an order of magnitude different than the Inventory disposal estimates for the entire time series.
27 EPA conducted a literature review between 2020 and 2022 to investigate other sources of industrial food waste
28 and annual waste disposal quantities. As a result of this effort, EPA decided to revise the food waste disposal factor
29 in the 1990 to 2021 Inventory for select years. A waste disposal factor of 4.86 percent is used for 1990 to 2009 and
30 a revised factor of 6 percent is used for 2010 to the current year. The 6 percent waste disposal factor is derived
31 from recent surveys of the food and beverage industry where approximately 94 percent of food waste generated is
32 repurposed (FWRA 2016). The 4.86% disposal factor is based on available data from a 1993 Report to Congress
33 (EPA 1993).
34 The composition of waste disposed of in industrial waste landfills is expected to be more consistent in terms of
35 composition and quantity than that disposed of in MSW landfills. The amount of waste landfilled is assumed to be
36 a fraction of production that is held constant over the time series as explained in Annex 3.14.
37 Landfill Cm recovery is not accounted for in industrial waste landfills and is believed to be minimal based on
38 available data collected under EPA's GHGRP for industrial waste landfills (Subpart TT), which shows that only one
39 of the 167 facilities, or 1 percent of facilities, have active gas collection systems (EPA 2022a). However, because
40 EPA's GHGRP is not a national database and comprehensive data regarding gas collection systems have not been
41 published for industrial waste landfills, assumptions regarding a percentage of landfill gas collection systems, or a
42 total annual amount of landfill gas collected for the non-reporting industrial waste landfills have not been made for
43 the Inventory methodology.
44 The amount of Cm oxidized by the landfill cover at industrial waste landfills was assumed to be 10 percent of the
45 Cm generated (IPCC 2006; Mancinelli and McKay 1985; Czepiel et al. 1996) for all years.
7-12 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
Box 7-3: Nationwide Municipal Solid Waste Data Sources
Municipal solid waste (MSW) generated in the United States can be managed through a variety of methods.
MSW that is not recycled, composted, combusted with energy recovery, or digested is assumed to be landfilled.
In addition to these management pathways, waste or excess food from the food manufacturing and processing
sector may be disposed through the sewerage network, used for animal feed, land application, donated for
human consumption, and rendered or recycled into biofuels in the case of animal by-products, fats, oils and
greases.
There have been three main sources for nationwide solid waste management data in the United States that the
Inventory has used (see Annex 3.14, Box A-3 for comparison of estimates from these data sources):
• The BioCycle and Earth Engineering Center of Columbia University's SOG in America surveys [no longer
published];
• The EPA's Advancing Sustainable Materials Management: Facts and Figures reports; and
• The EREF's MSW Generation in the United States reports.
The SOG surveys and, most recently EREF, collected state-reported data on the amount of waste generated and
the amount of waste managed via different management options: landfilling, recycling, composting, and
combustion. These data sources used a 'bottom-up' method. The survey asked for actual tonnages instead of
percentages in each waste category (e.g., residential, commercial, industrial, construction and demolition,
organics, tires) for each waste management option. If such a breakdown was not available, the survey asked for
total tons landfilled. The data were adjusted for imports and exports across state lines so that the principles of
mass balance were adhered to for completeness, whereby the amount of waste managed did not exceed the
amount of waste generated. The SOG and EREF reports present survey data aggregated to the state level.
The EPA Advancing Sustainable Materials Management: Facts and Figures report characterizes national post-
consumer municipal solid waste (MSW) generation and management using a top-down materials flow (mass
balance) methodology. It captures an annual snapshot of MSW generation and management in the United
States for specific products. Data are gathered from U.S. Government (e.g., U.S. Census Bureau and U.S.
Department of Commerce), state environmental agencies, industry and trade groups, and sampling studies. The
materials flow methodology develops MSW waste generation estimates of quantities of MSW products in the
marketplace (using product sales and replacement data) and assessing waste generation by component material
based on product lifespans. The data are used to estimate tons of materials and products generated, recycled,
combusted with energy recovery, managed via other food waste management pathways, or landfilled
nationwide. MSW that is not recycled or composted is assumed to be combusted or landfilled, except for
wasted food, which uses a different methodology and includes nine different management pathways. The 2018
Facts and Figures Report (EPA 2020) uses a methodology that expanded the number of management pathways
to include: animal feed; bio-based materials/biochemical processing (i.e., rendering); co-digestion/anaerobic
digestion; composting/aerobic processes; combustion; donation; land application; landfill; and
sewer/wastewater treatment.
In this Inventory, emissions from solid waste management are presented separately by waste management
option, except for recycling of waste materials. Emissions from recycling are attributed to the stationary
combustion of fossil fuels that may be used to power on-site recycling machinery and are presented in the
stationary combustion chapter in the Energy sector, although the emissions estimates are not called out
separately. Emissions from solid waste disposal in landfills and the composting of solid waste materials are
presented in the Landfills and Composting sections in the Waste sector of this report. Emissions from anaerobic
digesters are presented in three different sections depending on the digester category. Emissions from on-farm
digesters are included in the Agriculture sector; emissions from digesters at wastewater treatment plants
emissions from stand-alone digesters are presented in separate sections in the Waste sector of this report. In
the United States, almost all incineration of MSW occurs at waste-to-energy (WTE) facilities or industrial
Waste 7-13
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
facilities where useful energy is recovered, and thus emissions from waste incineration are accounted for in the
Incineration chapter of the Energy sector of this report.
Uncertainty
Several types of uncertainty are associated with the estimates of Cm emissions from MSW and industrial waste
landfills when the FOD method is applied directly for 1990 to 2004 in the Waste Model and, to some extent, in the
GHGRP methodology. The approach used in the MSW emission estimates assumes that the Cm generation
potential (L0) and the rate of decay that produces Cmfrom MSW, as determined from several studies of Cm
recovery at MSW landfills, are representative of conditions at U.S. MSW landfills. When this top-down approach is
applied at the nationwide level, the uncertainties are assumed to be less than when applying this approach to
individual landfills and then aggregating the results to the national level. In other words, the FOD method as
applied in this Inventory is not facility-specific modeling and while this approach may over- or underestimate CFU
generation at some landfills if used at the facility-level, the result is expected to balance out because it is being
applied nationwide.
There is a high degree of uncertainty associated with the FOD model, particularly when a homogeneous waste
composition and hypothetical decomposition rates are applied to heterogeneous landfills (IPCC 2006). There is less
uncertainty in EPA's GHGRP data because this methodology is facility-specific, uses directly measured Cm recovery
data (when applicable), and allows for a variety of landfill gas collection efficiencies, destruction efficiencies,
and/or oxidation factors to be used.
Uncertainty also exists in the scale-up factors (both 9 percent and 11 percent) applied for years 2005 to 2016 and
2017 to 2021, respectively, and in the back-casted emissions estimates for 2005 to 2009. As detailed in RTI
(2018a), limited information is available for landfills that do not report to the GHGRP. RTI developed an initial list
of landfills that do not report to the GHGRP with the intent of quantifying the total waste-in-place for these
landfills that would add up to the scale-up factor. Input was provided by industry, LMOP, and additional EPA
support. However, many gaps existed in the initial development of this Non-Reporting Landfills Database.
Assumptions were made for hundreds of landfills to estimate their waste-in-place and the subsequent scale-up
factors. The waste-in-place estimated for each landfill is likely not 100 percent accurate and should be considered
a reasonable estimate. Additionally, a simple methodology was used to back-cast emissions for 2005 to 2009 using
the GHGRP-reported emissions from 2010 to 2021. This methodology does not factor in annual landfill to landfill
changes in landfill CH4 generation and recovery. Because of this, an uncertainty factor of 25 percent is applied to
the scale-up factor and years (emission estimates) the scale-up factor is applied to.
Aside from the uncertainty in estimating landfill CH4 generation, uncertainty also exists in the estimates of the
landfill gas oxidized at MSW landfills. Facilities directly reporting to EPA's GHGRP can use oxidation factors ranging
from 0 to 35 percent, depending on their facility-specific CH4 flux. As recommended by the 2006 IPCC Guidelines
for managed landfills, a 10 percent default oxidation factor is applied in the Inventory for both MSW landfills
(those not reporting to the GHGRP and for the years 1990 to 2004 when GHGRP data are not available) and
industrial waste landfills regardless of climate, the type of cover material, and/or presence of a gas collection
system.
Another significant source of uncertainty lies with the estimates of CH4 recovered by flaring and gas-to-energy
projects at MSW landfills that are sourced from the Inventory's CH4 recovery databases (used for years 1990 to
2004). Four CH4 recovery databases are used to estimate nationwide CH4 recovery for MSW landfills for 1990 to
2009. The GHGRP MSW landfills database was added as a fourth recovery database starting with the 1990 to 2013
Inventory report (two years before the full GHGRP data set started being used for net CH4 emissions for the
Inventory). Relying on multiple databases for a complete picture introduces uncertainty because the coverage and
characteristics of each database differs, which increases the chance of double counting avoided emissions. The
methodology and assumptions that go into each database differ. For example, the flare database assumes the
7-14 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
midpoint of each flare capacity at the time it is sold and installed at a landfill; the flare may be achieving a higher
capacity, in which case the flare database would underestimate the amount of Cm recovered. Additionally, two
databases, the EIA database and flare vendor database, could no longer be updated for the entire time series due
to external factors. For example, the EIA database has not been updated since 2006 because the EIA stopped
collected landfill recovery data. The EIA database has, for the most part, been replaced by the GHGRP MSW
landfills database. The flare database was populated annually until 2015, but decreasing, voluntary participation
from flare vendors sharing their flare sales data for several years prior to 2015.
To avoid double counting and to use the most relevant estimate of Cm recovery for a given landfill, a hierarchical
approach is used among the four databases. GHGRP data and the EIA data are given precedence because facility
data were directly reported; the LFGE data are given second priority because Cm recovery is estimated from
facility-reported LFGE system characteristics; and the flare data are given the lowest priority because this database
contains minimal information about the flare, no site-specific operating characteristics, and includes smaller
landfills not included in the other three databases (Bronstein et al. 2012). The coverage provided across the
databases most likely represents the complete universe of landfill Cm gas recovery; however, the number of
unique landfills between the four databases does differ.
The 2006IPCC Guidelines default value of 10 percent for uncertainty in recovery estimates was used for two of the
four recovery databases in the uncertainty analysis where metering of landfill gas was in place (for about 64
percent of the Cm estimated to be recovered). This 10 percent uncertainty factor applies to the LFGE database; 12
percent to the EIA database; and 1 percent for the GHGRP MSW landfills dataset because of the supporting
information provided and rigorous verification process. For flaring without metered recovery data (the flare
database), a much higher uncertainty value of 50 percent is used. The compounding uncertainties associated with
the four databases in addition to the uncertainties associated with the FOD method and annual waste disposal
quantities leads to the large upper and lower bounds for MSW landfills presented in Table 7-5.
The lack of landfill-specific information regarding the number and type of industrial waste landfills in the United
States is a primary source of uncertainty with respect to the industrial waste generation and emission estimates.
The approach used here assumes that most of the organic waste disposed of in industrial waste landfills that
would result in CH4 emissions consists of waste from the pulp and paper and food processing sectors. However,
because waste generation and disposal data are not available in an existing data source for all U.S. industrial waste
landfills, a straight disposal factor is applied over the entire time series to the amount produced to determine the
amounts disposed. Industrial waste facilities reporting under EPA's GHGRP do report detailed waste stream
information, and these data have been used to improve, for example, the DOC value used in the Inventory
methodology for the pulp and paper sector. A 10 percent oxidation factor is also applied to CH4 generation
estimates for industrial waste landfills and carries the same amount of uncertainty as with the factor applied to
CH4 generation for MSW landfills.
The results of the 2006 IPCC Guidelines Approach 2 quantitative uncertainty analysis are summarized in Table 7-5.
There is considerable uncertainty for the MSW landfills estimates due to the many data sources used, each with its
own uncertainty factor.
Table 7-5: Approach 2 Quantitative Uncertainty Estimates for ChU Emissions from Landfills (MMT CO2
Eq. and Percent)
2021 Emission
Source
Gas
Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.)
(MMT C02
Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Total Landfills
ch4
122.6
99.0
154.8
-19%
26%
MSW
ch4
103.7
83.0
137.5
-20%
33%
Industrial
ch4
18.9
15.9
25.7
-16%
36%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
Individual uncertainty factors are applied to activity data and emission factors in the Monte Carlo analysis.
Waste 7-15
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
QA/QC and Verification
General quality assurance/quality control (QA/QC) procedures were applied consistent with the U.S. Inventory
QA/QC plan, which is in accordance with Vol. 1, Chapter 6 of 2006IPCC Guidelines (see Annex 8 for more details).
QA/QC checks are performed for the transcription of the published data set (e.g., EPA's GHGRP dataset) used to
populate the Inventory data set in terms of completeness and accuracy against the reference source. Additionally,
all datasets used for this category have been checked to ensure they are of appropriate quality and are
representative of U.S. conditions. The primary calculation spreadsheet is tailored from the 2006 IPCC Guidelines
waste model and has been verified previously using the original, peer-reviewed IPCC waste model. All model input
values and calculations were verified by secondary QA/QC review. Stakeholder engagements sessions in 2016 and
2017 were used to gather input on methodological improvements and facilitate an external expert review on the
methodology, activity data, and emission factors.
Category-specific checks include the following:
• Evaluation of the secondary data sources used as inputs to the Inventory dataset to ensure they are
appropriately collected and are reliable;
• Cross-checking the data (activity data and emissions estimates) with previous years to ensure the data are
reasonable, and that any significant variation can be explained through the activity data;
• Conducting literature reviews to evaluate the appropriateness of country-specific emission factors (e.g.,
DOC values, precipitation zones with respect to the application of the k values) given findings from recent
peer-reviewed studies; and
• Reviewing secondary datasets to ensure they are nationally complete and supplementing where
necessary (e.g., using a scale-up factor to account for emissions from landfills that do not report to EPA's
GHGRP).
A primary focus of the QA/QC checks in past Inventories was to ensure that Cm recovery estimates were not
double-counted and that all LFGE projects and flares were included in the respective project databases. QA/QC
checks performed in the past for the recovery databases were not performed in this Inventory, because new data
were not added to the recovery databases in this Inventory year.
For the GHGRP data, EPA verifies annual facility-level reports through a multi-step process (e.g., combination of
electronic checks and manual reviews by staff) to identify potential errors and ensure that data submitted to EPA
are accurate, complete, and consistent.3 Based on the results of the verification process, EPA follows up with
facilities to resolve mistakes that may have occurred. The post-submittals checks are consistent with several
general and category-specific QC procedures, including range checks, statistical checks, algorithm checks, and year-
to-year checks of reported data and emissions. For the MSW Landfills sector, under Subpart HH of the GHGRP,
MSW Landfills with gas collection are required to report emissions from their site using both a forward- (using a
first order decay model as a basis) and back-calculating (using parameters specific to the landfill itself, such as
measured recovery and collection efficiency of the landfill gas) methodology. Details on the forward- and back-
calculation approach can be found in Annex 3.14 and 40 CFR Subpart HH of Part 98. Reporters can choose which of
these two methodologies they believe best represents the emissions at their landfill and are required to submit
that value as their total Subpart HH emissions. Facilities are generally not expected to switch between the two
equations each year, as the emissions calculated using each method can vary greatly and can have a significant
effect on emission trends for that landfill, and potentially the entire MSW Landfill sector under the GHGRP. Key
checks are in place to assure that emissions are trending in a sensible way year over year for each reporting
landfill.
3 See https://www.epa.gov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
7-16 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Recalculations Discussion
Revisions to the individual facility reports submitted to EPA's GHGRP can be made at any time and a portion of
facilities have revised their reports since 2010 for various reasons, resulting in changes to the total net Cm
emissions for MSW landfills. Each Inventory year, the back-casted emissions for 2005 to 2009 will be recalculated
using the most recently verified data from the GHGRP. Changes in these data result in changes to the back-casted
emissions. The impact of the revisions to the GHGRP Subpart HH annual greenhouse gas reports resubmitted for
2010 to 2021 slightly increased or decreased total Subpart HH reported net emissions up to 0.5 percent in the
years the Subpart HH data are applied (i.e., 2005 to 2020). The resubmissions resulted in annual increases ranging
from 0.1 percent to 0.3 percent to the net MSW emissions between 2005 to 2009, no net emission changes for
2010 to 2015, and a slight decrease averaging -0.15 percent of emissions is observed between 2016 to 2019. A 0.5
percent increase is observed for 2020. Between 2005 to 2020, on average, the impact or change was very small
(less than 0.1% percent) in emissions across all reporters. A change in net Subpart HH reported emissions results in
the same percentage change in the Inventory emissions for that year.
The revision to the industrial food waste disposal factor from 4.86 percent to 6 percent increased net industrial
emissions between 2010 to 2020 from a low of 2.1 percent in 2011 to a high of 10.9 percent in 2020. Combined,
these two recalculations increased net landfill emissions for all years between 2005 to 2020. Emissions increased
by less than 1 percent between 2005 to 2014 (low of 0.3 percent in 2005 and a high of 0.8 percent in 2014) and up
to 1.9 percent between 2015 to 2020 (low of 1.0 percent in 2015 to a high of 1.9 percent in 2020).
In addition, for the current Inventory, estimates of CO2 equivalent emissions totals of CH4 emissions from landfills
have been revised to apply the 100-year global warming potentials (GWPs) provided in the IPCC Fifth Assessment
Report (AR5) (IPCC 2013). AR5 GWP values differ slightly from those presented in the IPCC Fourth Assessment
Report (AR4) (IPCC 2007) (used in the previous inventories). The GWP of CH4 has increased from 25 to 28, leading
to an overall increase in CCh-equivalent CH4 emissions. The AR5 GWPs have been applied across the entire time
series for consistency. Compared to the previous Inventory which applied 100-year GWP values from AR4, the
change in CH4 emissions was a 12 percent increase for each year of the time series. Further discussion on this
update and the overall impacts of updating the inventory GWPs to reflect the IPCC Fifth Assessment Report can be
found in Chapter 9, Recalculations and Improvements.
Planned Improvements
EPA received recommendations from industry stakeholders regarding the DOC values and decay rates (k values)
required to be used in the GHGRP calculations. Stakeholders have suggested that newer, more up-to-date default
values considering recent trends in the composition of waste disposed in MSW landfills for both k and DOC in the
GHGRP should be developed and reflected in the 2005 and later years of the Inventory. In response, EPA
developed a multivariate analysis using publicly available Subpart HH GHGRP data, solving for optimized DOC and k
values across the more than 1,100 landfills reporting to the program. The results of this analysis could help inform
a current GHGRP rulemaking (87 FR 36920) where changes could be made to the default DOC and k values
contained within Subpart HH, which could then be carried over to the Inventory emissions estimates for MSW
landfills upon promulgation of any revisions to 40 CFR part 98. This potential improvement may be long-term.
With respect to the scale-up factor, EPA received comments on revisions made to the scale-up for the 1990 to
2020 inventory from a total waste-in-place approach to a time-based threshold of 50 years. Commenters noted
that this time-based threshold approach does not adjust for the non-linearity of methane production of landfill
gas. In response, EPA will further investigate how best to account for emissions from MSW landfills that do not
report to the GHGRP, including using the FOD model for these landfills based on estimated annual waste disposed
for this subset of landfills between 2005 to 2021, reverting to the total waste-in-place approach, or modifying the
time-based threshold approach. Any methodological revisions to accounting for emissions from this subset of
landfills will be made in the future (1990 to 2022) Inventory.
Waste 7-17
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Relatedly, EPA will periodically assess the impact to the waste-in-place and emissions data from GHGRP facilities
that have resubmitted annual reports during any reporting years, are new reporting facilities, and from facilities
that have stopped reporting to the GHGRP to ensure national estimates are as complete as possible. Facilities may
stop reporting to the GHGRP when they meet the "off-ramp" provisions (reported less than 15,000 metric tons of
CO2 equivalent emissions for 3 consecutive years or less than 25,000 metric tons of CO2 equivalent emissions for 5
consecutive years). If warranted, EPA will revise the scale-up factor to reflect newly acquired information to ensure
completeness of the Inventory. EPA considered public comments received on the 1990-2019 Inventory specific to
using a time-based threshold to calculate the scale-up factor instead of a total waste-in-place approach. The
rationale supporting the comments was that older, closed landfills with large quantities of waste-in-place are
driving up the scale-up factor but have little impact on total methane generation. EPA assessed two time-based
scenarios for developing the scale-up factor - one scenario looking at the past 30 years of waste disposed, and the
second looking at the past 50 years of waste disposed. The 50-year time-based threshold was applied and resulted
in the 11 percent scale-up factor used between 2017 and 2021.
EPA is planning to account for unmanaged landfills in Puerto Rico and other U.S. Territories to the landfill
emissions estimates. Data limitations for historical waste received at these sites make this challenging. Presently,
emissions from managed sites in Puerto Rico and Guam are accounted for in 2005 to present as part of the GHGRP
Subpart HH dataset.
Additionally, with the recent publication of the 2019 Refinement to the 2006IPCC Guidelines for National
Greenhouse Gas Inventories (IPCC 2019), EPA will begin to update applicable emission factors, methodologies, and
assumptions underlying emission estimates for landfills and make any applicable changes during the next (1990 to
2022) Inventory cycle per the 2019 Refinement.
Box 7-4: Overview of U.S. Solid Waste Management Trends
As shown in Figure 7-4 and Figure 7-5 landfilling of MSW is currently and has been the most common waste
management practice. A large portion of materials in the waste stream are recovered for recycling and
composting, which is becoming an increasingly prevalent trend throughout the country. Materials that are
composted and recycled would have previously been disposed in a landfill.
Figure 7-4: Management of Municipal Solid Waste in the United States, 2018
Management of MSW in the United States
7-18 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Note: 2018 is the latest year of available data. Data taken from Table 35 of EPA (2020a). MSW to WTE is combustion
with energy recovery (WTE = waste-to-energy).
Source: EPA (2020b)
Figure 7-5: MSW Management Trends from 1990 to 2018
160
Landfilling
\
V < **
Recycling
140
120
. KM
1 80
^ 60
40
_ ~ Composting Other Food Management
0 *" (2018 only)
^ ¦$* ^ ^ ^ $ ¦$* $ $ <$'•$' $ $ ^ $ 4^ ¦#' ¦$
Recycling — — Composting > Other Food Man eg em ent Combustion with Energy Recovery — — — Lsndfiling
Note: 2018 is the latest year of available data. Only one year of data (2018) is available for the "Other Food Management"
category.
Source: EPA (2020b). The EPA Advancing Sustainable Materials Management reports only present data for select years,
thus several reports were used in the compilation of this figure. All data were taken from Table 35 in EPA 2020b for 1990,
2000, 2015, 2017 and 2018. Data were taken from Table 35 in EPA (2019) for 2010 and 2016. Data were taken from EPA
(2018) for 2014. Data were taken from Table 35 of EPA (2016b) for 2012 and 2013. Data were taken from Table 30 of EPA
(2014) for 2008 and 2011. The reports with data available for years prior to EPA (2012) can be provided upon request but
are no longer on the EPA's Advancing Sustainable Materials Management web site.4
Table 7-6 presents the national-level material composition of waste disposed across typical MSW landfills in
the United States over time. It is important to note that the actual composition of waste entering each landfill
will vary from that presented in Table 7-6.
Understanding how the waste composition changes over time, specifically for the degradable waste types
(i.e., those types known to generate Cm as they break down in a modern MSW landfill), is important for
estimating greenhouse gas emissions. Increased diversion of degradable materials so that they are not
disposed of in landfills reduces the Cm generation potential and Cm emissions from landfills. For certain
degradable waste types (i.e., paper and paperboard), the amounts discarded have decreased over time due
to an increase in waste diversion through recycling and composting (see Table 7-6 and Figure 7-6). As shown
in Figure 7-6, the diversion of food scraps has been consistently low since 1990 because most cities and
counties do not practice curbside collection of these materials, although the quantity has been slowly
increasing in recent years. Neither Table 7-6 nor Figure 7-6 reflect the frequency of backyard composting of
yard trimmings and food waste because this information is largely not collected nationwide and is hard to
estimate.
4 See https://www.epa.gov/facts-and-figures-about-materials-waste-and-recvcling/advancing-sustainable-materials-
management.
Waste 7-19
-------
Table 7-6: Materials Discarded in the Municipal Waste Stream by Waste Type from 1990
to 2018 (Percent)
Waste Type
1990
2005
2015
2016
2017
2018
Paper and Paperboard
30.0%
24.7%
13.3%
12.7%
13.1%
11.8%
Glass
6.0%
5.8%
5.0%
4.9%
4.9%
5.2%
Metals
7.2%
7.9%
9.5%
9.8%
9.9%
9.5%
Plastics
9.5%
16.4%
18.9%
18.9%
19.2%
18.5%
Rubber and Leather
3.2%
2.9%
3.3%
3.4%
3.5%
3.4%
Textiles
2.9%
5.3%
7.7%
8.0%
8.0%
7.7%
Wood
6.9%
7.5%
8.0%
8.8%
8.7%
8.3%
Other
1.4%
1.8%
2.2%
2.2%
2.2%
2.0%
Food Scraps
13.6%
18.5%
22.0%
22.1%
22.0%
24.1%
Yard Trimmings
17.6%
7.0%
7.8%
6.9%
6.2%
7.2%
Miscellaneous Inorganic Wastes
1.7%
2.2%
2.3%
2.3%
2.3%
2.3%
Source: EPA (2020b)
Figure 7-6: Percent of Degradable Materials Diverted from Landfills from 1990 to 2018
(Percent)
Note: The data shown in this chart are for recycling of paper and paperboard, composting of food scraps and yard
trimmings, and alternative management pathways for the Other Food Management (non-disposal) category. The Other
Food Management (non-disposal) category is a new addition and only one year of data are available for 2018 (28
percent of the food waste generated was beneficially reused or managed using a method that was not landfilling,
recycling, or composting). The Other Food Management pathways include animal feed, bio-based
materials/biochemical processing, co-digestion/anaerobic digestion, donation, land application, and sewer/wastewater
treatment.
Source: EPA (2020b). The EPA Advancing Sustainable Materials reports only present data for select years, thus several
reports were used in the compilation of this figure. All data were taken from Table 35 in EPA (2020b) for 1990, 2000,
2015, 2017 and 2018. Data were taken from Table 35 in EPA (2019) for 2010 and 2016. Data were taken from EPA
(2018) for 2014. Data were taken from Table 35 of EPA (2016b) for 2012 and 2013. Data were taken from Table 30 of
EPA (2014) for 2008 and 2011. The reports with data available for years prior to EPA (2012) can be provided upon
request, but are not longer on the EPA's Advancing Sustainable Materials Management website.5
7-20 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
7.2 Wastewater Treatment and Discharge
(CRF Source Category 5D)
Wastewater treatment and discharge processes are sources of anthropogenic methane (Cm) and nitrous oxide
(N2O) emissions. Wastewater from domestic and industrial sources is treated to remove soluble organic matter,
suspended solids, nutrients, pathogenic organisms, and chemical contaminants.6 Treatment of domestic
wastewater may either occur on site, most commonly through septic systems, or off site at centralized treatment
systems, most commonly at publicly owned treatment works (POTWs). In the United States, approximately 17
percent of domestic wastewater is treated in septic systems or other on-site systems, while the rest is collected
and treated centrally (U.S. Census Bureau 2019). Treatment of industrial wastewater may occur at the industrial
plant using package or specially designed treatment plants or be collected and transferred off site for co-treatment
with domestic wastewater in centralized treatment systems.
Centralized Treatment. Centralized wastewater treatment systems use sewer systems to collect and transport
wastewater to the treatment plant. Sewer collection systems provide an environment conducive to the formation
of Cm, which can be substantial depending on the configuration and operation of the collection system (Guisasola
et al. 2008). Recent research has shown that at least a portion of CH4 formed within the collection system enters
the centralized system where it contributes to CH4 emissions from the treatment system (Foley et al. 2015).
The treatment plant may include a variety of processes, ranging from physical separation of material that readily
settles out (typically referred to as primary treatment), to treatment operations that use biological processes to
convert and remove contaminants (typically referred to as secondary treatment), to advanced treatment for
removal of targeted pollutants, such as nutrients (typically referred to as tertiary treatment). Not all wastewater
treatment plants conduct primary treatment prior to secondary treatment, and not all plants conduct advanced or
tertiary treatment (EPA 1998a).
Soluble organic matter is generally removed using biological processes in which microorganisms consume the
organic matter for maintenance and growth. Microorganisms can biodegrade soluble organic material in
wastewater under aerobic or anaerobic conditions, where the latter condition produces CH4. The resulting biomass
(sludge) is removed from the effluent prior to discharge to the receiving stream and may be further biodegraded
under aerobic or anaerobic conditions, such as anaerobic sludge digestion. Sludge can be produced from both
primary and secondary treatment operations. Some wastewater may also be treated using constructed (or semi-
natural) wetland systems, though this is much less common in the United States and represents a relatively small
portion of wastewater treated centrally (<0.1 percent) (ERG 2016). Constructed wetlands are a coupled anaerobic-
aerobic system and may be used as the primary method of wastewater treatment, or are more commonly used as
a final treatment step following settling and biological treatment. Constructed wetlands develop natural processes
that involve vegetation, soil, and associated microbial assemblages to trap and treat incoming contaminants (IPCC
2014). Constructed wetlands do not produce secondary sludge (sewage sludge).
The generation of N2O may also result from the treatment of wastewater during both nitrification and
denitrification of the nitrogen (N) present, usually in the form of urea, proteins, and ammonia. Ammonia N is
converted to nitrate (NO3) through the aerobic process of nitrification. Denitrification occurs under
anoxic/anaerobic conditions, whereby anaerobic or facultative organisms reduce oxidized forms of nitrogen (e.g.,
nitrite, nitrate) in the absence of free oxygen to produce nitrogen gas (N2). Nitrous oxide is generated as a by-
5 See https://www.epa.gov/facts-and-figures-about-materials-waste-and-recycling/advancing-sustainable-materials-
management.
6 Throughout the Inventory, emissions from domestic wastewater also include any commercial and industrial wastewater
collected and co-treated with domestic wastewater.
Waste 7-21
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
product of nitrification, or as an intermediate product of denitrification. No matter where N20 is formed it is
typically stripped (i.e., transferred from the liquid stream to the air) in aerated parts of the treatment process.
Stripping also occurs in non-aerated zones at rates lower than in aerated zones.
On-site Treatment. The vast majority of on-site systems in the United States are septic systems composed of a
septic tank, generally buried in the ground, and a soil dispersion system. Solids and dense materials contained in
the incoming wastewater (influent) settle in the septic tank as sludge. Floatable material (scum) is also retained in
the tank. The sludge that settles on the bottom of the tank undergoes anaerobic digestion. Partially treated water
is discharged in the soil dispersal system. The solid fraction accumulates and remains in the tank for several years,
during which time it degrades anaerobically. The gas produced from anaerobic sludge digestion (mainly CFU and
biogenic CO2) rises to the liquid surface and is typically released through vents. The gas produced in the effluent
dispersal system (mainly N2O and biogenic CO2) is released through the soil.
Discharge. Dissolved Cm and N2O that is present in wastewater discharges to aquatic environments has the
potential to be released (Short et al. 2014; Short et al. 2017), and the presence of organic matter or nitrogen in
wastewater discharges is generally expected to increase CH4 and N2O emissions from these aquatic environments.
Where organic matter is released to slow-moving aquatic systems, such as lakes, estuaries, and reservoirs, CH4
emissions are expected to be higher. Similarly, in the case of discharge to nutrient-impacted or hypoxic waters,
N2O emissions can be significantly higher.
In summary, the principal factor in determining the Cm generation potential of wastewater is the amount of
degradable organic material in the wastewater. Common parameters used to measure the organic component of
the wastewater are the biochemical oxygen demand (BOD) and chemical oxygen demand (COD). Under the same
conditions, wastewater with higher COD (or BOD) concentrations will generally yield more Cm than wastewater
with lower COD (or BOD) concentrations. BOD represents the amount of oxygen that would be required to
completely consume the organic matter contained in the wastewater through aerobic decomposition processes,
while COD measures the total material available for chemical oxidation (both biodegradable and non-
biodegradable). The BOD value is most commonly expressed in milligrams of oxygen consumed per liter of sample
during 5 days of incubation at 20°C, or BODs. Throughout the rest of this chapter, the term "BOD" refers to BODs.
Because BOD is an aerobic parameter, it is preferable to use COD to estimate CH4 production, since Cm is
produced only in anaerobic conditions. Where present, biogas recovery and flaring operations reduce the amount
of Cm generated that is actually emitted. Per IPCC guidelines (IPCC 2019), emissions from anaerobic sludge
digestion, including biogas recovery and flaring operations, where the digester's primary use is for treatment of
wastewater treatment solids, are reported under Wastewater Treatment. The principal factor in determining the
N2O generation potential of wastewater is the amount of N in the wastewater. The variability of N in the influent
to the treatment system, as well as the operating conditions of the treatment system itself, also impact the N2O
generation potential. The methods and underlying data sources to estimate emissions from are described in
further detail in the "Methodology and Time Series Consistency" section below for treatment of domestic and
industrial wastewater.
Overall, treatment of wastewater emitted 42.0 MMT CO2 Eq. in 2021. Methane (CH4) emissions from domestic
wastewater treatment and discharge were estimated to be 11.9 MMT CO2 Eq. (424 kt CH4) and 2.0 MMT CO2 Eq.
(72 kt CH4), respectively, totaling 13.9 MMT CO2 Eq. (496 kt CH4) in 2021. Emissions remained fairly steady from
1990 through 2002 but have decreased since that time due to decreasing percentages of wastewater being treated
in anaerobic systems, generally including reduced use of on-site septic systems and central anaerobic treatment
systems (EPA 1992,1996, 2000, and 2004a; U.S. Census Bureau 2019). In 2021, CH4 emissions from industrial
wastewater treatment and discharge were estimated to be 6.6 MMT CO2 Eq. (237 kt CH4) and 0.5 MMT CO2 Eq. (19
kt CH4), respectively, totaling 7.2 MMT CO2 Eq. (256 kt CH4). Industrial emissions from wastewater treatment have
generally increased across the time series through 1999 and then fluctuated up and correspond with production
changes from the pulp and paper manufacturing, meat and poultry processing, fruit and vegetable processing,
starch-based ethanol production, petroleum refining, and brewery industries. Industrial wastewater emissions
have generally seen an uptick since 2016. Table 7-7 and Table 7-8 provide CH4 emission estimates from domestic
and industrial wastewater treatment.
7-22 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
With respect to N2O, emissions from domestic wastewater treatment and discharge in 2021 were estimated to be
16.2 MMT CO2 Eq. (61 kt N20) and 4.2 MMT C02 Eq. (16 kt N20), respectively, totaling 20.4 MMT C02 Eq. (77 kt
N2O). Nitrous oxide emissions from wastewater treatment processes gradually increased across the time series
because of increasing U.S. population and protein consumption. In 2021, N2O emissions from industrial
wastewater treatment and discharge were estimated to be 0.4 MMT CO2 Eq. (1.5 kt N2O) and 0.1 MMT CO2 Eq.
(0.3 kt N2O), respectively, totaling 0.5 MMT CO2 Eq. (1.7 kt N2O). Industrial emission sources have gradually
increased across the time series with production changes associated with the treatment of wastewater from the
pulp and paper manufacturing, meat and poultry processing, petroleum refining, and brewery industries. Table 7-7
and Table 7-8 provide N2O emission estimates from domestic wastewater treatment.
Table 7-7: ChU and N2O Emissions from Domestic and Industrial Wastewater Treatment
(MMT COz Eq.)
Activity
1990
2005
2017
2018
2019
2020
2021
ch4
22.7
22.7
21.5
21.4
21.2
21.3
21.1
Domestic Treatment
15.1
14.6
12.6
12.3
11.9
12.1
11.9
Domestic Effluent
1.4
1.4
2.0
2.0
2.0
2.0
2.0
Industrial Treatment3
5.5
6.1
6.4
6.5
6.6
6.6
6.6
Industrial Effluent3
0.7
0.6
0.6
0.6
0.6
0.5
0.5
n2o
14.8
18.1
20.6
21.2
21.3
20.9
20.9
Domestic Treatment
10.5
13.7
15.7
16.2
16.4
16.1
16.2
Domestic Effluent
3.9
3.9
4.4
4.5
4.5
4.3
4.2
Industrial Treatment15
0.3
0.4
0.4
0.4
0.5
0.4
0.4
Industrial Effluentb
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Total
37.5
40.7
42.2
42.5
42.5
42.2
42.0
a Industrial activity for CH4 includes the pulp and paper manufacturing, meat and poultry
processing, fruit and vegetable processing, starch-based ethanol production, petroleum refining,
and breweries industries.
b Industrial activity for N20 includes the pulp and paper manufacturing, meat and poultry
processing, starch-based ethanol production, and petroleum refining.
Note: Totals may not sum due to independent rounding.
Table 7-8: ChU and N2O Emissions from Domestic and Industrial Wastewater Treatment (kt)
Activity
1990
2005
2017
2018
2019
2020
2021
ch4
811
809
770
763
755
761
753
Domestic Treatment
539
521
449
438
426
433
424
Domestic Effluent
49
49
72
73
73
72
72
Industrial Treatment3
196
216
229
232
236
237
237
Industrial Effluent3
27
22
20
20
20
19
19
n2o
56
68
78
80
80
79
79
Domestic T reatment
40
52
59
61
62
61
61
Domestic Effluent
15
15
17
17
17
16
16
Industrial Treatment15
1
1
1
2
2
1
1
Industrial Effluentb
+
+
+
+
+
+
+
+ Does not exceed 0.5 kt.
a Industrial activity for CH4 includes the pulp and paper manufacturing, meat and poultry processing,
fruit and vegetable processing, starch-based ethanol production, petroleum refining, and
breweries industries.
b Industrial activity for N20 includes the pulp and paper manufacturing, meat and poultry
processing, starch-based ethanol production, and petroleum refining.
Note: Totals by gas may not sum due to independent rounding.
Waste 7-23
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Methodology and Time-Series Consistency
The methodologies presented in IPCC (2019) form the basis of the Cm and N2O emission estimates for both
domestic and industrial wastewater treatment and discharge.7 Domestic wastewater treatment follows the IPCC
Tier 2 methodology for key pathways, while domestic wastewater discharge follows IPCC Tier 2 discharge
methodology and emission factors. Default factors from IPCC (2019) or IPCC (2006) are used when there are
insufficient U.S.-specific data to develop a U.S.-specific factor, though IPCC default factors are often based in part
on data from or representative of U.S. wastewater treatment systems. Industrial wastewater treatment follows
IPCC Tier 1 and wastewater treatment discharge follows Tier 1 or Tier 2 methodologies, depending on the industry.
EPA will continue to implement the Tier 2 discharge methodology for more industries as data are investigated and
time and resource constraints allow (see the Planned Improvements section below). Similar to domestic
wastewater, IPCC default factors are used when there are insufficient U.S.-specific data to develop a U.S.-specific
factor.
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2021. In the following cases, the source used to capture activity data changed over the time series. EPA
transitioned to these newer data sources to continue estimating emissions in a way that ensured both accuracy
and continuity. For example:
• Starch-based ethanol production data: the source used for 1990 to 2017 production was no longer
available after 2017. A new, publicly available source was identified and is used for production in 2015-
2021. However, this source does not have sufficient data for the earlier timeseries. EPA confirmed with
experts familiar with the sources that combining these two sources to populate the time series was
accurate (ERG 2019; Lewis 2019) and does not present any significant discontinuities in the time series.
• Brewery production data: the source used for production changed in 2007 to publish craft brewery
production broken out by size but does not include data prior to 2007. Therefore, rather than estimating
total production data prior to 2007 with this source, another data source was used to ensure accuracy of
production data through the time series (ERG 2018b).
Refer to the Recalculations section below for details on updates implemented to improve accuracy, consistency
and/or completeness of the time series.
Domestic Wastewater CH4 Emission Estimates
Domestic wastewater CFU emissions originate from both septic systems and from centralized treatment systems.
Within these centralized systems, Cm emissions can arise from aerobic systems that liberate dissolved Cm that
formed within the collection system or that are designed to have periods of anaerobic activity (e.g., constructed
wetlands and facultative lagoons), anaerobic systems (anaerobic lagoons and anaerobic reactors), and from
anaerobic sludge digesters when the captured biogas is not completely combusted. Emissions will also result from
the discharge of treated effluent from centralized wastewater plants to waterbodies where carbon accumulates in
sediments (typically slow-moving systems, such as lakes, reservoirs, and estuaries). The systems with emissions
estimates are:
• Septic systems (A);
• Centralized treatment aerobic systems (B), including aerobic systems (other than constructed wetlands)
(Bl), constructed wetlands only (B2), and constructed wetlands used as tertiary treatment (B3);
7 IPCC (2019) updates, supplements, and elaborates the 2006 IPCC Guidelines where gaps or out-of-date science have been
identified. EPA used these methodologies to improve completeness and include sources of greenhouse gas emissions that have
not been estimated prior to the 1990-2019 Inventory, such as N20 emissions from industrial wastewater treatment, and to
improve emission estimates for other sources, such as emissions from wastewater discharge and centralized wastewater
treatment.
7-24 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 • Centralized anaerobic systems (C);
2 • Anaerobic sludge digesters (D); and
3 • Centralized wastewater treatment effluent (E).
4 Methodological equations for each of these systems are presented in the subsequent subsections; total domestic
5 Cm emissions are estimated as follows:
6 Equation 7-4: Total Domestic ChU Emissions from Wastewater Treatment and Discharge
7 Total Domestic CH4 Emissions from Wastewater Treatment and Discharge (kt) = A+ B + C+ D + E
8 Table 7-9 presents domestic wastewater CH4 emissions for both septic and centralized systems, including
9 anaerobic sludge digesters and emissions from centralized wastewater treatment effluent, in 2021.
10 Table 7-9: Domestic Wastewater ChU Emissions from Septic and Centralized Systems (2021,
11 kt, MMT CO2 Eq. and Percent)
CH4 Emissions (kt)
CH4 Emissions
(MMT CO? Eq.)
% of Domestic
Wastewater CH4
Septic Systems (A)
223
6.2
45.0
Centrally-Treated Aerobic Systems (B)
74
2.1
14.8
Centrally-Treated Anaerobic Systems (C)
119
3.3
24.1
Anaerobic Sludge Digesters (D)
8
0.2
1.6
Centrally-Treated Wastewater Effluent (E)
72
2.0
14.5
Total 496 13.9 100
12 Emissions from Septic Systems:
13 Methane emissions from septic systems were estimated by multiplying the U.S. population by the percent of
14 wastewater treated in septic systems (about 17 percent in 2021; U.S. Census Bureau 2019) and an emission factor
15 and then converting the result to kt/year.
16 U.S. population data were taken from historic U.S. Census Bureau national population totals data and include the
17 populations of the United States and Puerto Rico (U.S. Census Bureau 2002; U.S. Census Bureau 2011; U.S. Census
18 Bureau 2021a and 2021b; Instituto de Estadisticas de Puerto Rico 2021). Population data for American Samoa,
19 Guam, Northern Mariana Islands, and the U.S. Virgin Islands were taken from the U.S. Census Bureau International
20 Database (U.S. Census Bureau 2022). Table 7-12 presents the total U.S. population for 1990 through 2021. The
21 fraction of the U.S. population using septic systems or centralized treatment systems is based on data from the
22 American Housing Surveys (U.S. Census Bureau 2019).
23 Methane emissions for septic systems are estimated as follows:
24 Equation 7-5: ChU Emissions from Septic Systems
25 Emissions from Septic Systems (U.S. Specific) = A
26 = USpop X (Tseptic) X (EFseptic) X 1/109 X 365.25
27 Table 7-10: Variables and Data Sources for ChU Emissions from Septic Systems
Variable
Variable Description
Units
Inventory Years: Source of
Value
USpop
U.S. population3
Persons
United States and Puerto
Rico:
1990-1999: US Census Bureau
(2002); Instituto de
Estadisticas de Puerto Rico
(2021)
Waste 7-25
-------
Inventory Years: Source of
Variable
Variable Description
Units
Value
2000-2009: U.S. Census
Bureau (2011)
2010-2019: U.S. Census
Bureau (2021a)
2020-2021: U.S. Census
Bureau (2021b)
U.S. Territories other than
Puerto Rico:
1990-2021: U.S. Census
Bureau (2022)
Odd years from 1989 through
2019: U.S. Census Bureau
(2019)
Tseptic
Percent treated in septic systems3
%
Data for intervening years
obtained by linear
interpolation
2020-2021: Forecasted from
the rest of the time series
EFseptic
Methane emission factor - septic systems
(10.7)
g CH4/capita/day
1990-2021: Leverenz et al.
(2010)
1/109
Conversion factor
g to kt
Standard conversion
365.25
Conversion factor
Days in a year
Standard conversion
1 3 Value of activity data varies over the Inventory time series.
2 Emissions from Centrally Treated Aerobic and Anaerobic Systems:
3 Methane emissions from POTWs depend on the total organics in wastewater. Table 7-12 presents the total
4 organically degradable material in wastewater, or TOW, for 1990 through 2021. The TOW was determined using
5 BOD generation rates per capita weighted average both with and without kitchen scraps as well as an estimated
6 percent of housing units that utilize kitchen garbage disposals. Households with garbage disposals (with kitchen
7 scraps or ground up food scraps) typically have wastewater with higher BOD than households without garbage
8 disposals due to increased organic matter contributions (ERG 2018a). The equations are as follows:
9 Equation 7-6: Total Wastewater BODs Produced per Capita (U.S.-Specific [ERG 2018a])
10 BODgenrate (kg/capita/day)= BODwithoutscrap x (1 - %kitchen disposal) + BODwithscraps x (%kitchen disposal)
11
12 Equation 7-7: Total Organically Degradable Material in Domestic Wastewater (IPCC 2019
13 [Eq. 6.3])
14 TOW (Gg/year) = USpop x BODgenrate x 365.25 x 1/106
15 Table 7-11: Variables and Data Sources for Organics in Domestic Wastewater
Variable
Variable Description
Units
Inventory Years: Source of
Value
BODgen rate
Total wastewater BOD produced per
capita
kg/capita/day
1990-2021: Calculated
BODwjthout scrap
Wastewater BOD produced per capita
without kitchen scraps3
kg/capita/day
1990-2003: Metcalf & Eddy
(2003)
2004-2013: Linear
interpolation
2014-2021: Metcalf & Eddy
(2014)
BODwjth scraps
Wastewater BOD produced per capita
with kitchen scraps3
kg/capita/day
% kitchen disposal
Percent of housing units with kitchen
%
1990-2013: U.S. Census
7-26 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
kitchen disposal3
Bureau (2013)
2014-2021: Forecasted from
the rest of the time series
TOW
Total wastewater BOD Produced per
Capita3
Gg BOD/year
1990-2021: Calculated, ERG
(2018a)
United States and Puerto
Rico:
1990-1999: US Census
Bureau (2002); Instituto de
Estadisticas de Puerto Rico
(2021)
2000-2009: U.S. Census
USpop
U.S. population3
Persons
Bureau (2011)
2010-2019: U.S. Census
Bureau (2021a)
2020-2021: U.S. Census
Bureau (2021b)
U.S. Territories other than
Puerto Rico:
1990-2021: U.S. Census
Bureau (2022)
365.25
Conversion factor
Days in a year
Standard conversion
1/106
Conversion factor
kg to Gg
Standard conversion
a Value of activity data varies over the Inventory time series.
Table 7-12: U.S. Population (Millions) and Domestic Wastewater TOW (kt)
Activity
1990
2005
2017
2018
2019
2020
2021
Population
253
300
329
330
332
335
336
TOW
8,131
ii 9/624 /
9,894
9,958
10,019
10,132
10,159
Sources: U.S. Census Bureau (2002); U.S. Census Bureau (2011); U.S. Census Bureau (2021a and
2021b); Instituto de Estadisticas de Puerto Rico (2021); U.S. Census Bureau (2022); ERG (2018a).
Methane emissions from POTWs were estimated by multiplying the total organics in centrally treated wastewater
(total BODs) produced per capita in the United States by the percent of wastewater treated centrally, or percent
collected (about 83 percent in 2021), the correction factor for additional industrial BOD discharged to the sewer
system, the relative percentage of wastewater treated by aerobic systems (other than constructed wetlands),
constructed wetlands only, and anaerobic systems, and the emission factor8 for aerobic systems, constructed
wetlands only, and anaerobic systems. Methane emissions from constructed wetlands used as tertiary treatment
were estimated by multiplying the flow from treatment to constructed wetlands, wastewater BOD concentration
entering tertiary treatment, constructed wetlands emission factor, and then converting to kt/year.
In the United States, the removal of sludge9 from wastewater reduces the biochemical oxygen demand of the
wastewater that undergoes aerobic treatment. The amount of this reduction (S) is estimated using the default IPCC
(2019) methodology and multiplying the amount of sludge removed from wastewater treatment in the United
States by the default factors in IPCC (2019) to estimate the amount of BOD removed based on whether the
treatment system has primary treatment with no anaerobic sludge digestion (assumed to be zero by expert
8 Emission factors are calculated by multiplying the maximum CH4-producing capacity of domestic wastewater (B0, 0.6 kg
CH4/kg BOD) and the appropriate methane correction factors (MCF) for aerobic (0.03) and anaerobic (0.8) systems (IPCC 2019,
Table 6.3) and constructed wetlands (0.4) (IPCC 2014, Table 6.4).
9 Throughout this document, the term "sludge" refers to the solids separated during the treatment of municipal wastewater.
The definition includes domestic septage. "Biosolids" refers to treated sewage sludge that meets the EPA pollutant and
pathogen requirements for land application and surface disposal.
Waste 7-27
-------
1 judgment), primary treatment with anaerobic sludge digestion, or secondary treatment without primary
2 treatment. The organic component removed from anaerobic wastewater treatment and the amount of Cm
3 recovered or flared from both aerobic and anaerobic wastewater treatment were set equal to the IPCC default of
4 zero.
5 The methodological equations for Cm emissions from aerobic and anaerobic systems are:
6 Equation 7-8: Total Domestic ChU Emissions from Centrally Treated Aerobic Systems
7 Emissions from Centrally Treated Aerobic Systems (other than Constructed Wetlands) (Bl) + Emissions
8 from Centrally Treated Aerobic Systems (Constructed Wetlands Only) (B2) + Emissions from Centrally
9 Treated Aerobic Systems (Constructed Wetlands used as Tertiary Treatment) (B3) = B
10 where,
11 Equation 7-9: Total Organics in Centralized Wastewater Treatment [IPCC 2019 (Eq. 6.3A)]
12 TOWcentralized (Gg BOD/year) = TOW X Tcentralized X Icollected
13
14 Table 7-13: Variables and Data Sources for Organics in Centralized Domestic Wastewater
Variable
Variable Description Units
Inventory Years: Source of Value
Centrally Treated Organics (Gg BOD/year)
TOWcentrauzed
Total organics in centralized
wastewater treatment3
Gg BOD/year
1990-2021: Calculated
TOW
Total wastewater BOD Produced per
Capita3
Gg BOD/year
1990-2021: Calculated, ERG (2018a)
Tcentralized
Percent collected3
%
1990-2019: U.S. Census Bureau (2019)
Data for intervening years obtained by
linear interpolation
2020-2021: Forecasted from the rest
of the time series
Icollected
Correction factor for additional
industrial BOD discharged (1.25)
No units
1990-2021: IPCC (2019) Eq. 6.3a
15 a Value of this activity data varies over the time series.
16
17 Equation 7-10: Organic Component Removed from Aerobic Wastewater Treatment (IPCC
18 2019 [Eq. 6.3B])
19 Saerobic (Gg/year) = Smass x [(% aerobic w/primary x Krem,aer_Prim) + (% aerobic w/out primary x Krem,aer_noprim)
20 + (%aerobic+digestion x Krem,aer .digest)] X 1000
21
22 Equation 7-11: Emissions from Centrally Treated Aerobic Systems (other than Constructed
23 Wetlands) (IPCC 2019 [Eq. 6.1])
24 Bl (ktCH4/year)
25 = [(TOWcENTRALIZEd) X (% aerobiCOTCw) - Saerobic] X EFaerobic - Raerobic
26 Table 7-14: Variables and Data Sources for ChU Emissions from Centrally Treated Aerobic
27 Systems (Other than Constructed Wetlands)
Variable
Variable Description
Units
Inventory Years: Source of Value
Emissions from Centrally Treated Aerobic Systems (Other than Constructed Wetlands) (kt CH4/year)
Saerobic
Organic component removed from
aerobic wastewater treatment3
Gg
BOD/year
1990-2021: Calculated
Smass
Raw sludge removed from wastewater
treatment as dry mass3
Tg dry
weight/year
1988: EPA (1993c); EPA (1999)
7-28 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Variable
Variable Description
Units
Inventory Years: Source of Value
1990-1995: Calculated based on
sewage sludge production change
per year EPA (1993c); EPA (1999);
Beecher et al. (2007)
1996: EPA (1999)
2004: Beecher et al. (2007)
Data for intervening years obtained
by linear interpolation
2005-2017: Interpolated
2018: NEBRA (2022), as described in
ERG (2022)
2019-2021: Forecasted from the rest
of the time series.
Methodology for estimating sludge
generated from the U.S. territories
provided in ERG (2022).
% aerobicoTcw
Percent of flow to aerobic systems, other
than wetlands3
%
1990,1991: Set equal to 1992
1992, 1996, 2000, 2004: EPA (1992,
1996, 2000, 2004a), respectively
Data for intervening years obtained
by linear interpolation.
2005-2021: Forecasted from the rest
of the time series
% aerobic
w/primary
Percent of aerobic systems with primary
treatment and no anaerobic sludge
digestion (0)
%
% aerobic w/out
primary
Percent of aerobic systems without
primary treatment3
%
%aerobic+digestion
Percent of aerobic systems with primary
and anaerobic sludge digestion3
%
Krem,aer_prim
Sludge removal factor for aerobic
treatment plants with primary treatment
(mixed primary and secondary sludge,
untreated or treated aerobically) (0.8)
kg BOD/kg
sludge
1990-2021: IPCC (2019) Table 6.6a
Krem,aer_noprim
Sludge removal factor for aerobic
wastewater treatment plants without
separate primary treatment (1.16)
kg BOD/kg
sludge
Krem,aer_digest
Sludge removal factor for aerobic
treatment plants with primary treatment
and anaerobic sludge digestion (mixed
primary and secondary sludge, treated
anaerobically) (1)
kg BOD/kg
sludge
EF aerobic
Emission factor - aerobic systems (0.018)
kg CH4/kg
BOD
1990-2021: IPCC (2019) Table 6.3
Raerobic
Amount CH4 recovered or flared from
aerobic wastewater treatment (0)
kg CH4/year
1990-2021: IPCC (2019) Eq. 6.1
1000
Conversion factor
metric tons
to kilograms
Standard conversion
1 a Value of this activity data varies over the time series.
2 Constructed wetlands exhibit both aerobic and anaerobic treatment (partially anaerobic treatment) but are
3 referred to in this chapter as aerobic systems. Constructed wetlands may be used as the sole treatment unit at a
4 centralized wastewater treatment plant or may serve as tertiary treatment after simple settling and biological
5 treatment. Emissions from all constructed wetland systems were included in the estimates of emissions from
6 centralized wastewater treatment plant processes and effluent from these plants. Methane emissions equations
7 from constructed wetlands used as sole treatment were previously described. Methane emissions from
8 constructed wetlands used as tertiary treatment were estimated by multiplying the flow from treatment to
9 constructed wetlands, wastewater BOD concentration entering tertiary treatment, constructed wetlands emission
10 factor, and then converting to kt/year.
Waste 7-29
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
For constructed wetlands, an IPCC default emission factor for surface flow wetlands was used. This is the most
conservative factor for constructed wetlands and was recommended by IPCC (2014) when the type of constructed
wetland is not known. A median BODs concentration of 9.1 mg/Lwas used for wastewater entering constructed
wetlands used as tertiary treatment based on U.S. secondary treatment standards for POTWs. This median value is
based on plants generally utilizing simple settling and biological treatment (EPA 2013). Constructed wetlands do
not have secondary sludge removal.
Equation 7-12: Emissions from Centrally Treated Aerobic Systems (Constructed Wetlands
Only) [IPCC 2014 (Eq. 6.1)]
B2 (ktCH4/year)
= [(TOWcentralized) X (%aerobiccw)] X (EFcw)
Equation 7-13: Emissions from Centrally Treated Aerobic Systems (Constructed Wetlands
used as Tertiary Treatment) (U.S. Specific)
B3 (ktCH4/year)
= [(POTW_flow_CW) X (BODcwjnf) X 3.785 X (EFcw)] X 1/106 X 365.25
Table 7-15: Variables and Data Sources for ChU Emissions from Centrally Treated Aerobic
Systems (Constructed Wetlands)
Variable Variable Description
Units
Inventory Years: Source of Value
Emissions from Constructed Wetlands Only (kt CH4/year)
TOWcentralized
Total organics in centralized
wastewater treatment3
Gg
BOD/year
1990-2021: Calculated
% aerobiccw
Flow to aerobic systems,
constructed wetlands used as sole
treatment / total flow to POTWs.3
%
1990,1991: Set equal to 1992
1992, 1996, 2000, 2004, 2008, 2012:
EPA (1992, 1996, 2000, 2004a,
2008b, and 2012)
Data for intervening years obtained
by linear interpolation.
2013-2021: Forecasted from the rest
of the time series
EFcw
Emission factor for constructed
wetlands (0.24)
kg CH4/kg
BOD
1990-2021: IPCC (2014)
Emissions from Constructed Wetlands used as Tertiary Treatment (kt CH4/year)
POTW_flow_CW
Wastewater flow to POTWs that
use constructed wetlands as
tertiary treatment3
MGD
1990,1991: Set equal to 1992
1992, 1996, 2000, 2004, 2008, 2012:
EPA (1992, 1996, 2000, 2004a,
2008b, and 2012)
Data for intervening years obtained
by linear interpolation.
2013-2021: Forecasted from the rest
of the time series
BODcw.inf
BOD concentration in wastewater
entering the constructed wetland
(9.1)
mg/L
1990-2021: EPA (2013)
3.785
Conversion factor
liters to
gallons
Standard conversion
EFcw
Emission factor for constructed
wetlands (0.24)
kg CH4/kg
BOD
1990-2021: IPCC (2014)
1/106
Conversion factor
kg to kt
Standard conversion
365.25
Conversion factor
Days in a
year
Standard conversion
a Value of this activity data varies over the time series.
7-30 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Data sources and methodologies for centrally treated anaerobic systems are similar to those described for aerobic
2 systems, other than constructed wetlands. See discussion above.
3 Equation 7-14: Emissions from Centrally Treated Anaerobic Systems [IPCC 2019 (Eq. 6.1)]
4 C (ktCH4/year)
5 — [(TOWcENTRALIZEd) X (% anaerobic) — Sanaerobic] X EFanaerobic — Ranaerobic
6 Table 7-16: Variables and Data Sources for ChU Emissions from Centrally Treated Anaerobic
7 Systems
Variable
Variable Description
Units
Inventory Years: Source of Value
Emissions from Centrally Treated Anaerobic Systems (kt CH4/year)
TOWcentrauzed
Total organics in centralized
wastewater treatment3
Gg
BOD/year
1990-2021: Calculated
% anaerobic
Percent centralized wastewater that
is anaerobically treated3
%
1990,1991: Set equal to 1992
1992, 1996, 2000, 2004: EPA
(1992, 1996, 2000, 2004a),
respectively
Data for intervening years
obtained by linear interpolation.
2005-2021: Forecasted from the
rest of the time series
Sanaerobic
Organic component removed from
anaerobic wastewater treatment (0)
Gg/year
1990-2021: IPCC (2019) Table 6.3
EF anaerobic
Emission factor for anaerobic
reactors/deep lagoons (0.48)
kg CH4/kg
BOD
Ranaerobic
Amount CH4 recovered or flared
from anaerobic wastewater
treatment (0)
kg CH4/year
8 3 Value of this activity data varies over the time series.
9 Emissions from Anaerobic Sludge Digesters:
10 Total Cm emissions from anaerobic sludge digesters were estimated by multiplying the wastewater influent flow
11 to POTWs with anaerobic sludge digesters, the cubic feet of digester gas generated per person per day divided by
12 the flow to POTWs, the fraction of CH4 in biogas, the density of CH4, one minus the destruction efficiency from
13 burning the biogas in an energy/thermal device and then converting the results to kt/year.
14 Equation 7-15: Emissions from Anaerobic Sludge Digesters (U.S. Specific)
15 D (ktCH4/year)
16 = [(POTW_flow_AD) x (biogas gen)/(100)] x 0.0283 x (FRAC_CH4) x 365.25 x (662) x (1-DE) x 1/109
17 Table 7-17: Variables and Data Sources for Emissions from Anaerobic Sludge Digesters
Variable
Variable Description
Units
Inventory years: Source of
Value
Emissions from Anaerobic Sludge Digesters (kt CH4/year)
POTW_flow_AD
POTW Flow to Facilities with Anaerobic
Sludge Digesters3
MGD
1990,1991: Set equal to 1992
1992, 1996, 2000, 2004: EPA
(1992, 1996, 2000, and 2004a),
respectively
Data for intervening years
obtained by linear interpolation.
2005-2021: Forecasted from the
rest of the time series
Waste 7-31
-------
Variable
Variable Description
Units
Inventory years: Source of
Value
biogas gen
Gas Generation Rate (1.0)
ft3/ca pita/day
1990-2021: Metcalf & Eddy
(2014)
100
Per Capita POTW Flow (100)
gal/capita/day
1990-2021: Ten-State Standards
(2004)
0.0283
Conversion factor
ft3 to m3
Standard Conversion
frac_ch4
Proportion of Methane in Biogas (0.65)
No units
1990-2021: Metcalf & Eddy
(2014)
365.25
Conversion factor
Days in a year
Standard conversion
662
Density of Methane (662)
g CH4/m3 CH4
1990-2021: EPA (1993a)
DE
Destruction Efficiency (99% converted
to fraction)
No units
1990-2021: EPA (1998b); CAR
(2011); Sullivan (2007); Sullivan
(2010); and UNFCCC (2012)
1/109
Conversion factor
g to kt
Standard conversion
1 3 Value of this activity data varies over the time series.
2 Emissions from Discharge of Centralized Treatment Effluent:
3 Methane emissions from the discharge of wastewater treatment effluent were estimated by multiplying the total
4 BOD of the discharged wastewater effluent by an emission factor associated with the location of the discharge.
5 The BOD in treated effluent was determined by multiplying the total organics in centrally treated wastewater by
6 the percent of wastewater treated in primary, secondary, and tertiary treatment, and the fraction of organics
7 remaining after primary treatment (one minus the fraction of organics removed from primary treatment,
8 secondary treatment, and tertiary treatment).
9 Equation 7-16: Emissions from Centrally Treated Systems Discharge (U.S.-Specific)
10 E (kt CH4/year)
11 = (TOWrlE X EFrle) + (TOWother X EFother)
12 where,
13 Equation 7-17: Total Organics in Centralized Treatment Effluent (IPCC 2019 [Eq. 6.3D])
14 TOWEFFtreat.cENTRALizED (Gg BOD/year)
15 = [TOWcentralized X % primary X (l-TOWrem,PRiMARY)] + [TOWcentralized X % secondary X (1-
16 TOWrem,SECONDARY)] + [TOWCENTRALIZED X % tertiary X (l-TOWrem,TERTIARY)]
17 Equation 7-18: Total Organics in Effluent Discharged to Reservoirs, Lakes, or Estuaries
18 (U.S.-Specific)
19 TOWrle (Gg BOD/year)
20 = TOWEFFtreat.CENTRALIZED X PerCentRLE
21 Equation 7-19: Total Organics in Effluent Discharged to Other Waterbodies (U.S.-Specific)
22 TOWother (Gg BOD/year)
23 = TOWEFFtreat.CENTRALIZED X PerCentother
24 Table 7-18: Variables and Data Sources for ChU Emissions from Centrally Treated Systems
25 Discharge
Variable
Variable Description
Units
Source of Value
TOW EFFtreat.CENTRALIZED
Total organics in centralized treatment effluent3
Gg
BOD/year
1990-2021:
Calculated
TOWcentralized
Total organics in centralized wastewater treatment3
Gg
BOD/year
1990-2021:
Calculated
7-32 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Variable
Variable Description
Units
Source of Value
% primary
Percent of primary domestic centralized treatment3
%
1990,1991: Set
equal to 1992.
1992, 1996, 2000,
2004, 2008, 2012:
EPA (1992, 1996,
2000, 2004a, 2008,
and 2012),
respectively
Data for
intervening years
obtained by linear
interpolation.
2013-2021:
Forecasted from
the rest of the time
series
% secondary
Percent of secondary domestic centralized treatment3
%
% tertiary
Percent of tertiary domestic centralized treatment3
%
TOWrem.PRIMARY
Fraction of organics removed from primary domestic
centralized treatment (0.4)
No units
1990-2021: IPCC
(2019) Table 6.6B
TOWrem.SECONDARY
Fraction of organics removed from secondary domestic
centralized treatment (0.85)
No units
TOWrem.TERTIARY
Fraction of organics removed from tertiary domestic
centralized treatment (0.90)
No units
TOWrle
Total organics in effluent discharged to reservoirs, lakes, and
estuaries3
Gg
BOD/year
1990-2021:
Calculated
TOWother
Total organics in effluent discharge to other waterbodies3
Gg
BOD/year
EFrle
Emission factor (discharge to reservoirs/lakes/estuaries)
(0.114)
kg CH4/kg
BOD
1990-2021: IPCC
(2019) Table 6.8
EFother
Emission factor (discharge to other waterbodies) (0.021)
kg CH4/kg
BOD
PercentRLE
% discharged to reservoirs, lakes, and estuaries3
%
1990-2010: Set
equal to 2010
2010: ERG (2021a)
2011: Obtained by
linear interpolation
2012: ERG (2021a)
2013-2021: Set
equal to 2012
Percentother
% discharged to other waterbodies3
%
1 a Value of this activity data varies over the time series.
2 Industrial Wastewater CH4 Emission Estimates
3 Industrial wastewater Cm emissions originate from on-site treatment systems, typically comprised of biological
4 treatment operations. The collection systems at an industrial plant are not as extensive as domestic wastewater
5 sewer systems; therefore, it is not expected that dissolved Cm will form during collection. However, some
6 treatment systems are designed to have anaerobic activity (e.g., anaerobic reactors or lagoons), or may
7 periodically have anaerobic conditions form (facultative lagoons or large stabilization basins). Emissions will also
8 result from discharge of treated effluent to waterbodies where carbon accumulates in sediments (typically slow-
9 moving systems, such as lakes, reservoirs, and estuaries).
10 Industry categories that are likely to produce significant Cm emissions from wastewater treatment were identified
11 and included in the Inventory. The main criteria used to identify U.S. industries likely to generate Cm from
12 wastewater treatment are whether an industry generates high volumes of wastewater, whether there is a high
13 organic wastewater load, and whether the wastewater is treated using methods that result in Cm emissions. The
Waste 7-33
-------
1 top six industries that meet these criteria are pulp and paper manufacturing; meat and poultry processing;
2 vegetables, fruits, and juices processing; starch-based ethanol production; petroleum refining; and breweries.
3 Wastewater treatment and discharge emissions for these sectors for 2021 are displayed in Table 7-19 below.
4 Further discussion of wastewater treatment for each industry is included below.
5 Table 7-19: Total Industrial Wastewater ChU Emissions by Sector (2021, MMT CO2 Eq. and
6 Percent)
CH4 Emissions
% of Industrial
Industry
(MMT CO? Eq.)
Wastewater CH4
Meat & Poultry
5.7
78.9
Pulp & Paper
0.8
11.6
Fruit & Vegetables
0.2
3.3
Ethanol Refineries
0.2
2.3
Breweries
0.1
2.2
Petroleum Refineries
0.2
1.6
Total
7.2
100
Note: Totals may not sum due to independent rounding.
7 Emissions from Industrial Wastewater Treatment Systems:
8 Equation 7-20 presents the general IPCC equation (Equation 6.4, IPCC 2019) to estimate methane emissions from
9 each type of treatment system used for each industrial category.
10 Equation 7-20: Total ChU Emissions from Industrial Wastewater
11 CH4 (industrial sector) = [(TOWi - Si) x EF -Ri]
12 where,
13 Cm (industrial sector) = Total CH4 emissions from industrial sector wastewater treatment (kg/year)
14 i = Industrial sector
15 TOWi = Total organics in wastewater for industrial sector / (kg COD/year)
16 Si = Organic component removed from aerobic wastewater treatment for industrial
17 sector / (kg COD/year)
18 EF = System-specific emission factor (kg Cm/kg COD)
19 Ri = Methane recovered for industrial sector / (kg CFU/year)
20 Equation 7-21 presents the general IPCC equation to estimate the total organics in wastewater (TOW) for each
21 industrial category.
22 Equation 7-21: TOW in Industry Wastewater Treatment Systems
23 TOWi = Pi X Wi X CODi
24 where,
25
26 TOWi = Total organically degradable material in wastewater for industry / (kg COD/yr)
27 i = Industrial sector
28 Pi = Total industrial product for industrial sector / (t/yr)
29 Wi = Wastewater generated (m3/t product)
30 CODi = Chemical oxygen demand (industrial degradable organic component in wastewater) (kg
31 COD/m3)
32 The annual industry production is shown in Table 7-20, and the average wastewater outflow and the organics
33 loading in the outflow is shown in Table 7-21.
7-34 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
For some industries, U.S.-specific data on organics loading is reported as BOD rather than COD. In those cases, an
industry-specific COD:BOD ratio is used to convert the organics loading to COD.
The amount of organics treated in each type of wastewater treatment system was determined using the percent of
wastewater in the industry that is treated on site and whether the treatment system is anaerobic, aerobic or
partially anaerobic. Table 7-22 presents the industrial wastewater treatment activity data used in the calculations
and described in detail in ERG (2008a), ERG (2013a), ERG (2013b), and ERG (2021a). For Cm emissions, wastewater
treated in anaerobic lagoons or reactors was categorized as "anaerobic", wastewater treated in aerated
stabilization basins or facultative lagoons were classified as "ASB" (meaning there may be pockets of anaerobic
activity), and wastewater treated in aerobic systems such as activated sludge systems were classified as
"aerobic/other."
The amount of organic component removed from aerobic wastewater treatment as a result of sludge removal
(Saerobic) was either estimated as an industry-specific percent removal, if available, or as an estimate of sludge
produced by the treatment system and IPCC default factors for the amount of organic component removed (Krem),
using one of the following equations. Table 7-23 presents the sludge variables used for industries with aerobic
wastewater treatment operations (i.e., pulp and paper, fruit/vegetable processing, and petroleum refining).
Equation 7-22: Organic Component Removed from Aerobic Wastewater Treatment - Pulp,
Paper, and Pa per board
COD/yr)
% removal w/primary = Percent reduction of organics in pulp and paper wastewater associated with
sludge removal from primary treatment (%)
Equation 7-23: Organic Component Removed from Aerobic Treatment Plants
Spuip.asb = TOWpuip x % removal w/primary
where,
TOWpuip
Organic component removed from pulp and paper wastewater during primary
treatment before treatment in aerated stabilization basins (Gg COD/yr)
Total organically degradable material in pulp and paper wastewater (Gg
Saerobic — Smass X Krem xlO"6
where,
Sr
K,
Saerobic
10-6
'mass
.rem
Organic component removed from fruit and vegetable or petroleum refining wastewater
during primary treatment before treatment in aerated stabilization basins (Gg COD/yr)
Raw sludge removed from wastewater treatment as dry mass (kg sludge/yr)
Sludge factor (kg BOD/kg sludge)
Conversion factor, kilograms to Gigagrams
Raw sludge removed from wastewater treatment as dry mass (kg sludge/yr)
Sludge production from primary sedimentation (kg sludge/m3)
Sludge production from secondary aerobic treatment (kg sludge/m3)
Production (t/yr)
Wastewater Outflow (m3/t)
Sprim
Saer
P
w
Waste 7-35
-------
1 Default emission factors10 from IPCC (2019) were used. Information on methane recovery operations varied by
2 industry. See industry descriptions below.
3 Table 7-20: U.S. Pulp and Paper, Meat, Poultry, Vegetables, Fruits and Juices, Ethanol,
4 Breweries, and Petroleum Refining Production (MMT)
Meat
Poultry
Vegetables,
(Live Weight
(Live Weight
Fruits and
Ethanol
Petroleum
Year
Pulp and Paper3
Killed)
Killed)
Juices
Production
Breweries
Refining
1990
83.6
27.3
14.6
40.8
2.5
23.9
702.4
2005
92.4
31.4
25.1
45.3
11.7
23.1
818.6
2017
80.3
35.4
28.9
42.4
47.6
21.8
933.5
2018
78.7
36.4
29.4
42.3
48.1
21.5
951.7
2019
76.3
37.4
30.1
41.8
47.1
21.1
940.0
2020
74.7
37.8
30.5
40.6
41.6
21.1
806.5
2021
73.6
38.1
30.5
39.4
44.8
21.2
857.3
a Pulp and paper production is the sum of market pulp production plus paper and paperboard production.
Sources: Pulp and Paper - FAO (2022a) and FAO (2022b); Meat, Poultry, and Fruits and Vegetables - USDA (2022a
and 2022b), ERG (2022); Ethanol - Cooper (2018) and RFA (2022a and 2022b); Breweries - Beer Institute (2011)
and TTB (2022); Petroleum Refining - EIA (2022).
5 Table 7-21: U.S. Industrial Wastewater Characteristics Data (2021)
Industry
Wastewater
Wastewater
Wastewater
Outflow (m3/ton)
BOD (g/L)
COD (kg/m3)
COD:BOD Ratio
Pulp and Paper
See Table 7-25
0.3
-
2.5
Meat Processing
5.3
2.8
-
3
Poultry Processing
12.5
1.5
-
3
Fruit/Vegetable Processing
See Table 7-26
-
1.5
Ethanol Production - Wet Mill
10a
1.5
-
2
Ethanol Production - Dry Mill
1.25a
3b
-
2
Petroleum Refining
0.8
-
0.45
2.5
Breweries - Craft
3.09
-
17.6
1.67
Breweries - NonCraft
1.94
-
17.6
1.67
a Units are gallons per gallons ethanol produced.
b Units are COD (g/L).
Sources: Pulp and Paper (BOD, COD:BOD) - Malmberg (2018); Meat and Poultry (Outflow, BOD) - EPA (2002);
Meat and Poultry (COD:BOD) - EPA (1997a); Fruit/Vegetables (Outflow, BOD) - CAST (1995), EPA (1974), EPA
(1975); Fruit/Vegetables (COD:BOD) - EPA (1997a); Ethanol Production - Wet Mill (Outflow) - Donovan (1996),
NRBP (2001), Ruocco (2006a); Ethanol Production - Wet Mill (BOD) - White and Johnson (2003); Ethanol
Production - Dry Mill (Outflow and COD) - Merrick (1998), Ruocco (2006a); Ethanol Production (Dry and Wet,
COD:BOD) - EPA (1997a); Petroleum Refining (Outflow) - ERG (2013b); Petroleum Refining (COD) - Benyahia et
al. (2006); Petroleum Refining (COD:BOD) - EPA (1982); Breweries - Craft BIER (2017); ERG (2018b); Breweries -
NonCraft ERG (2018b); Brewers Association (2016a); Breweries (Craft and NonCraft; COD and COD:BOD) -
Brewers Association (2016b).
10 Emission factors are calculated by multiplying the maximum CH4-producing capacity of wastewater (B0, 0.25 kg CH4/kg COD)
and the appropriate methane correction factors (MCF) for aerobic (0), partially anaerobic (0.2), and anaerobic (0.8) systems
(IPCC 2019), Table 6.3.
7-36 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
l Table 7-22: U.S. Industrial Wastewater Treatment Activity Data
%
% Treated Aerobically
Industry
Wastewater
% Treated
% Treated
% Treated
in ASBs
% Treated
Treated On
Anaerobically
Aerobically
in Other
Site
Aerobic
Pulp and Paperb
60
5.2
75.9
38.5
37.4
Meat Processing
33
33a
33
0
33
Poultry Processing
25
25a
25
0
25
Fruit/Vegetable
Processing
11
0
11
5.5
5.5
Ethanol Production -
Wet Mill
33.3
33.3
66.7
0
0
Ethanol Production -
Dry Mill
75
75
25
0
0
Petroleum Refining
62.1
0
62.1
23.6
38.5
Breweries-Craft
0.5
0.5
0
0
0
Breweries - NonCraft
100
99
1
0
1
2 a Wastewater is pretreated in anaerobic lagoons prior to aerobic treatment.
3 b Remaining onsite treated in other treatment assumed to be non-emissive and not shown here.
4 Note: Due to differences in data availability and methodology, zero values in the table are for calculation purposes only and
5 may indicate unavailable data.
6 Sources: ERG (2008b); ERG (2013a); ERG (2013b); ERG (2021a).
7 Table 7-23: Sludge Variables for Aerobic Treatment Systems
Industry
Variable Pulp and Fruit/Vegetable Petroleum
Paper Processing Refining
Organic reduction associated with sludge removal (%) 58
Sludge Production (kg/m3)
Primary Sedimentation 0.15
Aerobic Treatment 0.096 0.096
Sludge Factor (kg BOD/kg dry mass sludge)
Aerobic Treatment w/Primary Sedimentation and No Anaerobic
Sludge Digestion 0.8
Aerobic Treatment w/out Primary Sedimentation 1.16
8 Sources: Organic reduction (pulp) - ERG (2008a); Sludge production - Metcalf & Eddy (2003); Sludge factors - IPCC (2019),
9 Table 6.6a.
10 Emissions from Discharge of Industrial Wastewater Treatment Effluent:
11 Methane emissions from discharge of industrial wastewater treatment effluent are estimated via a Tier 1 method
12 for all industries except for pulp, paper, and paperboard. Emissions from discharge of pulp, paper, and paperboard
13 treatment effluent is estimated via a Tier 2 method and is described in the industry-specific data section. Tier 1
14 emissions from effluent are estimated by multiplying the total organic content of the discharged wastewater
15 effluent by an emission factor associated with the discharge:
16 Equation 7-25: ChU Emissions from Industrial Wastewater Treatment Discharge
17 CH4 EffluentiND = TOWeffluent,ind X EFeffluent
18 where,
19 Cm EffluentiND = CH4 emissions from industrial wastewater discharge for inventory year (kg Cm/year)
20 TOWeffluent.ind = Total organically degradable material in wastewater effluent from industry for inventory
21 year (kg COD/year or kg BOD/year)
Waste 7-37
-------
1 EFeffluent = Tier 1 emission factor for wastewater discharged to aquatic environments (0.028 kg
2 Cm/kg COD or 0.068 kg CH4/kg BOD) (IPCC 2019)
3 The COD or BOD in industrial treated effluent (TOWeffluent.ind) was determined by multiplying the total organics in
4 the industry's untreated wastewater that is treated on site by an industry-specific percent removal where available
5 or a more general percent removal based on biological treatment for other industries. Table 7-22 presents the
6 percent of wastewater treated onsite, while Table 7-24 presents the fraction of TOW removed during treatment.
7 Equation 7-26: TOW in Industrial Wastewater Effluent
8 TOWeffluent.ind = TOWind * %onsite * (1 - TOWrem)
9 where,
10 TOWeffluent.ind
11
12 TOWind
13 %onsite
14 TOWrem
15
16 Table 7-24: Fraction of TOW Removed During Treatment by Industry
Industry
TOWrem
Source
Pulp, Paper, and Paperboard
0.905
Malmberg (2018)
Red Meat and Poultry
0.85
IPCC (2019), Table 6.6b
Fruits and Vegetables
0.85
IPCC (2019), Table 6.6b
Ethanol Production
Biomethanator T reatment
0.90
ERG (2008a), ERG (2006b)
Other Treatment
0.85
IPCC (2019), Table 6.6b
Petroleum Refining
0.93
Kenari, Sarrafzadeh, and Tavakoli (2010)
Breweries
0.85
IPCC (2019), Table 6.6b
17 Discussion of Industry-Specific Data:
18 Pulp, Paper, and Paperboard Manufacturing Wastewater Treatment. Wastewater treatment for the pulp, paper,
19 and paperboard manufacturing (hereinafter referred to as "pulp and paper") industry typically includes
20 neutralization, screening, sedimentation, and flotation/hydrocycloning to remove solids (World Bank 1999;
21 Nemerow and Dasgupta 1991). Secondary treatment (storage, settling, and biological treatment) mainly consists of
22 lagooning. About 60 percent of pulp and paper mills have on-site treatment with primary treatment and about half
23 of these also have secondary treatment (ERG 2008). In the United States, primary treatment is focused on solids
24 removal, equalization, neutralization, and color reduction (EPA 1993b). The vast majority of pulp and paper mills
25 with on-site treatment systems use mechanical clarifiers to remove suspended solids from the wastewater. About
26 10 percent of pulp and paper mills with treatment systems use settling ponds for primary treatment and these are
27 more likely to be located at mills that do not perform secondary treatment (EPA 1993b).
28 Approximately 42 percent of the BOD passes on to secondary treatment, which consists of activated sludge,
29 aerated stabilization basins, or non-aerated stabilization basins. Pulp and paper mill wastewater treated using
30 anaerobic ponds or lagoons or unaerated ponds were classified as anaerobic (with an MCF of 0.8). Wastewater
31 flow treated in systems with aerated stabilization basins or facultative lagoons was classified as partially anaerobic
32 (with an MCF of 0.2, which is the 2006 IPCC Guidelines-suggested MCF for shallow lagoons). Wastewater flow
33 treated in systems with activated sludge systems or similarly aerated biological systems was classified as aerobic.
34 A time series of CFU emissions for 1990 through 2021 was developed based on paper and paperboard production
35 data and market pulp production data. Market pulp production values were available directly for 1998, 2000
36 through 2003, and 2010 through 2020. Where market pulp data were unavailable, a percent of woodpulp that is
= Total organically degradable material in wastewater effluent from industry for inventory
year (kg COD/year or kg BOD/year)
= Total organics in untreated wastewater for industry for inventory year (kg COD/year)
= Percent of industry wastewater treated on site (%)
= Fraction of organics removed during treatment
7-38 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
market pulp was applied to woodpulp production values from FAOSTAT to estimate market pulp production (FAO
2022a). The percent of woodpulp that is market pulp for 1990 to 1997 was assumed to be the same as 1998,1999
was interpolated between values for 1998 and 2000, 2000 through 2009 were interpolated between values for
2003 and 2010, and 2021 was forecasted from the rest of the time series. A time series of the overall wastewater
outflow in units of cubic meters of wastewater per ton of total production (i.e., market pulp plus woodpulp) is
presented in Table 7-25. Data for 1990 through 1994 varies based on data outlined in ERG (2013a) to reflect
historical wastewater flow. Wastewater generation rates for 1995, 2000, and 2002 were estimated from the 2014
American Forest and Paper Association (AF&PA) Sustainability Report (AF&PA 2014). Wastewater generation rates
for 2004, 2006, 2008, 2010, 2012, and 2014 were estimated from the 2016 AF&PA Sustainability Report (AF&PA
2016). Data for 2005 and 2016 were obtained from the 2018 AF&PA Sustainability Report (AF&PA 2018), data for
2018 were obtained from the 2020 AF&PA Sustainability Report (AF&PA 2020), and data for 2020 were obtained
from a 2022 AF&PA sustainability update (AF&PA 2022). Data for intervening years were obtained by linear
interpolation, while 2021 was set equal to 2020. The average BOD concentration in raw wastewater was estimated
to be 0.4 grams BOD/liter for 1990 to 1998, while 0.3 grams BOD/liter was estimated for 2014 through 2021 (EPA
1997b; EPA 1993b; World Bank 1999; Malmberg 2018). Data for intervening years were obtained by linear
interpolation.
Table 7-25: Wastewater Outflow (m3/ton) for Pulp, Paper, and Paperboard Mills
Wastewater Outflow
Year (m3/ton)
1990
68
2005
43
2017
39
2018
40
2019
39
2020
39
2021
39
Sources: ERG (2013a), AF&PA (2014),
AF&PA (2016), AF&PA (2018), AF&PA
(2020); AF&PA (2022)
Pulp, Paper, and Paperboard Wastewater Treatment Effluent. Methane emissions from pulp, paper, and
paperboard wastewater treatment effluent were estimated by multiplying the total BOD of the discharged
wastewater effluent by an emission factor associated with the location of the discharge.
Equation 7-27: Emissions from Pulp and Paper Discharge (U.S. Specific)
Emissions from Pulp and Paper Discharge (U.S. Specific, kt CH4/year)
— (TOWRLE.pulp X EFrle) "I" (TOWother,pulp X EFother)
Equation 7-28: Total Organics in Pulp and Paper Effluent Discharged to Reservoirs, Lakes, Or
Estuaries (U.S. Specific)
T0Wrle,Puip (Gg BOD/year)
= TOWeffluent,ind X PercentRLE.puip
Equation 7-29: Total Organics in Pulp and Paper Effluent Discharged to Other Waterbodies
(U.S. Specific)
TOWother.puip (Gg BOD/year)
= TOWeffluent.ind X Percentother.puip
Waste 7-39
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
where,
TOWrle,Puip = Total organics in pulp, paper, and paperboard manufacturing wastewater treatment
effluent discharged to reservoirs, lakes, or estuaries (Gg BOD/year)
EFrle = Emission factor (discharge to reservoirs/lakes/estuaries) (0.114 kg Cm/kg BOD) (IPCC
2019)
TOWother.puip = Total organics in pulp, paper, and paperboard manufacturing wastewater treatment
effluent discharged to other waterbodies (Gg BOD/year)
EFother = Emission factor (discharge to other waterbodies) (0.021 kg Cm/kg BOD) (IPCC 2019)
TOWeffluent.ind = Total organically degradable material in pulp, paper, and paperboard manufacturing
wastewater effluent for inventory year (Gg BOD/year)
PercentRLE.puip = Percent of wastewater effluent discharged to reservoirs, lakes, and estuaries (ERG
2021b)
Percentother.puip = Percent of wastewater effluent discharged to other waterbodies (ERG 2021b)
The percent of pulp, paper, and paperboard wastewater treatment effluent routed to reservoirs, lakes, or
estuaries (3 percent) and other waterbodies (97 percent) were obtained from discussions with NCASI (ERG 2021b).
Data for 2019 were assumed the same as the rest of the time series due to lack of available data. Default emission
factors for reservoirs, lakes, and estuaries (0.114 kg Cm/kg BOD) and other waterbodies (0.021 kg CFU/kg BOD)
were obtained from IPCC (2019).
Meat and Poultry Processing. The meat and poultry processing industry makes extensive use of anaerobic lagoons
in sequence with screening, fat traps, and dissolved air flotation when treating wastewater on site. Although all
meat and poultry processing facilities conduct some sort of treatment on site, about 33 percent of meat processing
operations (EPA 2002) and 25 percent of poultry processing operations (U.S. Poultry 2006) perform on-site
treatment in anaerobic lagoons. The IPCC default emission factor of 0.2 kg Cm/kg COD for anaerobic lagoons were
used to estimate the Cm produced from these on-site treatment systems.
Vegetables, Fruits, and Juices Processing. Treatment of wastewater from fruits, vegetables, and juices processing
includes screening, coagulation/settling, and biological treatment (lagooning). The flows are frequently seasonal,
and robust treatment systems are preferred for on-site treatment. About half of the operations that treat and
discharge wastewater use lagoons intended for aerobic operation, but the large seasonal loadings may develop
limited anaerobic zones. In addition, some anaerobic lagoons may also be used (Nemerow and Dasgupta 1991).
Wastewater treated in partially anaerobic systems were assigned the IPCC default emission factor of 0.12 kg
Cm/kg BOD. Outflow and BOD data, presented in Table 7-26, were obtained from CAST (1995) for apples, apricots,
asparagus, broccoli, carrots, cauliflower, cucumbers (for pickles), green peas, pineapples, snap beans, and spinach;
EPA (1974) for potato and citrus fruit processing; and EPA (1975) for all other commodities.
Table 7-26: Wastewater Outflow (m3/ton) and BOD Production (g/L) for U.S. Vegetables,
Fruits, and Juices Production
Wastewater Outflow Organic Content in Untreated
Commodity (mB/ton) Wastewater (g BOD/L)
Vegetables
Potatoes 10.27 1.765
Other Vegetables 9.85 0.751
Fruit
Apples 9.08 8.16
Citrus Fruits 10.11 0.317
Non-citrus Fruits 12.59 1.226
Grapes (for wine) 2.78 1.831
Sources: CAST (1995); EPA (1974); EPA (1975).
7-40 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Ethanol Production. Ethanol, or ethyl alcohol, is produced primarily for use as a fuel component, but is also used in
industrial applications and in the manufacture of beverage alcohol. Ethanol can be produced from the
fermentation of sugar-based feedstocks (e.g., molasses and beets), starch- or grain-based feedstocks (e.g., corn,
sorghum, and beverage waste), and cellulosic biomass feedstocks (e.g., agricultural wastes, wood, and bagasse).
Ethanol can also be produced synthetically from ethylene or hydrogen and carbon monoxide. However, synthetic
ethanol comprises a very small percent of ethanol production in the United States. Currently, ethanol is mostly
made from sugar and starch crops, but with advances in technology, cellulosic biomass is increasingly used as
ethanol feedstock (DOE 2013).
Ethanol is produced from corn (or other sugar or starch-based feedstocks) primarily by two methods: wet milling
and dry milling. Historically, the majority of ethanol was produced by the wet milling process, but now the majority
is produced by the dry milling process. The dry milling process is cheaper to implement and is more efficient in
terms of actual ethanol production (Rendleman and Shapouri 2007). The wastewater generated at ethanol
production facilities is handled in a variety of ways. Dry milling facilities often combine the resulting evaporator
condensate with other process wastewaters, such as equipment wash water, scrubber water, and boiler blowdown
and anaerobically treat this wastewater using various types of digesters. Wet milling facilities often treat their
steepwater condensate in anaerobic systems followed by aerobic polishing systems. Wet milling facilities may treat
the stillage (or processed stillage) from the ethanol fermentation/distillation process separately or together with
steepwater and/or wash water. Methane generated in anaerobic sludge digesters is commonly collected and
either flared or used as fuel in the ethanol production process (ERG 2006b).
About 33 percent of wet milling facilities and 75 percent of dry milling facilities treat their wastewater
anaerobically. A default emission factor of 0.2 kg CHVkg COD for anaerobic treatment was used to estimate the
CH4 produced from these on-site treatment systems. The amount of CH4 recovered through the use of
biomethanators was estimated, and a 99 percent destruction efficiency was used. Biomethanators are anaerobic
reactors that use microorganisms under anaerobic conditions to reduce COD and organic acids and recover biogas
from wastewater (ERG 2006b). For facilities using biomethanators, approximately 90 percent of BOD is removed
during on-site treatment (ERG 2006b, 2008). For all other facilities, the removal of organics was assumed to be
equivalent to secondary treatment systems, or 85 percent (IPCC 2019).
Petroleum Refining. Petroleum refining wastewater treatment operations have the potential to produce CH4
emissions from anaerobic wastewater treatment. EPA's Office of Air and Radiation performed an Information
Collection Request (ICR) for petroleum refineries in 2011.11 Facilities that reported using non-aerated surface
impoundments or other biological treatment units (trickling filter, rotating biological contactor), which have the
potential to lead to anaerobic conditions, were assigned the IPCC default emission factor of 0.05 kg CH4/kg COD. In
addition, the wastewater generation rate was determined to be 26.4 gallons per barrel of finished product, or 0.8
m3/ton (ERG 2013b).
Breweries. Since 2010, the number of breweries has increased from less than 2,000 to more than 8,000 (Brewers
Association 2021). This increase has primarily been driven by craft breweries, which have increased by over 250
percent during that period. Craft breweries were defined as breweries producing less than six million barrels of
beer per year, and non-craft breweries produce greater than six million barrels. With their large amount of water
use and high strength wastewater, breweries generate considerable CH4 emissions from anaerobic wastewater
treatment. However, because many breweries recover their CH4, their emissions are much lower.
The Alcohol and Tobacco Tax and Trade Bureau (TTB) provides total beer production in barrels per year for
different facility size categories from 2007 to the present (TTB 2022). For years prior to 2007 where TTB data were
not readily available, the Brewers Almanac (Beer Institute 2011) was used, along with an estimated percent of craft
and non-craft breweries based on the breakdown of craft and non-craft for the years 2007 through 2020.
11 Available online at https://www.epa.gov/stationarv-sources-air-pollution/comprehensive-data-collected-petroleum-refining-
sector.
Waste 7-41
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
To determine the overall amount of wastewater produced, data on water use per unit of production and a
wastewater-to-water ratio were used from the Benchmarking Report (Brewers Association 2016a) for both craft
and non-craft breweries. Since brewing is a batch process, and different operations have varying organic loads,
full-strength brewery wastewater can vary widely on a day-to-day basis. However, the organic content of brewery
wastewater does not substantially change between craft and non-craft breweries. Some breweries may collect and
discharge high strength wastewater from particular brewing processes (known as "side streaming") to a POTW,
greatly reducing the organics content of the wastewater that is treated on site. Subsequently, the MCF for
discharge to a POTW was assumed to be zero (ERG 2018b).
Breweries may treat some or all of their wastewater on site prior to discharge to a POTW or receiving water. On-
site treatment operations can include physical treatment (e.g., screening, settling) which are not expected to
contribute to Cm emissions, or biological treatment, which may include aerobic treatment or pretreatment in
anaerobic reactors (ERG 2018b). The IPCC default emission factor of 0.2 kg Cm/kg COD for anaerobic treatment
and 0 for aerobic treatment were used to estimate the Cm produced from these on-site treatment systems (IPCC
2006). The amount of Cm recovered through anaerobic wastewater treatment was estimated, and a 99 percent
destruction efficiency was used (ERG 2018b; Stier J. 2018). Very limited activity data are available on the number
of U.S. breweries that are performing side streaming or pretreatment of wastewater prior to discharge.
Domestic Wastewater N2O Emission Estimates
Domestic wastewater N2O emissions originate from both septic systems and POTWs. Within these centralized
systems, N2O emissions can result from aerobic systems, including systems like constructed wetlands. Emissions
will also result from discharge of centrally treated wastewater to waterbodies with nutrient-impacted/eutrophic
conditions. The systems with emission estimates are:
• Septic systems (A);
• Centralized treatment aerobic systems (B), including aerobic systems (other than constructed wetlands)
(Bl), constructed wetlands only (B2), and constructed wetlands used as tertiary treatment (B3);
• Centralized anaerobic systems (C); and
• Centralized wastewater treatment effluent (D).
Methodological equations for each of these systems are presented in the subsequent subsections; total domestic
N2O emissions are estimated as follows:
Equation 7-30: Total Domestic N2O Emissions from Wastewater Treatment and Discharge
Total Domestic N2O Emissions from Wastewater Treatment and Discharge (kt) = A+ B + C + D
Table 7-27 presents domestic wastewater N2O emissions for both septic and centralized systems, including
emissions from centralized wastewater treatment effluent, in 2021.
Table 7-27: Domestic Wastewater N2O Emissions from Septic and Centralized Systems
(2021, kt, MMT CO2 Eq. and Percent)
N20 Emissions (kt)
N20 Emissions
(MMT CO? Eq.)
% of Domestic
Wastewater N20
Septic Systems
3
0.8
3.8
Centrally-Treated Aerobic Systems
58
15.4
75.5
Centrally-Treated Anaerobic Systems
+
+
+
Centrally-Treated Wastewater Effluent
16
4.2
20.7
Total
77
20.4
100
+ Does not exceed 0.5 kt or 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
7-42 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Emissions from Septic Systems:
Nitrous oxide emissions from domestic treatment depend on the nitrogen present, in this case, in the form of
protein. Per capita protein consumption (kg protein/person/year) was determined by multiplying per capita annual
food availability data and its protein content. Those data are then adjusted using a factor to account for the
fraction of protein actually consumed. The methodological equations are:
Equation 7-31: Annual per Capita Protein Supply (U.S. Specific)
ProteinsuppLY (kg/person/year)
= ProteinPercapita/1000 x 365.25
Equation 7-32: Consumed Protein [IPCC 2019 (Eq. 6.10A)]
Protein (kg/person/year)
= ProteinsuppLY x FPC
Table 7-28: Variables and Data Sources for Protein Consumed
Variable
Variable Description
Units
Inventory Years: Source of
Value
Protein
ProteinsuppLY
Annual per capita protein supply3
kg/person/year
1990-2021: Calculated
PrOteinper capita
Daily per capita protein supply3
g/person/day
1990-2021: USDA (2021b)
1000
Conversion factor
gto kg
Standard conversion
365.25
Conversion factor
Days in a year
Standard conversion
FPC
Fraction of Protein Consumed3
kg protein
consumed / kg
protein available
1990-2010: USDA (2021)
2011-2019: FAO (2022c)
and scaling factor
2020-2021: Forecasted
from the rest of the time
series
a Value of this activity data varies over the Inventory time series.
Nitrous oxide emissions from septic systems were estimated by multiplying the U.S. population by the percent of
wastewater treated in septic systems (about 17 percent in 2021; U.S. Census Bureau 2019), consumed protein per
capita (kg protein/person/year), the fraction of N in protein, the correction factor for additional nitrogen from
household products, the factor for industrial and commercial co-discharged protein into septic systems, the factor
for non-consumed protein added to wastewater and an emission factor and then converting the result to kt/year.
All factors obtained from IPCC (2019).
U.S. population data were taken from historic U.S. Census Bureau national population totals data and include the
populations of the United States and Puerto Rico (U.S. Census Bureau 2002; U.S. Census Bureau 2011; U.S. Census
Bureau 2021a and 2021b, Instituto de Estadisticas de Puerto Rico 2021). Population data for American Samoa,
Guam, Northern Mariana Islands, and the U.S. Virgin Islands were taken from the U.S. Census Bureau International
Database (U.S. Census Bureau 2022). Table 7-12 presents the total U.S. population for 1990 through 2021. The
fraction of the U.S. population using septic systems, as well as centralized treatment systems (see below), is based
on data from American Housing Survey (U.S. Census Bureau 2019). The methodological equations are:
Equation 7-33: Total Nitrogen Entering Septic Systems (IPCC 2019 [Eq. 6.10])
TNdom.septic (kg N/year)
= (USPOP X TsEPTIc) X Protein X FnPR X NhH X FNON-CON_septic X FlND-COM_septic
Waste 7-43
-------
1 Equation 7-34: Emissions from Septic Systems (IPCC 2019 [Eq. 6.9])
2 A (kt N20/year)
3 = TNdom. septic X (EFseptic) X 44/28 X 1/106
4 Table 7-29: Variables and Data Sources for N2O Emissions from Septic System
Variable
Variable Description
Units
Inventory Years: Source of
Value
Emissions from Septic Systems
TNdom septic
Total nitrogen entering septic systems
kg N/year
1990-2021: Calculated
USpop
U.S. population3
Persons
United States and Puerto
Rico:
1990-1999: US Census
Bureau 2002; Instituto de
Estadisticas de Puerto Rico
2021
2000-2009: U.S. Census
Bureau 2011
2010-2019: U.S. Census
Bureau (2021a)
2020-2021: U.S. Census
Bureau (2021b)
U.S. Territories other than
Puerto Rico:
1990-2021: U.S. Census
Bureau (2022)
Tseptic
Percent treated in septic systems3
%
Odd years from 1989
through 2019: U.S. Census
Bureau (2019)
Data for intervening years
obtained by linear
interpolation
2020-2021: Forecasted
from the rest of the time
series
Fnpr
Fraction of nitrogen in protein (0.16)
kg N/kg protein
1990-2021: IPCC (2019) Eq.
6.10
Nhh
Additional nitrogen from household products (1.17)
No units
1990-2021: IPCC (2019)
Table 6.10a
FNON-CON_septic
Factor for Non-Consumed Protein Added to
Wastewater (1.13)
No units
F|ND-COM_septic
Factor for Industrial and Commercial Co-Discharged
Protein, septic systems (1)
No units
1990-2021: IPCC (2019)
EFseptic
Emission factor, septic systems (0.0045)
kg N20-N/kg N
1990-2021: IPCC (2019)
Table 6.8a
44/28
Conversion factor
Molecular
weight ratio of
N20 to N2
Standard conversion
1/106
Conversion factor
kg to kt
Standard conversion
5 a Value of this activity data varies over the Inventory time series.
6 Emissions from Centrally Treated Aerobic and Anaerobic Systems:
7 Nitrous oxide emissions from POTWs depend on the total nitrogen entering centralized wastewater treatment. The
8 total nitrogen entering centralized wastewater treatment was estimated by multiplying the U.S. population by the
9 percent of wastewater collected for centralized treatment (about 83 percent in 2021), the consumed protein per
10 capita, the fraction of N in protein, the correction factor for additional N from household products, the factor for
7-44 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 industrial and commercial co-discharged protein into wastewater treatment, and the factor for non-consumed
2 protein added to wastewater.
3 Equation 7-35: Total Nitrogen Entering Centralized Systems (IPCC 2019 [Eq. 10])
4 TNdom.central (kg N/year)
5 = (USpop X Tcentralized) X Protein X Fnpr X Nhh X Fnon-con X Find-com
6 Table 7-30: Variables and Data Sources for Non-Consumed Protein and Nitrogen Entering
7 Centralized Systems
Inventory Years: Source
Variable
Variable Description
Units
of Value
United States and
Puerto Rico:
1990-1999: U.S. Census
Bureau (2002); Instituto
de Estadisticas de
Puerto Rico (2021)
2000-2009: U.S. Census
USpop
U.S. population3
Persons
Bureau 2011
2010-2019: U.S. Census
Bureau (2021a)
2020-2021: U.S. Census
Bureau (2021b)
U.S. Territories other
than Puerto Rico:
1990-2021: U.S. Census
Bureau (2022)
Odd years from 1989
through 2019: U.S.
Census Bureau (2019)
Data for intervening
Tcentralized
Percent collected3
%
years obtained by linear
interpolation
2020-2021: Forecasted
from the rest of the
time series
Protein
Consumed protein per capita3
kg/person/year
1990-2021: Calculated
1990-2021: IPCC (2019),
Fnpr
Fraction of nitrogen in protein (0.16)
kg N/kg protein
Eq. 6.10
Factor for additional nitrogen from household
Nhh
products (1.17)
No units
1990-2021: IPCC (2019),
Fnon-con
Factor for U.S. specific non-consumed protein
(1.13)
No units
Table 6.10a
Factor for Industrial and Commercial Co-
No units
1990-2021: IPCC (2019)
Find-com
Discharged Protein (1.25)
Table 6.11
8 a Value of this activity data varies over the Inventory time series.
9 Nitrous oxide emissions from POTWs were estimated by multiplying the total nitrogen entering centralized
10 wastewater treatment, the relative percentage of wastewater treated by aerobic systems (other than constructed
11 wetlands) and anaerobic systems, aerobic systems with constructed wetlands as the sole treatment, the emission
12 factor for aerobic systems and anaerobic systems, and the conversion from l\h to N2O.
13 Table 7-34 presents the data for U.S. population, population served by centralized wastewater treatment plants,
14 available protein, and protein consumed. The methodological equations are:
Waste 7-45
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Equation 7-36: Total Domestic N2O Emissions from Centrally Treated Aerobic Systems
Emissions from Centrally Treated Aerobic Systems (other than Constructed Wetlands) (Bl) + Emissions
from Centrally Treated Aerobic Systems (Constructed Wetlands Only) (B2) + Emissions from Centrally
Treated Aerobic Systems (Constructed Wetlands used as Tertiary Treatment) (B3) = B (ktN20/year)
where,
Equation 7-37: Emissions from Centrally Treated Aerobic Systems (other than Constructed
Wetlands) (IPCC 2019 [Eq. 6.9])
Bl (kt N20/year)
= [(TNdom.central) X (% aerobicoTcw)] X EFaerobic X 44/28 X 1/106
Table 7-31: Variables and Data Sources for N2O Emissions from Centrally Treated Aerobic
Systems (Other than Constructed Wetlands)
Variable
Variable Description
Units
Inventory Years: Source
of Value
Emissions from Centrally Treated Aerobic Systems (Other than Constructed Wetlands) (kt N20/year)
TNdom central
Total nitrogen entering centralized systems3
kg N/year
1990-2021: Calculated
% aerobicoTcw
Flow to aerobic systems, other than constructed
wetlands only / total flow to POTWsa
%
1990,1991: Set equal to
1992
1992, 1996, 2000, 2004:
EPA (1992, 1996, 2000,
2004a), respectively
Data for intervening
years obtained by linear
interpolation.
2005-2021: Forecasted
from the rest of the time
series
EF aerobic
U.S.-specific emission factor - aerobic systems
(0.015)
kg N20-N/kg N
1990-2021: IPCC (2022)
44/28
Conversion factor
Molecular
weight ratio of
N20 to N2
Standard conversion
1/106
Conversion factor
kg to kt
Standard conversion
a Value of this activity data varies over the Inventory time series.
Nitrous oxide emissions from constructed wetlands used as sole treatment include similar data and processes as
aerobic systems other than constructed wetlands. See description above. Nitrous oxide emissions from
constructed wetlands used as tertiary treatment were estimated by multiplying the flow to constructed wetlands
used as tertiary treatment, wastewater N concentration entering tertiary treatment, constructed wetlands
emission factor, and converting to kt/year.
Equation 7-38: Emissions from Centrally Treated Aerobic Systems (Constructed Wetlands
Only) (IPCC 2014 [Eq. 6.9])
B2 (ktN20/year)
= [(TNdom.central) X (%aerobiccw)] X EFcw X 44/28 X 1/106
Equation 7-39: Emissions from Centrally Treated Aerobic Systems (Constructed Wetlands
used as Tertiary Treatment) (U.S.-Specific)
B3 (kt N20/year)
= [(POTW_flow_CW) X (Ncw.inf) X 3.785 X (EFcw)] X 1/106 X 365.25
7-46 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Table 7-32: Variables and Data Sources for N2O Emissions from Centrally Treated Aerobic
2 Systems (Constructed Wetlands)
Variable
Variable Description
Units
Inventory Years: Source of
Value
Emissions from Constructed Wetlands Only (kt N20/year)
TNdom central
Total nitrogen entering centralized treatment3
kg N/year
1990-2021: Calculated
% aerobiccw
Flow to aerobic systems, constructed wetlands
used as sole treatment / total flow to POTWsa
%
1990,1991: Set equal to
1992
1992, 1996, 2000, 2004,
2008, 2012: EPA (1992,
1996, 2000, 2004a, 2008b,
and 2012)
Data for intervening years
obtained by linear
interpolation.
2013-2021: Forecasted
from the rest of the time
series
EFcw
Emission factor for constructed wetlands
(0.0013)
kg N20-N/kg N
1990-2021: IPCC (2014)
Table 6.7
44/28
Conversion factor
Molecular
weight ratio of
N20 to N2
Standard conversion
1/106
Conversion factor
kg to kt
Standard conversion
Emissions from Constructed Wetlands used as Tertiary Treatment (kt N20/year)
POTW_flow_CW
Wastewater flow to POTWs that use constructed
wetlands as tertiary treatmenta
MGD
1990,1991: Set equal to
1992
1992, 1996, 2000, 2004,
2008, 2012: EPA (1992,
1996, 2000, 2004a, 2008b,
and 2012)
Data for intervening years
obtained by linear
interpolation.
2013-2021: Forecasted
from the rest of the time
series
Ncw.inf
BOD concentration in wastewater entering the
constructed wetland (25)
mg/L
1990-2021: Metcalf & Eddy
(2014)
3.785
Conversion factor
liters to gallons
Standard conversion
EFcw
Emission factor for constructed wetlands
(0.0013)
kg N20-N/kg N
1990-2021: IPCC (2014)
Table 6.7
1/106
Conversion factor
mg to kg
Standard conversion
365.25
Conversion factor
Days in a year
Standard conversion
3 a Value of this activity data varies over the Inventory time series.
4 Data sources and methodologies are similar to those described for aerobic systems, other than constructed
5 wetlands. See discussion above.
6 Equation 7-40: Emissions from Centrally Treated Anaerobic Systems (IPCC 2019 [Eq. 6.9])
7 C (kt N20/year)
8 = [(TNdom.central) X (% anaerobic)] X EFanaerobic X 44/28 X 1/106
Waste 7-47
-------
1 Table 7-33: Variables and Data Sources for N2O Emissions from Centrally Treated Anaerobic
2 Systems
Variable
Variable Description
Units
Inventory Years: Source of
Value
Emissions from Centrally Treated Anaerobic Systems
TNdom_central
Total nitrogen entering centralized
treatment3
kg N/year
1990-2021: Calculated
% anaerobic
Percent centralized wastewater that
is anaerobically treated3
%
1990,1991: Set equal to 1992
1992, 1996, 2000, 2004: (EPA
1992, 1996, 2000, 2004a),
respectively
Data for intervening years
obtained by linear
interpolation.
2005-2021: Forecasted from
the rest of the time series
EF anaerobic
Emission factor for anaerobic
reactors/deep lagoons (0)
kg N20-N/kg N
1990-2021: IPCC (2019) Table
6.8A
44/28
Conversion factor
Molecular weight
ratio of N20 to N2
Standard conversion
1/106
Conversion factor
mg to kg
Standard conversion
3 3 Value of this activity data varies over the Inventory time series.
4
5 Table 7-34: U.S. Population (Millions) Fraction of Population Served by Centralized
6 Wastewater Treatment (percent), Protein Supply (kg/person-year), and Protein Consumed
7 (kg/person-year)
Year
Population
Centralized WWT
Population (%)
Protein Supply
Protein Consumed
1990
253
75.6
43.1
33.2
2005
300
78.8
44.9
34.7
2017
329
82.1
44.7
34.5
2018
330
82.9
45.5
35.1
2019
332
83.6
45.4
35.0
2020
335
82.7
44.6
34.4
2021
336
83.0
44.6
34.4
Sources: Population - U.S. Census Bureau 2002; U.S. Census Bureau 2011; U.S.
Census Bureau (2021a and 2021b); Instituto de Estadisticas de Puerto Rico (2021);
U.S. Census Bureau (2022); WWTP Population - U.S. Census Bureau (2019);
Available Protein - USDA (2021), FAO (2022c); Protein Consumed - FAO (2022c).
8 Emissions from Discharge of Centralized Treatment Effluent:
9 Nitrous oxide emissions from the discharge of wastewater treatment effluent were estimated by multiplying the
10 total nitrogen in centrally treated wastewater effluent by the percent of wastewater treated in primary,
11 secondary, and tertiary treatment and the fraction of nitrogen remaining after primary, secondary, or tertiary
12 treatment and then multiplying by the percent of wastewater volume routed to waterbodies with nutrient-
13 impaired/eutrophic conditions and all other waterbodies (ERG 2021a) and emission factors for discharge to
14 impaired waterbodies and other waterbodies from IPCC (2019). The methodological equations are:
15
7-48 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Equation 7-41: Emissions from Centrally Treated Systems Discharge (U.S.-Specific)
2 D (kt N20/year)
3 = [(Neffluent.imp X EFimp) + (Nefluent,nonimp X EFnonimp)] X 44/28 X 1/106
4 where,
5 Equation 7-42: Total Organics in Centralized Treatment Effluent (IPCC 2019 [Eq. 6.8])
6 Neffulent.dom (kg N/year)
7 = [TNdom.central12 X % primary X (l-Nrem,PRiMARY)] + [TNdom.central X % secondary X (l-Nrem,sEcoNDARY)] +
8 [TNdOM.CENTRAL X % tertiary X (l-Nrem,TERTIARY)]
9 Equation 7-43: Total Nitrogen in Effluent Discharged to Impaired Waterbodies (U.S.-
10 Specific)
11 Neffluent.imp (kg N/year)
12 = (Neffulent.dom X PercentiMp)/1000
13 Equation 7-44: Total Nitrogen in Effluent Discharged to Nonimpaired Waterbodies (U.S.-
14 Specific)
15 Neffluent.nonimp (kg N year)
16 = (Neffluent.dom X PercentNONiMp)/1000
17 Table 7-35: Variables and Data Sources for N2O Emissions from Centrally Treated Systems
18 Discharge
Variable
Variable Description
Units
Source of Value
Neffulent.dom
Total organics in centralized treatment effluent3
kg N/year
1990-2021: Calculated
44/28
Conversion factor
Molecular
weight ratio of
N20 to N2
Standard conversion
1/106
Conversion factor
kg to kt
Standard conversion
TNdom central
Total nitrogen entering centralized treatment3
kg N/year
1990-2021: Calculated
1000
Conversion factor
kg to kt
Standard Conversion
% primary
Percent of primary domestic centralized treatment3
%
1990,1991: Set equal to
1992.
1992, 1996, 2000,
2004, 2008, 2012: EPA
(1992, 1996, 2000,
2004a, 2008, and
2012), respectively
Data for intervening
years obtained by
linear interpolation.
2013-2021: Forecasted
from the rest of the
time series
% secondary
Percent of secondary domestic centralized treatment3
%
% tertiary
Percent of tertiary domestic centralized treatment3
%
Nrem.PRIMARY
Fraction of nitrogen removed from primary domestic
centralized treatment (0.1)
No units
1990-2021: IPCC (2019)
Table 6.10c
Nrem.SECONDARY
Fraction of nitrogen removed from secondary domestic
centralized treatment (0.4)
No units
Nrem.TERTIARY
Fraction of nitrogen removed from tertiary domestic
centralized treatment (0.9)
No units
12 See emissions from centrally treated aerobic and anaerobic systems for methodological equation calculating TNDom_central.
Waste 7-49
-------
Variable
Variable Description
Units
Source of Value
Neffluent.imp
Total nitrogen in effluent discharged to impaired waterbodies
kg N/year
1990-2021: Calculated
Neffluent.nonimp
Total nitrogen in effluent discharged to nonimpaired
waterbodies
kg N/year
EFimp
Emission factor (discharge to impaired waterbodies) (0.19)
kg N20-N/kg N
1990-2021: IPCC (2019)
Table 6.8a
EF|\]ONIMPr
Emissions factor (discharge to nonimpaired waterbodies)
(0.005)
kg N20-N/kg N
PercentiMP
Percent of wastewater discharged to impaired waterbodies3
%
1990-2010: Set equal to
2010
2010: ERG (2021a)
2011: Obtained by
linear interpolation
2012: ERG (2021a)
2013-2021: Set equal to
2012
PercentNONiMP
Percent of wastewater discharged to nonimpaired
waterbodies3
%
1 a Value for this activity data varies over the Inventory time series.
2 Industrial Wastewater N2O Emission Estimates
3 Nitrous oxide emission estimates from industrial wastewater are estimated according to the methodology
4 described in the 2019 Refinement. U.S. industry categories that are likely to produce significant N2O emissions
5 from wastewater treatment were identified based on whether they generate high volumes of wastewater,
6 whether there is a high nitrogen wastewater load, and whether the wastewater is treated using methods that
7 result in N2O emissions. The top four industries that meet these criteria and were added to the inventory are meat
8 and poultry processing; petroleum refining; pulp and paper manufacturing; and breweries (ERG 2021a).
9 Wastewater treatment and discharge emissions for these sectors for 2021 are displayed in Table 7-36 below. Table
10 7-20 contains production data for these industries.
11 Table 7-36: Total Industrial Wastewater N2O Emissions by Sector (2021, MMT CO2 Eq. and
12 Percent)
N20 Emissions
% of Industrial
Industry
(MMT CO? Eq.)
Wastewater N20
Meat & Poultry
0.2
47.7
Petroleum Refineries
0.1
29.8
Pulp & Paper
0.1
21.7
Breweries
+
0.8
Total
0.5
100
+ Does not exceed 0.5 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
13 Emissions from Industrial Wastewater Treatment Systems:
14 More recent research has revealed that emissions from nitrification or nitrification-denitrification processes at
15 wastewater treatment, previously judged to be a minor source, may in fact result in more substantial emissions
16 (IPCC 2019). N2O is generated as a by-product of nitrification, or as an intermediate product of denitrification.
17 Therefore, N2O emissions are primarily expected to occur from aerobic treatment systems. To estimate these
18 emissions, the total nitrogen entering aerobic wastewater treatment for each industry must be calculated. Then,
19 the emission factor provided by the 2019 Refinement is applied to the portion of wastewater that undergoes
20 aerobic treatment.
21 The total nitrogen that enters each industry's wastewater treatment system is a product of the total amount of
22 industrial product produced, the wastewater generated per unit of product, and the nitrogen expected to be
23 present in each meter cubed of wastewater (IPCC equation 6.13).
7-50 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Equation 7-45: Total Nitrogen in Industrial Wastewater
2 TNINDi = PtXWtX TNi
3 where,
4 TN iNDi — total nitrogen in wastewater for industry /' for inventory year, kg TN/year
5 / = industrial sector
6 Pi = total industrial product for industrial sector /for inventory year, t/year
7 Wi = wastewater generated per unit of production for industrial sector/for inventory year,
8 m3/t product
9 Tni = total nitrogen in untreated wastewater for industrial sector/for inventory year, kg TN/m3
10 For the four industries of interest, the total production and the total volume of wastewater generated has already
11 been calculated for Cm emissions. For these new N2O emission estimates, the total nitrogen in the untreated
12 wastewater was determined by multiplying the annual industry production, shown in Table 7-20, by the average
13 wastewater outflow, shown in Table 7-23, and the nitrogen loading in the outflow shown in Table 7-37.
14 Table 7-37: U.S. Industrial Wastewater Nitrogen Data
Industry
Wastewater Total N
(kg N/ m3)
Source for Total N
Pulp and Paper
0.30a
Cabrera (2017)
Meat Processing
0.19
IPCC (2019), Table 6.12
Poultry Processing
0.19
IPCC (2019), Table 6.12
Petroleum Refining
0.051
Kenari et al. (2010)
Breweries - Craft
0.055
IPCC (2019), Table 6.12
Breweries - NonCraft
0.055
IPCC (2019), Table 6.12
15 a Units are kilograms N per air-dried metric ton of production.
16 Nitrous oxide emissions from industry wastewater treatment are calculated by applying an emission factor to the
17 percent of wastewater (and therefore nitrogen) that undergoes aerobic treatment (IPCC Equation 6.11).
18 Equation 7-46: N2O Emissions from Indsutrial Wastewater Treatment Plants
19 N20 PlantsIND = [£j(7y x EFx TNINDiJ\ x —
20 where,
21 N2O PlantsiND = N2O emissions from industrial wastewater treatment plants for inventory year, kg
22 INhO/year
23 TN iNDi — total nitrogen in wastewater from industry /' for inventory year, kg N/year
24 Ti,j = degree of utilization of treatment/discharge pathway or system j, for each industry / for
25 inventory year
26 / = industrial sector
27 j = each treatment/discharge pathway or system
28 EFi.j = emission factor for treatment/discharge pathway or system j, kg N20-N/kg N. 0.015 kg
29 N20-N/kg N (IPCC 2022)
30 44/28 = conversion of kg N2O-N into kg N2O
31 For each industry, the degree of utilization (Ti,j)—the percent of wastewater that undergoes each type of
32 treatment-was previously determined for CFU emissions and presented in Table 7-22.
33 Emissions from Industrial Wastewater Treatment Effluent:
34 Nitrous oxide emissions from industrial wastewater treatment effluent are estimated by multiplying the total
35 nitrogen content of the discharged wastewater effluent by an emission factor associated with the location of the
36 discharge. Where wastewater is discharged to aquatic environments with nutrient-impacted/eutrophic conditions
Waste 7-51
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
(i.e., water bodies which are rich in nutrients and very productive in terms of aquatic animal and plant life), or
environments where carbon accumulates in sediments such as lakes, reservoirs, and estuaries, the additional
organic matter in the discharged wastewater is expected to increase emissions.
Equation 7-47: N2O Emissions from Industrial Wastewater Treatment Effluent
N2O EffluentiND = Neffluent,ind X EFeffluent X 44/28
where,
N2O EffluentiND = N2O emissions from industrial wastewater discharge for inventory year (kg INhO/year)
Neffluent.ind = Total nitrogen in industry wastewater effluent discharged to aquatic environments for
inventory year (kg N/year)
EFeffluent = Tier 1 emission factor for wastewater discharged to aquatic environments (0.005 kg
INhO-N/kg N) (IPCC 2019)
44/28 = Conversion of kg N2O-N into kg N2O
The total N in treated effluent was determined through use of a nutrient estimation tool developed by EPA's Office
of Water (EPA 2019). The Nutrient Tool uses known nutrient discharge data within defined industrial sectors or
subsectors, as reported on Discharge Monitoring Reports, to estimate nutrient discharges for facilities within that
sector or subsector that do not have reported nutrient discharges but are likely to discharge nutrients. The
estimation considers, within each sector or subsector, elements such as the median nutrient concentration and
flow, as well as the percent of facilities within the sector or subsector that have reported discharges. Data from
2018 are available for the pulp, paper, and paperboard, meat and poultry processing, and petroleum refining
industries. To complete the time series, an industry-specific percent removal of nitrogen was calculated using the
total nitrogen in untreated wastewater. See Table 7-38.
Because data for breweries was not available, the removal of nitrogen was assumed to be equivalent to secondary
treatment, or 40 percent (IPCC 2019). The Tier 1 emission factor (0.005 kg INhO/kg N) from IPCC (2019) was used.
Table 7-38: Industrial Wastewater Nitrogen Discharged in 2018 by Sector (kg N)
Industry-Specific
Industry N EffluentiND (kg N) N Removal Factor
Meat & Poultry
12,078,919
0.082
Petroleum Refineries
1,698,953
0.045
Pulp & Paper
18,809,623
1.08
Breweries3
1,604,878
NA
a Nitrogen discharged by breweries was estimated as 60 percent of
untreated wastewater nitrogen.
Source: ERG (2021a).
Uncertainty
The overall uncertainty associated with both the 2021 CFU and N2O emission estimates from wastewater
treatment and discharge was calculated using the 2006 IPCC Guidelines Approach 2 methodology (IPCC 2006).
Uncertainty associated with the parameters used to estimate CH4 emissions include that of numerous input
variables used to model emissions from domestic wastewater and emissions from wastewater from pulp and
paper manufacturing, meat and poultry processing, fruits and vegetable processing, ethanol production,
petroleum refining, and breweries. Uncertainty associated with the parameters used to estimate N2O emissions
include that of numerous input variables used to model emissions from domestic wastewater and emissions from
wastewater from pulp and paper manufacturing, meat and poultry processing, petroleum refining, and breweries.
Uncertainty associated with centrally treated constructed wetlands parameters including U.S. population served by
constructed wetlands, and emission and conversion factors are from IPCC (2014), whereas uncertainty associated
7-52 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
with POTW flow to constructed wetlands and influent BOD and nitrogen concentrations were based on expert
judgment (ERG 2021b).
The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 7-39. For 2021, methane
emissions from wastewater treatment were estimated to be between 15.1 and 27.8 MMT CO2 Eq. at the 95
percent confidence level (or in 19 out of 20 Monte Carlo Stochastic Simulations). This indicates a range of
approximately 29 percent below to 32 percent above the 2021 emissions estimate of 21.1 MMT CO2 Eq. Nitrous
oxide emissions from wastewater treatment were estimated to be between 13.8 and 61.2 MMT CO2 Eq., which
indicates a range of approximately 34 percent below to 193 percent above the 2021 emissions estimate of 20.9
MMTCChEq.
Table 7-39: Approach 2 Quantitative Uncertainty Estimates for 2021 Emissions from
Wastewater Treatment (MMT CO2 Eq. and Percent)
Source
Gas
2021 Emission Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Wastewater T reatment
ch4
21.1
15.1
27.8
-29%
+32%
Domestic
ch4
13.9
9.2
19.7
-34%
+42%
Industrial
ch4
7.2
4.2
11.3
-42%
+58%
Wastewater T reatment
n2o
20.9
13.8
61.2
-34%
+193%
Domestic
n2o
20.4
12.8
60.4
-37%
+195%
Industrial
n2o
0.5
0.5
1.4
-0.4%
+202%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
QA/QC and Verification
General QA/QC procedures were applied to activity data, documentation, and emission calculations consistent
with the U.S. Inventory QA/QC plan, which is in accordance with Vol. 1 Chapter 6 of the 2006IPCC Guidelines (see
Annex 8 for more details). This effort included a general or Tier 1 analysis, including the following checks:
• Checked for transcription errors in data input;
• Ensured references were specified for all activity data used in the calculations;
• Checked a sample of each emission calculation used for the source category;
• Checked that parameter and emission units were correctly recorded and that appropriate conversion
factors were used;
• Checked for temporal consistency in time series input data for each portion of the source category;
• Confirmed that estimates were calculated and reported for all portions of the source category and for all
years;
• Investigated data gaps that affected trends of emission estimates; and
• Compared estimates to previous estimates to identify significant changes.
Calculation-related QC (category-specific, Tier 2) was performed for a portion of the domestic wastewater
treatment discharges methodology, which included assessing available activity data to ensure the most complete
publicly data set was used and checking historical trends in the data to assist determination of best methodology
for filling in the time series for data that are not available annually.
All transcription errors identified were corrected and documented. The QA/QC analysis did not reveal any systemic
inaccuracies or incorrect input values.
Waste 7-53
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Recalculations Discussion
Population data were updated using the same and latest data sources as the state-level emissions Inventory to
create consistency across inventory estimates. These changes affected the entire timeseries, except 2000. Protein
data were updated to reflect available protein values available for 2011, 2013, and 2018 through 2020 (FAO
2022c). Pulp, paper, and paperboard production data were updated to reflect revised values for 2020 (FAO 2022a).
Pulp, paper, and paperboard wastewater outflow data were updated to reflect new available values for 2020
which affected 2019 and 2020 (AF&PA 2022). Updated red meat production values for 2020, were updated based
on revised data (USDA 2022a; USDA 2022c). Fruits and vegetables production values were updated for the time
series (ERG 2022). Ethanol production values for 2015 and 2020 were based on revised data (RFA 2022a; RFA
2022b). Petroleum refining production values for 2020 were revised based on EIA (2022). In addition, EPA revised
the domestic sludge generation methodology to estimate the sludge generation from U.S. Territories and update
the time series to include new 2018 values (ERG 2022).
In addition, for the current Inventory, estimates of CCh-equivalent total CFU and N2O emissions from wastewater
treatment and discharge have been revised to reflect the 100-year global warming potentials (GWPs) provided in
the IPCC Fifth Assessment Report (AR5) (IPCC 2013). AR5 GWP values differ slightly from those presented in the
IPCC Fourth Assessment Report (AR4) (IPCC 2007) (used in the previous Inventories). The GWP of CH4 has increased
from 25 to 28, leading to an overall increase in CCh-equivalent CFU emissions while the GWP for N2O decreased
from 298 to 265 leading to a decrease in CCh-equivalent N2O emissions. The AR5 GWPs have been applied across
the entire time series for consistency. Further discussion on this update and the overall impacts of updating the
Inventory GWP values to reflect the IPCC Fifth Assessment Report can be found in Chapter 9, Recalculation and
Improvements.
Compared to the previous Inventory which applied 100-year GWP values from AR4, the cumulative effect of all
these recalculations had a minor impact on the overall wastewater treatment emission estimates:
• Domestic wastewater treatment and discharge CH4 emissions increased on average 13.9 percent over the
timeseries, with the smallest increase of 11.4 percent (1.7 MMT CO2 Eq.) in 1995 and largest increase of
19.9 percent (2.3 MMT C02 Eq.) in 2019.
• Domestic wastewater treatment and discharge N2O emissions decreased an average 11.0 percent over
the timeseries, with the smallest decrease in 8.9 percent (2.0 MMT CO2 Eq.) in 2019 to the largest
decrease of 11.0 percent (2.6 MMT CO2 Eq.) in 2020.
• Industrial wastewater treatment and discharge CFU emissions increased on average 12.1 percent over the
timeseries, with the smallest increase of 11.3 percent (0.7 MMT CO2 Eq.) in 2020 and largest increase of
12.3 percent (0.77 MMT C02 Eq.) in 2017.
• Industrial wastewater treatment and discharge N2O emissions decreased an average 11.1 percent over
the timeseries, with the smallest decrease of 11.1 percent (0.04 MMT CO2 Eq.) in 1991 to the largest
decrease of 11.5 percent (0.06 MMT CO2 Eq.) in 2020.
Over the time series, the total emissions on average increased by 1.1 percent from the previous Inventory. The
changes ranged from the smallest increase, 0.4 percent (0.2 MMT CO2 Eq.), in 2004, to the largest decrease, 2.4
percent (1.0 MMT CO2 Eq.), in 2019.
Planned Improvements
EPA notes the following improvements may be implemented or investigated within the next two or three
inventory cycles pending time and resource constraints:
• Investigate anaerobic sludge digester and biogas data compiled by the Water Environment Federation
(WEF) in collaboration with other entities as a potential source of updated activity data;
7-54 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
o Due to lack of these data, the United States continues to use another method for estimating
biogas produced. This method uses the standard 100 gallons/capita/day wastewater generation
factor for the United States (Ten-State Standards). However, based on stakeholder input, some
regions of the United States use markedly less water due to water conservation efforts so EPA
plans to investigate updated sources for this method as well.
EPA notes the following improvements will continue to be investigated as time and resources allow, but there are
no immediate plans to implement them until data are available or identified:
• Investigate additional sources for estimating wastewater volume discharged and discharge location for
both domestic and industrial sources. For domestic wastewater, the goal would be to provide additional
data points along the time series, while the goal for industrial wastewater would be to update the Tier 1
discharge methodology to a Tier 2 methodology.
• Investigate additional sources for domestic wastewater treatment type in place data.
• Continue to review whether sufficient data exist to develop U.S.-specific Cm or N2O emission factors for
domestic wastewater treatment systems, including whether emissions should be differentiated for
systems that incorporate biological nutrient removal operations; and
• Investigate additional data sources for improving the uncertainty of the estimate of N entering municipal
treatment systems.
• Evaluate the use of POTW BOD effluent discharge data from ICIS-NPDES.13 Currently only half of POTWs
report organics as BODsso EPA would need to determine a hierarchy of parameters to appropriately sum
all loads. Using these data could potentially improve the current methane emission estimates from
domestic discharge.
• Evaluate the use of POTW N effluent discharge data from ICIS-NPDES. Currently only about 80 percent of
POTWs report a form of N so EPA would need to determine an appropriate method to scale to the total
POTW population. EPA is aware of a method for industrial sources and plans to determine if this method
is appropriate for domestic sources.
7.3 Composting (CRF Source Category 5B1)
Composting of organic waste, such as food waste, garden (yard) and park waste, and wastewater treatment sludge
and/or biosolids, is common in the United States. Composting reduces the amount of methane-generating waste
entering landfills, destroys pathogens in the waste, sequesters carbon, and provides a source of organic matter.
Composting can also generate a saleable product and reduce the need for chemical fertilizers when the end
product is used as a fertilizer or soil amendment. This source category assumes all composting facilities are
commercial, large-scale anaerobic windrow composting facilities with yard trimmings as the main waste stream
composted (BioCycle 2017). Facilities using aerobic composting methods (e.g., aerated static piles, in-vessel
composting) are operational in the United States, however national estimates of the material processed by these
facilities are not readily available and therefore not included. Residential backyard composting is also not included
in this source category.
Composting naturally converts a large fraction of the degradable organic carbon in the waste material into carbon
dioxide (CO2) through aerobic processes without anthropogenic influence. With anthropogenic influences (e.g., at
commercial or large on-site composting operations), anaerobic conditions can be created in sections of the
compost pile when there is excessive moisture or inadequate aeration (or mixing) of the compost pile, resulting in
the formation of methane (CH4). Methane in aerobic sections of a windrow pile are generally oxidized by
13 ICIS-NPDES refers to EPA's Integrated Compliance Information System - National Pollutant Discharge Elimination System.
Waste 7-55
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
microorganisms, which convert the Cm to CO2 emissions. Even though CO2 emissions are generated, they are not
included in net greenhouse gas emissions for composting. Consistent with the 2006IPCC Guidelines, net CChflux
from carbon stock changes in waste material are estimated and reported under the LULUCF sector. The estimated
Cm released into the atmosphere ranges from less than 1 percent to a few percent of the initial C content in the
material (IPCC 2006). Depending on how well the compost pile is managed, nitrous oxide (N2O) emissions can also
be produced. The formation of N2O depends on the initial nitrogen content of the material and is mostly due to
nitrogen oxide (NOx) denitrification during the thermophilic and secondary mesophilic stages of composting
(Cornell 2007). Emissions vary and range from less than 0.5 percent to 5 percent of the initial nitrogen content of
the material (IPCC 2006). Animal manures are typically expected to generate more N2O than, for example, yard
waste, however data are limited.
From 1990 to 2021, the amount of waste composted in the United States increased from 3,810 kt to 22,946 kt (see
Table 7-42). There was some fluctuation in the amount of waste composted between 2006 to 2009 where a peak
of 20,063 kt composted was observed in 2008, which decreased to 18,838 kt composted the following year,
presumably driven by the economic crisis of 2009 (data not shown). Since 2009, the amount of waste composted
has gradually increased, and when comparing 2010 to 2021, a 25 percent increase in waste composted is
observed. Emissions of CH4 and N2O from composting from 2010 to 2021 have increased by the same percentage.
In 2021, CH4 emissions from composting (see Table 7-40 and Table 7-41) were 2.6 MMT CO2 Eq. (92 kt), and N2O
emissions from composting were 1.8 MMT CO2 Eq. (7 kt), representing consistent emissions trends over the past
several years. Composted material primarily includes yard trimmings (grass, leaves, and tree and brush trimmings)
and food scraps from the residential and commercial sectors (such as grocery stores; restaurants; and school,
business, and factory cafeterias). The composted waste quantities reported here do not include small-scale
backyard composting and agricultural composting mainly due to the lack of consistent and comprehensive national
data. Additionally, it is assumed that backyard composting tends to be a more naturally managed process with less
chance of generating anaerobic conditions and CH4 and N2O emissions. Agricultural composting is accounted for in
Volume 4, Chapter 5 (Cropland) of this Inventory, as most agricultural composting operations are assumed to land-
apply the resultant compost to soils.
The growth in composting since the 1990s and specifically over the past decade may be attributable to the
following factors: (1) the enactment of legislation by state and local governments that discouraged or banned the
disposal of yard trimmings and/or food waste in landfills, (2) an increase in yard trimming collection and yard
trimming drop off sites operated by local solid waste management districts/divisions,, (3) an increased awareness
of the environmental benefits of composting, and (4) loans or grant programs to establish or expand composting
infrastructure.
Most bans or diversion laws on the disposal of yard trimmings were initiated in the early 1990s by state or local
governments (U.S. Composting Council 2010). California, for example, enacted a waste diversion law for organics
including yard trimmings and food scraps in 1999 (AB939) that required jurisdictions to divert 50 percent of the
waste stream by 2000, or be subjected to fines. Currently, 20 states representing up to 42 percent of the nation's
population have enacted legislation banning yard waste from landfill disposal (U.S. Composting Council 2022).
Additional initiatives at the metro and municipal level also exist across the United States. Roughly 4,713
composting facilities exist in the United States with most (57.2 percent) composting yard trimmings only (BioCycle
2017).
In the last decade, bans and diversions for food waste have also become more common. As of 2022, eight states
(California, Connecticut, Massachusetts, New Jersey, New York, Orgon, Vermont, Washington) and seven local
governments (Austin, TX; Boulder, CO; Hennepin County, MN; Portland, OR; New York City, NY; San Francisco, CA;
Seattle, WA) had implemented organic waste bans or mandatory recycling laws to help reduce organic waste
entering landfills, with most having taken effect after 2013 (U.S. Composting Council 2022). In most cases, organic
waste reduction in landfills is accomplished by following recycling guidelines, donating excess food for human
consumption, or by sending waste to organics processing facilities (Harvard Law School and CET 2019). An example
of an organic waste ban as implemented by California is the California Mandatory Recycling Law (AB1826), which
requires companies to comply with organic waste recycling procedures if they produce a certain amount of organic
7-56 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
waste and took effect on January 1, 2015 (Harvard Law School and CET 2019). In 2017, BioCycle released a report
in which 27 of 43 states that responded to their organics recycling survey noted that food waste (collected
residential, commercial, institutional, and industrial food waste) was recycled via anaerobic digestion and/or
composting. These 27 states reported an estimated total of 1.8 million tons of food waste diverted from landfills in
2016 (BioCycle 2018b). A growing number of initiatives to encourage households and businesses to compost or
beneficially reuse food waste also exist.
Table 7-40: ChU and N2O Emissions from Composting (MMT CO2 Eq.)
Activity
1990
2005
2017
2018
2019
2020
2021
ch4
0.4
2.1
2.7
2.5
2.5
2.6
2.6
n2o
0.3
1.5
! 1.9
1.8
1.8
1.8
1.8
Total
0.7
3.6
4.7
4.3
4.3
4.4
4.4
Note: Totals by gas may not sum due to independent rounding.
ible 7-41: CH4 and N2O Emissions from Composting (kt)
Activity
1990
2005
2017
2018
2019
2020
2021
ch4
15
75
98
90
91
92
92
n2o
1
6
7
7
7
7
7
Methane and N2O emissions from composting depend on factors such as the type of waste composted, the
amount and type of supporting material (such as wood chips and peat) used, temperature, moisture content (e.g.,
wet and fluid versus dry and crumbly), and aeration during the composting process.
The emissions shown in Table 7-40 and Table 7-41 were estimated using the IPCC default (Tier 1) methodology
(IPCC 2006), which is the product of an emission factor and the mass of organic waste composted (note: no CH4
recovery is expected to occur at composting operations in the emission estimates presented):
Equation 7-48: Greenhouse Gas Emission Calculation for Composting
Per IPCC Tier 1 methodology defaults, the emission factors for CH4 and N2O assume a moisture content of 60
percent in the wet waste (IPCC 2006). While the moisture content of composting feedstock can vary significantly
by type, composting as a process ideally proceeds between 40 to 65 percent moisture (University of Maine 2016;
Cornell 1996).
Estimates of the quantity of waste composted (M, wet weight as generated) are presented in Table 7-42 for select
years. Estimates of the quantity composted for 1990 and 2005 were taken from EPA's Advancing Sustainable
Materials Management: Facts and Figures 2015 (EPA 2018); estimates of the quantities composted for 2017 to
2018 were taken from EPA's Advancing Sustainable Materials Management: 2018 Tables and Figures (EPA 2020a);
the estimate of the quantity composted for 2019 to 2021 were extrapolated using the 2018 quantity composted
and a ratio of the U.S. population growth between 2018 to 2019, 2019 to 2020, and 2020 to 2021, respectively
(U.S. Census Bureau 2021 and U.S. Census Bureau 2022). Estimates of waste composted by commercial facilities in
Methodology
E: = M X EF:
where,
Ei
M
EFi
Cm or N2O emissions from composting, kt Cl-Uor N2O
mass of organic waste composted in kt
emission factor for composting, 41 Cl-U/kt of waste treated (wet basis) and
0.31 INhO/kt of waste treated (wet basis) (IPCC 2006)
designates either Cl-Uor N2O
Waste 7-57
-------
1 Puerto Rico were provided for select years by EPA Region 2 (Kijanka 2020). This inventory includes waste
2 composted in Puerto Rico for 2017, 2018, and/or 2019 from three facilities in Puerto Rico, ranging from
3 approximately 1,200 kt to a high of 15,000 kt. The average waste composted for these years was used as the
4 annual amount composted for the respective facility for years the facility was operational. The annual quantity of
5 composted waste in Puerto Rico was forecasted for 2020 and 2021 using available data from prior years, assumed
6 metro area population data near where each facility is located, and the Microsoft FORECAST function to obtain
7 annual composting estimates. Puerto Rico waste composition estimates for 2020 and 2021. Efforts are made each
8 inventory year to fill historical and current data gaps for Puerto Rico's waste composting estimates.
9 Table 7-42: U.S. Waste Composted (kt)
Activity
1990
2005
2017
2018
2019
2020
2021
Waste Composted
3,810
18,655
24,501
22,594
22,698
22,918
22,946
10 Uncertainty
11 The major uncertainty drivers are the assumption that all composting emissions come from commercial windrow
12 facilities and the use of default emission factors (IPCC 2006) which is tied to a homogenous mixture of waste
13 processed across the country (largely yard trimmings). Data presented by BioCycle (BioCycle 2017) confirm most
14 composting operations use the windrow method and yard trimmings are the largest share of material composted
15 across the country, but there are other composting methods used and waste characteristics will vary at a facility
16 level. Additionally, there are composting operations in Puerto Rico and U.S. territories that are not explicitly
17 included in the national quantity of material composted as reported in the EPA Sustainable Materials Management
18 Reports because the methodological scope does not include Puerto Rico and U.S. territories. EPA took steps to
19 include emissions from Puerto Rico and U.S. Territories beginning in the 1990 to 2020 inventory and will continue
20 to seek out additional data in future Inventories.
21 The estimated uncertainty from the 2006 IPCC Guidelines is ±58 percent for the Tier 1 methodology and considers
22 the individual emission factors applied to the default emission factors and activity data.
23 Emissions from composting in 2021 were estimated to range between 1.8 and 7.0 MMT CO2 Eq., which indicates a
24 range of 58 percent below to 58 percent above the 2021 emission estimate of each gas (see Table 7-43).
25 Table 7-43: Tier 1 Quantitative Uncertainty Estimates for Emissions from Composting (MMT
26 CO2 Eq. and Percent)
Source
Gas
2021 Emission Estimate
Uncertainty Range Relative to Emission Estimate
(MMT CO? Eq.)
(MMT CO?
Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
ch4
2.6
1.1
4.1
-58%
+58%
Composting
n2o
1.8
0.8
2.9
-58%
+58%
Total
4.4
1.8
7.0
-58%
+58%
27 QA/QC and Verification
28 General QA/QC procedures were applied to data gathering and input, documentation, and calculations consistent
29 with the U.S. Inventory QA/QC Plan, which is in accordance with Vol. 1 Chapter 6 of the 2006 IPCC Guidelines (see
30 Annex 8 for more details). No errors were found for the current Inventory.
7-58 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Recalculations Discussion
The U.S. population estimate for 2020 was revised with current U.S. Census Bureau data (U.S. Census Bureau
2022). Because the 2020 composting estimates are extrapolated based on population growth, this recalculation
also resulted in a nominal increase (1 percent or 145 kt) in the quantity of material composted for 2020 compared
to that in the 1990 to 2020 Inventory report.
In addition, for the current Inventory, CCh-equivalent estimates of total Cm and N2O emissions from composting
have been revised to apply the 100-year global warming potentials (GWPs) provided in the IPCC Fifth Assessment
Report (AR5) (IPCC 2013). AR5 GWP values differ slightly from those presented in the IPCC Fourth Assessment
Report (AR4) (IPCC 2007) (used in the previous inventories). The AR5 GWPs have been applied across the entire
time series for consistency. The GWP of CH4 has increased from 25 to 28, leading to an overall increase in CO2-
equivalent CFU emissions while the GWP for N2O decreased from 298 to 265 leading to a decrease in CO2-
equivalent N2O emissions. Compared to the previous Inventory which applied 100-year GWP values from AR4, the
change in CC>2-equivalent CFU emissions was a 12 percent increase for each year of the time series, while the
change in CC>2-equivalent N2O emissions was an 11 percent decrease for each year of the time series. Further
discussion on this update and the overall impacts of updating the inventory GWPs to reflect the IPCC Fifth
Assessment Report can be found in Chapter 9, Recalculations and Improvements. The net impact from these
updates was an average annual 1 percent increase in composting emissions for the time series.
Planned Improvements
EPA recently completed a literature search on emission factors and composting systems and management
techniques that were documented in a draft technical memorandum. The purpose of this literature review was to
compile all published emission factors specific to various composting systems and composted materials in the
United States to determine whether the emission factors used in the current methodology can be revised or
expanded to account for geographical differences and/or differences in composting systems used. For example,
outdoor composting processes in arid regions typically require the addition of moisture compared to similar
composting processes in wetter climates. In general, there is a lack of facility-specific data on the management
techniques and sum of material composted to enable the incorporate of different emission factors. EPA will
continue to seek out more detailed data on composting facilities to enable this improvement in the future.
Relatedly, EPA has received comments during previous Inventory cycles recommending that calculations for the
composting sector be based on waste subcategories (i.e., leaves, grass and garden debris, food waste) and
category-specific moisture contents. At this time, EPA is not aware of any available datasets which would enable
estimations to be performed at this level of granularity. EPA will continue to search for data which could lead to
the development of subcategory-specific composting emission factors to be used in future Inventory cycles.
EPA has put significant work into its Excess Food Opportunities Map dataset, including the compilation of
composting facilities and feedstock accepted across the country. Additionally, the EPA's 2018 Wasted Food Report
(EPA 2020b) includes estimates of composted waste for individual sectors (e.g., food and beverage manufacturing,
restaurants/food services, hospitals, correctional facilities, office buildings). Estimates are provided for one year,
2018. The Inventory compilation team plans to review this report's estimates in comparison to the EPA's Facts and
Figures report to identify sectors that are not duplicated in the Facts and Figures reports, and develop a
methodology to generate estimates for all years in the Inventory time series (1990 through 2021).
EPA will also continue to seek out activity data including processing capacity and years of operation for commercial
composting facilities in Puerto Rico (for additional years), Guam, and other U.S. Territories for inclusion in a future
Inventory.
Waste 7-59
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
7.4 Anaerobic Digestion at Biogas Facilities
(CRF Source Category 5B2)
Anaerobic digestion is a series of biological processes in the absence of oxygen in which microorganisms break
down organic matter, producing biogas and digestate. The biogas primarily consists of Cm, biogenic CO2, and trace
amounts of other gases such as N2O (IPCC 2006) and is often combusted to produce heat and power, or further
processed into renewable natural gas or for use as a transportation fuel. Digester gas contains approximately 65
percent CFU (a normal range is 55 percent to 65 percent) and approximately 35 percent CO2 (WEF 2012; EPA 1993).
Methane emissions may result from a fraction of the biogas that is lost during the process due to leakages and
other unexpected events (0 to 10 percent of the amount of CFU generated, IPCC 2006), collected biogas that is not
completely combusted, and entrained gas bubbles and residual gas potential in the digestate. Carbon dioxide
emissions are biogenic in origin and should be reported as an informational item in the Energy Sector (IPCC 2006).
Volume 5 Chapter 4 of the 2006 IPCC Guidelines notes that at biogas plants where unintentional CH4 emissions are
flared, CFU emissions are likely to be close to zero.
Anaerobic digesters differ based on the operating temperature, feedstock type and moisture content, and mode of
operation. The operating temperature dictates the microbial communities that live in the digester. Mesophilic
microbes are present at temperatures ranging from 85 to 100 degrees Fahrenheit while thermophilic microbes
thrive at temperatures ranging from 122 to 140 degrees Fahrenheit (WEF 2012). Digesters may process one or
more types of feedstock, including food waste; municipal wastewater solids; livestock manure; industrial
wastewater and residuals; fats, oils, and grease; and other types of organic waste streams. Co-digestion (multiple
feedstocks) is employed to increase methane production in cases where an organic matter type does not break
down easily. In co-digestion, various organic wastes are decomposed in a singular anaerobic digester by using a
combination of wastewater solids or manure and food waste from restaurants or food processing industry, a
combination of manure and waste from energy crops or crop residues (EPA 2016), or alternative combinations of
feedstock. The moisture content of the feedstock (wet or dry) impacts the amount of biogas generation. Wet
anaerobic digesters process feedstock with a solids content of less than 15 percent while dry anaerobic digesters
process feedstock with a solids content greater than 15 percent (EPA 2020). Digesters may also operate in batch or
continuous mode, which affects the feedstock loading and removal. Batch anaerobic digesters are manually loaded
with feedstock all at once and then manually emptied while continuous anaerobic digesters are continuously
loaded and emptied with feedstock (EPA 2020).
The three main categories of anaerobic digestion facilities included in national greenhouse gas inventories include
the following:
• Anaerobic digestion at biogas facilities, or stand-alone digesters, can be industry-dedicated digesters that
process waste from on industry or industrial facility (typically food of beverage waste from
manufacturing), or multi-source digesters that process feedstocks from various sources (e.g., municipal
food scraps, manure, food processing waste). Some stand-alone digesters also co-digest other organics
such as yard waste.
• On-farm digesters manage organic matter and reduce odor generated by farm animals or crops. On-farm
digesters are found mainly at dairy, swine, and poultry farms where there is the highest potential for
methane production to energy conversion. On-farm digesters may also accept food waste as feedstock for
co-digestion.
• Digesters at water resource recovery facilities (WRRF) produce biogas through the treatment and
reduction of wastewater solids. Some WRRF facilities may also accept and co-digest food waste.
This section focuses on stand-alone anaerobic digestion at biogas facilities. Emissions from on-farm digesters are
included Chapter 5 (Agriculture) and AD facilities at WRRFs are included in Section 7.2 (Wastewater Treatment).
7-60 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
From 1990 to 2021, the estimated amount of waste managed by stand-alone digesters in the United States
increased from approximately 786 kt to 8,263 kt, an increase of 951 percent. As described in the Uncertainty
section, no data sources present the annual amount of waste managed by these facilities prior to 2015 when the
EPA began a comprehensive data collection survey. Thus, the emission estimates between 1990 and 2014, and for
2019 to 2021 are general estimates extrapolated from data collected for years 2015 to 2018. The steady increase
in the amount of waste processed over the time series is likely driven by increasing interest in using waste as a
renewable energy source and other organics diversion goals.
In 2021, emissions from stand-alone anaerobic digestion at biogas facilities were approximately 0.2 MMT CO2 Eq.
(6 kt) (see Table 7-44 and Table 7-45).
Table 7-44: ChU Emissions from Anaerobic Digestion at Biogas Facilities (MMT CO2 Eq.) from
1990-2021
Activity
1990
2005
2017
2018
2019
2020
2021
CH4 Generation
+
0.1
0.2
0.2
0.2
0.2
0.2
CH4 Recovery
(+)
(+)
(+)
(+)
(+)
(+)
(+)
CH4 Emissions
+
+
0.2
0.2
0.2
0.2
0.2
+ Absolute value does not exceed 0.05 MMT.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values.
Table 7-45: ChU Emissions from Anaerobic Digestion at Biogas Facilities (kt) from 1990-2021
Activity
1990
2005
2017
2018
2019
2020
2021
CH4 Generation
1
2
7
7
7
7
7
CH4 Recovery
(+)
(+)
(+)
(+)
(+)
(+)
(+)
CH4 Emissions
1
2
6
6
6
6
6
+ Does not exceed 0.5 kt.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values.
Methodology
Methane emissions from anaerobic digestion depend on factors such as the type of waste managed, the amount
and type of supporting material (such as wood chips and peat) used, temperature, moisture content (e.g., wet and
fluid versus dry and crumbly), and aeration during the digestion process.
The emissions presented in Table 7-44 were estimated using the IPCC default (Tier 1) methodology (Volume 5,
Chapter 4, IPCC 2006) given in Equation 7-49 below, which is the product of an emission factor and the mass of
organic waste processed. Only Cm emissions are estimated because N2O emissions are considered negligible (IPCC
2006). Some Tier 2 data are available (annual quantity of waste digested) for the later portion of the time series
(2015 and later).
Equation 7-49: Methane Emissions Calculation for Anaerobic Digestion
CH4 Emissions = x EFt) x 10~3 — R
where,
Cm Emissions =
Mi
EF
i =
R
Waste 7-61
total CH4 emissions in inventory year, Gg CH4
mass of organic waste treated by biological treatment type /', Gg, see Table 7-46
emission factor for treatment /', g CFU/kg waste treated, 0.8 Mg/Gg CH4
anaerobic digestion
total amount of CH4 recovered in inventory year, Gg CH4
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Equation 7-50: Recovered Methane Estimation for Anaerobic Digestion
minutes
1
x Biogas CH4 Density x CCH4 x x (1 — DE)
R = Biogas x 0.0283 x
year
where,
662
65%
1/109
0.99
Biogas
0.0283
525,600
the annual amount of biogas produced, standard cubic feet per minute (scfm)
conversion factor cubic meter/cubic feet
minutes per year
Cm density in biogas (EPA 1993), g Cm/m3 CFU
Cch4, concentration of CFU in the biogas (WEF 2012; EPA 1993)
conversion factor, grams to kt
destruction efficiency for combustion unit
Per IPCC Tier 1 methodology defaults, the emission factor for CH4 assumes a moisture content of 60 percent in the
wet waste (IPCC 2006). Both liquid and solid wastes are processed by stand-alone digesters and the moisture
content entering a digester may be higher. One emission factor recommended by the 2006 IPCC Guidelines (0.8
Mg/Gg CH4) is applied for the entire time series (IPCC 2006 Volume 5, Chapter 4, Table 4.1).
The annual quantity of waste digested is sourced from recent EPA surveys of anaerobic digestion facilities (EPA
2018, 2019, and 2021). The EPA was granted the authority to survey anaerobic digestion facilities that process food
waste annually through an Information Collection Request (ICR No. 2533.01). The scope includes stand-alone and
co-digestion facilities (on-farm and water resource recovery facilities [WRRF]). Three reports with survey results
have been published to date:
• Anaerobic Digestion Facilities Processing Food Waste in the United States in 2015: Survey Results (EPA
2018)
• Anaerobic Digestion Facilities Processing Food Waste in the United States in 2016: Survey Results (EPA
2019)
• Anaerobic Digestion Facilities Processing Food Waste in the United States in (2017 & 2018): Survey Results
(EPA 2021)
These reports present aggregated survey data including the annual quantity of waste processed by digester type
(i.e., stand-alone, on-farm, and WRRF); waste types accepted; biogas generation and end use; and more. The
aggregated data presented in the EPA reports are underestimates of the actual amount of processed waste and
biogas produced because (1) surveys rarely achieve a 100 percent response rate and some fraction of facilities in
each survey year did not respond to the survey; (2) EPA focused this survey on facilities that process food waste,
and there may be additional operational digesters that are not located on farms or at wastewater treatment
plants; and (3) EPA has done due diligence to identify all stand-alone digesters that process food waste but may
not have identified all facilities across the United States and its territories. The amount of waste digested as
reported in the survey reports were assumed to be in wet weight; the majority of stand-alone digesters were
found to be wet and mesophilic (EPA 2019).
The annual quantity of waste digested at stand-alone digesters for 1990 to 2014 (only 1990 and 2005 are shown in
Table 7-46) was estimated by multiplying the count of estimated operating facilities (as presented in Table 7-47) by
the weighted average of waste digested in 2015 and 2016 collected through EPA's survey data (EPA 2018; EPA
2019). Masked survey responses of food and non-food waste processed were shared with the Inventory team by
the EPA team leading the EPA AD Data Collection Surveys. This provided an accurate count of the number of
facilities that provided annual quantities of digested waste, which matters for the weighted average. The weighted
average applied to the current inventory is calculated as follows for 1990 to 2014:
Equation 7-51: Weighted Average of Waste Processed
Weighted Average Waste Processed =
0^2016 X ^"aC2016 + ^2015 X FaC201s)
(Fac 2016 + Fac 2ois)
7-62 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
where,
W = total waste processed in the respective survey year, food and non-food waste (short tons).
Fac = the number of facilities that reported an amount of waste processed in the respective
survey year. Note the number of facilities that provided an annual quantity of waste
processed data was internally shared and differs from the total number of facilities that
responded to the EPA surveys as presented in EPA (2018, 2019).
Estimates of the quantity of waste digested (M, wet weight as generated) are presented in for select years and the
number of facilities that reported annual quantities of waste digested to the EPA survey were 45 and 44 in 2015
and 2016, respectively (using masked facility data provided by the EPA AD survey data collection team).
Estimates of the quantity of waste digested for 1990 to 2014 are calculated by multiplying the weighted average of
waste digested from 2015 and 2016 survey data (216,494 short tons) by the count of operating facilities in each
year. This calculation assumes that each facility operates continuously from the first year of operation for the
remainder of the time series. Additional efforts will be made to quantify the number of operating facilities and
estimates of the total waste digested by year for future Inventories as described in the Planned Improvements
section. Estimates of the quantity digested for 2015 and 2016 were taken from EPA's AD survey data (EPA 2018;
EPA 2019, respectively). The estimate of waste digested for 2019 through 2021 were extrapolated using the
average of the waste digested from the 2017 and 2018 survey data (EPA 2021) as a proxy. The average did not
include data from 2015 and 2016 because there is a drop in the amount of waste digested by nearly 1 million tons
between 2016 and 2017. The quantities digested between 2015 and 2016 are similar, and quantities digested
between 2017 and 2018 are similar. Estimates for 2019 to 2021 will be updated as future EPA survey reports are
published.
Table 7-46: U.S. Waste Digested (kt) from 1990-2021
Activity
1990
2005
2017
2018
2019
2020
2021
Waste Digested3
786
h 2,357
8,206
8,320
8,263
8,263
8,263
a The amount of waste digested primarily consists of food waste. The amount processed for all years is likely
an underestimate because the estimates were developed from survey data provided by operating facilities
for 2015 to 2018 (EPA 2018; EPA 2019; EPA 2021). Facilities that did not respond to the EPA surveys are not
included and all years except 2015 to 2018 are estimated using assumptions regarding the number of
operating facilities and the weighted average of waste digested. Additionally, the liquid portion of the waste
digested in 2015 and 2016 are not included due to limited information on the specific waste types to
perform the unit conversion to kt. EPA converted liquid waste to tons for 2018 and 2019 using a conversion
factor of 3.8 pounds per gallon (EPA 2021). The weighted average of waste digested in 2015 and 2016 (as
reported in EPA 2018 and 2019) is used as the average for 1990 to 2014, and the average waste digested as
reported in EPA (2021) is used as a proxy for years 2019 to 2021.
The estimated count of operating facilities is calculated by summing the count of digesters that began operating by
year over the time series. The year a digester began operating is sourced from EPA (2021). This assumes all
facilities are in operation from their first year of operation throughout the remainder of the time series, including
facilities prior to 1990. This is likely an overestimate of facilities operating per year but does not necessarily
translate to an overestimate in the amount of waste processed because a weighted average of waste processed for
the surveyed facilities is applied to these years. The number of facilities in 1990 to 2014 are directly used in
calculating the emissions, while the directly reported annual amount of waste processed from the survey data are
used for 2015 to 2021.
Waste 7-63
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Table 7-47: Estimated Number of Stand-Alone AD Facilities Operating3 from 1990-2021
Year
1990
2005
2017
2018
2019
2020
2021
Estimated Count of Operational Facilities
4
12
68
68
68
68
68
a The count of operational facilities was visually estimated from Figure 5 in EPA (2021), which presents the count of the
first year of digester operation. The number of operational facilities by year is assumed to be the cumulative total from
the prior year. This method assumes all facilities are operating from 1990, or their first year of operation, to 2020. The
number of facilities operating between 2015 to 2018 are equal to the number of facilities surveyed by EPA (EPA 2018,
2019, and 2021). The number of facilities operating in 2019 and 2020 are assumed to be the same as the last survey
report data year, i.e., 2018 as reported in EPA (2021). These assumptions are further discussed in the Methodology
and Time-Series Consistency section.
Activity data for the amount of biogas recovered (R in the emission calculation equation) is limited across the time
series. Currently, there are only four data points (2015, 2016, 2017, and 2018) represented for the entire sector, as
reported in the EPA AD Data Collection Survey reports (EPA 2018, 2019, and 2021). The total quantity of collected
biogas from the survey respondents is reported in standard cubic feet per minute (scfm) as shown in Table 7-48.
Volume 5, Chapter 4 of the 2006IPCC Guidelines notes that only emissions from flaring can be reported under the
waste sector. The top three known uses of the biogas generated by stand-alone digesters are combined heat and
power (CHP), the production of electricity that is sold to the grid, and using the biogas to fuel boilers and furnaces
to heat the digestor and other facility spaces (EPA 2018; EPA 2019). Thus, no biogas is assumed to be flared.
Table 7-48: Estimated Biogas Produced and Methane Recovered from Anaerobic Digestion at
Biogas Facilities Operating from 1990-2021"
Activity
1990
2005
2017
2018
2019
2020
Total Biogas Produced (scfm)b
767
2,301
6,402
7,282
6,842
6,842
R, recovered CH4 from biogas (kt)c
(0.05)
(0.14)
(0.41)
(0.47)
(0.49)
(0.49)
a Total biogas produced in standard cubic feet per minute (scfm) was reported in aggregate in the EPA survey data (EPA 2018, 2019, 2021) for 2C
2018. The quantities presented in this table are likely underestimates because not all operational facilities provided a survey response to the EF
Data Collection Surveys.
b Data for all years in the time series except for 2015 and 2016 are extrapolated using the average of the total biogas collected between 2015 to
divided by the average number of survey responses to generate a weighted average estimate of biogas collected per facility, which is then mull
by the total facility count (as shown in Table 7-47).
c The quantity of CH4 recovered from the biogas produced is estimated for all years except 2015 to 2018, which are taken from EPA (2018), EPA |
and EPA (2021).
Note: Parentheses indicate negative values.
Uncertainty
The methodology applied for the 1990 to 2014 emissions estimates should be considered a starting point to build
on in future years if additional historical data become available. Four years of facility-provided data are available
(2015 to 2018) while the rest of the time series is estimated based on an assumption of facility counts and the
2015 and 2016 weighted average annual waste digested as calculated from survey data. The major limitations, and
uncertainty drivers in the emissions estimates, are related to the uncertainty in assumptions to ensure
completeness across the time series and the limitations in the EPA AD survey data, as described below:
1. The EPA AD survey (EPA 2018; EPA 2019; EPA 2021) did not receive a 100 percent response rate, meaning
that the survey data represent a portion, albeit the majority, of stand-alone digesters, annual waste
processed, and biogas recovered. The methodology applied here did not attempt to estimate waste
digested by facilities that did not respond to the survey, which likely underestimates the quantity of waste
digested and CH4emissions.
2. The EPA AD survey data (EPA 2018; EPA 2019) present both food and non-food waste digested. The non-
food waste was reported as liquid (gallons) and solid (tons). The quantity of liquid waste managed is not
included in the estimated quantity of annual waste digested for 2015 and 2016, which is used as a proxy
for 1990 to 2014 because data on the waste types are not available to convert the quantity from gallons
7-64 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 to tons. This slightly underestimates the quantity of waste digested and Cmemissions. EPA (2021) did
2 convert the liquid waste managed to tons for 2017 and 2018 using a general conversion factor.
3 3. The assumption required to estimate the activity data for 1990 to 2014 may overestimate the number of
4 facilities in operation because it assumes that each facility operates from its start year for the entire time
5 series (i.e. facility closures are not taken into account). This introduces a large amount of uncertainty in
6 the estimates compared to years where there is directly reported survey data. It is unclear whether this
7 under- or over-estimates the quantity of waste digested and CFU emissions.
8 The estimated uncertainty from the 2006IPCC Guidelines is ±54 percent for the Approach 1 methodology.
9 Emissions from anaerobic digestion at stand-alone biogas facilities in 2021 were estimated to be between 0.1 and
10 0.3 MMT CO2 Eq., which indicates a range of 54 percent below to 54 percent above the 2021 emission estimate of
11 Cm (see Table 7-49). A ±20 percent uncertainty factor is applied to the annual amount of material digested (i.e.,
12 the activity data), which was developed with expert judgment (Bronstein 2021). A ±50 percent default uncertainty
13 factor is applied to the CH4 emission factor (IPCC 2006). Using the IPCC's error propagation equation (Equation 3.1
14 in IPCC 2006 Volume 1, Chapter 3), the combined uncertainty percentage is ±54 percent.
15 Table 7-49: Approach 1 Quantitative Uncertainty Estimates for Emissions from Anaerobic
16 Digestion (MMT CO2 Eq. and Percent)
Source
Gas
2021 Emission Estimate
Uncertainty Range Relative to Emission Estimate
(MMT CO? Eq.)
(MMT CO? Eq.)
(%)
Lower Upper
Lower Upper
Bound Bound
Bound Bound
Anaerobic Digestion
at Biogas Facilities
ch4
0.2
0.1 0.3
-54% +54%
17 QA/QC and Verification
18 General QA/QC procedures were applied to data gathering and input, documentation, and calculations consistent
19 with the U.S. Inventory QA/QC Plan, which is in accordance with Vol. 1, Chapter 6 of the 2006 IPCC Guidelines (see
20 Annex 8 for more details). No errors were found for the current Inventory.
21 Recalculations Discussion
22 For the current Inventory, estimates of C02-equivalent CH4 emissions from anaerobic digestion at biogas facilities
23 have been revised to reflect the 100-year global warming potentials (GWPs) provided in the IPCC Fifth Assessment
24 Report (AR5) (IPCC 2013). AR5 GWP values differ slightly from those presented in the IPCC Fourth Assessment
25 Report (AR4) (IPCC 2007) (used in previous Inventories). The AR5 GWPs have been applied across the entire time
26 series for consistency. The GWP of CH4 has increased from 25 to 28, leading to an overall increase in CO2-
27 equivalent CH4 emissions. Compared to the previous Inventory which applied 100-year GWP values from AR4, the
28 change in C02-equivalent CH4 emissions was a 12 percent increase for each year of the time series. Further
29 discussion on this update and the overall impacts of updating the Inventory GWPs to reflect the IPCC Fifth
30 Assessment Report can be found in Chapter 9, Recalculations and Improvements.
31 Planned Improvements
32 EPA will continue to incorporate updated survey data from future EPA AD Data Collection Surveys when the survey
33 data are published. These revisions will change the estimated emissions for 2019 to 2021.
Waste 7-65
-------
1 EPA will also re-assess how best to estimate annual waste processed using proxy data for years between the EPA
2 AD Data Collection Survey reports as needed (e.g., for 2019, 2020, 2021). The methodology described here
3 assumes the same average amount of waste is processed each year for 2019 through 2021.
4 EPA continues to seek out data sources to confirm the estimated number of operational facilities by year prior to
5 2015 and consider how best to estimate the quantity of waste processed per year by these facilities with the goal
6 of better estimating the annual quantity of waste digested between 1990 to 2014. Available data will also be
7 compiled where available for facilities that did not directly respond to the EPA AD Data Collection surveys for
8 completeness.
9 EPA will seek out data sources to confirm the amount of recovered biogas for years prior to 2015 (i.e., the years
10 prior to the EPA AD Data Collection Surveys). Currently, partial data of recovered biogas are available between
11 2015 to 2018 from the EPA AD Data Collection Surveys. The primary purpose of this improvement will be to
12 understand whether the range of recovered biogas from the survey data are reflective of earlier years in the time
13 series.
14 7.5 Waste Incineration (CRF Source
is Category 5C1)
16 As stated earlier in this chapter, carbon dioxide (CO2), nitrous oxide (N2O), and methane (CH4) emissions from the
17 combustion of waste are accounted for in the Energy sector rather than in the Waste sector because almost all
18 combustion of municipal solid waste (MSW) in the United States occurs at waste-to-energy facilities where useful
19 energy is recovered. Similarly, the Energy sector also includes an estimate of emissions from burning waste tires
20 and hazardous industrial waste, because virtually all of the combustion occurs in industrial and utility boilers that
21 recover energy. The combustion of waste in the United States in 2021 resulted in 12.8 MMT CO2 Eq. of emissions.
22 For more details on emissions from the combustion of waste, see Section 3.3 of the Energy chapter.
23 Additional sources of emissions from waste combustion include non-hazardous industrial waste incineration and
24 medical waste incineration. As described in Annex 5 of this report, data are not readily available for these sources
25 and emission estimates are not provided.
26 An analysis of the likely level of medical waste incineration emissions was conducted based on a 2009 study of
27 hospital/ medical/ infectious waste incinerator (HMIWI) facilities in the United States (RTI 2009). Based on that
28 study's information of waste throughput and an analysis of the fossil-based composition of the waste, it was
29 determined that annual greenhouse gas emissions for medical waste incineration would be below 500 kt CO2 Eq.
30 per year and considered insignificant for the purposes of Inventory reporting under the UNFCCC. More information
31 on this analysis is provided in Annex 5.
32 Furthermore, an analysis was conducted on the likely level of sewage sludge incineration emissions based on the
33 total amount of sewage sludge generated and assumed percent incineration. Based on assumed amount of sludge
34 incinerated and non-CC>2 factors for solid biomass it was determined that annual greenhouse gas emissions for
35 sewage sludge incineration would be below 500 kt CO2 Eq. per year and considered insignificant for the purposes
36 of Inventory reporting under the UNFCCC. More information on this analysis is provided in Annex 5.
7-66 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 7.6 Waste Sources of Precursor
2 Greenhouse Gases-TO BE UPDATED FOR
FINAL INVENTORY REPORT
In addition to the main greenhouse gases addressed above, waste generating and handling processes are also
sources of precursors to greenhouse gases. The reporting requirements of the UNFCCC14 request that information
be provided on precursor emissions, which include carbon monoxide (CO), nitrogen oxides (NOx), non-methane
volatile organic compounds (NMVOCs), and sulfur dioxide (SO2). These gases are not direct greenhouse gases, but
can indirectly impact Earth's radiative balance by altering the concentrations of other greenhouse gases (e.g.,
tropospheric ozone) and atmosphere aerosol (e.g., particulate sulfate). Total emissions of NOx, CO, NMVOCs, and
SO2 from waste sources for the years 1990 through 2021 are provided in Table 7-50.
Table 7-50: Emissions of NOx, CO, NMVOC, and SO2 from Waste (kt)
Gas/Source
1990
2005
2017
2018
2019
2020
2021
NOx
+
2
1
1
1
1
1
Landfills
+
2
1
1
1
1
1
Wastewater Treatment
+
0
0
0
0
0
0
Miscellaneous3
+
0
0
0
0
0
0
CO
1
7
6
5
5
5
5
Landfills
1
6
6
5
5
5
5
Wastewater Treatment
+
+
+
+
+
+
+
Miscellaneous3
+
0
0
0
0
0
0
NMVOCs
673
114
52
52
52
52
52
Wastewater Treatment
57
50
25
22
22
22
22
Miscellaneous3
557
44
22
20
20
20
20
Landfills
58
22
11
10
10
10
10
S02
+
1
1
1
1
1
1
Landfills
+
1
1
1
1
1
1
Wastewater Treatment
+
0
0
0
0
0
0
Miscellaneous3
+
0
0
0
0
0
0
+ Does not exceed 0.5 kt.
3 Miscellaneous includes TSDFs (Treatment, Storage, and Disposal Facilities under the Resource Conservation
and Recovery Act [42 U.S.C. § 6924, SWDA § 3004]) and other waste categories.
Note: Totals by gas may not sum due to independent rounding.
12
13
14
15
16
Methodology and Time-Series Consistency
Emission estimates for 1990 through 2021 were obtained from data published on the National Emissions Inventory
(NEI) Air Pollutant Emissions Trends Data website (EPA 2022a). For Table 7-50, NEI reported emissions of CO, NOx,
SO2, and NMVOCs are recategorized from NEI Tier 1/Tier 2 source categories to those more closely aligned with
IPCC categories, based on EPA (2003).15 NEI Tier 1 emission categories related to the IPCC waste sector include:
14 See http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
The NEI estimates and reports emissions from six criteria air pollutants (CAPS) and 187 hazardous air pollutants (HAPS) in
support of National Ambient Air Quality Standards. Reported NEI emission estimates are grouped into 60 sectors and 15 Tier 1
Waste 7-67
-------
1 Waste Disposal and Recycling (landfills; publicly owned treatment works; industrial wastewater; treatment,
2 storage, and disposal facilities; and other). As described in detail in the NEI Technical Support Documentation (TSD)
3 (EPA 2021), emissions are estimated through a combination of emissions data submitted directly to the EPA by
4 state, local, and tribal air agencies, as well as additional information added by the Agency from EPA emissions
5 programs, such as the emission trading program, Toxics Release Inventory (TRI), and data collected during rule
6 development or compliance testing.
7 Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
8 through 2021, which are described in detail in the NEI's TSD (EPA 2021). No quantitative estimates of uncertainty
9 were calculated for this source category.
source categories, which broadly cover similar source categories to those presented in this chapter. For this report, EPA has
mapped and regrouped emissions of greenhouse gas precursors (CO, NOx, S02, and NMVOCs) from NEI Tier 1/Tier 2 categories
to better align with IPCC source categories, and to ensure consistency and completeness to the extent possible. [See Annex 6.X
for more information on this mapping].
7-68 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 8. Other
2 The United States does not report any greenhouse gas emissions under the Intergovernmental Panel on Climate
3 Change (IPCC) "Other" sector.
Other 8-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
9. Recalculations and Improvements
Each year, many emission and sink estimates in the Inventory of U.S. Greenhouse Gas Emissions and Sinks 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. In this effort, the United States follows the 2006
IPCC Guidelines (IPCC 2006), which state, "Both methodological changes and refinements over time are an
essential part of improving inventory quality. It is good practice to change or refine methods when available data
have changed; the previously used method is not consistent with the IPCC guidelines for that category; a category
has become key; the previously used method is insufficient to reflect mitigation activities in a transparent manner;
the capacity for inventory preparation has increased; improved inventory methods become available; and/or for
correction of errors."
When methodological changes have been implemented, the previous Inventory's time series (i.e., 1990 to 2020) is
assessed and potentially recalculated to reflect the change, per guidance in IPCC (2006). Changes in historical data
are often the result of changes in statistical data supplied by other agencies, and these changes do not necessarily
impact the entire time series. In addition, the current Inventory updates GWPs for calculating CC>2-equivalent
emission estimates of non-CC>2 gases (Cm, N2O, HFCs, PFCs, SF6, and NF3) to reflect updated science. This inventory
has been revised to use the 100-year GWPs provided in the IPCC Fifth Assessment Report (AR5) (IPCC 2013). AR5
GWP values differ from those presented in the IPCC Fourth Assessment Report and used in the previous
Inventories as required by earlier UNFCCC reporting guidelines. Recent decisions under the UNFCCC1 require
Parties to use 100-year GWP values from the IPCC Fifth Assessment Report (AR5) for calculating CCh-equivalence in
their national reporting (IPCC 2013) by the end of 2024. In preparation for upcoming UNFCCC requirements2, this
report reflects CCh-equivalent greenhouse gas totals using 100-year AR5 GWP values. Note, all estimates provided
in sectoral chapters of this report are presented in both CO2 equivalents and unweighted units.
The results of all methodological changes and historical data updates made in the current Inventory are presented
in Figure 9-1, Table 9-3, and Table 9-4. Figure 9-1 presents the impact of recalculations by sector and on net total
emissions across the timeseries. Table 9-1 and Table 9-2 include the quantitative effects of methodological
changes as well as the impacts of updating GWPs from AR4 to AR5 in calculating CC>2-equivalent U.S. greenhouse
gas emissions by gas across the Energy, Industrial Processes and Product Use (IPPU), Agriculture, Land Use, Land
Use Change and Forestry, and Waste sectors. Table 9-3 summarizes the quantitative effect of all methodology and
data changes on U.S. greenhouse gas emissions by gas across the Energy, Industrial Processes and Product Use
(IPPU), Agriculture, and Waste sectors. Finally, Table 9-4 similarly summarizes the quantitative effect of
methodology and data changes on annual net fluxes from Land Use, Land-Use Change, and Forestry (LULUCF). The
1 See paragraphs 1 and 2 of the decision on common metrics adopted at the 27th UNFCCC Conference of Parties (COP27)
available online here: https://unfccc.int/sites/default/files/resource/sbsta2022 L25a01E.pdf. The UNFCCC reporting guidelines
require use of the 100-year GWPs listed in table 8.A.1 in Annex 8.A of Chapter 8 of the Fifth Assessment Report of the
Intergovernmental Panel on Climate Change, excluding the value for fossil methane.
2 See Annex to decision 18/CMA.l available online at https://unfccc.int/sites/default/files/resource/CMA2018 03a02E.pdf.
Recalculations and Improvements 9-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
tables below present results relative to the previously published Inventory (i.e., the 1990 to 2020 report) in units of
million metric tons of carbon dioxide equivalent (MMT CO2 Eq.). To understand the details of any specific
recalculation or methodological improvement, see the Recalculations within each source/sink categories' section
found in Chapters 3 through 7 of this report. A discussion of Inventory improvements in response to review
processes is described in Annex 8.
The use of AR5 GWP values in this Inventory results in time-series recalculations for most inventory sources. In
Table 9-1 below, recalculations are presented including both the quantitative effect of the data and
methodological changes as well as the quantitative effect of the change in using the AR5 GWP.
The Inventory includes new categories not included in the previous Inventory that improve completeness of the
national estimates. Specifically, the current report includes CO2 emissions from substitution of ozone depleting
substances, and the reporting of CO2 from the biogenic components of municipal solid waste as a memo item.
The following source and sink categories underwent the most significant methodological and historical data
changes. A brief summary of the recalculations and/or improvements undertaken are provided for these
categories.
• Forest Land Remaining Forest Land: Changes in Forest Carbon Stocks (CO2). The methods used in the
current Inventory to compile estimates for forest ecosystem carbon stocks and stock changes and
harvested wood products (HWPs) from 1990 through 2021 are consistent with those used in the previous
(1990 through 2020) Inventory. Population estimates of carbon stocks and stock changes were compiled
using NFI data from each U.S. state and national estimates were compiled by summing over all states.
New NFI data in most states were incorporated in the latest Inventory which contributed to lower forest
land area estimates and carbon stocks, particularly in Alaska with new data from 2018 to 2021. Fire data
sources were also updated for Alaska through 2021 and this, combined with the new NFI data for the
years 2018 through 2021, resulted in substantial changes in carbon stocks. These changes can be
attributed to obtaining plot-level soil orders using the more refined gridded National Soil Survey
Geographic Database (gNATSGO) dataset (Soil Survey Staff 2020a, 2020b), rather than the Digital General
Soil Map of the United States (STATSG02) dataset which had been used in previous Inventories. This
resulted in a structural change in the soil carbon estimates for mineral and organic soils across the entire
time series, particularly in Alaska where new data on forest area was included for the years 2018 through
2021. Finally, recent land-use change in Alaska (since 2015) also contributed to variability in soil carbon
stocks and stock changes in recent years in the time series. New data included in the HWP time-series
result in a minor decrease (<1 percent) in carbon stocks in the HWP pools but a substantial increase (60
percent) in the carbon stock change estimates for Products in Use and to a lesser extent (2 percent) in
SWDS between the previous Inventory and the current Inventory. With the easing of the global pandemic
and the return of consumers to the marketplace, there was a rebound in the purchase and accumulation
of both paper and solid wood products. These changes resulted in an average annual increase in C stock
change losses of 31.9 MMT CO2 Eq. (4.4 percent), across the 1990 through 2020 time series, relative to
the previous Inventory. See Chapter 6, Section 6.2 for more information on recalculations.
• Wetlands Remaining Wetlands: Emissions from Flooded Land Remaining Flooded Land (CHa). The 1990
through 2021 Inventory uses the National Wetlands Inventory (NWI) as the primary data source for
flooded land surface area, whereas the 1990 through 2020 Inventory report used the National
Hydrography Data (NHD) as the primary geospatial data source. The NWI is far more detailed than the
NHD, resulting in increased emission estimates across the time series. The NWI also includes Alaska,
Hawaii, and Puerto Rico, which were not included in the 1990 through 2020 Inventory. Emissions from
reservoirs in Flooded Land Remaining Flooded Land were further increased by correcting the creation
date of several large reservoirs in South Dakota, North Dakota, Alabama, Arkansas, Georgia, and South
Carolina. These reservoirs were incorrectly classified as Land Converted to Flooded Land for a portion of
the 1990 through 2020 time series but are classified as Flooded Land Remaining Flooded Land throughout
the 1990 through 2021 Inventory time series. The 1990 through 2020 Inventory distinguished between
reservoirs and inundation areas. Inundation areas were defined as periodically flooded lands that
9-2 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
bordered a permanently flooded reservoir. The NWI includes both permanently and periodically flooded
lands, but does not consistently discriminate between them, therefore inundation areas and reservoirs
are consolidated into reservoirs for the 1990 through 2021 Inventory. The net effect of these
recalculations was an average annual increase in Cm emission estimates from reservoirs of 23.4 MMT CO2
Eq. (107.1 percent) over the time series.
• Biomass and Biofuel Consumption (CO2). The CO2 emissions associated with the biogenic components of
MSW combustion were added to this year's report as a memo item. The emissions were calculated based
on the same approach used to develop fossil CO2 emissions from the fossil components of MSW as
described in Section 3.3. The result of these changes was an increase in biogenic CO2 emissions reported
as a memo item relative to the previous Inventory. See Chapter 3, Section 3.10 for more information on
recalculations.
• Petroleum Systems (CHa). In this Inventory, an update that incorporates additional basin-level data from
GHGRP Subpart W was implemented for several emission sources in the onshore production segment,
including for pneumatic controllers, equipment leaks, chemical injection pumps, and storage tanks. For
each of these emission sources, EPA modified the calculation methodology to use GHGRP data to develop
basin-specific activity factors and/or emission factors. The combined impact of revisions to 2020
petroleum systems Cm emission estimates on a CC>2-equivalent basis, compared to the previous
Inventory, is an increase from 45.0 to 54.5 MMT CO2 Eq. (9.4 MMT CO2 Eq., or 20.9 percent). The
recalculations resulted in higher Cm emission estimates on average across the 1990 through 2020 time
series, compared to the previous Inventory, by 5.7 MMT CO2 Eq., or 12.0 percent. See Chapter 3, Section
3.6 for more information on recalculations.
• Land Converted to Grassland: Changes in all Ecosystem Carbon Stocks (CCh). Recalculations are associated
with new FIA data from 1990 to 2021 on biomass, dead wood and litter C stocks associated with
conversions from Cropland Converted to Grassland (woodlands), Other Land Converted to Grassland, and
Settlements Converted to Grassland; updated FIA data from 1990 to 2021 on biomass, dead wood and
litter C stocks from Forest Land Converted to Grassland; and updated estimates for mineral soils from
2016 to 2021 using the linear extrapolation method. As a result, Land Converted to Grassland has an
estimated increase in C stock changes of 2.9 MMT CO2 Eq. (23.2 percent) on average over the time series.
• Land Converted to Cropland: Changes in all Ecosystem Carbon Stocks (CC>2). Recalculations are associated
with new FIA data from 1990 to 2021 on biomass, dead wood and litter C stocks in Grassland Converted
to Cropland (i.e., woodland conversion to cropland), updated FIA data from 1990 to 2021 on biomass,
dead wood and litter C stocks in Forest Land Converted to Cropland, and updated estimates for mineral
soils from 2016 to 2021 using the linear extrapolation method. As a result, Land Converted to Cropland
has an estimated larger C loss of 2.6 MMT CO2 Eq. (4.9 percent) on average over the time series. See
Chapter 6, Section 6.5 for more information on recalculations.
• Natural Gas Systems (CHa). In this Inventory, an update that incorporates additional basin-level data from
GHGRP Subpart W was implemented for several emission sources in the onshore production segment,
including for pneumatic controllers, equipment leaks, chemical injection pumps, storage tanks, and liquids
unloading. For each of these emission sources, EPA modified the calculation methodology to use GHGRP
data to develop basin-specific activity factors and/or emission factors. The combined impact of revisions
to 2020 natural gas systems CH4 emissions, compared to the previous Inventory, is an increase from
184.7to 185.4 MMT CO2 Eq. (0.7 MMT CO2 Eq., or 0.4 percent). The recalculations resulted in an average
increase in the annual CH4 emission estimates across the 1990 through 2020 time series, compared to the
previous Inventory, of 2.6 MMT CO2 Eq., or 1.4 percent. See Chapter 3, Section 3.7 for more information
on recalculations.
• Fossil Fuel Combustion (CO2). Several updates to activity data and emission factors led to recalculations of
previous year results. The major updates include updated data from EIA sources (2022a) for energy
consumption statistics, industrial energy sector activity data, natural gas consumption, and petroleum
Recalculations and Improvements 9-3
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
statistics across the time series relative to the previous Inventory. The carbon content for propylene was
updated from 65.95 kg CCh/MMBtu to 67.77 kg CCh/MMBtu to reflect values used in the EPA Greenhouse
Gas Emission Factors Hub. Fuel consumption for the U.S. Territories provided by ElA's International
Energy Statistics (EIA 2022b) was updated across the time series. Updates were also made to the values of
natural gas used for ammonia production which led to changes in energy sector adjustments. Overall,
these revisions impacted estimates from the combustion of fossil fuels in a number of ways including
decreased petroleum emissions from the residential sector, decreased petroleum emissions from U.S.
Territories, increased natural gas emissions across all economic sectors, and decreased coal emissions
from U.S. Territories. These changes resulted in an average annual increase of 2.5 MMT CO2 Eq. (12
percent) in CO2 emissions from fossil fuel combustion relative to the previous Inventory. See Chapter 3,
Section 3.1 for more information on recalculations.
• Land Converted to Settlements: Changes in all Ecosystem Carbon Stocks (CC>2). Recalculations are
associated with new FIA data from 1990 to 2021 on biomass, dead wood and litter C stocks in Forest Land
Converted to Settlements and woodland conversion associated with Grassland Converted to Settlements,
and updated estimates for mineral and organic soils from 2016 to 2021 using the linear extrapolation
method. As a result, Land Converted to Settlements has an estimated larger C loss of 2.3 MMT CO2 Eq. on
average over the time series. This represents a 2.9 percent increase in C stock changes for Land Converted
to Settlements compared to the previous Inventory. See Chapter 6, Section 6.11 for more information on
recalculations.
• Forest Land Remaining Forest Land: Non-CCh Emissions from Forest Fires (CHa and N2O). The methods
used in the current (1990 through 2021) Inventory to compile estimates of non-CC>2 emissions from forest
fires represent a slight change relative to the previous (1990 through 2020) Inventory. The basic
components of calculating forest fire emissions (IPCC 2006) remain unchanged, but the WFEIS-based
estimates now include estimates of area burned from both MTBS and MODIS as well as two alternate fuel
models to improve consistency across the time series and accuracy with use of updated data. An
additional source of change leading to recalculations are recent and ongoing updates to the MTBS fire
records (i.e., including both most-recent as well as possible updates to past years' fires). The net result of
implementing the improvements listed above was an average annual increase of 2.2 MMT CO2 Eq., or
44.7 percent, in total non-CC>2 emissions from forest fires across the entire time series. See Chapter 6,
Section 6.2 for more information on recalculations.
9-4 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1
2
Figure 9-1: Impacts from Recalculations to U.S. Greenhouse Gas Emissions by Sector,
Including Quantitative Change Related to the Use of AR5 GWP Values
100 — Change in Net Total Emissions
¦ Agriculture
90 ¦ Energy
¦ Industrial Processes and Product Use
80 ¦ LULUCF Sector Net Total
¦ Waste
70
60
i-i-rHiH-rHi-ii-i-rH-nHi-Hi-Hf\j(NrsicNrsj(NcsJrslcN(NrvJ(Nrvj
-------
Petroleum Systems
(0.1)
(1.8)
(0.6)
(1.2)
0.2
(1.1)
(1.2)
Incineration of Waste
(+)
(+)
NC
NC
NC
(0.2)
(+)
Ammonia Production
1.4
1.1
1.4
0.5
0.1
0.3
1.0
Lime Production
NC
NC
NC
NC
NC
NC
NC
Other Process Uses of Carbonates
NC
NC
NC
NC
(1.4)
(1.4)
(+)
Urea Fertilization
NC
NC
(+)
(0.1)
(0.1)
(0.2)
(+)
Carbon Dioxide Consumption
NC
NC
NC
NC
NC
NC
NC
Urea Consumption for Non-Agricultural
Purposes
NC
NC
(+)
0.1
0.1
(0.2)
+
Liming
+
+
(+)
(+)
(0.2)
0.5
(+)
Coal Mining
NC
NC
0.1
0.1
+
+
+
Glass Production
(+)
(+)
(+)
+
+
+
(+)
Soda Ash Production
NC
NC
NC
NC
NC
NC
NC
Ferroalloy Production
NC
NC
NC
NC
NC
NC
+
Aluminum Production
NC
NC
NC
+
(+)
NC
+
Titanium Dioxide Production
NC
NC
NC
NC
NC
(0.1)
NC
Zinc Production
NC
NC
NC
NC
NC
(+)
NC
Phosphoric Acid Production
NC
NC
NC
NC
NC
(+)
NC
Lead Production
NC
NC
NC
+
+
(+)
+
Carbide Production and Consumption
NC
NC
NC
NC
+
+
+
Abandoned Oil and Gas Wells
+
+
+
+
+
+
+
Substitution of Ozone Depleting Substances
+*
+*
+*
+*
+*
+*
+*
Magnesium Production and Processing
NC
NC
NC
NC
NC
NC
NC
Biomass and Biofue1°
18.5
14.7
16.2
16.2
15.8
13.9
15.8
International Bunker Fuelsb
NC
NC
NC
NC
NC
NC
NC
CH4c
87.9
93.7
99.0
103.1
99.0
91.8
95.1
Enteric Fermentation
19.6
20.2
21.0
21.1
21.1
21.0
20.4
Natural Gas Systems
19.6
25.9
19.8
22.6
21.5
20.5
24.3
Landfills
21.2
16.2
14.7
15.0
15.4
15.4
16.9
Manure Management
4.2
5.9
6.9
7.1
7.0
7.1
5.7
Petroleum Systems
3.5
9.6
21.4
22.0
19.5
14.2
10.9
Coal Mining
11.6
7.7
6.6
6.4
5.6
5.0
8.4
Wastewater T reatment
2.4
2.5
3.1
3.1
3.1
3.1
2.6
Rice Cultivation
1.9
2.2
1.8
1.9
1.8
1.9
1.9
Stationary Combustion
1.0
0.9
0.9
1.0
1.1
0.8
1.0
Abandoned Oil and Gas Wells
1.2
1.3
1.3
1.3
1.3
1.3
1.3
Abandoned Underground Coal Mines
0.9
0.8
0.8
0.7
0.7
0.7
0.9
Mobile Combustion
0.7
0.4
0.4
0.4
0.4
0.4
0.5
Composting
+
0.2
0.3
0.3
0.3
0.3
0.2
Field Burning of Agricultural Residues
+
0.1
0.1
0.1
0.1
0.1
+
Petrochemical Production
+
+
+
+
+
+
+
Anaerobic Digestion at Biogas Facilities
+
+
+
+
+
+
+
Ferroalloy Production
+
+
+
+
+
+
+
Carbide Production and Consumption
+
+
+
+
+
+
+
Iron and Steel Production & Metallurgical Coke
Production
+
+
+
+
+
+
+
Incineration of Waste
+
+
+
+
+
+
+
International Bunker Fuelsb
+
+
+
+
+
+
+
N2Oc
(53.8)
(48.3)
(41.8)
(39.2)
(57.7)
(48.4)
(49.8)
Agricultural Soil Management
(37.7)
(33.0)
(29.6)
(26.8)
(47.1)
(36.9)
(35.2)
Stationary Combustion
(2.8)
(3.8)
(3.1)
(3.1)
(2.7)
(2.6)
(3.3)
Wastewater T reatment
(1.8)
(2.2)
(2.6)
(2.4)
(2.1)
(2.7)
(2.2)
Manure Management
(1.5)
(1.8)
(2.1)
(2.1)
(2.2)
(2.2)
(1.8)
Mobile Combustion
(6.2)
(4.3)
(1.6)
(1.6)
(1.0)
(1.3)
(4.2)
9-6 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
Nitric Acid Production
(1.3)
(1.3)
(1.0)
(1.1)
(1.1)
(1.0)
(1.3)
AdipicAcid Production
(1.7)
(0.8)
(0.8)
(1.2)
(0.6)
(0.9)
(0.9)
N20 from Product Uses
(0.5)
(0.5)
(0.5)
(0.5)
(0.5)
(0.5)
(0.5)
Composting
(+)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Caprolactam, Glyoxal, and Glyoxylic Acid
Production
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(+)
(0.2)
Incineration of Waste
(0.1)
(+)
(+)
(+)
(+)
(+)
(+)
Electronics Industry
+
(+)
(+)
(+)
(+)
(+)
(+)
Field Burning of Agricultural Residues
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Petroleum Systems
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Natural Gas Systems
(+)
(+)
(+)
(+)
(+)
(+)
(+)
International Bunker Fuelsb
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
HFCs, PFCs, SF6 and NF3
(8.2)
(8.1)
(9.4)
(9.2)
(9.2)
(9.0)
(9.3)
HFCs
(7.5)
(11.1)
(10.3)
(10.2)
(10.5)
(10.6)
(10.2)
Substitution of Ozone Depleting Substances
+
(7.8)
(9.4)
(9.5)
(9.8)
(10.1)
(6.5)
HCFC-22 Production
(7.5)
(3.2)
(0.8)
(0.5)
(0.6)
(0.3)
(3.7)
Electronics Industry
(+)
(+)
(0.1)
(0.1)
(0.1)
(0.1)
(+)
Magnesium Production and Processing
NC
NC
(+)
(+)
(+)
(+)
(+)
PFCs
(2.4)
(0.6)
(0.4)
(0.5)
(0.6)
(0.5)
(1.1)
Electronics Industry
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.4)
Aluminum Production
(2.2)
(0.3)
(0.1)
(0.2)
(0.3)
(0.2)
(0.7)
Substitution of Ozone Depleting Substances
NC
(+)
(+)
(+)
(+)
(+)
(+)
Electrical Transmission and Distribution
NC
(+)
+
NC
(+)
(+)
(+)
sf6
1.7
3.7
1.4
1.4
1.9
2.1
2.0
Electrical Transmission and Distribution
1.5
3.5
1.3
1.4
1.9
2.1
1.9
Magnesium Production and Processing
0.2
0.1
+
+
+
+
0.1
Electronics Industry
+
0.1
+
+
+
+
+
nf3
(+)
(0.1)
(+)
(+)
(+)
(+)
(+)
Electronics Industry
(+)
(0.1)
(+)
(+)
(+)
(+)
(+)
Total Gross Emissions
24.8
32.1
49.0
55.9
32.6
33.2
33.6
Percent Change in Total Emissions
0.4%
0.4%
0.8%
0.8%
0.5%
0.6%
0.5%
NC (No Change)
+ Absolute value does not exceed 0.05 MMT C02 Eq. or 0.05 percent.
* Indicates a new source for the current Inventory year. Emissions from new sources are captured in net emissions and
percent change totals.
a Emissions from International Bunker Fuels are not included in totals.
b Emissions from Biomass and Biofuel Consumption are not included specifically in summing Energy sector totals. Net
carbon fluxes from changes in biogenic carbon reservoirs are accounted for in the estimates for Land Use, Land-Use
Change, and Forestry.
CLULUCF emissions of CH4 and N20 are reported separately from gross emissions totals in Table 9-2. LULUCF emissions
include the CH4 and N20 emissions reported for Peatlands Remaining Peatlands, forest fires, drained organic soils,
grassland fires, and Coastal Wetlands Remaining Coastal Wetlands; CH4 emissions from Land Converted to Coastal
Wetlands; and N20 emissions from forest soils and settlement soils.
Notes: Net change in total emissions presented without LULUCF. Parentheses indicate negative values. Totals may not sum
due to independent rounding.
Recalculations and Improvements 9-7
-------
1 Table 9-2: Revisions to U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land
2 Use, Land-Use Change, and Forestry, Including Quantitative Change Related to the Use of
3 AR5 GWP Values (MMT COz Eq.)
Land-Use Category
1990
2005
2017
2018
2019
2020
Average
Annual
Change
Forest Land Remaining Forest Land
(46.1)
(21.5)
(25.1)
(28.3)
(6.2)
(41.8)
(30.5)
Changes in Forest Carbon Stocks3
(47.5)
(27.0)
(22.4)
(27.3)
(14.5)
(39.4)
(31.9)
Non-C02 Emissions from Forest Firesb
1.4
5.5
(2.7)
(0.9)
8.3
(2.3)
1.5
N20 Emissions from Forest Soilsc
(+)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(+)
Non-C02 Emissions from Drained Organic
Soilsd
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Land Converted to Forest Land
0.1
0.6
1.2
1.3
1.3
1.3
0.7
Changes in Forest Carbon Stocks6
0.1
0.6
1.2
1.3
1.3
1.3
0.7
Cropland Remaining Cropland
+
+
(+)
(+)
+
(+)
+
Changes in Mineral and Organic Soil
Carbon Stocks
+
+
(+)
(+)
+
(+)
+
Land Converted to Cropland
3.0
2.6
2.3
2.4
2.3
2.3
2.6
Changes in all Ecosystem Carbon Stocks'
3.0
2.6
2.3
2.4
2.3
2.3
2.6
Grassland Remaining Grassland
1.8
2.3
1.6
1.6
1.6
1.5
2.2
Changes in Mineral and Organic Soil
Carbon Stocks
1.8
2.3
1.6
1.6
1.6
1.5
2.2
Non-C02 Emissions from Grassland Fires8
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Land Converted to Grassland
(3.5)
(3.1)
(1.8)
(1.8)
(1.8)
(1.8)
(2.9)
Changes in all Ecosystem Carbon Stocks'
(3.5)
(3.1)
(1.8)
(1.8)
(1.8)
(1.8)
(2.9)
Wetlands Remaining Wetlands
26.8
25.9
25.9
25.9
25.9
26.0
26.1
Changes in Organic Soil Carbon Stocks in
Peatlands
NC
NC
NC
+
+
+
+
Changes in Biomass, DOM, and Soil
Carbon Stocks in Coastal Wetlands
(8.4)
(7.7)
(8.8)
(8.8)
(8.8)
(8.8)
(6.8)
CH4 Emissions from Coastal Wetlands
Remaining Coastal Wetlands
12.6
11.9
13.1
13.1
13.1
13.1
11.1
N20 Emissions from Coastal Wetlands
Remaining Coastal Wetlands
(3.6)
(3.6)
(3.7)
(3.7)
(3.7)
(3.7)
(3.6)
Non-C02 Emissions from Peatlands
Remaining Peatlands
(0.1)
(0.2)
(0.1)
(0.2)
(0.2)
(0.2)
(0.2)
CH4 Emissions from Flooded Land
Remaining Flooded Land
26.4
25.5
25.5
25.5
25.5
25.5
25.7
Land Converted to Wetlands
(3.9)
0.1
0.2
0.2
0.2
(+)
(0.9)
Changes in Biomass, DOM, and Soil
Carbon Stocks in Land Converted to
Coastal Wetlands
+
+
+
+
+
NC
+
CH4 Emissions from Land Converted to
Coastal Wetlands
+
+
+
+
+
+
+
Changes in Land Converted to Flooded
Land
(2.4)
+
0.1
0.1
0.1
(+)
(0.6)
CH4 Emissions from Land Converted to
Flooded Land
(1.5)
+
0.1
0.1
0.1
+
(0.3)
Settlements Remaining Settlements
(0.2)
(0.3)
(0.2)
(0.1)
0.1
(7.9)
(0.5)
Changes in Organic Soil Carbon Stocks
NC
NC
NC
NC
NC
NC
NC
Changes in Settlement Tree Carbon
Stocks
NC
NC
0.2
0.3
0.5
(6.9)
(0.2)
Changes in Yard Trimming and Food Scrap
Carbon Stocks in Landfills
NC
NC
NC
NC
NC
-0.6
(+)
9-8 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
N20 Emissions from Settlement Soilsh
(0.2)
(0.3)
(0.4)
(0.4)
(0.4)
(0.4)
(0.3)
Land Converted to Settlements
1.7
2.2
2.9
3.1
3.2
3.2
2.3
Changes in all Ecosystem Carbon Stocks'
1.7
-> ->
2.9
3.1
3.2
3.2
2.3
Change in LULUCF Total Net Flux'
(46.8)
(22.4)
(15.8)
(20.5)
(7.4)
(40.4)
(27.9)
Change in LULUCF Emissions'
26.5
31.1
22.9
24.6
33.8
23.1
27.0
Change in LULUCF Sector Net Totalk
(20.3)
8.7
7.0
4.1
26.4
(17.2)
(0.9)
Percent Change in LULUCF Total Net Flux
-2.4%
1.1%
0.9%
0.5%
3.6%
-2.3%
0.0%
NC (No Change)
+ Absolute value does not exceed 0.05 MMT C02 Eq. or 0.05 percent.
a Includes the net changes to carbon stocks stored in all forest ecosystem pools and harvested wood products.
b Estimates include CH4 and N20 emissions from fires on both Forest Land Remaining Forest Land and Land Converted to
Forest Land.
c Estimates include N20 emissions from N fertilizer additions on both Forest Land Remaining Forest Land and Land
Converted to Forest Land.
d Estimates include CH4 and N20 emissions from drained organic soils on both Forest Land Remaining Forest Land and Land
Converted to Forest Land.
e Includes the net changes to carbon stocks stored in all forest ecosystem pools.
f Includes changes in mineral and organic soil carbon stocks for all land use conversions to cropland, grassland, and
settlements, respectively. Also includes aboveground/belowground biomass, dead wood, and litter carbon stock changes
for conversion of forest land to cropland, grassland, and settlements, respectively.
g Estimates include CH4 and N20 emissions from fires on both Grassland Remaining Grassland and Land Converted to
Grassland.
h Estimates include N20 emissions from N fertilizer additions on both Settlements Remaining Settlements and Land
Converted to Settlements because it is not possible to separate the activity data at this time.
' LULUCF Carbon Stock Change includes any C stock gains and losses from all land use and land use conversion categories.
j LULUCF emissions include the CH4 and N20 emissions reported for Peatlands Remaining Peatlands, forest fires, drained
organic soils, grassland fires, and Coastal Wetlands Remaining Coastal Wetlands; CH4 emissions from Land Converted to
Coastal Wetlands; and N20 emissions from forest soils and settlement soils.
k The LULUCF Sector Net Total is the net sum of all LULUCF CH4 and N20 emissions to the atmosphere plus LULUCF net
carbon stock changes.
Notes: Parentheses indicate negative values. Totals may not sum due to independent rounding.
1
2 Table 9-3: Revisions to U.S. Greenhouse Gas Emissions, Excluding Quantitative Change
3 Related to the Use of AR5 GWP Values (MMT COz Eq.)
Average
Gas/Source
1990
2005
2017
2018
2019
2020
Annual
Change
co2
(1.0)
(5.2)
1.1
1.3
0.6
(1.3)
(2.4)
Fossil Fuel Combustion
(3.0)
(4.7)
(0.8)
0.5
1.1
2.2
(2.5)
Electric Power Sector
351.0
541.5
(47.9)
(59.3)
(207.0)
(132.5)
354.8
Transportation
(351.0)
(541.5)
48.1
60.0
207.8
133.5
(354.8)
Industrial
(1.3)
(0.7)
(1.4)
(0.6)
(0.2)
1.6
(0.8)
Residential
NC
NC
NC
+
+
(2.7)
(0.1)
Commercial
NC
NC
NC
+
+
1.6
0.1
U.S. Territories
(1.7)
(4.0)
0.5
0.5
0.6
0.6
(1.8)
Non-Energy Use of Fuels
0.2
(+)
0.2
0.6
0.8
(1.8)
0.2
Iron and Steel Production & Metallurgical Coke
Production
NC
NC
0.2
0.2
NC
(+)
+
Cement Production
NC
NC
NC
NC
NC
NC
NC
Natural Gas Systems
0.5
0.3
0.6
0.6
(+)
0.9
0.4
Petrochemical Production
NC
NC
NC
NC
NC
(0.2)
(+)
Petroleum Systems
(0.1)
(1.8)
(0.6)
(1.2)
0.2
(1.1)
(1.2)
Recalculations and Improvements 9-9
-------
Incineration of Waste
(+)
(+)
NC
NC
NC
(0.2)
(+)
Ammonia Production
1.4
1.1
1.4
0.5
0.1
0.3
1.0
Lime Production
NC
NC
NC
NC
NC
NC
NC
Other Process Uses of Carbonates
NC
NC
NC
NC
(1.4)
(1.4)
(0.1)
Urea Fertilization
NC
NC
(+)
(0.1)
(0.1)
(0.2)
(+)
Carbon Dioxide Consumption
NC
NC
NC
NC
NC
NC
NC
Urea Consumption for Non-Agricultural
Purposes
NC
NC
(+)
0.1
0.1
(0.2)
(+)
Liming
+
+
(+)
(+)
(0.2)
0.5
+
Coal Mining
NC
NC
0.1
0.1
+
+
+
Glass Production
(+)
(+)
(+)
+
+
+
(+)
Soda Ash Production
NC
NC
NC
NC
NC
NC
NC
Ferroalloy Production
NC
NC
NC
NC
NC
NC
+
Aluminum Production
NC
NC
NC
+
(+)
NC
+
Titanium Dioxide Production
NC
NC
NC
NC
NC
(0.1)
(+)
Zinc Production
NC
NC
NC
NC
NC
(+)
(+)
Phosphoric Acid Production
NC
NC
NC
NC
NC
(+)
(+)
Lead Production
NC
NC
NC
+
+
(+)
+
Carbide Production and Consumption
NC
NC
NC
NC
+
+
+
Abandoned Oil and Gas Wells
+
+
+
+
+
+
+
Substitution of Ozone Depleting Substances
NC*
NC*
NC*
NC*
NC*
NC*
NC*
Magnesium Production and Processing
(+)
(+)
+
+
+
+
+
Biomass and Biofuel Consumptiona
18.5
14.7
16.2
16.2
15.8
13.9
15.7
International Bunker Fuelsb
NC
NC
NC
NC
NC
NC
NC
<4C
(5.8)
10.0
19.4
22.5
18.7
13.8
9.4
Enteric Fermentation
NC
NC
NC
NC
NC
NC
NC
Natural Gas Systems
(3.9)
4.6
(0.1)
2.0
0.8
0.7
2.6
Landfills
NC
0.4
1.6
1.6
1.7
2.3
0.4
Manure Management
NC
NC
NC
NC
NC
NC
+
Petroleum Systems
(2.2)
4.6
16.5
17.4
14.7
9.4
5.7
Coal Mining
NC
+
+
+
(0.1)
+
(+)
Wastewater T reatment
(+)
0.1
0.8
0.9
0.9
0.9
0.3
Rice Cultivation
NC
NC
NC
NC
NC
(+)
(+)
Stationary Combustion
(+)
(+)
+
+
+
(0.1)
(+)
Abandoned Oil and Gas Wells
0.4
0.4
0.5
0.5
0.5
0.5
0.4
Abandoned Underground Coal Mines
NC
NC
NC
NC
NC
+
+
Mobile Combustion
(0.1)
(0.1)
0.1
0.1
0.1
0.1
+
Composting
NC
NC
NC
NC
(+)
+
+
Field Burning of Agricultural Residues
NC
NC
NC
NC
NC
NC
(+)
Petrochemical Production
NC
NC
NC
NC
NC
(+)
(+)
Anaerobic Digestion at Biogas Facilities
NC
NC
NC
NC
NC
NC
NC
Ferroalloy Production
NC
NC
NC
NC
NC
NC
+
Carbide Production and Consumption
NC
NC
NC
NC
NC
NC
+
Iron and Steel Production & Metallurgical Coke
Production
NC
NC
NC
NC
NC
(+)
(+)
Incineration of Waste
NC
NC
NC
NC
NC
(+)
(+)
International Bunker Fuelsb
NC
NC
NC
NC
NC
NC
NC
Oc
(3.9)
2.0
7.4
11.5
(7.1)
(1.2)
0.5
Agricultural Soil Management
(2.7)
1.7
6.8
10.7
(8.9)
(1.9)
0.2
Stationary Combustion
(+)
(+)
+
+
+
(+)
(+)
Wastewater T reatment
(+)
(+)
+
0.2
0.5
(+)
+
Manure Management
NC
NC
NC
NC
NC
NC
NC
Mobile Combustion
(1.2)
0.2
0.6
0.5
1.2
0.7
0.3
Nitric Acid Production
NC
NC
NC
NC
NC
NC
NC
9-10 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
AdipicAcid Production
NC
NC
NC
NC
NC
NC
NC
N20 from Product Uses
NC
NC
NC
NC
NC
NC
NC
Composting
NC
NC
NC
NC
(+)
+
+
Caprolactam, Glyoxal, and Glyoxylic Acid
Production
NC
NC
NC
NC
NC
0.1
+
Incineration of Waste
NC
NC
NC
NC
NC
(+)
(+)
Electronics Industry
+
+
+
+
+
(+)
+
Field Burning of Agricultural Residues
NC
NC
NC
NC
NC
NC
(+)
Petroleum Systems
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Natural Gas Systems
+
+
(+)
(+)
(+)
(+)
+
International Bunker Fuelsb
NC
NC
NC
NC
NC
NC
NC
HFCs, PFCs, SF6 and NF3
0.8
3.2
1.1
1.2
1.5
1.9
1.5
HFCs
+
(0.1)
(0.1)
(+)
(+)
+
(0.1)
Substitution of Ozone Depleting Substances
NC
(0.1)
(0.1)
(+)
(+)
(+)
(0.1)
HCFC-22 Production
NC
NC
NC
NC
NC
NC
NC
Electronics Industry
+
+
+
+
+
+
+
Magnesium Production and Processing
NC
NC
NC
NC
+
NC
+
PFCs
+
(+)
+
(0.1)
(0.2)
(0.1)
(+)
Electronics Industry
+
(+)
+
(+)
+
(+)
(+)
Aluminum Production
NC
+
+
(0.1)
(0.2)
(0.1)
(+)
Substitution of Ozone Depleting Substances
NC
NC
NC
NC
NC
NC
NC
Electrical Transmission and Distribution
NC
(+)
+
NC
NC
NC
(+)
sf6
0.8
3.3
1.2
1.3
1.7
1.9
1.6
Electrical Transmission and Distribution
0.8
3.2
1.2
1.3
1.7
1.9
1.6
Magnesium Production and Processing
4.8
2.1
0.3
0.3
0.1
0.1
2.3
Electronics Industry
(4.8)
(2.0)
(0.3)
(0.3)
(0.1)
(0.1)
(2.3)
nf3
NC
(+)
+
(+)
(+)
(+)
(+)
Electronics Industry
NC
(+)
+
(+)
(+)
(+)
(+)
Total Gross Emissions
(10.0)
9.9
29.1
36.5
13.8
13.1
9.1
Percentage Change in Total Emissions
-0.2%
0.1%
0.4%
0.5%
0.2%
0.2%
0.1%
NC (No Change)
+ Absolute value does not exceed 0.05 MMT C02 Eq. or 0.05 percent.
* Indicates a new source for the current Inventory year. Emissions from new sources are captured in net emissions and
percent change totals.
a Emissions from International Bunker Fuels are not included in totals.
b Emissions from Biomass and Biofuel Consumption are not included specifically in summing Energy sector totals. Net
carbon fluxes from changes in biogenic carbon reservoirs are accounted for in the estimates for Land Use, Land-Use
Change, and Forestry.
c LULUCF emissions of CH4 and N20 are reported separately from gross emissions totals in Table 9-2. LULUCF emissions
include the CH4 and N20 emissions reported for Peatlands Remaining Peatlands, forest fires, drained organic soils,
grassland fires, and Coastal Wetlands Remaining Coastal Wetlands; CH4 emissions from Land Converted to Coastal
Wetlands; and N20 emissions from forest soils and settlement soils.
Notes: Net change in total emissions presented without LULUCF. Parentheses indicate negative values. Totals may not sum
due to independent rounding.
1 Table 9-4: Revisions to U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land
2 Use, Land-Use Change, and Forestry, Excluding Quantitative Change Related to the Use of
3 AR5 GWP Values (MMT COz Eq.)
Average
Annual
Land-Use Category
1990
2005
2015
2016
2017
2018
Change
Forest Land Remaining Forest Land
(46.1)
(21.5)
(25.1)
(28.3)
(6.2)
(41.8)
(30.5)
Changes in Forest Carbon Stocks3
(47.5)
(27.0)
(22.4)
(27.3)
(14.5)
(39.4)
(31.9)
Non-C02 Emissions from Forest Firesb
1.4
5.5
(2.7)
(0.9)
8.3
(2.3)
1.5
Recalculations and Improvements 9-11
-------
N20 Emissions from Forest Soilsc
(+)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(+)
Non-C02 Emissions from Drained Organic
Soilsd
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Land Converted to Forest Land
0.1
0.6
1.2
1.3
1.3
1.3
0.7
Changes in Forest Carbon Stocks6
0.1
0.6
1.2
1.3
1.3
1.3
0.7
Cropland Remaining Cropland
+
+
(+)
(+)
+
(+)
+
Changes in Mineral and Organic Soil
Carbon Stocks
+
+
(+)
(+)
+
(+)
+
Land Converted to Cropland
3.0
2.6
2.3
2.4
2.3
2.3
2.6
Changes in all Ecosystem Carbon Stocks'
3.0
2.6
2.3
2.4
2.3
2.3
2.6
Grassland Remaining Grassland
1.8
2.3
1.6
1.6
1.6
1.5
2.2
Changes in Mineral and Organic Soil
Carbon Stocks
1.8
2.3
1.6
1.6
1.6
1.5
2.2
Non-C02 Emissions from Grassland Fires8
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Land Converted to Grassland
(3.5)
(3.1)
(1.8)
(1.8)
(1.8)
(1.8)
(2.9)
Changes in all Ecosystem Carbon Stocks'
(3.5)
(3.1)
(1.8)
(1.8)
(1.8)
(1.8)
(2.9)
Wetlands Remaining Wetlands
26.8
25.9
25.9
25.9
25.9
26.0
26.1
Changes in Organic Soil Carbon Stocks in
Peatlands
NC
NC
NC
+
+
+
+
Changes in Biomass, DOM, and Soil
Carbon Stocks in Coastal Wetlands
+
+
+
+
+
+
+
CH4 Emissions from Coastal Wetlands
Remaining Coastal Wetlands
+
(0.1)
+
+
+
+
+
N20 Emissions from Coastal Wetlands
Remaining Coastal Wetlands
0.4
0.5
0.5
0.5
0.5
0.5
0.5
Non-C02 Emissions from Peatlands
Remaining Peatlands
(+)
(+)
(+)
(+)
(+)
(+)
(+)
CH4 Emissions from Flooded Land
Remaining Flooded Land
26.4
25.5
25.5
25.5
25.5
25.5
25.7
Land Converted to Wetlands
(3.9)
0.1
0.2
0.2
0.2
(+)
(0.9)
Changes in Biomass, DOM, and Soil
Carbon Stocks in Land Converted to
Coastal Wetlands
+
+
+
+
+
NC
+
CH4 Emissions from Land Converted to
Coastal Wetlands
+
+
+
+
+
+
+
Changes in Land Converted to Flooded
Land
(2.4)
+
0.1
0.1
0.1
(+)
(0.6)
CH4 Emissions from Land Converted to
Flooded Land
(1.5)
+
0.1
0.1
0.1
+
(0.3)
Settlements Remaining Settlements
0.4
0.6
0.5
0.6
0.8
(7.2)
0.3
Changes in Organic Soil Carbon Stocks
NC
NC
NC
NC
NC
NC
NC
Changes in Settlement Tree Carbon
Stocks
NC
NC
0.2
0.3
0.5
(6.9)
(0.2)
Changes in Yard Trimming and Food Scrap
Carbon Stocks in Landfills
NC
NC
NC
NC
NC
-0.6
(+)
N20 Emissions from Settlement Soilsh
0.4
0.6
0.3
0.3
0.3
0.3
0.5
Land Converted to Settlements
1.7
2.2
2.9
3.1
3.2
3.2
2.3
Changes in all Ecosystem Carbon Stocks'
1.7
2.2
2.9
3.1
3.2
3.2
2.3
Change in LULUCF Total Net Flux'
(46.8)
(22.4)
(15.8)
(20.5)
(7.4)
(40.4)
(27.9)
Change in LULUCF Emissions'
24.6
30.7
22.4
23.8
32.3
23.4
25.9
Change in LULUCF Sector Net Totalk
(22.2)
8.3
6.6
3.3
24.9
(17.0)
(2.0)
Percent Change in LULUCF Total Net Flux
-2.6%
1.1%
0.4%
0.4%
3.4%
-2.2%
-0.1%
NC (No Change)
+ Absolute value does not exceed 0.05 MMT C02 Eq. or 0.05 percent.
9-12 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
a Includes the net changes to carbon stocks stored in all forest ecosystem pools and harvested wood products.
b Estimates include CH4 and N20 emissions from fires on both Forest Land Remaining Forest Land and Land Converted to
Forest Land.
c Estimates include N20 emissions from N fertilizer additions on both Forest Land Remaining Forest Land and Land
Converted to Forest Land.
d Estimates include CH4 and N20 emissions from drained organic soils on both Forest Land Remaining Forest Land and Land
Converted to Forest Land.
e Includes the net changes to carbon stocks stored in all forest ecosystem pools.
f Includes changes in mineral and organic soil carbon stocks for all land use conversions to cropland, grassland, and
settlements, respectively. Also includes aboveground/belowground biomass, dead wood, and litter carbon stock changes
for conversion of forest land to cropland, grassland, and settlements, respectively.
g Estimates include CH4 and N20 emissions from fires on both Grassland Remaining Grassland and Land Converted to
Grassland.
h Estimates include N20 emissions from N fertilizer additions on both Settlements Remaining Settlements and Land
Converted to Settlements because it is not possible to separate the activity data at this time.
' LULUCF Carbon Stock Change includes any C stock gains and losses from all land use and land use conversion categories.
' LULUCF emissions include the CH4 and N20 emissions reported for Peatlands Remaining Peatlands, forest fires, drained
organic soils, grassland fires, and Coastal Wetlands Remaining Coastal Wetlands; CH4 emissions from Land Converted to
Coastal Wetlands; and N20 emissions from forest soils and settlement soils.
k The LULUCF Sector Net Total is the net sum of all LULUCF CH4 and N20 emissions to the atmosphere plus LULUCF net
carbon stock changes.
Notes: Parentheses indicate negative values. Totals may not sum due to independent rounding.
Recalculations and Improvements 9-13
-------
2 References and Abbreviations
3 Executive Summary
4 BEA (2022) 2021 Comprehensive Revision of the National Income and Product Accounts: Current-dollar and "real"
5 GDP, 1929-2021. Bureau of Economic Analysis (BEA), U.S. Department of Commerce, Washington, D.C. Available
6 online at: http://www.bea.gOv/national/index.htm#gdp.
7 EIA (2022) Electricity Generation. Monthly Energy Review, November 2022. Energy Information Administration, U.S.
8 Department of Energy, Washington, D.C. DOE/EIA-0035(2019/11).
9 EIA (2021) Electricity in the United States. Electricity Explained. Energy Information Administration, U.S.
10 Department of Energy, Washington, D.C. Available online at:
11 https://www. eia.gov/energyexplained/index. php?page=electricity in the united states.
12 EPA (2022) "1970 - 2021 Average annual emissions, all criteria pollutants in MS Excel." National Emissions
13 Inventory (NEI) Air Pollutant Emissions Trends Data. Office of Air Quality Planning and Standards, February 2022.
14 Available online at: https://www.epa.gov/air-emissions-inventories/air-pollutant-emissions-trends-data.
15 EPA (2021a) Acid Rain Program Dataset 1996-2020. Office of Air and Radiation, Office of Atmospheric Programs,
16 U.S. Environmental Protection Agency, Washington, D.C.
17 EPA (2021b) Greenhouse Gas Reporting Program (GHGRP). 2020 Envirofacts. Subpart HH: Municipal Solid Waste
18 Landfills and Subpart TT: Industrial Waste Landfills. Available online at: https://www.epa.gov/enviro/greenhouse-
19 gas-customized-search.
20 EPA (1997) Compilation of Air Pollutant Emission Factors, AP-42. Office of Air Quality Planning and Standards, U.S.
21 Environmental Protection Agency. Research Triangle Park, NC. October 1997.
22 FHWA (1996 through 2021) Highway Statistics. Federal Highway Administration, U.S. Department of
23 Transportation, Washington, D.C. Report FHWA-PL-96-024-annual. Available online at:
24 http://www.fhwa.dot.gov/policy/ohpi/hss/hsspubs.htm.
25 IEA (2021) CO2 Emissions from Fossil Fuel Combustion - Overview. International Energy Agency. Available online
26 at: https://www.iea.org/subscribe-to-data-services/co2-emissions-statistics.
References and Abbreviations 9-1
-------
1 IPCC (2021) Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth
2 Assessment Report of the Intergovernmental Panel on Climate Change. [Masson-Delmotte, V., P. Zhai, A. Pirani, S.
3 L. Connors, C. Pean, S. Berger, N. Caud, Y. Chen, L Goldfarb, M. I. Gomis, M. Huang, K. Leitzell, E. Lonnoy, J.B.R.
4 Matthews, T. K. Maycock, T. Waterfield, O. Yelekgi, R. Yu and B. Zhou (eds.)]. Cambridge University Press. In Press.
5 IPCC (2019) 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories, e [Buendia, E.,
6 Guendehou S., Limmeechokachai B., Pipatti R., Rojas Y., Sturgiss R., Tanabe K., Wirth T., (eds.)]. Cambridge
7 University Press. In Press.
8 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
9 Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F., D. Qin, G.-K., Plattner, M.
10 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
11 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
12 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
13 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
14 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom
15 996 pp.
16 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
17 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
18 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
19 IPCC (1996) Climate Change 1995: The Science of Climate Change. Intergovernmental Panel on Climate Change.
20 [J.T. Houghton, LG. Meira Filho, B.A. Callander, N. Harris, A. Kattenberg, and K. Maskell (eds.)]. Cambridge
21 University Press. Cambridge, United Kingdom.
22 National Academies of Sciences, Engineering, and Medicine (2018) Improving Characterization of Anthropogenic
23 Methane Emissions in the United States. Washington, DC: The National Academies Press. Available online at:
24 https://doi.org/10.17226/24987.
25 National Research Council (2010) Verifying Greenhouse Gas Emissions: Methods to Support International Climate
26 Agreements. Washington, DC: The National Academies Press. Available online at: https://doi.org/10.17226/12883.
27 NOAA/ESRL (2023a) Trends in Atmospheric Carbon Dioxide. Available online at: https://gml.noaa.gov/ccgg/trends/.
28 05 January 2023.
29 NOAA/ESRL (2023b) Trends in Atmospheric Methane. Available online at: https://gml.noaa.gov/ccgg/trends ch4/.
30 05 January 2023.
31 NOAA/ESRL (2023c) Nitrous Oxide (N2O) hemispheric and global monthly means from the NOAA/ESRL
32 Chromatograph for Atmospheric Trace Species data from baseline observatories (Barrow, Alaska; Summit,
33 Greenland; Niwot Ridge, Colorado; Mauna Loa, Hawaii; American Samoa; South Pole). Available online at:
34 https://www.esrl.noaa.gov/gmd/ccgg/trends n2o/. 05 January 2022.
35 UNFCCC (2014) Report of the Conference of the Parties on its Nineteenth Session, Held in Warsaw from 11 to 23
36 November 2013. (FCCC/CP/2013/10/Add.3). January 31, 2014. Available online at:
37 http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
38 U.S. Census Bureau (2021) U.S. Census Bureau International Database (IDB). Available online at:
39 https://www.census.gov/programs-surveys/international-programs.html.
9-2 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
i Introduction
2 IPCC (2021) Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth
3 Assessment Report of the Intergovernmental Panel on Climate Change. [Masson-Delmotte, V., P. Zhai, A. Pirani, S.L
4 Connors, C. Pean, S. Berger, N. Caud, Y. Chen, L. Goldfarb, M.I. Gomis, M. Huang, K. Leitzell, E. Lonnoy, J.B.R.
5 Matthews, T.K. Maycock, T. Waterfield, O. Yelekgi, R. Yu, and B. Zhou (eds.)]. Cambridge University Press. In Press
6 IPCC (2014) Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group III to the Fifth
1 Assessment Report of the Intergovernmental Panel on Climate Change [Edenhofer, O., R. Pichs-Madruga, Y.
8 Sokona, J. Minx, E. Farahani, S. Kadner, K. Seyboth, A. Adler, I. Baum, S. Brunner, P. Eickemeier, B. Kriemann, J.
9 Savolainen, S. Schlomer, C. von Stechow, and T. Zwickel (eds.)]. Cambridge University Press, Cambridge, United
10 Kingdom and New York, NY, USA, 1435 pp.
11 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
12 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
13 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
14 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
15 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
16 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
17 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom
18 996 pp.
19 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
20 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
21 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
22 IPCC (2001) Climate Change 2001: The Scientific Basis. Intergovernmental Panel on Climate Change. [J.T. Houghton,
23 Y. Ding, D.J. Griggs, M. Noguer, P.J. van der Linden, X. Dai, C.A. Johnson, and K. Maskell (eds.)]. Cambridge
24 University Press. Cambridge, United Kingdom.
25 IPCC/TEAP (2005) Special Report: Safeguarding the Ozone Layer and the Global Climate System, Chapter 4:
26 Refrigeration. 2005. Available online at: https://www.ipcc.ch/site/assets/uploads/2018/03/sroc04-l.pdf.
27 NOAA (2017) Vital Signs of the Planet. Available online at: http://climate.nasa.gov/causes/. Accessed on 9 January
28 2017.
29 NOAA/ESRL (2023a) Trends in Atmospheric Carbon Dioxide. Available online at:
30 https://gml.noaa.gov/ccgg/trends/gr.html. February 2, 2023.
31 NOAA/ESRL (2023b) Trends in Atmospheric Methane. Available online at: https://gml.noaa.gov/ccgg/trends ch4/.
32 February 2, 2023.
33 NOAA/ESRL (2023c) Trends in Atmospheric Nitrous Oxide. Available online at:
34 https://gml.noaa.gov/ccgg/trends n2o/. February 2, 2023.
35 NOAA/ESRL (2023d) Trends in Atmospheric Sulfur Hexafluoride. Available online at:
36 https://gml.noaa.gov/ccgg/trends sf6/. February 2, 2023.
37 UNEP/WMO (1999) Information Unit on Climate Change. Framework Convention on Climate Change. Available
38 online at: http://unfccc.int.
39 UNFCCC (2014) Report of the Conference of the Parties on its nineteenth session, held in Warsaw from 11 to 23
40 November 2013. (FCCC/CP/2013/10/Add.3). January 31, 2014. Available online at:
41 http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
References 9-3
-------
1 USGCRP (2017) Climate Science Special Report: Fourth National Climate Assessment, Volume I. [Wuebbles, D.J.,
2 D.W. Fahey, K.A. Hibbard, D.J. Dokken, B.C. Stewart, and T.K. Maycock (eds.)]. U.S. Global Change Research
3 Program, Washington, DC, USA, 470 pp, doi: 10.7930/J0J964J6. Available online at:
4 https://science2017.globalchange.gov/.
5 WMO/UNEP (2018) Scientific Assessment of Ozone Depletion: 2018. Available online at:
6 https://csl.noaa.gov/assessments/ozone/2018.
7 Trends in Greenhouse Gas Emissions
8 BEA (2022) 2021 Comprehensive Revision of the National Income and Product Accounts: Current-dollar and "real"
9 GDP, 1929-2021. Bureau of Economic Analysis (BEA), U.S. Department of Commerce, Washington, D.C. Available
10 online at: http://www.bea.gOv/national/index.htm#gdp.
11 EIA (2022) Monthly Energy Review, November 2022. Energy Information Administration, U.S. Department of
12 Energy, Washington, D.C. DOE/EIA-0035(2022/11).
13 EIA (1991 through 2021) Fuel Oil and Kerosene Sales. Energy Information Administration, U.S. Department of
14 Energy. Washington, D.C. Available online at: http://www.eia.gov/petroleum/fueloilkerosene.
15 EIA (2018) "In 2017, U.S. electricity sales fell by the greatest amount since the recession" Available online at:
16 https://www. eia.gov/todayinenergy/detail. php?id=35612.
17 EPA (2022a) "Crosswalk of Precursor Gas Categories." U.S. Environmental Protection Agency. April 6, 2022.
18 EPA (2022b) Light-Duty Automotive Technology, Carbon Dioxide Emissions, and Fuel Economy Trends: 1975 - 2020.
19 Office of Transportation and Air Quality, U.S. Environmental Protection Agency. Available online at:
20 https://www.epa.gov/fuel-economy/trends-report.
21 EPA (2022c) 1970 - 2021 Average annual emissions, all criteria pollutants in MS Excel. National Emissions Inventory
22 (NEI) Air Pollutant Emissions Trends Data. Office of Air Quality Planning and Standards, November 2022. Available
23 online at: https://www.epa.gov/air-emissions-inventories/air-pollutant-emissions-trends-data.
24 IPCC (2021) Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth
25 Assessment Report of the Intergovernmental Panel on Climate Change. [Masson-Delmotte, V., P. Zhai, A. Pirani, S.
26 L. Connors, C. Pean, S. Berger, N. Caud, Y. Chen, L Goldfarb, M. I. Gomis, M. Huang, K. Leitzell, E. Lonnoy, J.B.R.
27 Matthews, T. K. Maycock, T. Waterfield, O. Yelekgi, R. Yu and B. Zhou (eds.)]. Cambridge University Press. In Press.
28 IPCC (2019) 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories. [Buendia, E.,
29 Guendehou S., Limmeechokachai B., Pipatti R., Rojas Y., Sturgiss R., Tanabe K., Wirth T., (eds.)]. Cambridge
30 University Press. In Press.
31 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
32 Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F., D. Qin, G.-K., Plattner, M.
33 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
34 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
35 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
36 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
37 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom
38 996 pp.
39 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
40 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
41 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
9-4 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 U.S. Census Bureau (2021) U.S. Census Bureau International Database (IDB). Available online at:
2 https://www.census.gov/programs-surveys/intemational-programs.html.
3 U.S. Department of Agriculture, National Agricultural Statistics Service (USDA/NASS) (2020) Farm Production
4 Expenditures Annual Summary. National Agricultural Statistics Service, U.S. Department of Agriculture, Washington
5 DC. Available online at: https://usda.library.cornell.edu/concern/publications/qz20ss48r?locale=en.
e Energy
7 EIA (2022) Monthly Energy Review, November 2022, Energy Information Administration, U.S. Department of
8 Energy, Washington, DC. DOE/EIA-0035(2019/11).
9 IEA (2022) Energy related CO2 emissions, 1990-2021, International Energy Agency, Paris. Available online at:
10 https://www.iea.org/data-and-statistics/charts/global-energy-related-co2-emissions-1990-2021.
11 Carbon Dioxide Emissions from Fossil Fuel Combustion
12 AAR (2008 through 2022) Railroad Facts. Policy and Economics Department, Association of American Railroads,
13 Washington, D.C. Private communication with Dan Keen.
14 AISI (2004 through 2021) Annual Statistical Report, American Iron and Steel Institute, Washington, D.C.
15 APTA (2007 through 2020) Public Transportation Fact Book. American Public Transportation Association,
16 Washington, D.C. Available online at: http://www.apta.com/resources/statistics/Pages/transitstats.aspx.
17 APTA (2006) Commuter Rail National Totals. American Public Transportation Association, Washington, D.C.
18 BEA (2022) Table 1.1.6. Real Gross Domestic Product, Chained 2012 Dollars. Bureau of Economic Analysis (BEA),
19 U.S. Department of Commerce, Washington, D.C. February 2021. Available online at:
20 https://apps.bea.gov/iTable/?reqid=19&step=3&isuri=l&select all years=0&nipa table list=6&series=a&first yea
21 r=1950&last year=1959&scale=-9&categories=survey&thetable=
22 BEA (1991 through 2015) Unpublished BE-36 survey data. Bureau of Economic Analysis, U.S. Department of
23 Commerce. Washington, D.C.
24 Benson, D. (2002 through 2004) Unpublished data. Upper Great Plains Transportation Institute, North Dakota State
25 University and American Short Line & Regional Railroad Association.
26 Browning (2022a) Addressing the Time Series Inconsistency in FHWA Data. Memorandum from ICF to Sarah
27 Roberts, Office of Transportation and Air Quality, U.S. Environmental Protection Agency. September 2022.
28 Browning (2022b) Updated Methodology for Estimating CH4 and N2O Emissions from Highway Vehicle Alternative
29 Fuel Vehicles. Memorandum from ICF to Sarah Roberts, Office of Transportation and Air Quality, U.S.
30 Environmental Protection Agency. November 2022.
31 Browning, L. (2020) GHG Inventory EF Development Using Certification Data. Technical Memo, September 2020.
32 Browning, L (2019) Updated On-highway Cm and N2O Emission Factors for GHG Inventory. Memorandum from ICF
33 to Sarah Roberts, Office of Transportation and Air Quality, U.S. Environmental Protection Agency. September 2019.
34 Browning, L (2018a) Updated Methodology for Estimating Electricity Use from Highway Plug-In Electric Vehicles.
35 Technical Memo, October 2018.
36 Browning, L (2018b) Updated Non-Highway CH4 and N2O Emission Factors for U.S. GHG Inventory. Technical
37 Memo, November 2018.
References 9-5
-------
1 Browning, L (2017) Updated Methodology for Estimating Cm and N2O Emissions from Highway Vehicle Alternative
2 Fuel Vehicles. Technical Memo, October 2017.
3 Coffeyville Resources Nitrogen Fertilizers (2012) Nitrogen Fertilizer Operations. Available online at:
4 http://coffevvillegroup.com/NitrogenFertilizerOperations/index.html.
5 Dakota Gasification Company (2006) CO2 Pipeline Route and Designation Information. Bismarck, ND.
6 DHS (2008) Email Communication. Elissa Kay, Department of Homeland Security and Joe Aamidor, ICF
7 International. January 11, 2008.
8 DLA Energy (2022) Unpublished data from the Fuels Automated System (FAS). Defense Logistics Agency Energy,
9 U.S. Department of Defense. Washington, D.C.
10 DOC (1991 through 2022) Unpublished Report of Bunker Fuel Oil Laden on Vessels Cleared for Foreign Countries.
11 Form-563. Foreign Trade Division, Bureau of the Census, U.S. Department of Commerce. Washington, D.C.
12 DOE (1991 through 2020) Transportation Energy Data Book. Edition 40. Office of Transportation Technologies,
13 Center for Transportation Analysis, Energy Division, Oak Ridge National Laboratory. ORNL-6978. Personal
14 Communication between Stacy Davis (DOE) and Deep Shah (ICF) for sharing selected tables from the pre-release
15 version.
16 DOE (2012) 2010 Worldwide Gasification Database. National Energy Technology Laboratory and Gasification
17 Technologies Council. Available online at:
18 http://www.netl.doe.gov/technologies/coalpower/gasification/worlddatabase/index.html. Accessed on 15 March
19 2012.
20 DOT (1991 through 2022) Airline Fuel Cost and Consumption. U.S. Department of Transportation, Bureau of
21 Transportation Statistics, Washington, D.C. DAI-10. Available online at: http://www.transtats.bts.gov/fuel.asp.
22 Eastman Gasification Services Company (2011) Project Data on Eastman Chemical Company's Chemicals-from-Coal
23 Complex in Kingsport, TN.
24 EIA (2022a) Monthly Energy Review, November 2022, Energy Information Administration, U.S. Department of
25 Energy, Washington, DC. DOE/EIA-0035 (2022/11).
26 EIA (2022b) International Energy Statistics 1980-2021. Energy Information Administration, U.S. Department of
27 Energy. Washington, D.C. Available online at: https://www.eia.gov/beta/international/.
28 EIA (2022c) Quarterly Coal Report: January - June 2022. Energy Information Administration, U.S. Department of
29 Energy. Washington, D.C. DOE/EIA-0121.
30 EIA (2022d) Form EIA-923 detailed data with previous form data (EIA-906/920), Energy Information Administration,
31 U.S. Department of Energy. Washington, DC. DOE/EIA. October 2022.
32 EIA (2022e) Electric Power Annual 2021. Energy Information Administration, U.S. Department of Energy.
33 Washington, D.C. Available online at: https://www.eia.gov/electricity/annual/.
34 EIA (2022e) Natural Gas Annual 2021. Energy Information Administration, U.S. Department of Energy. Washington,
35 D.C. DOE/EIA-0131(20).
36 EIA (2022f) Annual Coal Report 2021. Energy Information Administration, U.S. Department of Energy. Washington,
37 D.C. DOE/EIA-0584.
38 EIA (2022g) "Energy use in homes." Use of energy explained. Available online at:
39 https://www.eia.gov/energyexplained/use-of-energy/homes.php.
40 EIA (2020a) Glossary. Energy Information Administration, U.S. Department of Energy, Washington, D.C. Available
41 online at: https://www.eia.gov/tools/glossary/?id=electricity.
9-6 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 EIA (2020b) "Natural gas prices, production, consumption, and exports increased in 2019." Today in Energy.
2 Available online at: https://www.eia.gov/todavinenergy/detail.php?id=37892.
3 EIA (2018) "Both natural gas supply and demand have increased from year-ago levels." Today in Energy. Available
4 online at: https://www.eia.gov/todavinenergy/detail.php?id=37193.
5 EIA (1991 through 2022) Fuel Oil and Kerosene Sales. Energy Information Administration, U.S. Department of
6 Energy. Washington, D.C. Available online at: http://www.eia.gov/petroleum/fueloilkerosene.
7 EIA (2009a) Emissions of Greenhouse Gases in the United States 2008, Draft Report. Office of Integrated Analysis
8 and Forecasting, Energy Information Administration, U.S. Department of Energy. Washington, D.C. DOE-EIA-
9 0573(2009).
10 EIA (2009b) Manufacturing Consumption of Energy (MECS) 2006. U.S. Department of Energy, Energy Information
11 Administration, Washington, D.C. Released July 2009.
12 EIA (2008) Historical Natural Gas Annual, 1930 - 2008. Energy Information Administration, U.S. Department of
13 Energy. Washington, D.C.
14 EIA (2007) Personal Communication. Joel Lou, Energy Information Administration and Aaron Beaudette, ICF
15 International. Residual and Distillate Fuel Oil Consumption for Vessel Bunkering (Both International and Domestic)
16 for American Samoa, U.S. Pacific Islands, and Wake Island. October 24, 2007.
17 EIA (2002) Alternative Fuels Data Tables. Energy Information Administration, U.S. Department of Energy.
18 Washington, D.C. Available online at: https://www.eia.gov/renewable/.
19 EIA (2001) U.S. Coal, Domestic and International Issues. Energy Information Administration, U.S. Department of
20 Energy. Washington, D.C. March 2001.
21 EIA (1990-2001) State Energy Data System. Energy Information Administration, U.S. Department of Energy.
22 Washington, D.C.
23 Environment and Climate Change Canada (2022) Personal Communication between Environment and Climate
24 Change Canada and Vincent Camobreco for imported CO2. March 2022.
25 EPA (2022a) Acid Rain Program Dataset 1996-2021. Office of Air and Radiation, Office of Atmospheric Programs,
26 U.S. Environmental Protection Agency, Washington, D.C.
27 EPA (2020c) EPA Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2019: Updated Gasoline and Diesel
28 Fuel CO2 Emission Factors - Memo.
29 EPA (2022b) Light-Duty Automotive Technology, Carbon Dioxide Emissions, and Fuel Economy Trends: 1975 - 2020.
30 Office of Transportation and Air Quality, U.S. Environmental Protection Agency. Available online at:
31 https://www.epa.gov/fuel-economy/trends-report.
32 EPA (2022c) MOtor Vehicle Emissions Simulator (MOVES3). Office of Transportation and Air Quality, U.S.
33 Environmental Protection Agency, Washington, D.C. Available online at: https://www.epa.gov/moves.
34 EPA (2021c) The Emissions & Generation Resource Integrated Database (eGRID) 2019 Technical Support
35 Document. Clean Air Markets Division, Office of Atmospheric Programs, U.S. Environmental Protection Agency,
36 Washington, D.C. Available Online at: https://www.epa.gov/sites/default/files/2021-
37 02/documents/egrid2019 technical guide.pdf
38 EPA (2010) Carbon Content Coefficients Developed for EPA's Mandatory Reporting Rule. Office of Air and
39 Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
40 Erickson, T. (2003) Plains CO2 Reduction (PCOR) Partnership. Presented at the Regional Carbon Sequestration
41 Partnership Meeting Pittsburgh, Pennsylvania, Energy and Environmental Research Center, University of North
42 Dakota. November 3, 2003.
References 9-7
-------
1 FAA (2022) Personal Communication between FAA and John Steller, Mausami Desai, and Vincent Camobreco for
2 aviation emissions estimates from the Aviation Environmental Design Tool (AEDT). March 2022.
3 FHWA (1996 through 2021) Highway Statistics. Federal Highway Administration, U.S. Department of
4 Transportation, Washington, D.C. Report FHWA-PL-96-023-annual. Available online at:
5 http://www.fhwa.dot.gov/policy/ohpi/hss/hsspubs.htm.
6 FHWA (2015) Off-Highway and Public-Use Gasoline Consumption Estimation Models Used in the Federal Highway
1 Administration, Publication Number FHWA-PL-17-012. Available online at:
8 https://www.fhwa.dot.gov/policyinformation/pubs/pll7012.pdf.
9 Fitzpatrick, E. (2002) The Weyburn Project: A Model for International Collaboration.
10 FRB (2022) Industrial Production and Capacity Utilization. Federal Reserve Statistical Release, G.17, Federal
11 Reserve Board. Available online at: http://www.federalreserve.gov/releases/G17/tablel 2.htm.
12 Gaffney, J. (2007) Email Communication. John Gaffney, American PublicTransportation Association and Joe
13 Aamidor, ICF International. December 17, 2007..
14 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
15 Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F., D. Qin, G.K. Plattner, M.
16 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
17 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
18 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
19 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
20 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom,
21 996 pp.
22 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
23 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
24 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.Marland, G. and A. Pippin (1990) "United States Emissions
25 of Carbon Dioxide to the Earth's Atmosphere by Economic Activity." Energy Systems and Policy, 14(4):323.
26 SAIC/EIA (2001) Monte Carlo Simulations of Uncertainty in U.S. Greenhouse Gas Emission Estimates. Final Report.
TJ Prepared by Science Applications International Corporation (SAIC) for Office of Integrated Analysis and Forecasting,
28 Energy Information Administration, U.S. Department of Energy. Washington, D.C. June 22, 2001.
29 U.S. Aluminum Association (USAA) (2008 through 2021) U.S. Primary Aluminum Production. U.S. Aluminum
30 Association, Washington, D.C.
31 USAF (1998) Fuel Logistics Planning. U.S. Air Force: AFPAM23-221. May 1,1998.
32 U.S. Census Bureau (2001 through 2011) Current Industrial Reports Fertilizer Materials and Related Products:
33 Annual Summary. Available online at: https://www.census.gov/data/tables/time-series/econ/cir/mq325b.html.
34 United States Geological Survey (USGS) (2020a) 2020 Mineral Commodity Summaries: Aluminum. U.S. Geological
35 Survey, Reston, VA.
36 USGS (2021b) 2021 Mineral Commodity Summary: Titanium and Titanium Dioxide. U.S. Geological Survey, Reston,
37 VA.
38 USGS (2019) 2017Mineral Yearbook: Aluminum. U.S. Geological Survey, Reston, VA
39 USGS (2014 through 2021a) Mineral Industry Surveys: Silicon. U.S. Geological Survey, Reston, VA.
40 USGS (2014 through 2021b) Mineral Commodity Summary, Lead. U.S. Geological Survey, Reston, VA.
41 USGS (2014 through 2019) Minerals Yearbook: Nitrogen [Advance Release], Available online at:
42 http://minerals.usgs.gov/minerals/pubs/commodity/nitrogen/.
9-8 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 USGS (1991 through 2020) Minerals Yearbook - Iron and Steel Scrap. U.S. Geological Survey, Reston, VA.
2 USGS (1991 through 2015a) Minerals Yearbook: Manufactured Abrasives Annual Report. U.S. Geological Survey,
3 Reston, VA. Available online at: http://minerals.usgs.gov/minerals/pubs/commodity/abrasives/.
4 USGS (1991 through 2015b) Minerals Yearbook: Titanium. U.S. Geological Survey, Reston, VA.
5 USGS (1991 through 2015c) Minerals Yearbook: Silicon Annual Report. U.S. Geological Survey, Reston, VA. Available
6 online at: http://minerals.usgs.gov/minerals/pubs/commodity/silicon/.
7 USGS (1996 through 2013) Minerals Yearbook: Silicon. U.S. Geological Survey, Reston, VA.
8 USGS (1995 through 2013) Minerals Yearbook: Lead Annual Report. U.S. Geological Survey, Reston, VA.
9 USGS (1995,1998, 2000, 2001, 2002, 2007) Minerals Yearbook: Aluminum Annual Report. U.S. Geological Survey,
10 Reston, VA.
11 Stationary Combustion (excluding C02)
12 EIA (2022a) Monthly Energy Review, November 2022. Energy Information Administration, U.S. Department of
13 Energy. Washington, D.C. DOE/EIA-0035(2022/11).
14 EIA (2022b) International Energy Statistics 1980-2021. Energy Information Administration, U.S. Department of
15 Energy. Washington, D.C. Available online at: https://www.eia.gov/international/data/world.
16 EPA (2022a) Acid Rain Program Dataset 1996-2021. Office of Air and Radiation, Office of Atmospheric Programs,
17 U.S. Environmental Protection Agency, Washington, D.C.
18 EPA (2022b) MOtor Vehicle Emissions Simulator (MOVES3). Office of Transportation and Air Quality, U.S.
19 Environmental Protection Agency, Washington, D.C. Available online at: https://www.epa.gov/moves.
20 EPA (1997) Compilation of Air Pollutant Emission Factors, AP-42. Office of Air Quality Planning and Standards, U.S.
21 Environmental Protection Agency. Research Triangle Park, NC. October 1997.
22 FHWA (1996 through 2022) Highway Statistics. Federal Highway Administration, U.S. Department of
23 Transportation, Washington, D.C. Report FHWA-PL-96-023-annual. Available online at:
24 http://www.fhwa.dot.gov/policv/ohpi/hss/hsspubs.litni.
25 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
26 Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F., D. Qin, G.K. Plattner, M.
27 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
28 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
29 IPCC (2007). Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
30 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
31 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom,
32 996 pp.
33 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
34 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
35 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.SAIC/EIA (2001) Monte Carlo Simulations of Uncertainty in
36 U.S. Greenhouse Gas Emission Estimates. Final Report. Prepared by Science Applications International Corporation
37 (SAIC) for Office of Integrated Analysis and Forecasting, Energy Information Administration, U.S. Department of
38 Energy. Washington, D.C. June 22, 2001.
References 9-9
-------
1 Mobile Combustion (excluding C02)
2 AAR (2008 through 2022) Railroad Facts. Policy and Economics Department, Association of American Railroads,
3 Washington, D.C. Private communication with Dan Keen.
4 The Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation Model (GREET2022). Argonne
5 National Laboratory. Available online at: https://greet.es.anl.gov.
6 APTA (2007 through 2022) Public Transportation Fact Book. American Public Transportation Association,
7 Washington, D.C. Available online at: http://www.apta.com/resources/statistics/Pages/transitstats.aspx.
8 APTA (2006) Commuter Rail National Totals. American Public Transportation Association, Washington, D.C.
9 Available online at: http://www.apta.com/research/stats/rail/crsum.cfm.
10 BEA (1991 through 2015) Unpublished BE-36 survey data. Bureau of Economic Analysis, U.S. Department of
11 Commerce. Washington, D.C.
12 Benson, D. (2002 through 2004) Personal communication. Unpublished data developed by the Upper Great Plains
13 Transportation Institute, North Dakota State University and American Short Line & Regional Railroad Association.
14 Browning (2022a) Addressing the Time Series Inconsistency in FHWA Data. Memorandum from ICF to Sarah
15 Roberts, Office of Transportation and Air Quality, U.S. Environmental Protection Agency. September 2022.
16 Browning (2022b) Updated Methodology for Estimating Cm and N2O Emissions from Highway Vehicle Alternative
17 Fuel Vehicles. Memorandum from ICF to Sarah Roberts, Office of Transportation and Air Quality, U.S.
18 Environmental Protection Agency. November 2022.
19 Browning (2020a) GHG Inventory EF Development Using Certification Data. Memorandum from ICF to Sarah
20 Roberts, Office of Transportation and Air Quality, U.S. Environmental Protection Agency. September 2020.
21 Browning, L. (2020b). Updated Methane and Nitrous Oxide Emission Factors for Non-Road Sources and On-road
22 Motorcycles. Technical Memorandum from ICF International to Sarah Roberts, Office of Transportation and Air
23 Quality, U.S. Environmental Protection Agency, September 2020.
24 Browning, L. (2019) Updated On-highway Cm and N2O Emission Factors for GHG Inventory. Memorandum from ICF
25 to Sarah Roberts and Justine Geidosch, Office of Transportation and Air Quality, U.S. Environmental Protection
26 Agency. September 2019.
27 Browning, L. (2018a). Updated Methodology for Estimating Electricity Use from Highway Plug-In Electric Vehicles.
28 Technical Memorandum from ICF International to Sarah Roberts and Justine Geidosch, Office of Transportation
29 and Air Quality, U.S. Environmental Protection Agency. October 2018.
30 Browning, L. (2018b) Updated Non-Highway CH4 and N2O Emission Factors for U.S. GHG Inventory. Technical
31 Memorandum from ICF International to Sarah Roberts and Justine Geidosch, Office of Transportation and Air
32 Quality, U.S. Environmental Protection Agency. November 2018.
33 Browning, L. (2017) Updated Methodology for Estimating CH4 and N2O Emissions from Highway Vehicle Alternative
34 Fuel Vehicles. Technical Memorandum from ICF International to Sarah Roberts and Justine Geidosch, Office of
35 Transportation and Air Quality, U.S. Environmental Protection Agency. October 2017.
36 Browning, L. (2009) Personal communication with Lou Browning, "Suggested New Emission Factors for Marine
37 Vessels," ICF International.
38 Browning, L. (2005) Personal communication with Lou Browning, "Emission control technologies for diesel highway
39 vehicles specialist," ICF International.
40 DHS (2008) Email Communication. Elissa Kay, Department of Homeland Security and Joe Aamidor, ICF
41 International. January 11, 2008.
9-10 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 DLA Energy (2022) Unpublished data from the Defense Fuels Automated Management System (DFAMS). Defense
2 Energy Support Center, Defense Logistics Agency, U.S. Department of Defense. Washington, D.C.
3 DOC (1991 through 2022) Unpublished Report of Bunker Fuel Oil Laden on Vessels Cleared for Foreign Countries.
4 Form-563. Foreign Trade Division, Bureau of the Census, U.S. Department of Commerce. Washington, D.C.
5 DOE (1993 through 2022) Transportation Energy Data Book Edition 40. Office of Transportation Technologies,
6 Center for Transportation Analysis, Energy Division, Oak Ridge National Laboratory. Personal Communication
7 between Stacy Davis (DOE) and Deep Shah (ICF) for sharing selected tables from the pre-release version.
8 DOT (1991 through 2022) Airline Fuel Cost and Consumption. U.S. Department of Transportation, Bureau of
9 Transportation Statistics, Washington, D.C. DAI-10. Available online at: http://www.transtats.bts.gov/fuel.asp.
10 EIA (2022) Monthly Energy Review, February 2022, Energy Information Administration, U.S. Department of Energy,
11 Washington, D.C. DOE/EIA-0035(2022/02).
12 EIA (2022f) Natural Gas Annual 2021. Energy Information Administration, U.S. Department of Energy, Washington,
13 D.C. DOE/EIA-O131(ll).
14 EIA (1991 through 2022) Fuel Oil and Kerosene Sales. Energy Information Administration, U.S. Department of
15 Energy. Washington, D.C. Available online at: http://www.eia.gov/petroleum/fyeloilkerosene.
16 EIA (2016) "Table 3.1: World Petroleum Supply and Disposition." International Energy Annual. Energy Information
17 Administration, U.S. Department of Energy. Washington, D.C. Available online at:
18 https://www.eia.gov/cfapps/ipdbproiect/l EDIndex3.cfm?tid=5&pid=66&aid=l 3.
19 EIA (2011) Annual Energy Review 2010. Energy Information Administration, U.S. Department of Energy,
20 Washington, D.C. DOE/EIA-0384(2011). October 19, 2011.
21 EIA (2007) Personal Communication. Joel Lou, Energy Information Administration and Aaron Beaudette, ICF
22 International. Residual and Distillate Fuel Oil Consumption for Vessel Bunkering (Both International and Domestic)
23 for American Samoa, U.S. Pacific Islands, and Wake Island. October 24, 2007.
24 EIA (2002) Alternative Fuels Data Tables. Energy Information Administration, U.S. Department of Energy,
25 Washington, D.C. Available online at: http://www.eia.doe.gov/fuelrenewable.html.
26 EPA (2022a) Light-Duty Automotive Technology, Carbon Dioxide Emissions, and Fuel Economy Trends: 1975 - 2020.
27 Office of Transportation and Air Quality, U.S. Environmental Protection Agency. Available online at:
28 https://www.epa.gov/fuel-economy/trends-report.
29 EPA (2022b) Motor Vehicle Emissions Simulator (MOVES3). Office of Transportation and Air Quality, U.S.
30 Environmental Protection Agency. Available online at: https://www.epa.gov/moves.
31 EPA (2022c) Confidential Engine Family Sales Data Submitted to EPA by Manufacturers. Office of Transportation
32 and Air Quality, U.S. Environmental Protection Agency.
33 EPA (2022d) Annual Certification Test Results Report. Office of Transportation and Air Quality, U.S. Environmental
34 Protection Agency. Available online at: https://www.epa.gov/compliance-and-fuel-economy-data/annual-
35 certification test data-vehicles-and-engines.
36 EPA (2016g) "1970 - 2015 Average annual emissions, all criteria pollutants in MS Excel." National Emissions
37 Inventory (NEI) Air Pollutant Emissions Trends Data. Office of Air Quality Planning and Standards. Available online
38 at: https://www.epa.gov/air-emissions-inventories/air-pollutant-emissions-trends-data.
39 EPA (2004) Mobile6.2 Vehicle Emission Modeling Software. Office of Mobile Sources, U.S. Environmental
40 Protection Agency, Ann Arbor, Michigan.
41 EPA (1999a) Emission Facts: The History of Reducing Tailpipe Emissions. Office of Mobile Sources. May 1999. EPA
42 420-F-99-017. Available online at: https://www.epa.gov/nscep.
References 9-11
-------
1 EPA (1999b) Regulatory Announcement: EPA's Program for Cleaner Vehicles and Cleaner Gasoline. Office of Mobile
2 Sources. December 1999. EPA420-F-99-051. Available online at:
3 https://nepis.epa.gov/Exe/ZvPDF.cgi/P100 lZ9W.PDF?Dockev=P1001Z9W.PDF.
4 EPA (1998) Emissions of Nitrous Oxide from Highway Mobile Sources: Comments on the Draft Inventory of U.S.
5 Greenhouse Gas Emissions and Sinks, 1990-1996. Office of Mobile Sources, Assessment and Modeling Division,
6 U.S. Environmental Protection Agency. August 1998. EPA420-R-98-009.
7 EPA (1994a) Automobile Emissions: An Overview. Office of Mobile Sources. August 1994. EPA 400-F-92-007.
8 Available online at: https://www.epa.gov/nscep.
9 EPA (1994b) Milestones in Auto Emissions Control. Office of Mobile Sources. August 1994. EPA 400-F-92-014.
10 Available online at: https://www.epa.gov/nscep.
11 EPA (1993) Automobiles and Carbon Monoxide. Office of Mobile Sources. January 1993. EPA 400-F-92-005.
12 Available online at: https://www.epa.gov/nscep.
13 Esser, C. (2003 through 2004) Personal Communication with Charles Esser, Residual and Distillate Fuel Oil
14 Consumption for Vessel Bunkering (Both International and Domestic) for American Samoa, U.S. Pacific Islands, and
15 Wake Island.
16 FAA (2022) Personal Communication between FAA and John Steller, Mausami Desai and Vincent Camobreco for
17 aviation emission estimates from the Aviation Environmental Design Tool (AEDT). March 2022.
18 FHWA (1996 through 2021) Highway Statistics. Federal Highway Administration, U.S. Department of
19 Transportation, Washington, D.C. Report FHWA-PL-96-023-annual. Available online at:
20 http://www.fhwa.dot.gov/policv/ohpi/hss/hsspubs.htm.
21 FHWA (2015) Off-Highway and Public-Use Gasoline Consumption Estimation Models Used in the Federal Highway
22 Administration, Publication Number FHWA-PL-17-012. Available online at:
23 https://www.fhwa.dot.gov/policyinformation/pubs/pll7012.pdf.
24 Gaffney, J. (2007) Email Communication. John Gaffney, American PublicTransportation Association and Joe
25 Aamidor, ICF International. December 17, 2007.
26 HybridCars.com (2019). Monthly Plug-In Electric Vehicle Sales Dashboard, 2010-2018. Available online at
27 https://www.hybridcars.com/december-2017-dashboard/.
28 ICF (2006a) Revised Gasoline Vehicle EFs for LEV and Tier 2 Emission Levels. Memorandum from ICF International to
29 John Davies, Office of Transportation and Air Quality, U.S. Environmental Protection Agency. November 2006.
30 ICF (2006b) Revisions to Alternative Fuel Vehicle (AFV) Emission Factors for the U.S. Greenhouse Gas Inventory.
31 Memorandum from ICF International to John Davies, Office of Transportation and Air Quality, U.S. Environmental
32 Protection Agency. November 2006.
33 ICF (2004) Update of Methane and Nitrous Oxide Emission Factors for On-Highway Vehicles. Final Report to U.S.
34 Environmental Protection Agency. February 2004.
35 ICF (2017b) Updated Non-Highway CH4 and N2O Emission Factors for U.S. GHG Inventory. Memorandum from ICF
36 to Sarah Roberts and Justine Geidosch, Office of Transportation and Air Quality, U.S. Environmental Protection
37 Agency. October 2017.
38 Lipman, T. and M. Delucchi (2002) "Emissions of Nitrous Oxide and Methane from Conventional and Alternative
39 Fuel Motor Vehicles." Climate Change, 53:477-516.
40 SAE (2010) Utility Factor Definitions for Plug-In Hybrid Electric Vehicles Using Travel Survey Data. Society of
41 Automotive Engineers. Report J2841, Available online at: https://www.sae.org/standards/content/i2841 201009/.
42 Raillnc (2014 through 2022) Raillnc Short line and Regional Traffic Index. Carloads Originated Year-to-Date.
43 December 2022. Available online at: https://www.railinc.com/rportal/railinc-indexes.
9-12 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Santoni, G., B. Lee, E. Wood, S. Herndon, R. Miake-Lye, S. Wofsy, J. McManus, D. Nelson, M. Zahniser (2011)
2 Aircraft emissions of methane and nitrous oxide during the alternative aviation fuel experiment. Environ Sci
3 Technol. 2011 Aug 15; 45(16):7075-82.
4 U.S. Census Bureau (2000) Vehicle Inventory and Use Survey. U.S. Census Bureau, Washington, D.C. Database CD-
5 EC97-VIUS.
6 Whorton, D. (2006 through 2014) Personal communication, Class II and III Rail energy consumption, American
7 Short Line and Regional Railroad Association.
8 Zukowski, D. (2022), More electric buses join transit fleets as costs and technology improve, SmartCitiesDive,
9 January 31, 2022. Available at https://www.smartcitiesdive.com/news/mQre-electric~buses-arriving-in-city-transit-
10 fleets/617072/
11 Carbon Emitted from Non-Energy Uses of Fossil Fuels
12 ACC (2022a) "U.S. Resin Production & Sales 2021 vs. 2020." Available online at:
13 https://www.americanchemistry.com/chemistry-in-america/data-industry-statistics/statistics-on-the-plastic-
14 resins-industry
15 ACC (2022b) "Guide to the Business of Chemistry, 2022," American Chemistry Council. Available online at:
16 https://www.americanchemistry.com/chemistry-in-america/data-industry-statistics/resources/2022-guide-to-the-
17 business-of-chemistry
18 ACC (2021) "U.S. Resin Production & Sales 2020 vs. 2019." Available online at:
19 https://www.americanchemistry.com/chemistry-in-america/chemistry-in-everyday-products/plastics
20 ACC (2020) "U.S. Resin Production & Sales 2019 vs. 2018." Available online at:
21 https://www.americanchemistry.com/chemistry-in-america/chemistry-in-everyday-products/plastics.
22 ACC (2019) "U.S. Resin Production & Sales 2018 vs. 2017." Available online at:
23 https://www.americanchemistry.com/chemistry-in-america/chemistry-in-everyday-products/plastics.
24 ACC (2018) "U.S. Resin Production & Sales 2017 vs. 2016." Available online at:
25 https://www.americanchemistry.com/chemistry-in-america/chemistry-in-everyday-products/plastics.
26 ACC (2017) "U.S. Resin Production & Sales 2016 vs. 2015."
27 ACC (2016) "U.S. Resin Production & Sales 2015 vs. 2014."
28 ACC (2015) "PIPS Year-End Resin Statistics for 2014 vs. 2013: Production, Sales and Captive Use." Available online
29 at: https://www.americanchemistry.com/chemistry-in-america/data-industry-statistics/statistics-on-the-plastic-
30 resins-industry/resin-report-subscriptions.
31 ACC (2014) "U.S. Resin Production & Sales: 2013 vs. 2012," American Chemistry Council. Available online at:
32 http://www.americanchemistry.com/Jobs/EconomicStatistics/Plastics-Statistics/Production-and-Sales-Data-by-
33 Resin.pdf.
34 ACC (2013) "U.S. Resin Production & Sales: 2012 vs. 2011," American Chemistry Council. Available online at:
35 http://www.americanchemistry.com/Jobs/EconomicStatistics/Plastics-Statistics/Production-and-Sales-Data-by-
36 Resin.pdf.
37 ACC (2003-2011) "PIPS Year-End Resin Statistics for 2010: Production, Sales and Captive Use." Available online at:
38 http://www.americanchemistry.com/Jobs/EconomicStatistics/Plastics-Statistics/Production-and-Sales-Data-by-
39 Resin.pdf.
40 Bank of Canada (2022) Financial Markets Department Year Average of Exchange Rates. Available online at:
41 https://www.bankofcanada.ca/rates/exchange/annual-average-exchange-rates/#download.
References 9-13
-------
1 Bank of Canada (2021) Financial Markets Department Year Average of Exchange Rates. Available online at:
2 https://www.bankofcanada.ca/rates/exchange/annual-average-exchangerates/#download.
3 Bank of Canada (2020) Financial Markets Department Year Average of Exchange Rates. Available online at:
4 https://www.bankofcanada.ca/rates/exchange/annual-average-exchange-rates/#download.
5 Bank of Canada (2019) Financial Markets Department Year Average of Exchange Rates. Available online at:
6 https://www.bankofcanada.ca/rates/exchange/annual-average-exchange-rates/#download.
7 Bank of Canada (2018) Financial Markets Department Year Average of Exchange Rates. Available online at:
8 https://www.bankofcanada.ca/rates/exchange/annual-average-exchange-rates/.
9 Bank of Canada (2017) Financial Markets Department Year Average of Exchange Rates. Available online at:
10 https://www.bankofcanada.ca/rates/exchange/legacy-noon-and-closing-rates/.
11 Bank of Canada (2016) Financial Markets Department Year Average of Exchange Rates. Available online at:
12 https://www.bankofcanada.ca/rates/exchange/legacy-noon-and-closing-rates/.
13 Bank of Canada (2014) Financial Markets Department Year Average of Exchange Rates. Available online at:
14 https://www.bankofcanada.ca/rates/exchange/legacy-noon-and-closing-rates/.
15 Bank of Canada (2013) Financial Markets Department Year Average of Exchange Rates. Available online at:
16 https://www.bankofcanada.ca/rates/exchange/legacy-noon-and-closing-rates/.
17 Bank of Canada (2012) Financial Markets Department Year Average of Exchange Rates. Available online at:
18 https://www.bankofcanada.ca/rates/exchange/legacy-noon-and-closing-rates/.
19 CIAC (2022). 2022 Economic Review of Chemistry. Available online at: https://canadianchemistry.ca/wp-
20 content/uploads/2022/06/2022-Economic-Review-of-Chemistry23?.V tcmoved.pdf.
21 EIA (2022) Monthly Energy Review, September 2022. Energy Information Administration, U.S. Department of
22 Energy, Washington, D.C. DOE/EIA-0035 (2022/09).
23 EIA (2021) EIA Manufacturing Consumption of Energy (MECS) 2018. U.S. Department of Energy, Energy Information
24 Administration, Washington, D.C.
25 EIA (2020) Glossary. Energy Information Administration, U.S. Department of Energy, Washington, D.C. Available
26 online at: https://www. eia.gov/tools/glossary/index. php?id=N#nat Gas Liquids.
27 EIA (2019) Personal communication between EIA and ICF on November 11, 2019.
28 EIA (2017) EIA Manufacturing Consumption of Energy (MECS) 2014. U.S. Department of Energy, Energy Information
29 Administration, Washington, D.C.
30 EIA (2013) EIA Manufacturing Consumption of Energy (MECS) 2010. U.S. Department of Energy, Energy Information
31 Administration, Washington, D.C.
32 EIA (2010) EIA Manufacturing Consumption of Energy (MECS) 2006. U.S. Department of Energy, Energy Information
33 Administration, Washington, D.C.
34 EIA (2005) EIA Manufacturing Consumption of Energy (MECS) 2002. U.S. Department of Energy, Energy Information
35 Administration, Washington, D.C.
36 EIA (2001) EIA Manufacturing Consumption of Energy (MECS) 1998. U.S. Department of Energy, Energy Information
37 Administration, Washington, D.C.
38 EIA (1997) EIA Manufacturing Consumption of Energy (MECS) 1994. U.S. Department of Energy, Energy Information
39 Administration, Washington, D.C.
40 EIA (1994) EIA Manufacturing Consumption of Energy (MECS) 1991. U.S. Department of Energy, Energy Information
41 Administration, Washington, D.C.
9-14 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 EPA (2022) "Criteria pollutants National Tier 1 for 1970 - 2021." National Emissions Inventory (NEI) Air Pollutant
2 Emissions Trends Data. Office of Air Quality Planning and Standards, February 2022. Available online at:
3 https://www.epa.gov/air-emissions-inventories/air-pollutant-emissions-trends-data.EPA (2021) Resource
4 Conservation and Recovery Act (RCRA) Info, Biennial Report, GM Form (Section 2- Onsite Management) and WR
5 Form.
6 EPA (2019) Advancing Sustainable Materials Management: 2016 and 2017 Data Tables. Office of Land and
7 Emergency Management, U.S. Environmental Protection Agency. Washington, D.C. Available online at:
8 https://www.epa.gov/sites/production/files/2019-
9 ll/documents/2016 and 2017 facts and figures data tables O.pdf.
10 EPA (2018a) Advancing Sustainable Materials Management: Facts and Figures 2015, Assessing Trends in Material
11 Generation, Recycling and Disposal in the United States. Washington, D.C.
12 EPA (2018b) Resource Conservation and Recovery Act (RCRA) Info, Biennial Report, GM Form (Section 2- Onsite
13 Management) and WR Form.
14 EPA (2017) EPA's Pesticides Industry Sales and Usage, 2008 - 2012 Market Estimates. Available online at:
15 https://www.epa.gov/sites/production/files/2017-01/documents/pesticides-industry-sales-usage-2016 O.pdf.
16 Accessed September 2017.
17 EPA (2016a) Advancing Sustainable Materials Management: 2014 Facts and Figures Fact Sheet. Office of Solid
18 Waste and Emergency Response, U.S. Environmental Protection Agency, Washington, D.C. Available online at:
19 https://www.epa.gov/sites/production/files/2016-ll/documents/2014 smmfactsheet 508.pdf.
20 EPA (2016b) Resource Conservation and Recovery Act (RCRA) Info, Biennial Report, GM Form (Section 2- Onsite
21 Management) and WR Form.
22 EPA (2015) Resource Conservation and Recovery Act (RCRA) Info, Biennial Report, GM Form (Section 2- Onsite
23 Management) and WR Form.
24 EPA (2014a) Municipal Solid Waste in the United States: 2012 Facts and Figures. Office of Solid Waste and
25 Emergency Response, U.S. Environmental Protection Agency, Washington, D.C. Available online at:
26 https://www.epa.gov/sites/default/files/2015-09/documents/2012 msw dat tbls.pdf.
27 EPA (2014b) Chemical Data Access Tool (CDAT). U.S. Environmental Protection Agency, June 2014. Available online
28 at: https://edg.epa.gov/metadata/catalog/search/resource/details.page?uuid=%7B2D73C764-6919-4Q4D-8C9B-
29 61869B3330D6%7D. Accessed January 2015.
30 EPA (2013a) Municipal Solid Waste in the United States: 2011 Facts and Figures. Office of Solid Waste and
31 Emergency Response, U.S. Environmental Protection Agency, Washington, D.C. Available online at:
32 http://www.epa.gov/epaoswer/non-hw/muncpl/msw99.litm.
33 EPA (2013b) Resource Conservation and Recovery Act (RCRA) Info, Biennial Report, GM Form (Section 2- Onsite
34 Management) and WR Form.
35 EPA (2011) EPA's Pesticides Industry Sales and Usage, 2006 and 2007 Market Estimates. Available online at:
36 https://www.epa.gov/pesticides/pesticides-industry-sales-and-usage-2006-and-2007-market-estimates. Accessed
37 January 2012.
38 EPA (2009) Biennial Reporting System (BRS) Database. U.S. Environmental Protection Agency, Envirofacts
39 Warehouse. Washington, D.C. Available online at: https://www.epa.gov/enviro/br-search. Data for 2001-2007 are
40 current as of Sept. 9, 2009.
41 EPA (2004) EPA's Pesticides Industry Sales and Usage, 2000 and 2001 Market Estimates. Available online at:
42 https://nepis.epa.gov/Exe/Z.yPURL.cgi?Dockev=3000659P.TXT. Accessed September 2006.
43 EPA (2002) EPA's Pesticides Industry Sales and Usage, 1998 and 1999 Market Estimates, Table 3.6. Available online
44 at https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=200001G5.TXT. Accessed July 2003.
References 9-15
-------
1 EPA (2001) AP 42, Volume I, Fifth Edition. Chapter 11: Mineral Products Industry. Available online at:
2 http://www.epa.gov/ttn/chief/ap42/chll/index.html.
3 EPA (2000a) Biennial Reporting System (BRS). U.S. Environmental Protection Agency, Envirofacts Warehouse.
4 Washington, D.C. Available online at: https://www.epa.gov/enviro/br-search.
5 EPA (2000b) Toxics Release Inventory, 1998. U.S. Environmental Protection Agency, Office of Environmental
6 Information, Office of Information Analysis and Access, Washington, D.C. Available online at:
7 https://enviro.epa.gov/triexplorer/tri release.chemical.
8 EPA (1999) EPA's Pesticides Industry Sales and Usage, 1996-1997 Market Estimates. Available online at:
9 https://nepis.epa.gov/Exe/Z.yPURL.cgi?Dockev=2(XXX311L. TXT.
10 EPA (1998) EPA's Pesticides Industry Sales and Usage, 1994-1995 Market Estimates. Available online at:
11 http://www.epa.gov/oppbeadl/pestsales/95pestsales/market estimatesl995.pdf.
12 FEB (2013) Fiber Economics Bureau, as cited in C&EN (2013) Lackluster Year for Chemical Output: Production
13 stayed flat or dipped in most world regions in 2012. Chemical &Engineering News, American Chemical Society, 1
14 July. Available online at: http://www.cen-online.org.
15 FEB (2012) Fiber Economics Bureau, as cited in C&EN (2012) Too Quiet After the Storm: After a rebound in 2010,
16 chemical production hardly grew in 2011. Chemical & Engineering News, American Chemical Society, 2 July.
17 Available online at: http://www.cen-online.org.
18 FEB (2011) Fiber Economics Bureau, as cited in C&EN (2011) Output Ramps up in all Regions. Chemical Engineering
19 News, American Chemical Society, 4 July. Available online at: http://www.cen-online.org.
20 FEB (2010) Fiber Economics Bureau, as cited in C&EN (2010) Output Declines in U.S., Europe. Chemical &
21 Engineering News, American Chemical Society, 6 July. Available online at: http://www.cen-online.org.
22 FEB (2009) Fiber Economics Bureau, as cited in C&EN (2009) Chemical Output Slipped In Most Regions Chemical &
23 Engineering News, American Chemical Society, 6 July. Available online at: http://www.cen-online.org.
24 FEB (2007) Fiber Economics Bureau, as cited in C&EN (2007) Gains in Chemical Output Continue. Chemical &
25 Engineering News, American Chemical Society. July 2, 2007. Available online at: http://www.cen-online.org.
26 FEB (2005) Fiber Economics Bureau, as cited in C&EN (2005) Production: Growth in Most Regions Chemical &
27 Engineering News, American Chemical Society, 11 July. Available online at: http://www.cen-online.org.
28 FEB (2003) Fiber Economics Bureau, as cited in C&EN (2003) Production Inches Up in Most Countries, Chemical &
29 Engineering News, American Chemical Society, 7 July. Available online at: http://www.cen-online.org.
30 FEB (2001) Fiber Economics Bureau, as cited in ACS (2001) Production: slow gains in output of chemicals and
31 products lagged behind U.S. economy as a whole Chemical & Engineering News, American Chemical Society, 25
32 June. Available online at: http://pubs.acs.org/cen.
33 Financial Planning Association (2006) Canada/US Cross-Border Tools: US/Canada Exchange Rates. Available online
34 at: http://www.fpanet.org/global/planners/US Canada ex rates.cfm. Accessed on August 16, 2006.
35 Gosselin, Smith, and Hodge (1984) "Clinical Toxicology of Commercial Products." Fifth Edition, Williams & Wilkins,
36 Baltimore.
37 ICIS (2016) "Production issues force US melamine plant down" Available online at:
38 https://www.icis.com/resources/news/2016/05/03/9994556/production-issues-force-us-melamine-plant-down/.
39 ICIS (2008) "Chemical profile: Melamine" Available online at:
40 https://www.icis.com/resources/news/2008/12/01/9174886/chemical-profile-melamine/. Accessed November,
41 2017.
9-16 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 IISRP (2003) "IISRP Forecasts Moderate Growth in North America to 2007" International Institute of Synthetic
2 Rubber Producers, Inc. New Release. Available online at: http://www.iisrp.com/press-releases/20Q3-Press-
3 Releases/IIS RP-NA-Forecast-03-07.html.
4 IISRP (2000) "Synthetic Rubber Use Growth to Continue Through 2004, Says IISRP and RMA" International Institute
5 of Synthetic Rubber Producers press release.
6 INEGI (2006) Produccion bruta total de las unidades economicas manufactureras por Subsector, Rama, Subrama y
7 Clase de actividad. Available online at:
8 http://www.inegi.gob.mx/est/contenidos/espanol/proyectos/censos/ce2004/tb manufacturas.asp. Accessed on
9 August 15, 2006.
10 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
11 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
12 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
13 Marland, G., and R.M. Rotty (1984) "Carbon dioxide emissions from fossil fuels: A procedure for estimation and
14 results for 1950-1982," Tellus 36b:232-261.
15 NPRA (2002) North American Wax - A Report Card. Available online at:
16 http://www.npra.org/members/publications/papers/lybes/LW-02-126.pdf.
17 U.S. Census Bureau (2021) 2017 Economic Census. Available online at:
18 https://www.census.gov/data/tables/2017/econ/economic-census/naics-sector-31-33.html. Accessed October
19 2021.
20 U.S. Census Bureau (2014) 2012 Economic Census. Available online at:
21 http://www.censys.gov/econ/census/schedule/whats been released.html. Accessed November
22 2014.http://smpbffl.dsd.census.gov/TheDataWeb HotReport/servlet/HotReportEngineServlet?emailname=vh(S)b
23 oc&filename=mfgl.hrml&20071204152004.Var.NAICS2002=325611&forward=20071204152004.Var.NAICS2002
24 U.S. Census Bureau (2009) Soap and Other Detergent Manufacturing: 2007. Available online at:.
25 U.S. Census Bureau (2004) Soap and Other Detergent Manufacturing: 2002. Issued December 2004. EC02-31I-
26 325611 (RV). Available online at: http://www.census.gov/prod/ec02/ec0231i325611.pdf.
27 U.S. Census Bureau (1999) Soap and Other Detergent Manufacturing: 1997. Available online at:
28 http://www.census.gov/epcd/www/ec97stat.htm.
29 U.S. International Trade Commission (2022) "Interactive Tariff and Trade DataWeb: Quick Query." Available online
30 at: http://dataweb.usitc.gov/. Accessed September 2022.
31 USTMA (2022) "2021 U.S. Scrap Tire Management Summary." U.S. Tire Manufacturers Association, Washington,
32 DC. October 2022. Available online at:
33 https://www.ustires.org/sites/default/files/21%20US%20Scrap%20Tire%20Management%20Report%20101722.pd
34 f.
35 USTMA (2020) "2019 U.S. Scrap Tire Management Summary." U.S. Tire Manufacturers Association, Washington,
36 DC. October 2020. Available online at:
37 https://www.ustires.org/sites/default/files/2019%20USTMA%20Scrap%20Tire%20Management%20Summary%20
38 Report.pdf.
39 USTMA (2018) "2017 U.S. Scrap Tire Management Summary." U.S. Tire Manufacturers Association, Washington,
40 DC. July 2018. Available online at: https://www.tyrepress.com/wp-
41 content/uploads/2018/07/USTMA scraptire sornm 2017 07 11 2018.pdf.
42 USTMA (2016) "2015 U.S. Scrap Tire Management Summary." U.S. Tire Manufacturers Association. August 2016.
43 Available online at: https://www.ustires.org/sites/default/files/MAR 028 USTMA.pdf.
References 9-17
-------
1 USTMA (2014) "2013 U.S. Scrap Tire Management Summary." U.S. Tire Manufacturers Association. November
2 2014. Available online at: https://www.ustires.org/sites/default/files/MAR 027 USTMA.pdf.
3 USTMA (2013) "U.S. Scrap Tire Management Summary 2005-2009." U.S. Tire Manufacturers Association. October
4 2011; Updated September 2013. Available online at:
5 https://www.ustires.org/sites/default/files/MAR 025 USTMA.pdf.
6 USTMA (2012) "Scrap Tire Markets: Facts and Figures - Scrap Tire Characteristics." U.S. Tire Manufacturers
7 Association. Accessed 18 on January 2012.
8 Incineration of Waste
9 ArSova, Ljupka, Rob van Haaren, Nora Goldstein, Scott M. Kaufman, and Nickolas J. Themelis (2008) "16th Annual
10 BioCycle Nationwide Survey: The State of Garbage in America" BioCycle, JG Press, Emmaus, PA. December.
11 Bahor, B (2009) Covanta Energy's public review comments re: Draft Inventory of U.S. Greenhouse Gas Emissions
12 and Sinks: 1990-2007. Submitted via email on April 9, 2009 to Leif Hockstad, U.S. EPA.
13 De Soete, G.G. (1993) "Nitrous Oxide from Combustion and Industry: Chemistry, Emissions and Control." In A. R.
14 Van Amstel, (ed.) Proc. of the International Workshop Methane and Nitrous Oxide: Methods in National Emission
15 Inventories and Options for Control, Amersfoort, NL February 3-5,1993.
16 Energy Recovery Council (2018) Energy Recovery Council. 2018 Directory of Waste to Energy Facilities. Ted
17 Michaels and Karunya Krishnan. October 2018. Available online at: http://energyrecoverycouncil.org/wp-
18 content/uploads/2019/10/ERC-2018-directory.pdf.
19 Energy Recovery Council (2009) "2007 Directory of Waste-to-Energy Plants in the United States." Accessed on
20 September 29, 2009.
21 EIA (2017) MSW Incineration for Heating or Electrical Generation, December 2017, Energy Information
22 Administration, U.S. Department of Energy, Washington, DC. DOE/EIA-0035. Available online at:
23 https://www.eia.gov/opendata/?src=-f3.
24 EIA (2019) EIA St. Louis Federal Reserve's Economic Data (FRED) Consumer Price Index for All Urban Consumers:
25 Education and Communication (CPIEDUSL). Available online at: < https://www.eia.gov/opendata/excel/>
26 EPA (2021). Greenhouse Gas Reporting Program (GHGRP). 2021 Envirofacts. Available online at:
27 https://ghgdata.epa.gov/ghgp/main.do.
28 EPA (2020a) Advancing Sustainable Materials Management: 2018 Data Tables. Office of Land and Emergency
29 Management, U.S. Environmental Protection Agency. Washington, D.C. Available online at:
30 https://www.epa.gov/sites/production/files/2020-ll/documents/2018 ff fact sheet.pdf.
31 EPA (2020b). Greenhouse Gas Reporting Program (GHGRP). 2020 Envirofacts. Available online at:
32 https://ghgdata.epa.gov/ghgp/main.do.
33 EPA (2019) Advancing Sustainable Materials Management: 2016 and 2017 Data Tables. Office of Land and
34 Emergency Management, U.S. Environmental Protection Agency. Washington, D.C. Available online at:
35 https://www.epa.gov/sites/production/files/2019-
36 ll/documents/2016 and 2017 facts and figures data tables O.pdf.
37 EPA (2018a) Advancing Sustainable Materials Management: 2015 Data Tables. Office of Land and Emergency
38 Management, U.S. Environmental Protection Agency. Washington, D.C. Available online at:
39 https://www.epa.gov/sites/production/files/2018-
40 07/documents/smm 2015 tables and figures 07252018 fnl 508 O.pdf.
9-18 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 EPA (2018b) Greenhouse Gas Reporting Program Data. Washington, DC: U.S. Environmental Protection Agency.
2 Available online at: https://www.epa.gov/ghgreporting/ghg-reporting-program-data-sets.
3 EPA (2016) Advancing Sustainable Materials Management: 2014 Fact Sheet. Office of Land and Emergency
4 Management, U.S. Environmental Protection Agency. Washington, D.C. Available online at:
5 https://www.epa.gov/sites/production/files/2016-ll/documents/2014 smmfactsheet 508.pdf.
6 EPA (2015) Advancing Sustainable Materials Management: Facts and Figures 2013 - Assessing Trends in Material
1 Generation, Recycling and Disposal in the United States. Office of Solid Waste and Emergency Response, U.S.
8 Environmental Protection Agency. Washington, D.C. Available online at:
9 http://www3.epa.gov/epawaste/nonhaz/municipal/pubs/2013 advncng smm rpt.pdf.
10 EPA (2007, 2008, 2011, 2013, 2014) Municipal Solid Waste in the United States: Facts and Figures. Office of Solid
11 Waste and Emergency Response, U.S. Environmental Protection Agency. Washington, D.C. Available online at:
12 http://www.epa.gov/osw/nonhaz/municipal/msw99.htm.
13 E PA (2006) Solid Waste Management and Greenhouse Gases: A Life-Cycle Assessment of Emissions and Sinks.
14 Office of Solid Waste and Emergency Response, U.S. Environmental Protection Agency. Washington, D.C.
15 EPA (2000) Characterization of Municipal Solid Waste in the United States: Source Data on the 1999 Update. Office
16 of Solid Waste, U.S. Environmental Protection Agency. Washington, D.C. EPA530-F-00-024.
17 Goldstein, N. and C. Madtes (2001) "13th Annual BioCycle Nationwide Survey: The State of Garbage in America."
18 BioCycle, JG Press, Emmaus, PA. December 2001.
19 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
20 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.K. Plattner, M.
21 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
22 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
23 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
24 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
25 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom,
26 996 pp.
27 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
28 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
29 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
30 Kaufman, et al. (2004) "14th Annual BioCycle Nationwide Survey: The State of Garbage in America 2004" Biocycle,
31 JG Press, Emmaus, PA. January 2004.
32 Schneider, S. (2007) E-mail between Shelly Schneider of Franklin Associates (a division of ERG) and Sarah Shapiro of
33 ICF International, January 10, 2007.
34 Shin, D. (2014) Generation and Disposition of Municipal Solid Waste (MSW) in the United States-A National
35 Survey. Thesis. Columbia University, Department of Earth and Environmental Engineering, January 3, 2014.
36 Simmons, et al. (2006) "15th Nationwide Survey of Municipal Solid Waste Management in the United States: The
37 State of Garbage in America." BioCycle, JG Press, Emmaus, PA. April 2006.
38 USTMA (2022) "2021 U.S. Scrap Tire Management Summary." U.S. Tire Manufacturers Association, Washington,
39 DC. October 2022. Available online at:
40 https://www.ustires.org/sites/default/files/21%20US%20Scrap%20Tire%20Management%20Report%20101722.pd
41 f.
42 USTMA (2020) "2019 U.S. Scrap Tire Management Summary." U.S. Tire Manufacturers Association, Washington,
43 DC. October 2020. Available online at:
References 9-19
-------
1 https://www.ustires.org/sites/default/files/2019%20USTMA%20Scrap%20Tire%20Management%20Summary%20
2 Report.pdf.
3 USTMA (2018) "2017 U.S. Scrap Tire Management Summary." U.S. Tire Manufacturers Association, Washington,
4 DC. July 2018. Available online at: https://www.tvrepress.com/wp-
5 content/uploads/2018/07/USTMA scraptire summ 2017 07 11 2018.pdf.
6 USTMA (2016) "2015 U.S. Scrap Tire Management Summary." U.S. Tire Manufacturers Association. August 2016.
7 Available online at: https://www.ustires.org/sites/default/files/MAR 028 USTMA.pdf.
8 USTMA (2014) "2013 U.S. Scrap Tire Management Summary." U.S. Tire Manufacturers Association. November
9 2014. Available online at: https://www.ustires.org/sites/default/files/MAR 027 USTMA.pdf.
10 USTMA (2013) "U.S. Scrap Tire Management Summary 2005-2009." U.S. Tire Manufacturers Association. October
11 2011; Updated September 2013. Available online at:
12 https://www.ustires.org/sites/default/files/MAR 025 USTMA.pdf.
13 USTMA (2012a) "Rubber FAQs." U.S. Tire Manufacturers Association. Accessed on 19 November 2014.
14 USTMA (2012b) "Scrap Tire Markets: Facts and Figures - Scrap Tire Characteristics." U.S. Tire Manufacturers
15 Association. Accessed 18 on January 2012.
16 van Haaren, Rob, Themelis, N., and Goldstein, N. (2010) "The State of Garbage in America." BioCycle, October
17 2010. Volume 51, Number 10, pg. 16-23.
is Coal Mining
19 AAPG (1984) Coalbed Methane Resources of the United States. AAPG Studies in Geology Series #17.
20 Creedy, D.P. (1993) Methane Emissions from Coal Related Sources in Britain: Development of a Methodology.
21 Chemosphere, 26: 419-439.
22 DMME (2022) DGO Data Information System. Department of Mines, Minerals and Energy of Virginia. Available
23 online at https://www.dmme.virginia.gov/dgoinquiry/frmmain.aspx.
24 EIA (2022) Annual Coal Report 2021. Table 1. Energy Information Administration, U.S. Department of Energy.
25 Washington, D.C. DOE/EIA-0584.
26 El Paso (2009) Shoal Creek Mine Plan, El Paso Exploration & Production.
27 EPA (2022) Greenhouse Gas Reporting Program (GHGRP) 2020 Subpart FF: Underground Coal Mines.
28 EPA (2005) Surface Mines Emissions Assessment. Draft. U.S. Environmental Protection Agency.
29 EPA (1996) Evaluation and Analysis of Gas Content and Coal Properties of Major Coal Bearing Regions of the United
30 States. EPA/600/R-96-065. U.S. Environmental Protection Agency.
31 ERG (2022). Correspondence between ERG and Buchanan Mine.
32 Geological Survey of Alabama State Oil and Gas Board (GSA) (2022) Well Records Database. Available online at
33 http://www.gsa.state.al.us/ogb/database.aspx.
34 IEA (2022) Global coal production, 2018-2021, International Energy Agency, Paris, Licence: CC BY 4.0. Available
35 online at: https://www.iea.org/data-and-statistics/charts/global-coal-production-2018-2021.
36 IPCC (2019) 2019 Refinement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories. Calvo Buendia,
37 E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M., Ngarize S., Osako, A., Pyrozhenko, Y., Shermanau, P. and
38 Federici, S. (eds). Published: IPCC, Switzerland.
39 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
40 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
9-20 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
2 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
3 IPCC (2011) Use of Models and Facility-Level Data in Greenhouse Gas Inventories. Report of IPCC Expert Meeting on
4 Use of Models and Measurements in Greenhouse Gas Inventories 9-11 August 2010, Sydney, Australia. Eds:
5 Eggleston H.S., Srivastava N., Tanabe K., Baasansuren J., Fukuda M. IGES.
6 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
7 Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M. Manning, Z. Chen,
8 M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.)]. Cambridge University Press, Cambridge, United Kingdom
9 and New York, NY, USA, 996 pp.
10 JWR (2010) No. 4&7 Mines General Area Maps. Walter Energy: Jim Walter Resources.
11 King, Brian (1994) Management of Methane Emissions from Coal Mines: Environmental, Engineering, Economic and
12 Institutional Implication of Options. Neil and Gunter Ltd.
13 McElroy OVS (2022) Marshall County VAM Abatement Project Offset Verification Statement submitted to
14 California Air Resources Board, August 2022.
15 MSHA (2022) Data Transparency at MSHA. Mine Safety and Health Administration. Available online at
16 http://www.msha.gov/.
17 Mutmansky, Jan M. and Yanbei Wang (2000) Analysis of Potential Errors in Determination of Coal Mine Annual
18 Methane Emissions. Mineral Resources Engineering, 9(4).
19 Saghafi, Abouna (2013) Estimation of Fugitive Emissions from Open Cut Coal Mining and Measurable Gas Content.
20 13th Coal Operators' Conference, University of Wollongong, The Australian Institute of Mining and Metallurgy &
21 Mine Managers Association of Australia. 306-313.
22 USBM (1986) Results of the Direct Method Determination of the Gas Contents of U.S. Coal Basins. Circular 9067.
23 U.S. Bureau of Mines.
24 West Virginia Geological & Economic Survey (WVGES) (2022) Oil & Gas Production Data. Available online at
25 http://www.wvgs.wvnet.edu/www/datastat/datastat.htm.
26 Abandoned Underground Coal Mines
TJ CMM (2022) The International Coal Mine Methane Recovery and Utilization Project Database. Available online at:
28 https://www.globalmethane.org/resources/details.aspx?resourceid=1981
29 CMOP (2022) EPA's Coalbed Methane Outreach Program, Map of US Coal Mine Methane Current Projects and
30 Potential Opportunities. Available online at: https://www.epa.gov/cmop/map-us-coal-mine-methane-current-
31 projects-and-potential-opportunities
32 EPA (2004) Methane Emissions Estimates & Methodology for Abandoned Coal Mines in the U.S. Draft Final Report.
33 Washington, D.C. April 2004.
34 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
35 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
36 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
37 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
38 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
39 Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M. Manning, Z. Chen,
40 M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.)]. Cambridge University Press, Cambridge, United Kingdom
41 and New York, NY, USA, 996 pp.
References 9-21
-------
1 MSHA (2022) U.S. Department of Labor, Mine Health & Safety Administration, Mine Data Retrieval System.
2 Available online at: https://www.msha.gov/mine-data-retrieval-system
3 Petroleum Systems
4 API (1992) Global Emissions of Methane from Petroleum Sources. American Petroleum Institute, Health and
5 Environmental Affairs Department, Report No. DR140, February 1992.
6 BOEM (2022a) BOEM Platform Structures Online Query. Available online at:
7 https://www.data.boem.gov/Platform/PlatformStructures/Default.aspx.
8 BOEM (2022b) BOEM Oil and Gas Operations Reports - Part A (OGOR-A). Production Data for 1947 to 2021.
9 Download "Production Data" online at: https://www.data.boem.gov/Main/RawData.aspx.
10 BOEM (2022c) BOEM Oil and Gas Operations Reports - Part A (OGOR-A). Production Data for 1996 to 2021.
11 Available online at: https://www.data.boem.gov/Main/OGOR-A.aspx.
12 BOEM (2022d) BOEM Oil and Gas Operations Reports - Part B (OGOR-B). Flaring volumes for 1996 to 2021.
13 Available online at: https://www.data.boem.gov/Main/OGOR-B.aspx.
14 EIA (2022) Crude Oil Production. Energy Information Administration.
15 Enverus (2021) August 2021 Download. Enverus, Inc.
16 EPA (1997) Compilation of Air Pollutant Emission Factors, AP-42. Office of Air Quality Planning and Standards, U.S.
17 Environmental Protection Agency. Research Triangle Park, NC. October 1997.
18 EPA (1999) Estimates of Methane Emissions from the U.S. Oil Industry (Draft Report). Prepared by ICF International.
19 Office of Air and Radiation, U.S. Environmental Protection Agency. October 1999.
20 EPA (2017) 2017 Nonpoint Oil and Gas Emission Estimation Tool, Version 1.2. Prepared for U.S. Environmental
21 Protection Agency by Eastern Research Group, Inc. (ERG). October 2019.
22 EPA (2022) Greenhouse Gas Reporting Program. U.S. Environmental Protection Agency. Data reported as of August
23 12,2022.
24 EPA/GRI (1996) Methane Emissions from the Natural Gas Industry. Prepared by Radian. U.S. Environmental
25 Protection Agency. April 1996.
26 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
27 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
28 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
29 Natural Gas Systems
30 AHS (2021) U.S. Census Bureau's American Housing Survey (AHS). https://www.census.gov/programs-
31 surveys/ahs.html.
32 CBECS (2021) Energy Information Administration's Commercial Buildings Energy Consumption Survey (CBECS).
33 https://www.eia.gov/consumption/commercial.
34 CenSARA (2012) 2011 Oil and Gas Emission Inventory Enhancement Project for CenSARA States. Prepared by
35 ENVIRON International Corporation and Eastern Research Group, Inc. (ERG). Central States Air Resources Agencies
36 (CenSARA). December 2012.
37 Cusworth, D.H., Duren, R.M., Thorpe, A.K., Pandey S., Maasakkers, J.D., Aben, I., et al. (2021). Multisatellite
38 imaging of a gas well blowout enables quantification of total methane emissions. Geophysical Research Letters, 48,
39 e2020GL090864. https://doi.org/10.1029/2020GL090864
9-22 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 EIA (2022) Natural Gas Gross Withdrawals and Production. Energy Information Administration.
2 EIA (2022b) October 2021 Monthly Energy Review. Energy Information Administration.
3 https://www.eia.gov/totalenergy/data/monthly/archive/00352110.pdf.
4 Enverus (2021) August 2021 Download. Enverus, Inc.
5 EPA (1977) Atmospheric Emissions from Offshore Oil and Gas Development and Production. Office of Air Quality
6 Planning and Standards, Research Triangle Park, NC. PB272268. June 1977.
7 EPA (2020) MOVES3. https://www.epa.gov/moves/latest-version-motor-vehicle-emission-simulator-moves.
8 EPA (2021) Nonpoint Oil & Gas Emission estimation Tool. Data received via email in November 2021.
9 EPA (2022) Greenhouse Gas Reporting Program- Subpart W-Petroleum and Natural Gas Systems. Environmental
10 Protection Agency. Data reported as of August 12, 2022.
11 Fischer et al. (2018) "An Estimate of Natural Gas Methane Emissions from California Homes." Environmental
12 Science & Technology 2018, 52 (17), 10205-10213. https://pubs.acs.org/doi/10.1021/acs.est.8b03217.
13 GRI/EPA (1996) Methane Emissions from the Natural Gas Industry. Prepared by Harrison, M., T. Shires, J. Wessels,
14 and R. Cowgill, eds., Radian International LLC for National Risk Management Research Laboratory, Air Pollution
15 Prevention and Control Division, Research Triangle Park, NC. EPA-600/R-96-080a.
16 GTI (2001) Gas Resource Database: Unconventional Natural Gas and Gas Composition Databases. Second Edition.
17 GRI-01/0136.
18 GTI (2019) Classification of Methane Emissions from Industrial Meters, Vintage vs Modern Plastic Pipe, and Plastic-
19 lined Steel and Cast-Iron Pipe. June 2019. Gas Technology Institute and U.S. Department of Energy GTI Project
20 Number 22070. DOE project Number ED-FE0029061.
21 Illinois Office of Oil and Gas Resource Management (2020) State-level natural gas production quantities.
22 Indiana Division of Oil & Gas (2020) State-level natural gas production quantities.
23 Kansas Department of Health and Environment (2020) County-level produced water quantities.
24 Lamb, et al. (2015) "Direct Measurements Show Decreasing Methane Emissions from Natural Gas Local
25 Distribution Systems in the United States." Environmental Science & Technology, Vol. 49 5161-5169.
26 Lavoie et al. (2017) "Assessing the Methane Emissions from Natural Gas-Fired Power Plants and Oil Refineries."
27 Environmental Science & Technology. 2017 Mar 21;51(6):3373-3381. doi: 10.1021/acs.est.6b05531.
28 Maasakkers, Joannes D., Mark Omara, Ritesh Gautam, Alba Lorente, Sudhanshu Pandey, Paul Tol, Tobias Borsdorff,
29 Sander Houweling, Use Aben (2022). Reconstructing and quantifying methane emissions from the full duration of a
30 38-day natural gas well blowout using space-based observations. Remote Sensing of Environment.
31 https://doi.Org/10.1016/j.rse.2021.112755.Ohio Environmental Protection Agency (2020) Well-level produced
32 water quantities.
33 Oklahoma Department of Environmental Quality (2020) Well-level produced water quantities.
34 Pandey, S., Gautam, R., Houweling, S., van der Gon, H. D., Sadavarte, P., Borsdorff, T., et al. (2019). Satellite
35 observations reveal extreme methane leakage from a natural gas well blowout. Proceedings of the National
36 Academy of Sciences, 116, 26376-26381. https://doi.org/10.1073/pnas.1908712116
37 PHMSA (2022a) Gas Distribution Annual Data. Pipeline and Hazardous Materials Safety Administration, U.S.
38 Department of Transportation, Washington, DC. Available online at: https://www.phmsa.dot.gov/data-and-
39 statistics/pipeline/annual-report-mileage-gas-distribution-systems.
40 PHMSA (2022b) Underground Natural Gas STAR, Part C. Pipeline and Hazardous Materials Safety Administration,
41 U.S. Department of Transportation, Washington, DC. https://www.phmsa.dot.gov/data-and-statistics/pipeline/gas-
42 distribution-gas-gathering-gas-transmission-hazardous-liquids.
References 9-23
-------
1 West Virginia Department of Environmental Protection (2020) State-level natural gas production quantities.
2 Zimmerle et al. (2019) "Characterization of Methane Emissions from Gathering Compressor Stations." October
3 2019. Available at https://mountainscholar.org/handle/10217/195489.
4 Zimmerle, et al. (2015) "Methane Emissions from the Natural Gas Transmission and Storage System in the United
5 States." Environmental Science and Technology, Vol. 49 9374-9383.
6 Abandoned Oil and Gas Wells
7 Alaska Oil and Gas Conservation Commission, Available online at:
8 https://www.commerce.alaska.gov/web/aogcc/Data.aspx.
9 Arkansas Geological & Conservation Commission, "List of Oil & Gas Wells - Data From November 1,1936 to January
10 1,1955."
11 The Derrick's Handbook of Petroleum: A Complete Chronological and Statistical Review of Petroleum
12 Developments From 1859 to 1898 (V.l), (1898-1899) (V.2).
13 Enverus (2021) August 2021 Download. Enverus, Inc.
14 GRI/EPA (1996) Methane Emissions from the Natural Gas Industry. Prepared by Harrison, M., T. Shires, J. Wessels,
15 and R. Cowgill, eds., Radian International LLC for National Risk Management Research Laboratory, Air Pollution
16 Prevention and Control Division, Research Triangle Park, NC. EPA-600/R-96-080a.
17 Florida Department of Environmental Protection - Oil and Gas Program, Available online at:
18 https://floridadep.gov/water/oil-gas.
19 Geological Survey of Alabama, Oil & Gas Board, Available online at: https://www.gsa.state.al.us/ogb/.
20 GTI (2001) Gas Resource Database: Unconventional Natural Gas and Gas Composition Databases. Second Edition.
21 GRI-01/0136.
22 Interstate Oil and Gas Compact Commission (2021). IDLE AND ORPHAN OIL AND GAS WELLS: STATE AND
23 PROVINCIAL REGULATORY STRATEGIES 2021. Available online at:
24 https://iogcc.ok.gOv/sites/g/files/gmc836/f/iogcc idle and orphan wells 2021 final web.pdf.
25 Kang, et al. (2016) "Identification and characterization of high methane-emitting abandoned oil and gas wells."
26 PNAS, vol. 113 no. 48, 13636-13641, doi: 10.1073/pnas.l605913113.
27 Oklahoma Geological Survey. "Oklahoma Oil: Past, Present, and Future." Oklahoma Geology Notes, v. 62 no. 3,
28 2002 pp. 97-106.
29 Pennsylvania Department of Environmental Protection, Oil and Gas Reports - Oil and Gas Operator Well Inventory.
30 Available online at:
31 http://www.depreportingservices.state.pa.us/ReportServer/Pages/ReportViewer.aspx7/Oil Gas/OG Well Invento
32 ry.
33 Energy Sources of Precursor Greenhouse Gases
34 EPA (2022) "Criteria pollutants National Tier 1 for 1970 - 2021." National Emissions Inventory (NEI) Air Pollutant
35 Emissions Trends Data. Office of Air Quality Planning and Standards, February 2022. Available online at:
36 https://www.epa.gov/air-emissions-inventories/air-pollutant-emissions-trends-data.EPA (2021) Resource
37 Conservation and Recovery Act (RCRA) Info, Biennial Report, GM Form (Section 2- Onsite Management) and WR
38 Form.
9-24 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 EPA (2021) "2017 National Emissions Inventory (NEI) Technical Support Document (TSD)." Office of Air Quality
2 Planning and Standards, April 2021. Available online at: https://www.epa.gov/air-emissions-inventories/2017-
3 national-emissions-inventory-nei-technical-support-document-tsd.
4 EPA (2003) Email correspondence containing preliminary ambient air pollutant data. Office of Air Pollution and the
5 Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency. December 22, 2003.
6 International Bunker Fuels
7 Anderson, B.E., et al. (2011) Alternative Aviation Fuel Experiment (AAFEX), NASA Technical Memorandum, in press.
8 ASTM (1989) Military Specification for Turbine Fuels, Aviation, Kerosene Types, NATO F-34 (JP-8) and NATO F-35.
9 February 10,1989.
10 DHS (2008) Personal Communication with Elissa Kay, Residual and Distillate Fuel Oil Consumption (International
11 Bunker Fuels). Department of Homeland Security, Bunker Report. January 11, 2008.
12 DLA Energy (2022) Unpublished data from the Defense Fuels Automated Management System (DFAMS). Defense
13 Energy Support Center, Defense Logistics Agency, U.S. Department of Defense. Washington, D.C.
14 DOC (1991 through 2022) Unpublished Report of Bunker Fuel Oil Laden on Vessels Cleared for Foreign Countries.
15 Form-563. Foreign Trade Division, Bureau of the Census, U.S. Department of Commerce. Washington, D.C.
16 DOT (1991 through 2013) Fuel Cost and Consumption. Federal Aviation Administration, Bureau of Transportation
17 Statistics, U.S. Department of Transportation. Washington, D.C. DAI-10.
18 EIA (2022) Monthly Energy Review, November 2022, Energy Information Administration, U.S. Department of
19 Energy, Washington, D.C. DOE/EIA-0035(2022/11).
20 EPA (2020) EPA Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2019: Updated Gasoline and Diesel
21 Fuel CO2 Emission Factors - Memo.
22 FAA (2022) Personal Communication between FAA and John Steller, Mausami Desai, and Vincent Camobreco for
23 aviation emissions estimates from the Aviation Environmental Design Tool (AEDT). March 2022.
24 IPCC/UNEP/OECD/IEA (1997) Revised 1996 IPCC Guidelines for National Greenhouse Gas Inventories. 31
25 Intergovernmental Panel on Climate Change, United Nations Environment Programme, Organization for Economic
26 32 Co-Operation and Development, International Energy Agency, Paris, France.
27 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
28 Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F., D. Qin, G.K. Plattner, M.
29 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
30 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
31 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
32 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
33 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom,
34 996 pp.
35 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
36 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
37 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.USAF (1998) Fuel Logistics Planning. U.S. Air Force
38 pamphlet AFPAM23-221, May 1, 1998.
References 9-25
-------
1 Wood Biomass and Biofuel Consumption
2 EIA (2022a) Monthly Energy Review, November 2022. Energy Information Administration, U.S. Department of
3 Energy. Washington, D.C. DOE/EIA-0035(2022/11).
4 EIA (2022b) Biofuels explained: Use of biomass-based diesel fuel. Energy Information Administration, U.S.
5 Department of Energy. Washington, D.C. Available online at: https://www.eia.gov/energyexplained/biofuels/use-
6 of-biodiesel.php.
7 EPA (2022a) Acid Rain Program Dataset 1996-2021. Office of Air and Radiation, Office of Atmospheric Programs,
8 U.S. Environmental Protection Agency, Washington, D.C.
9 EPA (2022b). Greenhouse Gas Reporting Program (GHGRP). 2021 Envirofacts. Available online at:
10 https://ghgdata.epa.gov/ghgp/main.do.
11 EPA (2010) Carbon Content Coefficients Developed for EPA's Mandatory Reporting Rule. Office of Air and
12 Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
13 Lindstrom, P. (2006) Personal Communication. Perry Lindstrom, Energy Information Administration and Jean Kim,
14 ICF International.
is Energy Sources of Precursor Greenhouse Gases - TO BE
16 UPDATED FOR FINAL INVENTORY REPORT
17 EPA (2022) "Crosswalk of Precursor Gas Categories." U.S. Environmental Protection Agency. April 6, 2022.
18 EPA (2021a) "Criteria pollutants National Tier 1 for 1970 - 2020." National Emissions Inventory (NEI) Air Pollutant
19 Emissions Trends Data. Office of Air Quality Planning and Standards, March 2021. Available online at:
20 https://www.epa.gov/air-emissions-inventories/air-pollutant-emissions-trends-data.
21 EPA (2021b) "2017 National Emissions Inventory (NEI) Technical Support Document (TSD)." Office of Air Quality
22 Planning and Standards, April 2021. Available online at: https://www.epa.gov/air-emissions-inventories/2017-
23 national-emissions-inventorv-nei-technical-support-document-tsd.
24 EPA (1997) Compilation of Air Pollutant Emission Factors, AP-42. Office of Air Quality Planning and Standards, U.S.
25 Environmental Protection Agency. Research Triangle Park, NC. October 1997.
26 Industrial Processes and Product Use
27 EPA (2014) Greenhouse Gas Reporting Program. Developments on Publication of Aggregated Greenhouse Gas
28 Data, November 25, 2014. See http://www.epa.gov/ghgreporting/confidential-business-information-ghg-
29 reporting.
30 EPA (2002) Quality Assurance/Quality Control and Uncertainty Management Plan for the U.S. Greenhouse Gas
31 Inventory: Procedures Manual for Quality Assurance/Quality Control and Uncertainty Analysis, U.S. Greenhouse
32 Gas Inventory Program, U.S. Environmental Protection Agency, Office of Atmospheric Programs, EPA430-R-02-
33 007B, June 2002.
34 IPCC (2011) Use of Models and Facility-Level Data in Greenhouse Gas Inventories (Report of IPCC Expert Meeting
35 on Use of Models and Measurements in Greenhouse Gas Inventories 9-11 August 2010, Sydney, Australia) eds.:
36 Eggleston H.S., Srivastava N., Tanabe K., Baasansuren J., Fukuda M., Pub. IGES, Japan 2011.
9-26 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Cement Production
2 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
3 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
4 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
5 U.S. Bureau of Mines (1990 through 1993) Minerals Yearbook: Cement Annual Report. U.S. Department of the
6 Interior, Washington, D.C.
7 U.S. Environmental Protection Agency (EPA) (2015) Greenhouse Gas Reporting Program Report Verification.
8 Available online at https://www.epa.gov/sites/production/files/2015-
9 07/documents/ghgrp verification factsheet.pdf.
10 U.S. EPA (2022) Greenhouse Gas Reporting Program (GHGRP). Aggregation of Reported Facility Level Data under
11 Subpart H -National Level Clinker Production from Cement Production for Calendar Years 2014 through 2021.
12 Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington,
13 D.C.
14 United States Geological Survey (USGS) (2022a) 2020 Minerals Yearbook - Cement (Advance Release Tables). U.S.
15 Geological Survey, Reston, VA. July 2022.
16 USGS (2022b) Mineral Commodity Summaries: Cement. U.S. Geological Survey, Reston, VA. January 2022. Available
17 at: https://pubs.usgs.gov/periodicals/mcs2022/mcs2022-cement.pdf.
18 USGS (1995 through 2014) Minerals Yearbook - Cement. U.S. Geological Survey, Reston, VA.
19 Van Oss (2013a) 1990 through 2012 Clinker Production Data Provided by Hendrik van Oss (USGS) via email on
20 November 8, 2013.
21 Van Oss (2013b) Personal communication. Hendrik van Oss, Commodity Specialist of the U.S. Geological Survey
22 and Gopi Manne, Eastern Research Group, Inc. October 28, 2013.
23 Lime Product!
24 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
25 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
26 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
27 Males, E. (2003) Memorandum from Eric Males, National Lime Association to William N. Irving & Leif Hockstad,
28 Environmental Protection Agency. March 6, 2003.
29 Miner, R. and B. Upton (2002) Methods for estimating greenhouse gas emissions from lime kilns at kraft pulp mills.
30 Energy. Vol. 27 (2002), p. 729-738.
31 Seeger (2013) Memorandum from Arline M. Seeger, National Lime Association to Leif Hockstad, Environmental
32 Protection Agency. March 15, 2013.
33 U.S. Environmental Protection Agency (EPA) (2022) Greenhouse Gas Reporting Program (GHGRP). Aggregation of
34 Reported Facility Level Data under Subpart S-National Lime Production for Calendar Years 2010 through
35 2021. Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency,
36 Washington, D.C.
37 United States Geological Survey (USGS) (2022a) 2022 Mineral Commodities Summary: Lime. U.S. Geological Survey,
38 Reston, VA (January 2022).
39 USGS (2022b) Personal communication. Lori E. Apodaca, U.S. Geological Survey and Amanda Chiu, U.S.
40 Environmental Protection Agency. November 3, 2022.
Waste 27
-------
1 USGS (2021a) 2021 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA (January 2021).
2 USGS (2021b) 2020 Minerals Yearbook Annual Tables: Lime. U.S. Geological Survey, Reston, VA (August 2021).
3 USGS (2021c) (1992 through 2018) Minerals Yearbook: Lime. U.S. Geological Survey, Reston, VA (October 2021).
4 USGS (2020a) 2020 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA (January 2020).
5 USGS (2020b) (1992 through 2017) Minerals Yearbook: Lime. U.S. Geological Survey, Reston, VA (June 2020).
6 USGS (2020c) 2018 Minerals Yearbook Annual Tables: Lime. U.S. Geological Survey, Reston, VA (November 2020).
7 USGS (2019) 2016 Minerals Yearbook: Lime. U.S. Geological Survey, Reston, VA (August 2019).
8 USGS (2018a) 2018 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA (January 2018).
9 USGS (2018b) 2015 Minerals Yearbook: Lime. U.S. Geological Survey, Reston, VA (March 2018).
10 USGS (2012) 2012 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA (January 2012).
11 USGS (2011) 2011 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA (January 2011).
12 USGS (2010) 2010 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA (January 2010).
13 USGS (2008) 2008 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA (January 2008).
14 USGS (2007) 2007 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA (January 2007).
15 USGS (2002) 2002 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA (January 2002).
16 USGS (1996) 1996 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA (January 1996).
17 USGS (1991) 1991 Minerals Yearbook: Lime. U.S. Geological Survey, Reston, VA (1991).
is Glass Production
19 Federal Reserve (2022) Board of Governors of the Federal Reserve System (US), Industrial Production:
20 Manufacturing: Durable Goods: Glass and Glass Product (NAICS = 3272) [IPG3272N], retrieved from FRED, Federal
21 Reserve Bank of St. Louis. Available at: https://fred.stlouisfed.org/series/lPG3272N. Accessed on December 7,
22 2022.
23 Icenhour (2022) Expert judgment. Melissa Icenhour, RTI International. November 16, 2022.
24 IPCC (2006) 2006IPCCGuidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
25 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
26 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
27 U.S. Bureau of Mines (1991 and 1993a) Minerals Yearbook: Crushed Stone Annual Report. U.S. Department of the
28 Interior. Washington, D.C.
29 U.S. Department of Energy (DOE) (2002) Glass Industry of the Future: Energy and Environmental Profile of the U.S.
30 Glass Industry. Office of Industrial Technologies, U.S. Department of Energy. Washington, D.C.
31 U.S. Environmental Protection Agency (EPA) (2009) Technical Support Document for the Glass Manufacturing
32 Sector: Proposed Rule for Mandatory Reporting of Greenhouse Gases. U.S. Environmental Protection Agency,
33 Washington, D.C.
34 U.S. EPA (2022) Greenhouse Gas Reporting Program (GHGRP). Aggregation of Reported Facility Level Data under
35 Subpart N -National Glass Production for Calendar Years 2010 through 2021. Office of Air and Radiation, Office of
36 Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
37 United States Geological Survey (USGS) (1995 through 2015b) Minerals Yearbook: Soda Ash Annual Report. U.S.
38 Geological Survey, Reston, VA.
28 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 USGS (2017) Minerals Industry Surveys: Soda Ash in January 2017. U.S. Geological Survey, Reston, VA. March 2017.
2 USGS (2018) Mineral Industry Surveys: Soda Ash in February 2018. U.S. Geological Survey, Reston, VA. 2018.
3 USGS (2019) Mineral Industry Surveys: Soda Ash in December 2018. U.S. Geological Survey, Reston, VA. March
4 2019.
5 USGS (2020) Mineral Industry Surveys: Soda Ash in April 2020. U.S. Geological Survey, Reston, VA. July 2020.
6 USGS (2021) Mineral Industry Surveys: Soda Ash in April 2021. U.S. Geological Survey, Reston, VA. July 2021.
7 USGS (2022) Mineral Industry Surveys: Soda Ash in June 2022. U.S. Geological Survey, Reston, VA. November 2022.
8 Other Process Uses of Carbonates
9 AISI (2018 through 2021) Annual Statistical Report. American Iron and Steel Institute.
10 Kostick, D. S. (2012) Personal communication. Dennis S. Kostick, U.S. Geological Survey, Soda Ash Commodity
11 Specialist and Gopi Manne and Bryan Lange of Eastern Research Group, Inc. October 2012.
12 U.S. Bureau of Mines (1991 and 1993a) Minerals Yearbook: Crushed Stone Annual Report. U.S. Department of the
13 Interior. Washington, D.C.
14 U.S. Environmental Protection Agency (EPA) (2022). Greenhouse Gas Reporting Program (GHGRP). Dataset as of
15 August 12, 2022. Available online at: https://ghgdata.epa.gov/ghgp/
16 United States Geological Survey (USGS) (2017a) Mineral Industry Surveys: Soda Ash in January 2017. U.S.
17 Geological Survey, Reston, VA. March 2017.
18 USGS (2018) Mineral Industry Surveys: Soda Ash in February 2018. U.S. Geological Survey, Reston, VA. 2018.
19 USGS (2019) Mineral Industry Surveys: Soda Ash in April 2019. U.S. Geological Survey, Reston, VA. July 2019.
20 USGS (2020a) 2016 Minerals Yearbook: Stone, Crushed [Advanced Release]. U.S. Geological Survey, Reston, VA.
21 January 2020.
22 USGS (2020b) Mineral Industry Surveys: Soda Ash in April 2020. U.S. Geological Survey, Reston, VA. July 2020.
23 USGS (2020c) Minerals Yearbook 2017: Stone, Crushed [Advanced Data Release of the 2017 Annual Tables]. U.S.
24 Geological Survey, Reston, VA. August 2020.
25 USGS (2021a) 2017 Minerals Yearbook: Stone, Crushed [Advanced Release]. U.S. Geological Survey, Reston, VA.
26 June 2021.
27 USGS (2021b) 2020 Mineral Commodity Summaries: Stone (Crushed). U.S. Geological Survey, Reston, VA. January
28 2021.
29 USGS (2021c) Minerals Yearbook 2019: Soda Ash [Advanced Data Release of the 2019 Annual Tables]. U.S.
30 Geological Survey, Reston, VA. August 2021.
31 USGS (2021d) Mineral Industry Surveys: Soda Ash in April 2021. U.S. Geological Survey, Reston, VA. July 2021.
32 USGS (2022a) 2018 Minerals Yearbook: Stone, Crushed [Advanced Release]. U.S. Geological Survey, Reston, VA.
33 August 2022.
34 USGS (2022b) Mineral Industry Surveys: Soda Ash in August 2022. U.S. Geological Survey, Reston, VA. November
35 2022.
36 USGS (1995a through 2017) Minerals Yearbook: Crushed Stone Annual Report. U.S. Geological Survey, Reston, VA.
37 USGS (1994 through 2015b) Minerals Yearbook: Soda Ash Annual Report. U.S. Geological Survey, Reston, VA.
Waste 29
-------
1 Willett (2017) Personal communication, Jason Christopher Willett, U.S. Geological Survey and Mausami Desai and
2 John Steller, U.S. Environmental Protection Agency. March 9, 2017.
3 Willett (2022) Personal communication, Jason Christopher Willett, U.S. Geological Survey and Amanda Chiu, U.S.
4 Environmental Protection Agency. November 15, 2022.
5 Ammonia Production
6 ACC (2021) Business of Chemistry (Annual Data). American Chemistry Council, Arlington, VA.
7 Bark (2004) Coffeyville Nitrogen Plant. December 15, 2004. Available online at:
8 http://www.gasification.org/uploads/downloads/Conferences/2003/07BARK.pdf.
9 Coffeyville Resources Nitrogen Fertilizers (2012) Nitrogen Fertilizer Operations. Available online at:
10 http://coffevvillegroup.com/NitrogenFertilizerOperations/index.html.
11 Coffeyville Resources Nitrogen Fertilizers (2011) Nitrogen Fertilizer Operations. Available online at:
12 http://coffevvillegroup.com/NitrogenFertilizerOperations/index.html.
13 Coffeyville Resources Nitrogen Fertilizers (2010) Nitrogen Fertilizer Operations. Available online at:
14 http://coffevvillegroup.com/NitrogenFertilizerOperations/index.html.
15 Coffeyville Resources Nitrogen Fertilizers (2009) Nitrogen Fertilizer Operations. Available online at:
16 http://coffevvillegroup.com/NitrogenFertilizerOperations/index.html.
17 Coffeyville Resources Nitrogen Fertilizers, LLC (2005 through 2007a) Business Data. Available online at:
18 http://www.coffeyvillegroup.com/businessSnapshot.asp.
19 Coffeyville Resources Nitrogen Fertilizers (2007b) Nitrogen Fertilizer Operations. Available online at:
20 http://coffeyvillegroup.com/nitrogenMain.aspx.
21 Coffeyville Resources Energy, Inc. (CVR) (2012) CVR Energy, Inc. 2012 Annual Report. Available online at:
22 http://cvrenergy.com.
23 CVR (2013) CVR Energy, Inc. 2013 Annual Report. Available online at: http://cvrenergy.com.
24 CVR (2014) CVR Energy, Inc. 2014 Annual Report. Available online at: http://cvrenergy.com.
25 CVR (2015) CVR Energy, Inc. 2015 Annual Report. Available online at: http://cvrenergy.com.
26 CVR (2016) CVR Energy, Inc. 2016 CVI Annual Report on Form 10-K (Web). Available online at:
27 http://cvrenergy.com.
28 CVR (2017) CVR Energy, Inc. 2017 CVI Annual Report on Form 10-K (Web). Available online at:
29 http://cvrenergy.com.
30 CVR (2018) CVR Energy, Inc. 2018 CVI Annual Report on Form 10-K -Final. Available online at:
31 http://cvrenergy.com.
32 CVR (2019) CVR Energy, Inc. 2019 CVI Form 10-K - Final. Available online at: http://cvrenergy.com.
33 CVR (2020) CVR Energy, Inc. 2018 CVI Annual Report on Form 10-K -Final. Available online at:
34 http://cvrenergy.com.
35 CVR (2021) CVR Energy, Inc. 2021 CVI Annual Report on Form 10-K. Available online at: http://cvrenergy.com.EFMA
36 (2000) Best Available Techniques for Pollution Prevention and Control in the European Fertilizer Industry. Booklet
37 No. 5 of 8: Production of Urea and Urea Ammonium Nitrate. Available online at:
38 http://fertilizerseurope.com/site/index.php?id=390.
30 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
2 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
3 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
4 United States Census Bureau (2011) Current Industrial Reports Fertilizer Materials and Related Products: 2010
5 Summary. Available online at: http://www.census.gov/manufacturing/cir/historical data/roq325b/index.htrol.
6 U.S. Census Bureau (2010) Current Industrial Reports Fertilizer Materials and Related Products: 2009 Summary.
1 Available online at: http://www.census.gov/manufacturing/cir/historical data/mq325b/index.html.
8 U.S. Census Bureau (2009) Current Industrial Reports Fertilizer Materials and Related Products: 2008 Summary.
9 Available online at: http://www.census.gov/manufacturing/cir/historical data/mq325b/index.html.
10 U.S. Census Bureau (2008) Current Industrial Reports Fertilizer Materials and Related Products: 2007Summary.
11 Available online at: http://www.census.gov/cir/www/325/mq325b/mq325b075.xls.
12 U.S. Census Bureau (2007) Current Industrial Reports Fertilizer Materials and Related Products: 2006 Summary.
13 Available online at: http://www.census.gOv/industry/l/mq325bQ65.pdf.
14 U.S. Census Bureau (2006) Current Industrial Reports Fertilizer Materials and Related Products: 2005 Summary.
15 Available online at: http://www.census.gov/cir/www/325/mq325b.html.
16 U.S. Census Bureau (2004, 2005) Current Industrial Reports Fertilizer Materials and Related Products: Fourth
17 Quarter Report Summary. Available online at: http://www.census.gov/cir/www/325/mq325b.html.
18 U.S. Census Bureau (1998 through 2003) Current Industrial Reports Fertilizer Materials and Related Products:
19 Annual Reports Summary. Available online at: http://www.census.gov/cir/www/325/mq325b.html.
20 U.S. Census Bureau (1991 through 1994) Current Industrial Reports Fertilizer Materials Annual Report. Report No.
21 MQ28B. U.S. Census Bureau, Washington, D.C.
22 United States EIA (2022) Monthly Energy Review, November 2022, Energy Information Administration, U.S.
23 Department of Energy, Washington, DC. DOE/EIA-0035(2022/11).
24 United States Environmental Protection Agency (EPA) (2018) Greenhouse Gas Reporting Program. Aggregation of
25 Reported Facility Level Data under Subpart G -Annual Urea Production from Ammonia Manufacturing for Calendar
26 Years 2011-2016. Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection
27 Agency, Washington, D.C.
28 U.S. EPA (2022) Greenhouse Gas Reporting Program. Aggregation of Reported Facility Level Data under Subpart G -
29 Annual Urea Production from Ammonia Manufacturing for Calendar Years 2017-2021. Office of Air and Radiation,
30 Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
31 United States Geological Survey (USGS) (2022) 2022 Mineral Commodity Summaries: Nitrogen (Fixed) - Ammonia.
32 January 2022. Available online at: https://pubs.usgs.gov/periodicals/mcs2022/mcs2022-nitrogen.pdf.
33 USGS (1994-2009) Minerals Yearbook: Nitrogen. Available online at:
34 http://minerals.usgs.gov/minerals/pubs/commodity/nitrogen/.
35 Urea Consumption for Non-Agricultural Purposes
36 EFMA (2000) Best Available Techniques for Pollution Prevention and Control in the European Fertilizer Industry.
37 Booklet No. 5 of 8: Production of Urea and Urea Ammonium Nitrate.
38 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
39 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
40 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Waste 31
-------
1 TFI (2002) U.S. Nitrogen Imports/Exports Table. The Fertilizer Institute. Available online at:
2 http://www.tfi.org/statistics/usnexim.asp. August 2002.
3 United States Census Bureau (2001 through 2011) Current Industrial Reports Fertilizer Materials and Related
4 Products: Annual Summary. Available online at:
5 http://www.census.gov/manufacturing/cir/historical data/index.html.
6 United States Department of Agriculture (2012) Economic Research Service Data Sets, Data Sets, U.S. Fertilizer
7 Imports/Exports: Standard Tables. Available online at: http://www.ers.usda.gov/data-products/fertilizer-
8 importsexports/standard-tables.aspx.
9 United States Environmental Protection Agency (EPA) (2018) Greenhouse Gas Reporting Program. Aggregation of
10 Reported Facility Level Data under Subpart G -Annual Urea Production from Ammonia Manufacturing for Calendar
11 Years 2011-2016. Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection
12 Agency, Washington, D.C.
13 U.S. EPA (2022a) Greenhouse Gas Reporting Program. Aggregation of Reported Facility Level Data under Subpart G
14 -Annual Urea Production from Ammonia Manufacturing for Calendar Years 2017-2021. Office of Air and Radiation,
15 Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
16 U.S. EPA (2022b). Greenhouse Gas Reporting Program. Dataset as of August 13, 2022. Available online at:
17 https://ghgdata.epa.gov/ghgp/.
18 United States International Trade Commission (ITC) (2002) United States International Trade Commission
19 Interactive Tariff and Trade DataWeb, Version 2.5.0. Available online at: http://dataweb.usitc.gov/. August 2002.
20 United States Geological Survey (USGS) (1994 through 2021a) Minerals Yearbook: Nitrogen. Available online at:
21 http://minerals.usgs.gov/minerals/pubs/commodity/nitrogen/.
22 USGS (2022b) Minerals Commodity Summaries: Nitrogen (Fixed)-Ammonia. Available online at:
23 http://minerals.usgs.gov/minerals/pubs/commodity/nitrogen/.
24 Nitric Acid Production
25 Climate Action Reserve (CAR) (2013) Project Report. Available online at:
26 https://thereserve2.apx.com/myModule/rpt/myrpt.asp?r=lll. Accessed on 18 January 2013.
27 Desai (2012) Personal communication. Mausami Desai, U.S. Environmental Protection Agency, January 25, 2012.
28 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
29 Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F., D. Qin, G.-K. Plattner, M.
30 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
31 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
32 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
33 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
34 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom
35 996 pp.
36 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
37 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
38 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
39 United States Census Bureau (2010a) Current Industrial Reports. Fertilizers and Related Chemicals: 2009. 'Table 1:
40 Summary of Production of Principle Fertilizers and Related Chemicals: 2009 and 2008." June, 2010. MQ325B(08)-5.
41 Available online at: http://www.census.gov/manufacturing/cir/historical data/mq325b/index.html.
32 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 U.S. Census Bureau (2010b) Personal communication between Hilda Ward (of U.S. Census Bureau) and Caroline
2 Cochran (of ICF International). October 26, 2010 and November 5, 2010.
3 U.S. Census Bureau (2009) Current Industrial Reports. Fertilizers and Related Chemicals: 2008. 'Table 1: Shipments
4 and Production of Principal Fertilizers and Related Chemicals: 2004 to 2008." June, 2009. MQ325B(08)-5. Available
5 online at: http://www.census.gov/manufacturing/cir/historical data/mq325b/index.html.
6 U.S. Census Bureau (2008) Current Industrial Reports. Fertilizers and Related Chemicals: 2007. 'Table 1: Shipments
7 and Production of Principal Fertilizers and Related Chemicals: 2003 to 2007." June, 2008. MQ325B(07)-5. Available
8 online at: http://www.census.gov/manufacturing/cir/historical data/mq325b/index.html.
9 United States Environmental Protection Agency (EPA) (2022) Greenhouse Gas Reporting Program. Aggregation of
10 Reported Facility Level Data under Subpart V -National Nitric Acid Production for Calendar Years 2010 through
11 2021. Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency,
12 Washington, D.C.
13 U.S. EPA (2015) Greenhouse Gas Reporting Program Report Verification. Available online at
14 https://www.epa.gov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
15 U.S. EPA (2013) Draft Nitric Acid Database. U.S. Environmental Protection Agency, Office of Air and Radiation.
16 September 2010.
17 U.S. EPA (2012) Memorandum from Mausami Desai, U.S. EPA to Mr. Bill Herz, The Fertilizer Institute. November
18 26,2012.
19 U.S. EPA (2010) Available and Emerging Technologies for Reducing Greenhouse Gas Emissions from the Nitric Acid
20 Production Industry. Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency. Research
21 Triangle Park, NC. December 2010. Available online at: http://www.epa.gov/nsr/ghgdocs/nitricacid.pdf.
22 U.S. EPA (1998) Compilation of Air Pollutant Emission Factors, AP-42. Office of Air Quality Planning and Standards,
23 U.S. Environmental Protection Agency. Research Triangle Park, NC. February 1998.
24 Adipic Acid Production
25 ACC (2022) Business of Chemistry (Annual Data). American Chemistry Council, Arlington, VA.
26 C&EN (1995) "Production of Top 50 Chemicals Increased Substantially in 1994." Chemical & Engineering News,
27 73(15):17. April 10, 1995.
28 C&EN (1994) 'Top 50 Chemicals Production Rose Modestly Last Year." Chemical & Engineering News, 72(15):13.
29 April 11,1994.
30 C&EN (1993) 'Top 50 Chemicals Production Recovered Last Year." Chemical & Engineering News, 71(15):11. April
31 12,1993.
32 C&EN (1992) "Production of Top 50 Chemicals Stagnates in 1991." Chemical & Engineering News, 70(15): 17. April
33 13, 1992.
34 CMR (2001) "Chemical Profile: Adipic Acid." Chemical Market Reporter. July 16, 2001.
35 CMR (1998) "Chemical Profile: Adipic Acid." Chemical Market Reporter. June 15,1998.
36 CW (2005) "Product Focus: Adipic Acid." Chemical Week. May 4, 2005.
37 CW (1999) "Product Focus: Adipic Acid/Adiponitrile." Chemical Week, p. 31. March 10,1999.
38 Desai (2010, 2011) Personal communication. Mausami Desai, U.S. Environmental Protection Agency and Adipic
39 Acid Plant Engineers. 2010 and 2011.
40 ICIS (2007) "Adipic Acid." ICIS Chemical Business Americas. July 9, 2007.
Waste 33
-------
1 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
2 Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F., D. Qin, G.-K. Plattner, M.
3 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
4 Cambridge, United Kingdom and New York, NY, USA.
5 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
6 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
7 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom.
8 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
9 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
10 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
11 Reimer, R.A., Slaten, C.S., Seapan, M., Koch, T.A. and Triner, V.G. (1999) "Implementation of Technologies for
12 Abatement of N2O Emissions Associated with Adipic Acid Manufacture." Proceedings of the 2nd Symposium on
13 Non-CC>2 Greenhouse Gases (NCGG-2), Noordwijkerhout, The Netherlands, 8-10 Sept. 1999, Ed. J. van Ham et at.,
14 Kluwer Academic Publishers, Dordrecht, pp. 347-358.
15 Thiemens, M.H., and W.C. Trogler (1991) "Nylon production; an unknown source of atmospheric nitrous oxide."
16 Science 251:932-934.
17 United States Environmental Protection Agency (EPA) (2022) Greenhouse Gas Reporting Program. Subpart E Data.
18 Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington,
19 D.C. Available online at: https://www.epa.gov/ghgreporting/ghg-reporting-program-data-sets.
20 U.S. EPA (2019, 2020) Greenhouse Gas Reporting Program. Subpart E, S-CEMS, BB, CC, LL Data Set (XLSX) (Adipic
21 Acid Tab). Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency,
22 Washington, D.C. Available online at: https://www.epa.gov/ghgreporting/ghg-reporting-program-data-sets.
23 U.S. EPA (2015). Greenhouse Gas Reporting Program Report Verification. Available online at
24 https://www.epa.gov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
25 U.S. EPA (2014 through 2018) Greenhouse Gas Reporting Program. Subpart E, S-CEMS, BB, CC, LL Data Set (XLSX)
26 (Adipic Acid Tab). Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection
27 Agency, Washington, D.C. Available online at: http://www2.epa.gov/ghgreporting/ghg-reporting-program-data-
28 sets.
29 U.S. EPA (2010 through 2013) Analysis of Greenhouse Gas Reporting Program data - Subpart E (Adipic Acid), Office
30 of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
31 Caprolactam, Glyoxal and Glyoxylic Acid Production
32 American Chemistry Council (ACC) (2022) Business of Chemistry (Annual Data). American Chemistry Council,
33 Arlington, VA.
34 AdvanSix (2022) AdvanSix's Hopewell Facility Fact Sheet. Retrieved from:
35 https://www.advansix.com/hopewell/about-us/ on October 7, 2022.
36 BASF (2022) Welcome to BASF in Freeport Texas. Retrieved from https://www.basf.com/us/en/who-we-
37 are/organization/locations/featured-sites/Freeport.html on October 7, 2022.
38 ChemView (2021). Compilation of data submitted under TSCA in 2012 and 2016. Accessed April 2021. Available at
39 https://chemview.epa.gov/chemview.
40 Cline, D. (2019) Firm to Clean Up and Market Former Fibrant Site. The Augusta Chronicle. September 9, 2019.
41 Retrieved from https://www.augustachronicle.com.
34 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Cogburn, M.O. (2012). United States v. Emerald Carolina Chem., LLC. Consent Decree, Civil Action No. 3:12-cv-
2 00554. United States District Court for the Western District of North Carolina, Charlotte Division. Decided October
3 25, 2012. Available at https://casetext.com/case/united-states-v-emerald-carolina-chem.
4 Ecofys, et al. (2009). Methodology for the free allocation of emission allowances in the EU ETS post 2012: Sector
5 Report for the Chemical Industry. Prepared by Ecofys, Fraunhofer Institute for Systems and Innovation Research,
6 and Oko-lnstitut for the European Commission. November 2009. Available at
7 https://ec.europa.eu/clima/system/files/2016-ll/bm study-chemicals en.pdf.
8 Fior Markets (2018). Summary of Global Glyoxylic Acid Market by Manufacturers, Regions, Type and Application,
9 Forecast to 2023. July 2018. Available at: https://www.fiormarkets.com/report/global-glvoxylic-acid-market-bv-
10 manufacturers-regions-type-268394.html.
11 ICIS (2004) Chemical Profile - Caprolactam. January 5, 2004. Available online at:
12 https://www.icis.com/explore/resources/news/2005/12/02/547244/chemical-profile-caprolactam/.
13 ICIS (2006) Chemical Profile - Caprolactam. October 15, 2006. Available online at:
14 https://www.icis.com/explore/resources/news/2006/10/18/2016832/chemical-profile-caprolactam/.
15 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
16 Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F., D. Qin, G.-K. Plattner, M.
17 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
18 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
19 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
20 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
21 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom
22 996 pp.
23 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
24 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
25 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
26 Shaw Industries Group, Inc. (Shaw) (2015) "Shaw Carpet Recycling Facility Successfully Processes Nylon and
27 Polyester". July 13, 2015. Available online at: https://shawinc.com/Newsroom/Press-Releases/Shaw-Carpet-
28 Recycling-Facilitv-Successfully-Proces/.
29 Teles, J.H. et al. (2015). "Oxidation." Ullmann's Encyclopedia of Industrial Chemistry. Wiley-VCH Verlag GmbH &
30 Co. KGaA, Weinheim. 10.1002/14356007.al8_261.pub2.
31 Textile World (2000) "Evergreen Makes Nylon Live Forever". Textile World. October 1, 2000. Available online at:
32 https://www.textileworld.com/textile-world/textile-news/200Q/10/evergreen-makes-nylon-live-forever/.
33 Carbide Production and Consumption
34 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
35 Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F., D. Qin, G.-K. Plattner, M.
36 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
37 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
38 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
39 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
40 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom
41 996 pp.
Waste 35
-------
1 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
2 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
3 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
4 Environment and Climate Change Canada (ECCC) (2022), Personal Communication between Mausami Desai and
5 Amanda Chiu (EPA) and Genevieve Leblanc-Power (ECCC). April 12, 2022.
6 United States Census Bureau (2005 through 2021) USITC Trade DataWeb. Available online at:
7 http://dataweb.usitc.gov/.
8 United States Geological Survey (USGS) (2021) 2018 Minerals Yearbook: Abrasives, Manufactured [Advance
9 Release], October 2021. U.S. Geological Survey, Reston, VA. Available online at:
10 https://www.usgs.gov/centers/nmic/manufactured-abrasives-statistics-and-information.
11 USGS (1991a through 2021) Minerals Yearbook: Manufactured Abrasives Annual Report. U.S. Geological Survey,
12 Reston, VA. Available online at: https://www.usgs.gov/centers/national-minerals-information-
13 center/manufactured-abrasives-statistics-and-information.
14 USGS (1991b through 2021b) Minerals Yearbook: Silicon Annual Report. U.S. Geological Survey, Reston, VA.
15 Available online at: http://minerals.usgs.gov/minerals/pubs/commodity/silicon/.
16 Washington Mills (2021), North Grafton, MA. Available online at: https://www.washingtonmills.com/silicon-
17 carbide/sic-industries.
is Titanium Dioxide Production
19 Gambogi, J. (2002) Telephone communication. Joseph Gambogi, Commodity Specialist, U.S. Geological Survey and
20 Philip Groth, ICF International. November 2002.
21 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
22 Inventories Programme, Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
23 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
24 United States Geological Survey (USGS) (2022) Mineral Commodity Summaries: Titanium and Titanium Dioxide.
25 U.S. Geological Survey, Reston, Va. January 2022. USGS (1991 through 2021) Minerals Yearbook: Titanium. U.S.
26 Geological Survey, Reston, VA.
27 Soda Ash Production
28 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
29 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
30 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
31 United States Geological Survey (USGS) (2022a) Mineral Commodity Summary: Soda Ash. U.S. Geological Survey,
32 Reston, VA. Accessed September 2022.
33 USGS (2022b) Mineral Industry Surveys: Soda Ash in June 2022. U.S. Geological Survey, Reston, VA. Accessed
34 September 2022.
35 USGS (2021) Mineral Industry Surveys: Soda Ash in April 2021. U.S. Geological Survey, Reston, VA. Accessed
36 September 2021.
37 USGS (2020) Mineral Industry Surveys: Soda Ash in April 2020. U.S. Geological Survey, Reston, VA. Accessed
38 September 2020.
39 USGS (2019) Mineral Industry Surveys: Soda Ash in April 2019. U.S. Geological Survey, Reston, VA. Accessed August
40 2019.
36 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 USGS (2018a) Mineral Industry Surveys: Soda Ash in February 2018. U.S. Geological Survey, Reston, VA. Accessed
2 September 2018.
3 USGS (2017) Mineral Industry Surveys: Soda Ash in January 2017. U.S. Geological Survey, Reston, VA. March 2017.
4 USGS (2016) Mineral Industry Surveys: Soda Ash in November 2016. U.S. Geological Survey, Reston, VA. January
5 2017.
6 USGS (2015a) Mineral Industry Surveys: Soda Ash in July 2015. U.S. Geological Survey, Reston, VA. September
7 2015.
8 USGS (1994 through 2015b, 2018b) Minerals Yearbook: Soda Ash Annual Report. U.S. Geological Survey, Reston,
9 VA.
10 USGS (1995c) Trona Resources in the Green River Basin, Southwest Wyoming. U.S. Department of the Interior, U.S.
11 Geological Survey. Open-File Report 95-476. Wiig, Stephen, Grundy, W.D., Dyni, John R.
12 Petrochemical Production
13 ACC (2022a) Business of Chemistry (Annual Data). American Chemistry Council, Arlington, VA.
14 ACC (2022b) Personal communication. Martha Moore, American Chemistry Council and Amanda Chiu, U.S.
15 Environmental Protection Agency. September 27, 2022.
16 AN (2014) About Acrylonitrile: Production. AN Group, Washington, D.C. Available online at:
17 http://www.angroup.org/about/production.cfm.
18 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
19 Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F., D. Qin, G.-K. Plattner, M.
20 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
21 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
22 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
23 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
24 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press, Cambridge, United Kingdom,
25 996 pp.
26 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
27 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
28 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
29 Johnson, G. L. (2005 through 2010) Personal communication. Greg Johnson of Liskow & Lewis, on behalf of the
30 International Carbon Black Association (ICBA) and Caroline Cochran, ICF International. September 2010.
31 Johnson, G. L. (2003) Personal communication. Greg Johnson of Liskow & Lewis, on behalf of the International
32 Carbon Black Association (ICBA) and Caren Mintz, ICF International. November 2003.
33 United States Environmental Protection Agency (EPA) (2022) Greenhouse Gas Reporting Program. Aggregation of
34 Reported Facility Level Data under Subpart X -National Petrochemical Production for Calendar Years 2010 through
35 2021. Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency,
36 Washington, D.C.
37 U.S. EPA (2015) Greenhouse Gas Reporting Program Report Verification. Available online at
38 https://www.epa.gov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
39 U.S. EPA (2008) Technical Support Document for the Petrochemical Production Sector: Proposed Rule for
40 Mandatory Reporting of Greenhouse Gases. U.S. Environmental Protection Agency. September 2008.
Waste 37
-------
1 U.S. EPA (2000) Economic Impact Analysis for the Proposed Carbon Black Manufacturing NESHAP, U.S.
2 Environmental Protection Agency. Research Triangle Park, NC. EPA-452/D-00-003. May 2000.
3 HCFC-22 Production
4 ARAP (2010) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
5 Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. September 10, 2010.
6 ARAP (2009) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
7 Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. September 21, 2009.
8 ARAP (2008) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
9 Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. October 17, 2008.
10 ARAP (2007) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
11 Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. October 2, 2007.
12 ARAP (2006) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
13 Atmospheric Policy to Sally Rand of the U.S. Environmental Protection Agency. July 11, 2006.
14 ARAP (2005) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
15 Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. August 9, 2005.
16 ARAP (2004) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
17 Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. June 3, 2004.
18 ARAP (2003) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
19 Atmospheric Policy to Sally Rand of the U.S. Environmental Protection Agency. August 18, 2003.
20 ARAP (2002) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
21 Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. August 7, 2002.
22 ARAP (2001) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
23 Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. August 6, 2001.
24 ARAP (2000) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
25 Atmospheric Policy to Sally Rand of the U.S. Environmental Protection Agency. August 13, 2000.
26 ARAP (1999) Facsimile from Dave Stirpe, Executive Director, Alliance for Responsible Atmospheric Policy to
27 Deborah Ottinger Schaefer of the U.S. Environmental Protection Agency. September 23,1999.
28 ARAP (1997) Letter from Dave Stirpe, Director, Alliance for Responsible Atmospheric Policy to Elizabeth Dutrow of
29 the U.S. Environmental Protection Agency. December 23,1997.
30 EPA (2015) Greenhouse Gas Reporting Program Report Verification. Available online at
31 https://www.epa.gov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
32 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
33 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
34 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom
35 996 pp.
36 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
37 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
38 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
39 RTI (2008) "Verification of Emission Estimates of HFC-23 from the Production of HCFC-22:Emissions from 1990
40 through 2006." Report prepared by RTI International for the Climate Change Division. March 2008.
38 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 RTI (1997) "Verification of Emission Estimates of HFC-23 from the Production of HCFC-22: Emissions from 1990
2 through 1996." Report prepared by Research Triangle Institute for the Cadmus Group. November 25,1997; revised
3 February 16,1998.
4 UNFCCC (2014) Report of the Conference of the Parties on its nineteenth session, held in Warsaw from 11 to 23
5 November 2013. United Nations Framework Convention on Climate Change, Warsaw. (FCCC/CP/2013/10/Add.3).
6 January 31, 2014. Available online at: http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
7 Carbon Dioxide Consumption
8 ARI (1990 through 2010) CO2 Use in Enhanced Oil Recovery. Deliverable to ICF International under Task Order 102,
9 July 15, 2011.
10 ARI (2007) CO2-EOR: An Enabling Bridge for the Oil Transition. Presented at "Modeling the Oil Transition—a
11 DOE/EPA Workshop on the Economic and Environmental Implications of Global Energy Transitions." Washington,
12 D.C. April 20-21, 2007.
13 ARI (2006) CO2-EOR: An Enabling Bridge for the Oil Transition. Presented at "Modeling the Oil Transition—a
14 DOE/EPA Workshop on the Economic and Environmental Implications of Global Energy Transitions." Washington,
15 D.C. April 20-21, 2006.
16 Broadhead (2003) Personal communication. Ron Broadhead, Principal Senior Petroleum Geologist and Adjunct
17 faculty, Earth and Environmental Sciences Department, New Mexico Bureau of Geology and Mineral Resources,
18 and Robin Petrusak, ICF International. September 5, 2003.
19 COGCC (2014) Monthly CO2 Produced by County (1999-2009). Available online at:
20 http://cogcc.state.co.us/COGCCReports/production.aspx?id=MonthlyCQ2ProdByCounty. Accessed October 2014.
21 Denbury Resources Inc. (2002 through 2010) Annual Report: 2001 through 2009, Form 10-K. Available online at:
22 http://www.denburv.com/investor-relations/SEC-Filings/SEC-Filings-Details/default.aspx?Filingld=9823015.
23 Accessed September 2014.
24 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
25 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
26 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
27 New Mexico Bureau of Geology and Mineral Resources (2006) Natural Accumulations of Carbon Dioxide in New
28 Mexico and Adjacent Parts of Colorado and Arizona: Commercial Accumulation of CO2. Available online at:
29 http://geoinfo.nmt.edU/staff/broadhead/C02.html#commercial.
30 U.S. Environmental Protection Agency (EPA) (2022) Greenhouse Gas Reporting Program (GHGRP). Aggregation of
31 Reported Facility Level Data under Subpart PP -National Level CO2 Transferred for Food & Beverage Applications
32 for Calendar Years 2010 through 2021. Office of Air and Radiation, Office of Atmospheric Programs, U.S.
33 Environmental Protection Agency, Washington, D.C.
34 U.S. EPA (2015) Greenhouse Gas Reporting Program Report Verification. Available online at
35 https://www.epa.gov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
36 Phosphoric Acid Production
37 EFMA (2000) "Production of Phosphoric Acid." Best Available Techniques for Pollution Prevention and Control in
38 the European Fertilizer Industry. Booklet 4 of 8. European Fertilizer Manufacturers Association. Available online at:
39 http://www.efma.org/Publications/BAT%202000/Bat04/section04.asp.
40 Florida Institute of Phosphate Research (FIPR) (2003a) "Analyses of Some Phosphate Rocks." Facsimile Gary
41 Albarelli, Florida Institute of Phosphate Research, Bartow, Florida, to Robert Lanza, ICF International. July 29, 2003.
Waste 39
-------
1 FIPR (2003b) Florida Institute of Phosphate Research. Personal communication. Mr. Michael Lloyd, Laboratory
2 Manager, FIPR, Bartow, Florida, to Mr. Robert Lanza, ICF International. August 2003.
3 Golder Associates and M3 Engineering, Bayovar 12 Phosphate Project: Nl 43-101 Updated Pre-Feasibility Study,
4 Issued June 28, 2016. Available at:
5 https://www.sec.gov/Archives/edgar/data/1471603/000121716016000634/focusiune2016bayovar techrep.htm.
6 Accessed on October 7, 2020.
7 NCDENR (2013) North Carolina Department of Environment and Natural Resources, Title V Air Permit Review for
8 PCS Phosphate Company, Inc. - Aurora. Available online at:
9 http://www.ncair.org/permits/permit reviews/PCS rev 08282012.pdf. Accessed on January 25, 2013.
10 United States Geological Survey (USGS) (2022) Mineral Commodity Summaries: Phosphate Rock 2022. January
11 2022. U.S. Geological Survey, Reston, VA. Accessed September 2022. Available online at:
12 https://www.usgs.gov/centers/nmic/phosphate-rock-statistics-and-information
13 USGS (2021a) Mineral Commodity Summaries: Phosphate Rock 2021. January 2021. U.S. Geological Survey, Reston,
14 VA. Accessed August 2021. Available online at: https://www.usgs.gov/centers/nmic/phosphate-rock-statistics-and-
15 information.
16 USGS (2021b) Personal communication between Stephen Jasinski (USGS) and Amanda Chiu (EPA) on August 25,
17 2021.
18 USGS (2020) Mineral Commodity Summaries: Phosphate Rock 2020. January 2020. U.S. Geological Survey, Reston,
19 VA. Accessed September 2020. Available online at: https://www.usgs.gov/centers/nmic/phosphate-rock-statistics-
20 and-information.
21 USGS (2019) Mineral Commodity Summaries: Phosphate Rock 2019. February 2019. U.S. Geological Survey, Reston,
22 VA. Accessed August 2019. Available online at: https://www.usgs.gov/centers/nmic/phosphate-rock-statistics-and-
23 information.
24 USGS (2019b) Communication between Stephen Jasinski (USGS) and John Steller EPA on November 15, 2019.
25 USGS (2018) Mineral Commodity Summaries: Phosphate Rock 2018. January 2018. U.S. Geological Survey, Reston,
26 VA. Available online at: https://www.usgs.gov/centers/nmic/phosphate-rock-statistics-and-information.
27 USGS (2017) Mineral Commodity Summaries: Phosphate Rock 2017. January 2017. U.S. Geological Survey, Reston,
28 VA. Available online at: https://www.usgs.gov/centers/nmic/phosphate-rock-statistics-and-information.
29 USGS (2016) Mineral Commodity Summaries: Phosphate Rock 2016. January 2016. U.S. Geological Survey, Reston,
30 VA. Available online at: https://www.usgs.gov/centers/nmic/phosphate-rock-statistics-and-information.
31 USGS (1994 through 2015b) Minerals Yearbook. Phosphate Rock Annual Report. U.S. Geological Survey, Reston, VA.
32 USGS (2012) Personal communication between Stephen Jasinski (USGS) and Mausami Desai (EPA) on October 12,
33 2012.
34 Iron and Steel Production and Metallurgical Coke Production
35 American Coke and Coal Chemicals Institute (ACCCI) (2021) U.S. Coke Plants as of November 2021, ACCCI,
36 Washington, D.C. November 2021.
37 American Iron and Steel Institute (AISI) (2004 through 2022) Annual Statistical Report, American Iron and Steel
38 Institute, Washington, D.C.
39 Carroll (2016) Personal communication, Colin P. Carroll, Director of Environment, Health and Safety, American Iron
40 and Steel Institute and Mausami Desai, U.S. Environmental Protection Agency, December 2016.
40 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Carroll (2017) Personal communication, Colin P. Carroll, Director of Environment, Health and Safety, American Iron
2 and Steel Institute and John Steller, U.S. Environmental Protection Agency, November 2017.
3 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
4 Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F., D. Qin, G.-K. Plattner, M.
5 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
6 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
7 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
8 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
9 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United
10 Kingdom996 pp. IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
11 Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L.
12 Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
13 IPCC/UNEP/OECD/IEA (1995) "Volume 3: Greenhouse Gas Inventory Reference Manual. Table 2-2." IPCC Guidelines
14 for National Greenhouse Gas Inventories. Intergovernmental Panel on Climate Change, United Nations
15 Environment Programme, Organization for Economic Co-Operation and Development, International Energy
16 Agency. IPCC WG1 Technical Support Unit, United Kingdom.
17 Steiner (2008) Personal communication, Bruce Steiner, Technical Consultant with the American Iron and Steel
18 Institute and Mausami Desai, U.S. Environmental Protection Agency, November 2008.
19 Tuck (2020) Personal communication, Christopher Tuck, Commodity Specialist, U.S. Geological Survey and Amanda
20 Chiu, U.S. Environmental Protection Agency, November 2020.
21 United States Department of Energy (DOE) (2000) Energy and Environmental Profile of the U.S. Iron and Steel
22 Industry. Office of Industrial Technologies, U.S. Department of Energy. August 2000. DOE/EE-0229.EIA.
23 United States Energy Information Administration (EIA) (1998 through 2019) Quarterly Coal Report: October-
24 December, Energy Information Administration, U.S. Department of Energy, Washington, D.C.
25 U.S. EIA (2021 through 2022) Quarterly Coal Report: January - March, Energy Information Administration, U.S.
26 Department of Energy. Washington, D.C.
27 U.S. EIA (2020) Natural Gas Annual 2019. Energy Information Administration, U.S. Department of Energy.
28 Washington, D.C. September 2020.
29 U.S. EIA (2017b) Monthly Energy Review, December 2017, Energy Information Administration, U.S. Department of
30 Energy, Washington, D.C. DOE/EIA-0035(2015/12).
31 U.S. EIA (1992) Coal and lignite production. EIA State Energy Data Report 1992, Energy Information Administration,
32 U.S. Department of Energy, Washington, D.C.
33 United States Environmental Protection Agency (EPA) (2010) Carbon Content Coefficients Developed for EPA's
34 Mandatory Reporting Rule. Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental
35 Protection Agency, Washington, D.C.
36 U.S. EPA (2022). Greenhouse Gas Reporting Program. Dataset as of August 13, 2022. Available online at:
37 https://ghgdata.epa.gov/ghgp/.
38 United States Geological Survey (USGS) (2022) 2022 Mineral Commodities Summaries: Iron and Steel. U.S.
39 Geological Survey, Reston, VA. January 2022.
40 USGS (2021) 2021 Mineral Commodities Summaries: Iron and Steel. U.S. Geological Survey, Reston, VA. January
41 2021.
42 USGS (2020) 2020 USGS Minerals Yearbook - Iron and Steel Scrap (unreleased tables). U.S. Geological Survey,
43 Reston, VA.
Waste 41
-------
1 USGS (2019) 2019 USGS Minerals Yearbook - Iron and Steel Scrap (tables-only release). U.S. Geological Survey,
2 Reston, VA.
3 USGS (2018) 2018 USGS Minerals Yearbook - Iron and Steel Scrap (tables-only release). U.S. Geological Survey,
4 Reston, VA.
5 USGS (2017) 2017 USGS Minerals Yearbook - Iron and Steel. U.S. Geological Survey, Reston, VA.
6 USGS (1991 through 2017) USGS Minerals Yearbook - Iron and Steel Scrap. U.S. Geological Survey, Reston, VA.
7 Ferroalloy Pre tion
8 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
9 Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F., D. Qin, G.-K. Plattner, M.
10 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
11 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
12 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
13 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
14 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom
15 996 pp.
16 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
17 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
18 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
19 Onder, H., and E.A. Bagdoyan (1993) Everything You've Always Wanted to Know about Petroleum Coke. Allis
20 Mineral Systems.
21 United States Geological Survey (USGS) (2022a) 2022 Mineral Commodity Summaries: Silicon. U.S. Geological
22 Survey, Reston, VA. January 2022.
23 USGS (2022b) 2019 Minerals Yearbook: Ferroalloys (tables-only release). U.S. Geological Survey, Reston, VA. June
24 2022.
25 USGS (2013a) 2013 Minerals Yearbook: Chromium. U.S. Geological Survey, Reston, VA. March 2016.
26 USGS (1996 through 2022) Minerals Yearbook: Silicon. U.S. Geological Survey, Reston, VA.
27 Aluminum Production
28 EPA (2022) Greenhouse Gas Reporting Program (GHGRP). Envirofacts, Subpart: F Aluminum Production. Available
29 online at:
30 https://enviro.epa.gov/enviro/ef metadata html.ef metadata table?p table name=F SUBPART LEVEL INFORM
31 ATlON&p topic=GHG.
32 EPA (2015) Greenhouse Gas Reporting Program Report Verification. Available online at
33 https://www.epa.gov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
34 IPCC (2019) 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
35 Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change. [Calvo Buendia, E.,
36 Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M., Ngarize, S., Osako, A., Pyrozhenko, Y., Shermanau, P. and
37 Federici, S. (eds.)]. Hayama, Kanagawa, Japan.
38 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
39 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
40 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
42 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
2 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
3 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
4 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
5 USGS (2021) 2020 Mineral Commodity Summaries: Aluminum. U.S. Geological Survey, Reston, VA.
6 USGS (2020) Mineral Industry Surveys: Aluminum in December 2020. U.S. Geological Survey, Reston VA. December
7 2020
8 USGS (2020) 2019 Mineral Commodity Summaries: Aluminum. U.S. Geological Survey, Reston, VA.
9 USGS (2021) 2019 Mineral Commodity Summaries: Aluminum. U.S. Geological Survey, Reston, VA.
10 USGS (2022) Mineral Commodity Summaries 2022. U.S. Geological Survey, Reston VA. January 2022USGS (2019)
11 2017Mineral Yearbook: Aluminum. U.S. Geological Survey, Reston, VA.USGS (2007) 2006 Mineral Yearbook:
12 Aluminum. U.S. Geological Survey, Reston, VA.USGS (1995, 1998, 2000, 2001, 2002) Minerals Yearbook: Aluminum
13 Annual Report. U.S. Geological Survey, Reston, VA.
14 Magnesium Production and Processing
15 ARB (2015) "Magnesium casters successfully retool for a cleaner future." California Air Resources Board News
16 Release. Release # 15-07. February 5, 2015. Accessed October 2017. Available online at:
17 https://www.arb.ca.gov/newsrel/newsrelease.php?id=704.
18 Bartos S., C. Laush, J. Scharfenberg, and R. Kantamaneni (2007) "Reducing greenhouse gas emissions from
19 magnesium die casting." Journal of Cleaner Production, 15: 979-987, March.
20 EPA (2020) Envirofacts. Greenhouse Gas Reporting Program (GHGRP), Subpart T: Magnesium Production and
21 Processing. Available online at: https://www.epa.gov/enviro/greenhouse-gas-customized-search. Accessed on
22 October 2020.
23 Gjestland, H. and D. Magers (1996) "Practical Usage of Sulphur [Sulfur] Hexafluoride for Melt Protection in the
24 Magnesium Die Casting Industry." #13,1996 Annual Conference Proceedings, International Magnesium
25 Association. Ube City, Japan.
26 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
27 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
28 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
29 Kramer Deborah A. (2000) "Magnesium" U.S. Geological Survey Minerals Yearbook - 2000.
30 RAND (2002) RAND Environmental Science and Policy Center, "Production and Distribution of SF6 by End-Use
31 Applications" Katie D. Smythe. International Conference on SF6 and the Environment: Emission Reduction
32 Strategies. San Diego, CA. November 21-22, 2002.
33 USGS (1995 through 2022) Minerals Yearbook: Magnesium Annual Report. U.S. Geological Survey, Reston, VA.
34 Available online at: http://minerals.usgs.gOv/minerals/pubs/commoditv/magnesium/index.html#mis.
35 USGS (2010b) Mineral Commodity Summaries: Magnesium Metal. U.S. Geological Survey, Reston, VA. Available
36 online at: http://minerals.usgs.gov/minerals/pubs/commodity/magnesium/mcs-2010-mgmet.pdf.
37 USGS (2005b) Personal Communication between Deborah Kramer of the USGS and Jeremy Scharfenberg of ICF
38 Consulting.
Waste 43
-------
1
Lead Production
2 Battery Industry (2021). Clarios closing battery recycling center in Florence, South Carolina. Available online at:
3 https://batteryindustry.tech/clarios-closing-battery-recvcling-center-in-florence-south-carolina/. Accessed on
4 September 19, 2022.
5 Dutrizac, J.E., V. Ramachandran, and J.A. Gonzalez (2000) Lead-Zinc 2000. The Minerals, Metals, and Materials
6 Society.
7 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
8 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
9 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
10 Morris, D., F.R. Steward, and P. Evans (1983) Energy Efficiency of a Lead Smelter. Energy 8(5):337-349.
11 Sjardin, M. (2003) CO2 Emission Factors for Non-Energy Use in the Non-Ferrous Metal, Ferroalloys and Inorganics
12 Industry. Copernicus Institute. Utrecht, the Netherlands.
13 Ullman (1997) Ullman's Encyclopedia of Industrial Chemistry: Fifth Edition. Volume A5. John Wiley and Sons.
14 United States Geological Survey (USGS) (2022) 2021 Mineral Commodity Summary, Lead. U.S. Geological Survey,
15 Reston, VA. January 2022.
16 USGS (2022a) 2019 Minerals Yearbook, Lead - Advance Data Release. U.S. Geological Survey, Reston, VA. October
17 2022.
18 USGS (2022b) 2021 Mineral Commodity Summary, Lead. U.S. Geological Survey, Reston, VA. January 2022.
19 USGS (2021a) 2017 Minerals Yearbook, Lead - Advance Release. U.S. Geological Survey, Reston, VA. July 2021.
20 USGS (2021b) 2020 Mineral Commodity Summary, Lead. U.S. Geological Survey, Reston, VA. February 2021.
21
USGS
(2020)
2019 Mineral
Commodity Summary,
Lead.
U.S.
Geological
Survey,
Reston,
VA.
February 2020.
22
USGS
(2019)
2018 Mineral
Commodity Summary,
Lead.
U.S.
Geological
Survey,
Reston,
VA.
February 2019.
23
USGS
(2018)
2017 Mineral
Commodity Summary,
Lead.
U.S.
Geological
Survey,
Reston,
VA.
January 2018.
24
USGS
(2017)
2016 Mineral
Commodity Summary,
Lead.
U.S.
Geological
Survey,
Reston,
VA.
January 2017.
25
USGS
(2016)
2015 Mineral
Commodity Summary,
Lead.
U.S.
Geological
Survey,
Reston,
VA.
January 2016.
26
USGS
(2015)
2014 Mineral
Commodity Summary,
Lead.
U.S.
Geological
Survey,
Reston,
VA.
January 2015.
27
USGS
(2014)
2013 Mineral
Commodity Summary,
Lead.
U.S.
Geological
Survey,
Reston,
VA.
February 2014.
28 USGS (1995 through 2013) Minerals Yearbook: Lead Annual Report. U.S. Geological Survey, Reston, VA.
29 Zinc Production
30 American Zinc Recycling (AZR) (2021) Summary of Company History. Available online at https://azr.com/our-
31 history/. Accessed on March 16, 2021.
32 AZR (2020) Personal communication. Erica Livingston, American Zinc Recycling and Amanda Chiu, U.S.
33 Environmental Protection Agency. October 29, 2020.
34 American Zinc Products (AZP) (2021) American Zinc Products Marks First Anniversary of Zinc Production. Available
35 online at https://americanzincproducts.com/american-zinc-products-marks-first-anniversary-of-zinc-production/.
36 Accessed on March 1, 2022.
44 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Befesa (2022) Personal communication. Eric Hunsberger, Befesa Zinc US Inc. and Amanda Chiu, U.S. Environmental
2 Protection Agency. November 8, 2022.
3 Horsehead Corp. (2016) Form 10-K, Annual Report for the Fiscal Year Ended December 31, 2015. Available online
4 at: https://www.sec.gov/Archives/edgar/data/1385544/00Q119312516725704/d236839dl0k.htm. Submitted on
5 January 25, 2017.
6 Horsehead Corp. (2015) Form 10-K, Annual Report for the Fiscal Year Ended December 31, 2014. Available online
7 at: http://www.sec.gov/Archives/edgar/data/1385544/0001385544150000Q5/zinc-2014123110k.htm. Submitted
8 on March 2, 2015.
9 Horsehead Corp. (2014) Form 10-K, Annual Report for the Fiscal Year Ended December 31, 2013. Available online
10 at: http://www.sec.gov/Archives/edgar/data/1385544/000138554414000Q03/zinc-2013123110k.htm. Submitted
11 on March 13, 2014.
12 Horsehead Corp. (2013) Form 10-K, Annual Report for the Fiscal Year Ended December 31, 2012. Available online
13 at: http://www.sec.gov/Archives/edgar/data/1385544/000119312513110431/0001193125-13-110431-index.htm.
14 Submitted March 18, 2013.
15 Horsehead Corp. (2012a) Form 10-K, Annual Report for the Fiscal Year Ended December 31, 2011. Available online
16 at: http://www.sec.gov/Archives/edgar/data/1385544/00Q119312512107345/d293011dl0k.htm. Submitted on
17 March 9, 2012.
18 Horsehead Corp. (2012b) Horsehead's New Zinc Plant and its Impact on the Zinc Oxide Business. February 22, 2012.
19 Available online at: http://www.horsehead.net/downloadAttachmentNDO.php?ID=118. Accessed on September
20 10,2015.
21 Horsehead Corp. (2011) 10-K Annual Report for the Fiscal Year Ended December 31, 2010. Available online at:
22 http://google.brand.edgar-online.com/default.aspx?sym=zinc. Submitted on March 16, 2011.
23 Horsehead Corp. (2010a) 10-K Annual Report for the Fiscal Year Ended December 31, 2009. Available online at:
24 http://google.brand.edgar-online.com/default.aspx?sym=zinc. Submitted on March 16, 2010.
25 Horsehead Corp. (2010b) Horsehead Holding Corp. Provides Update on Operations at its Monaco, PA Plant. July 28,
26 2010. Available online at: http://www.horsehead.net/pressreleases.php?showall=no&news=&lD=65.
27 Horsehead Corp (2009) 10-K Annual Report for the Fiscal Year Ended December 31, 2008. Available online at:
28 https://www.sec.gov/Archives/edgar/data/1385544/000095015209002674/l35087aelQvk.htm. Submitted on
29 March 16, 2009.
30 Horsehead Corp (2008) 10-K Annual Report for the Fiscal Year Ended December 31, 2007. Available online at:
31 http://google.brand.edgar-online.com/default.aspx?sym=zinc. Submitted on March 31, 2008.
32 Horsehead Corp (2007) Registration Statement (General Form) S-l. Available online at http://google.brand.edgar-
33 online.com/default.aspx?sym=zinc. Submitted on April 13, 2007.
34 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
35 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
36 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
37 Nyrstar (2017) 2016 Clarksville Fact Sheet. Available online at:
38 http://www.nyrstar.eom/~/media/Files/N/Nyrstar/operations/melting/fact-sheet-clarksville-en.pdf. Accessed on
39 September 27, 2017.
40 PIZO (2012) Available online at http://pizotech.com/index.html. Accessed on October 10, 2012.
41 PIZO (2021) Personal communication. Thomas Rheaume, Arkansas Department of Energy and Environment and
42 Amanda Chiu, U.S. Environmental Protection Agency. February 16, 2021.
Waste 45
-------
1 Recycling Today (2020) "AZR to restart for zinc recycling plant in North Carolina." March 6, 2020.
2 https://www.recvclingtodav.com/article/american-zinc-recvcling-restarting-north-carolina-plant-2020/. Accessed
3 October 10, 2020.
4 Recycling Today (2017) "Horsehead announces corporate name change to American Zinc Recycling." May 3, 2017.
5 https://www.recyclingtodav.com/article/horsehead-changes-name-american-zinc-recycling/. Accessed September
6 19,2022.
7 Steel Dust Recycling (SDR) (2022) Personal communication. Jeremy Whitten, Steel Dust Recycling LLC and Amanda
8 Chiu, U.S. Environmental Protection Agency. October 10, 2022.
9 SDR (2021) Personal communication. Jeremy Whitten, Steel Dust Recycling LLC and Amanda Chiu, U.S.
10 Environmental Protection Agency. January 8, 2021.
11 SDR (2018) Personal communication. Jeremy Whitten, Steel Dust Recycling LLC and John Steller, U.S.
12 Environmental Protection Agency. October 25, 2018.
13 SDR (2017) Personal communication. Jeremy Whitten, Steel Dust Recycling LLC and John Steller, U.S.
14 Environmental Protection Agency. January 26, 2017.
15 SDR (2015) Personal communication. Jeremy Whitten, Steel Dust Recycling LLC and Gopi Manne, Eastern Research
16 Group, Inc. September 22, 2015.
17 SDR (2014) Personal communication. Art Rowland, Steel Dust Recycling LLC and Gopi Manne, Eastern Research
18 Group, Inc. December 9, 2014.
19 SDR (2013) Available online at http://steeldust.com/home.htm. Accessed on October 29, 2013.
20 SDR (2012) Personal communication. Art Rowland, Steel Dust Recycling LLC and Gopi Manne, Eastern Research
21 Group, Inc. October 5, 2012.
22 Sjardin (2003) CO2 Emission Factors for Non-Energy Use in the Non-Ferrous Metal, Ferroalloys and Inorganics
23 Industry. Copernicus Institute. Utrecht, the Netherlands.
24 United States Environmental Protection Agency (EPA) (1992) "Applications Analysis Report: Horsehead Resource
25 Development Company Inc., Flame Reactor Technology" EPA/540/A5-91/005. May 1992.
26 United States Geological Survey (USGS) (2022) 2022 Mineral Commodity Summary: Zinc. U.S. Geological Survey,
27 Reston, VA. January 2022. Available online at: https://pubs.usgs.gov/periodicals/mcs2022/mcs2022-zinc.pdf.
28 USGS (2021) 2021 Mineral Commodity Summary: Zinc. U.S. Geological Survey, Reston, VA. January 2021.
29 USGS (2020) 2020 Mineral Commodity Summary: Zinc. U.S. Geological Survey, Reston, VA. January 2020.
30 USGS (2019) 2019 Mineral Commodity Summary: Zinc. U.S. Geological Survey, Reston, VA. January 2019.
31 USGS (2018) 2018 Mineral Commodity Summary: Zinc. U.S. Geological Survey, Reston, VA. January 2018.
32 USGS (2017) 2017 Mineral Commodity Summary: Zinc. U.S. Geological Survey, Reston, VA. January 2017.
33 USGS (2016) 2016 Mineral Commodity Summary: Zinc. U.S. Geological Survey, Reston, VA. January 2016.
34 USGS (2015) 2015 Mineral Commodity Summary: Zinc. U.S. Geological Survey, Reston, VA. January 2015.
35 USGS (1995 through 2014) Minerals Yearbook: Zinc Annual Report. U.S. Geological Survey, Reston, VA.
36 Viklund-White (2000) The use ofLCAforthe environmental evaluation of the recycling of galvanized steel. ISIJ
37 International, Vol. 40. No. 3, pp 292-299.
46 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Electronics Industry
2 Burton, C.S., and R. Beizaie (2001) "EPA's PFC Emissions Model (PEVM) v. 2.14: Description and Documentation"
3 prepared for Office of Global Programs, U. S. Environmental Protection Agency, Washington, DC. November 2001.
4 Citigroup Smith Barney (2005) Global Supply/Demand Model for Semiconductors. March 2005.
5 DisplaySearch (2010) DisplaySearch Q4'09 Quarterly FPD Supply/Demand and Capital Spending Report.
6 DisplaySearch, LLC.
7 Doering, R. and Nishi, Y (2000) "Handbook of Semiconductor Manufacturing Technology", Marcel Dekker, New
8 York, USA, 2000.
9 EPA (2006) Uses and Emissions of Liquid PFC Heat Transfer Fluids from the Electronics Sector. U.S. Environmental
10 Protection Agency, Washington, DC. EPA-430-R-06-901.
11 EPA (2010) Technical Support Document for Process Emissions from Electronics Manufacture (e.g., Micro-Electro-
12 Mechanical Systems, Liquid Crystal Displays, Photovoltaics, and Semiconductors). U.S. Environmental Protection
13 Agency, Washington, DC.
14 EPA Greenhouse Gas Reporting Program (GHGRP) Envirofacts. Subpart I: Electronics Manufacture. Available online
15 at: https://enviro.epa.gov/facts/ghg/search.html.
16 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
17 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
18 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
19 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
20 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
21 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
22 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
23 IPCC (2019) 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
24 Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change. Calvo Buendia, E.,
25 Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M., Ngarize, S., Osako, A., Pyrozhenko, Y., Shermanau, P. and
26 Federici, S. (eds). Published: IPCC, Switzerland.
27 ITRS (2007, 2008, 2011, 2013) International Technology Roadmap for Semiconductors: 2006 Update, January 2007;
28 International Technology Roadmap for Semiconductors: 2007 Edition, January 2008; International Technology
29 Roadmap for Semiconductors: 2011, January 2012; Update, International Technology Roadmap for
30 Semiconductors: 2013 Edition, Available online at: https://www.semiconductors.org/resources/2007-intemational-
31 technology-roadmap-for-semiconductors-itrs/. These and earlier editions and updates are available online at:
32 https://www.semiconductors.org/resources/7fwp resource types=utilization-reports&fwp paged=2. Information
33 about the number of interconnect layers for years 1990-2010 is contained in Burton and Beizaie, 2001. PEVM is
34 updated using new editions and updates of the ITRS, which are published annually. SEMI - Semiconductor
35 Equipment and Materials Industry (2017) World Fab Forecast, August 2018 Edition.
36 Platzer, Michaela D. (2015) U.S. Solar Photovoltaic Manufacturing: Industry Trends, Global Competition, Federal
37 Support. Congressional Research Service. January 27, 2015. https://fas.org/sgp/crs/misc/R42509.pdf.
38 SEMI - Semiconductor Equipment and Materials Industry (2021) World Fab Forecast, June 2021 Edition.
39 SEMI - Semiconductor Equipment and Materials Industry (2018) World Fab Forecast, June 2018 Edition.
40 SEMI - Semiconductor Equipment and Materials Industry (2016) World Fab Forecast, May 2017 Edition.
41 SEMI - Semiconductor Equipment and Materials Industry (2013) World Fab Forecast, May 2013 Edition.
42 SEMI - Semiconductor Equipment and Materials Industry (2012) World Fab Forecast, August 2012 Edition.
Waste 47
-------
1 Semiconductor Industry Association (SIA) (2009-2011) STATS: SICAS Capacity and Utilization Rates Q1-Q4 2008, Ql-
2 Q4 2009, Q1-Q4 2010. Available online at:
3 http://www.semiconductors.org/industry statistics/semiconductor capacity utilization sicas reports/.
4 United States Census Bureau (USCB) (2011, 2012, 2014, 2015, 2016, 2017, 2018, 2019, 2020, 2021) Historical Data:
5 Quarterly Survey of Plant Capacity Utilization. Available online at: https://www.census.gov/programs-
6 surveys/qpc.html.
7 VLSI Research, Inc. (2012) Worldwide Silicon Demand. August 2012.
8 Substitution of Ozone Depleting Substances
9 Carrier (2023) New Carrier AquaSnap 30RC Air Cooled Chiller Helps Maximize Building Space while Delivering
10 Efficiency and Sustainability, January 10, 2023. Available online at:
11 https://www.carrier.com/commercial/en/us/news/news-article/new-carrier-aquasnap-30rc-air-cooled-chiller-
12 helps-maximize-building-space-while-delivering-efficiencv-and-sustainability.html
13 EPA (2022c) Summary of Updates to the Unitary Air-conditioning End-uses in the Vintaging Model. Prepared for
14 U.S. EPA's Stratospheric Protection Division by ICF under EPA Contract Number 68HERH19D0029.
15 EPA (2022d) Summary of Updates to the Road Transport and Modern Rail Car End-uses in the Vintaging Model.
16 Prepared for U.S. EPA's Stratospheric Protection Division by ICF under EPA Contract Number 68HERH19D0029.
17 EPA (2022e) Review of HCFC-22 Dry-shipped Condensing Units in the Residential Unitary Air Conditioning End-Use
18 in the Vintaging Model. Prepared for U.S. EPA's Stratospheric Protection Division by ICF under EPA Contract
19 Number 68HERH19D0029.
20 EPA (2018) EPA's Vintaging Model of ODS Substitutes: A Summary of the 2017 Peer Review. Office of Air and
21 Radiation. Document Number EPA-400-F-18-001. Available online at:
22 https://www.epa.gov/sites/production/files/2018-Q9/documents/epas-vintaging-model-of-ods-substitutes-peer-
23 review-factsheet.pdf.
24 Hu et al. (2022) U.S. non-CC>2 greenhouse gas (GHG) emissions for 2007 - 2020 derived from atmospheric
25 observations. American Geophysical Union. December 2022. Abstract available online at:
26 https://agu.confex.com/agu/fm22/meetingapp.cgi/Paper/1112748.
27 Hu et al. (2017) Considerable Contribution of the Montreal Protocol to Declining Greenhouse Gas Emissions from
28 the United States, Geophys. Res. Lett., 44, 8075-8083, doi:10.1002/2017GL074388, August 14, 2017.
29 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
30 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
31 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
32 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
33 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
34 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
35 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom
36 996 pp.
37 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
38 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
39 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
48 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
i Electrical Transmission and Distribution
2 EPA (2022) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020. EPA 430-R-22-003. Available online
3 at: https://www.epa.gov/ghgemissions/inventory-us-greenhouse-gas-emissions-and-sinks-1990-2020.
4 Harnisch and Eisenhauer, "Natural CF4 and SF6 on Earth," GEOPHYSICAL RESEARCH LETTERS, VOL. 25, N0.13,
5 PAGES 2401-2404, JULY 1,1998. https://agupubs.onlinelibrarv.wilev.com/doi/pdf/10.1029/98GL01779
6 HIFLD (2019) Federal Energy Regulatory Commission. Homeland Infrastructure Foundation-Level Data (HIFLD).
7 2019. Accessed March 2021. Available online at: https://hifld-geoplatform.opendata.arcgis.com/datasets/electric-
8 power-transmission-lines.
9 HIFLD (2020) Federal Energy Regulatory Commission. Homeland Infrastructure Foundation-Level Data (HIFLD).
10 2020. Accessed October 2021. Available online at: https://hifld-
11 geoplatform.opendata.arcgis.com/datasets/electric-power-transmission-lines/explore?showTable=true.
12 HIFLD (2021) Federal Energy Regulatory Commission. Homeland Infrastructure Foundation-Level Data (HIFLD).
13 2021. Accessed September 2022. Available online at: https://hifld-
14 geoplatform.opendata.arcgis.com/datasets/electric-power-transmission-lines.
15 Hu, L, Ottinger, D., Bogle, S., Montzka, S., DeCola, P., Dlugokencky, E., Andrews, A., Thoning, K., Sweeney, C.,
16 Dutton, G., Aepli, L., and Crotwell, A. (2022). "Declining, seasonal-varying emissions of sulfur hexafluoride from the
17 United States point to a new mitigation opportunity." EGUsphere [preprint]. Available online at:
18 https://doi.org/10.5194/egusphere-2022-862.
19 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
20 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
21 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
22 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
23 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
24 Assessment Report of the Intergovernmental Panel on Climate Change. S. Solomon, D. Qin, M. Manning, Z. Chen,
25 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.). Cambridge University Press. Cambridge, United Kingdom
26 996 pp.
27 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
28 Inventories Programme, The Intergovernmental Panel on Climate Change. H.S. Eggleston, L. Buendia, K. Miwa, T.
29 Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
30 IPCC (1996) Climate Change 1995: The Science of Climate Change. Intergovernmental Panel on Climate Change, J.T.
31 Houghton, L.G. Meira Filho, B.A. Callander, N. Harris, A. Kattenberg, and K. Maskell (eds.). Cambridge University
32 Press. Cambridge, United Kingdom.
33 Levin et al. (2010) "The Global SF6 Source Inferred from Long-term High Precision Atmospheric Measurements and
34 its Comparison with Emission Inventories." Atmospheric Chemistry and Physics, 10: 2655-2662.
35 Middleton, B. (2000) Cold Weather Applications of Gas Mixture (SF6/N2, SF6/CF4) Circuit Breakers: A User Utility's
36 Perspective [Conference Presentation], The US Environmental Protection Agency's Conference on SF6 and the
37 Environment: Emission Reduction Strategies, San Diego, CA, United States. Available online at:
38 https://www.epa.gov/sites/default/files/2016-02/documents/conf00 middleton.pdf
39 O'Connell, P., F. Heil, J. Henriot, G. Mauthe, H. Morrison, L. Neimeyer, M. Pittroff, R. Probst, J.P. Tailebois (2002)
40 SFs in the Electric Industry, Status 2000, CIGRE. February 2002.
41 Ottinger D, Averyt, M. & Harris, D. (2014). Trends in emissions offluorinated GHGs reported under the Greenhouse
42 Gas Reporting Program: Patterns and potential causes. Submitted to the Seventh International Symposium on
43 Non-C02 Greenhouse Gases (NCGG-7), Amsterdam, Netherlands.
Waste 49
-------
1 RAND (2004) 'Trends in SF6 Sales and End-Use Applications: 1961-2003," Katie D. Smythe. International Conference
2 on SFe and the Environment: Emission Reduction Strategies. RAND Environmental Science and Policy Center,
3 Scottsdale, AZ. December 1-3, 2004.
4 UDI (2017) 2017 UDI Directory of Electric Power Producers and Distributors, 125th Edition, Platts.
5 UDI (2013) 2013 UDI Directory of Electric Power Producers and Distributors, 121st Edition, Platts.
6 UDI (2010) 2010 UDI Directory of Electric Power Producers and Distributors, 118th Edition, Platts.
7 UDI (2007) 2007 UDI Directory of Electric Power Producers and Distributors, 115th Edition, Platts.
8 UDI (2004) 2004 UDI Directory of Electric Power Producers and Distributors, 112th Edition, Platts.
9 UDI (2001) 2001 UDI Directory of Electric Power Producers and Distributors, 109th Edition, Platts.
10 UNFCCC (2014) Report of the Conference of the Parties on its nineteenth session, held in Warsaw from 11 to 23
11 November 2013. United Nations Framework Convention on Climate Change, Warsaw. (FCCC/CP/2013/10/Add.3).
12 January 31, 2014. Available online at: http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
13 Nitrous Oxide from Product Use
14 CGA (2003) "CGA Nitrous Oxide Abuse Hotline: CGA/NWSA Nitrous Oxide Fact Sheet." Compressed Gas
15 Association. November 3, 2003.
16 CGA (2002) "CGA/NWSA Nitrous Oxide Fact Sheet." Compressed Gas Association. March 25, 2002.
17 Heydorn, B. (1997) "Nitrous Oxide—North America." Chemical Economics Handbook, SRI Consulting. May 1997.
18 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
19 Assessment Report of the Intergovernmental Panel on Climate Change. [Stocker, T.F., D. Qin, G.-K. Plattner, M.
20 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
21 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
22 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
23 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
24 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom
25 996 pp.
26 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
27 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L Buendia, K. Miwa, T.
28 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
29 Ottinger (2021) Personal communication. Deborah Ottinger, U.S. Environmental Protection Agency and Amanda
30 Chiu, U.S. Environmental Protection Agency. January 7, 2021.
31 Tupman, M. (2002) Personal communication. Martin Tupman, Airgas Nitrous Oxide and Laxmi Palreddy, ICF
32 International. July 3, 2002.
33 Industrial Processes and Product Use Sources of Precursor
34 Greenhouse Gases
35 EPA (2022) "Crosswalk of Precursor Gas Categories." U.S. Environmental Protection Agency. April 6, 2022.
36 EPA (2021a) "Criteria pollutants National Tier 1 for 1970 - 2021." National Emissions Inventory (NEI) Air Pollutant
37 Emissions Trends Data. Office of Air Quality Planning and Standards, March 2021. Available online at:
38 https://www.epa.gov/air-emissions-inventories/air-pollutant-emissions-trends-data.
50 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 EPA (2021b) "2017 National Emissions Inventory (NEI) Technical Support Document (TSD)." Office of Air Quality
2 Planning and Standards, April 2021. Available online at: https://www.epa.gov/air-emissions-inventories/2017-
3 national-emissions-inventory-nei-technical-support-document-tsd.
4 EPA (1997) Compilation of Air Pollutant Emission Factors, AP-42. Office of Air Quality Planning and Standards, U.S.
5 Environmental Protection Agency. Research Triangle Park, NC. October 1997.
e Agriculture
7 Enteric Fermentation
8 Archibeque, S. (2011) Personal Communication. Shawn Archibeque, Colorado State University, Fort Collins,
9 Colorado and staff at ICF International.
10 Crutzen, P.J., I. Aselmann, and W. Seiler (1986) Methane Production by Domestic Animals, Wild Ruminants, Other
11 Herbivores, Fauna, and Humans. Tellus, 38B:271-284.
12 Donovan, K. (1999) Personal Communication. Kacey Donovan, University of California at Davis and staff at ICF
13 International.
14 Doren, P.E., J. F. Baker, C. R. Long and T. C. Cartwright (1989) Estimating Parameters of Growth Curves of Bulls, J
15 Animal Science 67:1432-1445.
16 Enns, M. (2008) Personal Communication. Dr. Mark Enns, Colorado State University and staff at ICF International.
17 EPA (2002) Quality Assurance/Quality Control and Uncertainty Management Plan for the U.S. Greenhouse Gas
18 Inventory: Procedures Manual for Quality Assurance/Quality Control and Uncertainty Analysis, U.S. Greenhouse
19 Gas Inventory Program, U.S. Environmental Protection Agency, Office of Atmospheric Programs, EPA430-R-02-
20 007B, June 2002.
21 ERG (2021) Updated Other Animal Population Distribution Methodology. ERG, Lexington, MA.
22 ERG (2016) Development of Methane Conversion Rate Scaling Factor and Diet-Related Inputs to the Cattle Enteric
23 Fermentation Model for Dairy Cows, Dairy Heifers, and Feedlot Animals. ERG, Lexington, MA. December 2016.
24 Galyean and Gleghorn (2001) Summary of the 2000 Texas Tech University Consulting Nutritionist Survey. Texas
25 Tech University. Available online at http://www.depts.ttu.edu/afs/burnett center/progress reports/bcl2.pdf.
26 June 2009.
27 Holstein Association (2010) History of the Holstein Breed (website). Available online at:
28 http://www.holsteinusa.com/holstein breed/breed history, html. Accessed September 2010.
29 ICF (2006) Cattle Enteric Fermentation Model: Model Documentation. Prepared by ICF International for the
30 Environmental Protection Agency. June 2006.
31 ICF (2003) Uncertainty Analysis of 2001 Inventory Estimates of Methane Emissions from Livestock Enteric
32 Fermentation in the U.S. Memorandum from ICF International to the Environmental Protection Agency. May 2003.
33 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
34 Assessment Report of the Intergovernmental Panel on Climate Change. S. Solomon, D. Qin, M. Manning, Z. Chen,
35 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.). Cambridge University Press. Cambridge, United Kingdom
36 996 pp.
37 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
38 Inventories Programme, The Intergovernmental Panel on Climate Change. H.S. Eggleston, L. Buendia, K. Miwa, T.
39 Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
Waste 51
-------
1 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
2 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
3 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
4 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
5 IPCC (2019) 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The
6 Intergovernmental Panel on Climate Change. Calvo Buendia, E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M.,
7 Ngarize, S., Osako, A., Pyrozhenko, Y., Shermanau, P. and Federici, S. (eds). Hayama, Kanagawa, Japan.Johnson, D.
8 (2002) Personal Communication. Don Johnson, Colorado State University, Fort Collins, and ICF International.
9 Johnson, D. (1999) Personal Communication. Don Johnson, Colorado State University, Fort Collins, and David
10 Conneely, ICF International.
11 Kebreab E., K. A. Johnson, S. L. Archibeque, D. Pape, and T. Wirth (2008) Model for estimating enteric methane
12 emissions from United States dairy and feedlot cattle. J. Anim. Sci. 86: 2738-2748.
13 Lippke, H., T. D. Forbes, and W. C. Ellis. (2000) Effect of supplements on growth and forage intake by stocker steers
14 grazing wheat pasture. J. Anim. Sci. 78:1625-1635.
15 National Bison Association (1999) Total Bison Population—1999. Report provided during personal email
16 communication with Dave Carter, Executive Director, National Bison Association, July 19, 2011.
17 Pinchak, W.E., D. R. Tolleson, M. McCloy, L. J. Hunt, R. J. Gill, R. J. Ansley, and S. J. Bevers (2004) Morbidity effects
18 on productivity and profitability of stocker cattle grazing in the southern plains. J. Anim. Sci. 82:2773-2779.
19 Platter, W. J., J. D. Tatum, K. E. Belk, J. A. Scanga, and G. C. Smith (2003) Effects of repetitive use of hormonal
20 implants on beef carcass quality, tenderness, and consumer ratings of beef palatability. J. Anim. Sci. 81:984-996.
21 Preston, R.L (2010) What's The Feed Composition Value of That Cattle Feed? Beef Magazine, March 1, 2010.
22 Available at: http://beefmagazine.com/nutrition/feed-composition-tables/feed-composition-value-cattle-0301.
23 Skogerboe, T. L, L Thompson, J. M. Cunningham, A. C. Brake, V. K. Karle (2000) The effectiveness of a single dose
24 of doramectin pour-on in the control of gastrointestinal nematodes in yearling stocker cattle. Vet. Parasitology
25 87:173-181.
26 Soliva, C.R. (2006) Report to the attention of IPCC about the data set and calculation method used to estimate
27 methane formation from enteric fermentation of agricultural livestock population and manure management in
28 Swiss agriculture. On behalf of the Federal Office for the Environment (FOEN), Berne, Switzerland.
29 U.S. Department of Agriculture (USDA) (2022) Quick Stats: Agricultural Statistics Database. National Agriculture
30 Statistics Service, U.S. Department of Agriculture. Washington, D.C. Available online at
31 http://quickstats.nass.usda.gov/. Accessed July 2022.
32 U.S. Department of Agriculture (USDA) (2021a) Quick Stats: Agricultural Statistics Database. National Agriculture
33 Statistics Service, U.S. Department of Agriculture. Washington, D.C. Available online at
34 http://quickstats.nass.usda.gov/. Accessed May-June 2021.
35 USDA (2021b) Economic Research Service Dairy Data. Available online at: https://www.ers.usda.gov/data-
36 products/dairy-data/. Accessed May 2021.
37 USDA (2019) 1987,1992,1997, 2002, 2007, 2012, and 2017 Census of Agriculture. National Agriculture Statistics
38 Service, U.S. Department of Agriculture. Washington, D.C. Available online at:
39 https://www.nass.usda.gov/AgCensus/index.php. May 2019.
40 USDA (1996) Beef Cow/Calf Health and Productivity Audit (CHAPA): Forage Analyses from Cow/Calf Herds in 18
41 States. National Agriculture Statistics Service, U.S. Department of Agriculture. Washington, D.C. Available online at
42 http://www.aphis.usda.gov/vs/ceah/cahm. March 1996.
43 USDA:APHIS:VS (2010) Beef 2007-08, Part V: Reference of Beef Cow-calf Management Practices in the United
44 States, 2007-08. USDA-APHIS-VS, CEAH. Fort Collins, CO.
52 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 USDA:APHIS:VS (2002) Reference of 2002 Dairy Management Practices. USDA-APHIS-VS, CEAH. Fort Collins, CO.
2 Available online at http://www.aphis.usda.gov/vs/ceah/cahm.
3 USDA:APHIS:VS (1998) Beef '97, Parts l-IV. USDA-APHIS-VS, CEAH. Fort Collins, CO. Available online at
4 http://www.aphis.usda.gov/animal health/nahms/beefcowcalf/index.shtml#beef97.
5 USDA:APHIS:VS (1996) Reference of 1996 Dairy Management Practices. USDA-APHIS-VS, CEAH. Fort Collins, CO.
6 Available online at http://www.aphis.usda.gov/vs/ceah/cahm.
7 USDA:APHIS:VS (1994) Beef Cow/Calf Health and Productivity Audit. USDA-APHIS-VS, CEAH. Fort Collins, CO.
8 Available online at http://www.aphis.usda.gov/vs/ceah/cahm.
9 USDA:APHIS:VS (1993) Beef Cow/Calf Health and Productivity Audit. USDA-APHIS-VS, CEAH. Fort Collins, CO.
10 August 1993. Available online at http://www.aphis.usda.gov/vs/ceah/cahm.
11 Vasconcelos and Galyean (2007) Nutritional recommendations of feedlot consulting nutritionists: The 2007 Texas
12 Tech University Study. J. Anim. Sci. 85:2772-2781.
is Manure Management
14 ASAE (1998) ASAE Standards 1998, 45th Edition. American Society of Agricultural Engineers. St. Joseph, Ml.
15 Bryant, M.P., V.H. Varel, R.A. Frobish, and H.R. Isaacson (1976) In H.G. Schlegel (ed.)]; Seminar on Microbial Energy
16 Conversion. E. Goltz KG. Gottingen, Germany.
17 Bush, E. (1998) Personal communication with Eric Bush, Centers for Epidemiology and Animal Health, U.S.
18 Department of Agriculture regarding National Animal Health Monitoring System's (NAHMS) Swine '95 Study.
19 EPA (2021) AgSTAR Anaerobic Digester Database. Available online at: https://www.epa.gov/agstar/livestock-
20 anaerobic-digester-database. Accessed September 2021.
21 EPA (2008) Climate Leaders Greenhouse Gas Inventory Protocol Offset Project Methodology for Project Type
22 Managing Manure with Biogas Recovery Systems.
23 EPA (2005) National Emission Inventory—Ammonia Emissions from Animal Agricultural Operations, Revised Draft
24 Report. U.S. Environmental Protection Agency. Washington, D.C. April 22, 2005.
25 EPA (2002a) Development Document for the Final Revisions to the National Pollutant Discharge Elimination System
26 (NPDES) Regulation and the Effluent Guidelines for Concentrated Animal Feeding Operations (CAFOS). U.S.
27 Environmental Protection Agency. EPA-821-R-03-001. December 2002.
28 EPA (2002b) Cost Methodology for the Final Revisions to the National Pollutant Discharge Elimination System
29 Regulation and the Effluent Guidelines for Concentrated Animal Feeding Operations. U.S. Environmental
30 Protection Agency. EPA-821-R-03-004. December 2002.
31 EPA (1992) Global Methane Emissions from Livestock and Poultry Manure, Office of Air and Radiation, U.S.
32 Environmental Protection Agency. February 1992.
33 ERG (2021) Updated Other Animal Population Distribution Methodology. Memorandum to EPA from ERG.
34 ERG (2019) "Incorporation of USDA 2016 ARMS Dairy Data into the Manure Management Greenhouse Gas
35 Inventory." Memorandum to USDA OCE and EPA from ERG, December 2019.
36 ERG (2018) "Incorporation of USDA 2009 ARMS Swine Data into the Manure Management Greenhouse Gas
37 Inventory." Memorandum to USDA OCE and EPA from ERG, November 2018.
38 ERG (2010a) "Typical Animal Mass Values for Inventory Swine Categories." Memorandum to EPA from ERG. July 19,
39 2010.
Waste 53
-------
1 ERG (2010b) Telecon with William Boyd of USDA NRCS and Cortney Itle of ERG Concerning Updated VS and Nex
2 Rates. August 8, 2010.
3 ERG (2010c) "Updating Current Inventory Manure Characteristics new USDA Agricultural Waste Management Field
4 Handbook Values." Memorandum to EPA from ERG. August 13, 2010.
5 ERG (2008) "Methodology for Improving Methane Emissions Estimates and Emission Reductions from Anaerobic
6 Digestion System for the 1990-2007 Greenhouse Gas Inventory for Manure Management." Memorandum to EPA
7 from ERG. August 18, 2008.
8 ERG (2003a) "Methodology for Estimating Uncertainty for Manure Management Greenhouse Gas Inventory."
9 Contract No. GS-10F-0036, Task Order 005. Memorandum to EPA from ERG, Lexington, MA. September 26, 2003.
10 ERG (2003b) "Changes to Beef Calves and Beef Cows Typical Animal Mass in the Manure Management Greenhouse
11 Gas Inventory." Memorandum to EPA from ERG, October 7, 2003.
12 ERG (2001) Summary of development of MDP Factor for methane conversion factor calculations. ERG, Lexington,
13 MA. September 2001.
14 ERG (2000a) Calculations: Percent Distribution of Manure for Waste Management Systems. ERG, Lexington, MA.
15 August 2000.
16 ERG (2000b) Discussion of Methodology for Estimating Animal Waste Characteristics (Summary of Bo Literature
17 Review). ERG, Lexington, MA. June 2000.
18 Groffman, P.M., R. Brumme, K. Butterbach-Bahl, K.E. Dobbie, A.R. Mosier, D. Ojima, H. Papen, W.J. Parton, K.A.
19 Smith, and C. Wagner-Riddle (2000) "Evaluating annual nitrous oxide fluxes at the ecosystem scale." Global
20 Biogeochemical Cycles, 14(4): 1061-1070.
21 Hashimoto, A.G. (1984) "Methane from Swine Manure: Effect of Temperature and Influent Substrate Composition
22 on Kinetic Parameter (k)." Agricultural Wastes, 9:299-308.
23 Hashimoto, A.G., V.H. Varel, and Y.R. Chen (1981) "Ultimate Methane Yield from Beef Cattle Manure; Effect of
24 Temperature, Ration Constituents, Antibiotics and Manure Age." Agricultural Wastes, 3:241-256.
25 Hill, D.T. (1984) "Methane Productivity of the Major Animal Types." Transactions of the ASAE, 27(2):530-540.
26 Hill, D.T. (1982) "Design of Digestion Systems for Maximum Methane Production." Transactions of the ASAE,
27 25(l):226-230.
28 IPCC (2019) 2019 Refinement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
29 Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change. [CalvoBuendia, E.,
30 Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M., Ngarize S., Osako, A., Pyrozhenko, Y., Shermanau, P. and
31 Federici, S. (eds)]. Switzerland.
32 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
33 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
34 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
35 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.IPCC (2006) 2006 IPCC Guidelines for National
36 Greenhouse Gas Inventories. The National Greenhouse Gas Inventories Programme, The Intergovernmental Panel
37 on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa,
38 Japan.
39 Morris, G.R. (1976) Anaerobic Fermentation of Animal Wastes: A Kinetic and Empirical Design Fermentation. M.S.
40 Thesis. Cornell University.
41 National Bison Association (1999) Total Bison Population—1999. Report provided during personal email
42 communication with Dave Carter, Executive Director, National Bison Association July 19, 2011.
54 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Ott, S.L (2000) Dairy '96 Study. Stephen L. Ott, Animal and Plant Health Inspection Service, U.S. Department of
2 Agriculture. June 19, 2000.
3 Robertson, G. P. and P. M. Groffman (2015) Nitrogen transformations. Soil Microbiology, Ecology, and
4 Biochemistry, pages 421-446. Academic Press, Burlington, Massachusetts, USA.
5 Safley, L.M., Jr. (2000) Personal Communication. Deb Bartram, ERG and LM. Safley, President, Agri-Waste
6 Technology. June and October 2000.
7 Sweeten, J. (2000) Personal Communication. John Sweeten, Texas A&M University and Indra Mitra, ERG. June
8 2000.
9 UEP (1999) Voluntary Survey Results—Estimated Percentage Participation/Activity. Caged Layer Environmental
10 Management Practices, Industry data submissions for EPA profile development, United Egg Producers and National
11 Chicken Council. Received from John Thorne, Capitolink. June 2000.
12 USDA (2022) Quick Stats: Agricultural Statistics Database. National Agriculture Statistics Service, U.S. Department
13 of Agriculture. Washington, D.C. Available online at: http://quickstats.nass.usda.gov/.
14 USDA (2021a) Quick Stats: Agricultural Statistics Database. National Agriculture Statistics Service, U.S. Department
15 of Agriculture. Washington, D.C. Available online at: http://quickstats.nass.usda.gov/.
16 USDA (2021b) Chicken and Eggs 2020 Summary. National Agriculture Statistics Service, U.S. Department of
17 Agriculture. Washington, D.C. February 2021. Available online at:
18 http://www.nass.usda.gov/Publications/index.asp.
19 USDA (2021c) Poultry - Production and Value 2020 Summary. National Agriculture Statistics Service, U.S.
20 Department of Agriculture. Washington, D.C. April 2021. Available online at:
21 http://www.nass.usda.gov/Publications/index.asp.
22 USDA (2019b) Chicken and Eggs 2018 Summary. National Agriculture Statistics Service, U.S. Department of
23 Agriculture. Washington, D.C. February 2019. Available online at:
24 http://www.nass.usda.gov/Publications/index.php.
25 USDA (2019b) Poultry - Production and Value 2018 Summary. National Agriculture Statistics Service, U.S.
26 Department of Agriculture. Washington, D.C. April 2019. Available online at:
27 http://www.nass.usda.gov/Publications/index.php.
28 USDA (2019c) Chicken and Eggs 2013-2017 Summary. National Agriculture Statistics Service, U.S. Department of
29 Agriculture. Washington, D.C. June 2019. Available online at: http://www.nass.usda.gov/Publications/index.php.
30 USDA (2019d) 1987,1992,1997, 2002, 2007, 2012, and 2017 Census of Agriculture. National Agriculture Statistics
31 Service, U.S. Department of Agriculture. Washington, D.C. Available online at:
32 https://www.nass.usda.gov/AgCensus/index.php. May 2019.
33 USDA (2018) Poultry - Production and Value 2017 Summary. National Agriculture Statistics Service, U.S.
34 Department of Agriculture. Washington, D.C. April 2018. Available online at:
35 http://www.nass.usda.gov/Publications/index.php.
36 USDA (2017) Poultry - Production and Value 2016 Summary. National Agriculture Statistics Service, U.S.
37 Department of Agriculture. Washington, D.C. April 2017. Available online at:
38 http://www.nass.usda.gov/Publications/index.php.
39 USDA (2016) Poultry - Production and Value 2015 Summary. National Agriculture Statistics Service, U.S.
40 Department of Agriculture. Washington, D.C. April 2016. Available online at:
41 http://www.nass.usda.gov/Publications/index.php.
42 USDA (2015) Poultry - Production and Value 2014 Summary. National Agriculture Statistics Service, U.S.
43 Department of Agriculture. Washington, D.C. April 2015. Available online at:
44 http://www.nass.usda.gov/Publications/index.php.
Waste 55
-------
1 USDA (2014) Poultry - Production and Value 2013 Summary. National Agriculture Statistics Service, U.S.
2 Department of Agriculture. Washington, D.C. April 2014. Available online at:
3 http://www.nass. usda.gov/Publications/index.php.
4 USDA (2013a) Chicken and Eggs 2012 Summary. National Agriculture Statistics Service, U.S. Department of
5 Agriculture. Washington, D.C. February 2013. Available online at:
6 http://www.nass.usda.gov/Publications/index.php.
7 USDA (2013b) Poultry - Production and Value 2012 Summary. National Agriculture Statistics Service, U.S.
8 Department of Agriculture. Washington, D.C. April 2013. Available online at:
9 http://www.nass.usda.gov/Publications/index.php.
10 USDA (2012a) Chicken and Eggs 2011 Summary. National Agriculture Statistics Service, U.S. Department of
11 Agriculture. Washington, D.C. February 2012. Available online at:
12 http://www.nass.usda.gov/Publications/index.php.
13 USDA (2012b) Poultry - Production and Value 2011 Summary. National Agriculture Statistics Service, U.S.
14 Department of Agriculture. Washington, D.C. April 2012. Available online at:
15 http://www.nass.usda.gov/Publications/index.php.
16 USDA (2011a) Chicken and Eggs 2010 Summary. National Agriculture Statistics Service, U.S. Department of
17 Agriculture. Washington, D.C. February 2011. Available online at:
18 http://www.nass.usda.gov/Publications/index.php.
19 USDA (2011b) Poultry - Production and Value 2010 Summary. National Agriculture Statistics Service, U.S.
20 Department of Agriculture. Washington, D.C. April 2011. Available online at:
21 http://www.nass.usda.gov/Publications/index.php.
22 USDA (2010a) Chicken and Eggs 2009 Summary. National Agriculture Statistics Service, U.S. Department of
23 Agriculture. Washington, D.C. February 2010. Available online at:
24 http://www.nass.usda.gov/Publications/index.php.
25 USDA (2010b) Poultry - Production and Value 2009 Summary. National Agriculture Statistics Service, U.S.
26 Department of Agriculture. Washington, D.C. April 2010. Available online at:
27 http://www.nass.usda.gov/Publications/index.php.
28 USDA (2009a) Chicken and Eggs 2008 Summary. National Agriculture Statistics Service, U.S. Department of
29 Agriculture. Washington, D.C. February 2009. Available online at:
30 http://www.nass.usda.gov/Publications/index.php.
31 USDA (2009b) Poultry - Production and Value 2008 Summary. National Agriculture Statistics Service, U.S.
32 Department of Agriculture. Washington, D.C. April 2009. Available online at:
33 http://www.nass.usda.gov/Publications/index.php.
34 USDA (2009c) Chicken and Eggs - Final Estimates 2003-2007. National Agriculture Statistics Service, U.S.
35 Department of Agriculture. Washington, D.C. March 2009. Available online at:
36 https://www.nass.usda.gov/Publications/index.php.
37 USDA (2009d) Poultry Production and Value—Final Estimates 2003-2007. National Agriculture Statistics Service,
38 U.S. Department of Agriculture. Washington, D.C. May 2009. Available online at:
39 https://www.nass.usda.gov/Publications/index.php.
40 USDA (2008) Agricultural Waste Management Field Handbook, National Engineering Handbook (NEH), Part 651.
41 Natural Resources Conservation Service, U.S. Department of Agriculture.
42 USDA (2004a) Chicken and Eggs—Final Estimates 1998-2003. National Agriculture Statistics Service, U.S.
43 Department of Agriculture. Washington, D.C. April 2004. Available online at:
44 https://www.nass.usda.gov/Publications/index.php.
56 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 USDA (2004b) Poultry Production and Value—Final Estimates 1998-2002. National Agriculture Statistics Service,
2 U.S. Department of Agriculture. Washington, D.C. April 2004. Available online at:
3 https://www.nass. usda.gov/Publications/index.php.
4 USDA (1999) Poultry Production and Value—Final Estimates 1994-97. National Agriculture Statistics Service, U.S.
5 Department of Agriculture. Washington, D.C. March 1999. Available online at:
6 https://www.nass.usda.gov/Publications/index.php.
7 USDA (1998) Chicken and Eggs—Final Estimates 1994-97. National Agriculture Statistics Service, U.S. Department
8 of Agriculture. Washington, D.C. December 1998. Available online at:
9 https://www.nass.usda.gov/Publications/index.php.
10 USDA (1996) Agricultural Waste Management Field Handbook, National Engineering Handbook (NEH), Part 651.
11 Natural Resources Conservation Service, U.S. Department of Agriculture. July 1996.
12 USDA (1995a) Poultry Production and Value—Final Estimates 1988-1993. National Agriculture Statistics Service,
13 U.S. Department of Agriculture. Washington, D.C. March 1995. Available online at:
14 https://www.nass.usda.gov/Publications/index.php.
15 USDA (1995b) Chicken and Eggs—Final Estimates 1988-1993. National Agriculture Statistics Service, U.S.
16 Department of Agriculture. Washington, D.C. December 1995. Available online at:
17 https://www.nass.usda.gov/Publications/index.php.
18 USDA (1994) Sheep and Goats—Final Estimates 1989-1993. National Agriculture Statistics Service, U.S. Department
19 of Agriculture. Washington, D.C. January 31,1994. Available online at:
20 https://www.nass.usda.gov/Publications/index.php.
21 USDA APHIS (2003) Sheep 2001, Part I: Reference of Sheep Management in the United States, 2001 and Part IV:
22 Baseline Reference of 2001 Sheep Feedlot Health and Management. USDA-APHIS-VS. Fort Collins, CO. #N356.0702.
23 Available online at http://www.aphis.usda.gov/animal health/nahms/sheep/index.shtml#sheep2001.
24 USDA APHIS (2000) Layers '99—Part II: References of 1999 Table Egg Layer Management in the U.S. USDA-APHIS-
25 VS. Fort Collins, CO. Available online at
26 http://www.aphis.usda.gov/animal health/nahms/poultry/downloads/layers99/Layers99 dr Partll.pdf.
27 USDA APHIS (1996) Swine '95: Grower/Finisher Part II: Reference of 1995 U.S. Grower/Finisher Health &
28 Management Practices. USDA-APHIS-VS. Fort Collins, CO. Available online at:
29 http://www.aphis.usda.gov/animal health/nahms/swine/downloads/swine95/Swine95 dr Partll.pdf.
30 Rice Cultivation
31 Baicich, P. (2013) The Birds and Rice Connection. Bird Watcher's Digest. Available online at:
32 http://www.usarice.com/doclib/194/6867.pdf.
33 Brockwell, P.J., and R.A. Davis (2016) Introduction to time series and forecasting. Springer.
34 Cantens, G. (2004 through 2005) Personal Communication. Janet Lewis, Assistant to Gaston Cantens, Vice
35 President of Corporate Relations, Florida Crystals Company and ICF International.
36 Cheng, K., S.M. Ogle, W.J. Parton, G. Pan. (2014) "Simulating greenhouse gas mitigation potentials for Chinese
37 croplands using the DAYCENT ecosystem model." Global Change Biology 20:948-962.
38 Cheng, K., S.M. Ogle, W.J. Parton and G. Pan. (2013) "Predicting methanogenesis from rice paddies using the
39 DAYCENT ecosystem model." Ecological Modelling 261-262:19-31.
40 Del Grosso, S.J., S.M. Ogle, W.J. Parton, and F.J. Breidt (2010) "Estimating Uncertainty in N2O Emissions from U.S.
41 Cropland Soils." Global Biogeochemical Cycles, 24, GB1009, doi:10.1029/2009GB003544.
Waste 57
-------
1 Deren, C. (2002) Personal Communication and Dr. Chris Deren, Everglades Research and Education Centre at the
2 University of Florida and Caren Mintz, ICF International. August 15, 2002.
3 Fitzgerald, G.J., K. M. Scow & J. E. Hill (2000) "Fallow Season Straw and Rice Management Effects on Methane
4 Emissions in California Rice." Global biogeochemical cycles, 14 (3), 767-776.
5 Fleskes, J.P., Perry, W.M., Petrik, K.L., Spell, R., and Reid, F. (2005) Change in area of winter-flood and dry rice in
6 the northern Central Valley of California determined by satellite imagery. California Fish and Game, 91: 207-215.
7 Gonzalez, R. (2007 through 2014) Email correspondence. Rene Gonzalez, Plant Manager, Sem-Chi Rice Company
8 and ICF International.
9 Hardke, J.T. (2015) Trends in Arkansas rice production, 2014. B.R. Wells Arkansas Rice Research Studies 2014.
10 Norman, R.J. and Moldenhauer, K.A.K. (Eds.). Research Series 626, Arkansas Agricultural Experiment Station,
11 University of Arkansas.
12 Hardke, J. (2014) Personal Communication. Dr. Jarrod Hardke, Rice Extension Agronomist at the University of
13 Arkansas Rice Research and Extension Center and Kirsten Jaglo, ICF International. September 11, 2014.
14 Hardke, J. (2013) Email correspondence. Dr. Jarrod Hardke, Rice Extension Agronomist at the University of
15 Arkansas Rice Research and Extension Center and Cassandra Snow, ICF International. July 15, 2013.
16 Hardke, J.T., and Wilson, C.E. Jr., (2014) Trends in Arkansas rice production, 2013. B.R. Wells Arkansas Rice
17 Research Studies 2013. Norman, R.J., and Moldenhauer, K.A.K., (Eds.). Research Series 617, Arkansas Agricultural
18 Experiment Station, University of Arkansas.
19 Hardke, J.T., and Wilson, C.E. Jr., (2013) Trends in Arkansas rice production. B.R. Wells Arkansas Rice Research
20 Studies 2012. Norman, R.J., and Moldenhauer, K.A.K., (Eds.). Research Series 609, Arkansas Agricultural Experiment
21 Station, University of Arkansas.
22 Hollier, C. A. (ed), (1999) Louisiana rice production handbook. Louisiana State University Agricultural Center. LCES
23 Publication Number 2321. 116 pp.
24 Holzapfel-Pschorn, A., R. Conrad, and W. Seiler (1985) "Production, Oxidation, and Emissions of Methane in Rice
25 Paddies." FEMS Microbiology Ecology, 31:343-351.
26 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
27 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
28 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
29 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
30 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
31 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
32 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
33 Kirstein, A. (2003 through 2004, 2006) Personal Communication. Arthur Kirstein, Coordinator, Agricultural
34 Economic Development Program, Palm Beach County Cooperative Extension Service, FL and ICF International.
35 Klosterboer, A. (1997,1999 through 2003) Personal Communication. Arlen Klosterboer, retired Extension
36 Agronomist, Texas A&M University and ICF International. July 7, 2003.
37 Lindau, C.W. and P.K. Bollich (1993) "Methane Emissions from Louisiana First and Ratoon Crop Rice." Soil Science,
38 156:42-48.
39 Linquist, B.A., M.A. Adviento-Borbe, C.M. Pittelkow, C.v. Kessel, et al. (2012) Fertilizer management practices and
40 greenhouse gas emissions from rice systems: A quantitative review and analysis. Field Crops Research, 135:10-21.
41 Linscombe, S. (1999, 2001 through 2014) Email correspondence. Steve Linscombe, Professor with the Rice
42 Research Station at Louisiana State University Agriculture Center and ICF International.
58 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 LSU, (2015) Louisiana ratoon crop and conservation: Ratoon & Conservation Tillage Estimates. Louisiana State
2 University, College of Agriculture AgCenter. Online at: www.lsuagcenter.com.
3 Miller, M.R., Garr, J.D., and Coates, P.S., (2010) Changes in the status of harvested rice fields in the Sacramento
4 Valley, California: Implications for wintering waterfowl. Wetlands, 30: 939-947.
5 Neue, H.U., R. Wassmann, H.K. Kludze, W. Bujun, and R.S. Lantin (1997) "Factors and processes controlling
6 methane emissions from rice fields." Nutrient Cycling in Agroecosystems 49: 111-117.
7 Ogle, S.M., F.J. Breidt, M. Easter, S. Williams and K. Paustian. (2007) "An empirically based approach for estimating
8 uncertainty associated with modeling carbon sequestration in soils." Ecological Modelling 205:453-463.
9 Ogle, S.M., S. Spencer, M. Hartman, L. Buendia, L. Stevens, D. du Toit, J. Witi (2016) "Developing national baseline
10 GHG emissions and analyzing mitigation potentials for agriculture and forestry using an advanced national GHG
11 inventory software system." In Advances in Agricultural Systems Modeling 6, Synthesis and Modeling of
12 Greenhouse Gas Emissions and Carbon Storage in Agricultural and Forestry Systems to Guide Mitigation and
13 Adaptation, S. Del Grosso, LR. Ahuja and W.J. Parton (eds.), American Society of Agriculture, Crop Society of
14 America and Soil Science Society of America, pp. 129-148.
15 Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface Submodel: Description
16 and Testing". Glob. Planet. Chang. 19: 35-48.
17 Parton, W.J., D.S. Schimel, C.V. Cole, D.S. Ojima (1987) "Analysis of factors controlling soil organic matter levels in
18 Great Plains grasslands." Soil Science Society of America Journal 51:1173-1179.
19 Sass, R. L. (2001) CFU Emissions from Rice Agriculture. Good Practice Guidance and Uncertainty Management in
20 National Greenhouse Gas Inventories. 399-417. Available online at: http://www.ipcc-
21 nggip.iges.or.jp/public/gp/bgp/4 7 CH4 Rice Agriculture.pdf.
22 Sass, R.L, F.M. Fisher, P.A. Harcombe, and F.T. Turner (1990) "Methane Production and Emissions in a Texas Rice
23 Field." Global Biogeochemical Cycles, 4:47-68.
24 Sass, R.L, F.M. Fisher, S.T. Lewis, M.F. Jund, and F.T. Turner. (1994) "Methane emissions from rice fields: effect of
25 soil texture." Global Biogeochemical Cycles 8:135-140.
26 Schueneman, T. (1997,1999 through 2001) Personal Communication. Tom Schueneman, Agricultural Extension
27 Agent, Palm Beach County, FLand ICF International.
28 Slaton, N. (1999 through 2001) Personal Communication. Nathan Slaton, Extension Agronomist—Rice, University
29 of Arkansas Division of Agriculture Cooperative Extension Service and ICF International.
30 Stansel, J. (2004 through 2005) Email correspondence. Dr. Jim Stansel, Resident Director and Professor Emeritus,
31 Texas A&M University Agricultural Research and Extension Center and ICF International.
32 TAMU (2015) Texas Rice Crop Survey. Texas A&M AgriLIFE Research Center at Beaumont. Online at:
33 https://beaumont.tamu.edu/.
34 Texas Agricultural Experiment Station (2007 through 2014) Texas Rice Acreage by Variety. Agricultural Research
35 and Extension Center, Texas Agricultural Experiment Station, Texas A&M University System. Available online at:
36 http://beaumont.tamu.edu/CropSurvev/CropSurveyReport.aspx.
37 Texas Agricultural Experiment Station (2006) 2005 - Texas Rice Crop Statistics Report. Agricultural Research and
38 Extension Center, Texas Agricultural Experiment Station, Texas A&M University System, p. 8. Available online at:
39 http://beaumont.tamu.edu/eLibrary/TRRFReport default.htm.
40 University of California Cooperative Extension (UCCE) (2015) Rice Production Manual. Revised (2015) UCCE, Davis,
41 in collaboration with the California Rice Research Board.
42 USDA (2005 through 2015) Crop Production Summary. National Agricultural Statistics Service, Agricultural Statistics
43 Board, U.S. Department of Agriculture, Washington, D.C. Available online at: http://usda.mannlib.cornell.edu.
Waste 59
-------
1 USDA (2012) Summary of USDA-ARS Research on the Interrelationship of Genetic and Cultural Management
2 Factors That Impact Grain Arsenic Accumulation in Rice. News and Events. Agricultural Research Service, U.S.
3 Department of Agriculture, Washington, D.C. Available online at:
4 http://www.ars.usda.gov/is/pr/2012/120919.htm. September 2013.
5 USDA (2003) Field Crops, Final Estimates 1997-2002. Statistical Bulletin No. 982. National Agricultural Statistics
6 Service, Agricultural Statistics Board, U.S. Department of Agriculture, Washington, D.C. Available online at:
7 http://usda.mannlib.cornell.edu/usda/reports/general/sb/. September 2005.
8 USDA (1998) Field Crops Final Estimates 1992-1997. Statistical Bulletin Number 947 a. National Agricultural
9 Statistics Service, Agricultural Statistics Board, U.S. Department of Agriculture, Washington, D.C. Available online
10 at: http://usda.mannlib.cornell.edu/. July 2001.
11 USDA (1994) Field Crops Final Estimates 1987-1992. Statistical Bulletin Number 896. National Agricultural Statistics
12 Service, Agricultural Statistics Board, U.S. Department of Agriculture, Washington, D.C. Available online at:
13 http://usda.mannlib.cornell.edu/. July 2001.
14 USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory. Natural Resources Conservation Service,
15 Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
16 https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd 1422028.pdf.
17 van Bodegom, P.M., R. Wassmann, T.M. Metra-Corton (2001) "A process based model for methane emission
18 predictions from flooded rice paddies." Global Biogeochemical Cycles 15: 247-263.
19 Wang, J.J., S.K. Dodla, S. Viator, M. Kongchum, S. Harrison, S. D. Mudi, S. Liu, Z. Tian (2013) Agriculture Field
20 Management Practices and Greenhouse Gas Emissions from Louisiana Soils. Louisiana Agriculture, Spring 2013: 8-
21 9. Available online at: http://www.lsuagcenter.com/NR/rdonlyres/78D8B61A-96A8-49El-B2EF-
22 BAlD4CE4E698/93016/v56no2Spring2013.pdf.
23 Wassmann, R. H.U. Neue, R.S. Lantin, K. Makarim, N. Chareonsil5, LV. Buendia, and H. Rennenberg (2000a)
24 Characterization of methane emissions from rice fields in Asia II. Differences among irrigated, rainfed, and
25 deepwater rice." Nutrient Cycling in Agroecosystems, 58(1): 13-22.
26 Wassmann, R., R.S. Lantin, H.U. Neue, LV. Buendia, et al. (2000b) "Characterization of Methane Emissions from
27 Rice Fields in Asia. III. Mitigation Options and Future Research Needs." Nutrient Cycling in Agroecosy stems,
28 58(l):23-36.
29 Way, M.O., McCauley, G.M., Zhou, X.G., Wilson, L.T., and Morace, B. (Eds.), (2014) 2014 Texas Rice Production
30 Guidelines. Texas A&M AgriLIFE Research Center at Beaumont.
31 Wilson, C. (2002 through 2007, 2009 through 2012) Personal Communication. Dr. Chuck Wilson, Rice Specialist at
32 the University of Arkansas Cooperative Extension Service and ICF International.
33 Wilson, C.E. Jr., and Branson, J.W., (2006) Trends in Arkansas rice production. B.R. Wells Arkansas Rice Research
34 Studies 2005. Norman, R.J., Meullenet, J.-F., and Moldenhauer, K.A.K., (Eds.). Research Series 540, Arkansas
35 Agricultural Experiment Station, University of Arkansas.
36 Wilson, C.E. Jr., and Branson, J.W., (2005) Trends in Arkansas rice production. B.R. Wells Arkansas Rice Research
37 Studies 2004. Norman, R.J., Meullenet, J.-F., and Moldenhauer, K.A.K., (Eds.). Research Series 529, Arkansas
38 Agricultural Experiment Station, University of Arkansas.
39 Wilson, C.E. Jr., Runsick, S.K., and Mazzanti, R., (2010) Trends in Arkansas rice production. B.R. Wells Arkansas Rice
40 Research Studies 2009. Norman, R.J., and Moldenhauer, K.A.K., (Eds.). Research Series 581, Arkansas Agricultural
41 Experiment Station, University of Arkansas.
42 Wilson, C.E. Jr., Runsick, S.K., Mazzanti, R., (2009) Trends in Arkansas rice production. B.R. Wells Arkansas Rice
43 Research Studies (2008) Norman, R.J., Meullenet, J.-F., and Moldenhauer, K.A.K., (Eds.). Research Series 571,
44 Arkansas Agricultural Experiment Station, University of Arkansas.
60 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Wilson, C.E. Jr., and Runsick, S.K., (2008) Trends in Arkansas rice production. B.R. Wells Arkansas Rice Research
2 Studies 2007. Norman, R.J., Meullenet, J.-F., and Moldenhauer, K.A.K., (Eds.). Research Series 560, Arkansas
3 Agricultural Experiment Station, University of Arkansas.
4 Wilson, C.E. Jr., and Runsick, S.K., (2007) Trends in Arkansas rice production. B.R. Wells Arkansas Rice Research
5 Studies 2006. Norman, R.J., Meullenet, J.-F., and Moldenhauer, K.A.K., (Eds.). Research Series 550, Arkansas
6 Agricultural Experiment Station, University of Arkansas.
7 Yan, X., H. Akiyana, K. Yagi, and H. Akimoto (2009) "Global estimations of the inventory and mitigation potential of
8 methane emissions from rice cultivation conducted using the 2006 Intergovernmental Panel on Climate Change
9 Guidelines." Global Biogeochemical Cycles, 23, DOI: 0.1029/2008GB003299.
10 Young, M. (2013) Rice and Ducks. Ducks Unlimited, Memphis, TN. Available online at:
11 http://www.ducks.org/conservation/farm-bill/rice-and-ducks—by-matt-young.
12 Agricultural Soil Management
13 AAPFCO (2008 through 2022) Commercial Fertilizers: 2008-2017. Association of American Plant Food Control
14 Officials. University of Missouri. Columbia, MO.
15 AAPFCO (1995 through 2000a, 2002 through 2007) Commercial Fertilizers: 1995-2007. Association of American
16 Plant Food Control Officials. University of Kentucky. Lexington, KY.
17 Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
18 Cibrowski, P. (1996) Personal Communication. Peter Cibrowski, Minnesota Pollution Control Agency and Heike
19 Mainhardt, ICF Incorporated. July 29,1996.
20 Cheng, B., and D.M. Titterington (1994) "Neural networks: A review from a statistical perspective." Statistical
21 science 9: 2-30.
22 Claassen, R., M. Bowman, J. McFadden, D. Smith, and S. Wallander (2018) Tillage intensity and conservation
23 cropping in the United States, EIB 197. United States Department of Agriculture, Economic Research Service,
24 Washington, D.C.
25 CTIC (2004) 2004 Crop Residue Management Survey. Conservation Technology Information Center. Available at
26 http://www.ctic.purdue.edu/CRM/.
27 Del Grosso, S.J., T. Wirth, S.M. Ogle, W.J. Parton (2008) Estimating agricultural nitrous oxide emissions. EOS 89,
28 529-530.
29 Del Grosso, S.J., A.R. Mosier, W.J. Parton, and D.S. Ojima (2005) "DAYCENT Model Analysis of Past and
30 Contemporary Soil N2O and Net Greenhouse Gas Flux for Major Crops in the USA." Soil Tillage and Research, 83: 9-
31 24. doi: 10.1016/j.still.2005.02.007.
32 Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel (2001) "Simulated
33 Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the DAYCENT Model." In Schaffer, M., L. Ma,
34 S. Hansen, (eds.). Modeling Carbon and Nitrogen Dynamics for Soil Management. CRC Press. Boca Raton, Florida.
35 303-332.
36 Del Grosso, S.J., S.M. Ogle, W.J. Parton, and F.J. Breidt (2010) "Estimating Uncertainty in N2O Emissions from U.S.
37 Cropland Soils." Global Biogeochemical Cycles, 24, GB1009, doi:10.1029/2009GB003544.
38 Del Grosso, S.J., W.J. Parton, C.A. Keough, and M. Reyes-Fox. (2011) Special features of the DAYCENT modeling
39 package and additional procedures for parameterization, calibration, validation, and applications, in Methods of
40 Introducing System Models into Agricultural Research, L.R. Ahuja and Liwang Ma, editors, p. 155-176, American
41 Society of Agronomy, Crop Science Society of America, Soil Science Society of America, Madison, Wl. USA.
Waste 61
-------
1 Del Grosso, S. J., S. M. Ogle, C. Nevison, R. Gurung, W. J. Parton, C. Wagner-Riddle, W. Smith, W. Winiwarter, B.
2 Grant, M. Tenuta, E. Marx, S. Spencer, and S. Williams. 2022. A gap in nitrous oxide emission reporting complicates
3 long-term climate mitigation. Proceedings of the National Academy of Sciences 119:e2200354119.
4 Delgado, J.A., S.J. Del Grosso, and S.M. Ogle (2009) "15N isotopic crop residue cycling studies and modeling suggest
5 that IPCC methodologies to assess residue contributions to N2O-N emissions should be reevaluated." Nutrient
6 Cycling in Agroecosystems, DOI 10.1007/sl0705-009-9300-9.
7 Edmonds, L, N. Gollehon, R.L. Kellogg, B. Kintzer, L. Knight, C. Lander, J. Lemunyon, D. Meyer, D.C. Moffitt, and J.
8 Schaeffer (2003) "Costs Associated with Development and Implementation of Comprehensive Nutrient
9 Management Plans." Part 1. Nutrient Management, Land Treatment, Manure and Wastewater Handling and
10 Storage, and Recordkeeping. Natural Resource Conservation Service, U.S. Department of Agriculture.
11 EPA (2003) Clean Watersheds Needs Survey 2000—Report to Congress, U.S. Environmental Protection Agency.
12 Washington, D.C. Available online at: http://www.epa.gov/owm/mtb/cwns/2000rtc/toc.htm.
13 EPA (1999) Biosolids Generation, Use and Disposal in the United States. Office of Solid Waste, U.S. Environmental
14 Protection Agency. Available online at: http://biosolids.policy.net/relatives/18941.PDF.
15 EPA (1993) Federal Register. Part II. Standards for the Use and Disposal of Sewage Sludge; Final Rules. U.S.
16 Environmental Protection Agency, 40 CFR Parts 257, 403, and 503.
17 Firestone, M. K., and E.A. Davidson, Ed. (1989) Microbiological basis of NO and N2O production and consumption in
18 soil. Exchange of trace gases between terrestrial ecosystems and the atmosphere. New York, John Wiley & Sons.
19 Friedman, J.H. (2001) "Greedy function approximation: A gradient boosting machine." Ann. Statist. 29 (5) 1189 -
20 1232.
21 Hagen, S. C., G. Delgado, P. Ingraham, I. Cooke, R. Emery, J. P. Fisk, L. Melendy, T. Olson, S. Patti, N. Rubin, B. Ziniti,
22 H. Chen, W. Salas, P. Elias, and D. Gustafson. 2020. Mapping Conservation Management Practices and Outcomes in
23 the Corn Belt Using the Operational Tillage Information System (OpTIS) and the Denitrification-Decomposition
24 (DNDC) Model. Land 9:408.
25 ILENR (1993) Illinois Inventory of Greenhouse Gas Emissions and Sinks: 1990. Office of Research and Planning,
26 Illinois Department of Energy and Natural Resources. Springfield, IL.
27 IPCC (2014) 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands. The
28 Intergovernmental Panel on Climate Change. [T, Hiraishi, T. Krug, K. Tanabe, N. Srivastava, B. Jamsranjav, M.
29 Fukuda and T. Troxler (eds.)]. Hayama, Kanagawa, Japan.
30 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
31 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
32 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
33 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
34 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
35 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
36 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
37 Little, R. (1988) "Missing-data adjustments in large surveys." Journal of Business and Economic Statistics 6: 287-
38 296.
39 McFarland, M.J. (2001) Biosolids Engineering, New York: McGraw-Hill, p. 2.12.
40 McGill, W.B., and C.V. Cole (1981) Comparative aspects of cycling of organic C, N, S and P through soil organic
41 matter. Geoderma 26:267-286.
42 Metherell, A.K., L.A. Harding, C.V. Cole, and W.J. Parton (1993) "CENTURYSoil Organic Matter Model
43 Environment." Agroecosystem version 4.0. Technical documentation, GPSRTech. Report No. 4, USDA/ARS, Ft.
44 Collins, CO.
62 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 NEBRA (2007) A National Biosolids Regulation, Quality, End Use & Disposal Survey. North East Biosolids and
2 Residuals Association, July 21, 2007.
3 Noller, J. (1996) Personal Communication. John Noller, Missouri Department of Natural Resources and Heike
4 Mainhardt, ICF Incorporated. July 30,1996.
5 Ogle, S.M., F.J. Breidt, M. Easter, S. Williams and K. Paustian (2007) "Empirically-Based Uncertainty Associated with
6 Modeling Carbon Sequestration Rates in Soils." Ecological Modeling 205:453-463.
7 Oregon Department of Energy (1995) Report on Reducing Oregon's Greenhouse Gas Emissions: Appendix D
8 Inventory and Technical Discussion. Oregon Department of Energy. Salem, OR.
9 Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface Submodel: Description
10 and Testing." Glob. Planet. Chang. 19: 35-48.
11 Potter, C., S. Klooster, A. Huete, and V. Genovese (2007) Terrestrial carbon sinks for the United States predicted
12 from MODIS satellite data and ecosystem modeling. Earth Interactions 11, Article No. 13, DOI 10.1175/EI228.1.
13 Potter, C. S., J.T. Randerson, C.B. Fields, P.A. Matson, P.M. Vitousek, H.A. Mooney, and S.A. Klooster (1993)
14 "Terrestrial ecosystem production: a process model based on global satellite and surface data." Global
15 Biogeochemical Cycles 7:811-841.
16 PRISM Climate Group (2022) PRISM Climate Data, Oregon State University, http://prism.oregonstate.edu,
17 downloaded January 2022.
18 Pukelsheim, F. (1994) 'The 3-Sigma-Rule." American Statistician 48:88-91.
19 Ruddy B.C., D.L Lorenz, and D.K. Mueller (2006) County-level estimates of nutrient inputs to the land surface of
20 the conterminous United States, 1982-2001. Scientific Investigations Report 2006-5012. U.S Department of the
21 Interior.
22 Scheer, C., S.J. Del Grosso, W.J. Parton, D.W. Rowlings, P.R. Grace (2013) Modeling Nitrous Oxide Emissions from
23 Irrigated Agriculture: Testing DAYCENT with High Frequency Measurements, Ecological Applications, in press.
24 Available online at: http://dx.doi.Org/10.1890/13-0570.l.
25 Soil Survey Staff (2020) Gridded Soil Survey Geographic (gSSURGO) Database for the Conterminous United States.
26 United States Department of Agriculture, Natural Resources Conservation Service, Accessed February 2020
27 (FY2020 official release), Available online at https://gdg.sc.egov.usda.gov/.
28 Towery, D. (2001) Personal Communication. Dan Towery regarding adjustments to the CTIC (1998) tillage data to
29 reflect long-term trends, Conservation Technology Information Center, West Lafayette, IN, and Marlen Eve,
30 National Resource Ecology Laboratory, Fort Collins, CO. February 2001.
31 TVA (1991 through 1992a, 1993 through 1994) Commercial Fertilizers. Tennessee Valley Authority, Muscle Shoals,
32 AL.
33 USDA-ERS (2020) Agricultural Resource Management Survey (ARMS) Farm Financial and Crop Production Practices:
34 Tailored Reports. Available online at: https://www.ers.usda.gov/data-products/arms-farm-financial-and-crop-
35 production-practices/.
36 USDA-ERS (1997) Cropping Practices Survey Data —1995. Economic Research Service, United States Department of
37 Agriculture. Available online at: http://www.ers.usda.gov/data/archive/93018/.
38 USDA-NASS (2022) Quick Stats. National Agricultural Statistics Service, United States Department of Agriculture,
39 Washington, D.C., Accessed October 2022, http://quickstats.nass.usda.gov/.
40 USDA-NASS (2021) Published crop data layer. Available at https://nassgeodata.gmu.edu/CropScape/, Accessed July
41 2021, USDA-NASS, Washington, DC.
42 USDA-NASS (2017) 2017 Census of Agriculture. USDA National Agricultural Statistics Service, Complete data
43 available at http://www.nass.usda.gov/AgCensus.
Waste 63
-------
1 USDA-NASS (2012) 2012 Census of Agriculture. USDA National Agricultural Statistics Service, Complete data
2 available at http://www.nass.usda.gov/AgCensus.
3 USDA-NASS (2004) Agricultural Chemical Usage: 2003 Field Crops Summary. Report AgChl(04)a. National
4 Agricultural Statistics Service, U.S. Department of Agriculture, Washington, D.C. Available online at:
5 http://usda.mannlib.cornell.edu/reports/nassr/other/pcu-bb/agcs0504.pdfH.
6 USDA-NASS (1999) Agricultural Chemical Usage: 1998 Field Crops Summary. Report AgCHl(99). National
7 Agricultural Statistics Service, U.S. Department of Agriculture, Washington, DC. Available online at:
8 http://usda.mannlib.cornell.edu/reports/nassr/other/pcu-bb/agch0599.pdf.
9 USDA-NASS (1992) Agricultural Chemical Usage: 1991 Field Crops Summary. Report AgChl(92). National
10 Agricultural Statistics Service, U.S. Department of Agriculture, Washington, D.C. Available online at:
11 http://usda.mannlib.cornell.edu/reports/nassr/other/pcu-bb/agch0392.txtH.
12 USDA-NRCS (2012) Assessment of the Effects of Conservation Practices on Cultivated Cropland in the Upper
13 Mississippi River Basin. U.S. Department of Agriculture, Natural Resources Conservation Service,
14 https://www.nrcs.usda.gov/lnternet/FSE DQCUMENTS/stelprdbl042093.pdf.
15 USDA-NRCS (2018) CEAP Cropland Farmer Surveys. USDA Natural Resources Conservation Service.
16 https://www.nrcs.usda.gov/wps/portal/nrcs/detail/national/technical/nra/ceap/na/?cid=nrcsl43 014163.
17 USDA-NRCS (2020) Summary Report: 2017 National Resources Inventory. Natural Resources Conservation Service,
18 Washington, DC, and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
19 https://www.nrcs.usda.gov/wps/portal/nrcs/main/national/technical/nra/nri/results/.
20 USDA-NRCS (2022) Conversation practice on cultivated croplands: A comparison of CEAP I and CEAP II survey data
21 and modeling. United States Department of Agriculture, Natural Resources Conservation Service,
22 https://www.nrcs.usda.gov/sites/default/files/2022-09/CEAP-Croplands-
23 ConservationPracticesonCultivatedCroplands-Report-March2022.pdf.
24 USFS (2019) Forest Inventory and Analysis Program. United States Department of Agriculture, U.S. Forest Service,
25 https://www.fia.fs.fed.us/tools-data/default.asp.
26 Van Buuren, S. (2012) "Flexible imputation of missing data." Chapman & Hall/CRC, Boca Raton, FL
27 Wagner-Riddle, C., Congreves, K. A., Abalos, D., Berg, A. A., Brown, S. E., Ambadan, J. T., Gao, X. & Tenuta, M.
28 (2017) "Globally important nitrous oxide emissions from croplands induced by freeze-thaw cycles." Nature
29 Geosciences 10(4): 279-283.
30 Wisconsin Department of Natural Resources (1993) Wisconsin Greenhouse Gas Emissions: Estimates for 1990.
31 Bureau of Air Management, Wisconsin Department of Natural Resources, Madison, Wl.
32 Yang, L, Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J., Fry, J., Funk, M.,
33 Granneman, B., Liknes, G. C., Rigge, M. &Xian, G. (2018) "A new generation of the United States National Land
34 Cover Database: Requirements, research priorities, design, and implementation strategies." ISPRS Journal of
35 Photogrammetry and Remote Sensing 146:108-123.
36 Liming
37 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
38 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
39 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
40 Tepordei, V.V. (1994 through 2015) "Crushed Stone," In Minerals Yearbook. U.S. Department of the Interior/U.S.
41 Geological Survey. Washington, D.C. Available online at: http://minerals.usgs.gov/minerals/.
64 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Tepordei, V.V. (2003b) Personal communication. Valentin Tepordei, U.S. Geological Survey and ICF Consulting,
2 August 18, 2003.
3 USGS (2022) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the First Quarter of 2022, U.S.
4 Geological Survey, Reston, VA. Available online at:
5 http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis.
6 USGS (2021) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the Fourth Quarter of 2021, U.S.
7 Geological Survey, Reston, VA. Available online at:
8 http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis.
9 USGS (2020) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the Fourth Quarter of 2020, U.S.
10 Geological Survey, Reston, VA. Available online at:
11 http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis.
12 West, T.O., and A.C. McBride (2005) "The contribution of agricultural lime to carbon dioxide emissions in the
13 United States: dissolution, transport, and net emissions," Agricultural Ecosystems & Environment 108:145-154.
14 West, T.O. (2008) Email correspondence. Tristram West, Environmental Sciences Division, Oak Ridge National
15 Laboratory, U.S. Department of Energy and Nikhil Nadkarni, ICF International on suitability of liming emission
16 factor for the entire United States. June 9, 2008.
17 Willett, J.C. (2022d) Personal communication. Jason Willett. Preliminary data tables from "Crushed Stone," In 2021
18 Minerals Yearbook. U.S. Department of the Interior/U.S. Geological Survey. Washington, D.C. October, 2022.
19 Willett, J.C. (2022c) "Crushed Stone," In Minerals Yearbook 2020. U.S. Department of the Interior/U.S. Geological
20 Survey, Washington, D.C. Available online at: https://www.usgs.gov/centers/national-minerals-information-
21 center/crushed-stone-statistics-and-information. Accessed October 2022
22 Willett, J.C. (2022b) "Crushed Stone," In Minerals Yearbook 2019. U.S. Department of the Interior/U.S. Geological
23 Survey, Washington, D.C. Available online at: https://www.usgs.gov/centers/national-minerals-information-
24 center/crushed-stone-statistics-and-information. Accessed October 2022
25 Willett, J.C. (2022a) "Crushed Stone," In Minerals Yearbook 2018. U.S. Department of the Interior/U.S. Geological
26 Survey, Washington, D.C. Available online at: https://www.usgs.gov/centers/national-minerals-information-
27 center/crushed-stone-statistics-and-information. Accessed October 2022.
28 Willett, J.C. (2020a) "Crushed Stone," In Minerals Yearbook 2016. U.S. Department of the Interior/U.S. Geological
29 Survey, Washington, D.C. Available online at:
30 http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed November 2020.
31 Willett, J.C. (2017) "Crushed Stone," In Minerals Yearbook 2015. U.S. Department of the Interior/U.S. Geological
32 Survey, Washington, D.C. Available online at:
33 http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed November 2017.
34 Willett, J.C. (2016) "Crushed Stone," In Minerals Yearbook 2014. U.S. Department of the Interior/U.S. Geological
35 Survey, Washington, D.C. Available online at:
36 http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed September 2016.
37 Willett, J.C. (2015) "Crushed Stone," In Minerals Yearbook 2013. U.S. Department of the Interior/U.S. Geological
38 Survey, Washington, D.C. Available online at:
39 http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed September 2015.
40 Willett, J.C. (2014) "Crushed Stone," In Minerals Yearbook 2012. U.S. Department of the Interior/U.S. Geological
41 Survey, Washington, D.C. Available online at:
42 http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed September 2014.
43 Willett, J.C. (2013b) Personal Communication. Jason Willet, U.S. Geological Survey and ICF International.
44 September 9, 2013.
Waste 65
-------
1 Willett, J.C. (2013a) "Crushed Stone," In Minerals Yearbook 2011. U.S. Department of the Interior/U.S. Geological
2 Survey, Washington, D.C. Available online at:
3 http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed May 2013.
4 Willett, J.C. (2011a) "Crushed Stone," In Minerals Yearbook 2009. U.S. Department of the Interior/U.S. Geological
5 Survey, Washington, D.C. Available online at:
6 http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed August 2011.
7 Willett, J.C. (2011b) "Crushed Stone," In Minerals Yearbook 2010. U.S. Department of the Interior/U.S. Geological
8 Survey, Washington, D.C. Available online at:
9 http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed September 2012.
10 Willett, J.C. (2010) "Crushed Stone," In Minerals Yearbook 2008. U.S. Department of the Interior/U.S. Geological
11 Survey, Washington, D.C. Available online at:
12 http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed August 2010.
13 Willett, J.C. (2009) "Crushed Stone," In Minerals Yearbook 2007. U.S. Department of the Interior/U.S. Geological
14 Survey, Washington, D.C. Available online at:
15 http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed August 2009.
16 Willett, J.C. (2007a) "Crushed Stone," In Minerals Yearbook 2005. U.S. Department of the Interior/U.S. Geological
17 Survey, Washington, D.C. Available online at:
18 http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed August 2007.
19 Willett, J.C. (2007b) "Crushed Stone," In Minerals Yearbook 2006. U.S. Department of the Interior/U.S. Geological
20 Survey, Washington, D.C. Available online at:
21 http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed August 2008.
22 Urea Fertilization
23 AAPFCO (2008 through 2022) Commercial Fertilizers. Association of American Plant Food Control Officials.
24 University of Missouri. Columbia, MO.
25 AAPFCO (1995 through 2000a, 2002 through 2007) Commercial Fertilizers. Association of American Plant Food
26 Control Officials. University of Kentucky. Lexington, KY.
27 AAPFCO (2000b) 1999-2000 Commercial Fertilizers Data, ASCII files. Available from David Terry, Secretary, AAPFCO.
28 EPA (2000) Preliminary Data Summary: Airport Deicing Operations (Revised). EPA-821-R-00-016. August 2000.
29 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
30 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
31 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
32 Itle, C. (2009) Email correspondence. Cortney Itle, ERG and Tom Wirth, U.S. Environmental Protection Agency on
33 the amount of urea used in aircraft deicing. January 7, 2009.
34 TVA (1991 through 1994) Commercial Fertilizers. Tennessee Valley Authority, Muscle Shoals, AL.
35 TVA (1992b) Fertilizer Summary Data 1992. Tennessee Valley Authority, Muscle Shoals, AL.
36 Field Burning of Agricultural Residues
37 Akintoye, H.A., Agbeyi, E.O., and Olaniyan, A.B. (2005) 'The effects of live mulches on tomato (Lycopersicon
38 esculentum) yield under tropical conditions." Journal of Sustainable Agriculture 26: 27-37.
39 Bange, M.P., Milroy, S.P., and Thongbai, P. (2004) "Growth and yield of cotton in response to waterlogging." Field
40 Crops Research 88:129-142.
66 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Beyaert, R.P. (1996) The effect of cropping and tillage management on the dynamics of soil organic matter. PhD
2 Thesis. University of Guelph.
3 Bouquet, D.J., and Breitenbeck, G.A. (2000) "Nitrogen rate effect on partitioning of nitrogen and dry matter by
4 cotton." Crop Science 40:1685-1693.
5 Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer. Cantens, G.
6 (2004 through 2005) Personal Communication. Janet Lewis, Assistant to Gaston Cantens, Vice President of
7 Corporate Relations, Florida Crystals Company and ICF International.
8 Brouder, S.M., and Cassman, K.G (1990) "Root development of two cotton cultivars in relation to potassium uptake
9 and plant growth in a vermiculitic soil." Field Crops Res. 23: 187-203.
10 Costa, L.D., and Gianquinto, G. (2002) "Water stress and watertable depth influence yield, water use efficiency,
11 and nitrogen recovery in bell pepper: lysimeter studies." Aust. J. Agric. Res. 53: 201-210.
12 Crafts-Brandner, S.J., Collins, M., Sutton, T.G., and Burton, H.R. (1994) "Effect of leaf maleic hydrazide
13 concentration on yield and dry matter partitioning in burley tobacco (Nicotiana tabacum L.)." Field Crops Research
14 37: 121-128.
15 De Pinheiro Henriques, A.R., and Marcelis, L.F.M. (2000) "Regulation of growth at steady-state nitrogen nutrition in
16 lettuce (Lactuca sativa L.): Interactive effects of nitrogen and irradiance." Annals of Botany 86: 1073-1080.0.
17 Diaz-Perez, J.C., Silvoy, J., Phatak,S.C., Ruberson, J., and Morse, R. (2008) Effect of winter cover crops and co-till on
18 the yield of organically-grown bell pepper (Capsicum annum L). Acta Hort. 767:243-247.
19 Dua, K.L., and Sharma, V.K. (1976) "Dry matter production and energy contents often varieties of sugarcane at
20 Muzaffarnagar (Western Uttar Pradesh)." Tropical Ecology 17: 45-49.
21 Fritschi, F.B., Roberts, B.A., Travis, R.L, Rains, D.W., and Hutmacher, R.B. (2003) "Seasonal nitrogen concentration,
22 uptake, and partitioning pattern of irrigated Acala and Pima cotton as influenced by nitrogen fertility level." Crop
23 Science 44:516-527.
24 Gerik, T.J., K.L Faver, P.M. Thaxton, and K.M. El-Zik. (1996) "Late season water stress in cotton: I. Plant growth,
25 water use, and yield." Crop Science 36: 914-921.
26 Gibberd, M.R., McKay, A.G., Calder, T.C., and Turner, N.C. (2003) "Limitations to carrot (Daucus carota L.)
27 productivity when grown with reduced rates of frequent irrigation on a free-draining, sandy soil." Australian
28 Journal of Agricultural Research 54: 499-506.
29 Giglio, L., I. Csiszar, and C.O. Justice (2006) "Global distribution and seasonality of active fires as observed with the
30 Terra and Aqua Moderate Resolution Imaging Spectroradiometer (MODIS) sensors" J. Geophys. Res. Ill, G02016,
31 doi:10.1029/2005JG000142.
32 Halevy, J. (1976) "Growth rate and nutrient uptake of two cotton cultivars grown under irrigation." Agronomy
33 Journal 68: 701-705.
34 Halvorson, A.D., Follett, R.F., Bartolo, M.E., and Schweissing, F.C. (2002) "Nitrogen fertilizer use efficiency of
35 furrow-irrigated onion and corn." Agronomy Journal 94: 442-449.
36 Heitholt, J.J., Pettigrew, W.T., and Meredith, W.R. (1992) "Light interception and lint yield of narrow-row cotton."
37 Crop Science 32: 728-733.
38 Hollifield, C.D., Silvertooth, J.C., and Moser, H. (2000) "Comparison of obsolete and modern cotton cultivars for
39 irrigated production in Arizona." 2000 Arizona Cotton Report, University of Arizona College of Agriculture,
40 http://ag.arizona.edu/pubs/crops/azll70/.
41 Hopkinson, J.M. (1967) "Effects of night temperature on the growth of Nicotiana tabacum." Australian Journal of
42 Experimental Agriculture and Animal Husbandry 1: 78-82.
Waste 67
-------
1 Huett, D.O., and Dettman, E.B. (1991) Effect of nitrogen on growth, quality and nutrient uptake of cabbages grown
2 in sand culture. Australian Journal of Experimental Agriculture 29: 875-81.
3 Huett, D.O., and Dettman, B. (1989) "Nitrogen response surface models of zucchini squash, head lettuce and
4 potato." Plant and Soil 134: 243-254.
5 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
6 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
7 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
8 IPCC/UNEP/OECD/IEA (1997) Revised 1996 IPCC Guidelines for National Greenhouse Gas Inventories.
9 Intergovernmental Panel on Climate Change, United Nations Environment Programme, Organization for Economic
10 Co-Operation and Development, International Energy Agency, Paris, France.
11 Jacobs, J.L, Ward, G.N., and Kearney, G. (2004) "Effects of irrigation strategies and nitrogen fertilizer on turnip dry
12 matter yield, water use efficiency, nutritive characteristics and mineral content in western Victoria." Australian
13 Journal of Experimental Agriculture 44:13-26.
14 Jacobs, J.L, Ward, G.N., McDowell, A.M., and Kearney, G. (2002) "Effect of seedbed cultivation techniques, variety,
15 soil type and sowing time, on brassica dry matter yields, water use efficiency and crop nutritive characteristics in
16 western Victoria." Australian Journal of Experimental Agriculture 42: 945-952.
17 Jacobs, J.L, Ward, G.N., McDowell, A.M., and Kearney, G.A. (2001) "A survey on the effect of establishment
18 techniques, crop management, moisture availability and soil type on turnip dry matter yields and nutritive
19 characteristics in western Victoria." Australian Journal of Experimental Agriculture 41: 743-751.
20 Kage, H., Alt, C., and Stutzel, H. (2003) "Aspects of nitrogen use efficiency of cauliflower II. Productivity and
21 nitrogen partitioning as influenced by N supply." Journal of Agricultural Science 141: 17-29.
22 Kumar, A., Singh, D.P., and Singh, P. (1994) "Influence of water stress on photosynthesis, transpiration, water-use
23 efficiency and yield of Brassica juncea L" Field Crops Research 37: 95-101.
24 LANDFIRE (2008) Existing Vegetation Type Layer, LANDFIRE 1.1.0, U.S. Department of the Interior, Geological
25 Survey. Accessed 28 October 2010 at http://landfire.cr.usgs.gov/viewer/.
26 MacLeod, LB., Gupta, U.C., and Cutcliffe, J.A. (1971) "Effect of N, P, and K on root yield and nutrient levels in the
27 leaves and roots of rutabagas grown in a greenhouse." Plant and Soil 35: 281-288.
28 Mahrani, A., and Aharonov, B. (1964) "Rate of nitrogen absorption and dry matter production by upland cotton
29 grown under irrigation." Israel J. Agric. Res. 14: 3-9.
30 Marcussi, F.F.N., Boas, R.L.V., de Godoy, L.J.G., and Goto, R. (2004) "Macronutrient accumulation and partitioning
31 in fertigated sweet pepper plants." Sci. Agric. (Piracicaba, Braz.) 61: 62-68.
32 McCarty, J.L. (2011) "Remote Sensing-Based Estimates of Annual and Seasonal Emissions from Crop Residue
33 Burning in the Contiguous United States." Journal of the Air & Waste Management Association, 61:1, 22-34, DOI:
34 10.3155/1047-3289.61.1.22.
35 McCarty, J.L. (2010) Agricultural Residue Burning in the Contiguous United States by Crop Type and State.
36 Geographic Information Systems (GIS) Data provided to the EPA Climate Change Division by George Pouliot,
37 Atmospheric Modeling and Analysis Division, EPA. Dr. McCarty's research was supported by the NRI Air Quality
38 Program of the Cooperative State Research, Education, and Extension Service, USDA, under Agreement No.
39 20063511216669 and the NASA Earth System Science Fellowship.
40 McCarty, J.L. (2009) Seasonal and Interannual Variability of Emissions from Crop Residue Burning in the Contiguous
41 United States. Dissertation. University of Maryland, College Park.
42 McPharlin, I.R., Aylmore, P.M., and Jeffery, R.C. (1992) "Response of carrots (Daucus carota L.) to applied
43 phosphorus and phosphorus leaching on a Karrakatta sand, under two irrigation regimes." Australian Journal of
44 Experimental Agriculture 32:225-232.
68 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Mondino, M.H., Peterlin, O.A., and Garay, F. (2004) "Response of late-planted cotton to the application of growth
2 regulator (chlorocholine chloride, CYCOCEL 75)." Expl Agric. 40: 381-387.
3 Moustakas, N.K., and Ntzanis, H. (2005) "Dry matter accumulation and nutrient uptake in flue-cured tobacco
4 (Nicotiana tabacum L.)." Field Crops Research 94:1-13.
5 Peach, L., Benjamin, L.R., and Mead, A. (2000) "Effects on the growth of carrots (Daucus carota L.), cabbage
6 (Brassica oleracea var. capitata L.) and onion (Allium cepa L) of restricting the ability of the plants to intercept
7 resources." Journal of Experimental Botany 51: 605-615.
8 Pettigrew, W.T., and Meredith, W.R., Jr. (1997) "Dry matter production, nutrient uptake, and growth of cotton as
9 affected by potassium fertilization." J. Plant Nutr. 20:531-548.
10 Pettigrew, W.T., Meredith, W.R., Jr., and Young, LD. (2005) "Potassium fertilization effects on cotton lint yield,
11 yield components, and reniform nematode populations." Agronomy Journal 97: 1245-1251.
12 PRISM Climate Group (2015) PRISM Climate Data. Oregon State University. July 24, 2015. Available online at:
13 http://prism.oregonstate.edu.
14 Reid, J.B., and English, J.M. (2000) "Potential yield in carrots (Daucus carota L.): Theory, test, and an application."
15 Annals of Botany 85: 593-605.
16 Sadras, V.O., and Wilson, LJ. (1997) "Growth analysis of cotton crops infested with spider mites: II. Partitioning of
17 dry matter." Crop Science 37: 492-497.
18 Scholberg, J., McNeal, B.L., Jones, J.W., Boote, K.J., Stanley, C.D., and Obreza, T.A. (2000a) "Growth and canopy
19 characteristics of field-grown tomato." Agronomy Journal 92: 152-159.
20 Scholberg, J., McNeal, B.L., Boote, K.J., Jones, J.W., Locasio, S.J., and Olson, S.M. (2000b) "Nitrogen stress effects on
21 growth and nitrogen accumulation by field-grown tomato." Agronomy Journal 92:159-167.
22 Singels, A. and Bezuidenhout, C.N. (2002) "A new method of simulating dry matter partitioning in the Canegro
23 sugarcane model." Field Crops Research 78: 151 - 164.
24 Sitompul, S.M., Hairiah, K., Cadisch, G., and Van Noordwuk, M. (2000) "Dynamics of density fractions of macro-
25 organic matter after forest conversion to sugarcane and woodlots, accounted for in a modified Century model."
26 Netherlands Journal of Agricultural Science 48: 61-73.
27 Stirling, G.R., Blair, B.L, Whittle, P.J.L., and Garside, A.L. (1999) "Lesion nematode (Pratylenchus zeae) is a
28 component of the yield decline complex of sugarcane." In: Magarey, R.C. (Ed.), Proceedings of the First
29 Australasian Soilborne Disease Symposium. Bureau of Sugar Experiment Stations, Brisbane, pp. 15-16.
30 Tan, D.K.Y., Wearing, A.H., Rickert, K.G., and Birch, C.J. (1999) "Broccoli yield and quality can be determined by
31 cultivar and temperature but not photoperiod in south-east Queensland." Australian Journal of Experimental
32 Agriculture 39: 901-909.
33 Tadesse, T., Nichols, M.A., and Fisher, K.J. (1999) Nutrient conductivity effects on sweet pepper plants grown using
34 a nutrient film technique. 1. Yield and fruit quality. New Zealand Journal of Crop and Horticultural Science, 27: 229-
35 237.
36 Torbert, H.A., and Reeves, D.W. (1994) "Fertilizer nitrogen requirements for cotton production as affected by
37 tillage and traffic." Soil Sci. Soc. Am. J. 58:1416-1423.
38 USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory, Natural Resources Conservation Service,
39 Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
40 https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd 1422028.pdf.
41 USDA (2019) Quick Stats: U.S. & All States Data; Crops; Production and Area Harvested; 1990 - 2018. National
42 Agricultural Statistics Service, U.S. Department of Agriculture. Washington, D.C. U.S. Department of Agriculture,
43 National Agricultural Statistics Service. Washington, D.C., Available online at: http://quickstats.nass.usda.gov/.
Waste 69
-------
1 Valantin, M., Gary, C., Vaissiere, B.E., and Frossard, J.S. (1999) "Effect of fruit load on partitioning of dry matter and
2 energy in cantaloupe (Cucumis melo L)." Annals of Botany 84:173-181.
3 Wallach, D., Marani, A., and Kletter, E. (1978) 'The relation of cotton crop growth and development to final yield."
4 Field Crops Research 1: 283-294.
5 Wells, R., and Meredith, W.R., Jr. (1984) "Comparative growth of obsolete and modern cultivars. I. Vegetative dry
6 matter partitioning." Crop Science 24: 858-872.4.
7 Wiedenfels, R.P. (2000) "Effects of irrigation and N fertilizer application on sugarcane yield and quality." Field Crops
8 Research 43:101-108.
9 Land Use, Land-Use Change, and Forestry
10 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
11 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
12 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
13 UNFCCC (2014) Report of the Conference of the Parties on its nineteenth session, held in Warsaw from 11 to 23
14 November 2013. United Nations Framework Convention on Climate Change, Warsaw. (FCCC/CP/2013/10/Add.3).
15 January 31, 2014. Available online at: http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
is Representation of the U.S. Land Base
17 Alaska Department of Natural Resources (2006) Alaska Infrastructure 1:63,360. Available online at:
18 http://dnr.alaska.gov/SpatialUtilitv/SUC?cmd=extract&layerid=75.
19 Alaska Interagency Fire Management Council (1998) Alaska Interagency Wildland Fire Management Plan. Available
20 online at: http://agdc.usgs.gov/data/blm/fire/index.html.
21 Alaska Oil and Gas Conservation Commission (2009) Oil and Gas Information System. Available online at:
22 http://doa.alaska.gov/ogc/publicdb.html.
23 EIA (2011) Coal Production and Preparation Report Shapefile. Available online at: http://www.eia.gov/state/notes-
24 sources.cfmtfmaps.
25 ESRI (2008) ESRI Data & Maps. Redlands, CA: Environmental Systems Research Institute. [CD-ROM],
26 Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and J. Wickham. (2011) Completion of
27 the 2006 National Land Cover Database for the Conterminous United States, PE&RS, Vol. 77(9):858-864.
28 Homer, C., J. Dewitz, J. Fry, M. Coan, N. Hossain, C. Larson, N. Herold, A. McKerrow, J.N. VanDriel and J. Wickham.
29 (2007) Completion of the 2001 National Land Cover Database for the Conterminous United States,
30 Photogrammetric Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
31 Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
32 Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-
33 Representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, v.
34 81, no. 5, p. 345-354.
35 IPCC (2014) 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands.
36 Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds.). Published: IPCC,
37 Switzerland.
70 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 IPCC (2010) Revisiting the use of managed land as a proxy for estimating national anthropogenic emissions and
2 removals. [Eggleston HS, Srivastava N, Tanabe K, Baasansuren J, (eds.)]. Institute for Global Environmental Studies,
3 Intergovernmental Panel on Climate Change, Hayama, Kanagawa, Japan.
4 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
5 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
6 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
7 Jin, S., L. Yang, P. Danielson, C. Homer, J. Fry, and G. Xian. (2013) A comprehensive change detection method for
8 updating the National Land Cover Database to circa 2011. Remote Sensing of Environment, 132:159-175.
9 Nelson, M.D., Riitters, K.H., Coulston, J.W., Domke, G.M., Greenfield, E.J., Langner, LL, Nowak, D.J., O'Dea, C.B.,
10 Oswalt, S.N., Reeves, M.C. and Wear, D.N. (2020) Defining the United States land base: a technical document
11 supporting the USDA Forest Service 2020 RPA assessment. Gen. Tech. Rep. NRS-191., 191, pp.1-70.
12 NOAA Coastal Change Analysis Program (C-CAP) Regional Land Cover Database. Data collected 1995-present
13 Charleston, SC: National Oceanic and Atmospheric Administration (NOAA) Coastal Services Center. Data accessed
14 at: www.csc.noaa.gov/landcover.
15 Nusser, S.M. and J.J. Goebel (1997) "The national resources inventory: a long-term multi-resource monitoring
16 programme." Environmental and Ecological Statistics 4:181-204.
17 Ogle, S.M., G. Domke, W.A. Zurz, M.T. Rocha, T. Huffman, A. Swan, J.E. Smith, C. Woodall, T. Krug (2018)
18 Delineating managed land for reporting greenhouse gas emissions and removals to the United Nations Framework
19 Convention on Climate Change. Carbon Balance and Management 13:9.
20 .U.S. Census Bureau (2010) Topological^ Integrated Geographic Encoding and Referencing (TIGER) system
21 shapefiles. U.S. Census Bureau, Washington, D.C. Available online at: http://www.census.gov/geo/www/tiger.
22 U.S. Department of Agriculture (2015) County Data - Livestock, 1990-2014. U.S. Department of Agriculture,
23 National Agriculture Statistics Service, Washington, D.C.
24 U.S. Department of Agriculture, Forest Service. (2012) Timber Product Output (TPO) Reports. Knoxville, TN: U.S.
25 Department of Agriculture Forest Service, Southern Research Station.
26 U.S. Department of Interior (2005) Federal Lands of the United States. National Atlas of the United States, U.S.
27 Department of the Interior, Washington D.C. Available online at:
28 http://nationalatlas.gov/atlasftp.html?openChapters=chpbound#chpbound.
29 United States Geological Survey (USGS), Gap Analysis Program (2012) Protected Areas Database of the United
30 States (PADUS), version 1.3 Combined Feature Class. November 2012.
31 USGS (2012) Alaska Resource Data File. Available online at: http://ardf.wr.usgs.gov/.
32 USGS (2005) Active Mines and Mineral Processing Plants in the United States in 2003. U.S. Geological Survey,
33 Reston, VA.
34 Yang, L, Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J., Fry, J., Funk, M.,
35 Granneman, B., Liknes, G. C., Rigge, M. & Xian, G. (2018) A new generation of the United States National Land
36 Cover Database: Requirements, research priorities, design, and implementation strategies. ISPRS Journal of
37 Photogrammetry and Remote Sensing 146:108-123.
38 Forest Land Remaining Forest Land: Changes in Forest Carbon
39 Stocks
40 AF&PA (2006a and earlier) Statistical roundup. (Monthly). Washington, D.C. American Forest & Paper Association.
Waste 71
-------
1 AF&PA (2006b and earlier) Statistics of paper, paperboard and wood pulp. Washington, D.C. American Forest &
2 Paper Association.
3 AF&PA (2021) 2020 Statistics - Paper Industry - Annual Summary Data Through 2020. Washington, D.C.: American
4 Forest and Paper Association, 54 p.
5 Amichev, B.Y. and J.M. Galbraith (2004) "A Revised Methodology for Estimation of Forest Soil Carbon from Spatial
6 Soils and Forest Inventory Data Sets." Environmental Management 33(Suppl. 1):S74-S86.
7 Bechtold, W.A.; Patterson, P.L (2005) The enhanced forest inventory and analysis program—national sampling
8 design and estimation procedures. Gen. Tech. Rep. SRS-80. Asheville, NC: U.S. Department of Agriculture Forest
9 Service, Southern Research Station. 85 p.
10 Birdsey, R. (1996) "Carbon Storage for Major Forest Types and Regions in the Conterminous United States." In R.N.
11 Sampson and D. Hair, (eds.). Forest and Global Change, Volume 2: Forest Management Opportunities for
12 Mitigating Carbon Emissions. American Forests. Washington, D.C., 1-26 and 261-379 (appendices 262 and 263).
13 Coulston, J.W., Wear, D.N., and Vose, J.M. (2015) Complex forest dynamics indicate potential for slowing carbon
14 accumulation in the southeastern United States. Scientific Reports. 5: 8002.
15 Domke, G.M., J.E. Smith, and C.W. Woodall. (2011) Accounting for density reduction and structural loss in standing
16 dead trees: Implications for forest biomass and carbon stock estimates in the United States. Carbon Balance and
17 Management. 6:14.
18 Domke, G.M., Woodall, C.W., Smith, J.E., Westfall, J.A., McRoberts, R.E. (2012) Consequences of alternative tree-
19 level biomass estimation procedures on U.S. forest carbon stock estimates. Forest Ecology and Management. 270:
20 108-116.
21 Domke, G.M., Woodall, C.W., Walters, B.F., Smith, J.E. (2013) From models to measurements: comparing down
22 dead wood carbon stock estimates in the U.S. forest inventory. PLoS ONE 8(3): e59949.
23 Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., and Smith, J.E. (2016) A framework for estimating litter
24 carbon stocks in forests of the United States. Science of the Total Environment 557-558: 469-478.
25 Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., Nave, L., Swanston, C. (2017) Toward inventory-based
26 estimates of soil organic carbon in forests of the United States. Ecological Applications. 27(4), 1223-1235.
27 Domke, G.M., Walters, B.F., Smith, J.E., Woodall, C.W. 2022. Chapter 6: FIA Carbon Attributes. In Sampling and
28 estimation documentation for the Enhanced Forest Inventory and Analysis Program: 2022. Westfall, J.A., Coulston,
29 J.W., Moisen, G.G., Andersen, H.-E., eds. Gen. Tech. Rep. NRS-GTR-207, Madison, Wl: U.S. Department of
30 Agriculture, Forest Service, Northern Research Station. 129 p. https://doi.org/10.2737/NRS-GTR-207.
31 EPA (2006) Municipal solid waste in the United States: 2005 Facts and figures. Office of Solid Waste, U.S.
32 Environmental Protection Agency. Washington, D.C. (5306P) EPA530-R-06-011. Available online at:
33 http://www.epa.gov/msw/msw99.htm.
34 FAO (2021) Forest product statistics. Rome, Italy: FAO Forestry Division, fao.org/forestry/statistics/en. Accessed
35 August 16, 2021.
36 Frayer, W.E., and G.M. Furnival (1999) "Forest Survey Sampling Designs: A History." Journal of Forestry 97(12): 4-
37 10.
38 Freed, R. (2004) Open-dump and Landfill timeline spreadsheet (unpublished). ICF International. Washington, D.C.
39 Hair, D. (1958) "Historical forestry statistics of the United States." Statistical Bull. 228. U.S. Department of
40 Agriculture Forest Service, Washington, D.C.
41 Hair. D. and A.H. Ulrich (1963) The Demand and price situation for forest products - 1963. U.S. Department of
42 Agriculture Forest Service, Misc Publication No. 953. Washington, D.C.
72 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Harmon, M.E., C.W. Woodall, B. Fasth, J. Sexton, M. Yatkov. (2011) Differences between standing and downed
2 dead tree wood density reduction factors: A comparison across decay classes and tree species. Res. Paper. NRS-15.
3 Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 40 p.
4 Howard, J. L. and Liang, S. (2019) U.S. timber production, trade, consumption, and price statistics 1965 to 2017.
5 Res. Pap. FPL-RP-701. Madison, Wl: USDA, Forest Service, Forest Products Laboratory.
6 Howard, J. L. and Jones, K.C. (2016) U.S. timber production, trade, consumption, and price statistics 1965 to 2013.
7 Res. Pap. FPL-RP-679. Madison, Wl: USDA, Forest Service, Forest Products Laboratory.
8 Howard, J. L. (2007) U.S. timber production, trade, consumption, and price statistics 1965 to 2005. Res. Pap. FPL-
9 RP-637. Madison, Wl: USDA, Forest Service, Forest Products Laboratory.
10 Howard, J. L. (2003) U.S. timber production, trade, consumption, and price statistics 1965 to 2002. Res. Pap. FPL-
11 RP-615. Madison, Wl: USDA, Forest Service, Forest Products Laboratory. Available online at:
12 http://www.fpl.fs.fed.us/documnts/fplrp/fplrp615/fplrp615.pdf.
13 IPCC (2014) 2013 Supplement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands.
14 [Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M., and Troxler, T.G. (eds.)]. Switzerland.
15 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
16 Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M. Manning, Z. Chen,
17 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press, Cambridge, United Kingdom
18 and New York, NY, USA, 996 pp.
19 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
20 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
21 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
22 Jenkins, J.C., D.C. Chojnacky, L.S. Heath, and R.A. Birdsey (2003) "National-scale biomass estimators for United
23 States tree species." Forest Science 49(l):12-35.
24 Jandl, R., Rodeghiero, M., Martinez, C., Cotrufo, M. F., Bampa, F., van Wesemael, B., Harrison, R.B., Guerrini, I.A.,
25 deB Richter Jr., D., Rustad, L, Lorenz, K., Chabbi, A., Miglietta, F. (2014) Current status, uncertainty and future
26 needs in soil organic carbon monitoring. Science of the Total Environment, 468, 376-383.
27 Johnson, K. Domke, G.M., Russell, M.B., Walters, B.F., Horn, J., Peduzzi, A., Birdsey, R., Dolan, K., Huang, W. (2017)
28 Estimating aboveground live understory vegetation carbon in the United States. Environmental Research Letters.
29 Nelson, M.D., Riitters, K.H., Coulston, J.W., Domke, G.M., Greenfield, E.J., Langner, L.L., Nowak, D.J., O'Dea, C.B.,
30 Oswalt, S.N., Reeves, M.C. and Wear, D.N. (2020) Defining the United States land base: a technical document
31 supporting the USDA Forest Service 2020 RPA assessment. Gen. Tech. Rep. NRS-191., 191, pp.1-70.
32 Ogle, S. M., G. M. Domke, W. A. Kurz, M. T. Rocha, T. Huffman, A. Swan, J. E. Smith, C. W. Woodall, and T. Krug.
33 (2018) Delineating managed land for reporting national greenhouse gas emissions and removals to the United
34 Nations framework convention on climate change. Carbon Balance and Management 13:9.
35 O'Neill, K.P., Amacher, M.C., Perry, C.H. (2005) Soils as an indicator of forest health: a guide to the collection,
36 analysis, and interpretation of soil indicator data in the Forest Inventory and Analysis program. Gen. Tech. Rep. NC-
37 258. St. Paul, MN: U.S. Department of Agriculture, Forest Service, North Central Research Station. 53 p.
38 Oswalt, S.N., Smith, W.B., Miles, P.D. and Pugh, S.A. (2019) Forest resources of the United States, 2017: Atechnical
39 document supporting the Forest Service 2020 RPA Assessment. Gen. Tech. Rep. WO-97. Washington, DC: U.S.
40 Department of Agriculture, Forest Service, Washington Office., 97.
41 Perry, C.H., C.W. Woodall, and M. Schoeneberger (2005) Inventorying trees in agricultural landscapes: towards an
42 accounting of "working trees". In: "Moving Agroforestry into the Mainstream." Proc. 9th N. Am. Agroforestry
43 Conf., Brooks, K.N. and Folliott, P.F. (eds.). 12-15 June 2005, Rochester, MN [CD-ROM], Dept. of Forest Resources,
44 Univ. Minnesota, St. Paul, MN, 12 p. Available online at: http://cinram.umn.edu/afta2005/. (verified 23 Sept 2006).
Waste 73
-------
1 Russell, M.B.; D'Amato, A.W.; Schulz, B.K.; Woodall, C.W.; Domke, G.M.; Bradford, J.B. (2014) Quantifying
2 understory vegetation in the U.S. Lake States: a proposed framework to inform regional forest carbon stocks.
3 Forestry. 87: 629-638.
4 Russell, M.B.; Domke, G.M.; Woodall, C.W.; D'Amato, A.W. (2015) Comparisons of allometric and climate-derived
5 estimates of tree coarse root carbon in forests of the United States. Carbon Balance and Management. 10: 20.
6 Skog, K.E. (2008) Sequestration of carbon in harvested wood products for the United States. Forest Products
7 Journal 58:56-72.
8 Smith, J.E.; Heath, L.S.; Skog, K.E.; Birdsey, R.A. (2006) Methods for calculating forest ecosystem and harvested
9 carbon with standard estimates for forest types of the United States. Gen. Tech. Rep. NE-343. Newtown Square,
10 PA: U.S. Department of Agriculture, Forest Service, Northeastern Research Station. 216 p.
11 Smith, W. B., P. D. Miles, C. H. Perry, and S. A. Pugh (2009) Forest Resources of the United States, 2007. General
12 Technical Report WO-78, U.S. Department of Agriculture Forest Service, Washington Office.
13 Smith, J.E., L.S. Heath, and M.C. Nichols (2010) U.S. Forest Carbon Calculation Tool User's Guide: Forestland Carbon
14 Stocks and Net Annual Stock Change. General Technical Report NRS-13 revised, U.S. Department of Agriculture
15 Forest Service, Northern Research Station, 34 p.
16 Smith, J.E., Domke, G.M. and Woodall, C.W. (2022) Predicting downed woody material carbon stocks in forests of
17 the conterminous United States. Science of the Total Environment, 803, p.150061.
18 Soil Survey Staff (2020a) Gridded National Soil Survey Geographic (gNATSGO) Database for the Conterminous
19 United States. United States Department of Agriculture, Natural Resources Conservation Service. Available online
20 at https://nrcs.app.box.eom/v/soils.
21 Soil Survey Staff (2020b) Gridded National Soil Survey Geographic (gNATSGO) Database for Alaska. United States
22 Department of Agriculture, Natural Resources Conservation Service. Available online at
23 https://nrcs.app.box.eom/v/soils.
24 Steer, Henry B. (1948) Lumber production in the United States. Misc. Pub. 669, U.S. Department of Agriculture
25 Forest Service. Washington, D.C.
26 Ulrich, Alice (1985) U.S. Timber Production, Trade, Consumption, and Price Statistics 1950-1985. Misc. Pub. 1453,
27 U.S. Department of Agriculture Forest Service. Washington, D.C.
28 Ulrich, A.H. (1989) U.S. Timber Production, Trade, Consumption, and Price Statistics, 1950-1987. USDA
29 Miscellaneous Publication No. 1471, U.S. Department of Agriculture Forest Service. Washington, D.C., 77.
30 United Nations Framework Convention on Climate Change (2013) Report on the individual review of the inventory
31 submission of the United States of America submitted in 2012. FCCC/ARR/2012/USA. 42 p.
32 USDA Forest Service (2022a) Forest Inventory and Analysis National Program: Program Features. U.S. Department
33 of Agriculture Forest Service. Washington, D.C. Available online at: https://www.fia.fs.usda.gov/program-
34 features/index.php. Accessed 7 October 2022.
35 USDA Forest Service. (2022b) Forest Inventory and Analysis National Program: FIA Data Mart. U.S. Department of
36 Agriculture Forest Service. Washington, D.C. Available online at:
37 https://apps.fs.usda.gov/fia/datamart/datamart.html. Accessed on 07 October 2022.
38 USDA Forest Service. (2022c) Forest Inventory and Analysis National Program, FIA library: Field Guides, Methods
39 and Procedures. U.S. Department of Agriculture Forest Service. Washington, D.C. Available online at:
40 https://www.fia.fs.usda.gov/library/field-guides-methods-proc/index.php. Accessed on 07 October 2022.
41 USDA Forest Service (2022d) Forest Inventory and Analysis National Program, FIA library: Database
42 Documentation. U.S. Department of Agriculture, Forest Service, Washington Office. Available online at:
43 https://www.fia.fs.usda.gov/library/database-documentation/index.php. Accessed on 07 October 2022.
74 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 U.S. Census Bureau (1976) Historical Statistics of the United States, Colonial Times to 1970, Vol. 1. Washington,
2 D.C.
3 Wear, D.N., Coulston, J.W. (2015) From sink to source: Regional variation in U.S. forest carbon futures. Scientific
4 Reports. 5:16518.
5 Westfall, J.A., Woodall, C.W., Hatfield, M.A. (2013) A statistical power analysis of woody carbon flux from forest
6 inventory data. Climatic Change. 118: 919-931.
7 Woodall, C.W., Coulston, J.W., Domke, G.M., Walters, B.F., Wear, D.N., Smith, J.E., Anderson, H.-E., Clough, B.J.,
8 Cohen, W.B., Griffith, D.M., Hagan, S.C., Hanou, I.S.; Nichols, M.C., Perry, C.H., Russell, M.B., Westfall, J.A., Wilson,
9 B.T. (2015a) The U.S. Forest Carbon Accounting Framework: Stocks and Stock change 1990-2016. Gen. Tech. Rep.
10 NRS-154. Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 49 pp.
11 Woodall, C.W., LS. Heath, G.M. Domke, and M.C. Nichols (2011a) Methods and equations for estimating
12 aboveground volume, biomass, and carbon for trees in the U.S. forest inventory, 2010. Gen. Tech. Rep. NRS-88.
13 Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 30 p.
14 Woodall, C.W., Amacher, M.C., Bechtold, W.A., Coulston, J.W., Jovan, S., Perry, C.H., Randolph, K.C., Schulz, B.K.,
15 Smith, G.C., Tkacz, B., Will-Wolf, S. (2011b) "Status and future of the forest health indicators program of the United
16 States." Environmental Monitoring and Assessment. 177: 419-436.
17 Woodall, C.W., and V.J. Monleon (2008) Sampling protocol, estimation, and analysis procedures for the down
18 woody materials indicator of the FIA program. Gen. Tech. Rep. NRS-22. Newtown Square, PA: U.S. Department of
19 Agriculture, Forest Service, Northern Research Station. 68 p.
20 Woodall, C.W., Walters, B.F., Oswalt, S.N., Domke, G.M., Toney, C., Gray, A.N. (2013) Biomass and carbon
21 attributes of downed woody materials in forests of the United States. Forest Ecology and Management 305: 48-59.
22 Woodall, C.W., Walters, B.F., Coulston, J.W., D'Amato, A.W., Domke, G.M., Russell, M.B., Sowers, P.A. (2015b)
23 Monitoring network confirms land use change is a substantial component of the forest carbon sink in the eastern
24 United States. Scientific Reports. 5:17028.
25 Zhu, Zhiliang, and McGuire, A.D., eds. (2016) Baseline and projected future carbon storage and greenhouse-gas
26 fluxes in ecosystems of Alaska: U.S. Geological Survey Professional Paper 1826,196 p., Available online at:
27 http://dx.doi.org/10.3133/ppl826.
28 Forest Land Remaining Forest Land: Non-CG2 Emissions from
29 Forest Fires
30 Eidenshink, J., Schwind, B., Brewer, K., Zhu, Z.L, Quayle, B. and Howard, S. (2007) A project for monitoring trends
31 in burn severity. Fire ecology, 3(1), pp.3-21.
32 French, N.H.F., W.J. de Groot, L.K. Jenkins, B.M. Rogers, E.C. Alvarado, B. Amiro, B. de Jong, S. Goetz, E. Hoy, E.
33 Hyer, R. Keane, D. McKenzie, S.G. McNulty, B.E Law, R. Ottmar, D.R. Perez-Salicrup, J. Randerson, K.M. Robertson,
34 and M. Turetsky (2011) "Model comparisons for estimating carbon emissions from North American wildland fire."
35 Journal of Geophysical Research 116. 10.1029/2010JG001469
36 French, N.H.F., D. McKenzie, T. Erickson, B. Koziol, M. Billmire, K.A. Endsley, N.K.Y. Scheinerman, L Jenkins, M.E.
37 Miller, R. Ottmar, and S. Prichard (2014) "Modeling regional-scale fire emissions with the Wildland Fire Emissions
38 Information System." Earth Interactions 18, no. 16
39 Giglio, L., Boschetti, L., Roy, D. P., Humber, M. L, and Justice, C. O. (2018) The Collection 6 MODIS burned area
40 mapping algorithm and product. Remote Sensing of Environment, 217, 72-85.
Waste 75
-------
1 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
2 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
3 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
4 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
5 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
6 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
7 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
8 Larkin, N. K., S. Raffuse, and T. T. Strand (2014) Wildland fire emissions, carbon, and climate: U.S. emissions
9 inventories. For. Ecol. Manage. 317:61-69. doi:10.1016/j.foreco.2013.09.012.
10 MTBS Data Access: Fire Level Geospatial Data (2021, April - revised) MTBS Project (USDA Forest Service/U.S.
11 Geological Survey). Available online at: http://mtbs.gov/direct-download. Accessed on 21 April 2021.
12 Ogle, S. M., G. M. Domke, W. A. Kurz, M. T. Rocha, T. Huffman, A. Swan, J. E. Smith, C. W. Woodall, and T. Krug.
13 (2018) Delineating managed land for reporting national greenhouse gas emissions and removals to the United
14 Nations framework convention on climate change. Carbon Balance and Management 13:9.
is Forest Land Remaining Forest Land: N20 Emissions from Soils
16 Albaugh, T.J., Allen, H.L., Fox, T.R. (2007) Historical Patterns of Forest Fertilization in the Southeastern United
17 States from 1969 to 2004. Southern Journal of Applied Forestry, 31,129-137(9).
18 Binkley, D. (2004) Email correspondence regarding the 95 percent confidence interval for area estimates of
19 southern pine plantations receiving N fertilizer (±20%) and the rate applied for areas receiving N fertilizer (100 to
20 200 pounds/acre). Dan Binkley, Department of Forest, Rangeland, and Watershed Stewardship, Colorado State
21 University and Stephen Del Grosso, Natural Resource Ecology Laboratory, Colorado State University. September
22 19,2004.
23 Binkley, D., R. Carter, and H.L Allen (1995) Nitrogen Fertilization Practices in Forestry. In: Nitrogen Fertilization in
24 the Environment, P.E. Bacon (ed.), Marcel Decker, Inc., New York.
25 Briggs, D. (2007) Management Practices on Pacific Northwest West-Side Industrial Forest Lands, 1991-2005: With
26 Projections to 2010. Stand Management Cooperative, SMC Working Paper Number 6, College of Forest Resources,
27 University of Washington, Seattle.
28 Fox, T.R., H. L. Allen, T.J. Albaugh, R. Rubilar, and C.A. Carlson (2007) Tree Nutrition and Forest Fertilization of Pine
29 Plantations in the Southern United States. Southern Journal of Applied Forestry, 31, 5-11.
30 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
31 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
32 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
33 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
34 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
35 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
36 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
37 USDA Forest Service (2001) U.S. Forest Facts and Historical Trends. FS-696. U.S. Department of Agriculture Forest
38 Service, Washington, D.C. Available online at: http://www.fia.fs.fed.us/library/ForestFactslVletric.pdf.
76 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Forest Land Remaining Forest Land: Drained Organic Soils
2 IPCC (2014) 2013 Supplement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands,
3 Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds.). Published: IPCC,
4 Switzerland.
5 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
6 Inventories Programme, The Intergovernmental Panel on Climate Change, H.S. Eggleston, L Buendia, K. Miwa, T.
7 Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
8 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
9 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
10 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
11 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
12 STATSG02 (2016) Soil Survey Staff, Natural Resources Conservation Service, United States Department of
13 Agriculture. U.S. General Soil Map (STATSG02). Available online at https://sdmdataaccess.sc.egov.usda.gov.
14 Accessed 10 November 2016.
15 USDA Forest Service (2022b) Forest Inventory and Analysis National Program: FIA Data Mart. U.S. Department of
16 Agriculture Forest Service. Washington, DC; 2015. Available online at
17 https://apps.fs.usda.gov/fia/datamart/datamart.html. Accessed 30 March 2022.
is Land Converted to Forest Land
19 Birdsey, R. (1996) "Carbon Storage for Major Forest Types and Regions in the Conterminous United States." In R.N.
20 Sampson and D. Hair, (eds.). Forest and Global Change, Volume 2: Forest Management Opportunities for
21 Mitigating Carbon Emissions. American Forests. Washington, D.C., 1-26 and 261-379 (appendices 262 and 263).
22 Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
23 Domke, G.M., J.E. Smith, and C.W. Woodall. (2011) Accounting for density reduction and structural loss in standing
24 dead trees: Implications for forest biomass and carbon stock estimates in the United States. Carbon Balance and
25 Management. 6:14.Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., Nave, L, Swanston, C. (2017) Toward
26 inventory-based estimates of soil organic carbon in forests of the United States. Ecological Applications. 27(4),
27 1223-1235.
28 Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., and Smith, J.E. (2016) A framework for estimating litter
29 carbon stocks in forests of the United States. Science of the Total Environment 557-558: 469-478.
30 Domke, G.M., Woodall, C.W., Walters, B.F., Smith, J.E. (2013) From models to measurements: comparing down
31 dead wood carbon stock estimates in the U.S. forest inventory. PLoS ONE 8(3): e59949.
32 Harmon, M.E., C.W. Woodall, B. Fasth, J. Sexton, M. Yatkov. (2011) Differences between standing and downed
33 dead tree wood density reduction factors: A comparison across decay classes and tree species. Res. Paper. NRS-15.
34 Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 40 p.
35 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
36 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
37 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
38 Jenkins, J.C., D.C. Chojnacky, L.S. Heath, and R.A. Birdsey (2003) "National-scale biomass estimators for United
39 States tree species." Forest Science 49(l):12-35.0gle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003)
40 "Uncertainty in estimating land use and management impacts on soil organic carbon storage for U.S.
41 agroecosystems between 1982 and 1997." Global Change Biology 9:1521-1542.
Waste 77
-------
1 Ogle, S.M., F.J. Breidt, and K. Paustian. (2006) "Bias and variance in model results due to spatial scaling of
2 measurements for parameterization in regional assessments." Global Change Biology 12:516-523.
3 Smith, J.E.; Heath, L.S.; Skog, K.E.; Birdsey, R.A. (2006) Methods for calculating forest ecosystem and harvested
4 carbon with standard estimates for forest types of the United States. Gen. Tech. Rep. NE-343. Newtown Square,
5 PA: U.S. Department of Agriculture, Forest Service, Northeastern Research Station. 216 p.
6 USDA Forest Service (2022b) Forest Inventory and Analysis National Program: FIA Data Mart. U.S. Department of
7 Agriculture Forest Service. Washington, D.C. Available online at:
8 https://apps.fs.usda.gov/fia/datamart/datamart.htmlAccessed on 07 October 2022.
9 USDA Forest Service (2022c) Forest Inventory and Analysis National Program, FIA library: Field Guides, Methods
10 and Procedures. U.S. Department of Agriculture Forest Service. Washington, D.C. Available online at:
11 https://www.fia.fs.usda.gov/library/field-guides-methods-proc/index.php. Accessed on 07 October 2022.
12 USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory, Natural Resources Conservation Service,
13 Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
14 https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd 1422028.pdf.
15 USDA-NRCS (1997) "National Soil Survey Laboratory Characterization Data," Digital Data, Natural Resources
16 Conservation Service, U.S. Department of Agriculture. Lincoln, NE.
17 Woodall, C.W., LS. Heath, G.M. Domke, and M.C. Nichols (2011a) Methods and equations for estimating
18 aboveground volume, biomass, and carbon for trees in the U.S. forest inventory, 2010. Gen. Tech. Rep. NRS-88.
19 Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 30 p.
20 Woodall, C.W., and V.J. Monleon (2008) Sampling protocol, estimation, and analysis procedures for the down
21 woody materials indicator of the FIA program. Gen. Tech. Rep. NRS-22. Newtown Square, PA: U.S. Department of
22 Agriculture, Forest Service, Northern Research Station. 68 p.
23 Woodall, C.W., Walters, B.F., Coulston, J.W., D'Amato, A.W., Domke, G.M., Russell, M.B., Sowers, P.A. (2015b)
24 Monitoring network confirms land use change is a substantial component of the forest carbon sink in the eastern
25 United States. Scientific Reports. 5:17028.
26 Woodall, C.W., Walters, B.F., Oswalt, S.N., Domke, G.M., Toney, C., Gray, A.N. (2013) Biomass and carbon
27 attributes of downed woody materials in forests of the United States. Forest Ecology and Management 305: 48-59.
28 Yang, L, Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J., Fry, J., Funk, M.,
29 Granneman, B., Liknes, G. C., Rigge, M. & Xian, G. (2018) A new generation of the United States National Land
30 Cover Database: Requirements, research priorities, design, and implementation strategies. ISPRS Journal of
31 Photogrammetry and Remote Sensing 146:108-123.
32 Cropland Remaining Cropland
33 Armentano, T. V., and E.S. Menges (1986) Patterns of change in the carbon balance of organic soil-wetlands of the
34 temperate zone. Journal of Ecology 74: 755-774.
35 Brady, N.C. and R.R. Weil (1999) The Nature and Properties of Soils. Prentice Hall. Upper Saddle River, NJ, 881.
36 Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
37 Cheng, B., and D.M. Titterington (1994) "Neural networks: A review from a statistical perspective." Statistical
38 science 9: 2-30.
39 Claassen, R., M. Bowman, J. McFadden, D. Smith, and S. Wallander (2018) Tillage intensity and conservation
40 cropping in the United States, EIB 197. United States Department of Agriculture, Economic Research Service,
41 Washington, D.C.
78 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Conant, R. T., K. Paustian, and E.T. Elliott (2001) "Grassland management and conversion into grassland: effects on
2 soil carbon." Ecological Applications 11: 343-355.
3 CTIC (2004) National Crop Residue Management Survey: 1989-2004. Conservation Technology Information Center,
4 Purdue University, Available online at: http://www.ctic.purdue.edu/CRM/.
5 Daly, C., R.P. Neilson, and D.L Phillips (1994) "A Statistical-Topographic Model for Mapping Climatological
6 Precipitation Over Mountainous Terrain." Journal of Applied Meteorology 33:140-158.
7 Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel (2001) "Simulated
8 Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the DAYCENT Model." In Modeling Carbon
9 and Nitrogen Dynamics for Soil Management, Schaffer, M., L Ma, S. Hansen, (eds.). CRC Press, Boca Raton, Florida,
10 pp. 303-332.
11 Del Grosso, S.J., S.M. Ogle, W.J. Parton (2011) Soil organic matter cycling and greenhouse gas accounting
12 methodologies, Chapter 1, pp 3-13 DOI: 10.1021/bk-2011-1072.ch001. In: Understanding Greenhouse Gas
13 Emissions from Agricultural Management, L. Guo, A. Gunasekara, L. McConnell (eds.). American Chemical Society,
14 Washington, D.C.
15 Edmonds, L, R. L. Kellogg, B. Kintzer, L Knight, C. Lander, J. Lemunyon, D. Meyer, D.C. Moffitt, and J. Schaefer
16 (2003) "Costs associated with development and implementation of Comprehensive Nutrient Management Plans."
17 Part I—Nutrient management, land treatment, manure and wastewater handling and storage, and recordkeeping.
18 Natural Resources Conservation Service, U.S. Department of Agriculture. Available online at:
19 http://www.nrcs.usda.gov/technical/land/pubs/cnmpl.html.
20 EPA (2022) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020. U.S. Environmental Protection
21 Agency, EPA 430-R-22-003. https://www.epa.gov/ghgemissions/draft-inventory-us-greenhouse-gas-emissions-
22 and-sinks-1990-2020.
23 Euliss, N., and R. Gleason (2002) Personal communication regarding wetland restoration factor estimates and
24 restoration activity data. Ned Euliss and Robert Gleason of the U.S. Geological Survey, Jamestown, ND, to Stephen
25 Ogle of the National Resource Ecology Laboratory, Fort Collins, CO. August 2002.
26 Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and Wickham, J. (2011) Completion of the 2006
27 National Land Cover Database for the Conterminous United States, PE&RS, Vol. 77(9):858-864.
28 Griscom, B. W., Adams, J., Ellis, P. W., Houghton, R. A., Lomax, G., Miteva, D. A., Schlesinger, W. H., Shoch, D., Siikamaki, J.
29 V., Smith, P., Woodbury, P., Zganjar, C., Blackman, A., Campari, J., Conant, R. T., Delgado, C., Elias, P., Gopalakrishna, T.,
30 Hamsik, M. R., Herrero, M., Kiesecker, J., Landis, E., Laestadius, L., Leavitt, S. M., Minnemeyer, S., Polasky, S., Potapov, P.,
31 Putz, F. E., Sanderman, J., Silvius, M., Wollenberg, E. & Fargione, J. (2017) "Natural climate solutions." Proceedings of the
32 National Academy of Sciences of the United States of America 114(44): 11645-11650.
33 Hijmans, R.J., S.E. Cameron, J.L. Parra, P.G. Jones and A. Jarvis (2005) Very high resolution interpolated climate
34 surfaces for global land areas. International Journal of Climatology 25: 1965-1978.
35 Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A., VanDriel, J.N., and
36 Wickham, J. (2007) Completion of the 2001 National Land Cover Database for the Conterminous United States.
37 Photogrammetric Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
38 Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
39 Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-
40 Representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, v.
41 81, no. 5, p. 345-354.
42 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
43 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
44 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Waste 79
-------
1 IPCC (2003) Good Practice Guidance for Land Use, Land-Use Change, and Forestry. The Intergovernmental Panel on
2 Climate Change, National Greenhouse Gas Inventories Programme, J. Penman, et al., eds. August 13, 2004.
3 Available online at: http://www.ipcc-nggip.iges.or.ip/public/gpglulucf/gpglulucf.htm.
4 Lai, R., Kimble, J. M., Follett, R. F. & Cole, C. V. (1998) The potential of U.S. cropland to sequester carbon and
5 mitigate the greenhouse effect. Chelsea, Ml: Sleeping Bear Press, Inc.
6 Little, R. (1988) "Missing-data adjustments in large surveys." Journal of Business and Economic Statistics 6: 287-
7 296.
8 McGill, W.B., and C.V. Cole (1981) Comparative aspects of cycling of organic C, N, S and P through soil organic
9 matter. Geoderma 26:267-286.
10 Metherell, A.K., L.A. Harding, C.V. Cole, and W.J. Parton (1993) "CENTURY Soil Organic Matter Model
11 Environment." Agroecosystem version 4.0. Technical documentation, GPSR Tech. Report No. 4, USDA/ARS, Ft.
12 Collins, CO.
13 Mesinger, F., G. DiMego, E. Kalnay, K. Mitchell, P. C. Shafran, W. Ebisuzaki, D. Jovic, J. Woollen, E. Rogers, E. H.
14 Berbery, M. B. Ek, Y. Fan, R. Grumbine, W. Higgins, H. Li, Y. Lin, G. Manikin, D. Parrish, and W. Shi (2006) North
15 American regional reanalysis. Bulletin of the American Meteorological Society 87:343-360.
16 NRCS (1999) Soil Taxonomy: A basic system of soil classification for making and interpreting soil surveys, 2nd
17 Edition. Agricultural Handbook Number 436, Natural Resources Conservation Service, U.S. Department of
18 Agriculture, Washington, D.C.
19 NRCS (1997) "National Soil Survey Laboratory Characterization Data," Digital Data, Natural Resources Conservation
20 Service, U.S. Department of Agriculture. Lincoln, NE.
21 NRCS (1981) Land Resource Regions and Major Land Resource Areas of the United States, USDA Agriculture
22 Handbook 296, United States Department of Agriculture, Natural Resources Conservation Service, National Soil
23 Survey Cente., Lincoln, NE, pp. 156.
24 Ogle, S. M., Alsaker, C., Baldock, J., Bernoux, M., Breidt, F. J., McConkey, B., Regina, K. & Vazquez-Amabile, G. G.
25 (2019) "Climate and Soil Characteristics Determine Where No-Till Management Can Store Carbon in Soils and
26 Mitigate Greenhouse Gas Emissions." Scientific Reports 9(1): 11665.
27 Ogle, S.M., F.J. Breidt, M. Easter, S. Williams, K. Killian, and K. Paustian (2010) "Scale and uncertainty in modeled
28 soil organic carbon stock changes for U.S. croplands using a process-based model." Global Change Biology 16:810-
29 820.
30 Ogle, S.M., F.J. Breidt, M. Easter, S. Williams and K. Paustian (2007) "Empirically-Based Uncertainty Associated with
31 Modeling Carbon Sequestration Rates in Soils." Ecological Modeling 205:453-463.
32 Ogle, S.M., F.J. Breidt, and K. Paustian (2006) "Bias and variance in model results due to spatial scaling of
33 measurements for parameterization in regional assessments." Global Change Biology 12:516-523.
34 Ogle, S. M., et al. (2005) "Agricultural management impacts on soil organic carbon storage under moist and dry
35 climatic conditions of temperate and tropical regions." Biogeochemistry 72: 87-121.
36 Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and management
37 impacts on soil organic carbon storage for U.S. agroecosystems between 1982 and 1997." Global Change Biology
38 9:1521-1542.
39 Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface Submodel: Description
40 and Testing". Glob. Planet. Chang. 19: 35-48.
41 Parton, W.J., D.S. Ojima, C.V. Cole, and D.S. Schimel (1994) "A General Model for Soil Organic Matter Dynamics:
42 Sensitivity to litter chemistry, texture and management," in Quantitative Modeling of Soil Forming Processes.
43 Special Publication 39, Soil Science Society of America, Madison, Wl, 147-167.
80 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Parton, W.J., D.S. Schimel, C.V. Cole, D.S. Ojima (1987) "Analysis of factors controlling soil organic matter levels in
2 Great Plains grasslands." Soil Science Society of America Journal 51:1173-1179.
3 Parton, W.J., J.W.B. Stewart, C.V. Cole. (1988) "Dynamics of C, N, P, and S in grassland soils: a model."
4 Biogeochemistry 5:109-131.
5 Paustian, K., et al. (1997a) "Agricultural soils as a sink to mitigate CO2 emissions." Soil Use and Management 13:
6 230-244.
7 Paustian, K., et al. (1997b) Management controls on soil carbon. In Soil organic matter in temperate
8 agroecosystems: long-term experiments in North America (Paul E.A., K. Paustian, and C.V. Cole, eds.). Boca Raton,
9 CRC Press, pp. 15-49.
10 Potter, C. S., J.T. Randerson, C.B. Fields, P.A. Matson, P.M. Vitousek, H.A. Mooney, and S.A. Klooster (1993)
11 "Terrestrial ecosystem production: a process model based on global satellite and surface data." Global
12 Biogeochemical Cycles 7:811-841.
13 Potter, C., S. Klooster, A. Huete, and V. Genovese (2007) Terrestrial carbon sinks for the United States predicted
14 from MODIS satellite data and ecosystem modeling. Earth Interactions 11, Article No. 13, DOI 10.1175/EI228.1.
15 PRISM Climate Group (2018) PRISM Climate Data, Oregon State University, http://prism.oregonstate.edu,
16 downloaded 18 July 2018.
17 Pukelsheim, F. (1994) 'The 3-Sigma-Rule." American Statistician 48:88-91
18 Soil Survey Staff (2016) State Soil Geographic (STATSGO) Database for State. Natural Resources Conservation
19 Service, United States Department of Agriculture. Available online at:
20 http://www.ncgc.nrcs.usda.gov/products/datasets/statsgo/index.html.
21 Spencer, S., S.M. Ogle, F.J. Breidt, J. Goebel, and K. Paustian. (2011) "Designing a national soil carbon monitoring
22 network to support climate change policy: a case example for US agricultural lands." Greenhouse Gas Management
23 & Measurement 1: 167-178.
24 Towery, D. (2001) Personal Communication. Dan Towery regarding adjustments to the CTIC (1998) tillage data to
25 reflect long-term trends, Conservation Technology Information Center, West Lafayette, IN, and Marlen Eve,
26 National Resource Ecology Laboratory, Fort Collins, CO. February 2001.
27 USDA-ERS (2018) Agricultural Resource Management Survey (ARMS) Farm Financial and Crop Production Practices:
28 Tailored Reports. Available online at: https://www.ers.usda.gov/data-products/arms-farm-financial-and-crop-
29 production-practices/.
30 USDA-ERS (1997) Cropping Practices Survey Data —1995. Economic Research Service, United States Department of
31 Agriculture. Available online at: http://www.ers.usda.gov/data/archive/93018/.
32 USDA-FSA (2015) Conservation Reserve Program Monthly Summary - September 2015. U.S. Department of
33 Agriculture, Farm Service Agency, Washington, D.C. Available online at: https://www.fsa.usda.gov/Assets/USDA-
34 FSA-Public/usdafiles/Conservation/PDF/sep2015summary.pdf.
35 USDA-NASS (2017) 2017 Census of Agriculture. USDA National Agricultural Statistics Service, Complete data
36 available at http://www.nass.usda.gov/AgCensus.
37 USDA-NASS (2012) 2012 Census of Agriculture. USDA National Agricultural Statistics Service, Complete data
38 available at http://www.nass.usda.gov/AgCensus.
39 USDA-NASS (2004) Agricultural Chemical Usage: 2003 Field Crops Summary. Report AgChl(04)a. National
40 Agricultural Statistics Service, U.S. Department of Agriculture, Washington, D.C. Available online at:
41 http://usda.mannlib.cornell.edu/reports/nassr/other/pcu-bb/agcs0504.pdfH.
Waste 81
-------
1 USDA-NASS (1999) Agricultural Chemical Usage: 1998 Field Crops Summary. Report AgCHl(99). National
2 Agricultural Statistics Service, U.S. Department of Agriculture, Washington, DC. Available online at:
3 http://usda.mannlib.cornell.edu/reports/nassr/other/pcu-bb/agch0599.pdf.
4 USDA-NASS (1992) Agricultural Chemical Usage: 1991 Field Crops Summary. Report AgChl(92). National
5 Agricultural Statistics Service, U.S. Department of Agriculture, Washington, D.C. Available online at:
6 http://usda.mannlib.cornell.edu/reports/nassr/other/pcu-bb/agch0392.txtH.
7 USDA-NRCS (2012) Assessment of the Effects of Conservation Practices on Cultivated Cropland in the Upper
8 Mississippi River Basin. U.S. Department of Agriculture, Natural Resources Conservation Service,
9 https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/stelprdb 1042093.pdf.
10 USDA-NRCS (2018a) Summary Report: 2015 National Resources Inventory. Natural Resources Conservation Service,
11 Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
12 https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd 1422028.pdf.
13 USDA-NRCS (2018b) CEAP Cropland Farmer Surveys. USDA Natural Resources Conservation Service.
14 https://www.nrcs.usda.gov/wps/portal/nrcs/detail/national/technical/nra/ceap/na/?cid=nrcsl43 014163.
15 USDA-NRCS (2000) Digital Data and Summary Report: 1997 National Resources Inventory. Revised December 2000.
16 Resources Inventory Division, Natural Resources Conservation Service, United States Department of Agriculture,
17 Beltsville, MD.
18 Van Buuren, S. (2012) "Flexible imputation of missing data." Chapman & Hall/CRC, Boca Raton, FL
19 Yang, L, Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J., Fry, J., Funk, M.,
20 Granneman, B., Liknes, G. C., Rigge, M. &Xian, G. (2018) "A new generation of the United States National Land
21 Cover Database: Requirements, research priorities, design, and implementation strategies." ISPRS Journal of
22 Photogrammetry and Remote Sensing 146:108-123.
23 Zomer RJ, Trabucco A, Bossio DA, van Straaten O, Verchot LV (2008) Climate Change Mitigation: A Spatial Analysis
24 of Global Land Suitability for Clean Development Mechanism Afforestation and Reforestation. Agric. Ecosystems
25 and Envir. 126: 67-80.
26 Zomer RJ, Bossio DA, Trabucco A, Yuanjie L, Gupta DC & Singh VP (2007) Trees and Water: Smallholder
27 Agroforestry on Irrigated Lands in Northern India. Colombo, Sri Lanka: International Water Management Institute.
28 pp 45. (IWMI Research Report 122).
29 Land Converted to Cropland
30 Sampson and D. Hair, (eds.). Forest and Global Change, Volume 2: Forest Management Opportunities for
31 Mitigating Carbon Emissions. American Forests. Washington, D.C., 1-26 and 261-379 (appendices 262 and 263).
32 Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
33 Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel (2001) "Simulated
34 Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the DAYCENT Model." In Modeling Carbon
35 and Nitrogen Dynamics for Soil Management, Schaffer, M., L. Ma, S. Hansen, (eds.). CRC Press, Boca Raton, Florida,
36 pp. 303-332.
37 Del Grosso, S.J., S.M. Ogle, W.J. Parton (2011) "Soil organic matter cycling and greenhouse gas accounting
38 methodologies." Chapter 1, pp 3-13 DOI: 10.1021/bk-2011-1072.ch001. In: Understanding Greenhouse Gas
39 Emissions from Agricultural Management (L. Guo, A. Gunasekara, L. McConnell. Eds.), American Chemical Society,
40 Washington, D.C.
41 Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel (2001) "Simulated
42 Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the DAYCENT Model." In Schaffer, M., L. Ma,
82 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 S. Hansen, (eds.); Modeling Carbon and Nitrogen Dynamics for Soil Management. CRC Press. Boca Raton, Florida.
2 303-332.
3 Domke, G.M., J.E. Smith, and C.W. Woodall. (2011) "Accounting for density reduction and structural loss in
4 standing dead trees: Implications for forest biomass and carbon stock estimates in the United States". Carbon
5 Balance and Management 6:14.
6 Domke, G.M., et al. (2013) "From models to measurements: comparing down dead wood carbon stock estimates in
7 the U.S. forest inventory." PLoS ONE 8(3): e59949.
8 Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., and Smith, J.E. (2016) "A framework for estimating litter
9 carbon stocks in forests of the United States." Science of the Total Environment 557-558: 469-478.
10 Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and Wickham, J. (2011) "Completion of the
11 2006 National Land Cover Database for the Conterminous United States." PE&RS, Vol. 77(9):858-864.
12 Harmon, M.E., C.W. Woodall, B. Fasth, J. Sexton, M. Yatkov. (2011) Differences between standing and downed
13 dead tree wood density reduction factors: A comparison across decay classes and tree species. Res. Paper. NRS-15.
14 Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 40 p.
15 Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A., VanDriel, J.N., and Wickham,
16 J. (2007) "Completion of the 2001 National Land Cover Database for the Conterminous United States."
17 Photogrammetric Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
18 Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
19 Megown, K. (2015) "Completion of the 2011 National Land Cover Database for the conterminous United States-
20 Representing a decade of land cover change information." Photogrammetric Engineering and Remote Sensing 81:
21 345-354.
22 Houghton, R.A., et al. (1983) "Changes in the carbon content of terrestrial biota and soils between 1860 and 1980:
23 a net release of CO2 to the atmosphere." Ecological Monographs 53: 235-262.
24 Houghton, R. A. and Nassikas, A. A. (2017) "Global and regional fluxes of carbon from land use and land cover
25 change 1850-2015." Global Biogeochemical Cycles 31(3): 456-472.
26 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
27 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
28 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
29 Jenkins, J.C., D.C. Chojnacky, L.S. Heath, and R.A. Birdsey (2003) "National-scale biomass estimators for United
30 States tree species." Forest Science 49(l):12-35.
31 Metherell, A.K., L.A. Harding, C.V. Cole, and W.J. Parton (1993) CENTURY Soil Organic Matter Model Environment.
32 Agroecosystem version 4.0. Technical documentation, GPSR Tech. Report No. 4, USDA/ARS, Ft. Collins, CO.
33 Ogle, S.M., F.J. Breidt, M. Easter, S. Williams, K. Killian, and K. Paustian (2010) "Scale and uncertainty in modeled
34 soil organic carbon stock changes for U.S. croplands using a process-based model." Global Change Biology 16:810-
35 820.
36 Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and management
37 impacts on soil organic carbon storage for U.S. agroecosystems between 1982 and 1997." Global Change Biology
38 9:1521-1542.
39 Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface Submodel: Description
40 and Testing". Glob. Planet. Chang. 19: 35-48.
41 Parton, W.J., D.S. Ojima, C.V. Cole, and D.S. Schimel (1994) "A General Model for Soil Organic Matter Dynamics:
42 Sensitivity to litter chemistry, texture and management," in Quantitative Modeling of Soil Forming Processes.
43 Special Publication 39, Soil Science Society of America, Madison, Wl, 147-167.
Waste 83
-------
1 Parton, W.J., D.S. Schimel, C.V. Cole, D.S. Ojima (1987) "Analysis of factors controlling soil organic matter levels in
2 Great Plains grasslands." Soil Science Society of America Journal 51:1173-1179.
3 Parton, W.J., J.W.B. Stewart, C.V. Cole. (1988) "Dynamics of C, N, P, and S in grassland soils: a model."
4 Biogeochemistry 5:109-131.
5 PRISM Climate Group (2018) PRISM Climate Data, Oregon State University, http://prism.oregonstate.edu,
6 downloaded 18 July 2018.
7 Smith, J.E.; Heath, L.S.; Skog, K.E.; Birdsey, R.A. (2006) Methods for calculating forest ecosystem and harvested
8 carbon with standard estimates for forest types of the United States. Gen. Tech. Rep. NE-343. Newtown Square,
9 PA: U.S. Department of Agriculture, Forest Service, Northeastern Research Station. 216 p.
10 Tubiello, F. N., et al. (2015) "The Contribution of Agriculture, Forestry and other Land Use activities to Global
11 Warming, 1990-2012." Global Change Biology 21:2655-2660.
12 USDA Forest Service. (2022) Forest Inventory and Analysis National Program: FIA Data Mart. U.S. Department of
13 Agriculture Forest Service. Washington, D.C. Available online at:
14 https://apps.fs.usda.gov/fia/datamart/datamart.html. Accessed on 07 October 2022.
15 USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory. Natural Resources Conservation Service,
16 Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
17 https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd 1422028.pdf.
18 Woodall, C.W., and V.J. Monleon (2008) Sampling protocol, estimation, and analysis procedures for the down
19 woody materials indicator of the FIA program. Gen. Tech. Rep. NRS-22. Newtown Square, PA: U.S. Department of
20 Agriculture, Forest Service, Northern Research Station. 68 p.
21 Woodall, C.W., LS. Heath, G.M. Domke, and M.C. Nichols (2011) Methods and equations for estimating
22 aboveground volume, biomass, and carbon for trees in the U.S. forest inventory, 2010. Gen. Tech. Rep. NRS-88.
23 Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 30 p.
24 Yang, L, Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J., Fry, J., Funk, M.,
25 Granneman, B., Liknes, G. C., Rigge, M. & Xian, G. (2018) "A new generation of the United States National Land
26 Cover Database: Requirements, research priorities, design, and implementation strategies." ISPRS Journal of
27 Photogrammetry and Remote Sensing 146:108-123.
28 Grassland Remaining Grassland: Soil Carbon Stock Changes
29 Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
30 Del Grosso, S.J., S.M. Ogle, W.J. Parton (2011) Soil organic matter cycling and greenhouse gas accounting
31 methodologies, Chapter 1, pp 3-13 DOI: 10.1021/bk-2011-1072.ch001. In: Understanding Greenhouse Gas
32 Emissions from Agricultural Management (L. Guo, A. Gunasekara, L. McConnell. Eds.), American Chemical Society,
33 Washington, D.C.
34 Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel (2001) "Simulated
35 Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the DAYCENT Model." In Modeling Carbon
36 and Nitrogen Dynamics for Soil Management, Schaffer, M., L. Ma, S. Hansen, (eds.). CRC Press, Boca Raton, Florida,
37 pp. 303-332.
38 Edmonds, L, R. L. Kellogg, B. Kintzer, L. Knight, C. Lander, J. Lemunyon, D. Meyer, D.C. Moffitt, and J. Schaefer
39 (2003) "Costs associated with development and implementation of Comprehensive Nutrient Management Plans."
40 Part I—Nutrient management, land treatment, manure and wastewater handling and storage, and recordkeeping.
41 Natural Resources Conservation Service, U.S. Department of Agriculture. Available online at:
42 http://www.nrcs.usda.gov/technical/land/pubs/cnmpl.html.
84 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 EPA (1999) Biosolids Generation, Use and Disposal in the United States. Office of Solid Waste, U.S. Environmental
2 Protection Agency. Available online at: http://biosolids.policy.net/relatives/18941.PDF.
3 Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L, Barnes, C., Herold, N., and Wickham, J. (2011) Completion of
4 the 2006 National Land Cover Database for the Conterminous United States, PE&RS, Vol. 77(9):858-864.
5 Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A., VanDriel, J.N., and
6 Wickham, J. (2007) Completion of the 2001 National Land Cover Database for the Conterminous United States.
7 Photogrammetric Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
8 Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
9 Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-
10 Representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, v.
11 81, no. 5, p. 345-354.
12 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
13 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
14 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
15 Kellogg, R.L, C.H. Lander, D.C. Moffitt, and N. Gollehon (2000) Manure Nutrients Relative to the Capacity of
16 Cropland and Pastureland to Assimilate Nutrients: Spatial and Temporal Trends for the United States. U.S.
17 Department of Agriculture, Washington, D.C. Publication number nps00-0579.
18 Metherell, A.K., LA. Harding, C.V. Cole, and W.J. Parton (1993) "CENTURYSoil Organic Matter Model
19 Environment." Agroecosystem version 4.0. Technical documentation, GPSRTech. Report No. 4, USDA/ARS, Ft.
20 Collins, CO.
21 NEBRA (2007) A National Biosolids Regulation, Quality, End Use & Disposal Survey. North East Biosolids and
22 Residuals Association. July 21, 2007.
23 Nusser, S.M. and J.J. Goebel (1997) The national resources inventory: a long-term multi-resource monitoring
24 programme. Environmental and Ecological Statistics 4:181-204.
25 Ogle, S.M., F.J. Breidt, M. Easter, S. Williams, K. Killian, and K. Paustian (2010) "Scale and uncertainty in modeled
26 soil organic carbon stock changes for U.S. croplands using a process-based model." Global Change Biology 16:810-
27 820.
28 Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and management
29 impacts on soil organic carbon storage for U.S. agroecosystems between 1982 and 1997." Global Change Biology
30 9:1521-1542.
31 Parton, W.J., D.S. Ojima, C.V. Cole, and D.S. Schimel (1994) "A General Model for Soil Organic Matter Dynamics:
32 Sensitivity to litter chemistry, texture and management," in Quantitative Modeling of Soil Forming Processes.
33 Special Publication 39, Soil Science Society of America, Madison, Wl, 147-167.
34 Parton, W.J., D.S. Schimel, C.V. Cole, D.S. Ojima (1987) "Analysis of factors controlling soil organic matter levels in
35 Great Plains grasslands." Soil Science Society of America Journal 51:1173-1179.
36 Parton, W.J., J.W.B. Stewart, C.V. Cole. (1988) "Dynamics of C, N, P, and S in grassland soils: a model."
37 Biogeochemistry 5:109-131.
38 Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface Submodel: Description
39 and Testing". Glob. Planet. Chang. 19: 35-48.PRISM Climate Group, Oregon State University,
40 http://prism.oregonstate.edu, created 24 July 2015.
41 PRISM Climate Group (2018) PRISM Climate Data, Oregon State University, http://prism.oregonstate.edu,
42 downloaded 18 July 2018.
Waste 85
-------
1 United States Bureau of Land Management (BLM) (2014) Rangeland Inventory, Monitoring, and Evaluation
2 Reports. Bureau of Land Management. U.S. Department of the Interior. Available online at:
3 http://www.blm.gov/wo/st/en/prog/more/rangeland management/rangeland inventory.html.
4 USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory. Natural Resources Conservation Service,
5 Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
6 https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd 1422028.pdf.
7 USDA Forest Service. (2022) Forest Inventory and Analysis National Program: FIA Data Mart. U.S. Department of
8 Agriculture Forest Service. Washington, D.C. Available online at:
9 https://apps.fs.usda.gov/fia/datamart/datamart.html. Accessed on 07 October 2022.
10 Yang, L, Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J., Fry, J., Funk, M.,
11 Granneman, B., Liknes, G. C., Rigge, M. & Xian, G. (2018) "A new generation of the United States National Land
12 Cover Database: Requirements, research priorities, design, and implementation strategies." ISPRS Journal of
13 Photogrammetry and Remote Sensing 146:108-123.
14 Grassland Remaining Grassland: Non-C02 Emissions from
is Grassland Fires
16 Anderson, R.C. Evolution and origin of the Central Grassland of North America: climate, fire and mammalian
17 grazers. Journal of the Torrey Botanical Society 133: 626-647.
18 Andreae, M.O. and P. Merlet (2001) Emission of trace gases and aerosols from biomass burning. Global
19 Biogeochemical Cycles 15:955-966.
20 Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
21 Chapin, F.S., S.F. Trainor, O. Huntington, A.L Lovecraft, E. Zavaleta, D.C. Natcher, A.D. McGuire, J.L Nelson, L. Ray,
22 M. Calef, N. Fresco, H. Huntington, T.S. Rupp, L. DeWilde, and R.L Naylor (2008) Increasing wildfires in Alaska's
23 Boreal Forest: Pathways to potential solutions of a wicked problem. Bioscience 58:531-540.
24 Daubenmire, R. (1968) Ecology of fire in grasslands. Advances in Ecological Research 5:209-266.
25 Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and Wickham, J. (2011) Completion of
26 the 2006 National Land Cover Database for the Conterminous United States, PE&RS, Vol. 77(9):858-864.
27 Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A., VanDriel, J.N., and Wickham,
28 J. (2007) Completion of the 2001 National Land Cover Database for the Conterminous United States.
29 Photogrammetric Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
30 Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
31 Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-
32 Representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, v. 81,
33 no. 5, p. 345-354.
34 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
35 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
36 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
37
38 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
39 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
40 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
41 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
42
86 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Ogle, S.M., S. Spencer, M. Hartman, L Buendia, L. Stevens, D. du Toit, J. Witi (2016) "Developing national baseline
2 GHG emissions and analyzing mitigation potentials for agriculture and forestry using an advanced national GHG
3 inventory software system." In Advances in Agricultural Systems Modeling 6, Synthesis and Modeling of
4 Greenhouse Gas Emissions and Carbon Storage in Agricultural and Forestry Systems to Guide Mitigation and
5 Adaptation, S. Del Grosso, LR. Ahuja and W.J. Parton (eds.), American Society of Agriculture, Crop Society of
6 America and Soil Science Society of America, pp. 129-148.
7 Nusser, S.M. and J.J. Goebel (1997) The national resources inventory: a long-term multi-resource monitoring
8 programme. Environmental and Ecological Statistics 4:181-204.
9 USDA-NRCS (2015) Summary Report: 2012 National Resources Inventory, Natural Resources Conservation Service,
10 Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa. Available
11 online at: http://www.nrcs.usda.gov/lnternet/FSE DOCU ME NTSZnrcseprd396218.pdf.
12 land Converted to Grassland
13 Asner, G.P., Archer, S., Hughes, R.F., Ansley, R.J. and Wessman, C.A. (2003) "Net changes in regional woody
14 vegetation cover and carbon storage in Texas drylands, 1937-1999." Global Change Biology 9(3): 316-335.
15 Birdsey, R. (1996) "Carbon Storage for Major Forest Types and Regions in the Conterminous United States." In R.N.
16 Sampson and D. Hair, (eds.). Forest and Global Change, Volume 2: Forest Management Opportunities for
17 Mitigating Carbon Emissions. American Forests. Washington, D.C., 1-26 and 261-379 (appendices 262 and 263).
18 Breshears, D.D., Knapp, A.K., Law, D.J., Smith, M.D., Twidwell, D. and Wonkka, C.L., 2016. Rangeland Responses to
19 Predicted Increases in Drought Extremity. Rangelands, 38(4), pp. 191-196.
20 Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
21 Del Grosso, S.J., S.M. Ogle, W.J. Parton. (2011) Soil organic matter cycling and greenhouse gas accounting
22 methodologies, Chapter 1, pp 3-13 DOI: 10.1021/bk-2011-1072.ch001. In: Understanding Greenhouse Gas
23 Emissions from Agricultural Management (L. Guo, A. Gunasekara, L. McConnell. Eds.), American Chemical Society,
24 Washington, D.C.
25 Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel (2001) "Simulated
26 Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the DAYCENT Model." In Modeling Carbon
27 and Nitrogen Dynamics for Soil Management (Schaffer, M., L Ma, S. Hansen, (eds.). CRC Press, Boca Raton, Florida,
28 pp. 303-332.
29 Domke, G.M., J.E. Smith, and C.W. Woodall. (2011) Accounting for density reduction and structural loss in standing
30 dead trees: Implications for forest biomass and carbon stock estimates in the United States. Carbon Balance and
31 Management. 6:14.
32 Domke, G.M., et al. (2013) From models to measurements: comparing down dead wood carbon stock estimates in
33 the U.S. forest inventory. PLoS ONE 8(3): e59949.
34 Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., and Smith, J.E. (2016) A framework for estimating litter
35 carbon stocks in forests of the United States. Science of the Total Environment 557-558: 469-478.
36 Epstein, H.E., Gill, R.A., Paruelo, J.M., Lauenroth, W.K., Jia, G.J. and Burke, I.C. (2002) The relative abundance of
37 three plant functional types in temperate grasslands and shrublands of North and South America: effects of
38 projected climate change. Journal of Biogeography, 29(7), pp.875-888.
39 Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L, Barnes, C., Herold, N., and Wickham, J. (2011) Completion of
40 the 2006 National Land Cover Database for the Conterminous United States, PE&RS, Vol. 77(9):858-864.
41 Harmon, M.E., C.W. Woodall, B. Fasth, J. Sexton, M. Yatkov. (2011) Differences between standing and downed
42 dead tree wood density reduction factors: A comparison across decay classes and tree species. Res. Paper. NRS-15.
43 Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 40 p.
Waste 87
-------
1 Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A., VanDriel, J.N., and Wickham,
2 J. (2007) Completion of the 2001 National Land Cover Database for the Conterminous United States.
3 Photogrammetric Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
4 Homer, C.G., Dewitz, J.A., Yang, L, Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
5 Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-
6 Representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, v. 81,
7 no. 5, p. 345-354.
8 Houghton, R.A., et al. (1983) "Changes in the carbon content of terrestrial biota and soils between 1860 and 1980:
9 a net release of CO2 to the atmosphere." Ecological Monographs 53: 235-262.
10 Houghton, R. A. and Nassikas, A. A. (2017) "Global and regional fluxes of carbon from land use and land cover
11 change 1850-2015." Global Biogeochemical Cycles 31(3): 456-472.
12 Huang, C.Y., Asner, G.P., Martin, R.E., Barger, N.N. and Neff, J.C. (2009) "Multiscale analysis of tree cover and
13 aboveground carbon stocks in pinyon-juniper woodlands." Ecological Applications 19(3): 668-681.
14 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
15 Inventories Programme, The Intergovernmental Panel on Climate Change, [H.S. Eggleston, L. Buendia, K. Miwa, T
16 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
17 Jenkins, J.C., D.C. Chojnacky, L.S. Heath, and R.A. Birdsey (2003) "National-scale biomass estimators for United
18 States tree species." Forest Science 49(l):12-35.
19 Jurena, P.N. and Archer, S., (2003) Woody plant establishment and spatial heterogeneity in grasslands. Ecology,
20 84(4), pp.907-919.
21 Lenihan, J.M., Drapek, R., Bachelet, D. and Neilson, R.P., (2003) Climate change effects on vegetation distribution,
22 carbon, and fire in California. Ecological Applications, 13(6), pp.1667-1681.
23 Metherell, A.K., L.A. Harding, C.V. Cole, and W.J. Parton (1993) "CENTURYSoil Organic Matter Model
24 Environment." Agroecosystem version 4.0. Technical documentation, GPSRTech. Report No. 4, USDA/ARS, Ft.
25 Collins, CO.
26 Ogle, S.M., F.J. Breidt, M. Easter, S. Williams, K. Killian, and K. Paustian (2010) "Scale and uncertainty in modeled
27 soil organic carbon stock changes for U.S. croplands using a process-based model." Global Change Biology 16:810-
28 820.
29 Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and management
30 impacts on soil organic carbon storage for U.S. agroecosystems between 1982 and 1997." Global Change Biology
31 9:1521-1542.
32 Parton, W.J., D.S. Ojima, C.V. Cole, and D.S. Schimel (1994) "A General Model for Soil Organic Matter Dynamics:
33 Sensitivity to litter chemistry, texture and management," in Quantitative Modeling of Soil Forming Processes.
34 Special Publication 39, Soil Science Society of America, Madison, Wl, 147-167.
35 Parton, W.J., D.S. Schimel, C.V. Cole, D.S. Ojima (1987) "Analysis of factors controlling soil organic matter levels in
36 Great Plains grasslands." Soil Science Society of America Journal 51:1173-1179.
37 Parton, W.J., J.W.B. Stewart, C.V. Cole (1988) "Dynamics of C, N, P, and S in grassland soils: a model."
38 Biogeochemistry 5:109-131.
39 Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface Submodel: Description
40 and Testing". Glob. Planet. Chang. 19: 35-48.
41 PRISM Climate Group (2018) PRISM Climate Data, Oregon State University, http://prism.oregonstate.edu,
42 downloaded 18 July 2018.
88 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Scholes, R.J. and Archer, S.R. (1997) Tree-grass interactions in savannas 1. Annual review of Ecology and
2 Systematics, 28(1), pp.517-544.
3 Sims, P.L, Singh, J.S. and Lauenroth, W.K. (1978) The structure and function often western North American
4 grasslands: I. Abiotic and vegetational characteristics. The Journal of Ecology, pp.251-285.
5 Smith, J.E.; Heath, L.S.; Skog, K.E.; Birdsey, R.A. (2006) Methods for calculating forest ecosystem and harvested
6 carbon with standard estimates for forest types of the United States. Gen. Tech. Rep. NE-343. Newtown Square,
7 PA: U.S. Department of Agriculture, Forest Service, Northeastern Research Station. 216 p.
8 Tarhouni, M., et al. (2016) Measurement of the aboveground biomass of some rangeland species using a digital
9 non-destructive technique. Botany Letters 163(3):281-287.
10 Tubiello, F. N., et al. (2015) "The Contribution of Agriculture, Forestry and other Land Use activities to Global
11 Warming, 1990-2012." Global Change Biology 21:2655-2660.
12 United States Bureau of Land Management (BLM) (2014) Rangeland Inventory, Monitoring, and Evaluation
13 Reports. Bureau of Land Management. U.S. Department of the Interior. Available online at:
14 http://www.blm.gov/wo/st/en/prog/more/rangeland management/rangeland inventory.html.
15 USDA Forest Service. (2022) Forest Inventory and Analysis National Program: FIA Data Mart. U.S. Department of
16 Agriculture Forest Service. Washington, D.C. Available online at:
17 https://apps.fs.usda.gov/fia/datamart/datamart.html. Accessed on 07 October 2022.
18 USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory. Natural Resources Conservation Service,
19 Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
20 https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd 1422028.pdf.
21 Woodall, C.W., and V.J. Monleon (2008) Sampling protocol, estimation, and analysis procedures for the down
22 woody materials indicator of the FIA program. Gen. Tech. Rep. NRS-22. Newtown Square, PA: U.S. Department of
23 Agriculture, Forest Service, Northern Research Station. 68 p.
24 Woodall, C.W., LS. Heath, G.M. Domke, and M.C. Nichols. (2011) Methods and equations for estimating
25 aboveground volume, biomass, and carbon for trees in the U.S. forest inventory, 2010. Gen. Tech. Rep. NRS-88.
26 Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 30 p.
27 Yang, L, Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J., Fry, J., Funk, M.,
28 Granneman, B., Liknes, G. C., Rigge, M. & Xian, G. (2018) "A new generation of the United States National Land
29 Cover Database: Requirements, research priorities, design, and implementation strategies." ISPRS Journal of
30 Photogrammetry and Remote Sensing 146:108-123.
31 Wetlands Remaining Wetlands: C02, CH4, and N20 Emissions
32 from Peatlands Remaining Peatlands
33 Apodaca, L. (2011) Email correspondence. Lori Apodaca, Peat Commodity Specialist, USGS and Emily Rowan, ICF
34 International. November.
35 Apodaca, L. (2008) E-mail correspondence. Lori Apodaca, Peat Commodity Specialist, USGS and Emily Rowan, ICF
36 International. October and November.
37 Cleary, J., N. Roulet and T.R. Moore (2005) "Greenhouse gas emissions from Canadian peat extraction, 1990-2000:
38 A life-cycle analysis." Ambio 34:456-461.
39 Division of Geological & Geophysical Surveys (DGGS), Alaska Department of Natural Resources (1997-2015)
40 Alaska's Mineral Industry Report (1997-2014). Alaska Department of Natural Resources, Fairbanks, AK. Available
41 online at http://www.dggs.dnr.state.ak.us/pubs/pubs?reqtype=minerals.
Waste 89
-------
1 IPCC (2014) Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth
2 Assessment Report of the Intergovernmental Panel on Climate Change. R.K. Pachauri and LA. Meyer (eds.). IPCC,
3 Geneva, Switzerland.
4 IPCC (2013) 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands.
5 Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds.). Published: IPCC,
6 Switzerland.
7 IPCC (2007) Climate Change 2007: Synthesis Report. Contribution of Working Groups I, II and III to the Fourth
8 Assessment Report (AR4) of the IPCC. The Intergovernmental Panel on Climate Change, R.K. Pachauri, A. Resinger
9 (eds.). Geneva, Switzerland.
10 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
11 Inventories Programme, The Intergovernmental Panel on Climate Change. H.S. Eggleston, L. Buendia, K. Miwa, T.
12 Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
13 Szumigala, D.J. (2011) Phone conversation. Dr. David Szumigala, Division of Geological and Geophysical Surveys,
14 Alaska Department of Natural Resources and Emily Rowan, ICF International. January 18, 2011.
15 Szumigala, D.J. (2008) Phone conversation. Dr. David Szumigala, Division of Geological and Geophysical Surveys,
16 Alaska Department of Natural Resources and Emily Rowan, ICF International. October 17, 2008.
17 USGS (1991-2018) Minerals Yearbook: Peat (1994-2018). United States Geological Survey, Reston, VA. Available
18 online at httpi//rriinerals.usgs.gov/rninerals/pubs/connnoditv/peat/index.html.
19 USGS (2022a) Minerals Yearbook: Peat (2019) Tables-only release. United States Geological Survey, Reston, VA.
20 Available online at https://www.usgs.gov/centers/nmic/peat statistics-arid-information.
21 USGS (2022b) Minerals Yearbook: Peat (2020) Tables-only release. United States Geological Survey, Reston, VA.
22 Available online at https://www.usgs.gov/centers/nmic/peat-statistics-and-information.
23 USGS (2022c) Mineral Commodity Summaries: Peat (1996-2021). United States Geological Survey, Reston, VA.
24 Available online at https://www.usgs.gov/centers/nmic/peat-statistics-and-information.
25 Wetlands Remaining Coastal Wetlands: Emissions and
26 Removals from Coastal Wetlands Remaining Coastal
27 Wetlands
28 Bianchi, T. S., Allison, M. A., Zhao, J., Li, X., Comeaux, R. S., Feagin, R. A., & Kulawardhana, R. W. (2013) Historical
29 reconstruction of mangrove expansion in the Gulf of Mexico: linking climate change with carbon sequestration in
30 coastal wetlands. Estuarine, Coastal and Shelf Science 119: 7-16.
31 Byrd, K. B., Ballanti, L. R., Thomas, N. M., Nguyen, D. K., Holmquist, J. R., Simard, M., Windham-Myers, L., Schile, L.
32 M., Parker, V. T.,... and Castaneda-Moya, E. (2017) Biomass/Remote Sensing dataset: 30m resolution tidal marsh
33 biomass samples and remote sensing data for six regions in the conterminous United States: U.S. Geological Survey
34 data release, https://doi.org/10.5066/F77943K8.
35 Byrd, K. B., Ballanti, L., Thomas, N., Nguyen, D., Holmquist, J.R., Simard, M., and Windham-Myers, L. (2018) A
36 remote sensing-based model of tidal marsh aboveground carbon stocks for the conterminous United States. ISPRS
37 Journal of Photogrammetry and Remote Sensing 139: 255-271.
38 Byrd, K. B., Ballanti, L., Thomas, N., Nguyen, D., Holmquist, J.R., Simard, M., and Windham-Myers, L. (2020)
39 Corrigendum to "A remote sensing-based model of tidal marsh aboveground carbon stocks for the conterminous
40 United States". ISPRS Journal of Photogrammetry and Remote Sensing 166: 63-67.
41 Callaway, J. C., Borgnis, E. L., Turner, R. E. & Milan, C. S. (2012a) Carbon sequestration and sediment accretion in
90 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 San Francisco Bay tidal wetlands. Estuaries and Coasts 35(5): 1163-1181.
2 Callaway, J. C., Borgnis, E. L., Turner, R. E., Milan, C. S., Goodfriend, W., & Richmond, S. (2012b) "Wetland Sediment
3 Accumulation at Corte Madera Marsh and Muzzi Marsh". San Francisco Bay Conservation and Development
4 Commission.
5 Church, T. M., Sommerfield, C. K., Velinsky, D. J., Point, D., Benoit, C., Amouroux, D. & Donard, O. F. X. (2006)
6 Marsh sediments as records of sedimentation, eutrophication and metal pollution in the urban Delaware Estuary.
7 Marine Chemistry 102(1-2): 72-95.
8 Couvillion, B. R., Barras, J. A., Steyer, G. D., Sleavin, W., Fischer, M., Beck, H., & Heckman, D. (2011) Land area
9 change in coastal Louisiana (1932 to 2010) (pp. 1-12). U.S. Department of the Interior, U.S. Geological Survey.
10 Couvillion, B. R., Fischer, M. R., Beck, H. J. and Sleavin, W. J. (2016) Spatial Configuration Trends in Coastal
11 Louisiana from 1986 to 2010. Wetlands 1-13.
12 Craft, C. B., & Richardson, C. J. (1998) Recent and long-term organic soil accretion and nutrient accumulation in the
13 Everglades. Soil Science Society of America Journal 62(3): 834-843.
14 Crooks, S., Findsen, J., Igusky, K., Orr, M. K. and Brew, D. (2009) Greenhouse Gas Mitigation Typology Issues Paper:
15 Tidal Wetlands Restoration. Report by PWA and SAIC to the California Climate Action Reserve.
16 Crooks, S., Rybczyk, J., O'Connell, K., Devier, D. L, Poppe, K., Emmett-Mattox, S. (2014) Coastal Blue Carbon
17 Opportunity Assessment for the Snohomish Estuary: The Climate Benefits of Estuary Restoration. Report by
18 Environmental Science Associates, Western Washington University, EarthCorps, and Restore America's Estuaries.
19 DeLaune, R. D., & White, J. R. (2012) Will coastal wetlands continue to sequester carbon in response to an increase
20 in global sea level?: A case study of the rapidly subsiding Mississippi river deltaic plain. Climatic Change, 110(1),
21 297-314.
22 Holmquist, J. R., Windham-Myers, L, Bliss, N., Crooks, S., Morris, J. T., Megonigal, J. P. & Woodrey, M. (2018)
23 Accuracy and Precision of Tidal Wetland Soil Carbon Mapping in the Conterminous United States. Scientific reports
24 8(1): 9478.
25 Hu, Z., Lee, J. W., Chandran, K., Kim, S. and Khanal, S. K. (2012) N2O Emissions from Aquaculture: A Review.
26 Environmental Science & Technology 46(12): 6470-6480.
27 Hussein, A. H., Rabenhorst, M. C. & Tucker, M. L. (2004) Modeling of carbon sequestration in coastal marsh soils.
28 Soil Science Society of America Journal 68(5): 1786-1795.
29 IPCC (2000) Good Practice Guidance and Uncertainty Management in National Greenhouse Gas Inventories.
30 Quantifying Uncertainties in Practice, Chapter 6. Penman, J., Kruger, D., Galbally, I., Hiraishi, T., Nyenzi, B.,
31 Emmanuel, S., Buendia, L., Hoppaus, R., Martinsen, T., Meijer, J., Miwa, K. and Tanabe, K. (eds). Institute of Global
32 Environmental Strategies (IGES), on behalf of the Intergovernmental Panel on Climate Change (IPCC): Hayama,
33 Japan.
34 IPCC (2003) Good Practice Guidance for Land Use, Land-Use Change and Forestry. LUCF Sector Good Practice
35 Guidance, Chapter 3. Penman, J., Gytarsky, M., Hiraishi, T., Krug, T., Kruger, D., Pipatti, R., Buendia, L., Miwa, K.,
36 Ngara, T., Tanabe, K. and Wagner, F. (eds). Institute of Global Environmental Strategies (IGES), on behalf of the
37 Intergovernmental Panel on Climate Change (IPCC): Hayama, Japan.
38 IPCC (2006) IPCC Guidelines for National Greenhouse Gas Inventories. Prepared by the National Greenhouse Gas
39 Inventories Programme, Eggleston H.S., Buendia L, Miwa K., Ngara T. and Tanabe K. (eds). IGES, Japan.
40 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
41 Assessment Report of the Intergovernmental Panel on Climate Change. Stocker, T., Qin, D., Plattner, G.-K., Tignor,
42 M. Allen, S.K., Boschung, J., Nauels, A., Xia, Y., Bex, V. and Midgley, P.M. (eds.). Cambridge University Press,
43 Cambridge, United Kingdom and New York, NY, USA.
44 IPCC (2014) 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands.
Waste 91
-------
1 Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds.). Published: IPCC,
2 Switzerland.
3 Kearney, M. S. & Stevenson, J. C. (1991) Island land loss and marsh vertical accretion rate evidence for historical
4 sea-level changes in Chesapeake Bay. Journal of Coastal Research 7(2): 403-415.
5 Koster, D., Lichter, J., Lea, P. D., & Nurse, A. (2007) Historical eutrophication in a river-estuary complex in mid-
6 coast Maine. Ecological Applications 17(3): 765-778.
7 Lu, M & Megonigal, J. P. (2017) Final Report for RAE Baseline Assessment Project. Memo to Silvestrum Climate
8 Associates by Smithsonian Environmental Research Center, Maryland.
9 Lynch, J. C. (1989) Sedimentation and nutrient accumulation in mangrove ecosystems of the Gulf of Mexico. M.S.
10 thesis, Univ. of Southwestern Louisiana, Lafayette, LA.
11 Marchio, D. A., Savarese, M., Bovard, B., & Mitsch, W. J. (2016) Carbon sequestration and sedimentation in
12 mangrove swamps influenced by hydrogeomorphic conditions and urbanization in Southwest Florida. Forests 7:
13 116-135.
14 McCombs, J. W., Herold, N. D., Burkhalter, S. G. and Robinson C. J. (2016) Accuracy Assessment of NOAA Coastal
15 Change Analysis Program 2006-2010 Land Cover and Land Cover Change Data. Photogrammetric Engineering &
16 Remote Sensing. 82:711-718.
17 Merrill, J. Z. (1999) Tidal Freshwater Marshes as Nutrient Sinks: particulate Nutrient Burial and Denitrification.
18 Ph.D. Dissertation, University of Maryland, College Park, MD, 342 pp.
19 National Marine Fisheries Service (2022). Fisheries of the United States, 2020. U.S. Department of Commerce,
20 NOAA Current Fishery Statistics No. 2020. Available at: https://www.fisheries.noaa.gov/ national/sustainable-
21 fisheries/fisheries-united-states
22 National Oceanic and Atmospheric Administration, Office for Coastal Management (2020) Coastal Change Analysis
23 Program (C-CAP) Regional Land Cover. Charleston, SC: NOAA Office for Coastal Management. Accessed October
24 2020 at .
25 Noe, G. B., Hupp, C. R., Bernhardt, C. E., & Krauss, K. W. (2016) Contemporary deposition and long-term
26 accumulation of sediment and nutrients by tidal freshwater forested wetlands impacted by sea level rise. Estuaries
27 and Coasts 39(4): 1006-1019.
28 Orson, R. A., Simpson, R. L., & Good, R. E. (1990) Rates of sediment accumulation in a tidal freshwater marsh.
29 Journal of Sedimentary Research 60(6): 859-869.
30 Orson, R., Warren, R. & Niering, W. (1998) Interpreting sea level rise and rates of vertical marsh accretion in a
31 southern New England tidal salt marsh. Estuarine, Coastal and Shelf Science 47(4): 419-429.
32 Roman, C., Peck, J., Allen, J., King, J. & Appleby, P. (1997) Accretion of a New England (USA) salt marsh in response
33 to inlet migration, storms, and sea-level rise. Estuarine, Coastal and Shelf Science 45(6): 717-727.
34 Villa, J. A. & Mitsch W. J. (2015) Carbon sequestration in different wetland plant communities of Southwest Florida.
35 International Journal for Biodiversity Science, Ecosystems Services and Management 11: 17-28
36 Weston, N. B., Neubauer, S. C., Velinsky, D. J., & Vile, M. A. (2014) Net ecosystem carbon exchange and the
37 greenhouse gas balance of tidal marshes along an estuarine salinity gradient. Biogeochemistry 120:163-189.
92 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Wetlands Remaining Wetlands: Flooded Land Remaining
2 Flooded Land
3 Abril, G., Gu'erin, F., Richard, S., Delmas, R., Galy-Lacaux, C., Gosse, P., et al., 2005. Carbon dioxide and methane
4 emissions and the carbon budget of a 10-year old tropical reservoir (Petit Saut, French Guiana). Global
5 Biogeochem. Cycles 19 (GB4007), 1-16. https://doi.org/10.1029/2005GB00245?.
6 Barros, N., Cole, J.J., Tranvik, L.J., Prairie, Y.T., Bastviken, D., Huszar, V.L.M., et al., 2011. Carbon emission from
7 hydroelectric reservoirs linked to reservoir age and latitude. Nat. Geosci. 4 (9), 593-596.
8 https://doi.org/10.1038/ngeol211.
9 Davis, D. W. (1973) Louisiana Canals and Their Influence on Wetland Development. Louisiana State University and
10 Agricultural & Mechanical College. LSU Historical Dissertations and Theses. 2386., Louisiana State University.
11 IPCC (2003) Good Practice Guidance for Land Use, Land-Use Change and Forestry. LUCF Sector Good Practice
12 Guidance, Chapter 3. Penman, J., Gytarsky, M., Hiraishi, T., Krug, T., Kruger, D., Pipatti, R., Buendia, L, Miwa, K.,
13 Ngara, T., Tanabe, K. and Wagner, F. (eds). Institute of Global Environmental Strategies (IGES), on behalf of the
14 Intergovernmental Panel on Climate Change (IPCC): Hayama, Japan.
15 IPCC (2006) IPCC Guidelines for National Greenhouse Gas Inventories. Prepared by the National Greenhouse Gas
16 Inventories Programme, Eggleston H.S., Buendia L, Miwa K., Ngara T. and Tanabe K. (eds). IGES, Japan.
17 IPCC. (2013) 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands.
18 Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds). In: IPCC,
19 Switzerland.
20 IPCC (2019) 2019 Refinement to the 2006 Guidelines for National Greenhouse Gas Inventories. Wetlands, Chapter
21 7. Lovelock, C. E., Evans, C., Barros, N., Prairie, Y. T., Aim, J., Bastviken, D., Beaulieu, J. J., Garneau, M., Harby, A.,
22 Harrison, J. A., Pare, David, Raadal, Hanne Lerche, Sherman, B., Zhang, Chengyi, Ogle, S. M.
23 Lehner B, Reidy Liermann C, Revenga C, Vorosmarty C, Fekete B, Crouzet P, Doll P, et al. (2011b) Global Reservoir
24 and Dam Database, Version 1 (GRanDvl): Dams, Revision 01. In: Palisades, NY: NASA Socioeconomic Data and
25 Applications Center (SEDAC).
26 Prairie, Y. T., et al. (2017) The GHG Reservoir Tool (G-res) User guide. UNESCO/IHA research project on the GHG
27 status of freshwater reservoirs. Joint publication of the UNESCO Chair in Global Environmental Change and the
28 International Hydropower Association: 38.
29 Teodoru, C.R., Bastien, J., Bonneville, M.C., Del Giorgio, P.a., Demarty, M., Garneau, M., et al., 2012. The net
30 carbon footprint of a newly created boreal hydroelectric reservoir. Global Biogeochem. Cycles 26 (GB2016), 1-14.
31 https://doi.org/10.1029/2011GB0Q4187.
32 Land Converted to Wetlands: Emissions and Removals from
33 Land Converted to Vegetated Coastal Wetlands
34 Bianchi, T. S., Allison, M. A., Zhao, J., Li, X., Comeaux, R. S., Feagin, R. A., & Kulawardhana, R. W. (2013) Historical
35 reconstruction of mangrove expansion in the Gulf of Mexico: linking climate change with carbon sequestration in
36 coastal wetlands. Estuarine, Coastal and Shelf Science 119: 7-16.
37 Byrd, K. B., Ballanti, L. R., Thomas, N. M., Nguyen, D. K., Holmquist, J. R., Simard, M., Windham-Myers, L., Schile, L.
38 M., Parker, V. T.,... and Castaneda-Moya, E. (2017) Biomass/Remote Sensing dataset: 30m resolution tidal marsh
39 biomass samples and remote sensing data for six regions in the conterminous United States: U.S. Geological Survey
40 data release, https://doi.org/10.5066/F77943K8.
41 Byrd, K. B., Ballanti, L., Thomas, N., Nguyen, D., Holmquist, J.R., Simard, M., and Windham-Myers, L. (2018) A
Waste 93
-------
1 remote sensing-based model of tidal marsh aboveground carbon stocks for the conterminous United States. ISPRS
2 Journal of Photogrammetry and Remote Sensing 139: 255-271.
3 Byrd, K. B., Ballanti, L., Thomas, N., Nguyen, D., Holmquist, J.R., Simard, M., and Windham-Myers, L. (2020)
4 Corrigendum to "A remote sensing-based model of tidal marsh aboveground carbon stocks for the conterminous
5 United States". ISPRS Journal of Photogrammetry and Remote Sensing 166: 63-67.
6 Callaway, J. C., Borgnis, E. L., Turner, R. E. & Milan, C. S. (2012a) Carbon sequestration and sediment accretion in
7 San Francisco Bay tidal wetlands. Estuaries and Coasts 35(5): 1163-1181.
8 Callaway, J. C., Borgnis, E. L., Turner, R. E., Milan, C. S., Goodfriend, W., & Richmond, S. (2012b) "Wetland Sediment
9 Accumulation at Corte Madera Marsh and Muzzi Marsh". San Francisco Bay Conservation and Development
10 Commission.
11 Church, T. M., Sommerfield, C. K., Velinsky, D. J., Point, D., Benoit, C., Amouroux, D. & Donard, O. F. X. (2006)
12 Marsh sediments as records of sedimentation, eutrophication and metal pollution in the urban Delaware Estuary.
13 Marine Chemistry 102(1-2): 72-95.
14 Craft, C. B., & Richardson, C. J. (1998) Recent and long-term organic soil accretion and nutrient accumulation in the
15 Everglades. Soil Science Society of America Journal 62(3): 834-843.
16 Crooks, S., Rybczyk, J., O'Connell, K., Devier, D.L., Poppe, K., Emmett-Mattox, S. (2014) Coastal Blue Carbon
17 Opportunity Assessment for the Snohomish Estuary: The Climate Benefits of Estuary Restoration. Report by
18 Environmental Science Associates, Western Washington University, EarthCorps, and Restore America's Estuaries.
19 Hussein, A. H., Rabenhorst, M. C. & Tucker, M. L. (2004) Modeling of carbon sequestration in coastal marsh soils.
20 Soil Science Society of America Journal 68(5): 1786-1795.
21 IPCC (2019) Refinement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories, Volume 4:
22 Agriculture, Forestry, and Other Land Use. Calvo Buendia, E., Tanabe K., Kranjc, A., Baasansuren, J., Fukuda, M.,
23 Ngarize, S., Osako, A., Pyrozhenko, Y., Shermanau, P., & Federici, S. (eds). IPCC, Switzerland.
24 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. Prepared by the National Greenhouse
25 Gas Inventories Programme, H.S.Eggleston, L. Buendia, K. Miwa, T. Ngara & K. Tanabe (eds). IGES, Japan.
26 IPCC (2014) 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands.
TJ Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds.). Published: IPCC,
28 Switzerland.
29 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
30 Assessment Report of the Intergovernmental Panel on Climate Change. Stocker, T., Qin, D., Plattner, G.-K., Tignor,
31 M. Allen, S.K., Boschung, J., Nauels, A., Xia, Y., Bex, V. and Midgley, P.M. (eds.). Cambridge University Press,
32 Cambridge, United Kingdom and New York, NY, USA.
33 IPCC (2003) Good Practice Guidance for Land Use, Land-Use Change and Forestry. LUCF Sector Good Practice
34 Guidance, Chapter 3. Penman, J., Gytarsky, M., Hiraishi, T., Krug, T., Kruger, D., Pipatti, R., Buendia, L, Miwa, K.,
35 Ngara, T., Tanabe, K. & F. Wagner (eds). Institute of Global Environmental Strategies (IGES), on behalf of the
36 Intergovernmental Panel on Climate Change (IPCC): Hayama, Japan.
37 IPCC (2000) Good Practice Guidance and Uncertainty Management in National Greenhouse Gas Inventories.
38 Quantifying Uncertainties in Practice, Chapter 6. Penman, J and Kruger, D and Galbally, I and Hiraishi, T and Nyenzi,
39 B and Emmanuel, S and Buendia, L and Hoppaus, R and Martinsen, T and Meijer, J and Miwa, K and Tanabe, K
40 (eds). Institute of Global Environmental Strategies (IGES), on behalf of the Intergovernmental Panel on Climate
41 Change (IPCC): Hayama, Japan.
42 Kearney, M. S. & Stevenson, J. C. (1991) Island land loss and marsh vertical accretion rate evidence for historical
43 sea-level changes in Chesapeake Bay. Journal of Coastal Research 7(2): 403-415.
44 Koster, D., Lichter, J., Lea, P. D., & Nurse, A. (2007) Historical eutrophication in a river-estuary complex in mid-
94 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 coast Maine. Ecological Applications 17(3): 765-778.
2 Lu, M & Megonigal, J.P. (2017) Final Report for RAE Baseline Assessment Project. Memo to Silvestrum Climate
3 Associates by Smithsonian Environmental Research Center, Maryland.
4 Lynch, J. C., Sedimentation and nutrient accumulation in mangrove ecosystems of the Gulf of Mexico, M.S. thesis,
5 Univ. of Southwestern Louisiana, Lafayette, La., 1989.
6 Marchio, D.A., Savarese, M., Bovard, B., & Mitsch, W.J. (2016) Carbon sequestration and sedimentation in
7 mangrove swamps influenced by hydrogeomorphic conditions and urbanization in Southwest Florida. Forests 7:
8 116-135.
9 McCombs, J.W., Herold, N.D., Burkhalter, S.G. and Robinson C.J., (2016) Accuracy Assessment of NOAA Coastal
10 Change Analysis Program 2006-2010 Land Cover and Land Cover Change Data. Photogrammetric Engineering &
11 Remote Sensing. 82:711-718.
12 Merrill, J. Z. (1999) Tidal Freshwater Marshes as Nutrient Sinks: particulate Nutrient Burial and Denitrification.
13 Ph.D. Dissertation, University of Maryland, College Park, MD, 342pp.
14 National Oceanic and Atmospheric Administration, Office for Coastal Management (2020) Coastal Change Analysis
15 Program (C-CAP) Regional Land Cover. Charleston, SC: NOAA Office for Coastal Management. Accessed October
16 2020 at .
17 Noe, G. B., Hupp, C. R., Bernhardt, C. E., & Krauss, K. W. (2016) Contemporary deposition and long-term
18 accumulation of sediment and nutrients by tidal freshwater forested wetlands impacted by sea level rise. Estuaries
19 and Coasts 39(4): 1006-1019.
20 Orson, R. A., Simpson, R. L., & Good, R. E. (1990) Rates of sediment accumulation in a tidal freshwater marsh.
21 Journal of Sedimentary Research 60(6): 859-869.
22 Orson, R., Warren, R. & Niering, W. (1998) Interpreting sea level rise and rates of vertical marsh accretion in a
23 southern New England tidal salt marsh. Estuarine, Coastal and Shelf Science 47(4): 419-429.
24 Roman, C., Peck, J., Allen, J., King, J. & Appleby, P. (1997) Accretion of a New England (USA) salt marsh in response
25 to inlet migration, storms, and sea-level rise. Estuarine, Coastal and Shelf Science 45(6): 717-727.
26 Villa, J. A. & Mitsch W. J. (2015) "Carbon sequestration in different wetland plant communities of Southwest
27 Florida". International Journal for Biodiversity Science, Ecosystems Services and Management 11:17-28.
28 Weston, N. B., Neubauer, S. C., Velinsky, D. J., & Vile, M. A. (2014) Net ecosystem carbon exchange and the
29 greenhouse gas balance of tidal marshes along an estuarine salinity gradient. Biogeochemistry 120:163-189.
30 Land Converted to Wetlands: Land Converted to Flooded Land
31 Abril, G., Gu'erin, F., Richard, S., Delmas, R., Galy-Lacaux, C., Gosse, P., et al., 2005. Carbon dioxide and methane
32 emissions and the carbon budget of a 10-year old tropical reservoir (Petit Saut, French Guiana). Global
33 Biogeochem. Cycles 19 (GB4007), 1-16. https://doi.org/10.1029/2005GB002457.
34 Barros, N., Cole, J.J., Tranvik, L.J., Prairie, Y.T., Bastviken, D., Huszar, V.L.M., et al., 2011. Carbon emission from
35 hydroelectric reservoirs linked to reservoir age and latitude. Nat. Geosci. 4 (9), 593-596.
36 https://doi.org/10.1038/ngeol211.
37
38 IPCC (2003) Good Practice Guidance for Land Use, Land-Use Change and Forestry. LUCF Sector Good Practice
39 Guidance, Chapter 3. Penman, J., Gytarsky, M., Hiraishi, T., Krug, T., Kruger, D., Pipatti, R., Buendia, L, Miwa, K.,
40 Ngara, T., Tanabe, K. and Wagner, F. (eds). Institute of Global Environmental Strategies (IGES), on behalf of the
41 Intergovernmental Panel on Climate Change (IPCC): Hayama, Japan.
42 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. Prepared by the National Greenhouse
Waste 95
-------
1 Gas Inventories Programme, H.S.Eggleston, L Buendia, K. Miwa, T. Ngara & K. Tanabe (eds). IGES, Japan.
2 IPCC. (2013) 2013 Supplement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands.
3 Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds). In: IPCC,
4 Switzerland.
5 IPCC (2019) 2019 Refinement to the 2006 Guidelines for National Greenhouse Gas Inventories. Wetlands, Chapter
6 7. Lovelock, C. E., Evans, C., Barros, N., Prairie, Y. T., Aim, J., Bastviken, D., Beaulieu, J. J., Garneau, M., Harby, A.,
7 Harrison, J. A., Pare, David, Raadal, Hanne Lerche, Sherman, B., Zhang, Chengyi, Ogle, S. M.
8 Lehner B, Reidy Liermann C, Revenga C, Vorosmarty C, Fekete B, Crouzet P, Doll P, et al. (2011b) Global Reservoir
9 and Dam Database, Version 1 (GRanDvl): Dams, Revision 01. In: Palisades, NY: NASA Socioeconomic Data and
10 Applications Center (SEDAC).
11 Prairie, Y. T., et al. (2017) The GHG Reservoir Tool (G-res) User guide. UNESCO/IHA research project on the GHG
12 status of freshwater reservoirs. Joint publication of the UNESCO Chair in Global Environmental Change and the
13 International Hydropower Association: 38.
14 Teodoru, C.R., Bastien, J., Bonneville, M.C., Del Giorgio, P.a., Demarty, M., Garneau, M., et al., 2012. The net
15 carbon footprint of a newly created boreal hydroelectric reservoir. Global Biogeochem. Cycles 26 (GB2016), 1-14.
16 https://doi.org/10.1029/2011GB004187.
17 Settlements Remaining Settlements: Soil Carbon Stock
is Changes
19 AAPFCO (2016 through 2022) Commercial Fertilizers: 2013-2017. Association of American Plant Food Control
20 Officials. University of Missouri. Columbia, MO.
21 Armentano, T. V., and E.S. Menges (1986) Patterns of change in the carbon balance of organic soil-wetlands of the
22 temperate zone. Journal of Ecology 74: 755-774.
23 Brady, N.C. and R.R. Weil (1999) The Nature and Properties of Soils. Prentice Hall. Upper Saddle River, NJ, 881.
24 Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
25 Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L, Barnes, C., Herold, N., and J. Wickham. (2011) Completion of
26 the 2006 National Land Cover Database for the Conterminous United States, PE&RS 77(9):858-864.
27 Homer, C., J. Dewitz, J. Fry, M. Coan, N. Hossain, C. Larson, N. Herold, A. McKerrow, J.N. VanDriel and J. Wickham.
28 (2007) Completion of the 2001 National Land Cover Database for the Conterminous United States.
29 Photogrammetric Engineering and Remote Sensing 73(4): 337-341.
30 Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
31 Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-
32 Representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing
33 81(5):345-354.
34 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
35 Inventories Programme, The Intergovernmental Panel on Climate Change. H.S. Eggleston, L. Buendia, K. Miwa, T.
36 Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
37 NRCS (1999) Soil Taxonomy: A basic system of soil classification for making and interpreting soil surveys, 2nd
38 Edition. Agricultural Handbook Number 436, Natural Resources Conservation Service, U.S. Department of
39 Agriculture, Washington, D.C.
40 Nusser, S.M. and J.J. Goebel (1997) The national resources inventory: a long-term multi-resource monitoring
41 programme. Environmental and Ecological Statistics 4:181-204.
96 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) Uncertainty in estimating land use and management
2 impacts on soil organic carbon storage for U.S. agroecosystems between 1982 and 1997. Global Change Biology
3 9:1521-1542.
4 Soil Survey Staff (2011) State Soil Geographic (STATSGO) Database for State. Natural Resources Conservation
5 Service, United States Department of Agriculture. Available online at:
6 http://www.ncgc.nrcs.usda.gov/products/datasets/statsgo/index.html.
7 USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory, Natural Resources Conservation Service,
8 Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
9 https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd 1422028.pdf.
10 Yang, L, Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J., Fry, J., Funk, M.,
11 Granneman, B., Liknes, G. C., Rigge, M. & Xian, G. (2018) A new generation of the United States National Land
12 Cover Database: Requirements, research priorities, design, and implementation strategies. ISPRS Journal of
13 Photogrammetry and Remote Sensing 146:108-123.
14 Settlements Remaining Settlements: Changes in Carbon Stocks
15 in Settlement Trees
16 deVries, R.E. (1987) A Preliminary Investigation of the Growth and Longevity of Trees in Central Park. M.S. thesis,
17 Rutgers University, New Brunswick, NJ.
18 Fleming, L.E. (1988) Growth Estimation of Street Trees in Central New Jersey. M.S. thesis, Rutgers University, New
19 Brunswick, NJ.
20 Frelich, L.E. (1992) Predicting Dimensional Relationships for Twin Cities Shade Trees. University of Minnesota,
21 Department of Forest Resources, St. Paul, MN, p. 33.
22 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
23 Inventories Programme, The Intergovernmental Panel on Climate Change. H.S. Eggleston, L. Buendia, K. Miwa, T.
24 Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
25 MRLC (2013) National Land Cover Database 2001 (NLCD2001). Available online at:
26 http://www.mrlc.gov/nlcd2001.php. Accessed August 2013.
27 Nowak, D.J. (1986) Silvics of an Urban Tree Species: Norway maple (Acer platanoides L). M.S. thesis, College of
28 Environmental Science and Forestry, State University of New York, Syracuse, NY.
29 Nowak, D.J. (1994) Atmospheric carbon dioxide reduction by Chicago's urban forest. In: Chicago's Urban Forest
30 Ecosystem: Results of the Chicago Urban Forest Climate Project. E.G. McPherson, D.J. Nowak, and R.A. Rowntree
31 (eds.). General Technical Report NE-186. U.S. Department of Agriculture Forest Service, Radnor, PA. pp. 83-94.
32 Nowak, D.J. (2012) Contrasting natural regeneration and tree planting in 14 North American cities. Urban Forestry
33 and Urban Greening. 11: 374-382.
34 Nowak, D.J. and D.E. Crane (2002) Carbon storage and sequestration by urban trees in the United States.
35 Environmental Pollution 116(3):381-389.
36 Nowak, D.J. and E. Greenfield (2010) Evaluating the National Land Cover Database tree canopy and impervious
37 cover estimates across the conterminous United States: A comparison with photo-interpreted estimates.
38 Environmental Management. 46: 378-390.
39 Nowak, D.J. and E.J. Greenfield (2018a) U.S. urban forest statistics, values and projections. Journal of Forestry.
40 116(2): 164-177
Waste 97
-------
1 Nowak, D.J. and E.J. Greenfield (2018b) Declining urban and community tree cover in the United States. Urban
2 Forestry and Urban Greening. 32:32-55.
3 Nowak, D.J., D.E. Crane, J.C. Stevens, and M. Ibarra (2002) Brooklyn's Urban Forest. General Technical Report NE-
4 290. U.S. Department of Agriculture Forest Service, Newtown Square, PA.
5 Nowak, D.J., R.E. Hoehn, D.E. Crane, J.C. Stevens, J.T. Walton, and J. Bond (2008) A ground-based method of
6 assessing urban forest structure and ecosystem services. Arboric. Urb. For. 34(6): 347-358.
7 Nowak, D.J., E.J. Greenfield, R.E. Hoehn, and E. Lapoint (2013) Carbon storage and sequestration by trees in urban
8 and community areas of the United States." Environmental Pollution 178: 229-236.
9 Nowak, D.J. A.R. Bodine, R.E. Hoehn, C.B. Edgar, D.R. Hartel, T.W. Lister, T.J. Brandeis (2016) Austin's Urban Forest,
10 2014. USDA Forest Service, Northern Research Station Resources Bulletin. NRS-100. Newtown Square, PA. 55 p.
11 Nowak, D.J. A.R. Bodine, R.E. Hoehn, C.B. Edgar, G. Riley, D.R. Hartel, K.J. Dooley, S.M. Stanton, M.A. Hatfield, T.J.
12 Brandeis, T.W. Lister (2017) Houston's Urban Forest, 2015. USDA Forest Service, Southern Research Station
13 Resources Bulletin. SRS-211. Newtown Square, PA. 91 p.
14 Smith, W.B. and S.R. Shifley (1984) Diameter Growth, Survival, and Volume Estimates for Trees in Indiana and
15 Illinois. Research Paper NC-257. North Central Forest Experiment Station, U.S. Department of Agriculture Forest
16 Service, St. Paul, MN.
17 U.S. Department of Interior (2018) National Land Cover Database 2011 (NLCD2011). Accessed online August 16,
18 2018. Available online at: https://www.mrlc.gov/nlcdll leg.php.
19 Settlements Remaining Settlements: N20 Emissions from Soils
20 Brakebill, J.W. and Gronberg, J.M. (2017) County-Level Estimates of Nitrogen and Phosphorus from Commercial
21 Fertilizer for the Conterminous United States, 1987-2012. U.S. Geological Survey,
22 https://doi.org/10.5066/F7H41PKX.
23 Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
24 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
25 Inventories Programme, The Intergovernmental Panel on Climate Change. H.S. Eggleston, L. Buendia, K. Miwa, T.
26 Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
27 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
28 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
29 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
30 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
31 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
32 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
33 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom
34 996 pp.
35 Soil Survey Staff (2016) State Soil Geographic (STATSGO) Database for State. Natural Resources Conservation
36 Service, United States Department of Agriculture. Available online at:
37 http://www.ncgc.nrcs.usda.gov/products/datasets/statsgo/index.html.
38 USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory, Natural Resources Conservation Service,
39 Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
40 https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd 1422028.pdf.
98 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Settlements Remaining Settlements: Changes in Yard
2 Trimmings and Food Scrap Carbon Stocks in Landfills
3 Barlaz, M.A. (2008) "Re: Corrections to Previously Published Carbon Storage Factors." Memorandum to Randall
4 Freed, ICF International. February 28, 2008.
5 Barlaz, M.A. (2005) "Decomposition of Leaves in Simulated Landfill." Letter report to Randall Freed, ICF Consulting.
6 June 29, 2005.
7 Barlaz, M.A. (1998) "Carbon Storage during Biodegradation of Municipal Solid Waste Components in Laboratory-
8 Scale Landfills." Global Biogeochemical Cycles 12:373-380.
9 De la Cruz, F.B. and M.A. Barlaz (2010) "Estimation of Waste Component Specific Landfill Decay Rates Using
10 Laboratory-Scale Decomposition Data" Environmental Science & Technology 44:4722- 4728.
11 Eleazer, W.E., W.S. Odle, Y. Wang, and M.A. Barlaz (1997) "Biodegradability of Municipal Solid Waste Components
12 in Laboratory-Scale Landfills." Environmental Science & Technology 31:911-917.
13 EPA (2020) Advancing Sustainable Materials Management: Facts and Figures 2018. U.S. Environmental Protection
14 Agency, Office of Solid Waste and Emergency Response, Washington, D.C. Available online at
15 https://www.epa.gov/smm/advancing-sustainable-materials-management-facts-and-figures-report.
16 EPA (2019) Advancing Sustainable Materials Management: Facts and Figures. U.S. Environmental Protection
17 Agency, Office of Solid Waste and Emergency Response, Washington, D.C. Available online at
18 https://www.epa.gov/smm/advancing-sustainable-materials-management-facts-and-figures-report.
19 EPA (2016) Municipal Solid Waste Generation, Recycling, and Disposal in the United States: Facts and Figures. U.S.
20 Environmental Protection Agency, Office of Solid Waste and Emergency Response, Washington, D.C. Available
21 online at https://archive.epa.gov/epawaste/nonhaz/municipal/web/html/msw99.html.
22 EPA (1995) Compilation of Air Pollutant Emission Factors. U.S. Environmental Protection Agency, Office of Air
23 Quality Planning and Standards, Research Triangle Park, NC. AP-42 Fifth Edition. Available online at
24 http://www3.epa.gov/ttnchiel/ap42/.
25 EPA (1991) Characterization of Municipal Solid Waste in the United States: 1990 Update. U.S. Environmental
26 Protection Agency, Office of Solid Waste and Emergency Response, Washington, D.C. EPA/530-SW-90-042.
27 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
28 Inventories Programme, The Intergovernmental Panel on Climate Change. H.S. Eggleston, L. Buendia, K. Miwa, T.
29 Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
30 IPCC (2003) Good Practice Guidance for Land Use, Land-Use Change, and Forestry. The Intergovernmental Panel on
31 Climate Change, National Greenhouse Gas Inventories Programme, J. Penman et al. (eds.). Available online at
32 http://www.ipcc-nggip.iges.or.ip/public/gpglulucf/gpglulucf.htm.
33 Oshins, C. and D. Block (2000) "Feedstock Composition at Composting Sites." Biocycle 41(9):31-34.
34 Tchobanoglous, G., H. Theisen, and S.A. Vigil (1993) Integrated Solid Waste Management, 1st edition. McGraw-Hill,
35 NY. Cited by Barlaz (1998) "Carbon Storage during Biodegradation of Municipal Solid Waste Components in
36 Laboratory-Scale Landfills." Global Biogeochemical Cycles 12:373-380.
37 Land Converted to Settlements
38 Birdsey, R. (1996) "Carbon Storage for Major Forest Types and Regions in the Conterminous United States." In R.N.
39 Sampson and D. Hair, (eds.). Forest and Global Change, Volume 2: Forest Management Opportunities for
40 Mitigating Carbon Emissions. American Forests. Washington, D.C., 1-26 and 261-379 (appendices 262 and 263).
Waste 99
-------
1 Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer. Domke, G.M.,
2 Perry, C.H., Walters, B.F., Woodall, C.W., and Smith, J.E. (2016) A framework for estimating litter carbon stocks in
3 forests of the United States. Science of the Total Environment 557-558: 469-478.
4 Domke, G.M., J.E. Smith, and C.W. Woodall. (2011) Accounting for density reduction and structural loss in standing
5 dead trees: Implications for forest biomass and carbon stock estimates in the United States. Carbon Balance and
6 Management. 6:14.
7 Domke, G.M., Woodall, C.W., Walters, B.F., Smith, J.E. (2013) From models to measurements: comparing down
8 dead wood carbon stock estimates in the U.S. forest inventory. PLoS ONE 8(3): e59949.
9 Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., and Smith, J.E. (2016) A framework for estimating litter
10 carbon stocks in forests of the United States. Science of the Total Environment 557-558: 469-478.
11 Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L, Barnes, C., Herold, N., and Wickham, J. (2011) Completion of
12 the 2006 National Land Cover Database for the Conterminous United States, PE&RS, Vol. 77(9):858-864.
13 Harmon, M.E., C.W. Woodall, B. Fasth, J. Sexton, M. Yatkov. (2011) Differences between standing and downed
14 dead tree wood density reduction factors: A comparison across decay classes and tree species. Res. Paper. NRS-15.
15 Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 40 p.
16 Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A., VanDriel, J.N., and
17 Wickham, J. (2007) Completion of the 2001 National Land Cover Database for the Conterminous United States.
18 Photogrammetric Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
19 Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
20 Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-
21 Representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, v.
22 81, no. 5, p. 345-354.
23 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
24 Inventories Programme, The Intergovernmental Panel on Climate Change, H.S. Eggleston, L. Buendia, K. Miwa, T
25 Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
26 Jenkins, J.C., D.C. Chojnacky, L.S. Heath, and R.A. Birdsey (2003) "National-scale biomass estimators for United
27 States tree species." Forest Science 49(l):12-35.
28 Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and management
29 impacts on soil organic carbon storage for U.S. agroecosystems between 1982 and 1997." Global Change Biology
30 9:1521-1542.
31 Ogle, S.M., F.J. Breidt, and K. Paustian (2006) "Bias and variance in model results due to spatial scaling of
32 measurements for parameterization in regional assessments." Global Change Biology 12:516-523.
33 Schimel, D.S. (1995) "Terrestrial ecosystems and the carbon cycle." Global Change Biology 1: 77-91.
34 Smith, J.E.; Heath, L.S.; Skog, K.E.; Birdsey, R.A. (2006) Methods for calculating forest ecosystem and harvested
35 carbon with standard estimates for forest types of the United States. Gen. Tech. Rep. NE-343. Newtown Square,
36 PA: U.S. Department of Agriculture, Forest Service, Northeastern Research Station. 216 p.
37 Tubiello, F. N., et al. (2015) "The Contribution of Agriculture, Forestry and other Land Use activities to Global
38 Warming, 1990-2012." Global Change Biology 21:2655-2660.
39 USDA Forest Service. (2022) Forest Inventory and Analysis National Program: FIA Data Mart. U.S. Department of
40 Agriculture Forest Service. Washington, D.C. Available online at:
41 https://apps.fs.usda.gov/fia/datamart/datamart.html. Accessed on 07 October 2022.
42 USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory, Natural Resources Conservation Service,
43 Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
44 https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd 1422028.pdf.
100 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 USDA-NRCS (1997) "National Soil Survey Laboratory Characterization Data," Digital Data, Natural Resources
2 Conservation Service, U.S. Department of Agriculture. Lincoln, NE.
3 Woodall, C.W., LS. Heath, G.M. Domke, and M.C. Nichols. (2011) Methods and equations for estimating
4 aboveground volume, biomass, and carbon for trees in the U.S. forest inventory, 2010. Gen. Tech. Rep. NRS-88.
5 Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 30 p.
6 Woodall, C.W., and V.J. Monleon (2008) Sampling protocol, estimation, and analysis procedures for the down
7 woody materials indicator of the FIA program. Gen. Tech. Rep. NRS-22. Newtown Square, PA: U.S. Department of
8 Agriculture, Forest Service, Northern Research Station. 68 p.
9 Yang, L, Jin, S., Danielson, P., Homer, C., Gass, L, Bender, S. M., Case, A., Costello, C., Dewitz, J., Fry, J., Funk, M.,
10 Granneman, B., Liknes, G. C., Rigge, M. & Xian, G. (2018) A new generation of the United States National Land
11 Cover Database: Requirements, research priorities, design, and implementation strategies. ISPRS Journal of
12 Photogrammetry and Remote Sensing 146:108-123
13 Waste
14 Landfills
15 40 CFR Part 60, Subpart WWW (2005) Standards of Performance for Municipal Solid Waste Landfills, 60.750-
16 60.759, Code of Federal Regulations, Title 40. Available online at: https://www.ecfr.gov/current/title-40/chapter-
17 l/subehapter-C/part-60/subpart-WWW.
18 40 CFR Part 258, Subtitle D of RCRA (2012) Criteria for Municipal Solid Waste Landfills, 258.1—258.75, Code of
19 Federal Regulations, Title 40. Available online at: https://www.ecfr.gov/cgi-bin/text-idx?node=pt40.25.258.
20 ATSDR (2001). Chapter 2: Landfill Gas Basics. In Landfill Gas Primer - An Overview for Environmental Health
21 Professionals. Figure 2-1, pp. 5-6. https://www.atsdr.cdc.gov/HAC/landfill/PDFs/Landfill_2001_ch2mod.pdf.
22 BioCycle (2010) "The State of Garbage in America" By L. Arsova, R. Van Haaren, N. Goldstein, S. Kaufman, and N.
23 Themelis. BioCycle. December 2010. Available online at: https://www.biocycle.net/2010/10/26/the-state-of-
24 garbage-in america-4/.
25 BioCycle (2006) "The State of Garbage in America" By N. Goldstein, S. Kaufman, N. Themelis, and J. Thompson Jr.
26 BioCycle. April 2006. Available online at: https://www.biocycle.net/2006/04/21/the-state-of-garbage-in-america-
27 2/.
28 Bronstein, K., Coburn, J., and R. Schmeltz (2012) "Understanding the EPA's Inventory of U.S. Greenhouse Gas
29 Emissions and Sinks and Mandatory GHG Reporting Program for Landfills: Methodologies, Uncertainties,
30 Improvements and Deferrals." Prepared for the U.S. EPA International Emissions Inventory Conference, August
31 2012, Tampa, Florida. Available online at:
32 https://www3.epa.gov/ttnchiel/conference/ei20/session3/kbronstein.pdf.
33 Business for Social Responsibility (BSR) (2014). Analysis of U.S. Food Waste Among Food Manufacturers, Retailers,
34 and Restaurants. Available online at: http://www.foodwastealliance.org/wp-
35 content/uploads/2014/ll/FWRA BSR Tier3 FINAL.pdf.
36 BSR (2013) Analysis of U.S. Food Waste Among Food Manufacturers, Retailers, and Restaurants. Available online
37 at: http://www.foodwastealliance.org/wp-content/uploads/2013/06/FWRA BSR Tier2 FINAL.pdf.
38 Czepiel, P., B. Mosher, P. Crill, and R. Harriss (1996) "Quantifying the Effect of Oxidation on Landfill Methane
39 Emissions." Journal of Geophysical Research, 101(D11):16721-16730. Dou, Z.; Ferguson, J. D.; Galligan, D. T.; Kelly,
40 A. M.; Finn, S. T.; Giegengack, R. (2016) "Assessing U.S. food wastage and opportunities for reduction." Global Food
41 Security Volume 8, March 2016, Pages 19-26. https://doi.Org/10.1016/i.gfs.2016.02.001.
Waste 101
-------
1 EIA (2007) Voluntary Greenhouse Gas Reports for EIA Form 1605B (Reporting Year 2006). Available online at:
2 https://www. eia.gov/eriviron ment/pdfpages/0608sl2009)index. php.
3 EPA (2022a) Greenhouse Gas Reporting Program (GHGRP). 2021 Amazon S3 Data. Subpart HH: Municipal Solid
4 Waste Landfills and Subpart TT: Industrial Waste Landfills. Accessed on August 13, 2022.
5 EPA (2022b) Landfill Methane Outreach Program (LMOP). 2022 Landfill and Project Level Data. August 2022.
6 Available online at: https://www.epa.gov/lmop/landfill-gas-energy-proiect-data.
7 EPA (2020a) Wasted Food Measurement Methodology Scoping Memo. July 2020. Available online at
8 https://www.epa.gov/sites/production/files/202Q-
9 06/documents/food measurement methodology scoping memo-6-18-20.pdf.
10 EPA (2020b) Advancing Sustainable Materials Management: Facts and Figures 2018. December 2020. Available
11 online at: https://www.epa.gov/sites/production/files/2020-ll/documents/2018 tables and figures fnl 508.pdf.
12 EPA (2020c) Advancing Sustainable Materials Management: Facts and Figures 2016 and 2017. November 2019.
13 Available online at: https://www.epa.gov/sites/default/files/2021-
14 01/documents/2018_tables_and_figures_dec_2020_fnl_508.pdf.
15 EPA (2018) Advancing Sustainable Materials Management: Facts and Figures 2015. July 2018. Available online at:
16 https://www.epa.gov/sites/production/files/2018-
17 07/documents/smm 2015 tables and figures 07252018 fnl 508 O.pdf.
18 EPA (2016a) Industrial and Construction and Demolition Landfills. Available online at:
19 https://www.epa.gov/landfills/industrial-and-construction-and-demolition-cd-landfills.
20 EPA (2016b) Advancing Sustainable Materials Management: Facts and Figures 2014. December 2016. Available
21 online at: https://www.epa.gov/sites/production/files/2016-ll/documents/2014 smm tablesfigures 508.pdf.
22 EPA (2014) Advancing Sustainable Materials Management: Facts and Figures 2014. February 2014. Available online
23 at: https://www.epa.gov/sites/production/files/2015-09/documents/2012 msw dat tbls.pdf.
24 EPA (2008) Compilation of Air Pollution Emission Factors, Publication AP-42, Draft Section 2.4 Municipal Solid
25 Waste Landfills. October 2008.
26 EPA (1993) Anthropogenic Methane Emissions in the United States, Estimates for 1990: Report to Congress, U.S.
27 Environmental Protection Agency, Office of Air and Radiation. Washington, D.C. EPA/430-R-93-003. April 1993.
28 EPA (1988) National Survey of Solid Waste (Municipal) Landfill Facilities, U.S. Environmental Protection Agency.
29 Washington, D.C. EPA/530-SW-88-011. September 1988.
30 EREF (The Environmental Research & Education Foundation) (2016). Municipal Solid Waste Management in the
31 United States: 2010 & 2013.
32 ERG (2021) Production Data Supplied by ERG for 1990-2020 for Pulp and Paper, Fruits and Vegetables, and Meat.
33 June 29, 2021.
34 Food Waste Reduction Alliance (FWRA) (2016) Analysis of U.S. Food Waste Among Food Manufacturers, Retailers,
35 and Restaurants. A joint project by the Food Marketing Institute, the Grocery Manufacturers Association, and the
36 National Restaurant Association. Available online at: https://foodwastealliance.org/wp-
37 content/uploads/2020/05/FWRA-Food-Waste-Survev-2016-Report Final.pdf.
38 IPCC (2019) 2019 Refinement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories. Available online
39 at https://www.ipcc.ch/report/2019-refinement-to-the-2006-ipcc-guidelines-for-national-greenhouse-gas-
40 inventories/.
41 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
42 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
102 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
2 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
3 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
4 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
5 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom
6 996 pp.
7 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
8 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
9 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
10 Mancinelli, R. and C. McKay (1985) "Methane-Oxidizing Bacteria in Sanitary Landfills." Proc. First Symposium on
11 Biotechnical Advances in Processing Municipal Wastes for Fuels and Chemicals, Minneapolis, MN, 437-450. August.
12 RTI (2018a) Methodological changes to the scale-up factor used to estimate emissions from municipal solid waste
13 landfills in the Inventory. Memorandum prepared by K. Bronstein and M. McGrath for R. Schmeltz (EPA). March 22,
14 2018.
15 RTI (2018b) Comparison of industrial waste data reported under Subpart TT and the Solid Waste chapter of the
16 GHG Inventory. Memorandum prepared by K. Bronstein, B. Jackson, and M. McGrath for R, Schmeltz (EPA).
17 October 12, 2018.
18 RTI (2017) Methodological changes to the methane emissions from municipal solid waste landfills as reflected in
19 the public review draft of the 1990-2015 Inventory. Memorandum prepared by K. Bronstein and M. McGrath for R.
20 Schmeltz (EPA). March 31, 2017.
21 RTI (2011) Updated Research on Methane Oxidation in Landfills. Memorandum prepared by K. Weitz (RTI) for R.
22 Schmeltz (EPA). January 14, 2011.
23 Waste Business Journal (WBJ) (2016) Directory of Waste Processing & Disposal Sites 2016.
24 WBJ (2010) Directory of Waste Processing & Disposal Sites 2010.
25 WTO (2017) "China's import ban on solid waste queried at import licensing meeting". World Trade Organization.
26 Published October 3, 2017. Available online at:
27 https://www.wto.org/english/news e/news!7 e/impl 03octl7 e.htro
28 Wastewater Treatment and Discharge
29 AF&PA (2022) "AF&PA Members Achieve Progress on Water Stewardship Goal for 2020." American Forest & Paper
30 Association. Available online at: https://www.afandpa.org/statistics-resources/afpa-members-achieve-progress-
31 water-stewardship-goal-2020. Accessed July 2022.
32 AF&PA (2020) "2020 AF&PA Sustainability Report: Advancing the sustainability of an essential industry." American
33 Forest & Paper Association. Available online at: https://www.afandpa.org/sites/default/files/2021-07/2020 AF-
34 PA-Sustainability-Report.pdf. Accessed June 2021.
35 AF&PA (2018) "2018 AF&PA Sustainability Report: Advancing U.S. Paper and Wood Products Industry Sustainability
36 Performance." American Forest & Paper Association. Available online at: http://sustainability.afandpa.org/wp-
37 content/yploads/2018/06/2018SustainabilityReport PAGES.pdf. Accessed July 2019.
38 AF&PA (2016) "2016 AF&PA Sustainability Report: Advancing U.S. Paper and Wood Products Industry Sustainability
39 Performance." American Forest & Paper Association.
40 AF&PA (2014) "2014 AF&PA Sustainability Report." American Forest & Paper Association.
41 Beecher et al. (2007) "A National Biosolids Regulation, Quality, End Use & Disposal Survey, Preliminary Report."
42 Northeast Biosolids and Residuals Association, April 14, 2007. Available online at:
Waste 103
-------
1 https://staticl.squarespace.eom/static/54806478e4b0dc44el698e88/t/5480c7a2e4b0787f2c73ad81/1417725858
2 575/NtlBiosldsRpt~AppD-FINAL.pdf. Accessed August 2021.
3 Beer Institute (2011) Brewers Almanac. Available online at: http://www.beerinstitute.org/multimedia/brewers-
4 almanac.
5 Benyahia, F., M. Abdulkarim, A. Embaby, and M. Rao. (2006) Refinery Wastewater Treatment: A true Technological
6 Challenge. Presented at the Seventh Annual U.A.E. University Research Conference.
7 BIER (2017) Beverage Industry Environmental Roundtable. 2016 Trends and Observations. Available online at:
8 https://www.bieroundtable.com/benchmarking-coeu. Accessed April 2018.
9 Brewers Association (2021) Statistics: Number of Breweries. Available online at:
10 https://www.brewersassociation.org/statistics-and-data/national-beer-stats/. Accessed August 2021.
11 Brewers Association (2017). 2016 Sustainability Benchmarking Update. Available online at:
12 https://www.brewersassociation.org/best-practices/sustainabilitv/sustainability-benchmarking-tools. Accessed
13 April 2018.
14 Brewers Association (2016a) 2015 Sustainability Benchmarking Report. Available online at:
15 https://www.brewersassociation.org/best-practices/sustainabilitv/sustainability-benchmarking-tools. Accessed
16 March 2018.
17 Brewers Association (2016b) Wastewater Management Guidance Manual. Available online at:
18 https://www.brewersassociation.org/educational-publications/wastewater-management-guidance-manual.
19 Accessed September 2017.
20 Cabrera (2017) "Pulp Mill Wastewater: Characteristics and Treatment." Biological Wastewater Treatment and
21 Resource Recovery. InTech. pp. 119-139.
22 CAST (1995) Council for Agricultural Science and Technology. Waste Management and Utilization in Food
23 Production and Processing. U.S.A. October 1995. ISBN 1-887383-02-6. Available online at: https://www.cast-
24 science.org/publication/waste-management-and-utilization-in-food-production-and-processing/.
25 Climate Action Reserve (CAR) (2011) Landfill Project Protocol V4.0, June 2011. Available online at:
26 http://www.climateactionreserve.org/how/protocols/us-landfill/.
27 Cooper (2018) Email correspondence. Geoff Cooper, Renewable Fuels Association to Kara Edquist, ERG. "Wet Mill
28 vs. Dry Mill Ethanol Production." May 18, 2018.
29 DOE (2013) U.S. Department of Energy Bioenergy Technologies Office. Biofuels Basics. Available online at:
30 http://energy.gov/eere/bioenergv/biofuels-basics. Accessed September 2013.
31 Donovan (1996) Siting an Ethanol Plant in the Northeast. C.T. Donovan Associates, Inc. Report presented to
32 Northeast Regional Biomass Program (NRBP). (April). Available online at:
33 https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.614.856&rep=repl&type=pdf. Accessed October
34 2006.
35 EIA (2022) Energy Information Administration. U.S. Refinery and Blender Net Production of Crude Oil and
36 Petroleum Products (Thousand Barrels). Available online at:
37 https://www.eia.gov/dnav/pet/pet pnp refp dc nus mbbl ro.htro. Accessed July 2022.
38 EPA (2019) Preliminary Effluent Guidelines Program Plan 14. EPA-821-R-19-005. Office of Water, U.S.
39 Environmental Protection Agency. Washington, DC. October 2019. Available online at:
40 https://www.epa.gov/sites/production/files/2019-10/documents/prelim-eg-plar ' ; > ^ I9.pdf. Accessed July
41 2020.
42 EPA (2013) U.S. Environmental Protection Agency. Report on the Performance of Secondary Treatment
43 Technology. EPA-821-R-13-001. Office of Water, U.S. Environmental Protection Agency. Washington, D.C. March
104 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 2013. Available online at: https://www.epa.gov/sites/production/files/2015-
2 11/documents/npdes secondary treatment report march2Q13.pdf.
3 EPA (2012) U.S. Environmental Protection Agency. Clean Watersheds Needs Survey 2012 - Report to Congress. U.S.
4 Environmental Protection Agency, Office of Wastewater Management. Washington, D.C. Available online at:
5 https://www.epa.gOv/cwns/clean-watersheds-needs-survey-cwns-2012-report-and-data#access. Accessed
6 February 2016.
7 EPA (2008) U.S. Environmental Protection Agency. Clean Watersheds Needs Survey 2008 - Report to Congress. U.S.
8 Environmental Protection Agency, Office of Wastewater Management. Washington, D.C. Available online at:
9 https://www.epa.gov/cwns/clean-watersheds-needs-survey-cwns-2008-report-and-data. Accessed December
10 2015.
11 EPA (2004a) U.S. Environmental Protection Agency. Clean Watersheds Needs Survey 2004 - Report to Congress.
12 U.S. Environmental Protection Agency, Office of Wastewater Management. Washington, D.C. Available online at:
13 https://www.epa.gov/cwns/clean-watersheds-needs-syrvev-cwns-report-congress-2004.
14 EPA (2004b) Technical Development Document for the Final Effluent Limitations Guidelines and Standards for the
15 Meat and Poultry Products Point Source Category (40 CFR 432). Office of Water. EPA-821-R-04-011, Washington
16 DC, July.
17 EPA (2002) U.S. Environmental Protection Agency. Development Document for the Proposed Effluent Limitations
18 Guidelines and Standards for the Meat and Poultry Products Industry Point Source Category (40 CFR 432). EPA-
19 821-B-01-007. Office of Water, U.S. Environmental Protection Agency. Washington, D.C. January 2002.
20 EPA (2000) U.S. Environmental Protection Agency. Clean Watersheds Needs Survey 2000 - Report to Congress.
21 Office of Wastewater Management, U.S. Environmental Protection Agency. Washington, D.C. Available online at:
22 https://www.epa.gov/cwns/clean-watersheds-needs-survev-cwns-2000-report-and-data. Accessed July 2007.
23 EPA (1999) U.S. Environmental Protection Agency. Biosolids Generation, Use and Disposal in the United States.
24 Office of Solid Waste and Emergency Response, U.S. Environmental Protection Agency. Washington, D.C. EPA530-
25 R-99-009. September 1999.
26 EPA (1998a) U.S. Environment Protection Agency. How Wastewater Treatment Works...The Basics. EPA F98-002.
27 Office of Water. Washington D.C. May 1998. Available online at: https://www3.epa.gov/npdes/pubs/bastre.pdf.
28 EPA (1998b) U.S. Environmental Protection Agency. "AP-42 Compilation of Air Pollutant Emission Factors." Chapter
29 2.4, Table 2.4-3, page 2.4-13. Available online at: https://www.epa.gov/sites/default/files/2020-
30 10/docuroents/c02s04,pdf.
31 EPA (1997a) U.S. Environmental Protection Agency. Estimates of Global Greenhouse Gas Emissions from Industrial
32 and Domestic Wastewater Treatment. EPA-600/R-97-091. Office of Policy, Planning, and Evaluation, U.S.
33 Environmental Protection Agency. Washington, D.C. September 1997.
34 EPA (1997b) U.S. Environmental Protection Agency. Supplemental Technical Development Document for Effluent
35 Guidelines and Standards (Subparts B & E). EPA-821-R-97-011. Office of Water, U.S. Environmental Protection
36 Agency. Washington, D.C. October 1997.
37 EPA (1996) U.S. Environmental Protection Agency. 1996 Clean Water Needs Survey Report to Congress.
38 Assessment of Needs for Publicly Owned Wastewater Treatment Facilities, Correction of Combined Sewer
39 Overflows, and Management of Storm Water and Nonpoint Source Pollution in the United States. Office of
40 Wastewater Management, U.S. Environmental Protection Agency. Washington, D.C.
41 EPA (1993a) U.S. Environmental Protection Agency, "Anthropogenic Methane Emissions in the U.S.: Estimates for
42 1990, Report to Congress." Office of Air and Radiation, Washington, DC. April 1993.
Waste 105
-------
1 EPA (1993b) U.S. Environmental Protection Agency. Development Document for the Proposed Effluent Limitations
2 Guidelines and Standards for the Pulp, Paper and Paperboard Point Source Category. EPA-821-R-93-019. Office of
3 Water, U.S. Environmental Protection Agency. Washington, D.C. October 1993.
4 EPA (1993c) Standards for the Use and Disposal of Sewage Sludge. 40 CFR Part 503.
5 EPA (1992) U.S. Environmental Protection Agency. Clean Watersheds Needs Survey 1992 - Report to Congress.
6 Office of Wastewater Management, U.S. Environmental Protection Agency. Washington, D.C.
7 EPA (1982) U.S. Environmental Protection Agency. Development Document for Effluent Limitations Guidelines and
8 standards for the Petroleum Refining. United States Environmental Protection Agency, Office of Water. EPA-440/1-
9 82-014. Washington D.C. October 1982.
10 EPA (1975) U.S. Environmental Protection Agency. Development Document for Interim Final and Proposed Effluent
11 Limitations Guidelines and New Source Performance Standards for the Fruits, Vegetables, and Specialties Segment
12 of the Canned and Preserved Fruits and Vegetables Point Source Category. United States Environmental Protection
13 Agency, Office of Water. EPA-440/1-75-046. Washington D.C. October 1975.
14 EPA (1974) U.S. Environmental Protection Agency. Development Document for Effluent Limitations Guidelines and
15 New Source Performance Standards for the Apple, Citrus, and Potato Processing Segment of the Canned and
16 Preserved Fruits and Vegetables Point Source Category. Office of Water, U.S. Environmental Protection Agency,
17 Washington, D.C. EPA-440/l-74-027-a. March 1974.
18 ERG (2022) Draft Memorandum: Improvements to the 1990-2021 Wastewater Treatment and Discharge
19 Greenhouse Gas Inventory. December 2022.
20 ERG (2021a) Revised Memorandum: Improvements to the 1990-2019 Wastewater Treatment and Discharge
21 Greenhouse Gas Inventory. March 2021.
22 ERG (2021b) Draft Memorandum: Improvements to the 1990-2020 Wastewater Treatment and Discharge
23 Greenhouse Gas Inventory. July 2021.
24 ERG (2019). Memorandum: Recommended Improvements to the 1990-2018 Wastewater Greenhouse Gas
25 Inventory. August 2019.
26 ERG (2018a) Memorandum: Updates to Domestic Wastewater BOD Generation per Capita. August 2018.
27 ERG (2018b) Memorandum: Inclusion of Wastewater Treatment Emissions from Breweries. July 2018.
28 ERG (2016) Revised Memorandum: Recommended Improvements to the 1990-2015 Wastewater Greenhouse Gas
29 Inventory. November 2016.
30 ERG (2013a) Memorandum: Revisions to Pulp and Paper Wastewater Inventory. October 2013.
31 ERG (2013b) Memorandum: Revisions to the Petroleum Refinery Wastewater Inventory. October 2013.
32 ERG (2008a) Memorandum: Planned Revisions of the Industrial Wastewater Inventory Emission Estimates for the
33 1990-2007 Inventory. 10 August 2008.
34 ERG (2008b) Memorandum: Estimation of Onsite Treatment at Industrial Facilities and Review of Wastewater
35 Characterization Data. 15 April 2008.
36 ERG (2006a) Memorandum: Recommended Improvements to EPA's Wastewater Inventory for Industrial
37 Wastewater. Prepared for Melissa Weitz, EPA. 11 August 2006.
38 ERG (2006b) Memorandum: Assessment of Greenhouse Gas Emissions from Wastewater Treatment of U.S. Ethanol
39 Production Wastewaters. Prepared for Melissa Weitz, EPA. 10 October 2006.
40 FAO (2022a) FAOSTAT-Forestry Database. Available online at: http://www.fao.Org/faostat/en/#data/FQ. Accessed
41 July 2022.
106 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 FAO (2022b) "Pulp and Paper Capacities Report." United States. From 1998 - 2003, 2000 - 2005, 2001 - 2006,
2 2002 - 2007, 2003 - 2008, 2010 - 2015, 2011 - 2016, 2012 - 2017, 2013 - 2018, 2014 - 2019, 2015 - 2020, 2016 -
3 2021, 2017 - 2022, 2018 - 2023, 2019 - 2024, 2020-2025 reports. Available online at:
4 http://www.fao.org/forestrv/statistics/80571/en/. Accessed July 2022.
5 FAO (2022c) FAOSTAT-Food Balance Sheets. Available online at: http://www.fao.Org/faostat/en/#data/FBS.
6 Accessed August 2022.
7 Foley et al. (2015) N2O and CHa Emission from Wastewater Collection and Treatment Systems: State of the Science
8 Report and Technical Report. GWRC Report Series. IWA Publishing, London, UK.
9 Great Lakes-Upper Mississippi River Board of State and Provincial Public Health and Environmental Managers.
10 (2004) Recommended Standards for Wastewater Facilities (Ten-State Standards).
11 Guisasola et al. (2008) Methane formation in sewer systems. Water Research 42(6-7): 1421-1430.
12 Instituto de Estadisticas de Puerto Rico. (2021). Population of Puerto Rico from 1990-1999 from "Estimados
13 anuales poblacionales de los municipios desde 1950." Accessed February 2021. Available online at:
14 https://censo.estadisticas.pr/EstimadosPoblacionales
15 IPCC (2022) Emission factor database: Emission Factor Detail (ID:625621). The Intergovernmental Panel on Climate
16 Change. Available online at: https://www.ipcc-nggip.iges.or.jp/EFDB/ef_detail.php
17 IPCC (2019) 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
18 Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change. [CalvoBuendia, E.,
19 Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M., Ngarize S., Osako, A., Pyrozhenko, Y., Shermanau, P. and
20 Federici, S. (eds)]. Switzerland.
21 IPCC (2014) 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands.
22 [Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds.)]. Published:
23 IPCC, Switzerland.
24 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
25 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
26 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
27 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
28 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
29 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
30 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom
31 996 pp.
32 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
33 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
34 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
35 Kenari et. al (2010) An Investigation on the Nitrogen Content of a Petroleum Refinery Wastewater and its Removal
36 by Biological Treatment. Journal of Environmental Health, Sciences, and Engineering. 7(1): 391-394.Leverenz, H.L,
37 G. Tchobanoglous, and J.L. Darby (2010) "Evaluation of Greenhouse Gas Emissions from Septic Systems." Water
38 Environment Research Foundation. Alexandria, VA.
39 Lewis, A. (2019) Email correspondence. Ann Lewis, RFA to Kara Edquist, ERG. "Wet Mill vs Dry Mill Ethanol
40 Production." August 20, 2019.
41 Malmberg, B. (2018) Draft Pulp and Paper Information for Revision of EPA Inventory of U.S. Greenhouse Gas
42 Emissions and Sinks, Waste Chapter. National Council for Air and Stream Improvement, Inc. Prepared for Rachel
43 Schmeltz, EPA. June 13, 2018.
44 McFarland (2001) Biosolids Engineering, New York: McGraw-Hill, p. 2.12.
Waste 107
-------
1 Merrick (1998) Wastewater Treatment Options for the Biomass-to-Ethanol Process. Report presented to National
2 Renewable Energy Laboratory (NREL). Merrick & Company. Subcontract No. AXE-8-18020-01. October 22,1998.
3 Metcalf & Eddy, Inc. (2014) Wastewater Engineering: Treatment and Resource Recovery, 5th ed. McGraw Hill
4 Publishing.
5 Metcalf & Eddy, Inc. (2003) Wastewater Engineering: Treatment, Disposal and Reuse, 4th ed. McGraw Hill
6 Publishing.
7 NEBRA (2022) "U.S. National Biosolids Data." Northeast Biosolids and Residuals Associations. Available online at:
8 https://staticl.squarespace.eom/static/601837dlc67bcc4elbll862f/t/62f4f5fbae32804dd9f51ef6/166022092535
9 6/National BiosolidsDataSummary NBDP 20220811.pdf
10 Nemerow, N.L and A. Dasgupta (1991) Industrial and Hazardous Waste Treatment. Van Nostrand Reinhold. NY.
11 ISBN 0-442-31934-7.
12 NRBP (2001) Northeast Regional Biomass Program. An Ethanol Production Guidebook for Northeast States.
13 Washington, D.C. (May 3).
14 Rendleman, C.M. and Shapouri, H. (2007) New Technologies in Ethanol Production. USDA Agricultural Economic
15 Report Number 842.
16 RFA (2022a) Renewable Fuels Association. Annual U.S. Fuel Ethanol Production. Available online at:
17 https://ethanolrfa.org/statistics/annual-ethanol-production. Accessed July 2022.
18 RFA (2022b) Renewable Fuels Association. Monthly Grain Use for U.S. Ethanol Production Report. Available online
19 at: https://ethanolrfa.org/statistics/feedstock-use-co-product-output. Accessed July 2021.
20 Ruocco (2006a) Email correspondence. Dr. Joe Ruocco, Phoenix Bio-Systems to Sarah Holman, ERG. "Capacity of
21 Bio-Methanators (Dry Milling)." October 6, 2006.
22 Ruocco (2006b) Email correspondence. Dr. Joe Ruocco, Phoenix Bio-Systems to Sarah Holman, ERG. "Capacity of
23 Bio-Methanators (Wet Milling)." October 16, 2006.
24 Short et al. (2017) Dissolved Methane in the Influent of Three Australian Wastewater Treatment Plants Fed by
25 Gravity Sewers. Sci Total Environ 599-600: 85-93.
26 Short et al. (2014) Municipal Gravity Sewers: an Unrecognised Source of Nitrous Oxide. Sci Total Environ 468-469:
27 211-218.
28 Stier, J. (2018) Personal communications between John Stier, Brewers Association Sustainability Mentor and Amie
29 Aguiar, ERG. Multiple dates.
30 Sullivan (SCS Engineers) (2010) The Importance of Landfill Gas Capture and Utilization in the U.S. Presented to
31 SWICS, April 6, 2010. Available online at: https://www.scsengineers.com/scs-white-papers/the-importance-of-
32 landfill-gas-capture arid-utilization in -the-u-s/.
33 Sullivan (SCS Engineers) (2007) Current MSW Industry Position and State of the Practice on Methane Destruction
34 Efficiency in Flares, Turbines, and Engines. Presented to Solid Waste Industry for Climate Solutions (SWICS). July
35 2007. Available online at: https://www.scsengineers.com/wp-
36 content/uploads/2015/03/Sullivan LFG Destruction Efficiency White Paper.pdf.
37 TTB (2022) Alcohol and Tobacco Tax and Trade Bureau. Beer Statistics. Available online at:
38 https://www.ttb.gov/beer/beer-stats.shtml. Accessed July 2021.
39 UNFCCC (2012) CDM Methodological tool, Project emissions from flaring (Version 02.0.0). EB 68 Report. Annex 15.
40 Available online at: http://cdm.unfccc.int/methodologies/PAmethodologies/tools/am-tool-06-vl.pdf/history view.
41 U.S. Census Bureau (2022) International Database. Available online at: https://www.census.gov/data-
42 tools/demo/idb/#/trends?YR ANIM=2020&dashPages=DASH&FIP5 SINGLE=US&COUNTRY YEAR=2020&menu=tre
108 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 ndsViz&TREND RANGE=1990,2021&TREND STEP=1&TREND ADD YRS=&FIPS=AQ,GQ,CQ,RQ,VQ&measures=POP.
2 Accessed July 2022.
3 U.S. Census Bureau (2021a). Annual Estimates of the Resident Population for the United States, Regions, States,
4 and Puerto Rico: April 1, 2010 to July 1,2020. Available online at: https://www.census.gov/data/tables/time-
5 series/demo/popest/2010s-state-total.html
6 U.S. Census Bureau (2021b). International Database. Available online at:
7 https://www.census.gov/data/tables/time-series/demo/popest/2020s-national-total.html. Accessed July 2022.
8 U.S. Census Bureau (2019) "American Housing Survey." Table 1A-4: Selected Equipment and Plumbing-All Housing
9 Units. From 1989,1991, 1993, 1995, 1997, 1999, 2001, 2003, 2005, 2007, and 2009 reports. Table C-04-AO
10 Plumbing, Water, and Sewage Disposal-All Occupied Units. From 2011, 2013, 2015, 2017, and 2019 reports.
11 Available online at http://www.census.gov/programs-surveys/ahs/data.html. Accessed August 2021.
12 U.S. Census Bureau (2013) "American Housing Survey." Table 1A-4: Selected Equipment and Plumbing-All Housing
13 Units. From 1989,1991, 1993, 1995, 1997, 1999, 2001, 2003, 2005, 2007, and 2009 reports. Table C-04-AO
14 Plumbing, Water, and Sewage Disposal-All Occupied Units. From 2011, and 2013 reports. Available online at
15 http://www.census.gov/programs-surveys/ahs/data.html. Accessed May 2020.
16 U.S. Census Bureau, Population Division. (2011). Table 1. Intercensal Estimates of the Resident Population for the
17 United States, Regions, States, and Puerto Rico: April 1, 2000 to July 1, 2010 (ST-EST00INT-01), Release Date:
18 September 2011. Available online at: https://www2.census.gov/programs-survevs/popest/datasets/2QOO-
19 2010/intercensal/state/st-est00int- alldata.csv
20 U.S. Census Bureau, Population Division (2002). Table CO-EST2001-12-00 - Time Series of Intercensal State
21 Population Estimates: April 1,1990 to April 1, 2000.Available online at: https://www2.census.gov/programs-
22 survevs/popest/tables/1990-2000/intercensal/st- co/co-est2001-12-00.pdf
23 USDA (2022a) U.S. Department of Agriculture. National Agricultural Statistics Service. Washington, D.C. Available
24 online at: https://www.nass.usda.gov/Publications/ and https://quickstats.nass.usda.gov/. Accessed September
25 2022
26 USDA (2022b) U.S. Department of Agriculture. National Agricultural Statistics Service. Vegetables 2021 Summary.
27 Available online at: https://usda.librarv.cornell.edu/concern/publications/02870v86p?locale=en. Accessed
28 September 2022.
29 USDA (2021) U.S. Department of Agriculture. Economic Research Service. Nutrient Availability. Washington D.C.
30 Available online at: https://www.ers.usda.gov/data-products/food-availabilitv-per-capita-data-system/food-
31 availability-per-capita-data-system. Accessed June 2021.
32 U.S. Poultry (2006) Email correspondence. John Starkey, USPOULTRY to D. Bartram, ERG. 30 August 2006.
33 White and Johnson (2003) White, P.J. and Johnson, LA. Editors. Corn: Chemistry and Technology. 2nd ed. AACC
34 Monograph Series. American Association of Cereal Chemists. St. Paul, MN.
35 World Bank (1999) Pollution Prevention and Abatement Handbook 1998, Toward Cleaner Production. The
36 International Bank for Reconstruction and Development/The WORLDBANK. 1818 H Street, N.W. Washington, DC.
37 20433, USA. ISBN 0-8213-3638-X.
38 Composting
39 BioCycle (2018a) Organic Waste Bans and Recycling Laws to Tackle Food Waste. Prepared by E. Broad Lieb, K.
40 Sandson, L. Macaluso, and C. Mansell. Available online at: https://www.biocycle.net/2018/09/ll/organic-waste-
41 bans-recycling-laws-tackle-food-waste/.
Waste 109
-------
1 BioCycle (2018b). State Food Waste Recycling Data Collection, Reporting Analysis. Prepared by Nora Goldstein.
2 Available online at: http://compostcolab.wpengine.com/wp-content/uploads/2018/ll/State-Food-Waste-
3 Recycling-Data-Colleetion~Reporting-Analysis.pdf.
4 BioCycle (2010) The State of Garbage in America. Prepared by Rob van Haaren, Nickolas Themelis and Nora
5 Goldstein. Available online at http://www.biocvcle.net/images/art/1010/bcl01016 s.pdf.
6 BioCycle (2017) The State of Organics Recycling in the U.S. Prepared by Nora Goldstein. Available online at
7 http://www.biocycle.net/17 10 06 1/0001/BioCycle StateQfOrganicsUS.pdf.
8 Cornell Composting (1996). Monitoring Compost Moisture. Cornell Waste Management Institute. Available online
9 at: http://compost.css.cornell.edu/monitor/monitormoisture.html.
10 Cornell Waste Management Institute (2007) The Science of Composting. Available online at
11 http://cwmi.css.cornell.edu/chapterl.pdf.
12 EPA (2020a) Advancing Sustainable Materials Management: 2018 Tables and Figures. Office of Solid Waste and
13 Emergency Response, U.S. Environmental Protection Agency, Washington, D.C. Available online at:
14 https://www.epa.gov/sites/default/files/2021-01/documents/2018 tables and figures dec 2020 fnl 508.pdf.
15 EPA (2020b) 2018 Wasted Food Report. November 2020. Available online at
16 https://www.epa.gov/sites/default/files/202Q-ll/documents/2018 wasted food report.pdf.
17 EPA (2018) Advancing Sustainable Materials Management: 2015 Tables and Figures. Office of Solid Waste and
18 Emergency Response, U.S. Environmental Protection Agency, Washington, D.C. Available online at
19 https://www.epa.gov/sites/production/files/2018-
20 07/documents/smm 2015 tables and figures 07252018 fnl 508 O.pdf.
21 EPA (2016) Advancing Sustainable Materials Management: Facts and Figures 2014. Office of Solid Waste and
22 Emergency Response, U.S. Environmental Protection Agency, Washington, D.C. Available online at
23 https://www.epa.gov/sites/production/files/2016-ll/documents/2014 smm tablesfigures 508.pdf.
24 EPA (2014) Municipal Solid Waste in the United States: 2012 Facts and Figures. Office of Solid Waste and
25 Emergency Response, U.S. Environmental Protection Agency, Washington, D.C. Available online at
26 https://www.epa.gov/sites/default/files/2015-09/documents/2012 msw fs.pdf.
27 Harvard Law School and Center for EcoTechnology (CET) (2019) Bans and Beyond: Designing and Implementing
28 Organic Waste Bans and Mandatory Organics Recycling Laws. Prepared by Katie Sandson and Emily Broad Leib,
29 Harvard Law School Food Law and Policy Clinic, with input from Lorenzo Macaluso and Coryanne Mansell, Center
30 for EcoTechnology (CET). Available online at https://wastedfood.cetonline.org/wp-
31 content/uploads/2019/07/Harvard-Law-School-FLPC-Center-for-EcoTechnology-CET-Qrganic-Waste-Bans-
32 Toolkit.pdf.
33 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
34 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
35 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
36 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
37 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
38 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
39 M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom
40 996 pp.
41 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. Volume 5: Waste, Chapter 4: Biological
42 Treatment of Solid Waste, Table 4.1. The National Greenhouse Gas Inventories Programme, The
43 Intergovernmental Panel on Climate Change, H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.).
44 Hayama, Kanagawa, Japan. Available online at https://www.ipcc-
45 nggip.iges.or.ip/public/2006gl/pdf/5 Volume5/V5 4 Ch4 Bio Treat.pdf.
110 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Institute for Local Self-Reliance (ISLR) (2014). State of Composting in the US: What, Why, Where & How. Available
2 at http://ilsr.org/wp-content/uploads/2014/07/state-of-composting-in-us.pdf.
3 Kijanka (2020) Email correspondence. Kenin Kijanka, EPA Region 2 to Rachel Schmeltz, EPA HQ. "Puerto Rico
4 Composting Operations." November 13, 2020.
5 Northeast Recycling Council (NERC) (2020) Disposal Bans & Mandatory Recycling in the United States. Available
6 online at https://nerc.org/documents/disposal%20bans%20mandatorv%20recycling%20united%20states.pdf.
7 University of Maine (2016). Compost Report Interpretation Guide. Soil Testing Lab. Available online at:
8 https://umaine.edu/soiltestinglab/wp-content/uploads/sites/227/2016/07/Compost-Report-lnterpretation-
9 Guide.pdf.
10 U.S. Census Bureau (2021) Table 1. Annual Estimates of the Resident Population for the United States, Regions,
11 States, the District of Columbia, and Puerto Rico: April 1, 2010 to July 1, 2019; April 1, 2020; and July 1, 2020 (NST-
12 EST2020). Available online at https://www.census.gov/programs-surveys/popest/technical-
13 documentation/research/evaluation-estimates/2020-evaluation-estimates/2010s-totals-national.html.
14 U.S. Census Bureau, Population Division (2022) Table 1. Annual Estimates of the Resident Population for the United
15 States, Regions, States, the District of Columbia, and Puerto Rico: April 1, 2020 to July 1, 2021 (NST-EST2021-POP).
16 Available online at https://www.census.gov/data/datasets/time-series/demo/popest/2020s-national-total.html.
17 U.S. Composting Council (2010) Yard Trimmings Bans: Impact and Support. Prepared by Stuart Buckner, Executive
18 Director, U.S., Composting Council. Available online at
19 https://cdn.vmaws.com/www.compostingcouncil.org/resource/resmgr/images/advocacy/Yard-Trimmings-Ban-
20 lmpacts-a.pdf.
21 U.S. Composting Council (2022) State and City Organics Bans, as of June 2021. Accessed on September 29, 2022.
22 Available at https://www.compostingcouncil.org/page/organicsbans.
23 Anaerobic Digestion at Biogas Facilities
24 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. Volume 5: Waste, Chapter 4: Biological
25 Treatment of Solid Waste, Table 4.1. The National Greenhouse Gas Inventories Programme, The Intergovernmental
26 Panel on Climate Change, H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Hayama, Kanagawa,
27 Japan. Available online at https://www.ipcc-
28 nggip.iges.or.ip/public/2006gl/pdf/5 Volume5/V5 4 Ch4 Bio Treat.pdf.
29 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
30 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
31 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
32 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
33 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
34 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
35 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom
36 996 pp.
37 Bronstein, Kate (2021). Expert Judgement Uncertainty of quantity of materials digested. RTI International, Solid
38 Waste Management GHG Expert.
39 EPA (2021) Anaerobic Digestion Facilities Processing Food Waste in the United States (2017 & 2018): Survey
40 Results. January 2021 EPA/903/S-21/001. Available online at https://www.epa.gov/sites/default/files/2021-
41 02/documents/2021 final ad report feb 2 with links.pdf.
42 EPA (2020) Types of Anaerobic Digesters: Common Ways to Describe Digesters. Available online at
43 https://www.epa.gov/anaerobic-digestion/types-anaerobic-digesters.
Waste 111
-------
1 EPA (2019) Anaerobic Digestion Facilities Processing Food Waste in the United States in 2016: Survey Results.
2 September 2019 EPA/903/S-19/001. Available online at https://www.epa.gov/sites/production/files/2018-
3 08/documents/ad data report final 508 compliant no password.pdf.
4 EPA (2018) Anaerobic Digestion Facilities Processing Food Waste in the United States in 2015: Survey Results. May
5 2018 EPA/903/S-18/001. Available online at https://www.epa.gov/sites/production/files/2019-
6 09/documents/ad data report vlO - 508 comp vl.pdf.
7 EPA (2016) Frequently Asked Questions About Anaerobic Digestion. Available online at
8 https://www.epa.goV/anaerobic-digestion/frequent-questions-about-anaerobic-digestion#codigestion.
9 EPA (1993) Anthropogenic Methane Emissions in the U.S.: Estimates for 1990, Report to Congress. Office of Air and
10 Radiation, Washington, DC. April 1993.
11 Water Environment Federation (WEF) (2012) What Every Operator Should Know about Anaerobic Digestion.
12 Available online at https://www.wef.org/globalassets/assets-wef/direct-download-librarv/public/operator-
13 essentials/wet-operator-essentials—anaerobic-digestion—decl2.pdf.
14 Waste Incineration
15 RTI (2010) Hospital/Medical/lnfectious Waste Incinerators: Summary of Requirements for Revised or New Section
16 lll(d)/129 State Plans Following Amendments to the Emission Guidelines. Available online at
17 https://nepis.epa.gov/Exe/ZvPDF.cgi/P1009ZW6.PDF?Dockev=P1009ZW6.PDF.
is Waste Sources of Precursor Greenhouse Gas Emissions - TO
19 BE UPDATED FOR FINAL INVENTORY REPORT
20 EPA (2022) "Criteria pollutants National Tier 1 for 1970 - 2021." National Emissions Inventory (NEI) Air Pollutant
21 Emissions Trends Data. Office of Air Quality Planning and Standards, February 2022. Available online at:
22 https://www.epa.gov/air-emissions-inventories/air-pollutant-emissions-trends-data.EPA (2021) Resource
23 Conservation and Recovery Act (RCRA) Info, Biennial Report, GM Form (Section 2- Onsite Management) and WR
24 Form.
25 EPA (2021) "2017 National Emissions Inventory (NEI) Technical Support Document (TSD)." Office of Air Quality
26 Planning and Standards, April 2021. Available online at: https://www.epa.gov/air-emissions-inventories/2017-
27 national-emissions-inventorv-nei-technical-support-document-tsd.
28 EPA (2003) Email correspondence containing preliminary ambient air pollutant data. Office of Air Pollution and the
29 Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency. December 22, 2003.
so Recalculations and Improvements
31 EIA (2022a) Monthly Energy Review, November 2022. Energy Information Administration, U.S. Department of
32 Energy. Washington, D.C. DOE/EIA-0035(2022/11).
33 EIA (2022b) International Energy Statistics 1980-2021. Energy Information Administration, U.S. Department of
34 Energy. Washington, D.C. Available online at: https://www.eia.gov/beta/international/.
35 EPA (2019Motor Vehicle Emissions Simulator (MOVES). Office of Transportation and Air Quality, U.S.
36 Environmental Protection Agency. Available online at: https://www.epa.gov/moves.
37 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
38 Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M.
112 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
1 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
2 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
3 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
4 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
5 M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom
6 996 pp.
7 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
8 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
9 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
10 Soil Survey Staff (2020a) Gridded National Soil Survey Geographic (gNATSGO) Database for the Conterminous
11 United States. United States Department of Agriculture, Natural Resources Conservation Service. Available online
12 at https://nrcs.app.box.eom/v/soils.
13 Soil Survey Staff (2020b) Gridded National Soil Survey Geographic (gNATSGO) Database for Alaska. United States
14 Department of Agriculture, Natural Resources Conservation Service. Available online at
15 https://nrcs.app.box.eom/v/soils.
16 STATSG02 (2016) Soil Survey Staff, Natural Resources Conservation Service, United States Department of
17 Agriculture. U.S. General Soil Map (STATSG02). Available online at https://sdmdataaccess.sc.egov.usda.gov.
18 Accessed 10 November 2016.
19
Waste 113
-------
i Abbreviations
ABS Acrylonitrile butadiene styrene
AC Air conditioner
ACC American Chemistry Council
AEDT FAA Aviation Environmental Design Tool
AEO Annual Energy Outlook
AER All-electric range
AF&PA American Forest and Paper Association
AFEAS Alternative Fluorocarbon Environmental
Acceptability Study
AFOLU Agriculture, Forestry, and Other Land Use
AFV Alternative fuel vehicle
AGA American Gas Association
AGR Acid gas removal
AHEF Atmospheric and Health Effect Framework
AHRI Air-Conditioning, Heating, and Refrigeration
Institute
AIM Act American Innovation and Manufacturing Act
AISI American Iron and Steel Institute
ALU Agriculture and Land Use
ANGA American Natural Gas Alliance
ANL Argonne National Laboratory
APC American Plastics Council
API American Petroleum Institute
APTA American Public Transportation Association
AR4 IPCC Fourth Assessment Report
AR5 IPCC Fifth Assessment Report
AR6 IPCC Sixth Assessment Report
ARI Advanced Resources International
ARMA Autoregressive moving-average
ARMS Agricultural Resource Management Surveys
ASAE American Society of Agricultural Engineers
ASLRRA American Short-line and Regional Railroad
Association
ASR Annual Statistical Report
ASTM American Society for Testing and Materials
AZR American Zinc Recycling
BCEF Biomass conversion and expansion factors
BEA Bureau of Economic Analysis, U.S. Department
of Commerce
BIER Beverage Industry Environmental Roundtable
BLM Bureau of Land Management
BoC Bureau of Census
BOD Biological oxygen demand
BOD5 Biochemical oxygen demand over a 5-day
period
BOEM Bureau of Ocean Energy Management
BOEMRE Bureau of Ocean Energy Management,
Regulation and Enforcement
BOF Basic oxygen furnace
BRS Biennial Reporting System
BSEE Bureau of Safety and Environmental
Enforcement
BTS Bureau of Transportation Statistics, U.S.
Department of Transportation
Btu British thermal unit
C Carbon
C&D Construction and demolition waste
C&EN Chemical and Engineering News
CAAA Clean Air Act Amendments of 1990
CAFOS Concentrated Animal Feeding Operations
CaO Calcium oxide
CAPP Canadian Association of Petroleum Producers
CARB California Air Resources Board
CBI Confidential business information
C-CAP Coastal Change Analysis Program
CDAT Chemical Data Access Tool
CEAP USDA-NRCS Conservation Effects Assessment
Program
CEFM Cattle Enteric Fermentation Model
CEMS Continuous emission monitoring system
CFC Chlorofluorocarbon
CFR Code of Federal Regulations
CGA Compressed Gas Association
CH4 Methane
CHAPA California Health and Productivity Audit
CHP Combined heat and power
CI Confidence interval
CIGRE International Council on Large Electric Systems
CKD Cement kiln dust
CLE Crown Light Exposure
CMA Chemical Manufacturer's Association
CMM Coal mine methane
CMOP Coalbed Methane Outreach Program
CMR Chemical Market Reporter
CNG Compressed natural gas
CO Carbon monoxide
C02 Carbon dioxide
COD Chemical oxygen demand
COGCC Colorado Oil and Gas Conservation Commission
CONUS Continental United States
CRF Common Reporting Format
CRM Component ratio method
CRP Conservation Reserve Program
CSRA Carbon Sequestration Rural Appraisals
CTIC Conservation Technology Information Center
CVD Chemical vapor deposition
CWNS Clean Watershed Needs Survey
d.b.h Diameter breast height
DE Digestible energy
114 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
DESC Defense Energy Support Center-DoD's Defense g
Logistics Agency G&B
DFAMS Defense Fuels Automated Management System GaAs
DGGS Division of Geological & Geophysical Surveys GCV
DHS Department of Homeland Security GDP
DLA DoD's Defense Logistics Agency GEI
DM Dry matter GHG
DOC Degradable organic carbon GHGRP
DOC U.S. Department of Commerce GIS
DoD U.S. Department of Defense GJ
DOE U.S. Department of Energy GOADS
DOI U.S. Department of the Interior GOM
DOM Dead organic matter GPG
DOT U.S. Department of Transportation GRI
DRE Destruction or removal efficiencies GSAM
DRI Direct Reduced Iron GTI
EAF Electric arc furnace GWP
EDB Aircraft Engine Emissions Databank ha
EDF Environmental Defense Fund HBFC
EER Energy economy ratio HC
EF Emission factor HCFC
EFMA European Fertilizer Manufacturers Association HCFO
EJ Exajoule HDDV
EGR Exhaust gas recirculation HDGV
EGU Electric generating unit HDPE
EIA Energy Information Administration, U.S. HF
Department of Energy HFC
EIIP Emissions Inventory Improvement Program HFO
EOR Enhanced oil recovery HFE
EPA U.S. Environmental Protection Agency HHV
EREF Environment Research & Education Foundation HMA
ERS Economic Research Service HMIWI
ETMS Enhanced Traffic Management System HTF
EV Electric vehicle HTS
EVI Enhanced Vegetation Index HVAE
FAA Federal Aviation Administration HWP
FAO Food and Agricultural Organization IBF
FAOSTAT Food and Agricultural Organization database IC
FAS Fuels Automated System ICAO
FCCC Framework Convention on Climate Change ICBA
FEB Fiber Economics Bureau ICE
FEMA Federal Emergency Management Agency ICR
FERC Federal Energy Regulatory Commission IEA
FGD Flue gas desulfurization IFO
FHWA Federal Highway Administration IGES
FIA Forest Inventory and Analysis IISRP
FIADB Forest Inventory and Analysis Database
FIPR Florida Institute of Phosphate Research ILENR
FOD First order decay
FOEN Federal Office for the Environment IMO
FOKS Fuel Oil and Kerosene Sales IPAA
FQSV First-quarter of silicon volume IPCC
FSA Farm Service Agency IPPU
FTP Federal Test Procedure ITC
Gram
Gathering and boosting
Gallium arsenide
Gross calorific value
Gross domestic product
Gulfwide Emissions Inventory
Greenhouse gas
EPA's Greenhouse Gas Reporting Program
Geographic Information Systems
Gigajoule
Gulf Offshore Activity Data System
Gulf of Mexico
Good Practice Guidance
Gas Research Institute
Gas Systems Analysis Model
Gas Technology Institute
Global warming potential
Hectare
Hydrobromofluorocarbon
Hydrocarbon
Hydrochlorofluorocarbon
Hydrochlorofluoroolefin
Heavy duty diesel vehicle
Heavy duty gas vehicle
High density polyethylene
Hydraulically fractured
Hydrofluorocarbon
Hydrofluoroolefin
Hydrofluoroether
Higher Heating Value
Hot Mix Asphalt
Hospital/medical/infectious waste incinerator
Heat Transfer Fluid
Harmonized Tariff Schedule
High Voltage Anode Effects
Harvested wood product
International bunker fuels
Integrated Circuit
International Civil Aviation Organization
International Carbon Black Association
Internal combustion engine
Information Collection Request
International Energy Agency
Intermediate Fuel Oil
Institute of Global Environmental Strategies
International Institute of Synthetic Rubber
Products
Illinois Department of Energy and Natural
Resources
International Maritime Organization
Independent Petroleum Association of America
Intergovernmental Panel on Climate Change
Industrial Processes and Product Use
U.S. International Trade Commission
References 9-1
-------
ITRS International Technology Roadmap for MRLC
Semiconductors
JWR Jim Walters Resources MRV
KCA Key category analysis MSHA
kg Kilogram MSW
kt Kiloton MT
kWh Kilowatt hour MTBE
LDPE Low density polyethylene MTBS
LDT Light-duty truck MVAC
LDV Light-duty vehicle MY
LEV Low emission vehicles N20
LFG Landfill gas NA
LFGTE Landfill gas-to-energy NACWA
LHV Lower Heating Value NAHMS
LKD Lime kiln dust NAICS
LLDPE Linear low density polyethylene NAPAP
LMOP EPA's Landfill Methane Outreach Program
LNG Liquefied natural gas NARR
LPG Liquefied petroleum gas(es) NAS
LTO Landing and take-off
LULUCF Land Use, Land-Use Change, and Forestry NASA
LVAE Low Voltage Anode Effects NASF
M&R Metering and regulating NASS
MARPOL International Convention for the Prevention of NC
Pollution from Ships NCASI
MC Motorcycle
MCF Methane conversion factor NCV
MCL Maximum Contaminant Levels ND
MCFD Thousand cubic feet per day NE
MDI Metered dose inhalers NEH
MDP Management and design practices NEI
MECS EIA Manufacturer's Energy Consumption Survey NEMA
MEMS Micro-electromechanical systems NEMS
MER Monthly Energy Review NESHAP
MGO Marine gas oil
MgO Magnesium oxide NEU
MJ Megajoule NEV
MLRA Major Land Resource Area NF3
mm Millimeter NFI
MMBtu Million British thermal units NGL
MMCF Million cubic feet NID
MMCFD Million cubic feet per day NIR
MMS Minerals Management Service NLA
MMT Million metric tons NLCD
MMTCE Million metric tons carbon equivalent NMOC
MMT C02 Million metric tons carbon dioxide equivalent NMVOC
Eq. NMOG
MODIS Moderate Resolution Imaging NO
Spectroradiometer N02
MoU Memorandum of Understanding NOx
MOVES U.S. EPA's Motor Vehicle Emission Simulator NOAA
model
MPG Miles per gallon NOF
NPDES
Multi-Resolution Land Characteristics
Consortium
Monitoring, reporting, and verification
Mine Safety and Health Administration
Municipal solid waste
Metric ton
Methyl Tertiary Butyl Ether
Monitoring Trends in Burn Severity
Motor vehicle air conditioning
Model year
Nitrous oxide
Not applicable; Not available
National Association of Clean Water Agencies
National Animal Health Monitoring System
North American Industry Classification System
National Acid Precipitation and Assessment
Program
North American Regional Reanalysis Product
National Academies of Sciences, Engineering,
and Medicine
National Aeronautics and Space Administration
National Association of State Foresters
USDA's National Agriculture Statistics Service
No change
National Council of Air and Stream
Improvement
Net calorific value
No data
Not estimated
National Engineering Handbook
National Emissions Inventory
National Electrical Manufacturers Association
National Energy Modeling System
National Emission Standards for Hazardous Air
Pollutants
Non-Energy Use
Neighborhood Electric Vehicle
Nitrogen trifluoride
National forest inventory
Natural gas liquids
National inventory of Dams
National Inventory Report
National Lime Association
National Land Cover Dataset
Non-methane organic compounds
Non-methane volatile organic compound
Non-methane organic gas
Not occurring
Nitrogen dioxide
Nitrogen oxides
National Oceanic and Atmospheric
Administration
Not on feed
National Pollutant Discharge Elimination System
9-2 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
-------
NPP Net primary productivity PU
NPRA National Petroleum and Refiners Association PVC
NRBP Northeast Regional Biomass Program PV
NRC National Research Council QA/QC
NRCS Natural Resources Conservation Service QBtu
NREL National Renewable Energy Laboratory R&D
NRI National Resources Inventory RECs
NSCEP National Service Center for Environmental RCRA
Publications RFA
NSCR Non-selective catalytic reduction RFS
NSPS New source performance standards RMA
NWS National Weather Service RPA
OAG Official Airline Guide RTO
OAP EPA Office of Atmospheric Programs SAE
OAQPS EPA Office of Air Quality Planning and Standards SAGE
ODP Ozone depleting potential SAIC
ODS Ozone depleting substances SAN
OECD Organization of Economic Co-operation and SAR
Development SCR
OEM Original equipment manufacturers SCSE
OGJ Oil & Gas Journal SDR
OGOR Oil and Gas Operations Reports SEC
OH Hydroxyl radical SEMI
OMS EPA Office of Mobile Sources
ORNL Oak Ridge National Laboratory SF6
OSHA Occupational Safety and Health Administration SIA
OTA Office of Technology Assessment SiC
OTAQ EPA Office of Transportation and Air Quality SICAS
OVS Offset verification statement SNAP
PADUS Protected Areas Database of the United States SNG
PAH Polycyclic aromatic hydrocarbons S02
PCA Portland Cement Association SOC
PCC Precipitate calcium carbonate SOG
PDF Probability Density Function SOHIO
PECVD Plasma enhanced chemical vapor deposition SSURGO
PET Polyethylene terephthalate STMC
PET Potential evapotranspiration SULEV
PEVM PFC Emissions Vintage Model SWANA
PFC Perfluorocarbon SWDS
PFPE Perfluoropolyether SWICS
PHEV Plug-in hybrid vehicles TA
PHMSA Pipeline and Hazardous Materials Safety TAM
Administration TAME
PI Productivity index TAR
PLS Pregnant liquor solution TBtu
POTW Publicly Owned Treatment Works TDN
ppbv Parts per billion (109) by volume TEDB
ppm Parts per million TFI
ppmv Parts per million (106) by volume TIGER
pptv Parts per trillion (1012) by volume
PRCI Pipeline Research Council International TJ
PRP Pasture/Range/Paddock TLEV
PS Polystyrene TMLA
PSU Primary Sample Unit TOW
Polyurethane
Polyvinyl chloride
Photovoltaic
Quality Assurance and Quality Control
Quadrillion Btu
Research and Development
Reduced Emissions Completions
Resource Conservation and Recovery Act
Renewable Fuels Association
Renewable Fuel Standard
Rubber Manufacturers' Association
Resources Planning Act
Regression-through-the-origin
Society of Automotive Engineers
System for assessing Aviation's Global Emissions
Science Applications International Corporation
Styrene Acrylonitrile
IPCC Second Assessment Report
Selective catalytic reduction
South central and southeastern coastal
Steel dust recycling
Securities and Exchange Commission
Semiconductor Equipment and Materials
Industry
Sulfur hexafluoride
Semiconductor Industry Association
Silicon carbide
Semiconductor International Capacity Statistics
Significant New Alternative Policy Program
Synthetic natural gas
Sulfur dioxide
Soil Organic Carbon
State of Garbage survey
Standard Oil Company of Ohio
Soil Survey Geographic Database
Scrap Tire Management Council
Super Ultra Low Emissions Vehicle
Solid Waste Association of North America
Solid waste disposal sites
Solid Waste Industry for Climate Solutions
Treated anaerobically (wastewater)
Typical animal mass
Tertiary amyl methyl ether
IPCC Third Assessment Report
Trillion Btu
Total digestible nutrients
Transportation Energy Data Book
The Fertilizer Institute
Topological^ Integrated Geographic Encoding
and Referencing survey
Terajoule
Traditional low emissions vehicle
Total Manufactured Layer Area
Total organics in wastewater
References 9-3
-------
TPO Timber Product Output VAIP
TRI Toxic Release Inventory
TSDF Hazardous waste treatment, storage, and VAM
disposal facility VKT
TTB Tax and Trade Bureau VMT
TVA Tennessee Valley Authority VOCs
UAN Urea ammonium nitrate VS
UDI Utility Data Institute WBJ
UFORE U.S. Forest Service's Urban Forest Effects model WEF
UG Underground (coal mining) WERF
U.S. United States WFF
U.S. ITC United States International Trade Commission
UEP United Egg Producers WGC
ULEV Ultra low emission vehicle WIP
UNEP United Nations Environmental Programme WMO
UNFCCC United Nations Framework Convention on WMS
Climate Change WRRF
USAA U.S. Aluminum Association WTE
USAF United States Air Force WW
USDA United States Department of Agriculture WWTP
USFS United States Forest Service ZEVs
USGS United States Geological Survey
1
USITC U.S. International Trade Commission
EPA's Voluntary Aluminum Industrial
Partnership
Ventilation air methane
Vehicle kilometers traveled
Vehicle miles traveled
Volatile organic compounds
Volatile solids
Waste Business Journal
Water Environment Federation
Water Environment Research Federation
World Fab Forecast (previously WFW, World
Fab Watch)
World Gas Conference
Waste-in-place
World Meteorological Organization
Waste management systems
Water resource recovery facilities
Waste-to-energy
Wastewater
Wastewater treatment plant
Zero emissions vehicles
9-4 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2021
------- |