1990-2020
if\ nr>A United States
Environmental Protection
Agency
EPA430-R-22-003
Inventory of
U.S. Greenhouse Gas
Emissions and Sinks
-------
Front cover photo credit for cow and digester: Vanguard Renewables.
-------
HOW TO OBTAIN COPIES
You can electronically download this document on the U.S. EPA's homepage at
https://www.epa.gov/ghgemissions/inventory-us-greenhouse-gas-emissions-and-sinks.
All data tables of this document for the full time series 1990 through 2020, inclusive, will be made available with
the final report published by April 15, 2022 at the internet site mentioned above.
RECOMMENDED CITATION
EPA (2022) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020. U.S. Environmental Protection
Agency, EPA 430-R-22-003. https://www.epa.gov/ghgemissions/draft-inventorv-us-greenhouse-gas-emissions-
and-sinks-1990-2020.
FOR FURTHER INFORMATION
Contact Ms. Mausami Desai, Environmental Protection Agency, (202) 343-9381, desai.mausami@epa.gov,
or Mr. Vincent Camobreco, Environmental Protection Agency, (202) 564-9043, camobreco.vincent@epa.gov.
For more information regarding climate change and greenhouse gas emissions, see the EPA web site at
https://www.epa.gov/ghgemissions.
-------
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 Programs, 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 CH4 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 Erin McDuffie (AAAS Science & Technology Policy Fellow hosted by the Office of Atmospheric
Programs) for her advice and review in areas such as fugitive oil and gas estimates and assessing uncertainty.
Other EPA offices and programs also contributed data, analysis, and technical review for this report. The Office of
Atmospheric Program's Greenhouse Gas Reporting Program staff facilitated aggregation and review of facility-level
data for use in the Inventory, in particular confidential business information data. The Office of Transportation and
Air Quality and the Office of Air Quality Planning and Standards provided analysis and review for several of the
source categories addressed in this report. The Office of Research and Development conducted field research to
support compiling 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.
-------
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, Ram Gurung, Ernie Marx, Stephen Williams and Guhan Dheenadayalan Sivakami)
for compiling the inventories for Cm 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), 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 would also like to 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, Mollie Carroll, Claire Trevisan, Kyle Herdegen, Deep Shah, Lou Browning, Sarah Whitlock, Erin Asher,
Lynn Socha, Neha Vaingankar, Emily Golla, Katie O'Malley, Emily Adkins, Tyler Brewer, Johanna Garfinkel, Zeyu Hu,
Alex Da Silva, Shubh Jain, Alida Monaco, Kenny Yerardi, Eliza Puritz, Annie Roberts, Ajo Rabemiarisoa, and Bikash
Acharya 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. Cortney Itle, Kara Edquist, Amber Allen, Tara
Stout, Spencer Sauter, 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, and Libby
Robinson for their analytical support in development of IPPU CO2, CH4, and N2O emissions; and Tiffany Moore,
Cassy Becker, and Jason Goldsmith for their analytical support on disaggregating industrial sector fossil fuel
combustion emissions. We would also like to thank Leslie Pearce, Jeff Coburn, and others at RTI for supporting
analyses for certain oil and gas fugitive sources.
-------
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 covered a 30-day
period from February 15 through March 17, 2022, and comments received during the public review period were
posted to the docket EPA-HQ-OAR-2022-0001. Comments received after the closure of the public comment period
were 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 2022.
-------
Table of Contents
TABLE OF CONTENTS VI
LIST OF TABLES, FIGURES, BOXES, AND EQUATIONS IX
EXECUTIVE SUMMARY ES-1
ES.l Background Information ES-2
ES.2 Recent Trends in U.S. Greenhouse Gas Emissions and Sinks ES-4
ES.3 Overview of Sector Emissions and Trends ES-16
ES.4 Other Information ES-20
1. INTRODUCTION 1-1
1.1 Background Information 1-3
1.2 National Inventory Arrangements 1-11
1.3 Inventory Process 1-13
1.4 Methodology and Data Sources 1-16
1.5 Key Categories 1-17
1.6 Quality Assurance and Quality Control (QA/QC) 1-23
1.7 Uncertainty Analysis of Emission Estimates 1-25
1.8 Completeness 1-28
1.9 Organization of Report 1-29
2. TRENDS IN GREENHOUSE GAS EMISSIONS 2-1
2.1 Trends in U.S. Greenhouse Gas Emissions and Sinks 2-1
2.2 Emissions by Economic Sector 2-27
2.3 Precursor Greenhouse Gas Emissions (CO, NOx, NMVOCs, and SO2) 2-39
3. ENERGY 3-1
3.1 Fossil Fuel Combustion (CRF Source Category 1A) 3-7
3.2 Carbon Emitted from Non-Energy Uses of Fossil Fuels (CRF Source Category 1A) 3-49
3.3 Incineration of Waste (CRF Source Category 1A5) 3-57
3.4 Coal Mining (CRF Source Category lBla) 3-61
3.5 Abandoned Underground Coal Mines (CRF Source Category lBla) 3-68
vi Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
3.6 Petroleum Systems (CRF Source Category lB2a) 3-72
3.7 Natural Gas Systems (CRF Source Category lB2b) 3-88
3.8 Abandoned Oil and Gas Wells (CRF Source Categories lB2a and lB2b) 3-108
3.9 International Bunker Fuels (CRF Source Category 1: Memo Items) 3-113
3.10 Wood Biomass and Biofuels Consumption (CRF Source Category 1A) 3-118
3.11 Energy Sources of Precursor Greenhouse Gas Emissions 3-122
4. INDUSTRIAL PROCESSES AND PRODUCT USE 4-1
4.1 Cement Production (CRF Source Category 2A1) 4-10
4.2 Lime Production (CRF Source Category 2A2 and 2H3) 4-15
4.3 Glass Production (CRF Source Category 2A3) 4-21
4.4 Other Process Uses of Carbonates (CRF Source Category 2A4) 4-25
4.5 Ammonia Production (CRF Source Category 2B1) 4-30
4.6 Urea Consumption for Non-Agricultural Purposes 4-35
4.7 Nitric Acid Production (CRF Source Category 2B2) 4-38
4.8 Adipic Acid Production (CRF Source Category 2B3) 4-43
4.9 Caprolactam, Glyoxal and Glyoxylic Acid Production (CRF Source Category 2B4) 4-47
4.10 Carbide Production and Consumption (CRF Source Category 2B5) 4-51
4.11 Titanium Dioxide Production (CRF Source Category 2B6) 4-55
4.12 Soda Ash Production (CRF Source Category 2B7) 4-58
4.13 Petrochemical Production (CRF Source Category 2B8) 4-61
4.14 HCFC-22 Production (CRF Source Category 2B9a) 4-70
4.15 Carbon Dioxide Consumption (CRF Source Category 2B10) 4-73
4.16 Phosphoric Acid Production (CRF Source Category 2B10) 4-77
4.17 Iron and Steel Production (CRF Source Category 2C1) and Metallurgical Coke Production 4-81
4.18 Ferroalloy Production (CRF Source Category 2C2) 4-92
4.19 Aluminum Production (CRF Source Category 2C3) 4-96
4.20 Magnesium Production and Processing (CRF Source Category 2C4) 4-104
4.21 Lead Production (CRF Source Category 2C5) 4-110
4.22 Zinc Production (CRF Source Category 2C6) 4-113
4.23 Electronics Industry (CRF Source Category 2E) 4-119
4.24 Substitution of Ozone Depleting Substances (CRF Source Category 2F) 4-136
4.25 Electrical Transmission and Distribution (CRF Source Category 2G1) 4-152
4.26 Nitrous Oxide from Product Uses (CRF Source Category 2G3) 4-161
4.27 Industrial Processes and Product Use Sources of Precursor Gases 4-164
5. AGRICULTURE 5-1
vii
-------
5.1 Enteric Fermentation (CRF Source Category 3A) 5-4
5.2 Manure Management (CRF Source Category 3B) 5-11
5.3 Rice Cultivation (CRF Source Category 3C) 5-20
5.4 Agricultural Soil Management (CRF Source Category 3D) 5-27
5.5 Liming (CRF Source Category 3G) 5-46
5.6 Urea Fertilization (CRF Source Category 3H) 5-49
5.7 Field Burning of Agricultural Residues (CRF Source Category 3F) 5-51
6. LAND USE, LAND-USE CHANGE, AND FORESTRY 6-1
6.1 Representation of the U.S. Land Base 6-9
6.2 Forest Land Remaining Forest Land (CRF Category 4A1) 6-24
6.3 Land Converted to Forest Land (CRF Source Category 4A2) 6-47
6.4 Cropland Remaining Cropland (CRF Category 4B1) 6-54
6.5 Land Converted to Cropland (CRF Category 4B2) 6-66
6.6 Grassland Remaining Grassland (CRF Category 4C1) 6-73
6.7 Land Converted to Grassland (CRF Category 4C2) 6-85
6.8 Wetlands Remaining Wetlands (CRF Category 4D1) 6-92
6.9 Land Converted to Wetlands (CRF Source Category 4D2) 6-130
6.10 Settlements Remaining Settlements (CRF Category 4E1) 6-152
6.11 Land Converted to Settlements (CRF Category 4E2) 6-172
6.12 Other Land Remaining Other Land (CRF Category 4F1) 6-178
6.13 Land Converted to Other Land (CRF Category 4F2) 6-179
7. WASTE 7-1
7.1 Landfills (CRF Source Category 5A1) 7-4
7.2 Wastewater Treatment and Discharge (CRF Source Category 5D) 7-20
7.3 Composting (CRF Source Category 5B1) 7-53
7.4 Anaerobic Digestion at Biogas Facilities (CRF Source Category 5B2) 7-58
7.5 Waste Incineration (CRF Source Category 5C1) 7-64
7.6 Waste Sources of Precursor Greenhouse Gases 7-64
8. OTHER 8-1
9. RECALCULATIONS AND IMPROVEMENTS 9-1
10. REFERENCES AND ABBREVIATIONS 10-1
viii Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
List of Tables, Figures, Boxes, and
Equations
Tables
Table ES-1: Global Warming Potentials (100-Year Time Horizon) Used in this Report ES-3
Table ES-2: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks (MMT CO2 Eq.) ES-4
Table ES-3: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks by IPCC Sector/Category (MMT CO2 Eq.)
ES-16
Table ES-4: U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use, Land-Use Change, and
Forestry (MMT C02 Eq.) ES-19
Table ES-5: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO2 Eq.) ES-21
Table ES-6: U.S. Greenhouse Gas Emissions with Electricity-Related Emissions Distributed by Economic Sector
(MMT CO2 Eq.) ES-22
Table ES-7: Recent Trends in Various U.S. Data (Index 1990 = 100) ES-23
Table 1-1: Global Atmospheric Concentration, Rate of Concentration Change, and Atmospheric Lifetime of
Selected Greenhouse Gases 1-5
Table 1-2: Global Warming Potentials and Atmospheric Lifetimes (Years) Used in this Report 1-10
Table 1-3: Comparison of 100-Year GWP values 1-11
Table 1-4: Key Categories for the United States (1990 and 2020) 1-18
Table 1-5: Estimated Overall Inventory Quantitative Uncertainty for 1990 (MMT CO2 Eq. and Percent) 1-26
Table 1-6: Estimated Overall Inventory Quantitative Uncertainty for 2020 (MMT CO2 Eq. and Percent) 1-27
Table 1-7: Quantitative Assessment of Trend Uncertainty (MMT CO2 Eq. and Percent) 1-28
Table 1-8: IPCC Sector Descriptions 1-29
Table 1-9: List of Annexes 1-30
Table 2-1: RecentTrends in U.S. Greenhouse Gas Emissions and Sinks (MMT CO2 Eq.) 2-3
Table 2-2: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks (kt) 2-5
Table 2-3: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks by IPCC Sector/Category (MMT CO2 Eq.).. 2-8
Table 2-4: Emissions from Energy (MMT CO2 Eq.) 2-10
Table 2-5: CO2 Emissions from Fossil Fuel Combustion by End-Use Sector (MMT CO2 Eq.) 2-13
Table 2-6: Emissions from Industrial Processes and Product Use (MMT CO2 Eq.) 2-18
Table 2-7: Emissions from Agriculture (MMT CO2 Eq.) 2-21
Table 2-8: U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use, Land-Use Change, and Forestry
(MMT CO2 Eq.) 2-23
Table 2-9: Emissions from Waste (MMT CO2 Eq.) 2-26
Table 2-10: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO2 Eq. and Percent of Total in
2020) 2-28
ix
-------
Table 2-11: Electric Power-Related Greenhouse Gas Emissions (MMT CO2 Eq.) 2-32
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 2020 2-33
Table 2-13: Transportation-Related Greenhouse Gas Emissions (MMT CO2 Eq.) 2-36
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) 2-40
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 2020 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: Approach 2 Quantitative Uncertainty Estimates for CFU and N2O Emissions from Energy-Related
Stationary Combustion, Including Biomass (MMT CO2 Eq. and Percent) 3-44
Table 3-19: Approach 2 Quantitative Uncertainty Estimates for CFU and N2O Emissions from Mobile Sources (MMT
CO2 Eq. and Percent) 3-47
Table 3-20: CO2 Emissions from Non-Energy Use Fossil Fuel Consumption (MMT CO2 Eq. and Percent) 3-50
Table 3-21: Adjusted Consumption of Fossil Fuels for Non-Energy Uses (TBtu) 3-50
Table 3-22: 2020 Adjusted Non-Energy Use Fossil Fuel Consumption, Storage, and Emissions 3-51
Table 3-23: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Non-Energy Uses of Fossil Fuels
(MMT CO2 Eq. and Percent) 3-54
Table 3-24: Approach 2 Quantitative Uncertainty Estimates for Storage Factors of Non-Energy Uses of Fossil Fuels
(Percent) 3-54
Table 3-25: CO2, CFU, and N2O Emissions from the Incineration of Waste (MMT CO2 Eq.) 3-57
x Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 3-26: CO2, CFU, and N2O Emissions from the Incineration of Waste (kt) 3-57
Table 3-27: Municipal Solid Waste Incinerated (Metric Tons) 3-58
Table 3-28: Calculated Fossil CO2 Content per Ton Waste Incinerated (kg CCh/Short Ton Incinerated) 3-59
Table 3-29: CO2 Emissions from Combustion of Tires (MMT CO2 Eq.) 3-59
Table 3-30: Approach 2 Quantitative Uncertainty Estimates for CO2 and N2O from the Incineration of Waste (MMT
CO2 Eq. and Percent) 3-60
Table 3-31: Coal Production (kt) 3-61
Table 3-32: CFU Emissions from Coal Mining (MMT CO2 Eq.) 3-62
Table 3-33: CFU Emissions from Coal Mining (kt) 3-62
Table 3-34: CO2 Emissions from Coal Mining (MMT CO2 Eq.) 3-65
Table 3-35: CO2 Emissions from Coal Mining (kt) 3-65
Table 3-36: Approach 2 Quantitative Uncertainty Estimates for CFU and CO2 Emissions from Coal Mining (MMT CO2
Eq. and Percent) 3-67
Table 3-37: CFU Emissions from Abandoned Coal Mines (MMT CO2 Eq.) 3-69
Table 3-38: CFU Emissions from Abandoned Coal Mines (kt) 3-69
Table 3-39: Number of Gassy Abandoned Mines Present in U.S. Basins in 2020, Grouped by Class According to
Post-Abandonment State 3-71
Table 3-40: Approach 2 Quantitative Uncertainty Estimates for CFU Emissions from Abandoned Underground Coal
Mines (MMT CO2 Eq. and Percent) 3-72
Table 3-41: Total Greenhouse Gas Emissions (CO2, CFU, and N2O) from Petroleum Systems (MMT CO2 Eq.) 3-74
Table 3-42: CFU Emissions from Petroleum Systems (MMT CO2 Eq.) 3-74
Table 3-43: CFU Emissions from Petroleum Systems (kt CH4) 3-75
Table 3-44: CO2 Emissions from Petroleum Systems (MMT CO2) 3-75
Table 3-45: CO2 Emissions from Petroleum Systems (kt CO2) 3-75
Table 3-46: N2O Emissions from Petroleum Systems (Metric Tons CO2 Eq.) 3-75
Table 3-47: N2O Emissions from Petroleum Systems (Metric Tons N2O) 3-76
Table 3-48: Approach 2 Quantitative Uncertainty Estimates for CFU and CO2 Emissions from Petroleum Systems
(MMT CO2 Eq. and Percent) 3-79
Recalculations of CO2 in Petroleum Systems (MMT CO2) 3-81
Recalculations of CH4 in Petroleum Systems (MMT CO2 Eq.) 3-81
HF Oil Well Completions National CO2 Emissions (kt CO2) 3-82
HF Oil Well Completions National CH4 Emissions (Metric Tons CH4) 3-82
Produced Water National CH4 Emissions (Metric Tons CH4) 3-82
Tanks National CO2 Emissions (kt CO2) 3-83
Tanks National CFU Emissions (MT CH4) 3-83
Pneumatic Controller National CH4 Emissions (Metric Tons CH4) 3-83
Associated Gas Flaring National CO2 Emissions (kt CO2) 3-84
Table
3-49
Table
3-50
Table
3-51
Table
3-52
Table
3-53
Table
3-54
Table
3-55
Table
3-56
Table
3-57
xi
-------
Table 3-58: Associated Gas Flaring National Cm Emissions (Metric Tons Cm) 3-84
Table 3-59: Associated Gas Venting National Cm Emissions (Metric Tons CFU) 3-84
Table 3-60: Chemical Injection Pump National Cm Emissions (Metric Tons CFU) 3-85
Table 3-61: Offshore Production National CFU Emissions (Metric Tons Cm) 3-85
Table 3-62: Gas Engine National CFU Emissions (Metric Tons Cm) 3-85
Table 3-63: National Oil Well Counts 3-86
Table 3-64: Refining National CO2 Emissions (kt CO2) 3-86
Table 3-65: Quantity of CO2 Captured and Extracted for EOR Operations (kt CO2) 3-88
Table 3-66: Geologic Sequestration Information Reported Under GHGRP Subpart RR 3-88
Table 3-67: Total Greenhouse Gas Emissions (Cm, CO2, and N2O) from Natural Gas Systems (MMT CO2 Eq.) 3-91
Table 3-68: CFU Emissions from Natural Gas Systems (MMT CO2 Eq.) 3-92
Table 3-69: CFU Emissions from Natural Gas Systems (kt) 3-92
Table 3-70: CO2 Emissions from Natural Gas Systems (MMT) 3-92
Table 3-71: CO2 Emissions from Natural Gas Systems (kt) 3-92
Table 3-72: N2O Emissions from Natural Gas Systems (Metric Tons CO2 Eq.) 3-93
Table 3-73: N2O Emissions from Natural Gas Systems (Metric Tons N2O) 3-93
Table 3-74: Approach 2 Quantitative Uncertainty Estimates for CFU and Non-combustion CO2 Emissions from
Natural Gas Systems (MMT CO2 Eq. and Percent) 3-96
Table 3-75: Recalculations of CO2 in Natural Gas Systems (MMT CO2) 3-98
Table 3-76: Recalculations of CH4 in Natural Gas Systems (MMT CO2 Eq.) 3-98
Table 3-77: Well Blowout National CFU Emissions (Metric Tons CH4) 3-99
Table 3-78: Produced Water National CH4 Emissions (Metric Tons CH4) 3-99
Table 3-79: Production Segment Pneumatic Controller National Emissions (Metric Tons CH4) 3-100
Table 3-80: Gas Engine National Emissions (MetricTons CH4) 3-100
Table 3-81: Miscellaneous Production Flaring National Emissions (Metric Tons CH4) 3-100
Table 3-82: Miscellaneous Production Flaring National Emissions (kt CO2) 3-101
Table 3-83: Production Storage Tanks National Emissions (Metric Tons CH4) 3-101
Table 3-84: Production Storage Tanks National Emissions (kt CO2) 3-101
Table 3-85: Gathering Stations Sources National CH4 Emissions (Metric Tons CH4) 3-101
Table 3-86: Gathering Pipeline Leak National CH4 Emissions (Metric Tons CH4) 3-102
Table 3-87: Gathering Pipeline Blowdowns National CH4 Emissions (Metric Tons CH4) 3-102
Table 3-88: National Gas Well Counts 3-102
Table 3-89: Processing Segment Flares National CO2 Emissions (kt CO2) 3-103
Table 3-90: Processing Segment Flares National CH4 Emissions (Metric Tons CH4) 3-103
Table 3-91: Underground Storage Well Leak National CH4 Emissions (Metric Tons CH4) 3-103
Table 3-92: Transmission Station Reciprocating Compressors National CH4 Emissions (Metric Tons CH4) 3-104
xii Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 3-93: Transmission Pipeline Venting National Cm Emissions (MetricTons Cm) 3-104
Table 3-94: Natural Gas STAR and Methane Challenge Emission Reductions (Metric Tons Cm Reduction) 3-105
Table 3-95: Post-Meter Segment National Cm Emissions (Metric Tons Cm) 3-106
Table 3-96: Post-Meter Segment National CO2 Emissions (kt CO2) 3-106
Table 3-97: Cm Emissions from Abandoned Oil and Gas Wells (MMT CO2 Eq.) 3-108
Table 3-98: Cm Emissions from Abandoned Oil and Gas Wells (kt) 3-108
Table 3-99: CO2 Emissions from Abandoned Oil and Gas Wells (MMT CO2) 3-109
Table 3-100: CO2 Emissions from Abandoned Oil and Gas Wells (kt) 3-109
Table 3-101: Abandoned Oil Wells Activity Data, CH4 and CO2 Emissions (kt) 3-110
Table 3-102: Abandoned Gas Wells Activity Data, CH4 and CO2 Emissions (kt) 3-110
Table 3-103: Approach 2 Quantitative Uncertainty Estimates for Cm and CO2 Emissions from Abandoned Oil and
Gas Wells (MMT C02 Eq. and Percent) 3-111
Table 3-104: CO2, CH4, and N2O Emissions from International Bunker Fuels (MMT CO2 Eq.) 3-114
Table 3-105: CO2, CH4, and N2O Emissions from International Bunker Fuels (kt) 3-114
Table 3-106: Aviation Jet Fuel Consumption for International Transport (Million Gallons) 3-116
Table 3-107: Marine Fuel Consumption for International Transport (Million Gallons) 3-116
Table 3-108: CO2 Emissions from Wood Consumption by End-Use Sector (MMT CO2 Eq.) 3-118
Table 3-109: CO2 Emissions from Wood Consumption by End-Use Sector (kt) 3-119
Table 3-110: CO2 Emissions from Ethanol Consumption (MMT CO2 Eq.) 3-119
Table 3-111: CO2 Emissions from Ethanol Consumption (kt) 3-119
Table 3-112: CO2 Emissions from Biodiesel Consumption (MMT CO2 Eq.) 3-120
Table 3-113: CO2 Emissions from Biodiesel Consumption (kt) 3-120
Table 3-114: Woody Biomass Consumption by Sector (Trillion Btu) 3-120
Table 3-115: Ethanol Consumption by Sector (Trillion Btu) 3-120
Table 3-116: Biodiesel Consumption by Sector (Trillion Btu) 3-121
Table 3-117: NOx, CO, NMVOC, and SO2 Emissions from Energy-Related Activities (kt) 3-122
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) 4-11
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) 4-13
Table 4-6: CO2 Emissions from Lime Production (MMT CO2 Eq. and kt) 4-16
Table 4-7: Gross, Recovered, and Net CO2 Emissions from Lime Production (kt) 4-16
Table 4-8: High-Calcium- and Dolomitic-Quicklime, High-Calcium- and Dolomitic-Hydrated, and Dead-Burned-
Dolomite Lime Production (kt) 4-18
xiii
-------
Table 4-9: Adjusted Lime Production (kt) 4-18
Table 4-10: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Lime Production (MMT CO2 Eq.
and Percent) 4-20
Table 4-11: CO2 Emissions from Glass Production (MMT CO2 Eq. and kt) 4-22
Table 4-12: Limestone, Dolomite, and Soda Ash Used in Glass Production (kt) and Average Annual Production
Index for Glass and Glass Product Manufacturing 4-23
Table 4-13: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Glass Production (MMT CO2 Eq.
and Percent) 4-24
Table 4-14: CO2 Emissions from Other Process Uses of Carbonates (MMT CO2 Eq.) 4-26
Table 4-15: CO2 Emissions from Other Process Uses of Carbonates (kt) 4-27
Table 4-16: Limestone and Dolomite Consumption (kt) 4-28
Table 4-17: Soda Ash Consumption Not Associated with Glass Manufacturing (kt) 4-28
Table 4-18: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Other Process Uses of
Carbonates (MMT CO2 Eq. and Percent) 4-29
Table 4-19: CO2 Emissions from Ammonia Production (MMT CO2 Eq.) 4-31
Table 4-20: CO2 Emissions from Ammonia Production (kt) 4-32
Table 4-21: Ammonia Production, Recovered CO2 Consumed for Urea Production, and Urea Production (kt).... 4-33
Table 4-22: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Ammonia Production (MMT
CO2 Eq. and Percent) 4-34
Table 4-23: CO2 Emissions from Urea Consumption for Non-Agricultural Purposes (MMT CO2 Eq.) 4-36
Table 4-24: CO2 Emissions from Urea Consumption for Non-Agricultural Purposes (kt) 4-36
Table 4-25: Urea Production, Urea Applied as Fertilizer, Urea Imports, and Urea Exports (kt) 4-37
Table 4-26: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Urea Consumption for Non-
Agricultural Purposes (MMT CO2 Eq. and Percent) 4-37
Table 4-27: N2O Emissions from Nitric Acid Production (MMT CO2 Eq. and kt N2O) 4-39
Table 4-28: Nitric Acid Production (kt) 4-42
Table 4-29: Approach 2 Quantitative Uncertainty Estimates for N2O Emissions from Nitric Acid Production (MMT
CO2 Eq. and Percent) 4-42
Table 4-30: N2O Emissions from Adipic Acid Production (MMT CO2 Eq. and kt N2O) 4-44
Table 4-31: Adipic Acid Production (kt) 4-46
Table 4-32: Approach 2 Quantitative Uncertainty Estimates for N2O Emissions from Adipic Acid Production (MMT
CO2 Eq. and Percent) 4-46
Table 4-33: N2O Emissions from Caprolactam Production (MMT CO2 Eq. and kt N2O) 4-49
Table 4-34: Caprolactam Production (kt) 4-50
Table 4-35: Approach 2 Quantitative Uncertainty Estimates for N2O Emissions from Caprolactam, Glyoxal and
Glyoxylic Acid Production (MMT CO2 Eq. and Percent) 4-51
Table 4-36: CO2 and Cm Emissions from Silicon Carbide Production and Consumption (MMT CO2 Eq.) 4-52
Table 4-37: CO2 and Cm Emissions from Silicon Carbide Production and Consumption (kt) 4-52
xiv Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 4-38: Production and Consumption of Silicon Carbide (Metric Tons) 4-54
Table 4-39: Approach 2 Quantitative Uncertainty Estimates for Cm and CO2 Emissions from Silicon Carbide
Production and Consumption (MMT CO2 Eq. and Percent) 4-54
Table 4-40: CO2 Emissions from Titanium Dioxide (MMT CO2 Eq. and kt) 4-56
Table 4-41: Titanium Dioxide Production (kt) 4-57
Table 4-42: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Titanium Dioxide Production
(MMT CO2 Eq. and Percent) 4-58
Table 4-43: CO2 Emissions from Soda Ash Production (MMT CO2 Eq. and kt CO2) 4-59
Table 4-44: Trona Ore Used in Soda Ash Production (kt) 4-60
Table 4-45: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Soda Ash Production (MMT
CO2 Eq. and Percent) 4-61
Table 4-46: CO2 and Cm Emissions from Petrochemical Production (MMT CO2 Eq.) 4-63
Table 4-47: CO2 and Cm Emissions from Petrochemical Production (kt) 4-63
Table 4-48: Production of Selected Petrochemicals (kt) 4-66
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) 4-67
Table 4-50: HFC-23 Emissions from HCFC-22 Production (MMT C02 Eq. and kt HFC-23) 4-71
Table 4-51: HCFC-22 Production (kt) 4-72
Table 4-52: Approach 2 Quantitative Uncertainty Estimates for HFC-23 Emissions from HCFC-22 Production (MMT
CO2 Eq. and Percent) 4-72
Table 4-53: CO2 Emissions from CO2 Consumption (MMT CO2 Eq. and kt) 4-73
Table 4-54: CO2 Production (kt CO2) and the Percent Used for Non-EOR Applications 4-75
Table 4-55: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from CO2 Consumption (MMT CO2
Eq. and Percent) 4-76
Table 4-56: CO2 Emissions from Phosphoric Acid Production (MMT CO2 Eq. and kt) 4-77
Table 4-57: Phosphate Rock Domestic Consumption, Exports, and Imports (kt) 4-79
Table 4-58: Chemical Composition of Phosphate Rock (Percent by Weight) 4-79
Table 4-59: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Phosphoric Acid Production
(MMT CO2 Eq. and Percent) 4-80
Table 4-60: CO2 Emissions from Metallurgical Coke Production (MMT CO2 Eq.) 4-82
Table 4-61: CO2 Emissions from Metallurgical Coke Production (kt) 4-82
Table 4-62: CO2 Emissions from Iron and Steel Production (MMT CO2 Eq.) 4-83
Table 4-63: CO2 Emissions from Iron and Steel Production (kt) 4-83
Table 4-64: CH4 Emissions from Iron and Steel Production (MMT CO2 Eq.) 4-83
Table 4-65: CH4 Emissions from Iron and Steel Production (kt) 4-83
Table 4-66: Material Carbon Contents for Metallurgical Coke Production 4-85
Table 4-67: Production and Consumption Data for the Calculation of CO2 Emissions from Metallurgical Coke
Production (Thousand Metric Tons) 4-86
xv
-------
Table 4-68: Production and Consumption Data for the Calculation of CO2 Emissions from Metallurgical Coke
Production (Million ft3) 4-86
Table 4-69: Material Carbon Contents for Iron and Steel Production 4-87
Table 4-70: Cm Emission Factors for Sinter and Pig Iron Production 4-87
Table 4-71: CO2 Emission Factors for Sinter Production, Direct Reduced Iron Production, and Pellet Production 4-88
Table 4-72: Production and Consumption Data for the Calculation of CO2 and CFU Emissions from Iron and Steel
Production (Thousand Metric Tons) 4-89
Table 4-73: Production and Consumption Data for the Calculation of CO2 Emissions from Iron and Steel Production
(Million ft3 unless otherwise specified) 4-89
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) 4-91
Table 4-75: CO2 and CFU Emissions from Ferroalloy Production (MMT CO2 Eq.) 4-93
Table 4-76: CO2 and CFU Emissions from Ferroalloy Production (kt) 4-93
Table 4-77: Production of Ferroalloys (Metric Tons) 4-94
Table 4-78: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Ferroalloy Production (MMT
CO2 Eq. and Percent) 4-95
Table 4-79: CO2 Emissions from Aluminum Production (MMT CO2 Eq. and kt) 4-97
Table 4-80: PFC Emissions from Aluminum Production (MMT CO2 Eq.) 4-97
Table 4-81: PFC Emissions from Aluminum Production (kt) 4-98
Table 4-82: Summary of HVAE Emissions 4-100
Table 4-83: Summary of LVAE Emissions 4-101
Table 4-84: Production of Primary Aluminum (kt) 4-102
Table 4-85: Approach 2 Quantitative Uncertainty Estimates for CO2 and PFC Emissions from Aluminum Production
(MMT CO2 Eq. and Percent) 4-103
Table 4-86: SF6, HFC-134a, FK 5-1-12 and CO2 Emissions from Magnesium Production and Processing (MMT CO2
Eq.) 4-104
Table 4-87: SF6, HFC-134a, FK 5-1-12 and CO2 Emissions from Magnesium Production and Processing (kt) 4-105
Table 4-88: SF6 Emission Factors (kg SF6 per metric ton of magnesium) 4-107
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) 4-109
Table 4-90: CO2 Emissions from Lead Production (MMT CO2 Eq. and kt) 4-111
Table 4-91: Lead Production (Metric Tons) 4-112
Table 4-92: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Lead Production (MMT CO2 Eq.
and Percent) 4-112
Table 4-93: CO2 Emissions from Zinc Production (MMT CO2 Eq. and kt) 4-115
Table 4-94: Zinc Production (Metric Tons) 4-115
Table 4-95: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Zinc Production (MMT CO2 Eq.
and Percent) 4-118
Table 4-96: PFC, HFC, SF6, NF3, and N2O Emissions from Electronics Industry (MMT CO2 Eq.) 4-122
xvi Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 4-97: PFC, HFC, SF6, NF3, and N2O Emissions from Semiconductor Manufacture (Metric Tons) 4-123
Table 4-98: F-HTF Emissions from Electronics Manufacture by Compound Group (kt CO2 Eq.) 4-123
Table 4-99: Approach 2 Quantitative Uncertainty Estimates for HFC, PFC, SF6, NF3 and N2O Emissions from
Electronics Manufacture (MMT CO2 Eq. and Percent) 4-134
Table 4-100: Emissions of HFCs and PFCs from ODS Substitutes (MMT CO2 Eq.) 4-137
Table 4-101: Emissions of HFCs and PFCs from ODS Substitution (Metric Tons) 4-137
Table 4-102: Emissions of HFCs and PFCs from ODS Substitutes (MMT CO2 Eq.) by Sector 4-138
Table 4-103: Approach 2 Quantitative Uncertainty Estimates for HFC and PFC Emissions from ODS Substitutes
(MMT CO2 Eq. and Percent) 4-141
Table 4-104
Table 4-105
Table 4-106
Table 4-107
Table 4-108
CO2 Eq.)
U.S. HFC Supply (MMT C02 Eq.) 4-143
Averaged U.S. HFC Demand (MMT C02 Eq.) 4-145
U.S. Emissions of HFC-32, HFC-125, HFC-134a and HFC-143a (Gg) 4-147
Percentage Differences between EPA and NOAA HFC Emission Estimates 4-149
SFs and CF4 Emissions from Electric Power Systems and Electrical Equipment Manufacturers (MMT
4-153
Table 4-109: SF6 and CF4 Emissions from Electric Power Systems and Electrical Equipment Manufacturers (kt) 4-153
Table 4-110: Transmission Mile Coverage (Percent) and Regression Coefficients (kg per mile) 4-157
Table 4-111: Approach 2 Quantitative Uncertainty Estimates for SF6 and CF4 Emissions from Electrical Transmission
and Distribution (MMT CO2 Eq. and Percent) 4-159
Table 4-112: N20 Production (kt) 4-161
Table 4-113: N2O Emissions from N2O Product Usage (MMT CO2 Eq. and kt) 4-161
Table 4-114: Approach 2 Quantitative Uncertainty Estimates for N2O Emissions from N2O Product Usage (MMT
CO2 Eq. and Percent) 4-163
Table 4-115: NOx, CO, NMVOC, and SO2 Emissions from Industrial Processes and Product Use (kt) 4-164
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: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from Enteric Fermentation (MMT CO2
Eq. and Percent) 5-9
Table 5-6: CH4 and N2O Emissions from Manure Management (MMT CO2 Eq.) 5-13
Table 5-7: CH4 and N2O Emissions from Manure Management (kt) 5-13
Table 5-8: Approach 2 Quantitative Uncertainty Estimates for CH4 and N2O (Direct and Indirect) Emissions from
Manure Management (MMT CO2 Eq. and Percent) 5-17
Table 5-9: IPCC (2006) Implied Emission Factor Default Values Compared with Calculated Values for CH4 from
Manure Management (kg/head/year) 5-18
Table 5-10: CH4 Emissions from Rice Cultivation (MMT CO2 Eq.) 5-21
Table 5-11: CH4 Emissions from Rice Cultivation (kt) 5-22
xvii
-------
Table 5-12: Rice Area Harvested (1,000 Hectares) 5-24
Table 5-13: Average Ratooned Area as Percent of Primary Growth Area (Percent) 5-25
Table 5-14: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from Rice Cultivation (MMT CO2 Eq.
and Percent) 5-27
Table 5-15: N2O Emissions from Agricultural Soils (MMT CO2 Eq.) 5-30
Table 5-16: N2O Emissions from Agricultural Soils (kt) 5-30
Table 5-17: Direct N2O Emissions from Agricultural Soils by Land Use Type and N Input Type (MMT CO2 Eq.).... 5-30
Table 5-18: Indirect N2O Emissions from Agricultural Soils (MMT CO2 Eq.) 5-31
Table 5-19: Quantitative Uncertainty Estimates of N2O Emissions from Agricultural Soil Management in 2020
(MMT CO2 Eq. and Percent) 5-44
Table 5-20: Emissions from Liming (MMT CO2 Eq.) 5-46
Table 5-21: Emissions from Liming (MMT C) 5-47
Table 5-22: Applied Minerals (MMT) 5-48
Table 5-23: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Liming (MMT CO2 Eq. and
Percent) 5-49
Table 5-24: CO2 Emissions from Urea Fertilization (MMT CO2 Eq.) 5-49
Table 5-25: CO2 Emissions from Urea Fertilization (MMT C) 5-49
Table 5-26: Applied Urea (MMT) 5-50
Table 5-27: Quantitative Uncertainty Estimates for CO2 Emissions from Urea Fertilization (MMT CO2 Eq. and
Percent) 5-51
Table 5-28: CH4 and N2O Emissions from Field Burning of Agricultural Residues (MMT CO2 Eq.) 5-52
Table 5-29: CH4, N2O, CO, and NOx Emissions from Field Burning of Agricultural Residues (kt) 5-53
Table 5-30: Agricultural Crop Production (kt of Product) 5-56
Table 5-31: U.S. Average Percent Crop Area Burned by Crop (Percent) 5-57
Table 5-32: Parameters for Estimating Emissions from Field Burning of Agricultural Residues 5-58
Table 5-33: Greenhouse Gas Emission Ratios and Conversion Factors 5-59
Table 5-34: Approach 2 Quantitative Uncertainty Estimates for CH4 and N2O Emissions from Field Burning of
Agricultural Residues (MMT CO2 Eq. and Percent) 5-59
Table 6-1: Emissions and Removals (Net Flux) from Land Use, Land-Use Change, and Forestry (MMT CO2 Eq.).... 6-3
Table 6-2: Emissions and Removals from Land Use, Land-Use Change, and Forestry by Gas (MMT CO2 Eq.) 6-5
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)
6-10
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 6-23
xviii Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 C02 (MMT per Year) Emissions from Forest Fires in the Conterminous 48 States and
Alaska3 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) 6-37
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 6-39
Table 6-16: Non-CC>2 Emissions from Forest Fires (kt)a 6-39
Table 6-17: Quantitative Uncertainty Estimates of Non-C02 Emissions from Forest Fires (MMT CO2 Eq. and
Percent)3 6-40
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) 6-41
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.) 6-44
Table 6-21: Non-C02 Emissions from Drained Organic Forest Soilsa b (kt) 6-44
Table 6-22: States identified as having Drained Organic Soils, Area of Forest on Drained Organic Soils, and
Sampling Error 6-45
Table 6-23: Quantitative Uncertainty Estimates for Non-C02 Emissions on Drained Organic Forest Soils (MMT CO2
Eq. and Percent)3 6-46
Table 6-24: Net CO2 Flux from Forest C Pools in Land Converted to Forest Land by Land Use Change Category (MMT
CO2 Eq.) 6-48
Table 6-25: Net C Flux from Forest C Pools in Land Converted to Forest Land by Land Use Change Category (MMT
C) 6-49
Table 6-26: Quantitative Uncertainty Estimates for Forest C Pool Stock Changes (MMT CO2 Eq. per Year) in 2020
from Land Converted to Forest Land by Land Use Change 6-52
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) 6-53
Table 6-28: Net CO2 Flux from Soil C Stock Changes in Cropland Remaining Cropland (MMT CO2 Eq.) 6-55
Table 6-29: Net CO2 Flux from Soil C Stock Changes in Cropland Remaining Cropland (MMT C) 6-56
Table 6-30: Approach 2 Quantitative Uncertainty Estimates for Soil C Stock Changes occurring within Cropland
Remaining Cropland (MMT CO2 Eq. and Percent) 6-64
xix
-------
Table 6-31: Area of Managed Land in Cropland Remaining Cropland that is not included in the current Inventory
(Thousand Hectares) 6-65
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.) 6-67
Table 6-33: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes in Land Converted to
Cropland (MMT C) 6-67
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) 6-71
Table 6-35: Area of Managed Land in Land Converted to Cropland that is not included in the current Inventory
(Thousand Hectares) 6-72
Table 6-36: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes in Grassland Remaining
Grassland (MMT C02 Eq.) 6-74
Table 6-37: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes in Grassland Remaining
Grassland (MMT C) 6-74
Table 6-38: Approach 2 Quantitative Uncertainty Estimates for C Stock Changes Occurring Within Grassland
Remaining Grassland (MMT CO2 Eq. and Percent) 6-79
Table 6-39: Area of Managed Land in Grassland Remaining Grassland in Alaska that is not included in the current
Inventory (Thousand Hectares) 6-81
Table 6-40: CH4 and N2O Emissions from Biomass Burning in Grassland (MMT CO2 Eq.) 6-82
Table 6-41: CH4, N2O, CO, and NOx Emissions from Biomass Burning in Grassland (kt) 6-82
Table 6-42: Thousands of Grassland Hectares Burned Annually 6-83
Table 6-43: Uncertainty Estimates for Non-C02 Greenhouse Gas Emissions from Biomass Burning in Grassland
(MMT CO2 Eq. and Percent) 6-84
Table 6-44: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for Land Converted to
Grassland (MMT C02 Eq.) 6-85
Table 6-45: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for Land Converted to
Grassland (MMT C) 6-86
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) 6-90
Table 6-47: Area of Managed Land in Land Converted to Grassland in Alaska that is not included in the current
Inventory (Thousand Hectares) 6-91
Table 6-48: Emissions from Peatlands Remaining Peatlands (MMT CO2 Eq.) 6-94
Table 6-49: Emissions from Peatlands Remaining Peatlands (kt) 6-94
Table 6-50: Peat Production of Lower 48 States (kt) 6-95
Table 6-51: Peat Production of Alaska (Thousand Cubic Meters) 6-95
Table 6-52: Peat Production Area of Lower 48 States (Hectares) 6-96
Table 6-53: Peat Production Area of Alaska (Hectares) 6-96
Table 6-54: Peat Production (Hectares) 6-97
Table 6-55: Approach 2 Quantitative Uncertainty Estimates for CO2, CH4, and N2O Emissions from Peatlands
Remaining Peatlands (MMT CO2 Eq. and Percent) 6-98
xx Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 6-56: Emissions and Removals from Coastal Wetlands Remaining Coastal Wetlands (MMT CO2 Eq.) 6-101
Table 6-57: Net CO2 Flux from C Stock Changes in Vegetated Coastal Wetlands Remaining Vegetated Coastal
Wetlands (MMT C02 Eq.) 6-102
Table 6-58: Net CO2 Flux from C Stock Changes in Vegetated Coastal Wetlands Remaining Vegetated Coastal
Wetlands (MMT C) 6-102
Table 6-59: CFU Emissions from Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands (MMT CO2
Eq. and kt CH4) 6-102
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) 6-103
Table 6-61
Table 6-62
Table 6-63
Table 6-64
Aboveground Biomass Carbon Stocks for Vegetated Coastal Wetlands (t C ha 6-103
Root to Shoot Ratios for Vegetated Coastal Wetlands 6-103
Annual Soil Carbon Accumulation Rates for Vegetated Coastal Wetlands (t C ha 1 yr 6-104
IPCC Approach 1 Quantitative Uncertainty Estimates for C Stock Changes and CH4 Emissions occurring
within Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands in 2020 (MMT CO2 Eq. and Percent). 6-
105
Table 6-65: Net CO2 Flux from C Stock Changes in Vegetated Coastal Wetlands Converted to Unvegetated Open
Water Coastal Wetlands (MMT CO2 Eq.) 6-107
Table 6-66: Net CO2 Flux from C Stock Changes in Vegetated Coastal Wetlands Converted to Unvegetated Open
Water Coastal Wetlands (MMT C) 6-107
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) 6-109
Table 6-68: CO2 Flux from C Stock Changes from Unvegetated Open Water Coastal Wetlands Converted to
Vegetated Coastal Wetlands (MMT CO2 Eq.) 6-110
Table 6-69: CO2 Flux from C Stock Changes from Unvegetated Open Water Coastal Wetlands Converted to
Vegetated Coastal Wetlands (MMT C) 6-111
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 2020 (MMT CO2 Eq. and Percent) 6-113
Table 6-71: N2O Emissions from Aquaculture in Coastal Wetlands (MMT CO2 Eq. and kt N2O) 6-114
Table 6-72: Approach 1 Quantitative Uncertainty Estimates for N2O Emissions from Aquaculture Production in
Coastal Wetlands in 2020 (MMT CO2 Eq. and Percent) 6-115
Table 6-73: CFU Emissions from Flooded Land Remaining Flooded Land —Reservoirs (MMT CO2 Eq.) 6-117
Table 6-74: CFU Emissions from Flooded Land Remaining Flooded Land —Reservoirs (kt CH4) 6-117
Table 6-75: Surface and Downstream CH4 Emissions (kt CH4) from Reservoirs and Associated Inundation Areas in
Flooded Land Remaining Flooded Land in 2020 6-118
Table 6-76: IPCC (2019) Default CFU Emission Factors for Surface Emission from Reservoirs and Associated
Inundation Areas in Flooded Land Remaining Flooded Land 6-120
Table 6-77: National Totals of Reservoirs and Associated Inundation Area Surface Area (millions of ha) in Flooded
Land Remaining Flooded Land 6-121
Table 6-78: State breakdown of Reservoirs and Associated Inundation Area Surface Area (millions of ha) in Flooded
Land Remaining Flooded Land 6-121
xx i
-------
Table 6-79: Approach 2 Quantitative Uncertainty Estimates for CFU Emissions from Reservoirs and Associated
Inundation Areas in Flooded Land Remaining Flooded Land 6-122
Table 6-80: CFU Emissions from Other Constructed Waterbodies in Flooded Land Remaining Flooded Land (MMT
C02 Eq.) 6-124
Table 6-81: CFU Emissions from Other Constructed Waterbodies in Flooded Land Remaining Flooded Land (kt CH4)
6-124
Table 6-82: CFU Emissions (kt CH4) from Other Constructed Waterbodies in Flooded Land Remaining Flooded Land
in 2020 6-125
Table 6-83: IPCC (2019) Default CH4 Emission Factors for Surface Emissions from Other Constructed Waterbodies in
Flooded Land Remaining Flooded Land 6-126
Table 6-84: National Surface Area (ha) Totals in Flooded Land Remaining Flooded Land - Other Constructed
Waterbodies 6-127
Table 6-85: State Totals of Surface Area (ha) in Flooded Land Remaining Flooded Land— Other Constructed
Waterbodies 6-128
Table 6-86: Approach 2 Quantitative Uncertainty Estimates for CFU Emissions from Other Constructed
Waterbodies in Flooded Land Remaining Flooded Land 6-129
Table 6-87: Net CO2 Flux from C Stock Changes in Land Converted to Vegetated Coastal Wetlands (MMT CO2 Eq.) ..
6-132
Table 6-88: Net CO2 Flux from C Stock Changes in Land Converted to Vegetated Coastal Wetlands (MMT C)... 6-132
Table 6-89: CFU Emissions from Land Converted to Vegetated Coastal Wetlands (MMT CO2 Eq. and kt CH4).... 6-133
Table 6-90: Approach 1 Quantitative Uncertainty Estimates for C Stock Changes occurring within Land Converted
to Vegetated Coastal Wetlands in 2020 (MMT CO2 Eq. and Percent) 6-135
Table 6-91: CFU Emissions from Reservoirs and Inundation Areas in Land Converted to Flooded Land (MMT CO2
Eq.) 6-138
Table 6-92: CFU Emissions from Reservoirs and Inundation Areas in Land Converted to Flooded Land (kt CH4). 6-138
Table 6-93: CO2 Emissions from Reservoirs and Inundation Areas in Land Converted to Flooded Land (MMT CO2) ....
6-138
Table 6-94: CO2 Emissions from Reservoirs and Inundation Areas in Land Converted to Flooded Land (MMT C) 6-138
Table 6-95: Methane and CO2 Emissions (kt) from Reservoirs and Associated Inundation Areas in Land Converted to
Flooded Land in 2020 6-139
Table 6-96: IPCC (2019) Default CFU and CO2 Emission Factors for Surface Emissions from Reservoirs and
Associated Inundation Areas in Land Converted to Flooded Land 6-140
Table 6-97: National Totals of Reservoir and Associated Inundation Areas Surface Area (thousands of ha) in Land
Converted to Flooded Land 6-142
Table 6-98: State breakdown of Reservoirs and Associated Inundation Area Surface Area (thousands of ha) in Land
Converted to Flooded Land 6-143
Table 6-99: Approach 2 Quantitative Uncertainty Estimates for CH4 and CO2 Emissions from Reservoirs and
Associated Inundation Areas in Land Converted to Flooded Land 6-144
Table 6-100: CFU Emissions from Other Constructed Waterbodies in Land Converted to Flooded Land (MMT CO2
Eq.) 6-146
Table 6-101: CFU Emissions from Other Constructed Waterbodies in Land Converted to Flooded Land (kt CH4) 6-146
xxii Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 6-102: CO2 Emissions from Other Constructed Waterbodies in Land Converted to Flooded Land (MMT CO2)....
6-146
Table 6-103: CO2 Emissions from Other Constructed Waterbodies in Land Converted to Flooded Land (MMT C)
6-146
Table 6-104: CFU and CO2 Emissions (MT CO2 Eq.) from Other Constructed Waterbodies in Land Converted to
Flooded Land in 2020 6-146
Table 6-105: IPCC Default Methane and CO2 Emission Factors for Other Constructed Waterbodies in Land
Converted to Flooded Land 6-148
Table 6-106: National Surface Area (ha) Totals of Other Constructed Waterbodies in Land Converted to Flooded
Land 6-149
Table 6-107: State Surface Area (ha) Totals of Other Constructed Waterbodies in Land Converted to Flooded Land
6-149
Table 6-108: Approach 2 Quantitative Uncertainty Estimates for CH4 and CO2 Emissions from Other Constructed
Waterbodies in Land Converted to Flooded Land 6-151
Table 6-109: Net CO2 Flux from Soil C Stock Changes in Settlements Remaining Settlements (MMT CO2 Eq.).... 6-152
Table 6-110: Net CO2 Flux from Soil C Stock Changes in Settlements Remaining Settlements (MMT C) 6-153
Table 6-111: Thousands of Hectares of Drained Organic Soils in Settlements Remaining Settlements 6-153
Table 6-112: Uncertainty Estimates for CO2 Emissions from Drained Organic Soils in Settlements Remaining
Settlements (MMT CO2 Eq. and Percent) 6-154
Table 6-113: Area of Managed Land in Settlements Remaining Settlements that is not included in the current
Inventory (Thousand Hectares) 6-154
Table 6-114: Net Flux from Trees in Settlements Remaining Settlements (MMT CO2 Eq. and MMT C)a 6-155
Table 6-115: 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) 6-158
Table 6-116: Estimated Annual C Sequestration (Metric Tons C/Year), Tree Cover (Percent), and Annual C
Sequestration per Area of Tree Cover (kg C/m2/year) for settlement areas in United States by State and the District
of Columbia (2020) 6-160
Table 6-117: Approach 2 Quantitative Uncertainty Estimates for Net CO2 Flux from Changes in C Stocks in
Settlement Trees (MMT CO2 Eq. and Percent) 6-162
Table 6-118: N2O Emissions from Soils in Settlements Remaining Settlements (MMT CO2 Eq. and kt N2O) 6-163
Table 6-119: Quantitative Uncertainty Estimates of N2O Emissions from Soils in Settlements Remaining Settlements
(MMT CO2 Eq. and Percent) 6-165
Table 6-120: Net Changes in Yard Trimmings and Food Scrap Carbon Stocks in Landfills (MMT CO2 Eq.) 6-166
Table 6-121: Net Changes in Yard Trimmings and Food Scrap Carbon Stocks in Landfills (MMT C) 6-167
Table 6-122: Moisture Contents, C Storage Factors (Proportions of Initial C Sequestered), Initial C Contents, and
Decay Rates for Yard Trimmings and Food Scraps in Landfills 6-169
Table 6-123: C Stocks in Yard Trimmings and Food Scraps in Landfills (MMT C) 6-170
Table 6-124: Approach 2 Quantitative Uncertainty Estimates for CO2 Flux from Yard Trimmings and Food Scraps in
Landfills (MMT CO2 Eq. and Percent) 6-170
Table 6-125: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for Land Converted to
Settlements (MMT C02 Eq.) 6-173
xxiii
-------
Table 6-126: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for Land Converted to
Settlements (MMTC) 6-173
Table 6-127: 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) 6-176
Table 6-128: Area of Managed Land in Settlements Remaining Settlements that is not included in the current
Inventory (Thousand Hectares) 6-178
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) 7-19
Table 7-7: CFU and N2O Emissions from Domestic and Industrial Wastewater Treatment (MMT CO2 Eq.) 7-22
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 (2020, kt, MMT CO2 Eq. and
Percent) 7-24
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 BODs Produced (kt) 7-26
Table 7-13: Variables and Data Sources for Organics in Centralized Domestic Wastewater 7-27
Table 7-14: Variables and Data Sources for CFU Emissions from Centrally Treated Aerobic Systems (Other than
Constructed Wetlands) 7-27
Table 7-15: Variables and Data Sources for CFU Emissions from Centrally Treated Aerobic Systems (Constructed
Wetlands) 7-29
Table 7-16: Variables and Data Sources for CFU Emissions from Centrally Treated Anaerobic Systems 7-30
Table 7-17: Variables and Data Sources for Emissions from Anaerobic Sludge Digesters 7-30
Table 7-18: Variables and Data Sources for CFU Emissions from Centrally Treated Systems Discharge 7-31
Table 7-19: Total Industrial Wastewater CH4 Emissions by Sector (2020, MMT CO2 Eq. and Percent) 7-33
Table 7-20: U.S. Pulp and Paper, Meat, Poultry, Vegetables, Fruits and Juices, Ethanol, Breweries, and Petroleum
Refining Production (MMT) 7-35
U.S. Industrial Wastewater Characteristics Data (2020) 7-35
U.S. Industrial Wastewater Treatment Activity Data 7-36
Sludge Variables for Aerobic Treatment Systems 7-36
Fraction of TOW Removed During Treatment by Industry 7-37
Wastewater Outflow (m3/ton) for Pulp, Paper, and Paperboard Mills 7-38
Wastewater Outflow (m3/ton) and BOD Production (g/L) for U.S. Vegetables, Fruits, and Juices
Production 7-39
Table 7-21
Table 7-22
Table 7-23
Table 7-24
Table 7-25
Table 7-26
xxiv Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 7-27: Domestic Wastewater N2O Emissions from Septic and Centralized Systems (2020, kt, MMT CO2 Eq. and
Percent) 7-41
Table 7-28: Variables and Data Sources for Protein Consumed 7-42
Table 7-29: Variables and Data Sources for N2O Emissions from Septic System 7-42
Table 7-30: Variables and Data Sources for Non-Consumed Protein and Nitrogen Entering Centralized Systems 7-43
Table 7-31: Variables and Data Sources for N2O Emissions from Centrally Treated Aerobic Systems (Other than
Constructed Wetlands) 7-44
Table 7-32: Variables and Data Sources for N2O Emissions from Centrally Treated Aerobic Systems (Constructed
Wetlands) 7-45
Table 7-33: Variables and Data Sources for N2O Emissions from Centrally Treated Anaerobic Systems 7-45
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) 7-46
Table 7-35: Variables and Data Sources for N2O Emissions from Centrally Treated Systems Discharge 7-47
Table 7-36: Total Industrial Wastewater N2O Emissions by Sector (2020, MMT CO2 Eq. and Percent) 7-48
Table 7-37: U.S. Industrial Wastewater Nitrogen Data 7-49
Table 7-38: Industrial Wastewater Nitrogen Discharged in 2018 by Sector (kg N) 7-50
Table 7-39: Approach 2 Quantitative Uncertainty Estimates for 2020 Emissions from Wastewater Treatment (MMT
CO2 Eq. and Percent) 7-51
Table 7-40: CFU and N2O Emissions from Composting (MMT CO2 Eq.) 7-55
Table 7-41: CFU and N2O Emissions from Composting (kt) 7-55
Table 7-42: U.S. Waste Composted (kt) 7-56
Table 7-43: Approach 1 Quantitative Uncertainty Estimates for CFU and N2O Emissions from Composting (MMT
CO2 Eq. and Percent) 7-56
Table 7-44: CFU Emissions from Anaerobic Digestion at Biogas Facilities (MMT CO2 Eq.) from 1990-2020 7-59
Table 7-45: CFU Emissions from Anaerobic Digestion at Biogas Facilities (kt) from 1990-2020 7-59
Table 7-46: U.S. Waste Digested (kt) from 1990-2020 7-61
Table 7-47: Estimated Number of Stand-Alone AD Facilities Operating3 from 1990-2020 7-61
Table 7-48: Estimated Biogas Produced and Methane Recovered from Anaerobic Digestion at Biogas Facilities
Operating from 1990-2020a 7-62
Table 7-49: Approach 1 Quantitative Uncertainty Estimates for Emissions from Anaerobic Digestion (MMT CO2 Eq.
and Percent) 7-63
Table 7-50: Emissions of NOx, CO, NMVOC, and SO2 from Waste (kt) 7-65
Table 9-1: Revisions to U.S. Greenhouse Gas Emissions (MMT CO2 Eq.) 9-4
Table 9-2: Revisions to U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use, Land-Use Change,
and Forestry (MMT CO2 Eq.) 9-6
Figures
Figure ES-1: U.S. Greenhouse Gas Emissions and Sinks by Gas ES-5
XXV
-------
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: 2020 U.S. Greenhouse Gas Emissions by Gas (Percentages based on MMT CO2 Eq.) ES-7
Figure ES-4: 2020 Sources of CO2 Emissions ES-8
Figure ES-5: 2020 CO2 Emissions from Fossil Fuel Combustion by Sector and Fuel Type ES-9
Figure ES-6: 2020 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: 2020 Sources of CFU Emissions ES-13
Figure ES-9: 2020 Sources of N2O Emissions ES-14
Figure ES-10: 2020 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: 2020 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: 2020 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 1-24
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: U.S. Greenhouse Gas Emissions and Sinks by IPCC Sector 2-8
Figure 2-4: Trends in Energy Sector Greenhouse Gas Sources 2-10
Figure 2-5: Trends in CO2 Emissions from Fossil Fuel Combustion by End-Use Sector and Fuel Type 2-14
Figure 2-6: Trends in End-Use Sector Emissions of CO2 from Fossil Fuel Combustion 2-15
Figure 2-7: Electric Power Generation (Billion kWh) and Emissions (MMT CO2 Eq.) 2-16
Figure 2-8: Trends in Industrial Processes and Product Use Sector Greenhouse Gas Sources 2-18
Figure 2-9: Trends in Agriculture Sector Greenhouse Gas Sources 2-20
Figure 2-10: Trends in Emissions and Removals (Net CO2 Flux) from Land Use, Land-Use Change, and Forestry.. 2-23
Figure 2-11: Trends in Waste Sector Greenhouse Gas Sources 2-26
Figure 2-12: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors 2-27
Figure 2-13: U.S. Greenhouse Gas Emissions with Electricity-Related Emissions Distributed to Economic Sectors
2-33
Figure 2-14: Trends in Transportation-Related Greenhouse Gas Emissions 2-36
Figure 2-15: U.S. Greenhouse Gas Emissions Per Capita and Per Dollar of Gross Domestic Product 2-39
Figure 3-1: 2020 Energy Sector Greenhouse Gas Sources 3-2
xxvi Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure 3-2: Trends in Energy Sector Greenhouse Gas Sources 3-2
Figure 3-3: 2020 U.S. Fossil Carbon Flows 3-3
Figure 3-4: 2020 U.S. Energy Use by Energy Source 3-10
Figure 3-5: Annual U.S. Energy Use 3-11
Figure 3-6: 2020 CO2 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 (1950-2020, Index Normal =
100) 3-12
Figure 3-8: Annual Deviations from Normal Cooling Degree Days for the United States (1950-2020, 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, Onroad VMT, and Total Sector CO2 Emissions 3-24
Figure 3-15: Sales-Weighted Fuel Economy of New Passenger Cars and Light-Duty Trucks, 1990-2020 3-26
Figure 3-16: Sales of New Passenger Cars and Light-Duty Trucks, 1990-2020 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: 2020 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. HFC Consumption (MMT C02 Eq.) 4-144
Figure 4-4: U.S. Emissions of HFC-32, HFC-125, and HFC-143a 4-147
Figure 4-5: U.S. Emissions of HFC-134a 4-148
Figure 5-1: 2020 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 5-23
Figure 5-4: Sources and Pathways of N that Result in N2O Emissions from Agricultural Soil Management 5-29
Figure 5-5: Croplands, 2015 Annual Direct N2O Emissions Estimated Using the Tier 3 DayCent Model 5-31
Figure 5-6: Grasslands, 2015 Annual Direct N2O Emissions Estimated Using the Tier 3 DayCent Model 5-32
Figure 5-7: Croplands, 2015 Annual Indirect N2O Emissions from Volatilization Using the Tier 3 DayCent Model
5-33
Figure 5-8: Grasslands, 2015 Annual Indirect N2O Emissions from Volatilization Using the Tier 3 DayCent Model
5-34
Figure 5-9: Croplands, 2015 Annual Indirect N2O Emissions from Leaching and Runoff Using the Tier 3 DayCent
Model 5-34
-------
Figure 5-10: Grasslands, 2015 Annual Indirect N2O Emissions from Leaching and Runoff Using the Tier 3 DayCent
Model 5-35
Figure 6-1: 2020 LULUCF Chapter Greenhouse Gas Sources and Sinks 6-2
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 2020 6-13
Figure 6-4: Changes in Forest Area by Region for Forest Land Remaining Forest Land in the conterminous United
States and Alaska (1990-2020) 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 (1990-2020) 6-31
Figure 6-6: Total Net Annual Soil C Stock Changes for Mineral Soils under Agricultural Management within States,
2015, Cropland Remaining Cropland 6-57
Figure 6-7: Total Net Annual Soil C Stock Changes for Organic Soils under Agricultural Management within States,
2015, Cropland Remaining Cropland 6-58
Figure 6-8: Total Net Annual Soil C Stock Changes for Mineral Soils under Agricultural Management within States,
2015, Grassland Remaining Grassland 6-75
Figure 6-9: Total Net Annual Soil C Stock Changes for Organic Soils under Agricultural Management within States,
2015, Grassland Remaining Grassland 6-76
Figure 6-10: U.S. reservoirs (black polygons) in the Flooded Land Remaining Flooded Land category in 2020. Colors
represent climate zone used to derive IPCC default emission factors 6-117
Figure 6-11: Total CH4 Emissions (Downstream + Surface) from Reservoirs and Associated Inundation Areas in
Flooded Land Remaining Flooded Land (kt CH4) 6-118
Figure 6-12: Example of a Reservoir and Associated Inundation Area 6-121
Figure 6-13: CFU Emissions (kt CH4) from Other Constructed Waterbodies in Flooded Land Remaining Flooded Land
in 2020 6-126
Figure 6-14: Surface Area (hectares) of Other Constructed Waterbodies in Flooded Land Remaining Flooded Land in
2020 6-128
Figure 6-15: U.S. Reservoirs (black polygons) in the Land Converted to Flooded Land Category in 2020 6-137
Figure 6-16: 2020 A) CFU and B) CO2 Emissions from U.S. Reservoirs and Inundation Areas in Land Converted to
Flooded Land 6-139
Figure 6-17: Example of a Reservoir and Associated Inundation Area in Land Converted to Flooded Land 6-142
Figure 6-18: Number of dams built per year from 1990-2020 6-143
Figure 6-19: CFU and CO2 Emissions (MT CO2 Eq.) from Other Constructed Waterbodies in Land Converted to
Flooded Land in 2020 6-147
Figure 6-20: Surface Area (ha) of Other Constructed Waterbodies in Land Converted to Flooded Land 6-149
Figure 7-1: 2020 Waste Sector Greenhouse Gas Sources 7-1
Figure 7-2: Trends in Waste Chapter 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 7-18
xxviii Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 9-4
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: Use of IPCC Reference Approach to support Verification of Emissions from Fossil Fuel Combustion 1-25
Box 2-1: Methodology for Aggregating Emissions by Economic Sector 2-30
Box 2-2: Trends in Various U.S. Greenhouse Gas Emissions-Related Data 2-38
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 3-52
Box 3-6: Carbon Dioxide Transport, Injection, and Geological Storage 3-87
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 5-3
Box 5-2: Surrogate Data Method 5-25
Box 5-3: Tier 1 vs. Tier 3 Approach for Estimating N2O Emissions 5-36
Box 5-4: Surrogate Data Method 5-37
Box 5-5: Comparison of the Tier 2 U.S. Inventory Approach and IPCC (2006) Default Approach 5-47
Box 5-6: Comparison of Tier 2 U.S. Inventory Approach and IPCC (2006) Default Approach 5-55
Box 6-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals 6-8
Box 6-2: Preliminary Estimates of Land Use in U.S. Territories 6-23
Box 6-3: CO2 Emissions from Forest Fires 6-31
Box 6-4: Surrogate Data Method 6-59
Box 6-5: Tier 3 Approach for Soil C Stocks Compared to Tier 1 or 2 Approaches 6-61
Box 7-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals, including
Relationship to Greenhouse Gas Reporting Data 7-3
xx ix
-------
Box 7-2: Description of a Modern, Managed Landfill in the United States 7-4
Box 7-3: Nationwide Municipal Solid Waste Data Sources 7-12
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 3-65
Equation 3-2: Estimating CO2 Emissions From Drained Methane Flared Or Catalytically Oxidized 3-66
Equation 3-3: Decline Function to Estimate Venting Abandoned Mine Methane Emissions 3-70
Equation 3-4: Decline Function to Estimate Flooded Abandoned Mine Methane Emissions 3-70
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)....
4-17
Equation 4-3: 2006 IPCC Guidelines Tier 2 Emission Factor for Lime Production, Dolomitic Lime (Equation 2.9).. 4-17
Equation 4-4: 2006 IPCC Guidelines Tier 3: N2O Emissions From Nitric Acid Production (Equation 3.6) 4-41
Equation 4-5: 2006 IPCC Guidelines Tier 2: N2O Emissions From Adipic Acid Production (Equation 3.8) 4-45
Equation 4-6: 2006 IPCC Guidelines Tier 1: N2O Emissions From Caprolactam Production (Equation 3.9) 4-49
Equation 4-7: 2006 IPCC Guidelines Tier 1: Emissions from Carbide Production (Equation 3.11) 4-53
Equation 4-8: 2006 IPCC Guidelines Tier 1: CO2 Emissions from Titanium Production (Equation 3.12) 4-56
Equation 4-9: CO2 Emissions from Phosphoric Acid Production 4-78
Equation 4-10: CO2 Emissions from Coke, Pig Iron, EAF Steel, and BOF Steel Production, based on 2006 IPCC
Guidelines Tier 2 Methodologies 4-84
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) 4-84
Equation 4-12: 2006 IPCC Guidelines Tier 1: CO2 Emissions for Ferroalloy Production (Equation 4.15) 4-93
Equation 4-13: 2006 IPCC Guidelines Tier 1: CFU Emissions for Ferroalloy Production (Equation 4.18) 4-93
Equation 4-14: CF4 Emissions Resulting from Low Voltage Anode Effects 4-101
Equation 4-15: 2006 IPCC Guidelines Tier 1: CO2 Emissions From Lead Production (Equation 4.32) 4-111
Equation 4-16: 2006 IPCC Guidelines Tier 1: CO2 Emissions From Zinc Production (Equation 4.33) 4-116
Equation 4-17: Waelz Kiln CO2 Emission Factor for Zinc Produced 4-116
Equation 4-18: Waelz Kiln CO2 Emission Factor for EAF Dust Consumed 4-117
Equation 4-19: Total Emissions from Electronics Industry 4-131
Equation 4-20: Total Emissions from Semiconductor Manufacturing 4-131
Equation 4-21: Total Emissions from MEMS Manufacturing 4-134
Equation 4-22: Total Emissions from PV Manufacturing 4-134
Equation 4-23: Estimation for SF6 Emissions from Electric Power Systems 4-154
Equation 4-24: Regression Equation for Estimating SF6 Emissions of Non-Reporting Facilities 4-157
xxx Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Equation 4-25: N2O Emissions from Product Use 4-162
Equation 5-1: Elemental C or N Released through Oxidation of Crop Residues 5-54
Equation 5-2: Emissions from Crop Residue Burning 5-55
Equation 5-3: Estimation of Greenhouse Gas Emissions from Fire 5-55
Equation 6-1: Net State Annual Carbon Sequestration 6-160
Equation 6-2: Total C Stock for Yard Trimmings and Food Scraps in Landfills 6-169
Equation 6-3: C Stock Annual Flux for Yard Trimmings and Food Scraps in Landfills 6-169
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 7-11
Equation 7-4: Total Domestic CFU Emissions from Wastewater Treatment and Discharge 7-24
Equation 7-5: CFU Emissions from Septic Systems 7-25
Equation 7-6: Total Wastewater BODs Produced per Capita (U.S.-Specific [ERG 2018a]) 7-25
Equation 7-7: Total Organically Degradable Material in Domestic Wastewater (IPCC 2019 [Eq. 6.3]) 7-25
Equation 7-8: Total Domestic CFU Emissions from Centrally Treated Aerobic Systems 7-27
Equation 7-9: Total Organics in Centralized Wastewater Treatment [IPCC 2019 (Eq. 6.3A)] 7-27
Equation 7-10: Organic Component Removed from Aerobic Wastewater Treatment (IPCC 2019 [Eq. 6.3B]) 7-27
Equation 7-11: Emissions from Centrally Treated Aerobic Systems (other than Constructed Wetlands) (IPCC 2019
[Eq. 6.1]) 7-27
Equation 7-12: Emissions from Centrally Treated Aerobic Systems (Constructed Wetlands Only) [IPCC 2014 (Eq.
6.1)] 7-29
Equation 7-13: Emissions from Centrally Treated Aerobic Systems (Constructed Wetlands used as Tertiary
Treatment) (U.S. Specific) 7-29
Equation 7-14: Emissions from Centrally Treated Anaerobic Systems [IPCC 2019 (Eq. 6.1)] 7-29
Equation 7-15: Emissions from Anaerobic Sludge Digesters (U.S. Specific) 7-30
Equation 7-16: Emissions from Centrally Treated Systems Discharge (U.S.-Specific) 7-31
Equation 7-17: Total Organics in Centralized Treatment Effluent (IPCC 2019 [Eq. 6.3D]) 7-31
Equation 7-18: Total Organics in Effluent Discharged to Reservoirs, Lakes, or Estuaries (U.S.-Specific) 7-31
Equation 7-19: Total Organics in Effluent Discharged to Other Waterbodies (U.S.-Specific) 7-31
Equation 7-20: Total CFU Emissions from Industrial Wastewater 7-33
Equation 7-21: TOW in Industry Wastewater Treatment Systems 7-33
Equation 7-22: Organic Component Removed from Aerobic Wastewater Treatment - Pulp, Paper, and Paperboard
7-34
Equation 7-23: Organic Component Removed from Aerobic Treatment Plants 7-34
Equation 7-24: Raw Sludge Removed from Wastewater Treatment as Dry Mass 7-34
Equation 7-25: CFU Emissions from Industrial Wastewater Treatment Discharge 7-36
Equation 7-26: TOW in Industrial Wastewater Effluent 7-37
xxxi
-------
Equation 7-27: Emissions from Pulp and Paper Discharge (U.S. Specific) 7-38
Equation 7-28: Total Organics in Pulp and Paper Effluent Discharged to Reservoirs, Lakes, Or Estuaries (U.S.
Specific) 7-38
Equation 7-29: Total Organics in Pulp and Paper Effluent Discharged to Other Waterbodies (U.S. Specific) 7-38
Equation 7-30: Total Domestic N2O Emissions from Wastewater Treatment and Discharge 7-41
Equation 7-31: Annual per Capita Protein Supply (U.S. Specific) 7-41
Equation 7-32: Consumed Protein [IPCC 2019 (Eq. 6.10A)] 7-41
Equation 7-33: Total Nitrogen Entering Septic Systems (IPCC 2019 [Eq. 10]) 7-42
Equation 7-34: Emissions from Septic Systems (IPCC 2019 [Eq. 6.9]) 7-42
Equation 7-35: Total Nitrogen Entering Centralized Systems (IPCC 2019 [Eq. 10]) 7-43
Equation 7-36: Total Domestic N2O Emissions from Centrally Treated Aerobic Systems 7-44
Equation 7-37: Emissions from Centrally Treated Aerobic Systems (other than Constructed Wetlands) (IPCC 2019
[Eq. 6.9]) 7-44
Equation 7-38: Emissions from Centrally Treated Aerobic Systems (Constructed Wetlands Only) (IPCC 2014 [Eq.
6.9]) 7-44
Equation 7-39: Emissions from Centrally Treated Aerobic Systems (Constructed Wetlands used as Tertiary
Treatment) (U.S.-Specific) 7-44
Equation 7-40: Emissions from Centrally Treated Anaerobic Systems (IPCC 2019 [Eq. 6.9]) C (kt N20/year) 7-45
Equation 7-41: Emissions from Centrally Treated Systems Discharge (U.S.-Specific) 7-46
Equation 7-42: Total Organics in Centralized Treatment Effluent (IPCC 2019 [Eq. 6.8]) 7-46
Equation 7-43: Total Nitrogen in Effluent Discharged to Impaired Waterbodies (U.S.-Specific) 7-47
Equation 7-44: Total Nitrogen in Effluent Discharged to Nonimpaired Waterbodies (U.S.-Specific) 7-47
Equation 7-45: Total Nitrogen in Industrial Wastewater 7-48
Equation 7-46: N2O Emissions from Indsutrial Wastewater Treatment Plants 7-49
Equation 7-47: N2O Emissions from Industrial Wastewater Treatment Effluent 7-49
Equation 7-48: Greenhouse Gas Emission Calculation for Composting 7-55
Equation 7-49: Methane Emissions Calculation for Anaerobic Digestion 7-59
Equation 7-50: Recovered Methane Estimation for Anaerobic Digestion 7-59
Equation 7-51: Weighted Average of Waste Processed 7-60
xxxii Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Executive Summary
An emissions inventory that identifies and quantifies a country's anthropogenic1 sources and sinks of greenhouse
gases 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 2020. 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.
Executive Summary ES-1
-------
Box ES-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 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 the 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 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. 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).
ES.l Background Information
Greenhouse gases absorb infrared radiation, thereby trapping heat in the atmosphere and making the planet
warmer. The most important greenhouse gases directly emitted by humans include carbon dioxide (CO2), methane
(CH4), nitrous oxide (N2O), and several fluorine-containing halogenated substances (HFCs, PFCs, SF6 and NF3).
Although 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.
ES-2 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
concentrations. From the pre-industrial era (i.e., ending about 1750) to 2020, concentrations of these greenhouse
gases have increased globally by 47.9,168.4, and 23.3 percent, respectively (IPCC 2013; NOAA/ESRL 2022a, 2022b,
2022c). 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, GWP-weighted emissions are provided in million metric tons
of CO2 equivalent (MMT CO2 Eq.).8,9 Estimates for all gases in this Executive Summary are presented in units of
MMT CO2 Eq. Emissions by gas in unweighted mass kilotons are provided in the Trends and sector chapters of this
report and in the Common Reporting Format (CRF) tables that are also part of the submission to the UNFCCC.
UNFCCC reporting guidelines for national inventories require the use of 100-year GWP values from the IPCC Fourth
Assessment Report (AR4) (IPCC 2007) to ensure that national greenhouse gas inventories reported by all nations
are comparable.10 All estimates are provided throughout the report in both CO2 equivalents and unweighted units.
A comparison of emission estimates using the 100-year AR4 GWP values versus the IPCC Fifth Assessment Report
(AR5) (IPCC 2013) and the IPCC Sixth Assessment Report (AR6) (IPCC 2021) GWP values can be found in Chapter 1
and, in more detail, in Annex 6.1 of this report. The GWP values used in this report are listed below in Table ES-1.
The UNFCCC will require countries to shift to use AR5 100-year GWP values in 2024, when countries submit their
first reports using updated reporting guidelines under the Paris Agreement.11
Table ES-1: Global Warming Potentials (100-Year Time Horizon) Used in this Report
Gas
GWP
C02
1
CH4a
25
N20
298
HFCs
up to 14,800
PFCs
up to 12,200
sf6
22,800
nf3
17,200
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 (2007).
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 http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
11 See https://unfccc.int/process-and-meetines/transparencv-and-reporting/reporting-and-review-under-the-paris-agreement.
Executive Summary ES-3
-------
ES.2 Recent Trends in U.S. Greenhouse Gas
Emissions and Sinks
In 2020, total gross U.S. greenhouse gas emissions were 5,981.4 million metric tons of carbon dioxide equivalent
(MMT CO2 Eq).12 Total U.S. emissions have decreased by 7.3 percent from 1990 to 2020, down from a high of 15.7
percent above 1990 levels in 2007. Emissions decreased from 2019 to 2020 by 9.0 percent (590.4 MMT CO2 Eq.).
Net emissions (including sinks) were 5,222.4 MMT CO2 Eq. in 2020. Overall, net emissions decreased 10.6 percent
from 2019 to 2020 and decreased 21.4 percent from 2005 levels as shown in Table ES-2. The sharp decline in
emissions from 2019 to 2020 is largely due to the impacts of the coronavirus (COVID-19) pandemic on travel and
economic activity. However, the decline also reflects the combined impacts of long-term trends in many factors,
including population, economic growth, energy markets, technological changes including energy efficiency, and the
carbon intensity of energy fuel choices. Between 2019 and 2020, the decrease in total greenhouse gas emissions
was driven largely by a 10.5 percent decrease in CO2 emissions from fossil fuel combustion, including a 13.3
percent decrease in transportation sector emissions from less travel due to the COVID-19 pandemic and a 10.4
percent decrease in emissions in the electric power sector. The decrease in electric power sector emissions was
due to a decrease in electricity demand of about 2.5 percent and also reflects the continued shift from coal to less
carbon intensive natural gas and renewables.
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 2020. 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
2016
2017
2018
2019
2020
C02
5,122.5
6,137.6
5,251.8
5,211.0
5,376.7
5,259.1
4,715.7
CH4a
780.8
697.5
657.6
663.8
671.1
668.8
650.4
N2Oa
450.5
453.3
449.2
444.6
457.7
456.8
426.1
HFCs
46.5
127.4
168.3
171.1
171.0
175.9
178.8
PFCs
24.3
6.7
4.4
4.2
4.8
4.6
4.4
sf6
28.8
11.8
6.0
5.9
5.7
5.9
5.4
nf3
+
0.5
0.6
0.6
0.6
0.6
0.6
Total Gross Emissions (Sources)
6,453.5
7,434.8
6,537.9
6,501.0
6,687.5
6,571.7
5,981.4
LULUCF Emissions3
31.4
41.3
35.4
45.5
39.8
30.3
53.2
ch4
27.2
30.9
28.3
34.0
30.7
25.5
38.1
n2o
4.2
10.5
7.1
11.5
9.1
4.8
15.2
LULUCF Carbon Stock Change/C02b
(892.0)
(831.1)
(862.0)
(826.7)
(809.0)
(760.8)
(812.2)
LULUCF Sector NetTotalc
(860.6)
(789.8)
(826.6)
(781.2)
(769.3)
(730.5)
(758.9)
Net Emissions (Sources and Sinks)
5,592.8
6,645.0
5,711.2
5,719.8
5,918.2
5,841.2
5,222.4
+ 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
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.
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.
ES-4 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 net C stock
changes.
Notes: Total emissions presented without LULUCF. Net emissions presented with LULUCF. Totals may not sum due to
independent rounding. Parentheses indicate negative values or sequestration.
Figure ES-1: U.S. Greenhouse Gas Emissions and Sinks by Gas
¦ HFCs, PFCs, SFe and NF3 ¦ Net Emissions (including LULUCF sinks)
9,000 g Nitrous Oxide
¦ Methane
8,000 B Carbon Dioxide
¦ Net CO2 Flux from LULUCF=
7,000
6,000
e 5,000
LU
8 4,000
I-
21
E 3,000
2,000
1,000
0
-1,000
cti o->
0s! CTl O"! Cl
O O O O O O O O O O ¦>—1 ¦>—• ¦»-1 ¦»—1 ¦>—1 t—I ¦»-1 ¦»-1 • ¦>—1
00000000000000000000
rMrMfMrMfNrMrvjrMrMrMrMfNjfMrMrvjrMrMrsjfMfM
a 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."
Executive Summary ES-5
-------
Figure ES-2: Annual Percent Change in Gross U.S. Greenhouse Gas Emissions and Sinks
Relative to the Previous Year
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
(i.e., 1990 to 2019) to ensure that the trend is accurate.
Below are categories with recalculations resulting in an average change over the time series of greater than 2.5
MMTCCh Eq.
• Natural Gas Systems (CH4)
• Land Converted to Grassland: Changes in all Ecosystem Carbon Stocks (CO2)
• Wastewater Treatment (N2O)
• Manure Management (CH4)
In addition, the Inventory includes new categories not included in the previous Inventory that improve
completeness of the national estimates. Specifically, the current report includes CH4 emissions from post-meter
uses (i.e., includes leak emissions from residential and commercial appliances, industrial facilities and power
plants, and natural gas fueled vehicles), fugitive CO2 emissions from coal mining, CO2 emissions from land
converted to flooded land (i.e., lands converted to use as reservoirs and other constructed water bodies), CH4
emissions from land remaining and land converted to flooded land, and PFC (CF4) emissions from electrical
transmission and distribution.
In each Inventory, the results of all methodological changes and historical data updates and 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
category description(s) within the relevant report chapter (i.e., Energy chapter (Chapter 3), the Industrial Process
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,
ES-6 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
"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 U.S. emissions in 2020, weighted
by global warming potential. The primary greenhouse gas emitted by human activities in the United States was
CO2, representing 78.8 percent of total greenhouse gas emissions. The largest source of CO2, and of overall
greenhouse gas emissions, was fossil fuel combustion primarily from transportation and power generation.
Methane (CH4) emissions account for 10.9 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 were the major sources of N2O emissions. Ozone depleting substance substitute emissions was the
primary contributor to aggregate hydrofluorocarbon (HFC) emissions. Perfluorocarbon (PFC) emissions were
primarily attributable to electronics manufacturing and primary aluminum production. Electrical transmission and
distribution systems accounted for most sulfur hexafluoride (SFs) emissions. The electronics industry is the only
source of nitrogen trifluoride (NF3) emissions.
Figure ES-3: 2020 U.S. Greenhouse Gas Emissions by Gas (Percentages based on MMT CO2
Eq.)
3.2%
HFCs, PFCs, SFs and NFs
From 1990 to 2020, total emissions of CO2 decreased by 406.8 MMT CO2 Eq. (7.9 percent), total emissions of Cm
decreased by 130.4 MMT CO2 Eq. (16.7 percent), and emissions of N2O decreased by 24.4 MMT CO2 Eq. (5.4
percent). During the same period, emissions of fluorinated greenhouse gases including HFCs, PFCs, SF6, and NF3
rose by 89.5 MMT CO2 Eq. (89.8 percent). From 1990 to 2020, emissions of HFCs increased by 132.2 MMT CO2 Eq.
(284.3 percent) and NF3 emissions increased by 0.6 MMT CO2 Eq. (1,195.3 percent), while emissions of PFCs
decreased by 19.8 MMT CO2 Eq. (81.8 percent) and SF6 emissions decreased by 23.4 MMT CO2 Eq. (81.3 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 13.6 percent of total emissions in 2020 (as reflected in
Executive Summary ES-7
-------
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.13
Since the Industrial Revolution (i.e., about 1750), global atmospheric concentrations of CO2 have risen 47.9 percent
(IPCC 2013; NOAA/ESRL 2022a), principally due to the combustion of fossil fuels for energy. Globally, an estimated
31,500 MMT of CO2 were added to the atmosphere through the combustion of fossil fuels in 2019, of which the
United States accounted for 15.4 percent.14
Within the United States, fossil fuel combustion accounted for 92.1 percent of CO2 emissions in 2020.
Transportation was the largest emitter of CO2 in 2020 followed by electric power generation. There are 26
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: 2020 Sources of CO2 Emissions
Fossil Fuel Combustion
Non-Energy Use of Fuels
Iron and Steel Production
Cement Production
Natural Gas Systems
Other Industrial Processes
Petroleum Systems
Petrochemical Production
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.
Note: 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 are included in Other Industrial Processes. Emissions from Abandoned Oil
and Gas Wells and Coal Mining are included in Other Energy.
As the largest source of U.S. greenhouse gas emissions, CO2 from fossil fuel combustion has accounted for 75.3
percent of GWP-weighted total U.S. gross emissions across the time series. Between 1990 and 2020, CO2 emissions
from fossil fuel combustion decreased from 4,731.2 MMT CO2 Eq. to 4,342.7 MMT CO2 Eq., an 8.2 percent total
13 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."
14 Global C02 emissions from fossil fuel combustion were taken from International Energy Agency C02 Emissions from Fossil
Fuels Combustion Overview. See https://webstore.iea.org/co2-emissions-from-fuel-combustion-2020-highlights (IEA 2021). The
publication has not yet been updated to include complete global 2020 data.
ES-8 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
decrease. Conversely, CO2 emissions from fossil fuel combustion decreased by 1,409.4 MMT CO2 Eq. from 2005
levels, a decrease of 24.5 percent. From 2019 to 2020, these emissions decreased by 509.7 MMT CO2 Eq. (10.5
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. Between 2019 and 2020, reduced economic activity and decreased travel due
to the COVID-19 pandemic 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: 2020 CO2 Emissions from Fossil Fuel Combustion by Sector and Fuel Type
2,500
2,000
S" 1,500
fS
0
u
t-
1 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)
227
23
Coal
Natural Gas
I Geothermal
I Petroleum
U.S. Territories
Commercial
Residential
1,439
1,572
Industrial Electric Power Transportation
Executive Summary ES-9
-------
Figure ES-6: 2020 End-Use Sector Emissions of CO2 from Fossil Fuel Combustion
2,000 ¦ Direct Fossil Fuel Combustion
Indirect Fossil Fuel Combustion
1,500
o
^ 1,000
500
U.S. Territories
Commercial
1,577
Residential
Industrial
Transportation
Transportation End-Use Sector. Transportation activities accounted for 36.2 percent of U.S. CO2 emissions from
fossil fuel combustion in 2020, with the largest contributors being passenger vehicles (38.5 percent), followed by
freight trucks (26.3 percent) and light-duty trucks (18.9 percent). 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 2020, total transportation CO2 emissions increased due, in large part, to
increased demand for travel15 as 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 2019 to 2020, transportation CO2
emissions decreased 13.3 percent, primarily as a result of the COVID-19 pandemic and associated restrictions that
led to less travel. While an increased demand for travel has led to generally increasing CO2 emissions since 1990,
improvements in average new vehicle fuel economy since 2005 has slowed the rate of increase of CO2 emissions.
In 2020, petroleum-based products supplied 94.5 percent of the energy consumed for transportation, primarily
from gasoline consumption in automobiles and other highway vehicles (57.3 percent), diesel fuel for freight trucks
(26.5 percent), jet fuel for aircraft (9.6 percent), and natural gas, residual fuel, aviation gasoline, and liquefied
petroleum gases (1.0 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 27.1 percent of CO2 emissions
from fossil fuel combustion in 2020. Approximately 65.2 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 22.0 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 2019 to 2020, total energy use in the industrial sector decreased by 4.4
percent partially as a result of reductions in economic and manufacturing activity due to the COVID-19 pandemic.
Residential and Commercial End-Use Sectors. The residential and commercial end-use sectors accounted for 19.8
and 16.3 percent, respectively, of CO2 emissions from fossil fuel combustion in 2020. The residential and
commercial sectors relied heavily on electricity for meeting energy demands, with 63.3 and 67.9 percent,
respectively, of their emissions attributable to electricity use for lighting, heating, cooling, and operating
15 VMT estimates are based on data from FHWA Highway Statistics Table VM-1 (FHWA 1996 through 2021). In 2007 and 2008
light-duty VMT decreased 3.0 percent and 2.3 percent, respectively. Note that the decline in light-duty VMT from 2006 to 2007
is due at least in part to a change in FHWA's methods for estimating VMT. In 2011, FHWA changed its methods for estimating
VMT by vehicle class, which led to a shift in VMT and emissions among on-road vehicle classes in the 2007 to 2020 time period.
In absence of these method changes, light-duty VMT growth between 2006 and 2007 would likely have been higher.
ES-10 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 7.6 percent since 1990.
Total direct and indirect emissions from the commercial sector have decreased by 7.7 percent since 1990. From
2019 to 2020, a decrease in heating degree days (9.4 percent) reduced energy demand for heating in the
residential and commercial sectors. This was partially offset by a 1.5 percent increase in cooling degree days
compared to 2019, which impacted demand for air conditioning in the residential and commercial sectors. This,
combined with people staying home in response to the COVID-19 pandemic, resulted in a 1.7 percent increase in
residential sector electricity use. From 2019 to 2020, the COVID-19 pandemic reduced economic and
manufacturing activity which contributed to 5.4 percent lower energy use in the commercial sector.
Electric Power. The United States relies on electricity to meet a significant portion of its energy demands.
Electricity generators used 31.2 percent of U.S. energy from fossil fuels and emitted 33.1 percent of the CO2 from
fossil fuel combustion in 2020. 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 91.4 percent of all coal
consumed for energy in the United States in 2020.17 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 19.9 percent in 2020.18 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-one-year period to represent 39.5 percent of electric power generation in 2020. Wind
and solar generation (in kWh) represented 0.1 percent of electric power generation in 1990 and increased over the
thirty-one-year period to represent 11.1 percent of electric power generation in 2020. Economic impacts of the
COVID-19 pandemic, combined with a warmer winter, led to a decrease in electricity use of about 2.5 percent in
2020, and the trend of decreased coal use and increased use of natural gas and renewable energy continued.
Between 2019 and 2020, coal electricity generation dropped by 19.9 percent, natural gas generation increased by
2.9 percent, and renewable energy generation increased by 7.9 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 20.9 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 19.2 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).
Executive Summary ES-11
-------
Figure ES-7: Electric Power Generation and Emissions
3,500
3,000
2,500 -J
LLJ
O
U
2,000
V)
c
o
1,500 "g
£
LLJ
"(O
1,000 °
500
0
Other significant CO2 trends included the following:
• Carbon dioxide emissions from natural gas and petroleum systems increased by 24.0 MMT CO2 Eq, (57.9
percent) from 1990 to 2020. 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
overtime.
• Carbon dioxide emissions from iron and steel production and metallurgical coke production have
decreased by 67.0 MMT CO2 Eq. (64.0 percent) from 1990 through 2020. 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 9.0 percent between 1990
and 2020. 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 25 over a
100-year time frame based on the IPCC Fourth Assessment Report estimate (IPCC 2007). Over the last two hundred
and fifty years, the concentration of CH4 in the atmosphere increased by 168.4 percent (IPCC 2013; NOAA/ESRL
2022b). 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).
4,500
4,000
3,500
3,000
fc 2,500
c
QJ
a3 2,000
%
a.
y 1,500
4-"
U
jy
UJ
1,000
500
Nuclear Generation (Billion kWh)
Renewable Generation (Billion kWh)
Petroleum Generation (Billion kWh)
Coal Generation (Billion kWh)
Natural Gas Generation (Billion kWh)
I Total Emissions (MMT CO2 Eq.) [Right Axis]
o th r\i co
CTi Ci Ci Q\
m vd co a* o ¦>-
o o o o O 1-1 r-
000 0 0000000000000000
ES-12 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure ES-8: 2020 Sources of ChU Emissions
Enteric Fermentation
Natural Gas Systems
Landfills
Manure Management
Other Energy
Coal Mining
LULUCF Emissions
Wastewater Treatment
Rice Cultivation
Stationary Combustion
Other Waste
Field Burning of Agricultural Residues
Other Industrial Processes
175
CO2
CH4
N2O
HFCs, PFCs, SFe and NFa
80 100
MMT CO2 Eq.
180
Note: Methane emissions from Abandoned Oil and Gas Wells, Underground Coal Mines, Incineration of Waste, and Mobile
Combustion are included in Other Energy. Methane emissions from anaerobic digestion at biogas facilities and composting
are included in Other Waste. Methane emissions from Carbide Production and Consumption, Ferroalloy Production, Iron and
Steel Production, and Petrochemical Production are included in Other Industrial Processes. 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. Cm emissions include the following:
• Enteric fermentation was the largest anthropogenic source of CH4 emissions in the United States in 2020,
accounting for 175.2 MMT CO2 Eq. of Cm (26.9 percent of total Cm emissions) and representing an
increase of 11.7 MMT CO2 Eq. (7.2 percent) since 1990. This increase in emissions from 1990 to 2020
generally follows the increasing trends in cattle populations.
• Natural gas systems were the second largest anthropogenic source category of Cm emissions in the
United States in 2020, accounting for 164.9 MMT CO2 Eq. of Cm (25.4 percent of total CH4 emissions).
Emissions decreased by 30.6 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 2020,
accounting for 109.3 MMT CO2 Eq. (16.8 percent of total CH4 emissions) and representing a decrease of
67.2 MMT CO2 Eq. (38.1 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 Emissions
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 nearly 300 times more powerful than CO2 at trapping heat in
the atmosphere over a 100-year time frame (IPCC 2007). Since 1750, the global atmospheric concentration of N2O
has risen by 23.3 percent (IPCC 2013; NOAA/ESRL 2022c). 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.
Executive Summary ES-13
-------
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: 2020 Sources of N2O Emissions
Agricultural Soil Management
Wastewater T reatment
Stationary Combustion
Manure Management
Mobile Combustion
LULUCF Emissions
Nitric Acid Production
Adipic Acid Production
Other Industrial Processes
Composting
Other Energy
Field Burning of Agricultural Residues
316
CO2
CH4
NiO
HFCs, PFCs, SFe and NFa
20
MMT COz Eq.
Note: Nitrous oxide emissions from Petroleum Systems, Natural Gas Systems, and Incineration of Waste are included in Other
Energy. Nitrous oxide emissions from Caprolactam, Glyoxal, and Glyoxylic Acid Production, Electronics Industry, and Product
Uses are included in Other Industrial Processes. 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 2020, accounting for 316.2
MMT CO2 Eq. (74.2 percent of N2O emissions) and 5.3 percent of total greenhouse gas emissions in the
United States. These emissions increased by 0.2 MMT CO2 Eq. (0.1 percent) from 1990 to 2020, but have
fluctuated during that period due to annual variations in weather patterns, fertilizer use, and crop
production.
• Wastewater treatment, both domestic and industrial, was the second largest anthropogenic source of
N2O emissions in 2020, accounting for 23.5 MMT CO2 Eq. (5.5 percent of N2O emissions) and 0.4 percent
of total greenhouse gas emissions in the United States in 2020. Emissions from wastewater treatment
increased by 6.9 MMT CO2 Eq. (41.8 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 19.7 MMT CO2 Eq. (4.6 percent of N2O
emissions) and 0.3 percent of total greenhouse gas emissions in the United States in 2020. These
emissions increased by 5.7 MMT CO2 Eq. (41.2 percent) from 1990 to 2020. 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.
• Nitrous oxide emissions from mobile combustion decreased by 27.2 MMT CO2 Eq. (61.0 percent) from
1990 to 2020, primarily as a result of national vehicle emissions standards and emission control
technologies for on-road vehicles.
ES-14 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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. One HFC, HFC-23, is 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: 2020 Sources of HFCs, PFCs, SFe, and NF3 Emissions
of Ozone Depleting Substances | 176
Electronics Industry | HFCs, PFCs, SFe, and NF3 as a
Portion of All Emissions
Electrical Transmission and Distribution
HCFC-22 Production
ICO2
Aluminum Production | ¦ CH4
IN2O
Magnesium Production and Processing I ¦ HFCs, PFCs, SFe and NF3
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 (93.2 percent) in 2020 and have
been consistently increasing, from small amounts in 1990 to 176.3 MMT CO2 Eq. in 2020. This increase
was in large part the result of efforts to phase out CFCs and other ODS in the United States.
• PFC, HFC, SFs, and NF3 emissions from the electronics industry have increased by 24.7 percent from 1990
to 2020, reflecting the competing influences of industrial growth and the adoption of emission reduction
technologies. Within that time span, emissions peaked at 9.0 MMT CO2 Eq. in 1999, the initial year of
EPA's PFC Reduction/Climate Partnership for the Semiconductor Industry, and have since declined to 4.4
MMT CO2 Eq. in 2020 (a 50.9 percent decrease relative to 1999).
• Sulfur hexafluoride emissions from electric power transmission and distribution systems decreased by
83.6 percent (19.4 MMT CO2 Eq.) from 1990 to 2020. 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.
• Emissions from HCFC-22 production were 2.1 MMT CO2 Eq. in 2020, a 95.4 percent decrease from 1990
emissions. 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.
Executive Summary ES-15
-------
• PFC emissions from aluminum production decreased by 92.2 percent (19.8 MMT CO2 Eq.) from 1990 to
2020, due to both industry emission reduction efforts and lower domestic aluminum production.
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 to promote comparability across countries. Over the thirty-one-year period of 1990 to 2020, total
emissions from the Industrial Processes and Product Use, and Agriculture sectors grew by, 30.2 MMT CO2 Eq. (8.7
percent), and 42.8 MMT CO2 Eq. (7.8 percent), respectively. Emissions from the Energy and Waste sectors
decreased by 486.5 MMT CO2 Eq. (9.1 percent) and 58.6 MMT CO2 Eq. (27.4 percent) respectively. Over the same
period, net carbon (C) sequestration in the LULUCF sector decreased by 79.8 MMT CO2 (9.0 percent decrease in
total net C sequestration), while emissions from the LULUCF sector (i.e., CH4 and N2O) increased by 21.8 MMT CO2
Eq. (69.6 percent).
Figure ES-11: U.S. Greenhouse Gas Emissions and Sinks by IPCC Sector/Category
9 000 B LULUCF (emissions) ¦ Agriculture
¦ Waste ¦ Energy
8 000 I " Industrial Processes and Product Use ¦ LULUCF (removals)
¦ Net Emissions (including LULUCF sinks)
7,000
6,000
& 5,000
LU
8 4,000
1-
s 3,000
2,000
1,000
0
-1,000
Oi-HfNn^-Lo^or^cocnoi-i(Nro^-mv£)r^cocT>Oi-trsjn^t-mvDrvooavo
0"i Ol ^ 0~> 0"> O O O O O O O O O O '—1 1—1 •>—1 •>—1 '—1 t—I *—1 1—1 ¦<—1 1—1 CsJ
ocr»cr*cncriCT*cr>o->CT>CTiooooooooooooooooooooo
1,—1,—I,—I,—ii—1-«—ii—I,—I,—ir\irM(Mr\ir\irMr\ir\ir\ir\ir\ir\ir\io>ir\ir\ir\ir\i(N(Nr\i
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
2016
2017
2018
2019
2020
Energy
5,341.1
6,319.8
5,413.1
5,372.7
5,539.5
5,409.8
4,854.7
Industrial Processes and Product Use
346.2
365.9
369.0
369.4
373.4
379.5
376.4
Agriculture
551.9
573.6
601.9
603.2
616.7
622.9
594.7
Waste
214.2
175.6
153.9
155.7
157.9
159.6
155.6
Total Gross Emissions3 (Sources)
6,453.5
7,434.8
6,537.9
6,501.0
6,687.5
6,571.7
5,981.4
LULUCF Sector Net Totalb
(860.6)
(789.8)
(826.6)
(781.2)
(769.3)
(730.5)
(758.9)
Net Emissions (Sources and Sinks)c
5,592.8
6,645.0
5,711.2
5,719.8
5,918.2
5,841.2
5,222.4
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 net carbon stock
changes in units of MMT C02 Eq.
c Net emissions with LULUCF.
ES-16 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Notes: Total emissions presented without LULUCF. Net emissions presented with LULUCF. Totals may not sum due to
independent rounding. Parentheses indicate negative values or sequestration.
Energy
The Energy chapter contains emissions of all greenhouse gases resulting from stationary and mobile energy
activities including fuel combustion and fugitive fuel emissions, and the use of fossil fuels for non-energy purposes.
Energy-related activities, primarily fossil fuel combustion, accounted for the vast majority of U.S. CO2 emissions for
the period of 1990 through 2020. Energy-related activities are also responsible for CFU and N2O emissions (41.4
percent and 9.6 percent of total U.S. emissions of each gas, respectively). Overall, emission sources in the Energy
chapter account for a combined 81.2 percent of total U.S. greenhouse gas emissions in 2020.
In 2020, 78.8 percent of the energy used in the United States (on a Btu basis) was produced through the
combustion of fossil fuels. The remaining 21.2 percent came from other energy sources, such as hydropower,
biomass, nuclear, wind, and solar energy (see Figure ES-12).
Figure ES-12: 2020 U.S. Energy Consumption by Energy Source (Percent)
Nuclear Electric Power
Industrial Processes and Product Use
The Industrial Processes and Product Use (IPPU) chapter contains information on greenhouse gas emissions
generated and emitted as the byproducts of non-energy-related industrial processes, which involve the chemical
or physical transformation of raw materials and can release waste gases such as CO2, Cm, N2O, and fluorinated
gases (e.g., HFC-23). These processes 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. Most of these industries also emit CO2 from fossil fuel combustion which, in
line with IPCC sectoral definitions, is included in the Energy Sector.
This chapter also contains information on the release of HFCs, PFCs, SF6, and NF3 and other fluorinated compounds
used in industrial manufacturing processes and by end-consumers (e.g., residential and mobile air conditioning).
These industries include electronics industry, electric power transmission and distribution, and magnesium metal
production and processing. In addition, N2O is used in and emitted by electronics industry and anesthetic and
aerosol applications, and CO2 is consumed and emitted through various end-use applications. In 2020, emissions
Executive Summary ES-17
-------
resulting from use of the substitution of ODS (e.g., chlorofluorocarbons [CFCs]) by end-consumers was the largest
source of IPPU emissions and accounted for 176.3 MMT CO2 Eq, or 46.8 percent of total IPPU emissions.
IPPU activities are responsible for 3.5, 0.1, and 5.5 percent of total U.S. CO2, CFU, and N2O emissions respectively as
well as for all U.S. emissions of fluorinated gases including HFCs, PFCs, SF6 and NF3. Overall, emission sources in the
IPPU chapter accounted for 6.3 percent of U.S. greenhouse gas emissions in 2020.
Agriculture
The Agriculture chapter contains information on anthropogenic emissions from agricultural activities (except fuel
combustion, which is addressed in the Energy chapter, and some agricultural CO2, CFU, and N2O fluxes, which are
addressed in the Land Use, Land-Use Change, and Forestry chapter).
Agricultural activities contribute directly to emissions of greenhouse gases through a variety of processes,
including the following source categories: agricultural soil management, enteric fermentation in domestic
livestock, livestock manure management, rice cultivation, urea fertilization, liming, and field burning of agricultural
residues.
In 2020, agricultural activities were responsible for emissions of 549.7 MMT CO2 Eq., or 9.9 percent of total U.S.
greenhouse gas emissions. Methane, N2O, and CO2 are greenhouse gases emitted by agricultural activities.
Methane emissions from enteric fermentation and manure management represented 26.9 percent and 9.2
percent of total CH4 emissions from anthropogenic activities, respectively, in 2020. Agricultural soil management
activities, such as application of synthetic and organic fertilizers, deposition of livestock manure, and growing N-
fixing plants, were the largest contributors to U.S. N2O emissions in 2020, accounting for 74.2 percent of total N2O
emissions. 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.
Land Use, Land-Use Change, and Forestry
The LULUCF chapter contains emissions and removals of CO2 and emissions of CFU and N2O from managed lands in
the United States. Consistent with the 2006IPCC Guidelines, emissions and removals from managed lands are
considered to be anthropogenic, while emissions and removals from unmanaged lands are considered to be
natural.20 The share of managed land in the U.S. is approximately 95 percent of total land included in the
Inventory.21 More information on the definition of managed land used in the Inventory is provided in Chapter 6.
Overall, the Inventory results show that managed land is a net sink for CO2 (C sequestration). The primary drivers
of fluxes on managed lands include forest management practices, tree planting in urban areas, the management of
agricultural soils, lands remaining and lands converted to reservoirs and other constructed waterbodies, landfilling
of yard trimmings and food scraps, and activities that cause changes in C stocks in coastal wetlands. The main
drivers for forest C sequestration include forest growth and increasing forest area (i.e., afforestation), as well as a
net accumulation of C stocks in harvested wood pools. The net sequestration in Settlements Remaining
Settlements, which occurs predominantly from urban forests (i.e., Settlement Trees) and landfilled yard trimmings
and food scraps, is a result of net tree growth and increased urban forest area, as well as long-term accumulation
of yard trimmings and food scraps carbon in landfills.
20 See http://www.ipcc-nggiD.iges.or.jP/public/2CX36gl/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.
ES-18 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
The LULUCF sector in 2020 resulted in a net increase in C stocks (i.e., net CO2 removals) of 812.2 MMT CO2 Eq.22
The removals of C offset 13.6 percent of total (i.e., gross) greenhouse gas emissions in 2020. Emissions of CH4 and
N2O from LULUCF activities in 2020 were 53.2 MMT CO2 Eq. and represent 0.9 percent of total greenhouse gas
emissions.23 Carbon dioxide removals from C stock changes are presented in Table ES-4 along with CH4 and N2O
emissions for LULUCF source categories.
Between 1990 and 2020, total C sequestration in the LULUCF sector decreased by 9.0 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. The overall net flux from LULUCF (i.e., net sum of all Cm and
N2O emissions to the atmosphere plus net carbon stock changes in units of MMT CO2 eq.) resulted in a removal of
758.9 MMT C02 Eq. in 2020.
Flooded lands were the largest source of CH4 emissions from the LULUCF sector in 2020, totaling 19.9 MMT CO2
Eq. (797 kt of CH4). Forest fires were the second largest source and resulted in Cm emissions of 13.6 MMT CO2 Eq.
(545 kt of CH4), followed by Coastal Wetlands Remaining Coastal Wetlands with CH4 emissions of 3.8 MMT CO2 Eq.
(154 kt of CH4).
Forest fires were the largest source of N2O emissions from the LULUCF sector in 2020, totaling 11.7 MMT CO2 Eq.
(39 kt of N2O). Nitrous oxide emissions from fertilizer application to settlement soils in 2020 totaled 2.5 MMT CO2
Eq. (8 kt of N2O).
Table ES-4: U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use, Land-
Use Change, and Forestry (MMT CO2 Eq.)
Land-Use Category
1990
2005
2016
2017
2018
2019
2020
Forest Land Remaining Forest Land3
(769.7)
(674.0)
(717.3)
(670.1)
(664.6)
(631.8)
(642.2)
Land Converted to Forest Landb
(98.6)
(99.1)
(99.5)
(99.5)
(99.5)
(99.5)
(99.5)
Cropland Remaining Cropland
(23.2)
(29.0)
(22.7)
(22.3)
(16.6)
(14.5)
(23.3)
Land Converted to Cropland0
51.8
52.0
54.1
54.3
54.0
53.9
54.4
Grassland Remaining Grasslandd
7.1
9.4
8.6
9.9
10.3
13.1
5.1
Land Converted to Grassland0
(3.1)
(37.0)
(22.6)
(22.7)
(22.4)
(21.5)
(24.1)
Wetlands Remaining Wetlandse
14.7
17.2
15.8
15.9
15.9
15.9
15.8
Land Converted to Wetlandse
7.1
1.2
0.6
0.6
0.6
0.6
0.6
Settlements Remaining Settlements'
(107.6)
(113.5)
(121.5)
(125.3)
(124.9)
(124.5)
(123.7)
Land Converted to Settlements0
60.8
82.8
77.8
77.9
78.0
77.9
77.9
LULUCF Carbon Stock Change/C02g
(892.0)
(831.1)
(862.0)
(826.7)
(809.0)
(760.8)
(812.2)
LULUCF CH4 and N20 Emissions'1
31.4
41.3
35.4
45.5
39.8
30.3
53.2
ch4
27.2
30.9
28.3
34.0
30.7
25.5
38.1
n2o
4.2
10.5
7.1
11.5
9.1
4.8
15.2
LULUCF Sector Net Total1
(860.6)
(789.8)
(826.6)
(781.2)
(769.3)
(730.5)
(758.9)
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.
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.
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.
Executive Summary ES-19
-------
d Estimates include CH4 and N20 emissions from fires on both Grassland Remaining Grassland and Land Converted to
Grass/and.
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
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 net carbon stock
changes in units of MMT C02 Eq.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Waste
The Waste chapter contains emissions from waste management activities (except incineration of waste, which is
addressed in the Energy chapter). Landfills were the largest source of anthropogenic greenhouse gas emissions
from waste management activities, generating 109.3 MMT CO2 Eq. and accounting for 70.3 percent of total
greenhouse gas emissions from waste management activities, and 16.8 percent of total U.S. CH4 emissions.24
Additionally, wastewater treatment generated emissions of 41.8 MMT CO2 Eq. and accounted for 26.9 percent of
total Waste sector greenhouse gas emissions, 2.8 percent of U.S. CH4 emissions, and 5.5 percent of U.S. N2O
emissions in 2020. Emissions of CH4 and N2O from composting are also accounted for in this chapter, generating
emissions of 2.3 MMT CO2 Eq., and 2.0 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 155.6 MMT CO2 Eq., or 2.6 percent of
total U.S. greenhouse gas emissions in 2020.
ES.4 Other Information
Emissions by Economic Sector
Throughout the Inventory of U.S. Greenhouse Gas Emissions and Sinks report, emission estimates are grouped into
five sectors (i.e., chapters) defined by the IPCC: 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. For
more information on trends in the Land use, Land Use Change and Forestry sector, see Section ES.2 Recent Trends
in U.S. Greenhouse Gas Emissions and Sinks.
Figure ES-13 shows the trend in emissions by economic sector from 1990 to 2020, and Table ES-5 summarizes
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.
ES-20 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure E
2,500
2,000
t! 1-500
0
U
I-
1 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
2016
2017
2018
2019
2020
Transportation
1,526.4
1,975.5
1,828.0
1,845.2
1,874.7
1,874.3
1,627.6
Electric Power Industry
1,880.5
2,456.7
1,860.5
1,780.6
1,799.8
1,651.0
1,482.2
Industry
1,652.4
1,536.2
1,424.4
1,446.7
1,507.6
1,521.7
1,426.2
Agriculture
596.8
626.3
643.4
644.4
657.9
663.9
635.1
Commercial
427.1
405.4
426.9
428.5
444.2
452.1
425.3
Residential
345.1
371.0
327.8
329.9
377.4
384.2
362.0
U.S. Territories
25.1
63.7
26.8
25.8
25.8
24.6
23.0
Total Gross Emissions (Sources)
6,453.5
7,434.8
6,537.9
6,501.0
6,687.5
6,571.7
5,981.4
LULUCF Sector Net Total3
(860.6)
(789.8)
(826.6)
(781.2)
(769.3)
(730.5)
(758.9)
Net Emissions (Sources and Sinks)
5,592.8
6,645.0
5,711.2
5,719.8
5,918.2
5,841.2
5,222.4
a The LULUCF Sector Net Total is the net sum of all LULUCF CH4 and N20 emissions to the atmosphere plus net
carbon stock changes.
Notes: Total emissions presented without LULUCF. Total net emissions 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 (27.2 percent)
of total U.S. greenhouse gas emissions in 2020. Electric power accounted for the second largest portion (24.8
percent) of U.S. greenhouse gas emissions in 2020, while emissions from industry accounted for the third largest
portion (23.8 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 24.2 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 10.6 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 7.1 percent and 6.1 percent of emissions, respectively, and U.S. Territories accounted for 0.4
:-13: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors
Executive Summary ES-21
-------
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 2022 and Duffield 2006).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, CH4 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 (30.3 percent and 27.3 percent,
respectively) in 2020. The residential and commercial sectors contributed the next largest shares of total U.S.
greenhouse gas emissions in 2020 (15.4 and 15.4 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 2020.
Table ES-6: U.S. Greenhouse Gas Emissions with Electricity-Related Emissions Distributed
by Economic Sector (MMT CO2 Eq.)
Economic Sectors
1990
2005
2016
2017
2018
2019
2020
Industry
2,326.5
2,251.6
1,917.5
1,926.4
1,983.1
1,964.7
1,813.7
Transportation
1,529.6
1,980.3
1,832.4
1,849.6
1,879.5
1,879.1
1,632.4
Residential
957.6
1,247.2
999.9
964.3
1,036.7
984.1
923.1
Commercial
982.7
1,227.4
1,078.6
1,051.7
1,065.3
1,020.1
919.7
Agriculture
631.9
664.6
682.6
683.2
697.1
699.1
669.5
U.S. Territories
25.1
63.7
26.8
25.8
25.8
24.6
23.0
Total Gross Emissions (Sources)
6,453.5
7,434.8
6,537.9
6,501.0
6,687.5
6,571.7
5,981.4
LULUCF Sector Net Total3
(860.6)
(789.8)
(826.6)
(781.2)
(769.3)
(730.5)
(758.9)
Net Emissions (Sources and Sinks)
5,592.8
6,645.0
5,711.2
5,719.8
5,918.2
5,841.2
5,222.4
a The LULUCF Sector Net Total is the net sum of all LULUCF CH4 and N20 emissions to the atmosphere plus 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.
ES-22 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure ES-14: U.S. Greenhouse Gas Emissions with Electricity-Related Emissions Distributed
to Economic Sectors
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 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 0.2 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 1.4 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.
Table ES-7: Recent Trends in Various U.S. Data (Index 1990 = 100)
Avg. Annual Avg. Annual
Growth Rate Growth Rate
Variable 1990 2005 2016 2017 2018 2019 2020 since 1990a Since 2005a
Greenhouse Gas Emissions'5
Energy Usec
GDPd
Population6
100 115
100 119
100 159
100 118
101
101
104
102
93
116
116
120
119
109
189
193
199
203
196
128
129
129
131
132
-0.2% -1.4%
0.3% -0.5%
2.3% 1.4%
0.9% 0.8%
Executive Summary ES-23
-------
a Average annual growth rate.
b 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 (2021), 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 2020 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.
ES-24 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure ES-16: 2020 Key Categories (Approach 1 including LULUCF)*
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
Direct N2O Emissions from Agricultural Soil Management
CO2 Emissions from Stationary Combustion - Gas - Residential
CO2 Emissions from Stationary Combustion - Oil - Industrial
CO2 Emissions from Stationary Combustion - Gas - Commercial
CH4 Emissions from Enteric Fermentation: Cattle
CH4 Emissions from Natural Gas Systems
Emissions from ODS Substitutes: Refrigeration and Air Conditioning
Net Carbon Stock Change from Settlements Remaining Settlements
CO2 Emissions from Mobile Combustion: Aviation
CO2 Emissions from Non-Energy Use of Fuels
Net Carbon Stock Change from Land Converted to Forest Land
CH4 Emissions from Commercial Landfills
Net Carbon Stock Change from Land Converted to Settlements
CO2 Emissions from Stationary Combustion - Oil - Residential
CO2 Emissions from Mobile Combustion: Other
Net Carbon Stock Change from Land Converted to Cropland ¦
CO2 Emissions from Stationary Combustion - Oil - Commercial I
Indirect N2O Emissions from Applied Nitrogen I
CO2 Emissions from Stationary Combustion - Coal - Industrial
Fugitive Emissions from Coal Mining
CO2 Emissions from Cement Production
CH4 Emissions from Petroleum Systems
CO2 Emissions from Iron and Steel Production & Metallurgical Coke Production
CO2 Emissions from Natural Gas Systems
CH4 Emissions from Manure Management: Cattle
CO2 Emissions from Mobile Combustion: Railways
CO2 Emissions from Petroleum Systems
CO2 Emissions from Petrochemical Production
CH4 Emissions from Manure Management: Other Livestock
Net Carbon Stock Change from Land Converted to Grassland
CO2 Emissions from Mobile Combustion: Marine
Net Carbon Stock Change from Cropland Remaining Cropland
N2O Emissions from Domestic Wastewater Treatment II
CH4 Emissions from Flooded Lands Remaining Flooded Lands I
Emissions from Substitutes for Ozone Depleting Substances: Aerosols
CO2 Emissions from Stationary Combustion - Oil - U.S. Territories
CO2 Emissions from Stationary Combustion - Oil - Electricity Generation
CH4 Emissions fr om Rice Cultivation
N2O Emissions from Stationary Combustion - Coal - Electricity Generation
CH4 Emissions from Abandoned Oil and Gas Wells
Net Carbon Stock Change from Grassland Remaining Grassland
CH4 Emissions from Stationary Combustion - Residential
Key Categories as a Portion of
All Emissions
95.9%
I Key Categories LULUCF
Other Categories
I Key Categories
0 200 400 600 800 1,000 1,200
2020 Emissions (MMT CO2 Eq.)
a 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.
Executive Summary ES-25
-------
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 these 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 one of most
suitable greenhouse gases for such comparisons. The 2019 Refinement makes this conclusion on fluorinated
gases based on the lack of confounding natural sources, the potential uncertainties in bottom-up inventory
methods, the long atmospheric lifetimes of many of these gases, and the well-known loss mechanisms. Unlike
most other gases in the Inventory, since there are no known natural sources of hydrofluorocarbons (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. More information on findings from applying this guidance in comparing recent HFC emission studies
conducted by NOAA with modeled bottom-up emissions are included under the QA/QC and Verification
discussion in Chapter 4, Section 4.24 Substitution of Ozone Depleting Substances 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).
Finally, in addition to use of atmospheric concentration measurement data for comparison with Inventory data,
in this year's Inventory, information from top-down studies has been directly incorporated in the Natural Gas
Systems calculations to quantify emissions from well blowout events. For more information, see Recalculations
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.
ES-26 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Discussion section in 3.6 Natural Gas Systems.
Uncertainty Analysis of Emission 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, 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.
Executive Summary ES-27
-------
1. Introduction
This report presents estimates by the United States government of U.S. anthropogenic greenhouse gas emissions
and sinks for the years 1990 through 2020. 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 these emission 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
Inventory of anthropogenic greenhouse gas emissions and sinks provides a common and consistent mechanism
1 More information provided in the Global Warming Potentials section of this chapter on the use of IPCC Fourth Assessment
Report (AR4) 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
-------
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 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).7 The GHGRP applies to direct
greenhouse gas emitters, fossil fuel suppliers, industrial gas suppliers, and facilities that inject carbon dioxide
(C02) underground for sequestration or other reasons and requires reporting by over 8,000 sources or suppliers
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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
in 41 industrial categories.1 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 CO 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).
1.1 Background Information
Science
For over the past 200 years, the burning of fossil fuels such as coal and oil, along with deforestation, land-use
changes, and other activities have caused the concentrations of heat-trapping "greenhouse gases" to increase
significantly in our atmosphere (IPCC 2021). These gases in the atmosphere absorb some of the energy being
radiated from the surface of the Earth that would otherwise be lost to space, essentially acting like a blanket that
makes the Earth's surface warmer than it would be otherwise.
Greenhouse gases are necessary to life as we know it. Without greenhouse gases to create the natural heat-
trapping properties of the atmosphere, the planet's surface would be about 60 degrees Fahrenheit cooler than
present (USGCRP 2017). Carbon dioxide is also necessary for plant growth. With emissions from biological and
geological sources, there is a natural level of greenhouse gases that is maintained in the atmosphere. Human
emissions of greenhouse gases and subsequent changes in atmospheric concentrations alter the balance of energy
transfers between space and the earth system (IPCC 2021). A gauge of these changes is called radiative forcing,
which is a measure of a substance's total net effect on the global energy balance for which a positive number
represents a warming effect and a negative number represents a cooling effect (IPCC 2021). IPCC concluded in its
most recent scientific assessment report that it is "unequivocal that human influence has warmed the atmosphere,
ocean and land" (IPCC 2021).
As concentrations of greenhouse gases continue to increase in from man-made sources, the Earth's temperature is
climbing above past levels. The Earth's average land and ocean surface temperature has increased by about 2.0
degrees Fahrenheit from the 1850 to 1900 period to the decade of 2011 to 2020 (IPCC 2021). The last four decades
have each been the warmest decade successively at the Earth's surface since at least 1850 (IPCC 2021). Other
aspects of the climate are also changing, such as rainfall patterns, snow and ice cover, and sea level. If greenhouse
gas concentrations continue to increase, climate models predict that the average temperature at the Earth's
surface is likely to increase by up to 8.3 degrees Fahrenheit above 2011 to 2020 levels by the end of this century,
8 See http://www.epa.gov/ehgreporting and http://ghgdata.epa.gov/ghgp/main.do.
Introduction 1-3
-------
depending on future emissions and the responsiveness of the climate system (IPCC 2021), though the lowest
emission scenario would limit future warming to an additional 0.5 degrees (best estimate).
For further information on greenhouse gases, radiative forcing, and implications for climate change, see the recent
scientific assessment reports from the IPCC,9 the U.S. Global Change Research Program (USGCRP),10 and the
National Academies of Sciences, Engineering, and Medicine (NAS).11
Greenhouse Gases
Although the Earth's atmosphere consists mainly of oxygen and nitrogen, neither plays a significant role in
enhancing the greenhouse effect because both are essentially transparent to terrestrial radiation. The greenhouse
effect is primarily a function of the concentration of water vapor, carbon dioxide (CO2), methane (CH4), nitrous
oxide (N2O), and other trace gases in the atmosphere that absorb the terrestrial radiation leaving the surface of
the Earth (IPCC 2021).
Naturally occurring greenhouse gases include water vapor, CO2, CH4, N2O, and ozone (O3). Several classes of
halogenated substances that contain fluorine, chlorine, or bromine are also greenhouse gases, but they are, for the
most part, solely a product of industrial activities. Chlorofluorocarbons (CFCs) and hydrochlorofluorocarbons
(HCFCs) are halocarbons that contain chlorine, while halocarbons that contain bromine are referred to as
bromofluorocarbons (i.e., halons). As stratospheric ozone depleting substances, CFCs, HCFCs, and halons are
covered under the Montreal Protocol on Substances that Deplete the Ozone Layer. The UNFCCC defers to this
earlier international treaty. Consequently, Parties to the UNFCCC are not required to include these gases in
national greenhouse gas inventories.12 Some other fluorine-containing halogenated substances—
hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), sulfur hexafluoride (SFs), and nitrogen trifluoride (NF3)—do
not deplete stratospheric ozone but are potent greenhouse gases. These latter substances are addressed by the
UNFCCC and accounted for in national greenhouse gas inventories.
There are also several other substances that influence the global radiation budget but are short-lived and
therefore not well-mixed, leading to spatially variable radiative forcing effects. These substances include carbon
monoxide (CO), nitrogen dioxide (NO2), sulfur dioxide (SO2), and tropospheric (ground level) ozone (O3).
Tropospheric ozone is formed from chemical reactions in the atmosphere of precursor pollutants, which include
volatile organic compounds (VOCs, including CH4) and nitrogen oxides (NOx), in the presence of ultraviolet light
(sunlight).
Aerosols are extremely small particles or liquid droplets suspended in the Earth's atmosphere that are often
composed of sulfur compounds, carbonaceous combustion products (e.g., black carbon), crustal materials (e.g.,
dust) and other human-induced pollutants. They can affect the absorptive characteristics of the atmosphere (e.g.,
scattering incoming sunlight away from the Earth's surface, or, in the case of black carbon, absorb sunlight) and
can play a role in affecting cloud formation and lifetime, as well as the radiative forcing of clouds and precipitation
patterns.
Carbon dioxide, CFU, and N2O are continuously emitted to and removed from the atmosphere by natural processes
on Earth. Anthropogenic activities (such as fossil fuel combustion, cement production, land-use, land-use change,
and forestry, agriculture, or waste management), however, can cause additional quantities of these and other
greenhouse gases to be emitted or sequestered, thereby changing their global average atmospheric
concentrations. Natural activities such as respiration by plants or animals and seasonal cycles of plant growth and
decay are examples of processes that only cycle carbon or nitrogen between the atmosphere and organic biomass.
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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
414 ppma
1.879 ppmb
0.333 ppmc
10.27 pptd
85.5 ppt0
Rate of concentration change
2.32 ppm/yrf
7.91 ppb/yrf'g
0.97 ppb/yrf
0.32 ppt/yrf
0.81 ppt/yrf
Atmospheric lifetime (years)
See footnote11
11.8
109'
About l,000i
50,000
a The atmospheric C02 concentration is the 2020 annual average at the Mauna Loa, HI station (NOAA/ESRL 2021a). The global
atmospheric C02 concentration, computed using an average of sampling sites across the world, was 412 ppm in 2019.
b The values presented are global 2020 annual average mole fractions (NOAA/ESRL 2021b).
c The values presented are global 2020 annual average mole fractions (NOAA/ESRL 2021c).
d The values presented are global 2020 annual average mole fractions (NOAA/ESRL 2021d).
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 2020 and has fluctuated between 1.5 to
3.0 ppm per year over this period (NOAA/ESRL 2021a). The rate of concentration change for CH4, N20, and SF6, is the average
rate of change between 2007 and 2020 (NOAA/ESRL 2021b; NOAA/ESRL 2021c; NOAA/ESRL 2021d). 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 7.91 ppb/year (NOAA/ESRL 2021b).
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
geochemical and biological processes. Carbon dioxide concentrations in the atmosphere increased from
Introduction 1-5
-------
approximately 280 parts per million by volume (ppmv) in pre-industrial times to 416 ppmv in 2020, a 49 percent
increase (IPCC 2021; NOAA/ESRL 2021a).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 157 percent since 1750, from a pre-industrial value of about 730 ppb to 1,879 ppb in
202015 although the rate of increase decreased to near zero in the early 2000s, and has recently increased again to
about 7.91 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
23 percent since 1750, from a pre-industrial value of about 270 ppb to 333 ppb in 2020,16 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
ozone in the stratosphere has resulted in negative radiative forcing, representing an indirect effect 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 2020 annual average mole fraction (NOAA/ESRL 2021b).
16 This value is the global 2020 annual average (NOAA/ESRL 2021c).
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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 (IPCC 2013). 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 of HFCs (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, 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.
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
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
-------
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 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 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
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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
( MMT
MMT C02 Eq. = (kt of gas) x (GWP) x
MMT
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
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. Parties to the UNFCCC are required to use GWPs based
upon a 100-year time horizon from the IPCC Fourth Assessment Report (AR4), although other time horizon values
are available.
...the global warming potential values used by Parties included in Annex I to the Convention (Annex I
Parties) to calculate the carbon dioxide equivalence of anthropogenic emissions by sources and removals by
sinks of greenhouse gases shall be those listed in the column entitled "Global warming potential for given
time horizon" in table 2.14 of the errata to the contribution of Working Group I to the Fourth Assessment
Report of the Intergovernmental Panel on Climate Change, based on the effects of greenhouse gases over a
100-year time horizon...24
All estimates are provided throughout the report in both MMT CO2 equivalents and unweighted units. The
UNFCCC will require countries to shift to use of IPCC Fifth Assessment Report (AR5) (IPCC 2013) 100-year GWP
values in 2024, when countries submit their first reports using updated reporting guidelines under the Paris
Agreement.25 A comparison of emission estimates using the 100-year AR4 GWP values versus the AR5 GWP values
and the IPCC Sixth Assessment Report (AR6) (IPCC 2021) is outlined in Box 1-2 below and, in more detail, in Annex
6.1 of this report.
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
atmosphere.
23 Carbon comprises 12/44ths of carbon dioxide by weight.
24 Framework Convention on Climate Change; Available online at: http://unfccc.int/resource/docs/2Q13/copl9/eng/lQaQ3.pdf;
31 January 2014; Report of the Conference of the Parties at its nineteenth session; held in Warsaw from 11 to 23 November
2013; Addendum; Part two: Action taken by the Conference of the Parties at its nineteenth session; Decision 24/CP.19; Revision
of the UNFCCC reporting guidelines on annual inventories for Parties included in Annex I to the Convention; p. 2. (UNFCCC
2014).
25 See https://unfccc.int/process-and-meetings/transparencv-and-reporting/reporting-and-review-under-the-paris-agreement
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
25
n2o
114
298
HFC-23
270
14,800
HFC-32
4.9
675
HFC-41d
3.7
92
HFC-125
29
3,500
HFC-134a
14
1,430
HFC-143a
52
4,470
HFC-152a
1.4
124
HFC-227ea
34.2
3,220
HFC-236fa
240
9,810
cf4
50,000
7,390
c2f6
10,000
12,200
C3Fs
2,600
8,830
c-C4Fs
3,200
10,300
sf6
3,200
22,800
nf3
740
17,200
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 IPCCSixth 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 AR4 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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
required by the 2013 revision to the UNFCCC reporting guidelines for national inventories.26 Updated reporting
guidelines under the Paris Agreement which require the United States and other countries to shift to use of IPCC
Fifth Assessment Report (AR5) (IPCC 2013) 100-year GWP values (without feedbacks) take effect for national
inventory reporting in 2024.27 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 AR4
AR5 with
AR5 with
Gas
AR4 AR5a
feedbacks'1
AR6C
AR5
feedbacks'1 AR6C
C02
1
1
1
1
NC
NC
NC
CH4d
25
28
34
27
3
9
2
N20
298
265
298
273
(33)
NC
(25)
HFC-23
14,800
12,400
13,856
14,600
(2,400)
(944)
(200)
HFC-32
675
677
817
771
2
142
96
HFC-41
92
116
NA
135
24
NA
43
HFC-125
3,500
3,170
3,691
3,740
(330)
191
240
HFC-134a
1,430
1,300
1,549
1,526
(130)
119
96
HFC-143a
4,470
4,800
5,508
5,810
330
1,038
1,340
HFC-152a
124
138
167
164
14
43
40
HFC-227ea
3,220
3,350
3,860
3,600
130
640
380
HFC-236fa
9,810
8,060
8,998
8,690
(1,750)
(812)
(1,120)
cf4
7,390
6,630
7,349
7,380
(760)
(41)
(10)
c2f6
12,200
11,100
12,340
12,400
(1,100)
140
200
C3Fs
8,830
8,900
9,878
9,290
70
(1,048)
460
c-C4Fs
10,300
9,540
10,592
10,200
(760)
292
(100)
sf6
22,800
23,500
26,087
25,200
700
3,287
2,400
nf3
17,200
16,100
17,885
17,400
(1,100)
685
200
NA (Not Applicable)
NC (No Change)
a The GWPs presented here are the ones most consistent with the methodology used in the AR4 report.
b The GWP values presented here from the AR5 report 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 100-year GWPs from AR6 are prepublication values based on the Working Group 1 report published in August
2021. As the report is finalized for full publication, in the final editing process, these values may be updated in
corrigenda and EPA will update this analysis to reflect the final values.
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 (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
26 See http://unfccc.int/resource/docs/2013/copl9/ene/10a03.pdf.
27 See https://unfccc.int/process-and-meetings/transparencv-and-reporting/reporting-and-review-under-the-paris-agreement.
Introduction 1-11
-------
supplying data to, planning methodological approaches and improvements, reviewing, or preparing portions of the
U.S. Inventory—including federal and state government authorities, research and academic institutions, industry
associations, and private consultants.
Within EPA, the Office of Atmospheric Programs (OAP) is the lead office responsible for the emission 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 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 calculations,
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 IPCC good 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
Agriculture and
LULUCF Data Sources
Industrial Processes
and Product Use Data
Waste Data Sources
2. Emissions
Calculations
U.S. Environmental
Protection Agency
Other U.S.
Government Agencies
(USFS, NOAA, DOD,
USGS, FAA)
3. Inventory
Compilation
U.S. Environmental
Protection Agency
Inventory Compiler
4. Inventory
Submission
U.S. Department
of State
United Nations
Framework
Convention on
Climate Change
1-12 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Overview of Inventory Data Sources by Source and Sink Category
Energy
Agriculture and LULUCF IPPU Waste
Energy Information
Administration
EPA Office of Land and EmergencyEPA Greenhouse Gas Reporting EPA Greenhouse Gas
Management Program (GHGRP) Reporting Program (GHGRP)
U.S. Department of Commerce Alaska Department of Natural American Chemistry Council EPA Office of Land and
- Bureau of the Census Resources (ACC) Emergency Management
U.S. Department of Defense -
Defense Logistics Agency
Federal Highway
Administration
EPA Acid Rain Program
EPA Office of Transportation
and Air Quality MOVES Model
EPA Greenhouse Gas Reporting
Program (GHGRP)
U.S. Department of Labor -
Mine Safety and Health
Administration
National Oceanic and
Atmospheric Administration
(NOAA)
Association of American Plant
Food Control Officials (AAPFCO)
U.S. Census Bureau
U.S. Department of Agriculture
(USDA) Animal and Plant Health
Inspection Service (APHIS)
EPA Office of Research and
Development
USDA National Agricultural
Statistics Service and Agricultural
Research Service
U.S. Geological Survey (USGS)
National Minerals Information
Center
American Iron and Steel
Institute (AISI)
U.S. Aluminum Association
U.S. International Trade
Commission (USITC)
Air-Conditioning, Heating, and
Refrigeration Institute
Data from other U.S.
government agencies, research
studies, trade publications, and
industry associations
Data from research studies,
trade publications, and
industry associations
American Association of USDA U.S. Forest Service Forest
Railroads Inventory and Analysis Program
American Public Transportation USDA Natural Resource
Association Conservation Service (NRCS)
U.S. Department of Homeland USDA Economic Research Service
Security (ERS)
U.S. Department of Energy and USDA Farm Service Agency (FSA)
its National Laboratories
Federal Aviation Administration U.S. Geological Survey (USGS)
U.S. Department of
Transportation & Bureau of
Transportation Statistics
U.S. Department of the Interior
(DOI), Bureau of Land
Management (BLM)
Data from research studies, Data from research studies, trade
trade publications, and industrypublications, and industry
associations associations
Note: This table is not an exhaustive list of all data sources.
1.3 Inventory Process
This section describes EPA's approach to preparing the annual U.S. Inventory, which consists of a National
Inventory Report (NIR) and Common Reporting Format (CRF) tables. The inventory coordinator at EPA, with
support from the cross-cutting compilation staff, is responsible for aggregating all emission and removal estimates,
conducting the overall uncertainty analysis of Inventory emissions and trends over time, and ensuring consistency
and quality throughout the NIR and CRF tables. Emission 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 contractors
Introduction 1-13
-------
familiar with the sources. Each year, the coordinator overseas 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 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 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 assurance and quality control (QA/QC) checks,
and category-level uncertainty analyses.
The treatment of confidential business information (CBI) in the Inventory is based on EPA internal guidelines, as
well as regulations28 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.29 In the Inventory, EPA is publishing only data values that meet
the GHGRP aggregation criteria.30 Specific uses of aggregated facility-level data are described in the respective
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
28 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.
29 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.
30 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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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, 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. Similarly, the
recalculation analysis and key category analysis is completed in a separate 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 the efficient storage, sharing, and
archiving of electronic files.
National Inventory Report 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 from all sources
discussed in the chapters. The inventory coordinator then carries out a key category analysis for the Inventory,
consistent with the 2006IPCC Guidelines for National Greenhouse Gas Inventories, and in accordance with the
reporting requirements of the UNFCCC. Also at this time, the Introduction, Executive Summary, and Trends in
Greenhouse Gas Emissions chapters are drafted, to reflect the trends for the most recent year 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. Text boxes are also created to 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 match the specification of the UNFCCC
reporting guidelines for National Inventory Reports.
Common Reporting Format Table Compilation
The CRF tables are compiled from individual tables completed by each individual source or sink category lead,
which contain emissions and/or removals and activity data. The inventory coordinator integrates the category data
into the UNFCCC's "CRF Reporter" for the United States, assuring consistency across all sectoral tables. The
summary reports for emissions, methods, and emission factors used, the overview tables for completeness of
estimates, the recalculation tables, the notation key completion tables, and the emission trends tables are then
completed by the inventory coordinator. Internal automated quality checks on the CRF Reporter, as well as
reviews by the category leads, are completed for the entire time series of CRF tables before submission.
QA/QC and Uncertainty
QA/QC and uncertainty analyses are guided by the QA/QC and Inventory coordinators, who help maintain the
QA/QC plan and the overall uncertainty analysis procedures (see sections on QA/QC and Uncertainty, below). This
coordinator works closely with the Inventory coordinator and source and sink category leads to ensure that a
consistent QA/QC plan and uncertainty analysis is implemented across all inventory sources. The inventory QA/QC
plan, outlined in Section 1.7 and Annex 8, is consistent with the quality assurance procedures outlined by EPA and
IPCC good practices. The QA/QC and uncertainty findings also inform overall improvement planning, and specific
improvements are noted in the Planned Improvements sections of respective categories. QA processes are
outlined below.
Introduction 1-15
-------
Expert, Public, and UNFCCC Reviews
The compilation of the inventory includes a two-stage review process, in addition to international technical expert
review following submission of the report. During the first stage (the 30-day Expert Review period), a first draft of
sectoral chapters of the document are sent to a select list of technical experts outside of EPA who are not directly
involved in preparing estimates. The purpose of the Expert Review is to provide an objective review, encourage
feedback on the methodological and data sources used in the current Inventory, especially for sources which have
experienced any changes since the previous Inventory.
Once comments are received and addressed, the second stage, or second draft of the document is released for
public review by publishing a notice in the U.S. Federal Register and posting the entire draft Inventory document
on the EPA website. The Public Review period allows for a 30-day comment period and is open to the entire U.S.
public. Comments may require further discussion with experts and/or additional research, and specific Inventory
improvements requiring further analysis as a result of comments are noted in the relevant category's Planned
Improvement section. EPA publishes responses to comments received during both reviews with the publication of
the final report on its website.
Following completion and submission of the report to the UNFCCC, the report also undergoes review by an
independent international team of experts for adherence to UNFCCC reporting guidelines and IPCC guidance.31
Feedback from all review processes that contribute to improving inventory quality over time are described further
in Annex 8.
Final Submittal to UNFCCC and Document Publication
After the final revisions to incorporate any comments from the Expert Review and Public Review periods, EPA
prepares the final NIR and the accompanying CRF tables for electronic reporting. EPA, as the National Inventory
focal point, sends the official submission of the U.S. Inventory to the UNFCCC using the CRF Reporter software,
coordinating with the U.S. Department of State, the overall UNFCCC focal point. Concurrently, for timely public
access, the report is also published on EPA's website.32
1.4 Methodology and Data Sources
Emissions 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 and emission factors. Depending on the emission source 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 categories, and many of these methodologies continue to be improved and refined as new research and
data become available. This report uses the IPCC methodologies when applicable, and supplements them with
other available country-specific methodologies and data where possible. For examples, as noted earlier in this
chapter, this report does use supplements and refinements to 2006 IPCC Guidelines in estimating emissions and
31 See https://unfccc.int/process-and-meetings/transparencv-and-reporting/reporting-and-review-under-the-
convention/greenhouse-gas-inventories-annex-i-parties/review-process.
32 See https://www.epa.gov/ghgemissions/inventory-us-greenhouse-gas-emissions-and-sinks.
1-16 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
removals from coal mining, wastewater, Low Voltage Anode Effects (LVAE) during aluminum production, drained
organic soils, and management of wetlands, included flooded lands. Choices made regarding the methodologies
and data sources used are provided in conjunction with the discussion of each source category in the main body 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
of those respective source categories (e.g., Coastal Wetlands).
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."33 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 (or proxies) 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 2020, based on the key category analysis, excluding the LULUCF sector and uncertainty, 35 categories
accounted for 95 percent of emissions. However, only four categories account for 54 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 CFU from abandoned oil and gas wells were also identified as a key category. In the trend
analysis, 34 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 2020 emissions and sinks, 43 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,
33 See Chapter 4 Volume 1, "Methodological Choice and Identification of Key Categories" in IPCC (2006). See http://www.ipcc-
nggip. iges.or.jp/public/20Q6gl/index. html.
Introduction 1-17
-------
additional categories such as CO2 emissions from Grasslands Remaining Grasslands were also identified as a key
category. In the trend analysis, 41 categories were identified as key, and when considering uncertainties, 7
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.
Finally, in addition to conducting Approach 1 and 2 level and trend assessments as described above, a qualitative
assessment of the source and sinks categories was conducted to capture any additional key categories that were
not identified using the previously described quantitative approaches. For this inventory, no additional categories
were identified using qualitative criteria recommend by IPCC, but EPA continues to review its qualitative
assessment on an annual basis. Find more information regarding the overall key category analysis in Annex 1 to
this report.
Table 1-4: Key Cat
tegories for the United States (1990 and 2020)
Approach 1
Approach 2 (includes uncertainty)
2020
Level Trend Level Trend
Level Trend Level Trend
Emissions
CRF Source/Sink
Without Without With With
Without Without With With
(MMT
Categories
Gas
LULUCF LULUCF LULUCF LULUCF
LULUCF LULUCF LULUCF LULUCF
C02 Eq.)
Energy
C02 Emissions from
Mobile Combustion:
C02
• • • •
• • • •
1,333.8
Road
C02 Emissions from
Stationary
Combustion - Coal -
C02
• • • •
• • • •
788.2
Electricity
Generation
C02 Emissions from
Stationary
Combustion - Gas -
C02
• • • •
• • • •
634.3
Electricity
Generation
C02 Emissions from
Stationary
C02
485.5
Combustion - Gas -
Industrial
C02 Emissions from
Stationary
C02
256.4
Combustion - Gas -
Residential
C02 Emissions from
Stationary
C02
237.8
Combustion - Oil -
Industrial
C02 Emissions from
Stationary
Combustion - Gas -
C02
• • • •
• • •
173.9
Commercial
C02 Emissions from
Mobile Combustion:
C02
• • • •
• • • •
121.3
Aviation
C02 Emissions from
Non-Energy Use of
C02
• • • •
• • • •
121.0
Fuels
1-18 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
CRF Source/Sink
Categories
Gas
Approach 1
Level Trend Level Trend
Without Without With With
LULUCF LULUCF LULUCF LULUCF
Approach 2 (includes uncertainty)
Level Trend Level Trend
Without Without With With
LULUCF LULUCF LULUCF LULUCF
C02 Emissions from
Stationary
Combustion - Oil -
Residential
C02 Emissions from
Mobile Combustion:
Other3
C02 Emissions from
Stationary
Combustion - Oil -
Commercial
C02 Emissions from
Stationary
Combustion - Coal -
Industrial
C02 Emissions from
Natural Gas Systems
C02 Emissions from
Mobile Combustion:
Railways
C02 Emissions from
Petroleum Systems
C02 Emissions from
Mobile Combustion:
Marine
C02 Emissions from
Stationary
Combustion - Oil -
U.S. Territories
C02 Emissions from
Stationary
Combustion - Oil -
Electricity
Generation
C02 Emissions from
Mobile Combustion:
Military
C02 Emissions from
Coal Mining
C02 Emissions from
Stationary
Combustion - Coal -
Commercial
C02 Emissions from
Stationary
Combustion - Coal -
Residential
CH4 Emissions from
Natural Gas Systems
Fugitive Emissions
from Coal Mining
C02
C02
C02
C02
C02
C02
C02
C02
C02
C02
C02
C02
C02
C02
CH4
ch4
Introduction 1-19
-------
CRF 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
2020
Emissions
(MMT
C02 Eq.)
CH4 Emissions from
Petroleum Systems
CH4 Emissions from
Abandoned Oil and
Gas Wells
CH4 Emissions from
Stationary
Combustion -
Residential
N20 Emissions from
Stationary
Combustion - Coal -
Electricity
Generation
N20 Emissions from
Mobile Combustion:
Road
N20 Emissions from
Stationary
Combustion - Gas -
Electricity
Generation
N20 Emissions from
Stationary
Combustion -
Industrial
CH4
ch4
ch4
n2o
n2o
n2o
n2o
40.2
6.9
4.1
15.2
9.8
4.5
2.3
Industrial Processes and Product Use
C02 Emissions from
Cement Production
C02 Emissions from
Iron and Steel
Production &
Metallurgical Coke
Production
C02 Emissions from
Petrochemical
Production
Emissions from
Substitutes for
Ozone Depleting
Substances:
Refrigeration and Air
conditioning
Emissions from
Substitutes for
Ozone Depleting
Substances: Aerosols
40.7
37.7
30.0
137.7
HFCs,
PFCs
18.1
1-20 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Approach 1
Approach 2 (includes uncertainty)
2020
Level Trend Level
Trend
Level Trend Level Trend
Emissions
CRF Source/Sink
Without Without With
With
Without Without With With
(MMT
Categories
Gas
LULUCF LULUCF LULUCF
LULUCF
LULUCF LULUCF LULUCF LULUCF
C02 Eq.)
Emissions from
Substitutes for
Ozone Depleting
HFCs,
PFCs
•
•
15.
Substances: Foam
Blowing Agents
SF6 and CF4
Emissions from
sf6,
cf4
Electrical
• • •
•
• •
3.8
Transmission and
Distribution
HFC-23 Emissions
from HCFC-22
HFCs
• • •
•
• •
2.1
Production
PFC Emissions from
Aluminum
PFCs
• • •
•
1.7
Production
Agriculture
C02 Emissions from
Liming
C02
• •
2.4
CH4 Emissions from
Enteric
ch4
• • •
•
• • •
168.9
Fermentation: Cattle
CH4 Emissions from
Manure
ch4
• • •
•
• • •
33.5
Management: Cattle
CH4 Emissions from
Manure
Management: Other
ch4
• •
•
26.1
Livestock
CH4 Emissions from
Rice Cultivation
ch4
• •
• •
15.7
Direct N20 Emissions
from Agricultural Soil
n2o
• •
• •
271.7
Management
Indirect N20
Emissions from
n2o
• •
• •
44.6
Applied Nitrogen
Waste
CH4 Emissions from
Commercial Landfills
ch4
• • •
•
• • • •
94.2
CH4 Emissions from
Domestic
Wastewater
ch4
•
11.
Treatment
N20 Emissions from
Domestic
Wastewater
n2o
• •
• • •
23.0
Treatment
Land Use, Land-Use Change, and Forestry
Introduction 1-21
-------
Approach 1
Approach 2 (includes uncertainty)
2020
Level
Trend
Level
Trend
Level Trend Level Trend
Emissions
CRF Source/Sink
Without
Without
With
With
Without Without With With
(MMT
Categories
Gas
LULUCF
LULUCF
LULUCF
LULUCF
LULUCF LULUCF LULUCF LULUCF
C02 Eq.)
Net C02 Emissions
from Land Converted
C02
•
•
• •
77.9
to Settlements
Net C02 Emissions
from Land Converted
co2
•
• •
54.4
to Cropland
Net C02 Emissions
from Grassland
co2
• •
4.5
Remaining Grassland
Net C02 Emissions
from Cropland
co2
•
•
(23.3)
Remaining Cropland
Net C02 Emissions
from Land Converted
co2
•
•
• •
(24.1)
to Grassland
Net C02 Emissions
from Land Converted
co2
•
•
(99.5)
to Forest Land
Net C02 Emissions
from Settlements
Remaining
co2
•
•
• •
(126.1)
Settlements
Net C02 Emissions
from Forest Land
Remaining Forest
co2
•
•
• •
(668.1)
Land
CH4 Emissions from
Flooded Lands
Remaining Flooded
ch4
•
19.9
Lands
CH4 Emissions from
Forest Fires
ch4
•
•
13.6
N20 Emissions from
Forest Fires
n2o
•
•
11.7
Subtotal of Key Categories Without LULUCFb
5,793.6
Total Gross Emissions Without LULUCF
5,981.4
Percent of Total Without LULUCF
97%
Subtotal of Key Categories With LULUCFC
5,013.7
Total Net Emissions With LULUCF
5,222.4
Percent of Total With LULUCF
96%
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).
1-22 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
IPCC Good Practice Guidance and 2006 IPCC 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 1-23
-------
Figure 1-2: U.S. QA/QC Plan Summary
Data
Data
Calculating
Gathering
\Documentation
v Emissions
• Obtain data in
Contact reports
—I/• Clearly label
electronic
for non-electronic
parameters, units,
format (if
communications
and conversion
possible)
• Provide cell
factors
• Review
references for
• Review spreadsheet
spreadsheet
primary data
integrity
construction
elements
o Equations
o Avoid
• Obtain copies of
o Units
hardwiring
ail data sources
o Inputs and
o Use data
• List and location
outputs
validation
of any
• Develop automated
-t—'
(/)
>
o Protect cells
working/external
checkers for:
~ru
• Develop
spreadsheets
o Input ranges
c
<
automatic
• Document
o Calculations
>
checkers for:
assumptions
o Emission
o
+-»
o Outliers,
• Complete QA/QC
aggregation
c
c
values, or
• CRF and summary
checks
missing data
tab links
o Variable
types match
values
o Time series
consistency
• Maintain
tracking tab for
status of
gathering
efforts
• Check input
• Check citations in
• Reproduce
data for
spreadsheet and
calculations
transcription
text for accuracy
• Review time
errors
and style
series consistency
• Inspect
• Check reference
• Review changes
+->
IS>
automatic
docket for new
in
03
checkers
citations
data/consistency
C
<
• Identify
• Review
with IPCC
u
spreadsheet
documentation
methodology
a
modifications
for any data /
§
that could
methodology
provide
changes
additional
• Complete QA/QC
QA/QC checks
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
spreadsheet for
linking to a
master summary
spreadsheet
• Follow strict
version control
procedures
• Document
QA/QC
procedures
1-24 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Box 1-3: Use of IPCC Reference Approach to support Verification of Emissions from Fossil Fuel Combustion
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.
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.
Emissions 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
1.7
Emission
Estimates
Introduction 1-25
-------
relative contribution of individual input parameter uncertainties to the overall Inventory, its trends, and each
source and sink category.
The overall uncertainty estimate for total U.S. greenhouse gas emissions was developed using the IPCC Approach 2
uncertainty estimation methodology, 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 emission estimation
methods, and collect more detailed, measured, and representative data. Individual source 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)
1990 Emission
Uncertainty Range Relative to Emission
Standard
Estimate
Estimate3
Mean'5
Deviationb
Gas
(MMT CO?
Eq.)
(MMT CO
zEq.)
(%)
(MMT C02 Eq.)
Lower
Upper
Lower
Upper
Boundc
Boundc
Bound
Bound
C02
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.
1-26 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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.
Table 1-6: Estimated Overall Inventory Quantitative Uncertainty for 2020 (MMT CO2 Eq. and
Percent)
2020
Emission
Uncertainty Range Relative to Emission
Standard
Estimate
Estimate3
Mean'5
Deviationb
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
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.
Introduction 1-27
-------
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.
In addition to the estimates of uncertainty associated with the current and base year emission 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 emission trend is defined in this Inventory as the
percentage change in the emissions (or removal) estimated for the current year, relative to the emission (or
removal) estimated for the base year. The uncertainty associated with this emission 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 emission 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.)
(%)
(%)
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; 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.
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.
1.8 Completeness
This report, along with its accompanying CRF tables, serves as a thorough assessment of the anthropogenic
sources and sinks of greenhouse gas emissions for the United States for the time series 1990 through 2020. This
report is intended to be comprehensive and includes the vast majority of emissions and removals identified as
anthropogenic, consistent with IPCC and UNFCCC guidelines. In general, sources or sink categories not accounted
1-28 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
for in this Inventory are excluded because they are not occurring in the United States, or because data are
unavailable to develop an estimate and/or the categories were determined to be insignificant34 in terms of overall
national emissions per UNFCCC reporting guidelines.
The United States is continually working to improve upon the understanding of such sources and sinks currently
not included and seeking to find the data required to estimate related emissions and removals, focusing on
categories that are anticipated to be significant. As such improvements are implemented, new emission and
removal estimates are quantified and included in the Inventory, improving completeness of national estimates. For
a list of sources and sink categories not included and more information on significance of these categories, see
Annex 5 and the respective category sections in each sectoral chapter of this report.
In accordance with the revision of the UNFCCC reporting guidelines agreed to at the nineteenth Conference of the
Parties (UNFCCC 2014), this Inventory of U.S. Greenhouse Gas Emissions and Sinks is grouped into five sector-
specific chapters consistent with the UN Common Reporting Framework, listed below in Table 1-8. In addition,
chapters on Trends in Greenhouse Gas Emissions, Other information, and Recalculations and Improvements to be
considered as part of the U.S. Inventory submission are included.
Table 1-8: IPCC Sector Descriptions
Chapter (IPCC Sector) Activities Included
Energy Emissions of all greenhouse gases resulting from stationary and mobile energy
Within each chapter, emissions are identified by the anthropogenic activity that is the source or sink of the
greenhouse gas emissions being estimated (e.g., coal mining). Overall, the following organizational structure is
consistently applied throughout this report:
Chapter/IPCC Sector: Overview of emissions and trends for each IPCC defined sector.
CRF Source or Sink Category: Description of category pathway and emission/removal trends based on IPCC
methodologies, consistent with UNFCCC reporting guidelines.
Methodology: Description of analytical methods (e.g., from 2006 IPCC Guidelines, or country-specific methods)
employed to produce emission estimates and identification of data references, primarily for activity data and
emission factors.
34 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."
1.9 Organization of Report
Land Use, Land-Use
Change, and Forestry
Waste
Industrial Processes and
Product Use
Agriculture
activities including fuel combustion and fugitive fuel emissions, and non-energy
use of fossil fuels.
Emissions resulting from industrial processes and product use of greenhouse
gases.
Emissions from agricultural activities except fuel combustion, which is
addressed under Energy.
Emissions and removals of C02, and emissions of CH4, and N20 from land use,
land-use change and forestry.
Emissions from waste management activities.
Introduction 1-29
-------
Uncertainty and Time-Series Consistency: A discussion and quantification of the uncertainty in emission estimates
and a discussion of time-series consistency.
QA/QC and Verification: A discussion on steps taken to QA/QC and verify the emission estimates, consistent with
the U.S. QA/QC plan, and any key QC findings.
Recalculations Discussion: A discussion of any data or methodological changes that necessitate a recalculation of
previous years' emission estimates, and the impact of the recalculation on the emission estimates, if applicable.
Planned Improvements: A discussion on any category-specific planned improvements, if applicable.
Special attention is given to C02 from fossil fuel combustion relative to other sources because of its share of
emissions and its dominant influence on emission trends. For example, each energy consuming end-use sector
(i.e., residential, commercial, industrial, and transportation), as well as the electricity generation sector, is
described individually. Additional information for certain source categories and other topics is also provided in
several Annexes listed in Table 1-9.
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
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. Complete List of Source Categories
6.4. Constants, Units, and Conversions
6.5. Chemical Formulas
6.6. Greenhouse Gas Precursors: Cross-Walk of NEI categories to the Inventory
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
1-30 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
Introduction 1-31
-------
2. Trends in Greenhouse Gas Emissions
2.1 Trends in U.S. Greenhouse Gas
Emissions and Sinks
In 2020, total gross U.S. greenhouse gas emissions were 5,981.4 million metric tons carbon dioxide equivalent
(MMT CO2 Eq).1 Total U.S. emissions have decreased by 7.3 percent from 1990 to 2020, down from a high of 15.7
percent above 1990 levels in 2007. Emissions decreased from 2019 to 2020 by 9.0 percent (590.4 MMT CO2 Eq.).
Net emissions (i.e., including sinks) were 5,222.4 MMT CO2 Eq. in 2020. Overall, net emissions decreased 10.6
percent from 2019 to 2020 and decreased 21.4 percent from 2005 levels, as shown in Table 2-1. The sharp decline
in emissions from 2019 to 2020 is largely due to the impacts of the coronavirus (COVID-19) pandemic on travel and
economic activity; however, the decline also reflects the combined impacts of long-term trends in many factors,
including population, economic growth, energy markets, technological changes including energy efficiency, and the
carbon intensity of energy fuel choices. Between 2019 and 2020, the decrease in total greenhouse gas emissions
was driven largely by a 10.5 percent decrease in CO2 emissions from fossil fuel combustion, including a 13.3
percent decrease in transportation sector emissions from less travel due to the COVID-19 pandemic and a 10.4
percent decrease in the electric power sector. The decrease in electric power sector emissions was due to a
decrease in electricity demand of 2.5 percent since 2019 and also reflects the continued shift from coal to less
carbon intensive natural gas and renewables.
Figure 2-1 and Figure 2-2 illustrate the overall trend in total U.S. emissions and sinks by gas, annual changes, and
relative changes since 1990.
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
-------
Figure 2-1: U.S. Greenhouse Gas Emissions and Sinks by Gas
¦ HFCs, PFCs, SFe and NFb ¦ Net Emissions (including LULUCF sinks)
9,000 j Nitrous Oxide
¦ Methane
8,000 B Carbon Dioxide
¦ Net CO2 Flux from LULUCF=
7,000
6,000 . ' "
a- 5,000
IU
8 4,000
I-
s
S 3,000
2,000
1,000
-1,000
CTi CTi CTi
CTi Oi Oi
CJi Oi Oi Oi
CT\ CTi Q"! CJ>
fM (M (N fM
LT> VO l*». 00
fMfNirMfMfMfMrvJCM
Ln VD rv 00 CTi
(N fM (\ IN
a 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
Overall from 1990 to 2020, total emissions of CO2decreased by 406.8 MMT CO2 Eq. (7.9 percent), as total
emissions of methane (CH4) decreased by 130.4 MMT CO2 Eq. (16.7 percent), and total emissions of nitrous oxide
(N2O) decreased by 24.4 MMT CO2 Eq. (5.4 percent). During the same period, emissions of fluorinated gases
including hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), sulfur hexafluoride (SFs), and nitrogen trifluoride
(NF3) rose by 89.5 MMT CO2 Eq. (89.8 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 them have
2-2 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
extremely high global warming potentials (GWPs), and, in the cases of PFCs, SF6, and NF3, long atmospheric
lifetimes. Conversely, U.S. greenhouse gas emissions were partly offset by carbon (C) sequestration in managed
forests, trees in urban areas, agricultural soils, landfilled yard trimmings, and coastal wetlands. These were
estimated to offset 13.6 percent (812.2 MMT CO2 Eq.) of total emissions in 2020.
Table 2-1 provides information on trends in emissions and sinks from all U.S. anthropogenic sources in weighted
units of MMT CO2 Eq., while unweighted gas emissions and sinks in kilotons (kt) are provided in Table 2-2.
Table 2-1: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks (MMT CO2 Eq.)
Gas/Source
1990
2005
2016
2017
2018
2019
2020
CO?
5,122.5
6,137.6
5,251.8
5,211.0
5,376.7
5,259.1
4,715.7
Fossil Fuel Combustion
4,731.2
5,752.0
4,909.6
4,853.3
4,989.3
4,852.3
4,342.7
Transportation
1,468.9
1,858.6
1,757.6
1,780.0
1,812.8
1,813.8
1,572.0
Electric Power Sector
1,820.0
2,400.1
1,808.9
1,732.0
1,752.9
1,606.1
1,439.0
Industrial
853.7
851.5
792.7
790.4
814.1
816.1
766.3
Residential
338.6
358.9
292.8
293.4
338.2
341.4
315.8
Commercial
228.3
227.1
231.5
232.0
245.8
250.7
226.8
U.S. Territories
21.7
55.9
26.0
25.5
25.5
24.3
22.7
Non-Energy Use of Fuels
112.2
128.9
99.5
112.6
128.9
126.8
121.0
Cement Production
33.5
46.2
39.4
40.3
39.0
40.9
40.7
Iron and Steel Production &
Metallurgical Coke Production
104.7
70.1
43.6
40.6
42.6
43.1
37.7
Natural Gas Systems
31.9
24.9
29.8
31.1
32.4
38.7
35.4
Petroleum Systems
9.6
12.0
21.9
25.0
37.3
46.7
30.2
Petrochemical Production
21.6
27.4
28.1
28.9
29.3
30.7
30.0
Incineration of Waste
12.9
13.3
14.4
13.2
13.3
12.9
13.1
Ammonia Production
13.0
9.2
10.2
11.1
12.2
12.3
12.7
Lime Production
11.7
14.6
12.6
12.9
13.1
12.1
11.3
Other Process Uses of Carbonates
6.2
7.5
10.8
9.9
7.4
9.8
9.8
Urea Consumption for Non-
Agricultural Purposes
3.8
3.7
5.3
5.2
6.0
6.0
6.0
Urea Fertilization
2.4
3.5
4.7
4.9
5.0
5.1
5.3
Carbon Dioxide Consumption
1.5
1.4
4.6
4.6
4.1
4.9
5.0
Liming
4.7
4.3
3.1
3.1
2.2
2.4
2.4
Coal Mining
4.6
4.2
2.8
3.1
3.1
3.0
2.2
Glass Production
2.3
2.4
2.1
2.0
2.0
1.9
1.9
Aluminum Production
6.8
4.1
1.3
1.2
1.5
1.9
1.7
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
Zinc Production
0.6
1.0
0.8
0.9
1.0
1.0
1.0
Phosphoric Acid Production
1.5
1.3
1.0
1.0
0.9
0.9
0.9
Lead Production
0.5
0.6
0.5
0.5
0.5
0.5
0.5
Carbide Production and
Consumption
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Abandoned Oil and Gas Wells
+
+
+
+
+
+
+
Magnesium Production and
Processing
0.1
+
+
+
+
+
+
Wood Biomass, Ethanol, and
Biodiesel Consumptiona
219.4
230.7
316.9
312.7
319.8
317.2
291.6
International Bunker Fuelsb
103.6
113.3
116.7
120.2
122.2
116.1
69.6
CH4c
780.8
697.5
657.6
663.8
671.1
668.8
650.4
Enteric Fermentation
163.5
168.0
171.3
174.9
175.7
176.1
175.2
Natural Gas Systems
195.5
177.5
165.2
166.6
171.8
172.1
164.9
Landfills
176.6
131.5
107.9
109.2
111.7
113.6
109.3
Trends 2-3
-------
Manure Management
34.8
49.0
57.1
57.5
59.4
58.7
59.6
Coal Mining
96.5
64.1
53.8
54.8
52.7
47.4
41.2
Petroleum Systems
47.8
41.4
40.4
40.5
38.6
40.4
40.2
Wastewater Treatment
20.3
20.1
18.7
18.5
18.3
18.1
18.3
Rice Cultivation
16.0
18.0
15.8
14.9
15.6
15.1
15.7
Stationary Combustion
8.6
7.8
7.9
7.7
8.6
8.8
7.9
Abandoned Oil and Gas Wells
6.5
6.8
6.9
6.9
6.9
7.0
6.9
Abandoned Underground Coal
Mines
7.2
6.6
6.7
6.4
6.2
5.9
5.8
Composting
0.4
1.9
2.3
2.5
2.3
2.3
2.3
Mobile Combustion
6.5
4.0
2.6
2.6
2.5
2.5
2.2
Field Burning of Agricultural
Residues
0.4
0.4
0.4
0.4
0.4
0.4
0.4
Petrochemical Production
0.2
0.1
0.2
0.3
0.3
0.3
0.3
Anaerobic Digestion at Biogas
Facilities
+
+
0.2
0.2
0.2
0.2
0.2
Carbide Production and
Consumption
+
+
+
+
+
+
+
Ferroalloy Production
+
+
+
+
+
+
+
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
450.5
453.3
449.2
444.6
457.7
456.8
426.1
Agricultural Soil Management
316.0
313.8
330.8
328.3
338.9
345.3
316.2
Wastewater Treatment
16.6
20.3
22.8
23.2
23.5
23.4
23.5
Stationary Combustion
25.1
34.4
30.0
28.4
28.2
24.9
23.2
Manure Management
13.9
16.3
18.4
19.0
19.3
19.5
19.7
Mobile Combustion
44.6
41.4
21.1
20.1
19.2
20.0
17.4
Nitric Acid Production
12.1
11.3
10.1
9.3
9.6
10.0
9.3
Adipic Acid Production
15.2
7.1
7.1
7.5
10.5
5.3
8.3
N20 from Product Uses
4.2
4.2
4.2
4.2
4.2
4.2
4.2
Composting
0.3
1.7
2.0
2.2
2.0
2.0
2.0
Caprolactam, Glyoxal, and Glyoxylic
Acid Production
1.7
2.1
1.7
1.5
1.4
1.4
1.2
Incineration of Waste
0.5
0.4
0.4
0.4
0.4
0.4
0.4
Electronics Industry
+
0.1
0.2
0.3
0.3
0.2
0.3
Field Burning of Agricultural
Residues
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Petroleum Systems
+
+
+
+
+
+
+
Natural Gas Systems
+
+
+
+
+
+
+
International Bunker Fuelsb
0.9
1.0
1.0
1.1
1.1
1.0
0.6
HFCs
46.5
127.4
168.3
171.1
171.0
175.9
178.8
Substitution of Ozone Depleting
Substancesd
0.2
107.2
165.1
165.5
167.3
171.8
176.2
HCFC-22 Production
46.1
20.0
2.8
5.2
3.3
3.7
2.1
Electronics Industry
0.2
0.2
0.3
0.4
0.4
0.4
0.4
Magnesium Production and
Processing
NO
NO
0.1
0.1
0.1
0.1
0.1
PFCs
24.3
6.7
4.4
4.2
4.8
4.6
4.4
Electronics Industry
2.8
3.3
3.0
3.0
3.1
2.8
2.7
Aluminum Production
21.5
3.4
1.4
1.1
1.6
1.8
1.7
Substitution of Ozone Depleting
Substancesd
NO
+
+
+
0.1
0.1
0.1
Electrical Transmission and
Distribution
NO
+
+
+
NO
+
+
2-4 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
sf6
28.8
11.8
6.0
5.9
5.7
5.9
5.4
Electrical Transmission and
Distribution
23.2
8.3
4.1
4.2
3.8
4.2
3.8
Magnesium Production and
Processing
5.2
2.7
1.1
1.0
1.0
0.9
0.9
Electronics Industry
0.5
0.7
0.8
0.7
0.8
0.8
0.7
nf3
+
0.5
0.6
0.6
0.6
0.6
0.6
Electronics Industry
+
0.5
0.6
0.6
0.6
0.6
0.6
Total Gross Emissions (Sources)
6,453.5
7,434.8
6,537.9
6,501.0
6,687.5
6,571.7
5,981.4
LULUCF Emissionsc
31.4
41.3
35.4
45.5
39.8
30.3
53.2
ch4
27.2
30.9
28.3
34.0
30.7
25.5
38.1
n2o
4.2
10.5
7.1
11.5
9.1
4.8
15.2
LULUCF Carbon Stock Change8
LULUCF Sector Net Total'
(892.0)
(860.6)
(831.1)
(789.8)
(862.0)
(826.6)
(826.7)
(781.2)
(809.0)
(769.3)
(760.8)
(730.5)
(812.2)
(758.9)
Net Emissions (Sources and Sinks)
5,592.8
6,645.0
5,711.2
5,719.8
5,918.2
5,841.2
5,222.4
+ Does not exceed 0.05 MMT C02 Eq.
NO (Not Occurring)
a Emissions from Wood Biomass, Ethanol, 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.
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 net carbon
stock changes.
Notes: Total(gross) emissions presented without LULUCF. Net emissions presented with LULUCF. Totals may not sum
due to independent rounding. Parentheses indicate negative values or sequestration.
Table 2-2: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks (kt)
Gas/Source
1990
2005
2016
2017
2018
2019
2020
co2
5,122,496
6,137,603
5,251,758
5,210,958
5,376,657
5,259,144
4,715,691
Fossil Fuel Combustion
4,731,178
5,752,043
4,909,609
4,853,299
4,989,308
4,852,330
4,342,659
Transportation
1,468,944
1,858,552
1,757,638
1,779,977
1,812,761
1,813,755
1,572,034
Electric Power Sector
1,819,951
2,400,057
1,808,872
1,732,033
1,752,936
1,606,106
1,438,990
Industrial
853,707
851,522
792,743
790,402
814,096
816,107
766,317
Residential
338,578
358,898
292,773
293,410
338,218
341,400
315,846
Commercial
228,298
227,130
231,547
231,999
245,838
250,703
226,815
U.S. Territories
21,700
55,883
26,036
25,478
25,459
24,259
22,657
Non-Energy Use of Fuels
112,175
128,920
99,505
112,616
128,871
126,716
120,987
Cement Production
33,484
46,194
39,439
40,324
38,971
40,896
40,688
Iron and Steel Production &
Metallurgical Coke
Production
104,737
70,076
43,621
40,566
42,627
43,090
37,731
Natural Gas Systems
31,894
24,945
29,780
31,145
32,407
38,740
35,353
Petroleum Systems
9,600
11,994
21,922
25,027
37,306
46,686
30,156
Petrochemical Production
21,611
27,383
28,110
28,890
29,314
30,702
30,011
Trends 2-5
-------
Incineration of Waste
12,937
13,283
14,356
13,161
13,339
12,948
13,133
Ammonia Production
13,047
9,177
10,245
11,112
12,163
12,272
12,717
Lime Production
11,700
14,552
12,630
12,882
13,106
12,112
11,299
Other Process Uses of
Carbonates
6,233
7,459
10,813
9,869
7,351
9,848
9,794
Urea Consumption for Non-
Agricultural Purposes
3,784
3,653
5,330
5,182
6,030
6,044
5,983
Urea Fertilization
2,417
3,504
4,679
4,897
5,019
5,140
5,275
Carbon Dioxide Consumption
1,472
1,375
4,640
4,580
4,130
4,870
4,970
Liming
4,667
4,349
3,081
3,080
2,248
2,413
2,382
Coal Mining
4,606
4,170
2,848
3,067
3,067
2,951
2,169
Glass Production
2,291
2,432
2,119
2,011
1,989
1,938
1,857
Aluminum Production
6,831
4,142
1,334
1,205
1,451
1,880
1,748
Soda Ash Production
1,431
1,655
1,723
1,753
1,714
1,792
1,461
Ferroalloy Production
2,152
1,392
1,796
1,975
2,063
1,598
1,377
Titanium Dioxide Production
1,195
1,755
1,662
1,688
1,541
1,474
1,340
Zinc Production
632
1,030
838
900
999
1,026
1,008
Phosphoric Acid Production
1,529
1,342
998
1,025
937
909
938
Lead Production
516
553
500
513
513
527
495
Carbide Production and
Consumption
243
213
170
181
184
175
154
Abandoned Oil and Gas Wells
6
7
7
7
7
7
7
Magnesium Production and
Processing
129
3
3
3
2
1
1
Wood Biomass, Ethanol, and
Biodiesel Consumptiona
219,413
230,700
316,853
312,717
319,805
317,231
291,613
International Bunker Fuelsb
103,634
113,328
116,682
120,192
122,179
116,132
69,638
CH4c
31,233
27,898
26,304
26,550
26,844
26,753
26,017
Enteric Fermentation
6,539
6,722
6,853
6,998
7,028
7,046
7,007
Natural Gas Systems
7,821
7,100
6,609
6,662
6,871
6,885
6,596
Landfills
7,063
5,262
4,318
4,368
4,467
4,545
4,373
Manure Management
1,394
1,960
2,285
2,300
2,375
2,348
2,383
Coal Mining
3,860
2,565
2,154
2,191
2,109
1,895
1,648
Petroleum Systems
1,912
1,655
1,616
1,621
1,544
1,615
1,609
Wastewater T reatment
812
806
748
740
732
723
730
Rice Cultivation
640
720
631
596
623
602
630
Stationary Combustion
344
313
315
307
344
351
317
Abandoned Oil and Gas Wells
261
273
275
276
277
279
276
Abandoned Underground
Coal Mines
288
264
268
257
247
237
231
Composting
15
75
91
98
90
91
91
Mobile Combustion
259
161
105
102
99
99
88
Field Burning of Agricultural
Residues
15
17
17
17
17
17
17
Petrochemical Production
9
3
10
10
12
13
13
Anaerobic Digestion at Biogas
Facilities
1
2
7
6
6
6
6
Carbide Production and
Consumption
1
+
+
+
+
+
+
Ferroalloy Production
1
+
1
1
1
+
+
Iron and Steel Production &
Metallurgical Coke
Production
1
1
+
+
+
+
+
Incineration of Waste
+
+
+
+
+
+
+
International Bunker Fuelsb
7
5
4
4
4
4
3
2-6 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
N2Oc
1,512
1,521
1,507
1,492
1,536
1,533
1,430
Agricultural Soil Management
1,060
1,053
1,110
1,102
1,137
1,159
1,061
Wastewater T reatment
56
68
76
78
79
79
79
Stationary Combustion
84
115
101
95
95
84
78
Manure Management
47
55
62
64
65
65
66
Mobile Combustion
150
139
71
68
64
67
58
Nitric Acid Production
41
38
34
31
32
34
31
AdipicAcid Production
51
24
24
25
35
18
28
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
Incineration of Waste 2
Electronics Industry +
Field Burning of Agricultural
Residues 1
Petroleum Systems +
Natural Gas Systems +
International Bunker Fuelsb 3
HFCs M
Substitution of Ozone
Depleting Substancesd M
HCFC-22 Production 3
Electronics Industry M
Magnesium Production and
Processing NO
PFCs M
Electronics Industry M
Aluminum Production M
Substitution of Ozone
Depleting Substancesd NO
Electrical Transmission and
Distribution NO
SF6 1
Electrical Transmission and
Distribution 1
Magnesium Production and
Processing +
Electronics Industry +
NF3 +
Electronics Industry +
CO 130,085
NOx 21,712
SO, 20,935
NMVOCs 20,923
7
1
+
1
+
+
3
M
M
1
M
NO
M
M
M
6
1
1
1
+
+
3
M
M
+
M
+
M
M
M
5
1
1
1
+
+
4
M
M
+
M
+
M
M
M
+
+
+
+
66,912
17,191
13,196
13,309
+
+
+
+
35,882
8,686
2,906
9,855
+
+
+
+
34,752
8,296
2,303
9,483
5
1
1
1
+
+
4
M
M
+
M
+
M
M
M
NO
+
+
+
+
+
33,743
7,869
2,211
9,310
5
1
1
1
+
+
3
M
M
+
M
+
M
M
M
+
+
+
+
32,734
7,374
1,943
9,136
4
1
1
1
+
+
2
M
M
+
M
+
M
M
M
+
+
+
+
31,725
6,883
1,780
8,963
+ Does not exceed 0.5 kt.
M (Mixture of multiple gases)
NO (Not Occurring)
a Emissions from Wood Biomass, Ethanol, 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.
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 may not sum due to independent rounding. Parentheses indicate negative values or sequestration.
Trends 2-7
-------
Emissions of all gases can be summed from each source category into a set of five sectors defined by the
Intergovernmental Panel on Climate Change (IPCC). Figure 2-3 and Table 2-3 illustrate that over the thirty-one-year
period of 1990 to 2020, total emissions from the Energy and Waste sectors decreased by 486.5 MMT CO2 Eq. (9.1
percent) and 58.6 MMT CO2 Eq. (27.4 percent), respectively. Emissions from Industrial Processes and Product Use
and Agriculture grew by 30.2 MMT CO2 Eq. (8.7 percent) and 42.8 MMT CO2 Eq. (7.8 percent), respectively. Over
the same period, total C sequestration in the Land Use, Land-Use Change, and Forestry (LULUCF) sector decreased
by 79.8 MMT CO2 (9.0 percent decrease in total C sequestration), and emissions from the LULUCF sector increased
by 21.8 MMT C02 Eq. (69.6 percent).
Figure 2-3: U.S. Greenhouse Gas Emissions and Sinks by IPCC Sector
9 000 B LULUCF (emissions) ¦ Agriculture
¦ Waste ¦ Energy
8 000 B Industrial Processes and Product Use ¦ LULUCF (removals)
¦ Net Emissions (including LULUCF sinks)
7,000
6,000
ri- 5,000
LU
S 4,000
I—
E 3,000
2,000
1,000
0
-1,000
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
2016
2017
2018
2019
2020
Energy
5,341.1
6,319.8
5,413.1
5,372.7
5,539.4
5,409.8
4,854.7
Fossil Fuel Combustion
4,731.2
5,752.0
4,909.6
4,853.3
4,989.3
4,852.3
4,342.7
Natural Gas Systems
227.4
202.5
195.0
197.7
204.2
210.9
200.3
Non-Energy Use of Fuels
112.2
128.9
99.5
112.6
128.9
126.8
121.0
Petroleum Systems
57.4
53.4
62.3
65.6
75.9
87.1
70.4
Coal Mining
101.1
68.3
56.7
57.9
55.8
50.3
43.4
Stationary Combustion3
33.7
42.2
37.9
36.1
36.8
33.7
31.2
Mobile Combustion3
51.1
45.4
23.7
22.7
21.6
22.4
19.6
Incineration of Waste
13.4
13.7
14.8
13.6
13.8
13.4
13.5
Abandoned Oil and Gas Wells
6.5
6.8
6.9
6.9
6.9
7.0
6.9
Abandoned Underground Coal Mines
7.2
6.6
6.7
6.4
6.2
5.9
5.8
Industrial Processes and Product Use
346.2
365.9
369.0
369.4
373.4
379.5
376.4
Substitution of Ozone Depleting
Substances
0.2
107.2
165.1
165.5
167.3
171.8
176.3
Cement Production
33.5
46.2
39.4
40.3
39.0
40.9
40.7
2-8 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Iron and Steel Production &
Metallurgical Coke Production
104.8
70.1
43.6
40.6
42.6
43.1
37.7
Petrochemical Production
21.8
27.5
28.4
29.1
29.6
31.0
30.3
Ammonia Production
13.0
9.2
10.2
11.1
12.2
12.3
12.7
Lime Production
11.7
14.6
12.6
12.9
13.1
12.1
11.3
Other Process Uses of Carbonates
6.2
7.5
10.8
9.9
7.4
9.8
9.8
Nitric Acid Production
12.1
11.3
10.1
9.3
9.6
10.0
9.3
Adipic Acid Production
15.2
7.1
7.1
7.5
10.5
5.3
8.3
Urea Consumption for Non-
Agricultural Purposes
3.8
3.7
5.3
5.2
6.0
6.0
6.0
Carbon Dioxide Consumption
1.5
1.4
4.6
4.6
4.1
4.9
5.0
Electronics Industry
3.6
4.8
5.0
4.9
5.1
4.7
4.7
N20 from Product Uses
4.2
4.2
4.2
4.2
4.2
4.2
4.2
Electrical Transmission and
Distribution
23.2
8.4
4.1
4.2
3.8
4.2
3.8
Aluminum Production
28.3
7.6
2.7
2.3
3.1
3.6
3.4
HCFC-22 Production
46.1
20.0
2.8
5.2
3.3
3.7
2.1
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
Caprolactam, Glyoxal, and Glyoxylic
Acid Production
1.7
2.1
1.7
1.5
1.4
1.4
1.2
Zinc Production
0.6
1.0
0.8
0.9
1.0
1.0
1.0
Phosphoric Acid Production
1.5
1.3
1.0
1.0
0.9
0.9
0.9
Magnesium Production and
Processing
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
551.9
573.6
601.9
603.2
616.7
622.9
594.7
Agricultural Soil Management
316.0
313.8
330.8
328.3
338.9
345.3
316.2
Enteric Fermentation
163.5
168.0
171.3
174.9
175.7
176.1
175.2
Manure Management
48.8
65.3
75.5
76.5
78.7
78.2
79.2
Rice Cultivation
16.0
18.0
15.8
14.9
15.6
15.1
15.7
Urea Fertilization
2.4
3.5
4.7
4.9
5.0
5.1
5.3
Liming
4.7
4.3
3.1
3.1
2.2
2.4
2.4
Field Burning of Agricultural Residues
0.5
0.6
0.6
0.6
0.6
0.6
0.6
Waste
214.2
175.6
153.9
155.7
157.9
159.6
155.6
Landfills
176.6
131.5
107.9
109.2
111.7
113.6
109.3
Wastewater Treatment
36.9
40.5
41.5
41.7
41.8
41.5
41.8
Composting
0.7
3.5
4.3
4.6
4.3
4.3
4.3
Anaerobic Digestion at Biogas
Facilities
+
+
0.2
0.2
0.2
0.2
0.2
Total Gross Emissions'1 (Sources)
6,453.5
7,434.8
6,537.9
6,501.0
6,687.5
6,571.7
5,981.4
LULUCF Sector NetTotalc
(860.6)
(789.8)
(826.6)
(781.2)
(769.3)
(730.5)
(758.9)
Forest land
(868.3)
(773.0)
(816.7)
(769.6)
(764.2)
(731.4)
(741.7)
Cropland
28.6
23.0
31.4
32.0
37.4
39.4
31.0
Grassland
4.0
(27.6)
(14.0)
(12.7)
(12.1)
(8.4)
(19.0)
Wetlands
21.9
18.5
16.5
16.6
16.5
16.5
16.5
Settlements
(46.8)
(30.7)
(43.8)
(47.4)
(46.9)
(46.6)
(45.8)
Net Emission (Sources and Sinks)d
5,592.8
6,645.0
5,711.2
5,719.8
5,918.2
5,841.2
5,222.4
+ Does not exceed 0.05 MMT C02 Eq.
a Includes CH4 and N20 emissions from fuel combustion.
b Total emissions without LULUCF.
Trends 2-9
-------
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 Net emissions with LULUCF.
Notes: Total (gross) emissions presented without LULUCF. Net emissions presented with LULUCF. Totals may not sum due
to independent rounding. Parentheses indicate negative values or sequestration.
Energy
Energy-related activities, primarily fossil fuel combustion, accounted for the vast majority of U.S. CO2 emissions for
the period of 1990 through 2020. Fossil fuel combustion is the largest source of energy-related emissions, with CO2
being the primary gas emitted (see Figure 2-4). Due to their relative importance, fossil fuel combustion-related CO2
emissions are considered in detail in the Energy chapter (see Energy chapter).
In 2020, 78.8 percent of the energy used in the United States (on a Btu basis) was produced through the
combustion of fossil fuels. The remaining 21.2 percent came from other energy sources such as hydropower,
biomass, nuclear, wind, and solar energy. A discussion of specific trends related to CO2 as well as other greenhouse
gas emissions from energy use is presented here with more detail in the Energy chapter. Energy-related activities
are also responsible for CH4 and N2O emissions (41.4 percent and 9.6 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 2-4: Trends in Energy Sector Greenhouse Gas Sources
8,000
7,000
6,000
iff 5'000
o
u
4,000
3,000
2,000
1,000
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
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
Table 2-4: Emissions from Energy (MMT CO2 Eq.)
2
Gas/Source 1990
2005
2016 2017 2018 2019 2020
C02 4,902.4
5,935.4
5,078.0 5,038.3 5,204.3 5,080.4 4,544.5
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.
2-10 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Transportation
1,468.9
1,858.6
1,757.6
1,780.0
1,812.8
1,813.8
1,572.0
Electricity Generation
1,820.0
2,400.1
1,808.9
1,732.0
1,752.9
1,606.1
1,439.0
Industrial
853.7
851.5
792.7
790.4
814.1
816.1
766.3
Residential
338.6
358.9
292.8
293.4
338.2
341.4
315.8
Commercial
228.3
227.1
231.5
232.0
245.8
250.7
226.8
U.S. Territories
21.7
55.9
26.0
25.5
25.5
24.3
22.7
Non-Energy Use of Fuels
112.2
128.9
99.5
112.6
128.9
126.8
121.0
Natural Gas Systems
31.9
24.9
29.8
31.1
32.4
38.7
35.4
Petroleum Systems
9.6
12.0
21.9
25.0
37.3
46.7
30.2
Incineration of Waste
12.9
13.3
14.4
13.2
13.3
12.9
13.1
Coal Mining
4.6
4.2
2.8
3.1
3.1
3.0
2.2
Abandoned Oil and Gas Wells
+
+
+
+
+
+
+
Biomass-Wooda
215.2
206.9
216.0
211.9
220.0
217.6
202.1
Biofuels-Ethanola
4.2
22.9
81.2
82.1
81.9
82.6
71.8
International Bunker Fuelsb
103.6
113.3
116.7
120.2
122.2
116.1
69.6
Biofuels-Biodiesela
0.0
0.9
19.6
18.7
17.9
17.1
17.7
ch4
368.6
308.3
283.5
285.4
287.3
284.0
269.1
Natural Gas Systems
195.5
177.5
165.2
166.6
171.8
172.1
164.9
Coal Mining
96.5
64.1
53.8
54.8
52.7
47.4
41.2
Petroleum Systems
47.8
41.4
40.4
40.5
38.6
40.4
40.2
Stationary Combustion
8.6
7.8
7.9
1.7
8.6
00
00
7.9
Abandoned Oil and Gas Wells
6.5
6.8
6.9
6.9
6.9
7.0
6.9
Abandoned Underground Coal
7.2
6.6
6.7
6.4
6.2
5.9
5.8
Mines
Mobile Combustion
6.5
4.0
2.6
2.6
2.5
2.5
2.2
Incineration of Waste
+
+
+
+
+
+
+
International Bunker Fuelsb
0.2
0.1
0.1
0.1
0.1
0.1
0.1
n2o
70.1
76.1
51.5
49.0
47.9
45.3
41.1
Stationary Combustion
25.1
34.4
30.0
28.4
28.2
24.9
23.2
Mobile Combustion
44.6
41.4
21.1
20.1
19.2
20.0
17.4
Incineration of Waste
0.5
0.4
0.4
0.4
0.4
0.4
0.4
Petroleum Systems
+
+
+
+
+
+
+
Natural Gas Systems
+
+
+
+
+
+
+
International Bunker Fuelsb
0.9
1.0
1.0
1.1
1.1
1.0
0.6
Total
5,341.1
6,319.8
5,413.1
5,372.7
5,539.4
5,409.8
4,854.7
+ Does not exceed 0.05 MMT C02 Eq.
a Emissions from Wood 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.
Note: Totals may not sum due to independent rounding.
CO2 Emissions from Fossil Fuel Combustion
As the largest contributor to U.S. greenhouse gas emissions, CO2 from fossil fuel combustion has accounted for
approximately 75.3 percent of gross emissions across the time series. Within the United States, fossil fuel
combustion accounted for 92.1 percent of CO2 emissions in 2020. Emissions from this source category decreased
by 8.2 percent (388.5 MMT CO2 Eq.) from 1990 to 2020 and were responsible for most of the decrease in national
emissions during this period. Similarly, CO2 emissions from fossil fuel combustion decreased by 1,409.4 MMT CO2
Eq. from 2005 and by 1,003.0 MMT CO2 Eq. from 2010, representing decreases of 24.5 percent between 2005 and
2020 and 18.8 percent between 2010 and 2020. From 2019 to 2020, these emissions decreased by 10.5 percent
(509.7 MMT CO2 Eq.). Historically, changes in emissions from fossil fuel combustion have been the main factor
influencing U.S. emission trends.
Trends 2-11
-------
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. CO2 emissions from coal combustion
gradually increased between 1990 and 2007, then began to decrease at a faster rate from 2008 to 2020. CO2
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. Likewise,
warmer winters lead to a decrease in heating degree days and result in a decreased demand for heating fuel and
electricity for heat in the residential and commercial sectors, which leads to a decrease in emissions from reduced
fuel consumption. The decrease in 2020 emissions was due primarily to the COVID-19 pandemic reducing overall
demand for fossil fuels across all sectors, but it also reflects a continued shift from coal to natural gas and
renewables in the electric power sector.
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 A-22 in Annex 2.1 for more detail on the C
Content Coefficient of different fossil fuels).
Recent trends in CO2 emissions from fossil fuel combustion have been strongly influenced by trends in the electric
power sector, which from 1990 to 2017 accounted for the largest share of emissions from this source (see Figure
2-12). Electric power sector emissions are driven by the total amount of electricity generated to meet electricity
demand and the carbon intensity of the energy mix used to produce the electricity. From 1990 to 2005, power
sector CO2 emissions increased 31.9 percent with a 34.3 percent increase in generation (see Figure 2-7). From 2005
to 2020, power sector CO2 emissions dropped 40.0 percent while generation remained relatively flat (a 1.4 percent
decrease). The types of fuel consumed to produce electricity have shifted over time, impacting emission trends.
Electricity generation from lower carbon intensity renewable energy sources increased by 132.3 percent from 2005
to 2020 and natural gas generation increased by 122.2 percent while coal generation decreased by 61.5 percent
over the same time period (see Table 3-12 for more detail on electricity generation by source). The decrease in
coal-powered electricity generation and increase in natural gas and renewable energy electricity generation have
contributed to the 40.0 percent decrease in overall CO2 emissions from electric power generation from 2005 to
2020 (see Figure 2-7). Between 2019 and 2020, emissions from the electric power sector decreased 10.4 percent
due to a decrease in electric power generation of 2.9 percent and a decrease in the carbon intensity of the electric
power energy mix reflecting the continued shift in the share of electric power generation from coal to natural gas
and renewable energy.
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
2018. Emissions from petroleum consumption for transportation (including bunker fuels) have decreased by 11.9
percent since 2016; this trend can be primarily attributed to a 11.0 percent decrease in vehicle miles traveled
(VMT) from 2019 to 2020, due largely to the impacts of the coronavirus pandemic which limited travel in 2020.
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 10.5 percent decrease in total CO2 emissions from fossil fuel combustion
from 2019 to 2020 and a 11.5 percent reduction since 2016. Trends in carbon dioxide emissions from fossil fuel
combustion, separated by end-use sector, are presented in Table 2-5 and Figure 2-5 based on the underlying U.S.
energy consumer data collected by the U.S. Energy Information Administration (EIA). Figure 2-6 further describes
2-12 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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.
• 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
generators that produce electricity and/or useful thermal output primarily to support on-site industrial
activities in this sector.)
• 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.)
• Residential. ElA's fuel consumption data for the residential sector consist of living quarters for private
households.
• Commercial. ElA's fuel consumption data for the commercial sector consist of service-providing facilities
and equipment from private and public organizations and businesses. (EIA includes generators that
produce electricity and/or useful thermal output primarily to support the activities at commercial
establishments in this sector.)
Table 2-5: CO2 Emissions from Fossil Fuel Combustion by End-Use Sector (MMT CO2 Eq.)
End-Use Sector
1990
2005
2016
2017
2018
2019
2020
Transportation
1,472.0
1,863.3
1,761.8
1,784.3
1,817.4
1,818.5
1,576.7
Combustion
1,468.9
1,858.6
1,757.6
1,780.0
1,812.8
1,813.8
1,572.0
Electricity
3.0
4.7
4.2
4.3
4.7
4.7
4.7
Industrial
1,540.1
1,587.8
1,310.3
1,294.8
1,315.3
1,281.4
1,175.8
Combustion
853.7
851.5
792.7
790.4
814.1
816.1
766.3
Electricity
686.4
736.3
517.6
504.4
501.2
465.3
409.5
Residential
931.3
1,214.9
946.2
910.5
980.4
925.0
860.6
Combustion
338.6
358.9
292.8
293.4
338.2
341.4
315.8
Electricity
592.7
856.0
653.5
617.1
642.2
583.6
544.8
Commercial
766.0
1,030.1
865.2
838.2
850.7
803.2
706.8
Combustion
228.3
227.1
231.5
232.0
245.8
250.7
226.8
Electricity
537.7
803.0
633.6
606.2
604.9
552.5
480.0
U.S. Territories3
21.7
55.9
26.0
25.5
25.5
24.3
22.7
Total
4,731.2
5,752.0
4,909.6
4,853.3
4,989.3
4,852.3
4,342.7
Electric Power
1,820.0
2,400.1
1,808.9
1,732.0
1,752.9
1,606.1
1,439.0
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.
Trends 2-13
-------
Figure 2-5: Trends in CO2 Emissions from Fossil Fuel Combustion by End-Use Sector and Fuel
Type
Coal ¦ Geothermal ¦ Natural Gas ¦ Petroleum
U.S. Territories
Commercial
Residential
2,000
o
u
1,000
2,000
o
u
1,000
2,000
o
u
1,000
1990 1995 2000 2005 2010 2015 2020
Industrial
Electric Power
Transportation
1990 1995 2000 2005 2010 2015 2020
Note on Figure 2-5: Fossil Fuel Combustion for electric power also includes emissions of less than 0.5 MMT C02 Eq. from
geothermal-based generation.
2-14 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure 2-6: Trends in End-Use Sector Emissions of CO2 from Fossil Fuel Combustion
¦ Direct Fossil Fuel Combustion ¦ Indirect Fossil Fuel Combustion
2,000
cr 1,500
8 1,000
500
0
2,000
1,500
S 1,000
500
0
2,000
1,500
8 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
0
Industrial
Transportation
1990 1995 2000 2005 2010 2015 2020
1990 1995 2000 2005 2010 2015 2020
Electric power was the second largest emitter of CO2 in 2020 (surpassed by transportation); electric power
generators used 31.2 percent of U.S. energy from fossil fuels and emitted 33.1 percent of the CO2 from fossil fuel
combustion in 2020. 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 20.9 percent since 1990, and the carbon intensity of the electric
power sector, in terms of CO2 Eq. per QBtu input, has significantly decreased by 19.2 percent during that same
timeframe. This decoupling of electric power generation and the resulting CO2 emissions is shown below in Figure
2-7.
Trends 2-15
-------
Figure 2-7: 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)
I Total Emissions (MMT CO2 Eq.) [Right Axis]
3,500
3,000
2,500 _
2,000 H
z
in
c
z
1,500 1
E
UJ
"tD
"8
1,000
500
0
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 1,576.7 MMT CO2 Eq. in 2020 or 36.3 percent of
total CO2 emissions from fossil fuel combustion. The industrial end-use sector accounted for 27.1 percent of CO2
emissions from fossil fuel combustion when including allocated electricity emissions. The residential and
commercial end-use sectors accounted for 19.8 and 16.3 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 63.3 and 67.9 percent of emissions from the residential and commercial end-use sectors,
respectively.
Other Significant Trends in Energy
Other significant trends in emissions from energy source categories (Figure 2-5 and Figure 2-6) over the thirty-one-
year period from 1990 through 2020 included the following:
• Methane emissions from natural gas systems and petroleum systems (combined here) decreased 38.2
MMT CO2 Eq. (15.7 percent decrease from 1990 to 2020) or from 243.3 MMT CO2 Eq. in 1990 to 205.1
MMT CO2 Eq. in 2020. Natural gas systems CFU emissions decreased by 30.6 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. Petroleum systems CH4 emissions
decreased by 7.6 MMT CO2 Eq. (or 15.8 percent) since 1990. 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 24.0 MMT CO2 Eq. (57.9 percent) from 1990 to 2020. 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.
2-16 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
• Methane emissions from coal mining decreased by 55.3 MMT CO2 Eq. (57.3 percent) from 1990 through
2020, primarily due to a decrease in the number of active mines and annual coal production over the time
period.
• Nitrous oxide emissions from mobile combustion decreased by 27.2 MMT CO2 Eq. (61.0 percent) from
1990 through 2020, primarily as a result of national vehicle criteria pollutant emissions standards and
emission control technologies for on-road vehicles.
• Carbon dioxide emissions from non-energy uses of fossil fuels increased by 8.8 MMT CO2 Eq. (7.9 percent)
from 1990 through 2020. Emissions from non-energy uses of fossil fuels were 121.0 MMT CO2 Eq. in 2020,
which constituted 2.6 percent of total national CO2 emissions, approximately the same proportion as in
1990.
• Carbon dioxide emissions from incineration of waste (13.1 MMT CO2 Eq. in 2020) increased slightly by 0.2
MMT CO2 Eq. (1.5 percent) from 1990 through 2020, as the volume of scrap tires and other fossil C-
containing materials in waste increased.
Industrial Processes and Product Use
Greenhouse gases can be generated and emitted by industry in two different ways. First, they are generated and
emitted as the byproducts of many non-energy-related industrial activities. For example, industrial processes can
chemically or physically transform raw materials, which often release waste gases such as CO2, CFU, N2O, and
fluorinated gases (e.g., HFC-23). 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.
Second, industrial manufacturing processes and use by end-consumers also release HFCs, PFCs, SF6, and NF3 and
other fluorinated compounds. In addition to the use of HFCs and some PFCs as substitutes for ozone depleting
substances (ODS), fluorinated compounds such as HFCs, PFCs, SF6, NF3, and others are also emitted through use by
a number of other industrial sources in the United States. These industries include the electronics industry, electric
power transmission and distribution, and magnesium metal production and processing. In addition, N2O is used in
and emitted by the electronics industry and anesthetic and aerosol applications, and CO2 is consumed and emitted
through various end-use applications.
Emission sources in the Industrial Processes and Product Use (IPPU) chapter accounted for 6.3 percent of U.S.
greenhouse gas emissions in 2020. Emissions from the IPPU sector increased by 8.7 percent from 1990 to 2020.
Total emissions from IPPU remained relatively constant between 2019 and 2020, decreasing 0.8 percent due to
offsetting trends within the sector. Some industrial processes and product use categories experienced decreases
due to impacts from the COVID-19 pandemic (e.g., Iron and Steel Production and Lime Production), while other
categories experienced increases in emissions from 2019 to 2020 (e.g., Ammonia Production and the Substitution
of Ozone Depleting Substances). Figure 2-8 presents greenhouse gas emissions from IPPU by source category.
Trends 2-17
-------
Figure 2-8: 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
Oi-ifNro^-mvDrvoocn
o"« cr> cr> o"» o"» o"i
cno^cr.cr>a^cr«cr>aiaia^
(NtN(NOJ(N(N(N(N(N(NNlN(N(N(N(N(N(N(N(N(N
Table 2-6: Emissions from Industrial Processes and Product Use (MMT CO2 Eq.)
Gas/Source
1990
2005
2016
2017
2018
2019
2020
C02
213.0
194.4
166.0
164.7
165.1
171.2
163.6
Iron and Steel Production & Metallurgical Coke
Production
104.7
70.1
43.6
40.6
42.6
43.1
37.7
Iron and Steel Production
99.1
66.2
41.0
38.6
41.3
40.1
35.4
Metallurgical Coke Production
5.6
3.9
2.6
2.0
1.3
3.0
2.3
Cement Production
33.5
46.2
39.4
40.3
39.0
40.9
40.7
Petrochemical Production
21.6
27.4
28.1
28.9
29.3
30.7
30.0
Ammonia Production
13.0
9.2
10.2
11.1
12.2
12.3
12.7
Lime Production
11.7
14.6
12.6
12.9
13.1
12.1
11.3
Other Process Uses of Carbonates
6.2
7.5
10.8
9.9
7.4
9.8
9.8
Urea Consumption for Non-Agricultural
Purposes
3.8
3.7
5.3
5.2
6.0
6.0
6.0
Carbon Dioxide Consumption
1.5
1.4
4.6
4.6
4.1
4.9
5.0
Glass Production
2.3
2.4
2.1
2.0
2.0
1.9
1.9
Aluminum Production
6.8
4.1
1.3
1.2
1.5
1.9
1.7
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
Zinc Production
0.6
1.0
0.8
0.9
1.0
1.0
1.0
Phosphoric Acid Production
1.5
1.3
1.0
1.0
0.9
0.9
0.9
Lead Production
0.5
0.6
0.5
0.5
0.5
0.5
0.5
Carbide Production and Consumption
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Magnesium Production and Processing
0.1
+
+
+
+
+
+
ch4
0.3
0.1
0.3
0.3
0.3
0.4
0.3
Petrochemical Production
0.2
0.1
0.2
0.3
0.3
0.3
0.3
Carbide Production and Consumption
+
+
+
+
+
+
+
Ferroalloy Production
+
+
+
+
+
+
+
Iron and Steel Production & Metallurgical Coke
Production
+
+
+
+
+
+
+
Iron and Steel Production
+
+
+
+
+
+
+
2-18 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Metallurgical Coke Production
NO
NO
NO
NO
NO
NO
NO
n2o
33.3
24.9
23.4
22.7
26.0
21.1
23.3
Nitric Acid Production
12.1
11.3
10.1
9.3
9.6
10.0
9.3
AdipicAcid Production
15.2
7.1
7.1
7.5
10.5
5.3
8.3
N20 from Product Uses
4.2
4.2
4.2
4.2
4.2
4.2
4.2
Caprolactam, Glyoxal, and Glyoxylic Acid
Production
1.7
2.1
1.7
1.5
1.4
1.4
1.2
Electronics Industry
+
0.1
0.2
0.3
0.3
0.2
0.3
HFCs
46.5
127.4
168.3
171.1
171.0
175.9
178.8
Substitution of Ozone Depleting Substances3
0.2
107.2
165.1
165.5
167.3
171.8
176.2
HCFC-22 Production
46.1
20.0
2.8
5.2
3.3
3.7
2.1
Electronics Industry
0.2
0.2
0.3
0.4
0.4
0.4
0.4
Magnesium Production and Processing
NO
NO
0.1
0.1
0.1
0.1
0.1
PFCs
24.3
6.7
4.4
4.2
4.8
4.6
4.4
Electronics Industry
2.8
3.3
3.0
3.0
3.1
2.8
2.7
Aluminum Production
21.5
3.4
1.4
1.1
1.6
1.8
1.7
Substitution of Ozone Depleting Substances
NO
+
+
+
0.1
0.1
0.1
Electrical Transmission and Distribution
NO
+
+
+
NO
+
+
sf6
28.8
11.8
6.0
5.9
5.7
5.9
5.4
Electrical Transmission and Distribution
23.2
8.3
4.1
4.2
3.8
4.2
3.8
Magnesium Production and Processing
5.2
2.7
1.1
1.0
1.0
0.9
0.9
Electronics Industry
0.5
0.7
0.8
0.7
0.8
0.8
0.7
nf3
+
0.5
0.6
0.6
0.6
0.6
0.6
Electronics Industry
+
0.5
0.6
0.6
0.6
0.6
0.6
Total
346.2
365.9
369.0
369.4
373.4
379.5
376.4
+ Does not exceed 0.05 MMT C02 Eq.
a Small amounts of PFC emissions also result from this source.
NO (Not Occurring)
Note: Totals may not sum due to independent rounding.
Some significant trends in U.S. emissions from IPPU source categories over the thirty-one-year period from 1990
through 2020 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 176.3 MMT CO2 Eq. in 2020 and accounted for 46.8
percent of total IPPU emissions.
• Combined CO2 and CFU emissions from iron and steel production and metallurgical coke production
decreased by 12.4 percent from 2019 to 2020 to 37.7 MMT CO2 Eq. and have declined overall by 67.0
MMT CO2 Eq. (64.0 percent) from 1990 through 2020, 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 38.9 percent between 1990 and 2020 from
21.6 MMT CO2 Eq. to 30.0 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 (12.7 MMT CO2 Eq. in 2020) decreased by 2.5
percent (0.3 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 steadily 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 21.5 percent (7.2 MMT CO2 Eq.) from
1990 through 2020. They rose from 1990 through 2006 and then fell until 2009, due to a decrease in
Trends 2-19
-------
demand for construction materials during the economic recession. Since 2010, CO2 emissions from
cement production have risen 29.4 percent (9.2 MMT CO2 Eq.).
• Carbon dioxide emissions from lime production decreased by 6.7 percent (0.8 MMT CO2) from 2019 to
2020. Compared to 1990, CO2 emissions have decreased by about 3.4 percent. The trends in CO2
emissions from lime production are directly proportional to trends in lime production and since 2015,
fluctuation in lime production has been driven by demand from the steel making industry.
• PFC emissions from aluminum production decreased by 92.2 percent (19.8 MMT CO2 Eq.) from 1990 to
2020, due to both industry emission reduction efforts and lower domestic aluminum production.
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 were the primary greenhouse gases emitted by agricultural activities.
Carbon stock changes from agricultural soils are included in the LULUCF sector.
In 2020, agricultural activities were responsible for emissions of 594.7 MMT CO2 Eq., or 9.9 percent of total U.S.
greenhouse gas emissions. Methane emissions from enteric fermentation and manure management represented
26.9 percent and 9.2 percent of total CFU emissions from anthropogenic activities, respectively, in 2020.
Agricultural soil management activities, such as application of synthetic and organic fertilizers, deposition of
livestock manure, and growing N-fixing plants, were the largest contributors to U.S. N2O emissions in 2020,
accounting for 74.2 percent. 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 2-9 and Table 2-7 illustrate agricultural greenhouse gas emissions by source.
Figure 2-9: Trends in Agriculture Sector Greenhouse Gas Sources
2-20 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 2-7: Emissions from Agriculture (MMT CO2 Eq.)
Gas/Source
1990
2005
2016
2017
2018
2019
2020
CO?
7.1
7.9
7.8
8.0
7.3
7.6
7.7
Urea Fertilization
2.4
3.5
4.7
4.9
5.0
5.1
5.3
Liming
4.7
4.3
3.1
3.1
2.2
2.4
2.4
ch4
214.7
235.5
244.7
247.8
251.1
250.3
250.9
Enteric Fermentation
163.5
168.0
171.3
174.9
175.7
176.1
175.2
Manure Management
34.8
49.0
57.1
57.5
59.4
58.7
59.6
Rice Cultivation
16.0
18.0
15.8
14.9
15.6
15.1
15.7
Field Burning of Agricultural
Residues
0.4
0.4
0.4
0.4
0.4
0.4
0.4
n2o
330.1
330.3
349.4
347.5
358.4
365.0
336.1
Agricultural Soil Management
316.0
313.8
330.8
328.3
338.9
345.3
316.2
Manure Management
13.9
16.3
18.4
19.0
19.3
19.5
19.7
Field Burning of Agricultural
Residues
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Total
551.9
573.6
601.9
603.2
616.7
622.9
594.7
Note: Totals may not sum due to independent rounding.
Some significant trends in U.S. emissions from Agriculture source categories (Figure 2-9) over the thirty-one-year
period from 1990 through 2020 included the following:
• Agricultural soils are the largest anthropogenic source of N2O emissions in the United States, accounting
for 74.2 percent of N2O emissions in 2020 and 5.3 percent of total emissions in the United States in 2020.
Estimated emissions from this source in 2020 were 316.2 MMT CO2 Eq. Annual N2O emissions from
agricultural soils fluctuated between 1990 and 2020, although overall emissions were only 0.2 MMT CO2
Eq. or 0.1 percent higher in 2020 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 2020,
enteric fermentation CFU emissions were 26.9 percent of total CFU emissions (175.2 MMT CO2 Eq.), which
represents an increase of 11.7 MMT CO2 Eq. (7.2 percent) since 1990. This increase in emissions from
1990 to 2020 in enteric fermentation 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 2020,
consistent with an increase in beef cattle population over those same years.
• Manure management emissions increased 62.5 percent between 1990 and 2020. This encompassed an
increase of 71.0 percent for CFU, from 34.8 MMT CO2 Eq. in 1990 to 59.6 MMT CO2 Eq. in 2020; and an
increase of 41.2 percent for N2O, from 13.9 MMT CO2 Eq. in 1990 to 19.7 MMT CO2 Eq. in 2020. The
majority of the increase observed in CH4 resulted from swine and dairy cattle manure, where emissions
increased 44.2 and 122.0 percent, respectively, from 1990 to 2020. From 2019 to 2020, there was a 0.5
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 are characterized in the LULUCF sector. Estimated emissions from
these sources were 2.4 and 5.3 MMT CO2 Eq., respectively. Liming emissions decreased by 1.3 percent
relative to 2019 and decreased 2.3 MMT CO2 Eq. or 49.0 percent relative to 1990, while urea fertilization
emissions increased by 2.6 percent relative to 2019 and 2.9 MMT CO2 Eq. or 118.3 percent relative to
1990.
Trends 2-21
-------
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 2020 resulted in a net increase in C stocks (i.e., net CO2 removals) of 812.2 MMT CO2 Eq.
(Table 2-8).3 This represents an offset of 13.6 percent of total (i.e., gross) greenhouse gas emissions in 2020.
Emissions of Cm and N2O from LULUCF activities in 2020 were 53.2 MMT CO2 Eq. and represent 0.9 percent of
total greenhouse gas emissions.4 Between 1990 and 2020, total net C sequestration in the LULUCF sector
decreased by 9.0 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 2020, totaling 19.9
MMT CO2 Eq. (797 kt of CH4). Forest fires were the second largest source of CFU emissions from LULUCF in 2020,
totaling 13.6 MMT CO2 Eq. (545 kt of CH4). Coastal Wetlands Remaining Coastal Wetlands resulted in CFU
emissions of 3.8 MMT CO2 Eq. (154 kt of CH4). Grassland fires resulted in CFU emissions of 0.3 MMT CO2 Eq. (12 kt
of CH4). Land Converted to Wetlands, Drained Organic Soils, and Peatlands Remaining Peatlands resulted in CH4
emissions of less than 0.05 MMT CO2 Eq. each.
Forest fires were the largest source of N2O emissions from LULUCF in 2020, totaling 11.7 MMT CO2 Eq. (39 kt of
N2O). Nitrous oxide emissions from fertilizer application to settlement soils in 2020 totaled to 2.5 MMT CO2 Eq. (8
kt of N2O). Additionally, the application of synthetic fertilizers to forest soils in 2020 resulted in N2O emissions of
0.5 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 0.2 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 2-10 and Table 2-8 along with CH4 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.
2-22 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure 2-10: Trends in Emissions and Removals (Net CO2 Flux) from Land Use, Land-Use
Change, and Forestry
o
u
400
300
200 Q^ 0~> (Ji (Ji (Ji 0\ o o o o o
Lnvr>r^cocT*0'-ir\jro*Tu->uDf^coo^o
ooooo—i-»h»-i—
00000000 00000000
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'
2016
2017
2018
2019
2020
(717.3)
(725.6)
7.8
0.5
0.1
(99.5)
(99.5)
(22.7)
(22.7)
54.1
54.1
8.6
S.O
0.6
(22.6)
(22.6)
(670.1)
(688.3)
17.7
0.5
0.1
(99.5)
(99.5)
(22.3)
(22.3)
54.3
54.3
9.9
9.3
0.6
(22.7)
(22.7)
(664.6)
(677.1)
11.9
0.5
0.1
(99.5)
(99.5)
(16.6)
(16.6)
54.0
54.0
10.3
9.7
0.6
(22.4)
(22.4)
(631.8)
(634.8)
2.5
0.5
0.1
(99.5)
(99.5)
(14.5)
(14.5)
53.9
53.9
13.1
12.4
0.6
(21.5)
(21.5)
(642.2)
(668.1)
25.3
0.5
0.1
(99.5)
(99.5)
(23.3)
(23.3)
54.4
54.4
5.1
4.5
0.6
(24.1)
(24.1)
Trends 2-23
-------
Wetlands Remaining Wetlands
14.7
17.2
15.8
15.9
15.9
15.9
15.8
Changes in Organic Soil Carbon Stocks in
Peatlands
1.1
1.1
0.7
0.8
0.8
0.8
0.7
Non-C02 Emissions from Peatlands
Remaining Peatlands
+
+
+
+
+
+
+
Changes in Biomass, DOM, and Soil
Carbon Stocks in Coastal Wetlands
(8.5)
(7.6)
(8.8)
(8.8)
(8.8)
(8.8)
(8.8)
CH4 Emissions from Coastal Wetlands
Remaining Coastal Wetlands
3.7
3.8
3.8
3.8
3.8
3.8
3.8
N20 Emissions from Coastal Wetlands
Remaining Coastal Wetlands
0.1
0.2
0.1
0.1
0.2
0.2
0.2
CH4 Emissions from Flooded Land
Remaining Flooded Land
18.2
19.8
19.9
19.9
19.9
19.9
19.9
Land Converted to Wetlands
7.2
1.3
0.6
0.6
0.6
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
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Changes in Land Converted to Flooded
Land
3.9
0.3
0.3
0.3
0.3
0.3
0.3
CH4 Emissions from Land Converted to
Flooded Land
2.6
0.2
0.2
0.2
0.2
0.2
0.2
Settlements Remaining Settlements
(107.6)
(113.5)
(121.5)
(125.3)
(124.9)
(124.5)
(123.7)
Changes in Organic Soil Carbon Stocks
11.3
12.2
16.0
16.0
15.9
15.9
15.9
Changes in Settlement Tree Carbon
Stocks
(96.4)
(117.4)
(129.8)
(129.8)
(129.8)
(129.8)
(129.8)
N20 Emissions from Settlement Soilsh
2.0
3.1
2.2
2.3
2.4
2.4
2.5
Changes in Yard Trimming and Food
Scrap Carbon Stocks in Landfills
(24.5)
(11.4)
(10.0)
(13.8)
(13.4)
(13.1)
(12.2)
Land Converted to Settlements
60.8
82.8
77.8
77.9
78.0
77.9
77.9
Changes in all Ecosystem Carbon Stocks'
60.8
82.8
77.8
77.9
78.0
77.9
77.9
LULUCF Emissions'
31.4
41.3
35.4
45.5
39.8
30.3
53.2
ch4
27.2
30.9
28.3
34.0
30.7
25.5
38.1
n2o
4.2
10.5
7.1
11.5
9.1
4.8
15.2
LULUCF Carbon Stock Change'
(892.0)
(831.1)
(862.0)
(826.7)
(809.0)
(760.8)
(812.2)
LULUCF Sector Net Totalk
(860.6)
(789.8)
(826.6)
(781.2)
(769.3)
(730.5)
(758.9)
+ 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.
g Estimates include CH4 and N20 emissions from fires on both Grassland Remaining Grassland and Land Converted to
Grass/and.
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
2-24 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 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 2020 in emissions from LULUCF categories (Figure 2-10) over the thirty-one-
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 12.0 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 15.1 percent over the period from
1990 to 2020. 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 28.1 percent from 1990 to 2020 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 greenhouse gas emissions (see Figure 2-11 and Table
2-9). In 2020, landfills were the third-largest source of U.S. anthropogenic Cm emissions, generating 109.3 MMT
CO2 Eq. and accounting for 16.8 percent of total U.S. CH4 emissions.5 Additionally, wastewater treatment
generates emissions of 41.8 MMT CO2 Eq. and accounts for 26.9 percent of waste emissions, 2.8 percent of U.S.
Cm emissions, and 5.5 percent of U.S. N2O emissions. Emissions of CH4 and N2O from composting are also
accounted for in this chapter, generating emissions of 2.3 MMT CO2 Eq. and 2.0 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 155.6
MMT C02Eq„ or 2.6 percent of total U.S. greenhouse gas emissions in 2020.
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.
Trends 2-25
-------
Figure 2-11: Trends in Waste Sector Greenhouse Gas Sources
Table 2-9: Emissions from Waste (MMT CO2 Eq.)
Gas/Source
1990
2005
2016
2017
2018
2019
2020
ch4
197.3
153.6
129.1
130.3
132.4
134.1
130.0
Landfills
176.6
131.5
107.9
109.2
111.7
113.6
109.3
Wastewater Treatment
20.3
20.1
18.7
18.5
18.3
18.1
18.3
Composting
0.4
1.9
2.3
2.5
2.3
2.3
2.3
Anaerobic Digestion at
Biogas Facilities
+
+
0.2
0.2
0.2
0.2
0.2
n2o
16.9
22.0
24.8
25.4
25.5
25.4
25.6
Wastewater Treatment
16.6
20.3
22.8
23.2
23.5
23.4
23.5
Composting
0.3
1.7
2.0
2.2
2.0
2.0
2.0
Total
214.2
175.6
153.9
155.7
157.9
159.6
155.6
+ 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 2-11) over the thirty-one-year
period from 1990 through 2020 included the following:
• Net Cm emissions from landfills decreased by 67.2 MMT CO2 Eq. (38.1 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.
• CH4 and N2O emissions from wastewater treatment decreased by 2.0 MMT CO2 Eq. (10.0 percent) and
increased by 6.9 MMT CO2 Eq. (41.8 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 generally increased 3.6 MMT CO2 Eq. since 1990,
from 0.7 MMT CO2 Eq. to 4.3 MMT CO2 Eq. in 2020, 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
2-26 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
and food waste in landfills; (2) 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 (27.2 percent) of total U.S.
greenhouse gas emissions in 2020. Emissions from electric power accounted for the second largest portion (24.8
percent), while emissions from industry accounted for the third largest portion (23.8 percent) of total U.S.
greenhouse gas emissions in 2020. 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 24.2 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 10.6 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 CO2 sequestration is assigned to the LULUCF sector rather than the agriculture economic
sector. The commercial and residential sectors accounted for roughly 7.1 percent and 6.1 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 2-12 shows the trend in emissions by sector from 1990 to 2020.
Figure 2-12: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors
Trends 2-27
-------
Note: Emissions and removals from Land Use, Land Use Change, and Forestry are excluded from figure above. Excludes U.S.
Territories.
Table 2-10: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO2 Eq. and
Percent of Total in 2020)
Sector/Source
1990
2005
2016
2017
2018
2019
2020
Percent3
Transportation
1,526.4
1,975.5
1,828.0
1,845.2
1,874.7
1,874.3
1,627.6
27.2%
C02 from Fossil Fuel Combustion
1,468.9
1,858.6
1,757.6
1,780.0
1,812.8
1,813.8
1,572.0
26.3%
Substitution of Ozone Depleting
Substances
+
69.3
43.3
40.1
38.5
36.7
35.0
0.6%
Mobile Combustion15
45.7
37.5
16.7
15.5
14.3
15.0
12.5
0.2%
Non-Energy Use of Fuels
11.8
10.2
10.4
9.6
9.2
8.8
8.0
0.1%
Electric Power Industry
1,880.5
2,456.7
1,860.5
1,780.6
1,799.8
1,651.0
1,482.2
24.8%
C02 from Fossil Fuel Combustion
1,820.0
2,400.1
1,808.9
1,732.0
1,752.9
1,606.1
1,439.0
24.1%
Stationary Combustion15
20.9
30.9
27.4
25.9
25.6
22.3
21.0
0.4%
Incineration of Waste
13.4
13.7
14.8
13.6
13.8
13.4
13.5
0.2%
Other Process Uses of Carbonates
3.1
3.7
5.4
4.9
3.7
4.9
4.9
0.1%
Electrical Transmission and
Distribution
23.2
8.4
4.1
4.2
3.8
4.2
3.8
0.1%
Industry
1,652.4
1,536.2
1,424.4
1,446.7
1,507.6
1,521.7
1,426.2
23.8%
C02 from Fossil Fuel Combustion
810.3
800.7
752.6
750.6
774.3
776.4
727.2
12.2%
Natural Gas Systems
227.4
202.5
195.0
197.7
204.2
210.9
200.3
3.3%
Non-Energy Use of Fuels
97.0
111.2
88.4
102.8
119.4
117.7
112.7
1.9%
Petroleum Systems
57.4
53.4
62.3
65.6
75.9
87.1
70.4
1.2%
Coal Mining
101.1
68.3
56.7
57.9
55.8
50.3
43.4
0.7%
Cement Production
33.5
46.2
39.4
40.3
39.0
40.9
40.7
0.7%
Iron and Steel Production
104.8
70.1
43.6
40.6
42.6
43.1
37.7
0.6%
Substitution of Ozone Depleting
Substances
+
8.6
29.9
32.4
34.4
35.7
36.5
0.6%
Petrochemical Production
21.8
27.5
28.4
29.1
29.6
31.0
30.3
0.5%
Landfills (Industrial)
10.9
14.4
15.0
15.0
15.0
15.1
15.1
0.3%
Ammonia Production
13.0
9.2
10.2
11.1
12.2
12.3
12.7
0.2%
Lime Production
11.7
14.6
12.6
12.9
13.1
12.1
11.3
0.2%
Nitric Acid Production
12.1
11.3
10.1
9.3
9.6
10.0
9.3
0.2%
Adipic Acid Production
15.2
7.1
7.1
7.5
10.5
5.3
8.3
0.1%
Wastewater Treatment
6.0
6.5
6.6
6.7
6.8
6.9
6.9
0.1%
Abandoned Oil and Gas Wells
6.5
6.8
6.9
6.9
6.9
7.0
6.9
0.1%
Urea Consumption for Non-
Agricultural Purposes
3.8
3.7
5.3
5.2
6.0
6.0
6.0
0.1%
Abandoned Underground Coal
Mines
7.2
6.6
6.7
6.4
6.2
5.9
5.8
0.1%
Mobile Combustion15
3.9
6.1
5.6
5.8
6.0
6.1
5.7
0.1%
Carbon Dioxide Consumption
1.5
1.4
4.6
4.6
4.1
4.9
5.0
0.1%
Other Process Uses of Carbonates
3.1
3.7
5.4
4.9
3.7
4.9
4.9
0.1%
Electronics Industry
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
+%
2-28 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Caprolactam, Glyoxal, and
Glyoxylic Acid Production
1.7
2.1
1.7
1.5
1.4
1.4
1.2
+%
Zinc Production
0.6
1.0
0.8
0.9
1.0
1.0
1.0
+%
Phosphoric Acid Production
1.5
1.3
1.0
1.0
0.9
0.9
0.9
+%
Magnesium Production and
Processing
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
596.8
626.3
643.4
644.4
657.9
663.9
635.1
10.6%
N20 from Agricultural Soil
Management
316.0
313.8
330.8
328.3
338.9
345.3
316.2
5.3%
Enteric Fermentation
163.5
168.0
171.3
174.9
175.7
176.1
175.2
2.9%
Manure Management
48.8
65.3
75.5
76.5
78.7
78.2
79.2
1.3%
C02 from Fossil Fuel Combustion
43.4
50.8
40.2
39.8
39.8
39.7
39.1
0.7%
Rice Cultivation
16.0
18.0
15.8
14.9
15.6
15.1
15.7
0.3%
Urea Fertilization
2.4
3.5
4.7
4.9
5.0
5.1
5.3
0.1%
Liming
4.7
4.3
3.1
3.1
2.2
2.4
2.4
+%
Mobile Combustion15
1.5
1.8
1.4
1.4
1.4
1.4
1.3
+%
Field Burning of Agricultural
Residues
0.5
0.6
0.6
0.6
0.6
0.6
0.6
+%
Stationary Combustion15
+
+
+
+
+
+
+
+%
Commercial
427.1
405.4
426.9
428.5
444.2
452.1
425.3
7.1%
C02 from Fossil Fuel Combustion
228.3
227.1
231.5
232.0
245.8
250.7
226.8
3.8%
Landfills (Municipal)
165.7
117.2
93.0
94.2
96.7
98.6
94.2
1.6%
Substitution of Ozone Depleting
Substances
+
22.1
61.5
61.0
60.8
62.3
63.5
1.1%
Wastewater Treatment
30.9
34.0
34.9
35.0
35.0
34.5
34.8
0.6%
Composting
0.7
3.5
4.3
4.6
4.3
4.3
4.3
0.1%
Stationary Combustion15
1.5
1.4
1.5
1.5
1.6
1.6
1.5
+%
Anaerobic Digestion at Biogas
Facilities
+
+
0.2
0.2
0.2
0.2
0.2
+%
Residential
345.1
371.0
327.8
329.9
377.4
384.2
362.0
6.1%
C02 from Fossil Fuel Combustion
338.6
358.9
292.8
293.4
338.2
341.4
315.8
5.3%
Substitution of Ozone Depleting
Substances
0.2
7.2
30.4
31.9
33.7
37.1
41.2
0.7%
Stationary Combustion15
6.3
4.9
4.7
4.5
5.5
5.7
4.9
0.1%
U.S. Territories
25.1
63.7
26.8
25.8
25.8
24.6
23.0
0.4%
C02 from Fossil Fuel Combustion
21.7
55.9
26.0
25.5
25.5
24.3
22.7
0.4%
Non-Energy Use of Fuels
3.4
7.6
0.7
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)
6,453.5
7,434.8
6,537.9
6,501.0
6,687.5
6,571.7
5,981.4
100.0%
LULUCF Sector Net Totalc
(860.6)
(789.8)
(826.6)
(781.2)
(769.3)
(730.5)
(758.9)
(12.7%)
Net Emissions (Sources and Sinks)
5,592.8
6,645.0
5,711.2
5,719.8
5,918.2
5,841.2
5,222.4
87.3%
+ Does not exceed 0.05 MMT C02 Eq. or 0.05 percent.
a Percent of total (gross) emissions excluding emissions from LULUCF for 2020.
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 net carbon stock
changes.
Notes: Total emissions presented without LULUCF. Total net emissions presented with LULUCF. Totals may not sum due to
independent rounding. Parentheses indicate negative values or sequestration.
Trends 2-29
-------
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, CH4 and N2O emissions from the combustion of fossil fuels that
are included in the EIA electric power sector. Carbon dioxide, Cm, 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 CH4 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. CO2 emissions from fossil fuel combustion, and
Cm and N2O emissions from stationary and mobile combustion, are then apportioned to the Agriculture
economic sector based on agricultural fuel use.
The other IPCC Agriculture emission source categories apportioned to the Agriculture economic sector include
2-30 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
N O emissions from Agricultural Soils, CH from Enteric Fermentation, CH and N. O from Manure Management,
CH ¦ from Rice Cultivation, CO emissions from Liming and Urea Application, and CH and N. O from Field Burning
of Agricultural Residues.
The Residential economic sector includes CO emissions from the combustion of fossil fuels that are included in
the EIA residential fuel-consuming sector. Stationary combustion emissions of CH and N O 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 CO. emissions from the combustion of fossil fuels that are included in
the EIA commercial fuel-consuming sector. Emissions of CH ¦ and N O 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 CH from municipal landfills, CH from anaerobic digestion at biogas facilities, CH and N. O from domestic
wastewater treatment, and composting, are also included in the Commercial economic sector.
Emissions with Electricity Distributed to Economic Sectors
It is also useful to view greenhouse gas emissions from economic sectors with emissions related to electric power
distributed into end-use categories (i.e., emissions from electric power are allocated to the economic sectors in
which the electricity is used).
The generation, transmission, and distribution of electricity accounted for 24.8 percent of total U.S. greenhouse
gas emissions in 2020. Electric power-related emissions decreased by 21.2 percent since 1990 and by 10.2 percent
from 2019 to 2020, due to the impacts of the COVID-19 pandemic in 2020 compared to 2019 and continued fuel
switching in the electric power sector. Between 2019 to 2020, the consumption of natural gas for electric power
generation increased by 2.9, while the consumption of coal and petroleum decreased by 19.2 and 2.2 percent,
respectively, reflecting a continued shift from coal to natural gas for electricity generation.
From 2019 to 2020, electricity sales to the residential end-use sector increased by 1.7 percent. Alternatively,
electricity sales to the commercial end-use and industrial sectors decreased by 5.4 percent and 4.3 percent,
respectively. Overall, from 2019 to 2020, the amount of electricity retail sales (in kWh) decreased by 2.5 percent.
Table 2-11 provides a detailed summary of emissions from electric power-related activities.
Trends 2-31
-------
Table 2-11: Electric Power-Related Greenhouse Gas Emissions (MMT CO2 Eq.)
Gas/Fuel Type or Source
1990
2005
2016
2017
2018
2019
2020
CO?
1,836.0
2,417.1
1,828.6
1,750.1
1,770.0
1,624.0
1,457.0
Fossil Fuel Combustion
1,820.0
2,400.1
1,808.9
1,732.0
1,752.9
1,606.1
1,439.0
Coal
1,546.5
1,982.8
1,242.0
1,207.1
1,152.9
973.5
788.2
Natural Gas
175.4
318.9
545.0
505.6
577.4
616.0
634.3
Petroleum
97.5
98.0
21.5
18.9
22.2
16.2
16.2
Geothermal
0.5
0.5
0.4
0.4
0.4
0.4
0.4
Incineration of Waste
12.9
13.3
14.4
13.2
13.3
12.9
13.1
Other Process Uses of Carbonates
3.1
3.7
5.4
4.9
3.7
4.9
4.9
ch4
0.4
0.9
1.2
1.1
1.2
1.3
1.2
Stationary Sources3
0.4
0.9
1.2
1.1
1.2
1.3
1.2
Incineration of Waste
+
+
+
+
+
+
+
n2o
21.0
30.4
26.6
25.2
24.8
21.5
20.1
Stationary Sources3
20.5
30.1
26.2
24.8
24.4
21.1
19.7
Incineration of Waste
0.5
0.4
0.4
0.4
0.4
0.4
0.4
sf6
23.2
8.3
4.1
4.2
3.8
4.2
3.8
Electrical Transmission and Distribution
23.2
8.3
4.1
4.2
3.8
4.2
3.8
PFCs
+
+
+
+
+
+
+
Electrical Transmission and Distribution
+
+
+
+
+
+
+
Total
1,880.5
2,456.7
1,860.5
1,780.6
1,799.8
1,651.0
1,482.2
+ 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 industrial processes economic end-use sector.6
When emissions from electricity use are distributed among these economic end-use sectors, industrial activities
account for the largest share of total U.S. greenhouse gas emissions (30.3 percent), followed closely by emissions
from transportation (27.3 percent). Emissions from the residential and commercial sectors also increase
substantially when emissions from electricity are included (both 15.4 percent). In all economic end-use sectors
except agriculture, CO2 accounts for more than 77.5 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 2-13 shows the trend in these emissions by
sector from 1990 to 2020.
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.
2-32 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure 2-13: U.S. Greenhouse Gas Emissions with Electricity-Related Emissions Distributed
to Economic Sectors
Note: Emissions and removals from Land Use, Land-Use Change, and Forestry are excluded from figure above. Excludes U.S.
Territories.
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 2020
Sector/Gas
1990
2005
2016
2017
2018
2019
2020
Percent3
Industry
2,326.5
2,251.6
1,917.5
1,926.4
1,983.1
1,964.7
1,813.7
30.3%
Direct Emissions
1,652.4
1,536.2
1,424.4
1,446.7
1,507.6
1,521.7
1,426.2
23.8%
C02
1,163.3
1,143.6
1,056.1
1,072.4
1,127.9
1,148.8
1,066.3
17.8%
ch4
372.7
319.5
296.8
299.1
300.3
297.0
283.2
4.7%
n2o
40.1
33.6
31.3
30.7
34.1
29.3
31.0
0.5%
HFCs, PFCs, SF6 and NF3
76.3
39.5
40.1
44.4
45.3
46.6
45.6
0.8%
Electricity-Related
674.1
715.4
493.2
479.7
475.4
443.1
387.5
6.5%
C02
658.1
703.8
484.7
471.5
467.6
435.8
380.9
6.4%
ch4
0.2
0.3
0.3
0.3
0.3
0.3
0.3
+%
n2o
7.5
8.9
7.1
6.8
6.6
5.8
5.3
0.1%
sf6
8.3
2.4
1.1
1.1
1.0
1.1
1.0
+%
Transportation
1,529.6
1,980.3
1,832.4
1,849.6
1,879.5
1,879.1
1,632.4
27.3%
Direct Emissions
1,526.4
1,975.5
1,828.0
1,845.2
1,874.7
1,874.3
1,627.6
27.2%
C02
1,480.8
1,868.7
1,768.0
1,789.5
1,822.0
1,822.6
1,580.1
26.4%
ch4
5.7
2.9
1.6
1.5
1.4
1.4
1.2
+%
n2o
39.9
34.5
15.1
14.0
12.9
13.5
11.3
0.2%
HFCsb
+
69.3
43.3
40.1
38.5
36.7
35.0
0.6%
Electricity-Related
3.1
4.8
4.3
4.4
4.8
4.9
4.8
0.1%
C02
3.1
4.8
4.2
4.3
4.7
4.8
4.7
0.1%
ch4
+
+
+
+
+
+
+
+%
n2o
+
0.1
0.1
0.1
0.1
0.1
0.1
+%
sf6
+
+
+
+
+
+
+
+%
Residential
957.6
1,247.2
999.9
964.3
1,036.7
984.1
923.1
15.4%
Direct Emissions
345.1
371.0
327.8
329.9
377.4
384.2
362.0
6.1%
C02
338.6
358.9
292.8
293.4
338.2
341.4
315.8
5.3%
ch4
5.2
4.1
3.9
3.8
4.6
4.7
4.1
0.1%
n2o
1.0
0.9
0.8
0.8
0.9
0.9
0.8
+%
Trends 2-33
-------
sf6
0.2
7.2
30.4
31.9
33.7
37.1
41.2
0.7%
Electricity-Related
612.5
876.2
672.1
634.4
659.4
599.9
561.1
9.4%
C02
598.0
862.1
660.6
623.6
648.4
590.1
551.6
9.2%
ch4
0.1
0.3
0.4
0.4
0.4
0.5
0.5
+%
n2o
6.8
10.9
9.6
9.0
9.1
7.8
7.6
0.1%
sf6
7.5
3.0
1.5
1.5
1.4
1.5
1.4
+%
Commercial
982.7
1,227.4
1,078.6
1,051.7
1,065.3
1,020.1
919.7
15.4%
Direct Emissions
427.1
405.4
426.9
428.5
444.2
452.1
425.3
7.1%
C02
228.3
227.1
231.5
232.0
245.8
250.7
226.8
3.8%
ch4
181.9
134.3
109.2
110.3
112.3
113.9
109.6
1.8%
n2o
16.9
21.8
24.6
25.2
25.3
25.3
25.4
0.4%
HFCs
+
22.1
61.5
61.0
60.8
62.3
63.5
1.1%
Electricity-Related
555.6
822.0
651.7
623.2
621.1
567.9
494.4
8.3%
C02
542.5
808.7
640.6
612.6
610.8
558.6
486.0
8.1%
ch4
0.1
0.3
0.4
0.4
0.4
0.4
0.4
+%
n2o
6.2
10.2
9.3
00
00
8.6
7.4
6.7
0.1%
sf6
6.8
2.8
1.4
1.5
1.3
1.5
1.3
+%
Agriculture
631.9
664.6
682.6
683.2
697.1
699.1
669.5
11.2%
Direct Emissions
596.8
626.3
643.4
644.4
657.9
663.9
635.1
10.6%
C02
50.5
58.7
47.9
47.8
47.1
47.2
46.7
0.8%
ch4
214.8
235.7
244.8
247.9
251.2
250.5
251.0
4.2%
n2o
331.5
331.9
350.7
348.7
359.6
366.2
337.3
5.6%
Electricity-Related
35.2
38.3
39.2
38.8
39.2
35.2
34.4
0.6%
C02
34.3
37.7
38.5
38.1
38.5
34.6
33.8
0.6%
ch4
+
+
+
+
+
+
+
+%
n2o
0.4
0.5
0.6
0.5
0.5
0.5
0.5
+%
sf6
0.4
0.1
0.1
0.1
0.1
0.1
0.1
+%
U.S. Territories
25.1
63.7
26.8
25.8
25.8
24.6
23.0
0.4%
Total Gross Emissions
(Sources)
6,453.5
7,434.8
6,537.9
6,501.0
6,687.5
6,571.7
5,981.4
100.0%
LULUCF Sector NetTotalc
(860.6)
(789.8)
(826.6)
(781.2)
(769.3)
(730.5)
(758.9)
(12.7%)
Net Emissions (Sources and
Sinks)
5,592.8
6,645.0
5,711.2
5,719.8
5,918.2
5,841.2
5,222.4
87.3%
+ Does not exceed 0.05 MMT C02 Eq. or 0.05 percent.
a Percent of total (gross) emissions excluding emissions from LULUCF for year 2020.
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 net carbon stock
changes.
Notes: Total emissions presented without LULUCF. Net emissions 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 30.3 percent of U.S. greenhouse
gas emissions in 2020. 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 22.0 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.
2-34 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Transportation
When electricity-related emissions are distributed to economic end-use sectors, transportation activities
accounted for 27.2 percent of U.S. greenhouse gas emissions in 2020. The largest sources of transportation
greenhouse gas emissions in 2020 were passenger cars (37.8 percent); freight trucks (25.9 percent); light-duty
trucks, which include sport utility vehicles, pickup trucks, and minivans (19.3 percent); commercial aircraft (5.6
percent); pipelines (3.5 percent); rail (2.1 percent); ships and boats (2.0 percent); and other aircraft (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 2020, total transportation emissions from fossil fuel combustion increased by approximately 4.8
percent. From 1990 to 2019, emissions increased by 20.9 percent, followed by a decline of 13.3 percent from 2019
to 2020. The increase in transportation emissions from 1990 to 2019 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 47.5
percent from 1990 to 2019, as a result of a confluence of factors including population growth, economic growth,
urban sprawl, and periods of low fuel prices. The drop in transportation emissions from 2019 to 2020 was primarily
the result of less travel caused by the COVID-19 pandemic. During this period, the number of VMT by light-duty
motor vehicles decreased by 12.2 percent.
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 2013,7 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 56.1 percent. Light-
duty truck market share was about 56.1 percent of new vehicles in model year 2020 (EPA 2021a).
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 7.0 percent from 1990
to 2020.8 This rise in CO2 emissions, combined with an increase in HFCs from close to zero emissions in 1990 to
35.0 MMT CO2 Eq. in 2020, led to an increase in overall greenhouse gas emissions from transportation activities of
6.7 percent.9
7 VMT estimates are based on data from FHWA Highway Statistics Table VM-1 (FHWA 1996 through 2021). In 2007 and 2008
light-duty VMT decreased 3.9 percent and 2.3 percent, respectively. Note that the decline in light-duty VMT from 2006 to 2007
is due at least in part to a change in FHWA's methods for estimating VMT. In 2011, FHWA changed its methods for estimating
VMT by vehicle class, which led to a shift in VMT and emissions among on-road vehicle classes in the 2007 to 2018 time period.
In absence of these method changes, light-duty VMT growth between 2006 and 2007 would likely have been higher.
8 See previous footnote.
9 See previous footnote.
Trends 2-35
-------
Figure 2-
2,800
2,600
2,400
2,200
2,000
1,800
£ 1,600
IN
8 1,400
I-
I 1,200
1,000
800
600
400
200
0
OH f\jrOfiniCNODaiOH(Mf0^irii£lN03ffiO'H(NnTinUDN03ffiO
O C* C> C> C?» O. C1- O O O O O O O O O O *-if-t*—it-i T-t th ^ t-i *-t f>J
o\o*o*o*o*ooooooooo©o©ooooooooo
HrtHrtrtrtHHrtH(\i(\(N(\|fM(N(N(N(N(\fNN(N(N(NN(\(NN(N(N
Table 2-13: Transportation-Related Greenhouse Gas Emissions (MMT CO2 Eq.)
Gas/Vehicie
1990
2005
2016
2017
2018
2019
2020
Passenger Cars
639.6
691.7
763.2
760.6
770.2
763.1
617.7
C02
612.2
641.4
742.4
742.5
754.2
748.0
605.0
ch4
3.2
1.3
0.6
0.5
0.5
0.5
0.4
n2o
24.1
17.3
7.1
6.1
5.1
5,3
3.9
HFCs
0.0
31.7
13.2
11.4
10.4
9.3
8.3
Light-Duty Trucks
326,7
537.7
330.0
324.3
325.6
323.7
315.8
C02
312.2
490.0
305.8
302.4
305.2
304.0
297.8
ch4
1.7
0.8
0.2
0.2
0.2
0.2
0.2
n2o
12.8
13.6
2.9
2.4
2.1
2.5
2.2
HFCs
0.0
33.3
21.1
19.2
18.1
16.9
15.6
Medium- and Heavy-Duty Trucks
230.3
404.1
416.8
429.7
440.0
439.5
422.8
C02
229.3
399.4
408.2
420.8
430.7
429.9
412.9
ch4
0.3
0,1
0.1
0.1
0.1
0.1
0.1
n2o
0.7
1.2
2.9
3.1
3.3
3.4
3.4
HFCs
0.0
3.4
5.5
5.7
5.9
6.1
6.3
Buses
8.5
12.3
19.0
20.5
21.8
21.7
18.0
C02
8.4
11.8
18.4
19.8
21.1
21.1
17.4
14: Trends in Transportation-Related Greenhouse Gas Emissions10
Lubricants U Ships and Boats
Motorcycles Aircraft
Buses ¦ Medium- and Heavy-Duty Trucks
I Pipelines I Light-Duty Trucks
Rail ¦ Passenger Cars
rs.
m
vjD
Cr>
CO
in
r-*
hs
H
¦H
i—i
S?
GO
O"1 O*1
0 o
K CO
01
o o
CO CO
G\ cn
vO
r-v.
<»
T—1
o
rv.
o
i
oo
ca
i
T—1
¦
IB
o
in
cq.
o
cn
rvj
CO
Is-
m
CO
c»
1—1
c»
tH
¦
—
¦
10 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 for the 1990 through 2008 Inventory and apply to the 2007 to 2020 time period. This resulted in large changes in
VMT data by vehicle class, leading to a shift in emissions among on-road vehicle classes. This change in vehicle classification has
moved some smaller trucks and sport utility vehicles from the light truck category to the passenger vehicle category in this
Inventory.
2-36 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
ch4
+
0.2
0.1
0.1
0.1
0.1
+
n2o
+
+
0.1
0.1
0.2
0.2
0.1
HFCs
0.0
0.3
0.4
0.4
0.4
0.4
0.4
Motorcycles
1.7
1.6
3.9
3.8
3.9
3.7
3.3
C02
1.7
1.6
3.8
3.7
3.8
3.6
3.2
ch4
+
+
+
+
+
+
+
n2o
+
+
0.1
0.1
0.1
0.1
0.1
Commercial Aircraft3
110.9
133.9
121.5
129.2
130.8
135.4
92.1
C02
109.9
132.7
120.4
128.0
129.6
134.2
91.3
ch4
0.0
0.0
0.0
0.0
0.0
0.0
0.0
n2o
1.0
1.2
1.1
1.2
1.2
1.2
0.8
Other Aircraftb
78.1
59.6
47.4
45.6
44.6
45.6
31.0
C02
77.3
59.0
47.0
45.1
44.2
45.2
30.7
ch4
0.1
0.1
+
+
+
+
+
n2o
0.7
0.5
0.4
0.4
0.4
0.4
0.3
Ships and Boatsc
47.0
45.4
40.7
43.8
41.1
40.0
32.3
C02
46.3
44.3
37.1
39.9
36.9
35.5
27.6
ch4
0.4
0.4
0.4
0.4
0.4
0.4
0.4
n2o
0.3
0.3
0.2
0.2
0.2
0.2
0.2
HFCs
0.0
0.5
2.9
3.3
3.6
3.9
4.2
Rail
39.0
51.5
40.2
41.4
42.5
39.7
34.2
C02
38.5
50.8
39.6
40.7
41.8
39.1
33.7
ch4
0.1
0.1
0.1
0.1
0.1
0.1
0.1
n2o
0.3
0.4
0.3
0.4
0.4
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
32.4
39.2
41.3
49.9
57.9
57.1
C02
36.0
32.4
39.2
41.3
49.9
57.9
57.1
Lubricants
11.8
10.2
10.4
9.6
9.2
8.8
8.0
C02
11.8
10.2
10.4
9.6
9.2
00
00
8.0
Total Transportation
1,529.6
1,980.3
1,832.4
1,849.6
1,879.5
1,879.1
1,632.4
International Bunker FuelsS
54.8
44.7
35.0
34.6
32.5
26.4
22.7
Ethanol C02g
4.1
21.6
76.9
77.7
78.6
78.7
68.1
Biodisel C029
0.0
0.9
19.6
18.7
17.9
17.1
17.7
+ 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.
Trends 2-37
-------
Residential
The residential end-use sector, with electricity-related emissions distributed, accounts for 15.4 percent of U.S.
greenhouse gas emissions in 2020 and similarly, 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. Emissions from the residential 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. 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 15.4 percent of U.S.
greenhouse gas emissions in 2020 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 decreasing slightly.
Agriculture
The agriculture end-use sector accounts for 11.2 percent of U.S. greenhouse gas emissions in 2020 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 2020, agricultural soil
management was the largest source of N2O emissions, and enteric fermentation was the largest source of CH4
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 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 0.2 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 2-15). 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 1.4 percent since 2005. Fossil fuel consumption has also decreased at a slower rate than
2-38 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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)
Variable
1990
2005
2016
2017
2018
2019
2020
Avg. Annual
Change
Since 1990a
Avg. Annual
Change
Since 2005a
Greenhouse Gas Emissions'5
100
115
101
101
104
102
93
-0.2%
-1.4%
Energy Usec
100
119
116
116
120
119
109
0.3%
-0.5%
GDPd
100
159
189
193
199
203
196
2.3%
1.4%
Population6
100
118
128
129
129
131
132
0.9%
0.8%
a Average annual growth rate.
b GWP-weighted values.
c Energy-content-weighted values (EIA 2022).
d GDP in chained 2009 dollars (BEA 2021).
e U.S. Census Bureau (2021).
Figure 2-15: U.S. Greenhouse Gas Emissions Per Capita and Per Dollar of Gross Domestic
Product
Source: BEA (2021), U.S. Census Bureau (2021), and emission estimates in this report.
2.3 Precursor Greenhouse Gas Emissions
(CO, NOx, NMVOCs, and S02)
The reporting requirements of the UNFCCC11 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
11 See http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
Trends 2-39
-------
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
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.12 Precursor emission estimates for this report for 1990 through 2020 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 (2022) 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
2016
2017
2018
2019
2020
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
12 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/.
2-40 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
so2
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 j
466
509
509
509
509
Waste
+
1
1
1
1
1
1
Agriculture
NA
NA
NA
NA
NA
NA
NA
LULUCF
NA
NA J
NA
NA
NA
NA
NA
+ Does not exceed 0.5 kt.
NA (Not Available)
Note: Totals 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.
Trends 2-41
-------
3. Energy
Energy-related activities were the primary sources of U.S. anthropogenic greenhouse gas emissions, accounting for
81.2 percent of total greenhouse gas emissions on a carbon dioxide (CO2) equivalent basis in 2020.1 This included
96.4, 41.4, and 9.6 percent of the nation's CO2, methane (CH4), and nitrous oxide (N2O) emissions, respectively.
Energy-related CO2 emissions alone constituted 76.0 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 (5.2 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 31,500 million metric tons (MMT) of
CO2 were added to the atmosphere through the combustion of fossil fuels in 2020, 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 third 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, 1990-2020 - Charts Available at: https://www.iea.org/data-and-statistics/charts/global-energy-related-co2-
emissions-1990-2020 (IEA 2021).
Energy 3-1
-------
Figure 3-1: 2020 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
,R" 5,000
O
U
4,000
3,000
2,000
1,000
0
Incineration of Waste
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
("v. 00 2] *-l
in S n 5 - -
~ § S ^
s a 2 *-
§ S ° s? S o
kD kO -
o th cn m
(Ti Oi Oi O
(Ji
3-2 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure 3-3: 2020 U.S. Fossil Carbon Flows
Table 3-1 summarizes emissions from the Energy sector in units of MMT CO2 Eq., while unweighted gas emissions
in kilotons (kt) are provided in Table 3-2. Overall, emissions due to energy-related activities were 4,854.7 MMT CO2
Eq. in 2020,3 a decrease of 9.1 percent since 1990 and a decrease of 10.3 percent since 2019. The decrease in
2020 emissions was due primarily to the coronavirus (COVID-19) pandemic reducing overall demand for fossil fuels
across all sectors but it also reflects a continued shift from coal to natural gas and renewables in the electric power
sector.
Table 3-1: CO2, ChU, and N2O Emissions from Energy (MMT CO2 Eq.)
Gas/Source
1990
2005
2016
2017
2018
2019
2020
C02
4,902.4
5,935.4
5,078.0
5,038.3
5,204.3
5,080.4
4,544.5
Fossil Fuel Combustion
4,731.2
5,752.0
4,909.6
4,853.3
4,989.3
4,852.3
4,342.7
Transportation
1,468.9
1,858.6
1,757.6
1,780.0
1,812.8
1,813.8
1,572.0
Electricity Generation
1,820.0
2,400.1
1,808.9
1,732.0
1,752.9
1,606.1
1,439.0
Industrial
853.7
851.5
792.7
790.4
814.1
816.1
766.3
Residential
338.6
358.9
292.8
293.4
338.2
341.4
315.8
Commercial
228.3
227.1
231.5
232.0
245.8
250.7
226.8
U.S. Territories
21.7
55.9
26.0
25.5
25.5
24.3
22.7
Non-Energy Use of Fuels
112.2
128.9
99.5
112.6
128.9
126.8
121.0
Natural Gas Systems
31.9
24.9
29.8
31.1
32.4
38.7
35.4
Petroleum Systems
9.6
12.0
21.9
25.0
37.3
46.7
30.2
Incineration of Waste
12.9
13.3
14.4
13.2
13.3
12.9
13.1
Coal Mining
4.6
4.2
2.8
3.1
3.1
3.0
2.2
Abandoned Oil and Gas Wells
+
+
+
+
+
+
+
Biomass-Wooda
215.2
206.9
216.0
211.9
220.0
217.6
202.1
Biofuels-Ethanola
4.2
22.9
81.2
82.1
81.9
82.6
71.8
International Bunker Fuelsb
103.6
113.3
116.7
120.2
122.2
116.1
69.6
Biofuels-Biodiesela
0.0
0.9
19.6
18.7
17.9
17.1
17.7
ch4
368.6
308.3
283.5
285.4
287.3
284.0
269.1
3 Following the current reporting requirements under the UNFCCC, this Inventory report presents C02 equivalent values based
on the IPCC Fourth Assessment Report (AR4) GWP values. See the Introduction chapter for more information.
Energy 3-3
-------
Natural Gas Systems
195.5
177.5
165.2
166.6
171.8
172.1
164.9
Coal Mining
96.5
64.1
53.8
54.8
52.7
47.4
41.2
Petroleum Systems
47.8
41.4
40.4
40.5
38.6
40.4
40.2
Stationary Combustion
8.6
7.8
7.9
7.7
8.6
8.8
7.9
Abandoned Oil and Gas Wells
6.5
6.8
6.9
6.9
6.9
7.0
6.9
Abandoned Underground Coal
Mines
7.2
6.6
6.7
6.4
6.2
5.9
5.8
Mobile Combustion
6.5
4.0
2.6
2.6
2.5
2.5
2.2
Incineration of Waste
+
+
+
+
+
+
+
International Bunker Fuelsb
0.2
0.1
0.1
0.1
0.1
0.1
0.1
n2o
70.1
76.1
51.5
49.0
47.9
45.3
41.1
Stationary Combustion
25.1
34.4
30.0
28.4
28.2
24.9
23.2
Mobile Combustion
44.6
41.4
21.1
20.1
19.2
20.0
17.4
Incineration of Waste
0.5
0.4
0.4
0.4
0.4
0.4
0.4
Petroleum Systems
+
+
+
+
+
+
+
Natural Gas Systems
+
+
+
+
+
+
+
International Bunker Fuelsb
0.9
1.0
1.0
1.1
1.1
1.0
0.6
Total
5,341.1
6,319.8
5,413.1
5,372.7
5,539.4
5,409.8
4,854.7
+ Does not exceed 0.05 MMT C02 Eq.
a Emissions from Wood 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
2016
2017
2018
2019
2020
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
CH4
Natural Gas
Systems
Coal Mining
Petroleum Systems
Stationary
Combustion
4,902,396
4,731,178
112,175
31,894
9,600
12,937
4,606
6
215,186
4,227
103,634
0
14,744
7,821
3,860
1,912
344
5,935,361
5,752,043
128,920
24,945
11,994
13,283
4,170
7
206,901
22,943
113,328
856
12,331
7,100
2,565
1,655
313
5,078,027 5,038,320 5,204,305 5,080,437 4,544,464
4,909,609 4,853,299 4,989,308 4,852,330 4,342,659
99,505 112,616 128,871 126,776 120,987
29,780
21,922
14,356
2,848
7
215,955
81,250
116,682
19,648
11,342
6,609
2,154
1,616
315
31,145
25,027
13,161
3,067
7
211,925
82,088
120,192
18,705
11,417
6,662
2,191
1,621
307
32,407
37,306
13,339
3,067
7
219,951
81,917
122,179
17,936
11,492
6,871
2,109
1,544
344
38,740
46,686
12,948
2,951
7
217,574
82,578
116,132
17,080
11,360
6,885
1,895
1,615
351
35,353
30,156
13,133
2,169
7
202,088
71,847
69,638
17,678
10,766
6,596
1,648
1,609
317
3-4 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
261
288
259
7
235
84
150
2
+
273
264
161
5
255
115
139
1
+
275
268
105
4
173
101
71
1
+
276
257
102
4
164
95
68
1
+
277
247
99
4
161
95
64
1
+
279
237
99
4
152
84
67
1
+
276
231
3
138
78
58
1
+
+ Does not exceed 0.5 kt.
a Emissions from Wood Biomass, Ethanol, 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.
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.
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 2019) to ensure that the trend
is accurate. Key updates in this year's Inventory include updates to the Incineration of Waste methodology (e.g.,
new waste tonnage estimates data sources to replace proxied data, new GHGRP carbon emission factor, and new
MSW incineration activity data), updated emission factors for CFU and N2O from newer non-road gasoline and
diesel vehicles for emissions from Mobile Combustion, revisions to the Natural Gas Systems methodology (e.g.,
inclusion of post-meter emissions, adding well blowout emissions, and changes to methane reduction data
processing), changes to the Abandoned Oil and Gas Wells methodology to improve estimates of plugged wells,
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). The combined impact of these recalculations averaged 19.3 MMT CO2 Eq. (+0.3 percent)
per year across the time series. For more information on specific methodological updates, please see the
Recalculations Discussion section for each category in this chapter.
Energy 3-5
-------
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 3.6 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 3.6 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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
2016
2017
2018
2019
2020
C02
4,731.2
5,752.0
4,909.6
4,853.3
4,989.3
4,852.3
4,342.7
ch4
15.1
11.9
10.5
10.2
11.1
11.2
10.1
n2o
69.7
75.7
51.1
48.6
47.4
44.9
40.6
Total
4,815.9
5,839.6
4,971.2
4,912.1
5,047.8
4,908.4
4,393.4
Note: Totals may not sum due to independent rounding.
Table 3-4: CO2, ChU, and N2O Emissions from Fossil Fuel Combustion (kt)
Gas
1990
2005
2016
2017
2018
2019
2020
C02
4,731,178
5,752,043
4,909,609
4,853,299
4,989,308
4,852,330
4,342,659
ch4
603
474
420
409
443
450
406
n2o
234
254
171
163
159
151
136
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
2020, CO2 emissions from fossil fuel combustion decreased by 10.5 percent relative to the previous year (as shown
in Table 3-6). The decrease in CO2 emissions from fossil fuel consumption was a result of a 9.2 percent decrease in
fossil fuel energy use. This decrease in fossil fuel consumption was due primarily to the COVID-19 pandemic but
also reflects a continued shift from coal to natural gas and renewables. Carbon dioxide emissions from both
natural gas and coal consumption decreased in 2020. CO2 emissions from natural gas decreased by 38.0 MMT CO2
Eq., a 2.3 percent decrease from 2019. CO2 emissions from coal consumption decreased by 192.6 MMT CO2 Eq., an
18.7 percent decrease from 2019. The decrease in natural gas consumption and emissions in 2020 is observed
across all sectors except the Electric Power sector. This increase in the Electric Power sector is primarily driven by a
Energy 3-7
-------
continued shift away from coal consumption to natural gas. In 2020, CO2 emissions from fossil fuel combustion
were 4,342.7 MMT CO2 Eq., or 8.2 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
2016
2017
2018
2019
2020
Coal
1,719.8
2,113.7
1,310.7
1,270.0
1,211.6
1,028.2
835.6
Residential
3.0
0.8
0.0
0.0
0.0
0.0
0.0
Commercial
12.0
9.3
2.3
2.0
1.8
1.6
1.4
Industrial
157.8
117.8
63.2
58.7
54.4
49.5
43.0
Transportation
NO
NO
NO
NO
NO
NO
NO
Electric Power
1,546.5
1,982.8
1,242.0
1,207.1
1,152.9
973.5
788.2
U.S. Territories
0.5
3.0
3.3
2.3
2.6
3.6
3.1
Natural Gas
1,000.0
1,167.0
1,461.3
1,434.6
1,592.0
1,648.8
1,610.7
Residential
237.8
262.2
238.4
241.5
273.8
275.5
256.4
Commercial
142.0
162.9
170.5
173.2
192.5
192.9
173.9
Industrial
408.8
388.6
463.9
469.5
494.0
501.6
485.5
Transportation
36.0
33.1
40.1
42.3
50.9
58.9
58.1
Electric Power
175.4
318.9
545.0
505.6
577.4
616.0
634.3
U.S. Territories
NO
1.3
3.4
2.5
3.3
3.8
2.6
Petroleum
2,010.9
2,470.9
2,137.2
2,148.3
2,185.3
2,175.0
1,895.9
Residential
97.8
95.9
54.4
51.9
64.4
65.9
59.5
Commercial
74.3
54.9
58.7
56.8
51.5
56.2
51.6
Industrial
287.1
345.0
265.7
262.2
265.7
265.0
237.8
Transportation
1,432.9
1,825.5
1,717.6
1,737.7
1,761.8
1,754.8
1,514.0
Electric Power
97.5
98.0
21.5
18.9
22.2
16.2
16.2
U.S. Territories
21.2
51.6
19.4
20.6
19.6
16.9
16.9
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
4,731.2
5,752.0
4,909.6
4,853.3
4,989.3
4,852.3
4,342.7
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 2019 to 2020 trends were particularly impacted by the COVID-19 pandemic which generally led to a
reduction in demand for fossil fuels.
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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 2020 CO2 Emissions from Fossil Fuel
Combustion for Selected Fuels and Sectors (MMT CO2 Eq. and Percent)
Sector
Fuel Type
2016 to 2017
2017 to 2018
2018 to 2019
2019
to 2020
Total 2020
Transportation
Petroleum
20.1
1.2%
24.1
1.4%
-7.0
-0.4%
-240.9
-13.7%
1,514.0
Electric Power
Coal
-34.9
-2.8%
-54.2
-4.5%
-179.3
-15.6%
-185.4
-19.0%
788.2
Electric Power
Natural Gas
-39.4
-7.2%
71.8
14.2%
38.6
6.7%
18.2
3.0%
634.3
Industrial
Natural Gas
5.6
1.2%
24.5
5.2%
7.7
1.6%
-16.1
-3.2%
485.5
Residential
Natural Gas
3.1
1.3%
32.3
13.4%
1.7
0.6%
-19.1
-6.9%
256.4
Commercial
Natural Gas
2.6
1.6%
19.3
11.2%
0.4
0.2%
-19.1
-9.9%
173.9
Transportation
All Fuels3
22.3
1.3%
32.8
1.8%
1.0
0.1%
-241.7
-13.3%
1,572.0
Electric Power
All Fuels3
-76.8
-4.2%
20.9
1.2%
-146.8
-8.4%
-167.1
-10.4%
1,439.0
Industrial
All Fuels3
-2.3
-0.3%
23.7
3.0%
2.0
0.2%
-49.8
-6.1%
766.3
Residential
All Fuels3
0.6
0.2%
44.8
15.3%
3.2
0.9%
-25.6
-7.5%
315.8
Commercial
All Fuels3
0.5
0.2%
13.8
6.0%
4.9
2.0%
-23.9
-9.5%
226.8
All Sectors3
All Fuels3
-56.3
-1.1%
136.0
2.8%
-137.0
-2.7%
-509.7
-10.5%
4,342.7
+ Does not exceed 0.05 percent.
a Includes sector and fuel combinations not shown in this table.
As shown in Table 3-6, recent trends in CO2 emissions from fossil fuel combustion show a 1.1 percent decrease
from 2016 to 2017, a 2.8 percent increase from 2017 to 2018, a 2.7 percent decrease from 2018 to 2019, and a
10.5 percent decrease from 2019 to 2020. These changes contributed to an overall 11.5 percent decrease in CO2
emissions from fossil fuel combustion from 2016 to 2020.
Recent trends in CO2 emissions from fossil fuel combustion are largely driven by the electric power sector, which
until recently 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 2021d). Total electric power generation from all fossil
and non-fossil sources decreased by 1.0 percent from 2016 to 2017, increased by 3.6 percent from 2017 to 2018,
decreased by 1.3 percent from 2018 to 2019 and decreased by 2.9 percent from 2019 to 2020. Carbon dioxide
emissions decreased from 2019 to 2020 by 10.4 percent due to decreasing electric power generation from
petroleum and coal outweighing increases in natural gas generation. Carbon dioxide emissions from coal
consumption for electric power generation decreased by 36.5 percent since 2016, which can be largely attributed
to a shift to the use of less-CC>2-intensive natural gas to generate electricity and a rapid increase in renewable
energy capacity additions in the electric power sector in recent years.
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 7.0 percent
and 1.9 percent from 2016 to 2020, respectively. This trend can be partially attributed to a 1.0 percent increase in
heating degree days from 2016 to 2020, 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 2020. Emissions from petroleum consumption for transportation have
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
-------
decreased by 11.9 percent since 2016 and are primarily attributed to a 8.5 percent decrease in VMT over the same
time period. This decrease in VMT was largely due to the impacts of the COVID-19 pandemic which limited travel in
2020. 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 2019 to 2020 trends were largely driven by the COVID-19 pandemic which reduced economic activity and
caused changes in energy demand and supply patterns across different sectors in 2020. Reduced economic and
manufacturing activity resulted in lower energy use in the commercial and industrial sectors. More people working
from home combined with warmer temperatures in 2020 compared to 2019 resulted in a mixed impact on energy
use in the residential sector. People staying home in response to the COVID-19 pandemic combined with increased
summer cooling demand resulted in an increase in residential sector electricity use while lowered residential space
heating demand resulted in reduced natural gas use in the residential sector. Overall consumption of electricity in
the United States decreased in 2020 and the trend of decreased coal use and increased use of natural gas and
renewables continued. Reduced travel caused by the COVID-19 pandemic resulted in decreased energy use in the
transportation sector in 2020 compared to 2019, including decreased road transportation but in particular
decreased aviation travel.
In the United States, 78.8 percent of the energy used in 2020 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 35 percent of total U.S. energy used in 2020. Natural gas
and coal followed in order of fossil fuel energy demand importance, accounting for approximately 34 percent and
10 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 2020
was supplied by nuclear electric power (9 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: 2020 U.S. Energy Use by Energy Source
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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
Nuclear Electric Power
8.9%
Renewable Energy
12.4%
Coal
9.9%
-------
Figure 3-5; Annual U.S. Energy Use
120
§, 100
o
Q.
E
80
60
40
20
Total Energy
Renewable & Nuclear
o h w n j m ic
ffi Oi
CTt CT> CTi CTi CJi (Ji
8 8
coaio->-irMco"3-Ln<£>rN.coo,>0'<-irsjro"5d-mvofs-.coa^o
O'lC^OOOOOOOOOOi-Hi—ItHi—ItHtHtHi—li-HT-HfN
o^o^ooooooooooooooooooooo
HHtNNC\fN(M(NMfMCMfNCM(NCMNtM(NM(N(N(MN
Figure 3-6: 2020 CO2 Emissions from Fossil Fuel Combustion by Sector and Fuel Type
2,500
2,000
iS" 1,500
O
u
1,000
500
Relative Contribution by Fuel Type
<0.05%
(Geothermal)
227
Coal
Natural Gas
I Geothermal
I Petroleum
1,439
1,572
23
U.S. Territories
Commercial
Residential
Industrial
Electric Power
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.
Box 3-2: Weather and Non-Fossil Energy Effects on CO2 Emissions from Fossil Fuel Combustion Trends
The United States in 2020 experienced a warmer winter overall compared to 2019, as heating degree days
decreased 9.4 percent. Warmer winter conditions compared to 2019 impacted the amount of energy required
for heating, In 2020 heating degree days in the United States were 9.8 percent below normal (see Figure 3-7).
Cooling degree days increased by 1.5 percent compared to 2019, which increased demand for air conditioning in
the residential and commercial sector. Hotter summer conditions compared to 2019 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 14.0 percent above normal (see
Figure 3-8) (EIA 2021c).10 The combination of warmer winter and hotter summer conditions led to overall
residential and commercial energy consumption decreases of 7.5 and 9.5 percent, respectively relative to 2019.
Figure 3-7: Annual Deviations from Normal Heating Degree Days for the United States
(1950-2020, Index Normal = 100)
Normal
30 (3,916 Heating Degree Days)
ridiukiu
! 11 'tffi
§ -10 ^0/,° Confidence
-20
Note: Climatological normal data are highlighted in dark red. Statistical confidence interval for "normal" climatology period of
-30 1991 through 2020.
ofN^rvoooofNTrvocoorvj'^rvoooorvj^rvoGoorM^rvocoorM'^i-voooofMTr^Dooo
LnLnLnmmvovovjDwDor^r^r^r>«r^.cocococococr1criCTia^CTiooooo-rH-^i-^'-(-rHrsj
(^c^o^cricriC^cr.aicricriO^aiO^c^c^c^c^aicr.o^cT.criO^o^a.00000000000
Figure 3-8: Annual Deviations from Normal Cooling Degree Days for the United States
(1950-2020, Index Normal = 100)
40
30
ts 20
ro
E
5
^ 10
E
p
Normal
(1,514 cooling degree days)
99% Confidence _ ¦
1 °
1 -10
C
QJ
1 "20
-30
-40
pilPF'l ^
Note: Climatological normal data are highlighted dark blue. Statistical confidence interval for "normal" climatology period of
1991 through 2020,
1950
1952
1954
1956
1958
1960
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 2020 remained high at 93 percent. In
2020, nuclear power represented 21 percent of total electricity generation. Since 1990, the wind and solar power
sectors have shown strong growth (between an observed minimum of 89 percent annual electricity generation
growth to a maximum of 162 percent annual electricity generation growth) and have become relatively important
electricity sources. Between 1990 and 2020, renewable energy generation (in kWh) from solar and wind energy
have increased from 0.1 percent in 1990 to 11 percent in 2020 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 CFU and N2O from Stationary
Combustion).
Table 3-7: CO2 Emissions from Stationary Fossil Fuel Combustion (MMT CO2 Eq.)
Sector/Fuel Type
1990
2005
2016
2017
2018
2019
2020
Electric Power
1,820.0
2,400.1
1,808.9
1,732.0
1,752.9
1,606.1
1,439.0
Coal
1,546.5
1,982.8
1,242.0
1,207.1
1,152.9
973.5
788.2
Natural Gas
175.4
318.9
545.0
505.6
577.4
616.0
634.3
Fuel Oil
97.5
98.0
21.5
18.9
22.2
16.2
16.2
Geothermal
0.5
0.5
0.4
0.4
0.4
0.4
0.4
Industrial
853.7
851.5
792.7
790.4
814.1
816.1
766.3
Coal
157.8
117.8
63.2
58.7
54.4
49.5
43.0
Natural Gas
408.8
388.6
463.9
469.5
494.0
501.6
485.5
Fuel Oil
287.1
345.0
265.7
262.2
265.7
265.0
237.8
Commercial
228.3
227.1
231.5
232.0
245.8
250.7
226.8
Coal
12.0
9.3
2.3
2.0
1.8
1.6
1.4
Natural Gas
142.0
162.9
170.5
173.2
192.5
192.9
173.9
Fuel Oil
74.3
54.9
58.7
56.8
51.5
56.2
51.6
Residential
338.6
358.9
292.8
293.4
338.2
341.4
315.8
Coal
3.0
0.8
NO
NO
NO
NO
NO
Natural Gas
237.8
262.2
238.4
241.5
273.8
275.5
256.4
Fuel Oil
97.8
95.9
54.4
51.9
64.4
65.9
59.5
U.S. Territories
21.7
55.9
26.0
25.5
25.5
24.3
22.7
Coal
0.5
3.0
3.3
2.3
2.6
3.6
3.1
Natural Gas
NO
1.3
3.4
2.5
3.3
3.8
2.6
Fuel Oil
21.2
51.6
19.4
20.6
19.6
16.9
16.9
Total
3,262.2
3,893.5
3,152.0
3,073.3
3,176.5
3,038.6
2,770.6
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)." Data for both the generation and net summer capacity are from EIA (2019).
Energy 3-13
-------
Table 3-8: ChU Emissions from Stationary Combustion (MMT CO2 Eq.)
Sector/Fuel Type
1990
2005
2016
2017
2018
2019
2020
Electric Power
0.4
0.9
1.2
1.1
1.2
1.3
1.2
Coal
0.3
0.4
0.2
0.2
0.2
0.2
0.2
Fuel Oil
+
+
+
+
+
+
+
Natural gas
0.1
0.5
0.9
0.9
1.0
1.1
1.1
Wood
+
+
+
+
+
+
+
Industrial
1.8
1.7
1.6
1.5
1.5
1.5
1.4
Coal
0.4
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
0.2
0.2
0.2
Wood
1.0
1.0
1.0
1.0
1.0
1.0
1.0
Commercial
1.1
1.1
1.2
1.2
1.2
1.2
1.2
Coal
+
+
+
+
+
+
+
Fuel Oil
0.3
0.2
0.2
0.2
0.2
0.2
0.2
Natural gas
0.3
0.4
0.4
0.4
0.4
0.4
0.4
Wood
0.5
0.5
0.6
0.6
0.6
0.6
0.6
Residential
5.2
4.1
3.9
3.8
4.6
4.7
4.1
Coal
0.2
0.1
0.0
0.0
0.0
0.0
0.0
Fuel Oil
0.3
0.3
0.2
0.2
0.2
0.2
0.2
Natural Gas
0.5
0.6
0.5
0.5
0.6
0.6
0.6
Wood
4.1
3.1
3.2
3.1
3.7
3.9
3.3
U.S. Territories
+
0.1
+
+
+
+
+
Coal
+
+
+
+
+
+
+
Fuel Oil
+
0.1
+
+
+
+
+
Natural Gas
NO
+
+
+
+
+
+
Wood
NE
NE
NE
NE
NE
NE
NE
Total
8.6
7.8
7.9
7.7
8.6
8.8
7.9
+ Does not exceed 0.05 MMT C02 Eq.
NO (Not Occurring)
NE (Not Estimated)
Note: Totals may not sum due to independent rounding.
Table 3-9: N2O Emissions from Stationary Combustion (MMT CO2 Eq.)
Sector/Fuel Type
1990
2005
2016
2017
2018
2019
2020
Electric Power
20.5
30.1
26.2
24.8
24.4
21.1
19.7
Coal
20.1
28.0
22.4
21.2
20.3
16.7
15.2
Fuel Oil
0.1
0.1
+
+
+
+
+
Natural Gas
0.3
1.9
3.8
3.6
4.1
4.4
4.5
Wood
+
+
+
+
+
+
+
Industrial
3.1
2.9
2.6
2.5
2.5
2.5
2.3
Coal
0.7
0.6
0.3
0.3
0.3
0.2
0.2
Fuel Oil
0.5
0.5
0.4
0.4
0.4
0.4
0.3
Natural Gas
0.2
0.2
0.2
0.3
0.3
0.3
0.3
Wood
1.6
1.6
1.7
1.6
1.6
1.6
1.5
Commercial
0.4
0.3
0.3
0.3
0.3
0.3
0.3
Coal
0.1
+
+
+
+
+
+
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
1.0
0.9
0.8
0.8
0.9
0.9
0.8
Coal
+
+
0.0
0.0
0.0
0.0
0.0
Fuel Oil
0.2
0.2
0.1
0.1
0.2
0.2
0.2
3-14 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Natural Gas
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Wood
0.7
0.5
0.5
0.5
0.6
0.6
0.5
U.S. Territories
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Coal
+
+
+
+
+
+
+
Fuel Oil
0.1
0.1
+
0.1
+
+
+
Natural Gas
NO
+
+
+
+
+
+
Wood
NE
NE
NE
NE
NE
NE
NE
Total
25.1
34.4
30.0
28.4
28.2
24.9
23.2
+ Does not exceed 0.05 MMT C02 Eq.
NO (Not Occurring)
NE (Not Estimated)
Note: Totals may not sum due to independent rounding.
Fossil Fuel Combustion Emissions by Sector
Table 3-10 provides an overview of the CO2, CH4, and N2O emissions from fossil fuel combustion by sector,
including transportation, electric power, industrial, residential, commercial, and U.S. territories.
Table 3-10: CO2, ChU, and N2O Emissions from Fossil Fuel Combustion by Sector (MMT CO2
Eq.)
End-Use Sector
1990
2005
2016
2017
2018
2019
2020
Transportation
1,520.0
1,903.9
1,781.3
1,802.6
1,834.4
1,836.2
1,591.6
C02
1,468.9
1,858.6
1,757.6
1,780.0
1,812.8
1,813.8
1,572.0
ch4
6.5
4.0
2.6
2.6
2.5
2.5
2.2
n2o
44.6
41.4
21.1
20.1
19.2
20.0
17.4
Electric Power
1,840.9
2,431.0
1,836.2
1,757.9
1,778.5
1,628.5
1,460.0
C02
1,820.0
2,400.1
1,808.9
1,732.0
1,752.9
1,606.1
1,439.0
ch4
0.4
0.9
1.2
1.1
1.2
1.3
1.2
n2o
20.5
30.1
26.2
24.8
24.4
21.1
19.7
Industrial
858.6
856.2
796.9
794.5
818.2
820.1
770.1
C02
853.7
851.5
792.7
790.4
814.1
816.1
766.3
ch4
1.8
1.7
1.6
1.5
1.5
1.5
1.4
n2o
3.1
2.9
2.6
2.5
2.5
2.5
2.3
Residential
344.9
363.8
297.4
298.0
343.7
347.1
320.7
C02
338.6
358.9
292.8
293.4
338.2
341.4
315.8
ch4
5.2
4.1
3.9
3.8
4.6
4.7
4.1
n2o
1.0
0.9
0.8
0.8
0.9
0.9
0.8
Commercial
229.8
228.6
233.1
233.5
247.4
252.3
228.3
C02
228.3
227.1
231.5
232.0
245.8
250.7
226.8
ch4
1.1
1.1
1.2
1.2
1.2
1.2
1.2
n2o
0.4
0.3
0.3
0.3
0.3
0.3
0.3
U.S. Territories3
21.8
56.1
26.1
25.6
25.6
24.4
22.7
Total
4,815.9
5,839.6
4,971.2
4,912.1
5,047.8
4,908.4
4,393.4
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.
Other than greenhouse gases CO2, Cm, and N2O, gases emitted from stationary combustion include the
greenhouse gas precursors nitrogen oxides (NOx), CO, NMVOCs, and SO2. Methane and N2O emissions from
stationary combustion sources depend upon fuel characteristics, size and vintage of combustion device, along with
combustion technology, pollution control equipment, ambient environmental conditions, and operation and
maintenance practices. Nitrous oxide emissions from stationary combustion are closely related to air-fuel mixes
and combustion temperatures, as well as the characteristics of any pollution control equipment that is employed.
Energy 3-15
-------
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 Cm, 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
2016
2017
2018
2019
2020
Transportation
1,523.1
1,908.6
1,785.5
1,806.9
1,839.0
1,840.9
1,596.3
C02
1,472.0
1,863.3
1,761.8
1,784.3
1,817.4
1,818.5
1,576.7
ch4
6.5
4.0
2.6
2.6
2.5
2.5
2.2
n2o
44.6
41.4
21.1
20.1
19.2
20.0
17.4
Industrial
1,552.9
1,602.0
1,322.4
1,306.4
1,326.7
1,291.9
1,185.6
C02
1,540.1
1,587.8
1,310.3
1,294.8
1,315.3
1,281.4
1,175.8
ch4
2.0
2.0
1.9
1.9
1.9
1.9
1.8
n2o
10.8
12.2
10.1
9.8
9.5
8.6
7.9
Residential
944.4
1,230.9
960.8
924.3
995.3
938.8
873.5
C02
931.3
1,214.9
946.2
910.5
980.4
925.0
860.6
ch4
5.4
4.4
4.3
4.2
5.0
5.2
4.5
n2o
7.7
11.6
10.3
9.6
9.9
8.6
8.3
Commercial
773.7
1,041.9
876.3
848.8
861.1
812.5
715.3
C02
766.0
1,030.1
865.2
838.2
850.7
803.2
706.8
ch4
1.2
1.4
1.6
1.6
1.6
1.7
1.6
n2o
6.5
10.4
9.5
9.0
00
00
7.6
6.9
U.S. Territories3
21.8
56.1
26.1
25.6
25.6
24.4
22.7
Total
4,815.9
5,839.6
4,971.2
4,912.1
5,047.8
4,908.4
4,393.4
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 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.
3-16 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Electric Power Sector
The process of generating electricity is the largest stationary source of CO2 emissions in the United States,
representing 30.5 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 1.4 percent, respectively. Electric power also accounted for 33.1 percent of CO2
emissions from fossil fuel combustion in 2020. Methane and N2O from electric power represented 12.3 and 48.6
percent of total CH4 and N2O emissions from fossil fuel combustion in 2020, 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 20.7 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 minor increase (0.3 percent).
From 2008 to 2020, as electricity demand decreased by 0.4 percent, electric power sector emissions decreased by
39 percent, driven by a significant drop (31 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 32 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 20
percent in 2020.15 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 31-year period to represent 39 percent of electric power sector
generation in 2020 (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
U.S. in 2020 had an average carbon content of 14.43 MMT C/QBtu and 26.12 MMT C/QBtu respectively.
Table 3-12: Electric Power Generation by Fuel Type (Percent)
Fuel Type
1990
2005
2016
2017
2018
2019
2020
Coal
54.1%
51.1%
31.4%
30.9%
28.4%
24.2%
19.9%
Natural Gas
10.7%
17.5%
32.7%
30.9%
34.0%
37.3%
39.5%
Nuclear
19.9%
20.0%
20.6%
20.8%
20.1%
20.4%
20.5%
Renewables
11.3%
8.3%
14.7%
16.8%
16.8%
17.6%
19.5%
Petroleum
4.1%
3.0%
0.6%
0.5%
0.6%
0.4%
0.4%
Other Gases3
+%
0.1%
0.1%
0.1%
0.1%
0.1%
0.1%
Net Electricity Generation
(Billion kWh)b
2,905
3,902
3,917
3,877
4,017
3,963
3,849
+ 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
-------
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 2020, CO2 emissions from the electric power sector decreased by 10.4 percent relative to 2019. This decrease in
CO2 emissions was primarily driven by a decrease in coal and petroleum consumed to produce electricity in the
electric power sector as well as a decrease in electricity demand (2.5 percent reduction in retail sales).
Consumption of coal for electric power decreased by 19.2 percent while consumption of natural gas increased 2.9
percent from 2019 to 2020. 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 8 percent from
2019 to 2020 (see Table 3-12). The 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 2020, 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 64 percent while the cost of
coal (in $/MMBtu) increased by 66 percent (EIA 2021c). Also, between 1990 and 2020, renewable energy
generation (in kWh) from wind and solar energy increased from 0.1 percent of total generation in 1990 to 11
percent in 2020, 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 37 percent, from 2,713 billion kWh in 1990
to 3,718 billion kWh in 2020.
Figure 3-9: Fuels Used in Electric Power Generation and Total Electric Power Sector CO2
Emissions
50,000
40,000
f? 30,000
QJ
00
=>
>¦
20,000
c
LU
10,000
0
I Nuclear (TBtu)
Renewable Energy Sources (TBtu)
I Petroleum (TBtu)
Matural Gas (TBtu)
Coal (TBtu)
I Net Generation (Index from 1990) [Right Axis]
I Sector CO2 Emissions (Index from 1990) [Right Axis]
160
140
120
100
80
60
40
20
0
o
>
X
01
-o
sz
o ih c\J cn
O"! (7* _
OOOOOOOOOOOOOOOOOOOOO
CNlCMrvJCMCMCMCMCMrslrvJrvJCMCMCMCMCNCNC'JCMCNCM
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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure 3-
10: Electric Power Retail Sales by End-Use Sector
1,600
1,500
1,400
^ 1,300
| 1,200
s 1,100
1,000
900
800
o-HrsjroTrLO'sOr-.coc^O'- CJ\ CTi CT> CTi O O O O O O O O O O -»—« -»—< -»—< i ' -¦—< « f"NJ
c^c^c^c^c^c^c^c^c^c^ooooooooooooooooooooo
^^^^.H^^i^^^fMrMiNrNtNrMrMrNrMrNiNrNJMMfNiNiNrNjfNrvifN
In 2020, electricity sales to the residential and commercial end-use sectors, as presented in Figure 3-10, increased
by 1.7 percent and decreased by 5.4 percent relative to 2019, respectively. Electricity sales to the industrial sector
in 2020 decreased by approximately 4.3 percent relative to 2019. The sections below describe end-use sector
energy use in more detail. Overall, in 2020, the amount of electricity retail sales (in kWh) decreased by 2.5 percent
relative to 2019. These electricity sales trends between 2019 and 2020 were likely impacted by the COVID-19
pandemic as people staying at home more increased electricity sales in the residential sector while decreasing
sales in other sectors.
Industrial Sector
Industrial sector CO2, CFU, and N2O emissions accounted for 18,14, and 6 percent of CO2, CFU, and N2O emissions
from fossil fuel combustion, respectively in 2020. 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 2019 to 2020, total industrial production and manufacturing output decreased by 7.2 percent (FRB 2021).
Over this period, output decreased slightly across all production indices including Food, Nonmetallic Mineral
Products, Paper, Petroleum Refineries, Chemicals, and Primary Metals (see Figure 3-11). From 2019 to 2020, total
energy use in the industrial sector decreased by 5.3 percent partially as a result of reductions in economic and
manufacturing activity due to the COVID-19 pandemic. Due to the relative increases and decreases of individual
indices there was a decrease in natural gas and a decrease 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
-------
2020, CCh, CH4, and N:0 emissions from fossil fuel combustion and electricity use within the industrial end-use
sector totaled 1,185.6 MMT CO2 Eq., an 8.2 percent decrease from 2019 emissions.
Through EPA's Greenhouse Gas Reporting Program (GHGRP), specific industrial sector trends can be discerned
from the overall total EIA industrial fuel consumption data used for these calculations. For example, from 2019 to
2020, the underlying EIA data showed decreased consumption of coal and natural gas in the industrial sector. The
GHGRP data highlights that several industries contributed to these trends, including chemical manufacturing; pulp,
paper and print; food processing, beverages and tobacco; minerals manufacturing; and agriculture-forest-
fisheries.17
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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure 3-12: Fuels and Electricity Used in Industrial Sector, Industrial Output, and Total
Sector CO2 Emissions (Including Electricity)
4—'
CO
35,000
30,000
25,000
oj 20,000
en
Z>
>-
CD
!—
o h cm m
O^O^iO^O^i
180
160
140
120
100
80
60
40
20
0
OiHrMco^-miorvoociOiHOJcnTt-inor^ooGto
OOOOOOOOOOi-Hi-li-li-li-li-li-li-li-li-lfM
000000000000000000000
CNrslCNCNCNrslCNCNCNlrslCNCNrslCNCNCNlCNCNCNCNCN
O
3>
-------
UNFCCC, progress was made on certain fuel types for specific industries and has been included in the CRF tables
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 2020 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)
180
160
140
120
O
Ch
CTH
100
ui
>
80 |
c
60
40
20
0
In 2020 the residential and commercial sectors accounted for 7 and 5 percent of CO2 emissions from fossil fuel
combustion, respectively; 40 and 12 percent of CFU 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
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)
I Sector CO2 Emissions (Index vs. 1990) [Right Axis]
I Heating and Cooling Degree Days (Index vs. 1990) [Right Axis]
O i—I Csl CO
Ol CTi
(Nfo^rmvDr^cocTio
00000
(NNfMCNfNJCMfMfMfNlfMCMCNfMCMNNNNfMfMN
19 See https://www.epa.gov/ghgemissions/inventorv-us-greenhouse-gas-emissions-and-sinks.
3-22 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
energy use in the residential sector. In 2020, total emissions (CO2, CH4, and N2O) from fossil fuel combustion and
electricity use within the residential and commercial end-use sectors were 873.5 MMT CO2 Eq. and 715.3 MMT CO2
Eq., respectively. Total CO2, CH4, and N2O emissions from combined fossil fuel combustion and electricity use
within the residential and commercial end-use sectors decreased by 7.0 and 12.0 percent from 2019 to 2020,
respectively. A decrease in heating degree days (9.4 percent) reduced energy demand for heating in the residential
and commercial sectors. This was partially offset by a 1.5 percent increase in cooling degree days compared to
2019, which impacted demand for air conditioning in the residential and commercial sectors. This, combined with
people staying home in response to the COVID-19 pandemic, resulted in a 1.7 percent increase in residential sector
electricity use. From 2019 to 2020 the COVID-19 pandemic reduced economic activity which contributed to 9.6
percent lower direct energy use in the commercial sector. In addition, a shift toward energy efficient products and
more stringent energy efficiency standards for household equipment has contributed to a decrease in energy
demand in households (EIA 2020g), resulting in a decrease in energy-related emissions. In the long term, the
residential sector is also affected by population growth, migration trends toward warmer areas, and changes in
total housing units and building attributes (e.g., larger sizes and improved insulation).
In 2020, combustion emissions from natural gas consumption represented 81 and 77 percent of the direct fossil
fuel CO2 emissions from the residential and commercial sectors, respectively. Carbon dioxide emissions from
natural gas combustion in the residential and commercial sectors in 2020 decreased by 6.9 percent and 9.9
percent from 2019 to 2020, respectively.
U.S. Territories
Emissions from U.S. Territories are based on the fuel consumption in American Samoa, Guam, Puerto Rico, U.S.
Virgin Islands, Wake Island, and other outlying U.S. Pacific Islands. As described in the Methodology section of CO2
from Fossil Fuel Combustion, this data is collected separately from the sectoral-level data available for the general
calculations. As sectoral information is not available for U.S. Territories, CO2, CFU, and N2O emissions are not
presented for U.S. Territories in the tables above by sector, though the emissions will occur across all sectors and
sources including stationary, transportation and mobile combustion sources. Due to data availability limitations,
2020 energy consumption for U.S. Territories for petroleum is proxied to 2019 consumption data.
Transportation Sector and Mobile Combustion
This discussion of transportation emissions follows the alternative method of presenting combustion emissions by
allocating emissions associated with electricity generation to the transportation end-use sector, as presented in
Table 3-11. Table 3-10 presents direct CO2, CFU, and N2O emissions from all transportation sources (i.e., excluding
emissions allocated to electricity consumption in the transportation end-use sector).
The transportation end-use sector and other mobile combustion accounted for 1,596.4 MMT CO2 Eq. in 2020,
which represented 35 percent of CO2 emissions, 22 percent of CFU emissions, and 43 percent of N2O emissions
from fossil fuel combustion, respectively.20 Fuel purchased in the United States for international aircraft and
marine travel accounted for an additional 70.3 MMT CO2 Eq. in 2020; these emissions are recorded as international
bunkers and are not included in U.S. totals according to UNFCCC reporting protocols.
Transportation End-Use Sector
From 1990 to 2019, transportation emissions from fossil fuel combustion increased by 20.9 percent, followed by a
decline of 13.3 percent from 2019 to 2020. Overall, from 1990 to 2020, transportation emissions from fossil fuel
combustion increased by 4.8 percent. The increase in transportation emissions from fossil fuel combustion from
1990 to 2019 was due, in large part, to increased demand for travel (see Figure 3-14). The number of vehicle miles
traveled (VMT) by light-duty motor vehicles (passenger cars and light-duty trucks) increased 47.5 percent from
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.
Energy 3-23
-------
1990 to 2019,21 as a result of a confluence of factors including population growth, economic growth, urban sprawl,
and periods of low fuel prices. The drop in transportation emissions from fossil fuel combustion from 2019 to 2020
was primarily the result of the COVID-19 pandemic and associated restrictions, such as people working from home
and traveling less. During this period, the number of vehicle miles traveled (VMT) by light-duty motor vehicles
(passenger cars and light-duty trucks) decreased by 12.2 percent.
Commercial aircraft emissions decreased by 32 percent between 2019 and 2020 and have decreased 35 percent
since 2007 (FAA 2022 and DOT 1991 through 2021).22 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 24 percent from 1990 to
2019, followed by a reduction of 13 percent from 2019 to 2020. Annex 3.2 presents the total emissions from all
transportation and mobile sources, including CO2, N2O, Cm, and HFCs.
Figure 3-14: Fuels Used in Transportation Sector, Onroad VMT, and Total Sector CO2
Emissions
40,000
35,000
30,000
25,000
3 20,000
>•
2s
Gl O^i 0*i
in ko n co
Ci Ol C*
Oi O"* &) &)
0000
in in fx go cr» o
— 000 —
200
180
160
140
120
100
80
60
40
20
rNmrrm^rvoQcrio
000000000000000000000
fMfNfNfNrslfNf-NlfMfNfNfMf-Nlf-NlfMfNfNrsJf-Nlf-NlfMfN
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).
21 VMT estimates are based on data from FHWA Highway Statistics Table VM-1 (FHWA 1996 through 2021). In 2011, FHWA
changed its methods for estimating VMT by vehicle class, which led to a shift in VMT and emissions among on-road vehicle
classes in the 2007 to 2020 time period. In absence of these method changes, light-duty VMT growth between 1990 and 2020
would likely have been higher.
22 Commercial aircraft consists of passenger aircraft, cargo, and other chartered flights.
3-24 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Transportation Fossil Fuel Combustion CO2 Emissions
Domestic transportation CO2 emissions increased by 7 percent (104.7 MMT CO2) between 1990 and 2020, an
annualized increase of 0.2 percent. This includes a 24 percent increase in CO2 emissions between 1990 and 2019,
followed by a 13 percent decline between 2019 and 2020. Among domestic transportation sources in 2020, 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 27 percent, commercial aircraft 6 percent, and other
sources 10 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.23 Ethanol consumption by the
transportation sector has increased from 0.7 billion gallons in 1990 to 11.7 billion gallons in 2020, while biodiesel
consumption has increased from 0.01 billion gallons in 2001 to 1.9 billion gallons in 2020. For additional
information, see Section 3.10 on biofuel consumption at the end of this chapter and Table A-76 in Annex 3.2.
Carbon dioxide emissions from passenger cars and light-duty trucks increased from 924.5 MMT CO2 in 1990 to
1052.1 CO2 in 2019, then dropped to 902.8 MMT CO2 in 2020, due to the COVID-19 pandemic and associated
restrictions. Overall, CO2 emissions from passenger cars and light-duty trucks decreased 2 percent (-21.7 MMT
CO2) from 1990 to 2020. The 14 percent (127.6 MMT CO2) increase in CO2 emissions from passenger cars and light-
duty trucks from 1990 to 2019 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 2020). Carbon dioxide emissions from passenger cars
and light-duty trucks peaked at 1,154.7 MMT CO2 in 2004, and since then have declined about 22 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 2013,24 then grew at a faster rate until 2016 (2.6 percent from 2014 to 2015, and 2.5 percent
from 2015 to 2016). Between 2016 and 2019, the rate of light-duty VMT growth slowed to less than one percent
each year. In 2020, light-duty VMT declined by 12.2 percent from 2019 to 2020 due to the COVID-19 pandemic and
associated restrictions. Average new vehicle fuel economy has increased almost every year since 2005, while the
light-duty truck share decreased to about 33 percent in 2009 and has since varied from year to year between 36
and 56 percent. Since 2014, the light-duty truck share has slowly increased and is about 56 percent of new vehicles
sales in model year 2020 (EPA 2021b). See Annex 3.2 for data by vehicle mode and information on VMT and the
share of new vehicles (in VMT).
23 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.
24 VMT estimates are based on data from FHWA Highway Statistics Table VM-1 (FHWA 1996 through 2021). In 2007 and 2008
light-duty VMT decreased 3.0 percent and 2.3 percent, respectively. Note that the decline in light-duty VMT from 2006 to 2007
is due at least in part to a change in FHWA's methods for estimating VMT. In 2011, FHWA changed its methods for estimating
VMT by vehicle class, which led to a shift in VMT and emissions among on-road vehicle classes in the 2007 to 2020 time period.
In absence of these method changes, light-duty VMT growth between 2006 and 2007 would likely have been higher.
Energy 3-25
-------
Medium- and heavy-duty truck CO2 emissions increased by 80 percent from 1990 to 2020. This increase was largely
due to a substantial growth in medium- and heavy-duty truck VMT, which increased by 107 percent between 1990
and 2020.25
Carbon dioxide emissions from the domestic operation of commercial aircraft increased by 22 percent (24.3 MMT
CO2) from 1990 to 2019, followed by a decline of 32 percent (42.9 MMT CO2) from 2019 to 2020. Across all
categories of aviation, excluding international bunkers, CO2 emissions decreased by 4 percent (7.8 MMT CO2)
between 1990 and 2019, followed by a sharper decline of 32 percent (57.3 MMT CO2) between 2019 and 2020.26
Emissions from military aircraft decreased 70 percent between 1990 and 2020. 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 2020.
Transportation sources also produce Cm and N2O; these emissions are included in Table 3-14and Table 3-15 and in
the Cm and N2O from Mobile Combustion section. Annex 3.2 presents total emissions from all transportation and
mobile sources, including CO2, CH4, N2O, and HFCs.
Figure 3-15: Sales-Weighted Fuel Economy of New Passenger Cars and Light-Duty Trucks,
1990-2020
30
28
Source: EPA (2021a).
rN m <3-
g g ®
rNrsJfNfNfNrsJrNrsJrsJrsJrsJrsJrsJrslfNrslrNrsI
25 While FHWA data shows consistent growth in medium- and heavy-duty truck VMT over the 1990 to 2020 time period, part of
the growth reflects a method change for estimating VMT starting in 2007. This change in methodology in FHWA's VM-1 table
resulted in large changes in VMT by vehicle class, thus leading to a shift in VMT and emissions among on-road vehicle classes in
the 2007 to 2020 time period. During the time period prior to the method change (1990 to 2006), VMT for medium- and heavy-
duty trucks increased by 51 percent.
26 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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure 3-16: Sales of New Passenger Cars and Light-Duty Trucks, 1990-2020
Source: EPA (2021b).
Table 3-13: CO2 Emissions from Fossil Fuel Combustion in Transportation End-Use Sector
(MMT COz Eq.)
Fuel/Vehicle Type
1990
2005
2016a
2017a
2018a
2019a
2020a
Gasolineb
958.9
1,150.1
1,084.4
1,081.8
1,097.0
1,086.5
936.9
Passenger Cars
604.3
637.1
737.5
737.4
748.7
742.1
599.9
Light-Duty Trucks
300.6
463.5
291.7
288.2
290.9
289.0
283.4
Medium- and Heavy-Duty
Trucks0
37.7
33.8
40.0
40.9
41.9
40.1
39.6
Buses
0.3
0.4
0.9
0.9
1.0
1.0
0.8
Motorcycles
1.7
1.6
3.8
3.7
3.8
3.6
3.2
Recreational Boatsd
14.3
13.7
10.6
10.6
10.7
10.7
9.9
Distillate Fuel Oil (Diesel)b
274.6
472.1
461.1
474.9
486.6
484.1
455.0
Passenger Cars
7.9
4.3
4.2
4.3
4.3
4.5
3.5
Light-Duty Trucks
11.5
26.1
14.0
14.0
14.1
14.7
13.9
Medium- and Heavy-Duty
Trucks0
190.5
364.2
367.9
379.6
388.5
389.5
372.9
Buses
8.0
10.7
16.6
17.8
19.0
19.0
15.5
Rail
35.5
46.1
36.1
37.4
38.5
36.0
31.0
Recreational Boatsd
2.7
2.9
2.7
2.8
2.8
2.9
2.6
Ships and Non-Recreational
Boatse
6.8
8.4
10.9
10.0
9.3
7.5
7.6
International Bunker FuelsS
11.7
9.5
8.7
9.0
10.0
10.1
7.8
Jet Fuel
222.3
249.5
240.1
249.4
253.1
258.5
160.4
Commercial Aircraft5
109.9
132.7
120.4
128.0
129.6
134.2
91.3
Military Aircraft
35.7
19.8
12.5
12.5
12.1
12.1
10.7
General Aviation Aircraft
38.5
36.8
33.0
31.2
30.6
31.4
18.6
International Bunker FuelsS
38.2
60.2
74.1
77.8
80.9
80.8
39.8
International Bunker Fuels
from Commercial Aviation
30.0
55.6
70.8
74.5
77.7
77.6
36.7
Aviation Gasoline
3.1
2.4
1.4
1.4
1.5
1.6
1.4
General Aviation Aircraft
3.1
2.4
1.4
1.4
1.5
1.6
1.4
Energy 3-27
-------
Residual Fuel Oil
76.3
62.9
46.8
49.9
45.4
39.7
29.6
Ships and Non-Recreational
Boatse
22.6
19.3
12.9
16.5
14.0
14.5
7.5
International Bunker Fuels1
53.7
43.6
33.8
33.4
31.4
25.2
22.1
Natural GasJ
36.0
33.1
40.1
42.3
50.9
58.9
58.1
Passenger Cars
+
+
+
+
+
+
+
Light-Duty Trucks
+
+
+
+
+
+
+
Medium- and Heavy-Duty
Trucks
+
+
+
+
+
+
+
Buses
+
0.6
0.8
0.9
0.9
1.0
0.9
Pipeline11
36.0
32.4
39.2
41.3
49.9
57.9
57.1
LPG J
1.4
1.8
0.5
0.4
0.4
0.4
0.4
Passenger Cars
+
+
+
+
+
+
+
Light-Duty Trucks
0.2
0.3
0.1
0.1
0.1
0.1
0.1
Medium- and Heavy-Duty
Trucks0
1.1
1.3
0.3
0.3
0.3
0.3
0.3
Buses
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Electricity1
3.0
4.7
4.2
4.3
4.7
4.7
4.7
Passenger Cars
+
+
0.6
0.8
1.2
1.4
1.6
Light-Duty Trucks
+
+
0.1
0.1
0.2
0.2
0.4
Buses
+
+
+
+
+
+
+
Rail
3.0
4.7
3.5
3.4
3.3
3.1
2.6
Totalk
1,575.6
1,976.6
1,878.5
1,904.5
1,939.6
1,934.6
1,646.3
Total (Including Bunkers)'
1,472.0
1,863.3
1,761.8
1,784.3
1,817.4
1,818.5
1,576.7
Biofuels-Ethanol'
4.1
21.6
76.9
77.7
78.6
78.7
68.1
Biofuels-Biodiesel'
+
0.9
19.6
18.7
17.9
17.1
17.7
+ Does not exceed 0.05 MMT C02 Eq.
a In 2011, FHWA changed its methods for estimating vehicle miles traveled (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 for the 1990 through 2010 Inventory and apply to the 2007 through
2020 time period. This resulted in large changes in VMT and fuel consumption data by vehicle class, thus leading to a shift
in emissions among on-road vehicle classes.
b Gasoline and diesel highway vehicle fuel consumption estimates are based on data from FHWA Highway Statistics Table
MF-27 and VM-1 (FHWA 1996 through 2021). Data from Table VM-1 is used to estimate the share of consumption
between each on-road vehicle class. These fuel consumption estimates are combined with estimates of fuel shares by
vehicle type from DOE's TEDB Annex Tables A.l through A.6 (DOE 1993 through 2021).
c Includes medium- and heavy-duty trucks over 8,500 lbs.
d 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 2020.
e Note that large year over year fluctuations in emission estimates partially reflect nature of data collection for these
sources.
f 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,
s Commercial aircraft, as modeled in FAA's Aviation Environmental Design Tool (AEDT), consists of passenger aircraft,
cargo, and other chartered flights.
h Pipelines reflect C02 emissions from natural gas-powered pipelines transporting natural gas.
' 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 2020 time period.
k Includes emissions from rail electricity.
3-28 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
1 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 2020 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.
Mobile Fossil Fuel Combustion CH4 and N2O Emissions
Mobile combustion includes emissions of Cm and N2O from all transportation sources identified in the U.S.
Inventory with the exception of pipelines and electric locomotives;27 mobile sources also include non-
transportation sources such as construction/mining equipment, agricultural equipment, vehicles used off-road,
and other sources (e.g., snowmobiles, lawnmowers, etc.).28 Annex 3.2 includes a summary of all emissions from
both transportation and mobile sources. Table 3-14 and Table 3-15 provide mobile fossil fuel Cm and N2O emission
estimates in MMT CO2 Eq.29
Mobile combustion was responsible for a small portion of national CH4 emissions (0.3 percent) and was the fifth
largest source of national N2O emissions (4.1 percent). From 1990 to 2020, mobile source CH4 emissions declined
by 66 percent, to 2.2 MMT CO2 Eq. (88 kt CH4), due largely to emissions control technologies employed in on-road
vehicles since the mid-1990s to reduce CO, NOx, NMVOC, and Cm emissions. Mobile source emissions of N2O
decreased by 61 percent from 1990 to 2020, to 17.4 MMT CO2 Eq. (58 kt N2O). Earlier generation emissions control
technologies initially resulted in higher N2O emissions, causing a 29 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 70 percent decrease in mobile source N2O emissions from 1997 to 2020 (Figure 3-17).
Overall, Cm 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.
27 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.
28 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.
29 See Annex 3.2 for a complete time series of emission estimates for 1990 through 2020.
Energy 3-29
-------
Figure 3-17: Mobile Source ChU and N2O Emissions
Table 3-14: ChU Emissions from Mobile Combustion (MMT CO2 Eq.)
Fuel Type/Vehicle Type3
1990
2005
2016
2017
2018
2019
2020
Gasoline On-Roadb
5.2
2.2
0.9
0.8
0.7
0.8
0.6
Passenger Cars
3.2
1.3
0.6
0.5
0.5
0.5
0.4
Light-Duty Trucks
1.7
0.8
0.2
0.2
0.2
0.2
0.2
Medium- and Heavy-Duty Trucks
and Buses
0.3
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
Medium- and Heavy-Duty Buses
+
+
+
+
+
+
+
Alternative Fuel On-Road
+
0.2
0.1
0.1
0.1
0.1
+
Non-Roade
1.3
1.6
1.5
1.6
1.6
1.5
1.4
Ships and Boats
0.4
0.4
0.4
0.4
0.4
0.4
0.4
Railc
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Aircraft
0.1
0.1
+
+
+
+
+
Agricultural Equipment
0.1
0.2
0.1
0.1
0.1
0.1
0.1
Construction/Mining Equipment6
0.1
0.2
0.2
0.2
0.2
0.2
0.2
Other'
0.5
0.7
0.7
0.7
0.7
0.7
0.7
Total
6.5
4.0
2.6
2.6
2.5
2.5
2.2
+ 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.
c Rail emissions do not include emissions from electric powered locomotives. Class II and Class III diesel consumption
data for 2014 to 2017 is estimated by applying the historical average fuel usage per carload factor to the annual
number of carloads.
d Includes equipment, such as tractors and combines, as well as fuel consumption from trucks that are used off-road in
agriculture.
e Includes equipment, such as cranes, dumpers, and excavators, as well as fuel consumption from trucks that are used
off-road in construction.
f "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.
3-30 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
g 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% reduction factor is used, based on transportation diesel use (EIA 2022).
Notes: In 2011, FHWA changed its methods for estimating vehicle miles traveled (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 for the 1990 through 2010 Inventory and apply to the 2007
through 2020 time period. This resulted in large changes in VMT and fuel consumption data by vehicle class, thus
leading to a shift in emissions among on-road vehicle classes. Totals may not sum due to independent rounding.
Table 3-15: N2O Emissions from Mobile Combustion (MMT CO2 Eq.)
Fuel Type/Vehicle Type3
1990
2005
2016
2017
2018
2019
2020
Gasoline On-Roadb
37.5
31.9
10.2
8.7
7.3
7.9
6.2
Passenger Cars
24.1
17.3
7.0
6.0
5.1
5.2
3.9
Light-Duty Trucks
12.8
13.6
2.7
2.3
1.9
2.4
2.1
Medium- and Heavy-Duty Trucks
and Buses
0.5
0.9
0.3
0.3
0.2
0.2
0.2
Motorcycles
+
+
0.1
0.1
0.1
0.1
0.1
Diesel On-Roadb
0.2
0.3
2.7
3.0
3.3
3.3
3.4
Passenger Cars
+
+
0.1
0.1
0.1
0.1
+
Light-Duty Trucks
+
+
0.1
0.1
0.1
0.1
0.1
Medium- and Heavy-Duty Trucks
0.2
0.3
2.3
2.5
2.8
2.8
3.0
Medium- and Heavy-Duty Buses
+
+
0.3
0.3
0.3
0.3
0.3
Alternative Fuel On-Road
+
+
0.2
0.2
0.2
0.2
0.2
Non-Roads
6.9
9.2
8.0
8.3
8.4
8.5
7.6
Ships and Boats
0.3
0.3
0.2
0.2
0.2
0.2
0.2
Railc
0.3
0.3
0.3
0.3
0.3
0.3
0.2
Aircraft
1.7
1.7
1.5
1.6
1.6
1.6
1.1
Agricultural Equipment
1.4
1.6
1.3
1.2
1.2
1.2
1.2
Construction/Mining Equipment6
1.3
2.1
1.6
1.8
1.8
1.9
1.8
Other'
2.0
3.1
3.1
3.2
3.2
3.3
3.1
Total
44.6
41.3
21.1
20.1
19.2
20.0
17.4
+ 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.
c Rail emissions do not include emissions from electric powered locomotives. Class II and Class III diesel consumption data
for 2014 through 2017 is estimated by applying the historical average fuel usage per carload factor to the annual number
of carloads.
d Includes equipment, such as tractors and combines, as well as fuel consumption from trucks that are used off-road in
agriculture.
e Includes equipment, such as cranes, dumpers, and excavators, as well as fuel consumption from trucks that are used off-
road in construction.
f "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.
s 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).
Notes: In 2011, FHWA changed its methods for estimating vehicle miles traveled (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 for the 1990 through 2010 Inventory and apply to the 2007
through 2020 time period. This resulted in large changes in VMT and fuel consumption data by vehicle class, thus leading
to a shift in emissions among on-road vehicle classes. Totals may not sum due to independent rounding.
Energy 3-31
-------
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.30 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).31
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.32
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).33
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
adjustments, additional data were collected from AISI (2004 through 2021), Coffeyville (2012), U.S. Census
Bureau (2001 through 2011), EIA (2022a, 2021b, 2021e), 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
30 The IPCC Tier 3B methodology is used for estimating emissions from commercial aircraft.
31 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 22.7 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.
32 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.
33 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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
through 2020), USGS (2014 through 2021a), USGS (1991 through 2015b), USGS (2021b), USGS (1991
through 2020).34
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.35 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.36
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 (2021e) 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 2020), Benson
(2002 through 2004), DOE (1993 through 2017), EIA (2007), EIA (1991 through 2020), EPA (2021b), and
FHWA (1996 through 2021).37
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
purposes was subtracted from total fuel consumption. Data on non-fuel consumption were provided by
EIA (2021b).
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
34 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.
35 Natural gas energy statistics from EIA (2021d) are already adjusted downward to account for biogas in natural gas.
36 These adjustments are explained in greater detail in Annex 2.1.
37 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
-------
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).38 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 2021) 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 2020); residual and distillate fuel
use for civilian marine bunkers was obtained from DOC (1991 through 2020) for 1990 through 2001 and
2007 through 2020, and DHS (2008) for 2003 through 2006.39 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 (2021b) and USAF (1998).40
• 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
38 See International Bunker Fuels section in this chapter for a more detailed discussion.
39 Data for 2002 were interpolated due to inconsistencies in reported fuel consumption data.
40 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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
percent gasoline, diesel, and other (e.g., CNG, LPG) fuel consumption are estimated using data from
DOE (1993 through 2021). 41'42
• For non-road vehicles, activity data were obtained from AAR (2008 through 2021), APTA (2007
through 2021), APTA (2006), BEA (1991 through 2015), Benson (2002 through 2004), DLA Energy
(2021), DOC (1991 through 2020), DOE (1993 through 2021), DOT (1991 through 2021), EIA (2009a),
EIA (2021d), EIA (2002), EIA (1991 through 2020), EPA (2021b),43 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 2020. Due to data availability and sources, some adjustments outlined in the methodology above are not
applied consistently across the full 1990 to 2020 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.
41 Data from FHWA's Table VM-1 is used to estimate the share of fuel consumption between each on-road vehicle class. These
fuel consumption estimates are combined with estimates of fuel shares by vehicle type from DOE's TEDB Annex Tables A.l
through A.6 (DOE 1993 through 2021). In 2011, FHWA changed its methods for estimating data in the VM-1 table. 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 for the 1990 through 2010 Inventory and apply to the time period from 2007
through 2020. This resulted in large changes in VMT and fuel consumption data by vehicle class, thus leading to a shift in
emissions among on-road vehicle classes.
42 Transportation sector natural gas and LPG consumption are based on data from EIA (2020g). In previous Inventory years,
data from DOE (1993 through 2021) 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.
43 In 2014, EPA incorporated the NONROAD2008 model into MOVES2014 (EPA 2019). In 2021, EPA updated the MOVESV model
to MOVES3 (EPA 2021b). The current Inventory uses the Nonroad component of MOVES2014b for years 1999 through 2020.
Energy 3-35
-------
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
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
2016
2017
2018
2019
2020
Residential3
57.4
56.8
55.2
55.1
55.3
55.2
55.1
Commercial3
59.7
57.8
56.7
56.6
56.0
56.1
56.2
Industrial3
64.5
64.6
61.1
60.8
60.5
60.3
59.8
Transportation3
71.1
71.5
71.1
71.2
71.0
70.9
70.9
Electric Powerb
87.3
85.8
76.8
77.3
75.5
72.9
70.5
U.S. Territories0
72.3
72.6
71.0
71.1
70.5
70.9
71.8
All Sectors0
73.1
73.6
69.2
69.1
68.3
67.3
66.3
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 2020, was approximately 16.5 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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure 3-18: U.S. Energy Consumption and Energy-Related CO2 Emissions Per Capita and
Per Dollar GDP
Carbon intensity estimates were developed using nuclear and renewable energy data from EIA (2021b), 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
-------
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.44 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.45
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).46 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 2020 were estimated to be between 4,255.1 and 4,532.1 MMT CO2 Eq. at a 95
percent confidence level. This indicates a range of 2 percent below to 4 percent above the 2020 emission estimate
of 4,342.7 MMTCO2 Eq.
44 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.
45 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.
46 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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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)
2020 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
835.6
807.2
913.7
-3%
9%
Residential
NO
NO
NO
NO
NO
Commercial
1.4
1.3
1.6
-5%
15%
Industrial
43.0
40.9
49.8
-5%
16%
Transportation
NO
NO
NO
NO
NO
Electric Power
788.2
757.8
863.5
-4%
10%
U.S. Territories
3.1
2.7
3.7
-12%
19%
Natural Gasb
1,610.7
1,590.7
1,684.1
-1%
5%
Residential
256.4
249.0
274.4
-3%
7%
Commercial
173.9
169.0
186.1
-3%
7%
Industrial
485.5
469.6
521.3
-3%
7%
Transportation
58.1
56.4
62.1
-3%
7%
Electric Power
634.3
615.9
666.8
-3%
5%
U.S. Territories
2.6
2.3
3.1
-12%
17%
Petroleumb
1,895.9
1,781.5
2,010.4
-6%
6%
Residential
59.5
56.1
62.8
-6%
6%
Commercial
51.6
48.8
54.3
-5%
5%
Industrial
237.8
187.0
288.5
-21%
21%
Transportation
1,514.0
1,417.3
1,611.5
-6%
6%
Electric Power
16.2
15.3
17.6
-5%
9%
U.S. Territories
16.9
15.7
18.8
-7%
11%
Total (excluding Geothermal)b
4,342.3
4,254.5
4,531.5
-2%
4%
Geothermal
0.4
NE
NE
NE
NE
Electric Power
0.4
NE
NE
NE
NE
Total (including Geothermal)b'c
4,342.7
4,255.1
4,532.1
-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.
QA/QC and Verification
In order to ensure the quality of the CO2 emission estimates from fossil fuel 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 Tier 2 procedures that were implemented involved
checks specifically focusing on the activity data and methodology used for estimating CO2 emissions from fossil fuel
combustion in the United States. Emission totals for the different sectors and fuels were compared and trends
were investigated to determine whether any corrective actions were needed. Minor corrective actions were taken.
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. The reference
Energy 3-39
-------
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. EIA revised sector allocations of propane for 2019 for petroleum consumption and the heat
content of petroleum consumption, which impacted LPG by sector in 2019. Approximate heat rates for
electricity and the heat content of electricity were revised for petroleum, total fossil fuels, and
noncombustible renewable energy, which impacted electric power energy consumption by sector.
Additionally, EIA has updated its data reported for biofuels including updating the methodology used for
calculating consumption of Other Renewable Diesel.
• EPA also revised industrial HGL C contents to only include industrial propane consumption (excluding
residential and commercial propane consumption) in the updated weighted factor calculation to align
with ElA's revised heat contents and HGL fuel type categorization (EIA 2022a; ICF 2020). A discussion of
the methodology used to develop the C content coefficients is presented in Annex 2.2. This resulted in an
average annual increase of 0.2 percent in the weighted industrial HGLC contents.
All of the revisions discussed above resulted in the following impacts on emissions over time for petroleum:
• Petroleum emissions decreased by an average annual amount of 0.2 MMT CO2 Eq. (less than 0.05 percent
of petroleum emissions) from 1990 to 1999, which is mainly due to decreased emissions in the industrial
sector as a result of the update in the weighted industrial HGLC contents.
• Similarly, petroleum emissions decreased by an average annual amount of 0.3 MMT CO2 Eq. (less than
0.05 percent) from 2000 to 2007.
• Petroleum emissions decreased again by an average of annual amount of 1.8 MMT CO2 Eq. at the end of
the time-series from 2008 to 2019. In 2019, petroleum emissions by the residential sector increased by
4.4 MMT CO2 Eq. relative to the previous Inventory. Petroleum emissions by the industrial and
transportation sector decreased by 4.7 and 7.7 MMT CO2 Eq respectively. This change in 2019 is due to
ElA's revised sector allocations for propane and updates to biofuels data accounting.
• Across the time series, petroleum emissions from the transportation sector decreased by an average
annual amount of 0.8 MMT CO2 Eq. This decrease is due to updates to biofuels data by EIA.
Overall, these changes resulted in an average annual decrease of 0.7 MMT CO2 Eq. (less than 0.05 percent) in CO2
emissions from fossil fuel combustion for the period 1990 through 2019, relative to the previous Inventory.
However, there were bigger absolute changes across the time series as discussed above.
3-40 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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.47 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.48
An ongoing planned improvement is to develop improved estimates of domestic waterborne fuel consumption.
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. It may be possible to more accurately estimate
domestic fuel use and emissions by using detailed data on marine ship activity. The feasibility of using domestic
marine activity data to improve the estimates will continue to be investigated.
EPA is also evaluating the methods used to adjust for conversion of fuels and exports of CO2. EPA is exploring the
approach used to account for CO2 transport, injection, and geologic storage, as part of this there may be changes
made to accounting for CO2 exports.
Finally, another ongoing planned improvement is to evaluate data availability to update the carbon and heat
content of more fuel types accounted for in this Inventory. This update will impact consumption and emissions
across all sectors and will improve consistency with EIA data as carbon and heat contents of fuels will be accounted
for as annually variable and therefore improve accuracy across the time series. Some of the fuels considered in this
effort include petroleum coke, residual fuel, and woody biomass.
47 See https://unfccc.int/resource/docs/2013/copl9/ene/10a03.pdf#paee=2.
48 See http://www.ipcc-nggip.iges.or.lp/public/tb/TFI Technical Bulletin l.pdf.
Energy 3-41
-------
CH4and N20 from Stationary Combustion
Methodology and Time-Series Consistency
Methane and N2O emissions from stationary combustion were estimated by multiplying fossil fuel and wood
consumption data by emission factors (by sector and fuel type for industrial, residential, commercial, and U.S.
Territories; and by fuel and technology type for the electric power sector). The electric power sector utilizes a Tier
2 methodology, whereas all other sectors utilize a Tier 1 methodology. The activity data and emission factors used
are described in the following subsections.
More detailed information on the methodology for calculating emissions from stationary combustion, including
emission factors and activity data, is provided in Annex 3.1.
Industrial, Residential, Commercial, and U.S. Territories
National coal, natural gas, fuel oil, and wood consumption data were grouped by sector: industrial, commercial,
residential, and U.S. Territories. For the CFU and N2O emission estimates, consumption data for each fuel were
obtained from ElA's Monthly Energy Review (EIA 2022a). Because the United States does not include territories in
its national energy statistics, fuel consumption data for territories were provided separately by ElA's International
Energy Statistics (EIA 2022b).49 Fuel consumption for the industrial sector was adjusted to subtract out mobile
source construction and agricultural use, which is reported under mobile sources. Construction and agricultural
mobile source fuel use was obtained from EPA (2021) and FHWA (1996 through 2021). Estimates for wood biomass
consumption for fuel combustion do not include municipal solid waste, tires, etc., that are reported as biomass by
EIA. Non-CC>2 emissions from combustion of the biogenic portion of municipal solid waste and tires is included
under waste incineration (Section 3.2). Estimates for natural gas combustion do not include biogas, and therefore
non-CC>2 emissions from biogas are not included (see the Planned Improvements section, below). Tier 1 default
emission factors for the industrial, commercial, and residential end-use sectors were provided by the 2006IPCC
Guidelines for National Greenhouse Gas Inventories (IPCC 2006). U.S. Territories' emission factors were estimated
using the U.S. emission factors for the primary sector in which each fuel was combusted.
Electric Power Sector
The electric power sector uses a Tier 2 emission estimation methodology as fuel consumption for the electric
power sector by control-technology type was based on EPA's Acid Rain Program Dataset (EPA 2022). Total fuel
consumption in the electric power sector from EIA (2022a) was apportioned to each combustion technology type
and fuel combination using a ratio of fuel consumption by technology type derived from EPA (2022) data. The
combustion technology and fuel use data by facility obtained from EPA (2022) were only available from 1996 to
2020, so the consumption estimates from 1990 to 1995 were estimated by applying the 1996 consumption ratio by
combustion technology type from EPA (2022) to the total EIA (2022a) consumption for each year from 1990 to
1995.
Emissions were estimated by multiplying fossil fuel and wood consumption by technology-, fuel-, and country-
specific Tier 2 emission factors. The Tier 2 emission factors used are based in part on emission factors published by
49 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.
3-42 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
EPA, and EPA's Compilation of Air Pollutant Emission Factors, AP-42 (EPA 1997) for coal wall-fired boilers, residual
fuel oil, diesel oil and wood boilers, natural gas-fired turbines, and combined cycle natural gas units.50
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2020 as discussed below. As discussed in Annex 5, data are unavailable to include estimates of Cm and
N2O emissions from biomass use in Territories, but those emissions are assumed to be insignificant.
Uncertainty
Methane emission estimates from stationary sources exhibit high uncertainty, primarily due to difficulties in
calculating emissions from wood combustion (i.e., fireplaces and wood stoves). The estimates of CH4 and N2O
emissions presented are based on broad indicators of emissions (i.e., fuel use multiplied by an aggregate emission
factor for different sectors), rather than specific emission processes (i.e., by combustion technology and type of
emission control).
An 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.
The uncertainty estimation model for this source category was developed by integrating the CFU and N2O
stationary source inventory estimation models with the model for CO2 from fossil fuel combustion to realistically
characterize the interaction (or endogenous correlation) between the variables of these three models. About 55
input variables were simulated for the uncertainty analysis of this source category (about 20 from the CO2
emissions from fossil fuel combustion inventory estimation model and about 35 from the stationary source
inventory models).
In developing the uncertainty estimation model, uniform distribution was assumed for all activity-related input
variables and N2O emission factors, based on the SAIC/EIA (2001) report.51 For these variables, the uncertainty
ranges were assigned to the input variables based on the data reported in SAIC/EIA (2001).52 However, the CFU
emission factors differ from those used by EIA. These factors and uncertainty ranges are based on IPCC default
uncertainty estimates (IPCC 2006).
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3-18. Stationary
combustion CFU emissions in 2020 (including biomass) were estimated to be between 5.3 and 17.8 MMT CO2 Eq. at
a 95 percent confidence level. This indicates a range of 34 percent below to 125 percent above the 2020 emission
estimate of 7.9 MMT CO2 Eq.53 Stationary combustion N2O emissions in 2020 (including biomass) were estimated
to be between 17.6 and 35.0 MMT CO2 Eq. at a 95 percent confidence level. This indicates a range of 24 percent
below to 51 percent above the 2020 emission estimate of 23.2 MMT CO2 Eq.
50 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.
51 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.
52 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.
53 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 3-43
-------
Table 3-18: Approach 2 Quantitative Uncertainty Estimates for ChU and N2O Emissions from
Energy-Related Stationary Combustion, Including Biomass (MMT CO2 Eq. and Percent)
2020 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
Source
Gas
(MMT CO? Eq.)
(MMT CO?
Eq.)
(%)
Lower
Upper
Lower Upper
Bound
Bound
Bound Bound
Stationary Combustion
ch4
7.9
5.3
17.8
-34% +125%
Stationary Combustion
n2o
23.2
17.6
35.0
-24% +51%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
The uncertainties associated with the emission estimates of Cm and N2O are greater than those associated with
estimates of CO2 from fossil fuel combustion, which mainly rely on the carbon content of the fuel combusted.
Uncertainties in both Cm and N2O estimates are due to the fact that emissions are estimated based on emission
factors representing only a limited subset of combustion conditions. For the indirect greenhouse gases,
uncertainties are partly due to assumptions concerning combustion technology types, age of equipment, emission
factors used, and activity data projections.
QA/QC and Verification
In order to ensure the quality of the non-CC>2 emission estimates from stationary 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 Tier 2 procedures that were implemented involved
checks specifically focusing on the activity data and emission factor sources and methodology used for estimating
Cm, N2O, and the greenhouse gas precursors from stationary combustion in the United States. Emission totals for
the different sectors and fuels were compared and trends were investigated.
Recalculations Discussion
Methane and N2O emissions from stationary sources (excluding CO2) across the entire time series were revised due
to revised data from EIA (2022a) relative to the previous Inventory. EIA (2022a) revised approximate heat rates for
electricity and the heat content of electricity for petroleum and noncombustible renewable energy, which
impacted electric power energy consumption by sector.
EIA also revised sector allocations for propane for 2019, which impacted LPG by sector. The historical data changes
resulted in an average annual increase of 0.01 MMT CO2 Eq. (0.1 percent) in CH4 emissions, and an average annual
change of less than 0.05 MMT CO2 Eq. (less than 0.05 percent) in N2O emissions for the 1990 through 2019 period.
Planned Improvements
Several items are being evaluated to improve the CH4 and N2O emission estimates from stationary combustion and
to reduce uncertainty for U.S. Territories. Efforts will be taken to work with EIA and other agencies to improve the
quality of the U.S. Territories data. Because these data are not broken out by stationary and mobile uses, further
research will be aimed at trying to allocate consumption appropriately. In addition, the uncertainty of biomass
emissions will be further investigated because it was expected that the exclusion of biomass from the estimates
would reduce the uncertainty; and in actuality the exclusion of biomass increases the uncertainty. These
improvements are not all-inclusive but are part of an ongoing analysis and efforts to continually improve these
stationary combustion estimates from U.S. Territories.
3-44 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Other forms of biomass-based gas consumption include biogas. EPA will examine EIA and GHGRP data on biogas
collected and burned for energy use and determine if Cm and N2O emissions from biogas 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.
CH4 and N20 from Mobile Combustion
Methodology and Time-Series Consistency
Estimates of Cm and N2O emissions from mobile combustion were calculated by multiplying emission factors by
measures of activity for each fuel and vehicle type (e.g., light-duty gasoline trucks). Activity data included vehicle
miles traveled (VMT) for on-road vehicles and fuel consumption for non-road mobile sources. The activity data and
emission factors used in the calculations are described in the subsections that follow. A complete discussion of the
methodology used to estimate CH4 and N2O emissions from mobile combustion and the emission factors used in the
calculations is provided in Annex 3.2.
On-Road Vehicles
Estimates of Cm and N2O emissions from gasoline and diesel on-road vehicles are based on VMT and emission
factors (in grams of Cm and N2O per mile) by vehicle type, fuel type, model year, and emission control technology.
Emission estimates for alternative fuel vehicles (AFVs) are based on VMT and emission factors (in grams of CH4 and
N2O per mile) by vehicle and fuel type.54
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 gram per mile emissions of CO2, CO, HC, NOx,
and PM from vehicles under various conditions.55
Diesel on-road vehicle emission factors were developed by ICF (2006a). CH4 and N2O emissions factors for newer
(starting at 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 2021). For light-duty trucks, EPA used a curve fit
of 1999 through 2011 travel fractions for LDT1 and LDT2 (MOVES Source Type 31 for LDT1 and MOVES Source Type
32 for LDT2). For medium-duty vehicles, EPA used emission factors for light heavy-duty vocational trucks. For
54 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.
55 Additional information regarding the MOBILE model can be found online at https://www.epa.gov/moves/description-and-
history-mobile-highway-vehicle-emission-factor-model.
Energy 3-45
-------
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 Cm emission factors were corrected from GREET
values to be the same as Cm emission factors for diesel vehicles. GREET overestimated biodiesel Cm 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 2021).56
VMT estimates were then allocated from FHWA's vehicle categories to fuel-specific vehicle categories using the
calculated shares of vehicle fuel use for each vehicle category by fuel type reported in DOE (1993 through 2021)
and information on total motor vehicle fuel consumption by fuel type from FHWA (1996 through 2021). VMT for
AFVs were estimated based on Browning (2017 and 2018a). 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).
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 non-road mobile category for CH4 and N2O includes ships and boats, aircraft, locomotives, and off-road
sources (e.g., construction or agricultural equipment). For non-road sources, fuel-based emission factors are
applied to data on fuel consumption, following the IPCC Tier 1 approach, for locomotives, aircraft, ships and boats.
The Tier 2 approach would require separate fuel-based emissions factors by technology for which data are not
available. For some of the non-road categories, 2-stroke and 4-stroke technologies are broken out and have
separate emission factors; those cases could be considered a Tier 2 approach.
To estimate CH4 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 CH4 per kilogram of fuel
consumed).57 Activity data were obtained from AAR (2008 through 2021), APTA (2007 through 2021), Raillnc (2014
through 2021), APTA (2006), BEA (1991 through 2015), Benson (2002 through 2004), DLA Energy (2021), DOC
(1991 through 2020), DOE (1993 through 2021), DOT (1991 through 2021), EIA (2002, 2007, 2022), EIA (2021f), EIA
(1991 through 2020), EPA (2021b), Esser (2003 through 2004), FAA (2022), FHWA (1996 through 2021),58 Gaffney
56 The source of VMT data is FHWA Highway Statistics Table VM-1. In 2011, FHWA changed its methods for estimating data in
the VM-1 table. 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 for the 1990 through 2010 Inventory and apply to
the 2007 through 2020 time period. This resulted in large changes in VMT by vehicle class, thus leading to a shift in emissions
among on-road vehicle classes. For example, the category "Passenger Cars" has been replaced by "Light-duty Vehicles-Short
Wheelbase" and "Other 2 axle-4 Tire Vehicles" has been replaced by "Light-duty Vehicles, Long Wheelbase." This change in
vehicle classification has moved some smaller trucks and sport utility vehicles from the light truck category to the passenger
vehicle category in the current Inventory. These changes are reflected in a large drop in light-truck emissions between 2006
and 2007.
57 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.
58 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.
3-46 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
(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 2020 estimates of Cm 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 3.9 - Uncertainty Analysis of Emission Estimates. However, a much higher level of uncertainty is associated
with Cm and N2O emission factors due to limited emission test data, and because, unlike CO2 emissions, the
emission pathways of CH4 and N2O are highly complex.
Based on the uncertainty analysis, mobile combustion CH4 emissions from all mobile sources in 2020 were
estimated to be between 2.0 and 2.7 MMT CO2 Eq. at a 95 percent confidence level. This indicates a range of 8
percent below to 24 percent above the corresponding 2020 emission estimate of 2.2 MMT CO2 Eq. Mobile
combustion N2O emissions from mobile sources in 2020 were estimated to be between 16.0 and 20.7 MMT CO2
Eq. at a 95 percent confidence level. This indicates a range of 8 percent below to 19 percent above the
corresponding 2020 emission estimate of 17.4 MMT CO2 Eq.
Table 3-19: Approach 2 Quantitative Uncertainty Estimates for ChU and N2O Emissions from
Mobile Sources (MMT CO2 Eq. and Percent)
Source
Gas
2020 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
2.2
2.0
2.7
-8% +24%
Mobile Sources
n2o
17.4
16.0
20.7
-8% +19%
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 CH4 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
Energy 3-47
-------
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
Updates were made to Cm and N2O emission factors for newer non-road gasoline and diesel vehicles. Previously,
these emission factors were calculated using the updated 2006 IPCC Tier 3 guidance and the Nonroad component
EPA's MOVES2014b model. Updated factors are calculated using the Nonroad component of MOVES3 model. CH4
emission factors were calculated directly from MOVES3. N2O emission factors were calculated using MOVES-
Nonroad activity and emission factors in g/kWh by fuel type from the European Environment Agency. Updated
emission factors were developed using EPA engine certification data for non-road small and large spark-ignition
(SI) gasoline engines and compression-ignition diesel engines (model year 2011 and newer), as well as non-road
motorcycles (model year 2006 and newer), SI marine engines (model year 2011 and newer), and diesel marine
engines (model year 2000 and newer). Further refinements were made to the calculation of CH4 and N2O emission
factors for non-road equipment. In previous Inventories, average emission factors by non-road equipment type
and fuel type were applied to average engine power values. In the refined method, emission factors developed
from certification data were binned by engine power, and emissions were calculated for each horsepower bin,
non-road equipment type, and fuel type combination. These were then combined to determine emission factors
for a given non-road equipment and fuel type.
The collective result of these changes was a net increase in CFU emissions and a decrease in N2O emissions from
mobile combustion relative to the previous Inventory. Methane emissions increased by 11.9 percent and N2O
emissions decreased by 1.3 percent. Furthermore, the adoption of the MOVES3 model for this update does not
impact estimates of CO2 emissions from transportation and non-transportation mobile sources.
Previously, heavy-duty diesel buses were grouped with heavy-duty diesel trucks under the heavy-duty diesel
vehicle category. The updated approach calculates emissions from heavy-duty buses as a separate category. New
emission factors specific to buses have been developed from EPA certification data.
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2020 with one recent notable exception. An update by FHWA to the method for estimating on-road VMT
created an inconsistency in on-road CFU and N2O for the time periods 1990 to 2006 and 2007 to 2020. Details on
the emission trends and methodological inconsistencies through time are described in the Methodology section
above.
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.
3-48 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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),59
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.
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 3-20, fossil fuel emissions in 2020 from the non-energy uses of fossil fuels were 121.0 MMT CO2
Eq., which constituted approximately 2.6 percent of overall fossil fuel emissions. In 2020, the consumption of fuels
for non-energy uses (after the adjustments described above) was 5,570.6 TBtu (see Table 3-21). A portion of the C
in the 5,570.6 TBtu of fuels was stored (229.6 MMT CO2 Eq.), while the remaining portion was emitted (121.0 MMT
CO2 Eq.). Non-energy use emissions decreased by 4.6 percent from 2019 to 2020, mainly due to a decrease in
industrial fuel use (specifically in the coking coal industry) potentially caused by the COVID-19 pandemic. See
Annex 2.3 for more details.
59 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. Adjustments were made in the 1990 to 2019 Inventory report to HGL activity data, carbon content coefficients,
and heat contents HGL.
Energy 3-49
-------
Table 3-20: CO2 Emissions from Non-Energy Use Fossil Fuel Consumption (MMT CO2 Eq. and
Percent)
Year
1990
2005
2016
2017
2018
2019
2020
Potential Emissions
305.6
366.8
317.7
332.0
352.2
355.5
350.5
C Stored
193.4
237.9
218.2
219.4
223.4
228.8
229.6
Emissions as a % of Potential
37%
35%
31%
34%
37%
36%
35%
C Emitted
112.2
128.9
99.5
112.6
128.9
126.8
121.0
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 (2021b) (see Annex 2.1). Consumption values for industrial coking coal, petroleum coke,
other oils, and natural gas in Table 3-21 and Table 3-22 have been adjusted to subtract non-energy uses that are
included in the source categories of the Industrial Processes and Product Use chapter.60 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).
For the remaining non-energy uses, the quantity of C stored was estimated by multiplying the potential emissions
by a storage factor.
• For several fuel types—petrochemical feedstocks (including natural gas for non-fertilizer uses, HGL,
naphthas, other oils, still gas, special naphtha, and industrial other coal), asphalt and road oil, lubricants,
and waxes—U.S. data on C stocks and flows were used to develop C storage factors, calculated as the
ratio of (a) the C stored by the fuel's non-energy products to (b) the total C content of the fuel consumed.
A lifecycle approach was used in the development of these factors in order to account for losses in the
production process and during use. Because losses associated with municipal solid waste management
are handled separately in the Energy sector under the Incineration of Waste source category, the storage
factors do not account for losses at the disposal end of the life cycle.
• For industrial coking coal and distillate fuel oil, storage factors were taken from Marland and Rotty (1984).
• For the remaining fuel types (petroleum coke, miscellaneous products and other petroleum), IPCC (2006)
does not provide guidance on storage factors, and assumptions were made based on the potential fate of
C in the respective non-energy use products. Carbon dioxide emissions from carbide production are
implicitly accounted for in the storage factor calculation for the non-energy use of petroleum coke.
Table 3-21: Adjusted Consumption of Fossil Fuels for Non-Energy Uses (TBtu)
Year 1990 2005 2016 2017 2018 2019 2020
Industry 4,317.5 5,115.0 4,833.0 5,089.5 5,447.7 5,484.1 5,447.7
Industrial Coking Coal NO 80.4 89.6 113.0 124.8 113.4 78.8
60 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.
3-50 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
496.4
588.0
676.4
667.6
663.0
Asphalt & Road Oil
1,170.2
1,323.2
853.4
849.2
792.8
843.9
832.3
HGLa
1,217.7
1,609.9
2,127.9
2,193.3
2,506.5
2,550.3
2,656.5
Lubricants
186.3
160.2
135.1
124.9
122.0
118.3
107.4
Natural Gasolineb
117.5
95.4
53.1
81.7
105.3
155.0
163.6
Naphtha (<401 °F)
327.0
679.5
398.2
413.0
421.2
369.5
329.3
Other Oil (>401 °F)
663.6
499.5
204.6
242.9
219.1
212.1
195.5
Still Gas
36.7
67.7
166.1
163.8
166.9
158.7
145.4
Petroleum Coke
29.1
106.2
NO
NO
NO
NO
NO
Special Naphtha
101.1
60.9
89.0
95.3
87.0
89.5
80.7
Distillate Fuel Oil
7.0
16.0
5.8
5.8
5.8
5.8
5.8
Waxes
33.3
31.4
12.8
10.2
12.4
10.4
9.2
Miscellaneous Products
137.8
112.8
191.3
198.8
198.0
180.2
170.7
Transportation
176.0
151.3
154.4
142.0
137.0
131.3
119.3
Lubricants
176.0
151.3
154.4
142.0
137.0
131.3
119.3
U.S. Territories
50.8
114.9
10.5
3.5
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
9.5
2.4
2.5
2.6
2.6
Total
4,544.4
5,381.2
4,997.9
5,234.9
5,588.3
5,619.1
5,570.6
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.
Table 3-22: 2020 Adjusted Non-Energy Use Fossil Fuel Consumption, Storage, and Emissions
Sector/Fuel Type
Adjusted
Non-Energy Carbon Content Potential
Usea Coefficient Carbon
(TBtu) (MMTC/QBtu) (MMTC)
Storage Carbon Carbon Carbon
Factor Stored Emissions Emissions
(MMT C) (MMT C) (MMT COz Eq.)
Industry 5,447.7 NA 93.1 NA 62.4 30.7 112.7
Industrial Coking Coal 78.8 25.60 2.0 0.10 0.2 1.8 6.7
Industrial Other Coal 9.5 26.13 0.2 0.63 0.2 0.1 0.3
Natural Gas to
Chemical Plants 663.0 14.47 9.6 0.63 6.0 3.6 13.1
Asphalt & Road Oil 832.3 20.55 17.1 1.00 17.0 0.1 0.3
HGLb 2,656.5 16.77 44.5 0.63 27.9 16.7 61.1
Lubricants 107.4 20.20 2.2 0.09 0.2 2.0 7.2
Natural Gasolinec 163.6 18.24 3.0 0.63 1.9 1.1 4.1
Naphtha (<401° F) 329.3 18.55 6.1 0.63 3.8 2.3 8.4
Other Oil (>401° F) 195.5 20.17 3.9 0.63 2.5 1.5 5.4
Still Gas 145.4 17.51 2.5 0.63 1.6 1.0 3.5
Petroleum Coke NO 27.85 NO 0.30 NO NO NO
Special Naphtha 80.7 19.74 1.6 0.63 1.0 0.6 2.2
Distillate Fuel Oil 5.8 20.22 0.1 0.50 0.1 0.1 0.2
Waxes 9.2 19.80 0.2 0.58 0.1 0.1 0.3
Miscellaneous
Products 170.7 NO NO NO NO NO NO
Transportation 119.3 NA 2.4 NA 0.2 2.2 8.0
Lubricants 119.3 20.20 2.4 0.09 0.2 2.2 8.0
Energy 3-51
-------
U.S. Territories 3.6 NA 0.1 NA 0.0 0.1 0.2
Lubricants 1.0 20.20 + 0.09 + + 0.1
Other Petroleum
(Misc. Prod.) Z6 20.00 + 010 + + 0.2
Total 5,570.6 95^ 6^6 33^ 121.0
+ 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 3-20). 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 2021b),
Toxics Release Inventory, 1998 (EPA 2000b), Biennial Reporting System (EPA 2000a, 2009), Resource Conservation
and Recovery Act Information System (EPA 2013b, 2015, 2016b, 2018b, 2021a), 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, 2021a); the National
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); Financial Planning Association (2006); INEGI
(2006); the United States International Trade Commission (2021); Gosselin, Smith, and Hodge (1984); EPA's
Municipal Solid Waste (MSW) Facts and Figures (EPA 2013, 2014a, 2016a, 2018a, 2019); the Rubber
Manufacturers' Association (RMA2009, 2011, 2014, 2016, 2018); 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, 2021a); and the Guide to the Business of Chemistry (ACC 2021b). 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 2020 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.61 In this Inventory, C storage and C emissions from
61 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).
3-52 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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).62
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 3-22).
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.63 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
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.
Uncertainty
An uncertainty analysis was conducted to quantify the uncertainty surrounding the estimates of emissions and
storage factors from non-energy uses. This analysis, performed using @RISK software and the IPCC-recommended
Approach 2 methodology (Monte Carlo Stochastic Simulation technique), 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 presented below provide the 95 percent confidence interval, the range of values
within which emissions are likely to fall, for this source category.
62 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.
63 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 3-53
-------
As noted above, the non-energy use analysis is based on U.S.-specific storage factors for (1) feedstock materials
(natural gas, HGL, natural gasoline, naphthas, other oils, still gas, special naphthas, and other industrial coal), (2)
asphalt, (3) lubricants, and (4) waxes. For the remaining fuel types (the "other" category in Table 3-21 and Table
3-22) the storage factors were taken directly from IPCC (2006), where available, and otherwise assumptions were
made based on the potential fate of carbon in the respective NEU products. To characterize uncertainty, five
separate analyses were conducted, corresponding to each of the five categories. In all cases, statistical analyses or
expert judgments of uncertainty were not available directly from the information sources for all the activity
variables; thus, uncertainty estimates were determined using assumptions based on source category knowledge.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3-23 (emissions) and Table
3-24 (storage factors). Carbon emitted from non-energy uses of fossil fuels in 2020 was estimated to be between
76.3 and 180.2 MMT CO2 Eq. at a 95 percent confidence level. This indicates a range of 37 percent below to 49
percent above the 2020 emission estimate of 121.0 MMT CO2 Eq. The uncertainty in the emission estimates is a
function of uncertainty in both the quantity of fuel used for non-energy purposes and the storage factor.
Table 3-23: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Non-
Energy Uses of Fossil Fuels (MMT CO2 Eq. and Percent)
2020 Emission Estimate Uncertainty Range Relative to Emission Estimate3
Source Gas
(MMT CO? Eq.) (MMT C02 Eq.) (%)
Lower Upper Lower Upper
Bound Bound Bound Bound
Feedstocks C02 98.1 56.0 159.4 -43% +62%
Asphalt C02 0.3 0.1 0.6 -59% +121%
Lubricants C02 15.3 12.7 17.8 -17% +16%
Waxes C02 0.3 0.2 0.6 -26% +100%
Other CO2 TO 1A 81 -80% +15%
Total C02 121.0 76.3 180.2 -37% +49%
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.
Table 3-24: Approach 2 Quantitative Uncertainty Estimates for Storage Factors of Non-
Energy Uses of Fossil Fuels (Percent)
2020 Storage Factor Uncertainty Range Relative to Emission Estimate3
Source Gas
(%) (%) (%, Relative)
Lower Upper Lower Upper
Bound Bound Bound Bound
Feedstocks C02 62.6% 50.0% 73.1% -20% +17%
Asphalt C02 99.6% 99.1% 99.8% -0.5% +0.3%
Lubricants C02 9.2% 3.8% 17.6% -58% +92%
Waxes C02 57.8% 47.4% 67.6% -18% +17%
Other C02 11.5% 7.1% 83.3% -38% +622%
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 3-24, feedstocks 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-
appears to have 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
3-54 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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.
/erification
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 2019 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.
Recalculations Discussion
Several updates to activity data factors lead to recalculations of previous year results. The major updates are as
follows:
• EIA (2021b) updated energy consumption statistics across the time series relative to the previous
Inventory, which resulted in a slight decrease in emissions from 1990 to 2019.
• ACC (2021b) updated polyester fiber and acetic acid production in 2019, which resulted in a slight
decrease in emissions relative to the previous Inventory.
• U.S. International Trade Commission (2021) updated 2018 and 2019 import and export data, resulting in
fewer net exports relative to the previous Inventory.
• U.S. Census Bureau (2021) released new shipment data, which increased historical cleanser shipment
estimates from 2013 to 2019. Cleanser shipment data from 2013 to 2016 were updated to be linearly
interpolated between the 2012 and 2017 Economic Census values, and data from 2018 to 2019 were
proxied to the 2017 value.
Overall, these changes resulted in an average annual decrease of 0.6 MMT CO2 Eq. (0.5 percent) in carbon
emissions from non-energy uses of fossil fuels for the period 1990 through 2019, relative to the previous
Inventory.
Energy 3-55
-------
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
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.
3-56 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
3.3 Incineration of Waste (CRF Source
Category 1A5)
Incineration 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, incineration of MSW tends to occur at waste-to-energy facilities or industrial facilities where useful energy
is recovered, and thus emissions from waste incineration 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.
Incineration of waste results in conversion of the organic inputs to CO2. According to the 2006IPCC 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 incineration 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 incineration
estimate, though waste disposal practices for tires differ from MSW. Estimates on emissions from hazardous waste
incineration 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 27.6 million metric tons of MSW were incinerated in 2020 (EPA 2020b). Carbon dioxide emissions
from incineration of waste increased 1.5 percent since 1990, to an estimated 13.1 MMT CO2 (13,133 kt) in 2020.
Emissions across the time series are shown in Table 3-25 and Table 3-26.
Waste incineration is also a source of CFU and N2O emissions (De Soete 1993; IPCC 2006). Methane emissions from
the incineration of waste were estimated to be less than 0.05 MMT CO2 Eq. (less than 0.05 kt CH4) in 2020 and
have remained steady since 1990. Nitrous oxide emissions from the incineration of waste were estimated to be 0.4
MMT CO2 Eq. (1.4 kt N2O) in 2020 and have decreased by 13 percent since 1990. This decrease is driven by the
decrease in total MSW incinerated.
Table 3-25: CO2, ChU, and N2O Emissions from the Incineration of Waste (MMT CO2 Eq.)
Gas
1990
2005
2016
2017
2018
2019
2020
C02
12.9
13.3
14.4
13.2
13.3
12.9
13.1
ch4
+
+
+
+
+
+
+
n2o
0.5
0.4
0.4
0.4
0.4
0.4
0.4
Total
13.4
13.7
14.8
13.6
13.8
13.4
13.5
+ Does not exceed 0.05 MMT C02 Eq.
ble 3-26: CO2, ChU, and N2O Emissions from the Incineration of Waste (kt)
Gas
1990
2005
2016
2017
2018
2019
2020
C02
12,937
13,283
14,356
13,161
13,339
12,948
13,133
ch4
+
+
+
+
+
+
+
n2o
2
1
1
1
1
1
1
+ Does not exceed 0.05 kt.
Energy 3-57
-------
Methodology and Time-Series Consistency
Municipal Solid Waste Incineration
To determine both CO2 and non-CC>2 emissions from the incineration of waste, the tonnage of waste incinerated
and an estimated emissions factor are needed. Emission estimates from the incineration of tires are discussed
separately. Data for total waste incinerated 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 incineration tonnages based on data availability and accuracy throughout the time
series.
• 1990-2006: MSW incineration tonnages are from Biocycle incineration data. Tire incineration data from
RMA are removed to arrive at MSW incinerated without tires
• 2006-2010: MSW incineration tonnages are an average of Biocycle (with RMA tire data tonnage
removed), U.S. EPA Facts and Figures, EIA, and Energy Recovery Council data (with RMA tire data tonnage
removed).
• 2011-2020: MSW incineration tonnages are from EPA's GHGRP data.
Table 3-27 provides the estimated tons of MSW incinerated including and excluding tires.
Table 3-27: Municipal Solid Waste Incinerated (Metric Tons)
Year
Waste Incinerated
(excluding tires)
Waste Incinerated
(including tires)
1990
33,344,839
33,766,239
2005
26,486,414
28,631,054
2016
29,704,817
31,534,322
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
29,233,101
Sources: BioCycle, EPA Facts and Figures, ERC, GHGRP, EIA, RMA.
CO2 Emissions from MSW Excluding Scrap Tires
Fossil CO2 emission factors were calculated from EPA's GHGRP data for non-biogenic sources. Using GHGRP-
reported emissions for CH4and N2O and assumed emission factors, the tonnage of waste incinerated, excluding
tires, was derived. Methane and N2O emissions and assumed emission factors were used to estimate the amount
of MSW combusted in terms of energy content. The energy content of MSW combusted was then converted into
tonnage based on assumed MSW heating value. Two estimates were generated (one for CH4 and one for N2O) and
the two were averaged together. Dividing fossil CO2 emissions from GHGRP FLIGHT data for MSW combustors by
this estimated tonnage yielded an annual CO2 emission factor. As this data was only available following 2011, all
years prior use an average of the emission factors from 2011 through 2020.
Finally, CO2 emissions were calculated by multiplying the annual tonnage estimates, excluding tires, by the
calculated emissions factor. Calculated fossil CO2 emission factors are shown in Table 3-28.
3-58 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 3-28: Calculated Fossil CO2 Content per Ton Waste Incinerated (kg C02/Short Ton
Incinerated)
1990
2005
2016
2017
2018
2019
2020
C02 Emission Factors
367
367
381
360
361
363
377
CO2 Emissions from Scrap Tires
Scrap tires contain several types of synthetic rubber, carbon black, and synthetic fibers. Each type of synthetic
rubber has a discrete C content, and carbon black is 100 percent C. For synthetic rubber and carbon black in scrap
tires, information was obtained biannually from U.S. Scrap Tire Management Summary for 2005 through 2019 data
(RMA 2020). Information about scrap tire composition was taken from the Rubber Manufacturers' Association
internet site (RMA 2012a). Emissions of CO2 were calculated based on the amount of scrap tires used for fuel and
the synthetic rubber and carbon black content of scrap tires. The mass of incinerated material is multiplied by its C
content to calculate the total amount of carbon stored. 2020 values are proxied from 2019 data. More detail on
the methodology for calculating emissions from each of these waste incineration sources is provided in Annex 3.7.
Table 3-29 provides CO2 emissions from combustion of waste tires.
Table 3-29: CO2 Emissions from Combustion of Tires (MMT CO2 Eq.)
1990
2005
2016
2017
2018
2019
2020
Synthetic Rubber
0.3
1.6
1.4
1.3
1.3
1.2
1.2
C Black
0.4
2.0
1.7
1.6
1.5
1.5
1.5
Total
0.7
13.7
3.0
2.9
2.8
2.7
2.7
Non-CCh Emissions
Incineration of waste also results in emissions of Cl-Uand N2O. These emissions were calculated by multiplying the
total estimated mass of waste incinerated, including tires, by the respective emission factors. The emission factors
for Cm and N2O emissions per quantity of MSW combusted are default emission factors for the default
continuously-fed stoker unit MSW incineration technology type and were taken from IPCC (2006).
Uncertainty
An Approach 2 Monte Carlo analysis was performed to determine the level of uncertainty surrounding the
estimates of CO2 emissions and N2O emissions from the incineration of waste (given the very low emissions for
CH4, no uncertainty estimate was derived). IPCC Approach 2 analysis allows the specification of probability density
functions for key variables within a computational structure that mirrors the calculation of the Inventory estimate.
Statistical analyses or expert judgments of uncertainty were not available directly from the information sources for
most variables; thus, uncertainty estimates for these variables were determined using assumptions based on
source category knowledge and the known uncertainty estimates for the waste generation variables.
The uncertainties in the waste incineration emission estimates arise from both the assumptions applied to the data
and from the quality of the data. Key factors include reported CO2 emissions; N2O and Cl-Uemissions factors, and
tire synthetic rubber and black carbon contents. The highest levels of uncertainty surround the reported emissions
from GHGRP; the lowest levels of uncertainty surround variables that were determined by quantitative
measurements (e.g., combustion efficiency, C content of C black).
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3-30. Waste incineration
CO2 emissions in 2020 were estimated to be between 10.8 and 15.3 MMT CO2 Eq. at a 95 percent confidence level.
This indicates a range of 17 percent below to 17 percent above the 2020 emission estimate of 13.1 MMT CO2 Eq.
Also at a 95 percent confidence level, waste incineration N2O emissions in 2020 were estimated to be between 0.2
Energy 3-59
-------
and 1.0 MMT CO2 Eq. This indicates a range of 53 percent below to 162 percent above the 2020 emission estimate
of 0.4 MMT CO2 Eq.
Table 3-30: Approach 2 Quantitative Uncertainty Estimates for CO2 and N2O from the
Incineration of Waste (MMT CO2 Eq. and Percent)
2020 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
13.1
10.8
15.3
-17%
17%
Incineration of Waste
N20
0.4
0.2
1.0
-53%
162%
a Range of emission estimates predicted by Monte Carlo Simulation for a 95 percent confidence interval.
QA/QC and Verification
In order to ensure the quality of the emission estimates from waste incineration, 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 specifically focused on the emission factor and activity data
sources and methodology used for estimating emissions from incineration of waste. Trends across the time series
were analyzed to determine whether any corrective actions were needed. Corrective actions were taken to rectify
minor errors in the use of activity data.
Recalculations Discussion
Waste incineration estimates in the current Inventory were derived following a new methodology relying on
different data sources than previously used. Specifically:
• Waste tonnage estimates for 2006 through 2019 relied on several new data sources. Prior years relied on
proxied data from 2011 from Shin (2014).
• For 1990 through 2020, CO2 emissions were calculated with a new methodology using a carbon emission
factor calculated from EPA's GHGRP data. An emission factor for years prior to 2011 was estimated using
the average of 2011 through 2020 emission factors. The previous methodology relied on generation,
disposal, and incineration rates of synthetic fibers, plastics, and synthetic rubber and the accompanying
carbon contents to calculate CO2 emissions for incineration of these materials. The methodology for
estimating tire CO2 emissions did not change.
• Non-C02 emissions were calculated using the same IPCC (2006) default factor as previous years. However,
MSW incineration activity data changed based on the revisions to the methodology.
As a result of the changes in data and methodology, CO2 emissions in 2019 increased 13 percent relative to the
previous Inventory and there was an average annual increase of 20 percent over the time series. Non-C02
emissions for both CH4 and N2O increased by 30 percent relative to the prior Inventory. The observed change in
emissions is primarily due to the difference in MSW tonnages starting in 2010 and the revision of the CO2 emission
factor across the time series.
Planned Improvements
Currently, emission estimates for the biomass and biomass-based fuels source category included in this Inventory
are limited to woody biomass, ethanol, and biodiesel. EPA will incorporate emissions from biogenic components of
MSW to biomass and biomass-based fuels or waste incineration in future Inventory assessments.
3-60 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
3.4 Coal Mining (CRF Source Category
lBla)
Three types of coal mining-related activities release Cm and CO2 to the atmosphere: underground mining, surface
mining, and post-mining (i.e., coal-handling) activities. While surface coal mines account for the majority of U.S.
coal production, underground coal mines contribute the largest share of fugitive CH4 emissions (see Table 3-32 and
Table 3-33) due to the higher Cm content of coal in the deeper underground coal seams. In 2020,196
underground coal mines and 350 surface mines were operating in the United States (EIA 2021). In recent years, the
total number of active coal mines in the United States has declined. In 2020, the United States was the fifth largest
coal producer in the world (485 MMT), after China (3,764 MMT), India (760 MMT), Indonesia (564 MMT), and
Australia (493 MMT) (IEA 2021).
Table 3-31: 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
2005
586
334,399
789
691,447
1,398
1,025,846
2016
251
228,707
439
431,282
690
659,989
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
Fugitive CH4 Emissions
Underground coal mines liberate Cm from ventilation systems and from degasification systems. Ventilation
systems pump air through the mine workings to dilute noxious gases and ensure worker safety; these systems can
exhaust significant amounts of Cm to the atmosphere in low concentrations. Degasification systems are wells
drilled from the surface or boreholes drilled inside the mine that remove large, often highly concentrated volumes
of Cm before, during, or after mining. Some mines recover and use CH4 generated from ventilation and
degasification systems, thereby reducing emissions to the atmosphere.
Surface coal mines liberate CH4 as the overburden is removed and the coal is exposed to the atmosphere. Methane
emissions are normally a function of coal rank (a classification related to the percentage of carbon in the coal) and
depth. Surface coal mines typically produce lower-rank coals and remove less than 250 feet of overburden, so their
level of emissions is much lower than from underground mines.
In addition, CH4 is released during post-mining activities, as the coal is processed, transported, and stored for use.
Total CH4 emissions in 2020 were estimated to be 1,648 kt (41.2 MMT CO2 Eq.), a decline of approximately 57
percent since 1990 (see Table 3-32and Table 3-33). In 2020, underground mines accounted for approximately 76
percent of total emissions, surface mines accounted for 12 percent, and post-mining activities accounted for 12
percent. In 2020, total CH4 emissions from coal mining decreased by approximately 13 percent relative to the
previous year. This decrease was due to a decrease in annual coal production. The amount of CH4 recovered and
used in 2020 decreased by approximately 3 percent compared to 2019 levels.
Energy 3-61
-------
Table 3-32: ChU Emissions from Coal Mining (MMT CO2 Eq.)
Activity
1990
2005
2016
2017
2018
2019
2020
Underground (UG) Mining
74.2
42.0
40.7
40.7
38.9
34.5
31.4
Liberated
80.8
59.7
56.9
58.1
57.7
50.3
46.8
Recovered & Used
(6.6)
(17.7)
(16.2)
(17.4)
(18.8)
(15.8)
(15.4)
Surface Mining
10.8
11.9
6.8
7.2
7.0
6.4
4.9
Post-Mining (UG)
9.2
7.6
4.8
5.3
5.3
5.2
3.9
Post-Mining (Surface)
2.3
2.6
1.5
1.6
1.5
1.4
1.1
Total
96.5
64.1
53.8
54.8
52.7
47.4
41.2
Note: Parentheses indicate negative values.
ible 3-33: ChU Emissions from Coal Mining (kt)
Activity
1990
2005
2016
2017
2018
2019
2020
Underground (UG) Mining
2,968
1,682
1,629
1,626
1,556
1,379
1,257
Liberated
3,231
2,388
2,277
2,324
2,308
2,011
1,871
Recovered & Used
(263)
(706)
(648)
(698)
(752)
(633)
(614)
Surface Mining
430
475
273
290
280
255
194
Post-Mining (UG)
368
306
193
213
212
206
155
Post-Mining (Surface)
93
103
59
63
61
55
42
Total
3,860
2,565
2,154
2,191
2,109
1,895
1,648
Note: Parentheses indicate negative values.
Methodology and Time-Series Consistency
EPA uses an IPCC Tier 3 method for estimating CH4 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 CH4 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 CH4 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.
3-62 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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)64 (Subpart FF, "Underground Coal Mines"), data provided by the U.S. Mine Safety and Health
Administration (MSHA) (MSHA 2021), 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 2021).65 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.66
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
quarterly ventilation methane emissions (metric tons) are summed for each mine to develop mine-specific annual
ventilation emissions. For MSHA data, the average daily CH4 emission rate for each mine is determined using the
CH4 total for all data measurement events conducted during the calendar year and total duration of all data
measurement events (in days). The calculated average daily CH4 emission rate is then multiplied by 365 days to
estimate annual ventilation CH4 emissions for the MSHA dataset.
Step 1.2: Estimate CH4 Liberated from Degasification Systems
Particularly gassy underground mines also use degasification systems (e.g., wells or boreholes) to remove CH4
before, during, or after mining. This CH4 can then be collected for use or vented to the atmosphere. Twenty mines
used degasification systems in 2020 and 19 of these mines reported the CH4 removed through these systems to
EPA's GHGRP under Subpart FF (EPA 2021). Based on the weekly measurements reported to EPA's GHGRP,
degasification data summaries for each mine are added to estimate the CH4 liberated from degasification systems.
Thirteen of the 20 mines with degasification systems had operational CH4 recovery and use projects (see step 1.3
below).67
Degasification data reported to EPA's GHGRP by underground coal mines is the primary source of data used to
develop estimates of CH4 liberated from degasification systems. Data reported to EPA's GHGRP were used
exclusively to estimate CH4 liberated from degasification systems at 15 of the 20 mines that used degasification
systems in 2020. Data from state gas well production databases were used exclusively for a single mine and state
gas well production data were used to supplement GHGRP degasification data for the remaining four mines
(DMME 2021; GSA 2021; WVGES 2021).
For pre-mining wells, cumulative degasification volumes that occur prior to the well being mined through are
attributed to the mine in the inventory year in which the well is mined through.68 EPA's GHGRP does not require
64 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).
65 Underground coal mines report to EPA under Subpart FF of the GHGRP (40 CFR Part 98). In 2020, 71 underground coal mines
reported to the program.
66 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.
67 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.
68 A well is "mined through" when coal mining development or the working face intersects the borehole or well.
Energy 3-63
-------
gas production from virgin coal seams (coalbed methane) to be reported by coal mines under Subpart FF.69 Most
pre-mining wells drilled from the surface are considered coalbed methane wells prior to mine-through and
associated Cm emissions are reported under another subpart of the GHGRP (Subpart W, "Petroleum and Natural
Gas Systems"). As a result, GHGRP data must be supplemented to estimate cumulative degasification volumes that
occurred prior to well mine-through. There were five mines with degasification systems that include pre-mining
wells that were mined through in 2020. For four of these mines, GHGRP data were supplemented with historical
data from state gas well production databases (DMME 2021; ERG 2021; GSA 2021; WVGES 2021), as well as with
mine-specific information regarding the locations and dates on which the pre-mining wells were mined through
(JWR 2010; El Paso 2009; ERG 2021). State gas well production data were exclusively used for a single mine (GSA
2021).
Step 1.3: Estimate CH4 Recovered from Ventilation and Degasification Systems, and Utilized or
Destroyed (Emissions Avoided)
Thirteen mines had CH4 recovery and use projects in place in 2020, including one mine that had two recovery and
use projects. Thirteen of these projects involved degasification systems, in place at twelve mines, and one involved
a ventilation air methane abatement project (VAM). Eleven of these mines sold the recovered CH4to a pipeline,
including one that also used CH4 to fuel a thermal coal dryer. One mine used recovered CH4to heat mine
ventilation air (data were unavailable for estimating CH4 recovery at this mine). One mine destroyed the recovered
CH4 (VAM) using regenerative thermal oxidation (RTO) without energy recovery.
The CH4 recovered and used (or destroyed) at the thirteen mines described above are estimated using the
following methods:
• EPA's GHGRP data was exclusively used to estimate the CH4 recovered and used from seven of the 12
mines that deployed degasification systems in 2020. Based on weekly measurements, the GHGRP
degasification destruction data summaries for each mine are added together to estimate the CH4
recovered and used from degasification systems.
• State sales data were used to estimate CH4 recovered and used from the remaining five mines that
deployed degasification systems in 2020 (DMME 2021, ERG 2021, GSA 2021, and WVGES 2021). These five
mines intersected pre-mining wells in 2020. 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 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 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 2021) 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).
^ 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.
3-64 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 2020 were estimated to be 2,169 kt (2.2 MMT CO2 Eq.), a decline of approximately
53 percent since 1990. In 2020, underground mines accounted for approximately 89 percent of total fugitive CO2
emissions. In 2020, total fugitive CO2 emissions from coal mining decreased by approximately 27 percent relative
to the previous year. This decrease was due to a decrease in annual coal production.
Table 3-34: CO2 Emissions from Coal Mining (MMT CO2 Eq.)
Activity
1990
2005
2016
2017
2018
2019
2020
Underground (UG) Mining
4.2
3.6
2.5
2.7
2.7
2.6
1.9
Liberated
4.2
3.6
2.5
2.7
2.7
2.6
1.9
Recovered & Used
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Flaring
NO
NO
+
+
+
+
+
Surface Mining
0.4
0.6
0.3
0.4
0.4
0.3
0.2
Total
4.6
4.2
2.8
3.1
3.1
3.0
2.2
+ Does not exceed 0.05 MMT C02 Eq.
NO (Not Occurring)
Note: Parentheses indicate negative values.
Table 3-35: CO2 Emissions from Coal Mining (kt)
Activity
1990
2005
2016
2017
2018
2019
2020
Underground (UG) Mining
4,164
3,610
2,499
2,699
2,714
2,629
1,919
Liberated
4,171
3,630
2,483
2,690
2,712
2,633
1,926
Recovered & Used
(8)
(20)
(18)
(20)
(21)
(18)
(18)
Flaring
NO
NO
34
29
24
14
11
Surface Mining
443
560
349
368
353
322
249
Total
4,606
4,170
2,848
3,067
3,067
2,951
2,169
NO (Not Occurring)
Note: Parentheses indicate negative values.
Methodology and Time-Series Consistency
EPA uses an IPCC Tier 1 method for estimating fugitive CO2 emissions from underground coal mining and surface
mining (IPCC 2019). IPCC methods and data to estimate fugitive CO2 emissions from post-mining activities (for both
underground and surface coal mining) are currently not available.
Step 1: Underground Mining
EPA used the following overarching IPCC equation to estimate fugitive CO2 emissions from underground coal mines
(IPCC 2019):
Equation 3-1: Estimating Fugitive CO2 Emissions From Underground Mines
Total C02 from Underground Mines
= C02 from underground mining — Amount of C02 in gas recovered
+ C02 from methane flaring
Energy 3-65
-------
Step 1.1: Estimate Fugitive CO2 Emissions From Underground Mining
EPA estimated fugitive CO2 emissions from underground mining using the IPCC Tier 1 emission factor (5.9
m3/metric ton) and annual coal production from underground mines (EIA 2021). The underground mining default
emission factor accounts for all the fugitive CO2 likely to be emitted from underground coal mining. Therefore, the
amount of CO2 from coal seam gas recovered and utilized for energy is subtracted from underground mining
estimates in Step 2, below. Under IPCC methods, the CO2 emissions from gas recovered and utilized for energy use
(e.g., injected into a natural gas pipeline) are reported under other sectors of the Inventory (e.g., stationary
combustion of fossil fuel or oil and natural gas systems) and not under the coal mining sector.
Step 1.2: Estimate Amount of CO2 In Coal Seam Gas Recovered for Energy Purposes
EPA estimated fugitive CO2 emissions from coal seam gas recovered and utilized for energy purposes by using the
IPCC Tier 1 default emission factor (19.57 metric tons CCh/million cubic meters of coal bed methane (CBM)
produced) and quantity of coal seam gas recovered and utilized. Data on annual quantity of coal seam gas
recovered and utilized are available from GHGRP and state sales data (GHGRP 2021; DMME 2021; ERG 2021; GSA
2021; WVGES 2021). The quantity of coal seam gas recovered and destroyed without energy recovery (e.g., VAM
projects) is deducted from the total coal seam gas recovered quantity.
Step 1.3: Estimate Fugitive CO2 Emissions From Flaring
The IPCC method includes combustion CO2 emissions from gas recovered for non-energy uses (i.e., flaring, or
catalytic oxidation) under fugitive CO2 emission estimates for underground coal mining. In effect, these emissions,
though occurring through stationary combustion, are categorized as fugitive emissions in the Inventory. EPA
estimated CO2 emissions from methane flaring using the following equation:
Equation 3-2: Estimating CO2 Emissions From Drained Methane Flared Or Catalytically
Oxidized
C02 from flaring
= 0.98 x Volume of methane flared x Conversion Factor
x Stoichiometric Mass Factor
Currently there is only a single mine that reports catalytic oxidation of recovered methane through flaring without
energy use. Annual data for 2020 were obtained from the mine's offset verification statement (OVS) submitted to
the California Air Resources Board (CARB) (McElroy OVS 2021).
Step 2: Surface Mining
EPA estimated fugitive CO2 emissions from surface mining using the IPCC Tier 1 emission factor (0.44 m3/metric
ton) and annual coal production from surface mines (EIA 2021).
Uncertainty
A quantitative uncertainty analysis was conducted for the coal mining source category using the IPCC-
recommended Approach 2 uncertainty estimation methodology. Because emission estimates of CH4 from
underground ventilation systems were based on actual measurement data from EPA's GHGRP or from MSHA,
uncertainty is relatively low. A degree of imprecision was introduced because the ventilation air measurements
used were not continuous but rather quarterly instantaneous readings that were used to determine the average
annual emission rates. Additionally, the measurement equipment used can be expected to have resulted in an
average of 10 percent overestimation of annual CH4 emissions (Mutmansky & Wang 2000). Equipment
measurement uncertainty is applied to GHGRP data.
Estimates of CH4 liberated and recovered by degasification systems are relatively certain for utilized CH4 because of
the availability of EPA's GHGRP data and state gas sales information. Many of the liberation and recovery
estimates use data on wells within 100 feet of a mined area. However, uncertainty exists concerning the radius of
3-66 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
influence of each well. The number of wells counted, and thus the liberated Cm and avoided emissions, may vary if
the drainage area is found to be larger or smaller than estimated.
EPA's GHGRP requires weekly Cm monitoring of mines that report degasification systems, and continuous Cm
monitoring is required for Cm utilized on- or off-site. Since 2012, GHGRP data have been used to estimate Cm
emissions from vented degasification wells, reducing the uncertainty associated with prior MSHA estimates used
for this sub-source. Beginning in 2013, GHGRP data were also used for determining CH4 recovery and use at mines
without publicly available gas usage or sales records, which has reduced the uncertainty from previous estimation
methods that were based on information from coal industry contacts.
Surface mining and post-mining emissions are associated with considerably more uncertainty than underground
mines, because of the difficulty in developing accurate emission factors from field measurements. However, since
underground coal mining, as a general matter, results in significantly larger CH4 emissions due to production of
higher-rank coal and greater depth, and estimated emissions from underground mining constitute the majority of
estimated total coal mining CH4 emissions, the uncertainty associated with underground emissions is the primary
factor that determines overall uncertainty.
The major sources of uncertainty for estimates of fugitive CO2 emissions are the Tier 1IPCC default emission
factors used for underground mining (-50 percent to +100 percent) and surface mining (-67 percent to +200
percent) (IPCC 2019). Additional sources of uncertainty for fugitive CO2 emission estimates include ElA's annual
coal production data and data used for gas recovery projects, such as GHGRP data, state gas sales data, and VAM
estimates for the single mine that operates an active VAM project. Uncertainty ranges for these additional data
sources are already available, as these are the same data sources used for CH4 emission estimates.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3-36. Coal mining CH4
emissions in 2020 were estimated to be between 37.4 and 48.2 MMT CO2 Eq. at a 95 percent confidence level. This
indicates a range of 9.2 percent below to 17.1 percent above the 2020 emission estimate of 41.2 MMT CO2 Eq.
Coal mining fugitive CO2 emissions in 2020 were estimated to be between 0.7 and 3.8 MMT CO2 Eq. at a 95 percent
confidence level. This indicates a range of 68.3 percent below to 76.3 percent above the 2020 emission estimate of
2.2 MMT CO2 Eq.
Table 3-36: Approach 2 Quantitative Uncertainty Estimates for ChU and CO2 Emissions from
Coal Mining (MMT CO2 Eq. and Percent)
Source
Gas
2020 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
41.2
37.4
48.2
-9.2%
+17.1%
Coal Mining
C02
2.2
0.7
3.8
-68.3%
+76.3%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
In order to ensure the quality of the emission estimates for coal mining, 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 reported emissions data used for estimating fugitive emissions from
coal mining. Trends across the time series were analyzed to determine whether any corrective actions were
needed.
Emission estimates for coal mining rely in large part on data reported by coal mines to EPA's GHGRP. EPA verifies
annual facility-level reports through a multi-step process to identify potential errors and ensure that data
submitted to EPA are accurate, complete, and consistent. All reports submitted to EPA are evaluated by electronic
validation and verification checks. If potential errors are identified, EPA will notify the reporter, who can resolve
Energy 3-67
-------
the issue either by providing an acceptable response describing why the flagged issue is not an error or by
correcting the flagged issue and resubmitting their annual report. Additional QA/QC and verification procedures
occur for each GHGRP subpart. In 2021, a single facility resubmitted its 2020 annual Cm emissions data (i.e., mine
vent emissions) under subpart FF to correct data reporting issues in its initial submission.
Recalculations Discussion
State gas sales production values were updated for three mines as part of normal data updates conducted by
states. Data were updated for 2012 to 2014 and 2016 for one mine, for 2018 and 2019 for the second mine, and
for 2019 for the third mine. These changes resulted in slightly lower degasification Cm emissions and Cm
emissions avoided from underground mining for 2012 (0.04 percent) and 2016 (0.9 percent); and slightly higher
degasification and avoided emissions in the remaining years (0.02 to 1.0 percent) with the highest change in 2019
(1 percent). The change in both the degasification emissions and emissions avoided is less than 0.05 percent over
the 2012 to 2019 time series, compared to the previous Inventory.
Planned Improvements
EPA is assessing planned improvements for future reports, but at this time has no specific planned improvements
for estimating Cm and CO2 emissions from underground and surface mining and CH4 emissions from post-mining.
Updates on planned improvements will be included in the next Inventory submission.
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 7.2 to 10.8 MMT CO2 Eq. from 1990 to 2020, 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 (10.8 MMT CO2 Eq.) due to the large number of gassy
3-68 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
mine70 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 2020 there were three gassy mine closures. Gross abandoned mine
emissions decreased slightly from 8.5 MMT CO2 Eq. (341 kt CH4) in 2019 to 8.4 (335 kt CH4) MMT CO2 Eq. in 2020
(see Table 3-37 and Table 3-38). Gross emissions are reduced by Cm recovered and used at 45 mines, resulting in
net emissions in 2020 of 5.8 MMT CO2 Eq (231 kt CH4).
Table 3-37: ChU Emissions from Abandoned Coal Mines (MMT CO2 Eq.)
Activity 1990
2005
2016
2017
2018
2019
2020
Abandoned Underground Mines 7.2
Recovered & Used NO
8.4
(1.8)
9.5
(2.8)
9.2
(2.7)
8.9
(2.7)
8.5
(2.6)
8.4
(2.6)
Total 7.2
6.6
6.7
6.4
6.2
5.9
5.8
NO (Not Occurring)
Notes: Parentheses indicate negative values. Totals may not sum due to independent rounding.
able 3-38: ChU Emissions from Abandoned Coal Mines (kt)
Activity 1990
2005
2016
2017
2018
2019
2020
Abandoned Underground Mines 288
Recovered & Used NO
334
(70)
380
(112)
367
(109)
355
(107)
341
(104)
335
(104)
Total 288
264
268
257
247
237
231
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:
70 A mine is considered a "gassy" mine if it emits more than 100 thousand cubic feet of CH4 per day (100 mcfd).
Energy 3-69
-------
Equation 3-3: Decline Function to Estimate Venting Abandoned Mine Methane Emissions
q = qt (1 + &A0("1/fc)
where,
q
qi
b
Di
t
Gas flow rate at time t in million cubic feet per day (mmcfd)
Initial gas flow rate at time zero (t0), mmcfd
The hyperbolic exponent, dimensionless
Initial decline rate, 1/year
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
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 529 abandoned mines closed after 1972 produced Cm emissions greater than 100 mcfd when active. Further,
the status of 302 of the 529 mines (or 57 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 Cm flow to the atmosphere). The
remaining 43 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 3-39 presents the count of mines by post-abandonment state, based on EPA's probability
distribution analysis.
q = qie{-Dt)
where,
q
qi
D
t
Gas flow rate at time t in mmcfd
Initial gas flow rate at time zero (t0), mmcfd
Decline rate, 1/year
Elapsed time from t0 (years)
3-70 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 3-39: Number of Gassy Abandoned Mines Present in U.S. Basins in 2020, Grouped by
Class According to Post-Abandonment State
Basin
Sealed
Vented
Flooded
Total
Known
Unknown Total Mines
Central Appl.
43
25
50
118
144
262
Illinois
34
3
14
51
31
82
Northern Appl.
48
23
15
86
38
124
Warrior Basin
0
0
16
16
0
16
Western Basins
28
4
2
34
10
44
Total
153
55
97
305
223
528
Inputs to the decline equation require the average Cm 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 U.S., 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 2021). 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 CFU vented to determine the total CFU
liberation rate for all mines that closed between 1992 and 2020. 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 2020, 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.
Energy 3-71
-------
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) Cm 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 3-40. Annual abandoned
coal mine Cm emissions in 2020 were estimated to be between 4.5 and 6.9 MMT CO2 Eq. at a 95 percent
confidence level. This indicates a range of 22 percent below to 20 percent above the 2020 emission estimate of 5.8
MMT CO2 Eq. One of the reasons for the relatively narrow range is that mine-specific data is available for use in the
methodology for mines closed in 1972 and later years. Emissions from mines closed prior to 1972 have the largest
degree of uncertainty because no mine-specific Cm liberation rates at the time of abandonment exist.
Table 3-40: Approach 2 Quantitative Uncertainty Estimates for ChU Emissions from
Abandoned Underground Coal Mines (MMT CO2 Eq. and Percent)
Source Gas
2020 Emission Estimate
(MMT C02 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Abandoned Underground
CH4
Coal Mines
5.8
4.5
6.9
-21.9%
+19.5%
a Range of emission estimates predicted by Monte Carlo Simulation for a 95 percent confidence interval.
QA/QC and Verification
In order to ensure the quality of the emission estimates for abandoned coal mines, 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 reported emissions data used for estimating emissions from
abandoned coal mines. Trends across the time series were analyzed to determine whether any corrective actions
were needed.
Recalculations Discussion
No recalculations were performed for prior year estimates in the time series.
3.6 Petroleum Systems (CRF Source
Category lB2a)
This IPCC category (lB2a) is for fugitive emissions from petroleum systems, which per IPCC guidelines include
emissions from leaks, venting, and flaring. Methane emissions from petroleum systems are primarily associated
with onshore and offshore crude oil exploration, production, transportation, and refining operations. During these
activities, Cm is released to the atmosphere as emissions from leaks, venting (including emissions from operational
upsets), and flaring. Carbon dioxide emissions from petroleum systems are primarily associated with onshore and
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
3-72 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
(except for flaring) are accounted for in the fossil fuel combustion chapter (see Section 3). Emissions of N2O from
petroleum systems are primarily associated with flaring. Total greenhouse gas emissions (CH4, CO2, and N2O) from
petroleum systems in 2020 were 70.4 MMT CO2 Eq., an increase of 23 percent from 1990, primarily due to
increases in CO2 emissions. Total emissions increased by 23 percent from 2010 levels, and have decreased by 19
percent since 2019. Total CO2 emissions from petroleum systems in 2020 were 30.2 MMT CO2 (30,156 kt CO2), 3.1
times higher than in 1990. Total CO2 emissions in 2020 were 2.0 times higher than in 2010 and 35 percent lower
than in 2019. Total CH4 emissions from petroleum systems in 2020 were 40.2 MMT CO2 Eq. (1,609 kt CH4), a
decrease of 16 percent from 1990. Since 2010, total Cm emissions decreased by 4 percent; and since 2019, Cm
emissions decreased by 0.3 percent. Total N2O emissions from petroleum systems in 2020 were 0.04 MMT CO2 Eq.
(0.13 kt N2O), 2.5 times higher than in 1990,1.9 times higher than in 2010, and 18 percent lower than in 2019.
Since 1990, U.S. oil production has increased by 54 percent. In 2020, production was 106 percent higher than in
2010 and 8 percent lower than in 2019.
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 2020) to ensure
that the trend is representative of changes in emissions levels. Recalculations in petroleum systems in this year's
Inventory include:
• Updates to well counts, oil and gas production volumes, and produced water volumes using the most
recent data from Enverus and the United States Energy Information Administration (EIA)
• Recalculations due to Greenhouse Gas Reporting Program (GHGRP) submission revisions
The Recalculations Discussion section below provides more details on the updated methods.
Exploration. Exploration includes well drilling, testing, and completions. Exploration accounts for less than 1
percent of total CH4 emissions (including leaks, vents, and flaring) from petroleum systems in 2020. 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 92 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 30 times
higher than in 2020; and lowest in 2020. Emissions of CH4 from exploration decreased 27 percent from 2019 to
2020, due to a decrease in emissions from hydraulically fractured oil well completions without RECs or flaring.
Exploration accounts for 3 percent of total CO2 emissions (including leaks, vents, and flaring) from petroleum
systems in 2020. Emissions of CO2 from exploration in 2020 were 2.4 times higher than in 1990, and decreased by
64 percent from 2019, due to a large decrease in the number of hydraulically fractured oil well completions (by
50% from 2019). Emissions of CO2 from exploration were highest in 2014, over 4 times higher than in 2020.
Exploration accounts for 1 percent of total N2O emissions from petroleum systems in 2020. Emissions of N2O from
exploration in 2020 are 2.3 times higher than in 1990, and 59 percent lower than in 2019, due to the
abovementioned changes in hydraulically fractured oil well completions with flaring.
Production. Production accounts for 97 percent of total CH4 emissions (including leaks, vents, and flaring) from
petroleum systems in 2020. The predominant sources of emissions from production field operations are pneumatic
controllers, offshore oil platforms, equipment leaks, gas engines, produced water, chemical injection pumps, and
associated gas flaring. In 2020, these seven sources together accounted for 92 percent of the CH4 emissions from
production. Since 1990, CH4 emissions from production have decreased by 10 percent due to decreases in
emissions from offshore platforms and tanks. Overall, production segment CH4 emissions increased by less than
0.5 percent from 2019 levels due primarily to increased pneumatic controller emissions. Production emissions
account for 83 percent of the total CO2 emissions (including leaks, vents, and flaring) from petroleum systems in
2020. The principal sources of CO2 emissions are associated gas flaring, miscellaneous production flaring, and oil
tanks with flares. In 2020, these three sources together accounted for 97 percent of the CO2 emissions from
production. In 2020, CO2 emissions from production were 4.2 times higher than in 1990, due to increases in flaring
emissions from associated gas flaring, miscellaneous production flaring, and tanks. Overall, in 2020, production
segment CO2 emissions decreased by 36 percent from 2019 levels primarily due to decreases in associated gas
flaring and miscellaneous production flaring in the Permian and Williston Basins. Production emissions accounted
Energy 3-73
-------
for 65 percent of the total N2O emissions from petroleum systems in 2020. The principal sources of N2O emissions
are associated gas flaring, oil tanks with flares, miscellaneous production flaring, and offshore flaring. In 2020, N2O
emissions from production were 4.4 times higher than in 1990 and were 18 percent lower than in 2019.
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 less than 1 percent of total CH4 emissions from petroleum systems. Emissions from tanks, marine loading, and
truck loading operations account for 74 percent of CH4 emissions from crude oil transportation. Since 1990, CH4
emissions from transportation have increased by 32 percent. In 2020, CH4 emissions from transportation
decreased by 4 percent from 2019 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 74 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) Cm emissions from petroleum systems. This low share is because most of the CH4 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 14,13 and 9 percent, respectively. Fugitive CH4
emissions from refining of crude oil have increased by 15 percent since 1990, and decreased 13 percent from
2019; however, like the transportation subcategory, this increase has had little effect on the overall emissions of
Cm from petroleum systems. Crude oil refining processes and systems account for 14 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.71 Since 1990, refinery fugitive CO2 emissions
increased by 32 percent and have decreased by 15 percent from the 2019 levels, due to a decrease in flaring.
Flaring occurring at crude oil refining processes and systems accounts for 34 percent of total fugitive N2O
emissions from petroleum systems. In 2020, refinery fugitive N2O emissions increased by 39 percent since 1990,
and decreased by 15 percent from 2019 levels.
Table 3-41: Total Greenhouse Gas Emissions (CO2, ChU, and N2O) from Petroleum Systems
(MMT COz Eq.)
Activity
1990
2005
2016
2017
2018
2019
2020
Exploration
4.2
5.7
1.9
2.3
3.7
2.8
1.2
Production
49.1
43.1
55.4
58.5
67.4
78.0
63.9
Transportation
0.2
0.1
0.2
0.2
0.2
0.2
0.2
Crude Refining
4.0
4.5
4.8
4.6
4.6
6.0
5.1
Total
57.4
53.4
62.3
65.6
75.9
87.1
70.4
Note: Totals may not sum due to independent rounding.
Table 3-42: ChU Emissions from Petroleum Systems (MMT CO2 Eq.)
Activity
1990
2005
2016
2017
2018
2019
2020
Exploration
3.8
5.3
0.6
0.4
0.5
0.4
0.3
Production
43.1
35.2
38.8
39.0
37.1
38.8
38.9
Pneumatic Controllers
18.4
17.7
20.6
20.9
18.2
18.3
21.3
Offshore Production
00
00
6.5
5.1
5.1
4.9
4.9
4.8
Equipment Leaks
2.0
2.2
2.5
2.5
2.5
2.5
2.4
Gas Engines
2.0
1.8
2.2
2.2
2.3
2.3
2.2
Produced Water
2.3
1.6
2.0
2.1
2.3
2.4
2.2
Chemical Injection Pumps
1.2
1.7
2.0
2.0
2.0
2.0
1.9
71 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.
3-74 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Assoc Gas Flaring
0.5
0.4
0.7
1.0
1.7
1.9
1.0
Other Sources
CO
3.5
3.6
3.2
3.2
4.5
3.0
Crude Oil Transportation
0.2
0.1
0.2
0.2
0.2
0.2
0.2
Refining
0.7
0.8
0.8
0.8
0.8
0.9
0.8
Total
47.8
41.4
40.4
40.5
38.6
40.4
40.2
Note: Totals may not sum due to independent rounding.
Table 3-43: ChU Emissions from Petroleum Systems (kt ChU)
Activity
1990
2005
2016
2017
2018
2019
2020
Exploration
154
211
22
17
20
16
12
Production
1,725
1,408
1,552
1,562
1,484
1,553
1,557
Pneumatic Controllers
736
709
822
835
727
732
854
Offshore Production
353
259
204
204
196
196
193
Equipment Leaks
82
86
102
100
99
98
95
Gas Engines
82
71
90
89
92
94
89
Produced Water
91
62
81
84
93
98
89
Chemical Injection Pumps
47
68
82
81
80
79
76
Assoc Gas Flaring
20
14
29
38
68
77
42
Other Sources
313
139
142
130
129
179
120
Crude Oil Transportation
7
5
8
8
8
9
9
Refining
27
31
33
34
31
36
31
Total 1,912 1,655 1,616 1,621 1,544 1,615 1,609
Note: Totals may not sum due to independent rounding.
Table 3-44: CO2 Emissions from Petroleum Systems (MMT CO2)
Activity
1990
2005
2016
2017
2018
2019
2020
Exploration
0.4
0.4
1.4
1.9
3.2
2.4
0.9
Production
6.0
7.9
16.6
19.4
30.3
39.2
25.0
Transportation
+
+
+
+
+
+
+
Crude Refining
3.3
3.7
4.0
3.7
3.8
5.1
4.3
Total
9.6
12.0
21.9
25.0
37.3
46.7
30.2
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Table 3-45: CO2 Emissions from Petroleum Systems (kt CO2)
Activity
1990
2005
2016
2017
2018
2019
2020
Exploration
360
392
1,373
1,852
3,189
2,418
860
Production
5,955
7,874
16,555
19,449
30,296
39,187
24,969
Transportation
0.9
0.7
1.1
1.1
1.2
1.3
1.2
Crude Refining
3,284
3,728
3,994
3,725
3,820
5,080
4,326
Total
9,600
11,994
21,922
25,027
37,306
46,686
30,156
Note: Totals may not sum due to independent rounding.
Table 3-46: N2O Emissions from Petroleum Systems (Metric Tons CO2 Eq.)
Activity
1990
2005
2016
2017
2018
2019
2020
Exploration
179
193
700
811
1,503
1,017
419
Production
5,518
6,145
14,370
15,069
28,724
29,734
24,386
Transportation
NE
NE
NE
NE
NE
NE
NE
Crude Refining
9,130
10,363
11,582
10,801
10,786
14,905
12,730
Total
14,827
16,702
26,652
26,680
41,012
45,656
37,534
Energy 3-75
-------
NE (Not Estimated)
Note: Totals may not sum due to independent rounding.
Table 3-47: N2O Emissions from Petroleum Systems (Metric Tons N2O)
Activity
1990
2005
2016
2017
2018
2019
2020
Exploration
0.6
0.6
2.3
2.7
5.0
3.4
1.4
Production
18.5
20.6
48.2
50.6
96.4
99.8
81.8
Transportation
NE
NE
NE
NE
NE
NE
NE
Crude Refining
30.6
34.8
38.9
36.2
36.2
50.0
42.7
Total
49.8
56.0
89.4
89.5
137.6
153.2
126.0
NE (Not Estimated)
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
See Annex 3.5 for the full time series of emissions data, activity data, emission factors, and additional information
on methods and data sources.
Petroleum systems includes emission estimates for activities occurring in petroleum systems from the oil wellhead
through crude oil refining, including activities for crude oil exploration, production field operations, crude oil
transportation activities, and refining operations. Generally, emissions are estimated for each activity by
multiplying emission factors (e.g., emission rate per equipment or per activity) by corresponding activity data (e.g.,
equipment count or frequency of activity). Certain sources within petroleum refineries are developed using an
IPCC Tier 3 approach (i.e., all refineries in the nation report facility-level emissions data to the GHGRP, which are
included directly in the national emissions estimates here). Other estimates are developed with a Tier 2 approach.
Tier 1 approaches are not used.
EPA received stakeholder feedback on updates in the Inventory through EPA's stakeholder process on oil and gas
in the Inventory. Stakeholder feedback is noted below in Recalculations Discussion and Planned Improvements.
More information on the stakeholder process can be found online.72
Emission Factors. Key references for emission factors include Methane Emissions from the Natural Gas Industry by
the Gas Research Institute and EPA (GRI/EPA 1996), Estimates of Methane Emissions from the U.S. Oil Industry (EPA
1999), Compilation of Air Pollutant Emission Factors, AP-42 (EPA 1997), Global Emissions of Methane from
Petroleum Sources (API 1992), consensus of industry peer review panels, Bureau of Ocean Energy Management
(BOEM) reports, Nonpoint Oil and Gas Emission Estimation Tool (EPA 2017), and analysis of GHGRP data (EPA
2021).
Emission factors for hydraulically fractured (HF) oil well completions and workovers (in four control categories)
were developed using EPA's GHGRP data; year-specific data were used to calculate emission factors from 2016-
forward and the year 2016 emission factors were applied to all prior years in the time series. The emission factors
for all years for pneumatic controllers and chemical injection pumps were developed using GHGRP data for
reporting year 2014. The emission factors for tanks, 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.73 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
72 See https://www.epa.Eov/ghgemissions/natural-gas-and-petroleum-svstems.
73 See https://www.epa.gov/ghgemissions/natural-gas-and-petroleum-systems.
3-76 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 2020, 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 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
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 2021).
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
Energy 3-77
-------
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 CFU
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 2020.
Uncertainty
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 Estimates,74
EPA used Microsoft Excel's @ 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.
74 See https://www.epa.gov/ghgemissions/natural-gas-and-petroleum-systems.
3-78 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 3-48. 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 3-48: Approach 2 Quantitative Uncertainty Estimates for ChU and CO2 Emissions from
Petroleum Systems (MMT CO2 Eq. and Percent)
Source
Gas
2020 Emission Estimate
(MMT CO? Eq.)b
Uncertainty Range Relative to Emission Estimate3
(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
n2o
0.04
0.03
0.05
-22%
+26%
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.75
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 2021. 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.
75 See https://www.epa.gov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
Energy 3-79
-------
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.76 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.77
An updated version of the gridded inventory is being developed and will improve efforts to compare results of the
inventory with atmospheric studies.
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, stakeholder
feedback on updates under consideration, and new studies.
EPA did not make methodological updates for Petroleum Systems emission sources for the current Inventory.
However, for certain sources, CH4 and/or CO2 emissions changed by greater than 0.05 MMT CO2 Eq., comparing the
previous estimate for 2019 to the current (recalculated) estimate for 2019. The emissions changes were mostly
due to GHGRP data submission revisions and Enverus well count updates. These sources are discussed below and
include hydraulically fractured oil well completions, associated gas venting and flaring, production storage tanks,
pneumatic controllers, chemical injection pumps, gas engines, produced water, offshore production, and
refineries.
The combined impact of revisions to 2019 petroleum systems CH4 emission estimates, compared to the previous
Inventory, is an increase from 39.3 to 40.4 MMT CO2 Eq. (1.1 MMT CO2 Eq., or 3 percent). The recalculations
resulted in an average increase in CH4 emission estimates across the 1990 through 2019 time series, compared to
the previous Inventory, of 1.5 MMT CO2 Eq., or 4 percent, with the largest increase being in the estimate for 2012
(3.4 MMT CO2 Eq. or 8 percent) primarily due to the recalculations for hydraulically fractured oil well completions.
76 See https://www.epa.gov/eheemissions/eridded-2012-methane-emissions.
77 See https://www.epa.eov/eheemissions/us-greenhouse-eas-inventory-report-1990-2014.
3-80 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
The combined impact of revisions to 2019 petroleum systems CO2 emission estimates, compared to the previous
Inventory, is a decrease from 47.3 to 46.7 MMT CO2 (0.58 MMT CO2, or 1 percent). The recalculations resulted in
an average decrease in emission estimates across the 1990 through 2019 time series, compared to the previous
Inventory, of 0.02 MMT CO2 Eq„ or 0.4 percent with the largest changes being for 2019 primarily due to the
recalculations for associated gas flaring.
The combined impact of revisions to 2019 petroleum systems N2O emission estimates, compared to the previous
Inventory, is a decrease of 0.001 MMT CO2, Eq. or 3 percent. The emission changes were primarily driven by
reduction in flaring emissions from associated gas and offshore production flaring due to GHGRP data submission
revisions. The recalculations resulted in an average decrease in emission estimates across the 1990 through 2019
time series, compared to the previous Inventory, of 0.002 MMT CO2 Eq., or 9 percent.
In Table 3-49 and Table 3-50 below are categories in Petroleum Systems with updated methodologies or with
recalculations resulting in a change of greater than 0.05 MMT CO2 Eq., comparing the previous estimate for 2019
to the current (recalculated) estimate for 2019. For more information, please see the discussion below.
Table 3-49: Recalculations of CO2 in Petroleum Systems (MMT CO2)
Previous Estimate
Current Estimate
Current Estimate
Segment/Source
Year 2019,
2021 Inventory
Year 2019,
2022 Inventory
Year 2020,
2022 Inventory
Exploration
2.1
2.4
0.9
HF Oil Well Completions
2.1
2.4
0.9
Production
40.2
39.2
25.0
Tanks
6.1
6.7
6.5
Associated Gas Flaring
25.4
23.7
13.0
Transportation
+
+
+
Refining
5.0
5.1
4.3
Petroleum Systems Total
47.3
46.7
30.2
+ Does not exceed 0.05 MMT C02.
ible 3-50: Recalculations of CH4
in Petroleum Systems (MMT CO2 Eq.)
Previous Estimate
Current Estimate
Current Estimate
Segment/Source
Year 2019,
2021 Inventory
Year 2019,
2022 Inventory
Year 2020,
2022 Inventory
Exploration
0.3
0.4
0.3
HF Oil Well Completions
0.2
0.4
0.3
Production
35.7
38.8
38.9
Produced Water
2.1
2.4
2.2
Tanks
1.5
0.9
0.7
Pneumatic Controllers
17.5
18.3
21.3
Associated Gas Flaring
2.0
1.9
1.0
Associated Gas Venting
1.1
1.7
0.6
Chemical Injection Pumps
1.9
2.0
1.9
Offshore Production
5.0
4.9
4.8
Gas Engines
2.4
2.3
2.2
Transportation
0.2
0.2
0.2
Refining
0.9
0.9
0.8
Petroleum Systems Total
39.3
40.4
40.2
Exploration
HF Oil Well Completions (Recalculation with Updated Data)
Energy 3-81
-------
HF oil well completion CO2 emissions increased by an average of 24 percent across the time series and increased
by 16 percent in 2019, compared the to the previous Inventory. The emissions changes were due to GHGRP data
submission revisions and updated Enverus well completion counts.
Table 3-51: HF Oil Well Completions National CO2 Emissions (kt CO2)
Source
1990
2005
2016
2017
2018
2019
2020
HF Completions: Non-REC with Venting
3
4
+
+
+
+
+
HF Completions: Non-REC with Flaring
115
163
280
430
644
925
386
HF Completions: REC with Venting
NO
NO
+
+
+
+
+
HF Completions: REC with Flaring
NO
NO
1,053
1,385
2,512
1,489
472
Total Emissions
119
168
1,333
1,815
3,155
2,414
858
Previous Estimate
91
144
1,174
1,664
2,874
2,078
NA
+ Does not exceed 0.5 kt C02.
NA (Not Applicable)
NO (Not Occurring)
HF oil well completion CH4 emissions increased by an average of 27 percent across the time series and increased
by 53 percent in 2019, compared the to the previous Inventory. The emissions changes were due to GHGRP data
submission revisions and updated Enverus well completion counts.
Table 3-52: HF Oil Well Completions National CH4 Emissions (Metric Tons CH4)
Source
1990
2005
2016
2017
2018
2019
2020
HF Completions: Non-REC with Venting
142,812
202,077
8,034
2,882
169
805
819
HF Completions: Non-REC with Flaring
492
696
1,195
1,877
2,690
3,059
2,018
HF Completions: REC with Venting
NO
NO
3,695
4,127
4,892
5,071
5,716
HF Completions: REC with Flaring
NO
NO
5,584
6,499
10,362
6,127
2,024
Total Emissions
143,304
202,773
18,507
15,386
18,113
15,062
10,576
Previous Estimate
109,658
173,537
15,039
12,326
14,187
9,871
NA
NA (Not Applicable)
NO (Not Occurring)
Production
Produced Water (Recalculation with Updated Data)
Produced water CH4 emissions increased by an average of 2 percent across the time series and increased by 15
percent in 2019, compared the to the previous Inventory. The emissions changes were primarily due to
incorporating year-specific produced water volumes from Enverus and supplemented with updated data from the
NEI's O&G Tool for 6 states (IL, IN, KS, OK, PA, and WV).
Table 3-53: Produced Water National CH4 Emissions (Metric Tons CH4)
Source
1990
2005
2016
2017
2018
2019
2020
Low Pressure Oil Wells
Regular Pressure Oil Wells
20,273
71,118
13,855
48,603
18,073
63,403
18,661
65,464
20,599
72,263
21,680
76,055
19,658
68,964
Total
91,391
62,458
81,477
84,125
92,863
97,735
88,622
Previous Estimate
91,478
62,184
77,278
78,739
82,806
84,726
NA
NA (Not Applicable)
Tanks (Recalculation with Updated Data)
Tank CO2 emissions estimates decreased by an average of 0.2 percent across the 1990 to 2019 time series and
increased by 10 percent in 2019, compared to the previous inventory. The emission changes were due to GHGRP
data submission revisions.
3-82 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 3-54: Tanks National CO2 Emissions (kt CO2)
Source
1990
2005
2016
2017
2018
2019
2020
Large Tanks w/Flares
NO
2,440
4,441
4,247
6,130
6,625
6,486
Large Tanks w/VRU
NO
5
3
4
4
6
1
Large Tanks w/o Control
24
6
5
4
5
5
4
Small Tanks w/Flares
NO
2
10
11
8
9
13
Small Tanks w/o Flares
6
3
4
4
4
4
5
Malfunctioning Separator Dump
Valves
85
50
31
43
38
33
28
Total Emissions
115
2,505
4,494
4,313
6,189
6,682
6,537
Previous Estimate
116
2,517
4,546
4,364
6,278
6,098
NA
NA (Not Applicable)
NO (Not Occurring)
Tank Cm emissions estimates decreased by an average of 2 percent across the 1990 to 2019 time series and
decreased by 41 percent in 2019, compared to the previous inventory. The emission changes were due to GHGRP
data submission revisions.
Table 3-55: Tanks National ChU Emissions (MT CH4)
Source
1990
2005
2016
2017
2018
2019
2020
Large Tanks w/Flares
NO
2,303
3,994
5,310
6,879
4,324
4,268
Large Tanks w/VRU
NO
1,116
14,369
9,058
2,574
2,430
1,109
Large Tanks w/o Control
210,278
52,435
48,888
39,930
44,185
25,454
20,746
Small Tanks w/Flares
NO
15
17
63
22
29
33
Small Tanks w/o Flares
4,206
2,009
2,551
2,399
2,710
2,493
3,014
Malfunctioning Separator Dump
Valves
3,935
2,308
6,029
4,338
1,043
536
443
Total Emissions
218,419
60,186
75,848
61,098
57,412
35,266
29,613
Previous Estimate
219,476
60,489
76,086
61,658
58,848
59,965
NA
NA (Not Applicable)
NO (Not Occurring)
Pneumatic Controllers (Recalculation with Updated Data)
Pneumatic controller CH4 emission estimates increased by an average of 3 percent across the 1990 to 2019 time
series and increased by 5 percent in 2019, compared to the previous Inventory. The emission changes were due to
GHGRP data submission revisions and updated Enverus well counts.
Table 3-56: Pneumatic Controller National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2016
2017
2018
2019
2020
High Bleed
Low Bleed
Intermittent Bleed
689,395
47,052
NO
423,219
44,322
241,140
81,038
17,302
724,015
52,508
19,395
763,226
39,530
30,184
657,651
45,196
39,497
647,399
43,014
33,783
776,765
Total Emissions
736,447
708,680
822,355
835,129
727,365
732,092
853,562
Previous Estimate
792,075
672,769
785,023
799,496
693,976
699,488
NA
NA (Not Applicable)
Energy 3-83
-------
Associated Gas Flaring (Recalculation with Updated Data)
Associated gas flaring CO2 emission estimates decreased by an average of 3 percent across the time series and
decreased by 6 percent in 2019 in the current Inventory, compared to the previous Inventory. The emission
changes were due to GHGRP data submission revisions.
Table 3-57: Associated Gas Flaring National CO2 Emissions (kt CO2)
Source
1990
2005
2016
2017
2018
2019
2020
220 - Gulf Coast Basin (LA, TX)
225
124
405
749
651
713
798
360 - Anadarko Basin
102
63
1
62
79
18
10
395 - Williston Basin
969
1,243
6,090
6,909
11,140
14,762
8,052
430 - Permian Basin
2,844
1,971
2,215
3,141
6,711
7,227
3,558
"Other" Basins
944
507
322
384
624
990
624
Total Emissions
5,084
3,908
9,033
11,245
19,206
23,710
13,041
220 - Gulf Coast Basin (LA, TX)
227
121
404
744
643
584
NA
360 - Anadarko Basin
108
66
1
64
82
18
NA
395 - Williston Basin
987
1,263
6,091
6,908
11,140
16,572
NA
430 - Permian Basin
2,9S3
2,056
2,261
3,209
6,782
7,161
NA
"Other" Basins
935
505
324
387
641
1,021
NA
Previous Estimate
5,241
4,011
9,081
11,313
19,287
25,356
NA
NA (Not Applicable)
Associated gas flaring CH4 emission estimates decreased by an average of 4 percent across the time series and
decreased by 5 percent in 2019 in the current Inventory, compared to the previous Inventory. The emission
changes were due to GHGRP data submission revisions.
Table 3-58: Associated Gas Flaring National CH4 Emissions (Metric Tons ChU)
Source
1990
2005
2016
2017
2018
2019
2020
220 - Gulf Coast Basin (LA, TX)
886
490
1,576
2,949
2,480
2,995
3,710
360 - Anadarko Basin
447
274
4
268
348
88
21
395 - Williston Basin
2,665
3,419
16,945
20,707
37,756
43,637
22,954
430- Permian Basin
11,263
7,805
8,793
12,912
25,236
27,194
12,854
"Other" Basins
4,369
2,347
1,185
1,278
1,881
3,507
2,312
Total Emissions
19,630
14,335
28,503
38,115
67,701
77,422
41,850
220 - Gulf Coast Basin (LA, TX)
896
479
1,572
2,936
2,448
2,626
NA
360 - Anadarko Basin
472
288
4
277
358
87
NA
395 - Williston Basin
2,931
3,750
16,948
20,707
37,754
48,453
NA
430 - Permian Basin
11,815
8,143
8,972
13,189
25,511
27,016
NA
"Other" Basins
4,32S
2,335
1,193
1,290
1,932
3,614
NA
Previous Estimate
20,441
14,995
28,689
38,399
68,004
81,797
NA
NA (Not Applicable)
Associated Gas Venting (Recalculation with Updated Data)
Associated gas venting CH4 emission estimates increased by an average of 1 percent across the 1990 to 2019 time
series in the current Inventory, compared to the previous Inventory. The CH4 estimates increased by 63 percent in
2019, primarily due to Permian Basin data. The changes were due to GHGRP data submission revisions.
Table 3-59: Associated Gas Venting National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2016
2017
2018
2019
2020
220 - Gulf Coast Basin (LA, TX)
475
263
2,455
580
957
7,621
1,288
360 - Anadarko Basin
811
497
782
4,585
318
596
1,700
395 - Williston Basin
207
265
1,479
628
575
4,044
341
3-84 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
430- Permian Basin
"Other" Basins
Total Emissions
4,041
15,763
21,297
2,800
8,468
12,295
8,060
5,071
17,847
9,635
3,472
18,899
9,505
1,739
13,093
54,192
2,082
68,535
18,269
1,919
23,517
220 - Gulf Coast Basin (LA, TX)
480
257
2,449
579
944
18,139
NA
360 - Anadarko Basin
858
524
813
4,728
328
590
NA
395 - Williston Basin
211
269
1,479
628
575
10,855
NA
430 - Permian Basin
4,239
2,922
8,224
9,842
9,647
8,637
NA
"Other" Basins
15,613
8,424
5,104
3,503
1,788
3,830
NA
Previous Estimate
21,401
12,396
18,069
19,280
13,282
42,051
NA
NA (Not Applicable)
Chemical Injection Pumps (Recalculation with Updated Data)
Chemical injection pump Cm emission estimates increased by an average of 3 percent across the time series and
increased by 5 percent in 2019, compared to the previous Inventory. The emission changes were due to updated
Enverus well counts.
Table 3-60: Chemical Injection Pump National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2016
2017
2018
2019
2020
Chemical Injection Pump
Previous Estimate
46,758
50,806
67,685
64,259
81,936
78,351
80,728
77,061
79,793
76,014
79,128
75,182
76,284
NA
NA (Not Applicable)
Offshore Production (Recalculation with Updated Data)
Offshore production Cm emission estimates decreased by an average of 2 percent across the time series and
decreased by 3 percent in 2019, compared to the previous Inventory. The emission changes were due to updated
offshore complex counts.
Table 3-61: Offshore Production National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2016
2017
2018
2019
2020
GOM Federal Waters
302,936
219,285
189,595
186,606
182,662
181,724
178,496
GOM State Waters
5,657
665
108
96
60
71
61
Pacific Waters
22,609
17,659
5,008
5,052
3,794
3,370
4,262
Alaska State Waters
21,936
21,191
9,680
12,163
9,834
10,461
10,123
Total Emissions
353,138
258,801
204,391
203,917
196,349
195,626
192,943
Previous Estimate
373,650
260,994
205,958
205,008
199,063
200,720
NA
NA (Not Applicable)
Gas Engines (Recalculation with Updated Data)
Gas engine (combustion slip) Cm emission estimates decreased by an average of 3 percent across the time series
and decreased by 4 percent in 2019, compared to the previous Inventory. The emission changes were due to
updated Enverus well counts. Even though the well counts have increased across the time series, the 2019 gas
engine estimates are calculated using the ratio of 2019 to 1993 well counts. Since the 1993 well counts show a
larger increase (12 percent) compared to the 2019 well counts (5 percent), the gas engine estimates increased.
Table 3-62: Gas Engine National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2016
2017
2018
2019
2020
Total Gas Engine Emissions
Previous Estimate
81,916
87,854
71,348
73,659
89,735
94,771
89,471
94,311
91,693
96,338
93,556
97,828
89,471
NA
NA (Not Applicable)
Energy 3-85
-------
Well Counts (Recalculation with Updated Data)
EPA uses annual producing oil well counts as an input for estimates of emissions from multiple sources in the
Inventory, including exploration well testing, pneumatic controllers, chemical injection pumps, well workovers, and
equipment leaks. Annual well count data are obtained from Enverus for the entire time series during each
Inventory cycle. In addition, well counts for Illinois and Indiana were fully incorporated for this Inventory, based on
information available from state agencies or from EIA. Enverus does not contain well count data for Illinois and
Indiana. There are an average of approximately 25,200 oil wells for Illinois and 15,600 oil wells for Indiana, across
the time series. Annual well counts increased by an average of 7 percent across the 1990 to 2019 time series and
increased by 5 percent in 2019, compared to the previous Inventory.
Table 3-63: National Oil Well Counts
Source
1990
2005
2016
2017
2018
2019
2020
Oil Wells
Previous Estimate
520,364
506,730
482,007
447,683
568,640
543,759
560,258
534,806
553,769
527,544
549,153
521,771
529,419
NA
NA (Not Applicable)
In January 2022, EIA released an updated time series of national oil and gas well counts (covering 2000 through
2020). EIA estimates 936,984 total wells for year 2020. EPA's total well count for 2020 is 939,665. EPA well counts
are higher due to the inclusion of wells for Illinois and Indiana in the current Inventory. EIA does not include wells
for these two states. If these states are excluded from the well count comparison (i.e., well counts are compared
only for the states that are in both EIA and EPA datasets), EPA's well counts are about 2 percent lower than ElA's in
2020, in part due to well definitions. ElA's well counts include side tracks (i.e., secondary wellbore away from
original wellbore in order to bypass unusable formation, explore nearby formations, or other reasons),
completions, and recompletions, and therefore are expected to be higher than EPA's which include only producing
wells. Note, EPA and EIA use a different threshold for distinguishing between oil versus gas wells (EIA uses 6
mcf/bbl, while EPA uses 100 mcf/bbl), which results in EIA having a lower fraction of oil wells (e.g., 44 percent
versus EPA's 56 percent in 2020) and a higher fraction of gas wells (e.g., 56 percent versus EPA's 44 percent in
2020) than EPA.
Transportation
Recalculations for the transportation segment have resulted in an average decrease in calculated Cm and CO2
emissions over the time series from this segment of less than 0.2 percent, compared to the previous Inventory.
Refining
Recalculations due to resubmitted GHGRP data in the refining segment have resulted in an average decrease in
calculated Cm emissions over the time series from this segment of less than 0.1 percent and increased by 0.8
percent in 2019 in the current Inventory, compared to the previous Inventory. Additionally, EPA identified one
refinery that stopped reporting to GHGRP starting in 2013 due to meeting the criteria for cessation of reporting.
EPA used the refinery's 2012 reported annual emissions to gap fill for 2013 through 2020. This resulted in a very
minor increase in refinery Cm emissions compared to the previous Inventory (0.02 percent).
Refining CO2 emission estimates increased by an average of 0.1 percent across the time series and increased by 1
percent in 2019 in the current Inventory, compared to the previous Inventory.
Table 3-64: Refining National CO2 Emissions (kt CO2)
Source
1990
2005
2016
2017
2018
2019
2020
Refining
3,284
3,728
3,994
3,725
3,820
5,080
4,326
Previous Estimate
3,284
3,728
3,991
3,714
3,735
5,019
NA
NA (Not Applicable)
3-86 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Planned Improvements
Upcoming Data, and Additional Data that Could Inform the Inventory
EPA will assess new data received by the Methane Challenge Program 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 stakeholder comments.
EPA also continues to seek new data that could be used to assess or update the estimates in the Inventory. For
example, in recent years, stakeholder comments have highlighted areas where additional data that could inform
the Inventory are currently limited or unavailable:
• Tank measurements and tank and flaring malfunction and control efficiency data.
• Improved equipment leak data (activity and emissions).
• Activity data and emissions data for production facilities that do not report to GHGRP.
• Associated gas venting and flaring data on practices from 1990 through 2010.
• Onshore mud degassing.
• Refineries emissions data.
• Anomalous leak events information throughout the time series and for future years.
EPA received stakeholder feedback through comments on the public review draft of the current Inventory. Several
stakeholders asserted that methane emissions are undercounted in petroleum systems. A stakeholder comment
suggested developing the inventory using a strategy that combines information from satellites, aircraft-based
instruments, and ground-based sensors. Stakeholder feedback on the public review draft recommended use of
updated emission factors for pneumatic controllers.
EPA will continue to seek available data on these and other sources as part of the process to update the Inventory.
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
on the specific application.
In the Inventory, C02that 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-
Energy 3-87
-------
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 2020 and data reported for geologic sequestration from 2016 to
2020.
The amount of CO2 captured and extracted from natural and industrial sites for EOR applications in 2020 is
35,210 kt (35.2 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, and EPA received a public review comment in
support of updating the approach.
Table 3-65: Quantity of CO2 Captured and Extracted for EOR Operations (kt CO2)
Stage
2016
2017
2018
2019
2020
Quantity of C02 Captured and Extracted
for EOR Operations
46,700
49,600
48,400
52,100
35,210
Several facilities are reporting under GHGRP subpart RR (Geologic Sequestration of Carbon Dioxide). See Table
3-66 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 3-66: Geologic Sequestration Information Reported Under GHGRP Subpart RR
Stage
2016
2017
2018
2019
2020
Number of Reporting Facilities
1
3
5
5
6
Reported Annual C02Sequestered (kt)
3,091
5,958
7,662
8,332
6,765
Reported Annual C02 Emissions from
Equipment Leaks (kt)
10
10
11
16
74
3.7 Natural Gas Systems (CRF Source
Category lB2b)
The U.S. natural gas system encompasses hundreds of thousands of wells, hundreds of processing facilities, and
over a million miles of transmission and distribution pipelines. This IPCC category (lB2b) is for fugitive emissions
from natural gas systems, which per IPCC guidelines include emissions from leaks, venting, and flaring. Total
greenhouse gas emissions (CH4, CO2, and N2O) from natural gas systems in 2020 were 200.3 MMT CO2 Eq., a
decrease of 12 percent from 1990 and a decrease of 5 percent from 2019, both primarily due to decreases in CH4
3-88 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
emissions. From 2010, emissions increased by 4 percent, primarily due to increases in CO2 emissions. National total
dry gas production in the United States increased by 88 percent from 1990 to 2020, decreased by 1 percent from
2019 to 2020, and increased by 62 percent from 2009 to 2020. Of the overall greenhouse gas emissions (200.3
MMT CO2 Eq.), 82 percent are Cm emissions (164.9 MMT CO2 Eq.), 18 percent are CO2 emissions (35.4 MMT), and
less than 0.01 percent are N2O emissions (0.01 MMT CO2 Eq.).
Overall, natural gas systems emitted 164.9 MMT CO2 Eq. (6,596 kt CH4) of Cm in 2020, a 16 percent decrease
compared to 1990 emissions, and 4 percent decrease compared to 2019 emissions (see Table 3-68 and Table 3-69).
For non-combustion CO2, a total of 35.4 MMT CO2 Eq. (35,369 kt) was emitted in 2020, a 11 percent increase
compared to 1990 emissions, and a 9 percent decrease compared to 2019 levels. The 2020 N2O emissions were
estimated to be 0.01 MMT CO2 Eq. (0.03 kt N2O), a 105 percent increase compared to 1990 emissions, and a 15
percent decrease compared to 2019 levels.
The 1990 to 2019 emissions trend is not consistent across segments or gases. Overall, the 1990 to 2020 decrease in
Cm emissions is due primarily to the decrease in emissions from the following segments: distribution (70 percent
decrease), transmission and storage (29 percent decrease), processing (42 percent decrease), and exploration (93
percent decrease). Over the same time period, the production segment saw increased CH4 emissions of 43 percent
(with onshore production emissions increasing 24 percent, offshore production emissions decreasing 77 percent,
and gathering and boosting [G&B] emissions increasing 103 percent), and post-meter emissions increasing by 58
percent. The 1990 to 2020 increase in CO2 emissions is primarily due to an increase in CO2 emissions in the
production segment, where emissions from flaring have increased over time.
Methane and CO2 emissions from natural gas systems include those resulting from normal operations, routine
maintenance, and system upsets. Emissions from normal operations include natural gas engine and turbine
uncombusted exhaust, flaring, and leak emissions from system components. Routine maintenance emissions
originate from pipelines, equipment, and wells during repair and maintenance activities. Pressure surge relief
systems and accidents can lead to system upset emissions. Emissions of N2O from flaring activities are included in
the Inventory, with most of the emissions occurring in the processing and production segments. Note, CO2
emissions exclude all combustion emissions (e.g., engine combustion) except for flaring CO2 emissions. All
combustion CO2 emissions (except for flaring) are accounted for in Section 3.1 CO2 from Fossil Fuel Combustion.
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 2019) to
ensure that the trend is representative of changes in emissions. Recalculations in natural gas systems in this year's
Inventory include:
• Incorporation of an estimate for post-meter emissions
• Incorporation of estimates for large anomalous leak events
• Updated GasSTAR and Methane Challenge data
• Updated activity data for underground storage wells
• Updates to well counts using the most recent data from Enverus
• Recalculations due to Greenhouse Gas Reporting Program (GHGRP) submission revisions
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 completions. Emissions from exploration accounted for
less than 1 percent of CH4 emissions and of CO2 emissions from natural gas systems in 2020. Well completions
accounted for approximately 90 percent of CH4 emissions from the exploration segment in 2020, with the rest
resulting from well testing and drilling. Flaring emissions account for most of the CO2 emissions. Methane
emissions from exploration decreased by 93 percent from 1990 to 2020, with the largest decreases coming from
hydraulically fractured gas well completions without reduced emissions completions (RECs). Methane emissions
decreased 89 percent from 2019 to 2020 due to decreases in emissions from hydraulically fractured well
Energy 3-89
-------
completions with RECs and venting. Methane emissions were highest from 2005 to 2008. Carbon dioxide emissions
from exploration decreased by 68 percent from 1990 to 2020 and decreased 57 percent from 2019 to 2020 due to
decreases in hydraulically fractured gas well completions. Carbon dioxide emissions were highest from 2006 to
2008. Nitrous oxide emissions decreased 87 percent from 1990 to 2020 and decreased 57 percent from 2019 to
2020.
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 Cm emissions and 22
percent of CO2 emissions from natural gas systems in 2020. Emissions from gathering and boosting and pneumatic
controllers in onshore production accounted for most of the production segment CH4 emissions in 2020. Within
gathering and boosting, the largest sources of CH4 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 156 percent from 1990 to 2020, 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 7 percent from 2019 to 2020 due to decreases in emissions from pneumatic
controllers and from tanks in gathering and boosting. Carbon dioxide emissions from production increased by
approximately a factor of 2.6 from 1990 to 2020 due to increases in emissions at flare stacks in gathering and
boosting and miscellaneous onshore production flaring, and decreased 29 percent from 2019 to 2020 due
primarily to decreases in emissions from flare stacks and dehydrator vents at gathering and boosting stations .
Nitrous oxide emissions decreased less than lpercent from 1990 to 2020 and decreased 23 percent from 2019 to
2020. The decrease in N2O emissions from 1990 to 2020 and from 2018 to 2020 is primarily due to decreases in
emissions from flare stacks 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 72 percent of CO2 emissions from natural gas systems.
Methane emissions from processing decreased by 42 percent from 1990 to 2020 as emissions from compressors
(leaks and venting) and equipment leaks decreased; and decreased 2 percent from 2018 to 2020 due to decreased
emissions from centrifugal compressors. Carbon dioxide emissions from processing decreased by 10 percent from
1990 to 2020, due to a decrease in AGR emissions, and decreased 3 percent from 2019 to 2020 due to decreased
emissions from reciprocating compressors. Nitrous oxide emissions increased 39 percent from 2018 to 2019.
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 CH4 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
3-90 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
emissions from natural gas systems, while CO2 emissions from transmission and storage accounted for 6 percent of
the CO2 emissions from natural gas systems. CH4emissions from this source decreased by 29 percent from 1990 to
2020 due to reduced compressor station emissions (including emissions from compressors and leaks) and
increased 3 percent from 2019 to 2020 due to increased emissions from transmission compressors. CO2 emissions
from transmission and storage were 11.3 times higher in 2020 than in 1990, due to increased emissions from LNG
export terminals and LNG stations, and increased by 64 percent from 2019 to 2020, also due to LNG export
terminals. The quantity of LNG exported from the United States increased by a factor of 45 from 1990 to 2020, and
by 31 percent from 2019 to 2020. LNG emissions are about 1 percent of CH4 and 86 percent of CO2 emissions from
transmission and storage in year 2020. Nitrous oxide emissions from transmission and storage increased by 317
percent from 1990 to 2020 and increased 70 percent from 2019 to 2020.
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 1,316,800 miles of distribution mains in 2020, an increase of 372,643 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 2020 were 70 percent lower than 1990 levels and less than 1 percent lower than 2019
emissions. Distribution system CO2 emissions in 2020 were 69 percent lower than 1990 levels and less than 1
percent lower than 2019 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 8 percent of emissions from natural gas systems in 2020. Post-
meter Cm emissions increased by 58 percent from 1990 to 2020 and increased by 1 percent from 2019 to 2020,
due to increases in the number of residential houses using natural gas and increased natural gas 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 3-67. Total CH4 emissions for these same segments of natural gas systems are shown in MMT CO2 Eq.
(Table 3-68) and kt (Table 3-69). 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 2020, 2.6 MMT CO2 Eq. CH4 is subtracted from production segment emissions, 4.0 MMT CO2 Eq. CH4 is
subtracted from the transmission and storage segment, and 0.1 MMT CO2 Eq. CH4 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 3-67: Total Greenhouse Gas Emissions (CH4, CO2, and N2O) from Natural Gas Systems
(MMT COz Eq.)
Stage
1990
2005
2016
2017
2018
2019
2020
Exploration
3.3
10.4
0.9
1.7
2.7
2.1
0.3
Production
64.3
87.9
97.6
99.8
102.3
103.6
94.1
Processing
49.7
30.4
33.0
34.5
35.1
39.0
37.8
Transmission and Storage
57.4
39.7
38.7
37
39.0
40.8
42.7
Distribution
45.5
25.5
14.2
14.1
14.0
13.9
13.9
Post-Meter
7.2
8.6
10.7
10.6
11.1
11.4
11.5
Total
227.4
202.5
195.06
197.7
204.2
210.9
200.3
Note: Totals may not sum due to independent rounding.
Energy 3-91
-------
Table 3-68: ChU Emissions from Natural Gas Systems (MMT CO2 Eq.)
Stage
1990
2005
2016
2017
2018
2019
2020
Exploration
3.0
9.0
0.7
1.2
2.3
1.9
0.2
Production
61.3
83.4
90.1
92.6
93.8
92.8
86.4
Onshore Production
37.5
56.8
52.6
53.4
54.3
51.6
47.6
Gathering and Boosting
18.5
24.0
36.4
38.3
38.7
39.9
37.5
Offshore Production
4.3
1.8
0.8
0.7
0.8
0.8
1.0
Processing
21.3
11.6
11.2
11.5
12.1
12.6
12.4
Transmission and Storage
57.2
39.5
38.3
36.5
38.4
39.6
40.6
Distribution
45.5
25.5
14.2
14.1
14.0
13.9
13.9
Post-Meter
7.2
8.6
10.7
10.6
11.1
11.4
11.5
Total
195.5
177.5
165.2
166.6
171.8
172.1
164.9
Note: Totals may not sum due to independent rounding.
Table 3-69: ChU Emissions from Natural Gas Systems (kt)
Stage
1990
2005
2016
2017
2018
2019
2020
Exploration
119
358
27
49
94
75
8
Production
2,450
3,336
3,605
3,705
3,753
3,710
3,455
Onshore Production
1,542
2305
2,115
2,145
2,174
2,085
1,916
Gathering and Boosting
739
958
1,457
1,533
1,548
1,595
1,500
Offshore Production
170
73
32
26
31
30
39
Processing
853
463
447
460
483
505
494
Transmission and Storage
2,288
1,580
1,534
1,460
1,538
1,583
1,625
Distribution
1,819
1,018
569
564
559
555
554
Post-Meter
290
344
426
424
445
457
459
Total
7,819
7,100
6,609
6,662
6,871
6,885
6,596
Note: Totals may not sum due to independent rounding.
Table 3-70: CO2 Emissions from Natural Gas Systems (MMT)
Stage
1990
2005
2016
2017
2018
2019
2020
Exploration
0.3
1.4
0.2
0.4
0.3
0.2
0.1
Production
3.0
4.5
7.4
7.2
8.5
10.9
7.7
Processing
28.3
18.8
21.8
23.0
23.0
26.4
25.5
Transmission and Storage
0.2
0.2
0.3
0.5
0.5
1.2
2.0
Distribution
0.1
+
+
+
+
+
+
Post-Meter
+
+
+
+
+
+
+
Total
31.9
24.9
29.8
31.1
32.4
38.7
35.4
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Table 3-71: CO2 Emissions from Natural Gas Systems (kt)
Stage
1990
2005
2016
2017
2018
2019
2020
Exploration
297
1,434
190
444
336
220
95
Production
3,024
4,468
7,444
7,194
8,503
10,885
7,736
Processing
28,338
18,836
21,787
22,988
23,001
26,373
25,468
Transmission and Storage
180
176
340
499
547
1,242
2,036
Distribution
54
30
17
17
17
16
16
Post-Meter
1
1
2
2
2
2
2
Total
31,894
24,945
29,780
31,145
32,407
38,740
35,353
NE (Not Estimated)
Note: Totals may not sum due to independent rounding.
3-92 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 3-72: N2O Emissions from Natural Gas Systems (Metric Tons CO2 Eq.)
Stage
1990
2005
2016
2017
2018
2019
2020
Exploration
399
1,225
113
244
176
116
50
Production
4,318
5,795
8,889
4,306
4,669
5,585
4,293
Processing
NO
3,348
3,732
2,975
3,372
5,689
4,765
Transmission and Storage
257
309
382
462
234
630
1,070
Distribution
NO
NO
NO
NO
NO
NO
NO
Post-Meter
NO
NO
NO
NO
NO
NO
NO
Total
4,974
10,676
13,116
7,987
8,451
12,020
10,178
NO (Not Occurring)
Note: Totals may not sum due to independent rounding.
Table 3-73: N2O Emissions from Natural Gas Systems (Metric Tons N2O)
Stage
1990
2005
2016
2017
2018
2019
2020
Exploration
1.3
4.1
0.4
0.8
0.6
0.4
0.2
Production
14.5
19.4
29.8
14.4
15.7
18.7
14.4
Processing
NO
11.2
12.5
10.0
11.3
19.1
16.0
Transmission and Storage
0.9
1.0
1.3
1.6
0.8
2.1
3.6
Distribution
NO
NO
NO
NO
NO
NO
NO
Post-Meter
NO
NO
NO
NO
NO
NO
NO
Total
16.7
35.8
44.0
26.8
28.4
40.3
34.2
NO (Not Occurring)
Note: Totals may not sum due to independent rounding.
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 CH4, 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
2021a), 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.
Energy 3-93
-------
GHGRP Subpart W data were used to develop Cm, 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). CO2 emissions from post-
meter residential sources are included in fossil fuel combustion data.
Flaring N2O emissions were estimated for flaring sources using GHGRP data.
See Annex 3.6 for more detailed information on the methodology and data used to calculate CH4, CO2, and N2O
emissions from natural gas systems.
Activity Data. Activity data were taken from various published data sets, as detailed in Annex 3.6. Key activity data
sources include data sets developed and maintained by EPA's GHGRP (EPA 2020); Enverus (Enverus 2021); BOEM;
Federal Energy Regulatory Commission (FERC); EIA; the Natural Gas STAR and Methane Challenge Programs annual
data; Oil and Gas Journal; and PHMSA.
For a few sources, recent direct activity data are not available. For these sources, either 2019 data were used as a
proxy for 2020 data, or a set of industry activity data drivers was developed and used to calculate activity data over
the time series. Drivers include statistics on gas production, number of wells, system throughput, miles of various
kinds of pipe, and other statistics that characterize the changes in the U.S. natural gas system infrastructure and
operations. More information on activity data and drivers is available in Annex 3.6.
A complete list of references for emission factors and activity data by emission source is provided in Annex 3.6.
Calculating Net Emissions. For most sources, net emissions are calculated directly by applying emission factors to
activity data. Emission factors used in net emission approaches reflect technology-specific information, and take
into account regulatory and voluntary reductions. However, for production, transmission and storage, and
distribution, some sources are calculated using potential emission factors, and CH4 that is not emitted is deducted
from the total CH4 potential estimates. To take into account use of such technologies and practices that result in
lower emissions but are not reflected in "potential" emission factors, data are collected on both regulatory and
voluntary reductions. Regulatory actions addressed using this method include EPA National Emission Standards for
Hazardous Air Pollutants (NESHAP) regulations for dehydrator vents. Voluntary reductions included in the
Inventory are those reported to Natural Gas STAR and Methane Challenge for certain sources. Natural Gas STAR
and Methane Challenge reductions were reassessed for this Inventory, see the Recalculations Discussion for more
information.
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2020. GHGRP data available (starting in 2011) and other recent data sources have improved estimates of
3-94 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
emissions from natural gas systems. 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 2010 by interpolating activity data or emission factors or both between 1992 and 2011 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 detailed documentation of methodologies, please see Annex 3.5.
Through EPA's stakeholder process on oil and gas in the Inventory, EPA received stakeholder feedback on updates
under consideration for the Inventory. Stakeholder feedback is noted below in Recalculations Discussion and
Planned Improvements.
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 CFU
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).
Uncertainty
EPA has conducted a quantitative uncertainty analysis using the IPCC Approach 2 methodology (Monte Carlo
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.7S
EPA used Microsoft Excel's @ 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 3-74. Natural gas systems CH4 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
78 See https://www.epa.gov/ghgemissions/natural-gas-and-petroleum-systems.
Energy 3-95
-------
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 3-74: Approach 2 Quantitative Uncertainty Estimates for CH4 and Non-combustion CO2
Emissions from Natural Gas Systems (MMT CO2 Eq. and Percent)
Source
Gas
2020 Emission Estimate
(MMT CO? Eq.)b
Uncertainty Range Relative to Emission Estimate3
(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
n2o
+
+
+
-16%
+19%
+ Less than 0.05 MMT C02 Eq.
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 3-68 and Table 3-69.
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.79
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 2021. 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,
79 See https://www.epa.gov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
3-96 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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.80 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.81
An updated version of the gridded inventory is being developed and will improve efforts to compare results of the
Inventory with atmospheric studies.
Recalculations Discussion
EPA received information and data related to the emission estimates through GHGRP reporting, the annual
Inventory formal public notice periods, stakeholder feedback on updates under consideration, and new studies. In
September 2021, EPA released draft memoranda that discussed changes under consideration, and requested
stakeholder feedback on those changes. EPA then updated the memoranda to document the methodology
implemented in the current Inventory.82 Memoranda cited in the Recalculations Discussion below are: Inventory of
U.S. Greenhouse Gas Emissions and Sinks 1990-2020: Updates for Anomalous Events Including Well Blowout and
Well Release Emissions (Anomalous Events memo), Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-
2020: Updates for Activity Data (Activity Data memo), Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-
2020: Updates for Gas STAR and Methane Challenge Reductions (Reductions memo), and Inventory of U.S.
Greenhouse Gas Emissions and Sinks 1990-2020: Updates for Post-Meter Emissions (Post-Meter memo).
EPA thoroughly evaluated relevant information available and made several updates to the Inventory, including
adding well blowout emissions, using PHMSA data to update underground storage well counts, reassessing the Gas
STAR reductions data and incorporating Methane Challenge data, and incorporating post-meter emissions. These
changes are discussed in detail below. In addition, certain sources did not undergo methodological updates, but
Cm and/or CO2 emissions changed by greater than 0.05 MMT CO2 Eq., comparing the previous estimate for 2019
to the current (recalculated) estimate for 2019. For sources without methodological updates, the emissions
changes were mostly due to GHGRP data submission revisions and updates to well counts in the Enverus dataset.
The combined impact of revisions to 2019 natural gas systems CH4 emissions, compared to the previous Inventory,
is an increase from 167.7 to 178.4 MMT CO2 Eq. (10.7 MMT CO2 Eq., or 6 percent). The recalculations resulted in
an average increase in the annual CH4 emission estimates across the 1990 through 2019 time series, compared to
the previous Inventory, of 13.2 MMT CO2 Eq., or 8.1 percent.
The combined impact of revisions to 2019 natural gas systems CO2 emissions, compared to the previous Inventory,
is an increase from 37.2 MMT to 38.7 MMT, or 4 percent. The recalculations resulted in an average decrease in
80 See https://www.epa.gov/eheemissions/eridded-2012-methane-emissions.
81 See https://www.epa.eov/eheemissions/us-greenhouse-eas-inventory-report-1990-2014.
82 Stakeholder materials including draft and final memoranda for the current (i.e., 1990 to 2020) Inventory are available at
https://www.epa.gov/ehgemissions/natural-eas-and-petroleum-svstems.
Energy 3-97
-------
emission estimates across the 1990 through 2019 time series, compared to the previous Inventory, of 0.1 MMT
CO2 Eq„ or 0.3 percent.
The combined impact of revisions to 2019 natural gas systems N2O emissions, compared to the previous Inventory,
is an increase from 11.3 kt CO2 Eq. to 12.0 kt CO2 Eq., or 6 percent. The recalculations resulted in an average
decrease in emission estimates across the 1990 through 2019 time series, compared to the previous Inventory, of
1 percent.
In Table 3-75 and Table 3-76 below are categories in Natural Gas Systems with recalculations resulting in a change
of greater than 0.05 MMT CO2 Eq., comparing the previous estimate for 2019 to the current (recalculated)
estimate for 2019. No changes made to N2O estimates resulted in a change greater than 0.05 MMT CO2 Eq. For
more information, please see the Recalculations Discussion below.
Table 3-75: Recalculations of CO2 in Natural Gas Systems (MMT CO2)
Segment and Emission Sources with
Previous Estimate
Current Estimate
Current Estimate
Changes of Greater than 0.05 MMT C02
Year 2019,
Year 2019,
Year 2020,
due to Recalculations
2021 Inventory
2022 Inventory
2022 Inventory
Exploration
0.2
0.2
0.1
Production
11.0
10.9
7.7
Misc. Onshore Production Flaring
1.8
1.9
1.1
Large Tanks with Flares
0.7
0.6
0.6
Processing
24.8
26.4
25.5
Flares
8.3
9.8
7.9
Transmission and Storage
1.2
1.2
2.0
Distribution
+
+
+
Post-Meter
+
+
+
Total
37.2
38.7
35.4
+ Does not exceed 0.05 MMT C02.
)ble 3-76: 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
2019,
2021 Inventory
Current Estimate
Year 2019, 2022
Inventory
Current Estimate
Year 2020, 2022
Inventory
Exploration
0.5
1.9
0.2
Well Blowouts
0.0
1.3
0.0
Production
97.1
95.0
86.4
Produced Water
4.7
4.0
3.5
Pneumatic Controllers
28.2
25.6
23.8
Gas Engines
6.3
5.8
5.7
Miscellaneous Onshore Flaring
0.2
0.2
0.1
Small Tanks w/o Flares
0.5
0.5
0.3
G&B Station Sources
40.9
39.9
37.5
Gathering Pipeline Leaks
2.8
2.9
3.2
Gathering Pipeline Blowdowns
0.8
0.2
0.2
Processing
12.4
12.6
12.4
Flares
0.9
1.1
0.9
Transmission and Storage
43.7
43.4
40.6
Reciprocating Compressors (Transmission)
10.2
10.2
10.5
Wells (Storage)
0.4
0.3
0.3
Pipeline Venting
5.0
4.7
5.5
Distribution
14.0
13.9
13.9
Post-Meter
NA
11.4
11.5
Total
167.7
178.4
164.9
3-98 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
NA (Not Applicable)
Exploration
Well Blowouts (Methodological Update)
EPA added estimates for well blowout emissions into the Inventory for three discrete well blowout events, using
emission estimates calculated in Pandey et al. (2019), Cusworth et al. (2021), and Maasakkers et al. 2022).
Pandey et al. (2019) calculated emissions from a 20-day well blowout in Ohio occurring in February to March 2018
using data collected from Tropospheric Monitoring Instrument (TROPOMI).
Cusworth et al. (2021) calculated emissions from a 20-day well blowout (starting November 1, 2019) in Texas using
data collected from a combination of satellite instruments including TROPOMI, GHGSat-D, the Visible Infrared
Imaging Radiometer Suite (VIIRS) instrument, and the PRecursore IperSpettrale della Missione Applicativa
(PRISMA) satellite imaging spectrometer.
Maasakkers et al. (2022) calculated emissions from a 38-day well blowout (starting August 30, 2019) in Louisiana
using data collected from TROPOMI and VIIRS.
The Anomalous Events memo contains additional information on this update.
Table 3-77: Well Blowout National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2016
2017
2018
2019
2020
Gas Well Blowout
NO
NO
NO
NO
60,000
53,800
NO
Previous Estimate
NA
NA
NA
NA
NA
NA
NA
NA (Not Applicable)
NO (Not Occurring)
Production
Produced Water (Recalculation with Updated Data)
Produced water Cm emissions increased by an average of 16 percent across the 1990 to 2019 time series and
decreased by 14 percent in 2019, compared to the previous Inventory. These changes were due to updates to the
handling of Enverus data and NEI's O&G Tool data for six states (IL, IN, KS, PA, OK, and WV). The largest changes
occurred earlier in the time series (e.g., 1990 to 1999), where the estimate of the annual volume of produced
water increased by an average of 41 percent over the previous estimate. This change was primarily due to revised
data available from the NEI for OK. The revised NEI data were obtained for 2002, 2005, 2008, 2011, 2014, and
2016-2020 (EPA 2021f). For the missing years in the time-series, EPA estimated state-level produced water
volumes (for IL, IN, KS, PA, OK, and WV) using the average ratio of produced water to gas production calculated for
2002, 2005, 2008, 2011, and 2014. Revised produced water data for the remaining states is from Enverus (Enverus
2021).
Table 3-78: Produced Water National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2016
2017
2018
2019
2020
Gas Well Produced Water
Previous Estimate
121,867
82,250
153,709
139,453
142,777
154,394
145,965
157,488
150,073
188,601
160,548
187,070
131,322
NA
NA (Not Applicable)
Energy 3-99
-------
Pneumatic Controllers (Recalculation with Updated Data)
Pneumatic controller Cm emission estimates decreased by an average of 1.5 percent across the 1990 to 2019 time
series and decreased by 9 percent in 2019, compared to the previous Inventory. These changes were due to
GHGRP submission revisions.
Table 3-79: Production Segment Pneumatic Controller National Emissions (Metric Tons ChU)
Source
1990
2005
2016
2017
2018
2019
2020
Low Bleed
0
23,565
32,794
36,102
35,069
33,089
32,224
High Bleed
308,908
471,540
107,928
113,112
92,941
72,923
43,096
Intermittent Bleed
201,446
546,397
928,445
955,682
944,864
918,666
875,399
Total Emissions
510,354
1,041,503
1,069,168
1,104,896
1,072,874
1,024,678
950,718
Previous Estimate
482,334
1,062,685
1,063,791
1,103,082
1,075,645
1,126,531
NA
NA (Not Applicable)
Gas Engines (Recalculation with Updated Data)
Gas engine (combustion slip) Cm emissions increased by an average of 37 percent across the 1990 to 2019 time
series and increased by 81 percent in 2019, compared to the previous Inventory. These changes were due to
updates to well counts in the Enverus dataset.
Table 3-80: Gas Engine National Emissions (Metric Tons ChU)
Source
1990
2005
2016
2017
2018
2019
2020
Gas Engines
Previous Estimate
115,689
116,684
198,004
129,715
214,661
124,835
197,218
120,272
207,051
116,437
202,052
111,886
197,074
NA
NA (Not Applicable)
Miscellaneous Production Flaring (Recalculation with Updated Data)
Miscellaneous production flaring Cm emissions increased by an average of 6 percent across the 1990 to 2019 time
series and increased by 8 percent in 2019, compared to the previous Inventory. CO2 emissions for this source
increased across the 1990 to 2019 time series by an average of 1 percent and increased by 5 percent in 2019.
These changes were due to GHGRP submission revisions.
Table 3-81: Miscellaneous Production Flaring National Emissions (Metric Tons ChU)
Source
1990
2005
2016
2017
2018
2019
2020
Miscellaneous Flaring-Gulf
Coast Basin
NO
547
649
524
401
1,268
939
Miscellaneous Flaring-
Williston Basin
NO
+
+
107
65
9
30
Miscellaneous Flaring-
Permian Basin
NO
1,354
2,315
3,539
2,911
5,096
2,946
Miscellaneous Flaring-
Other Basins
NO
557
1,937
1,414
1,587
1,791
980
Total Emissions
NO
2,458
4,902
5,584
4,964
8,164
4,894
Previous Estimate
NO
2,269
4,849
5,552
5,029
7,680
NA
+ Does not exceed 0.5 metric tons.
NO (Not Occurring)
NA (Not Applicable)
3-100 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 3-82: Miscellaneous Production Flaring National Emissions (kt CO2)
Source
1990
2005
2016
2017
2018
2019
2020
Miscellaneous Flaring-
Gulf Coast Basin
NO
166
234
209
137
399
251
Miscellaneous Flaring-
Williston Basin
NO
+
+
10
6
4
4
Miscellaneous Flaring-
Permian Basin
NO
260
500
622
707
1,159
591
Miscellaneous Flaring-
Other Basins
NO
117
427
304
493
342
213
Total Emissions
NO
543
1,161
1,145
1,344
1,904
1,060
Previous Estimate
NO
543
1,162
1,152
1,388
1,820
NA
+ Does not exceed 0.5 kt.
NO (Not Occurring)
NA (Not Applicable)
Production Storage Tanks (Recalculation with Updated Data)
Methane emissions for small production storage tanks without flares decreased by an average of 0.5 percent
across the 1990 to 2019 time series and decreased by 15 percent in 2019, compared to the previous Inventory. The
large production storage tank with flares CO2 emissions estimate decreased by an average of 1 percent across the
time series and by 21 percent in 2019, compared to the previous Inventory. These changes were due to GHGRP
submission revisions.
Table 3-83: Production Storage Tanks National Emissions (Metric Tons ChU)
Source
1990
2005
2016
2017
2018
2019
2020
Small Tanks w/o Flares
Previous Estimate
10,144
10,180
7,760
7,763
22,584
22,520
16,123
16,013
16,753
16,668
18,591
21,951
13,613
NA
NA (Not Applicable)
Table 3-84: Production Storage Tanks National Emissions (kt CO2)
Source
1990
2005
2016
2017
2018
2019
2020
Large Tanks w/ Flares
292
367
1,107
1,085
779
573
552
Previous Estimate
293
369
1,114
1,090
781
723
NA
NA (Not Applicable)
Gathering and Boosting (G&B) Stations (Recalculation with Updated Data)
Methane emission estimates for sources at gathering and boosting stations decreased in the current Inventory by
less than 0.1 percent across the time series and decreased by 1 percent in 2019, compared to the previous
Inventory. The G&B sources with the largest decrease in CH4 emissions estimates for year 2019, compared to the
previous Inventory, are compressors (decrease of 3.5 kt, or 1 percent), gas engines (decrease of 4.8 kt, or 1
percent), and station blowdowns (decrease of 25 kt or 36 percent). Intermittent bleed pneumatic device CH4
emissions increased by 11 kt, or 6 percent in 2019, compared to the previous Inventory. These changes were due
to GHGRP submission revisions.
Table 3-85: Gathering Stations Sources National ChU Emissions (Metric Tons ChU)
Source 1990 2005 2016 2017 2018 2019 2020
Compressors 130,165 165,664 261,677 280,355 298,220 305,896 306,935
Station Blowdowns 20,517 26,113 41,247 63,852 78,548 43,865 44,881
Energy 3-101
-------
Intermittent Bleed
Pneumatic Devices
Gas Engines
Other Gathering Sources
Total Emissions
Previous Estimate
79,716 101,456 160,351 191,528 173,811 181,860 172,429
172,279 219,263 346,340 371,406 395,047 405,617 407,130
245,501 312,455 493,474 470,489 466,178 532,944 432,570
648,179 824,951 1,303,088 1,377,631 1,411,804 1,470,183 1,363,946
652,538 823,648 1,299,276 1,359,628 1,398,994 1,491,704 NA
NA (Not Applicable)
Gathering Pipeline Leaks
Gathering pipeline leak Cm emissions estimates increased by an average of 0.2 percent across the 1990 to 2019
time series and increased by 3 percent in 2019, compared to the previous inventory. The emission changes were
due to GHGRP submission revisions.
Table 3-86: Gathering Pipeline Leak National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2016
2017
2018
2019
2020
Pipeline Leaks
Previous Estimate
81,659
78,046
120,311
120,280
139,170
138,645
135,940
141,873
119,890
116,590
116,470
112,881
126,661
NA
NA (Not Applicable)
Gathering Pipeline Blowdowns
Gathering pipeline blowdowns Cm emissions estimates decreased by an average of 0.1 percent across the 1990 to
2019 time series and decreased by 73 percent in 2019, compared to the previous inventory. The emission changes
were due to GHGRP submission revisions.
Table 3-87: Gathering Pipeline Blowdowns National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2016
2017
2018
2019
2020
Pipeline Blowdowns
8,841
13,026
15,068
19,777
16,060
8,377
9,390
Previous Estimate
8,482
13,072
15,068
19,777
16,060
31,477
NA
NA (Not Applicable)
Well Counts (Recalculation with Updated Data)
EPA uses annual producing gas well counts as an input for estimates of emissions from multiple sources in the
Inventory, including exploration well testing, pneumatic controllers, chemical injection pumps, well workovers, and
equipment leaks. Annual well count data are obtained from Enverus for the entire time series during each
Inventory cycle. In addition, well counts for Illinois and Indiana were more fully incorporated for this Inventory,
based on information available from state agencies or from EIA. There are an average of 400 gas wells for Illinois
and 1,500 gas wells for Indiana, across the time series. Annual gas well counts increased by an average of 1 percent
across the 1990 to 2019 time series and by 1 percent in 2019, compared to the previous Inventory.
Table 3-88: National Gas Well Counts
Source
1990
2005
2016
2017
2018
2019
2020
Gas Wells
Previous Estimate
193,344
185,141
351,129
351,982
432,952
429,697
429,952
427,046
426,372
424,507
420,439
417,507
410,246
NA
NA (Not Applicable)
In January 2022, EIA released an updated time series of national oil and gas well counts (covering 2000 through
2020). EIA estimates 936,984 total wells for year 2020. EPA's total well count for 2020 is 939,665. EPA well counts
are higher due to the inclusion of wells for Illinois and Indiana in the current Inventory. EIA does not include wells
for these two states. If these states are excluded from the well count comparison (i.e., well counts are compared
only for the states that are in both EIA and EPA datasets), EPA's well counts are about 2 percent lower than ElA's in
3-102 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
2020, in part due to well definitions. ElA's well counts include side tracks (i.e., secondary wellbore away from
original wellbore in order to bypass unusable formation, explore nearby formations, or other reasons),
completions, and recompletions, and therefore are expected to be higher than EPA's which include only producing
wells. Note, EPA and EIA use a different threshold for distinguishing between oil versus gas wells (EIA uses 6
mcf/bbl, while EPA uses 100 mcf/bbl), which results in EIA having a lower fraction of oil wells (e.g., 44 percent
versus EPA's 56 percent in 2020) and a higher fraction of gas wells (e.g., 56 percent versus EPA's 44 percent in
2020) than EPA.
Processing
Flares (Recalculation with Updated Data)
Processing segment flare CO2 emission estimates increased by an average of less than 1 percent across the 1993 to
2019 time series and increased by 19 percent for 2019, compared to the previous Inventory. Processing segment
flare Cm emission estimates increased by nearly 3 percent across the 2011 to 2019 time series and by 24 percent
for 2019, compared to the previous Inventory. These changes were due to GHGRP submission revisions.
Table 3-89: Processing Segment Flares National CO2 Emissions (kt CO2)
Source 1990 2005 2016 2017 2018 2019 2020
Flares NO 3,517 5,123 5,590 6,176 9,837 7,879
Previous Estimate NO 3,517 5,246 5,726 6,394 8,257 NA
NA (Not Applicable)
NO (Not Occurring)
Table 3-90: Processing Segment Flares National ChU Emissions (Metric Tons ChU)
Source 1990 2005 2016 2017 2018 2019 2020
Flares NO NA 20,199 24,533 24,195 43,518 35,704
Previous Estimate NO NA 20,239 24,498 24,373 35,147 NA_
NA (Not Applicable)
NO (Not Occurring)
Transmission and Storage
Underground Storage Well Leaks (Methodological Update)
EPA updated the methodology for underground storage well leaks to use storage well count data from PHMSA
(PHMSA 2021b). The PHMSA storage well data were identified by stakeholders during the stakeholder process for
the previous Inventory. The Activity Data memo presents considerations for this update. PHMSA storage well
counts are used for 2017 forward, storage well counts for 1990 to 1992 are retained from the previous Inventory
methodology, and linear interpolation is applied from the 1992 to 2017 values to estimate intermediate years.
Underground storage well leak CH4 emission estimates decreased by an average of 11 percent for the 1990 to 2019
time series and by 27 percent in 2019, compared to the previous Inventory.
Table 3-91: Underground Storage Well Leak National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2016
2017
2018
2019
2020
Storage Well Leaks
Previous Estimate
13,565
13,565
12,295
14,910
32,891
34,716
11,483
13,632
11,434
15,439
11,326
15,495
11,255
NA
NA (Not Applicable)
Transmission Station Reciprocating Compressors (Recalculation with Updated Data)
Energy 3-103
-------
Methane emission estimates from reciprocating compressors at transmission compressor stations increased by an
average of 0.2 percent for 2011 to 2019, compared to the previous Inventory. This increase in the Cm emission
estimates was due to GHGRP submission revisions.
Table 3-92: Transmission Station Reciprocating Compressors National ChU Emissions (Metric
Tons CH4)
Source
1990
2005
2016
2017
2018
2019
2020
Transmission Station -
Reciprocating Compressors
NA
NA
347,178
349,784
375,187
409,709
419,480
Previous Estimate
NA
NA
345,224
347,830
373,233
406,453
NA
NA (Not Applicable)
Transmission Pipeline Venting (Recalculation with Updated Data)
Pipeline venting Cm emissions estimates increased by an average of 0.5 percent across the 2011 to 2019 time
series and decreased by 6 percent in 2019, compared to the previous Inventory. The emission changes were due to
GHGRP submission revisions.
Table 3-93: Transmission Pipeline Venting National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2016
2017
2018
2019
2020
Pipeline Venting
Previous Estimate
177,951
177,951
183,159
183,159
249,933
250,153
200,542
185,003
208,438
185,050
187,268
199,370
221,278
NA
NA (Not Applicable)
Distribution
There were no methodological updates to the distribution segment, and recalculations due to updated data
resulted in average decreases in calculated Cm and CO2 emissions over the time series of less than 1 percent.
Natural Gas STAR and Methane Challenge Reductions
EPA has reassessed the voluntary emission reductions reported under the Natural Gas STAR and Methane
Challenge programs for this Inventory. The latest reported data were paired with sources in the Inventory that use
potential emissions approaches and incorporated into the estimates (e.g., gas engines). In recent years, the
Inventory used 2013 Gas STAR reductions data for all years from 2013 forward. The Reductions memo provides the
full considerations for this update. As in previous Inventories, reductions data are only included in the Inventory if
the emission source uses "potential" emission factors, and for Natural Gas STAR reductions, short-term emission
reductions are assigned to the reported year only, while long-term emission reductions are assigned to the
reported year and every subsequent year in the time series. Voluntary emission reductions decreased by an
average of 55 percent across the 1990 to 2019 time series, compared to the previous Inventory.
In reviewing calculated net emissions on a source-by-source basis, it was determined that the updated
incorporation of voluntary program reductions data resulted in calculated negative emissions (i.e., the absolute
value of the reductions is greater than the potential emissions from that source) for certain sources in some years.
The sources with calculated negative net emissions (and years of negative emissions) include:
• Production segment
o Compressor blowdowns (2001-2020)
o Compressor starts (1994-2020)
o Dehydrator vents (2010-2011)
• Transmission segment
o Dehydrator vents (1997-2020)
o Pipeline leaks (1998-1999, 2007-2012, 2014, 2017-2018)
3-104 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
• Distribution segment
o Pipeline blowdowns (1997, 2005-2006)
o PRV releases (2002)
EPA removed Gas STAR reductions entirely for sources with more than ten years of negative calculated emissions
(production segment compressor blowdowns and compressor starts, and transmission segment dehydrator vents
and pipeline leaks). For the remaining sources with negative emissions (production segment dehydrator vents and
distribution segment pipeline blowdowns and PRV releases), calculated negative emissions occur for a maximum
of three years in the time series. EPA replaced the negative net emissions value with zero for the years of negative
net emissions for these sources.
In addition, as in previous Inventories, EPA has removed the reductions for years 1990 to 1992 as those are already
considered to be included in current emission factors.
Table 3-94: Natural Gas STAR and Methane Challenge Emission Reductions (Metric Tons ChU
Reduction)
Source
1990
2005
2016
2017
2018
2019
2020
Production
NA
71,220
88,780
100,364
82,782
84,380
84,380
Transmission and Storage
NA
72,856
115,408
123,082
143,493
153,828
153,828
Distribution
NA
6,605
4,209
4,547
4,987
3,825
3,825
Total
NA
150,681
208,397
227,993
231,262
242,033
242,033
Previous Estimate
NA
420,902
519,798
519,798
519,798
519,798
NA
NA (Not Applicable)
Post-Meter
The Inventory was updated to include an estimate for post-meter emissions. Post-meter emission factors are
presented in the 2019 Refinement to the 2006 Intergovernmental Panel on Climate Change (IPCC) Guidelines for
National Greenhouse Gas Inventories under natural gas systems (IPCC 2019). Post-meter emission sources include
certain leak emissions from residential and commercial appliances, industrial facilities and power plants, and
natural gas fueled vehicles. The specific sources within the post-meter estimate are as follows:
• Appliances in residential and commercial sectors—Leakage from house piping and natural gas appliances
such as furnaces, water heaters, stoves and ovens, and barbecues/grills.
• Leakage at industrial plants and power stations (EGUs) —Leakage from internal piping.
• Natural gas fueled vehicles—Emissions from vehicles with alternative fuels produced from natural gas
e.g., LNG, CNG, propane. Emissions for natural gas-fueled vehicles include releases from dead volumes
during fueling, emptying of gas cylinders of high-pressure interim storage units, for execution of pressure
tests and relaxation of residual pressure from vehicles' gas tanks, or decommissioning.
EPA's considerations for this source are documented in the Post-Meter memo. For each of the emission sources,
emissions are estimated by multiplying emission factors (e.g., emission rate per unit fuel consumption or per
natural gas household) by corresponding activity data (e.g., fuel consumption, or number of natural gas
households). The methodology and data sources used for each are discussed here.
For residential sources, EPA applied the Cm emission factor from Fischer et al. (2018), which is on an emission rate
per natural gas household basis. The Fischer et al. EF accounts for passive house leak emissions and appliance leak
emissions. Activity data used to estimate CFU emissions from residential post-meter sources are national counts of
natural gas households (i.e., households using natural gas for space heating, water heating, cooking, and other
purposes). EPA used national-level data on natural gas households from the U.S. Census Bureau's American
Housing Survey publications (AHS 2021). AHS data are published on a biennial basis and EPA estimated data for
missing time-series years using the average of data from the years immediately before and after the missing year.
Energy 3-105
-------
The residential post-meter emission factor captures combustion emissions for gas appliances along with unburned
methane emissions. To ensure there is no double-counting with residential natural gas combustion emissions (i.e.,
from stationary fuel combustion), EPA subtracted the Cm emissions for the residential natural gas combustion
source (see Section 3.1 CH4and N20 from Stationary Combustion) from the estimated residential post-meter
emissions.
For commercial post-meter emissions, EPA used the IPCC default Cm and CO2 emission factors (IPCC 2019) and
national data on commercial buildings, by fuel types and end use, from ElA's Commercial Buildings Energy
Consumption Survey (CBECS 2021). CBECS contains data on the number of commercial buildings that use natural
gas for specific end uses such as space heating, water heating, and cooking but does not indicate the number of
appliances at commercial buildings. The CBECS data are only available for 1992,1995,1999, 2003, 2012, and 2018.
EPA estimated national commercial appliance counts for these years by assuming one appliance of each type per
commercial building using natural gas for that appliance type. Using the estimated appliance counts and natural
gas commercial meter counts, EPA developed an average estimate of 1 appliance per commercial meter. EPA then
estimated annual commercial appliance counts for the time-series by applying the estimate of 1 appliance per
commercial meter to time-series data on natural gas commercial meter counts.
For industrial post-meter emissions, EPA used the IPCC default CH4 and CO2 emission factors (IPCC 2019) and
activity data on natural gas consumption in the industrial and EGU sectors from EIA (EIA 2021b).
For vehicle post-meter emissions, EPA used the IPCC default CH4 and CO2 emission factors (IPCC 2019) and
estimated the national natural gas fueled vehicle population based on data from EPA's Motor Vehicle Emission
Simulator model (MOVES) (EPA 2020).
In 2020, total CFU emissions from all post-meter sources were estimated to be 459.1 kilotons (11.5 MMT CO2 Eq.).
This represents a 35 percent increase from 1990 levels and a slight increase of 1 percent from the previous year.
Approximately 53 percent of all post-meter CFU emissions are from the industrial and EGUs sub-segment, 42
percent from the residential sub-segment, and approximately 5 percent from the commercial sub-segment.
Natural gas vehicles contribute less than 0.05 percent of total post-meter CH4 emissions. CO2 emissions from post-
meter are 2.2 kt, contributing less than 0.01 percent of total CO2 from natural gas systems Inventory. CO2
emissions from residential appliances are assumed to be captured by residential natural gas combustion source
and are not included under post-meter estimates.
Table 3-95: Post-Meter Segment National ChU Emissions (Metric Tons ChU)
Activity
1990
2005
2016
2017
2018
2019
2020
Residential
142,755
169,828
186,242
189,537
188,637
190,478
192,199
Commercial
16,945
20,792
21,899
22,000
22,073
22,206
22,508
Industrial and EGUs
130,251
153,837
218,144
212,931
234,483
243,838
244,333
Natural Gas Vehicles
+
7
21
24
27
30
32
Total
289,951
344,464
426,306
424,492
445,220
456,551
459,072
Previous Estimate
NE
NE
NE
NE
NE
NE
NE
+ Does not exceed 0.5 metric tons.
Note: Totals may not sum due to independent rounding.
Table 3-96: Post-Meter Segment National CO2 Emissions (kt CO2)
Activity
1990
2005
2016
2017
2018
2019
2020
Residential
IE
IE
IE
IE
IE
IE
IE
Commercial
0.1
0.2
0.2
0.2
0.2
0.2
0.2
Industrial and EGUs
1.1
1.3
1.8
1.8
1.9
2.0
2.0
Natural Gas Vehicles
+
+
+
+
+
+
+
Total
1.2
1.4
2.0
1.9
2.1
2.2
2.2
Previous Estimate
NE
NE
NE
NE
NE
NE
NE
+ Does not exceed 0.05 kt.
IE (Included Elsewhere). Due to calculation methodologies, residential post-meter C02 fugitive emissions are included
in the fossil fuel combustion values.
3-106 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
NE (Not Estimated)
Note: Totals may not sum due to independent rounding.
Planned Improvements
Post-Meter Fugitive Emissions
EPA received feedback on this update through the September 2021 Post-Meter Memo and the public review of the
current Inventory. EPA received comments suggesting that EPA delay the inclusion of post-meter estimates.
Stakeholders presented concerns on the use of Fischer et al. study data in developing national estimates for
residential post-meter sources. Stakeholders suggested that the Fischer et al. study, conducted in California, is not
representative of national activity. EPA reviewed other residential post-meter studies, including the Merrin and
Francisco (2019) study conducted in Boston and Indianapolis (refer to Post-Meter memo for more details). The
other studies reviewed covered only emissions from major appliances, whereas the Fischer et al. study covered
emissions from passive house leaks and gas appliances (both major and minor appliances). A stakeholder comment
also suggested that a phase out of pilot lights occurring over the past several decades should be reflected in the
time series.
EPA will continue to track studies that may include data that could be used to update the emission factor for
residential post-meter emissions, and also to use instead of IPCC default values for commercial, industrial, and
vehicle post-meter emissions. EPA may consider approaches to take into account changes in emissions rates over
the time series such as applying default IPCC factors for residential emissions for earlier years of the time series.
Transmission Station Counts
Stakeholder feedback suggested alternate approaches for calculating the annual number of transmission stations.
EPA will consider the update for the next (1990 through 2021) Inventory. Stakeholder feedback on the public
review draft recommended against use of proprietary data sources for this activity data set. EPA will consider using
the proprietary data sets for QA/QC of EPA's activity data estimates.
Upcoming Data, and Additional Data that Could Inform the Inventory
EPA will assess new data received by the EPA Methane Challenge Program on an ongoing basis, which may be used
to validate 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 emission estimates, 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 stakeholder comments.
EPA also continues to seek new data that could be used to assess or update the estimates in the Inventory. For
example, stakeholder comments have highlighted areas where additional data that could inform the Inventory are
currently limited or unavailable:
• Tank measurements and tank and flaring malfunction and control efficiency data.
• Improved equipment leak data (activity and emissions data).
• Activity data and emissions data for production facilities that do not report to GHGRP.
• Onshore mud degassing.
• Anomalous leak events information throughout the time series and for future years.
Emission sources for which calculated emissions are negative when Gas STAR or Methane Challenge reductions are
applied. See Recalculations Discussion section on Natural Gas STAR and Methane Challenge for the list of sources.
Energy 3-107
-------
EPA received stakeholder feedback through comments on the public review draft of the current Inventory. Several
stakeholders asserted that methane emissions are undercounted in natural gas systems. A stakeholder
commented suggested developing the inventory using a strategy that combines information from satellites,
aircraft-based instruments, and ground-based sensors. Stakeholder feedback on the public review draft
recommended use of updated emission factors for pneumatic controllers. A stakeholder suggested that current
emission factors underestimate emissions from combustion slip.
EPA will continue to seek available data on these and other sources as part of the process to update the Inventory.
3.8 Abandoned Oil and Gas Wells (CRF
Source Categories lB2a and lB2b)
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 3.0 million abandoned oil wells and 0.7 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 U.S. 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. Other groups have developed
estimates of the total number of orphaned wells. The Interstate Oil and Gas Compact Commission for example
estimates 92,198 orphaned wells in the U.S. (IOGCC 2021). State applications for grants to plug orphaned wells
indicate over 130,000 orphaned wells in the U.S. (Department of Interior 2022). 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 41 percent of the abandoned well population in the United States are plugged. This
fraction has increased over the time series (from around 23 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 219 kt CH4 and 4 kt CO2 in 2020. Emissions of both gases
increased by 2 percent from 1990, while the total population of abandoned oil wells increased 38 percent.
Abandoned gas wells. Abandoned gas wells emitted 57 kt CH4 and 3 kt CO2 in 2020. Emissions of both gases
increased by 25 percent from 1990, while the total population of abandoned gas wells increased 74 percent.
Table 3-97: ChU Emissions from Abandoned Oil and Gas Wells (MMT CO2 Eq.)
Activity
1990
2005
2016
2017
2018
2019
2020
Abandoned Oil Wells
5.4
5.5
5.5
5.5
5.5
5.5
5.5
Abandoned Gas Wells
1.1
1.3
1.4
1.4
1.4
1.5
1.4
Total
6.5
6.8
6.9
6.9
6.9
7.0
6.9
Note: Totals may not sum due to independent rounding.
Table 3-98: ChU Emissions from Abandoned Oil and Gas Wells (kt)
Activity 1990 2005 2016 2017 2018 2019 2020
Abandoned Oil Wells 215 222 218 219 220 221 219
3-108 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Abandoned Gas Wells
46
51
57 57 57 58 57
Total 261 273 275 276 277 279 276
Note: Totals may not sum due to independent rounding.
Table 3-99: CO2 Emissions from Abandoned Oil and Gas Wells (MMT CO2)
Activity 1990 2005 2016 2017 2018 2019 2020
Abandoned Oil Wells + + + + + + +
Abandoned Gas Wells + + + + + + +
Total + + + + + + +
+ Does not exceed 0.05 MMT C02.
Table 3-100: CO2 Emissions from Abandoned Oil and Gas Wells (kt)
Activity 1990 2005 2016 2017 2018 2019 2020
Abandoned Oil Wells 4 5 4 4 4 4 4
Abandoned Gas Wells 2 2 2 3 3 3 3
Total 6 7 7 7 7 7 7
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, 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 Cm 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 all other abandoned wells. EPA developed abandoned well CO2 emission factors
using the Cm 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 CH4, which
was derived through API TankCalc modeling runs. For abandoned gas wells, EPA used the Natural Gas Systems
default production segment CFU and CO2 gas content values (GRI/EPA 1996, GTI 2001) to develop a ratio of 0.044
MT CO2/MT Cm. The same respective emission factors are applied for each year of the time series.
EPA developed 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
in the Enverus dataset (see 2018 Abandoned Wells Memo for additional information on how the value was
calculated). 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. To calculate the number of wells not included in
the Enverus dataset, estimated abandoned well counts (oil, gas, and dry) for 1975 from historical data available at
the state-level (by subtracting total active wells from total drilled wells) and deducted abandoned well counts
developed using Enverus data for 1975 for the corresponding states. The resulting total number of abandoned
wells (i.e., not included in Enverus data) is then added to the annual abandoned well counts developed using
Enverus data for 1990 to 2020.
The total abandoned well population was then split into plugged and unplugged wells by assuming that all
abandoned wells were unplugged in 1950 and using year-specific Enverus data to calculate the fraction of plugged
abandoned wells (41 percent) in 2020 in that data set. Abandoned wells not included in the Enverus dataset were
Energy 3-109
-------
assumed to be unplugged. 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.83 The abandoned wells activity data methodology was also updated for this Inventory; see the
Recalculations Discussion section for more information.
Abandoned Oil Wells
Table 3-101: Abandoned Oil Wells Activity Data, ChU and CO2 Emissions (kt)
Source/Activity Data
1990
2005
2016
2017
2018
2019
2020
Plugged abandoned oil wells
(number of wells)
507,322
834,303
j 1,108,361
1,140,315
1,171,441
1,205,046
1,222,510
Unplugged abandoned oil
wells (number of wells)
1,660,257
1,758,160
1,761,684
1,768,403
1,772,730
1,779,690
1,762,226
Total Abandoned Oil Wells
2,167,579
2,592,463
2,870,046
2,908,718
2,944,171
2,984,736
2,984,736
Abandoned oil wells in
Appalachia (percent)
23%
21%
20%
20%
20%
20%
20%
Abandoned oil wells outside
of Appalachia (percent)
77%
79%
80%
80%
80%
80%
80%
CH4 from plugged
abandoned oil wells (kt)
0.37
0.56
0.70
0.73
0.74
0.77
0.78
CH4from unplugged
abandoned oil wells (kt)
214.6
221.3
217.7
218.6
219.1
219.9
217.8
Total ChUfrom Abandoned
Oil Wells (kt)
215.0
221.8
218.4
219.3
219.8
220.7
218.6
Total C02 from Abandoned
Oil Wells (kt) 4.4 4.5 4.4 4.4 4.5 4.5 4.4
Abandoned Gas Wells
Table 3-102: Abandoned Gas Wells Activity Data, ChU and CO2 Emissions (kt)
Source/Activity Data
1990
2005
2016
2017
2018
2019
2020
Plugged abandoned gas wells
(number of wells)
100,295
180,578
273,018
282,358
291,443
301,449
305,818
Unplugged abandoned gas
wells (number of wells)
328,226
380,540
433,948
437,880
441,037
445,200
440,831
Total Abandoned Gas Wells
428,521
561,119
706,966
720,238
732,480
746,649
746,649
Abandoned gas wells in
Appalachia (percent)
29%
25%
23%
23%
23%
23%
23%
Abandoned gas wells outside
of Appalachia (percent)
71%
75%
77%
77%
77%
77%
77%
CH4from plugged abandoned
gas wells (kt)
0.09
0.15
0.20
0.21
0.22
0.22
0.23
CH4from unplugged
abandoned gas wells (kt)
45.7
50.9
56.3
56.9
57.3
57.8
57.2
Total CH4 from Abandoned
Gas Wells (kt)
45.8
51.0
56.6
57.1
57.5
58.0
57.5
83 See https://www.epa.gov/ghgemissions/natural-gas-and-petroleum-systems.
3-110 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Total C02 from Abandoned
Gas Wells (kt) 2.0 2.2 2.5 2.5 2.5 2.5 2.5
Uncertainty
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
wells in year 2020, then applied the calculated bounds to both CH4 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 3-103 provide the 95 percent confidence bound within which actual
emissions from abandoned oil and gas wells are likely to fall for the year 2020, using the recommended IPCC
methodology. Abandoned oil well CFU emissions in 2020 were estimated to be between 0.9 and 16.2 MMT CO2 Eq.,
while abandoned gas well CH4 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 3-103: Approach 2 Quantitative Uncertainty Estimates for ChU and CO2 Emissions from
Abandoned Oil and Gas Wells (MMT CO2 Eq. and Percent)
2020 Emission Estimate Uncertainty Range Relative to Emission Estimate3
Source Gas
(MMT C02 Eq.)b (MMT CP2 Eq.) (%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Abandoned Oil Wells
ch4
5.5
0.9
16.2
-83%
+197%
Abandoned Gas Wells
ch4
1.4
0.2
4.3
-83%
+197%
Abandoned Oil Wells
C02
0.004
0.001
0.013
-83%
+197%
Abandoned Gas Wells
C02
0.003
0.0004
0.007
-83%
+197%
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 2020.
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
Energy 3-111
-------
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 2021.
Recalculations Discussion
EPA received information and data related to the emission estimates through feedback on updates under
consideration. In September 2021, EPA released a draft memorandum that discussed changes under consideration
and requested stakeholder feedback on those changes. EPA then updated the memorandum to document the
methodology implemented in the current Inventory.84 The memorandum cited in the Recalculations Discussion
below is, Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2020: Updates for Abandoned Oil and Gas
Wells (Abandoned Wells memo).
EPA updated the methodology to estimate abandoned wells activity data, including the population of abandoned
wells and the fraction of abandoned wells that are plugged, as discussed in the Abandoned Wells memo. EPA did
not update the emission factors. As in previous Inventories, the activity data methodology relies on Enverus data
to: (1) estimate the population of abandoned oil and gas wells over the time series (along with data from historical
references) and (2) estimate the fraction of abandoned wells that are plugged versus unplugged. This Inventory
was recalculated with modifications to the Enverus data processing. Modifications to both steps are discussed
here.
To estimate the population of abandoned oil wells and abandoned gas wells over the time series, EPA updated its
method to rely on the gas-to-oil ratio (GOR) and the production type field within Enverus data to classify
abandoned wells as oil versus gas wells. EPA used the production type field within the Enverus wells dataset only
to apportion dry wells to oil and gas wells. The production type field was used in the previous (2021) Inventory to
apportion dry wells, but for Inventories prior to the 2021 submission, only the GOR was used to assign abandoned
wells as oil and gas wells.
To estimate the fraction of plugged and unplugged abandoned wells, EPA used the updated plugging status
assignments for Enverus well status codes (see the Abandoned Wells memo) and assumed that all historical wells
that are not captured in the Enverus wells dataset are unplugged. EPA first analyzed the Enverus dataset and
determined that 58 percent of abandoned wells within Enverus are plugged. EPA then incorporated the historical
well population (approximately 1.2 million wells) and assumed all historical wells are unplugged, resulting in an
estimate of 41 percent of abandoned wells plugged (that percent is applied to year 2020 in the Inventory).
The Methodology and Time-Series Consistency section above includes tables with the updated plugged and
unplugged abandoned well counts, reflecting the updates discussed here.
EPA received stakeholder feedback on the updates. A stakeholder recommended that the production type field
within the Enverus dataset should be used to apportion wells that would otherwise be classified as dry wells
(based on using the gas-to-oil ratio) into the classification of abandoned oil or abandoned gas wells, which was
implemented in the final Inventory. The stakeholder suggested dry wells may be double counted and noted that
some dry holes are plugged. The stakeholder also indicated that it may not be appropriate to assume all
abandoned wells not captured within the Enverus dataset are unplugged. The stakeholder recommended using the
"western US" emission factor from theTownsend-Small et al. study for areas outside of Appalachia, instead of the
national average currently applied. Additionally, a stakeholder suggested that emissions from abandoned wells are
underestimated.
84 Stakeholder materials including draft and final memoranda for the current (i.e., 1990 to 2020) Inventory are available at
https://www.epa.gov/ehgemissions/natural-gas-and-petroleum-systems.
3-112 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
As an outcome of these revisions, calculated abandoned oil well Cm emissions decreased by an average of 6
percent across the time series and increased by 6 percent in 2019, compared to the values in the previous
Inventory. Abandoned gas well Cm emissions increased by an average of 5 percent across the time series and
increased by 6 percent in 2019, compared the to the previous Inventory.
Planned Improvements
EPA will continue to assess new data and stakeholder feedback on considerations (such as disaggregation of the
well population into regions other than Appalachia and non-Appalachia, and emission factor data from regions not
included in the measurement studies on which current emission factors are based) to improve the abandoned well
count estimates and emission factors.
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.85 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).86
Two transport modes are addressed under the IPCC definition of international bunker fuels: aviation and marine.87
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.88
85 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).
86 Note that the definition of international bunker fuels used by the UNFCCC differs from that used by the International Civil
Aviation Organization.
87 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).
88 Naphtha-type jet fuel was used in the past by the military in turbojet and turboprop aircraft engines.
Energy 3-113
-------
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 Cm is emitted by modern engines (Anderson et al. 2011), and as a result, Cm
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.
Overall, aggregate greenhouse gas emissions in 2020 from the combustion of international bunker fuels from both
aviation and marine activities were 70.3 MMT CO2 Eq., or 32.8 percent below emissions in 1990 (see Table 3-104
and Table 3-105). Emissions from international flights and international shipping voyages departing from the
United States have increased by 4.1 percent and decreased by 54.4 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.
Table 3-104: CO2, ChU, and N2O Emissions from International Bunker Fuels (MMT CO2 Eq.)
Gas/Mode
1990
2005
2016
2017
2018
2019
2020
CO?
103.6
113.3
116.7
120.2
122.2
116.1
69.6
Aviation
38.2
60.2
74.1
77.8
80.9
80.8
39.8
Commercial
30.0
55.5
70.8
74.5
77.7
77.6
36.7
Military
8.2
4.5
3.3
3.3
3.2
3.2
3.1
Marine
65.4
53.1
42.6
42.4
41.3
35.4
29.9
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
0.9
1.0
1.0
1.1
1.1
1.0
0.6
Aviation
0.4
0.6
0.7
0.7
0.8
0.8
0.4
Marine
0.5
0.4
0.3
0.3
0.3
0.3
0.2
Total
104.7
114.4
117.8
121.3
123.4
117.2
70.3
NO (Not Occurring)
Notes: Totals may not sum due to independent rounding. Includes aircraft cruise altitude emissions.
Table 3-105: CO2, ChU, and N2O Emissions from International Bunker Fuels (kt)
Gas/Mode
1990
2005
2016
2017
2018
2019
2020
CO?
103,634
113,328
116,682
120,192
122,179
116,132
69,638
Aviation
38,205
60,221
74,128
77,764
80,853
80,780
39,781
Marine
65,429
53,107
42,554
42,428
41,325
35,351
29,857
ch4
7
5
4
4
4
4
3
Aviation
NO
NO
NO
NO
NO
NO
NO
Marine
7
5
4
4
4
4
3
N20
3
3
3
4
4
3
2
Aviation
1
2
2
2
3
3
1
Marine
2
•; 1 BP
1
1
1
1
1
NO (Not Occurring)
Notes: Totals may not sum due to independent rounding. Includes aircraft cruise altitude emissions.
3-114 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Methodology and Time-Series Consistency
Emissions of CO2 were 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, 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).
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
bunker fuel use by the U.S. military.
Emission estimates for CH4 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 CFU 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 CH4 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 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 2021). 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 3-106. See Annex 3.8 for additional discussion of military data.
Energy 3-115
-------
Table 3-106: Aviation Jet Fuel Consumption for International Transport (Million Gallons)
Nationality
1990
2005
2016
2017
2018
2019
2020
U.S. and Foreign Carriers
3,155
5,858
7,452
7,844
8,178
8,170
3,859
U.S. Military
862
462
333
326
315
318
308
Total
4,017
6,321
7,785
8,171
8,493
8,488
4,167
Note: Totals may not sum due to independent rounding.
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 2020) for 1990 through 2001, 2007 through 2020, 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 (2021). 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 3-107.
Table 3-107: Marine Fuel Consumption for International Transport (Million Gallons)
Fuel Type
1990
2005
2016
2017
2018
2019
2020
Residual Fuel Oil
4,781
3,881
3,011
2,975
2,790
2,246
1,964
Distillate Diesel Fuel & Other
617
444
534
568
684
702
461
U.S. Military Naval Fuels
522
471
314
307
285
281
296
Total
5,920
4,796
3,858
3,850
3,759
3,229
2,721
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 2020.
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.89 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
89 See uncertainty discussions under section 3.1 C02 from Fossil Fuel Combustion.
3-116 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
associated with ground fuel estimates for 1997 through 2020, 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 CO2.90
There is also concern regarding the reliability of the existing DOC (1991 through 2020) data on marine vessel fuel
consumption reported at U.S. customs stations due to the significant degree of inter-annual variation.
/erification
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.
90 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 3-117
-------
Recalculations Discussion
The density for jet fuel was updated to 3.002 kilograms per gallon (EIA 2022) to improve consistency across
estimates and data sources. This revision resulted in an average annual change of less than 0.05 MMT CO2 Eq. in
total emissions from international bunker fuels.
Planned Improvements
EPA will evaluate data availability to update the sources for densities, energy contents, and emission factors
applied to estimate emissions from aviation and marine fuels. Many are from sources from the late 1990s, such as
IPCC/UNEP/OECD/IEA (1997). Potential sources with more recent data include the International Maritime
Organization (IMO) greenhouse gas emission inventory, International Air Transport Association (IATA)/ICAO
greenhouse gas reporting system (CORSIA), and the EPA Greenhouse Gas Reporting Program (GHGRP) Technical
Support Document for Petroleum Products. Specifically, EPA will evaluate data availability to support updating the
heat contents and carbon contents of jet fuel with input from EIA.
A longer-term effort is underway to consider the feasibility of including data from a broader range of domestic and
international sources for bunker fuels. Potential sources include the IMO greenhouse gas emission inventory, data
from the U.S. Coast Guard on vehicle operation currently used in criteria pollutant modeling, data from the
International Energy Agency (IEA), relevant updated FAA models to improve aviation bunker fuel estimates, and
researching newly available marine bunker data.
3.10 Wood Biomass and Biofuels
Consumption (CRF Source Category 1A)
The combustion of biomass fuels—such as wood, charcoal, and wood waste and biomass-based fuels such as
ethanol, biogas, and biodiesel—generates CO2 in addition to CH4 and N2O already covered in this chapter. In line
with the reporting requirements for inventories submitted under the UNFCCC, CO2 emissions from biomass
combustion have been estimated separately from fossil fuel CO2 emissions and are not directly included in the
energy sector contributions to U.S. totals. In accordance with IPCC methodological guidelines, any such emissions
are calculated by accounting for net carbon fluxes from changes in biogenic C reservoirs in wooded or crop lands.
For a more complete description of this methodological approach, see the Land Use, Land-Use Change, and
Forestry chapter (Chapter 6), which accounts for the contribution of any resulting CO2 emissions to U.S. totals
within the Land Use, Land-Use Change, and Forestry sector's approach.
Therefore, CO2 emissions from wood biomass and biofuel consumption are not included specifically in summing
energy sector totals. However, they are presented here for informational purposes and to provide detail on wood
biomass and biofuels consumption.
In 2020, total CO2 emissions from the burning of woody biomass in the industrial, residential, commercial, and
electric power sectors were approximately 202.1 MMT CO2 Eq. (202,088 kt) (see Table 3-108 and Table 3-109). As
the largest consumer of woody biomass, the industrial sector was responsible for 63.0 percent of the CO2
emissions from this source. The residential sector was the second largest emitter, constituting 23.3 percent of the
total, while the electric power and commercial sectors accounted for the remainder.
Table 3-108: CO2 Emissions from Wood Consumption by End-Use Sector (MMT CO2 Eq.)
End-Use Sector
1990
2005
2016
2017
2018
2019
2020
Industrial
135.3
136.3
138.3
135.4
134.4
132.1
127.2
Residential
59.8
44.3
45.8
44.3
54.1
56.1
47.2
Commercial
6.8
7.2
8.6
8.6
8.7
8.7
8.6
3-118 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Electric Power 133 lOl 23J. 23£ 2Z8 207 19.1
Total 215.2 206.9 216.0 211.9 220.0 217.6 202.1
Table 3-109: CO2 Emissions from Wood Consumption by End-Use Sector (kt)
End-Use Sector
1990
2005
2016
2017
2018
2019
2020
Industrial
135,348
136,269
138,339
135,386
134,417
132,069
127,242
Residential
59,808
44,340
45,841
44,257
54,070
56,135
47,177
Commercial
6,779
7,218
8,635
8,634
8,669
8,693
8,554
Electric Power
13,252
19,074
23,140
23,647
22,795
20,677
19,115
Total
215,186
206,901
215,955
211,925
219,951
217,574
202,088
Note: Totals may not sum due to independent rounding.
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 2020, the United States transportation sector consumed an estimated 994.6 trillion Btu of ethanol (95 percent
of total), and as a result, produced approximately 68.1 MMT CO2 Eq. (68,084 kt) (see Table 3-110 and Table 3-111)
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 3-110: CO2 Emissions from Ethanol Consumption (MMT CO2 Eq.)
End-Use Sector
1990
2005
2016
2017
2018
2019
2020
Transportation3
4.1
21.6
76.9
77.7
78.6
78.7
68.1
Industrial
0.1
1.2
1.8
1.9
1.4
1.6
1.6
Commercial
0.1
0.2
2.6
2.5
1.9
2.2
2.2
Total
4.2
22.9
81.2
82.1
81.9
82.6
71.8
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.
Table 3-111: CO2 Emissions from Ethanol Consumption (kt)
End-Use Sector
1990
2005
2016
2017
2018
2019
2020
Transportation3
4,059
21,616
76,903
77,671
78,603
78,739
68,084
Industrial
105
1,176
1,789
1,868
1,404
1,610
1,582
Commercial
63
151
2,558
2,550
1,910
2,229
2,182
Total
4,227
22,943
81,250
82,088
81,917
82,578
71,847
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.
The transportation sector is assumed to be responsible for all of the biodiesel consumption in the United States
(EIA 2022). Biodiesel is currently produced primarily from soybean oil, but it can be produced from a variety of
biomass feedstocks including waste oils, fats, and greases. Biodiesel for transportation use appears in low-level
blends (less than 5 percent) with diesel fuel, high-level blends (between 6 and 20 percent) with diesel fuel, and 100
percent biodiesel (EIA 2020).
In 2020, the United States consumed an estimated 239.4 trillion Btu of biodiesel, and as a result, produced
approximately 17.7 MMT CO2 Eq. (17,678 kt) (see Table 3-112 and Table 3-113) of CO2 emissions. Biodiesel
production and consumption has grown significantly since 2001 due to the favorable economics of blending
biodiesel into diesel and federal policies that have encouraged use of renewable fuels (EIA 2020). There was no
measured biodiesel consumption prior to 2001 EIA (2022).
Energy 3-119
-------
Table 3-112: CO2 Emissions from Biodiesel Consumption (MMT CO2 Eq.)
End-Use Sector
1990
2005
2016
2017
2018
2019
2020
Transportation3
NO
0.9
19.6
18.7
17.9
17.1
17.7
NO (Not Occurring)
a See Annex 3.2, Table A-76 for additional information on transportation consumption of these fuels.
Table 3-113: CO2 Emissions from Biodiesel Consumption (kt)
End-Use Sector
1990
2005
2016
2017
2018
2019
2020
Transportation3
NO
856
19,648
18,705
17,936
17,080
17,678
NO (Not Occurring)
a See Annex 3.2, Table A-76 for additional information on transportation consumption of these fuels.
Methodology and Time-Series Consistency
Woody biomass emissions were estimated by applying two gross heat contents from EIA (Lindstrom 2006) to U.S.
consumption data (EIA 2022) (see Table 3-114), provided in energy units for the industrial, residential, commercial,
and electric power sectors. One heat content (16.95 MMBtu/MT wood and wood waste) was applied to the
industrial sector's consumption, while the other heat content (15.43 MMBtu/MT wood and wood waste) was
applied to the consumption data for the other sectors. An EIA emission factor of 0.434 MT C/MT wood (Lindstrom
2006) was then applied to the resulting quantities of woody biomass to obtain CO2 emission estimates. The woody
biomass is assumed to contain black liquor and other wood wastes, have a moisture content of 12 percent, and
undergo complete combustion to be converted into CO2.
The amount of ethanol allocated across the transportation, industrial, and commercial sectors was based on the
sector allocations of ethanol-blended motor gasoline. The sector allocations of ethanol-blended motor gasoline
were determined using a bottom-up analysis conducted by EPA, as described in the Methodology section of Fossil
Fuel Combustion. Total U.S. ethanol consumption from EIA (2022) was allocated to individual sectors using the
same sector allocations as ethanol-blended motor gasoline. The emissions from ethanol consumption were
calculated by applying an emission factor of 18.67 MMT C/Qbtu (EPA 2010) to adjusted ethanol consumption
estimates (see Table 3-115). The emissions from biodiesel consumption were calculated by applying an emission
factor of 20.1 MMT C/Qbtu (EPA 2010) to U.S. biodiesel consumption estimates that were provided in energy units
(EIA 2022) (see Table 3-116).91
Table 3-114: Woody Biomass Consumption by Sector (Trillion Btu)
End-Use Sector
1990
2005
2016
2017
2018
2019
2020
Industrial
1,441.9
1,451.7
1,473.8
1,442.3
1,432.0
1,407.0
1,355.6
Residential
580.0
430.0
444.6
429.2
524.4
544.4
457.5
Commercial
65.7
70.0
83.7
83.7
84.1
84.3
83.0
Electric Power
128.5
185.0
224.4
229.3
221.1
200.5
185.4
Total
2,216.2
2,136.7
2,226.5
2,184.6
2,261.5
2,236.2
2,081.4
Note: Totals may not sum due to independent rounding.
ble 3-115: Ethanol Consumption by Sector (Trillion Btu)
End-Use Sector
1990
2005
2016
2017
2018
2019
2020
Transportation
59.3
315.8
1,123.4
1,134.6
1,148.2
1,150.2
994.6
Industrial
1.5
17.2
26.1
27.3
20.5
23.5
23.1
91 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.
3-120 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Commercial 09 Z2 37A 3T2 273 3Z6 31.9
Total 6L7 335.1 1,186.9 1,199.1 1,196.6 1,206.3 1,049.5
Note: Totals may not sum due to independent rounding.
Table 3-116: Biodiesel Consumption by Sector (Trillion Btu)
End-Use Sector 1990 2005 2016 2017 2018 2019 2020
Transportation NO 1L6 266.1 253.3 242.9 231.3 239.4
Total NO 1L6 266.1 253.3 242.9 231.3 239.4
NO (Not Occurring)
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2020.
Uncertainty
It is assumed that the combustion efficiency for woody biomass is 100 percent, which is believed to be an
overestimate of the efficiency of wood combustion processes in the United States. Decreasing the combustion
efficiency would decrease emission estimates for CO2. Additionally, the heat content applied to the consumption
of woody biomass in the residential, commercial, and electric power sectors is unlikely to be a completely accurate
representation of the heat content for all the different types of woody biomass consumed within these sectors.
Emission estimates from ethanol and biodiesel production are more certain than estimates from woody biomass
consumption due to better activity data collection methods and uniform combustion techniques.
Recalculations Discussion
EIA (2022) revised approximate heat rates for electricity and the heat content of electricity for noncombustible
renewable energy, which impacted wood energy consumption by the industrial sector from 2016 through 2019. In
addition, EIA (2022) revised its methodology for calculating renewable diesel fuel consumption which impacts
biofuel consumption. Between 2016 and 2019, revisions to biomass consumption resulted in an average annual
increase of 0.5 MMT CO2 Eq. (0.2 percent). Overall, revisions to biomass consumption resulted in an average
annual increase of 0.1 MMT CO2 Eq. (less than 0.05 percent) across the time series.
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, ethanol, and biodiesel. Additional forms of biomass-based fuel consumption include biogas, the
biogenic components of MSW, 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 (2022) 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. Sources of
estimates for the biogenic fraction of MSW will be examined, including EPA's GHGRP, EIA data, and EPA MSW
characterization data. 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
Energy 3-121
-------
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.92
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.93
Carbon dioxide emissions from biomass used in the electric power sector are calculated using woody biomass
consumption data from ElA's Monthly Energy Review (EIA 2022), whereas non-CC>2 biomass emissions from the
electric power sector are estimated by applying technology and fuel use data from EPA's Clean Air Market Acid
Rain Program dataset (EPA 2022) to fuel consumption data from EIA (2022). There were significant discrepancies
identified between the EIA woody biomass consumption data and the consumption data estimated using EPA's
Acid Rain Program Dataset (see the Methodology section for CH4 and N2O from Stationary Combustion). EPA will
continue to investigate this discrepancy in order to apply a consistent approach to both CO2 and non-CC>2 emission
calculations for woody biomass consumption in the electric power sector.
3.11 Energy Sources of Precursor
Greenhouse Gas Emissions
In addition to the main greenhouse gases addressed above, energy-related activities are also sources of
greenhouse gas precursors. The reporting requirements of the UNFCCC94 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 3-117.
Table 3-117: NOx, CO, NMVOC, and SO2 Emissions from Energy-Related Activities (kt)
Gas/Activity
1990
2005
2016
2017
2018
2019
2020
NOx
21,106
16,602
8,268
7,883
7,456
6,962
6,471
Fossil Fuel Combustion
20,885
16,153
7,595
7,246
6,819
6,325
5,834
Transportation
10,862
10,295
4,739
4,519
4,153
3,788
3,422
Industrial
2,559
1,515
890
859
859
859
859
Electric Power Sector
6,045
3,434
1,234
1,049
987
859
733
Commercial
671
490
440
537
537
537
537
92 See https://unfccc.int/resource/docs/2013/copl9/ene/10a03.pdf#paee=2.
93 See http://www.ipcc-neeip.iges.or.ip/public/tb/TFI Technical Bulletin l.pdf.
94 See http://unfccc.int/resource/docs/2013/copl9/ene/10a03.pdf.
3-122 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Residential
749
418
292
283
283
283
283
Petroleum and Natural Gas Systems
137
301
557
530
530
530
530
Incineration of Waste
82
128
80
71
71
71
71
Other Energy
2
20
37
35
35
35
35
International Bunker Fuelsa
1,953
1,699
1,464
1,475
1,456
1,290
1,019
CO
125,640
64,985
34,461
33,401
32,392
31,384
30,376
Fossil Fuel Combustion
124,360
63,263
32,479
31,634
30,626
29,617
28,609
Transportation
119,360
58,615
28,789
27,942
26,934
25,926
24,918
Residential
3,668
2,856
2,215
2,291
2,291
2,291
2,291
Industrial
797
1,045
771
736
736
736
736
Electric Power Sector
329
582
575
532
532
532
532
Commercial
205
166
128
133
133
133
133
Petroleum and Natural Gas Systems
299
294
560
546
546
546
546
Incineration of Waste
978
1,403
1,375
1,175
1,175
1,175
1,175
Other Energy
3
24
47
46
46
46
46
International Bunker Fuelsa
102
131
147
153
158
154
101
NMVOCs
12,612
7,345
6,022
5,664
5,491
5,318
5,145
Fossil Fuel Combustion
11,836
6,594
3,443
3,293
3,120
2,947
2,774
Transportation
10,932
5,724
2,873
2,728
2,555
2,382
2,209
Residential
686
518
322
319
319
319
319
Commercial
10
188
117
116
116
116
116
Industrial
165
120
101
101
101
101
101
Electric Power Sector
43
44
31
29
29
29
29
Petroleum and Natural Gas Systems
552
497
2,397
2,205
2,205
2,205
2,205
Incineration of Waste
222
241
121
109
109
109
109
Other Energy
2
13
62
57
57
57
57
International Bunker Fuelsa
57
6,594
49
50
50
46
34
S02
19,628
12,364
2,439
1,794
1,701
1,433
1,270
Fossil Fuel Combustion
19,200
12,159
2,327
1,686
1,594
1,326
1,163
Electric Power Sector
14,433
9,439
1,819
1,257
1,167
902
742
Industrial
3,221
1,574
389
342
342
342
342
Transportation
793
619
57
48
45
42
39
Commercial
589
370
43
28
28
28
28
Residential
165
158
18
12
12
12
12
Petroleum and Natural Gas Systems
387
177
85
82
82
82
82
Incineration of Waste
38
25
24
22
22
22
22
Other Energy
3
3
3
3
3
3
3
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).95 NEI Tier 1 emission categories related to the energy sector categories in
95 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
Energy 3-123
-------
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's TSD and on EPA's Air Pollutant Emission Trends website
(EPA 2021a; EPA 2021b). Updates to historical activity data are documented in NEI's TSD (EPA 2021b). No
quantitative estimates of uncertainty were calculated for this source category.
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.
3-124 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 2020, IPPU generated emissions of 376.4 million metric tons of CO2 equivalent (MMT CO2 Eq.), or 6.3 percent of
total U.S. greenhouse gas emissions.1 Carbon dioxide emissions from all industrial processes were 163.6 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 to the extent of which industries are occurring.
Industrial Processes and Product Use 4-1
-------
Eq. (163,571 kt CO2) in 2020, or 3.5 percent of total U.S. CO2 emissions. Methane emissions from industrial
processes resulted in emissions of approximately 0.3 MMT CO2 Eq. (14 kt CH4) in 2020, which was 0.1 percent of
U.S. Cm emissions. Nitrous oxide emissions from IPPU were 23.3 MMT CO2 Eq. (78 kt N2O) in 2020, or 5.5 percent
of total U.S. N2O emissions. In 2020 combined emissions of HFCs, PFCs, SF6, and NF3 totaled 189.2 MMT CO2 Eq.
Total emissions from IPPU in 2020 were 8.7 percent more than 1990 emissions. Total emissions from IPPU
remained relatively constant between 2019 and 2020, decreasing by 0.8 percent due to offsetting trends within
the sector. Some industrial processes and product use categories experienced decreases due to impacts from the
coronavirus (COVID-19) pandemic (e.g., iron and steel production and lime production), while other categories
experienced increases in emissions from 2019 to 2020 (e.g., ammonia production and the substitution of ozone
depleting substances). More information on emissions of greenhouse gas precursors emissions that also result
from IPPU are presented in Section 4.27 of this chapter.
Figure 4-1: 2020 Industrial Processes and Product Use Sector Greenhouse Gas Sources
Substitution of Ozone Depleting Substances
Cement Production
Iron and Steel Production & Metallurgical Coke Production
Petrochemical Production
Ammonia Production
Lime Production
Other Process Uses of Carbonates
Nitric Acid Production
Adipic Acid Production
Urea Consumption for Non-Agricultural Purposes
Carbon Dioxide Consumption
Electronics Industry
N2O from Product Uses
Electrical Transmission and Distribution
Aluminum Production
HCFC-22 Production
Glass Production
Soda Ash Production
Ferroalloy Production
Titanium Dioxide Production
Caprolactam, Glyoxal, and Glyoxylic Acid Production
Zinc Production
Phosphoric Acid Production
Magnesium Production and Processing
Lead Production
Carbide Production and Consumption
1176
Industrial Processes and Product Use as a
Portion of All Emissions
< 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 many chemical production sources have either decreased or
not changed significantly since 1990, with the exception of petrochemical production, ammonia production, urea
consumption for non-agricultural purposes, and carbon dioxide consumption, which has steadily increased.
Emissions from mineral sources have either increased (e.g., cement production) or not changed significantly (e.g.,
glass and lime production) since 1990 but largely follow economic cycles. Hydrofluorocarbon emissions from the
substitution of ODS have increased drastically since 1990 and are the largest source of IPPU emissions (46.8
percent in 2020), 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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure 4-2: Trends in Industrial Processes and Product Use Sector Greenhouse Gas Sources
I Electronics Industry
J Other Product Manufacture and Use
I Mineral Industry
I Metal Industry
I Chemical Industry
] Substitution of Ozone Depleting Substances
o-r-HCNro'^-LnvDrs.oocTi
O"* ^ Ol CJ1 0"> CT> 0"» Ol en
(N(N(MN(N(N(N(NOJ(N
(N fN CN fN
CM (N CM (N fM
Table 4-1 summarizes emissions for the IPPU chapter in MMT CO2 Eq. using IPCC Fourth Assessment Report (AR4)
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 native 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 products, chemical
production, metal production, 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 2019) to
ensure that the trend is accurate. Key updates to this year's inventory include revisions to the Glass Production
methodology to use more complete GHGRP activity data for the years 2010 through 2020; updated activity data
for Iron and Steel Production (e.g., updated coke production values, updated scrap steel consumption for EAF steel
production, scrap steel consumption for BOF steel production, and pellet consumption in blast furnace); 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 CO2 from Magnesium Production and Processing
(e.g., the inclusion of CO2 emissions from permanent mold, wrought, and anode production for the time series, the
inclusion of CO2 emissions from sand casting for the years 1990 through 2010) and Other Process Use of
Carbonates (e.g., moving CO2 emissions from the use of dolomite in primary magnesium metal production from
Other Process Uses of Carbonates to Magnesium Production and Processing). Together, these updates increased
greenhouse gas emissions an average of 0.7 MMT CO2 Eq. (0.2 percent) across the time series.
2 See http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
Industrial Processes and Product Use 4-3
-------
Table 4-1: Emissions from Industrial Processes and Product Use (MMT CO2 Eq.)
Gas/Source
1990
2005
2016
2017
2018
2019
2020
CO?
213.0
194.4
166.0
164.7
165.1
171.2
163.6
Iron and Steel Production &
Metallurgical Coke Production
104.7
70.1
43.6
40.6
42.6
43.1
37.7
Iron and Steel Production
99.1
66.2
41.0
38.6
41.3
40.1
35.4
Metallurgical Coke Production
5.6
3.9
2.6
2.0
1.3
3.0
2.3
Cement Production
33.5
46.2
39.4
40.3
39.0
40.9
40.7
Petrochemical Production
21.6
27.4
28.1
28.9
29.3
30.7
30.0
Ammonia Production
13.0
9.2
10.2
11.1
12.2
12.3
12.7
Lime Production
11.7
14.6
12.6
12.9
13.1
12.1
11.3
Other Process Uses of Carbonates
6.2
7.5
10.8
9.9
7.4
9.8
9.8
Urea Consumption for Non-
Agricultural Purposes
3.8
3.7
5.3
5.2
6.0
6.0
6.0
Carbon Dioxide Consumption
1.5
1.4
4.6
4.6
4.1
4.9
5.0
Glass Production
2.3
2.4
2.1
2.0
2.0
1.9
1.9
Aluminum Production
6.8
4.1
1.3
1.2
1.5
1.9
1.7
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
Zinc Production
0.6
1.0
0.8
0.9
1.0
1.0
1.0
Phosphoric Acid Production
1.5
1.3
1.0
1.0
0.9
0.9
0.9
Lead Production
0.5
0.6
0.5
0.5
0.5
0.5
0.5
Carbide Production and
Consumption
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Magnesium Production and
Processing
0.1
+
+
+
+
+
+
ch4
0.3
0.1
0.3
0.3
0.3
0.4
0.3
Petrochemical Production
0.2
0.1
0.2
0.3
0.3
0.3
0.3
Carbide Production and
Consumption
+
+
+
+
+
+
+
Ferroalloy Production
+
+
+
+
+
+
+
Iron and Steel Production &
Metallurgical Coke Production
+
+
+
+
+
+
+
Iron and Steel Production
+
+
+
+
+
+
+
Metallurgical Coke Production
0.0
0.0
0.0
0.0
0.0
0.0
0.0
n2o
33.3
24.9
23.4
22.7
26.0
21.1
23.3
Nitric Acid Production
12.1
11.3
10.1
9.3
9.6
10.0
9.3
AdipicAcid Production
15.2
7.1
7.1
7.5
10.5
5.3
8.3
N20 from Product Uses
4.2
4.2
4.2
4.2
4.2
4.2
4.2
Caprolactam, Glyoxal, and Glyoxylic
Acid Production
1.7
2.1
1.7
1.5
1.4
1.4
1.2
Electronics Industry
+
0.1
0.2
0.3
0.3
0.2
0.3
HFCs
46.5
127.4
168.3
171.1
171.0
175.9
178.8
Substitution of Ozone Depleting
Substances3
0.2
107.2
165.1
165.5
167.3
171.8
176.2
HCFC-22 Production
46.1
20.0
2.8
5.2
3.3
3.7
2.1
Electronics Industry
0.2
0.2
0.3
0.4
0.4
0.4
0.4
Magnesium Production and
Processing
0.0
0.0
0.1
0.1
0.1
0.1
0.1
PFCs
24.3
6.7
4.4
4.2
4.8
4.6
4.4
Electronics Industry
2.8
3.3
3.0
3.0
3.1
2.8
2.7
Aluminum Production
21.5
3.4
1.4
1.1
1.6
1.8
1.7
4-4 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Substitution of Ozone Depleting
Substances
0.0
+
+
+
0.1
0.1
0.1
Electrical Transmission and
Distribution
0.0
+
+
+
0.0
+
+
sf6
28.8
11.8
6.0
5.9
5.7
5.9
5.4
Electrical Transmission and
Distribution
23.2
8.3
4.1
4.2
3.8
4.2
3.8
Magnesium Production and
Processing
5.2
2.7
1.1
1.0
1.0
0.9
0.9
Electronics Industry
0.5
0.7
0.8
0.7
0.8
0.8
0.7
nf3
+
0.5
0.6
0.6
0.6
0.6
0.6
Electronics Industry
+
0.5
0.6
0.6
0.6
0.6
0.6
Total
346.2
365.9
369.0
369.4
373.4
379.5
376.4
Note: Totals may not sum due to independent rounding.
+ Does not exceed 0.05 MMT C02 Eq.
a Small amounts of PFC emissions also result from this source.
Table 4-2: Emissions from Industrial Processes and Product Use (kt)
Gas/Source
1990
2005
2016
2017
2018
2019
2020
co2
213,017
194,389
165,969
164,660
165,086
171,154
163,571
Iron and Steel Production &
Metallurgical Coke Production
104,737
70,076
43,621
40,566
42,627
43,090
37,731
Iron and Steel Production
99,129
66,156
40,979
38,587
41,345
40,084
35,407
Metallurgical Coke Production
5,608
3,921
2,643
1,978
1,282
3,006
2,324
Cement Production
33,484
46,194
39,439
40,324
38,971
40,896
40,688
Petrochemical Production
21,611
27,383
28,110
28,890
29,314
30,702
30,011
Ammonia Production
13,047
9,177
10,245
11,112
12,163
12,272
12,717
Lime Production
11,700
14,552
12,630
12,882
13,106
12,112
11,299
Other Process Uses of
Carbonates
6,233
7,459
10,813
9,869
7,351
9,848
9,794
Urea Consumption for Non-
Agricultural Purposes
3,784
3,653
5,330
5,182
6,030
6,044
5,983
Carbon Dioxide Consumption
1,472
1,375
4,640
4,580
4,130
4,870
4,970
Glass Production
2,291
2,432
2,119
2,011
1,989
1,938
1,857
Aluminum Production
6,831
4,142
1,334
1,205
1,451
1,880
1,748
Soda Ash Production
1,431
1,655
1,723
1,753
1,714
1,792
1,461
Ferroalloy Production
2,152
1,392
1,796
1,975
2,063
1,598
1,377
Titanium Dioxide Production
1,195
1,755
1,662
1,688
1,541
1,474
1,340
Zinc Production
632
1,030
838
900
999
1,026
1,008
Phosphoric Acid Production
1,529
1,342
998
1,025
937
909
938
Lead Production
516
553
500
513
513
527
495
Carbide Production and
Consumption
243
213
170
181
184
175
154
Magnesium Production and
Processing
129
3
3
3
2
1
1
ch4
11
4
11
11
13
15
14
Petrochemical Production
9
3
10
10
12
13
13
Carbide Production and
Consumption
1
+
+
+
+
+
+
Ferroalloy Production
1
+
1
1
1
+
+
Iron and Steel Production &
Metallurgical Coke Production
1
1
+
+
+
+
+
Iron and Steel Production
1
1
+
+
+
+
+
Metallurgical Coke Production
0
0
0
0
0
0
0
Industrial Processes and Product Use 4-5
-------
n2o
112
84
79
76
87
71
78
Nitric Acid Production
41
38
34
31
32
34
31
Adipic Acid Production
51
24
24
25
35
18
28
N20 from Product Uses
14
14
14
14
14
14
14
Caprolactam, Glyoxal, and
Glyoxylic Acid Production
6
7
6
5
5
5
4
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
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
Substances
NO
+
+
+
+
+
+
Electrical Transmission and
Distribution
NO
+
+
+
NO
+
+
sf6
1
1
+
+
+
+
+
Electrical Transmission and
Distribution
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 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 that 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
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 2006 IPCC Guidelines, these non-energy uses of
4-6 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
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.
Industrial Processes and Product Use 4-7
-------
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 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 Iron
and Steel Institute (AISI) (specified within each source category). The emission factors used include those derived
from the EPA's GHGRP and application of IPCC default factors. Descriptions of uncertainties and assumptions for
activity data and emission factors are included within the uncertainty discussion sections for each IPPU source
category.
3 See https://www.epa.gov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
4-8 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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).
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
-------
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. Because these "sintering" reactions are highly exothermic, they
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 2020). 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, California, Missouri,
and Florida were the leading cement-producing states in 2020 and accounted for almost 45 percent of total U.S.
production (USGS 2021). Clinker production in 2020 remained at relatively flat levels, compared to 2019 (EPA
2020; USGS 2021). In 2020, shipments of cement were essentially unchanged from 2019, and imports increased by
about 7 percent compared to 2019. In 2020, U.S. clinker production totaled 78,200 kilotons (EPA 2021). The
resulting CO2 emissions were estimated to be 40.7 MMT CO2 Eq. (40,688 kt) (see Table 4-3). In 2020 due to the
COVID-19 pandemic, production of cement was temporarily idled in many localities and countries in response to
the lockdowns imposed to limit the spread of COVID-19. Disruptions in the construction industry affected cement
demand, and several plant openings and expansions were delayed due to the COVID-19 pandemic. The U.S.
cement industry, however, showed no prolonged or widespread negative effects from the COVID-19 pandemic
(USGS 2021).
4-10 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 4-3: CO2 Emissions from Cement Production (MMT CO2 Eq. and kt)
Year
MMTCOz Eq.
kt
1990
33.5
33,484
2005
46.2
46,194
2016
2017
2018
2019
2020
39.4
40.3
39.0
40.9
40.7
39,439
40,324
38,971
40,896
40,688
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 22 percent. Emissions from cement production were at their lowest levels in 2009 (2009 emissions
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 2010, emissions have increased
by about 30 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.
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 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: 2006IPCCGuidelinesTier 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
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 2019, the percentage of facilities not using CEMS was 8 percent.
Methodology and Time-Series Consistency
Industrial Processes and Product Use 4-11
-------
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.
Small amounts of impurities (i.e., not calcium carbonate) may exist in the raw limestone used to produce clinker.
The proportion of these impurities is generally minimal, although a small amount (1 to 2 percent) of magnesium
oxide (MgO) may be desirable as a flux. Per the IPCC Tier 2 methodology, a correction for MgO is not used, since
the amount of MgO from carbonate is likely very small and the assumption of a 100 percent carbonate source of
CaO already yields an overestimation of emissions (IPCC 2006).
The 1990 through 2012 activity data for clinker production were obtained from USGS (Van Oss 2013a, Van Oss
2013b). Clinker production data for 2013 were also obtained from USGS (USGS 2014). USGS compiled the data (to
the nearest ton) through questionnaires sent to domestic clinker and cement manufacturing plants, including
facilities in Puerto Rico. Clinker production values in the current Inventory report utilize GHGRP data for the years
2014 through 2020 (EPA 2021). Clinker production data are summarized in Table 4-4. Details on how this GHGRP
data compares to USGS reported data can be found in the section on QA/QC and Verification.
Table 4-4: Clinker Production (kt)
Year Clinker
1990 64,355
2005 88,783
2016 75,800
2017 77,500
2018 74,900
2019 78,600
202 0 78,200
Notes: Clinker production from 1990
through 2020 includes Puerto Rico
(relevant U.S. Territories).
Methodological approaches were applied to the entire time series to ensure time-series consistency from 1990
through 2020. The methodology for cement production spliced activity data from two different sources: USGS for
1990 through 2013 and GHGRP starting in 2014. 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.
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 CO2 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)..."
4-12 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Uncertainty
The uncertainties contained in these estimates are primarily due to uncertainties in the lime content of clinker and
in the percentage of CKD recycled inside the cement kiln. Uncertainty is also associated with the assumption that
all calcium-containing raw materials are CaCC>3, when a small percentage likely consists of other carbonate and
non-carbonate raw materials. The lime content of clinker varies from 60 to 67 percent; 65 percent is used as a
representative value (Van Oss 2013a). This contributes to the uncertainty surrounding the emission factor for
clinker which has an uncertainty range of ±5 percent with uniform densities (Van Oss 2013b). The amount of CO2
from CKD loss can range from 1.5 to 8 percent depending upon plant specifications, and uncertainty was estimated
at ±3 percent with uniform densities (Van Oss 2013b). Additionally, some amount of CO2 is reabsorbed when the
cement is used for construction. As cement reacts with water, alkaline substances such as calcium hydroxide are
formed. During this curing process, these compounds may react with CO2 in the atmosphere to create calcium
carbonate. This reaction only occurs in roughly the outer 0.2 inches of the total thickness. Because the amount of
CO2 reabsorbed is thought to be minimal, it was not estimated. EPA assigned default uncertainty bounds of ±3
percent for clinker production, based on expert judgment (Van Oss 2013b).
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-5. Based on the
uncertainties associated with total U.S. clinker production, the CO2 emission factor for clinker production, and the
emission factor for additional CO2 emissions from CKD, 2020 CO2 emissions from cement production were
estimated to be between 38.3 and 43.1 MMT CO2 Eq. at the 95 percent confidence level. This confidence level
indicates a range of approximately 6 percent below and 6 percent above the emission estimate of 40.7 MMT CO2
Eq.
Table 4-5: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Cement
Production (MMT CO2 Eq. and Percent)
Source
Gas
2020 Emission Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower Upper
Lower Upper
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.
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).
EPA relied upon the latest guidance from the IPCC on the use of facility-level data in national inventories and
applied a category-specific QC process to compare activity data from EPA's GHGRP with existing data from USGS
surveys. This was to ensure time-series consistency of the emission estimates presented in the Inventory. Total
U.S. clinker production is assumed to have low uncertainty because facilities routinely measure this for economic
reasons and because both USGS and GHGRP take multiple steps to ensure that reported totals are accurate. EPA
verifies annual facility-level GHGRP reports through a multi-step process that is tailored to the reporting industry
(e.g., combination of electronic checks including range checks, statistical checks, algorithm checks, year-to-year
comparison checks, along with manual reviews involving outside data checks) to identify potential errors and
ensure that data submitted to EPA are accurate, complete, and consistent (EPA 2015). Based on the results of the
Industrial Processes and Product Use 4-13
-------
verification process, EPA follows up with facilities to resolve mistakes that may have occurred.10 Facilities are also
required to monitor and maintain records of monthly clinker production per section 98.84 of the GHGRP regulation
(40 CFR 98.84).
EPA's GHGRP requires all facilities producing Portland cement to report greenhouse gas emissions, including CO2
process emissions from each kiln, CO2 combustion emissions from each kiln, CH4 and N2O combustion emissions
from each kiln, and CO2, Cm, and N2O emissions from each stationary combustion unit other than kilns (40 CFR
Part 98 Subpart H). Source-specific quality control measures for the Cement Production category are included in
section 98.84, Monitoring and QA/QC Requirements.
As mentioned above, EPA compares GHGRP clinker production data to the USGS clinker production data. For the
year 2014 and 2020, USGS and GHGRP clinker production data showed a difference of approximately 1 percent. In
2018, the difference was approximately 3 percent. In 2015, 2016, 2017, and 2019, that difference was less than 1
percent between the two sets of activity data. This difference resulted in a difference in emissions compared to
USGS data of about 0.1 MMT CO2 Eq. in 2015, 2016, 2017, and 2019. The information collected by the USGS
National Minerals Information Center surveys continue to be an important data source.
Recalculations Discussion
No recalculations were performed for the 1990 through 2019 portion of the time series.
Planned Improvements
EPA is continuing to evaluate and analyze data reported under EPA's GHGRP that would be useful to improve the
emission estimates for the Cement Production source category. Most cement production facilities reporting under
EPA's GHGRP use Continuous Emission Monitoring Systems (CEMS) to monitor and report CO2 emissions, thus
reporting combined process and combustion emissions from kilns. In implementing further 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, in addition to category-specific QC methods recommended by the 2006 IPCC
Guidelines.11 EPA's long-term improvement plan includes continued assessment of the feasibility of using
additional GHGRP information beyond aggregation of reported facility-level clinker data, in particular
disaggregating the combined process and combustion emissions reported using CEMS, to separately present
national process and combustion emissions streams consistent with IPCC and UNFCCC guidelines. This long-term
planned analysis is still in development and has not been applied for this current Inventory.
Finally, in response to feedback from Portland Cement Association (PCA) during the Public Review comment period
of a previous Inventory, EPA plans to work with PCA to discuss additional long-term improvements to review
methods and data used to estimate CO2 emissions from cement production to account for organic material in the
raw material and to discuss the carbonation that occurs across the duration of the cement product. Work includes
identifying data and studies on the average carbon content for organic materials in kiln feed in the United States
and CO2 reabsorption rates via carbonation for various cement products. This information is not reported by
facilities subject to GHGRP reporting.
10 See GHGRP Verification Fact Sheet https://www.epa.gov/sites/production/files/2Q15-
07/documents/ghgrp verification factsheet.pdf.
11 See IPCC Technical Bulletin on Use of Facility-Specific Data in National Greenhouse Gas Inventories http://www.ipcc-
nggip.iges.or.jp/public/tb/TFl Technical Bulletin l.pdf.
4-14 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
4.2 Lime Production (CRF Source Category
2A2 and 2H3)
Lime is an important manufactured product with many industrial, chemical, and environmental applications. Lime
production involves three main processes: stone preparation, calcination, and hydration. Carbon dioxide (CO2) is
generated during the calcination stage, when limestone—consisting of calcium (CaCOs) and/or magnesium
(MgCC>3) carbonate—is roasted at high temperatures in a kiln to produce calcium oxide (CaO) and CO2. The CO2 is
given off as a gas and is normally emitted to the atmosphere.
CaCO3 —> CaO + C02
Some facilities, however, recover CO2 generated during the production process for use in sugar refining and
precipitated calcium carbonate (PCC) production.12 PCC is used as a filler or coating in the paper, food, and plastic
industries and is derived from reacting hydrated high-calcium quicklime with CO2, a production process that does
not result in net emissions of CO2 to the atmosphere. Emissions from fuels consumed for energy purposes during
the production of lime are included in the Energy chapter.
For U.S. operations, the term "lime" actually refers to a variety of chemical compounds. These include CaO, or
high-calcium quicklime; calcium hydroxide (Ca(OH)2), or hydrated lime; dolomitic quicklime ([CaOMgO]); and
dolomitic hydrate ([Ca(OH)2*MgO] or [Ca(OH)2*Mg(OH)2]).
The current lime market is approximately distributed across five end-use categories, as follows: metallurgical uses,
34 percent; environmental uses, 30 percent; chemical and industrial uses, 21 percent; construction uses, 11
percent; and refractory dolomite, 1 percent (USGS 2020b). The major uses are in steel making, flue gas
desulfurization (FGD) systems at coal-fired electric power plants, construction, and water treatment, as well as
uses in mining, pulp and paper and precipitated calcium carbonate manufacturing. Lime is also used as a CO2
scrubber, and there has been experimentation on the use of lime to capture CO2 from electric power plants. Both
lime (CaO) and limestone (CaCOs) can be used as a sorbent for FGD systems. Emissions from limestone
consumption for FGD systems are reported under Section 4.4 Other Process Uses of Carbonate Production (CRF
Source Category 2A4).
Emissions from lime production have increased and decreased over the time series depending on lime end-use
markets - primarily the steel making industry and FGD systems for utility and industrial plants - and also energy
costs. One significant change to lime end-use since 1990 has been the increase in demand for lime for FGD at coal-
fired electric power plants, which can be attributed to compliance with sulfur dioxide (SO2) emission regulations of
the Clean Air Act Amendments of 1990. Phase I went into effect on January 1,1995, followed by Phase II on
January 1, 2000. To supply lime for the FGD market, the lime industry installed more than 1.8 million tons per year
of new capacity by the end of 1995 (USGS 1996). The need for air pollution controls continued to drive the FGD
lime market, which had doubled between 1990 and 2019 (USGS 1991 and 2020d).
The U.S. lime industry temporarily shut down some individual gas-fired kilns and, in some case, entire lime plants
during 2000 and 2001, due to significant increases in the price of natural gas. Lime production continued to
decrease in 2001 and 2002, a result of lower demand from the steel making industry, lime's largest end-use
market, when domestic steel producers were affected by low priced imports and slowing demand (USGS 2002).
Emissions from lime production increased and then peaked in 2006 at approximately 30.3 percent above 1990
levels, due to strong demand from the steel and construction markets (road and highway construction projects),
before dropping to its lowest level in 2009 at approximately 2.5 percent below 1990 emissions, driven by the
economic recession and downturn in major markets including construction, mining, and steel (USGS 2007, 2008,
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.
Industrial Processes and Product Use 4-15
-------
2010). In 2010, the lime industry began to recover as the steel, FGD, and construction markets also recovered
(USGS 2011 and 2012). Fluctuation in lime production since 2015 has been driven largely by demand from the steel
making industry (USGS 2018b, 2019, 2020b, 2020c). In 2020, annual domestic lime production decreased due to
temporary plant closures as a result of the COVID-19 pandemic (USGS 2021c).
Lime production in the United States—including Puerto Rico—was reported to be 15,862 kilotons in 2020, a
decrease of about 6.1 percent compared to 2019 levels (USGS 2021b). Compared to 1990, lime production
increased by about 0.1 percent. At year-end 2020, 74 primary lime plants were operating in the United States,
including Puerto Rico according to the USGS MCS (USGS 2021a).13 Principal lime producing states were Missouri,
Alabama, Ohio, Texas, and Kentucky (USGS 2021a).
U.S. lime production resulted in estimated net CO2 emissions of 11.3 MMT CO2 Eq. (11,299 kt) (see Table 4-6 and
Table 4-7). Carbon dioxide emissions from lime production decreased by about 6.7 percent compared to 2019
levels. Compared to 1990, CO2 emissions have decreased by about 3.4 percent. The trends in CO2 emissions from
lime production are directly proportional to trends in production, which are described above.
Table 4-6: CO2 Emissions from Lime Production (MMT CO2 Eq. and kt)
Year MMT CP2 Eq. kt
1990 1L7 11,700
2005 14.6 14,552
2016 12.6 12,630
2017 12.9 12,882
2018 13.1 13,106
2019 12.1 12,112
2020 11.3 11,299
Table 4-7: Gross, Recovered, and Net CO2 Emissions from Lime Production (kt)
Year Gross Recovered3 Net Emissions
1990 11,959 259 11,700
2005 15,074 522 14,552
2016 13,000 370 12,630
2017 13,283 401 12,882
2018 13,609 503 13,106
2019 12,676 564 12,112
2020 11,875 576 11,299
Note: Totals may not sum due to independent rounding.
a For sugar refining and PCC production.
Methodology and Time-Series Consistency
To calculate emissions, the amounts of high-calcium and dolomitic lime produced were multiplied by their
respective emission factors using the Tier 2 approach from the 2006IPCC Guidelines. The emission factor is the
product of the stoichiometric ratio between CO2 and CaO, and the average CaO and MgO content for lime. The
13 In 2020, 71 operating primary lime facilities in the United States reported to the EPA Greenhouse Gas Reporting Program,
including three facilities that reported emission values of zero.
4-16 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
CaO and MgO content for lime is assumed to be 95 percent for both high-calcium and dolomitic lime (IPCC 2006).
The emission factors were calculated as follows:
Equation 4-2: 2006IPCCGuide/inesTier 2 Emission Factor for Lime Production, High-
Calcium Lime (Equation 2.9)
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
Equation 4-3: 2006IPCC Guide/inesTier 2 Emission Factor for Lime Production, Dolomitic
Lime (Equation 2.9)
EFDolomitic Lime — [(88.02 g/mole C02) 4 (96.39 g/mole CaO)] x (0.9500 CaO/lime) = 0.8675 g C02/g lime
Production was adjusted to remove the mass of chemically combined water found in hydrated lime, determined
according to the molecular weight ratios of H20 to (Ca(OH)2 and [Ca(OH)2*Mg(OH)2]) (IPCC 2006). These factors set
the chemically combined water content to 27 percent for high-calcium hydrated lime, and 30 percent for dolomitic
hydrated lime.
The 2006 IPCC Guidelines (Tier 2 method) also recommends accounting for emissions from lime kiln dust (LKD)
through application of a correction factor. LKD is a byproduct of the lime manufacturing process typically not
recycled back to kilns. LKD is a very fine-grained material and is especially useful for applications requiring very
small particle size. Most common LKD applications include soil reclamation and agriculture. Emissions from the
application of lime for agricultural purposes are reported in the Agriculture chapter under 5.5 Liming (CRF Source
Category 3G). Currently, data on annual LKD production is not readily available to develop a country-specific
correction factor. Lime emission estimates were multiplied by a factor of 1.02 to account for emissions from LKD
(IPCC 2006). See the Planned Improvements section associated with efforts to improve uncertainty analysis and
emission estimates associated with LKD.
Lime emission estimates were further adjusted to account for the amount of C02 captured for use in on-site
processes. All the domestic lime facilities are required to report these data to EPA under its GHGRP. The total
national-level annual amount of C02 captured for on-site process use was obtained from EPA's GHGRP (EPA 2021)
based on reported facility-level data for years 2010 through 2020. The amount of C02 captured/recovered for on-
site process use is deducted from the total gross emissions (i.e., from lime production and LKD). The net lime
emissions are presented in Table 4-6 and Table 4-7. GHGRP data on C02 removals (i.e., C02 captured/recovered)
was available only for 2010 through 2020. Since GHGRP data are not available for 1990 through 2009, IPCC
"splicing" techniques were used as per the 2006 IPCC Guidelines on time-series consistency (IPCC 2006, Volume 1,
Chapter 5).
Lime production data by type (i.e., high-calcium and dolomitic quicklime, high-calcium and dolomitic hydrated
lime, and dead-burned dolomite) for 1990 through 2020 (see Table 4-8) were obtained from U.S. Geological Survey
(USGS) Minerals Yearbook (USGS 1992 through 2021b) and are compiled by USGS to the nearest ton. Dead-burned
dolomite data are additionally rounded by USGS to no more than one significant digit to avoid disclosing company
proprietary data. Natural hydraulic lime, which is produced from CaO and hydraulic calcium silicates, is not
manufactured in the United States (USGS 2018a). Total lime production was adjusted to account for the water
content of hydrated lime by converting hydrate to oxide equivalent based on recommendations from the IPCC and
using the water content values for high-calcium hydrated lime and dolomitic hydrated lime mentioned above, and
is presented in Table 4-9 (IPCC 2006). The CaO and CaO*MgO contents of lime, both 95 percent, were obtained
from the IPCC (IPCC 2006). Since data for the individual lime types (high calcium and dolomitic) were not provided
prior to 1997, total lime production for 1990 through 1996 was calculated according to the three-year distribution
from 1997 to 1999.
Industrial Processes and Product Use 4-17
-------
Table 4-8: High-Calcium- and Dolomitic-Quicklime, High-Calcium- and Dolomitic-Hydrated,
and Dead-Burned-Dolomite Lime Production (kt)
High-Calcium Dolomitic High-Calcium Dolomitic Dead-Burned
Year Quicklime Quicklime Hydrated Hydrated Dolomite
1990 11,166 2,234 1,781 319 342
2005 14,100 2,990 2,220 474 200
2016 12,100 2,420 2,350 280 200
2017 12,200 2,650 2,360 276 200
2018 12,400 2,810 2,430 265 200
2019 11,300 2,700 2,430 267 200
2020 10,700 2,390 2,320 252 200
Table 4-9: Adjusted Lime Production (kt)
Year High-Calcium Dolomitic
1990 12,466 2,800
2005 15,721 3,522
2016 13,816 2,816
2017 13,923 3,043
2018 14,174 3,196
2019 13,074 3,087
2020 12,394 2,766
Note: Minus water content of hydrated lime.
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2020.
Uncertainty
The uncertainties contained in these estimates can be attributed to slight differences in the chemical composition
of lime products and CO2 recovery rates for on-site process use over the time series. Although the methodology
accounts for various formulations of lime, it does not account for the trace impurities found in lime, such as iron
oxide, alumina, and silica. Due to differences in the limestone used as a raw material, a rigid specification of lime
material is impossible. As a result, few plants produce lime with exactly the same properties.
In addition, a portion of the CO2 emitted during lime production will actually be reabsorbed when the lime is
consumed, especially at captive lime production facilities. As noted above, lime has many different chemical,
industrial, environmental, and construction applications. In many processes, CO2 reacts with the lime to create
calcium carbonate (e.g., water softening). Carbon dioxide reabsorption rates vary, however, depending on the
application. For example, 100 percent of the lime used to produce precipitated calcium carbonate reacts with CO2,
whereas most of the lime used in steel making reacts with impurities such as silica, sulfur, and aluminum
compounds. Quantifying the amount of CO2 that is reabsorbed would require a detailed accounting of lime use in
the United States and additional information about the associated processes where both the lime and byproduct
CO2 are "reused." Research conducted thus far has not yielded the necessary information to quantify CO2
4-18 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
reabsorption rates.14 Some additional information on the amount of CO2 consumed on site at lime facilities,
however, has been obtained from EPA's GHGRP.
In some cases, lime is generated from calcium carbonate byproducts at pulp mills and water treatment plants.15
The lime generated by these processes is included in the USGS data for commercial lime consumption. In the
pulping industry, mostly using the Kraft (sulfate) pulping process, lime is consumed in order to causticize a process
liquor (green liquor) composed of sodium carbonate and sodium sulfide. The green liquor results from the dilution
of the smelt created by combustion of the black liquor where biogenic carbon (C) is present from the wood. Kraft
mills recover the calcium carbonate "mud" after the causticizing operation and calcine it back into lime—thereby
generating CO2—for reuse in the pulping process. Although this re-generation of lime could be considered a lime
manufacturing process, the CO2 emitted during this process is mostly biogenic in origin and therefore is not
included in the industrial processes totals (Miner and Upton 2002). In accordance with IPCC methodological
guidelines, any such emissions are calculated by accounting for net C fluxes from changes in biogenic C reservoirs
in wooded or crop lands (see the Land Use, Land-Use Change, and Forestry chapter).
In the case of water treatment plants, lime is used in the softening process. Some large water treatment plants
may recover their waste calcium carbonate and calcine it into quicklime for reuse in the softening process. Further
research is necessary to determine the degree to which lime recycling is practiced by water treatment plants in the
United States.
Another uncertainty is the assumption that calcination emissions for LKD are around 2 percent. The National Lime
Association (NLA) has commented that the estimates of emissions from LKD in the United States could be closer to
6 percent. They also note that additional emissions (approximately 2 percent) may also be generated through
production of other byproducts/wastes (off-spec lime that is not recycled, scrubber sludge) at lime plants (Seeger
2013). Publicly available data on LKD generation rates, total quantities not used in cement production, and types of
other byproducts/wastes produced at lime facilities are limited. NLA compiled and shared historical emissions
information and quantities for some waste products reported by member facilities associated with generation of
total calcined byproducts and LKD, as well as methodology and calculation worksheets that member facilities
complete when reporting. There is uncertainty regarding the availability of data across the time series needed to
generate a representative country-specific LKD factor. Uncertainty of the activity data is also a function of the
reliability and completeness of voluntarily reported plant-level production data. Further research, including
outreach and discussion with NLA, and data is needed to improve understanding of additional calcination
emissions to consider revising the current assumptions that are based on IPCC guidelines. More information can be
found in the Planned Improvements section below.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-10. Lime CO2 emissions
for 2020 were estimated to be between 11.1 and 11.5 MMT CO2 Eq. at the 95 percent confidence level. This
confidence level indicates a range of approximately 2 percent below and 2 percent above the emission estimate of
11.3 MMT CO2 Eq.
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 4-19
-------
Table 4-10: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Lime
Production (MMT CO2 Eq. and Percent)
Source
Gas
2020 Emission Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower Upper
Lower
Upper
Bound Bound
Bound
Bound
Lime Production
C02
11.3
11.1 11.5
-2%
+2%
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 noted 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 associated with reporting on CO2
captured for onsite use applicable to lime manufacturing facilities can be found under Subpart S (Lime
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 2020).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 2019 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
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.jp/public/tb/TFI Technical Bulletin l.pdf.
4-20 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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.
4.3 Glass Production (CRF Source Category
2A3)
Glass production is an energy and raw-material intensive process that results in the generation of carbon dioxide
(CO2) from both the energy consumed in making glass and the glass production process itself. Emissions from fuels
consumed for energy purposes during the production of glass are included in the Energy sector.
Glass production employs a variety of raw materials in a glass-batch. These include formers, fluxes, stabilizers, and
sometimes colorants. The major raw materials (i.e., fluxes and stabilizers) that emit process-related CO2 emissions
during the glass melting process are limestone, dolomite, and soda ash. The main former in all types of glass is
silica (SiCh). Other major formers in glass include feldspar and boric acid (i.e., borax). Fluxes are added to lower the
temperature at which the batch melts. Most commonly used flux materials are soda ash (sodium carbonate,
Na2CC>3) and potash (potassium carbonate, K2O). Stabilizers make glass more chemically stable and keep the
finished glass from dissolving and/or falling apart. Commonly used stabilizing agents in glass production are
limestone (CaCOs), dolomite (CaCOsMgCOs), alumina (AI2O3), magnesia (MgO), barium carbonate (BaCOs),
strontium carbonate (SrCOs), lithium carbonate (Li2CC>3), and zirconia (ZrCh) (DOE 2002). Glass makers also use a
certain amount of recycled scrap glass (cullet), which comes from in-house return of glassware broken in the
production process or other glass spillage or retention, such as recycling or from cullet broker services.
The raw materials (primarily soda ash, limestone, and dolomite) release CO2 emissions in a complex high-
temperature chemical reaction during the glass melting process. This process is not directly comparable to the
calcination process used in lime manufacturing, cement manufacturing, and process uses of carbonates (i.e.,
limestone/dolomite use) but has the same net effect in terms of CO2 emissions (IPCC 2006).
The U.S. glass industry can be divided into four main categories: containers, flat (window) glass, fiber glass, and
specialty glass. The majority of commercial glass produced is container and flat glass (EPA 2009). The United States
is one of the major global exporters of glass. Domestically, demand comes mainly from the construction, auto,
bottling, and container industries. There are more than 1,700 facilities that manufacture glass in the United States,
with the largest companies being Corning, Guardian Industries, Owens-Illinois, and PPG Industries.19
The glass container sector is one of the leading soda ash consuming sectors in the United States. In 2020, glass
production accounted for 48 percent of total domestic soda ash consumption (USGS 2021). Emissions from soda
ash production are reported in 4.12 Soda Ash Production (CRF Source Category 2B7).
In 2020, 2,130 kilotons of soda ash, 1,334 kilotons of limestone, and 824 kilotons of dolomite were consumed for
glass production (USGS 2021; EPA 2021). Use of soda ash, limestone, and dolomite in glass production resulted in
aggregate CO2 emissions of 1.9 MMT CO2 Eq. (1,857 kt) (see Table 4-11). Overall, emissions have decreased by 19
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.
Industrial Processes and Product Use 4-21
-------
percent compared to 1990. Glass production and emissions decreased by about 4 percent compared to 2019
levels.
Emissions from glass production have remained relatively constant over the time series with some fluctuations
since 1990. In general, these fluctuations were related to the behavior of the export market and the U.S. economy.
Specifically, the extended downturn in residential and commercial construction and automotive industries
between 2008 and 2010 resulted in reduced consumption of glass products, causing a drop in global demand for
limestone/dolomite and soda ash and resulting in lower emissions. Some commercial food and beverage package
manufacturers are shifting from glass containers towards lighter and more cost-effective polyethylene
terephthalate (PET) based containers, putting downward pressure on domestic consumption of soda ash (USGS
1995 through 2015b). Due to the COVID-19 pandemic, glass production dropped in the spring of 2020 but mostly
rebounded by the end of the year (Federal Reserve 2021).
Table 4-11: CO2 Emissions from Glass Production (MMT CO2 Eq. and kt)
Year
MMT CO? Eq.
kt
1990
2.3
2,291
2005
2.4
2,432
2016
2.1
2,119
2017
2.0
2,011
2018
2.0
1,989
2019
1.9
1,938
2020
1.9
1,857
Methodology and Time-Series Consistency
Carbon dioxide emissions were calculated based on the 2006IPCC Guidelines Tier 3 method by multiplying the
quantity of input carbonates (limestone, dolomite, and soda ash) by the IPCC default carbonate-based emission
factor (in metric tons CCh/metric ton carbonate).
The methodology for estimating CO2 emissions from the use of soda ash for glass production remains unchanged
for 1990 to 2020. This methodology continues to assume that soda ash contains 100 percent sodium carbonate
(Na2CC>3), consistent with 2006 IPCC Guidelines and the previous methodology. For 1990 through 2020, data on
soda ash used for glass manufacturing were obtained from the U.S. Bureau of Mines (1991 and 1993a), the USGS
Minerals Yearbook: Soda Ash Annual Report (USGS 1995 through 2015b), and USGS Mineral Industry Surveys for
Soda Ash (USGS 2017 through 2021).
2010 through 2020
For this Inventory, the methodology for estimating CO2 emissions from the use of limestone and dolomite for glass
production for years 2010 through 2020 has changed to use new activity data reported to the U.S. EPA
Greenhouse Gas Reporting Program (GHGRP) on the quantities of limestone and dolomite used for glass
production (EPA 2021). USGS data on the quantity of soda ash used for glass production continues to be used
because it was obtained directly from the soda ash producers and includes use by smaller artisanal glass
operations, which are excluded in the GHGRP data.
GHGRP collects data from glass production facilities with greenhouse gas emissions greater than 25,000 metric
tons CO2 Eq. The reporting threshold is used to exclude artisanal glass operations that are expected to have much
lower greenhouse gas emissions than the threshold. These smaller facilities have not been accounted for yet for
this portion of the time series due to limited data. Facilities report the total quantity of each type of carbonate
(e.g., limestone, dolomite, soda ash) used in glass production each year to GHGRP, with data collection starting in
2010 (EPA 2021).
4-22 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Using the total quantities of each carbonate, EPA calculated the metric tons of emissions resulting from glass
production by multiplying the quantity of input carbonates (i.e., limestone, dolomite, and soda ash) by IPCC default
carbonate-based emission factors (in metric tons CCh/metric ton carbonate): limestone, 0.43971; dolomite,
0.47732; and soda ash, 0.41492 and by the average carbonate-based mineral mass fraction for each year. The
average carbonate-based mineral mass fractions from the GHGRP, averaged across 2010 through 2020, indicate
that the limestone used in glass production contained 98.6 percent calcium carbonate (CaCOs) and dolomite
contained 98.5 percent calcium magnesium carbonate (CaMg(CC>3)2). The previous methodology assumed that
limestone contained 100 percent CaCC>3 and dolomite contained 100 percent CaMg(CC>3)2. This methodology
continues to assume that soda ash contains 100 percent sodium carbonate (Na2COs), consistent with 2006 IPCC
Guidelines and the previous methodology.
1990 through 2009
Data from GHGRP on the quantity of limestone and dolomite used in glass production is not available for 1990
through 2009. USGS and GHGRP datasets for 2010 through 2020 showed inconsistent overlap, and using USGS
data for 1990 through 2009 would have introduced inconsistencies over the time series.
To address time-series consistency, total emissions from 1990 to 2009 were calculated using the Federal Reserve
Industrial Production Index for glass production in the United States as a surrogate for the total quantity of
carbonates used in glass production. The production index measures real output expressed as a percentage of real
output in a base year, which is currently 2017 (Federal Reserve 2021). Since January 1971, the Federal Reserve has
released the monthly glass production index for NAICS code 3272 (Glass and Glass Product Manufacturing) as part
of release G.17, "Industrial Production and Capacity Utilization" (Federal Reserve 2021). The monthly index values
for each year were averaged to calculate an average annual glass production index value. Total annual emissions
were calculated by taking a ratio of the average annual glass production index for each year, with a base year of
2017, and the calculated 2017 emissions based on GHGRP data.
Emissions from limestone and dolomite consumption were disaggregated from total annual emissions, using the
average percent contribution of each carbonate to total annual emissions for 2010 through 2020 based on GHGRP
data: 32.1 percent limestone and 19.0 percent dolomite. A comparison of the 1990 to 2009 methodology applied
to 2010 to 2020 and the calculated emissions based on GHGRP data of quantities of carbonates consumed for glass
production for 2010 to 2020 showed that these two methods are closely correlated. The methodology for
estimating CO2 emissions from the use of soda ash for glass production and data sources for the amount of soda
ash used in glass production are described above.
The amount of limestone, dolomite, and soda ash used in glass production each year and the annual average
Federal Reserve production indices for glass production are shown in Table 4-12.
Table 4-12: Limestone, Dolomite, and Soda Ash Used in Glass Production (kt) and Average
Annual Production Index for Glass and Glass Product Manufacturing
Activity
1990
2005
2016
2017
2018
2019
2020
Limestone
1,391
1,668
1,560
1,488
1,442
1,370
1,334
Dolomite
757
908
836
806
871
883
824
Soda Ash
3,177
3,050
2,510
2,360
2,280
2,220
2,130
Total
5,325
5,626
4,906
4,653
4,593
4,473
4,287
Production Index3
94.3
113.1
102.6
100
102.5
100
91.1
a Average Annual Production Index uses 2017 as the base year.
Note: Totals may not sum due to independent rounding.
As discussed above, methodological approaches were applied to the entire time series to ensure consistency in
emissions from 1990 through 2020. Consistent with the 2006 IPCC Guidelines, the overlap technique was applied
to compare USGS and GHGRP data sets for 2010 through 2020. To address the inconsistencies, adjustments were
made as described above.
Industrial Processes and Product Use 4-23
-------
Uncertainty
The methodology and activity data used in this Inventory reduced uncertainty for glass production, compared to
the previous Inventory. Uncertainty levels presented in this section in previous Inventories arose in part due to
variations in the chemical composition of limestone used in glass production. For example in addition to calcium
carbonate, limestone may contain smaller amounts of magnesia, silica, and sulfur, among other minerals (e.g.,
potassium carbonate, strontium carbonate and barium carbonate, and dead burned dolomite). The methodology
in this Inventory report uses GHGRP data on the average mass fraction of each mineral in the limestone and
dolomite used in glass production for each year from 2010-2020.
The data and methodology used in this Inventory report also reduce uncertainty associated with activity data. The
methodology uses the amount of limestone and dolomite used in glass manufacturing which is reported directly by
the glass manufacturers for years 2010 through 2020 and the amount of soda ash used in glass manufacturing
which is reported by soda ash producers for the full time series. The emissions from other carbonates reported to
GHGRP-barium carbonate (BaCOs), potassium carbonate (K2CO3), lithium carbonate (IJ2CO3), and strontium
carbonate (SrCOs)-are not included in these estimates.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-13. In 2020, glass
production CO2 emissions were estimated to be between 1.8 and 1.9 MMT CO2 Eq. at the 95 percent confidence
level. This indicates a range of approximately 2 percent below and 2 percent above the emission estimate of 1.9
MMTCO2 Eq.
Table 4-13: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Glass
Production (MMT CO2 Eq. and Percent)
Source
Gas
2020 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT CO? Eq.)
(MMT CO? Eq.)
(%)
Lower Upper
Lower Upper
Bound Bound
Bound Bound
Glass Production
C02
1.9
cn
T—1
00
T—1
-2% +2%
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 2015).20 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
For the current Inventory, a new methodology using more complete activity data from GHGRP for 2010 through
2020 and the industrial production index for glass and glass product manufacturing from the Federal Reserve for
20 GHGRP Report Verification Factsheet. See https://www.epa.gov/sites/production/files/2Q15-
07/documents/ghgrp verification factsheet.pdf.
4-24 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
1990 through 2009 to address time-series consistency were implemented and is described under the Methodology
and Time-Series Consistency section. The revised values for 1990 through 2019 resulted in increased emissions
estimates for all years. Across the time series, emissions increased by an average of 52 percent compared to the
previous Inventory. Annual emission increases during the time series ranged from an 18 percent increase in 2006
(373 kt) to a 91 percent increase in 1999 (1,238 kt).
Planned Improvements
EPA incorporated data from GHGRP on limestone and dolomite used for glass production into the emissions
estimates for the Glass Production source category for 1990 through 2020. EPA continues to evaluate and analyze
data reported under GHGRP that would be useful to improve the emission estimates for the Glass Production
source category, particularly the use of barium carbonate, potassium carbonate, lithium carbonate, and strontium
carbonate for glass production. EPA will also evaluate updates to uncertainty levels for the activity data and
mineral mass fraction values from EPA's GHGRP. This is a near-term planned improvement.
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 assess how to refine the methodology to
ensure complete national coverage of this category. 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 4.2, 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 4.2 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 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).
21 Limestone and dolomite are collectively referred to as limestone by the industry, and intermediate varieties are seldom
distinguished.
Industrial Processes and Product Use 4-25
-------
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 2017, the leading limestone producing states were Texas, Florida, Missouri,
Ohio, and Pennsylvania, which contributed 44 percent of the total U.S. output (USGS 2021a). 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 2021a). 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).
In 2020,15,346 kilotons (kt) of limestone, 4,374 kt of dolomite, and 2,310 kt of soda ash were consumed for these
emissive applications, which excludes consumption for the production of cement, lime, glass, and iron and steel
(Willett 2021, USGS 2021d). Limestone and dolomite consumption data for 2020 were not available in time for
publication and were estimated using 2019 values, as described in the Methodology and Time-Series Consistency
section below. Usage of limestone, dolomite and soda ash resulted in aggregate CO2 emissions of 9.8 MMT CO2 Eq.
(9,794 kt) (see Table 4-14 and Table 4-15). The 2019 and 2020 emissions increased over 30 percent compared to
2018, primarily as a result of increased limestone consumption attributed to sulfur oxide removal usage for FGD
systems and dolomite consumption attributed to flux stone. Disruptions in the mining and construction industries
associated with the COVID-19 pandemic led to decreased consumption of crushed stone in 2020; however, the
impacts on emissions from limestone and dolomite consumption are not able to be quantified without more
detailed information on consumption from the emissive sources in 2020 (USGS 2021b). Overall emissions have
increased 57 percent from 1990 through 2020.
Table 4-14: CO2 Emissions from Other Process Uses of Carbonates (MMT CO2 Eq.)
Other
Flux
Soda Ash
Miscellaneous
Year
Stone
FGD
Consumption3
Usesb
Total
1990
2.6
1.4
1.4
0.8
6.2
2005
2.6
3.0
1.3
0.5
7.5
2016
2.6
6.2
1.1
1.0
10.8
2017
2.4
5.6
1.1
0.8
9.9
2018
2.8
2.2
1.1
1.3
7.4
2019
4.8
3.5
1.0
0.5
9.8
2020
4.8
3.5
1.0
0.5
9.8
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.
4-26 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 4-15: CO2 Emissions from Other Process Uses of Carbonates (kt)
Other
Flux
Soda Ash
Miscellaneous
Year
Stone
FGD
Consumption3
Usesb
Total
1990
2,592
1,432
1,390
819
6,233
2005
2,649
2,973
1,305
533
7,459
2016
2,585
6,164
1,082
981
10,813
2017
2,441
5,598
1,058
771
9,869
2018
2,795
2,229
1,069
1,259
7,351
2019
4,811
3,537
1,036
463
9,848
2020
4,835
3,537
958
463
9,794
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.
Methodology and Time-Series Consistency
Carbon dioxide emissions were calculated based on the 2006IPCC Guidelines Tier 2 method by multiplying the
quantity of limestone or dolomite consumed by the emission factor for limestone or dolomite calcination,
respectively: 0.43971 metric ton CCh/metric ton carbonate for limestone and 0.47732 metric ton CCh/metric ton
carbonate for dolomite.22 This methodology was used for flux stone, flue gas desulfurization systems, chemical
stone, mine dusting or acid water treatment, and acid neutralization. Flux stone used during the production of iron
and steel was deducted from the Other Process Uses of Carbonates source category estimate and attributed to the
Iron and Steel Production source category estimate. Similarly, limestone and dolomite consumption for glass
manufacturing, cement, and lime manufacturing are excluded from this category and attributed to their respective
categories.
Consumption data for 1990 through 2019 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
2017, 2020a, 2020c), preliminary data for 2019 from USGS Crushed Stone Commodity Expert (Willett 2021),
American Iron and Steel Institute limestone and dolomite consumption data (AISI 2018 through 2020), and the U.S.
Bureau of Mines (1991 and 1993a), which are reported to the nearest ton. Limestone and dolomite consumption
data for 2020 were not available at the time of publication and were estimated using 2019 values. In addition, the
estimated values for limestone and dolomite consumption for flux stone used during the production of iron and
steel were adjusted down, using emissions data from the EPA's Greenhouse Gas Reporting Program (GHGRP)
subpart Q for the iron and steel sector to account for the impacts of the COVID-19 pandemic in 2020. GHGRP
process emissions data decreased by approximately 14 percent from 2019 to 2020 (EPA 2021). This adjustment
method is consistent with the method used in Section 4.17 (CRF Source Category 2C1) and Metallurgical Coke
Production. Similar data on 2020 emissions trends were not available for the other process uses included in this
section, which prevented the use of a similar approach.
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.
22 2006 IPCC Guidelines, Volume 3: Chapter 2, Table 2.1.
Industrial Processes and Product Use 4-27
-------
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
2016
2017
2018
2019
2020
Flux Stone
5,842
5,745
5,686
5,447
6,242
10,570
10,622
Limestone
5,237
2,492
3,415
4,216
4,891
6,222
6,248
Dolomite
605
3,254
2,270
1,230
1,351
4,348
4,374
FGD
3,258
6,761
14,019
12,732
5,068
8,045
8,045
Other Miscellaneous Uses
1,835
1,212
2,231
1,754
2,862
1,054
1,054
Total
10,935
13,719
21,935
19,932
14,172
19,668
19,720
Note: Totals may not sum due to independent rounding.
Excluding glass manufacturing which is reported under Section 4.2 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,
approximately 0.113 metric tons of carbon (or 0.415 metric tons of CO2) are released for every metric ton of soda
ash consumed. The activity data for soda ash consumption for 1990 to 2020 (see Table 4-17) were obtained from
the U.S. Geological Survey (USGS) Minerals Yearbook for Soda Ash (1994 through 2015b) and USGS Mineral
Industry Surveys for Soda Ash (USGS 2017a, 2018, 2019, 2020b, 2021d). Soda ash consumption data were collected
by the USGS from voluntary surveys of the U.S. soda ash industry.
Table 4-17: Soda Ash Consumption Not Associated with Glass Manufacturing (kt)
Activity
1990
2005
2016
2017
2018
2019
2020
Soda Asha
3,351
3,144
2,608
2,550
2,576
2,497
2,310
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).
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2020.
23 This approach was recommended by USGS, the data collection agency.
4-28 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Uncertainty
The uncertainty levels presented in this section account for uncertainty associated with activity data. Data on
limestone and dolomite consumption are collected by USGS through voluntary national surveys. USGS contacts the
mines (i.e., producers of various types of crushed stone) for annual sales data. Data on other carbonate
consumption are not readily available. The producers report the annual quantity sold to various end-users and
industry types. USGS estimates the historical response rate for the crushed stone survey to be approximately 70
percent, and the rest is estimated by USGS. Large fluctuations in reported consumption exist, reflecting year-to-
year changes in the number of survey responders. The uncertainty resulting from a shifting survey population is
exacerbated by the gaps in the time series of reports. The accuracy of distribution by end use is also uncertain
because this value is reported by the producer/mines and not the end user. Additionally, there is significant
inherent uncertainty associated with estimating withheld data points for specific end uses of limestone and
dolomite. Lastly, much of the limestone consumed in the United States is reported as "other unspecified uses;"
therefore, it is difficult to accurately allocate this unspecified quantity to the correct end-uses. EPA contacted the
USGS National Minerals Information Center Crushed Stone commodity expert to assess the current uncertainty
ranges associated with the limestone and dolomite consumption data compiled and published by USGS. During
this discussion, the expert confirmed that EPA's range of uncertainty was still reasonable (Willett 2017).
Uncertainty in the estimates also arises in part due to variations in the chemical composition of limestone. In
addition to calcium carbonate, limestone may contain smaller amounts of magnesia, silica, and sulfur, among
other minerals. The exact specifications for limestone or dolomite used as flux stone vary with the
pyrometallurgical process and the kind of ore processed.
For emissions from soda ash consumption, the primary source of uncertainty results from the fact that these
emissions are dependent upon the type of processing employed by each end-use. Specific emission factors for
each end-use are not available, so a Tier 1 default emission factor is used for all end-uses. Therefore, there is
uncertainty surrounding the emission factors from the consumption of soda ash. Additional uncertainty comes
from the reported consumption and allocation of consumption within sectors that is collected on a quarterly basis
by the USGS. Efforts have been made to categorize company sales within the correct end-use sector.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-18. Carbon dioxide
emissions from other process uses of carbonates in 2020 were estimated to be between 8.2 and 12.9 MMT CO2 Eq.
at the 95 percent confidence level. This indicates a range of approximately 19 percent below and 28 percent above
the emission estimate of 9.8 MMT CO2 Eq.
Table 4-18: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Other
Process Uses of Carbonates (MMT CO2 Eq. and Percent)
2020 Emission
Source
Gas
Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT CO? Eq.)
(MMT CO?
Eq.)
(%)
Lower
Upper
Lower Upper
Bound
Bound
Bound Bound
Other Process Uses
of Carbonates
C02
9.8
8.2
12.9
-19% +28%
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).
Industrial Processes and Product Use 4-29
-------
Recalculations Discussion
Emissions from carbonate consumption for magnesium metal production previously included in this chapter have
been moved from Other Process Uses of Carbonates to Section 4.20 Magnesium Production and Processing (CRF
Source Category 2C4) in the current Inventory, consistent with the 2006IPCC Guidelines. Emissions were removed
from this chapter for 1990 through 2001, resulting in approximately 50 to 70 kt CO2 reduction for these years.
Emissions previously included in this chapter for limestone consumption for sugar refining have been removed in
the current inventory, as it was determined that these emissions are already accounted for in the Lime Production
source category emissions. Emissions were removed from this chapter for 1990 through 2019, resulting in a range
of 0 to 1,500 kt CO2 reduction for these years.
Additionally, for the current Inventory, updated USGS data on limestone and dolomite consumption was available
for 2019, resulting in updated emissions estimates for that year. Compared to the previous Inventory, emissions
for 2019 increased by 32 percent (2,391 kt CO2 Eq.).
Planned Improvements
In response to comments received during previous Inventory reports from the UNFCCC, EPA has inquired to the
availability of ceramics and non-metallurgical magnesia data. The USGS notes that this data is not currently
reported by survey respondents. EPA continues to conduct outreach with other entities, but at this time, the
research has not yielded any alternative data on national levels of carbonates. This improvement remains ongoing,
and EPA plans to continue to update this Planned Improvements section in future reports as more information
becomes available.
EPA also plans to continue dialogue with USGS to assess uncertainty ranges for activity data used to estimate
emissions from other process use of carbonates. This planned improvement is currently planned as a medium-
term improvement.
4.5 Ammonia Production (CRF Source
Category 2B1)
Emissions of carbon dioxide (CO2) occur during the production of synthetic ammonia (NH3), primarily through the
use of natural gas, petroleum coke, or naphtha as a feedstock. The natural gas-, naphtha-, and petroleum coke-
based processes produce CO2 and hydrogen (H2), the latter of which is used in the production of ammonia. The
brine electrolysis process for production of ammonia does not lead to process-based CO2 emissions. Due to
national circumstances, emissions from fuels consumed for energy purposes during the production of ammonia
are accounted for in the Energy chapter. More information on this approach can be found in the Methodology
section below.
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 2020,16 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
2021).
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
4-30 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
catalyst. Only 30 to 40 percent of the Cm 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 + 3H2 -> 2NH3
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 [COfNHhh], which has a variety of agricultural and industrial applications.
The chemical reaction that produces urea is:
2nh3+ C02 -> 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 4.5 Urea Consumption for Non-
Agricultural Purposes of this chapter.
Total emissions of CO2 from ammonia production in 2020 were 12.7 MMT CO2 Eq. (12,717 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,
emissions from ammonia production have decreased by about 3 percent. Emissions in 2020 increased by about 4
percent from the 2019 levels.
Emissions from ammonia production have increased steadily since 2016, due to the addition of new ammonia
production facilities and new production units at existing facilities in 2016, 2017, and 2018. Agriculture continues
to drive demand for nitrogen fertilizers, comprising of approximately 88 percent of domestic ammonia
consumption. In 2020 during the COVID-19 pandemic, the fertilizer industry was considered part of the critical
chemical sector by the U.S. Department of Homeland Security. The COVID-19 pandemic stay-at-home orders
issued in March 2020 did not affect the fertilizer industry, and U.S. ammonia plants maintained full operations
(USGS 2021).
Table 4-19: CO2 Emissions from Ammonia Production (MMT CO2 Eq.)
Source
1990
2005
2016
2017
2018
2019
2020
Ammonia Production
13.0
9.2
10.2
11.1
12.2
12.3
12.7
Industrial Processes and Product Use 4-31
-------
Table 4-20: CO2 Emissions from Ammonia Production (kt)
Source
1990
2005
2016
2017
2018
2019
2020
Ammonia Production
13,047
9,177
10,245
11,112
12,163
12,272
12,717
Methodology and Time-Series Consistency
For this Inventory, CO2 emissions from the production of synthetic ammonia from natural gas feedstock are
estimated using a country-specific approach modified from the 2006IPCC Guidelines (IPCC 2006) Tier 1 and 2
methods. In the country-specific approach, emissions are not based on total fuel requirement per the 2006 IPCC
Guidelines due to data disaggregation limitations of energy statistics provided by the Energy Information
Administration (EIA). Data on total fuel use (including fuel used for ammonia feedstock and fuel used for energy)
for ammonia production are not known in the United States. EIA does not provide data broken out by industrial
category, only at the broad industry sector level. To estimate emissions, a country-specific emission factor is
developed and applied to national ammonia production to estimate ammonia-production emissions from
feedstock fuel use. Emissions from fuel used for energy at ammonia plants are included in the overall EIA Industrial
sector energy use and accounted for in the Energy chapter.
The country-specific approach uses a CO2 emission factor of 1.2 metric tons CCh/metric ton NH3, which is published
by the European Fertilizer Manufacturers Association (EFMA) and is based on natural gas-based ammonia
production technologies that are similar to those employed in the United States (EFMA 2000a). The EFMA reported
an emission factor range of 1.15 to 1.30 metric tons CO2 per metric ton NH3, with 1.2 metric tons CO2 per metric
ton NH3 as a typical value (EFMA 2000a). Technologies (e.g., catalytic reforming process, etc.) associated with this
factor are found to closely resemble those employed in the United States for use of natural gas as a feedstock. The
EFMA reference also indicates that more than 99 percent of the CH4 feedstock to the catalytic reforming process is
ultimately converted to CO2. This country-specific approach is compatible with the 2006 IPCC Guidelines as it is
based on the same scientific approach that the carbon in the fuel used to produce ammonia is released as CO2. The
CO2 emission factor is applied to the percent of total annual domestic ammonia production from natural gas
feedstock.
Emissions of CChfrom ammonia production are then adjusted to account for the use of some of the CO2 produced
from ammonia production as a raw material in the production of urea. The CO2 emissions reported for ammonia
production are reduced by a factor of 0.733 multiplied by total annual domestic urea production. This corresponds
to a stoichiometric CCh/urea factor of 44/60, assuming complete conversion of ammonia (NH3) and CO2 to urea
(IPCC 2006; EFMA 2000b).
All synthetic ammonia production and subsequent urea production are assumed to be from the same process-
conventional catalytic reforming of natural gas feedstock, with the exception of ammonia production from
petroleum coke feedstock at one plant located in Kansas. Annual ammonia and urea production are shown in
Table 4-21.
The implied CO2 emission factor for total ammonia production is a combination of the emission factors for
ammonia production from natural gas and from petroleum coke. Changes in the relative production of ammonia
from natural gas and petroleum coke will impact overall emissions and emissions per ton of total ammonia
produced. For example, between 2000 and 2001 there were increases in the amount of ammonia produced from
petroleum coke which caused increases in the implied emission factor across those years.
The CO2 emission factor for petroleum coke feedstock is 3.52 metric tons of CO2 per metric ton of NFhand is
applied to the percent of total annual domestic ammonia production from petroleum coke feedstock. The CO2
emission factor is based on an average of the ratio of ammonia production from petroleum coke for years 2010
through 2015 (ACC 2020) and the facility-specific CO2 emissions from the one ammonia production plant located in
Kansas that is manufacturing ammonia from petroleum coke feedstock for years 2010 through 2015 (EPA 2021b).
Ammonia and urea are assumed to be manufactured in the same manufacturing complex, as both the raw
materials needed for urea production are produced by the ammonia production process.
4-32 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
The methodology for ammonia produced from petroleum coke shifts in 2016 when the parent company of the
facility manufacturing ammonia from petroleum coke feedstock, CVR Energy, acquired a second plant that uses
natural gas as a feedstock. The amount of ammonia production reported by CVR Energy was no longer specific to
the use of petroleum coke as a feedstock. To adjust for this, beginning in 2016, the amount of CChfrom the
ammonia production plant located in Kansas that manufactured ammonia from petroleum coke feedstock (as
reported under EPA 2021b) is now being used, along with the emission factor of 3.52 metric tons of CO2 per metric
ton of NH3 to back-calculate the amount of ammonia produced through the use of petroleum coke as feedstock.
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 CChfrom Fossil Fuel
Combustion. See the Planned Improvements section on improvements of reporting fuel and feedstock CO2
emissions utilizing EPA's GHGRP data to improve consistency with 2006IPCC Guidelines.
Total ammonia production data for 2011 through 2020 were obtained from American Chemistry Council (ACC
2021). For years before 2011, ammonia production data (see Table 4-21) were obtained from Coffeyville Resources
(Coffeyville 2005, 2006, 2007a, 2007b, 2009, 2010, 2011, and 2012) and the Census Bureau of the U.S. Department
of Commerce (U.S. Census Bureau 1991 through 1994,1998 through 2011) as reported in Current Industrial
Reports Fertilizer Materials and Related Products annual and quarterly reports. Urea-ammonia nitrate production
from petroleum coke for 1990 through 2011 was obtained from Coffeyville Resources (Coffeyville 2005, 2006,
2007a, 2007b, 2009, 2010, 2011, and 2012) and from CVR Energy, Inc. Annual Report (CVR 2012 through 2015) for
2012 through 2015. Urea production data for 1990 through 2008 were obtained from the Minerals Yearbook:
Nitrogen (USGS 1994 through 2009). Urea production data for 2009 through 2010 were obtained from the U.S.
Census Bureau (U.S. Census Bureau 2010 and 2011). The U.S. Census Bureau ceased collection of urea production
statistics in 2011. Urea production values for the years 2011 through 2020 utilize GHGRP data (EPA 2018; EPA
2021a).
Table 4-21: Ammonia Production, Recovered CO2 Consumed for Urea Production, and Urea
Production (kt)
Total C02 Consumption
Year
Ammonia Production
for Urea Production
Urea Production
1990
15,425
5,463
7,450
2005
10,143
3,865
5,270
2016
12,305
5,419
7,390
2017
14,070
6,622
9,030
2018
16,010
7,847
10,700
2019
16,410
8,360
11,400
2020
16,855
8,433
11,500
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2020. The methodology for ammonia production spliced activity data from different sources: U. S. Census
Bureau data for 1990 through 2010, and ACC data beginning in 2011. 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
The uncertainties presented in this section are primarily due to how accurately the emission factor used represents
an average across all ammonia plants using natural gas feedstock. Uncertainties are also associated with ammonia
production estimates and the assumption that all ammonia production and subsequent urea production was from
Industrial Processes and Product Use 4-33
-------
the same process—conventional catalytic reforming of natural gas feedstock, with the exception of one ammonia
production plant located in Kansas that is manufacturing ammonia from petroleum coke feedstock. Uncertainty is
also associated with the representativeness of the emission factor used for the petroleum coke-based ammonia
process. It is also assumed that ammonia and urea are produced at co-located plants from the same natural gas
raw material. The uncertainty of the total urea production activity data, based on USGS Minerals Yearbook:
Nitrogen data, is a function of the reliability of reported production data and is influenced by the completeness of
the survey responses. EPA assigned a default uncertainty range of ±5 percent for both ammonia production and
the emission factor used for the petroleum coke-based ammonia process, consistent with the ranges in Section
3.2.3.2 of the 2006IPCC Guidelines, and ±10 percent for urea production, based on expert judgment.
Recovery of CO2 from ammonia production plants for purposes other than urea production (e.g., commercial sale,
etc.) has not been considered in estimating the CO2 emissions from ammonia production, as data concerning the
disposition of recovered CO2 are not available. Such recovery may or may not affect the overall estimate of CO2
emissions depending upon the end use to which the recovered CO2 is applied. Further research is required to
determine whether byproduct CO2 is being recovered from other ammonia production plants for application to
end uses that are not accounted for elsewhere; however, for reporting purposes, CO2 consumption for urea
production is provided in this chapter.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-22. Carbon dioxide
emissions from ammonia production in 2020 were estimated to be between 11.4 and 14.1 MMT CO2 Eq. at the 95
percent confidence level. This indicates a range of approximately 10 percent below and 11 percent above the
emission estimate of 12.7 MMT CO2 Eq.
Table 4-22: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from
Ammonia Production (MMT CO2 Eq. and Percent)
Source
Gas
2020 Emission Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower Upper
Lower Upper
Bound Bound
Bound Bound
Ammonia Production
C02
12.7
11.4 14.1
-10% +11%
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 to ammonia production emission
estimates 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 ammonia facilities can be found
under Subpart G (Ammonia Production) of the regulation (40 CFR Part 98).24 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.25 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.
More details on the greenhouse gas calculation, monitoring and QA/QC methods applicable to reporting of urea
produced at ammonia production facilities can be found under Section 4.5 Urea Consumption for Non-Agricultural
Purposes.
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.
4-34 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Planned Improvements
Future improvements involve continuing to evaluate and analyze data reported under EPA's GHGRP to improve the
emission estimates for the Ammonia Production source category, in particular new facility-level reporting data
from updated reporting requirements finalized in October of 2014 (79 FR 63750) and December 2016 (81 FR
89188)26 that include facility-level ammonia production data and feedstock consumption. The data were first
reported by facilities in 2018 and available post-verification in 2019 to assess for use in future Inventories, if the
data meet GHGRP CBI aggregation criteria. The data are still being evaluated and will be incorporated in future
Inventory reports, if possible. Particular attention will be made to ensure time-series consistency of the emission
estimates presented in future Inventory reports, along with application of appropriate category-specific QC
procedures consistent with IPCC and UNFCCC guidelines. For example, data reported in 2018 will reflect activity in
2017 and may not be representative of activity in prior years of the time series. This assessment is required as the
new GHGRP data associated with new requirements are only applicable starting with reporting for calendar year
2017, and thus are not available for all inventory years (i.e., 1990 through 2016) 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.27 Specifically, the planned improvements
include assessing the anticipated new data to update the emission factors to include both fuel and feedstock CO2
emissions to improve consistency with 2006 IPCC Guidelines, in addition to reflecting CO2 capture and storage
practices (beyond use of CO2 for urea production). Methodologies will also be updated if additional ammonia
production plants are found to use hydrocarbons other than natural gas for ammonia production. Due to limited
resources and ongoing data collection efforts, this planned improvement is still in development and is not
incorporated into this Inventory. This is a long-term planned improvement.
4.6 Urea Consumption for Non-
Agricultural Purposes
Urea is produced using ammonia (NH3) and carbon dioxide (CO2) as raw materials. All urea produced in the United
States is assumed to be produced at ammonia production facilities where both ammonia and CO2 are generated.
There were 35 plants producing ammonia in the United States in 2020, with two additional plants sitting idle for
the entire year (USGS 2021).
The chemical reaction that produces urea is:
2nh3+ C02 -> NH2COONH4 -> CO(NH2)2 +h2o
This section accounts for CO2 emissions associated with urea consumed exclusively for non-agricultural purposes.
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.
The industrial applications of urea include its use in adhesives, binders, sealants, resins, fillers, analytical reagents,
catalysts, intermediates, solvents, dyestuffs, fragrances, deodorizers, flavoring agents, humectants and
dehydrating agents, formulation components, monomers, paint and coating additives, photosensitive agents, and
surface treatments agents. In addition, urea is used for abating nitrogen oxide (NOx) emissions from coal-fired
power plants and diesel transportation motors.
Emissions of CO2 from urea consumed for non-agricultural purposes in 2020 were estimated to be 6.0 MMT CO2
Eq. (5,983 kt) and are summarized in Table 4-23 and Table 4-24. Net CO2 emissions from urea consumption for
26 See https://www.epa.gov/ehereportine/historical-rulemakines.
27 See http://www.ipcc-nggip.iees.or.lp/public/tb/TFI Technical Bulletin l.pdf.
Industrial Processes and Product Use 4-35
-------
non-agricultural purposes have increased by approximately 58 percent from 1990 to 2020 and decreased by
approximately 1.0 percent from 2019 to 2020.
Table 4-23: CO2 Emissions from Urea Consumption for Non-Agricultural Purposes (MMT CO2
Eq.)
Source 1990 2005 2016 2017 2018 2019 2020
Urea Consumption 3.8 3.7 5.3 5.2 6.0 6.0 6.0
Table 4-24: CO2 Emissions from Urea Consumption for Non-Agricultural Purposes (kt)
Source 1990 2005 2016 2017 2018 2019 2020
Urea Consumption 3,784 3,653 , 5,330 5,182 6,030 6,044 5,983
Methodology and Time-Series Consistency
Emissions of CO2 resulting from urea consumption for non-agricultural purposes are estimated by multiplying the
amount of urea consumed in the United States for non-agricultural purposes by a factor representing the amount
of CO2 used as a raw material to produce the urea. This method is based on the assumption that all of the carbon
in urea is released into the environment as CO2 during use, consistent with the Tier 1 method used to estimate
emissions from ammonia production in the 2006IPCC Guidelines (IPCC 2006) which states that the "CO2 recovered
[from ammonia production] for downstream use can be estimated from the quantity of urea produced where CO2
is estimated by multiplying urea production by 44/60, the stoichiometric ratio of CO2 to urea."
The amount of urea consumed for non-agricultural purposes in the United States is estimated by deducting the
quantity of urea fertilizer applied to agricultural lands, which is obtained directly from the Agriculture chapter (see
Table 5-25), from the total domestic supply of urea as reported in Table 4-25. The domestic supply of urea is
estimated based on the amount of urea produced plus urea imports and minus urea exports. A factor of 0.733 tons
of CO2 per ton of urea consumed is then applied to the resulting supply of urea for non-agricultural purposes to
estimate CO2 emissions from the amount of urea consumed for non-agricultural purposes. The 0.733 tons of CO2
per ton of urea emission factor is based on the stoichiometry of carbon in urea. This corresponds to a
stoichiometric CO2 to urea factor of 44/60, assuming complete conversion of carbon in urea to CChflPCC 2006;
EFMA2000).
Urea production data for 1990 through 2008 were obtained from the U.S. Geological Survey (USGS) Minerals
Yearbook: Nitrogen (USGS 1994 through 2009a). Urea production data for 2009 through 2010 were obtained from
the U.S. Census Bureau (2011). The U.S. Census Bureau ceased collection of urea production statistics in 2011.
Starting with the 1990 through 2017 Inventory report, EPA began utilizing urea production data from EPA's GHGRP
to estimate emissions. Urea production values in the current Inventory report utilize GHGRP data for the years
2011 through 2020 (EPA 2018; EPA 2021a; EPA 2021b).
Urea import data for 2020 were not available at the time of publication and were estimated using 2019 values.
Urea import data for 2013 to 2019 were obtained from the USGS Minerals Yearbook: Nitrogen (USGS 2021a). Urea
import data for 2011 and 2012 were taken from U.S. Fertilizer Import/Exports from the United States Department
of Agriculture (USDA) Economic Research Service Data Sets (U.S. Department of Agriculture 2012). USDA
suspended updates to this data after 2012. Urea import data for the previous years were obtained from the U.S.
Census Bureau Current Industrial Reports Fertilizer Materials and Related Products annual and quarterly reports for
1997 through 2010 (U.S. Census Bureau 2001 through 2011), The Fertilizer Institute (TFI 2002) for 1993 through
1996, and the United States International Trade Commission Interactive Tariff and Trade DataWeb (U.S. ITC 2002)
for 1990 through 1992 (see Table 4-25).
Urea export data for 2020 were not available at the time of publication and were estimated using 2019 values.
Urea export data for 2013 to 2019 were obtained from the USGS Minerals Yearbook: Nitrogen (USGS 2021a). Urea
4-36 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
export data for 1990 through 2012 were taken from U.S. Fertilizer Import/Exports from USDA Economic Research
Service Data Sets (U.S. Department of Agriculture 2012). USDA suspended updates to this data after 2012.
Table 4-25: Urea Production, Urea Applied as Fertilizer, Urea Imports, and Urea Exports (kt)
Year
Urea
Production
Urea Applied
as Fertilizer
Urea
Imports
Urea
Exports
Urea Consumed for Non-
Agricultural Purposes
1990
7,450
3,296
1,860
854
5,160
2005
5,270
4,779
5,026
536
4,981
2016
7,390
6,381
6,580
321
7,268
2017
9,030
6,678
5,510
795
7,067
2018
10,700
6,844
5,110
743
8,223
2019
11,400
7,009
4,410
559
8,242
2020
11,500
7,193
4,410
559
8,158
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2020. The methodology for urea consumption for non-agricultural purposes spliced activity data from
different sources: USGS data for 1990 through 2008, U. S. Census Bureau data for 2009 and 2010, and GHGRP data
beginning in 2011. Consistent with the 2006IPCC Guidelines, the overlap technique was applied to compare the
data sets for years where there was overlap, with findings that the data sets were consistent and adjustments
were not needed.
Uncertainty
There is limited publicly available data on the quantities of urea produced and consumed for non-agricultural
purposes. Therefore, the amount of urea used for non-agricultural purposes is estimated based on a balance that
relies on estimates of urea production, urea imports, urea exports, and the amount of urea used as fertilizer. The
primary uncertainties associated with this source category are associated with the accuracy of these estimates as
well as the fact that each estimate is obtained from a different data source. Because urea production estimates are
no longer available from the USGS, there is additional uncertainty associated with urea produced beginning in
2011. There is also uncertainty associated with the assumption that all of the carbon in urea is released into the
environment as CO2 during use.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-26. Carbon dioxide
emissions associated with urea consumption for non-agricultural purposes during 2020 were estimated to be
between 5.1 and 6.8 MMT CO2 Eq. at the 95 percent confidence level. This indicates a range of approximately 14
percent below and 14 percent above the emission estimate of 6.0 MMT CO2 Eq.
Table 4-26: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Urea
Consumption for Non-Agricultural Purposes (MMT CO2 Eq. and Percent)
Source Gas
2020 Emission Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Urea Consumption
for Non-Agricultural C02
6.0
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 4-37
-------
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).
More details on the greenhouse gas calculation, monitoring and QA/QC methods applicable to reporting of urea
production occurring at ammonia facilities can be found under Subpart G (Ammonia Manufacturing) of the
regulation (40 CFR Part 98).28 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.29 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
urea production data against external datasets including the USGS Minerals Yearbook data. The comparison shows
consistent trends in urea production overtime.
Recalculations Discussion
Based on updated quantities of urea applied for agricultural uses for 2014-2019, updated urea imports from USGS
for 2018 and 2019, and updated urea exports from USGS for 2018 and 2019, recalculations were performed for
2014 through 2019. Compared to the previous Inventory, CO2 emissions from urea consumption for non-
agricultural purposes increased by less than 1 percent (2 kt CO2) for 2014,1.6 percent (73 kt CO2) for 2015, 3.9
percent (198 kt CO2) for 2016, 3.1 percent (154 kt CO2) for 2017, and 3.0 percent (173 kt CO2) for 2018 and
decreased by 2.9 percent (178 kt CO2) for 2019.
Recalculations Discussion
No recalculations were performed for the 1990 through 2019 portion of the time series.
4.7 Nitric Acid Production (CRF Source
Category 2B2)
Nitrous oxide (N2O) is emitted during the production of nitric acid (HNO3), an inorganic compound used primarily
to make synthetic commercial fertilizers. Nitric acid is also a major component in the production of adipic acid—a
feedstock for nylon—and explosives. Virtually all of the nitric acid produced in the United States is manufactured
by the high-temperature catalytic oxidation of ammonia (EPA 1998). There are two different nitric acid production
methods: weak nitric acid and high-strength nitric acid. The first method utilizes oxidation, condensation, and
absorption to produce nitric acid at concentrations between 30 and 70 percent nitric acid. High-strength acid (90
percent or greater nitric acid) can be produced from dehydrating, bleaching, condensing, and absorption of the
weak nitric acid. Most U.S. plants were built between 1960 and 2000. As of 2020, there were 32 active nitric acid
production plants, including one high-strength nitric acid production plant in the United States (EPA 2010; EPA
2021).
28 See http://www.ecfr.gov/cgi-bin/text-idx7tpN/ecfrbrowse/Title40/40cfr98 main Q2.tpl.
29 See https://www.epa.eov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
4-38 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
The basic process technology for producing nitric acid has not changed significantly over time. During this process,
N2O is formed as a byproduct and is released from reactor vents into the atmosphere. Emissions from fuels
consumed for energy purposes during the production of nitric acid are included in the Energy chapter.
Nitric acid is made from the reaction of ammonia (NH3) with oxygen (O2) in two stages. The overall reaction is:
4NH3 + 802 -> 4HNO:i + 4H2
Currently, the nitric acid industry in the United States controls emissions of NO and NO2 (i.e., NOx), using a
combination of non-selective catalytic reduction (NSCR) and selective catalytic reduction (SCR) technologies. In the
process of destroying NOx, NSCR systems are also very effective at destroying N2O. Five nitric acid plants had NSCR
systems installed between 1964 and 1977, over half due to the finalization of the Nitric Acid Plant New Source
Performance Standards (NSPS) which went into effect in 1971. Four additional nitric acid plants had NSCR systems
installed between 2016 and 2018, as a result of EPA Consent Decrees to control NOx emissions more effectively.
NSCR systems are used in approximately one-third of the weak acid production plants. For N2O abatement, U.S.
facilities are using both tertiary (i.e., NSCR and SCR) and secondary controls (i.e., catalysts added to the ammonia
reactor to lessen potential N2O production).
Emissions from the production of nitric acid are generally directly proportional to the annual amount of nitric acid
produced because emissions are calculated as the product of the total annual production and plant-specific
emission factors. There are a few instances, however, where that relationship has not been directly proportional.
For example, in 2015 and 2019, nitric acid production decreased and emissions increased, compared to the
respective preceding years. N2O emissions for those years are calculated based on data from the GHGRP as
discussed in the Methodology section below. According to data from plants reporting to GHGRP, plant-specific
operations can affect the emission factor used, including: (1) site-specific fluctuations in ambient temperature and
humidity, (2) catalyst age and condition, (3) process changes, (4) the addition or removal of abatement
technologies, and (5) the number of nitric acid trains. Changes in those operating conditions for the years in
question (2015 and 2018) caused changes in emission factors and, therefore, the emissions to change
disproportionally to production in those years.
Nitrous oxide emissions from this source were estimated to be 9.3 MMT CO2 Eq. (31 kt of N2O) in 2020 (see Table
4-27). Emissions from nitric acid production have decreased by 23 percent since 1990, while production has
increased by 11 percent over the same time period (see Table 4-27). Emissions have decreased by 36 percent since
1997, the highest year of production in the time series. The primary use of nitric acid is to produce synthetic
fertilizers, and in 2020, the fertilizer industry was considered part of the critical chemical sector by the U.S.
Department of Homeland Security, which minimized the impact of the COVID-19 pandemic on nitric acid
production and emissions (USGS 2021).
Table 4-27: N2O Emissions from Nitric Acid Production (MMT CO2 Eq. and kt N2O)
Year
MMT CO? Eq.
kt N20
1990
12.1
41
2005
11.3
38
2016
10.1
34
2017
9.3
31
2018
9.6
32
2019
10.0
34
2020
9.3
31
Methodology and Time-Series Consistency
Emissions of N2O were calculated using the estimation methods provided by the 2006IPCC Guidelines and a
country-specific method utilizing EPA's GHGRP. The 2006 IPCC Guidelines Tier 2 method was used to estimate
Industrial Processes and Product Use 4-39
-------
emissions from nitric acid production for 1990 through 2009, and a country-specific approach similar to the IPCC
Tier 3 method was used to estimate N2O emissions for 2010 through 2020.
For this Inventory, EPA reviewed GHGRP facility-level information on the installation date of all N2O abatement
equipment (EPA 2021). Revisions to GHGRP reporting requirements were finalized in December 2016, and this
information was first reported by facilities in 2018 and available post-verification in 2019. EPA verified that all
reported N2O abatement equipment had already been incorporated into the estimation of N2O emissions from
nitric acid production over the full time series.
2010 through 2020
Process N2O emissions and nitric acid production data were obtained directly from EPA's GHGRP for 2010 through
2020 by aggregating reported facility-level data (EPA 2021).
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.30 All nitric acid facilities are required to
calculate process 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 by
directly measuring N2O emissions using monitoring equipment.31
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 for consistency and are shown in Table 4-28.
1990 through 2009
Using GHGRP data for 2010,32 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. At that time, EPA had initiated compilation of a nitric acid database to improve estimation of emissions from
30 See 40 CFR 98.2(i)(l) and 40 CFR 98.2(i)(2) for more information about these provisions.
31 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.
32 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).
4-40 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
this industry and obtained updated information on application of controls via review of permits and outreach with
facilities and trade associations. The research indicated recent installation of abatement technologies at additional
facilities.
Based on the available data, it was assumed that emission factors for 2010 would be more representative of
operating conditions in 1990 through 2009 than more recent years. Initial review of historical data indicates that
percent production with and without abatement can change over time and from year to year due to changes in
application of facility-level abatement technologies, maintenance of abatement technologies, and plant closures
and start-ups (EPA 2012, 2013; Desai 2012; CAR 2013). In this Inventory, EPA verified the installation dates of N2O
abatement technologies for all facilities based on GHGRP facility-level information (EPA 2021), as noted above. 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:
Equation 4-4: 2006IPCCGuide/inesTier 3: N2O Emissions From Nitric Acid Production
(Equation 3.6)
where,
Ei
Pi
EFweighted,i =
%Pc,i
EFc
%Punc,i =
EFunc =
i =
Ei — Pi X EFWelgfr):eCl:l
EFweighted,i = X EFc) + (%PUnc,i X EFunc)\
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). Publicly available information on plant-
level abatement technologies was used to estimate the shares of nitric acid production with and without
abatement for 2008 and 2009 (EPA 2012, 2013; Desai 2012; CAR 2013). In previous Inventory reports, EPA
conducted a review of operating permits to obtain more information on the use or installation of abatement
technologies for 1990 through 2007; therefore, the share of national production with and without abatement for
2008 was assumed to be constant for 1990 through 2007. As noted above, EPA used GHGRP facility-level
information to verify that all reported N2O abatement equipment had already been incorporated into the
estimation of N2O emissions from nitric acid production over the full time series (EPA 2021).
Industrial Processes and Product Use 4-41
-------
Table 4-28: Nitric Acid Production (kt)
Year kt_
1990 7,200
2005 6,710
2016 7,810
2017 7,780
2018 8,210
2019 8,080
2020 7,970
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2020. 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 2006IPCC 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
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 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, 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
submitted to EPA are accurate, complete, and consistent. The annual production reported by each nitric acid
facility under EPA's GHGRP and then aggregated to estimate national N2O emissions is assumed to have low
uncertainty. EPA assigned an uncertainty range of ±5 percent for facility-reported N2O emissions, consistent with
section 3.4.3.1 of the 2006 IPCC Guidelines, and ±2 percent for nitric acid production, consistent with section
3.3.3.2 of the 2006 IPCC Guidelines.
The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 4-29. Nitrous oxide
emissions from nitric acid production were estimated to be between 8.8 and 9.8 MMT CO2 Eq. at the 95 percent
confidence level. This indicates a range of approximately 5 percent below to 5 percent above the 2020 emissions
estimate of 9.3 MMT CO2 Eq.
Table 4-29: Approach 2 Quantitative Uncertainty Estimates for N2O Emissions from Nitric
Acid Production (MMT CO2 Eq. and Percent)
2020 Emission Estimate Uncertainty Range Relative to Emission Estimate3
(MMTCO; Eq.) (MMTCO; Eq.) (%)
4-42 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Nitric Acid Production
N20
9.3
8.8
9.8
-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 nitric acid facilities can be found under Subpart V: Nitric Acid
Production of the GHGRP regulation (40 CFR Part 98).33
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 that is tailored to
the Subpart (e.g., combination of electronic checks including range checks, statistical checks, algorithm checks,
year-to-year comparison checks, along with manual reviews) 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 (EPA 2015).34 EPA's review of observed trends
noted that while emissions have generally mirrored production, in 2015 and 2019 nitric acid production decreased
compared to the previous year and emissions increased. While review is ongoing, based on feedback from the
verification process to date, these changes are due to facility-specific changes (e.g., in the nitric production process
and management of abatement equipment).
Recalculations Discussion
No recalculations were performed for the 1990 through 2019 portion of the time series.
Planned Improvements
Pending resources, EPA is considering a near-term improvement to estimates and associated characterization of
uncertainty. In the short-term, with 10 years of EPA's GHGRP data, EPA anticipates completing updates of
category-specific QC procedures. EPA also anticipates making improvements to both qualitative and quantitative
uncertainty estimates.
4.8 Adipic Acid Production (CRF Source
Category 2B3)
Adipic acid is produced through a two-stage process during which nitrous oxide (N2O) is generated in the second
stage. Emissions from fuels consumed for energy purposes during the production of adipic acid are accounted for
in the Energy chapter. The first stage of manufacturing usually involves the oxidation of cyclohexane to form a
cyclohexanone/cyclohexanol mixture. The second stage involves oxidizing this mixture with nitric acid to produce
33 See Subpart V monitoring and reporting regulation http://www.ecfr.gov/cgi-bin/text-
idx?tpl=/ecfrbrowse/Title40/40cfr98 main Q2.tpl.
34 See GHGRP Verification Factsheet https://www.epa.gov/sites/production/files/2015-
07/documents/ghgrp verification factsheet.pdf.
Industrial Processes and Product Use 4-43
-------
adipic acid. Nitrous oxide is generated as a byproduct of the nitric acid oxidation stage and is emitted in the waste
gas stream (Thiemens and Trogler 1991). The second stage is represented by the following chemical reaction:
(iCH2)5CO(cyclohexanone) + (CH2)5CHOH (cyclohexanol) + wHN03
-» HOOC(CH2)4COOH(adipic acid) + xN20 + yH20
Process emissions from the production of adipic acid vary with the types of technologies and level of emission
controls employed by a facility. In 1990, two major adipic acid-producing plants had N2O abatement technologies
in place and, as of 1998, three major adipic acid production facilities had control systems in place (Reimer et al.
1999). In 2020, catalytic reduction, non-selective catalytic reduction (NSCR), and thermal reduction abatement
technologies were applied as N2O abatement measures at adipic acid facilities (EPA 2021).
Worldwide, only a few adipic acid plants exist. The United States, Europe, and China are the major producers, with
the United States accounting for the largest share of global adipic acid production capacity in recent years. In 2020,
the United States had two companies with a total of two adipic acid production facilities (one in Texas and one in
Florida), following the ceased operations of a third major production facility at the end of 2015 (EPA 2021).
Adipic acid is a white crystalline solid used in the manufacture of synthetic fibers, plastics, coatings, urethane
foams, elastomers, and synthetic lubricants. Commercially, it is the most important of the aliphatic dicarboxylic
acids, which are used to manufacture polyesters. Eighty-four percent of all adipic acid produced in the United
States is used in the production of nylon 6,6; 9 percent is used in the production of polyester polyols; 4 percent is
used in the production of plasticizers; and the remaining 4 percent is accounted for by other uses, including
unsaturated polyester resins and food applications (ICIS 2007). Food grade adipic acid is used to provide some
foods with a "tangy" flavor (Thiemens and Trogler 1991).
National adipic acid production has decreased by approximately 7 percent over the period of 1990 through 2020,
to approximately 700,000 metric tons (ACC 2021). Nitrous oxide emissions from adipic acid production were
estimated to be 8.3 MMT CO2 Eq. (28 kt N2O) in 2020 (see Table 4-30). Over the period 1990 through 2020,
facilities have reduced emissions by 45 percent due to the widespread installation of pollution control measures in
the late 1990s. The COVID-19 pandemic may have partially influenced the decrease in adipic acid production
between 2019 and 2020.
Significant changes in the amount of time that the N2O abatement device at one facility was in operation has been
the main cause of fluctuating emissions in recent years. These fluctuations are most evident for years where trends
in emissions and adipic acid production were not directly proportional: (1) between 2016 and 2017, (2) between
2017 and 2018, and (3) between 2019 and 2020. As noted above, changes in control measures and abatement
technologies at adipic acid production facilities, including maintenance of equipment, can result in annual emission
fluctuations. Little additional information is available on drivers of trends, and the amount of adipic acid produced
is not reported under EPA's GHGRP.
Table 4-30: N2O Emissions from Adipic Acid Production (MMT CO2 Eq. and kt N2O)
Year MMT CP2 Eq. kt N2Q
1990 15.2 51
2005 7.1 24
2016 7.1 24
2017 7.5 25
2018 10.5 35
2019 5.3 18
2020 8.3 28
4-44 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Methodology and Time-Series Consistency
Emissions are estimated using both Tier 2 and Tier 3 methods consistent with the 2006IPCC Guidelines. Due to
confidential business information (CBI), plant names are not provided in this section; therefore, the four adipic
acid-producing facilities that have operated over the time series will be referred to as Plants 1 through 4. Overall,
as noted above, the two currently operating facilities use catalytic reduction, NSCR, and thermal reduction
abatement technologies.
2010 through 2020
All emission estimates for 2010 through 2020 were obtained through analysis of GHGRP data (EPA 2010 through
2021), which is consistent with the 2006 IPCC Guidelines Tier 3 method. Facility-level greenhouse gas emissions
data were obtained from EPA's GHGRP for the years 2010 through 2020 (EPA 2010 through 2021) and aggregated
to national N2O emissions. Consistent with IPCC Tier 3 methods, all adipic acid production facilities are required to
either calculate N2O emissions using a facility-specific emission factor developed through annual performance
testing under typical operating conditions or directly measure N2O emissions using monitoring equipment.35
1990 through 2009
For years 1990 through 2009, which were prior to EPA's GHGRP reporting, for both Plants 1 and 2, emission
estimates were obtained directly from the plant engineers and account for reductions due to control systems in
place at these plants during the time series. These prior estimates are considered CBI and hence are not published
(Desai 2010, 2011). These estimates were based on continuous process monitoring equipment installed at the two
facilities.
For Plant 4,1990 through 2009 N2O emissions were estimated using the following Tier 2 equation from the 2006
IPCC Guidelines:
Equation 4-5: 2006IPCCGuide/inesTier 2: N2O Emissions From Adipic Acid Production
(Equation 3.8)
Eaa = Qaa x EFaa X (1 - [DF X UF])
where,
Eaa
= N2O emissions from adipic acid production, metric tons
Qaa
= Quantity of adipic acid produced, metric tons
EFaa
= Emission factor, metric ton INhO/metric ton adipic acid produced
DF
= N2O destruction factor
UF
= Abatement system utility factor
The adipic acid production is multiplied by an emission factor (i.e., N2O emitted per unit of adipic acid produced),
which has been estimated to be approximately 0.3 metric tons of N2O per metric ton of product (IPCC 2006). The
"N2O destruction factor" in the equation represents the percentage of N2O emissions that are destroyed by the
installed abatement technology. The "abatement system utility factor" represents the percentage of time that the
abatement equipment operates during the annual production period. Plant-specific production data for Plant 4
were obtained across the time series through personal communications (Desai 2010, 2011). The plant-specific
production data were then used for calculating emissions as described above.
For Plant 3, 2005 through 2009 emissions were obtained directly from the plant (Desai 2010, 2011). For 1990
through 2004, emissions were estimated using plant-specific production data and the IPCC factors as described
35 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.
Industrial Processes and Product Use 4-45
-------
above for Plant 4. Plant-level adipic acid production for 1990 through 2003 was estimated by allocating national
adipic acid production data to the plant level using the ratio of known plant capacity to total national capacity for
all U.S. plants (ACC 2020; CMR 2001,1998; CW 1999; C&EN 1992 through 1995). For 2004, actual plant production
data were obtained and used for emission calculations (CW 2005).
Plant capacities for 1990 through 1994 were obtained from Chemical & Engineering News, "Facts and Figures" and
"Production of Top 50 Chemicals" (C&EN 1992 through 1995). Plant capacities for 1995 and 1996 were kept the
same as 1994 data. The 1997 plant capacities were taken from Chemical Market Reporter, "Chemical Profile: Adipic
Acid" (CMR 1998). The 1998 plant capacities for all four plants and 1999 plant capacities for three of the plants
were obtained from Chemical Week, Product Focus: Adipic Acid/Adiponitrile (CW 1999). Plant capacities for the
year 2000 for three of the plants were updated using Chemical Market Reporter, "Chemical Profile: Adipic Acid"
(CMR 2001). For 2001 through 2003, the plant capacities for three plants were held constant at year 2000
capacities. Plant capacity for 1999 to 2003 for the one remaining plant was kept the same as 1998.
National adipic acid production data (see Table 4-31) from 1990 through 2020 were obtained from the American
Chemistry Council (ACC 2021).
Table 4-31: Adipic Acid Production (kt)
Year kt
1990 755
2005 865
2016 860
2017 830
2018 825
2019 810
2020 700
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2020. The methodology for adipic acid production spliced activity data from multiple sources: plant-
specific emissions data and publicly available plant capacity data for 1990 through 2009 and GHGRP emission data
starting in 2010. Consistent with the 2006IPCC 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
Uncertainty associated with N2O emission estimates includes the methods used by companies to monitor and
estimate emissions. While some information has been obtained through outreach with facilities, limited
information is available over the time series on these methods, abatement technology destruction and removal
efficiency rates, and plant-specific production levels. EPA assigned an uncertainty range of ±5 percent for facility-
reported N2O emissions, consistent with section 3.4.3.1 of the 2006 IPCC Guidelines.
The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 4-32. Nitrous oxide
emissions from adipic acid production for 2020 were estimated to be between 7.9 and 8.7 MMT CO2 Eq. at the 95
percent confidence level. These values indicate a range of approximately 5 percent below to 5 percent above the
2020 emission estimate of 8.3 MMT CO2 Eq.
Table 4-32: Approach 2 Quantitative Uncertainty Estimates for N2O Emissions from Adipic
Acid Production (MMT CO2 Eq. and Percent)
2020 Emission Estimate Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (MMT CP2 Eq.) (%)
4-46 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Adipic Acid Production
N20
8.3
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).36 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).37 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
Recalculations of adipic acid emissions were performed for the 2016 through 2019 portion of the time series due
to GHGRP resubmittals for those years. For years 2016 through 2018, the emissions increased by 0.4 MMT CO2 Eq.
(1.6 percent), 0.3 MMT CO2 Eq. (1.2 percent), and 0.6 MMT CO2 Eq. (1.8 percent), respectively. For year 2019, the
emissions decreased by 0.1 MMT CO2 Eq. (0.3 percent).
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 utilization of N2O abatement technology. The facility using direct measurement of N2O emissions has reported
the use of N2O abatement technology but is not required to report the date of installation.
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
36 See http://www.ecfr.gov/cgi-bin/text-idx7tpN/ecfrbrowse/Title40/40cfr98 main Q2.tpl.
37 See https://www.epa.eov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
Industrial Processes and Product Use 4-47
-------
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 N0/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 (N0H(S03NH4)2)
I
(N0H(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^NO. H2S04 (+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
(Cline 2019). As of 2020, two companies in the United States produced caprolactam at two facilities: AdvanSix
(formerly Honeywell) in Virginia (AdvanSix 2021) and BASF in Texas (BASF 2021).
Nitrous oxide emissions from caprolactam production in the United States were estimated to be 1.2 MMT CO2 Eq.
(4 kt N2O) in 2020 (see Table 4-33). National emissions from caprolactam production decreased by approximately
28 percent over the period of 1990 through 2020. Emissions in 2020 decreased by approximately 13 percent from
the 2019 levels. While this decrease could be related to the COVID-19 pandemic, caprolactam production has been
declining since 2013, with the largest decrease of 15 percent happening between 2016 and 2017.
4-48 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 4-33: N2O Emissions from Caprolactam Production (MMT CO2 Eq. and kt N2O)
Year
MMT C02 Eq.
kt N20
1990
1.7
6
2005
2.1
7
2016
2017
2018
2019
2020
1.7
1.5
1.4
1.4
1.2
6
5
5
5
4
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.
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 2020, as shown in this formula:
Equation 4-6: 2006IPCCGuide/inesTier 1: N2O Emissions From Caprolactam Production
(Equation 3.9)
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
generate one metric ton of N2O, resulting in an emission factor of 9.0 kg N2O per metric ton of caprolactam (IPCC
Methodology and Time-Series Consistency
En2o = EF x CP
where,
Annual N2O Emissions (kg)
N2O emission factor (default) (kg INhO/metric ton caprolactam produced)
Caprolactam production (metric tons)
Industrial Processes and Product Use 4-49
-------
2006). When applying the Tier 1 method, the 2006IPCC Guidelines state that it is good practice to assume that
there is no abatement of N2O emissions and to use the highest default emission factor available in the guidelines.
In addition, EPA did not find support for the use of secondary catalysts to reduce N2O emissions, such as those
employed at nitric acid plants.
The activity data for caprolactam production (see Table 4-34) from 1990 to 2020 were obtained from the American
Chemistry Council's Guide to the Business of Chemistry (ACC 2021). EPA will continue to analyze and assess
alternative sources of production data as a quality control measure.
Table 4-34: Caprolactam Production (kt)
Year kt
1990 626
2005 795
2016 640
2017 545
2018 530
2019 515
2020 450
Carbon dioxide and methane (CH4) emissions may also occur from the production of caprolactam, but currently the
IPCC does not have methodologies for calculating these emissions associated with caprolactam production.
Methodological approaches, consistent with IPCC guidelines, have been applied to the entire time series to ensure
consistency in emissions from 1990 through 2020.
Uncertainty
Estimation of emissions of N2O from caprolactam production can be treated as analogous to estimation of
emissions of N2O from nitric acid production. Both production processes involve an initial step of NH3 oxidation,
which is the source of N2O formation and emissions (IPCC 2006). Therefore, uncertainties for the default emission
factor values in the 2006 IPCC Guidelines are an estimate based on default values for nitric acid plants. In general,
default emission factors for gaseous substances have higher uncertainties because mass values for gaseous
substances are influenced by temperature and pressure variations and gases are more easily lost through process
leaks. The default values for caprolactam production have a relatively high level of uncertainty due to the limited
information available (IPCC 2006).
The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 4-35. Nitrous oxide
emissions from Caprolactam, Glyoxal and Glyoxylic Acid Production for 2020 were estimated to be between 0.8
and 1.6 MMT CO2 Eq. at the 95 percent confidence level. These values indicate a range of approximately 31
percent below to 32 percent above the 2020 emission estimate of 1.2 MMT CO2 Eq.
4-50 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 4-35: Approach 2 Quantitative Uncertainty Estimates for N2O Emissions from
Caprolactam, Glyoxal and Glyoxylic Acid Production (MMT CO2 Eq. and Percent)
Source
Gas
2020 Emission Estimate
(MMT C02 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower Upper
Bound Bound
Lower Upper
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.
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).
Recalculations Discussion
No recalculations were performed for the 1990 through 2019 portion of the time series.
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. EPA plans to share latest findings from
ongoing research for feedback during the next Inventory expert review cycle. During the Expert Review period for
the current Inventory report, EPA continued 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 (Si02> 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)
Industrial Processes and Product Use 4-51
-------
Carbon dioxide and Cm 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, Cm 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
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 2020, 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 (USGS 1991b through 2020).
Silicon carbide was manufactured by two facilities in the United States, one of which produced primarily non-
abrasive SiC (USGS 2020). USGS production values for the United States consists of SiC used for abrasives and for
metallurgical and other non-abrasive applications (USGS 2020). 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 2021). Consumption of SiC, however, decreased by
approximately 25 percent due to a sharp decline in imports (U.S. Census Bureau 2005 through 2021).
Carbon dioxide emissions from SiC production and consumption in 2020 were 0.2 MMT CO2 Eq. (154 kt CO2), which
are about 40 percent lower than emissions in 1990 (243 kt) (see Table 4-36 and Table 4-37). Approximately 59
percent of these emissions resulted from SiC production, while the remainder resulted from SiC consumption.
Methane emissions from SiC production in 2020 were 0.01 MMT CO2 Eq. (0.4 kt CH4) (see Table 4-36 and Table
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
2016
2017
2018
2019
2020
Production
C02
0.2
0.1
0.1
0.1
0.1
0.1
0.1
ch4
+
+
+
+
+
+
+
Consumption
C02
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Total
0.2
0.2
0.2
0.2
0.2
0.2
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
2016
2017
2018
2019
2020
Production
C02
170
92
92
92
92
92
92
ch4
1
+
+
+
+
+
+
Consumption
C02
73
121
78
90
93
84
62
+ Does not exceed 0.5 kt
4-52 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Methodology and Time-Series Consistency
Emissions of CO2 and CH4 from the production of SiC were calculated using the Tier 1 method provided by the 2006
IPCC Guidelines. Annual estimates of SiC production were multiplied by the default emission factors, as shown
below:
Equation 4-7: 2006IPCCGuide/inesTier 1: Emissions from Carbide Production (Equation
3.11)
Esc,C02 = EFsc,C02 X Qsc
(1 metric ton\
Esc,CH4 ~ EFsc,CH4 * Qsc *
/I metric ton\
V 1000 kg )
where,
Esc,co2 = CO2 emissions from production of SiC, metric tons
EFsc,co2 = Emission factor for production of SiC, metric ton CCh/metric ton SiC
Qsc = Quantity of SiC produced, metric tons
Esc,ch4 = Cm emissions from production of SiC, metric tons
EFsc,ch4 = Emission factor for production of SiC, kilogram CH4/metric ton SiC
Emission factors were taken from the 2006 IPCC Guidelines:
• 2.62 metric tons C02/metric ton SiC
• 11.6 kg Cl-U/metric ton SiC
Production data for metallurgical and other non-abrasive applications of SiC are not available; therefore, both CO2
and CH4 estimates for SiC are based solely upon production data for SiC for industrial abrasive applications.
Silicon carbide industrial abrasives production data for 1990 through 2017 were obtained from the U.S. Geological
Survey (USGS) Minerals Yearbook: Manufactured Abrasives (USGS 1991a through 2017). Production data for 2018
through 2020 were obtained from the Mineral Commodity Summaries: Abrasives (Manufactured) (USGS 2021).
Silicon carbide production data published by USGS have been rounded to the nearest 5,000 metric tons to avoid
disclosing company proprietary data. For the period 1990 through 2001, reported USGS production data include
production from a facility located in Canada that ceased operations in 2001. Using SiC data from Canada (UNFCCC
GHG Data Interface 2021), U.S. SiC production for 1990 through 2001 was revised to reflect only U.S. production.
SiC consumption for the entire time series is estimated using USGS consumption data (USGS 1991b through 2020)
and data from the U.S. International Trade Commission (USITC) database on net imports and exports of SiC(U.S.
Census Bureau 2005 through 2021) (see Table 4-38). Total annual SiC consumption (utilization) was estimated by
subtracting annual exports of SiC from the annual total of national SiC production and net imports.
Emissions of CO2 from SiC consumption for metallurgical uses were calculated by multiplying the annual utilization
of SiC for metallurgical uses (reported annually in the USGS Minerals Yearbook: Silicon) by the carbon content of
SiC (30.0 percent), which was determined according to the molecular weight ratio of SiC. Because USGS withheld
consumption data for metallurgical uses from publication for 2017 and 2018 due to concerns of disclosing
company-specific sensitive information, SiC consumption for 2017 and 2018 were estimated using 2016 values.
Emissions of C02from SiC consumption for other non-abrasive uses were calculated by multiplying the annual SiC
consumption for non-abrasive uses by the carbon content of SiC (30 percent). The annual SiC consumption for non-
abrasive uses was calculated by multiplying the annual SiC consumption (production plus net imports) by the
percentage used in metallurgical and other non-abrasive uses (50 percent) (USGS 1991a through 2017) and then
subtracting the SiC consumption for metallurgical use.
The petroleum coke portion of the total CO2 process emissions from silicon carbide 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
Industrial Processes and Product Use 4-53
-------
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
Production
Consumption
1990
65,000
132,465
2005
35,000
220,149
2016
35,000
142,104
2017
35,000
163,492
2018
35,000
168,526
2019
35,000
152,410
2020
35,000
113,736
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2020.
Uncertainty
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). There is also uncertainty associated
with the use or destruction of CFU 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.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-39. Silicon carbide
production and consumption CO2 emissions from 2020 were estimated to be between 9 percent below and 9
percent above the emission estimate of 0.15 MMT CO2 Eq. at the 95 percent confidence level. Silicon carbide
production CFU emissions were estimated to be between 9 percent below and 9 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
2020 Emission Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Silicon Carbide Production
and Consumption
C02
0.15
0.14
0.17
-9%
+9%
Silicon Carbide Production
ch4
+
+
+
-9%
+9%
+ 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).
4-54 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Recalculations Discussion
For the period 1990 through 2001, reported USGS production data included production from two facilities located
in Canada. Using SiC data from Canada (UNFCCC GHG Data Interface 2021),38 U.S. SiC production for 1990 through
2001 was recalculated to reflect only U.S. production. Using the recalculated production values, CO2 emissions
decreased by 25 to 127 kt CO2 per year, a decrease in emissions of between about 10 percent and 45 percent.
Estimates for CFU emissions decreased by about 0.1 to 0.5 kt per year, a decrease of between 20 percent and 50
percent.
Planned Improvements
EPA is initiating research for data on SiC consumption by end-use for consideration in updating emissions
estimates from SiC consumption and to account for emissions where they occur. This planned improvement is
subject to data availability and will be implemented in the medium- to long-term.
EPA has not integrated aggregated facility-level GHGRP information to inform estimates of CO2 and Cl-Ufrom SiC
production and consumption. The aggregated information (e.g., activity data and emissions) associated with silicon
carbide did not meet criteria to shield underlying confidential business information (CBI) from public disclosure.
EPA plans to examine the use of GHGRP silicon carbide emissions data for possible use in emission estimates
consistent with both Volume 1, Chapter 6 of the 2006IPCC Guidelines and the latest IPCC guidance on the use of
facility-level data in national inventories. This planned improvement is ongoing and has not been incorporated into
this Inventory report. This is a long-term planned improvement.
4.11 Titanium Dioxide Production (CRF
Source Category 2B6)
Titanium dioxide (TiCh) is manufactured using one of two processes: the chloride process and the sulfate process.
The chloride process uses petroleum coke and chlorine as raw materials and emits process-related carbon dioxide
(CO2). Emissions from fuels consumed for energy purposes during the production of titanium dioxide are
accounted for in the Energy chapter. The sulfate process does not use petroleum coke or other forms of carbon as
a raw material and does not emit CO2. The chloride process is based on the following chemical reactions and does
emit CO2:
2FeTi03 + 7Cl2 + 3C —> 2TiCl^ + 2FeCl^ + 3C02
2TiCl4 + 202 ~~* 2Ti02 ~l~ ^Cl2
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 TiC>2 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 TiC>2 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 2020, U.S. TiC>2 production totaled 1,000,000 metric tons
(USGS 2021a). Five plants produced TiC>2 in the United States in 2020.
Emissions of CO2 from titanium dioxide production in 2020 were estimated to be 1.3 MMT CO2 Eq. (1,340 kt CO2),
which represents an increase of 12 percent since 1990 (see Table 4-40). Compared to 2019, emissions from
titanium dioxide production decreased by 9 percent in 2020, due to a 9 percent decrease in production. Demand
38 The data were confirmed with Environment and Climate Change Canada.
Industrial Processes and Product Use 4-55
-------
for TiCh pigments decreased during the first half of 2020 due to restrictions implemented during the COVID-19
pandemic (USGS 2021a).
Table 4-40: CO2 Emissions from Titanium Dioxide (MMT CO2 Eq. and kt)
Year MMTCQ2Eq. kt_
1990 1.2 1,195
2005 1.8 1,755
2016 1.7 1,662
2017 1.7 1,688
2018 1.5 1,541
2019 1.5 1,474
2020 1.3 1,340
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 TiCh
Qtd = Quantity of TiCh 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 is chloride process for 1990 through
2017 (see Table 4-41) were obtained through the U.S. Geological Survey (USGS) Minerals Yearbook: Titanium
(USGS 1991 through 2020). Production data for 2018 through 2019 were obtained from the USGS Minerals
Yearbook: Titanium, advanced data release of the 2019 tables (USGS 2021b). Production data for 2020 were
4-56 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
obtained from the Minerals Commodity Summaries: Titanium and Titanium Dioxide (USGS 2021a).39 Data on the
percentage of total TiCh production capacity that is chloride process were not available for 1990 through 1993, so
data from the 1994 USGS Minerals Yearbook were used for these years. Because a sulfate process plant closed in
September 2001, the chloride process percentage for 2001 was estimated based on a discussion with Joseph
Gambogi (Gambogi 2002). By 2002, only one sulfate process plant remained online in the United States, and this
plant closed in 2004 (USGS 2005).
Table 4-41: Titanium Dioxide Production (kt)
Year
kt
1990
979
2005
1,310
2016 1,240
2017 1,260
2018 1,150
2019 1,100
2020 1,000
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2020.
Uncertainty
Each year, the USGS collects titanium industry data for titanium mineral and pigment production operations. If
TiC>2 pigment plants do not respond, production from the operations is estimated based on prior year production
levels and industry trends. Variability in response rates fluctuates from 67 to 100 percent of TiCh pigment plants
over the time series.
Although some TiC>2 may be produced using graphite or other carbon inputs, information and data regarding these
practices were not available. Titanium dioxide produced using graphite inputs, for example, may generate differing
amounts of CChper unit of TiCh produced as compared to that generated using petroleum coke in production.
While the most accurate method to estimate emissions would be to base calculations on the amount of reducing
agent used in each process rather than on the amount of TiC>2 produced, sufficient data were not available to do
so.
As of 2004, the last remaining sulfate-process plant in the United States closed. Since annual TiCh production was
not reported by USGS by the type of production process used (chloride or sulfate) prior to 2004 and only the
percentage of total production capacity by process was reported, the percent of total TiC>2 production capacity that
was attributed to the chloride process was multiplied by total TiC>2 production to estimate the amount of TiC>2
produced using the chloride process. Finally, the emission factor was applied uniformly to all chloride-process
production, and no data were available to account for differences in production efficiency among chloride-process
plants. In calculating the amount of petroleum coke consumed in chloride-process TiC>2 production, literature data
were used for petroleum coke composition. Certain grades of petroleum coke are manufactured specifically for
use in the TiC>2 chloride process; however, this composition information was not available.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-42. Titanium dioxide
consumption CO2 emissions from 2020 were estimated to be between 1.2 and 1.5 MMT CO2 Eq. at the 95 percent
39 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.
Industrial Processes and Product Use 4-57
-------
confidence level. This indicates a range of approximately 13 percent below and 13 percent above the emission
estimate of 1.3 MMT CO2 Eq.
Table 4-42: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Titanium
Dioxide Production (MMT CO2 Eq. and Percent)
Source
Gas
2020 Emission Estimate
(MMT C02 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower Upper
Lower Upper
Bound Bound
Bound 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.
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).
Recalculations Discussion
No recalculations were performed for the 1990 through 2019 portion of the time series.
Planned Improvements
EPA plans to examine the use of GHGRP titanium dioxide emissions and other data for possible use in emission
estimates consistent with both Volume 1, Chapter 6 of the 2006 IPCC Guidelines and the latest IPCC guidance on
the use of facility-level data in national inventories.40 This planned improvement is ongoing and has not been
incorporated into this Inventory report. This is a long-term planned improvement.
4.12 Soda Ash Production (CRF Source
Category 2B7)
Carbon dioxide (CO2) is generated as a byproduct of calcining trona ore to produce soda ash and is eventually
emitted into the atmosphere. In addition, CO2 may also be released when soda ash is consumed. Emissions from
soda ash consumption not associated with glass production are reported under Section 4.4 Other Process Uses of
Carbonates (CRF Category 2A4), and emissions from fuels consumed for energy purposes during the production
and consumption of soda ash are accounted for in the Energy chapter.
Calcining involves placing crushed trona ore into a kiln to convert sodium bicarbonate into crude sodium carbonate
that will later be filtered into pure soda ash. The emission of CO2 during trona-based production is based on the
following reaction:
2Na2C03 ¦ NaHC03 ¦ 2H20(Trona) -» 3Na2C03(Soda Ash) + 5H20 +C02
40 See http://www.ipcc-nggip.iges.or.jp/public/tb/TFI Technical Bulletin l.pdf.
4-58 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Soda ash (sodium carbonate, Na2COs) is a white crystalline solid that is readily soluble in water and strongly
alkaline. Commercial soda ash is used as a raw material in a variety of industrial processes and in many familiar
consumer products such as glass, soap and detergents, paper, textiles, and food. The largest use of soda ash is for
glass manufacturing. Emissions from soda ash used in glass production are reported under Section 4.2, Glass
Production (CRF Source Category 2A3). In addition, soda ash is used primarily to manufacture many sodium-based
inorganic chemicals, including sodium bicarbonate, sodium chromates, sodium phosphates, and sodium silicates
(USGS 2018b). Internationally, two types of soda ash are produced: natural and synthetic. The United States
produces only natural soda ash and is second only to China in total soda ash production. Trona is the principal ore
from which natural soda ash is made.
The United States represents about one-fifth of total world soda ash output (USGS 2021a). Only two states
produce natural soda ash: Wyoming and California. Of these two states, net emissions of CO2 from soda ash
production were only calculated for Wyoming, due to specifics regarding the production processes employed in
the state.41 Based on 2020 reported data, the estimated distribution of soda ash by end-use in 2020 (excluding
glass production) was chemical production, 54 percent; other uses, 15 percent; soap and detergent manufacturing,
11 percent; wholesale distributors (e.g., for use in agriculture, water treatment, and grocery wholesale), 10
percent; flue gas desulfurization, 6 percent; water treatment, 2 percent; and pulp and paper production, 2 percent
(USGS 2021b).42
U.S. natural soda ash is competitive in world markets because it is generally considered a better-quality raw
material than synthetically produced soda ash, and the majority of the world output of soda ash is made
synthetically. Although the United States continues to be a major supplier of soda ash, China surpassed the United
States in soda ash production in 2003, becoming the world's leading producer.
In 2020, CO2 emissions from the production of soda ash from trona ore were 1.5 MMT CO2 Eq. (1,461 kt CO2) (see
Table 4-43). Total emissions from soda ash production in 2020 decreased by approximately 18 percent compared
to emissions in 2019 primarily due to decreased global demand associated with the COVID-19 pandemic and have
increased by approximately 2 percent from 1990 levels.
Other than the significant decrease observed in 2020, emissions have remained relatively constant over the time
series with some fluctuations since 1990. In general, these fluctuations were related to the behavior of the export
market and the U.S. economy. The U.S. soda ash industry had continued a trend of increased production and value
through 2019 since experiencing a decline in domestic and export sales caused by adverse global economic
conditions in 2009.
Table 4-43: CO2 Emissions from Soda Ash Production (MMT CO2 Eq. and kt CO2)
Year MMT CP2 Eq. ktCP2
1990 1.4 1,431
2005 1.7 1,655
2016 1.7 1,723
2017 1.8 1,753
41 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.
42 Percentages may not add up to 100 percent due to independent rounding.
Industrial Processes and Product Use 4-59
-------
2018
2019
2020
1.7
1.5
1.8
1,714
1,792
1,461
Methodology and Time-Series Consistency
During the soda ash production process, trona ore is calcined in a rotary kiln and chemically transformed into a
crude soda ash that requires further processing. Carbon dioxide and water are generated as byproducts of the
calcination process. Carbon dioxide emissions from the calcination of trona ore can be estimated based on the
chemical reaction shown above. Based on this formula, which is consistent with an IPCC Tier 1 approach,
approximately 10.27 metric tons of trona ore are required to generate one metric ton of CO2, or an emission factor
of 0.0974 metric tons CO2 per metric ton of trona ore (IPCC 2006). Thus, the 15.0 million metric tons of trona ore
mined in 2020 for soda ash production (USGS 2021b) resulted in CO2 emissions of approximately 1.5 MMT CO2 Eq.
(1,461 kt).
Once produced, most soda ash is consumed in chemical production, with minor amounts used in soap production,
pulp and paper, flue gas desulfurization, and water treatment (excluding soda ash consumption for glass
manufacturing). As soda ash is consumed for these purposes, additional CO2 is usually emitted. Consistent with the
2006 IPCC Guidelines for National Greenhouse Gas Inventories, emissions from soda ash consumption in chemical
production processes are reported under Section 4.4 Other Process Uses of Carbonates (CRF Category 2A4).
Data is not currently available for the quantity of trona used in soda ash production. Because trona ore produced is
used primarily for soda ash production, EPA assumes that all trona produced was used in soda ash production. The
activity data for trona ore production (see Table 4-44) for 1990 through 2020 were obtained from the U.S.
Geological Survey (USGS) Minerals Yearbook for Soda Ash (1994 through 2015b) and USGS Mineral Industry
Surveys for Soda Ash (USGS 2016 through 2017, 2018a, 2019, 2020, 2021b). Soda ash production43 data were
collected by the USGS from voluntary surveys of the U.S. soda ash industry. EPA will continue to analyze and assess
opportunities to use facility-level data from EPA's GHGRP to improve the emission estimates for the Soda Ash
Production source category consistent with IPCC44 and UNFCCC guidelines.
Table 4-44: Trona Ore Used in Soda Ash Production (kt)
Year Use3
1990 14,700
2005 17,000
2016 17,700
2017 18,000
2018 17,600
2019 18,400
202 0 15,000
a Trona ore use is assumed to be
equal to trona ore production.
Methodological approaches were applied to the entire time series to ensure consistency in emissions estimates
from 1990 through 2020.
43 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.
44 See http://www.ipcc-nggip.iges.or.ip/public/tb/TFI Technical Bulletin l.pdf.
4-60 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Uncertainty
Emission estimates from soda ash production have relatively low associated uncertainty levels because reliable
and accurate data sources are available for the emission factor and activity data for trona-based soda ash
production. One source of uncertainty is the purity of the trona ore used for manufacturing soda ash. The emission
factor used for this estimate assumes the ore is 100 percent pure and likely overestimates the emissions from soda
ash manufacture. The average water-soluble sodium carbonate-bicarbonate content for ore mined in Wyoming
ranges from 85.5 to 93.8 percent (USGS 1995c).
EPA is aware of one facility producing soda ash from a liquid alkaline feedstock process, based on EPA's GHGRP.
Soda ash production data was collected by the USGS from voluntary surveys. A survey request was sent to each of
the five soda ash producers, all of which responded, representing 100 percent of the total production data (USGS
2020b).
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-45. Soda ash production
CO2 emissions for 2020 were estimated to be between 1.3 and 1.5 MMT CO2 Eq. at the 95 percent confidence
level. This indicates a range of approximately 9 percent below and 8 percent above the emission estimate of 1.5
MMTCCh Eq.
Table 4-45: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Soda Ash
Production (MMT CO2 Eq. and Percent)
Source
Gas
2020 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
1.5
1.3 1.5
-9% +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).
Recalculations Discussion
No recalculations were performed for the 1990 through 2019 portion of the time series.
4.13 Petrochemical Production (CRF Source
Category 2B8)
The production of some petrochemicals results in the release of carbon dioxide (CO2) and methane (CH4)
emissions. Petrochemicals are chemicals isolated or derived from petroleum or natural gas. Carbon dioxide
emissions from the production of acrylonitrile, carbon black, ethylene, ethylene dichloride, ethylene oxide, and
methanol, and CH4 emissions from the production of methanol and acrylonitrile are presented here and reported
under IPCC Source Category 2B8. The petrochemical industry uses primary fossil fuels (i.e., natural gas, coal,
petroleum, etc.) for non-fuel purposes in the production of carbon black and other petrochemicals. Emissions from
fuels and feedstocks transferred out of the system for use in energy purposes (e.g., indirect or direct process heat
Industrial Processes and Product Use 4-61
-------
or steam production) are currently accounted for in the Energy sector. The allocation and reporting of emissions
from feedstocks transferred out of the system for use in energy purposes to the Energy chapter is consistent with
the 2006IPCC Guidelines.
Worldwide, more than 90 percent of acrylonitrile (vinyl cyanide, C3H3N) is made by way of direct ammoxidation of
propylene with ammonia (NH3) and oxygen over a catalyst. This process is referred to as the SOHIO process,
named after the Standard Oil Company of Ohio (SOHIO) (IPCC 2006). The primary use of acrylonitrile is as the raw
material for the manufacture of acrylic and modacrylic fibers. Other major uses include the production of plastics
(acrylonitrile-butadiene-styrene [ABS] and styrene-acrylonitrile [SAN]), nitrile rubbers, nitrile barrier resins,
adiponitrile, and acrylamide. All U.S. acrylonitrile facilities use the SOHIO process (AN 2014). The SOHIO process
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 emissions 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:
Hf, C2H4 + H2
Small amounts of CH4 are also generated from the steam cracking process. In addition, CO2 and CH4 emissions are
also generated 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 CH4
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
4-62 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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).
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 2020 were 30.0 MMT CO2 Eq. (30,011 kt CO2) and 0.3
MMT CO2 Eq. (13 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 39
percent, and CH4 emissions increased by 43 percent. Emissions of CO2 in 2020 are 7 percent below the peak in
1999, and Cl-Uemissions in 2020 are 9 percent below the peak in 1997. Compared to 2019, CO2 emissions
decreased 2 percent in 2020, and Cm emissions decreased 5 percent. This decrease in emissions is due in part to
lower production as a result of the COVID-19 pandemic reducing demand and also a strong hurricane season that
temporarily shut down operations in Texas and Louisiana in 2020.
Table 4-46: CO2 and ChU Emissions from Petrochemical Production (MMT CO2 Eq.)
Year
1990
2005
2016
2017
2018
2019
2020
CO?
21.6
27.4
28.1
28.9
29.3
30.7
30.0
Carbon Black
3.4
4.3
3.2
3.3
3.4
3.3
2.6
Ethylene
13.1
19.0
19.6
20.0
19.4
20.7
20.7
Ethylene Dichloride
0.3
0.5
0.4
0.4
0.4
0.5
0.5
Ethylene Oxide
1.1
1.5
1.1
1.3
1.3
1.4
1.7
Acrylonitrile
1.2
1.3
1.0
1.0
1.3
1.0
0.9
Methanol
2.5
0.8
2.8
2.9
3.5
3.8
3.6
ch4
0.2
0.1
0.2
0.3
0.3
0.3
0.3
Acrylonitrile
+
+
+
+
+
+
+
Methanol
0.2
0.1
0.2
0.2
0.3
0.3
0.3
Total
21.8
27.5
28.4
29.1
29.6
31.0
30.3
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Table 4-47: CO2 and ChU Emissions from Petrochemical Production (kt)
Year
1990
2005
2016
2017
2018
2019
2020
CO?
21,611
27,383
28,110
28,890
29,314
30,702
30,011
Carbon Black
3,381
4,269
3,160
3,310
3,440
3,300
2,610
Ethylene
13,126
19,024
19,600
20,000
19,400
20,700
20,700
Ethylene Dichloride
254
455
447
412
440
503
456
Ethylene Oxide
1,123
1,489
1,100
1,250
1,300
1,370
1,680
Acrylonitrile
1,214
1,325
955
1,040
1,250
990
930
Methanol
2,513
821
2,848
2,878
3,484
3,839
3,635
ch4
9
3
10
10
12
13
13
Industrial Processes and Product Use 4-63
-------
Acrylonitrile
+
+
+
+
+
+
+
Methanol
9 1
3 |
10
10
12
13
12
+ Does not exceed 0.5 kt CH4.
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
Emissions of CO2 and CH4 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,45 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-CH4 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 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 2020
Carbon dioxide emissions and national production were aggregated directly from EPA's GHGRP dataset for 2010
through 2020 (EPA 2021). In 2020, data reported to the GHGRP included CO2 emissions of 2,610,000 metric tons
from carbon black production; 20,700,000 metric tons of CChfrom ethylene production; 456,000 metric tons of
CO2 from ethylene dichloride production; and 1,680,000 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 facilities46 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
45 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.
46 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.
4-64 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
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 2021).
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 1 CO2 and CH4
emission factors to estimate emissions for 1990 through 2019. 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 2020 were obtained from ACC's Business of Chemistry (ACC
2021). 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.
Industrial Processes and Product Use 4-65
-------
Methanol
Carbon dioxide and methane emissions from methanol production were estimated using the Tier 1 method in the
2006IPCC Guidelines. Annual methanol production data were used with IPCC default Tier 1 CO2 and Cm emission
factors to estimate emissions for 1990 through 2020. 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 2020 were obtained from the ACC's Business of Chemistry (ACC
2021). 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
2016
2017
2018
2019
2020
Carbon Black
1,310
1,650
1,190
1,240
1,280
1,210
990
Ethylene
16,500
24,000
26,600
27,800
30,500
32,400
33,500
Ethylene Dichloride
6,280
11,300
11,700
12,400
12,500
12,600
11,900
Ethylene Oxide
2,430
3,220
3,210
3,350
3,310
3,800
4,680
Acrylonitrile
1,214
1,325
955
1,040
1,250
990
930
Methanol
3,750
1,225
4,250
4,295
5,200
5,730
5,425
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 2006 IPCC Guidelines. According to the 2006 IPCC 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).47 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 2020. 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 2020. 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
47 See https://www.epa.gov/ghgreporting/historical-rulemakings.
4-66 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
through 2009 and GHGRP for 2010 through 2020. 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.
Uncertainty
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
2020 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT CO? Eq.)
(MMT CO? Eq.)
(%)
Lower Upper
Lower
Upper
Bound Bound
Bound
Bound
Petrochemical
Production
C02
30.0
28.4 31.7
-5%
+6%
Petrochemical
Production
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
Industrial Processes and Product Use 4-67
-------
98).48 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).49 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 2018; in 2018 through 2020, the GHGRP emissions have been about 20 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. For ethylene oxide, GHGRP emissions vary from 17 percent
less than the Tier 1 emissions to 20 percent more than the Tier 1 emissions, depending on the year.
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 500 kt/yr,
48 See http://www.ecfr.gov/cgi-bin/text-idx7tpN/ecfrbrowse/Title40/40cfr98 main Q2.tpl.
49 See https://www.epa.gov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
4-68 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
which is not significant enough to prioritize for inclusion in the report at 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 production quantity for 2019 was updated with the revised value in ACC's Business of Chemistry
(ACC 2021). This change resulted in less than a 0.3 percent (90 kt) decrease in total petrochemical emissions for
2019, compared to the previous Inventory.
Emissions from ethylene production in 2016 and emissions from carbon black production in 2017 were updated
and reduced slightly to be consistent with updated GHGRP data (EPA 2021). These changes resulted in a 0.7
percent (200 kt) decrease in total emissions from petrochemical production for 2016 and a 0.07 percent (20 kt)
decrease in total emissions from petrochemical production for 2017, compared to the previous Inventory.
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 to 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.
EPA plans to review USGS data to improve use of activity data to estimate emissions, consistent with the
methodological decision trees in 2006IPCC Guidelines. EPA also plans to review GHGRP emissions data for possible
use in estimates, consistent with both Volume 1, Chapter 6 of the 2006 IPCC Guidelines and the latest IPCC
guidance on the use of facility-level data in national inventories.50 This planned improvement is ongoing and has
not been incorporated into this Inventory report. This is a medium-term planned improvement and expected to be
completed by the next (i.e., 2023) Inventory submission.
50 See http://www.ipcc-nggiD.iges.or.jp/public/tb/TFl Technical Bulletin l.pdf.
Industrial Processes and Product Use 4-69
-------
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.51 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 2020. Emissions of HFC-23 from this activity in 2020 were
estimated to be 2.1 MMT CO2 Eq. (0.1 kt) (see Table 4-50). This quantity represents a 43 percent decrease from
2019 emissions and a 95 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 decrease from
2019 emissions was caused by both a decrease in the HFC-23 emission rate at one plant and a decrease 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.
51 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],
4-70 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 4-50: HFC-23 Emissions from HCFC-22 Production (MMT CCh Eq. and kt HFC-23)
Year
MMTCOz Eq.
kt HFC-23
1990
46.1
3
2005
20.0
1
2016
2017
2018
2019
2020
2.8
5.2
3.3
3.7
2.1
0.2
0.3
0.2
0.3
0.1
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 2020 were obtained through reports submitted by U.S. HCFC-22
production facilities to EPA's Greenhouse Gas Reporting Program (GHGRP). EPA's GHGRP mandates that all HCFC-
22 production facilities report their annual emissions of HFC-23 from HCFC-22 production processes and HFC-23
destruction processes. Previously, data were obtained by EPA through collaboration with an industry association
that received voluntarily reported HCFC-22 production and HFC-23 emissions annually from all U.S. HCFC-22
producers from 1990 through 2009. These emissions were aggregated and reported to EPA on an annual basis.
For the other three plants, the last of which closed in 1993, methods comparable to the Tier 1 method in the 2006
IPCC Guidelines were used. Emissions from these three plants have been calculated using the recommended
emission factor for unoptimized plants operating before 1995 (0.04 kg HCFC-23/kg HCFC-22 produced).
The five plants that have operated since 1994 measure (or, for the plants that have since closed, measured)
concentrations of HFC-23 as well as mass flow rates of process streams to estimate their generation of HFC-23.
Plants using thermal oxidation to abate their HFC-23 emissions monitor the performance of their oxidizers to verify
that the HFC-23 is almost completely destroyed. One plant that releases a small fraction of its byproduct HFC-23
periodically measures HFC-23 concentrations at process vents using gas chromatography. This information is
combined with information on quantities of products (e.g., HCFC-22) to estimate HFC-23 emissions.
To estimate 1990 through 2009 emissions, reports from an industry association were used that aggregated HCFC-
22 production and HFC-23 emissions from all U.S. HCFC-22 producers and reported them to EPA (ARAP 1997,1999,
2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, and 2010). To estimate 2010 through 2020
emissions, facility-level data (including both HCFC-22 production and HFC-23 emissions) reported through EPA's
GHGRP were analyzed. In 1997 and 2008, comprehensive reviews of plant-level estimates of HFC-23 emissions and
HCFC-22 production were performed (RTI1997; RTI 2008). The 1997 and 2008 reviews enabled U.S. totals to be
reviewed, updated, and where necessary, corrected, and also for plant-level uncertainty analyses (Monte-Carlo
simulations) to be performed for 1990,1995, 2000, 2005, and 2006. Estimates of annual U.S. HCFC-22 production
are presented in Table 4-51.
Industrial Processes and Product Use 4-71
-------
Table 4-51: HCFC-22 Production (kt)
Year kt
1990 139
2005 156
2012 96
2013-2020 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.
Uncertainty
The uncertainty analysis presented in this section was based on a plant-level Monte Carlo Stochastic Simulation for
2006. The Monte Carlo analysis used estimates of the uncertainties in the individual variables in each plant's
estimating procedure. This analysis was based on the generation of 10,000 random samples of model inputs from
the probability density functions for each input. A normal probability density function was assumed for all
measurements and biases except the equipment leak estimates for one plant; a log-normal probability density
function was used for this plant's equipment leak estimates. The simulation for 2006 yielded a 95-percent
confidence interval for U.S. emissions of 6.8 percent below to 9.6 percent above the reported total.
The relative errors yielded by the Monte Carlo Stochastic Simulation for 2006 were applied to the U.S. emission
estimate for 2020. The resulting estimates of absolute uncertainty are likely to be reasonably accurate because (1)
the methods used by the two remaining plants to estimate their emissions are not believed to have changed
significantly since 2006, and (2) although the distribution of emissions among the plants has changed between
2006 and 2020 (because one plant has closed), the plant that currently accounts for most emissions had a relative
uncertainty in its 2006 (as well as 2005) emissions estimate that was similar to the relative uncertainty for total
U.S. emissions. Thus, the closure of one plant is not likely to have a large impact on the uncertainty of the national
emission estimate.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-52. HFC-23 emissions
from HCFC-22 production were estimated to be between 2.0 and 2.3 MMT CO2 Eq. at the 95 percent confidence
level. This indicates a range of approximately 7 percent below and 10 percent above the emission estimate of 2.1
MMTCCh Eq.
Table 4-52: Approach 2 Quantitative Uncertainty Estimates for HFC-23 Emissions from
HCFC-22 Production (MMT CO2 Eq. and Percent)
2020 Emission Estimate Uncertainty Range Relative to Emission Estimate3
Source Gas
(MMT C02 Eq.) (MMT CP2 Eq.) (%)
Lower Upper Lower Upper
Bound Bound Bound Bound
HCFC-22 Production HFC-23 2.1 2.0 2.3 -7% +10%
a Range of emissions reflects 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
4-72 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
introduction of the IPPU chapter (see Annex 8 for more details). Under the 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 (EPA 2015).52 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.
The GHGRP also requires source-specific quality control measures for the HCFC-22 Production category. Under
EPA's GHGRP, HCFC-22 producers are required to (1) measure concentrations of HFC-23 and HCFC-22 in the
product stream at least weekly using equipment and methods (e.g., gas chromatography) with an accuracy and
precision of 5 percent or better at the concentrations of the process samples, (2) measure mass flows of HFC-23
and HCFC-22 at least weekly using measurement devices (e.g., flowmeters) with an accuracy and precision of 1
percent of full scale or better, (3) calibrate mass measurement devices at the frequency recommended by the
manufacturer using traceable standards and suitable methods published by a consensus standards organization,
(4) calibrate gas chromatographs at least monthly through analysis of certified standards, and (5) document these
calibrations.
Recalculations
The emissions estimate for 2011 was revised to exclude HFC-23 emissions from one plant that did not produce
HCFC-22. This revision resulted in a decrease in 2011 emissions of 459 kg HFC-23, about 0.07 percent of the
previous estimate.
4.15 Carbon Dioxide Consumption (CRF
Source Category 2B10)
Carbon dioxide (CO2) is used for a variety of commercial applications, including food processing, chemical
production, carbonated beverage production, and refrigeration, and is also used in petroleum production for
enhanced oil recovery (EOR). CO2 used for EOR is injected underground to enable additional petroleum to be
produced. For the purposes of this analysis, CO2 used in commercial applications other than EOR is assumed to be
emitted to the atmosphere. A further discussion of CO2 used in EOR is described in the Energy chapter in Box 3-6
titled "Carbon Dioxide Transport, Injection, and Geological Storage" and is not included in this section.
Carbon dioxide is produced from naturally-occurring CO2 reservoirs, as a byproduct from the energy and industrial
production processes (e.g., ammonia production, fossil fuel combustion, ethanol production), and as a byproduct
from the production of crude oil and natural gas, which contain naturally occurring CO2 as a component.
In 2020, the amount of CO2 produced and captured for commercial applications and subsequently emitted to the
atmosphere was 5.0 MMT CO2 Eq. (4,970 kt) (see Table 4-53). This is a 2 percent increase (100 kt) from 2019 levels
and is an increase of approximately 238 percent since 1990.
Table 4-53: CO2 Emissions from CO2 Consumption (MMT CO2 Eq. and kt)
Year MMT CP2 Eq. kt_
1990 1.5 1,472
52 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 4-73
-------
Year
MMTCOz Eq.
kt
2005
1.4
1,375
2016
2017
2018
2019
2020
4.6
4.6
4.1
4.9
5.0
4,640
4,580
4,130
4,870
4,970
Methodology and Time-Series Consistency
Carbon dioxide emission estimates for 1990 through 2020 were based on the quantity of CO2 extracted and
transferred for industrial applications (i.e., non-EOR end-uses). Some of the CO2 produced by these facilities is used
for EOR, and some is used in other commercial applications (e.g., chemical manufacturing, food production). It is
assumed that 100 percent of the CO2 production used in commercial applications other than EOR is eventually
released into the atmosphere.
For 2010 through 2020, data from EPA's GHGRP (Subpart PP) were aggregated from facility-level reports to
develop a national-level estimate for use in the Inventory (EPA 2021). Facilities report CO2 extracted or produced
from natural reservoirs and industrial sites, and CO2 captured from energy and industrial processes and transferred
to various end-use applications to EPA's GHGRP. This analysis includes only reported CO2 transferred to food and
beverage end-uses. EPA is continuing to analyze and assess integration of CO2 transferred to other end-uses to
enhance the completeness of estimates under this source category. Other end-uses include industrial applications,
such as metal fabrication. EPA is analyzing the information reported to ensure that other end-use data excludes
non-emissive applications and publication will not reveal CBI. Reporters subject to EPA's GHGRP Subpart PP are
also required to report the quantity of CO2 that is imported and/or exported. Currently, these data are not publicly
available through the GHGRP due to data confidentiality reasons and hence are excluded from this analysis.
Facilities subject to Subpart PP of EPA's GHGRP are required to measure CO2 extracted or produced. More details
on the calculation and monitoring methods applicable to extraction and production facilities can be found under
Subpart PP: Suppliers of Carbon Dioxide of the regulation, Part 98.53 The number of facilities that reported data to
EPA's GHGRP Subpart PP (Suppliers of Carbon Dioxide) for 2010 through 2020 is much higher (ranging from 44 to
53) than the number of facilities included in the Inventory for the 1990 to 2009 time period prior to the availability
of GHGRP data (4 facilities). The difference is largely due to the fact the 1990 to 2009 data includes only CO2
transferred to end-use applications from naturally occurring CO2 reservoirs and excludes industrial sites.
For 1990 through 2009, data from EPA's GHGRP are not available. For this time period, CO2 production data from
four naturally-occurring CO2 reservoirs were used to estimate annual CO2 emissions. These facilities were Jackson
Dome in Mississippi, Bravo and West Bravo Domes in New Mexico, and McCallum Dome in Colorado. The facilities
in Mississippi and New Mexico produced CO2 for use in both EOR and in other commercial applications (e.g.,
chemical manufacturing, food production). The fourth facility in Colorado (McCallum Dome) produced CO2 for
commercial applications only (New Mexico Bureau of Geology and Mineral Resources 2006).
Carbon dioxide production data and the percentage of production that was used for non-EOR applications for the
Jackson Dome, Mississippi facility were obtained from Advanced Resources International (ARI 2006, 2007) for 1990
53 See http://www.ecfr.gov/cgi-bin/text-idx7tpk/ecfrbrowse/Title40/40cfr98 main Q2.tpl.
2010 through 2020
1990 through 2009
4-74 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
to 2000, and from the Annual Reports of Denbury Resources (Denbury Resources 2002 through 2010) for 2001 to
2009 (see Table 4-54). Denbury Resources reported the average CO2 production in units of MMCF CO2 per day for
2001 through 2009 and reported the percentage of the total average annual production that was used for EOR.
Production from 1990 to 1999 was set equal to 2000 production, due to lack of publicly available production data
for 1990 through 1999. Carbon dioxide production data for the Bravo Dome and West Bravo Dome were obtained
from ARI for 1990 through 2009 (ARI1990 to 2010). Data for the West Bravo Dome facility were only available for
2009. The percentage of total production that was used for non-EOR applications for the Bravo Dome and West
Bravo Dome facilities for 1990 through 2009 were obtained from New Mexico Bureau of Geology and Mineral
Resources (Broadhead 2003; New Mexico Bureau of Geology and Mineral Resources 2006). Production data for the
McCallum Dome (Jackson County), Colorado facility were obtained from the Colorado Oil and Gas Conservation
Commission (COGCC) for 1999 through 2009 (COGCC 2014). Production data for 1990 to 1998 and percentage of
production used for EOR were assumed to be the same as for 1999, due to lack of publicly available data.
Table 4-54: CO2 Production (kt CO2) and the Percent Used for Non-EOR Applications
Year
Jackson Dome,
MS
C02 Production
(kt) (% Non-EOR)
Bravo Dome,
NM
C02 Production
(kt) (% Non-EOR)
West Bravo Dome,
NM
C02 Production
(kt) (% Non-EOR)
McCallum Dome,
CO
C02 Production
(kt) (% Non-EOR)
Total C02
Production
from Extraction
and Capture
Facilities (kt)
%
Non-
EOR3
1990
1,344 (100%)
63 (1%)
+
65 (100%)
NA
NA
2005
1,254 (27%)
58 (1%)
+
63 (100%)
NA
NA
2016
NA
NA
NA
NA
55,900b
8%
2017
NA
NA
NA
NA
59,900b
8%
2018
NA
NA
NA
NA
58,400b
7%
2019
NA
NA
NA
NA
61,300b
8%
2020
NA
NA
NA
NA
44,700b
11%
+ Does not exceed 0.5 percent.
NA (Not Available)
a Includes only food and beverage applications.
b For 2010 through 2020, the publicly available GHGRP data were aggregated at the national level based on GHGRP CBI
criteria.
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2020. The methodology for CO2 consumption spliced activity data from two different sources: Industry
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
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.
Industrial Processes and Product Use 4-75
-------
Based on the results of the verification process, EPA follows up with facilities to resolve mistakes that may have
occurred.54
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-55. Carbon dioxide
consumption CO2 emissions for 2020 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
2020 Emission Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Emission Estimate3
(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 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 CO2 Consumption can be found under Subpart PP (Suppliers of
Carbon Dioxide) of the regulation (40 CFR Part 98).55 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).56 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 2019 portion of the time series.
Planned Improvements
EPA will continue to evaluate the potential to include additional GHGRP data on other emissive end-uses to
improve the accuracy and completeness of estimates for this source category. Particular attention will be made to
ensuring 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
54 See https://www.epa.gov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
55 See http://www.ecfr.gov/cgi-bin/text-idxPtpk/ecfrbrowse/Title40/40cfr98 main Q2.tpl.
56 See https://www.epa.gov/sites/production/files/2015-07/documents/ghgrp verification factsheet.pdf.
57 See http://www.ipcc-nggip.iges.or.jp/public/tb/TFI Technical Bulletin l.pdf.
4-76 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
These improvements are still in process and will be incorporated into future Inventory reports. These are near-to
medium-term improvements.
4.16 Phosphoric Acid Production (CRF
Source Category 2BIO)
Phosphoric acid (H3PO4) is a basic raw material used in the production of phosphate-based fertilizers. Phosphoric
acid production from natural phosphate rock is a source of carbon dioxide (CO2) emissions, due to the chemical
reaction of the inorganic carbon (calcium carbonate) component of the phosphate rock.
Phosphate rock is mined in Florida and North Carolina, which account for more than 75 percent of total domestic
output, and in Idaho and Utah (USGS 2021a). It is used primarily as a raw material for wet-process phosphoric acid
production. The composition of natural phosphate rock varies, depending on the location where it is mined.
Natural phosphate rock mined in the United States generally contains inorganic carbon in the form of calcium
carbonate (limestone) and may also contain organic carbon.
The phosphoric acid production process involves chemical reaction of the calcium phosphate (CasfPCUh)
component of the phosphate rock with sulfuric acid (H2SO4) and recirculated phosphoric acid (H3PO4) (EFMA 2000).
The generation of CO2, however, is due to the associated limestone-sulfuric acid reaction, as shown below:
CaCO3 + //2S04 + H20 —* CaS04 ¦ 2H20 + C02
Total U.S. phosphate rock production used in 2020 was an estimated 24.0 million metric tons (USGS 2021a). Total
imports of phosphate rock to the United States in 2020 were 2.3 million metric tons (USGS 2021a). Between 2016
and 2019, most of the imported phosphate rock (85 percent) came from Peru, with 15 percent from Morocco
(USGS 2021a). All phosphate rock mining companies in the United States are vertically integrated with fertilizer
plants that produce phosphoric acid located near the mines. The phosphoric acid production facilities that use
imported phosphate rock are located in Louisiana.
Over the 1990 to 2020 period, domestic phosphate rock production has decreased by nearly 52 percent. Total CO2
emissions from phosphoric acid production were 0.9 MMT CO2 Eq. (938 kt CO2) in 2020 (see Table 4-56). Domestic
consumption of phosphate rock in 2020 was estimated to have increased 3 percent relative to 2019 levels. The
COVID-19 pandemic did not have a major effect on the domestic phosphate rock market as both the fertilizer
industry and related agricultural businesses were considered essential industries (USGS 2021a).
Table 4-56: CO2 Emissions from Phosphoric Acid Production (MMT CO2 Eq. and kt)
Year
MMT C02 Eq.
kt
1990
1.5
1,529
2005
1.3
1,342
2016
1.0
998
2017
1.0
1,025
2018
0.9
937
2019
0.9
909
2020
0.9
938
Industrial Processes and Product Use 4-77
-------
Methodology and Time-Series Consistency
The United States uses a country-specific methodology consistent with an IPCC Tier 1 approach to calculate
emissions from production of phosphoric acid from phosphate rock.58 Carbon dioxide emissions from production
of phosphoric acid from phosphate rock are estimated by multiplying the average amount of inorganic carbon
(expressed as CO2) contained in the natural phosphate rock as calcium carbonate by the amount of phosphate rock
that is used annually to produce phosphoric acid, accounting for domestic production and net imports for
consumption. The estimation methodology is as follows:
Equation 4-9: CO2 Emissions from Phosphoric Acid Production
Epa Cpr * Qpr
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
The CO2 emissions calculation methodology assumes that all of the inorganic C (calcium carbonate) content of the
phosphate rock reacts to produce CO2 in the phosphoric acid production process and is emitted with the stack gas.
The methodology also assumes that none of the organic C content of the phosphate rock is converted to CO2 and
that all of the organic C content remains in the phosphoric acid product.
From 1993 to 2004, the U.S. Geological Survey (USGS) Mineral Yearbook: Phosphate Rock disaggregated phosphate
rock mined annually in Florida and North Carolina from phosphate rock mined annually in Idaho and Utah, and
reported the annual amounts of phosphate rock exported and imported for consumption (see Table 4-57). For the
years 1990 through 1992, and 2005 through 2019, only nationally aggregated mining data was reported by USGS.
For the years 1990,1991, and 1992, the breakdown of phosphate rock mined in Florida and North Carolina, and
the amount mined in Idaho and Utah, are approximated using data reported by USGS for the average share of U.S.
production in those states from 1993 to 2004. For the years 2005 through 2016 and years 2017 through 2020, the
same approximation method is used, but the share of U.S. production based on production capacity in those states
were obtained from the USGS commodity specialist for phosphate rock (USGS 2012; USGS 2021b). Data for
domestic sales or consumption of phosphate rock, exports of phosphate rock (primarily from Florida and North
Carolina), and imports of phosphate rock for consumption for 1990 through 2010 were obtained from USGS
Minerals Yearbook: Phosphate Rock (USGS 1994 through 2015b), and from USGS Minerals Commodity Summaries:
Phosphate Rock (USGS 2016 through 2021a). From 2004 through 2019, the USGS reported no exports of phosphate
rock from U.S. producers (USGS 2021a).
The carbonate content of phosphate rock varies depending upon where the material is mined. Composition data
for domestically mined and imported phosphate rock were provided by the Florida Institute of Phosphate
Research, now known as the Florida Industrial and Phosphate Research Institute (FIPR 2003a). Phosphate rock
mined in Florida contains approximately 1 percent inorganic C, and phosphate rock imported from Morocco
contains approximately 1.46 percent inorganic C. Calcined phosphate rock mined in North Carolina and Idaho
contains approximately 0.41 percent and 0.27 percent inorganic C, respectively (see Table 4-58). Similar to the
phosphate rock mined in Morocco, phosphate rock mined in Peru contains approximately 5 percent CO2 (Golder
Associates and M3 Engineering 2016).
Carbonate content data for phosphate rock mined in Florida are used to calculate the CO2 emissions from
consumption of phosphate rock mined in Florida and North Carolina (more than 75 percent of domestic
production), and carbonate content data for phosphate rock mined in Morocco and Peru are used to calculate CO2
emissions from consumption of imported phosphate rock. The CO2 emissions calculation assumes that all of the
58 The 2006 IPCC Guidelines do not provide a method for estimating process emissions (C02) from Phosphoric Acid Production.
4-78 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
domestic production of phosphate rock is used in uncalcined form. As of 2006, the USGS noted that one phosphate
rock producer in Idaho produces calcined phosphate rock; however, no production data were available for this
single producer (USGS 2006). The USGS confirmed that no significant quantity of domestic production of
phosphate rock is in the calcined form (USGS 2012).
Table 4-57: Phosphate Rock Domestic Consumption, Exports, and Imports (kt)
Location/Year
1990
2005
2016
2017
2018
2019
2020
U.S. Domestic Consumption
49,800
35,200
26,700
26,300
23,300
23,400
24,000
FL and NC
42,494
28,160
21,360
20,510
18,170
18,250
18,720
ID and UT
7,306
7,040
5,340
5,790
5,130
5,150
5,280
Exports—FL and NC
6,240
0
0
0
0
0
0
Imports
451
2,630
1,590
2,470
2,770
2,140
2,300
Total U.S. Consumption
44,011
37,830
28,290
28,770
26,070
25,540
26,300
Note: Totals may not sum due to independent rounding.
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)
Methodological approaches were applied to the entire time series to ensure consistency in emissions estimates
from 1990 through 2020.
Uncertainty
Phosphate rock production data used in the emission calculations were developed by the USGS through monthly
and semiannual voluntary surveys of the active phosphate rock mines during 2019. Prior to 2006, USGS provided
the data disaggregated regionally; however, beginning in 2006, only total U.S. phosphate rock production was
reported. Regional production for 2020 was estimated based on regional production data from 2017 to 2020 and
multiplied by regionally-specific emission factors. There is uncertainty associated with the degree to which the
estimated 2019 regional production data represents actual production in those regions. Total U.S. phosphate rock
production data are not considered to be a significant source of uncertainty because all the domestic phosphate
rock producers report their annual production to the USGS. Data for exports of phosphate rock used in the
emission calculations are reported to the USGS by phosphate rock producers and are not considered to be a
significant source of uncertainty. Data for imports for consumption are based on international trade data collected
by the U.S. Census Bureau. These U.S. government economic data are not considered to be a significant source of
uncertainty.
An additional source of uncertainty in the calculation of CO2 emissions from phosphoric acid production is the
carbonate composition of phosphate rock, as the composition of phosphate rock varies depending upon where the
material is mined and may also vary over time. The Inventory relies on one study (FIPR 2003a) of chemical
composition of the phosphate rock; limited data are available beyond this study. Another source of uncertainty is
the disposition of the organic carbon content of the phosphate rock. A representative of FIPR indicated that in the
phosphoric acid production process the organic C content of the mined phosphate rock generally remains in the
phosphoric acid product, which is what produces the color of the phosphoric acid product (FIPR 2003b). Organic
carbon is therefore not included in the calculation of CO2 emissions from phosphoric acid production.
Industrial Processes and Product Use 4-79
-------
A third source of uncertainty is the assumption that all domestically-produced phosphate rock is used in
phosphoric acid production and used without first being calcined. Calcination of the phosphate rock would result
in conversion of some of the organic C in the phosphate rock into CO2. However, according to air permit
information available to the public, at least one facility has calcining units permitted for operation (NCDENR 2013).
Finally, USGS indicated that in 2020 less than 5 percent of domestically-produced phosphate rock was used to
manufacture elemental phosphorus and other phosphorus-based chemicals, rather than phosphoric acid (USGS
2021a). According to USGS, there is only one domestic producer of elemental phosphorus, in Idaho, and no data
were available concerning the annual production of this single producer. Elemental phosphorus is produced by
reducing phosphate rock with coal coke, and it is therefore assumed that 100 percent of the carbonate content of
the phosphate rock will be converted to CO2 in the elemental phosphorus production process. The calculation for
CO2 emissions assumes that phosphate rock consumption, for purposes other than phosphoric acid production,
results in CChemissions from 100 percent of the inorganic carbon content in phosphate rock, but none from the
organic carbon content.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-59. 2020 phosphoric acid
production CO2 emissions were estimated to be between 0.8 and 1.2 MMT CO2 Eq. at the 95 percent confidence
level. This indicates a range of approximately 18 percent below and 20 percent above the emission estimate of 0.9
MMTCO2 Eq.
Table 4-59: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from
Phosphoric Acid Production (MMT CO2 Eq. and Percent)
Source
Gas
2020 Emission Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Phosphoric Acid Production
C02
0.9
-------
(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.59 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 iron60 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 2020, approximately eleven integrated iron and steel steelmaking facilities utilized BOFs to refine and produce
steel from iron, and raw steel was produced at 98 facilities across the United States. Approximately 30 percent of
steel production was attributed to BOFs and 70 percent to EAFs (USGS 2021). 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-
located with integrated iron and steel facilities (ACCCI 2021). In the United States, 6 states account for roughly 51
percent of total raw steel production: Indiana, Alabama, Tennessee, Kentucky, Mississippi, and Arkansas (AISI
2021).
Total annual production of crude steel in the United States was fairly constant between 2000 and 2008 and ranged
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
demand caused by the global economic downturn (particularly from the automotive industry), crude steel
production in the United States sharply decreased to 65,459,000 tons in 2009. Crude steel production was fairly
constant from 2011 through 2014, and after a dip in production from 2014 to 2015, crude steel production has
slowly and steadily increased for the past few years. The United States was the fourth largest producer of raw steel
59 See http://www.ipcc-nggip.iges.or.jp/public/tb/TFl Technical Bulletin l.pdf.
60 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 4-81
-------
in the world, behind China, India and Japan, accounting for approximately 3.9 percent of world production in 2020
(AISI 2004 through 2021).
The majority of CO2 emissions from the iron and steel production process come from the use of metallurgical coke
in the production of pig iron and from the consumption of other process byproducts, with lesser amounts emitted
from the use of carbon-containing flux and from the removal of carbon from pig iron used to produce steel.
According to the 2006IPCC Guidelines, the production of metallurgical coke from coking coal is considered to be an
energy use of fossil fuel, and the use of coke in iron and steel production is considered to be an industrial process
source. The 2006 IPCC Guidelines suggest that emissions from the production of metallurgical coke should be
reported separately in the Energy sector, while emissions from coke consumption in iron and steel production
should be reported in the Industrial Processes and Product Use sector. The approaches and emission estimates for
both metallurgical coke production and iron and steel production, however, are presented here because much of
the relevant activity data is used to estimate emissions from both metallurgical coke production and iron and steel
production. For example, some byproducts (e.g., coke oven gas) of the metallurgical coke production process are
consumed during iron and steel production, and some byproducts of the iron and steel production process (e.g.,
blast furnace gas) are consumed during metallurgical coke production. Emissions associated with the consumption
of these byproducts are attributed at the point of consumption. Emissions associated with the use of conventional
fuels (e.g., natural gas, fuel oil) for electricity generation, heating and annealing, or other miscellaneous purposes
downstream of the iron and steelmaking furnaces are reported in the Energy chapter.
Metallurgical Coke Production
Emissions of CO2 from metallurgical coke production in 2020 were 2.3 MMT CO2 Eq. (2,324 kt CO2) (see Table 4-60
and Table 4-61). Emissions decreased by 23 percent from 2019 to 2020 and have decreased by 59 percent since
1990. Coke production in 2020 was about 20 percent lower than in 2019 and 63 percent below 1990 (EIA 2021,
AISI 2021).
Significant activity data for 2020 were not available in time for publication of this report and were estimated using
2019 values adjusted based on GHGRP emissions data, as described in the Methodology and Time-Series
Consistency section below.
Table 4-60: CO2 Emissions from Metallurgical Coke Production (MMT CO2 Eq.)
Gas 1990 2005 2016 2017 2018 2019 2020
CO2 ^6 3J) Z6 Z0 13 ^0 23_
Table 4-61: CO2 Emissions from Metallurgical Coke Production (kt)
Gas 1990 2005 2016 2017 2018 2019 2020
CO; 5,608 4 3,921 2,643 1,978 1,282 3,006 2,324
Iron and Steel Production
Emissions of CO2 and CH4 from iron and steel production in 2020 were 35.4 MMT CO2 Eq. (35,386 kt) and 0.0066
MMT CO2 Eq. (0.3 kt CH4), respectively (see Table 4-62 through Table 4-65), totaling 35.4 MMT CO2 Eq. Emissions
from iron and steel production decreased by 12 percent from 2019 to 2020 and have decreased by 64 percent
since 1990, due to restructuring of the industry, technological improvements, and increased scrap steel utilization.
Carbon dioxide emission estimates include emissions from the consumption of carbonaceous materials in the blast
furnace, EAF, and BOF, as well as blast furnace gas and coke oven gas consumption for other activities at the steel
mill.
Significant activity data for 2020 were not available in time for publication of this report and were estimated using
2019 values adjusted based on GHGRP emissions data, as described in the Methodology and Time-Series
Consistency section below.
4-82 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
In 2020, domestic production of pig iron decreased by 18 percent from 2019 levels. Overall, domestic pig iron
production has declined since the 1990s. Pig iron production in 2020 was 62 percent lower than in 2000 and 63
percent below 1990. Carbon dioxide emissions from iron production have decreased by 82 percent (37.3 MMT CO2
Eq.) since 1990. Carbon dioxide emissions from steel production have decreased by 29 percent (2.3 MMT CO2 Eq.)
since 1990, while overall CO2 emissions from iron and steel production have declined by 64 percent (63.7 MMT
CO2 Eq.) from 1990 to 2020.
Table 4-62: CO2 Emissions from Iron and Steel Production (MMT CO2 Eq.)
Source/Activity Data
1990
2005
2016
2017
2018
2019
2020
Sinter Production
2.4
1.7
0.9
0.9
0.9
0.9
0.8
Iron Production
45.7
17.7
9.9
8.2
9.6
9.4
8.4
Pellet Production
1.8
1.5
0.9
0.9
0.9
0.9
0.8
Steel Production
8.0
9.4
6.9
6.2
5.8
5.8
5.6
Other Activities3
41.2
35.9
22.5
22.4
24.1
23.2
19.8
Total
99.1
66.2
41.0
38.6
41.3
40.1
35.4
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.
Table 4-63: CO2 Emissions from Iron and Steel Production (kt)
Source/Activity Data
1990
2005
2016
2017
2018
2019
2020
Sinter Production
2,448
1,663
877
869
937
876
750
Iron Production
45,706
17,661
9,928
8,237
9,581
9,360
8,416
Pellet Production
1,817
1,503
869
867
924
878
752
Steel Production
7,964
9,395
6,854
6,218
5,754
5,812
5,650
Other Activities3
41,194
35,934
22,451
22,396
24,149
23,158
19,838
Total
99,129
66,156
40,979
38,587
41,345
40,084
35,407
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.
Table 4-64: ChU Emissions from Iron and Steel Production (MMT CO2 Eq.)
Source/Activity Data
1990
2005
2016
2017
2018
2019
2020
Sinter Production
+
+
+
+
+
+
+
+ Does not exceed 0.05 MMT C02 Eq.
Table 4-65: ChU Emissions from Iron and Steel Production (kt)
Source/Activity Data
1990
2005
2016
2017
2018
2019
2020
Sinter Production
0.9
0.6
0.3
0.3
0.3
0.3
0.3
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
Industrial Processes and Product Use 4-83
-------
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 2006IPCC Guidelines Tier 2 Methodologies
Em- —
^(<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: 2006IPCC Guidelines Tier 1: Emissions from Sinter, Direct Reduced Iron, and
Pellet Production (Equations 4.6,4.7, and 4.8)
ES,P = Qs x EFSJ)
Ed,co2 = Qd x EFd£02
Ep,co2 = Qp x EFp C02
where,
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
A significant number of activity data that serve as inputs to emissions calculations were unavailable for 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. GHGRP process emissions data
decreased by approximately 14 percent from 2019 to 2020 (EPA 2021), and this percentage decrease was applied
to all 2020 activity data estimated with 2019 values.
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
4-84 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
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) (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 2021) and through personal
communications with AISI (Steiner 2008) (see Table 4-68). Coke plant consumption and production data from the
AISI Annual Statistical Report 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.
Industrial Processes and Product Use 4-85
-------
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 2016 2017 2018 2019 2020
Metallurgical Coke Production
Coking Coal Consumption at Coke Plants 35,269 21,259 14,955 15,910 16,635 16,261 13,076
Coke Production at Coke Plants 25,054 4 15,167 10,755 11,746 12,525 11,676 9,392
Coke Breeze Production 2,645 *4 1,594 1,122 1,193 1,248 1,220 981
Coal Tar Production 1,058 j 638 449 477 499 488 392
Table 4-68: Production and Consumption Data for the Calculation of CO2 Emissions from
Metallurgical Coke Production (Million ft3)
Source/Activity Data 1990 2005 2016 2017 2018 2019 2020
Metallurgical Coke Production
Coke Oven Gas Production 250,767 114,213 74,807 74,997 80,750 77,692 66,554
Natural Gas Consumption 599 2,996 2,077 2,103 2,275 2,189 1,875
Blast Furnace Gas Consumption 24,602 4,460 3,741 3,683 4,022 3,914 3,353
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
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 pig iron production was deducted from the "Other Process
Uses of Carbonates" source category (CRF Source Category 2A4) to avoid double-counting.
4-86 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 Cm, 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 Cm 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 CH4 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.
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.
Industrial Processes and Product Use 4-87
-------
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 2021) and through personal communications with AISI (Steiner 2008) (see
Table 4-72). Data from the AISI Annual Statistical Report 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.
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); however, 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. Also, 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). Some data from the AISI Annual Statistical Report 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 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 2021) 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 2021) and through personal communications with AISI (Steiner 2008).
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). 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). Some data from the AISI
Annual Statistical Report on natural gas consumption were withheld for 2020, so the 2019 values were used as
4-88 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
2016
2017
2018
2019
2020
Sinter Production
12,239
8,315
4,385
4,347
4,687
4,378
3,751
Direct Reduced Iron Production
517
1,303
C
C
C
C
C
Pellet Production
60,563
50,096
28,967
28,916
30,793
29,262
25,067
Pig Iron Production
Coke Consumption
24,946
13,832
7,124
7,101
7,618
7,291
6,246
Pig Iron Production
49,669
37,222
22,293
22,395
24,058
22,302
18,320
Direct Injection Coal
Consumption
1,485
2,573
1,935
2,125
2,569
2,465
2,112
EAF Steel Production
EAF Anode and Charge Carbon
Consumption
67
1,127
1,120
1,127
1,133
1,137
1,118
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
52,589
55,825
58,904
61,172
51,349
BOF Steel Production
Pig Iron Consumption
47,307
34,400
C
C
C
C
C
Scrap Steel Consumption
14,713
11,400
C
C
C
C
C
Flux Consumption
576
582
408
408
408
363
311
BOF Steel Production
43,973
42,705
25,888
25,788
27,704
26,591
21,383
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 2016 2017 2018 2019 2020
Pig Iron Production
Natural Gas Consumption
Fuel Oil Consumption
(thousand gallons)
Coke Oven Gas
Consumption
Blast Furnace Gas
Production
EAF Steel Production
Natural Gas Consumption
BOF Steel Production
Coke Oven Gas
Consumption
Other Activities
Industrial Processes and Product Use 4-89
56,273
163,397
22,033
1,439,380
15,905
3,851
59,844
16,170
16,557
1,299,980
19,985
524
38,396 38,142 40,204 37,934 32,496
6,124 4,352 3,365 2,321 1,988
12,404 12,459 13,337 12,926 11,073
811,005 808,499 871,860 836,033 716,182
3,915 8,105 8,556 9,115 7,808
367
374
405
389
333
-------
Coke Oven Gas
Consumption
224,883
97,132
62,036
62,164
67,008
64,377
55,148
Blast Furnace Gas
Consumption
1,414,778
1,295,520
807,264
804,816
867,838
832,119
712,829
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2020.
Uncertainty
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
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 2006 IPCC Guidelines. During EPA's discussion with
4-90 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 2006IPCC 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
2020 Emission Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Metallurgical Coke & Iron
and Steel Production
C02
35.4
31.4
44.2
-17%
+17%
Metallurgical Coke & Iron
ch4
-21%
+23%
and Steel Production
+ 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.
/erification
For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
Chapter 6 of the 2006 IPCC Guidelines, see the QA/QC and Verification Procedures section in the introduction of
the IPPU chapter.
Recalculations Discussion
Recalculations were performed for the year 2019 with updated values for coke production at coke plants, scrap
steel consumption for EAF steel production, scrap steel consumption for BOF steel production, and pellet
consumption in blast furnaces from EIA, USGS, and AISI. These updates resulted in emissions increases of 114
percent from metallurgical coke production (1.6 MMT CO2), less than 1 percent from iron production (87 kt CO2),
1.2 percent from pellet production (11 kt CO2), and less than 1 percent from steel production (42 kt CO2).
Planned Improvements
Significant activity data for 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 2020 data and other estimation
approaches if 2020 data is not available for the next Inventory report. EPA will update the calculations for the
2023 Inventory submission if new data becomes available.
Future improvements involve improving activity data and emission factor sources for CO2 and CFU emissions
estimations from pellet production. EPA will also evaluate and analyze data reported under EPA's GHGRP to
improve the emission estimates for this and other 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
Industrial Processes and Product Use 4-91
-------
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.61 This remains a 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 and is not expected until a future (i.e., 2024) Inventory submission.
EPA also received comments during the Expert Review cycle of a previous (i.e., 1990 through 2016) Inventory on
recommendations to improve the description of the iron and steel industry and emissive processes. EPA began
incorporating some of these recommendations into a previous Inventory (i.e., 1990 through 2016) and will require
some additional time to implement other substantive changes.
4.18 Ferroalloy Production (CRF Source
Category 2C2)
Carbon dioxide (CO2) and methane (CH4) are emitted from the production of several ferroalloys. Ferroalloys are
composites of iron (Fe) and other elements such as silicon (Si), manganese (Mn), and chromium (Cr). Emissions
from fuels consumed for energy purposes during the production of ferroalloys are accounted for in the Energy
chapter. Emissions from the production of two types of ferrosilicon (25 to 55 percent and 56 to 95 percent silicon),
silicon metal (96 to 99 percent silicon), and miscellaneous alloys (32 to 65 percent silicon) have been calculated.
Emissions from the production of ferrochromium and ferromanganese are not included because of the small
number of manufacturers of these materials in the United States. Government information disclosure rules
prevent the publication of production data for these production facilities. Additionally, production of
ferrochromium in the United States ceased in 2009 (USGS 2013).
Similar to emissions from the production of iron and steel, CO2 is emitted when metallurgical coke is oxidized
during a high-temperature reaction with iron and the selected alloying element. Due to the strong reducing
environment, CO is initially produced and eventually oxidized to CO2. A representative reaction equation for the
production of 50 percent ferrosilicon (FeSi) is given below:
Fe203 + 2Si02 + 7C —> 2FeSi + 7C0
While most of the carbon contained in the process materials is released to the atmosphere as CO2, a percentage is
also released as CFU and other volatiles. The amount of CH4 that is released is dependent on furnace efficiency,
operation technique, and control technology.
Ferroalloys are used to alter the material properties of the steel. Ferroalloys are produced in conjunction with the
iron and steel industry, often at co-located facilities, and production trends closely follow that of the iron and steel
industry. As of 2018,11 facilities in the United States produce ferroalloys (USGS 2021b). Emissions of CO2 from
ferroalloy production in 2020 were 1.4 MMT CO2 Eq. (1,377 kt CO2) (see Table 4-75 and Table 4-76), which is a 36
61 See http://www.ipcc-nggiD.iges.or.jp/public/tb/TFl Technical Bulletin l.pdf.
4-92 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
percent reduction since 1990. Emissions of Cm from ferroalloy production in 2020 were 0.01 MMT CO2 Eq. (0.4 kt
CH4), which is a 43 percent decrease since 1990. The decrease in emissions since 1990 can largely be attributed to
two facility shutdowns in 2018 and one facility shutdown in 2020. Additionally, the COVID-19 pandemic and lower
priced imported ferrosilicon had an impact on ferroalloy production in 2020 (USGS 2021a).
Table 4-75: CO2 and ChU Emissions from Ferroalloy Production (MMT CO2 Eq.)
Gas 1990 2005 2016 2017 2018 2019 2020
C02 2.2 1.4 1.8 2.0 2.1 1.6 1.4
CH4 + + + + + + +
Total Z2 1A 1.8 2.0 2.1 1.6 1.4
+ Does not exceed 0.05 MMT C02 Eq.
Table 4-76: CO2 and ChU Emissions from Ferroalloy Production (kt)
Gas 1990 2005 2016 2017 2018 2019 2020
C02 2,152 1,392 1,796 1,975 2,063 1,598 1,377
CH4 1 + 1 1 1 + + _
+ Does not exceed 0.5 kt
Methodology and Time-Series Consistency
Emissions of CO2 and CH4 from ferroalloy production were calculated62 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 CH4 emissions are as follows:
Equation 4-12: 2006 IPCC Guidelines Tier 1: CO2 Emissions for Ferroalloy Production
(Equation 4.15)
Eco2 = YSMPi x EFi)
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)
Ech, = Y^MPi X EFi)
i
where,
Ech4 = CH4 emissions, kg
MP, = Production of ferroalloy type /', metric tons
EFi = Generic emission factor for ferroalloy type /', kg Cl-U/metric ton specific ferroalloy product
62 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.
Industrial Processes and Product Use 4-93
-------
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 CFU 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 Cm/metric ton of alloy produced.
• Ferrosilicon, 56 to 95 percent Si: 4.0 metric tons CCh/metric ton alloy produced, 1.0 kg Cm/metric ton of
alloy produced.
• Silicon Metal: 5.0 metric tons CCh/metric ton metal produced, 1.2 kg Cm/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.
Ferroalloy production data for 1990 through 2020 (see Table 4-77) were obtained from the U.S. Geological Survey
(USGS) through the Minerals Yearbook: Silicon (USGS 1996 through 2013, 2021c) and the Mineral Industry Surveys:
Silicon (USGS 2014, 2015, 2016b, 2017, 2018b, 2019, 2020). 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/total ferroalloy production) were used with the total silicon materials production quantity to estimate
the production quantity by ferroalloy product type for 2011 through 2020 (USGS 2017, 2018b, 2019, 2020, 2021c).
Table 4-77: Production of Ferroalloys (Metric Tons)
Year Ferrosilicon Ferrosilicon Silicon Metal Misc. Alloys
25%-55% 56%-95% 32-65%
1990 321,385 109,566 145,744 72,442
2005 123,000 86,100 148,000 NA
2016 165,282 145,837 159,881 NA
2017 181,775 160,390 175,835 NA
2018 189,846 167,511 183,642 NA
2019 147,034 129,736 142,229 NA
2020 126,681 111,778 122,541 NA
NA (Not Available) for product type, aggregated along with ferrosilicon (25-55% Si)
4-94 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2020.
Uncertainty
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, 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.63 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
of reducing agent used in the process, rather than the amount of ferroalloys produced. These data, however, were
not available, and are also often considered confidential business information.
Emissions of CFU from ferroalloy production will vary depending on furnace specifics, such as type, operation
technique, and control technology. Higher heating temperatures and techniques such as sprinkle charging would
reduce CFU emissions; however, specific furnace information was not available or included in the CH4 emission
estimates.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-78. Ferroalloy
production CO2 emissions from 2020 were estimated to be between 1.2 and 1.6 MMT CO2 Eq. at the 95 percent
confidence level. This indicates a range of approximately 13 percent below and 13 percent above the emission
estimate of 1.4 MMT CO2 Eq. Ferroalloy production CH4 emissions were estimated to be between a range of
approximately 12 percent below and 13 percent above the emission estimate of 0.01 MMT CO2 Eq.
Table 4-78: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from
Ferroalloy Production (MMT CO2 Eq. and Percent)
Source
Gas
2020 Emission Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Ferroalloy Production
C02
1.4
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.
63 Emissions and sinks of biogenic carbon are accounted for in the Land Use, Land-Use Change, and Forestry chapter.
Industrial Processes and Product Use 4-95
-------
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.
Recalculations Discussion
No recalculations were performed for the 1990 through 2019 portion of the time series.
Planned Improvements
Pending available resources and prioritization of improvements for more significant sources, EPA will continue to
evaluate and analyze data reported under EPA's GHGRP that would be useful to improve the emission estimates
and category-specific QC procedures for the Ferroalloy Production source category. Given the small number of
facilities and reporting thresholds, particular attention will be made to ensure completeness and 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.64 This is a long-term planned improvement, and EPA is still assessing the possibility of incorporating this
improvement into the Inventory. This improvement has not been included in the current Inventory report.
4.19 Aluminum Production (CRF Source
Category 2C3)
Aluminum is a lightweight, malleable, and corrosion-resistant metal that is used in many manufactured products,
including aircraft, automobiles, bicycles, and kitchen utensils. As of recent reporting, the United States was the
ninth65 largest producer of primary aluminum, with approximately 1.5 percent of the world total production (USGS
2020). The United States was also a major importer of primary aluminum. The production of primary aluminum —
in addition to consuming large quantities of electricity—results in process-related emissions of carbon dioxide
(CO2) and two perfluorocarbons (PFCs): perfluoromethane (CF4) and perfluoroethane (C2F6).
Carbon dioxide is emitted during the aluminum smelting process when alumina (aluminum oxide, AI2O3) is reduced
to aluminum using the Hall-Heroult reduction process. The reduction of the alumina occurs through electrolysis in
a molten bath of natural or synthetic cryolite (NasAIFs). The reduction cells contain a carbon (C) lining that serves
as the cathode. Carbon is also contained in the anode, which can be a C mass of paste, coke briquettes, or
prebaked C blocks from petroleum coke. During reduction, most of this C is oxidized and released to the
atmosphere as CO2.
Process emissions of CO2 from aluminum production were estimated to be 1.7 MMT CO2 Eq. (1,748 kt) in 2020 (see
Table 4-79). The C anodes consumed during aluminum production consist of petroleum coke and, to a minor
extent, coal tar pitch. The petroleum coke portion of the total CO2 process emissions from aluminum production is
64 See http://www.ipcc-nggiD.iges.or.jp/public/tb/TFl Technical Bulletin l.pdf.
65 Based on the U.S. USGS (2020) 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/mcs2021/mcs2021-aluminum.pdf
4-96 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
considered to be a non-energy use of petroleum coke and is accounted for here and not under the CO2 from Fossil
Fuel Combustion source category of the Energy sector. Similarly, the coal tar pitch portion of these CO2 process
emissions is accounted for here.
Table 4-79: CO2 Emissions from Aluminum Production (MMT CO2 Eq. and kt)
Year
MMT CO? Eq.
kt
1990
6.8
6,831
2005
4.1
4,142
2016
1.3
1,334
2017
1.2
1,205
2018
1.5
1,451
2019
1.9
1,880
2020
1.7
1,748
In addition to CO2 emissions, the aluminum production industry is also a source of PFC emissions. During the
smelting process, when the alumina ore content of the electrolytic bath falls below critical levels required for
electrolysis, rapid voltage increases occur, which are termed High Voltage Anode Effects (HVAEs) HVAEs cause C
from the anode and fluorine from the dissociated molten cryolite bath to combine, thereby producing fugitive
emissions of CF4 and C2F6. In general, the magnitude of emissions for a given smelter and level of production
depends on the frequency and duration of these anode effects. As the frequency and duration of the anode effects
increase, emissions increase. Another type of anode effect, Low Voltage Anode Effects (LVAEs), became a concern
in the early 2010s as the aluminum industry increasingly began to use cell technologies with higher amperage and
additional anodes (IPCC 2019). LVAEs emit CF4 and are included in PFC emission totals from 2006 forward.
Since 1990, emissions of CF4 and C2F6 have both declined by 92 percent to 1.4 MMT CO2 Eq. of CF4 (0.2 kt) and 0.3
MMT CO2 Eq. of C2F6 (0.02 kt) in 2020, respectively, as shown in Table 4-80 and Table 4-81. This decline is due both
to reductions in domestic aluminum production and to actions taken by aluminum smelting companies to reduce
the frequency and duration of anode effects. These actions include technology and operational changes such as
employee training, use of computer monitoring, and changes in alumina feeding techniques. Since 1990, aluminum
production has declined by 75 percent, while the combined CF4 and C2F6 emission rate (per metric ton of
aluminum produced) has been reduced by 69 percent. PFC emissions decreased by approximately 5 percent
between 2019 and 2020 due to decreases in aluminum production in 2020 for multiple factors, including
shutdowns and economic (supply chain) disruptions from the COVID-19 pandemic.
Table 4-80: PFC Emissions from Aluminum Production (MMT CO2 Eq.)
Year
cf4
c2f6
Total
1990
17.9
3.5
21.5
2005
2.9
0.6
3.4
2016
1.0
0.4
1.4
2017
0.7
0.4
1.1
2018
1.2
0.4
1.6
2019
1.4
0.4
1.8
2020
1.4
0.3
1.7
Note: Totals may not sum due to
independent rounding.
Industrial Processes and Product Use 4-97
-------
Table 4-81: PFC Emissions from Aluminum Production (kt)
Year CF4 C2F6
1990 2.4 0.29
2005
0.4
0.05
2016
2017
2018
2019
2020
0.1
0.1
0.2
0.2
0.2
0.04
0.03
0.03
0.03
0.02
In 2020, U.S. primary aluminum production totaled approximately 1.012 million metric tons, a 4 percent decrease
from 2019 production levels (USAA 2020). In 2020, three companies managed production at seven operational
primary aluminum smelters in six states. Two smelters operated at full capacity during 2020, while four smelters
operated at reduced capacity (USGS 2021). One smelter operated at reduced capacity until it was idled in July.
Domestic smelters were operating at about 49 percent of capacity of 1.79 million tons per year at year end 2020
(USGS 2021).
The COVID-19 pandemic impacted domestic aluminum production and imports indirectly and directly, and neither
USGS nor USAA sources have stated projections for the production year 2021.
Process CO2 and PFC (i.e., CF4 and C2F6) emission estimates from primary aluminum production for 2010 through
2020 are available from EPA's GHGRP Subpart F (Aluminum Production) (EPA 2021). Under EPA's GHGRP, facilities
began reporting primary aluminum production process emissions (for 2010) in 2011; as a result, GHGRP data (for
2010 through 2020) are available to be incorporated into the Inventory. EPA's GHGRP mandates that all facilities
that contain an aluminum production process must report: CF4 and C2F6 emissions from anode effects in all
prebake and S0derberg electrolysis cells, CO2 emissions from anode consumption during electrolysis in all prebake
and S0derberg cells, and all CO2 emissions from onsite anode baking. To estimate the process emissions, EPA's
GHGRP uses the process-specific equations detailed in Subpart F (aluminum production).66 These equations are
based on the Tier 2/Tier 3 IPCC (2006) methods for primary aluminum production, and Tier 1 methods when
estimating missing data elements. It should be noted that the same methods (i.e., 2006 IPCC Guidelines) were used
for estimating the emissions prior to the availability of the reported GHGRP data in the Inventory. Prior to 2010,
aluminum production data were provided through EPA's Voluntary Aluminum Industrial Partnership (VAIP).
As previously noted, the use of petroleum coke for aluminum 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.
Process CO2 Emissions from Anode Consumption and Anode Baking
Carbon dioxide emission estimates for the years prior to the introduction of EPA's GHGRP in 2010 were estimated
using 2006 IPCC Guidelines methods, but individual facility reported data were combined with process-specific
66 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.
4-98 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
Methodology and Time-Series Consistency
-------
emissions modeling. These estimates were based on information previously gathered from EPA's Voluntary
Aluminum Industrial Partnership (VAIP) program, U.S. Geological Survey (USGS) Mineral Commodity reviews, and
The Aluminum Association (USAA) statistics, among other sources. Since pre- and post-GHGRP estimates use the
same methodology, emission estimates are comparable across the time series.
Most of the CO2 emissions released during aluminum production occur during the electrolysis reaction of the C
anode, as described by the following reaction:
2AI2O3 + 3C -> 4A1 + 3C02
For prebake smelter technologies, CO2 is also emitted during the anode baking process. These emissions can
account for approximately 10 percent of total process CO2 emissions from prebake smelters.
Depending on the availability of smelter-specific data, the CO2 emitted from electrolysis at each smelter was
estimated from: (1) the smelter's annual anode consumption, (2) the smelter's annual aluminum production and
rate of anode consumption (per ton of aluminum produced) for previous and/or following years, or (3) the
smelter's annual aluminum production and IPCC default CO2 emission factors. The first approach tracks the
consumption and carbon content of the anode, assuming that all C in the anode is converted to CO2. Sulfur, ash,
and other impurities in the anode are subtracted from the anode consumption to arrive at a C consumption figure.
This approach corresponds to either the IPCC Tier 2 or Tier 3 method, depending on whether smelter-specific data
on anode impurities are used. The second approach interpolates smelter-specific anode consumption rates to
estimate emissions during years for which anode consumption data are not available. This approach avoids
substantial errors and discontinuities that could be introduced by reverting to Tier 1 methods for those years. The
last approach corresponds to the IPCC Tier 1 method (IPCC 2006) and is used in the absence of present or historic
anode consumption data.
The equations used to estimate CO2 emissions in the Tier 2 and 3 methods vary depending on smelter type (IPCC
2006). For Prebake cells, the process formula accounts for various parameters, including net anode consumption,
and the sulfur, ash, and impurity content of the baked anode. For anode baking emissions, the formula accounts
for packing coke consumption, the sulfur and ash content of the packing coke, as well as the pitch content and
weight of baked anodes produced. For S0derberg cells, the process formula accounts for the weight of paste
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 2020 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-
Industrial Processes and Product Use 4-99
-------
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
S
AE
F
D
CF4 or C2F6, kg/MT aluminum
Slope coefficient, PFC/AE
Anode effect, minutes/cell-day
Anode effect frequency per cell-day
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.
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 or USAA, 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 MMTCQ2 Eq.
1990 21.5
2005 3.4
2016 1.4
2017 1.0
2018 1.6
2019 1.7
2020 1.6
4-100 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Low Voltage Anode Effects
LVAE emissions of CF4 were estimated for 2006 through 2020 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.67 The following equation was used to
estimate LVAE PFC emissions:
Equation 4-14: CF4 Emissions Resulting from Low Voltage Anode Effects
LVAE ECF4 = LVAEEFCF4 X MP
where,
LVAE Ecf4 = LVAE emissions of CF4 from aluminum production, kg CF4
LVAE EFcf4 = LVAE emission factor for CF4 (default by cell technology type)
MP = metal production by cell technology type, tons Al.
Once LVAE emissions were estimated, they were then combined with HVAE emissions estimates to calculate total
PFC emissions from aluminum production.
Table 4-83: Summary of LVAE Emissions
Year
MMT CO? Eq.
2006
0.1
2016
0.1
2017
0.1
2018
0.1
2019
0.1
2020
0.1
Production Data
Between 1990 and 2009, production data were provided under the VAIP by 21 of the 23 U.S. smelters that
operated during at least part of that period. For the non-reporting smelters, production was estimated based on
the difference between reporting smelters and national aluminum production levels as reported to USGS, with
allocation to specific smelters based on reported production capacities (USGS 1990 through 2009).
National primary aluminum production data for 2020, 2019, and 2018 were obtained via the 2020 USGS Mineral
Industry Surveys, and the 2021 USGS Mineral Commodity Summaries. For 1990 through 2001, and 2006 (see Table
4-84) data were obtained from USGS Mineral Industry Surveys: Aluminum Annual Report (USGS 1995,1998, 2000,
2001, 2002, 2007). For 2002 through 2005, and 2007 through 2017, national aluminum production data were
obtained from the USAA's Primary Aluminum Statistics (USAA 2004 through 2006, 2008 through 2017).
67 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.
Industrial Processes and Product Use 4-101
-------
Table 4-84: Production of Primary Aluminum (kt)
Year kt
1990 4,048
2005 2,478
2016 818
2017 741
2018 891
2019 1,093
2020 1,012
Methodological approaches were applied to the entire time-series to ensure time-series consistency from 1990
through 2020.
Uncertainty
Uncertainty was estimated for the CO2, CF4, and C2F6 emission values reported by each individual facility to EPA's
GHGRP, taking into consideration the uncertainties associated with aluminum production, anode effect minutes,
and slope factors. The uncertainty bounds used for these parameters were established based on information
collected under the VAIP and held constant through 2020. Uncertainty surrounding the reported CO2, CF4, and C2F6
emission values were determined to have a normal distribution with uncertainty ranges of approximately 6
percent below to 6 percent above, 16 percent below to 16 percent above, and 20 percent below to 20 percent
above their 2020 emission estimates, respectively.
For LVAE, since emission values were not reported through EPA's GHGRP but estimated instead through a Tier 1
methodology, the uncertainty analysis examined uncertainty associated with primary capacity data as well as
technology-specific emission factors. Uncertainty for each facility's primary capacity, reported in the USGS
Yearbook, was estimated to have a Pert Beta distribution with an uncertainty range of 10 percent below to 7
percent above the capacity estimates based on the uncertainty of reported capacity data, the number of years
since the facility reported new capacity data, and uncertainty in capacity utilization. Uncertainty was applied to
LVAE emission factors according to technology using the uncertainty ranges provided in the 2019 Refinement to
the 2006IPCC Guidelines. An uncertainty range for Horizontal Stud S0derberg (HSS) technology was not provided
in the 2019 Refinement to the 2006 IPCC Guidelines due to insufficient data, so a normal distribution and
uncertainty range of ±99 percent was applied for that technology based on expert judgment. A Monte Carlo
analysis was applied to estimate the overall uncertainty of the CO2, CF4, and C2F6 emission estimates for the U.S.
aluminum industry as a whole, and the results are provided below.
The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 4-85. Aluminum
production-related CO2 emissions were estimated to be between 1.71 and 1.79 MMT CO2 Eq. at the 95 percent
confidence level. This indicates a range of approximately 2 percent below to 2 percent above the emission
estimate of 1.75 MMT CO2 Eq. Also, production-related CF4 emissions were estimated to be between 1.29 and 1.50
MMT CO2 Eq. at the 95 percent confidence level. This indicates a range of approximately 7 percent below to 8
percent above the emission estimate of 1.39 MMT CO2 Eq. Aluminum production-related C2F6 emissions were
estimated to be between 0.25 and 0.32 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 0.29 MMT CO2 Eq. Finally,
Aluminum production-related aggregated PFCs emissions were estimated to be between 1.57 and 1.79 MMT CO2
Eq. at the 95 percent confidence level. This indicates a range of approximately 6 percent below to 7 percent above
the emission estimate of 1.68 MMT CO2 Eq.
4-102 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 4-85: Approach 2 Quantitative Uncertainty Estimates for CO2 and PFC Emissions from
Aluminum Production (MMT CO2 Eq. and Percent)
Source
Gas
2020 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
1.75
1.71
1.79
-2%
2%
Aluminum Production
cf4
1.39
1.29
1.50
-7%
8%
Aluminum Production
c2f6
0.29
0.25
0.32
-11%
11%
Aluminum Production
PFCs
1.68
1.57
1.79
-6%
7%
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).68 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
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 USAA with the total U.S. capacity reported for this
specific year, based on the USGS yearbook. 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 (<6 percent) than the production capacity reported that year. In practice, this is most likely explained by the
degree of uncertainty in the USAA annual production reporting, and the differences in process efficiencies,
measurements and methods used by each facility to obtain the CO2, which cannot be completely homogenized
throughout the reporting facilities.
Following Expert review comments, the total primary aluminum production estimates were updated to reflect data
reported to the USGS (as detailed in Production Data section above) for the year 2018, 2019 and 2020, whereas
previously, production estimates from the U.S. Aluminum Association were used for these specific 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 these recent years, all companies, who were sent the
surveys, responded, thus making USGS data the most accurate available. These data source modifications did not
lead to differences in the greenhouse gas emissions calculations for these specific years.
68 GHGRP Report Verification Factsheet. See https://www.epa.gov/sites/production/files/2Q15-
07/documents/ghgrp verification factsheet.pdf.
Industrial Processes and Product Use 4-103
-------
Planned Improvements
EPA will further investigate the sources of historical total primary aluminum production estimates for the earlier
years in the timeseries and potentially update historical estimates to aim for increased consistency throughout the
timeseries. As part of this planned improvement, EPA will review whether historical estimates are broken down
into smelter specific production estimates, which are the basis for calculating smelter PFC (for non-partners) and
CO2 emissions (for all facilities) for the 1990 through 2009 time series (years preceding GHGRP reporting).
4.20 Magnesium Production and Processing
(CRF Source Category 2C4)
The magnesium metal production and casting industry uses sulfur hexafluoride (SFs) as a cover gas to prevent the
rapid oxidation of molten magnesium in the presence of air. Sulfur hexafluoride has been used in this application
around the world for more than thirty years. A dilute gaseous mixture of SF6 with dry air and/or carbon dioxide
(CO2) is blown over molten magnesium metal to induce and stabilize the formation of a protective crust. A small
portion of the SF6 reacts with the magnesium to form a thin molecular film of mostly magnesium oxide and
magnesium fluoride. The amount of SF6 reacting in magnesium production and processing is considered to be
negligible and thus all SF6 used is assumed to be emitted into the atmosphere. Alternative cover gases, such as
AM-cover™ (containing HFC-134a), Novec™ 612 (FK-5-1-12) and dilute sulfur dioxide (SO2) systems can and are
being used by some facilities in the United States. However, many facilities in the United States are still using
traditional SF6 cover gas systems. Carbon dioxide is also released during primary magnesium production if
carbonate based raw materials, such as dolomite, are used. During the processing of these raw materials to
produce magnesium, calcination occurs which results in a release of CO2 emissions.
The magnesium industry emitted 0.9 MMT CO2 Eq. (0.04 kt) of SF6, 0.1 MMT CO2 Eq. (0.04 kt) of HFC-134a, and
0.001 MMT CO2 Eq. (0.9 kt) of CO2 in 2020. This represents a decrease of approximately 2 percent from total 2019
emissions (see Table 4-86 and Table 4-87) and a decrease in SF6 emissions by 1 percent. In 2020, total HFC-134a
emissions decreased from 0.066 MMT CO2 Eq. to 0.058 MMT CO2 Eq., or a 13 percent decrease as compared to
2019 emissions. FK 5-1-12 emissions in 2020 were consistent with 2019. The emissions of the carrier gas, CO2,
decreased from 1.40 kt in 2019 to 0.94 kt in 2020, or 33 percent. These decreases are likely attributed to
decreasing production levels between 2019 and 2020. For the first time this year CO2 emissions from the use of
dolomite in primary production are included under Magnesium Production and Processing. Previously, these
emissions had been included under Other Process Uses of Carbonates. This inclusion resulted in a significant
increase in CO2 emissions from 1990 through 2001, the time period during which it is known that dolomite was
used in primary production, as compared to previously compiled Inventories. Additional information related to this
update is provided below.
Table 4-86: SF6, HFC-134a, FK 5-1-12 and CO2 Emissions from Magnesium Production and
Processing (MMT CO2 Eq.)
Year
1990
2005
2016
2017
2018
2019
2020
sf6
5.2
2.7
1.1
1.0
1.0
0.9
0.9
HFC-134a
0.0
0.0
0.1
0.1
0.1
0.1
0.1
C02
0.1
+
+
+
+
+
+
FK 5-1-12"
0.0
0.0
+
+
+
+
+
Total
5.3
2.7
1.2
1.1
1.1
0.9
0.9
+ Does not exceed 0.05 MMT C02 Eq.
a Emissions of FK 5-1-12 are not included in totals.
4-104 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 4-87: SF6, HFC-134a, FK 5-1-12 and CO2 Emissions from Magnesium Production and
Processing (kt)
Year 1990 2005 2016 2017 2018 2019 2020
SFe 0.2 0.1 + + + + +
HFC-134a 0.0 0.0 0.1 0.1 0.1 + +
C02 128.5 3.3 2.8 3.3 1.6 1.4 0.9
FK 5-1-12° 00 00 + + + +
+ Does not exceed 0.05 kt
a Emissions of FK 5-1-12 are not included in totals.
Methodology and Time-Series Consistency
Emission estimates for the magnesium industry incorporate information provided by industry participants in EPA's
SFe 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 2020 (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
SFe 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
Industrial Processes and Product Use 4-105
-------
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.69 For 1990
through 1998, production was estimated to be equal to the production capacity of the facility.
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
69 See https://www.ipcc-nggip.iges.or.jP/public/2CX36gl/pdf/3 Volume3/V3 4 Ch4 Metal lndustry.pdf.
4-106 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
end of 2001, the sole magnesium production plant operating in the United States that produced magnesium metal
using a dolomitic process that resulted in the release of CO2 emissions ceased its operations (USGS 1995b through
2020).
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.
2011 through 2020
For 2011 through 2020, for the primary and secondary producers, GHGRP-reported cover and carrier gases
emissions data were used. For sand and die casting, some emissions data was obtained through EPA's GHGRP.
Additionally, in 2018 a new GHGRP reporter began reporting permanent mold emissions. The balance of the
emissions for this industry segment was estimated based on previous Partner reporting (i.e., for Partners that did
not report emissions through EPA's GHGRP) or were estimated by multiplying emission factors by the amount of
metal produced or consumed. Partners who did not report through EPA's GHGRP were assumed to have continued
to emit SFs at the last reported level, which was from 2010 in most cases, unless publicly available sources
indicated that these facilities have closed or otherwise eliminated SF6 emissions from magnesium production (ARB
2015). Many Partners that did report through the GHGRP showed increases in SF6 emissions driven by increased
production related to a continued economic recovery after the 2008 recession. One Partner in particular reported
an anonymously large increase in SF6 emissions from 2010 to 2011, further driving increases in emissions between
the two time periods of inventory estimates. All Partners were assumed to have continued to consume magnesium
at the last reported level. Where the total metal consumption estimated for the Partners fell below the U.S. total
reported by USGS, the difference was multiplied by the emission factors discussed in the section above, i.e., non-
partner emission factors. For the other types of production and processing (i.e., permanent mold, wrought, and
anode casting), emissions were estimated by multiplying the industry emission factors with the metal production
Industrial Processes and Product Use 4-107
-------
or consumption statistics obtained from USGS (USGS 2020). USGS data for 2020 were not yet available at the time
of the analysis, so the 2019 values were held constant through 2020 as an estimate.
Emissions of carrier gases for permanent mold, wrought, and anode processes were estimated using an approach
consistent with the 1999 through 2010 time series.
Methodological approaches were applied to the entire time series to ensure time-series consistency from 1990
through 2020. 2006IPCC Guidance methodologies were used throughout the timeseries, mainly either a Tier 2 or
Tier 3 approach depending on available data. Additionally, in this Inventory, steps were taken to ensure time-series
consistency for CO2 emissions. These steps are further highlights in the recalculations discussion below.
Uncertainty
Uncertainty surrounding the total estimated emissions in 2020 is attributed to the uncertainties around SF6, HFC-
134a, and CO2 emission estimates. To estimate the uncertainty surrounding the estimated 2020 SF6 emissions from
magnesium production and processing, the uncertainties associated with three variables were estimated: (1)
emissions reported by magnesium producers and processors for 2020 through EPA's GHGRP, (2) emissions
estimated for magnesium producers and processors that reported via the Partnership in prior years but did not
report 2020 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 2006 IPCC 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 2019 and 2020.
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 0.84 and 1.00 MMT CO2 Eq.
at the 95 percent confidence level. This indicates a range of approximately 9 percent below to 9 percent above the
2020 emission estimate of 0.92 MMT CO2 Eq. The uncertainty estimates for 2020 are slightly higher to the
uncertainty reported for 2019 in the previous Inventory. This increase in uncertainty is attributed to the increased
uncertainty around the emissions data that was estimated for reporters that did not report in 2020 or, in some
cases, dating back to 2010. The longer the time period for which EPA needs to estimate emissions the larger the
associated uncertainty with those emission estimates will be.
4-108 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
2020 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
0.92
0.84 1.00
-9% 9%
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).70 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
Sand casting and permanent mold casting volumes were updated based on the release of an updated USGS
Minerals Yearbook (USGS 2020). Primary production SF6 emissions were set equal to zero from 2016 through 2020
because of a confirmation that the facility transitioned completed to HFC-134a in 2016. Additionally, one facility's
reported GHGRP emissions were revised in 2017 due to additional information provided on emissions from HFC-
134a and CO2. Lastly, a correction was made for a die casting facility for 2016. This facility did not report in 2016
and previously 2016 SF6 emissions were held constant at 2015 levels. This approach was updated to estimate 2016
SFe emissions through interpolation between 2015 and 2017.
Three changes were made in this Inventory in relation to CO2 emissions. First, CO2 emissions were added for
permanent mold, wrought, and anode production throughout the time series. Second, it was discovered that CO2
emissions from sand casting were not included from 1990 through 2010. These emissions were added in this
Inventory. Lastly, CO2 emissions from the use of dolomite in primary production from 1990 to 2001 are now
reported under Magnesium Production and Processing instead of elsewhere in the inventory, which is consistent
with the 2006 IPCC Guidelines. The methods used to implement these changes are described above.
Planned Improvements
Cover gas research conducted over the last decade has found that SF6 used for magnesium melt protection can
have degradation rates on the order of 20 percent in die casting applications (Bartos et al. 2007). Current emission
estimates assume (per the 2006 IPCC Guidelines) that all SF6 utilized is emitted to the atmosphere. Additional
70 GHGRP Report Verification Factsheet. See https://www.epa.gov/sites/production/files/2Q15-
07/documents/ghgrp verification factsheet.pdf.
Industrial Processes and Product Use 4-109
-------
research may lead to a revision of the 2006IPCC Guidelines to reflect this phenomenon and until such time,
developments in this sector will be monitored for possible application to the Inventory methodology.
Additional emissions are generated as byproducts from the use of alternate cover gases, which are not currently
accounted for. Research on this topic is developing, and as reliable emission factors become available, these
emissions will be incorporated into the Inventory.
4.21 Lead Production (CRF Source Category
2C5)
In 2020, lead was produced in the United States only using secondary production processes. Until 2014, lead
production in the United States involved both primary and secondary processes—both of which emit carbon
dioxide (CO2) (Sjardin 2003). Emissions from fuels consumed for energy purposes during the production of lead are
accounted for in the Energy chapter.
Primary production of lead through the direct smelting of lead concentrate produces CO2 emissions as the lead
concentrates are reduced in a furnace using metallurgical coke (Sjardin 2003). Primary lead production, in the form
of direct smelting, previously occurred at a single smelter in Missouri. This primary lead smelter was closed at the
end of 2013. In 2014, the smelter processed a small amount of residual lead during demolition of the site (USGS
2015).
Similar to primary lead production, CO2 emissions from secondary lead production result when a reducing agent,
usually metallurgical coke, is added to the smelter to aid in the reduction process. Carbon dioxide emissions from
secondary production also occur through the treatment of secondary raw materials (Sjardin 2003). Secondary
production primarily involves the recycling of lead acid batteries and post-consumer scrap at secondary smelters.
Secondary lead production has increased in the United States over the past decade, while primary lead production
has decreased to production levels of zero. In 2020, secondary lead production accounted for 100 percent of total
lead production. The lead-acid battery industry accounted for about 92 percent of the reported U.S. lead
consumption in 2020 (USGS 2021).
In 2020, U.S. primary lead production remained at production levels of zero, and secondary lead production in the
United States decreased by approximately 6 percent compared to 2019, due to the COVID-19 pandemic, the
resulting quarantine-related restrictions, and a decrease in demand for lead (USGS 2021). Secondary lead
production in 2020 is 19 percent higher than in 1990 (USGS 1994 and 2021). The United States has become more
reliant on imported refined lead, owing to the closure of the last primary lead smelter in 2013. Exports of spent
starting-lighting-ignition (SLI) batteries have been generally decreasing since 2014and were 12 percent lower in the
first 9 months of 2020 compared to the same time period in 2014 (USGS 2015 through 2020). In the first 9 months
of 2020,19.7 million spent SLI lead-acid batteries were exported, slightly less than that in the same time period in
2019 (USGS 2021).
In 2020, U.S. lead production totaled 1,100,000 metric tons (USGS 2021). The resulting emissions of CChfrom 2020
lead production were estimated to be 0.5 MMT CO2 Eq. (495 kt) (see Table 4-90).
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 2020 (USGS 2021).
4-110 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 4-90: CO2 Emissions from Lead Production (MMT CO2 Eq. and kt)
Year
MMTCOz Eq.
kt
1990
0.5
516
2005
0.6
553
2016
2017
2018
2019
2020
0.5
0.5
0.5
0.5
0.5
500
513
513
527
495
After a steady increase in total emissions from 1995 to 2000, total emissions decreased between 2000 and 2013 (8
percent decline across the time period), exhibited a single year decrease of 16 percent between 2013 and 2014,
gradually increased between 2014 and 2019, and are currently 4 percent lower than 1990 levels.
Methodology and Time-Series Consistency
The methods used to estimate emissions for lead production71 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)
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 (3.1 Fossil Fuel Combustion (CRF Source Category 1A)) and Annex 2.1, Methodology for Estimating
Emissions of CO2 from Fossil Fuel Combustion.
71 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.
C 02 Emissions = (DS x EFDS) + (5 x EFS)
where,
DS
S
EFds
EFs
Lead produced by direct smelting, metric ton
Lead produced from secondary materials
Emission factor for direct smelting, metric tons CCh/metric ton lead product
Emission factor for secondary materials, metric tons CCh/metric ton lead product
Industrial Processes and Product Use 4-111
-------
The 1990 through 2020 activity data for primary and secondary lead production (see Table 4-91) were obtained
from the U.S. Geological Survey (USGS 1995 through 2021).
Table 4-91: Lead Production (Metric Tons)
Year
Primary
Secondary
1990
404,000
922,000
2005
143,000
1,150,000
2016
0
1,110,000
2017
0
1,140,000
2018
0
1,140,000
2019
0
1,170,000
2020
0
1,100,000
Methodological approaches discussed below were applied to applicable years to ensure time-series consistency in
emissions from 1990 through 2020.
Uncertainty
Uncertainty associated with lead production relates to the emission factors and activity data used. The direct
smelting emission factor used in primary production is taken from Sjardin (2003) who averaged the values
provided by three other studies (Dutrizac et al. 2000; Morris et al. 1983; Ullman 1997). For secondary production,
Sjardin (2003) added a CO2 emission factor associated with battery treatment. The applicability of these emission
factors to plants in the United States is uncertain. There is also a smaller level of uncertainty associated with the
accuracy of primary and secondary production data provided by the USGS which is collected via voluntary surveys;
the uncertainty of the activity data is a function of the reliability of reported plant-level production data and the
completeness of the survey response.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-92. Lead production CO2
emissions in 2020 were estimated to be between 0.4 and 0.6 MMT CO2 Eq. at the 95 percent confidence level. This
indicates a range of approximately 15 percent below and 16 percent above the emission estimate of 0.5 MMT CO2
Eq.
Table 4-92: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Lead
Production (MMT CO2 Eq. and Percent)
Source Gas
2020 Emission Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower Upper
Lower Upper
Bound Bound
Bound Bound
Lead Production C02
0.5
0.4 0.6
-15% +16%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
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.
4-112 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Initial review of activity data show that EPA's GHGRP Subpart R lead production data and resulting emissions are
fairly consistent with those reported by USGS. EPA is still reviewing available GHGRP data, reviewing QC analysis to
understand differences in data reporting (i.e., threshold implications), and assessing the possibility of including this
planned improvement in future Inventory reports (see Planned Improvements section below). Currently, GHGRP
data are used for QA purposes only.
Recalculations Discussion
Emissions for 2019 were revised from 0.5 MMT CO2 Eq. (540 kt) to 0.5 MMT CO2 Eq. (527 kt) based on revised
USGS data for secondary lead production.
Planned Improvements
Pending resources and prioritization of improvements for more significant sources, EPA will continue to evaluate
and analyze data reported under EPA's GHGRP that would be useful to improve the emission estimates and
category-specific QC for the Lead Production source category, in particular considering completeness of reported
lead production given the reporting threshold. Particular attention will be made to ensuring 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.72
4.22 Zinc Production (CRF Source Category
2C6)
Zinc production in the United States consists of both primary and secondary processes. Of the primary and
secondary processes currently used in the United States, only the electrothermic and Waelz kiln secondary
processes result in non-energy carbon dioxide (CO2) emissions (Viklund-White 2000). Emissions from fuels
consumed for energy purposes during the production of zinc are accounted for in the Energy chapter.
The majority of zinc produced in the United States is used for galvanizing. Galvanizing is a process where zinc
coating is applied to steel in order to prevent corrosion. Zinc is used extensively for galvanizing operations in the
automotive and construction industry. Zinc is also used in the production of zinc alloys and brass and bronze alloys
(e.g., brass mills, copper foundries, and copper ingot manufacturing). Zinc compounds and dust are also used, to a
lesser extent, by the agriculture, chemicals, paint, and rubber industries.
Production of zinc can be conducted with a range of pyrometallurgical (e.g., electrothermic furnace, Waelz kiln,
flame reactor, batch retorts, Pinto process, and PIZO process) and hydrometallurgical (e.g., hydrometallurgical
recovery, solvent recovery, solvent extraction-electrowinning, and electrolytic) processes. Hydrometallurgical
production processes are assumed to be non-emissive since no carbon is used in these processes (Sjardin 2003).
Primary production in the United States is conducted through the electrolytic process, while secondary techniques
include the electrothermic and Waelz kiln processes, as well as a range of other processes. Worldwide primary zinc
production also employs a pyrometallurgical process using an Imperial Smelting Furnace; however, this process is
not used in the United States (Sjardin 2003).
72 See http://www.ipcc-nggiD.iges.or.jp/public/tb/TFl Technical Bulletin l.pdf.
Industrial Processes and Product Use 4-113
-------
In the electrothermic process, roasted zinc concentrate and secondary zinc products enter a sinter feed where
they are burned to remove impurities before entering an electric retort furnace. Metallurgical coke is added to the
electric retort furnace as a carbon-containing reductant. This concentration step, using metallurgical coke and high
temperatures, reduces the zinc oxides and produces vaporized zinc, which is then captured in a vacuum
condenser. This reduction process also generates non-energy CO2 emissions.
ZnO + C -» Zn(gas) + C02 (Reaction 1)
ZnO + CO -» Zn(gas) + C02 (Reaction 2)
In the Waelz kiln process, electric arc furnace (EAF) dust, which is captured during the recycling of galvanized steel,
enters a kiln along with a reducing agent (typically carbon-containing metallurgical coke). When kiln temperatures
reach approximately 1,100 to 1,200 degrees Celsius, zinc fumes are produced, which are combusted with air
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 2020, the only companies in the United States that used emissive technology to produce secondary zinc
products were American Zinc Recycling (AZR) (formerly "Horsehead Corporation") and Steel Dust Recycling (SDR).
PIZO Operating Company, LLC (PIZO) operated a secondary zinc production facility that processed EAF dust in
Blytheville, AR from 2009 to 2012.
For AZR, EAF dust is recycled in Waelz kilns at their Calumet, IL; Palmerton, PA; Rockwood, TN; and Barnwell, SC
facilities. The 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), most of which was transported to their Monaca, PA facility where the products were
smelted into refined zinc using electrothermic technology. In April 2014, AZR permanently closed their Monaca
smelter. This was replaced by their 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. The current capacity of the
new facility is 155,000 short tons. 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 at the new Mooresboro facility. The new facility is
reported to have significantly lower greenhouse gas and other air emissions than the Monaca smelter (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 2020 were estimated to be 1.0 MMT CO2 Eq. (1,008 kt CO2) (see Table
4-93). All 2020 CO2 emissions resulted from secondary zinc production processes. Emissions from zinc production
4-114 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
in the United States have increased overall since 1990 due to a gradual shift from non-emissive primary production
to emissive secondary production. In 2020, emissions were estimated to be 60 percent higher than they were in
1990. Emissions decreased 2 percent from 2019 levels. Due largely to the COVID-19 pandemic, a decrease in both
the demand for zinc and zinc prices led to a decrease in global zinc mine production in most producing countries,
including the United States. 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 pandemic. (USGS 2021).
Table 4-93: CO2 Emissions from Zinc Production (MMT CO2 Eq. and kt)
Year MMTCQ2Eq. kt
1990 06 632
2005 1.0 1,030
2016 0.8 838
2017 0.9 900
2018 1.0 999
2019 1.0 1,026
2020 1.0 1,008
In 2020, United States primary and secondary refined zinc production were estimated to total 150,000 metric tons
(USGS 2021) (see Table 4-94). Domestic zinc mine production decreased in 2020, owing partially to a decrease in
production at the Red Dog Mine in Alaska and the closure of the Pend Oreille Mine in Washington State in July
2019. Primary zinc production (primary slab zinc) in 2018 is used as an estimate for 2019 and 2020 due to the lack
of available data. Secondary zinc production in 2020 increased by 250 percent compared to 2019 and was largely
influenced by the reopening of the idled AZR secondary zinc refinery in Mooresboro, NC in March 2020 (USGS
2021; AZP 2021). Secondary zinc production from the reopened facility was estimated by subtracting estimated
primary zinc production from the total zinc production value obtained from the USGS Minerals Yearbook: Zinc.
Table 4-94: Zinc Production (Metric Tons)
Year Primary Secondary Total
1990
262,704
95,708
358,412
2005
191,120
156,000
347,120
2016
111,000
15,000
126,000
2017
117,000
15,000
132,000
2018
101,000
15,000
116,000
2019
101,000
14,000
115,000
2020
101,000
49,000
150,000
Industrial Processes and Product Use 4-115
-------
Methodology and Time-Series Consistency
The methods used to estimate non-energy CO2 emissions from zinc production73 using the electrothermic primary
production and Waelz kiln secondary production processes are based on Tier 1 methods from the 2006IPCC
Guidelines (IPCC 2006). The Tier 1 equation used to estimate emissions from zinc production is as follows:
Equation 4-16: 2006IPCCGuide/inesTier 1: CO2 Emissions From Zinc Production (Equation
4.33)
Eco2 ~ Zn x EFdefault
where,
Eco2 = CO2 emissions from zinc production, metric tons
Zn = Quantity of zinc produced, metric tons
EFdefauit = Default emission factor, metric tons CCh/metric ton zinc produced
The Tier 1 emission factors provided by IPCC for Waelz kiln-based secondary production were derived from
metallurgical coke consumption factors and other data presented in Vikland-White (2000). These coke
consumption factors as well as other inputs used to develop the Waelz kiln emission factors are shown below. IPCC
does not provide an emission factor for electrothermic processes due to limited information; therefore, the Waelz
kiln-specific emission factors were also applied to zinc produced from electrothermic processes. Starting in 2014,
refined zinc produced in the United States used hydrometallurgical processes and is assumed to be non-emissive.
For Waelz kiln-based production, IPCC recommends the use of emission factors based on EAF dust consumption, if
possible, rather than the amount of zinc produced since the amount of reduction materials used is more directly
dependent on the amount of EAF dust consumed. Since only a portion of emissive zinc production facilities
consume EAF dust, the emission factor based on zinc production is applied to the non-EAF dust consuming
facilities, while the emission factor based on EAF dust consumption is applied to EAF dust consuming facilities.
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
t-j ^ aelz KlL~h ~~ ¦ ¦ ^ ^ . -— . .
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); 2011 from the 201110-k (Horsehead Corp. 2012a);
73 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.
4-116 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
2012 from the 2012 10-k (Horsehead Corp. 2013); and 2013 from the 2013 10-k (Horsehead Corp. 2014).
Metallurgical coke consumption levels for 2014 and later were zero due to the closure of the AZR (formerly
"Horsehead Corporation") Monaca, PA electrothermic furnace facility. 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
EFgAp Qtic? ^ ^
metric tons EAF Dust metric tons coke metric tons C metric tons EAF Dust
Metallurgical coke consumption for EAF dust consuming facilities for 1990 through 2020 were calculated based on
the values of EAF dust consumed. The values of EAF dust consumed for AZR, SDR, and PIZO are explained below.
The total amount of EAF dust consumed by AZR at their Waelz kilns was available from AZR (formerly "Horsehead
Corporation") financial reports for years 2006 through 2015 (Horsehead 2007, 2008, 2010a, 2011, 2012a, 2013,
2014, 2015, and 2016) and from AZR for 2016, 2017, 2018, and 2019 (AZR 2020). EAF dust consumption for 2020
was not available at the time of publication and were estimated using 2019 values. 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 AZR'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 2020 (SDR 2012, 2014, 2015, 2017, 2018, 2021). 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, and 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.
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 the current Inventory report, EPA
sought expert judgment from the USGS mineral commodity expert to assess approaches for splitting total
production into primary and secondary values. For years 2016 through 2020, only one facility produced primary
Industrial Processes and Product Use 4-117
-------
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 2020.
Uncertainty
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 (1) an EAF dust consumption value reported annually by AZR/Horsehead Corporation as
part of its financial reporting to the Securities and Exchange Commission (SEC) and provided by AZR, and (2) an EAF
dust consumption value obtained from the Waelz kiln facility operated in Alabama by Steel Dust Recycling LLC.
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 is 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.
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.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-95. Zinc production CO2
emissions from 2020 were estimated to be between 0.8 and 1.2 MMT CO2 Eq. at the 95 percent confidence level.
This indicates a range of approximately 19 percent below and 20 percent above the emission estimate of 1.0 MMT
CO2 Eq.
Table 4-95: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Zinc
Production (MMT CO2 Eq. and Percent)
Source
Gas 2020 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT CO? Eq.)
(MMT C02 Eq.)
(%)
Lower Upper
Lower Upper
Bound Bound
Bound Bound
Zinc Production
C02 1.0
-------
Recalculations Discussion
No recalculations were made impacting emissions for the 1990 through 2019 portion of the time series.
Planned Improvements
Pending resources and prioritization of improvements for more significant sources, EPA will continue to evaluate
and analyze data reported under EPA's GHGRP that would be useful to improve the emission estimates and
category-specific QC for the Zinc Production source category, in particular considering completeness of reported
zinc production given the reporting threshold. Given the small number of facilities in the United States, particular
attention will be made to risks for disclosing CBI and ensuring 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.74 This is a long-term planned improvement,
and EPA is still assessing the possibility of including this improvement in future Inventory reports.
4.23 Electronics Industry (CRF Source
Category 2E)
The electronics industry uses multiple greenhouse gases in its manufacturing processes. In semiconductor
manufacturing, these include long-lived fluorinated greenhouse gases used for plasma etching and chamber
cleaning (CRF Source Category 2E1), fluorinated heat transfer fluids used for temperature control and other
applications (CRF Source Category 2E4), and nitrous oxide (N2O) used to produce thin films through chemical vapor
deposition and in other applications (reported under CRF Source Category 2H3). Similar to semiconductor
manufacturing, the manufacturing of micro-electro-mechanical systems (MEMS) devices (reported under CRF
Source Category 2E5 Other) and photovoltaic (PV) cells (CRF Source Category 2E3) requires the use of multiple
long-lived fluorinated greenhouse gases for various processes.
The gases most commonly employed in the electronics industry are trifluoromethane (hydrofluorocarbon (HFC)-23
or CHF3), perfluoromethane (CF4), perfluoroethane (C2F6), nitrogen trifluoride (NF3), and sulfur hexafluoride (SFs),
although other fluorinated compounds such as perfluoropropane (C3F8) and perfluorocyclobutane (c-C4Fs) are also
used. The exact combination of compounds is specific to the process employed.
In addition to emission estimates for these seven commonly used fluorinated gases, this Inventory contains
emissions estimates for N2O and other HFCs and unsaturated, low-GWP PFCs including CsFs, C4F6, HFC-32, HFC-41,
and HFC-134a. These additional HFCs and PFCs are emitted from etching and chamber cleaning processes in much
smaller amounts, accounting for 0.02 percent of emissions (in CO2 Eq.) from these processes.
For semiconductors, a single 300 mm silicon wafer that yields between 400 to 600 semiconductor products
(devices or chips) may require more than 100 distinct fluorinated-gas-using process steps, principally to deposit
and pattern dielectric films. Plasma etching (or patterning) of dielectric films, such as silicon dioxide and silicon
nitride, is performed to provide pathways for conducting material to connect individual circuit components in each
device. The patterning process uses plasma-generated fluorine atoms, which chemically react with exposed
dielectric film to selectively remove the desired portions of the film. The material removed as well as undissociated
74 See http://www.ipcc-nggiD.iges.or.jp/public/tb/TFl Technical Bulletin l.pdf.
Industrial Processes and Product Use 4-119
-------
fluorinated gases flow into waste streams and, unless emission abatement systems are employed, into the
atmosphere. Plasma enhanced chemical vapor deposition (PECVD) chambers, used for depositing dielectric films,
are cleaned periodically using fluorinated and other gases. During the cleaning cycle the gas is converted to
fluorine atoms in plasma, which etches away residual material from chamber walls, electrodes, and chamber
hardware. Undissociated fluorinated gases and other products pass from the chamber to waste streams and,
unless abatement systems are employed, into the atmosphere.
In addition to emissions of unreacted gases, some fluorinated compounds can also be transformed in the plasma
processes into different fluorinated compounds which are then exhausted, unless abated, into the atmosphere.
For example, when C2F6 is used in cleaning or etching, CF4 is typically generated and emitted as a process
byproduct. In some cases, emissions of the byproduct gas can rival or even exceed emissions of the input gas, as is
the case for NF3 used in remote plasma chamber cleaning, which often generates CF4 as a byproduct.
Besides dielectric film etching and PECVD chamber cleaning, much smaller quantities of fluorinated gases are used
to etch polysilicon films and refractory metal films like tungsten.
Nitrous oxide is used in manufacturing semiconductor devices to produce thin films by CVD and nitridation
processes as well as for N-doping of compound semiconductors and reaction chamber conditioning (Doering
2000).
Liquid perfluorinated compounds are also used as heat transfer fluids (F-HTFs) for temperature control, device
testing, cleaning substrate surfaces and other parts, and soldering in certain types of semiconductor
manufacturing production processes. Leakage and evaporation of these fluids during use is a source of fluorinated
gas emissions (EPA 2006). Unweighted F-HTF emissions consist primarily of perfluorinated amines,
hydrofluoroethers, perfluoropolyethers (specifically, PFPMIEs), and perfluoroalkylmorpholines. One percent or 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 10,000.75
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 2016 to
2020. The rapid growth of the electronics industry and the increasing complexity (growing number of layers and
functions)76 of electronic products led to an increase in emissions of 153 percent between 1990 and 1999, when
75 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.
76 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.
4-120 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
emissions peaked at 9.1 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 48 percent
decline from 1999 to 2020. 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 32 percent between 1990 and 2020. Total emissions from semiconductor manufacture in 2020
were slightly higher than 2019 emissions, increasing by less than 1 percent, largely due to a large increase in N2O
emissions.
For 2020, total GWP-weighted emissions of all fluorinated greenhouse gases and N2O from deposition, etching,
and chamber cleaning processes in the U.S. semiconductor industry were estimated to be 4.7 MMT CO2 Eq. This is
a decrease in emissions from 1999 of 49 percent, and an increase in emissions from 1990 of 30 percent. These
trends are driven by the above stated reasons.
Emissions from all fluorinated greenhouse gases from photovoltaic cells and MEMS manufacturing, 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 SFe,77 and were equivalent to only 0.096 percent to 0.233 percent of the total reported emissions from
electronics manufacturing in 2011 to 2020. F-GHG emissions, the primary type of emissions for MEMS, ranged
from 0.0003 to 0.0107 MMT CO2 Eq. from 1991 to 2020. 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 GWP-weighted emissions from manufacturing of photovoltaic cells were estimated to range from 0.0003
MMT CO2 Eq. to 0.0235 MMT CO2 Eq. from 1998 to 2020 and were equivalent to between 0.003 percent to 0.496
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 0.0222 MMT CO2 Eq. from 1998 to 2020. 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.78
Emissions of F-HTFs, grouped by HFCs, PFCs or SF6 are presented in Table 4-96. Table 4-98 shows F-HTF emissions
in tons by compound group based on reporting to EPA's Greenhouse Gas Reporting Program (GHGRP) by
semiconductor manufacturers during years 2014 through 2020. 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. An analysis of the data reported to EPA's GHGRP indicates that F-HTF emissions account for
anywhere between 13 percent and 19 percent of total annual emissions (F-GHG, N2O and F-HTFs) from
77 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.
78 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.
Industrial Processes and Product Use 4-121
-------
semiconductor manufacturing.79 Table 4-98 shows F-HTF emissions in tons by compound group based on reporting
to EPA's GHGRP during years 2014 through 2020.80
Table 4-96: PFC, HFC, SFe, NF3, and N2O Emissions from Electronics Industry (MMT CO2 Eq.)
Year
1990
2005
2016
2017
2018
2019
2020
cf4
c2f6
CsFs
C4Fs
HFC-23
SF6
nf3
C4F6
CsFs
CH2F2
ch3f
CH2FCF3
0.8
2.0
+
0.0
0.2
0.5
+
+
+
+
+
+
1.1
2.0
0.1
0.1
0.2
0.7
0.5
+
+
+
+
+
1.5
1.2
0.1
0.1
0.3
0.8
0.6
+
+
+
+
+
1.6
1.2
0.1
0.1
0.4
0.7
0.6
+
+
+
+
+
1.7
1.1
0.1
0.1
0.4
0.8
0.6
+
+
+
+
+
1.6
1.0
0.1
0.1
0.4
0.7
0.6
+
+
+
+
+
1.7
0.9
0.1
0.1
0.4
0.7
0.6
+
+
+
+
+
Total Semiconductors
3.6
4.6
4.7
4.6
4.8
4.4
4.4
CF4
c2f6
CsFs
C4Fs
HFC-23
SF6
nf3
0.0
0.0
0.0
0.0
0.0
0.0
0.0
+
+
0.0
+
+
+
+
+
+
0.0
+
+
+
+
+
+
0.0
+
+
+
+
+
+
0.0
+
+
+
+
+
+
0.0
+
+
+
+
Total MEMS
0.0
cf4
c2f6
C4Fs
HFC-23
SF6
nf3
0.0
0.0
0.0
0.0
0.0
0.0
+
+
+
+
0.0
0.0
+
+
+
+
0.0
0.0
+
+
+
+
0.0
0.0
+
+
+
+
0.0
0.0
+
+
+
+
0.0
0.0
+
+
+
+
0.0
0.0
Total PV
0.0
+
+
+
+
+
+
N20 (Semiconductors)
+
0.1
0.2
0.3
0.3
0.2
0.3
N20 (MEMS)
0.0
+
+
+
+
+
+
N20 (PV)
0.0
+
+
+
+
+
+
Total N20
+
0.1
0.2
0.3
0.3
0.2
0.3
HFC, PFC and SF6 F-HTFs
0.0
+
+
+
+
+
+
Total Electronics Industry
3.6
4.8
5.0
4.9
5.1
4.7
4.7
+ Does not exceed 0.05 MMT C02 Eq.
79 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.
80 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.
4-122 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 4-97: PFC, HFC, SFe, NF3, and N2O Emissions from Semiconductor Manufacture (Metric
Tons)
Year
1990
2005
2016
2017
2018
2019
2020
cf4
114.8
146.2
208.6
219.8
234.7
219.0
224.5
c2f6
160.0
161.7
99.5
97.6
92.9
79.1
70.3
CsFs
0.4
9.0
14.3
11.7
12.1
10.1
9.0
C4Fs
0.0
11.4
5.4
5.8
6.0
5.7
5.7
HFC-23
14.6
13.7
23.2
25.7
26.5
25.5
26.5
sf6
21.7
30.7
35.7
30.0
33.4
32.4
31.8
nf3
2.8
28.5
33.2
32.8
34.1
33.2
36.1
C4F6
0.7
0.9
1.0
0.9
0.8
0.9
0.8
CsFs
0.4
0.6
0.5
0.8
0.5
1.2
0.4
ch2f2
0.7
0.9
0.9
1.1
0.9
1.0
1.1
CH3F
1.5
2.0
1.9
2.3
3.0
2.5
2.8
CH2FCF3
+
+
+
+
+
+
+
N20
120.2
412.0
789.8
911.3
852.0
781.6
993.1
+ Does not exceed 0.05 MT.
Table 4-98: F-HTF Emissions from Electronics Manufacture by Compound Group (kt CO2 Eq.)
Year
2014
2015
2016
2017
2018
2019
2020
HFCs
3.3
3.0
4.1
3.6
2.7
1.1
0.9
PFCs
1.6
2.8
2.6
9.1
10.0
8.4
1.8
sf6
20.7
12.8
11.4
16.6
13.2
6.0
12.8
HFEs
4.8
4.2
7.5
2.9
4.6
1.3
5.3
PFPMIEs
182.2
208.1
173.7
148.5
183.0
171.7
149.9
Perfluoalkylromorpholines
108.3
81.5
75.7
52.3
58.6
56.4
60.9
Perfluorotrialkylamines
490.4
438.9
386.7
383.9
410.7
363.6
379.8
Total F-HTFs
811.4
751.4
661.7
616.9
682.9
608.4
611.3
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.
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 PFC81 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)82—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 2020 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 2020. 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 2020. The
81 In the context of the EPA Partnership and PEVM, PFC refers to perfluorocompounds, not perfluorocarbons.
82 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 4-123
-------
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
2020. 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 (<3 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
manufacturing in DisplaySearch in 2009 to the total capacity of c-Si PV manufacturing reported in the
Congressional Research Service report in 2012. Capacities were held constant for non-reporters for 2012 to 2020.
Average emissions per MW from the GHGRP reporter in 2015 and 2016 were then applied to the total capacity
prior to 2015. Emissions for 2014 from the GHGRP reporter that reported in 2015 and 2016 were scaled to the
number of months open in 2014. For 1998 through 2020, emissions per MW (capacity) from the GHGRP reporter
were applied to the non-reporters. For 2017 through 2020, there are no reported PV emissions. Therefore,
emissions were estimated using the EPA-derived emission factor and estimated manufacturing capacity from non-
reporters only.
HTFs
Facility emissions of F-HTFs from semiconductor manufacturing are reported to EPA under its GHGRP and are
available for the years 2011 through 2020. EPA estimates the emissions of F-HTFs from non-reporting
semiconductor facilities by calculating the ratio of GHGRP-reported fluorinated HTF emissions to GHGRP reported
F-GHG emissions from etching and chamber cleaning processes, and then multiplying this ratio by the F-GHG
emissions from etching and chamber cleaning processes estimated for non-reporting facilities. Fluorinated HTF use
in semiconductor manufacturing is assumed to have begun in the early 2000s and to have gradually displaced
other HTFs (e.g., de-ionized water and glycol) in semiconductor manufacturing (EPA 2006). For time-series
consistency, EPA interpolated the share of F-HTF emissions to F-GHG emissions between 2000 (at 0 percent) and
4-124 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
2011 (at 22 percent) and applied these shares to the unadjusted F-GHG emissions during those years to estimate
the fluorinated HTF emissions.
Semiconductors
1990 through 1994
From 1990 through 1994, Partnership data were unavailable, and emissions were modeled using PEVM (Burton
and Beizaie 2001).83 The 1990 to 1994 emissions are assumed to be uncontrolled, since reduction strategies such
as chemical substitution and abatement were yet to be developed.
PEVM is based on the recognition that fluorinated greenhouse gas emissions from semiconductor manufacturing
vary with: (1) the number of layers that comprise different kinds of semiconductor devices, including both silicon
wafer and metal interconnect layers, and (2) silicon consumption (i.e., the area of semiconductors produced) for
each kind of device. The product of these two quantities, Total Manufactured Layer Area (TMLA), constitutes the
activity data for semiconductor manufacturing. PEVM also incorporates an emission factor that expresses
emissions per unit of manufactured layer-area. Emissions are estimated by multiplying TMLA by this emission
factor.
PEVM incorporates information on the two attributes of semiconductor devices that affect the number of layers:
(1) linewidth technology (the smallest manufactured feature size),84 and (2) product type (discrete, memory or
logic).85 For each linewidth technology, a weighted average number of layers is estimated using VLSI product-
specific worldwide silicon demand data in conjunction with complexity factors (i.e., the number of layers per
Integrated Circuit (IC) specific to product type (Burton and Beizaie 2001; ITRS 2007). PEVM derives historical
consumption of silicon (i.e., square inches) by linewidth technology from published data on annual wafer starts
and average wafer size (VLSI Research, Inc. 2012).
The emission factor in PEVM is the average of four historical emission factors, each derived by dividing the total
annual emissions reported by the Partners for each of the four years between 1996 and 1999 by the total TMLA
estimated for the Partners in each of those years. Over this period, the emission factors varied relatively little (i.e.,
the relative standard deviation for the average was 5 percent). Since Partners are believed not to have applied
significant emission reduction measures before 2000, the resulting average emission factor reflects uncontrolled
emissions and hence may be use here to estimate 1990 through 1994 emissions. The emission factor is used to
estimate U.S. uncontrolled emissions using publicly available data on world (including U.S.) silicon consumption.
As it was assumed for this time period that there was no consequential adoption of fluorinated-gas-reducing
measures, a fixed distribution of fluorinated-gas use was assumed to apply to the entire U.S. industry to estimate
gas-specific emissions. This distribution was based upon the average fluorinated-gas purchases made by
semiconductor manufacturers during this period and the application of IPCC default emission factors for each gas
(Burton and Beizaie 2001).
83 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.
84 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).
85 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 4-125
-------
PEVM only addressed the seven main F-GHGs (CF4, C2F6, C3F8, C-C4F8, HFC-23, SF6, and NF3) used in semiconductor
manufacturing. Through reporting under Subpart I of EPA's GHGRP, data on other F-GHGs (C4F6, CsFs, HFC-32, HFC-
41, HFC-134a) used in semiconductor manufacturing became available and EPA was therefore able to extrapolate
this data across the entire 1990 to 2020 timeseries. To estimate emissions for these "other F-GHGs", 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 and then these shares were applied to all years prior to reported data from
Subpart I (1990 through 2010) and to the emissions from non-reporters from 2011 to 2020.
To estimate N2O emissions, it was assumed the proportion of N2O emissions estimated for 1995 (discussed below)
remained constant for the period of 1990 through 1994.
1995 through 1999
For 1995 through 1999, total U.S. emissions were extrapolated from the total annual emissions reported by the
Partners (1995 through 1999). Partner-reported emissions are considered more representative (e.g., in terms of
capacity utilization in a given year) than PEVM-estimated emissions and are used to generate total U.S. emissions
when applicable. The emissions reported by the Partners were divided by the ratio of the total capacity of the
plants operated by the Partners and the total capacity of all of the semiconductor plants in the United States; this
ratio represents the share of capacity attributable to the Partnership. This method assumes that Partners and non-
Partners have identical capacity utilizations and distributions of manufacturing technologies. Plant capacity data is
contained in the World Fab Forecast (WFF) database and its predecessors, which is updated quarterly. Gas-specific
emissions were estimated using the same method as for 1990 through 1994.
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.
For this time period, the N2O emissions were estimated using an emission factor that was applied to the annual,
total U.S. TMLA manufactured. The emission factor was developed using a regression-through-the-origin (RTO)
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.86 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
86 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.
4-126 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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).87,88,89
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.
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.90 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
87 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.
88 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.
89 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.
90 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.
Industrial Processes and Product Use 4-127
-------
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.91 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
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,92 if a site-specific DRE was indicated), and the fab-wide DREs reported in
2014.93 To adjust emissions for facilities that abated emissions in 2011 through 2013, EPA first estimated
91 GaAs and Si technologies refer to the wafer on which devices are manufactured, which use the same PFCs but in different
ways.
92 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.
93 If abatement information was not available for 2014 or the reported incorrectly in 2014, data from 2015 or 2016 was
substituted.
4-128 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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.94
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).95 In a
refinement of the method used to estimate emissions for the non-Partner population for prior years, different
emission factors were developed for different subpopulations of fabs, disaggregated by wafer size (200 mm and
300 mm). For each of these groups, a subpopulation-specific emission factor was obtained using a regression-
through-the-origin (RTO) model: facility-reported aggregate emissions of seven F-GHGs (CF4, C2F6, C3F8, C-C4F8,
CHF3, SFs and NF3)96 were regressed against the corresponding TMLA to estimate an aggregate F-GHG emissions
factor (CO2 Eq./MSI TMLA), and facility-reported N2O emissions were regressed against the corresponding TMLA to
estimate a N2O emissions factor (CO2 Eq./MSI TMLA). For each subpopulation, the slope of the RTO model is the
emission factor for that subpopulation. Information on the use of point-of-use abatement by non-reporting fabs
was not available; thus, EPA conservatively assumed that non-reporting facilities did not use point-of-use
abatement.
For 2011 and 2012, estimates of TMLA relied on the capacity utilization of the fabs published by the U.S. Census
Bureau's Historical Data Quarterly Survey of Plant Capacity Utilization (USCB 2011, 2012). Similar to the
assumption for 2007 through 2010, facilities with only R&D activities were assumed to utilize only 20 percent of
their manufacturing capacity. All other facilities in the United States are assumed to utilize the average percent of
the manufacturing capacity without distinguishing whether fabs produce discrete products or logic products.
Non-reporting fabs were then broken out into subpopulations by wafer size (200 mm and 300 mm), using
information available through the WFF. The appropriate emission factor was applied to the total TMLA of each
subpopulation of non-reporting facilities to estimate the GWP-weighted emissions of that subpopulation.
Gas-specific, GWP-weighted emissions for each subpopulation of non-reporting facilities were estimated using the
corresponding reported distribution of gas-specific, GWP-weighted emissions from which the aggregate emission
factors, based on GHGRP-reported data, were developed. Estimated in this manner, the non-reporting population
94 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.
95 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.)
96 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 4-129
-------
accounted for 4.9 and 5.0 percent of U.S. emissions in 2011 and 2012, respectively. The GHGRP-reported emissions
and the calculated non-reporting population emissions are summed to estimate the total emissions from
semiconductor manufacturing.
2013 and 2014
For 2013 and 2014, as for 2011 and 2012, F-GHG and N2O emissions data received through EPA's GHGRP were
aggregated, by gas, across all semiconductor-manufacturing GHGRP reporters to calculate gas-specific emissions
for the GHGRP-reporting segment of the U.S. industry. However, for these years WFF data was not available.
Therefore, an updated methodology that does not depend on the WFF derived activity data was used to estimate
emissions for the segment of the industry that are not covered by EPA's GHGRP. For the facilities that did not
report to the GHGRP (i.e., which are below EPA's GHGRP reporting threshold or are R&D facilities), emissions were
estimated based on the proportion of total U.S. emissions attributed to non-reporters for 2011 and 2012. EPA used
a simple averaging method by first estimating this proportion for both F-GHGs and N2O for 2011, 2012, and 2015
through 2020, resulting in one set of proportions for F-GHGs and one set for N2O, and then applied the average of
each set to the 2013 and 2014 GHGRP reported emissions to estimate the non-reporters' emissions. Fluorinated
gas-specific, GWP-weighted emissions for non-reporters were estimated using the corresponding reported
distribution of gas-specific, GWP-weighted emissions reported through EPA's GHGRP for 2013 and 2014.
GHGRP-reported emissions in 2013 were adjusted to capture changes to the default emission factors and default
destruction or removal efficiencies used for GHGRP reporting, affecting the emissions trend between 2013 and
2014. EPA used the same method to make these adjustments as described above for 2011 and 2012 GHGRP data.
2015 through 2020
Similar to the methods described above for 2011 and 2012, and 2013 and 2014, EPA relied upon emissions data
reported directly through the GHGRP. For 2015 through 2020, EPA took an approach similar to the one used for
2011 and 2012 to estimate emissions 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. However, in a change from
previous years, EPA was able to develop new annual emission factors for 2015 through 2020 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
4-130 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 2006IPCC 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, 2015 and 2016 were
obtained from the U.S. Census Bureau's Historical Data Quarterly Survey of Plant Capacity Utilization (USCB 2011,
2012, 2015, and 2016).
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).
Uncertainty
A quantitative uncertainty analysis of this source category was performed using the IPCC-recommended Approach
2 uncertainty estimation methodology, the Monte Carlo Stochastic Simulation technique. The Monte Carlo
Stochastic Simulation was performed on the total emissions estimate from the Electronics Industry, represented in
equation form as:
Equation 4-19: Total Emissions from Electronics Industry
Total Emissions (Et) = Semiconductors F-GHG and N2O Emissions (Esemi)+ MEMS F-GHG and N2O Emissions
(Emems) + PV F-GHG and N2O Emissions (Epv) + HFC, PFC and SF6 F-HTFs Emissions (Ehtf)
The uncertainty in the total emissions for the Electronics Industry, presented in Table 4-99 below, results from the
convolution of four distributions of emissions, namely from semiconductors manufacturing, MEMS manufacturing,
PV Manufacturing and emissions of Heat Transfer Fluids. The approaches for estimating uncertainty in each of the
sources are described below:
Semiconductors Manufacture Emission Uncertainty
The Monte Carlo Stochastic Simulation was performed on the emissions estimate from semiconductor
manufacturing, represented in equation form as:
Equation 4-20: Total Emissions from Semiconductor Manufacturing
Semiconductors F-GHG and N2O Emissions (Esemi) = GHGRP Reported F-GHG Emissions (ER,F-GHG,semi) + Non-
Reporters' Estimated F-GHG Emissions (ENR,F-GHG,semi) + GHGRP Reported N2O Emissions (ER.Nzo.semi) + Non-
Reporters' Estimated N2O Emissions (ENR,N20,semi)
Industrial Processes and Product Use 4-131
-------
The uncertainty in Esemi results from the convolution of four distributions of emissions, ER,F-GHG,semi ER,N2o,semi Enr,f-
GHG/Semi and ENR,N2o,semi. The approaches for estimating each distribution and combining them to arrive at the
reported 95 percent confidence interval (CI) for Esemi are described in the remainder of this section.
The uncertainty estimate of Er, F-GHG,semi, or GHGRP-reported F-GHG emissions, is developed based on gas-specific
uncertainty estimates of emissions for two industry segments, one processing 200 mm or less wafers and one
processing 300 mm wafers. Uncertainties in emissions for each gas and industry segment are based on an
uncertainty analysis conducted during the assessment of emission estimation methods for the Subpart I
rulemaking in 2012 (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).97 This
assessment relied on facility-specific gas information by gas and wafer size, and incorporated uncertainty
associated with both emission factors and gas consumption quantities. The 2012 analysis did not consider the use
of abatement.
For the industry segment that manufactured 200 mm wafers, estimates of uncertainty at a 95 percent CI ranged
from ±29 percent for C3F8 to ±10 percent for CF4. For the corresponding 300 mm industry segment, estimates of
uncertainty at the 95 percent CI ranged from ±36 percent for C4F8 to ±16 percent for CF4. For gases for which
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
97 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.
4-132 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
inputs in a separate Monte Carlo simulation to estimate the uncertainty around the TMLA of both individual
facilities as well as the total non-reporting TMLA of each sub-population.
The uncertainty around the emission factors for non-reporting facilities is dependent on the uncertainty of the
total emissions (MMT CO2 Eq. units) and the TMLA of each reporting facility in that category. For each wafer size
for reporting facilities, total emissions were regressed on TMLA (with an intercept forced to zero) for 10,000
emission and 10,000 TMLA values in a Monte Carlo simulation, which results in 10,000 total regression coefficients
(emission factors). The 2.5th and the 97.5th percentile of these emission factors are determined, and the bounds
are assigned as the percent difference from the estimated emission factor.
The next step in estimating the uncertainty in emissions of reporting and non-reporting facilities in semiconductor
manufacture is convolving the distribution of reported emissions, emission factors, and TMLA using Monte Carlo
simulation. For this Monte Carlo simulation, the distributions of the reported F-GHG gas- and wafer size-specific
emissions are assumed to be normally distributed, and the uncertainty bounds are assigned at 1.96 standard
deviations around the estimated mean. The were some instances, though, where departures from normality were
observed for variables, including for the distributions of the gas- and wafer size-specific N2O emissions, TMLA, and
non-reporter emission factors, both for F-GHGs and N2O. As a result, the distributions for these parameters were
assumed to follow a pert beta distribution.
MEMS Manufacture Emission Uncertainty
The Monte Carlo Stochastic Simulation was performed on the emissions estimate from MEMS manufacturing,
Industrial Processes and Product Use 4-133
-------
represented in equation form as:
Equation 4-21: Total Emissions from MEMS Manufacturing
MEMS F-GHG and N2O Emissions (Emems) = GHGRP Reported F-GHG Emissions (Er,f-ghg,mems) + GHGRP
Reported N2O Emissions (Er.nzo, mems)
Emissions from MEMS manufacturing are only quantified for GHGRP reporters. MEMS manufacturers that report
to the GHGRP all report the use of 200 mm wafers. Some MEMS manufacturers report using abatement
equipment. Therefore, the estimates of uncertainty at the 95 percent CI for each gas emitted by MEMS
manufacturers are set equal to the gas-specific uncertainties for manufacture of 200mm semiconductor wafers
with partial abatement. The same assumption is applied for uncertainty levels for GHGRP reported MEMS N2O
emissions (Er,n2o,mems).
PV Manufacture Emission Uncertainty
The Monte Carlo Stochastic Simulation was performed on the emissions estimate from PV manufacturing,
represented in equation form as:
Equation 4-22: Total Emissions from PV Manufacturing
PV F-GHG and N2O Emissions (Epv) = Non-Reporters' Estimated F-GHG Emissions (Enr,f-ghg,pv) + Non-
Reporters' Estimated N2O Emissions (Enr.nzo.pv)
Emissions from PV manufacturing are only estimated for non-GHGRP reporters. There were no reported emissions
from PV manufacturing in GHGRP in 2020. The "Non-Reporters' Estimated F-GHG Emissions" term was estimated
using an emission factor developed using emissions from reported data in 2015 and 2016 and total non-reporters'
capacity. Due to a lack of information and data and because they represent similar physical and chemical
processes, the uncertainty at the 95 percent I level for non-reporter PV capacity is assumed to be the same as the
uncertainty in non-reporter TMLA for semiconductor manufacturing. Similarly, the uncertainty for the PV
manufacture emission factors are assumed to be the same as the uncertainties in emission factors used for non-
reporters in semiconductor manufacture.
Heat Transfer Fluids Emission Uncertainty
There is a lack of data related to the uncertainty of emission estimates of heat transfer fluids used for electronics
manufacture. Therefore, per the 2006IPCC Guidelines (IPCC 2006, Volume 3, Chapter 6), uncertainty bounds of 20
percent were applied to the segments of PFCs, HFCs and SF6 at national levels.
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. The
emissions estimate for total U.S. F-GHG, N2O, and HTF emissions from electronics manufacturing were estimated
to be between 4.45 and 5.03 MMT CO2 Eq. at a 95 percent confidence level. This range represents 6 percent below
to 6 percent above the 2020 emission estimate of 4.74 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)
2020 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
4.74
4.45
5.03
-6% 6%
4-134 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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).98 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
Emissions from 2015 through 2020 were updated to reflect updated emissions reporting in EPA's GHGRP, relative
to the previous 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.
• In the dataset used to estimate photovoltaics manufacturing capacity from 2000 to 2009, a correction was
made to the formula which sums annual capacity across all producers. This resulted in slight changes to
emissions estimates for the years where this dataset is used.
• Previously, all N2O emissions were attributed solely to semiconductor manufacturing. For this inventory
update, EPA revised the N2O estimates by assigning emissions to the specific types of electronics
manufacturing (i.e., semiconductor, photovoltaic cells, and MEMS). N2O estimates are now reported with
subtotals for each product type within the electronics industry.
• EPA revised the individual gases reported for semiconductor manufacturing to remove the "Other F-
GHGs" category and replace it with separate totals for each individual gas. Similarly, EPA also updated the
MEMS and photovoltaic cells estimates to show disaggregated totals for each individual HFC and PFC
compound.
• A GHGRP fab that had previously been identified as a MEMS fab was determined to have produced
photovoltaics. Their F-GHG emissions were removed from the MEMS totals and added to the PV totals.
• Previously, F-GHG emissions in 2016 from a PV manufacturer reporting through the GHGRP were held
constant for 2017 through the most recent Inventory year. EPA determined that this manufacturer
ceased operations in 2016, so their reported emissions were changed to zero for 2017 and beyond.
98 GHGRP Report Verification Factsheet. See https://www.epa.gov/sites/production/files/2Q15-
07/documents/ghgrp verification factsheet.pdf.
Industrial Processes and Product Use 4-135
-------
• 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 0.02 MMT CO2 Eq. (0.44 percent) across
the time series (1990 through 2019).
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.
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 2014. 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 2020. 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) and perfluorocarbons (PFCs) 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." 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
99 [42 U.S.C § 7671, CAA Title VI],
4-136 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 and PFCs used as substitutes for ODSs are provided in Table 4-100 and Table 4-101.100
Table 4-100: Emissions of HFCs and PFCs from ODS Substitutes (MMT CO2 Eq.)
Gas
1990
2005
2016
2017
2018
2019
2020
HFC-23
0.0
+
+
+
+
+
+
HFC-32
0.0
0.3
4.6
5.3
6.0
6.8
7.7
HFC-125
+
9.0
46.9
50.1
53.7
58.4
63.5
HFC-134a
+
80.1
69.1
64.7
62.1
60.9
59.5
HFC-143a
+
9.4
28.2
28.0
27.7
27.8
27.9
HFC-236fa
0.0
1.2
1.3
1.2
1.2
1.1
1.1
cf4
0.0
+
+
+
0.1
0.1
0.1
Others3
0.2
7.2
14.9
16.1
16.6
16.8
16.6
Total
0.2
107.2
165.1
165.5
167.3
171.8
176.3
+ 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 C6Fi4.
Note: Totals may not sum due to independent rounding.
Table 4-101: Emissions of HFCs and PFCs from ODS Substitution (Metric Tons)
Gas
1990
2005
2016
2017
2018
2019
2020
HFC-23
0
1
2
2
2
2
2
HFC-32
0
397
6,791
7,832
8,937
10,077
11,374
HFC-125
+
2,580
13,399
14,308
15,335
16,682
18,153
HFC-134a
+
56,029
48,337
45,264
43,419
42,558
41,590
HFC-143a
+
2,093
6,320
6,264
6,188
6,230
6,234
HFC-236fa
0
118
129
124
118
112
108
cf4
0
2
6
6
7
7
7
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.101 In 1993, the use of HFCs in foam production began, and
100 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.
101 R-404A contains HFC-125, HFC-143a, and HFC-134a.
Industrial Processes and Product Use 4-137
-------
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.
The use and subsequent emissions of HFCs and PFCs as ODS substitutes has been increasing from small amounts in
1990 to 176.3 MMT CO2 Eq. emitted in 2020. 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 and PFCs as ODS substitutes by end-use sector for 1990 through 2020. 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 and PFCs as ODS substitutes in 2020 include refrigeration and air-
conditioning (137.7 MMT CO2 Eq., or approximately 78 percent), aerosols (18.1 MMT CO2 Eq., or approximately 10
percent), and foams (15.5 MMT CO2 Eq., or approximately 9 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 (34.3 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 and PFCs from ODS Substitutes (MMT CO2 Eq.) by Sector
Sector
1990
2005
2016
2017
2018
2019
2020
Refrigeration/Air Conditioning
+
89.7
126.4
126.9
129.3
133.3
137.7
Commercial Refrigeration
+
15.0
42.8
41.4
40.3
41.1
41.6
Domestic Refrigeration
+
0.2
1.3
1.3
1.4
1.3
1.3
Industrial Process
Refrigeration
+
1.9
11.6
12.9
14.1
15.3
16.5
Transport Refrigeration
+
1.6
5.9
6.4
6.9
7.4
7.9
Mobile Air Conditioning
+
67.7
37.4
33.7
31.5
29.3
27.1
Residential Stationary Air
Conditioning
+
1.4
21.6
24.8
28.2
31.6
35.7
Commercial Stationary Air
Conditioning
+
2.0
5.8
6.3
6.8
7.2
7.6
Aerosols
0.2
10.7
19.6
18.6
17.4
17.8
18.1
Foams
+
4.0
14.7
15.6
16.1
16.0
15.5
Solvents
+
1.7
1.9
1.9
2.0
2.0
2.0
Fire Protection
+
1.1
2.4
2.5
2.6
2.8
2.8
Total
0.2
107.2
165.1
165.5
167.3
171.8
176.3
+ 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
4-138 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
are HFC-134a, R-410A,102 R-404A, and R-507A.103 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-454B104 in the future. 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.
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 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.
102 R-410A contains HFC-32 and HFC-125.
103 R-507A, also called R-507, contains HFC-125 and HFC-143a.
104 R_454B contains HFC-32 and HFO-1234yf.
Industrial Processes and Product Use 4-139
-------
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
a low-GWP option and 2-BTP is being considered. As fire protection equipment is tested or deployed, emissions of
these fire protection agents occur.
Methodology and Time-Series Consistency
A detailed Vintaging Model of ODS-containing equipment and products was used to estimate the actual—versus
potential—emissions of various ODS substitutes, including HFCs and PFCs. The name of the model refers to the fact
that it tracks the use and emissions of various compounds for the annual "vintages" of new equipment that enter
service in each end-use. The Vintaging Model predicts ODS and ODS substitute use in the United States based on
modeled estimates of the quantity of equipment or products sold each year containing these chemicals and the
amount of the chemical required to manufacture and/or maintain equipment and products over time. Emissions
for each end-use were estimated by applying annual leak rates and release profiles, which account for the lag in
emissions from equipment as they leak over time. By aggregating the data for 78 different end-uses, the model
produces estimates of annual use and emissions of each compound. Further information on the Vintaging Model is
contained in Annex 3.9.
Methodological approaches were applied to the entire time series to ensure time-series consistency from 1990
through 2020.
Uncertainty
Given that emissions of ODS substitutes occur from thousands of different kinds of equipment and from millions of
point and mobile sources throughout the United States, emission estimates must be made using analytical tools
such as the Vintaging Model or the methods outlined in IPCC (2006). Though the model is more comprehensive
than the IPCC default methodology, significant uncertainties still exist with regard to the levels of equipment sales,
equipment characteristics, and end-use emissions profiles that were used to estimate annual emissions for the
various compounds.
The uncertainty analysis quantifies the level of uncertainty associated with the aggregate emissions across the 78
end-uses in the Vintaging Model. In order to calculate uncertainty, functional forms were developed to simplify
some of the complex "vintaging" aspects of some end-use sectors, especially with respect to refrigeration and air-
conditioning, and to a lesser degree, fire extinguishing. These sectors calculate emissions based on the entire
lifetime of equipment, not just equipment put into commission in the current year, thereby necessitating
simplifying equations. The functional forms used variables that included growth rates, emission factors, transition
from ODSs, change in charge size as a result of the transition, disposal quantities, disposal emission rates, and
either stock (e.g., number of air conditioning units in operation) for the current year or ODS consumption before
transition to alternatives began (e.g., in 1985 for most end-uses). Uncertainty was estimated around each variable
within the functional forms based on expert judgment, and a Monte Carlo analysis was performed.
The most significant sources of uncertainty for the ODS Substitutes source category include the total stock of
refrigerant installed in industrial process refrigeration and cold storage equipment, as well as the charge size for
technical aerosols using HFC-134a.
4-140 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-103. Substitution of
ozone depleting substances HFC and PFC emissions were estimated to be between 170.3 and 200.8 MMT CO2 Eq.
at the 95 percent confidence level. This indicates a range of approximately 3.4 percent below to 14.0 percent
above the emission estimate of 176.3 MMT CO2 Eq.
Table 4-103: Approach 2 Quantitative Uncertainty Estimates for HFC and PFC Emissions
from ODS Substitutes (MMT CO2 Eq. and Percent)
2020 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
176.3
170.3
200.8
-3.4%
+14.0%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
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. Category specific QC findings are described below.
The QA/QC and verification process for individual gases and sources in the Vintaging Model includes review against
up-to-date market information, including equipment stock estimates, leak rates, and sector transitions to new
chemicals and technologies. In addition, comparisons against published emission and consumption sources by gas
and by source are performed when available as described further below. Independent peer reviews of the
Vintaging Model are periodically performed, including one conducted in 2017 (EPA 2018), to confirm Vintaging
Model estimates and identify updates. The HFCs and PFCs within the unspecified mix of HFCs and PFCs are
modelled and verified individually in the same process as all other gases and sources in the Vintaging Model. For
the purposes of reporting emissions to protect Confidential Business Information (CBI), some HFCs and PFCs are
grouped into an unspecified mix.
Comparison of Reported Consumption to Modeled Consumption of HFCs
Data from EPA's Greenhouse Gas Reporting Program (GHGRP)105 was also used to perform quality control as a
reference scenario check on the modeled net supply of HFCs, which in turn affects the modeled emissions from
this source category as specified in 2006 IPCC Guidelines for National Greenhouse Gas Inventories. To do so,
consumption patterns demonstrated through data reported under GHGRP Subpart OO (Suppliers of Industrial
Greenhouse Gases) and Subpart QQ (Importers and Exporters of Fluorinated Greenhouse Gases Contained in Pre-
Charged Equipment or Closed-Cell Foams) were compared to the modeled demand for new saturated HFCs used as
ODS substitutes from the Vintaging Model. The collection of data from suppliers of HFCs enables EPA to calculate
the reporters' aggregated net supply-the sum of the quantities of chemical produced or imported into the United
States less the sum of the quantities of chemical transformed (used as a feedstock in the production of other
105 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.
Industrial Processes and Product Use 4-141
-------
chemicals), destroyed, or exported from the United States.106This allows for an overall quality control check on
the modeled demand for new chemical in the Vintaging Model as a proxy for total amount supplied, which is
similar to net supply, as an input to the emission calculations in the model.
GHGRP data is not used directly to estimate emissions of ODS Substitutes because it does not include complete,
publishable information on the sectors or end-uses in which that chemical will be used, so it does not provide the
data that would be needed to calculate the source or time that chemical is emitted. Reports to the GHGRP on
production and bulk import (Subpart 00) do not currently include any information on expected end-uses. Reports
on fluorinated gases used in equipment and foams (Subpart QQ) do include information on the type of product
imported or exported. However, this information is confidential and has not been determined to be publishable at
the end-use (i.e., product) level. Irrespective of that, the information would not capture the entire market in the
United States, unless it could be determined that for any given product there is no domestic production.
Reported Net Supply (GHGRP Top-Down Estimate)
Under EPA's GHGRP, suppliers (i.e., producers, importers, and exporters) of HFCs under Subpart 00 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). 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 2011 through 2020 from all companies that reported AIM-listed HFCs, though not all species
were reported in each reporting year.
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.107 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.108 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. 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.
There are eleven saturated HFC species modeled in the Vintaging Model: HFC-23, HFC-32, HFC-125, HFC-134a,
HFC-143a, HFC-152a, HFC-227ea, HFC-236fa, HFC-245fa, HFC-365mfc, and HFC-43-10mee. While some amounts of
less-used saturated HFCs, including isomers of those included in the Vintaging Model, are reportable under EPA's
106 Chemical that is exported, transformed, or destroyed—unless otherwise imported back to the United States—will never be
emitted in the United States.
107 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.
108 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.
4-142 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
GHGRP, the data are believed to represent an amount comparable to the modeled estimates as a quality control
check.
Comparison Results and Discussion
Comparing the estimates of consumption from these two approaches (i.e., reported and modeled) ultimately
supports and improves estimates of emissions, as noted in the 2006IPCC Guidelines (which refer to fluorinated
greenhouse gas consumption based on supplies as "potential emissions"):
[W]hen considered along with estimates of actual emissions, the potential emissions approach can assist
in validation of completeness of sources covered and as a QC check by comparing total domestic
consumption as calculated in this 'potential emissions approach' per compound with the sum of all
activity data of the various uses (IPCC 2006).
Table 4-104 and Figure 4-3 compare the published net supply of saturated HFCs in MMT CO2 Eq. as determined
from Subpart 00 (supply of HFCs in bulk) and Subpart QQ (supply of HFCs in products and foams) of EPA's GHGRP
for the years 2011 through 2020 (EPA 2021a; EPA 2022a) and the chemical demand as calculated by the Vintaging
Model for the same time series.
Table 4-104: U.S. HFC Supply (MMT COz Eq.)
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
Reported Net Supply
(GHGRP)
257
260
307
294
318
271
322
340
344
344
Industrial GHG Suppliers
250
242
290
269
292
243
290
306
314
309
HFCs in Products and
Foams
7
18
17
25
26
28
32
34
30
35
Modeled Supply (Vintaging
Model)
270
274
279
283
282
285
276
280
273
270
Percent Difference
5%
6%
-9%
-4%
-11%
5%
-14%
-18%
-21%
-22%
Industrial Processes and Product Use 4-143
-------
Figure 4-3: U.S. HFC Consumption (MMT CO2 Eq.)
¦ Reported Imports in Products and Foams
¦ Modeled Consumption
¦ Reported Bulk Supply
2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
400 -|
As shown, the estimates from the Vintaging Model are lower than the GHGRP estimates by an average of 8.3
percent across the time series (i.e., 2011 through 2020), with the difference growing to an average of 19 percent
over the last four years (2017 through 2020). Potential reasons for the differences between the reported and
modeled data include:
• The Vintaging Model does not include every saturated HFC that is reported to EPA's GHGRP. Potential
improvements in the modeling could include investigation of what sources use and emit such chemicals—
which are not necessarily used as ODS substitutes—and to add them into the Inventory. However, the
additional reported HFCs represent a small fraction of total HFC use for this source category, both in
GWP-weighted and unweighted terms, and as such, it is not expected that the additional HFCs reported to
EPA are a major driver for the difference between the two sets of estimates. To the extent lower-GWP
isomers were used in lieu of the modeled chemicals (e.g., HFC-134 instead of HFC-134a), lower CO2 Eq.
amounts in the GHGRP data compared to the modeled estimates would be expected.
• Because 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, there could be a temporal discrepancy
when comparing data. A potential improvement would be to incorporate a time lag into the model, which
would require obtaining data on the movement of supplies through the point of actual use. Because the
GHGRP data and the Vintaging Model estimates generally increase over time (although some year-to-year
variations exist), EPA would expect the modeled estimates to be slightly lower than the corresponding
GHGRP data due to this temporal effect. Regulations under the AIM Act require the reporting of chemical
supplies held at the close of the calendar year; such reports may help investigate this possible factor.
• An additional temporal effect can result from the stockpiling of chemicals by suppliers and distributors.
Suppliers might decide 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. Based on information
collected by the EPA at the time, such stockpiling behavior was seen during ODS phasedowns, but it is
unclear if such behavior 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.
4-144 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
This effect may be a major reason why there is a divergence in the comparison above, with the GHGRP
data in 2017 through 2020 significantly higher than the modeled data. Similar to above, improvements of
the model methodology to incorporate a temporal factor could be investigated. Information on U.S. HFC
stockpiles could be used to assess this possible source of discrepancy; however, this data is not collected
from suppliers under the GHGRP. Future reporting under the AIM Act may provide useful information in
evaluating this issue. Under EPA's GHGRP, all facilities that produce HFCs are required to report their
quantities, whereas importers or exporters of HFCs or pre-charged equipment and closed-cell foams that
contain HFCs are only required to report if either their total imports or their total exports of greenhouse
gases are greater than or equal to 25,000 metric tons of CO2 Eq. per year. Thus, some imports or exports
may not be accounted for in the GHGRP data. In 2021, some companies below the reporting threshold for
imports and exports reported to the GHGRP, including data from as early as 2011, for AIM-listed HFCs as
part of data collection efforts for the U.S. production and consumption baselines; this data is included in
the totals presented above. Future reporting under the AIM Act, if released, would likewise be included in
the reported totals in the future.
• There could be underreporting to the GHGRP. EPA routinely reviews import data provided by U.S.
Customs and Border Protection (CBP) to verify reported supply data and identify facilities that may be
subject to the GHGRP. Based on this review and other information, there appeared to be companies that
imported or exported more than 25,000 metric tons CO2 Eq. of HFCs annually that had not reported
imports or exports to the GHGRP. Continued enactment and enforcement of the AIM Act is expected to
minimize any such information gaps.
• In some years, imports and exports may be greater than consumption because the excess is being used to
increase chemical or equipment stockpiles as discussed above; in other years, the opposite may hold true.
Similarly, relocation of manufacturing facilities or recovery from the recessions and the COVID-19
pandemic could contribute to variability in imports or exports. Averaging net supplies over multiple years
can minimize the impact of such fluctuations. For example, when the 2012 and 2013 net additions to the
supply are averaged, as shown in Table 4-105, the percent difference between the consumption estimates
decreases compared to the 2012-only and 2013-only estimates.
Table 4-105: Averaged U.S. HFC Demand (MMT CCh Eq.)
2011-
2012-
2013-
2014-
2015-
2016-
2017-
2018-
2019-
2012
2013
2014
2015
2016
2017
2018
2019
2020
Avg.
Avg.
Avg.
Avg.
Avg.
Avg.
Avg.
Avg.
Avg.
Reported Net
Supply (GHGRP)
259
284
301
306
295
297
331
342
344
Modeled Demand
(Vintaging Model)
272
277
281
283
283
280
278
276
271
Percent Difference
5%
-2%
-6%
-8%
-4%
-6%
-16%
-19%
-21%
• The Vintaging Model does not reflect the dynamic nature of reported HFC consumption, with significant
differences seen in each year. Whereas the Vintaging Model projects demand increasing or decreasing
slowly, with some annual fluctuations, actual consumption for specific chemicals or equipment may vary
over time and could even switch from positive to negative (indicating more chemical exported,
transformed, and destroyed than produced and imported in a given year). Furthermore, consumption as
calculated in the Vintaging Model is a function of demand not met by recovery of HFCs from equipment
that is being disposed. If, in any given year, a significant number of units are disposed, there will be a large
amount of additional recovery in that year that can cause an unexpected and not modeled decrease in
demand and thus a decrease in consumption. On the other hand, if market, economic, or other factors
cause less than expected disposal or recovery, actual supply would decrease, and hence consumption
would increase to meet that demand not satisfied by recovered quantities, increasing the GHGRP
amounts. EPA has published reclamation data, which would encompass a portion of the refrigerant
Industrial Processes and Product Use 4-145
-------
recovered annually. This data could be reviewed to determine if it can be used to improve the modeling of
these factors.
• The Vintaging Model is used to estimate the emissions that occur in the United States. As such, all
equipment or products that contain ODSs or alternatives, including saturated HFCs, are assumed to
consume and emit chemicals equally as like equipment or products originally produced in the United
States. The GHGRP data from Subpart 00 (industrial greenhouse gas suppliers) includes HFCs produced or
imported and used to fill or manufacture products that are then exported from the United States. The
Vintaging Model estimates of demand and supply are not meant to incorporate such chemical. Likewise,
chemicals may be used outside the United States to create products or charge equipment that is then
imported to and used in the United States. The Vintaging Model estimates of demand and supply are
meant to capture this chemical, as it will lead to emissions inside the United States. The GHGRP data from
Subpart QQ (supply of HFCs in products) accounts for most of these differences; however, the scope of
Subpart QQ does not cover all such equipment or products and the chemical contained therein.
Depending on whether the United States is a net importer or net exporter of such chemical, this factor
may account for some of the difference shown above or might lead to a further discrepancy.
One factor, however, would only lead to modeled estimates to be even higher than the estimates shown and
hence for some years possibly higher than GHGRP data:
• Saturated HFCs are also known to be used and emitted from other sources, such as electronics
manufacturing and magnesium production and processing. The Vintaging Model estimates here do not
include the amount of HFCs used for these applications, but rather only the amount used for applications
that traditionally were served by ODSs. Nonetheless, EPA expects the quantities of HFCs used for
electronics and magnesium production to be very small compared to the ODS substitute use for the years
analyzed. EPA estimates that electronics and magnesium production respectively consumed 0.8 MMT CO2
Eq. and 0.1 MMT CO2 Eq. of HFCs in 2019, which is much less than the ODS substitute sector in that year
(170 MMT CO2 Eq.) (U.S. EPA 2021b).
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
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 most other 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 from Hu et al. (2017) were used in this comparison. This provides a quality
check on the modeled emissions reported above. Hu et al. (2017) provided similar comparisons; here the EPA data
used in Hu et al. was updated to reflect the current Inventory estimates. Potential Inventory updates identified due
to the comparison with atmospheric data are noted in the Planned Improvements section below.
4-146 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Comparison of Results
Table 4-106 lists the emissions NOAA derived for the contiguous United States from atmospheric measurements as
described in Hu et al. (2017) and those from EPA's Vintaging Model. Figure 4-4 and Figure 4-5 below show this
information graphically. Specifically, the data compared are emissions of HFC-32, HFC-125 and HFC-143a (Table
4-106 and Figure 4-4) and emissions of HFC-134a (Table 4-106 and Figure 4-5) for the years covered in Hu et al.,
i.e., 2008 through 2014. In the Supplemental Information, Hu et al. (2017) provided uncertainty results
representing one standard deviation of the spread of several inversion calculations. These are provided in the
tables and figures below. There is also uncertainty in the EPA results. Overall, the uncertainty in EPA's total
Substitution of ODS emissions range from -3.4 percent to 14.0 percent (95 percent confidence interval), as shown
above. The nature of the model and the uncertainty analysis, however, does not allow EPA to provide specific
uncertainties to each species and hence comparisons below are to the EPA estimates without consideration of the
uncertainty involved in those estimates.
Table 4-106: U.S. Emissions of HFC-32, HFC-125, HFC-134a and HFC-143a (Gg)
Gas
Source
2008
2009
2010
2011
2012
2013
2014
HFC-32
NOAA
EPA
1.65±0.34
1.22
2.12±0.44
1.56
2.87±0.44
2.17
3.33±0.66
2.78
3.75±0.43
3.44
4.26±0.44
4.19
5.05±0.86
5.00
HFC-125
NOAA
EPA
7.05±1.68
5.02
6.52±1.52
6.05
7.91±1.37
7.22
7.92±1.29
8.34
7.79±0.85
9.37
8.79±1.05
10.41
9.77±1.40
11.43
HFC-134a
NOAA
EPA
49.14111.05
60.43
42.11±9.59
62.27
49.81±6.46
62.32
40.45±6.90
59.35
37.63±3.23
56.34
40.80±5.19
53.20
42.81±5.97
52.06
HFC-143a
NOAA
EPA
4.94±1.22
3.42
4.07±1.13
3.99
4.95±0.94
4.52
3.97±0.59
4.99
3.65±0.31
5.40
4.18±0.63
5.75
5.34±0.84
6.01
Note: NOAA uncertainty values represent one standard deviation
Figure 4-4: U.S. Emissions of HFC-32, HFC-125, and HFC-143a
¦HFC-32 (EPA)
HFC-125 (EPA)
• HFC-143a (EPA)
•HFC-32 (NOAA)
HFC-125 (NOAA)
¦ HFC-143a (NOAA)
2008
2009
2010
2011
2012
2013
2014
12 1
10 -
8 -
c. J
(5 b-
4 -
2 -
Industrial Processes and Product Use 4-147
-------
Figure 4-5: U.S. Emissions of HFC-134a
---HFC-134a(EPA) HFC-134a (NOAA)
2008 2009 2010 2011 2012 2013 2014
70 n
30 -
20 -
10 -
0 J 1 1 1 1 1 1 1 i
As shown, modeled estimates of HFC-32 are comparable with those derived from atmospheric measurements,
with only small differences (in Gg y"1). The modeled estimates for 2011 to 2014 lie within the one standard
deviation (1 s.d.) uncertainty range of the atmospherically derived estimates after 2010, and both datasets show a
similar trend of year-on-year increasing emissions, reaching ~5 Gg y1 by 2014. On the other hand, modeled
emissions of HFC-134a were consistently higher than those seen through atmospheric measurements, well above
the one standard deviation uncertainty. While the mean values from NOAA show year-to-year variability, the data
may suggest a slight downward trend in HFC-134a emissions through this entire period, like the modeled result;
however, confidence in the trend from atmospheric measurements is limited because the magnitude of
uncertainties are similar to the overall change and because increasing or decreasing trends of the mean values do
not persist for more than two years. The magnitude and time-dependence of the differences for both HFC-125 and
HFC-143a were similar, as results from this inventory model were lower in 2008 through 2010 and higher in 2011
through 2014, compared to the means estimated by NOAA. Considering the uncertainty ranges, the modeled
results for HFC-125 agree for the years 2009 to 2011, and those for HFC-143a agree for 2009 to 2010 and 2014.
The modeled results trend upward year-on-year for both gases, although the increase is smaller for HFC-143a. In
the NOAA estimates, no secular trend is discernable from 2008 to 2014 for HFC-143a considering the annual mean
uncertainties of approximately 18 percent. NOAA results for HFC-125 have similar uncertainty magnitudes but may
suggest a small increase in emissions over time, particularly in the latter years during this time interval.
Table 4-107 shows the differences in the emissions results from EPA's Vintaging Model and the mean results from
NOAA for those years where modeled estimates were not within the given 1 s.d. uncertainty range in the NOAA
results. Years when modeled estimates are within the uncertainty range reported by NOAA are not shown as those
differences are assumed to be insignificant. Because the uncertainty represents several inversion calculations, a
formally estimated 2 s.d. uncertainty range is not available. Instead, EPA considers twice the uncertainty estimated
by NOAA, which represents a range that is larger than the actual 2 s.d. from the spread of inversions. Emissions
differences found to be outside that range are shown in bold in the table, indicating more attention may be
warranted to understand these results. Comparing the results from the individual gases shows changes over time,
for example:
4-148 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
a. For HFC-32, while the difference was greater than 20 percent prior to 2011 compared to the mean
values, the difference from the 1 s.d. amounts averaged only 0.16 Gg during these three years. These
differences are insignificant at the twice uncertainty level. Results were within the 1 s.d. uncertainty
range of the NOAA estimates starting in 2011.
b. For HFC-125, the difference was within the uncertainty range of the NOAA estimates from 2009 to
2011, but greater than 15 percent of the mean values for 2008 (model results lower) and after 2011
(model results higher). All results were within the twice uncertainty range.
c. For HFC-134a, the differences ranged from 22 percent (in 2014) to 50 percent (in 2012), with all
modeled estimates higher than the NOAA estimates even when the 1 s.d. uncertainty ranges were
considered. For this gas, only the 2008, 2010, and 2014 estimates were within the NOAA estimates at
twice the uncertainty.
d. For HFC-143a, the modeled results were within the uncertainty range in 2009 to 2010 and again in
2014. The 2008 and 2011 model results were within the twice uncertainty range. For 2008, the
modeled results were below the uncertainty range by 31 percent compared to the mean value,
whereas for 2011 to 2013 the modeled results were above the uncertainty ranges, by an average of 37
percent compared to the mean values.
Table 4-107: Percentage Differences between EPA and NOAA HFC Emission Estimates
Year
HFC-32
HFC-125
HFC-134a
HFC-143a
2008
-0.43 (-26%)
-2.0 (-29%)
11.3 (23%)
-1.5 (-31%)
2009
-0.56 (-26%)
20.2 (48%)
2010
-0.70 (-25%)
12.5 (25%)
2011
18.9 (47%)
1.0 (26%)
2012
1.6 (20%)
18.7 (50%)
1.7 (48%)
2013
1.6 (18%)
12.4 (30%)
1.6 (37%)
2014
1.7 (17%)
9.3 (22%)
Average
-0.57 (-15%)
0.7 (7%)
14.8 (35%)
0.7 (20%)
Average of Absolute Values
26%
17.3 (21%)
14.8 (35%)
14.6 (36%)
Notes: Differences smaller than the 1 s.d. uncertainty on the annual NOAA-based estimates
are not shown. Differences greater than 2 s.d. shown in bold font. Uncertainties
associated with the Vintaging model have not been estimated by compound and year so
are not included and could imply fewer differences than shown in this table.
Discussion and Areas for Additional Research
The following are potential contributing factors to the variation between the results and possible ways these could
inform changes to the model that would reduce the differences seen.
• When examining the NOAA estimates with twice the uncertainties provided, only a few of these larger
differences from EPA model results are identified. The uncertainties in the NOAA estimates are primarily
driven by the frequency and spatial density of the atmospheric sampling, and the transport model
simulations. There is also inherent uncertainty in the consistency of the setup of each gas
chromatography measurement taken-e.g., variation in calibration, impurities in the carrier gas used,
among others (Barwick 1999); however, that uncertainty is likely less than 1 percent for HFC-125, HFC-
134a, and HFC-143a, and less than 5 percent for HFC-32. Given the magnitude of the uncertainties relative
to the size of any apparent emission changes, and the limited time period of the analysis, overall trends in
most of the gases are hard to discern with confidence except in the case of HFC-32. Although NOAA
estimates are derived from thousands of individual sample analyses (approximately 5,000 per year),
continued analysis and additional years will enable a better understanding of any secular trends in the
NOAA-derived estimates, and hence whether the modeled results are showing similar changes over time.
Industrial Processes and Product Use 4-149
-------
As discussed above, there is also uncertainty in the EPA estimates. Although these are not available by
individual species, these uncertainties may also explain some of the differences seen.
• A thorough discussion of the uncertainties and influencing factors in the NOAA estimates is provided in Hu
et al. (2017). That study notes that emissions estimated from inverse modeling of atmospheric data can
depend on assumed prior emission distributions and magnitudes, and accordingly the quoted
uncertainties on the NOAA results have been augmented to include these influences. In general, in a
region where there are fewer atmospheric observations, the NOAA results will inherently tend towards
the prior and be impacted by neighboring regions and populations (NOAA/EPA 2020). If the emissions or
emissions per person (depending on which prior is used) are significantly different in these areas
compared to the nearby areas, derived emissions for these regions can be biased.
• Uncertainty in atmospheric emission estimates is influenced by the number of NOAA's atmospheric
sampling sites, which changed between 2008 and 2014. Uncertainties were greatest in 2008 and 2009—
i.e., early on in the North American sampling program (Hu et al. 2017)—due to a fewer number of tower
sites and available measurements in those startup years. This may help explain why none of the EPA
results for 2008 were within one standard deviation of the NOAA estimates, but all were within twice the
uncertainty range. Also, changes in the number and location of measurement sites within the air sampling
network containing over 25 sites can lead to biases in the year-to-year emission estimates. During the
2008 to 2014 period addressed by Hu et al. (2017), measurements at four network sites began only after
2008, while observations at two others were terminated. Uncertainties related to network changes were
estimated with separate inversion runs in which sites were removed from the analysis and differences
ascertained. These influences contribute to the uncertainties quoted on the NOAA estimates, as do the
uncertainties related to meteorological models.
• The Vintaging Model estimated emissions for the entire United States, including all 50 states and
territories. Conversely, NOAA limits scope to the contiguous 48 states and the District of Columbia
(NOAA/EPA 2020). In that regard, EPA would expect the model to estimate slightly higher emissions than
those reported by NOAA, by roughly 2 percent based on population data (U.S. Census 2021). Activity data
for Hawaii, Alaska and territories could be researched. If available, calculations to reduce the bottom-up
results could be made and the results compared again to the NOAA results.
• For HFC-125 and HFC-143a, the EPA model suggests lower emissions in 2008 and higher in 2012 to 2013
relative to the atmosphere-derived estimates. Further research into the refrigeration market might
improve the agreement in the estimates for these two gases. As stated in the Introduction above,
emissions from the large retail food end-use (e.g., supermarkets), which uses both these gases, were
estimated to have the second highest contribution to the overall HFC emissions. Research in this industry
on the shift away from blends such as R-404A (which contains both HFC-125 and HFC-143a) or success in
lowering emission rates could be used to improve the bottom-up model.
• The modeled emissions of HFC-32 agreed well with the atmospheric inversion results in absolute terms,
and both data sets showing the same year-on-year increasing trend. Irrespective of the uncertainties,
slightly lower model results might imply that the model assumed a higher than actual use of "dry-charge"
residential AC equipment in lieu of R-410A (a 50:50 by mass ratio of HFC-32 and HFC-125). It might also
mean the actual emissions from R-410A equipment were slightly higher than modeled.
• The modeled inventory results for HFC-134a are complicated by an assumed decrease in emissions from
motor vehicle air conditioning (due to previous shifts towards lower charge sizes and emission rates, as
well as the on-going transition to HFO-1234yf) with concurrent increases in other sectors, such as for
foam blowing given the HCFC bans in foam blowing and other uses. Even though the NOAA results may
also suggest an increase from 2012 to 2014, the magnitudes of uncertainties prevent a robust conclusion
of emission increases over this period. While the inter-annual changes in the NOAA mean values for this
gas are small compared to the uncertainties, they are not inconsistent with the slow rate of increase
followed by a slow rate of decrease seen in the modeled emissions during 2008 through 2014. If the
model is overestimating the increased use in foam blowing and/or underestimating the decrease in
4-150 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
emissions from the motor vehicle air conditioning end-use, that might account for some of the differences
seen.
• In addition to its use as an ODS substitute, HFC-134a is used in a cover gas to prevent oxidation during
magnesium metal production and in semiconductor manufacturing. EPA's Vintaging Model does not
include these possible emission sources of HFC-134a, which, if included, would increase the difference
between the model-based result and NOAA's. The use and emissions of HFC-134a from these sources are
small (see above and elsewhere in this Inventory), so this would not be a significant contribution to the
comparison above.
• There are data limitations inherent in the bottom-up model. As described above, emissions are estimated
by applying assumed emission profiles to multiple end-uses, each of which can have thousands or millions
of individual uses in the United States. In some cases where equipment stocks or sales are unknown,
estimates are made using an average growth rate and by taking the most recent year where the starting
stock or sales of equipment is known, then accounting for equipment lifetimes, and subsequently
estimating the amount of equipment in future and/or preceding years where a value was not available.
Such assumptions are evident in the approximately constant slope of the EPA emission estimates for HFC-
32, HFC-125, and HFC-143a, compared to the more varying nature found in NOAA's mean results. Except
for HFC-32, which shows year-on-year increases across both sets of estimates, trends in the NOAA-
derived emissions are typically small relative to uncertainty magnitudes in measured data. Future work
could look at whether these variations might be consistent with other factors that influence emissions,
such as equipment installations, sales, or retirements, which could vary from year to year.
Using a Tier 2 bottom-up modeling methodology to estimate emissions requires assumptions and expert
judgment. Comparing the Vintaging Model's estimates to GHGRP-reported estimates of supply shown above and
emissions estimates derived from atmospheric measurements, particularly for more widely used chemicals, can
help validate the model but it is expected that the model will have limitations. 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. Likewise, these comparisons show reasonable agreement with atmospheric measurement derivations
of emissions, though certain chemicals and during certain years differences can be significant. Hence, areas for
further research that may improve the modeling are highlighted above. Despite the strengths and weaknesses of
three independent approaches for estimating emissions of these HFCs, the reasonable agreement noted here in
most instances provides added confidence in EPA's understanding of total U.S. emissions for these chemicals and
how they've change over time and, furthermore, has helped identify areas for potential improvement in the
future.
Recalculations Discussion
For the current Inventory, updates to the Vintaging Model included updating market size, manufacturing loss rate,
disposal loss rate, and post-life emission rate assumptions for various PU foam end-uses to align with market
research (EPA 2021c). Growth rates for window units were updated to align with sales data for Energy Star- and
non-Energy Star-certified units and a transition to HFC-32 was implemented beginning in 2015 to reflect
manufacturer transitions (EPA 2022b). In addition, HFC consumption for MDIs was updated to align with an
analysis of MDI sales in the United States (EPA 2022c). Together, these updates decreased ODS substitute
emissions on average by 0.03 MMT CO2 Eq. (0.004 percent) between 1990 and 2019.
Planned Improvements
Future improvements to the Vintaging Model are planned for the Fire Suppression and Aerosols sectors.
Specifically, streaming agent fire suppression lifetimes, market size, and growth rates 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
Industrial Processes and Product Use 4-151
-------
application-specific allowances for MDIs. EPA expects these revisions to be prepared for the 2023 or 2024
Inventory submission.
EPA has identified several updates to the Vintaging Model based on regularly published or released data that will
be implemented in the Vintaging Model on an annual basis, including updating growth rates for residential and
commercial unitary air-conditioning to align with annual sales estimates published by AHRI and for window units to
align with sales data released by EPA's Energy Star Program. In addition, as future application-specific allocations
for MDIs are granted, EPA will ensure the Vintaging Model is in alignment. Implemented updates are expected to
have a lagging effect on Inventory estimates (i.e., 2021 data published in 2022 will appear in the 2023 Inventory
submission) and will therefore not be explicitly discussed in the Recalculations Discussion.
As discussed above, future reporting under the AIM Act may provide useful information for verification purposes
and possible improvements to the Vintaging Model. 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 based on the comparison with atmospheric data.
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. Slightly lower model
results for HFC-32 might imply that the model assumed a higher than actual use of "dry-charge" residential AC
equipment in lieu of R-410A; EPA plans to investigate the amount of "dry-charge" AC imports during the 2023
Inventory cycle. 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, staring with the higher-
emitted HFCs identified above, in a future (i.e., 2024 or 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 GWP-weighted 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 3.8 MMT CO2 Eq. (0.2 kt) in 2020. This quantity represents an 84
percent decrease from the estimate for 1990 (see Table 4-108 and Table 4-109). 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.
4-152 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 2020. A recent examination of the SF6 emissions reported by electric
power systems to EPA's GHGRP revealed that SF6 emissions from reporters have decreased by 48 percent from
2011 to 2020,109 with much of the reduction seen from utilities that are not participants in the Partnership. These
utilities may be making relatively large reductions in emissions as they take advantage of relatively large and/or
inexpensive emission reduction opportunities (i.e., "low hanging fruit," such as replacing major leaking circuit
breakers) that Partners have already taken advantage of under the voluntary program (Ottinger et al. 2014). Total
emissions from electrical transmission and distribution in 2020 were lower than 2019 emissions, decreasing by
11.7 percent. The decrease in emissions may be attributed to a decrease in the average emission rate reported to
the GHGRP in 2020.
Table 4-108: SF6 and CF4 Emissions from Electric Power Systems and Electrical Equipment
Manufacturers (MMT CO2 Eq.)
Electrical
Electric Power
Equipment
Year
Systems
Manufacturers
Total
1990
22.8
0.3
23.2
2005
7.7
0.7
8.4
2016
3.8
0.2
4.1
2017
3.9
0.3
4.2
2018
3.6
0.3
3.8
2019
3.9
0.3
4.2
2020
3.3
0.5
3.8
Note: Totals may not sum due to independent rounding.
Table 4-109: SF6 and CF4 Emissions from Electric Power Systems and Electrical Equipment
Manufacturers (kt)
Year
SF6 Emissions
CF4 Emissions
1990
1.0
NO
2005
0.4
0.00032
2016
0.2
0.00004
2017
0.2
+
2018
0.2
NO
2019
0.2
0.00006
2020
0.2
0.00002
+ Does not exceed 0.000005 kt.
NO (Not Occurring)
109 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
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.
Industrial Processes and Product Use 4-153
-------
Methodology and Time-Series Consistency
The estimates of emissions from Electrical Transmission and Distribution are comprised of emissions from electric
power systems and emissions from the manufacture of electrical equipment. The methodologies for estimating
both sets of emissions are described below.
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.110 (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)111
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 13.6 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
110 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.
111 Nameplate capacity is defined as the amount of SF6 within fully charged electrical equipment.
4-154 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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,
atmospheric studies confirm that the downward trend in estimated global emissions between 1995 and 1998 was
real (see the Uncertainty discussion below).
1999 through 2020 Emissions from Electric Power Systems
Emissions from electric power systems from 1999 to 2020 were estimated based on: (1) reporting from utilities
participating in EPA's SF6 Emission Reduction Partnership for Electric Power Systems (Partners), which began in
1999; (2) reporting from utilities covered by EPA's GHGRP, which began in 2012 for emissions occurring in 2011
(GHGRP-Only Reporters); and (3) the relationship between utilities' reported emissions and their transmission
miles as reported in the 2001, 2004, 2007, 2010, 2013, and 2016 Utility Data Institute (UDI) Directories of Electric
Power Producers and Distributors (UDI 2001, 2004, 2007, 2010, 2013, and 2017), and 2019 and 2020 Homeland
Infrastructure Foundation-Level Data (HIFLD) (HIFLD 2019 and 2020), which was applied to the electric power
systems that do not report to EPA (Non-Reporters). Total U.S. Transmission mileage was interpolated between
2016 and 2019 to estimate transmission mileage of electric power systems in 2017 and 2018. (Transmission miles
are defined as the miles of lines carrying voltages above 34.5 kV).
Partners
Over the period from 1999 to 2020, Partner utilities, which for inventory purposes are defined as utilities that
either currently are or previously have been part of the Partnership,112 represented 49 percent, on average, of
total U.S. transmission miles. Partner utilities estimated their emissions using a Tier 3 utility-level mass balance
approach (IPCC 2006). If a Partner utility did not provide data for a particular year, emissions were interpolated
between years for which data were available or extrapolated based on Partner-specific transmission mile growth
rates. In 2012, many Partners began reporting their emissions (for 2011 and later years) through EPA's GHGRP
(discussed further below) rather than through the Partnership. In 2020, approximately 1 percent of the total
emissions attributed to Partner utilities were reported through Partnership reports. Approximately 99 percent of
the total emissions attributed to Partner utilities were reported and verified through EPA's GHGRP. Partners
without verified 2020 data accounted for less than 1 percent of the total emissions attributed to Partner
utilities.113
The GHGRP program has an "offramp" provision (40 CFR Part 98.2(i)) that exempts facilities from reporting under
certain conditions. If reported total greenhouse gas emissions are below 15,000 metric tons of carbon dioxide
equivalent (MT CO2 Eq.) for three consecutive years or below 25,000 MT CO2 Eq. for five consecutive years, the
facility may elect to discontinue reporting. GHGRP reporters that have off-ramped are extrapolated for three years
of non-reporting using a utility-specific transmission mile growth rate. After three consecutive years of non-
reporting, they are treated as non-reporters, as described in the section below on non-reporters. Partners that
have years of non-reporting between reporting years are gap filled by interpolating between reported values.
112 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.
113 Only data reported as of August 7, 2021 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.
Industrial Processes and Product Use 4-155
-------
GHGRP-Only Reporters
EPA's GHGRP requires users of SF6 in electric power systems to report emissions if the facility has a total SF6
nameplate capacity that exceeds 17,820 pounds. (This quantity is the nameplate capacity that would result in
annual SF6 emissions equal to 25,000 metric tons of C02 equivalent at the historical emission rate reported under
the Partnership.) As under the Partnership, electric power systems that report their SF6 emissions under EPA's
GHGRP are required to use the Tier 3 utility-level mass-balance approach. Many Partners began reporting their
emissions through EPA's GHGRP in 2012 (reporting emissions for 2011 and later years) because their nameplate
capacity exceeded the reporting threshold. Some Partners who did not report through EPA's GHGRP continued to
report through the Partnership.
In addition, many non-Partners began reporting to EPA for the first time through its GHGRP in 2012. Non-Partner
emissions reported and verified under EPA's GHGRP were compiled to form a new category of reported data
(GHGRP-Only Reporters). GHGRP-Only Reporters accounted for 17 percent of U.S. transmission miles and 24
percent of estimated U.S. emissions from electric power system in 2020.114
Emissions for GHGRP-only reporters that off-ramp are extrapolated for three years of non-reporting using a utility-
specific annual transmission mile growth rate. After three consecutive years of non-reporting, they are treated as
non-reporters, and emissions are subsequently estimated based on the methodology described below.
Non-Reporters
Emissions from Non-Reporters (i.e., utilities other than Partners and GHGRP-Only Reporters) in every year since
1999 were estimated using the results of a regression analysis that correlated emissions from reporting utilities
(using verified data from both Partners and GHGRP-Only Reporters) with their transmission miles.115 As noted
above, non-Partner emissions were reported to the EPA for the first time through its GHGRP in 2012 (representing
2011 emissions). This set of reported data was of particular interest because it provided insight into the emission
rate of non-Partners, which previously was assumed to be equal to the historical (1999) emission rate of Partners.
Specifically, emissions were estimated for Non-Reporters as follows:
• Non-Reporters, 1999 to 2011: First, the 2011 emission rates (per kg nameplate capacity and per
transmission mile) reported by Partners and GHGRP-Only Reporters were reviewed to determine whether
there was a statistically significant difference between these two groups. Transmission mileage data for
2011 was reported through GHGRP, with the exception of transmission mileage data for Partners that did
not report through GHGRP, which was obtained from UDI. It was determined that there is no statistically
significant difference between the emission rates of Partners and GHGRP-Only reporters; therefore,
Partner and GHGRP-Only reported data for 2011 were combined to develop regression equations to
estimate the emissions of Non-Reporters. Historical emissions from Non-Reporters were estimated by
linearly interpolating between the 1999 regression coefficient (based on 1999 Partner data) and the 2011
regression coefficient.
• Non-Reporters, 2012 to Present: The emissions data from Partners and by GHGRP-Only Reporters were
combined to develop regression equations for 2012. This was repeated for 2013 through 2020 using
Partner and GHGRP-Only Reporter data for each year.
o The 2020 regression equation for reporters was developed based on the emissions reported by a
subset of Partner utilities and GHGRP-Only utilities who reported non-zero emissions and non-zero
114 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
regression equations (discussed further below). These emissions are already implicitly accounted for in the relationship
between transmission miles and emissions.
115 In the United States, SF6 is contained primarily in transmission equipment rated above 34.5 kV.
4-156 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
transmission miles (representing approximately 62 percent of total U.S. transmission miles). The
regression equation for 2020 is:
Equation 4-24: Regression Equation for Estimating SF6 Emissions of Non-Reporting Facilities
Emissions (kg) = 0.186 x Transmission Miles
Table 4-110 below shows the percentage of transmission miles covered by reporters (i.e., associated with reported
data) and the regression coefficient for 1999 (the first year data was reported), and for 2011 through present (the
years with GHGRP reported data). The coefficient decreased between 2016 and 2020.
Table 4-110: Transmission Mile Coverage (Percent) and Regression Coefficients (kg per
1999
2005
2016
2017
2018
2019
2020
Percentage of Miles Covered by Reporters
50%
50%
72%
73%
72%
66%
62%
Regression Coefficient3
0.71
0.35
0.21
0.25
0.21
0.23
0.19
a Regression coefficient for emissions is calculated utilizing transmission miles as the explanatory variable and emissions
as the response variable. The equation utilizes a constant intercept of zero. When calculating the regression coefficient,
outliers are also removed from the analysis when the standard residual for that reporter exceeds the value 3.0.
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 and 2020, non-reporter
transmission mileage was derived by subtracting reported transmission mileage data from the total U.S.
transmission mileage from 2019 and 2020 HIFLD Data (HIFLD 2019 and 2020). 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 2019 was calculated to be 0.9 percent, as
transmission miles increased by approximately 19,900 miles.
• The annual transmission mile growth rate for 2019 through 2020 was calculated to be 0.06 percent, as
transmission miles increased by approximately 420 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.
Total Industry Emissions
As a final step, total electric power system emissions from 1999 through 2020 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, and the non-
reporting utilities' emissions (determined using the regression equations).
Industrial Processes and Product Use 4-157
-------
1990 through 2020 Emissions from Manufacture of Electrical Equipment
Three different methods were used to estimate 1990 to 2020 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 2020 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.
Methodological approaches were applied to the entire time series to ensure time-series consistency from 1990
through 2020.
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
uncertainty of all Partner-reported data was estimated to be 6.0 percent. The uncertainty associated with
extrapolated or interpolated emissions from non-reporting Partners was assumed to be 20 percent.
4-158 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
For GHGRP-Only Reporters, reported SF6 data was assumed to have an uncertainty of 20 percent.116 Based on a
Monte Carlo analysis, the cumulative uncertainty of all GHGRP-Only reported data was estimated to be 8.5
percent.
There are two sources of uncertainty associated with the regression equations used to estimate emissions in 2019
from Non-Reporters: (1) uncertainty in the coefficients (as defined by the regression standard error estimate), and
(2) the uncertainty in total transmission miles for Non-Reporters. Uncertainties were also estimated regarding (1)
estimates of SF6 and CF4 emissions from OEMs reporting to EPA's GHGRP, and (2) the assumption on the percent
share of OEM emissions from OEMs reporting to EPA's GHGRP.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-111. Electrical
Transmission and Distribution SF6 and CF4 emissions were estimated to be between 3.2 and 4.5 MMT CO2 Eq. at
the 95 percent confidence level. This indicates a range of approximately 16 percent below and 18 percent above
the emission estimate of 3.8 MMT CO2 Eq.
Table 4-111: Approach 2 Quantitative Uncertainty Estimates for SF6 and CF4 Emissions from
Electrical Transmission and Distribution (MMT CO2 Eq. and Percent)
2020 Emission
Estimate
Uncertainty Range Relative to 2018 Emission Estimate3
Source
Gas
(MMT CO? Eq.)
(MMT C02
Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
sf6
Electrical Transmission
and
3.8
3.2
4.5
-16%
+18%
and Distribution
cf4
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
In addition to the uncertainty quantified above, there is uncertainty associated with using global SF6 sales data to
estimate U.S. emission trends from 1990 through 1999. However, the trend in global emissions implied by sales of
SFs appears to reflect the trend in global emissions implied by changing SF6 concentrations in the atmosphere. That
is, emissions based on global sales declined by 29 percent between 1995 and 1998 (RAND 2004), and emissions
based on atmospheric measurements declined by 17 percent over the same period (Levin et al. 2010).
Several pieces of evidence indicate that U.S. SF6 emissions were reduced as global emissions were reduced. First,
the decreases in sales and emissions coincided with a sharp increase in the price of SF6 that occurred in the mid-
1990s and that affected the United States as well as the rest of the world. A representative from DILO, a major
manufacturer of SF6 recycling equipment, stated that most U.S. utilities began recycling rather than venting SF6
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.
/erification
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).117 Based on the results
of the verification process, EPA follows up with facilities to resolve mistakes that may have occurred. The post-
116 Uncertainty is assumed to be higher for the GHGRP-Only category, because 2011 is the first year that those utilities have
reported to EPA.
117 GHGRP Report Verification Factsheet. See https://www.epa.gov/sites/production/files/2Q15-
07/documents/ghgrp verification factsheet.pdf.
Industrial Processes and Product Use 4-159
-------
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
The historical emissions estimated for this source category have undergone the following revisions for the period
1990 through 2019.
• GHGRP report resubmissions: Historical estimates for the period 2015 through 2019 were updated
relative to the previous report based on revisions to reported historical data in EPA's GHGRP. In addition,
EPA identified two facilities that merged with another reporting facility and another facility who reported
under one GHGRP ID in 2011 and switched their ID in subsequent years. Estimation methodologies were
revised for these four facilities.
• Transmission mileage update: Historical estimates for total transmission mileage relied on a growth rate
of UDI data from 2012 to 2017 to estimate total transmission mileage for 2019 and 2020. EPA used HIFLD
data to replace 2019 data and interpolated transmission mileage between 2016 and 2019 to estimate
2017 and 2018 total transmission mileage.
• CF4 emissions from OEMs: Previous inventories did not capture the emissions of CF4from OEMs. EPA used
GHGRP data to calculate CF4 emissions from 2011 through 2019 and used an average ratio of SF6
emissions to CF4 emissions in 2011 through 2013 to estimate CF4 emissions from 1991 through 2010.
As a result of the recalculations, SF6 emissions from electrical transmission and distribution decreased by 1.20
percent for 2019 relative to the previous report, and SF6 nameplate capacity decreased by 2.5 percent for 2019
relative to the previous report. On average, SF6 emission estimates for the entire time series decreased by
approximately 0.2 percent per year.
Planned Improvements
EPA plans to more closely examine the methodology used to estimate non-reporter emissions. The current
methodology uses a reporter emissions rate to estimate non-reporter emissions. However, the preliminary results
of research conducted by the National Oceanic Atmospheric Administration (Hu 2021) indicate that U.S. emissions
of SF6 are significantly higher than what is being estimated in the current inventory for emissions of SF6 from all
sources. Because emissions from non-reporting electric power systems are a significant source of uncertainty in
the current U.S. SF6 inventory, EPA will investigate whether the methodology for determining the emission rate for
non-reporters should be revised.
Additionally, as the information on the type of new and retiring equipment is collected through GHGRP reporting,
EPA expects this data to provide insight into the relative importance of the two types of equipment as potential
emission sources. Historically, hermetically sealed pressure equipment has been considered to be a relatively small
source of SF6 in the United States; however, better estimating its potential source of emissions upon end-of-life
(i.e., disposal emissions) is an area for further analysis.
4-160 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 2020 was approximately 15 kt (see Table 4-112).
Table 4-112: N2O Production (kt)
Year
kt
1990
16
2005
15
2016
15
2017
15
2018
15
2019
15
2020
15
Nitrous oxide emissions were 4.2 MMT CO2 Eq. (14 kt N2O) in 2020 (see Table 4-113). 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 4-113: N2O Emissions from N2O Product Usage (MMT CO2 Eq. and kt)
Year MMT CP2 Eq. Irt
1990 4.2 14
2005 4.2 14
2016 4.2 14
2017 4.2 14
2018 4.2 14
2019 4.2 14
2020 4.2 14
Industrial Processes and Product Use 4-161
-------
Methodology and Time-Series Consistency
Emissions from N2O product uses were estimated using the following equation:
Equation 4-25: N2O Emissions from Product Use
Epu = X Sa x ERa)
a
where,
Sa
ER;
a
P
N2O emissions from product uses, metric tons
Total U.S. production of N2O, metric tons
specific application
Share of N2O usage by application a
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 2019, 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 has changed
only slightly over the past decade. 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 unavailability of data, production estimates for years 2004 through 2019 were held constant at the
2003 value.
4-162 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 unavailability of data, the share of total quantity of N2O usage data for years 2004 through 2019 was
assumed to equal the 2003 value. The emissions rate 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 emissions rate for all other subcategories was obtained from communication with a N2O
industry expert (Tupman 2002). The emissions rate for the medical/dental subcategory was obtained from the
2006IPCC Guidelines.
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2020.
Uncertainty
The overall uncertainty associated with the 2020 N2O emission estimate from N2O product usage was calculated
using the 2006 IPCC Guidelines (2006) Approach 2 methodology. Uncertainty associated with the parameters used
to estimate N2O emissions include production data, total market share of each end use, and the emission factors
applied to each end use, respectively.
The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 4-114. Nitrous oxide
emissions from N2O product usage were estimated to be between 3.2 and 5.2 MMT CO2 Eq. at the 95 percent
confidence level. This indicates a range of approximately 24 percent below to 24 percent above the emission
estimate of 4.2 MMT CO2 Eq.
Table 4-114: Approach 2 Quantitative Uncertainty Estimates for N2O Emissions from N2O
Product Usage (MMT CO2 Eq. and Percent)
Source
Gas
2020 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT CO? Eq.)
(MMT CO? Eq.)
(%)
Lower Upper
Lower Upper
Bound Bound
Bound Bound
N20 from Product Uses
N20
4.2
3.2 5.2
-24% +24%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
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 2006 IPCC Guidelines, see the QA/QC and Verification Procedures section in the introduction of
the IPPU chapter.
Recalculations Discussion
No recalculations were performed for the 1990 through 2019 portion of the time series.
Planned Improvements
EPA recently initiated an evaluation of alternative production statistics for cross-verification and updating time-
series activity data, emission factors, assumptions, etc., and a reassessment of N2O product use subcategories that
accurately represent trends. This evaluation includes conducting a literature review of publications and research
Industrial Processes and Product Use 4-163
-------
that may provide additional details on the industry. This work remains ongoing, and thus far no additional sources
of data have been found to update this category.
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
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 UNFCCC118 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
2020 are reported in Table 4-115.
Table 4-115: NOx, CO, NMVOC, and SO2 Emissions from Industrial Processes and Product
Use (kt)
Gas/Source
1990
2005
2016
2017
2018
2019
2020
NOx
592
572
402
397
397
397
397
Mineral Industry
246
329
221
220
220
220
220
Other Industrial Processes3
105
125
80
80
80
80
80
Metal Industry
88
60
61
60
60
60
60
Chemical Industry
152
55
39
37
37
37
37
Product Usesb
1
3
1
1
1
1
1
118 See http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
4-164 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
CO
4,129
1,557
1,075
1,007
1,007
1,007
1,007
Metal Industry
2,395
752
468
425
425
425
425
Other Industrial Processes3
608
420
316
311
311
311
311
Mineral Industry
49
194
179
163
163
163
163
Chemical Industry
1,073
189
110
107
107
107
107
Product Usesb
5
2
1
1
1
1
1
NMVOCs
7,638
5,849
3,776
3,767
3,767
3,767
3,767
Product Usesb
5,216
3,851
2,721
2,696
2,696
2,696
2,696
Other Industrial Processes3
1,720
1,708
940
958
958
958
958
Chemical Industry
575
213
69
68
68
68
68
Mineral Industry
16
32
24
24
24
24
24
Metal Industry
111
45
22
20
20
20
20
S02
1,307
831
466
509
509
509
509
Other Industrial Processes3
123
226
186
243
243
243
243
Chemical Industry
269
228
104
101
101
101
101
Mineral Industry
250
215
91
87
87
87
87
Metal Industry
659
158
83
77
77
77
77
Product Usesb
6
3
2
1
1
1
1
+ 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 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 4-115, 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).119 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 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's TSD 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 2021b). A
quantitative uncertainty analysis was not performed.
119 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 4-165
-------
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: 2020 Agriculture Sector Greenhouse Gas Emission Sources
Agricultural Soil Management
Enteric Fermentation
Manure Management
Rice Cultivation
Urea Fertilization
Liming
Field Burning of Agricultural Residues
180 200
MMT CO2 Eq.
In 2020, the Agriculture sector was responsible for emissions of 594.7 MMT CO2 Eq.,1 or 9.9 percent of total U.S.
greenhouse gas emissions. Methane emissions from enteric fermentation and manure management represent
26.9 percent and 9.2 percent of total CFU emissions from anthropogenic activities, respectively. Of all domestic
animal types, beef and dairy cattle were the largest emitters of CH4. 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 74.2 percent. Rice
cultivation and field burning of agricultural residues were minor sources of CH4. Manure management and field
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 Fourth Assessment Report (AR4) GWP values. See the
Introduction chapter for more information.
Agriculture 5-1
-------
burning of agricultural residues were also small sources of N2O emissions. Urea fertilization and liming accounted
for 0.1 percent and 0.05 percent of total CO2 emissions from anthropogenic activities, respectively.
Table 5-1 and Table 5-2 present emission estimates for the Agriculture sector. Between 1990 and 2020, CChand
Cm emissions from agricultural activities increased by 8.1 percent and 16.9 percent, respectively, while N2O
emissions from agricultural activities fluctuated from year to year, but increased by 1.8 percent overall. Trends in
sources of agricultural emissions over the 1990 to 2020 time series are shown in Figure 5-2.
Figure 5-2: Trends in Agriculture Sector Greenhouse Gas Emission Sources
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 2019)
to ensure that the trend is accurate. This year's notable updates include (1) Enteric Fermentation: updated to use
Cattle Enteric Fermentation Model (CEFM) for all years; (2) Manure Management: Updated animal population
data, updated state animal population distribution methodology, updated animal population data, and updated
MCF for pastures to align with IPCC 2019 guidance; (3) Urea Fertilization: Updates to fertilizer consumption data
and calculation formula; (4) Liming: using recently acquired limestone and dolomite data. In total, the
improvements made to the Agriculture sector in this Inventory decreased greenhouse gas emissions by 5.8 MMT
CO2 Eq. (0.9 percent) in 2019. For more information on specific methodological updates, please see the
Recalculations discussions within the respective source category sections of this chapter.
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). In addition, 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 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 agriculture activities are occurring. See Annex 5 for more information on
EPA's assessment of the sources not included in this Inventory.
5-2 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 5-1: Emissions from Agriculture (MMT CO2 Eq.)
Gas/Source
1990
2005
2016
2017
2018
2019
2020
C02
7.1
7.9
7.8
8.0
7.3
7.6
7.7
Urea Fertilization
2.4
3.5
4.7
4.9
5.0
5.1
5.3
Liming
4.7
4.3
3.1
3.1
2.2
2.4
2.4
ch4
214.7
235.5
244.7
247.8
251.1
250.3
250.9
Enteric Fermentation
163.5
168.0
171.3
174.9
175.7
176.1
175.2
Manure Management
34.8
49.0
57.1
57.5
59.4
58.7
59.6
Rice Cultivation
16.0
18.0
15.8
14.9
15.6
15.1
15.7
Field Burning of Agricultural Residues
0.4
0.4
0.4
0.4
0.4
0.4
0.4
n2o
330.1
330.3
349.4
347.5
358.4
365.0
336.1
Agricultural Soil Management
316.0
313.8
330.8
328.3
338.9
345.3
316.2
Manure Management
13.9
16.3
18.4
19.0
19.3
19.5
19.7
Field Burning of Agricultural Residues
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Total
551.9
573.6
601.9
603.2
616.7
622.9
594.7
Note: Totals may not sum due to independent rounding.
ible 5-2: Emissions from Agriculture (kt)
Gas/Source
1990
2005
2016
2017
2018
2019
2020
C02
7,084
7,854
7,761
7,977
7,267
7,553
7,657
Urea Fertilization
2,417
3,504
4,679
4,897
5,019
5,140
5,275
Liming
4,667
4,349
3,081
3,080
2,248
2,413
2,382
ch4
8,587
9,419
9,787
9,911
10,043
10,013
10,036
Enteric Fermentation
6,539
6,722
6,853
6,998
7,028
7,046
7,007
Manure Management
1,394
1,960
2,285
2,300
2,375
2,348
2,383
Rice Cultivation
640
720
631
596
623
602
630
Field Burning of Agricultural Residues
15
17
17
17
17
17
17
n2o
1,108
1,108
1,173
1,166
1,203
1,225
1,128
Agricultural Soil Management
1,060
1,053
1,110
1,102
1,137
1,159
1,061
Manure Management
47
55
62
64
65
65
66
Field Burning of Agricultural Residues
1
1
1
1
1
1
1
Note: Totals may not sum due to independent rounding.
Box 5-1: 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 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, follows this standardized format, and provides an explanation of the application of methods used to
calculate emissions from agricultural activities.
Agriculture 5-3
-------
5.1 Enteric Fermentation (CRF Source
Category 3A)
Methane is produced as part of normal digestive processes in animals. During digestion, microbes resident in an
animal's digestive system ferment food consumed by the animal. This microbial fermentation process, referred to
as enteric fermentation, produces Cm as a byproduct, which can be exhaled or eructated by the animal. The
amount of Cm produced and emitted by an individual animal depends primarily upon the animal's digestive
system, and the amount and type of feed it consumes.2
Ruminant animals (e.g., cattle, buffalo, sheep, goats, and camels) are the major emitters of Cm because of their
unique digestive system. Ruminants possess a rumen, or large "fore-stomach," in which microbial fermentation
breaks down the feed they consume into products that can be absorbed and metabolized. The microbial
fermentation that occurs in the rumen enables them to digest coarse plant material that non-ruminant animals
cannot. Ruminant animals, consequently, have the highest Cm emissions per unit of body mass among all animal
types.
Non-ruminant animals (e.g., swine, horses, and mules and asses) also produce Cm emissions through enteric
fermentation, although this microbial fermentation occurs in the large intestine. These non-ruminants emit
significantly less Cm on a per-animal-mass basis than ruminants because the capacity of the large intestine to
produce Cm is lower.
In addition to the type of digestive system, an animal's feed quality and feed intake also affect Cm emissions. In
general, lower feed quality and/or higher feed intake leads to higher Cm emissions. Feed intake is positively
correlated to animal size, growth rate, level of activity and production (e.g., milk production, wool growth,
pregnancy, or work). Therefore, feed intake varies among animal types as well as among different management
practices for individual animal types (e.g., animals in feedlots or grazing on pasture).
Methane emission estimates from enteric fermentation are provided in Table 5-3 and Table 5-4. Total livestock Cm
emissions in 2020 were 175.2 MMT CO2 Eq. (7,007 kt). Beef cattle remain the largest contributor of CH4 emissions
from enteric fermentation, accounting for 72 percent in 2020. Emissions from dairy cattle in 2020 accounted for 25
percent, and the remaining emissions were from horses, sheep, swine, goats, American bison, mules and asses.3
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).
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 of 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.
5-4 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 5-3: ChU Emissions from Enteric Fermentation (MMT CO2 Eq.)
Livestock Type
1990
2005
2016
2017
2018
2019
2020
Beef Cattle
118.5
124.7
122.6
125.8
126.0
126.5
125.3
Dairy Cattle
38.7
36.8
42.5
42.9
43.4
43.3
43.6
Swine
2.0
2.3
2.6
2.7
2.8
2.9
2.9
Horses
1.0
1.7
1.4
1.3
1.2
1.1
1.1
Sheep
2.6
1.4
1.2
1.2
1.2
1.2
1.2
Goats
0.6
0.7
0.6
0.6
0.6
0.6
0.6
American Bison
0.1
0.4
0.4
0.4
0.4
0.4
0.4
Mules and Asses
+
0.1
0.1
0.1
0.1
0.1
0.1
Total
163.5
168.0
171.3
174.9
175.7
176.1
175.2
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Table 5-4: ChU Emissions from Enteric Fermentation (kt)
Livestock Type
1990
2005
2016
2017
2018
2019
2020
Beef Cattle
4,742
4,986
4,905
5,033
5,042
5,062
5,013
Dairy Cattle
1,547
1,473
1,700
1,715
1,737
1,732
1,744
Swine
81
92
105
108
110
115
116
Horses
40
70
54
51
48
46
43
Sheep
102
55
48
47
47
47
47
Goats
23
26
24
24
24
25
25
American Bison
4
17
15
15
15
16
16
Mules and Asses
1
2
3
3
3
3
3
Total 6,539 6,722 6,853 6,998 7,028 7,046 7,007
Note: Totals may not sum due to independent rounding.
From 1990 to 2020, emissions from enteric fermentation have increased by 7.2 percent. From 2019 to 2020,
emissions decreased by 0.5 percent, largely driven by a decrease in cattle populations. While emissions generally
follow trends in cattle populations, over the long term there are exceptions. For example, while dairy cattle
emissions increased 12.7 percent over the entire time series, the population has declined by 3.6 percent, and milk
production increased 46 percent (USDA 2021a). These trends indicate that while emissions per head are
increasing, emissions per unit of product (i.e., meat, milk) are decreasing.
Generally, from 1990 to 1995 emissions from beef cattle increased and then decreased from 1996 to 2004. These
trends were mainly due to fluctuations in beef cattle populations and increased digestibility of feed for feed lot
cattle. Beef cattle emissions generally increased from 2004 to 2007, as beef cattle populations increased, and an
extensive literature review indicated a trend toward a decrease in feed digestibility for those years. Beef cattle
emissions decreased again from 2007 to 2014, as populations again decreased, but increased from 2015 to 2019,
consistent with another increase in population over those same years. Emissions and populations slightly declined
in 2020.
Emissions from dairy cattle generally trended downward from 1990 to 2004, along with an overall dairy cattle
population decline during the same period. Similar to beef cattle, dairy cattle emissions rose from 2004 to 2007
due to population increases and a decrease in feed digestibility (based on an analysis of more than 350 dairy cow
diets used by producers across the United States). Dairy cattle emissions have continued to trend upward since
2007, generally in line with dairy cattle population changes.
Regarding trends in other animals, populations of sheep have steadily declined, with an overall decrease of 54
percent since 1990. Horse populations are 8 percent greater than they were in 1990, but their numbers have been
declining by an average of 4 percent annually since 2007. Goat populations increased by about 20 percent through
2007, steadily decreased through 2012, then increased again, by about 1 percent annually, through 2020. Swine
Agriculture 5-5
-------
populations have trended upward through most of the time series, increasing 43 percent from 1990 to 2020. The
population of American bison more than quadrupled over the 1990 to 2020 time period, while the population of
mules and asses increased by a factor of five.
Methodology and Time-Series Consistency
Livestock enteric fermentation emission estimate methodologies fall into two categories: cattle and other
domesticated animals. Cattle, due to their large population, large size, and particular digestive characteristics,
account for the majority of enteric fermentation Cm emissions from livestock in the United States. A more detailed
methodology (i.e., IPCC Tier 2) was therefore applied to estimate emissions for all cattle. Emission estimates for
other domesticated animals (horses, sheep, swine, goats, American bison, and mules and asses) were estimated
using the IPCC Tier 1 approach, as suggested by the 2006 IPCC Guidelines (see the Planned Improvements section).
While the large diversity of animal management practices cannot be precisely characterized and evaluated,
significant scientific literature exists that provides the necessary data to estimate cattle emissions using the IPCC
Tier 2 approach. The Cattle Enteric Fermentation Model (CEFM), developed by EPA and used to estimate cattle Cm
emissions from enteric fermentation using IPCC's Tier 2 method, incorporates this information and other analyses
of livestock population, feeding practices, and production characteristics.
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. See Annex 3.10 for more detailed
information on the methodology and data used to calculate Cm emissions from enteric fermentation. In addition,
variables and the resulting emissions are also available at the state level in Annex 3.10.
Inventory Methodology for Cattle
National cattle population statistics were disaggregated into the following cattle sub-populations:
• Dairy Cattle
o Calves
o Heifer Replacements
o Cows
• Beef Cattle
o Calves
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
5-6 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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.
To estimate CFU emissions, the population was divided into state, age, sub-type (i.e., dairy cows and replacements,
beef cows and replacements, heifer and steer stockers, heifers and steers in feedlots, bulls, beef calves 4 to 6
months, and dairy calves 4 to 6 months), and production (i.e., pregnant, lactating) groupings to more fully capture
differences in Cm emissions from these animal types. The transition matrix was used to simulate the age and
weight structure of each sub-type on a monthly basis in order to more accurately reflect the fluctuations that
occur throughout the year. Cattle diet characteristics were then used in conjunction with Tier 2 equations from
IPCC (2006) to produce Cm emission factors for the following cattle types: dairy cows, beef cows, dairy
replacements, beef replacements, steer stockers, heifer stockers, steer feedlot animals, heifer feedlot animals,
bulls, and calves. To estimate emissions from cattle, monthly population data from the transition matrix were
multiplied by the calculated emission factor for each cattle type. More details are provided in Annex 3.10.
Non-Cattle Livestock
Emission estimates for other animal types were based on average emission factors (Tier 1 default IPCC emission
factors) representative of entire populations of each animal type. Methane emissions from these animals
accounted for a minor portion of total Cm emissions from livestock in the United States from 1990 through 2020.
Additionally, the variability in emission factors for each of these other animal types (e.g., variability by age,
production system, and feeding practice within each animal type) is less than that for cattle.
Annual livestock population data for 1990 to 2020 for sheep; swine; goats; horses; mules and asses; and American
bison were obtained for available years from USDA-NASS (USDA 2021a; USDA 2019). Horse, goat, and mule and ass
4 Due to inconsistencies in the 2003 literature values, the 2002 values were used for 2003 as well.
Agriculture 5-7
-------
population data were available for 1987,1992,1997, 2002, 2007, 2012, and 2017 (USDA 2019); the remaining
years between 1990 and 2020 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). 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 2020 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 of the
United Nations (FAO) 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 now 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 5-5.
Based on this analysis, enteric fermentation Cm emissions in 2020 were estimated to be between 155.9 and 206.7
MMT CO2 Eq. at a 95 percent confidence level, which indicates a range of 11 percent below to 18 percent above
the 2020 emission estimate of 175.2 MMT CO2 Eq.
5-8 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 5-5: Approach 2 Quantitative Uncertainty Estimates for ChU Emissions from Enteric
Fermentation (MMT CO2 Eq. and Percent)
2020 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
175.2
155.9 206.7
-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 2020 estimates.
c The overall uncertainty calculated in 2003, and applied to the 2020 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.
QA/QC and Verification
In order to ensure the quality of the emission estimates from enteric fermentation, the 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. Category-specific or Tier 2 QA procedures included
independent review of emission estimate methodologies from previous inventories.
As part of the quality assurance process, average implied emissions factors for U.S. dairy and beef cattle were
developed based on CEFM output and compared to emission factors for other countries provided by IPCC (2006).
This comparison is discussed in further detail in Annex 3.10.
Over the past few years, particular importance has been placed on harmonizing the data exchange between the
enteric fermentation and manure management source categories. The current Inventory now utilizes the transition
matrix from the CEFM for estimating cattle populations and weights for both source categories, and the CEFM is
used to output volatile solids and nitrogen excretion estimates using the diet assumptions in the model in
conjunction with the energy balance equations from the IPCC (2006). This approach facilitates the QA/QC process
for both of these source categories.
Recalculations Discussion
In the previous Inventory, 1990 to 2017 estimates were retained from the 1990 through 2017 Inventory, and 2018
and 2019 estimates were based on a simplified approach that used emission factors and extrapolated population
estimates for all animals. For the current Inventory, the CEFM was used for cattle for all years, resulting in different
estimates for 2018 and 2019 than the prior Inventory. For non-cattle livestock in the current Inventory, updated
Tier 1 estimates were calculated for 2018 and 2019, yielding different results than the simplified approach used for
these years in the prior Inventory.
For cattle, there were also changes to emissions resulting from activity data changes, including:
• The USDA published minor data revisions that EPA incorporated into the CEFM:
o Calf birth data were revised for 2013, 2016, and 2017;
o Dairy cow milk production values were updated for several states for 2013 through 2017;
o Cattle populations were revised for various states, depending on the year and cattle type, from 2014
through 2017;
o Cattle on feed population data were updated for 2014 through 2017;
Agriculture 5-9
-------
o Slaughter data were revised for 2017.
• EPA updated average milk fat in the CEFM from a constant IPCC default value of 4 percent to annual
average values of monthly milk fat values from USDA's Economic Research Services dairy data (USDA
2021b). These values ranged from 3.7 percent to 4.1 percent across the time series and are more
representative of U.S. livestock industry.
"Other" (non-cattle) livestock emissions were impacted by the following changes made between the current and
prior Inventories:
• USDA published updated animal population data that impacted 2018 emissions for poultry, sheep, and
swine.
• The manure management inventory updated how "other" state USDA combined populations were
distributed to their respective states (ERG 2021). See Section 5.2. These changes impacted emissions for
1990 through 2019 for both sheep and poultry, as well as 2010 through 2016 for swine.
All of these recalculations impacted the overall emission estimates between 0.5 and 1.4 percent over the time
series.
Planned Improvements
Regular annual data reviews and updates are necessary to maintain an emissions inventory that reflects the
current base of knowledge. EPA conducts the following list of regular annual assessments of data availability when
updating the estimates to extend time series each year:
• Further research to improve the estimation of dry matter intake (as gross energy intake) using data from
appropriate production systems;
• Updating input variables that are from older data sources, such as beef births by month, beef and dairy
annual calving rates, and beef cow lactation rates;
• Investigating the availability of data for dairy births by month, to replace the current assumption that
births are evenly distributed throughout the year;
• Investigating the availability of annual data for the DE, Ym, and crude protein values of specific diet and
feed components for grazing and feedlot animals;
• Further investigation on additional sources or methodologies for estimating DE for dairy cattle, given the
many challenges in characterizing dairy cattle diets;
• Further evaluation of the assumptions about weights and weight gains for beef cows, such that trends
beyond 2007 are updated, rather than held constant; and
• Further evaluation of the estimated weight for dairy cows (i.e., 1,500 lbs) that is based solely on Holstein
cows as mature dairy cow weight is likely slightly overestimated, based on knowledge of the breeds of
dairy cows in the United States.
Depending upon the outcome of ongoing investigations, future improvement efforts for enteric fermentation
could include some of the following options which are additional to the regular updates, and may or may not have
implications for regular updates once addressed:
• Potentially updating to a Tier 2 methodology for other animal types (i.e., sheep, swine, goats, horses);
efforts to move to Tier 2 will consider the emissions significance of livestock types;
• Investigation of methodologies and emission factors for including enteric fermentation emission
estimates from poultry;
• Comparison of the current CEFM with other models that estimate enteric fermentation emissions for
quality assurance and verification;
5-10 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
• Investigation of recent research implications suggesting that certain parameters in enteric models may be
simplified without significantly diminishing model accuracy; and
• Recent changes that have been implemented to the CEFM warrant an assessment of the current
uncertainty analysis; therefore, a revision of the quantitative uncertainty surrounding emission estimates
from this source category will be initiated. EPA plans to perform this uncertainty analysis following the
completed updates to the CEFM.
EPA is continuously investigating these recommendations and potential improvements and working with USDA and
other experts to utilize the best available data and methods for estimating emissions. Many of these
improvements are major updates and may take multiple years to implement in full.
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 (NH4+ 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
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.
Agriculture 5-11
-------
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. A very 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
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.
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 2020 were 59.6 MMT CO2 Eq. (2,383 kt); in 1990,
emissions were 34.8 MMT CO2 Eq. (1,394 kt). This represents a 71 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 44 and 122 percent, respectively. From
2019 to 2020, there was a 1 percent increase in total CH4 emissions from manure management, mainly due to an
increase in swine populations as well as an update to milk fat content which increased the average volatile solids
excretion for dairy cows.
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 2020, total N2O emissions from manure management were estimated to be 19.7 MMT CO2 Eq. (66 kt); in 1990,
emissions were 13.9 MMT CO2 Eq. (47 kt). These values include both direct and indirect N2O emissions from
manure management. Nitrous oxide emissions have increased since 1990. Small changes in N2O emissions from
individual animal groups exhibit the same trends as the animal group populations, with the overall net effect that
N2O emissions showed a 41 percent increase from 1990 to 2020 and a 0.9 percent increase from 2019 to 2020.
Overall shifts toward liquid systems have driven down the emissions per unit of nitrogen excreted as dry manure
handling systems have greater aerobic conditions that promote N2O emissions.
5-12 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 5-6 and Table 5-7 provide estimates of Cm and N2O emissions from manure management by animal
category.7
Table 5-6: ChU and N2O Emissions from Manure Management (MMT CO2 Eq.)
Gas/Animal Type
1990
2005
2016
2017
2018
2019
2020
CH4a
34.8
49.0
57.1
57.5
59.4
58.7
59.6
Dairy Cattle
14.3
23.6
30.8
31.2
32.0
30.9
31.7
Swine
15.5
20.3
21.1
21.0
22.0
22.3
22.4
Poultry
3.3
3.2
3.3
3.4
3.5
3.6
3.5
Beef Cattle
1.6
1.7
1.7
1.7
1.8
1.8
1.8
Horses
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Sheep
0.1
0.1
+
+
+
+
+
Goats
+
+
+
+
+
+
+
American Bison
+
+
+
+
+
+
+
Mules and Asses
+
+
+
+
+
+
+
N2Ob
13.9
16.3
18.4
19.0
19.3
19.5
19.7
Beef Cattle
5.9
7.2
8.5
8.9
9.1
9.2
9.4
Dairy Cattle
5.2
5.4
6.0
6.1
6.1
6.1
6.1
Swine
1.2
1.6
1.9
2.0
2.0
2.1
2.1
Poultry
1.4
1.6
1.6
1.6
1.7
1.7
1.7
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
48.8
65.3
75.5
76.5
78.7
78.2
79.2
+ 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.
Table 5-7: ChU and N2O Emissions from Manure Management (kt)
Gas/Animal Type
1990
2005
2016
2017
2018
2019
2020
CH4a
1,394
1,960
2,285
2,300
2,375
2,348
2,383
Dairy Cattle
572
943
1,232
1,248
1,278
1,237
1,269
Swine
621
812
846
840
882
891
895
Poultry
131
130
134
136
139
144
142
Beef Cattle
63
67
68
70
70
71
71
Horses
4
5
3
3
3
3
3
Sheep
3
2
2
2
2
2
2
Goats
+
+
+
+
+
+
+
American Bison
+
+
+
+
+
+
+
Mules and Asses
+
+
+
+
+
+
+
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
-------
N2Ob
Beef Cattle
Dairy Cattle
Swine
Poultry
Sheep
Horses
Goats
Mules and Asses
American Bisonc
+ 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 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 CH4 and N2O emissions from manure management.
See Annex 3.11 for more detailed information on the methodology (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 CH4 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 CH4 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
47
55
62
64
65
65
66
20
24
28
30
30
31
31
17
18
20
20
21
20
21
4
5
6
7
7
7
7
5
5
5
5
6
6
6
+
1
1
1
1
1
1
+
+
+
+
+
+
+
r..'-' ~
+
+
+
+
+
+
+
+
+
+
+
+
+
+
NA
NA
NA
NA
NA
NA
NA
5-14 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
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 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) were 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.
Agriculture 5-15
-------
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 (EFmnoff/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.8
• 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
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, and the emission factor for runoff and leaching (EFmnoff/ieach, in kg N2O per kg N), and the conversion
8 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.
5-16 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 data were from another data source, EPA updated with the more recent data
source to reflect the best available current data. 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.
Refer to the Recalculations section below for details on updates implemented to improve accuracy, consistency,
and/or completeness of the time series.
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 Cm 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 Cm 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 2020 emission estimates.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 5-8. Manure management
Cm emissions in 2020 were estimated to be between 48.8 and 71.5 MMT CO2 Eq. at a 95 percent confidence level,
which indicates a range of 18 percent below to 20 percent above the actual 2020 emission estimate of 59.6 MMT
CO2 Eq. At the 95 percent confidence level, N2O emissions were estimated to be between 16.5 and 24.4 MMT CO2
Eq. (or approximately 16 percent below and 24 percent above the actual 2020 emission estimate of 19.7 MMT CO2
Eq.).
Table 5-8: Approach 2 Quantitative Uncertainty Estimates for CH4 and N2O (Direct and
Indirect) Emissions from Manure Management (MMT CO2 Eq. and Percent)
2020 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
59.6
48.8 71.5
-18% +20%
Manure Management
n2o
19.7
16.5 24.4
-16% +24%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence
interval.
Agriculture 5-17
-------
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 Cm 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 5-9
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 Cm emissions.
Table 5-9: 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
2016
2017
2018
2019
2020
Dairy Cattle
48-112
29.3
53.0
65.4
66.0
67.3
65.6
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
12.1
11.6
12.0
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
NA (Not Applicable)
s Ranges reflect 2006 IPCC Guidelines (Volume 4, Table 10.14) default emission factors for North America across
different climate zones.
5-18 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
The manure management emission estimates include the following recalculations:
• EPA revised the methodology for population distribution to states where USDA population data are
withheld due to disclosure concerns (ERG 2021). The following animal population estimates were
impacted:
o Poultry in 1990 through 2018 (for several states that varied over the time series),
o Sheep in 1990 through 2018 (for several states that varied over the time series),
o Swine in 2010 through 2016 (for ID and WA).
• EPA updated the MCF for pasture to align with updated guidance from IPCC (2019).
• USDA updated raw animal population data which affected the following populations:
o Swine in 2018 for select states,
o Poultry in 2018 for select states,
o Sheep in 2018 for select states.
• The following data were updated in the CEFM which impacted emissions in the manure inventory (see
Section 5.1 for more details):
o Milk fat data for dairy and beef cows were updated for 1990 through 2018 which affected VS and
Nex for those animals and years,
o Crude protein data were updated which affected Nex for feedlot animals for 2015 through 2018.
o Annual calf birth and cattle population data were updated which impacted cattle populations for
2013 through 2018.
o USDA revised cattle populations for various states which changed cattle populations from 2014
through 2018.
The cumulative effect of these recalculations had a medium impact on the overall manure management emission
estimates. Over the time series:
• The average total emissions decreased by 4 percent from the previous Inventory. The changes ranged
from the smallest decrease, 2.6 percent (2.0 MMT CO2 Eq.), in 2017, to the largest decrease, 4.6 percent
(2.4 MMTCO2 Eq.), in 1990.
• The average Cm emissions decreased by 5 percent from the previous Inventory. The changes ranged
from the smallest decrease 3.7 percent (2.3 MMT CO2 Eq.), in 2018, to the largest decrease, 6.2 percent
(2.3 MMT CO2 Eq.), in 1990.
• The average N2O emissions decreased by 0.2 percent from the previous Inventory. The changes ranged
from the smallest decrease 0.8 percent (0.2 MMT CO2 Eq.), in 2018, to the largest increase, 1.8 percent
(0.3 MMT CO2 Eq.), in 2017.
Planned Improvements
Regular annual data reviews and updates are necessary to maintain an emissions inventory that reflects the
current base of knowledge. EPA conducts the following list of regular annual assessments of data availability when
updating the estimates to extend time series each year. EPA is actively pursuing the following updates for the
either the 2023 or 2024 Inventory submission:
Agriculture 5-19
-------
• 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:
• Revising the anaerobic digestion estimates to estimate CH4 emissions reductions due to the use of
anaerobic digesters (the Inventory currently estimates only emissions from anaerobic digestion systems).
• Investigating the updated IPCC 2019 Refinement default N2O emissions factor for anaerobic digesters.
Historically, EPA has not estimated N2O emissions from digesters as the default guidance was no
emissions. Incorporating AgSTAR data for N2O emissions, like CFU emissions, is a longer-term goal for EPA.
• Investigating updates to the current AD MCFs based on IPCC (2019).
• Determining if there are revisions to the U.S.-specific method for calculating liquid systems for MCFs
based on updated guidance from the IPCC 2019 Refinement.
• Investigating improved emissions estimate methodologies for swine pit systems with less than one month
of storage (the recently updated swine WMS data included this WMS category).
• Improving collaboration with the Enteric Fermentation source category estimates. For future Inventories,
it may be beneficial to have the CEFM and Manure Management calculations in the same model, as they
rely on much of the same activity data and on each other's outputs to properly calculate emissions.
• Revising the uncertainty analysis to address changes that have been implemented to the CH4 and N2O
estimates. EPA plans to align the timing of the updated Manure Management uncertainty analysis with
the uncertainty analysis for Enteric Fermentation.
EPA is actively pursuing the following updates but notes that implementation may be based on available resources
and data availability:
• Updating the Bo data used in the Inventory, as data become available. EPA is conducting outreach with
counterparts from USDA as to available data and research on Bo.
• Comparing CH4 and N2O emission estimates with estimates from other models and more recent studies
and compare the results to the Inventory.
• Comparing manure management emission estimates with on-farm measurement data to identify
opportunities for improved estimates.
• Comparing VS and Nex data to literature data to identify opportunities for improved estimates.
5.3 Rice Cultivation (CRF Source Category
3C)
Most of the world's rice is grown on flooded fields (Baicich 2013) that create anaerobic conditions leading to CH4
production through a process known as methanogenesis. Approximately 60 to 90 percent of the CH4 produced by
methanogenic bacteria in flooded rice fields is oxidized in the soil and converted to CO2 by methanotrophic
bacteria. The remainder is emitted to the atmosphere (Holzapfel-Pschorn et al. 1985; Sass et al. 1990) or
transported as dissolved CH4 into groundwater and waterways (Neue et al. 1997). Methane is transported to the
atmosphere primarily through the rice plants, but some CH4 also escapes via ebullition (i.e., bubbling through the
water) and to a much lesser extent by diffusion through the water (van Bodegom et al. 2001).
5-20 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Water management is arguably the most important factor affecting Cm emissions in rice cultivation, and improved
water management has the largest potential to mitigate emissions (Yan et al. 2009). Upland rice fields are not
flooded, and therefore do not produce CFU, but large amounts of CFUcan be emitted in continuously irrigated
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 exudates9 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 CFU emissions, particularly the use of fertilizers with sulfate (Wassmann et al.
2000b; Linquist et al. 2012), which can reduce CH4 emissions. 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 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 2020 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 5-10, Table 5-11, and Figure 5-3). Most emissions occur in Arkansas, California, Louisiana, Mississippi,
Missouri and Texas. In 2020, CH4 emissions from rice cultivation were 15.7 MMT CO2 Eq. (630 kt). Annual emissions
fluctuate between 1990 and 2020, 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 2020 are 2 percent lower
than emissions in 1990.
Table 5-10: ChU Emissions from Rice Cultivation (MMT CO2 Eq.)
State
1990
2005
2016
2017
2018
2019
2020
Arkansas
5.4
7.9
NE
NE
NE
NE
NE
California
3.3
3.4
NE
NE
NE
NE
NE
Florida
+
+
NE
NE
NE
NE
NE
Illinois
+
+
NE
NE
NE
NE
NE
Kentucky
+
+
NE
NE
NE
NE
NE
9 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 and use to stimulate more
production. The amount of root exudate produced by a rice plant over a growing season varies among rice varieties.
Agriculture 5-21
-------
Louisiana
2.6
2.8
NE
NE
NE
NE
NE
Minnesota
+
0.1
NE
NE
NE
NE
NE
Mississippi
1.1
1.4
NE
NE
NE
NE
NE
Missouri
0.6
1.1
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
3.0
1.3
NE
NE
NE
NE
NE
Total
16.0
18.0
15.8
14.9
15.6
15.1
15.7
+ Does not exceed 0.05 MMT C02 Eq.
NE (Not Estimated). State-level emissions are not estimated for 2016 through 2020 in this Inventory
because data are unavailable. 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.
Table 5-11: ChU Emissions from Rice Cultivation (kt)
State
1990
2005
2016
2017
2018
2019
2020
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
631
596
623
602
630
+ Does not exceed 0.5 kt.
NE (Not Estimated). State-level emissions are not estimated for 2016 through 2020 in this Inventory
because data are unavailable. 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.
5-22 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure 5-3: Annual CHU Emissions from Rice Cultivation, 2015
¦ >20
Note: Only riational-scaie emissions are estimated for 2016 through 2020 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 CH4 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 CH4, as well as CH4 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 CH4 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 CHa 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 crops and rotations, land uses, as well as
organic soils or cobbly, gravelly, and shaley mineral soiis.
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
Agriculture 5-23
-------
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).10
Rice cultivation areas are based on cropping 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 currently available through 2015
(USDA-NRCS 2018). The current Inventory only uses NRI data through 2015 because newer data are not available,
but will be incorporated when additional years of data are released by USDA-NRCS. The harvested rice areas in
each state are presented in Table 5-12.
Table 5-12: Rice Area Harvested (1,000 Hectares)
State/Crop
1990
2005
2016
2017
2018
2019
2020
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). State-level area data are not available for 2016 through 2020 but will be
added in a future Inventory with release of new NRI survey data.
Note: Totals may not sum due to independent rounding.
The Southeastern states have sufficient growing periods for a ratoon crop in some years (Table 5-13). 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.
10 See http://www.nrel.colostate.edu/proiects/ALUsoftware/.
5-24 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 5-13: Average Ratooned Area as Percent of Primary Growth Area (Percent)
State
1990-2015
Arkansas3
1%
California
0%
Florida15
49%
Louisiana0
32%
Mississippi3
1%
Missouri3
1%
Texasd
45%
a 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
for Texas, Louisiana, Florida, Missouri, or Mississippi. For these 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 2020 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.11 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
11 See https://quickstats.nass.usda.gov/.
Agriculture 5-25
-------
been compiled using the inventory methods described in this section. The model to extend the time series is
given by
Y=xp+ e,
where Y is the response variable (e.g., Cm emissions), xp is the surrogate data that is used to predict the
missing emissions data, 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 observed data for 1990 to 2015 using standard statistical techniques, and these estimates are used to
predict the missing emissions data for 2016 to 2020.
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 Cm
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).
In order to ensure time-series consistency, the same methods are applied from 1990 to 2015, and a linear
extrapolation method is used to approximate emissions for the remainder of the 2016 to 2020 time series based
on the emissions data from 1990 to 2015. This extrapolation method is consistent with data splicing methods in
IPCC (2006).
Uncertainty
Sources of uncertainty in the Tier 3 method include management practices, uncertainties in model structure (i.e.,
algorithms and parameterization), and variance associated with the NRI sample. Sources of uncertainty in the IPCC
(2006) Tier 1 method include the emission factors, management practices, and variance associated with the NRI
sample. A Monte Carlo analysis was used to propagate uncertainties in the Tier 1 and 3 methods. For 2016 to 2020,
there is additional uncertainty propagated through the Monte Carlo analysis associated with the surrogate data
method (See Box 5-2 for information about propagating uncertainty with the surrogate data method). The
uncertainties from the Tier 1 and 3 approaches are combined to produce the final Cm emissions estimate using
simple error propagation (IPCC 2006). Additional details on the uncertainty methods are provided in Annex 3.12.
Rice cultivation CFU emissions in 2020 were estimated to be between 4.0 and 27.5 MMT CO2 Eq. at a 95 percent
confidence level, which indicates a range of 75 percent below to 75 percent above the 2020 emission estimate of
15.7 MMT CO2 Eq. (see Table 5-14).
5-26 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 5-14: Approach 2 Quantitative Uncertainty Estimates for ChU Emissions from Rice
Cultivation (MMT CO2 Eq. and Percent)
Source
Inventory
Method
Gas
2020 Emission
Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.)
(MMT C02
Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Rice Cultivation
Tier 3
ch4
13.2
1.5
24.9
-88%
+88%
Rice Cultivation
Tier 1
ch4
2.5
1.3
3.7
-48%
+48%
Rice Cultivation
Total
ch4
15.7
4.0
27.5
-75%
+75%
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. Quality control measures include 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.
Model results are compared to field measurements to verify if results adequately represent Cm emissions. The
comparisons included over 17 long-term experiments, representing about 238 combinations of management
treatments across all the sites. A statistical relationship was developed to assess uncertainties in the model
structure, adjusting the estimates for model bias and assessing precision in the resulting estimates (methods are
described in Ogle et al. 2007). See Annex 3.12 for more information.
Recalculations Discussion
No recalculations were done in this Inventory.
Planned Improvements
A key planned improvement for rice cultivation is to fill several gaps in the management activity including
compiling new data on water management, organic amendments and ratooning practices in rice cultivation
systems. This improvement is expected to be completed for the next Inventory, but may not be prioritized
considering overall improvements to make best use of available resources.
5.4 Agricultural Soil Management (CRF
Source Category 3D)
Nitrous oxide is naturally produced in soils through the microbial processes of nitrification and denitrification that
is driven by the availability of mineral nitrogen (N) (Firestone and Davidson 1989).12 Mineral N is made available in
soils through decomposition of soil organic matter and plant litter, as well as asymbiotic fixation of N from the
12 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.
Agriculture 5-27
-------
atmosphere.13 Several agricultural activities increase mineral N availability in soils that lead to direct N2O
emissions at the site of a management activity (see Figure 5-4) (Mosier et al. 1998). These activities include
synthetic N fertilization; application of managed livestock manure; application of other organic materials such as
biosolids (i.e., treated sewage sludge); deposition of manure on soils by domesticated animals in pastures, range,
and paddocks (PRP) (i.e., unmanaged manure); retention of crop residues (N-fixing legumes and non-legume crops
and forages); and drainage of organic soils14 (i.e., Histosols) (IPCC 2006). Additionally, agricultural soil management
activities, including irrigation, drainage, tillage practices, cover crops, and fallowing of land, can influence N
mineralization from soil organic matter and levels of asymbiotic N fixation. Indirect emissions of N2O occur when N
is transported from a site and is subsequently converted to N2O; there are two pathways for indirect emissions: (1)
volatilization and subsequent atmospheric deposition of applied/mineralized N, and (2) surface runoff and leaching
of applied/mineralized N into groundwater and surface water.15 Direct and indirect emissions from agricultural
lands are included in this section (i.e., cropland and grassland as defined in Section 6.1 Representation of the U.S.
Land Base). Nitrous oxide emissions from Forest Land and Settlements soils are found in Sections 6.2 and 6.10,
respectively.
13 Asymbiotic N fixation is the fixation of atmospheric N2 by bacteria living in soils that do not have a direct relationship with
plants.
14 Drainage of organic soils in former wetlands enhances mineralization of N-rich organic matter, thereby increasing N20
emissions from these soils.
15 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.
5-28 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure 5-4: Sources and Pathways of N that Result in N2O Emissions from Agricultural Soil
Management
Sources and Pathways of N thai Result in N2O Emissions from Agricultural Soil Management
pr
Synthetic N Fertilizers
Synthetic N fertilizer applied to soil
II II 1 I!
\
Organic
Amendments
Includes both commercial and
non-co/nmercislfertilizers (i.e.,
animal manure compost,
sewage sludge tankage, etc.)
\
Urine and Dung from
Grazing Animals
Manure deposited on pasture range
and paddock
\
Crop Residues
Indudes above- and beiowground
residues for all crops (norvN andfsl-
fixing (and from perennial forage
crops and pastures following r^iewal
\
Mineralization of
Soil Organic Matter
Includes N converted to mineral form
upon decomposition of soil organic
matter
\
Asymbiotic Fixation
Fixation of atmospheric Nj by bacteria
living in soils that do not have a direct
relati onship with plants
N Rows:
N Inputs to
Managed Soils
Direct N2O
Emissions
N Volatilization
and Deposition
Indirect N20
Emissions
This graphic illustrates the sources and pathways of nitrogen that result
in direct and indirect N20 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.
Histosol
Cultivation
Agricultural soils produce the majority of N2O emissions in the United States. Estimated emissions in 2020 are
316.2 MMT CO2 Eq. (1,061 kt) (see Table 5-15 and Table 5-16). Annual N2O emissions from agricultural soils are 0.1
percent greater in 2020 compared to 1990, but emissions fluctuated between 1990 and 2020 due to inter-annual
variability largely associated with weather patterns, synthetic fertilizer use, and crop production. From 1990 to
2020, cropland accounted for 68 percent of total direct emissions on average from agricultural soil management,
while grassland accounted for 32 percent. On average, 78 percent of indirect emissions are from croplands and 22
percent from grasslands. Estimated direct and indirect N2O emissions by sub-source category are shown in Table
5-17 and Table 5-18.
Agriculture 5-29
-------
Table 5-15: N2O Emissions from Agricultural Soils (MMT CO2 Eq.)
Activity
1990
2005
2016
2017
2018
2019
2020
Direct
272.6
272.9
282.0
280.8
286.4
290.9
271.7
Cropland
186.0
183.8
191.0
190.6
195.2
196.5
187.4
Grassland
86.6
89.1
91.0
90.3
91.3
94.4
84.3
Indirect
43.5
40.9
48.9
47.4
52.5
54.4
44.6
Cropland
34.2
31.6
38.9
37.4
42.4
43.9
35.4
Grassland
9.2
9.3
10.0
10.0
10.1
10.5
9.2
Total
316.0
313.8
330.8
328.3
338.9
345.3
316.2
Notes: Estimates after 2015 are based on a data splicing method (See Methodology section). Totals
may not sum due to independent rounding.
Table 5-16: N2O Emissions from Agricultural Soils (kt)
Activity
1990
2005
2016
2017
2018
2019
2020
Direct
915
916
946
942
961
976
912
Cropland
624
617
641
640
655
659
629
Grassland
291
299
305
303
306
317
283
Indirect
146
137
164
159
176
182
150
Cropland
115
106
130
126
142
147
119
Grassland
31
31
34
34
34
35
31
Total
1,060
1,053
1,110
1,102
1,137
1,159
1,061
Notes: Estimates after 2015 are based on a data splicing method (See Methodology section). Totals may
not sum due to independent rounding.
Table 5-17: Direct N2O Emissions from Agricultural Soils by Land Use Type and N Input Type
(MMT COz Eq.)
Activity
1990
2005
2016
2017
2018
2019
2020
Cropland
185.9
183.7
191.0
190.5
195.1
196.5
187.3
Mineral Soils
182.1
180.0
187.6
187.1
191.7
193.1
183.9
Synthetic Fertilizer
58.9
60.1
65.7
65.4
66.8
67.1
63.8
Organic Amendment3
12.8
13.1
14.7
14.5
14.4
14.3
13.9
Residue Nb
40.7
41.0
41.2
41.2
42.3
42.7
40.8
Mineralization and
Asymbiotic Fixation
69.8
65.8
66.0
66.1
68.2
68.9
65.4
Drained Organic Soils
3.8
3.7
3.4
3.4
3.4
3.4
3.4
Grassland
86.7
89.2
91.0
90.3
91.3
94.4
84.3
Mineral Soils
84.2
86.7
88.5
87.8
88.8
91.9
81.9
Synthetic Fertilizer
+
+
+
+
+
+
+
PRP Manure
14.5
13.6
13.0
13.0
13.1
13.4
12.4
Managed Manurec
+
+
+
+
+
+
+
Biosolids (i.e., treated
Sewage Sludge)
0.2
0.5
0.6
0.6
0.6
0.7
0.7
Residue Nd
29.8
30.9
31.6
31.3
31.7
32.8
29.0
Mineralization and
Asymbiotic Fixation
39.6
41.8
43.3
42.9
43.4
45.0
39.8
Drained Organic Soils
2.5
2.4
2.5
2.5
2.5
2.5
2.5
Total
272.6
272.9
282.0
280.8
286.4
290.9
271.7
+ 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).
bCropland residue N inputs include N in unharvested legumes as well as crop residue N.
c Managed manure inputs include managed manure and daily spread manure amendments that are applied
to grassland soils.
5-30 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
d Grassland residue N inputs include N in ungrazed legumes as well as ungrazed grass residue N.
Notes: Estimates after 2015 are based on a data splicing method {See Methodology section). Totals may not
sum due to independent rounding.
Table 5-18: Indirect N2O Emissions from Agricultural Soils (MMT CO2 Eq.)
Activity
1990
2005
2016
2017
2018
2019
2020
Cropland
34.2
31.6
38.9
37.4
42.4
43.9
35.4
Volatilization & Atm.
Deposition
6.5
7.3
8.1
7.9
8.0
8.0
7.6
Surface Leaching & Run-Off
27.8
24.4
30.8
29.5
34.4
35.9
27.8
Grassland
9.2
9.3
10.0
10.0
10.1
10.5
9.2
Volatilization & Atm.
Deposition
3.6
3.7
3.5
3.6
3.6
3.7
3.5
Surface Leaching & Run-Off
5.6
5.6
6.5
6.4
6.5
6.8
5.7
Total
43.5
40.9
48.9
47.4
52.5
54.4
44.6
Notes: Estimates after 2015 are based on a data splicing method (See Methodology section). Totals may not
sum due to independent rounding.
Figure 5-5 and Figure 5-6 show regional patterns for direct NzO 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.
Figure 5-5: Croplands, 2015 Annual Direct N2O Emissions Estimated Using the Tier 3
DayCent Model
Note: Only national-scale emissions are estimated for 2016 to 2020 using a splicing method, and therefore the fine-scale
emission patterns in this map are based on Inventory data from 2015.
Agriculture 5-31
-------
Direct N2O emissions from croplands occur throughout all of the cropland regions but tend to be high in the
Midwestern Corn Belt Region (Illinois, Iowa, Indiana, Ohio, southern Minnesota and Wisconsin, and eastern
Nebraska), where a large portion of the land is used for growing highly fertilized corn and N-fixing soybean crops
(see Figure 5-5), Kansas, South Dakota and North Dakota have relatively high emissions from large areas of crop
production that are found in the Great Plains region. 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
through Nebraska, Snake River Valley in Idaho and the Central Valley in California, Direct emissions are low in
many parts of the eastern United States because only a small portion of land is cultivated, and in many western
states where rainfall and access to irrigation water are limited.
Direct emissions from grasslands are more evenly distributed throughout the United States (see Figure 5-6), but
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. However, there are relatively large emissions from local
areas in the eastern United States, particularly Kentucky and Tennessee, in addition to areas in Missouri and Iowa,
where there can be higher rates of Pasture/Range/Paddock (PRP) manure N additions on a relatively small amount
of pasture. These areas have greater stocking rates of livestock per unit of area, compared to other regions of the
United States.
Figure 5-6: Grasslands, 2015 Annual Direct N2O Emissions Estimated Using the Tier 3
DayCent Model
Note: Only national-scale emissions are estimated for 2016 to 2020 using a splicing method, and therefore the fine-scale
emission patterns in this map are based on Inventory data from 2015.
Indirect N2O emissions from volatilization in croplands have a similar pattern as the direct N2O emissions with
higher emissions in the Midwestern Corn Belt, Lower Mississippi River Basin and Great Plains. Indirect N2O
emissions from volatilization in grasslands are higher in the Southeastern United States, along with portions of the
Mid-Atlantic and southern Iowa, The higher emissions in this region are mainly due to large additions of PRP
manure N on relatively small but productive pastures that support intensive grazing, which in turn, stimulates NH3
volatilization.
5-32 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Indirect N: 0 emissions from surface runoff and leaching of applied/mineralized N in croplands is highest in the
Midwestern Corn Belt. There are also relatively high emissions associated with N management in the Lower
Mississippi River Basin, Piedmont region of the Southeastern United States and the Mid-Atlantic states. In addition,
areas of high emissions occur in portions of the Great Plains that have relatively large areas of irrigated croplands
with high leaching rates of applied/mineralized N. Indirect N:0 emissions from surface runoff and leaching of
applied/mineralized N in grasslands are higher in the eastern United States and coastal Northwest region. These
regions have greater precipitation and higher levels of leaching and runoff compared to arid to semi-arid regions in
the Western United States.
Figure 5-7: Croplands, 2015 Annual Indirect N2O Emissions from Volatilization Using the
Tier 3 DayCent Model
Note: Only national-scale emissions are estimated for 2016 to 2020 using a splicing method, and therefore the fine-scale
emission patterns in this map are based on Inventory data from 2015.
Agriculture 5-33
-------
Figure 5-8: Grasslands, 2015 Annual Indirect N2O Emissions from Volatilization Using the
Tier 3 DayCent Model
Note: Only national-scale emissions are estimated for 2016 to 2020 using a splicing method, and therefore the fine-scale
emission patterns in this map are based on Inventory data from 2015.
Figure 5-9: Croplands, 2015 Annual Indirect N2O Emissions from Leaching and Runoff Using
the Tier 3 DayCent Model
Note: Only national-scale emissions are estimated for 2016 to 2020 using a splicing method, and therefore the fine-scale
emission patterns in this map are based on Inventory data from 2015.
5-34 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure 5-10: Grasslands, 2015 Annual Indirect N2O Emissions from Leaching and Runoff
Using the Tier 3 DayCent Model
Note: Only national-scale emissions are estimated for 2016 to 2020 using a splicing method, and therefore the fine-scale
emission patterns in this map are based on Inventory data from 2015.
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 consistent with the managed land concept (IPCC 2006),
including direct and indirect N2O emissions from asymbiotic fixation16 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.
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 IPCC Tier 1 and 3 approaches, along with application of a splicing method for latter
16 N 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.
Agriculture 5-35
-------
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.
The Tier 3 approach is based on the crop and land use histories recorded in the USDA National Resources Inventory
(NRI) (USDA-NRCS 2018a). The NRI is a statistically-based sample of all non-federal land,17 and includes 349,464
points 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 175,527 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 points 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 2015 (USDA-NRCS
2018a).
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 hot moments of N2O
emissions (Wagner-Riddle et al. 2017). 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.
17 The NRI survey does include sample points on federal lands, but the program does not collect data from those sample
locations.
5-36 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
DayCent is used to estimate N2O emissions associated with production of 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 N2O emissions from other crops or rotations with other crops,18 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 from 2016 to 2020 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 2015 emissions that are derived using the Tier 3 method.
Surrogate data for these regression models includes corn and soybean yields from USDA-NASS statistics,19 and
weather data from the PRISM Climate Group (PRISM 2018). For the Tier 1 method, a linear-time series model is
used to estimate emissions from 2016 to 2020 without surrogate data. See Box 5-4 for more information about the
splicing method. Emission estimates for 2016 to 2020 will be recalculated in future Inventory reports when new
NRI data are available.
Box 5-4: Surrogate Data 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 modeled 1990 to 2015 emissions data,
which 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 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 from the emissions data for 1990 to 2015 using standard statistical techniques, and these
estimates are used in the model described above to predict the missing emissions data for 2016 to 2020.
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
18 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.
19 See https://quickstats.nass.usda.gov/.
Agriculture 5-37
-------
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 from 1990 to 2015). 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 from
2016 to 2020. Furthermore, the 95 percent confidence intervals are estimated using the 3 sigma rules assuming
a unimodal density (Pukelsheim 1994).
Tier 3 Approach for Mineral Cropland Soils
The DayCent biogeochemical model (Parton et al. 1998; Del Grosso et al. 2001 and 2011) is used to estimate direct
N2O emissions from mineral cropland soils that are managed for production of a wide variety of crops (see list in
previous section) based on the crop histories in the 2015 NRI (USDA-NRCS 2018a). Crops simulated by DayCent are
grown on approximately 85 percent of total cropland area in the United States. The model simulates net primary
productivity (NPP) using the NASA-CASA production algorithm MODIS Enhanced Vegetation Index (EVI) products,
MOD13Q1 and MYD13Q120 (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 Geographic Database (SSURGO) (Soil Survey Staff 2019). DayCent is used to
estimate direct N2O emissions due to mineral N available from the following sources: (1) application of synthetic
fertilizers; (2) application of livestock manure; (3) 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 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 from 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 specific NRI locations using
predictive mean matching methods that are 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. 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 artificial neural network. There are six complete imputations of
the management activity data using these methods.
20 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.
5-38 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 data collection 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). The 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 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).
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 2015. 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.
Nitrous oxide emissions 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 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). Further
elaboration on the methodology and data used to estimate N2O emissions from mineral soils are described in
Annex 3.12.
Agriculture 5-39
-------
In order to ensure time-series consistency, the DayCent model is applied from 1990 to 2015, and a linear
extrapolation method is used to approximate emissions for the remainder of the time series, 2016 to 2020, based
on the pattern in emissions data from 1990 to 2015 (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 mineral N
added to the soil, or 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. 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:
• 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 2016,
county-level fertilizer used on-farms is adjusted based on annual fluctuations in total U.S. fertilizer sales
(AAPFCO 2013 through 2021).21 After subtracting the portion of fertilizer applied to crops and grasslands
simulated by DayCent (see Tier 3 Approach for Mineral Cropland Soils and Direct N2O Emissions from
Grassland Soils sections for information on data sources), the remainder of the total fertilizer used on farms is
assumed to be applied to crops that are not simulated by DayCent.
• 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
21 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).
5-40 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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, which are
converted from mass of fertilizer to units of N using average organic fertilizer N content, which range between
2.3 to 4.2 percent across the time series (TVA 1991 through 1994; AAPFCO 1995 through 2021). Commercial
fertilizers do include dried manure and biosolids (i.e., treated sewage sludge), but the amounts are removed
from the commercial fertilizer data to avoid double counting22 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 2018a), crop production yield statistics (USDA-NASS 2019), 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 amount of soil mineral N from applied fertilizers and crop residues is 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 2015, and a linear
extrapolation method is used to approximate emissions from 2016 to 2020 based on the emission patterns
between 1990 and 2015 (See Box 5-4), with the exception of crop residue N, in which the data splicing method is
only used for 2019 and 2020. 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 for Direct N2O Emissions from Mineral Grassland Soils
As with N2O emissions from croplands, the Tier 3 process-based 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 2018a) 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. Mineral 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
22 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 2021), 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.
Agriculture 5-41
-------
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 61 percent of total PRP manure N in aggregate across the
country.23 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 grasslands. 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.
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 through the estimate 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 2015 based
on the Tier 1 and 3 methods, with the exception of biosolids (discussed below). In order to ensure time-series
consistency, emissions from 2016 to 2020 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 2015 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 2016 to 2020 will be recalculated in a future Inventory when the activity data are updated. Biosolids
application data are compiled through 2020 in this Inventory, and therefore soil N2O emissions and confidence
intervals are estimated using the Tier 1 method for all years in the time series 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 2015 NRI (USDA-NRCS 2018a) using soils data from the Soil Survey Geographic Database
(SSURGO) (Soil Survey Staff 2019). Temperature data from the PRISM Climate Group (PRISM 2018) 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). Annual NRI data are only available between 1990 and 2015, but the time series was adjusted using
data from the Forest Inventory and Analysis Program (USFS 2019) in order to estimate emissions from 2016 to
23 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.
5-42 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
2018. The land representation data have not been updated for this Inventory, so the amount of drained organic
soils is assumed to be the same in 2019 and 2020 as the estimated areas in 2018, and consequently the emissions
in 2019 and 2020 are also assumed to be the same as 2018. Further elaboration on the methodology and data
used to estimate N2O emissions from organic soils are described in Annex 3.12.
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 5-15 and Table 5-16).
Indirect N2O Emissions Associated with Nitrogen Management in Cropland and
Grasslands
Indirect N2O emissions occur when mineral 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, mineralization of soil
organic matter and residue, including N incorporated into crops and forage from symbiotic N fixation, 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.
Tier 1 and 3 Approaches for Indirect N2O Emissions from Atmospheric Deposition of Volatilized N
The Tier 3 DayCent model and IPCC (2006) Tier 1 methods are combined to estimate the amount of N that is
volatilized and eventually emitted as N2O. DayCent is used to estimate N volatilization for land areas whose direct
emissions are simulated with DayCent (i.e., most commodity and some specialty crops and most grasslands). The N
inputs included are the same as described for direct N2O emissions in the Tier 3 Approach for Mineral Cropland
Soils and Direct N2O Emissions from Grassland Soils sections. Nitrogen volatilization from all other areas is
estimated using the Tier 1 method with default IPCC fractions for N subject to volatilization (i.e., N inputs on
croplands not simulated by DayCent, PRP manure N excreted on federal grasslands, and biosolids [i.e., treated
sewage sludge] application on grasslands).
The IPCC (2006) default emission factor is multiplied by the amount of volatilized N generated from both DayCent
and Tier 1 methods to estimate indirect N2O emissions occurring following re-deposition of the volatilized N (see
Table 5-18). Further elaboration on the methodology and data used to estimate indirect N2O emissions are
described in Annex 3.12.
Tier 1 and 3 Approaches for Indirect N2O Emissions from Leaching/Runoff
As with the calculations of indirect emissions from volatilized N, the Tier 3 DayCent model and IPCC (2006) Tier 1
method are combined to estimate the amount of N that is subject to leaching and surface runoff into water bodies,
and eventually emitted as N2O. DayCent is used to simulate the amount of N transported from lands in the Tier 3
Approach. Nitrogen transport from all other areas is estimated using the Tier 1 method and the IPCC (2006) default
factor for the proportion of N subject to leaching and runoff associated with N applications on croplands that are
Agriculture 5-43
-------
not simulated by DayCent, applications of biosolids on grasslands, and PRP manure N excreted on federal
grasslands.
For both the DayCent Tier 3 and IPCC (2006) Tier 1 methods, nitrate leaching is assumed to be an insignificant
source of indirect N2O in cropland and grassland systems in arid regions, as discussed in IPCC (2006). In the United
States, the threshold for significant nitrate leaching is based on the potential evapotranspiration (PET) and rainfall
amount, similar to IPCC (2006), and is assumed to be negligible in regions where the amount of precipitation does
not exceed 80 percent of PET (Note: All irrigated systems are assumed to have significant amounts of leaching of N
even in drier climates).
For leaching and runoff data estimated by the Tier 3 and Tier 1 approaches, the IPCC (2006) default emission factor
is used to estimate indirect N2O emissions that occur in groundwater and waterways (see Table 5-18). Further
elaboration on the methodology and data used to estimate indirect N2O emissions are described in Annex 3.12.
In order to ensure time-series consistency, indirect soil N2O emissions are estimated using the Tier 1 and 3
approaches from 1990 to 2015 and then a linear extrapolation data splicing method, described in Box 5-4, is
applied to estimate emissions from 2016 to 2020 based on the emission patterns from 1990 to 2015. Linear
extrapolation is a standard data splicing method for estimating emissions at the end of a time series (IPCC 2006).
As with the direct N2O emissions, the time series will be recalculated in a future Inventory report when new
activity data are incorporated into the analysis.
Uncertainty
Uncertainty is estimated for each of the following five components of N2O emissions from agricultural soil
management: (1) direct emissions simulated by DayCent; (2) the components of indirect emissions (N volatilized
and leached or runoff) simulated by DayCent; (3) direct emissions estimated with the IPCC (2006) Tier 1 method;
(4) the components of indirect emissions (N volatilized and leached or runoff) estimated with the IPCC (2006) Tier
1 method; and (5) indirect emissions estimated with the IPCC (2006) Tier 1 method. Uncertainty in direct emissions
as well as the components of indirect emissions that are estimated from DayCent are derived from a Monte Carlo
Analysis (consistent with IPCC Approach 2), addressing uncertainties in model inputs and structure (i.e., algorithms
and parameterization) (Del Grosso et al. 2010). For 2016 to 2020, there is additional uncertainty propagated
through the Monte Carlo Analysis associated with the splicing method (See Box 5-4).
Simple error propagation methods (IPCC 2006) are used to derive confidence intervals for direct emissions
estimated with the IPCC (2006) Tier 1 method, the proportion of volatilization and leaching or runoff estimated
with the IPCC (2006) Tier 1 method, and indirect N2O emissions. Uncertainty in the splicing method is also included
in the error propagation for 2016 to 2020 (see Box 5-4). Additional details on the uncertainty methods are
provided in Annex 3.12.
Table 5-19 shows the combined uncertainty for direct soil N2O emissions. The estimated emissions ranges from 33
percent below to 33 percent above the 2020 emission estimate of 271.7 MMT CO2 Eq. The combined uncertainty
for indirect soil N2O emissions ranges from 67 percent below to 145 percent above the 2020 estimate of 44.6 MMT
CO2 Eq.
Table 5-19: Quantitative Uncertainty Estimates of N2O Emissions from Agricultural Soil
Management in 2020 (MMT CO2 Eq. and Percent)
2020 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
271.7
183.2
360.2
-33% 33%
Indirect Soil N20 Emissions
n2o
44.6
14.9
109.0
-67% 145%
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
5-44 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
uncertainty will be included in future Inventory reports, which is a standard data splicing method for estimating
emissions at the end of a time series (IPCC 2006).
Additional uncertainty is associated with an incomplete estimation of N2O emissions from managed croplands and
grasslands in Hawaii and Alaska. The Inventory currently includes the N2O emissions from managed manure and
PRP N additions in Alaska and Hawaii, and drained organic soils in Hawaii. Land areas used for agriculture in Alaska
and Hawaii are small relative to major crop commodity states in the conterminous United States, so the emissions
are likely to be small for the other sources of N (e.g., synthetic fertilizer and crop residue inputs), which are not
currently included in the Inventory.
/erification
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. DayCent results for N2O emissions and NO3" leaching are compared with field data
representing various cropland and grassland systems, soil types, and climate patterns (Del Grosso et al. 2005; Del
Grosso et al. 2008), and further evaluated by comparing the model results to emission estimates produced using
the IPCC (2006) Tier 1 method for the same sites. Nitrous oxide measurement data for cropland are available for
64 sites representing 796 different combinations of fertilizer treatments and cultivation practices, and
measurement data for grassland are available for 13 sites representing 36 different management treatments.
Nitrate leaching data are available for 12 sites, representing 279 different combinations of fertilizer treatments and
tillage practices. In general, DayCent predicted N2O emission and nitrate leaching for these sites reasonably well.
See Annex 3.12 for more detailed information about the comparisons.
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. Small input errors were found in the amount of synthetic and
managed manure N applied to soils in the Tier 3 model simulations. Corrections were made by adjusting the
amount of N applied to the maximum available by state and N2O emissions were reduced based on the latest N
Fertilizer sales data and managed manure N available for application to soils. Databases containing input data,
emission factors, and calculations required for the Tier 1 method have been checked and updated as needed. In
particular, the amount of synthetic and managed manure N included in the Tier 1 analysis was also adjusted in this
process so that the total synthetic and managed manure N was equal to the amounts reported by the activity data
sources. Links between spreadsheets have also been checked, updated, and corrected when necessary.
Recalculations Discussion
Three improvements have been implemented in this Inventory leading to the need for recalculations. Updated
synthetic N fertilizer sales data were available for 2015 and new sales data for 2016 were published, both
incorporated into Tier 3 and Tier 1 analyses (AAPFCO 2021). Additionally, updates to the time series of PRP N and
manure N available for application to soils were incorporated into the analysis. The surrogate data method was
also applied to re-estimate N2O emissions from N fertilizer applications for 2016 to 2020. Finally, errors in the
previous Inventory were corrected where the amount of synthetic fertilizer or managed manure N were over-
applied in the Tier 3 analysis resulting in more N than was available for application. These changes resulted in an
average increase in emissions of 0.2 percent from 1990 to 2019 relative to the previous Inventory.
Planned Improvements
A key improvement for a future Inventory will be to incorporate additional management activity data from the
USDA-NRCS Conservation Effects Assessment Project survey. This survey has compiled new data in recent years
that will be available for the Inventory analysis by next year. The latest land use data will also be incorporated from
the USDA National Resources Inventory and related management data from USDA-ERS ARMS surveys.
Agriculture 5-45
-------
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. Model development is
underway to represent the influence of enhanced efficiency fertilizers, which include stabilized fertilizers (e.g.,
nitrification inhibitors and urease inhibitors), slow-release fertilizers (e.g., methylene urea or sulfur coated urea),
and controlled release fertilizers (e.g., polymer-coated fertilizers), on N2O emissions. Experimental study sites will
continue to be added for quantifying model structural uncertainty. Studies that have continuous (daily)
measurements of N2O (e.g., Scheer et al. 2013) will be given priority. Other suggested improvements identified
through public review are being evaluated for future Inventory submissions.
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 5.5). Alaska and Hawaii are not included for all sources in the current Inventory for agricultural soil
management, with the exception of N2O emissions from drained organic soils in croplands and grasslands for
Hawaii, managed manure N and PRP N additions for grasslands in Alaska and Hawaii. There is also an improvement
based on updating the Tier 1 emission factor for N2O emissions from drained organic soils by using the revised
factor in the 2013 Supplement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands
(IPCC 2013).
In addition, there is a planned improvement associated with implementation of the Tier 1 method. Specifically, soil
N2O emissions will be estimated and reported for N mineralization from soil organic matter decomposition that is
accelerated with Forest Land Converted to Cropland and Grassland Converted to Cropland. A review of available
data on biosolids (i.e., treated sewage sludge) application will also be undertaken to improve the distribution of
biosolids application on croplands, grasslands and settlements.
These improvements are expected to be completed for the next Inventory (i.e., 2023 submission to the UNFCCC,
1990 through 2021 Inventory). However, the timeline may be extended if there are insufficient resources to fund
all or part of these planned improvements.
5.5 Liming (CRF Source Category 3G)
Crushed limestone (CaCOs) and dolomite (CaMg(CC>3)2) are added to soils by land managers to increase soil pH
(i.e., to reduce acidification). Carbon dioxide emissions occur as these compounds react with hydrogen ions in
soils. The rate of degradation of applied limestone and dolomite depends on the soil conditions, soil type, climate
regime, and whether limestone or dolomite is applied. Emissions from limestone and dolomite that are used in
industrial processes (e.g., cement production, glass production, etc.) are reported in the IPPU chapter. Emissions
from liming of soils have fluctuated between 1990 and 2020 in the United States, ranging from 2.2 MMT CO2 Eq. to
6.0 MMT CO2 Eq. across the entire time series. In 2020, liming of soils in the United States resulted in emissions of
2.4 MMT CO2 Eq. (0.6 MMT C), representing a 49 percent decrease in emissions since 1990 (see Table 5-20 and
Table 5-21). The trend is driven by variation in the amount of limestone and dolomite applied to soils over the time
period.
Table 5-20: Emissions from Liming (MMT CO2 Eq.)
Source
1990
2005
2016
2017
2018
2019
2020
Limestone
4.1
3.9
2.8
2.9
2.0
2.2
2.2
Dolomite
0.6
0.4
0.3
0.2
0.2
0.2
0.2
Total
4.7
4.3
3.1
3.1
2.2
2.4
2.4
Note: Totals may not sum due to independent rounding.
5-46 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 5-21: Emissions from Liming (MMT C)
Source 1990
2005
2016 2017 2018 2019 2020
Limestone 1.1
Dolomite 0.2
1.1
0.1
0.8 0.8 0.6 0.6 0.6
0.1 0.1 0.1 0.1 0.1
Total 1.3
1.2
0.8 0.8 0.6 0.7 0.6
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
Carbon dioxide emissions from application of limestone and dolomite to soils were estimated using a Tier 2
methodology consistent with IPCC (2006). The annual amounts of limestone and dolomite, which are applied to
soils (see Table 5-22), were multiplied by CO2 emission factors from West and McBride (2005). These country-
specific emission factors (0.059 metric ton C/metric ton limestone, 0.064 metric ton C/metric ton dolomite) are
lower than the IPCC default emission factors because they account for the portion of carbonates that are
transported from soils through hydrological processes and eventually deposited in ocean basins (West and
McBride 2005). This analysis of lime dissolution is based on studies in the Mississippi River basin, where the vast
majority of lime application occurs in the United States (West 2008). Moreover, much of the remaining lime
application is occurring under similar precipitation regimes, and so the emission factors are considered a
reasonable approximation for all lime application in the United States (West 2008) (See Box 5-5).
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), as well as preliminary data that will
eventually be published in the Minerals Yearbook for the latter part of the time series (Willett 2019, 2020b, 2021a,
2021b). Data for the final year of the inventory is based on the Mineral Industry Surveys, as discussed below (USGS
2021). 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 2020 U.S. emission estimate from liming of soils is 2.4 MMT CO2
Eq. using the country-specific factors. In contrast, emissions would be estimated at 4.8 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
Agriculture 5-47
-------
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, 2019, and 2020 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 2019 and 2020
data, 2018 and 2019 fractions were applied to the 2019 and 2020 estimates of total crushed stone, respectively.
The basis for these estimates is from the USGS Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the
Fourth Quarter of 2020 and First Quarter of 2021 (USGS 2020; USGS 2021).
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.
Table 5-22: Applied Minerals (MMT)
Mineral
1990
2005
2016
2017
2018
2019
2020
Limestone
19.0
18.1
13.0
13.4
9.4
10.1
9.9
Dolomite
2.4
1.9
1.1
0.8
0.9
1.0
1.0
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2019. In addition, the same methods are applied throughout the time series, and the activity data are
extended in the last two years of the time series based on proportions of specified, unspecified and estimated
agricultural limestone and dolomite so that estimates are consistent with the previous year's data. These years
will be recalculated when additional data are available on the amounts of limestone and dolomite that are used for
agricultural purposes.
Uncertainty
Uncertainty regarding the amount of limestone and dolomite applied to soils was estimated at ±15 percent with
normal densities (Tepordei 2003; Willett 2013b). Analysis of the uncertainty associated with the emission factors
included the fraction of lime dissolved by nitric acid versus the fraction that reacts with carbonic acid, and the
portion of bicarbonate that leaches through the soil and is transported to the ocean. Uncertainty regarding the
time associated with leaching and transport was not addressed in this analysis, but is assumed to be a relatively
small contributor to the overall uncertainty (West 2005). The probability distribution functions for the fraction of
lime dissolved by nitric acid and the portion of bicarbonate that leaches through the soil were represented as
triangular distributions between ranges of zero and 100 percent of the estimates. The uncertainty surrounding
these two components largely drives the overall uncertainty.
A Monte Carlo (Approach 2) uncertainty analysis was applied to estimate the uncertainty in CO2 emissions from
liming. The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 5-23. Carbon
dioxide emissions from carbonate lime application to soils in 2020 were estimated to be between -0.26 and 4.73
MMT CO2 Eq. at the 95 percent confidence level. This confidence interval represents a range of 111 percent below
to 97 percent above the 2020 emission estimate of 2.4 MMT CO2 Eq. Note that there is a small probability of a
negative emissions value leading to a net uptake of CO2 from the atmosphere. Net uptake occurs due to the
dominance of the carbonate lime dissolving in carbonic acid rather than nitric acid (West and McBride 2005).
5-48 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 5-23: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Liming
(MMT CO2 Eq. and Percent)
Source
Gas
2020 Emission Estimate
(MMT C02 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Liming
C02
2.4
(0.26) 4.73
-111% 97%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
QA/QC and Verification
A source-specific QA/QC plan for liming has been developed and implemented, consistent with the U.S. Inventory
QA/QC plan outlined in Annex 8. The quality control effort focused on the Tier 1 procedures for this Inventory. A
small error was found in results for Florida and Georgia due to incorrect cell references in the spreadsheet, which
was corrected. No other errors were found.
Recalculations Discussion
Limestone and dolomite application data for 2019 were updated with the recently acquired data from Willett, J.C.
(2021a), rather than approximated by a ratio method, which was used in the previous Inventory. There were also
corrections in cell references for Florida and Georgia. With these revisions, the emissions decreased by 1.2 percent
for 2019 relative to the previous Inventory.
5.6 Urea Fertilization (CRF Source Category
3H)
The use of urea (COfNFhh) as a fertilizer leads to greenhouse gas emissions through the release of C02that was
fixed during the production of urea. In the presence of water and urease enzymes, urea that is applied to soils as
fertilizer is converted into ammonium (NhV), hydroxyl ion (OH), and bicarbonate (HCO3 ). The bicarbonate then
evolves into CO2 and water. Emissions from urea fertilization in the United States were 5.3 MMT CO2 Eq. (1.5 MMT
C) in 2020 (Table 5-24 and Table 5-25). Carbon dioxide emissions have increased by 118 percent between 1990 and
2020 due to an increasing amount of urea that is applied to soils. The variation in emissions across the time series
is driven by differences in the amounts of fertilizer applied to soils each year. Carbon dioxide emissions associated
with urea that is used for non-agricultural purposes are reported in the IPPU chapter (Section 4.6).
Table 5-24: CO2 Emissions from Urea Fertilization (MMT CO2 Eq.)
Source
1990
2005
2016
2017
2018
2019
2020
Urea Fertilization
2.4
3-5
4.7
4.9
5.0
5.1
5.3
Table 5-25: CO2 Emissions from Urea Fertilization (MMT C)
Source
1990
2005
2016
2017
2018
2019
2020
Urea Fertilization
0.7
1.0
1.3
1.3
1.4
1.4
1.4
Agriculture 5-49
-------
Methodology and Time-Series Consistency
Carbon dioxide emissions from the application of urea to agricultural soils were estimated using the IPCC (2006)
Tier 1 methodology. The method assumes that C in the urea is released after application to soils and converted to
CO2. The annual amounts of urea applied to croplands (see Table 5-26) were derived from the state-level fertilizer
sales data provided in Commercial Fertilizer reports (TVA 1991,1992,1993,1994; AAPFCO 1995 through 2021).24
These amounts were multiplied by the default IPCC (2006) emission factor (0.20 metric tons of C per metric ton of
urea), which is equal to the C content of urea on an atomic weight basis. The calculations were made using a
Monte Carlo analysis as described in the Uncertainty section below.
Fertilizer sales data are reported in fertilizer years (July previous year through June current year) so a calculation
was performed to convert the data to calendar years (January through December). According to monthly fertilizer
use data (TVA 1992b), 35 percent of total fertilizer used in any fertilizer year is applied between July and December
of the previous calendar year, and 65 percent is applied between January and June of the current calendar year.
Fertilizer sales data for the 2017 through 2020 fertilizer years were not available for this Inventory. Therefore, urea
application in the 2017 through 2020 fertilizer years were estimated using a linear, least squares trend of
consumption over the data from the previous five years (2012 through 2016) at the state scale. A trend of five
years was chosen as opposed to a longer trend as it best captures the current inter-state and inter-annual
variability in consumption. State-level estimates of CO2 emissions from the application of urea to agricultural soils
were summed to estimate total emissions for the entire United States. The fertilizer year data is then converted
into calendar year (Table 5-26) data using the method described above.
Table 5-26: Applied Urea (MMT)
1990
2005
2016
2017
2018
2019
2020
Urea Fertilizer3
3.3
00
'sT
6.4
6.7
6.8
7.0
7.2
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 category.
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2020. In addition, the same methods are applied in all years and the activity data are extended using a
data splicing method with a linear extrapolation based on the last four years of urea fertilization data to ensure
consistency in the time series. These years will be recalculated when additional data are available on urea
fertilization.
Uncertainty
An Approach 2 Monte Carlo analysis is conducted as described by the IPCC (2006). The largest source of
uncertainty is the default emission factor, which assumes that 100 percent of the C in CO(NH2)2 applied to soils is
emitted as CO2. The uncertainty surrounding this factor incorporates the possibility that some of the C may not be
emitted to the atmosphere, and therefore the uncertainty range is set from 50 percent emissions to the maximum
emission value of 100 percent using a triangular distribution. In addition, urea consumption data have uncertainty
that is represented as a normal density. Due to the highly skewed distribution of the resulting emissions from the
Monte Carlo uncertainty analysis, the estimated emissions are based on the analytical solution to the equation,
and the confidence interval is approximated based on the values at 2.5 and 97.5 percentiles.
24 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.
5-50 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Carbon dioxide emissions from urea fertilization of agricultural soils in 2020 are estimated to be between 3.02 and
5.44 MMT CO2 Eq. at the 95 percent confidence level. This indicates a range of 43 percent below to 3 percent
above the 2020 emission estimate of 5.3 MMT CO2 Eq. (Table 5-27).
Table 5-27: Quantitative Uncertainty Estimates for CO2 Emissions from Urea Fertilization
(MMT CO2 Eq. and Percent)
Uncertainty Range Relative to Emission
Source Gas
2020 Emission Estimate
Estimate3
(MMT C02 Eq.)
(MMT C02 Eq.)
(%)
Lower Upper
Lower Upper
Bound Bound
Bound Bound
Urea Fertilization C02
5.3
3.02 5.44
-43% +3%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence
interval.
There are additional uncertainties that are not quantified in this analysis. There is uncertainty surrounding the
assumptions underlying conversion of fertilizer years to calendar years. These uncertainties are negligible over
multiple years because an over- or under-estimated value in one calendar year is addressed with a corresponding
increase or decrease in the value for the subsequent year. In addition, there is uncertainty regarding the fate of C
in urea that is incorporated into solutions of urea ammonium nitrate (UAN) fertilizer. Emissions of CO2 from UAN
applications to soils are not estimated in the current Inventory (see Planned Improvements).
/erification
A source-specific QA/QC plan for Urea Fertilization has been developed and implemented, consistent with the U.S.
Inventory QA/QC plan. The UNFCCC expert review recommended using the analytical solution from the Tier 1
method for the estimate of CO2 emissions from Urea Fertilization. This recommendation has been adopted for
inventory reporting.
Recalculations Discussion
In the previous inventory, an error existed with the formula for predicting the total urea fertilizer consumption
data for 2019. The error was corrected. New fertilization data were available for 2016, which was updated, and
this also led to an update in the linear extrapolation from 2017 to 2020 (AAPFCO 2021). These modifications
resulted in changes ranging from -4.1 percent and -0.1 percent for 2014 to 2019.
Planned Improvements
A key planned improvement is to incorporate Urea Ammonium Nitrate (UAN) in the estimation of Urea CO2
emissions. Activity data for UAN have been identified, but additional information is needed to fully incorporate this
type of fertilizer into the analysis, which will be completed in a future Inventory.
5.7 Field Burning of Agricultural Residues
(CRF Source Category 3F)
Crop production creates large quantities of agricultural crop residues, which farmers manage in a variety of ways.
For example, crop residues can be left in the field and possibly incorporated into the soil with tillage; collected and
used as fuel, animal bedding material, supplemental animal feed, or construction material; composted and applied
Agriculture 5-51
-------
to soils; transported to landfills; or burned in the field. Field burning of crop residues is not considered a net source
of CO2 emissions because the C released to the atmosphere as CO2 during burning is reabsorbed during the next
growing season by the crop. However, crop residue burning is a net source of CH4, N2O, CO, and NOx, which are
released during combustion.
In the United States, field burning of agricultural residues commonly occurs in southeastern states, the Great
Plains, and the Pacific Northwest (McCarty 2011). The primary crops that are managed with residue burning
include corn, cotton, lentils, rice, soybeans, sugarcane and wheat (McCarty 2009). In 2020, CH4 and N2O emissions
from field burning of agricultural residues were 0.4 MMT CO2 Eq. (17 kt) and 0.2 MMT CO2 Eq. (1 kt), respectively
(Table 5-28 and Table 5-29). Annual emissions of CH4 and N2O have increased from 1990 to 2020 by 14 percent and
16 percent, respectively. The increase in emissions over time is partly due to higher yielding crop varieties with
larger amounts of residue production and fuel loads, but also linked with an increase in the area burned for some
of the crop types.
Table 5-28: ChU and N2O Emissions from Field Burning of Agricultural Residues (MMT CO2
Eq.)
Gas/Crop Type
1990
2005
2016
2017
2018
2019
2020
ch4
0.4
0.4
0.4
0.4
0.4
0.4
0.4
Maize
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Rice
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Wheat
0.1
0.1
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
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Maize
+
+
0.1
0.1
0.1
0.1
0.1
Rice
+
+
+
+
+
+
+
Wheat
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Barley
+
+
+
+
+
+
+
Oats
+
+
+
+
+
+
+
Other Small Grains
+
+
+
+
+
+
+
Sorghum
+
+
+
+
+
+
+
Cotton
+
+
+
+
+
+
+
Grass Hay
+
+
+
+
+
+
+
Legume Hay
+
+
+
+
+
+
+
Peas
+
+
+
+
+
+
+
5-52 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Sunflower
+
+
+
+
+
+
+
Tobacco
+
+
+
+
+
+
+
Vegetables
+
+
+
+
+
+
+
Chickpeas
+
+
+
+
+
+
+
Dry Beans
+
+
+
+
+
+
+
Lentils
+
+
+
+
+
+
+
Peanuts
+
+
+
+
+
+
+
Soybeans
+
+
+
+
+
+
+
Potatoes
+
+
+
+
+
+
+
Sugarbeets
+
+
+
+
+
+
+
Total
0.5
0.6
0.6
0.6
0.6
0.6
0.6
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Table 5-29: ChU, N2O, CO, and NOx Emissions from Field Burning of Agricultural Residues
(kt)
Gas/Crop Type
1990
2005
2016
2017
2018
2019
2020
ch4
15
17
17
17
17
17
17
Maize
2
4
5
5
5
5
5
Rice
3
3
2
3
2
3
2
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
+
+
+
+
+
+
+
Agriculture 5-53
-------
Legume Hay
+
+
+
+
+
+
+
Peas
+
+
+
+
+
+
+
Sunflower
+
+
+
+
+
+
+
Tobacco
+
+
+
+
+
+
+
Vegetables
+
+
+
+
+
+
+
Chickpeas
+
+
+
+
+
+
+
Dry Beans
+
+
+
+
+
+
+
Lentils
+
+
+
+
+
+
+
Peanuts
+
+
+
+
+
+
+
Soybeans
+
+
+
+
+
+
+
Potatoes
+
+
+
+
+
+
+
Sugarbeets
+
+
+
+
+
+
+
CO
315
363
340
339
338
337
336
NOx
13
15
14
14
14
14
14
+ Does not exceed 0.5 kt.
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
A country-specific Tier 2 method is used to estimate greenhouse gas emissions from field burning of agricultural
residues from 1990 to 2014 (for more details comparing the country-specific approach to the IPCC (2006) default
approach, see Box 5-6), and a data splicing method with a linear extrapolation is applied to complete the emissions
time series from 2015 to 2020. The following equation is used to estimate the amounts of C and N released
(Ri, where iis C or N) from burning.
Equation 5-1: Elemental C or N Released through Oxidation of Crop Residues
Rt = CP x RCR x DMF x Ftx FB x CE
FB =
where,
Crop Production (CP)
Residue: Crop Ratio (RCR)
Dry Matter Fraction (DMF)
Fraction C or N (Fj)
Fraction Burned (FB)
Combustion Efficiency (CE)
Area Burned (AB)
Crop Area Harvested (CAH)
AB
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 C or N released with respect to the total amount of C or N
available in the burned material, respectively, unitless
Total area of crop burned, by state, ha
Total area of crop harvested, by state, ha
Crop production data are available by state and year from USDA (2019) for twenty-one crops that are burned in
the conterminous United States, including maize, rice, wheat, barley, oats, other small grains, sorghum, cotton,
grass hay, legume hay, peas, sunflower, tobacco, vegetables, chickpeas, dry beans, lentils, peanuts, soybeans,
5-54 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
potatoes, and sugarbeets.25 Crop area data are based on the 2015 National Resources Inventory (NRI) (USDA-NRCS
2018). In order to estimate total crop production, the crop yield data from USDA Quick Stats crop yields is
multiplied by the NRI crop areas. The production data for the crop types are presented in Table 5-30. Alaska and
Hawaii are not included in the current analysis, but there is a planned improvement to estimate residue burning
emissions for these two states in a future Inventory.
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) = emission ratio by gas, g CH4-C or CO-C/g C released, or g N2O-N or NOx-
N/g N released
Conversion Factor (CF) = 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)
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 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 (MbX Cf) x Get x 10~6
where,
Area Burned (AB) = Total area of crop burned (ha)
Mass of Fuel (Mb x Cf) = IPCC (2006) default carbon fractions with fuel biomass consumption U.S.-
Specific Values using NASS Statistics26 (metric tons dry matter burnt
ha"1)
Emission Factor (Gef) = IPCC (2006) emission factor (g kg1 dry matter burnt)
25 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).
26 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 5-55
-------
The IPCC (2006) Tier 1 method approach resulted in 33 percent lower emissions of Cm and 53 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,
N2O and NOx, compared to IPCC (2006) approach that is based on dry matter rather than elemental
composition.
Table 5-30: Agricultural Crop Production (kt of Product)
Crop 1990 2005 2013 2014
Maize 296,065 371,256 436,565 453,524
Rice 9,543 11,751 10,894 12,380
Wheat 79,805 68,077 67,388 62,602
Barley 9,281 5,161 4,931 5,020
Oats 5,969 2,646 1,806 2,042
Other Small Grains 2,651 2,051 1,902 2,492
Sorghum 23,687 14,382 18,680 18,436
Cotton 4,605 6,106 3,982 4,396
Grass Hay 44,150 49,880 45,588 46,852
Legume Hay 90,360 91,819 79,669 82,844
Peas 51 660 599 447
Sunflower 1,015 1,448 987 907
Tobacco 1,154 337 481 542
Vegetables 0 1,187 1,844 2,107
Chickpeas 0 5 0 0
Dry Beans 467 1,143 1,110 1,087
Lentils 0 101 72 76
Peanuts 1,856 2,176 2,072 2,735
Soybeans 56,612 86,980 94,756 110,560
Potatoes 18,924 20,026 20,234 19,175
Sugarbeets 24,951 25,635 31,890 31,737
Note: The amount of crop production has not been compiled for 2015 to 2020
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 2014), 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
5-56 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 often, and avoid models that appear to
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 5-31 shows the resulting percentage of crop residue burned at the
national scale by crop type. State-level estimates are also available upon request.
Table 5-31: U.S. Average Percent Crop Area Burned by Crop (Percent)
Crop
1990
2005
2013
2014
Maize
+%
+%
+%
+%
Rice
8%
8%
4%
6%
Wheat
1%
2%
2%
1%
Barley
1%
+%
1%
1%
Oats
1%
1%
2%
1%
Other Small Grains
1%
1%
1%
1%
Sorghum
1%
1%
1%
1%
Cotton
1%
1%
1%
1%
Grass Hay
+%
+%
+%
+%
Legume Hay
+%
+%
+%
+%
Peas
+%
+%
+%
+%
Sunflower
+%
+%
+%
+%
Tobacco
2%
2%
3%
3%
Vegetables
0%
+%
+%
+%
Chickpeas
0%
1%
0%
0%
Dry Beans
1%
1%
+%
+%
Lentils
0%
+%
+%
+%
Peanuts
3%
3%
3%
3%
Soybeans
+%
+%
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 5-32). 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 5-32 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 5-33).
Agriculture 5-57
-------
Table 5-32: 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
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.
5-58 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Tomatoes: Scholberg et al. (2000a,b); Akintoye et al. (2005); values for AGR-N and BGR-N are from
Grains.
Table 5-33: 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
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).
For this Inventory, new activity data on the burned areas have not been analyzed for 2015 to 2020. To complete
the emissions time series, a linear extrapolation of the trend is applied to estimate the emissions in the last five
years of the inventory. Specifically, a linear regression model with autoregressive moving-average (ARMA) errors is
used to estimate the trend in emissions over time from 1990 through 2014, and the trend is used to approximate
the Cm, N2O, CO and NOx for the last six years in the time series from 2015 to 2020 (Brockwell and Davis 2016).
The Tier 2 method described previously will be applied to recalculate the emissions for the last six years in the time
series (2015 to 2020) in a future Inventory.
In order to ensure time-series consistency, the same method is applied from 1990 to 2014, and a linear
extrapolation method is used to approximate emissions for the remainder of the time series based on the
emissions data from 1990 to 2014. This extrapolation method is consistent with data splicing methods in IPCC
(2006).
Uncertainty
Emissions are estimated using a linear regression model with autoregressive moving-average (ARMA) errors for
2020. The linear regression ARMA model produced estimates of the upper and lower bounds to quantify
uncertainty (Table 5-34), and the results are summarized in Table 5-34. Methane emissions from field burning of
agricultural residues in 2020 are between 0.35 and 0.50 MMT CO2 Eq. at a 95 percent confidence level. This
indicates a range of 18 percent below and 18 percent above the 2020 emission estimate of 0.4 MMT CO2 Eq.
Nitrous oxide emissions are between 0.16 and 0.22 MMT CO2 Eq., or approximately 17 percent below and 17
percent above the 2020 emission estimate of 0.2 MMT CO2 Eq.
Table 5-34: Approach 2 Quantitative Uncertainty Estimates for ChU and N2O Emissions from
Field Burning of Agricultural Residues (MMT CO2 Eq. and Percent)
2020 Emission
Source
Gas
Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.)
(MMT C02
Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Field Burning of Agricultural
Residues
ch4
0.4
0.35
0.50
-18%
18%
Field Burning of Agricultural
Residues
n2o
0.2
0.16
0.22
-17%
17%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
Due to data limitations, there are additional uncertainties in agricultural residue burning, particularly the potential
omission of burning associated with Kentucky bluegrass (produced on farms for turf grass installation) and
sugarcane (see Annex 5 on sugarcane).
Agriculture 5-59
-------
QA/QC and Verification
A source-specific QA/QC plan for field burning of agricultural residues is implemented with Tier 1 analyses,
consistent with the U.S. Inventory QA/QC plan outlined in Annex 8. The previous Inventory included a term for
burning efficiency that is not found in the IPCC/UNEP/OECD/IEA (1997) method. This term has been removed
based on a QA/QC initiated by the UN Expert Review Team. In addition, the combustion efficiency term has been
set to 90 percent to be consistent with the Tier 1 method in IPCC/UNEP/OECD/IEA (1997).
Recalculations Discussion
No recalculations have been conducted for this source category.
Planned Improvements
A key planned improvement is to estimate the emissions associated with field burning of agricultural residues in
the states of Alaska and Hawaii. In addition, a new method is in development that will directly link agricultural
residue burning with the Tier 3 methods that are used in several other source categories, including Agricultural Soil
Management, Cropland Remaining Cropland, and Land Converted to Cropland chapters of the Inventory. The
method is based on simulating burning events directly within the DayCent process-based model framework using
information derived from remote sensing fire products as described in the Methodology section. This
improvement will lead to greater consistency in the methods for across sources, ensuring mass balance of C and N
in the Inventory analysis.
As previously noted in this chapter, remote sensing data were used in combination with a resource survey to
estimate non-C02 emissions and these data did not allow identification of burning of sugarcane (see Annex 5). EPA
has received feedback on this category/crop type, which includes average estimates of emissions of sugarcane
burning found in academic literature. EPA is assessing this information identified in feedback, other available
activity data, and an updated methodology, as part of Inventory improvements which EPA plans to implement for
the 2023 submission.
5-60 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Land Usej
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 CFUemissions 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 CH4 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 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
-------
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 2020 the land use, land-use change, and forestry (LULUCF) sector resulted in a net increase in C stocks (i.e., net
CO2 removals) of 812.2 MMT CO2 Eq.2 This represents an offset of approximately 13.6 percent of total (i.e., gross)
greenhouse gas emissions in 2020. Emissions of CH4 and N2O from LULUCF activities in 2020 were 38.1 and 15.2
MMT CO2 Eq., respectively, and combined represent 0.9 percent of total greenhouse gas emissions.3 In 2020 the
overall net flux from LULUCF resulted in a removal of 758.9 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 2020
time series are shown in Figure 6-2.
Figure 6-1: 2020 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-C02 Emissions from Drained Organic Soils
CFU Emissions from Land Converted to Coastal Wetlands
CH4 Emissions from Land Converted to Flooded Land
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
N011-CO2 Emissions from Flooded Land Remaining Flooded Land
Non-CCh Emissions from Forest Fires
Land Converted to Cropland
Land Converted to Settlements
(250) (200) (150) (100) (50) 0 50 100
MMT CO2 Eq.
Note: Parentheses in horizontal axis indicate net sequestration.
(668.1)
Carbon Stock Change
I Non-CCh Emissions
l< 0.51
l< 0.51
l< 0.5|
l< 0.51
l< 0.5|
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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure 6-2; Trends in Emissions and Removals (Net CO2 Flux) from Land Use, Land-Use
Change, and Forestry
400
300
I Land Converted to Settlements
I Land Converted to Cropland
Grassland Remaining Grassland
Land Converted to Wetlands
Wetlands Remaining Wetlands
Cropland Remaining Cropland
Land Converted to Grassland
Land Converted to Forest Land
I Settlements Remaining Settlements
I Forest Land Remaining Forest Land
I Net Emissions (Sources and Sinks)
-tH q nj CO rJ ffl H
m ^ ^ fN P! rM m
O -300
u
-400
-500
-600
-700
-800
-900
-1,000
^0 1^1
CO fN |<
^ rM m
u"i co ro
qx d) c\ O) O)
Table 6-1: 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 Fires8
Land Converted to Grassland
Changes in all Ecosystem Carbon Stocks'
Wetlands Remaining Wetlands
2017
2018
2019
2020
(670.1)
(688.3)
17.7
0.5
0.1
(99.5)
(99.5)
(22.3)
(22.3)
54.3
54.3
9.9
9.3
0.6
(22.7)
(22.7)
15.9
(664.6)
(677.1)
11.9
0.5
0.1
(99.5)
(99.5)
(16.6)
(16.6)
54.0
54.0
10.3
9.7
0.6
(22.4)
(22.4)
15.9
(631.8)
(634.8)
2.5
0.5
0.1
(99.5)
(99.5)
(14.5)
(14.5)
53.9
53.9
13.1
12.4
0.6
(21.5)
(21.5)
15.9
(642.2)
(668.1)
25.3
0,5
0.1
(99.5)
(99.5)
(23.3)
(23.3)
54.4
54.4
5.1
4.5
0,6
(24.1)
(24.1)
15.8
Land Use, Land-Use Change, and Forestry 6-3
-------
Changes in Organic Soil Carbon Stocks in
Peatlands
1.1
1.1
0.7
0.8
0.8
0.8
0.7
Non-C02 Emissions from Peatlands
Remaining Peatlands
+
+
+
+
+
+
+
Changes in Biomass, DOM, and Soil
Carbon Stocks in Coastal Wetlands
(8.5)
(7.6)
(8.8)
(8.8)
(8.8)
(8.8)
(8.8)
CH4 Emissions from Coastal Wetlands
Remaining Coastal Wetlands
3.7
3.8
3.8
3.8
3.8
3.8
3.8
N20 Emissions from Coastal Wetlands
Remaining Coastal Wetlands
0.1
0.2
0.1
0.1
0.2
0.2
0.2
CH4 Emissions from Flooded Land
Remaining Flooded Land
18.2
19.8
19.9
19.9
19.9
19.9
19.9
Land Converted to Wetlands
7.2
1.3
0.6
0.6
0.6
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
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Changes in Land Converted to Flooded
Land
3.9
0.3
0.3
0.3
0.3
0.3
0.3
CH4 Emissions from Land Converted to
Flooded Land
2.6
0.2
0.2
0.2
0.2
0.2
0.2
Settlements Remaining Settlements
(107.6)
(113.5)
(121.5)
(125.3)
(124.9)
(124.5)
(123.7)
Changes in Organic Soil Carbon Stocks
11.3
12.2
16.0
16.0
15.9
15.9
15.9
Changes in Settlement Tree Carbon
Stocks
(96.4)
(117.4)
(129.8)
(129.8)
(129.8)
(129.8)
(129.8)
N20 Emissions from Settlement Soilsh
2.0
3.1
2.2
2.3
2.4
2.4
2.5
Changes in Yard Trimming and Food
Scrap Carbon Stocks in Landfills
(24.5)
(11.4)
(10.0)
(13.8)
(13.4)
(13.1)
(12.2)
Land Converted to Settlements
60.8
82.8
77.8
77.9
78.0
77.9
77.9
Changes in all Ecosystem Carbon Stocks'
60.8
82.8
77.8
77.9
78.0
77.9
77.9
LULUCF Emissions'
31.4
41.3
35.4
45.5
39.8
30.3
53.2
ch4
27.2
30.9
28.3
34.0
30.7
25.5
38.1
n2o
4.2
10.5
7.1
11.5
9.1
4.8
15.2
LULUCF Carbon Stock Change1'
(892.0)
(831.1)
(862.0)
(826.7)
(809.0)
(760.8)
(812.2)
LULUCF Sector Net Totalk
(860.6)
(789.8)
(826.6)
(781.2)
(769.3)
(730.5)
(758.9)
+ 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.
g Estimates include CH4 and N20 emissions from fires on both Grassland Remaining Grassland and Land Converted to
Grass/and.
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
6-4 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Coastal Wetlands, Flooded Land Remaining Flooded Land, and Land Converted to Flooded Land; and N20 emissions from
Forest Soils and Settlement Soils.
j 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 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 Cm 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 9.0 percent between
1990 and 2020. 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 2020, 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 2020 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, included for the first time in this year's estimates, was the largest source of
Cm emissions from LULUCF in 2020, totaling 19.9 MMT CO2 Eq. (797kt of CH4). Forest fires resulted in CH4
emissions of 13.6MMT CO2 Eq. (545kt of CH4). Coastal Wetlands Remaining Coastal Wetlands resulted in CH4
emissions of 3.8MMT 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. (7 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 2020, totaling 11.7 MMT CO2 Eq. (39 kt of
N2O). Nitrous oxide emissions from fertilizer application to settlement soils in 2020 totaled to 2.5 MMT CO2 Eq. (8
kt of N2O). This represents an increase of 23.2 percent since 1990. Additionally, the application of synthetic
fertilizers to forest soils in 2020 resulted in N2O emissions of 0.5 MMT CO2 Eq. (2 kt of N2O). Nitrous oxide
emissions from fertilizer application to forest soils have increased by 455.1 percent since 1990, but still account for
a relatively small portion of overall emissions. Grassland fires resulted in N2O emissions of 0.3 MMT CO2 Eq. (1 kt of
N2O). Coastal Wetlands Remaining Coastal Wetlands resulted in N2O emissions of 0.2 MMT CO2 Eq. (1 kt of N2O).
Drained Organic Soils on forest lands resulted in N2O emissions of 0.1 MMT CO2 Eq. (less than 0.5 kt of N2O), and
Peatlands Remaining Peatlands resulted in N2O emissions of less than 0.05 MMT CO2 Eq.
Table 6-2: Emissions and Removals from Land Use, Land-Use Change, and Forestry by Gas
(MMT COz Eq.)
Gas/Land-Use Category
1990
2005
2016
2017
2018
2019
2020
Carbon Stock Change (C02)a
(892.0)
(831.1)
(862.0)
(826.7)
(809.0)
(760.8)
(812.2)
Forest Land Remaining Forest Land
(774.0)
(687.3)
(725.6)
(688.3)
(677.1)
(634.8)
(668.1)
Land Converted to Forest Land
(98.6)
(99.1)
(99.5)
(99.5)
(99.5)
(99.5)
(99.5)
Cropland Remaining Cropland
(23.2)
(29.0)
(22.7)
(22.3)
(16.6)
(14.5)
(23.3)
Land Converted to Cropland
51.8
52.0
54.1
54.3
54.0
53.9
54.4
Grassland Remaining Grassland
6.9
8.7
8.0
9.3
9.7
12.4
4.5
Land Converted to Grassland
(3.1)
(37.0)
(22.6)
(22.7)
(22.4)
(21.5)
(24.1)
Wetlands Remaining Wetlands
(7.4)
(6.5)
(8.0)
(8.0)
(8.0)
(8.0)
(8.1)
Land Converted to Wetlands
4.3
0.8
0.3
0.3
0.3
0.3
0.3
Settlements Remaining Settlements
(109.6)
(116.6)
(123.8)
(127.7)
(127.3)
(127.0)
(126.1)
Land Converted to Settlements
60.8
82.8
77.8
77.9
78.0
77.9
77.9
ch4
27.2
30.9
28.3
34.0
30.7
25.5
38.1
Forest Land Remaining Forest Land:
2.3
6.5
3.9
9.5
6.2
1.1
13.6
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
-------
Forest Firesb
Forest Land Remaining Forest Land:
Drained Organic Soilsd
+
+
+
+
+
+
+
Grassland Remaining Grassland:
Grassland Firesc
0.1
0.3
0.3
0.3
0.3
0.3
0.3
Wetlands Remaining Wetlands: Flooded
Land Remaining Flooded Land
18.2
19.8
19.9
19.9
19.9
19.9
19.9
Wetlands Remaining Wetlands: Coastal
Wetlands Remaining Coastal Wetlands
3.7
3.8
3.8
3.8
3.8
3.8
3.8
Wetlands Remaining Wetlands:
Peatlands Remaining Peatlands
+
+
+
+
+
+
+
Land Converted to Wetlands: Land
Converted to Flooded Lands
2.6
0.2
0.2
0.2
0.2
0.2
0.2
Land Converted to Wetlands: Land
Converted to Coastal Wetlands
0.2
0.2
0.2
0.2
0.2
0.2
0.2
n2o
4.2
10.5
7.1
11.5
9.1
4.8
15.2
Forest Land Remaining Forest Land:
Forest Firesb
1.8
6.3
3.9
8.2
5.7
1.3
11.7
Forest Land Remaining Forest Land:
Forest Soils'
0.1
0.5
0.5
0.5
0.5
0.5
0.5
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
0.2
0.2
0.2
Wetlands Remaining Wetlands:
Peatlands Remaining Peatlands
+
+
+
+
+
+
+
Settlements Remaining Settlements:
Settlement Soilse
2.0
3.1
2.2
2.3
2.4
2.4
2.5
LULUCF Carbon Stock Change3
(892.0)
(831.1)
(862.0)
(826.7)
(809.0)
(760.8)
(812.2)
LULUCF Emissions^
31.4
41.3
35.4
45.5
39.8
30.3
53.2
LULUCF Sector Net Totalh
(860.6)
(789.8)
(826.6)
(781.2)
(769.3)
(730.5)
(758.9)
+ 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.
g 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 net carbon
stock changes in units of MMT C02 Eq.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
6-6 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 6-3: Emissions and Removals from Land Use, Land-Use Change, and Forestry by Gas
(kt)
Gas/Land-Use Category
1990
2005
2016
2017
2018
2019
2020
Carbon Stock Change (C02)a
(892,027)
(831,126)
(862,045)
(826,667)
(809,026)
(760,820)
(812,176)
Forest Land Remaining Forest Land
(773,993)
(687,271)
(725,571)
(688,301)
(677,101)
(634,824)
(668,057)
Land Converted to Forest Land
(98,585)
(99,068)
(99,454)
(99,523)
(99,518)
(99,520)
(99,521)
Cropland Remaining Cropland
(23,176)
(29,002)
(22,731)
(22,293)
(16,597)
(14,544)
(23,335)
Land Converted to Cropland
51,784
52,032
54,107
54,273
53,975
53,935
54,380
Grassland Remaining Grassland
6,940
8,734
7,958
9,308
9,670
12,425
4,497
Land Converted to Grassland
(3,141)
(36,951)
(22,553)
(22,693)
(22,397)
(21,485)
(24,101)
Wetlands Remaining Wetlands
(7,399)
(6,549)
(8,046)
(7,954)
(7,994)
(8,034)
(8,084)
Land Converted to Wetlands
4,329
807
254
258
265
271
279
Settlements Remaining Settlements
(109,567)
(116,642)
(123,794)
(127,679)
(127,299)
(126,977)
(126,128)
Land Converted to Settlements
60,793
82,784
77,784
77,938
77,970
77,932
77,895
ch4
1,088
1,235
1,131
1,359
1,226
1,022
1,522
Forest Land Remaining Forest Land:
Forest Firesb
92
260
154
381
249
45
545
Forest Land Remaining Forest Land:
Drained Organic Soilsd
1
1
1
1
1
1
1
Grassland Remaining Grassland:
Grassland Firesc
3
13
11
12
12
12
12
Wetlands Remaining Wetlands:
Flooded Land Remaining Flooded
Land
729
792
797
797
797
797
797
Wetlands Remaining Wetlands:
Coastal Wetlands Remaining Coastal
Wetlands
149
151
153
153
153
153
154
Wetlands Remaining Wetlands:
Peatlands Remaining Peatlands
+
+
+
+
+
+
+
Land Converted to Wetlands: Land
Converted to Flooded Lands
103
9
7
7
7
7
7
Land Converted to Wetlands: Land
Converted to Coastal Wetlands
10
10
8
8
7
7
7
n2o
14
35
24
39
31
16
51
Forest Land Remaining Forest Land:
Forest Firesb
6
21
13
27
19
4
39
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
Wetlands Remaining Wetlands:
Peatlands Remaining Peatlands
+
+
+
+
+
+
+
Settlements Remaining Settlements:
Settlement Soilse
7
10
8
8
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.
Land Use, Land-Use Change, and Forestry 6-7
-------
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 may not sum due to independent rounding. Parentheses indicate net sequestration.
Each year, some emission and sink estimates in the LULUCF 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 emissions and
sinks 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 2019)
to ensure that the trend is accurate. Of the updates implemented for this Inventory, the most significant include
(1) Flooded Land Remaining Flooded Land and Land Converted to Flooded Land: new categories included for the
first time based on new guidance in the 2019 Refinement to the 2006IPCC Guidelines for National GHG Inventories,
(2) Forest Lands: use of new data from the National Forest Inventory (NFI), compiling population estimates of
carbon stocks and stock changes using NFI data from each U.S. state and summing over all states to obtain the
national estimates, refined estimates in the Digital General Soil Map, and new data on area burned from the
Monitoring Trends in Burn Severity (MTBS) data product; and (3) Coastal Wetlands: an updated NOAA report on
fisheries data was released in 2021 and was used in estimating N2O emissions from aquaculture. Together, these
updates for 2019 decreased total sequestration of CO2 by 51.6 MMT CO2 Eq. (5.7 percent) and decreased total
non-CC>2 emissions by 16.2 MMT CO2 Eq. (81.5 percent), compared to the previous Inventory (i.e., 1990 to 2019).
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. 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,
^ See http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
6-8 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
and this chapter, follows 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
anthropogenic drivers. Guidelines for factoring out natural emissions and removals may be developed in the
future, but currently the managed land proxy is considered the most practical approach for conducting an
inventory in this sector (IPCC 2010). This section of the Inventory has been developed in order to comply with this
guidance.
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,
which has not changed much over the time series of the Inventory (Table 6-5). In 2020, the United States had a
total of 282 million hectares of managed Forest Land (0.03 percent decrease compared to 1990). There are 162
million hectares of cropland (7.2 percent decrease compared to 1990), 337 million hectares of managed Grassland
(0.01 percent increase compared to 1990), 39 million hectares of managed Wetlands (1.8 percent increase
compared to 1990), 45 million hectares of Settlements (34 percent increase compared to 1990), and 22 million
hectares of managed Other Land (2.4 percent increase compared to 1990) (Table 6-5).
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 http://www.fia.fs.fed.us/tools-data/default.asp.
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.
Land Use, Land-Use Change, and Forestry 6-9
-------
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 There are
also discrepancies in the inventory emissions data and the land representation section because new FIA data were
used in the inventory analysis, but were not incorporated into the land representation analysis due to timing of
data availability and resources to complete the analysis. The land representation analysis will incorporate the new
time series of FIA data int the next Inventory. In addition, planned improvements are under development to
estimate C stock changes and greenhouse gas emissions on all managed land and 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
and eastern portions of the country, as well as coastal regions. Settlements are more concentrated along the
coastal margins and in the eastern states.
Table 6-4: Managed and Unmanaged Land Area by Land-Use Categories for All 50 States
(Thousands of Hectares)
Land Use Categories
1990
2005
2016
2017
2018
2019a
2020a
Managed Lands
886,515
886,513
886,513
886,513
886,513
886,513
886,513
Forest
281,621
281,681
281,796
281,652
281,546
281,546
281,546
Croplands
174,471
165,727
161,933
161,933
161,933
161,933
161,933
Grasslands
336,840
337,621
336,657
336,781
336,863
336,863
336,863
Settlements
33,446
40,469
44,795
44,797
44,797
44,797
44,797
Wetlands
38,422
39,017
39,089
39,108
39,132
39,132
39,132
Other
21,715
21,997
22,243
22,243
22,243
22,243
22,243
Unmanaged Lands
49,681
49,684
49,683
49,683
49,683
49,683
49,683
Forest
9,243
8,829
8,208
8,208
8,208
8,208
8,208
Croplands
0
0
0
0
0
0
0
Grasslands
25,530
25,962
26,608
26,608
26,608
26,608
26,608
Settlements
0
0
0
0
0
0
0
Wetlands
4,166
4,166
4,165
4,165
4,165
4,165
4,165
Other
10,742
10,727
10,701
10,701
10,701
10,701
10,701
Total Land Areas
936,196
936,196
936,196
936,196
936,196
936,196
936,196
Forest
290,864
290,510
290,004
289,860
289,754
289,754
289,754
Croplands
174,471
165,727
161,933
161,933
161,933
161,933
161,933
Grasslands
362,370
363,583
363,266
363,389
363,471
363,471
363,471
Settlements
33,446
40,469
44,795
44,797
44,797
44,797
44,797
Wetlands
42,589
43,183
43,254
43,273
43,297
43,297
43,297
Other
32,457
32,725
32,944
32,944
32,944
32,944
32,944
a Land use data were not updated in this Inventory and the data for 2019 and 2020 were assumed to be the same as in 2018.
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. 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 the next Inventory.
12 These "managed area" discrepancies also occur in the Common Reporting Format (CRF) tables submitted to the UNFCCC.
6-10 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 6-5: Land Use and Land-Use Change for the U.S. Managed Land Base for All 50 States
(Thousands of Hectares)
Land Use & Land-Use
Change Categories3
1990
2005
2016
2017
2018
2019
2020
Total Forest Land
281,621
281,681
281,796
281,652
281,546
281,546
281,546
FF
280,393
280,207
280,529
280,380
280,274
280,274
280,274
CF
169
167
134
135
135
135
135
GF
919
1,162
989
992
992
992
992
WF
77
28
25
25
25
25
25
SF
12
24
26
26
26
26
26
OF
50
93
93
93
93
93
93
Total Cropland
174,471
165,727
161,933
161,933
161,933
161,933
161,933
CC
162,163
150,304
148,885
148,884
148,884
148,884
148,884
FC
182
86
58
58
58
58
58
GC
11,738
14,820
12,609
12,609
12,609
12,609
12,609
WC
118
178
104
104
104
104
104
SC
75
100
99
99
99
99
99
OC
195
239
179
179
179
179
179
Total Grassland
336,840
337,621
336,657
336,781
336,863
336,863
336,863
GG
327,446
315,161
316,408
316,502
316,622
316,622
316,622
FG
593
560
553
583
545
545
545
CG
8,237
17,523
16,600
16,600
16,600
16,600
16,600
WG
176
542
308
308
308
308
308
SG
43
509
346
346
346
346
346
OG
345
3,328
2,442
2,442
2,442
2,442
2,442
Total Wetlands
38,422
39,017
39,089
39,108
39,132
39,132
39,132
WW
37,860
37,035
37,616
37,634
37,658
37,658
37,658
FW
83
59
54
54
54
54
54
CW
132
566
440
440
440
440
440
GW
297
1,187
836
836
836
836
836
SW
0
38
25
25
25
25
25
OW
50
133
118
118
118
118
118
Total Settlements
33,446
40,469
44,795
44,797
44,797
44,797
44,797
SS
30,585
31,522
38,210
38,210
38,210
38,210
38,210
FS
310
549
539
541
541
541
541
CS
1,237
3,602
2,452
2,452
2,452
2,452
2,452
GS
1,255
4,499
3,352
3,352
3,352
3,352
3,352
WS
4
61
46
46
46
46
46
OS
54
235
197
197
197
197
197
Total Other Land
21,715
21,997
22,243
22,243
22,243
22,243
22,243
00
20,953
18,231
19,007
19,007
19,007
19,007
19,007
FO
41
70
90
90
90
90
90
CO
301
590
678
678
678
678
678
GO
391
2,965
2,331
2,331
2,331
2,331
2,331
WO
26
121
121
121
121
121
121
SO
2
20
16
16
16
16
16
Grand Total
886,515
886,513
886,513
886,513
886,513
886,513
886,513
a The abbreviations are "F" for Forest Land, "C" for Cropland, "G" for Grassland, "W" for Wetlands, "S" for Settlements,
and "0" for Other Lands. Lands remaining in the same land-use category are identified with the land-use abbreviation
given twice (e.g., "FF" is Forest Land Remaining Forest Land), and land-use change categories are identified with the
previous land use abbreviation followed by the new land-use abbreviation (e.g., "CF" is Cropland Converted to Forest
Land).
Notes: All land areas reported in this table are considered managed. A planned improvement is underway to deal with
an exception for Wetlands, which based on the definitions for the current U.S. Land Representation assessment
includes both managed and unmanaged lands. U.S. Territories have not been classified into land uses and are not
included in the U.S. Land Representation Assessment. See the Planned Improvements section for discussion on plans
to include territories in future Inventories. In addition, C stock changes are not currently estimated for the entire land
Land Use, Land-Use Change, and Forestry 6-11
-------
base, which leads to discrepancies between the managed land area data presented here and in the subsequent
sections of the Inventory (see land use chapters e.g., Forest Land Remaining Forest Land for more information).
Totals may not sum due to independent rounding.
6-12 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure 6-3: Percent of Total Land Area for Each State in the General Land Use Categories for
2020
Forest Lands
Croplands
n< 10
~ 11 - 30
~ 31-50
¦ > 50
~ <10
~ 11 - 30
¦ 31 - 50
¦ > 50
Grasslands
Wetlands
T
Land Use, Land-Use Change, and Forestry 6-13
-------
Methodology and Time-Series Consistency
IPCC Approaches for Representing Land Areas
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 changes of area between
categories 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 derived 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 and peatlands 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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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. 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,15 while definitions of Cropland, Grassland, and Settlements are based on the NRI.16The 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 (Oswalt et al. 2014).
• 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,17 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
enrolled in conservation reserve programs (i.e., set-asides18) 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
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 See http://www.fia.fs.fed.us/librarv/field-guides-methods~proc/docs/2Q15/Core-FIA-FG-7.pdf. page 22.
16 See https://www.nrcs.usda.gov/wps/portal/nrcs/main/national/technical/nra/nri/.
17 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.
18 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.
Land Use, Land-Use Change, and Forestry 6-15
-------
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.19 Savannas, deserts, and tundra are
considered Grassland.20 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. Managed Wetlands are those where the water level is
artificially changed, or 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).
• 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 are estimated for
Land Converted to Other Land during the first 20 years following conversion to account for legacy effects.
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
19 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.
20 2006 IPCC Guidelines do not include provisions to separate desert and tundra as land-use categories.
6-16 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 derived 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
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 2015 from the NRI. The land use
patterns are assumed to remain the same from 2016 through 2020 for this Inventory, but the time series will be
updated when new data are integrated into the land representation analysis.
Land Use, Land-Use Change, and Forestry 6-17
-------
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 conterminous United States in addition to the
including southeast and south-central coastal Alaska. 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 14 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 2015 through 2018; 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. In addition, FIA only records data for forest land across the land base in the
conterminous United States and Alaska.21 Consequently, gaps exist in the land representation when the datasets
are combined, such as federal grassland 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, and 2016 in the conterminous
United States (Yang et al. 2018; Fry et al. 2011; Homer et al. 2007, 2015), and also for Alaska in 2001 and 2011 and
Hawaii in 2001. A Land Cover Change Product is also available for Alaska from 2001 to 2011. 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, 2016 for the conterminous United States and 2011 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 derived from the NLCD
products were used in combination with the NRI database to represent land use and land-use change for federal
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.
21 FIA does collect some data on non-forest land use, but these are held in regional databases versus the national database.
The status of these data is being investigated.
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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 do 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. 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
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
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
-------
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 finally 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
adjustments. For example, if 50 percent of the land use change to Forest Land is associated with Grassland
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
24 The exception is cropland and settlement areas in the NRI, which are classified as managed, regardless of the managed land
base derived 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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
on the specific land-use changes in the past,27 and 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 a given 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 an under- or over-prediction 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 increases or decreases
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 the 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
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
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.
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.
Land Use, Land-Use Change, and Forestry 6-21
-------
• 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 acres (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).
/erification
The land base derived 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 derived 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 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 derived 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 NRI, FIA and NLCD. There is only a 0.4
6-22 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
No recalculations were performed for the 1990 through 2019 portion of the time series, thus the land use areas for
2020 are assumed the same as 2019.
Planned Improvements
The next (i.e., 1990 through 2021) Inventory will be improved by using new NRI, FIA and possibly NLCD data to
update the time series for land representation, providing consistency between the total area of managed land in
the land representation section and the remainder of the Inventory. 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, 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
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.
Land Use, Land-Use Change, and Forestry 6-23
-------
The NOAA C-CAP product is not used directly in the land representation analysis, however, so a planned
improvement for the next (i.e., 1990 through 2021) Inventory is to reconcile the coastal wetlands data from the C-
CAP product with the wetlands area data provided in the NRI, FIA and NLCD. 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.
• Dead wood, which includes all non-living woody biomass either standing, lying on the ground (but not
including litter), or in the soil.
• Litter, which includes all duff, humus, and fine woody debris above the mineral soil and includes woody
fragments with diameters of up to 7.5 cm.
• 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.
In addition, there are two harvested wood pools included when estimating C flux:
• Harvested wood products (HWP) in use.
• HWP in solid waste disposal sites (SWDS).
Forest Carbon Cycle
Carbon is continuously cycled among the previously defined C storage pools and the atmosphere as a result of
biogeochemical processes in forests (e.g., photosynthesis, respiration, decomposition, and disturbances such as
fires or pest outbreaks) and anthropogenic activities (e.g., harvesting, thinning, and replanting). As trees
photosynthesize and grow, C is removed from the atmosphere and stored in living tree biomass. As trees die and
otherwise deposit litter and debris on the forest floor, C is released to the atmosphere and is also transferred to
the litter, dead wood, and soil pools by organisms that facilitate decomposition.
The net change in forest C is not equivalent to the net flux between forests and the atmosphere because timber
harvests do not cause an immediate flux of all harvested biomass C to the atmosphere. Instead, harvesting
transfers a portion of the C stored in wood to a "product pool." Once in a product pool, the C is emitted over time
as CO2 in the case of decomposition and as CO2, CH4, N2O, CO, and NOxwhen the wood product combusts. The rate
of emission varies considerably among different product pools. For example, if timber is harvested to produce
6-24 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
energy, combustion releases C immediately, and these emissions are reported for information purposes in the
Energy sector while the harvest (i.e., the associated reduction in forest C stocks) and subsequent combustion are
implicitly estimated in the Land Use, Land-Use Change, and Forestry (LULUCF) sector (i.e., the portion of harvested
timber combusted to produce energy does not enter the HWP pools). Conversely, if timber is harvested and used
as lumber in a house, it may be many decades or even centuries before the lumber decays and C is released to the
atmosphere. If wood products are disposed of in SWDS, the C contained in the wood may be released many years
or decades later or may be stored almost permanently in the SWDS. These latter fluxes, with the exception of Cm
from wood in SWDS, which is included in the Waste sector, are also estimated in the LULUCF sector.
Net Change in Carbon Stocks within Forest Land of the United States
This section describes the general method for quantifying the net changes in C stocks in the five C storage pools
and two harvested wood pools (a more detailed description of the methods and data is provided in Annex 3.13).
The underlying methodology for determining C stock and stock change relies on data from the national forest
inventory (NFI) conducted by the Forest Inventory and Analysis (FIA) program within the USDA Forest Service. The
annual NFI is implemented across all U.S. forest lands within the conterminous 48 states and Alaska and
inventories have been initiated in Hawaii and some of the U.S. Territories. The methods for estimation and
monitoring are continuously improved and these improvements are reflected in the C estimates (Domke et al.
2016; Domke et al. 2017). First, the total C stocks are estimated for each C storage pool at the individual NFI plot,
next the annual net changes in C stocks for each pool are estimated, and then the changes in stocks are summed
for all pools to estimate total net flux at the population level (e.g., U.S. state). Changes in C stocks from
disturbances, such natural disturbances (e.g., wildfires, insects/disease, wind) or harvesting, are included in the net
changes (See Box 6-3 for more information). For instance, an inventory conducted after a fire implicitly includes
only the C stocks remaining on the NFI plot. The IPCC (2006) recommends estimating changes in C stocks from
forest lands according to several land-use types and conversions, specifically Forest Land Remaining Forest Land
and Land 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 carbon 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 2021 (which includes data collected through 2020 - note that the ongoing COVID 19
pandemic has resulted in delays in data collection in many states) were used in this Inventory. Since area estimates
for some land use categories were not updated in the Land Representation in the current Inventory there are
differences in the area estimates reported in this section and those reported in Section 6.1 Representation of the
U.S. Land Base. The NFIs from each of the conterminous 48 states (CONUS; USDA Forest Service 2022a, 2022b) and
Alaska comprise an estimated 282 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 2020 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 HI
was not included in this section of the current Inventory there are small differences in the area estimates reported
Land Use, Land-Use Change, and Forestry 6-25
-------
in this section and those reported in Section 6.1 Representation of the U.S. Land Base.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, and 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 land removed from
production because it does 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 CONUS and Alaska represented here 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 across the 1990 to 2020 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 about double 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.
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.
31 The term "biomass density" refers to the mass of live vegetation per unit area. It is usually measured on a dry-weight basis. A
carbon fraction of 0.5 is used to convert dry biomass to C (USDA Forest Service 2022d).
6-26 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure 6-4: Changes in Forest Area by Region for Forest Land Remaining Forest Land in the
conterminous United States and Alaska (1990-2020)
100-|
« 90H
o
v_
CO
aj 80-
c
o
ro
CD
w
0
1
o
u.
70-
ro 60-
50-
1 South
¦North
Pacific
Coast
Rocky
Mountain
40 | i i i i | i i i i | i i i i | i i i i | i ii I | I I I I |
1990 1995 2000 2005 2010 2015 2020
Year
South
Forest Carbon Stocks and Stock Change
In Forest Land Remaining Forest Land, 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 uptake (i.e., net sequestration
or accumulation) of C each year from 1990 through 2020. 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). Maintaining current harvesting practices and regeneration activities on these 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
Land Use, Land-Use Change, and Forestry 6-27
-------
C stocks in the forest ecosystem and harvested wood pools associated with Forest Land Remaining Forest Land
were estimated to result in net uptake of 668.1 MMT CO2 Eq. (182.2 MMT C) in 2020 (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 584.4 MMT CO2 Eq. (159.4 MMT C) in 2020 (Table 6-8 and Table 6-9). The majority of this uptake in 2020,
398.7 MMT CO2 Eq. (108.7 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 198 MT C ha 1 from 1990 to 2020. 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 2020. 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. The distribution of carbon in forest ecosystems in Alaska is substantially different
from forests in the CONUS. In Alaska, more than 11 percent of forest ecosystem C is stored in the litter carbon pool
whereas in the CONUS 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) which is why there are substantial C losses in
this pool during severe fire years (Figure 6-5, Table A-227).
The estimated net uptake of C in HWP was 83.6 MMT CO2 Eq. (22.8 MMT C) in 2020 (Table 6-8, Table 6-9, Table A-
210, and Table A-211). The majority of this uptake, 63.6 MMT CO2 Eq. (17.3 MMT C), was from wood and paper in
SWDS. Products in use were an estimated 20.0 MMT CO2 Eq. (5.5 MMT C) in 2020.
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
2016
2017
2018
2019
2020
Forest Ecosystem
(650.2)
(581.2)
(630.4)
(588.1)
(583.0)
(546.0)
(584.4)
Aboveground Biomass
(462.5)
(416.3)
(432.7)
(407.7)
(406.6)
(393.1)
(398.7)
Belowground Biomass
(94.2)
(84.2)
(86.3)
(80.9)
(80.8)
(78.1)
(79.1)
Dead Wood
(96.8)
(96.8)
(106.4)
(99.8)
(102.0)
(97.0)
(101.5)
Litter
0.6
16.0
(3.1)
(1.9)
1.3
22.8
(1.9)
Soil (Mineral)
3.0
(0.3)
(5.6)
(1.1)
4.1
(0.6)
(4.1)
Soil (Organic)
(0.9)
(0.3)
3.0
2.5
0.3
(0.7)
0.2
Drained Organic Soil3
0.8
0.8
0.8
0.8
0.8
0.8
0.8
Harvested Wood
(123.8)
(106.0)
(95.1)
(100.2)
(94.1)
(88.8)
(83.6)
Products in Use
(54.8)
(42.6)
(30.4)
(34.9)
(29.0)
(24.4)
(20.0)
SWDS
(69.0)
(63.4)
(64.8)
(65.3)
(65.1)
(64.5)
(63.6)
Total Net Flux
(774.0)
(687.3)
(725.6)
(688.3)
(677.1)
(634.8)
(668.1)
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
6-28 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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. 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 6.10 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. 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.
Table 6-9: Net C Flux from Forest Ecosystem Pools in Forest Land Remaining Forest Land
and Harvested Wood Pools (MMT C)
Carbon Pool
1990
2005
2016
2017
2018
2019
2020
Forest Ecosystem
(177.3)
(158.5)
(171.9)
(160.4)
(159.0)
(148.9)
(159.4)
Aboveground Biomass
(126.1)
(113.5)
(118.0)
(111.2)
(110.9)
(107.2)
(108.7)
Belowground Biomass
(25.7)
(23.0)
(23.5)
(22.1)
(22.0)
(21.3)
(21.6)
Dead Wood
(26.4)
(26.4)
(29.0)
(27.2)
(27.8)
(26.5)
(27.7)
Litter
0.2
4.4
(0.9)
(0.5)
0.3
6.2
(0.5)
Soil (Mineral)
0.8
(0.1)
(1.5)
(0.3)
1.1
(0.2)
(1.1)
Soil (Organic)
(0.3)
(0.1)
0.8
0.7
0.1
(0.2)
0.1
Drained Organic Soil3
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Harvested Wood
(33.8)
(28.9)
(25.9)
(27.3)
(25.7)
(24.2)
(22.8)
Products in Use
(14.9)
(11.6)
(8.3)
(9.5)
(7.9)
(6.6)
(5.5)
SWDS
(18.8)
(17.3)
(17.7)
(17.8)
(17.8)
(17.6)
(17.3)
Total Net Flux
(211.1)
(187.4)
(197.9)
(187.7)
(184.7)
(173.1)
(182.2)
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.
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 6.10 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. 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.
Land Use, Land-Use Change, and Forestry 6-29
-------
Stock estimates for forest ecosystem and harvested wood C storage pools are presented in Table 6-10. Together,
the estimated aboveground biomass and soil C pools account for a large proportion of total forest ecosystem C
stocks. Forest land area estimates are also provided in Table 6-10, but these do not precisely match those in
Section 6.1 Representation of the U.S. Land Base for Forest Land Remaining Forest Land. This is because the forest
land area estimates in Table 6-10 only include managed forest land in the conterminous 48 states and Alaska while
the area estimates in Section 6.1 include all managed forest land in Hawaii. Differences also exist because forest
land area estimates are based on the latest NFI data through 2020 and woodland areas previously included as
forest land have been separated and included in the Grassland categories in this Inventory.32
Table 6-10: Forest Area (1,000 ha) and C Stocks in Forest Land Remaining Forest Land and
Harvested Wood Pools (MMT C)
1990
2005
2017
2018
2019
2020
2021
Forest Area (1,000 ha)
282,585
282,250
282,352
282,312
282,177
282,061
281,951
Carbon Pools (MMT C)
Forest Ecosystem
53,148
55,721
57,687
57,848
58,007
58,156
58,316
Aboveground Biomass
12,062
13,874
15,250
15,361
15,472
15,579
15,688
Belowground Biomass
2,375
2,743
3,019
3,041
3,064
3,085
3,106
Dead Wood
2,060
2,460
2,787
2,814
2,842
2,868
2,896
Litter
3,838
3,834
3,815
3,816
3,815
3,809
3,810
Soil (Mineral)
25,458
25,452
25,458
25,458
25,457
25,457
25,459
Soil (Organic)
7,355
7,358
7,357
7,357
7,357
7,357
7,357
Harvested Wood
1,895
2,353
2,618
2,645
2,671
2,695
2,718
Products in Use
1,249
1,447
1,506
1,515
1,523
1,530
1,536
SWDS
646
906
1,112
1,129
1,147
1,165
1,182
Total C Stock
55,043
58,074
60,305
60,493
60,678
60,851
61,034
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. 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.
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 2020 requires estimates of C stocks for 2020 and
2021.
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 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 (1990-2020)
20 n
8 l
i" CT3
¦S (j
l-
I
0_H
-20-
40-
-60-
-80-
-100-
-120-
-140-
¦&5 £ -160-
c
•S P
2 ro
i-5
iS to
c
o
-180-
-200-
-220
I 1 1
1990
I I | I I I I | I i I
1995 2000
I | i i i I | i
2005 2010
Year
i i | i
2015
2020
• All forest ecosystem pools
Aboveground biomass
¦ Belowground biomass
Dead wood
¦ Litter
Soil (mineral)
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 all 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 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-C02 greenhouse gas emissions from forest
fires that are also quantified in a separate section beiow as required by IPCC Guidance and UNFCCC reporting
requirements.
Emissions estimates are developed consistent with 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-CCh emissions. Estimated CO2 emissions for fires on forest
lands in the conterminous 48 states and in Alaska for 2020 are 237 MMT CO2 per year (Table 6-11). This
estimate is an embedded component of the net annual forest C stock change estimates provided previously
Land Use, Land-Use Change, and Forestry 6-31
-------
(i.e., Table 6-9), but this separate approach to estimate 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 C02 (MMT per Year) Emissions from Forest Fires in the
Conterminous 48 States and Alaska3
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
11.2
26.0
37.1
2005
33.9
93.5
127.4
2016
73.1
5.7
78.8
2017
154.8
10.0
164.8
2018
108.5
6.7
115.2
2019
27.0
55.8
82.7
2020
236.8
0.6
237.4
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 IPCC (2006). Forest ecosystem C stocks and net annual C
stock change were determined according to the stock-difference method for the CONUS, 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 report 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 with 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, HI was not
included in this section of the current Inventory so that also contributes to small 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 CONUS 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 2020. 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. To implement the gain-loss approach in Alaska, forest land conditions in Alaska were
6-32 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
observed on NFI plots from 2004 to 2020. 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 2020. First, carbon stocks for each forest ecosystem carbon pool were predicted for the year 2016 for all
base intensity NFI plot locations (representing approximately 2,403 ha) in coastal southeast and southcentral
Alaska and for 1/5 intensity plots in interior Alaska (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 by 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 2020. 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 and Domke et al. (In prep) 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 2020. 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 and 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
conducted by the FIA program (Frayer and 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-panel design, with 20 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
Land Use, Land-Use Change, and Forestry 6-33
-------
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 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 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.
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 United States 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
6-34 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 separately the emissions from drained organic forest soils,
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
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 information purposes in the Energy sector, but the non-CC>2 emissions
associated with biomass energy are included in the Energy sector emissions (see Chapter 3). EPA includes HWP
within the forest chapter because that is the source of wood that goes into the HWP estimates. EPA includes HWP
within the forest chapter because that is 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:
(IA) annual change of C in wood and paper products in use in the United States,
(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.
Land Use, Land-Use Change, and Forestry 6-35
-------
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 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 2020 net annual change for
forest C stocks was estimated to be between -744.6 and -592.2 MMT CO2 Eq. around a central estimate of-668.1
MMT CO2 Eq. at a 95 percent confidence level. This includes a range of-657.5 to -511.4 MMT CO2 Eq. around a
central estimate of-584.4 MMT CO2 Eq. for forest ecosystems and -106.4 to -63.1 MMT CO2 Eq. around a central
estimate of-83.6 MMT C02 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)
2020 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
(584.4)
(657.5)
(511.4)
-12.5%
12.5%
Harvested Wood Products'5
C02
(83.6)
(106.4)
(63.1)
-27.3%
24.5%
Total Forest
C02
(668.1)
(744.6)
(592.2)
-11.5%
11.4%
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
As discussed above, the FIA program has conducted consistent forest surveys based on extensive statistically-
based sampling of most of the forest land in the conterminous United States, 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 are archived and are publicly available on the Internet (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; FAQ 2021). Factors to convert wood and
6-36 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
paper to units of C are based on estimates by industry and 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 Cm emissions from landfills based on EPA (2006) data are reasonable in
comparison to Cm 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 2020 are consistent with those used in the previous (1990 through 2019)
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 increases in forest land area estimates and carbon
stocks, particularly in Alaska where new data from 2018 to 2020 (with the exception of litter and soil) were
included (Table 6-13). Fire data sources were also updated for AK through 2020 and this combined with the new
NFI data for the years 2018 through 2020 resulted in substantial changes in carbon stocks and stock changes. In
2019, in particular, an estimated 646,276 ha of forest land burned in AK—the fifth largest fire year in the time-
series— which resulted in substantial differences in the carbon stock and stock change estimates reported in the
previous (i.e., 1990 through 2019) Inventory and those in the current Inventory. Additionally, this report does not
include separate emission estimates for prescribed fires (a change from recent annual reports) because the data
records do not specify the fire origins allowing for separation of wild and prescribed fire emissions. Soil carbon
stocks increased in the latest Inventory relative to the previous Inventory and this change can be attributed to
refinements in the Digital General Soil Map of the United States (STATSG02) dataset where soil orders may have
changed in the updated data product (Table 6-13). This resulted in a structural change in the soil organic carbon
estimates for mineral and organic soils across the entire time series, particularly in AK where new data on forest
area was included for the years 2018 through 2020 (Table 6-8). Finally, recent land use change in AK (since 2015)
also contributed to variability in soil carbon stocks and stock changes in recent years in the time series which
contributed to differences in estimates in the 2021 Inventory and the current Inventory. New data in the HWP
time-series result in a minor decrease (< 1 percent) in carbon stocks in the HWP pools but a substantial decrease
(38 percent) in the carbon stock change estimates for Products in Use and to a lesser extent (7 percent) in SWDS
between the previous Inventory and the current Inventory. The new HWP data suggest a continued decline in
paper products in use over time due to changes in consumer behavior (i.e., more use of electronic information
sources) and a small drop in solid wood products in the last year due to a downturn in the economy associated
with the global pandemic.
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)
Forest Area (1000 ha)
Carbon Pools (MMT C)
Forest
Aboveground Biomass
Belowground Biomass
2020 Estimate,
Previous Inventory
279,289
55,933
15,260
3,103
2020 Estimate,
Current Inventory
282,061
58,156
15,579
3,085
2021 Estimate,
Current Inventory
281,951
58,316
15,688
3,106
Land Use, Land-Use Change, and Forestry 6-37
-------
Dead Wood
Litter
Soil (Mineral)
Soil (Organic)
Harvested Wood
Products in Use
SWDS
2,852
3,638
25,147
5,933
2,699
1,532
1,167
2,868
3,809
25,457
7,357
2,695
1,530
1,165
2,896
3,810
25,459
7,357
2,718
1,536
1,182
Total Stock
58,632
60,851
61,034
Note: Totals may not sum due to independent rounding.
Table 6-14: Recalculations of Net C Flux from Forest Ecosystem Pools in Forest Land
Remaining Forest Land and Harvested Wood Pools (MMT C)
Carbon Pool (MMT C)
2019 Estimate,
Previous Inventory
2019 Estimate,
Current Inventory
2020 Estimate,
Current Inventory
Forest
(159.1)
(148.9)
(159.4)
Aboveground Biomass
(107.4)
(107.2)
(108.7)
Belowground Biomass
(24.3)
(21.3)
(21.6)
Dead Wood
(27.1)
(26.5)
(27.7)
Litter
(0.1)
6.2
(0.5)
Soil (Mineral)
(0.7)
(0.2)
(1.1)
Soil (Organic)
0.3
(0.2)
0.1
Drained organic soil
0.2
0.2
0.2
Harvested Wood
(29.6)
(24.2)
(22.8)
Products in Use
(10.7)
(6.6)
(5.5)
SWDS
(18.9)
(17.6)
(17.3)
Total Net Flux
(188.7)
(173.1)
(182.2)
Note: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Planned Improvements
Reliable estimates of forest C stocks and changes across the diverse ecosystems of the United States require a high
level of investment in both annual monitoring and associated analytical techniques. Development of improved
monitoring/reporting techniques is a continuous process that occurs simultaneously with annual Inventory
submissions. Planned improvements can be broadly assigned to the following categories: development of a robust
estimation and reporting system, individual C pool estimation, coordination with other land-use categories, and
annual inventory data incorporation.
While this Inventory submission includes C change by Forest Land Remaining Forest Land and Land Converted to
Forest Land and C stock changes for all IPCC pools in these two categories, there are many improvements that are
still necessary. The estimation approach used for the CONUS in the current Inventory for the forest land category
operates at the state scale, whereas previously the western United States and southeast and southcentral coastal
Alaska operated at a regional scale. While this is an improvement over previous Inventories and led to improved
estimation and separation of land use categories in the current Inventory, research is 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 easier 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
6-38 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
QA/QC in reporting and further improve transparency, completeness, consistency, accuracy, and availability of
data used in reporting. Finally, a combination of approaches were 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 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 and western Oklahoma, all other states in
the CONUS 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 Alaska will be used in the years ahead
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 detection and
attribution across the entire reporting period and all managed forest land in the United States. Leveraging this
auxiliary information will aid not only the interior Alaska effort but the entire inventory system. In addition to fully
inventorying all managed forest land in the United States, the more intensive sampling of fine woody debris, litter,
and SOC on a subset of FIA plots continues and will substantially improve resolution of C pools (i.e., greater sample
intensity; 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. 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 2020, emissions from this source
were estimated to be 13.7 MMT CO2 Eq. of CH4 and 11.7 MMT CO2 Eq. of N2O (Table 6-15; kt units provided in
Table 6-16). The estimates of non-CC>2 emissions from forest fires are for the conterminous 48 states and all
managed forest land in Alaska (Ogle et al. 2018).
Table 6-15: N011-CO2 Emissions from Forest Fires (MMT CO2 Eq.)a
Gas
1990
2005
2016
2017
2018
2019
2020
ch4
2.3
6.5
3.9
9.5
6.2
1.1
13.6
n2o
1.8
6.3
3.9
8.2
5.7
1.3
11.7
Total
4.1
12.8
7.8
17.7
11.9
2.5
25.3
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
Table 6-16: N011-CO2 Emissions from Forest Fires (kt)a
Gas 1990 2005 2016 2017 2018 2019 2020
Land Use, Land-Use Change, and Forestry 6-39
-------
CH4 92 260 154 381 249 45 545
n2o
6
21
13
27
19
4
39
CO
2,589
7,284
3,775
8,591
5,457
1,095
11,739
NOx
47
120
87
167
119
30
224
a These estimates include Non-C02 Emissions from Forest Fires on Forest Land Remaining
Forest Land and Land Converted to Forest Land.
Methodology and Time-Series Consistency
Non-CC>2 emissions from forest fires—primarily CFU and N2O emissions—were calculated consistent with IPCC
(2006) methodology, which included U.S.-specific data and models on area, fuel, consumption, and emission. The
annual estimates were calculated by the Wildland Fire Emissions Inventory System (WFEIS, French et al. 2011,
2014) with area burned based on Monitoring Trends in Burn Severity (MTBS, Eidenshink et al. 2007) or MODIS
burned area mapping (MODIS MCD64A1, Giglio et al. 2018) data. The MTBS data available for this report (MTBS
2021) included fires through 2018 with only a partial set of the 2019 fires included with the data. The MODIS-
based records include 2001 through 2020. Emissions reported here originate from MTBS data for the 1990 to 2018
interval, and the 2019 and 2020 emissions are based on MODIS burned areas. All other parts of calculations—fuels,
fire characteristics, and emissions—are via WFEIS and therefore identical throughout the 1990 to 2020 interval.
Note that N2O emissions are not included in WFEIS calculations; the emissions provided here are based on the
average N2O to CO2 ratio of 0.000166 following Larkin et al. (2014). See Emissions from Forest Fires in Annex 3.13
for further details on all fire-related emissions calculations for forests. Consistent data sources, data processing,
and calculation methods were applied to the entire time series to ensure time-series consistency from 1990
through 2020.
Uncertainty
In order to quantify the uncertainties for non-C02 emissions from forest fires, a Monte Carlo (IPCC Approach 2)
sampling approach was employed to propagate uncertainties in per-fire quantities of fuel and forest area burned.
See Annex 3.13 for the quantities and assumptions employed to define and propagate uncertainty. The results of
the Approach 2 quantitative uncertainty analysis are summarized in Table 6-17.
Table 6-17: Quantitative Uncertainty Estimates of Non-C02 Emissions from Forest Fires
(MMT CO2 Eq. and Percent)3
Source
Gas
2020 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
13.6
8.6
19.3
-37%
42%
Non-C02 Emissions from
Forest Fires
n2o
11.7
7.6
16.3
-35%
39%
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.
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 estimating non-C02 emissions from forest fires included checking input data, documentation,
and calculations to ensure data were properly handled through the inventory process and results were consistent
with values expected from those calculations. The QA/QC procedures did not reveal any inaccuracies or incorrect
input values.
6-40 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Recalculations Discussion
The methods used in the current (1990 through 2020) Inventory to compile estimates of non-CC>2 emissions from
forest fires represent a change relative to the previous (1990 through 2019) Inventory. The basic components of
calculating forest fire emissions (IPCC 2006) remain unchanged, but the WFEIS-based estimates for fuel,
combustion, and allocation of emissions provide both increased specificity (for site and fire) and more consistent
application of these factors. An additional source of change recalculation are recent updates to the MTBS fire
records (post-2000 fires). These recalculations resulted in a 10 percent increase in average annual emissions over
the 2014 to 2018 interval as compared to the previous Inventory (interval represents years with emissions
estimated via both past and current methods for comparison).
Planned Improvements
Continuing improvements are planned for developing better fire and site-specific estimates for forest fires. The
focus will be on addressing three aspects of reporting: best use of WFEIS, better resolution of uncertainty, and
identification of forest area burned but not included in the MTBS records.
N20 Emissions from N Additions to Forest Soils
Of the synthetic nitrogen (N) fertilizers applied to soils in the United States, no more than one percent is applied to
forest soils. Application rates are similar to those occurring on cropland soils, but in any given year, only a small
proportion of total forested land receives N fertilizer. This is because forests are typically fertilized only twice
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 the off-site location. 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 2020
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 2020
were 0.5 MMT CO2 Eq. (1.5 kt) and have increased by 455 percent from 1990 to 2020. 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
2016
2017
2018
2019
2020
Direct N20 Fluxes from Soils
MMTCO2 Eq.
0.1
0.3
0.3
0.3
0.3
0.3
0.3
kt N20
+
1.2
1.2
1.2
1.2
1.2
1.2
Indirect N20 Fluxes from Soils
MMTCO2 Eq.
+
0.1
0.1
0.1
0.1
0.1
0.1
kt N20
+
+
+
+
+
+
+
Total
MMT CO? Eq.
0.1
0.5
0.5
0.5
0.5
0.5
0.5
kt N20
+
1.5
1.5
1.5
1.5
1.5
1.5
+ Does not exceed 0.05 MMT C02 Eq. or 0.5 kt.
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.
Land Use, Land-Use Change, and Forestry 6-41
-------
Note: 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
2020, 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
area and the portion of fertilized area are multiplied to obtain annual area estimates of fertilized Douglas-fir
stands. Similar to the Southeast, data are not available for 2005 through 2020, so data from 2004 are used for
these years. The annual area estimates are multiplied by the typical rate used in this region (200 lbs. N per acre) to
estimate total N applied (Briggs 2007), and the total N applied to forests is multiplied by the IPCC (2006) default
emission factor of one percent to estimate direct N2O emissions.
For indirect emissions, the volatilization and leaching/runoff N fractions for forest land are calculated using the
IPCC default factors of 10 percent and 30 percent, respectively. 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. 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.
The same method is applied in all years of this Inventory to ensure time-series consistency from 1990 through
2020.
Uncertainty
The amount of N2O emitted from forests depends not only on N inputs and fertilized area, but also on a large
number of variables, including organic C availability, oxygen gas partial pressure, soil moisture content, pH,
temperature, and tree planting/harvesting cycles. The effect of the combined interaction of these variables on N2O
flux is complex and highly uncertain. IPCC (2006) does not incorporate any of these variables into the default
methodology, except variation in estimated fertilizer application rates and estimated areas of forested land
receiving N fertilizer. All forest soils are treated equivalently under this methodology. Furthermore, only
applications of synthetic N fertilizers to forest are captured in this inventory, so applications of organic N fertilizers
are not estimated. However, the total quantity of organic N inputs to soils in the United States is included in the
inventory for Agricultural Soil Management (Section 5.4) and Settlements Remaining Settlements (Section 6.10).
Uncertainties exist in the fertilization rates, annual area of forest lands receiving fertilizer, and the emission
factors. Fertilization rates are assigned a default level34 of uncertainty at ±50 percent, and area receiving fertilizer
is assigned a ±20 percent according to expert knowledge (Binkley 2004). The uncertainty ranges around the 2004
activity data and emission factor input variables are directly applied to the 2020 emission estimates. IPCC (2006)
34 Uncertainty is unknown for the fertilization rates so a conservative value of ±50 percent is used in the analysis.
6-42 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
provided estimates for the uncertainty associated with direct and indirect N2O emission factor for synthetic N
fertilizer application to soils.
Uncertainty is quantified using simple error propagation methods (IPCC 2006). The results of the quantitative
uncertainty analysis are summarized in Table 6-19. Direct N2O fluxes from soils in 2020 are estimated to be
between 0.1 and 1.1 MMT CO2 Eq. at a 95 percent confidence level. This indicates a range of 59 percent below and
211 percent above the emission estimate of 0.3 MMT CO2 Eq. for 2020. Indirect N2O emissions in 2020 are 0.1
MMT CO2 Eq. and have a range are between 0.02 and 0.4 MMT CO2 Eq., which is 86 percent below to 238 percent
above the emission estimate for 2020.
Table 6-19: Quantitative Uncertainty Estimates of N2O Fluxes from Soils in Forest Land
Remaining Forest Land and Land Con verted to Forest Land (MMT CO2 Eq. and Percent)
Source
Gas
2020 Emission Estimate
(MMT C02 Eq.)
Uncertainty Range Relative to Emission Estimate
(MMT C02 Eq.) (%)
Forest Land Remaining Forest
Lower
Upper
Lower
Upper
Land
Bound
Bound
Bound
Bound
Direct N20 Fluxes from Soils
N20
0.3
0.1
1.1
-59%
+211%
Indirect N20 Fluxes from Soils
n2o
0.1
+
0.4
-86%
+238%
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding
QA/QC and Verification
The spreadsheet containing fertilizer applied to forests and calculations for N2O and uncertainty ranges are
checked and verified based on the sources of these data.
Recalculations Discussion
No recalculations were performed for the 1990 to 2019 estimates.
C02, CH4, and N20 Emissions from Drained Organic Soils35
Drained organic soils on forest land are identified separately from other forest soils largely because mineralization
of the exposed or partially dried organic material results in continuous CO2 and N2O emissions (IPCC 2006). In
addition, the 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands
(IPCC 2014) calls for estimating CH4 emissions from these drained organic soils and the ditch networks used to
drain them.
Organic soils are identified on the basis of thickness of organic horizon and percent organic matter. All organic soils
are assumed to have originally been wet, and drained organic soils are further characterized by drainage or the
process of artificially lowering the soil water table, which exposes the organic material to drying and the associated
emissions described in this section. The land base considered here is drained inland organic soils that are
coincident with forest area as identified by the NFI of the USDA Forest Service (USDA Forest Service 2022b).
The estimated area of drained organic soils on forest land is 70,849 ha and did not change over the time series
based on the data used to compile the estimates in the current Inventory. These estimates are based on
permanent plot locations of the NFI (USDA Forest Service 2022b) coincident with mapped organic soil locations
(STATSG02 2016), which identifies forest land on organic soils. Forest sites that are drained are not explicitly
35 Estimates of C and 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.
Land Use, Land-Use Change, and Forestry 6-43
-------
identified in the data, but for this estimate, planted forest stands on sites identified as mesic or xeric (which are
identified in USDA Forest Service 2022c, d) are labeled "drained organic soil" sites.
Land use, region, and climate are broad determinants of emissions as are more site-specific factors such as
nutrient status, drainage level, exposure, or disturbance. Current data are limited in spatial precision and thus lack
site specific details. At the same time, corresponding emissions factor data specific to U.S. forests are similarly
lacking. Tier 1 estimates are provided here following IPCC (2014). Total annual non-CC>2 emissions on forest land
with drained organic soils in 2020 are estimated as 0.8 MMT CO2 Eq. per year (Table 6-20; kt units provided in
6-21).
The Tier 1 methodology provides methods to estimate C emission as CO2 from three pathways: direct emissions
primarily from mineralization; indirect, or off-site, emissions associated with dissolved organic carbon releasing
CO2 from drainage waters; and emissions from (peat) fires on organic soils. Data about forest fires specifically
located on drained organic soils are not currently available; as a result, no corresponding estimate is provided
here. Non-CC>2 emissions provided here include CH4 and N2O. 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. Emission of N2O can be significant from these drained organic soils
in contrast to the very low emissions from wet organic soils.
Table 6-20: N011-CO2 Emissions from Drained Organic Forest Soilsa'b (MMT CO2 Eq.)
Source
1990
2005
2016
2017
2018
2019
2020
ch4
+
+
+
+
+
+
+
n2o
0.1
0.1
0.1
0.1
0. 1
0.1
0.1
Total
0.8
0.8
0.8
0.8
0.8
0.8
0.8
+ 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 C and 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.
Table 6-21: Non-C02 Emissions from Drained Organic Forest Soilsa'b (kt)
Source
1990
2005
2016
2017
2018
2019
2020
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 C and 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.
Methodology and Time-Series Consistency
The Tier 1 methods for estimating CO2, CFU and N2O emissions from drained inland organic soils on forest lands
follow IPCC (2006), with extensive updates and additional material presented in the 2013 Supplement to the 2006
IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands (IPCC 2014). With the exception of quantifying
area of forest on drained organic soils, which is user-supplied, all quantities necessary for Tier 1 estimates are
provided in Chapter 2, Drained Inland Organic Soils of IPCC (2014).
Estimated area of drained organic soils on forest land is 70,849 ha based on analysis of the permanent NFI of the
USDA Forest Service and did not change over the time series. The most recent plot data per state within the
6-44 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
inventories were used in a spatial overlay with the STATSG02 (2016) soils data, and forest plots coincident with the
soil order histosol were selected as having organic soils. Information specific to identifying "drained organic" are
not in the inventory data so an indirect approach was employed here. Specifically, artificially regenerated forest
stands (inventory field STDORGCD=l) on mesic or xeric sites (inventory field 11
-------
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 less significant digits in the resulting estimates, relative to past values. The
total non-CC>2 emissions in 2020 from drained organic soils on Forest Land Remaining Forest Land and Land
Converted to Forest Land were estimated to be between -0.006 and 0.162 MMT CO2 Eq. around a central estimate
of 0.073 MMT CO2 Eq. at a 95 percent confidence level.
Table 6-23: Quantitative Uncertainty Estimates for Non-C02 Emissions on Drained Organic
Forest Soils (MMT CO2 Eq. and Percent)3
2020 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
-109%
+123%
+ 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.
QA/QC and Verification
IPCC (2014) guidance cautions of a possibility of double counting some of these emissions. Specifically, the off-site
emissions of dissolved organic C from drainage waters may be double counted if soil C stock and change is based
on sampling and this C is captured in that sampling. Double counting in this case is unlikely since plots identified as
drained were treated separately in this chapter. Additionally, some of the non-CC>2 emissions may be included in
either the Wetlands or sections on N2O emissions from managed soils. These paths to double counting emissions
are unlikely here because these issues are taken into consideration when developing the estimates and this
chapter is the only section directly including such emissions on forest land.
Recalculations Discussion
No recalculations were performed for the 1990 through 2019 estimates.
Planned Improvements
Additional data will be compiled to update estimates of forest areas on drained organic soils as new reports are
made available and new geospatial products become available.
6-46 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 40
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 2020 was -99.5 MMT
C02 Eq. (-27.1 MMT C) (Table 6-24 and Table 6-25).
Mineral soil C stocks increase 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 conversion to
forest in Alaska and Hawaii were not included. Since area estimates for some land use categories were not updated in the Land
Representation in the current Inventory there are 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, 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.
Land Use, Land-Use Change, and Forestry 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.)
Land Use/Carbon Pool
1990
2005
2016
2017
2018
2019
2020
Cropland Converted to Forest Land
(39.6)
(39.5)
(39.5)
(39.6)
(39.6)
(39.6)
(39.6)
Aboveground Biomass
(23.0)
(23.0)
(23.0)
(23.0)
(23.0)
(23.0)
(23.0)
Belowground Biomass
(4.4)
(4.4)
(4.4)
(4.4)
(4.4)
(4.4)
(4.4)
Dead Wood
(5.0)
(5.0)
(5.0)
(5.0)
(5.0)
(5.0)
(5.0)
Litter
(6.9)
(6.9)
(6.9)
(6.9)
(6.9)
(6.9)
(6.9)
Mineral Soil
(0.3)
(0.3)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Grassland Converted to Forest Land
(11.5)
(11.6)
(11.8)
(11.8)
(11.8)
(11.8)
(11.8)
Aboveground Biomass
(5.9)
(6.0)
(6.1)
(6.1)
(6.1)
(6.1)
(6.1)
Belowground Biomass
(1.0)
(1.0)
(1.0)
(1.0)
(1.0)
(1.0)
(1.0)
Dead Wood
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
Litter
(3.8)
(3.8)
(3.9)
(3.9)
(3.9)
(3.9)
(3.9)
Mineral Soil
0.2
0.3
0.3
0.3
0.3
0.3
0.3
Other Land Converted to Forest Land
(10.1)
(10.7)
(10.9)
(10.9)
(10.9)
(10.9)
(10.9)
Aboveground Biomass
(4.7)
(4.8)
(4.9)
(4.9)
(4.9)
(4.9)
(4.9)
Belowground Biomass
(0.8)
(0.9)
(0.9)
(0.9)
(0.9)
(0.9)
(0.9)
Dead Wood
(1.4)
(1.4)
(1.4)
(1.4)
(1.4)
(1.4)
(1.4)
Litter
(2.5)
(2.5)
(2.6)
(2.6)
(2.6)
(2.6)
(2.6)
Mineral Soil
(0.6)
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
Settlements Converted to Forest Land
(34.2)
(34.1)
(34.1)
(34.1)
(34.1)
(34.1)
(34.1)
Aboveground Biomass
(20.9)
(20.9)
(20.8)
(20.8)
(20.8)
(20.8)
(20.8)
Belowground Biomass
(4.0)
(4.0)
(4.0)
(4.0)
(4.0)
(4.0)
(4.0)
Dead Wood
(3.9)
(3.9)
(3.9)
(3.9)
(3.9)
(3.9)
(3.9)
Litter
(5.4)
(5.3)
(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
(3.2)
(3.2)
(3.2)
(3.2)
(3.2)
(3.2)
(3.2)
Aboveground Biomass
(1.4)
(1.4)
(1.4)
(1.4)
(1.4)
(1.4)
(1.4)
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
(1.2)
(1.2)
(1.2)
(1.2)
(1.2)
(1.2)
(1.2)
Mineral Soil
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Total Aboveground Biomass Flux
(55.9)
(55.9)
(56.1)
(56.1)
(56.1)
(56.1)
(56.1)
Total Belowground Biomass Flux
(10.5)
(10.5)
(10.5)
(10.5)
(10.5)
(10.5)
(10.5)
Total Dead Wood Flux
(11.7)
(11.7)
(11.8)
(11.8)
(11.8)
(11.8)
(11.8)
Total Litter Flux
(19.8)
(19.8)
(19.9)
(19.9)
(19.9)
(19.9)
(19.9)
Total Mineral Soil Flux
(0.8)
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
Total Flux
(98.6)
(99.1)
(99.5)
(99.5)
(99.5)
(99.5)
(99.5)
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. 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. Since area estimates for some land use categories were not updated in the Land Representation in the current
Inventory there are differences in the area estimates reported in this section and those reported in Section 6.1
Representation of the U.S. Land Base. 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.
6-48 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 6-25: NetCFIuxfrom Forest C Pools in Land Converted to Forest Land by Land Use
Change Category (MMT C)
Land Use/Carbon Pool
1990
2005
2016
2017
2018
2019
2020
Cropland Converted to Forest
(10.8)
(10.8)
(10.8)
(10.8)
(10.8)
(10.8)
(10.8)
Aboveground Biomass
(6.3)
(6.3)
(6.3)
(6.3)
(6.3)
(6.3)
(6.3)
Belowground Biomass
(1.2)
(1.2)
(1.2)
(1.2)
(1.2)
(1.2)
(1.2)
Dead Wood
(1.4)
(1.4)
(1.4)
(1.4)
(1.4)
(1.4)
(1.4)
Litter
(1.9)
(1.9)
(1.9)
(1.9)
(1.9)
(1.9)
(1.9)
Mineral Soil
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Grassland Converted to Forest
(3.1)
(3.2)
(3.2)
(3.2)
(3.2)
(3.2)
(3.2)
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
(2.7)
(2.9)
(3.0)
(3.0)
(3.0)
(3.0)
(3.0)
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
(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
(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)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
Mineral Soil
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Total Aboveground Biomass Flux
(15.2)
(15.3)
(15.3)
(15.3)
(15.3)
(15.3)
(15.3)
Total Belowground Biomass Flux
(2.9)
(2.9)
(2.9)
(2.9)
(2.9)
(2.9)
(2.9)
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)
(5.4)
(5.4)
(5.4)
(5.4)
(5.4)
Total Mineral Soil Flux
(0.2)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
Total Flux
(26.9)
(27.0)
(27.1)
(27.1)
(27.1)
(27.1)
(27.1)
+ 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. 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. Since area estimates for some land use categories were not updated in the Land Representation
in the current Inventory there are differences in the area estimates reported in this section and those reported in Section
6.1 Representation of the U.S. Land Base 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.
Land Use, Land-Use Change, and Forestry 6-49
-------
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 population estimates. 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 stocks removed due to land use conversion from Croplands and Grasslands were used in the year of
conversion on individual plots. All annual NFI plots available through August 2021 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 2020. 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
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. 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
6-50 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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.
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.
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 2020 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 2020 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 2020 C stock
change estimate of-99.5 MMT CO2 Eq.
Land Use, Land-Use Change, and Forestry 6-51
-------
Table 6-26: Quantitative Uncertainty Estimates for Forest C Pool Stock Changes (MMT CO2
Eq. per Year) in 2020 from Land Converted to Forest Land by Land Use Change
Land Use/Carbon Pool
2020 Flux
Uncertainty Range Relative to Flux Range1
a
Estimate
(MMT C02 Eq.)
(MMT C02
Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Cropland Converted to Forest Land
(39.6)
(48.2)
(30.9)
-22%
22%
Aboveground Biomass
(23.0)
(31.4)
(14.5)
-37%
37%
Belowground Biomass
(4.4)
(5.5)
(3.4)
-24%
24%
Dead Wood
(5.0)
(6.2)
(3.8)
-24%
24%
Litter
(6.9)
(8.0)
(5.8)
-16%
16%
Mineral Soils
(0.2)
(0.5)
0.1
-134%
134%
Grassland Converted to Forest Land
(11.8)
(14.3)
(9.4)
-21%
20%
Aboveground Biomass
(6.1)
(7.5)
(4.7)
-23%
23%
Belowground Biomass
(1.0)
(1.3)
(0.7)
-28%
28%
Dead Wood
(1.1)
(1.3)
(1.0)
-13%
13%
Litter
(3.9)
(4.4)
(3.4)
-14%
14%
Mineral Soils
0.3
(0.1)
0.6
-136%
136%
Other Lands Converted to Forest Land
(10.9)
(13.2)
(8.5)
-22%
22%
Aboveground Biomass
(4.9)
(7.0)
(2.8)
-43%
43%
Belowground Biomass
(0.9)
(1.3)
(0.4)
-50%
50%
Dead Wood
(1.4)
(2.0)
(0.8)
-40%
40%
Litter
(2.6)
(3.2)
(2.0)
-25%
25%
Mineral Soils
(1.1)
(1.9)
(0.4)
-66%
66%
Settlements Converted to Forest Land
(34.1)
(40.6)
(27.6)
-19%
19%
Aboveground Biomass
(20.8)
(27.0)
(14.6)
-30%
30%
Belowground Biomass
(4.0)
(5.3)
(2.7)
-33%
33%
Dead Wood
(3.9)
(5.0)
(2.7)
-30%
30%
Litter
(5.3)
(6.2)
(4.4)
-17%
17%
Mineral Soil
(0.1)
(0.1)
+
-44%
44%
Wetlands Converted to Forest Land
(3.2)
(3.4)
(3.0)
-5%
5%
Aboveground Biomass
(1.4)
(1.5)
(1.2)
-10%
10%
Belowground Biomass
(0.3)
(0.3)
(0.2)
-12%
12%
Dead Wood
(0.4)
(0.4)
(0.3)
-11%
11%
Litter
(1.2)
(1.3)
(1.1)
-5%
5%
Mineral Soils
0.0
0.0
0.0
NA
NA
Total: Aboveground Biomass
(56.1)
(66.9)
(45.4)
-19%
19%
Total: Belowground Biomass
(10.5)
(12.3)
(8.8)
-17%
17%
Total: Dead Wood
(11.8)
(13.6)
(10.0)
-15%
15%
Total: Litter
(19.9)
(21.6)
(18.4)
-8%
8%
Total: Mineral Soils
(1.1)
(1.7)
(0.6)
-50%
50%
Total: Lands Converted to Forest Lands
(99.5)
(110.7)
(88.3)
-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.
/erification
See QA/QC and Verification sections under Forest Land Remaining Forest Land and for mineral soil estimates
Cropland Remaining Cropland.
6-52 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 available
through August 2020 were used in this Inventory. This is the third year that remeasurement data from the annual
NFI were available throughout the CONUS (with the exception of Wyoming and western Oklahoma) 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 less than 1 percent in 2019 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. In the previous Inventory, Grasslands Converted to Forest Land represented the largest transfer and
uptake of C across the land use conversion categories. In this Inventory, Cropland Converted to Forest Land
represented the largest transfer and uptake of C across the land use change categories followed by Settlements
Converted to Forest Land (Table 6-27).
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
and Carbon pool (MMT C)
2019 Estimate,
Previous Inventory
2019 Estimate,
Current Inventory
2020 Estimate,
Current Inventory
Cropland Converted to Forest Land
(10.9)
(10.8)
(10.8)
Aboveground Biomass
(6.3)
(6.3)
(6.3)
Belowground Biomass
(1.2)
(1.2)
(1.2)
Dead Wood
(1.4)
(1.4)
(1.4)
Litter
(1.9)
(1.9)
(1.9)
Mineral soil
(0.1)
(0.1)
(0.1)
Grassland Converted to Forest Land
(2.9)
(3.2)
(3.2)
Aboveground Biomass
(1.3)
(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)
(3.0)
(3.0)
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
(9.4)
(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)
(0.3)
(0.3)
Land Use, Land-Use Change, and Forestry 6-53
-------
Mineral soil
0.0
0.0
0.0
Total Aboveground Biomass Flux
(15.0)
(15.3)
(15.3)
Total Belowground Biomass Flux
(2.9)
(2.9)
(2.9)
Total Dead Wood Flux
(3.2)
(3.2)
(3.2)
Total Litter Flux
(5.6)
(5.4)
(5.4)
Total SOC (mineral) Flux
(0.3)
(0.3)
(0.3)
Total Flux
(27.0)
(27.1)
(27.1)
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.
(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. 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
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.
6-54 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
(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).
Cropland in Alaska is not included in the Inventory, but is a relatively small amount of U.S. cropland area
(approximately 28,700 hectares). 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). This leads
to a small discrepancy between the managed area in Cropland Remaining Cropland (see Table 6-31 in Planned
Improvements for more details on the land area discrepancies) and the cropland area included in the Inventory
analysis. Improvements are underway to include croplands in Alaska as part of future C inventories.
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 2020, mineral soils are estimated to
sequester 56.2 MMT CO2 Eq. from the atmosphere (15.3 MMT C). This rate of C storage in mineral soils represents
about a 3 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 2020, which is a 6 percent decrease compared to 1990. In total,
United States agricultural soils in Cropland Remaining Cropland sequestered approximately 23.3 MMT CO2 Eq. (6.4
MMT C) in 2020.
Table 6-28: Net CO2 Flux from Soil C Stock Changes in Cropland Remaining Cropland (MMT
COz Eq.)
Soil Type
1990
2005
2016
2017
2018
2019
2020
Mineral Soils
(58.2)
(62.4)
(54.3)
(55.1)
(49.4)
(47.4)
(56.2)
Organic Soils
35.0
33.4
31.6
32.8
32.8
32.9
32.9
Total Net Flux
(23.2)
(29.0)
(22.7)
(22.3)
(16.6)
(14.5)
(23.3)
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
39 N20 emissions from drained organic soils are included in the Agricultural Soil Management section of the Agriculture chapter
of the Inventory.
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 6-55
-------
Table 6-29: Net CO2 Flux from Soil C Stock Changes in Cropland Remaining Cropland (WAT
C)
Soil Type
1990
2005
2016
2017
2018
2019
2020
Mineral Soils
(15.9)
(17.0)
(14.8)
(15.0)
(13.5)
(12.9)
(15.3)
Organic Soils
9.5
9.1
8.6
8.9
8.9
9.0
9.0
Total Net Flux
(6.3)
(7.9)
(6.2)
(6.1)
(4.5)
(4.0)
(6.4)
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., 2020 is 0.7 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
series with a small decline associated with the land base declining for Cropland Remaining Cropland on organic
soils since 1990.
The spatial variability in the 2015 annual soil organic C stock changes41 are displayed in Figure 6-6 and Figure 6-7
for mineral and organic soils, respectively. Isolated areas with high rates of C accumulation occur throughout the
agricultural land base in the United States, but there are more concentrated areas. In particular, higher rates of net
C accumulation in mineral soils occur in the Corn Belt region, which is the region with the largest amounts of
conservation tillage, along with moderate rates of CRP enrollment. The regions with the highest rates of emissions
from drainage of organic soils occur in the Southeastern Coastal Region (particularly Florida), upper Midwest and
Northeast surrounding the Great Lakes, and isolated areas along the Pacific Coast (particularly California), which
coincides with the largest concentrations of organic soils in the United States that are used for agricultural
production.
41 Only national-scale emissions are estimated for 2016 to 2020 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.
6-56 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure 6-6: Total Net Annual Soil C Stock Changes for Mineral Soils under Agricultural
Management within States, 2015, Cropland Remaining Cropland
~ -1 to 1
Note: Only national-scale soil organic C stock changes are estimated for 2016 to 2020 in the current Inventory
using a surrogate data method, and therefore the fine-scale emission patterns in this map are based on
inventory data from 2015. Negative values represent a net increase in soil organic C stocks, and positive values
represent a net decrease in soil organic C stocks.
Land Use, Land-Use Change, and Forestry 6-57
-------
Figure 6-7: Total Net Annual Soil C Stock Changes for Organic Soils under Agricultural
Management within States, 2015, Cropland Remaining Cropland
¦ >40
Note: Only natiorial-scale soil organic C stock changes are estimated for 2016 to 2019 in the current Inventory
using a surrogate data method, and therefore the fine-scale emission patterns in this map are based on
inventory data from 2015.
Methodology and Time-Series Consistency
The following section includes a description of the methodology used to estimate changes in soil organic C stocks
for Cropland Remaining Cropland, including (1) agricultural land-use and management activities on mineral soils;
and (2) agricultural land-use and management activities on organic soils. Carbon dioxide emissions and removals42
due to changes in mineral soil organic C stocks are estimated using a Tier 3 method for the majority of annual
crops (Ogle et al. 2010). A Tier 2 IPCC method is used for the remaining crops not included in the Tier 3 method
(see list of crops in the Mineral Soil Carbon Stock Changes section below) (Ogle et al. 2003, 2006). In addition, a
Tier 2 method is used for very gravelly, cobbly, or shaley soils (i.e., classified as soils that have greater than 35
percent of soil volume comprised of gravel, cobbles, or shale, regardless of crop). Emissions from organic soils are
estimated using a Tier 2 IPCC method. While a combination of Tier 2 and 3 methods are used to estimate C stock
changes across most of the time series, a surrogate data method has been applied to estimate stock changes in the
last few years of the Inventory. Stock change estimates based on surrogate data will be recalculated in a future
Inventory report using the Tier 2 and 3 methods when data become available.
Soil organic C stock changes on non-federal lands are estimated for Cropland Remaining Cropland (as well as
agricultural land falling into the IPCC categories Land Converted to Cropland, Grassland Remaining Grassland, and
Land Converted to Grassland) according to land-use histories recorded in the USDA NRI survey (USDA-NRCS
2018a). The NRI is a statistically-based sample of all non-federal land, and includes approximately 489,178 survey
locations in agricultural land for the conterminous United States and Hawaii. Each survey location is associated
with an "expansion factor" that allows scaling of C stock changes from NRI survey locations to the entire country
42 Removals occur through uptake of C02 into crop and forage biomass that is later incorporated into soil C pools.
6-58 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
(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 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 2020 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 2020 will be recalculated in future Inventories with an updated time series of
activity data.
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
43 See https://quickstats.nass.usda.gov/.
Land Use, Land-Use Change, and Forestry 6-59
-------
(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 2020.
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
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).
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.
6-60 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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.
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
Land Use, Land-Use Change, and Forestry 6-61
-------
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.
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
2020 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 2020 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 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
6-62 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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.
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 2020 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.
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 2020 will be recalculated in a future inventory when new activity data are
incorporated into the analysis.
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 for further
discussion). For 2016 to 2020, 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
Land Use, Land-Use Change, and Forestry 6-63
-------
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 317 percent below
to 317 percent above the 2020 stock change estimate of -23.3 MMT CO2 Eq. The large relative uncertainty around
the 2020 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 with this splicing method.
Table 6-30: Approach 2 Quantitative Uncertainty Estimates for Soil C Stock Changes
occurring within Cropland Remaining Cropland (MMT CO2 Eq. and Percent)
Source
2020 Flux Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Flux Estimate3
(MMT C02 Eq.) (%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Mineral Soil C Stocks: Cropland Remaining
Cropland, Tier 3 Inventory Methodology
Mineral Soil C Stocks: Cropland Remaining
Cropland, Tier 2 Inventory Methodology
Organic Soil C Stocks: Cropland Remaining
Cropland, Tier 2 Inventory Methodology
(51.4)
(4.9)
32.9
(122.4)
(11.9)
13.9
19.7
2.1
51.9
-138%
-144%
-58%
+138%
+144%
+58%
Combined Uncertainty for Flux associated
with Agricultural Soil Carbon Stock Change in
Cropland Remaining Cropland
(23.3)
(97.2)
50.5
-317%
+317%
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
assess uncertainties in the predictive capability of the model (Ogle et al. 2007). The comparisons include 72 long-
term experiment sites and 142 NRI soil monitoring network sites, with 948 observations across all of the sites (see
Annex 3.12 for more information).
Recalculations Discussion
There are no recalculations in the time series from the previous Inventory.
6-64 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Planned Improvements
A key improvement for a future Inventory will be to incorporate additional management activity data from the
USDA-NRCS Conservation Effects Assessment Project survey. This survey has compiled new data in recent years
that will be available for the Inventory analysis by next year. The latest land use data will also be incorporated from
the USDA National Resources Inventory and related management data from USDA-ERS ARMS surveys.
There are several other planned improvements underway related to the plant production module. Crop
parameters associated with temperature effects on plant production will be further improved in DayCent with
additional model calibration. Senescence events following grain filling in crops, such as wheat, are being modified
based on recent model algorithm development, and will be incorporated. There will also be further testing and
parameterization of the DayCent model to reduce the bias in model predictions for grasslands, which was
discovered through model evaluation by comparing output to measurement data from 72 experimental sites and
142 NRI soil monitoring network sites (See QA/QC and Verification section).
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 are used in the Field Burning of Agricultural Residues source category
(see Section 5.7). This improvement will more accurately represent the C inputs to the soil that are associated with
residue burning.
A review of available data on biosolids (i.e., treated sewage sludge) application will be undertaken to improve the
distribution of biosolids application on croplands, grasslands and settlements.
In the future, the Inventory will include an analysis of C stock changes in Alaska for cropland, using the Tier 2
method for mineral and organic soils that is described earlier in this section. This analysis will initially focus on land
use change, which typically has a larger impact on soil organic C stock changes than management practices, but
will be further refined over time to incorporate management data. See Table 6-31 for the amount of managed area
in Cropland Remaining Cropland that is not included in the Inventory, which is less than one thousand hectares per
year. This includes the area in Alaska and also other miscellaneous cropland areas, such as aquaculture.
Many of these improvements are expected to be completed for the 1990 through 2021 Inventory (i.e., 2023
submission to the UNFCCC). However, the timeline may be extended if there are insufficient resources to fund all
or part of these planned improvements.
Table 6-31: Area of Managed Land in Cropland Remaining Cropland that is not included in
the current Inventory (Thousand Hectares)
Area (Thousand Hectares)
Year
Managed Land Inventory
Not Included in
Inventory
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
162,163 162,134
161,721 161,692
161,252 161,223
159,449 159,420
157,732 157,703
157,054 157,025
156,409 156,380
155,767 155,738
152,016 151,987
151,135 151,105
150,981 150,952
150,471 150,442
150,175 150,146
150,843 150,814
150,645 150,616
29
29
29
29
29
29
29
29
29
29
29
29
29
29
29
Land Use, Land-Use Change, and Forestry 6-65
-------
2005
150,304
150,275
29
2006
149,791
149,762
29
2007
150,032
150,003
29
2008
149,723
149,694
29
2009
149,743
149,714
29
2010
149,343
149,314
29
2011
148,844
148,815
29
2012
148,524
148,495
29
2013
149,018
148,989
29
2014
149,492
149,463
29
2015
148,880
148,851
29
2016
ND
ND
ND
2017
ND
ND
ND
2018
ND
ND
ND
2019
ND
ND
ND
2020
ND
ND
ND
Note: NRI data are not available after 2015, and so these years are
designated as ND (No data).
6.5 Land Converted to Cropland (CRF
Category 4B2)
Land Converted to Cropland includes all cropland in an inventory year that had been in another land use(s) during
the previous 20 years (USDA-NRCS 2018), and used to produce food or fiber, or forage that is harvested and used
as feed (e.g., hay and silage). For example, grassland or forest land converted to cropland during the past 20 years
would be reported in this category. Recently converted lands are retained in this category for 20 years as
recommended by IPCC (2006). This Inventory includes all croplands in the conterminous United States and Hawaii,
but does not include a minor amount of Land Converted to Cropland in Alaska. Some miscellaneous croplands are
also not included in the Inventory due to limited understanding of greenhouse gas dynamics in management
systems (e.g., aquaculture). Consequently, there is a discrepancy between the total amount of managed area in
Land Converted to Cropland (see Section 6.1 Representation of the U.S. Land Base) and the cropland area included
in the Inventory. Improvements are underway to include croplands in Alaska and miscellaneous croplands in future
C inventories (see Table 6-35 in the Planned Improvements section for more details on the land area
discrepancies).
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).
The 2006 IPCC Guidelines recommend reporting changes in biomass, dead organic matter and soil organic C stocks
with land use change. All soil organic C stock changes are estimated and reported for Land Converted to Cropland,
but reporting of C stock changes for aboveground and belowground biomass, dead wood, and litter pools is limited
to Forest Land Converted to Cropland.46
46 Changes in biomass C stocks are not currently reported land use conversions to cropland except for Forest Land Converted to
Cropland, but this is a planned improvement for a future Inventory. Note: changes in dead organic matter are assumed to
negligible for other land use conversions to cropland, except Forest Land.
6-66 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Forest Land Converted to Cropland is the largest source of emissions from 1990 to 2020, accounting for
approximately 87 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 2020 led to CO2 emissions to the atmosphere of 54.4 MMT CO2 Eq. (14.8 MMT
C), including 28.2 MMT CO2 Eq. (7.7 MMT C) from aboveground biomass C losses, 5.5 MMT CO2 Eq. (1.5 MMT C)
from belowground biomass C losses, 5.5 MMT CO2 Eq. (1.5 MMT C) from dead wood C losses, 8.0 MMT CO2 Eq.
(2.2 MMT C) from litter C losses, 3.5 MMT C02 Eq. (0.9 MMT C) from mineral soils and 3.8 MMT C02 Eq. (1.0 MMT
C) from drainage and cultivation of organic soils. Emissions in 2020 are 5 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
2016
2017
2018
2019
2020
Grassland Converted to Cropland
6.9
7.5
8.5
8.7
8.5
8.4
8.8
Mineral Soils
4.1
4.0
5.2
5.4
5.1
5.1
5.5
Organic Soils
2.7
3.5
3.3
3.3
3.3
3.3
3.3
Forest Land Converted to Cropland
46.3
46.6
47.3
47.3
47.3
47.3
47.3
Aboveground Live Biomass
27.4
27.7
28.2
28.2
28.2
28.2
28.2
Belowground Live Biomass
5.3
5.4
5.5
5.5
5.5
5.5
5.5
Dead Wood
5.4
5.4
5.5
5.5
5.5
5.5
5.5
Litter
7.7
7.8
8.0
8.0
8.0
8.0
8.0
Mineral Soils
0.4
0.2
0.1
0.1
0.1
0.1
0.2
Organic Soils
0.1
0.1
+
+
+
+
+
Other Lands Converted to Cropland
(2.2)
(2.9)
(2.1)
(2.2)
(2.2)
(2.3)
(2.3)
Mineral Soils
(2.3)
(2.9)
(2.1)
(2.2)
(2.2)
(2.3)
(2.3)
Organic Soils
0.2
0.1
0.0
0.0
0.0
0.0
0.0
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
+
+
+
+
+
+
+
Wetlands Converted to Cropland
0.8
0.9
0.5
0.6
0.6
0.6
0.6
Mineral Soils
0.3
0.3
0.2
0.2
0.2
0.2
0.2
Organic Soils
0.6
0.6
0.3
0.3
0.4
0.4
0.4
Aboveground Live Biomass
27.4
27.7
28.2
28.2
28.2
28.2
28.2
Belowground Live Biomass
5.3
5.4
5.5
5.5
5.5
5.5
5.5
Dead Wood
5.4
5.4
5.5
5.5
5.5
5.5
5.5
Litter
7.7
7.8
8.0
8.0
8.0
8.0
8.0
Total Mineral Soil Flux
2.3
1.3
3.3
3.4
3.1
3.0
3.5
Total Organic Soil Flux
3.7
4.3
3.7
3.7
3.7
3.7
3.8
Total Net Flux
51.8
52.0
54.1
54.3
54.0
53.9
54.4
Does not exceed 0.05 MMT C02 Eq.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
fable 6-33: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes in
.and Converted to Cropland (MMT C)
1990
2005
2016
2017
2018
2019
2020
Grassland Converted to Cropland
1.9
2.0
2.3
2.4
2.3
2.3
2.4
Mineral Soils
1.1
1.1
1.4
1.5
1.4
1.4
1.5
Organic Soils
0.7
1.0
0.9
0.9
0.9
0.9
0.9
Forest Land Converted to Cropland
12.6
12.7
12.9
12.9
12.9
12.9
12.9
Aboveground Live Biomass
7.5
7.6
7.7
7.7
7.7
7.7
7.7
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
Land Use, Land-Use Change, and Forestry 6-67
-------
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
+
+
0.0
0.0
0.0
0.0
0.0
Settlements Converted to Cropland
+
+
+
+
+
+
+
Mineral Soils
+
+
+
+
+
+
+
Organic Soils
+
+
+
+
+
+
+
Wetlands Converted to Cropland
0.2
0.3
0.1
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
7.5
7.6
7.7
7.7
7.7
7.7
7.7
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
Total Mineral Soil Flux
0.6
0.4
0.9
0.9
0.8
0.8
0.9
Total Organic Soil Flux
1.0
1.2
1.0
1.0
1.0
1.0
1.0
Total Net Flux
14.1
14.2
14.8
14.8
14.7
14.7
14.8
+ Does not exceed 0.05 MMT C.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
Methodology and Time-Series Consistency
The following section includes a description of the methodology used to estimate C stock changes for Land
Converted to Cropland, including (1) loss of aboveground and belowground biomass, dead wood and litter C with
conversion of forest lands to croplands, as well as (2) the impact from all land use conversions to cropland on
mineral and soil organic C stocks.
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. 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 2020).
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
6-68 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 2020 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 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
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).
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 2020, 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 2020 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 for additional discussion of the Tier 3 methodology for
mineral soils.
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 2015).
48 See https://quickstats.nass.usda.gov/.
Land Use, Land-Use Change, and Forestry 6-69
-------
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 2020 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 2020 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 2020 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 2020 will be recalculated.
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.
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 2020) 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 will be recalculated
in future Inventories when new NRI data are available.
Uncertainty
The uncertainty analysis for biomass, dead wood and litter C losses with Forest Land Converted to Cropland is
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 (Also see Annex 3.12 for
further discussion). 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 2020, 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 95
percent below to 95 percent above the 2020 stock change estimate of 54.4 MMT CO2 Eq. The large relative
6-70 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
uncertainty in the 2020 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)
2020 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
8.8
(25.5)
43.1
-390%
390%
Mineral Soil C Stocks: Tier 3
1.1
(33.0)
35.2
-3006%
3,006%
Mineral Soil C Stocks: Tier 2
4.4
1.3
7.4
-70%
70%
Organic Soil C Stocks: Tier 2
3.3
0.8
5.8
-75%
75%
Forest Land Converted to Cropland
47.3
8.8
85.8
-81%
81%
Aboveground Live Biomass
28.2
(7.6)
64.0
-127%
127%
Belowground Live Biomass
5.5
(1.5)
12.5
-127%
127%
Dead Wood
5.5
(1.5)
12.5
-127%
127%
Litter
8.0
(2.2)
18.2
-127%
143%
Mineral Soil C Stocks: Tier 2
0.2
(0.1)
0.4
-134%
134%
Organic Soil C Stocks: Tier 2
+
(0.1)
0.1
-1852%
1852%
Other Lands Converted to Cropland
(2.3)
(3.7)
(0.8)
-64%
64%
Mineral Soil C Stocks: Tier 2
(2.3)
(3.7)
(0.8)
-64%
64%
Organic Soil C Stocks: Tier 2
+
+
+
+
+
Settlements Converted to Cropland
(0.1)
(0.3)
+
-117%
117%
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
0.6
+
1.3
-97%
97%
Mineral Soil C Stocks: Tier 2
0.2
+
0.5
-107%
107%
Organic Soil C Stocks: Tier 2
0.4
(0.2)
1.0
-142%
142%
Total: Land Converted to Cropland
54.4
2.8
106.0
-95%
95%
Aboveground Live Biomass
28.2
(7.6)
64.0
-127%
127%
Belowground Live Biomass
5.5
(1.5)
12.5
-127%
127%
Dead Wood
5.5
(1.5)
12.5
-127%
127%
Litter
8.0
(2.2)
18.2
-127%
127%
Mineral Soil C Stocks: Tier 3
1.1
(33.0)
35.2
-3006%
3,006%
Mineral Soil C Stocks: Tier 2
2.3
(1.1)
5.7
-147%
147%
Organic Soil C Stocks: Tier 2
3.8
1.2
6.3
-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.
Uncertainty is also associated with lack of reporting of agricultural biomass and dead organic matter C stock
changes. Biomass C stock changes are likely minor in perennial crops, such as orchards and nut plantations, given
the small amount of change in land that is used to produce these commodities in the United States. In contrast,
agroforestry practices, such as shelterbelts, riparian forests and intercropping with trees, may have led to larger
changes in biomass C stocks at least in some regions of the United States. However, there are currently no datasets
to evaluate the trends. Changes in dead organic matter C stocks are assumed to be negligible with conversion of
land to croplands with the exception of forest lands, which are included in this analysis. This assumption will be
further explored in a future Inventory.
/erification
See the QA/QC and Verification section in Cropland Remaining Cropland for information on QA/QC steps.
Land Use, Land-Use Change, and Forestry 6-71
-------
Recalculations Discussion
Recalculations are associated with new FIA data from 1990 to 2020 on biomass, dead wood and litter C stocks in
Forest Land Converted to Cropland, and updated estimates for mineral soils from 2016 to 2020 using the linear
extrapolation method. As a result, Land Converted to Cropland has an estimated smaller C loss of 0.13 MMT CO2
Eq. on average over the time series. This represents a 1 percent decrease in C stock changes for Land Converted to
Grassland compared to the previous Inventory.
Planned Improvements
Planned improvements are underway to include an analysis of C stock changes in Alaska for cropland, using the
Tier 2 method for mineral and organic soils that is described earlier in this section. This analysis will initially focus
on land use change, which typically has a larger impact on soil organic C stock changes than management
practices, but will be further refined over time to incorporate management data that drive C stock changes on
long-term cropland. See Table 6-35 for the amount of managed area in Land Converted to Cropland that is not
included in the Inventory, which is less than one thousand hectares per year. This includes the area in Alaska and
other miscellaneous cropland areas, such as aquaculture. Additional planned improvements are discussed in the
Planned Improvements section of Cropland Remaining Cropland.
Table 6-35: Area of Managed Land in Land Converted to Cropland that is not included in the
current Inventory (Thousand Hectares)
Area (Thousand Hectares)
Not Included in
Year
Managed Land
Inventory
Inventory
1990
12,308
12,308
<1
1991
12,654
12,654
<1
1992
12,943
12,943
<1
1993
14,218
14,218
<1
1994
15,400
15,400
<1
1995
15,581
15,581
<1
1996
15,888
15,888
<1
1997
16,073
16,073
<1
1998
17,440
17,440
<1
1999
17,819
17,819
<1
2000
17,693
17,693
<1
2001
17,600
17,600
<1
2002
17,487
17,487
<1
2003
16,257
16,257
<1
2004
15,317
15,317
<1
2005
15,424
15,424
<1
2006
15,410
15,410
<1
2007
14,923
14,923
<1
2008
14,399
14,399
<1
2009
13,814
13,814
<1
2010
13,905
13,905
<1
2011
14,186
14,186
<1
2012
14,429
14,429
<1
2013
13,752
13,752
<1
2014
13,050
13,050
<1
2015
13,049
13,049
<1
6-72 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
2016
2017
2018
2019
2020
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
ND
Note: NRI data are not available after 2015 so these years are designated as ND (No data).
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 grasslands),
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.49
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). The current Inventory includes all
grasslands in the conterminous United States and Hawaii, but does not include approximately 50 million hectares
of Grassland Remaining Grassland in Alaska. This leads to a discrepancy with the total amount of managed area in
Grassland Remaining Grassland (see Table 6-39 in Planned Improvements for more details on the land area
discrepancies) and the grassland area included in the Inventory analysis.
In Grassland Remaining Grassland, there has been considerable variation in C stocks between 1990 and 2020.
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 2020 led to net CO2 emissions to the atmosphere
of 4.5 MMT CO2 Eq. (1.2 MMT C), including 0.2 MMT CO2 Eq. (0.1 MMT C) from net losses of aboveground biomass
C, 0.1 MMT CO2 Eq. (<0.05 MMT C) from net losses in belowground biomass C, 2.3 MMT CO2 Eq. (0.6 MMT C) from
net losses in dead wood C, 0.2 MMT CO2 Eq. (0.1 MMT C) from net gains in litter C, 3.3 MMT CO2 Eq. (0.9 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 35 percent lower in 2020 compared to
1990, but as noted previously, stock changes are highly variable from 1990 to 2020, with an average annual change
of 7.2 MMT CO2 Eq. (2.0 MMT C).
49 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 6-73
-------
Table 6-36: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes
Grassland Remaining Grassland (MMT CO2 Eq.)
in
Soil Type
1990
2005
2016
2017
2018
2019
2020
Aboveground Live Biomass
0.1
0.1
0.2
0.2
0.2
0.2
0.2
Belowground Live Biomass
+
0.1
0.1
0.1
0.1
0.1
0.1
Dead Wood
2.8
2.7
2.4
2.4
2.4
2.3
2.3
Litter
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Mineral Soils
(2.2)
0.8
0.1
1.4
1.8
4.6
(3.3)
Organic Soils
6.3
5.2
5.4
5.4
5.4
5.4
5.4
Total Net Flux
6.9
8.7
8.0
9.3
9.7
12.4
4.5
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net
sequestration.
Table 6-37: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes in
Grassland Remaining Grassland (MMT C)
Soil Type
1990
2005
2016
2017
2018
2019
2020
Aboveground Live Biomass
+
+
0.1
0.1
0.1
0.1
0.1
Belowground Live Biomass
+
+
+
+
+
+
+
Dead Wood
0.8
0.7
0.7
0.7
0.6
0.6
0.6
Litter
+
+
(o.i)
(o.i)
(0.1)
(0.1)
(0.1)
Mineral Soils
(0.6)
0.2
+
0.4
0.5
1.2
(0.9)
Organic Soils
1.7
1.4
1.5
1.5
1.5
1.5
1.5
Total Net Flux
1.9
2.4
2.2
2.5
2.6
3.4
1.2
+ Does not exceed 0.05 MMT C
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net
sequestration.
The spatial variability in soil organic C stock changes for 201550 is displayed in Figure 6-8 for mineral soils and in
Figure 6-9 for organic soils. Although relatively small on a per-hectare basis, grassland soils gained C in isolated
areas that mostly occurred in pastures of the eastern United States. For organic soils, the regions with the highest
rates of emissions coincide with the largest concentrations of organic soils used for managed grassland, including
the Southeastern Coastal Region (particularly Florida), upper Midwest and Northeast, and a few isolated areas
along the Pacific Coast.
50 Only national-scale emissions are estimated for 2016 to 2020 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.
6-74 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure 6-8: Total Net Annual Soil C Stock Changes for Mineral Soils under Agricultural
Management within States, 2015, Grassland Remaining Grassland
Note: Only national-scale soil organic C stock changes are estimated for 2016 to 2020 in the current Inventory using a
surrogate data method, and therefore the fine-scale emission patterns in this map are based on inventory data from
2015. Negative values represent a net increase in soil organic C stocks, and positive values represent a net decrease in
soil organic C stocks.
Land Use, Land-Use Change, and Forestry 6-75
-------
Figure 6-9: Total Net Annual Soil C Stock Changes for Organic Soils under Agricultural
Management within States, 2015, Grassland Remaining Grassland
¦ > 40
Note: Only national-scale soil organic carbon stock changes are estimated for 2016 to 2020 in the current Inventory
using a surrogate data method, and therefore the fine-scale emission patterns in this map are based on inventory
data from 2015.
Methodology and Time-Series Consistency
The following section includes a description of the methodology used to estimate C stock changes for Grassland
Remaining Grassland, including (1) aboveground and belowground biomass, dead wood and litter C for woodlands,
as well as (2) soil organic C stocks for mineral and organic soils.
Biomass, Dead Wood and Litter Carbon Stock Changes
The methodology is consistent with IPCC (2006). Woodlands are lands that do not meet the definition of forest
land or agroforestry (see Section 6.1 Representation of the U.S. Land Base), but include woody vegetation with C
storage in aboveground and belowground biomass, dead wood and litter C (IPCC 2006) as described in the Forest
Land Remaining Forest Land section. Carbon stocks and net annual C stock change were determined according to
the stock-difference method for the CONUS, which involved applying C estimation factors to annual forest
inventories across time to obtain C stocks and then subtracting the values between years to estimate the stock
changes. The methods for estimating carbon stocks and stock changes for woodlands in Grassland Land Remaining
Grassland are consistent with those in the Forest Land Remaining Forest Land section and are described in Annex
3.13. All annual National Forest Inventory (NFI) plots available in the public FIA database (USDA Forest Service
2020) were used in the current Inventory. While the NFI is an all-lands inventory, only those plots that meet the
definition of forest land are typically measured. However, in some cases, particularly in the Central Plains and
Southwest United States, woodlands have been measured as part of the survey. This analysis is limited to those
plots and is not considered a comprehensive assessment of trees outside of forest land that meet the definition of
grassland. The same methods are applied from 1990 to 2020 in order to ensure time-series consistency.
6-76 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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.
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 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.51
A surrogate data method is used to estimate soil organic C stock changes from 2016 to 2020 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. Stock change estimates for 2016 to 2020 will be recalculated in future Inventories when the activity
data time series is updated.
Tier 3 Approach. Mineral soil organic C stocks and stock changes for Grassland Remaining Grassland are estimated
using the DayCent biogeochemical52 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., 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 and 95 percent confidence
51 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 2015).
52 Biogeochemical cycles are the flow of chemical elements and compounds between living organisms and the physical
environment.
Land Use, Land-Use Change, and Forestry 6-77
-------
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.
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 2020 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). Future Inventories will be updated with new
activity data, and the time series will be recalculated for 2016 to 2020 (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.
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 2020 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 2020 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 for further
discussion).
6-78 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Organic Soil Carbon Stock Changes
Annual C emissions from drained organic soils in Grassland Remaining Grassland are estimated using the Tier 2
method in IPCC (2006), which utilizes country-specific C loss rates (Ogle et al. 2003) rather than default IPCC rates.
For more information, see the Cropland Remaining Cropland section for organic soils and Annex 3.12.
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 2020) 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 will be recalculated
in future Inventories when new NRI data are available.
Uncertainty
The uncertainty analysis for biomass, dead wood and litter C losses with Forest Land Converted to Grassland is
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.
Uncertainty analysis 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 Grassland Remaining Grassland is
estimated using a Monte Carlo approach, which is also described in the Cropland Remaining Cropland section. For
2016 to 2020, there is additional uncertainty propagated through the Monte Carlo Analysis associated with the
surrogate data method.
Uncertainty estimates are presented in Table 6-38 for each subsource (i.e., 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 from
the Tier 2 and 3 approaches are combined using the simple error propagation methods provided by the IPCC
(2006), i.e., 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 Grassland Remaining Grassland ranges from more than 3,256
percent below and above the 2020 stock change estimate of 4.5 MMT CO2 Eq. The large relative uncertainty is
mostly due to large uncertainty in the Tier 3 method and variation in soil organic C stock changes that is not
explained by the surrogate data method, leading to high prediction error.
Table 6-38: Approach 2 Quantitative Uncertainty Estimates for C Stock Changes Occurring
Within Grassland Remaining Grassland (MMT CO2 Eq. and Percent)
2020 Flux Estimate Uncertainty Range Relative to Flux Estimate3
Source
(MMT C02 Eq.) (MMT CP2 Eq.) (%)
Lower Upper Lower Upper
Bound Bound Bound Bound
Woodland Biomass:
Aboveground live biomass 0.2 0.2 0.2 -31% 31%
Land Use, Land-Use Change, and Forestry 6-79
-------
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 45 (142.0) 150.9 -3,256% 3,256%
+ 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.
Uncertainty is also associated with a lack of reporting on biomass, dead wood and litter C stock changes for
agroforestry systems. Changes in biomass and dead organic matter C stocks are assumed to be negligible in other
grasslands, largely comprised of herbaceous biomass, although there are certainly significant changes at sub-
annual time scales across seasons.
/erification
See the QA/QC and Verification section in Cropland Remaining Cropland.
Recalculations Discussion
Recalculations are associated with updated estimates for mineral soils from 2016 to 2020 using the linear
extrapolation method, in addition to a correction in the estimation of biomass C. The correction is associated with
foliage estimates for woodlands that had been based on values for non-woodland foliage in the previous
Inventory. As a result of these new data, Grassland Remaining Grassland has a smaller loss of C compared to the
previous Inventory, estimated at a reduction in C loss of 1.3 MMT CO2 Eq., or 32 percent decrease in C loss, on
average over the time series for Grassland Remaining Grassland compared to the previous Inventory.
Planned Improvements
Grasslands in Alaska are not currently included in the Inventory. This is a significant planned improvement and
estimates are expected to be available in a future Inventory contingent on funding availability. Table 6-39 provides
information on the amount of managed area in Alaska that is Grassland Remaining Grassland, which includes
about 50 million hectares per year. For information about other improvements, see the Planned Improvements
section in Cropland Remaining Cropland.
0.1
0.1
-16%
16%
1.8
2.8
-22%
22%
(0.4)
+
-104%
104%
0.1
2.3
(0.2)
(2.3) (148.4) 143.9 -6,479% 6,479%
(0.9) (9.9) 8.1 -986% 986%
(0.2) (0.3) (0.1) -50% 50%
5.4 1.2 9.6 -77% 77%
6-80 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 6-39: Area of Managed Land in Grassland Remaining Grassland in Alaska that is not
included in the current Inventory (Thousand Hectares)
Area (Thousand Hectares)
Year
Managed Land
Inventory
Not Included in
Inventory
1990
327,446
277,406
50,040
1991
326,959
276,918
50,040
1992
326,462
276,422
50,040
1993
324,524
274,484
50,040
1994
322,853
272,813
50,040
1995
322,015
271,975
50,040
1996
321,164
271,123
50,040
1997
320,299
270,259
50,040
1998
318,214
268,174
50,040
1999
317,341
267,301
50,040
2000
316,242
266,202
50,040
2001
315,689
265,649
50,040
2002
315,232
265,192
50,040
2003
315,442
265,403
50,039
2004
315,459
265,421
50,038
2005
315,161
265,123
50,038
2006
314,841
264,804
50,037
2007
314,786
264,749
50,036
2008
314,915
264,878
50,037
2009
315,137
265,099
50,037
2010
314,976
264,942
50,035
2011
314,662
264,627
50,035
2012
314,466
264,413
50,053
2013
315,301
265,239
50,062
2014
316,242
266,180
50,062
2015
316,287
266,234
50,053
2016
ND
ND
ND
2017
ND
ND
ND
2018
ND
ND
ND
2019
ND
ND
ND
2020
ND
ND
ND
Note: NRI data are not available after 2015, and these years are designated as ND (No data).
Additionally, a review of available data on biosolids (i.e., treated sewage sludge) application will be undertaken to
improve the distribution of biosolids application on croplands, grasslands and settlements.
Non-C02 Emissions from Grassland Fires (CRF Source Category
4C1)
Fires are common in grasslands, and are thought to have been a key feature shaping the evolution of the grassland
vegetation in North America (Daubenmire 1968; Anderson 2004). Fires can occur naturally through lightning
strikes, but are also an important management practice to remove standing dead vegetation and improve forage
for grazing livestock. Woody and herbaceous biomass will be oxidized in a fire, although in this section the current
Land Use, Land-Use Change, and Forestry 6-81
-------
focus is primarily on herbaceous biomass.53 Biomass burning emits a variety of trace gases including non-CC>2
greenhouse gases such as CH4 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 nearly 300
percent since 1990. In 2020, CH4 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 2020 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
2016
2017
2018
2019
2020
ch4
0.1
0.3
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
2016
2017
2018
2019
2020
ch4
3
13
11
12
12
12
12
n2o
+
1
1
1
1
1
1
CO
84
358
324
345
331
341
334
NOx
5
21
19
21
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 2020, 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
2014.54 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).
53 A planned improvement is underway to incorporate woodland tree biomass into the Inventory.
54 See http://www.mtbs.gov.
6-82 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 6-42: Thousands of Grassland Hectares Burned Annually
Year
Thousand Hectares
1990
317
2005
1,343
2016
NE
2017
NE
2018
NE
2019
NE
2020
NE
Notes: Burned area was not estimated
(NE) for 2015 to 2020 but will be
updated in a future Inventory.
Burned area for the year 2014 is
estimated to be 1,659 thousand
hectares.
For 1990 to 2014, the total area of grassland burned is multiplied by the IPCC default factor for grassland biomass
(4.1 tonnes dry matter per ha) (IPCC 2006) to estimate the amount of combusted biomass. A combustion factor of
1 is assumed in this Inventory, and the resulting biomass estimate is multiplied by the IPCC default grassland
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
matter) and NOx (3.9 g NOx per kg dry matter) (IPCC 2006). The Tier 1 analysis is implemented in the Agriculture
and Land Use National Greenhouse Gas Inventory (ALU) software (Ogle et al. 2016).55
A linear extrapolation of the trend in the time series is applied to estimate emissions for 2015 to 2020 because
new activity data have not been compiled for these years. Specifically, a linear regression model with
autoregressive moving-average (ARMA) errors (Brockwell and Davis 2016) is used to estimate the trend in
emissions over time from 1990 to 2014, and the trend is used to approximate the 2015 to 2020 emissions. The Tier
1 method described previously will be applied to recalculate the 2015 to 2020 emissions in a future Inventory.
The same methods are applied from 1990 to 2014, and a data splicing method is used to extend the time series
from 2015 to 2020 ensuring a consistent time series of emissions data. The trend extrapolation is a standard data
splicing method for estimating emissions at the end of a time series if activity data are not available (IPCC 2006).
Uncertainty
Emissions are estimated using a linear regression model with ARMA errors for 2015 to 2020. The linear regression
ARMA model produced estimates of the upper and lower bounds of the emission estimate and the results are
summarized in Table 6-43. Methane emissions from Biomass Burning in Grassland for 2020 are estimated to be
between approximately 0.0 and 0.7 MMT CO2 Eq. at a 95 percent confidence level. This indicates a range of 100
percent below and 145 percent above the 2020 emission estimate of 0.3 MMT CO2 Eq. Nitrous oxide emissions are
estimated to be between approximately 0.0 and 0.8 MMT CO2 Eq., or approximately 100 percent below and 145
percent above the 2020 emission estimate of 0.3 MMT CO2 Eq.
55 See http://www.nrel.colostate.edu/proiects/ALUsoftware/.
Land Use, Land-Use Change, and Forestry 6-83
-------
Table 6-43: Uncertainty Estimates for Non-C02 Greenhouse Gas Emissions from Biomass
Burning in Grassland (MMT CO2 Eq. and Percent)
Source
Gas
2020 Emission Estimate
(MMT CO? Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMTCOzEq.) (%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Grassland Burning
Grassland Burning
ch4
n2o
0.3
0.3
+ 0.7
+ 0.8
-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.
Uncertainty is also associated with lack of reporting of emissions from biomass burning in grassland of Alaska.
Grassland burning emissions could be relatively large in this region of the United States, and therefore extending
this analysis to include Alaska is a planned improvement for the Inventory. There is also uncertainty due to lack of
reporting combustion of woody biomass, and this is another planned improvement.
/erification
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. Quality control identified problems with input data for common reporting format
tables in the spreadsheets, which have been corrected.
Recalculations Discussion
There are no recalculations in the time series from the previous Inventory.
Planned Improvements
A data splicing method is applied to estimate emissions in the latter part of the time series, which introduces
additional uncertainty in the emissions data. Therefore, a key improvement for the next Inventory will be to
update the time series with new activity data from the Monitoring Trends in Burn Severity program and recalculate
the emissions. Two other planned improvements have been identified for this source category, including a)
incorporation of country-specific grassland biomass factors, and b) extending the analysis to include Alaska. In the
current Inventory, biomass factors are based on a global default for grasslands that is provided by the IPCC (2006).
There is considerable variation in grassland biomass, however, which would affect the amount of fuel available for
combustion in a fire. Alaska has an extensive area of grassland and includes tundra vegetation, although some of
the areas are not managed. There has been an increase in fire frequency in boreal forest of the region (Chapin et
al. 2008), and this may have led to an increase in burning of neighboring grassland areas. There is also an effort
under development to incorporate grassland fires into DayCent model simulations. Both improvements are
expected to reduce uncertainty and produce more accurate estimates of non-C02 greenhouse gas emissions from
grassland burning.
6-84 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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).56 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. This Inventory includes all grasslands
in the conterminous United States and Hawaii, but does not include Land Converted to Grassland in Alaska.
Consequently, there is a discrepancy between the total amount of managed area for Land Converted to Grassland
(see Table 6-47 in Planned Improvements) and the grassland area included in the inventory analysis.
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, but these C stock changes are not estimated for
other land use conversions to grassland.57
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 2020 of 11.6, 2.1, and 4.6 MMT CO2 Eq. (3.2, 0.6, and 1.3 MMT C), respectively. Land
use and management of mineral soils in Land Converted to Grassland led to an increase in soil organic C stocks,
estimated at 43.9 MMT CO2 Eq. (12.0 MMT C) in 2020. The gains are primarily associated with conversion of Other
Land, which have relatively low soil organic C stocks, to Grassland that tend to have conditions suitable for storing
larger amounts of C in soils, and also due to conversion of Cropland to Grassland that leads to less intensive
management of the soil. Drainage of organic soils for grassland management led to CO2 emissions to the
atmosphere of 1.8 MMT CO2 Eq. (0.5 MMT C). The total net C stock change in 2020 for Land Converted to
Grassland is estimated as a gain of 24.1 MMT CO2 Eq. (6.6 MMT C), which represents an increase in C stock change
of 584 percent compared to the initial reporting year of 1990.
Table 6-44: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for
Land Converted to Grassland (MMT CO2 Eq.)
1990 2005 2016 2017 2018 2019 2020
Cropland Converted to
Grassland (18.3) (23.5) (17.8) (18.0) (18.0) (17.4) (19.7)
Mineral Soils (18.9) (25.0) (19.1) (19.4) (19.3) (18.7) (21.0)
56 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.
57 Changes in biomass C stocks are not currently reported for other conversions to grassland (other than forest land), but this is
a planned improvement for a future Inventory. Note: changes in dead organic matter are assumed to 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 6-85
-------
Organic Soils
0.6
1.5
1.4
1.4
1.3
1.3
1.3
Forest Land Converted to
Grassland
19.4
19.4
18.1
18.1
18.1
18.1
18.1
Aboveground Live Biomass
12.8
12.6
11.6
11.6
11.6
11.6
11.6
Belowground Live Biomass
2.3
2.2
2.1
2.1
2.1
2.1
2.1
Dead Wood
(0.4)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
Litter
4.7
4.8
4.6
4.6
4.6
4.6
4.6
Mineral Soils
(0.1)
(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
(4.2)
(31.7)
(22.2)
(22.1)
(21.9)
(21.5)
(21.8)
Mineral Soils
(4.2)
(31.7)
(22.3)
(22.2)
(21.9)
(21.6)
(21.9)
Organic Soils
+
+
0.1
0.1
0.1
0.1
0.1
Settlements Converted to
Grassland
(0.2)
(1.4)
(0.9)
(1.0)
(0.9)
(0.9)
(1.0)
Mineral Soils
(0.2)
(1.4)
(0.9)
(1.0)
(0.9)
(0.9)
(1.0)
Organic Soils
+
+
+
+
+
+
+
Wetlands Converted to
Grassland
0.1
0.2
0.3
0.3
0.3
0.3
0.2
Mineral Soils
+
+
+
+
+
+
+
Organic Soils
0.1
0.2
0.3
0.2
0.2
0.2
0.2
Aboveground Live Biomass
12.8
12.6
11.6
11.6
11.6
11.6
11.6
Belowground Live Biomass
2.3
2.2
2.1
2.1
2.1
2.1
2.1
Dead Wood
(0.4)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
Litter
4.7
4.8
4.6
4.6
4.6
4.6
4.6
Total Mineral Soil Flux
(23.4)
(58.2)
(42.4)
(42.5)
(42.2)
(41.3)
(43.9)
Total Organic Soil Flux
0.8
1.9
1.9
1.9
1.9
1.8
1.8
Total Net Flux
(3.1)
(37.0)
(22.6)
(22.7)
(22.4)
(21.5)
(24.1)
+ Does not exceed 0.05 MMT C02 Eq.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
Table 6-45: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for
Land Converted to Grassland (MMT C)
1990
2005
2016
2017
2018
2019
2020
Cropland Converted to Grassland
(5.0)
(6.4)
(4.8)
(4.9)
(4.9)
(4.7)
(5.4)
Mineral Soils
(5.2)
(6.8)
(5.2)
(5.3)
(5.3)
(5.1)
(5.7)
Organic Soils
0.2
0.4
0.4
0.4
0.4
0.4
0.4
Forest Land Converted to
Grassland
5.3
5.3
4.9
4.9
4.9
4.9
4.9
Aboveground Live Biomass
3.5
3.4
3.2
3.2
3.2
3.2
3.2
Belowground Live Biomass
0.6
0.6
0.6
0.6
0.6
0.6
0.6
Dead Wood
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(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
(1.1)
(8.6)
(6.1)
(6.0)
(6.0)
(5.9)
(5.9)
Mineral Soils
(1.2)
(8.6)
(6.1)
(6.1)
(6.0)
(5.9)
(6.0)
Organic Soils
+
+
+
+
+
+
+
Settlements Converted to
Grassland
+
(0.4)
(0.3)
(0.3)
(0.3)
(0.2)
(0.3)
Mineral Soils
+
(0.4)
(0.3)
(0.3)
(0.3)
(0.2)
(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
6-86 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Aboveground Live Biomass
3.5
3.4
3.2
3.2
3.2
3.2
3.2
Belowground Live Biomass
0.6
0.6
0.6
0.6
0.6
0.6
0.6
Dead Wood
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Litter
1.3
1.3
1.3
1.3
1.3
1.3
1.3
Total Mineral Soil Flux
(6.4)
(15.9)
(11.6)
(11.6)
(11.5)
(11.3)
(12.0)
Total Organic Soil Flux
0.2
0.5
0.5
0.5
0.5
0.5
0.5
Total Net Flux
(0.9)
(10.1)
(6.2)
(6.2)
(6.1)
(5.9)
(6.6)
+ Does not exceed 0.05 MMT C.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
Methodology and Time-Series Consistency
The following section includes a description of the methodology used to estimate C stock changes for Land
Converted to Grassland, including (1) loss of aboveground and belowground biomass, dead wood and litter C with
conversion of Forest Land Converted to Grassland, as well as (2) the impact from all land use conversions to
grassland on mineral and organic soil organic C stocks.
Biomass, Dead Wood, and Litter Carbon Stock Changes
A Tier 3 method is applied to estimate biomass, dead wood and litter C stock changes for Forest Land Converted to
Grassland. 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 2020).
There are limited data on the herbaceous grassland C stocks following conversion so default biomass estimates
(IPCC 2006) for grasslands are used to estimate C stock changes (Note: litter and dead wood C stocks are assumed
to be zero following conversion 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 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
Land Use, Land-Use Change, and Forestry 6-87
-------
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 2020 in order to ensure
time-series consistency. See Annex 3.13 for more information about reference C density estimates for forest land.
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 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 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 2020 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. Stock change estimates for 2016 to 2020 will be recalculated in future Inventories when
the times series of activity data is updated.
Tier 3 Approach. Mineral soil organic C stocks and stock changes are estimated using the DayCent
biogeochemical58 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
58 Biogeochemical cycles are the flow of chemical elements and compounds between living organisms and the physical
environment.
6-88 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 2020 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). Future Inventories will be updated with new
activity data, and the time series will be recalculated for 2016 to 2020 (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, time series of
activity data will be updated in a future Inventory, and emissions from 2016 to 2020 will be recalculated.
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). As with the Tier 3 method, time series of activity data
will be updated in a future Inventory, and emissions from 2016 to 2020 will be recalculated.
Uncertainty
The uncertainty analysis for biomass, dead wood and litter C losses with Forest Land Converted to Grassland is
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 2020, 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
Land Use, Land-Use Change, and Forestry 6-89
-------
(2006), as discussed in the previous paragraph. The combined uncertainty for total C stocks in Land Converted to
Grassland ranges from 153 percent below to 153 percent above the 2020 stock change estimate of 24.1 MMT CO2
Eq. The large relative uncertainty around the 2020 stock change estimate is partly due to large uncertainties in
biomass and dead organic matter C losses with Forest Land Conversion to Grassland. The large relative uncertainty
is also partly due to variation in soil organic C stock changes that is not explained by the surrogate data method,
leading to high prediction error with the splicing 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)
Source
2020 Flux Estimate3
Uncertainty Range Relative to Flux Estimate3
(MMT C02 Eq.)
(MMT C02
Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Cropland Converted to Grassland
(19.7)
(50.3)
10.9
-155%
155%
Mineral Soil C Stocks: Tier 3
(17.3)
(47.7)
13.2
-176%
176%
Mineral Soil C Stocks: Tier 2
(3.8)
(7.0)
(0.6)
-85%
85%
Organic Soil C Stocks: Tier 2
1.3
+
2.7
-103%
103%
Forest Land Converted to Grassland
18.1
4.9
31.3
-73%
73%
Aboveground Live Biomass
11.6
(0.5)
23.7
-104%
104%
Belowground Live Biomass
2.1
(0.1)
4.2
-103%
104%
Dead Wood
(0.3)
(0.1)
+
-78%
100%
Litter
4.6
(0.2)
9.4
-104%
104%
Mineral Soil C Stocks: Tier 2
+
(0.2)
0.1
-348%
348%
Organic Soil C Stocks: Tier 2
0.2
+
0.4
-114%
114%
Other Lands Converted to Grassland
(21.8)
(37.3)
(6.3)
-71%
71%
Mineral Soil C Stocks: Tier 2
(21.9)
(37.4)
(6.3)
-71%
71%
Organic Soil C Stocks: Tier 2
0.1
+
0.2
-155%
155%
Settlements Converted to Grassland
(1.0)
(1.7)
(0.2)
-76%
76%
Mineral Soil C Stocks: Tier 2
(1.0)
(1.7)
(0.2)
-75%
75%
Organic Soil C Stocks: Tier 2
+
+
+
-292%
292%
Wetlands Converted to Grasslands
0.2
+
0.5
-116%
116%
Mineral Soil C Stocks: Tier 2
+
(0.1)
0.1
-882%
882%
Organic Soil C Stocks: Tier 2
0.2
+
0.5
-116%
116%
Total: Land Converted to Grassland
(24.1)
(60.9)
12.7
-153%
153%
Aboveground Live Biomass
11.6
(0.5)
23.7
-104%
104%
Belowground Live Biomass
2.1
(0.1)
4.2
-103%
104%
Dead Wood
(0.3)
(0.1)
+
-78%
100%
Litter
4.6
(0.2)
9.4
-104%
104%
Mineral Soil C Stocks: Tier 3
(17.3)
(47.7)
13.2
-176%
176%
Mineral Soil C Stocks: Tier 2
(26.6)
(42.5)
(10.8)
-60%
60%
Organic Soil C Stocks: Tier 2
1.8
0.4
3.2
-77%
77%
+ 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.
Uncertainty is also associated with a lack of reporting on biomass, dead wood and litter C stock changes for
agroforestry systems. However, there are currently no datasets to evaluate the trends. Changes in biomass and
dead organic matter C stocks are assumed to be negligible with the exception of forest lands, which are included in
this analysis in other grasslands. This assumption will be further explored in a future Inventory.
/erification
See the QA/QC and Verification section in Cropland Remaining Cropland and Grassland Remaining Grassland for
information on QA/QC steps.
6-90 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Recalculations Discussion
Recalculations are associated with new FIA data from 1990 to 2020 on biomass, dead wood and litter C stocks in
Forest Land Converted to Grassland, and updated estimates for mineral soils from 2016 to 2020 using the linear
extrapolation method. As a result, Land Converted to Grassland has an estimated smaller gain in C of 2.93 MMT
CO2 Eq. on average over the time series. This represents a 15 percent decrease in C stock changes for Land
Converted to Grassland compared to the previous Inventory.
Planned Improvements
The amount of biomass C that is lost abruptly or the slower changes that continue to occur over a decade or longer
with Forest Land Converted to Grasslands will be further refined in a future Inventory. The current values are
estimated based on the amount of C before conversion and an estimated level of C left after conversion based on
limited plot data from the FIA and published literature for the Western United States and Great Plains Regions. The
amount of C left after conversion will be further investigated with additional data collection, particularly in the
Western United States and Great Plains, including tree biomass, understory biomass, dead wood and litter C pools.
In addition, biomass C stock changes will be estimated for Cropland Converted to Grassland, and other land use
conversions to grassland, to the extent that data are available.
An additional planned improvement for the Land Converted to Grassland category is to develop an inventory of C
stock changes for grasslands in Alaska. Table 6-47 provides information on the amount of managed area in Alaska
that is Land Converted to Grassland, which is as high as 54 thousand hectares in 2011.59 Note that areas of Land
Converted to Grassland in Alaska for 1990 to 2001 are classified as Grassland Remaining Grassland because land
use change data are not available until 2002. For information about other improvements, see the Planned
Improvements section in Cropland Remaining Cropland and Grassland Remaining Grassland.
Table 6-47: Area of Managed Land in Land Converted to Grassland's Alaska that is not
included in the current Inventory (Thousand Hectares)
Area (Thousand Hectares)
Not Included in
Year
Managed Land
Inventory
Inventory
1990
9,394
9,394
0
1991
9,485
9,485
0
1992
9,691
9,691
0
1993
11,566
11,566
0
1994
13,378
13,378
0
1995
13,994
13,994
0
1996
14,622
14,622
0
1997
15,162
15,162
0
1998
19,052
19,052
0
1999
19,931
19,931
0
2000
20,859
20,859
0
2001
21,968
21,968
0
2002
22,395
22,392
3
59 All of the Land Converted to Grassland according to the land representation is included in the inventory from 1990 through
2001 for the conterminous United States. However, there are no data to evaluate land use change in Alaska for this time
period, and so the balance of the managed area that may be converted to grassland in these years is included in Grassland
Remaining Grassland section. This gap in land use change data for Alaska will be addressed in a future Inventory.
Land Use, Land-Use Change, and Forestry 6-91
-------
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
22,015
22,557
22,460
22,718
22,450
22,685
22,608
22,664
22,805
22,643
21,472
20,195
20,242
ND
ND
ND
ND
ND
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
ND
ND
ND
ND
ND
7
10
13
16
21
24
26
29
54
47
33
33
33
ND
ND
ND
ND
ND
Note: NRI data are not available after 2015, and these years are
designated as ND (No data).
6.8 Wetlands Remaining Wetlands (CRF
Category 4D1)
Wetlands Remaining Wetlands includes all wetland in an Inventory year that had been classified as wetland for the
previous 20 years, and in this Inventory the flux estimates include Peatlands, Coastal Wetlands, and Flooded Land.
Managed peatlands are peatlands that have been cleared and drained for the production of peat. The production
cycle of a managed peatland has three phases: land conversion in preparation for peat extraction (e.g., clearing
surface biomass, draining), extraction (which results in the emissions reported under Peatlands Remaining
Peatlands), and abandonment, restoration, rewetting, or conversion of the land to another use.
Carbon dioxide emissions from the removal of biomass and the decay of drained peat constitute the major
greenhouse gas flux from managed peatlands. Managed peatlands may also emit Cm and N2O. The natural
production of Cm is largely reduced but not entirely shut down when peatlands are drained in preparation for
peat extraction (Strack et al. 2004 as cited in the 2006IPCC Guidelines). Drained land surface and ditch networks
contribute to the Cm flux in peatlands managed for peat extraction. Methane emissions were considered
insignificant under the IPCC Tier 1 methodology (IPCC 2006), but are included in the emissions estimates for
Peatlands Remaining Peatlands consistent with the 2013 Supplement to the 2006 IPCC Guidelines for National
Greenhouse Gas Inventories: Wetlands (IPCC 2013). Nitrous oxide emissions from managed peatlands depend on
site fertility. In addition, abandoned and restored peatlands continue to release greenhouse gas emissions.
Although methodologies are provided for rewetted organic soils (which includes rewetted/restored peatlands) in
IPCC (2013) guidelines, information on the areal extent of rewetted/restored peatlands in the United States is
6-92 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
Peatlands Remaining Peatlands
Emissions from Managed Peatlands
-------
currently unavailable. The current Inventory estimates CO2, CFUand N2O emissions from peatlands managed for
peat extraction in accordance with IPCC (2006 and 2013) guidelines.
CO2, 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 methodologies estimate only on-site N2O
and Cm emissions, since off-site N2O estimates are complicated by the risk of double-counting emissions from
nitrogen fertilizers added to horticultural peat, and off-site CH4 emissions are not relevant given the non-energy
uses of peat, so methodologies are not provided in IPCC (2013) guidelines.
On-site emissions from managed peatlands occur as the land is cleared of vegetation and the underlying peat is
exposed to sun and weather. 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 soils contain significant amounts of organic nitrogen in inactive form.
Draining land in preparation for peat extraction allows bacteria to convert the nitrogen into nitrates which leach to
the surface where they are reduced to N2O, and 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 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 were estimated to be 0.7 MMT CO2 Eq. in 2020 (see Table
6-48 and Table 6-49) comprising 0.7 MMT C02 Eq. (708 kt) of C02, 0.004 MMT C02 Eq. (0.15 kt) of CH4 and 0.0006
MMT CO2 Eq. (0.002 kt) of N2O. Total emissions in 2020 were 6.2 percent less than total emissions in 2019.
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 2020. 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. Nitrous oxide emissions showed a decreasing trend from 1990 until 1995, followed by an increasing
trend through 2001. Nitrous oxide emissions decreased between 2001 and 2006, followed by a leveling off
Land Use, Land-Use Change, and Forestry 6-93
-------
between 2008 and 2010, and a general decline between 2011 and 2020. Methane emissions decreased from 1990
until 1995, followed by an increasing trend through 2000, a period of fluctuation through 2010, and a general
decline between 2010 and 2020 (emissions rose slightly from 2016 to 2017 but resumed the downward trend
since).
Table 6-48: Emissions from Peatlands Remaining Peatlands (MMT CO2 Eq.)
Gas
1990
2005
2016
2017
2018
2019
2020
C02
1.1
1.1
0.7
0.8
0.8
0.8
0.7
Off-site
1.0
1.0
0.7
0.8
0.7
0.7
0.7
On-site
0.1
0.1
+
0.1
0.1
+
+
CH4 (On-site)
+
+
+
+
+
+
+
N20 (On-site)
+
+
+
+
+
+
+
Total
1.1
1.1
0.7
0.8
0.8
0.8
0.7
+ Does not exceed 0.05 MMT C02 Eq.
Note: These numbers are based on U.S. production data in accordance with Tier 1 guidelines, which
does not take into account imports, exports, and stockpiles (i.e., apparent consumption). Off-site N20
emissions are not estimated to avoid double-counting N20 emitted from the fertilizer that the peat is
mixed with prior to horticultural use (see IPCC 2006). Totals may not sum due to independent
rounding.
Table 6-49: Emissions from Peatlands Remaining Peatlands (kt)
Gas
1990
2005
2016
2017
2018
2019
2020
co2
1,055
1,101
733
829
792
755
708
Off-site
985
1,030
686
774
741
706
662
On-site
70
71
47
55
51
49
46
CH4 (On-site)
+
+
+
+
+
+
+
N20 (On-site)
+
+
+
+
+
+
+
+ Does not exceed 0.5 kt
Note: These numbers are based on U.S. production data in accordance with Tier 1 guidelines, which
does not take into account imports, exports, and stockpiles (i.e., apparent consumption). Off-site N20
emissions are not estimated to avoid double-counting N20 emitted from the fertilizer that the peat is
mixed with prior to horticultural use (see IPCC 2006). Totals 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 lower 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 2017; USGS 2021a; USGS 2021b). 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; and USGS estimates data
for non-respondents on the basis of prior-year production levels (Apodaca 2011).
The Alaska estimates 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
6-94 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 lower 48 states and
Alaska, large variations in peat production can also result from variations in precipitation and the subsequent
changes in moisture conditions, since unusually wet years can hamper peat production. The methodology
estimates Alaska emissions separately from lower 48 emissions because the state conducts its own mineral survey
and reports 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 but with volume-specific
C fraction conversion factors from IPCC (2006).60 Peat production was not reported for 2015 in Alaska's Mineral
Industry 2014 report (DGGS 2015); and reliable data are not available beyond 2012, so Alaska's peat production in
2013 through 2019 (reported in cubic yards) was assumed to be equal to the 2012 value.
Consistent with IPCC (2013) guidelines, off-site CO2 emissions from dissolved organic carbon were estimated based
on the total area of peatlands managed for peat extraction, which is calculated from production data using the
methodology described in the On-Site CO2 Emissions section below. Carbon dioxide emissions from dissolved
organic C were estimated by multiplying the area of peatlands by the default emission factor for dissolved organic
C provided in IPCC (2013).
The apparent consumption of peat, which includes production plus imports minus exports plus the decrease in
stockpiles, in the United States is over time the amount of domestic peat production. However, consistent with the
Tier 1 method whereby only domestic peat production is accounted for when estimating off-site emissions, off-site
CO2 emissions from the use of peat not produced within the United States are not included in the Inventory. The
United States has largely imported peat from Canada for horticultural purposes; in 2018, imports of Sphagnum
moss (nutrient-poor) peat from Canada represented 96 percent of total U.S. peat imports (USGS 2021a). Most peat
produced in the United States is reed-sedge peat, generally from southern states, which is classified as nutrient-
rich by IPCC (2006). Higher-tier calculations of CO2 emissions from apparent consumption would involve
consideration of the percentages of peat types stockpiled (nutrient-rich versus nutrient-poor) as well as the
percentages of peat types imported and exported.
Table 6-50: Peat Production of Lower 48 States (kt)
Type of Deposit
1990
2005
2016
2017
2018
2019
2020
Nutrient-Rich
595.1
657.6
388.1
423.3
415.0
410.4
417.1
Nutrient-Poor
55.4
27.4
52.9
74.7
62.0
45.6
12.9
Total Production
692.0
685.0
441.0
498.0
477.0
456.0
430.0
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).
Table 6-51: Peat Production of Alaska (Thousand Cubic Meters)
1990
2005
2016
2017
2018
2019
2020
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).
On-site CO2 Emissions
IPCC (2006) suggests basing the calculation of on-site emission estimates on the area of peatlands managed for
peat extraction differentiated by the nutrient type of the deposit (rich versus poor). Information on the area of
60 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).
Land Use, Land-Use Change, and Forestry 6-95
-------
land managed for peat extraction is currently not available for the United States, but consistent with IPCC (2006),
an average production rate for the industry was applied to derive an area estimate. In a mature industrialized peat
industry, such as exists in the United States and Canada, the vacuum method can extract up to 100 metric tons per
hectare per year (Cleary et al. 2005 as cited in IPCC 2006).61 The area of land managed for peat extraction in the
lower 48 states of the United States was estimated using nutrient-rich and nutrient-poor production data and the
assumption that 100 metric tons of peat are extracted from a single hectare in a single year, see Table 6-52. The
annual land area estimates were then multiplied by the IPCC (2013) default emission factor in order to calculate
on-site CO2 emission estimates.
Production data are not available by weight for Alaska. In order to calculate on-site emissions resulting from
Peatlands Remaining Peatlands in Alaska, the production data by volume were converted to weight using annual
average bulk peat density values, and then converted to land area estimates using the assumption that a single
hectare yields 100 metric tons, see Table 6-53. The IPCC (2006) on-site emissions equation also includes a term
that accounts for emissions resulting from the change in C stocks that occurs during the clearing of vegetation
prior to peat extraction. Area data on land undergoing conversion to peatlands for peat extraction is also
unavailable for the United States. However, USGS records show that the number of active operations in the United
States has been declining since 1990; therefore, it seems reasonable to assume that no new areas are being
cleared of vegetation for managed peat extraction. Other changes in C stocks in living biomass on managed
peatlands are also assumed to be zero under the Tier 1 methodology (IPCC 2006 and 2013).
Table 6-52: Peat Production Area of Lower 48 States (Hectares)
1990a
2005
2016
2017
2018
2019
2020
Nutrient-Rich
5,951
6,576
3,881
4,233
4,150
4,104
4,171
Nutrient-Poor
554
274
529
747
620
456
129
Total Production
6,920
6,850
4,410
4,980
4,770
4,560
4,300
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.
Notes: Calculated using peat production values in Table 6-50, an assumed yield of 100 metric tons per hectare
per year.
Table 6-53: Peat Production Area of Alaska (Hectares)
1990
2005
2016
2017
2018
2019
2020
Nutrient-Rich
0
0
0
0
0
0
0
Nutrient-Poor
286
104
201
333
212
212
212
Total Production
286
104
201
333
212
212
212
Sources: Calculated using peat production values in Table 6-51, an assumed yield of 100 metric tons
per hectare per year.
On-site N2O Emissions
IPCC (2006) suggests basing the calculation of on-site N2O emission estimates on the area of nutrient-rich
peatlands managed for peat extraction. These area data are not available directly for the United States, but the on-
site CO2 emissions methodology above details the calculation of area data from production data. In order to
estimate N2O emissions, the area of nutrient-rich Peatlands Remaining Peatlands was multiplied by the
appropriate default emission factor taken from IPCC (2013).
61 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).
6-96 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
On-site CHi Emissions
IPCC (2013) also suggests basing the calculation of on-site Cm emission estimates on the total area of peatlands
managed for peat extraction. Area data is derived using the calculation from production data described in the On-
site CO2 Emissions section above. In order to estimate CH4 emissions from drained land surface, the area of
Peatlands Remaining Peatlands was multiplied by the emission factor for direct CH4 emissions taken from IPCC
(2013). In order to estimate Cm emissions from drainage ditches, the total area of peatland was multiplied by the
default fraction of peatland area that contains drainage ditches, and the appropriate emission factor taken from
IPCC (2013). See Table 6-54 for the calculated area of ditches and drained land.
Table 6-54: Peat Production (Hectares)
1990
2005
2016
2017
2018
2019
2020
Lower 48 States
Area of Drained Land
6,574
6,508
4,190
4,731
4,532
4,332
4,085
Area of Ditches
346
343
221
249
239
228
215
Total Production
6,920
6,850
4,410
4,980
4,770
4,560
4,300
Alaska
Area of Drained Land
272
99
191
317
202
202
202
Area of Ditches
14
5
10
17
11
11
11
Total Production
286
104
201
333
212
212
212
Sources: Calculated using peat production values in Tables Table 6-50 and Table 6-51, an assumed yield of 100 metric tons
per hectare per year, and an assumed value of 5 percent ditch area.
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2020. 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 estimates 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 2020, 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 lower 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.
Land Use, Land-Use Change, and Forestry 6-97
-------
• 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 2020 were estimated to be between 0.6 and 0.8 MMT CO2 Eq. at
the 95 percent confidence level. This indicates a range of 17 percent below to 17 percent above the 2020 emission
estimate of 0.71 MMT CO2 Eq. Methane emissions from Peatlands Remaining Peatlands in 2020 were estimated to
be between 0.002 and 0.007 MMT CO2 Eq. This indicates a range of 58 percent below to 77 percent above the
2020 emission estimate of 0.004 MMT CO2 Eq. Nitrous oxide emissions from Peatlands Remaining Peatlands in
2020 were estimated to be between 0.0003 and 0.0009 MMT CO2 Eq. at the 95 percent confidence level. This
indicates a range of 53 percent below to 53 percent above the 2020 emission estimate of 0.0006 MMT CO2 Eq.
Table 6-55: Approach 2 Quantitative Uncertainty Estimates for CO2, Cm, and N2O Emissions
from Peatlands Remaining Peatlands (MMT CO2 Eq. and Percent)
2020 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
-17%
17%
Peatlands Remaining Peatlands
ch4
+
+
+
-58%
77%
Peatlands Remaining Peatlands
n2o
+
+
+
-53%
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.
QA/QC and Verification
A QA/QC analysis was performed to review input data and calculations, and no issues were identified. In addition,
the emission trends were analyzed to ensure they reflected activity data trends.
Recalculations Discussion
The lower 48 states peat production estimates for Peatlands Remaining Peatlands were updated using the Peat
section of the Mineral Commodity Summaries 2021. The 2021 edition updated 2018 and 2019 peat production
data and provided peat type production estimates for 2020. The updated data lowered previously estimated
emissions for 2018 and 2019 by 0.4 percent and 2.9 percent versus estimated emissions for 2018 and 2019 in the
previous (i.e., 1990 through 2019) Inventory for Peatlands Remaining Peatlands.
Although Alaska peat production data for 2015 through 2020 were unavailable, 2014 data are available in the
Alaska's Mineral Industry 2014 report. However, the reported values represented an apparent 98 percent
decrease in production since 2012. Due to the uncertainty of the most recent data, 2013, 2014, 2015, 2016, 2017,
2018, 2019, and 2020 values were assumed to be equal to the 2012 value, seen in the Alaska's Mineral Industry
2013 report. If updated Alaska data are available for the next Inventory cycle, this will result in a recalculation in
the next (i.e., 1990 through 2021) Inventory report.
Planned Improvements
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 undergoing peat extraction.
During the next Inventory cycle, efforts are planned 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
6-98 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
future publication of data has been discontinued. In addition, edits to the trends and methodology sections are
planned based on expert review comments.
The USGS has not published a Minerals Yearbook since the 2018 Advance Release. If more recent versions of the
Minerals Yearbook are made available, these updated data will be included in the next inventory cycle.
Correspondence with USGS indicated the Minerals Yearbook publications undergo a lengthy editing process and
may delay the release of the publications (Brioche 2021).
The implied emission factors will be calculated and included in this chapter for future Inventories. Currently, the
N2O emissions calculation uses different land areas than the CO2 and Cm emission calculations, so estimating the
implied emission factor per total land area is not appropriate and are not generated in the CRF tables. The
inclusion of implied emission factors in this chapter will provide another method of QA/QC and verification.
Coastal Wetlands Remaining Coastal Wetlands
This Inventory recognizes Wetlands as a "land-use that includes land covered or saturated for all or part of the
year, in addition to areas of lakes, reservoirs, and rivers." Consistent with ecological definitions of wetlands,62 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 and the gain-
loss method for biomass and DOM. Presently, this Inventory does not calculate the lateral flux of carbon to or from
any land use. Lateral transfer of organic carbon to coastal wetlands and to marine sediments within U.S. waters is
the subject of ongoing scientific investigation.
The United States recognizes both Vegetated Wetlands and Unvegetated Open Water as Coastal Wetlands. Per
guidance provided by the Wetlands Supplement, sequestration of carbon into biomass, DOM and soil carbon 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 stock losses occur when Vegetated
Coastal Wetlands are converted to Unvegetated Open Water Coastal Wetlands.
This Inventory includes all privately-owned and publicly-owned coastal wetlands (i.e., mangroves and tidal marsh)
along the oceanic shores on the conterminous United States, but does not include Coastal Wetlands Remaining
Coastal Wetlands in Alaska or Hawaii. Seagrasses are not currently included within the Inventory due to insufficient
data on distribution, change through time and carbon (C) stocks or C stock changes as a result of anthropogenic
influence.
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,
2) Carbon stock changes on Vegetated Coastal Wetlands Converted to Unvegetated Open Water Coastal
Wetlands,
62 See https://water.usBS.gov/nwsum/WSP2425/definitions.html; accessed August 2021.
Land Use, Land-Use Change, and Forestry 6-99
-------
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, belowground, 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 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 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. 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.
Mangrove forests that are less than 5 m are reported under Coastal Wetlands. All other non-drained, intact coastal
marshes are intended to be 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 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)63 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 -3.7 to -4.8 MMT
CO2 Eq. across the majority of the time series, however, between 2006 and 2010 they were a net source of
emissions (ranging from of 5.2 to 5.5 MMT CO2 Eq.), resulting from large loss of vegetated coastal wetlands to
open water due to hurricanes (Table 6-56). Recognizing removals of C02to soil of 10.2 MMT CO2 Eq. and CFU
emissions of 3.8 MMT CO2 Eq. in 2020, Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands are a
net sink of 6.4 MMT CO2 Eq. Loss of coastal wetlands, primarily in the Mississippi Delta as a result of hurricane
impacts and sediment diversion and other human impacts, recognized as Vegetated Coastal Wetlands Converted
to Unvegetated Coastal Wetlands, drive an emission of 1.5 MMT CO2 Eq. over the past five years, primarily from
63 See https://coast.noaa.gov/digitalcoast/tools/lca.html; accessed August 2021.
6-100 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
soils. Building of new wetlands from open water, recognized as Unvegetated Coastal Wetlands Converted to
Vegetated Coastal, results each year in removal of 0.03 MMT CO2 Eq. Aquaculture is a minor industry in the United
States, resulting in an emission of N2O across the time series of between 0.1 to 0.2 MMT CO2 Eq. In all, Coastal
Wetlands are a net sink of 4.8 MMT CO2 Eq. in 2020.
Table 6-56: Emissions and Removals from Coastal Wetlands Remaining Coastal Wetlands
(MMT COz Eq.)
Land Use/Carbon Pool
1990
2005
2016
2017
2018
2019
2020
Vegetated Coastal Wetlands
Remaining Vegetated Coastal
Wetlands
(6.5)
(6.4)
(6.4)
(6.4)
(6.4)
(6.4)
(6.4)
Biomass C Flux
(+)
0.1
(+)
(+)
(+)
(+)
(+)
Soil C Flux
(10.2)
(10.2)
(10.2)
(10.2)
(10.2)
(10.2)
(10.2)
Net CH4 Flux
3.7
3.8
3.8
3.8
3.8
3.8
3.8
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
(+)
(+)
0
0
0
0
0
Soil C Flux
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Net N20 Flux from Aquaculture in
Coastal Wetlands
0.1
0.2
0.1
0.1
0.2
0.2
0.2
Total Biomass C Flux
+
0.1
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Total Dead Organic Matter C Flux
(+)
(+)
+
+
+
+
+
Total Soil C Flux
(8.5)
(7.7)
(8.7)
(8.7)
(8.7)
(8.7)
(8.7)
Total CH4 Flux
3.7
3.8
3.8
3.8
3.8
3.8
3.8
Total N20 Flux
0.1
0.2
0.1
0.1
0.2
0.2
0.2
Total Flux
(4.6)
(3.7)
(4.8)
(4.8)
(4.8)
(4.8)
(4.8)
+ Absolute value does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Emissions and Removals from Vegetated Coastal Wetlands
Remaining Vegetated Coastal Wetlands
The conterminous United States currently has 2.98 million hectares of intertidal Vegetated Coastal Wetlands
Remaining Vegetated Coastal Wetlands comprised of tidally influenced palustrine emergent marsh (660,448 ha),
palustrine scrub shrub (133,148 ha) and estuarine emergent marsh (1,894,045 ha), estuarine scrub shrub (94,110
ha) and estuarine forested wetlands (195,619 ha). Mangroves fall under both estuarine forest and estuarine scrub
shrub categories depending upon height. Dwarf mangroves, found in subtropical states along the Gulf of Mexico,
do not attain the height status to be recognized as Forest Land, and are therefore always classified within
Vegetated Coastal Wetlands. Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands are found in
cold temperate (53,973 ha), warm temperate (896,253 ha), subtropical (1,964,383 ha) and Mediterranean (62,762
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 biomass C stocks along with soils. Methane emissions from
Land Use, Land-Use Change, and Forestry 6-101
-------
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 below summarize nationally aggregated biomass and soil C stock changes and Cm
emissions on Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands. Intact Vegetated Coastal
Wetlands Remaining Vegetated Coastal Wetlands hold a total biomass C stock of 88.8 MMT C. Removals from
biomass C stocks in 2020 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 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 2020, a value that has remained relatively constant
across the reporting period. For Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands, methane
emissions of 3.8 of MMT CO2 Eq. (154 kt CH4) in 2020 (Table 6-59) offset C removals resulting in a net removal of
6.4 MMT CO2 Eq. in 2020; 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 Cm 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
2016
2017
2018
2019
2020
Biomass Flux
(+)
0.1
(+)
(+)
(+)
(+)
(+)
Soil Flux
(10.2)
(10.2)
(10.2)
(10.2)
(10.2)
(10.2)
(10.2)
Total C Stock Change
(10.2)
(10.2)
i (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
2016
2017
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
2016
2017
2018
2019
2020
Methane Emissions (MMT C02 Eq.)
3.7
3.8
3.8
3.8
3.8
3.8
3.8
Methane Emissions (kt CH4)
149
151
153
153
153
153
154
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 Cm 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 2020.
6-102 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Biomass Carbon Stock Changes
Above- and below ground 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 2020 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 content but is differentiated based on climate zone. Aboveground biomass carbon 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 carbon
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 (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 carbon stock. Above- and belowground values were summed to obtain total biomass carbon 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 2016 2017 2018 2019 2020
Vegetated Coastal Wetlands
Remaining Vegetated Coastal
Wetlands
Vegetated Coastal Wetlands
Converted to Unvegetated Open
Water Coastal Wetlands
Unvegetated Open Water Coastal
Wetlands Converted to
Vegetated Coastal Wetlands
2,985,512
1,720
2,988,258 2,972,368 2,972,634 2,974,900 2,976,166 2,977,432
2,515 1,488 1,488 1,488 1,488 1,488
953 1,775 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
3.17
2.24
4.69
Palustrine Emergent Wetland
3.25
3.17
2.24
4.69
Estuarine Forested Wetland
3.05
3.10
17.83
3.44
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).
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
Land Use, Land-Use Change, and Forestry 6-103
-------
Palustrine Emergent Wetland
Estuarine Forested Wetland
Estuarine Scrub/Shrub Wetland
Estuarine Emergent Wetland
Source: All values from IPCC (2014).
Soil Carbon Stock Changes
Soil C stock changes are estimated for Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands for
both mineral and organic soils. Soil C stock changes, stratified by climate zones and wetland classes, are derived
from a synthesis of peer-reviewed literature (Table 6-63; 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. 2012a&b; Bianchi et al. 2013; Crooks et al. 2014; Weston et al. 2014; Villa & Mitsch
2015; Marchio et al. 2016; Noe et al. 2016).
Tier 2 level estimates of soil C removals associated with annual soil C accumulation on managed Vegetated Coastal
Wetlands Remaining Vegetated Coastal Wetlands were developed with country-specific soil C removal factors
multiplied by activity data of land area for Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands.
The methodology follows Eq. 4.7, Chapter 4 of the Wetlands Supplement, and is applied to the area of Vegetated
Coastal Wetlands Remaining Vegetated Coastal Wetlands on an annual basis. To estimate soil C stock changes, no
differentiation is made between organic and mineral soils since currently no statistical evidence supports
disaggregation (Holmquist et al. 2018).
Table 6-63: Annual Soil Carbon Accumulation Rates for Vegetated Coastal Wetlands (t C ha1
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
1.01
0.82
0.87
0.85
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)64
1.15
1.15
2.11
2.11
1.15
1.15
2.11
2.11
3.65
0.96
3.65
3.65
3.63
3.63
3.63
3.63
Soil 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 Cm 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
64 See https://github.com/Smithsonian/Coastal-Wetland-NGGI-Data-Public; accessed August 2021.
6-104 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 Cm 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, where the largest uncertainty for any one soil C stock referenced using
published literature values for a community class; uncertainty approaches provide that if multiple values are
available for a single parameter, the highest uncertainty value should be applied to the propagation of errors; IPCC
2000). Uncertainty for root to shoot ratios, which are used for quantifying belowground biomass, are derived from
the 2013 Wetlands Supplement. Uncertainties for Cm 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 Cm 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 Cm) and taking the square root of
that total.
Uncertainty estimates are presented in Table 6-64 for each subsource (i.e., soil C, biomass C and Cm emissions).
The combined uncertainty across all subsources is +/-36.6 percent, which is primarily driven by the uncertainty in
the Cm estimates because there is high variability in CH4 emissions when the salinity is less than 18 ppt. In 2020,
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
Wetiandsin 2020 (MMT CO2 Eq. and Percent)
Source
Gas
2020 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)
-17.8%
17.8%
CH4 emissions
ch4
3.8
2.7
5.0
-29.8%
29.8%
Total Flux
(6.4)
(8.7)
(4.0)
-36.6%
36.6%
Note: Parentheses indicate net sequestration. Totals may not sum due to independent rounding.
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 QA/QC assessment. Acceptance of final datasets into archive and
dissemination are contingent upon the product compilation being compliant with mandatory QA/QC requirements
(McCombs et al. 2016). QA/QC and verification of soil C stock datasets have been provided by the Smithsonian
Environmental Research Center and Coastal Wetland Inventory team leads who reviewed summary tables against
reviewed sources. 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 this Inventory. A team of two evaluated and verified there were no computational
errors within the calculation worksheets. Soil and biomass C stock change data are based upon peer-reviewed
literature and CH4 emission factors derived from the Wetlands Supplement.
Recalculations Discussion
No recalculations were needed for the current Inventory.
Land Use, Land-Use Change, and Forestry 6-105
-------
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 stock and biomass estimates for
coastal wetlands.65 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 updated changes in soil C stocks for estuarine emergent wetlands.
Work is currently underway to examine the feasibility of incorporating seagrass soil and biomass C stocks into the
Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands estimates. Additionally, investigation into
quantifying the distribution, area, and emissions resulting from impounded waters (i.e., coastal wetlands where
tidal connection to the ocean has been restricted or eliminated completely) is underway.
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 2020, 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 aquafer 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 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 2020, there were 1,488 ha of Vegetated Coastal Wetlands Converted to Unvegetated Open Water Coastal
Wetlands across all wetland types and climates, which resulted in 1.5 MMT CO2 Eq. (0.4 MMT C) and 0.06 MMT
65 See https://serc.si.edu/coastalcarbon; accessed August 2021.
6-106 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
C02 Eq. (0.02 MMT C) lost through soil and biomass, respectively, while DOM C stock loss was present it was
minimal (Table 6-60, 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
2016
2017
2018
2019
2020
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.
Table 6-66: Net CO2 Flux from C Stock Changes in Vegetated Coastal Wetlands Converted to
Unvegetated Open Water Coastal Wetlands (MMT C)
Year
1990
2005
2016
2017
2018
2019
2020
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.
Methodology and Time-Series Consistency
The following section includes a brief description of the methodology used to estimate changes in soil, biomass
and DOM C stocks for Vegetated Coastal Wetlands Converted to Unvegetated Open Water Coastal Wetlands.
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2020.
Biomass Carbon Stock Changes
Biomass C stock changes for palustrine and estuarine marshes are estimated for Vegetated Coastal Wetlands
Converted to Unvegetated Open Water Coastal Wetlands on lands below the elevation of high tides (taken to be
mean high water spring tide elevation) within the U.S. Land Representation 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. C-CAP areas are calculated at the state/territory level and summed according to
climate zone to national values. Publicly-owned and privately-owned lands are represented. Trends in land cover
change are extrapolated to 1990 and 2020 from these datasets. The C-CAP database provides peer reviewed
country-specific mapping to support IPCC Approach 3 quantification of coastal wetland distribution, including
conversion to and from open water. Biomass C stocks are not sensitive to soil organic content but are
differentiated based on climate zone. Non-forested aboveground biomass C stock 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 carbon 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 (201766; Table 6-61). Aboveground 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
66 See https://github.com/Smithsonian/Coastal-Wetland-NGGI-Data-Public; accessed August 2021.
Land Use, Land-Use Change, and Forestry 6-107
-------
disaggregation of a community class. Root to shoot ratios from the Wetlands Supplement were used to account for
belowground biomass, which were multiplied by the aboveground carbon stock (Table 6-62; IPCC 2014). Above-
and belowground values were summed to obtain total biomass carbon stocks. Conversion to open water results in
emissions of all biomass 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 in each climate zone by its mean biomass.
Dead Organic Matter
Dead organic matter (DOM) carbon stocks, which include litter and dead wood stocks for subtropical estuarine
forested wetlands, are an emission from Vegetated Coastal Wetlands Converted to Unvegetated Open Water
Coastal Wetlands across all years in the time series. Data on DOM carbon stocks are not currently available for
either palustrine or estuarine scrub/shrub wetlands for any climate zone. Data for estuarine forested wetlands in
other climate zones are not included since there is no estimated loss of these forests to unvegetated open water
coastal wetlands across any year based on C-CAP data. For subtropical estuarine forested wetlands, Tier 1
estimates of 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 2020 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 CFU 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 carbon 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, where the largest uncertainty for any one soil C stock referenced using published literature
values for a community class; if multiple values are available for a single parameter, the highest uncertainty value
should be applied to the propagation of errors; 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.
6-108 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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.
Details on the emission/removal trends and methodologies through time are described in more detail in the
Introduction and Methodology sections.
Uncertainty estimates are presented in Table 6-67 for each subsource (i.e., soil C, biomass C, and DOM emissions).
The combined uncertainty across all subsources is +/- 32.0 percent, which is driven by the uncertainty in the soil C
estimates. In 2020, 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
2020 Flux Estimate
Uncertainty Range Relative to Flux Estimate
(MMT CO? Eq.)
(MMT CO
zEq.)
(%)
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%
Note: Totals may not sum due to independent rounding.
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
No recalculations were needed for the current Inventory.
Planned Improvements
The depth of soil carbon 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. 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.
Land Use, Land-Use Change, and Forestry 6-109
-------
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
mapping could provide a more refined regional Approach 2-3 land representation to support the national-scale
assessment provided by C-CAP.
An approach for calculating the fraction of remobilized coastal wetland soil C returned to the atmosphere as CO2 is
currently under review and may be included in future reports.
Research by USGS is investigating higher resolution mapping approaches to quantify conversion of coastal
wetlands is also underway. Such approaches may form the basis for a full Approach 3 land representation
assessment in future years. C-CAP data harmonization with the National Land Cover Dataset (NLCD) will be
incorporated into a future iteration of the Inventory.
Stock Changes from Unvegetated Open Water Coastal
Wetlands Converted to Vegetated Coastal Wetlands
Open water within the U.S. land base, as described in Section 6.1 Representation of the U.S. Land Base, is
recognized as Coastal Wetlands within this Inventory. The appearance of vegetated tidal wetlands on lands
previously recognized as open water reflects either the building of new vegetated marsh through sediment
accumulation or the transition from other lands uses through an intermediary open water stage as flooding
intolerant plants are displaced and then replaced by wetland plants. Biomass, DOM and soil C accumulation on
Unvegetated Open Water Coastal Wetlands Converted to Vegetated Coastal Wetlands begins with vegetation
establishment.
Within the United States, conversion of Unvegetated Open Water Coastal Wetlands to Vegetated Coastal
Wetlands is predominantly due to engineered activities, which include active restoration of wetlands (e.g.,
wetlands restoration in San Francisco Bay), dam removals or other means to reconnect sediment supply to the
nearshore (e.g., Atchafalaya Delta, Louisiana, Couvillion et al. 2011). Wetlands restoration projects have been
ongoing in the United States since the 1970s. Early projects were small, a few hectares in size. By the 1990s,
restoration projects, each hundreds of hectares in size, were becoming common in major estuaries. In several
coastal areas e.g., San Francisco Bay, Puget Sound, Mississippi Delta and south Florida, restoration activities are in
planning and implementation phases, each with the goal of recovering tens of thousands of hectares of wetlands.
In 2020, 2,406 ha of unvegetated open water coastal wetlands were converted to vegetated coastal wetlands
across all wetland types and climates, which has steadily increased over the reporting period (Table 6-59). This
resulted in 0.007 MMT CO2 Eq. (0.002 MMT C) and 0.1 MMT CO2 Eq. (0.03 MMT C) sequestered in soil and
biomass, respectively (Table 6-68 and Table 6-69). The soil C stock has increased during the Inventory's reporting
period, likely due to increasing vegetated coastal wetland restoration over time. While DOM C stock increases are
present they are minimal in the early part of the time series and zero in the later because there are no conversions
from unvegetated open water coastal wetlands to subtropical estuarine forested wetlands between 2011 and
2016 (and by proxy through 2020), and that is the only coastal wetland type where DOM data is currently
available.
Throughout the reporting period, the amount of Open Water Coastal Wetlands Converted to Vegetated Coastal
Wetlands has increased over time, reflecting the increase in engineered restoration activities mentioned above.
Table 6-68: CO2 Flux from C Stock Changes from Unvegetated Open Water Coastal Wetlands
Converted to Vegetated Coastal Wetlands (MMT CO2 Eq.)
Year
1990
2005
2016
2017
2018
2019
2020
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
(+)
(+)
(+)
(+)
(+)
(+)
(+)
6-110 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Total C Stock Change (+) (0.1) (0.1) (0.1) (0.1) (0.1) (0.1)
+ Absolute value does not exceed 0.05 MMT C02 Eq.
Note: Parentheses indicate net sequestration. Totals may not sum due to independent rounding.
Table 6-69: CO2 Flux from C Stock Changes from Unvegetated Open Water Coastal Wetlands
Year
1990
2005
2016
2017
2018
2019
2020
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.
Note: Parentheses indicate net sequestration. Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
The following section includes a brief description of the methodology used to estimate changes in soil, biomass
and DOM C stocks, and Cm emissions for Unvegetated Open Water Coastal Wetlands Converted to Vegetated
Coastal Wetlands.
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2020.
Biomass Carbon Stock Changes
Quantification of regional coastal wetland biomass C stock changes for palustrine and estuarine marsh vegetation
are presented for Unvegetated Open Water Coastal Wetlands Converted to Vegetated Coastal Wetlands on lands
below the elevation of high tides (taken to be mean high water spring tide elevation) according to the national
LiDAR dataset, the national network of tide gauges and land use histories recorded in the 1996, 2001, 2005, 2011,
and 2016 NOAA C-CAP surveys. C-CAP areas are calculated at the state/territory level and summed according to
climate zone to national values. Privately-owned and publicly-owned lands are represented. Trends in land cover
change are extrapolated to 1990 and 2020 from these datasets (Table 6-58). C-CAP provides peer reviewed
country-level mapping of coastal wetland distribution, including conversion to and from open water. Biomass C
stock is not sensitive to soil organic content but differentiated based on climate zone. Data for non-forested
wetlands are derived from a national assessment combining field plot data and aboveground biomass mapping by
remote sensing (Table 6-61; Byrd et al. 2017; Byrd et al. 2018; Byrd et al. 2020). The aboveground biomass carbon
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 (201767). Aboveground 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. Root to shoot ratios from the
Wetlands Supplement were used to account for belowground biomass, which were multiplied by the aboveground
carbon stock (Table 6-62; IPCC 2014). Above- and belowground values were summed to obtain total biomass
carbon stocks.
Conversion of open water to Vegetated Coastal Wetlands results in the establishment of a standing biomass C
stock; therefore, stock changes that occur are calculated by multiplying the C-CAP derived area gained that year in
each climate zone by its mean biomass. While the process of revegetation of unvegetated open water wetlands
67 See https://github.com/Smithsonian/Coastal-Wetland-NGGI-Data-Public; accessed August 2021.
Land Use, Land-Use Change, and Forestry 6-111
-------
can take many years to occur, it is assumed in the calculations that the total biomass is reached in the year of
conversion.
Dead Organic Matter
Dead organic matter (DOM) carbon stocks, which include litter and dead wood stocks, are included for subtropical
estuarine forested wetlands for Vegetated Coastal Wetlands Converted to Unvegetated Open Water Coastal
Wetlands across all years. Tier 1 default or country-specific data on DOM are not currently available for either
palustrine or estuarine scrub/shrub wetlands for any climate zone. Data for estuarine forested wetlands in other
climate zones are not included since there is no estimated loss of these forests to unvegetated open water coastal
wetlands across any year based on C-CAP data. Tier 1 estimates of subtropical estuarine forested wetland 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 2020 time series. Dead organic matter removals are calculated by multiplying the C-
CAP derived area gained that year by its Tier 1 DOM C stock. Similar to biomass C stock gains, gains in DOM can
take many years to occur, but for this analysis, the total DOM stock is assumed to accumulate during the first year
of conversion.
Soil Carbon Stock Change
Soil C stock changes are estimated for Unvegetated Open Water Coastal Wetlands Converted to Vegetated Coastal
Wetlands. Country-specific soil C removal factors associated with soil C accretion, stratified by climate zones and
wetland classes, are derived from a synthesis of peer-reviewed literature and updated this year based upon
refined review of the dataset (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). Soil C stock changes are stratified based upon wetland class (Estuarine, Palustrine) and subclass
(Emergent Marsh, Scrub Shrub). For soil C stock change no differentiation is made for soil type (i.e., mineral,
organic). Soil C removal factors were developed from literature references that provided soil C removal factors
disaggregated by climate region and vegetation type by salinity range (estuarine or palustrine) as identified using
NOAA C-CAP as described above (see Table 6-63 for values).
Tier 2 level estimates of C stock changes associated with annual soil C accumulation in Vegetated Coastal Wetlands
were developed using country-specific soil C removal factors multiplied by activity data on Unvegetated Coastal
Wetlands converted to Vegetated Coastal Wetlands. The methodology follows Eq. 4.7, Chapter 4 of the Wetlands
Supplement, and is applied to the area of Unvegetated Coastal Wetlands converted to Vegetated Coastal Wetlands
on an annual basis.
Soil Methane Emissions
A Tier 1 assumption has been applied that salinity conditions are unchanged and hence Cm emissions are assumed
to be zero with conversion of Vegetated Open Water Coastal Wetlands to Vegetated Coastal Wetlands.
Uncertainty
Underlying uncertainties in estimates of soil and biomass C stock changes include uncertainties associated with
country-specific (Tier 2) literature values of these C stocks and assumptions that underlie the methodological
approaches applied and uncertainties linked to interpretation of remote sensing data. Uncertainty specific to
coastal wetlands include differentiation of palustrine and estuarine community classes that 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 applied to each, respectively (i.e., applying approach for asymmetrical
errors, where the largest uncertainty for any one soil C stock referenced using published literature values for a
community class; uncertainty approaches provide that if multiple values are available for a single parameter, the
highest uncertainty value should be applied to the propagation of errors; IPCC 2000). For aboveground biomass C
stocks, the mean standard error was very low and largely influenced by error in estimated map area (Byrd et al.
6-112 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
2018). Uncertainty for root to shoot ratios, which are used for quantifying 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 subsource (i.e., soil C, biomass C and DOM emissions).
The combined uncertainty across all subsources is +/-33.4 percent. In 2020, the total C flux was -0.1 MMT CO2 Eq.,
with lower and upper estimates of-0.1 and -0.07 MMT CO2 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 2020 (MMT CO2 Eq. and Percent)
Source
2020 Flux Estimate
(MMT CO? Eq.)
Uncertainty Range
(MMT CO? Eq.)
Relative to Flux Estimate
(%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Biomass C Stock Flux
Dead Organic Matter C Stock Flux
Soil C Stock Flux
(0.1)
0
(0.007)
(0.12)
0
(0.008)
(0.08)
0
(0.005)
-20.0%
-25.8%
-17.8%
20.0%
25.8%
17.8%
Total Flux
(0.1)
(0.14)
(0.07)
-33.4%
33.4%
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
No recalculations were needed for the current 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 published data quantifying soil C stock
and biomass in coastal wetlands. Reference values for soil and biomass C stocks will be updated as new data
emerge. Refined error analysis combining land cover change, soil and biomass C stock estimates will be updated at
those times.
Land Use, Land-Use Change, and Forestry 6-113
-------
The USGS is investigating higher resolution mapping approaches to quantify conversion of coastal wetlands. Such
approaches may form the basis for a full Approach 3 land representation assessment in future years. C-CAP data
harmonization with the National Land Cover Dataset (NLCD) will be incorporated into a future iteration of the
inventory.
N20 Emissions from Aquaculture in Coastal Wetlands
Shrimp and fish cultivation in coastal areas increases nitrogen loads resulting in direct emissions of N2O. Nitrous
oxide is generated and emitted as a byproduct of the conversion of ammonia (contained in fish urea) to nitrate
through nitrification and nitrate to N2 gas through denitrification (Hu et al. 2012). Nitrous oxide emissions can be
readily estimated from data on fish production (IPCC 2014).
Aquaculture production in the United States has fluctuated slightly from year to year, with resulting N2O emissions
increasing from 0.1 in 1990 to upwards of 0.2 MMT CO2 Eq. between 1992 and 2010, and reducing again to 0.1
MMT CO2 Eq. between 2015 and 2020 (Table 6-71). Aquaculture production data were updated through 2018;
data through 2020 are not yet available and in this analysis are held constant with 2018 emissions of 0.2 MMT CO2
Eq. (0.5 Kt N2O).
Table 6-71: N2O Emissions from Aquaculture in Coastal Wetlands (MMT CO2 Eq. and kt N2O)
Year
1990
2005
2016
2017
2018
2019
2020
Emissions (MMT C02 Eq.)
0.1
0.2
0.1
0.1
0.2
0.2
0.2
Emissions (kt N20)
0.4
0.6
0.5
0.5
0.5
0.5
0.5
Methodology and Time-Series Consistency
The methodology to estimate N2O emissions from Aquaculture in Coastal Wetlands follows the Tier 1 guidance in
the Wetlands Supplement by applying country-specific fisheries production data and the IPCC Tier 1 default
emission factor.
Each year NOAA Fisheries document the status of U.S. marine fisheries in the annual report of Fisheries of the
United States (National Marine Fisheries Service 2021), from which activity data for this analysis is derived.68 The
fisheries report has been produced in various forms for more than 100 years, primarily at the national level, on
U.S. recreational catch and commercial fisheries landings and values. In addition, data are reported on U.S.
aquaculture production, the U.S. seafood processing industry, imports and exports offish-related products, and
domestic supply and per capita consumption of fisheries products. Within the aquaculture chapter, the mass of
production for catfish, striped bass, tilapia, trout, crawfish, salmon and shrimp are reported. While some of these
fisheries are produced on land and some in open water cages within coastal wetlands, all have data on the
quantity of food stock produced, which is the activity data that is applied to the IPCC Tier 1 default emissions
factor to estimate emissions of N2O from aquaculture. It is not apparent from the data as to the amount of
aquaculture occurring above the extent of high tides on river floodplains. While some aquaculture occurs on
coastal lowland floodplains, this is likely a minor component of tidal aquaculture production because of the need
for a regular source of water for pond flushing. The estimation of N2O emissions from aquaculture is not sensitive
to salinity using IPCC approaches and as such the location of aquaculture ponds within the boundaries of coastal
wetlands does not influence the calculations.
Other open water shellfisheries for which no food stock is provided, and thus no additional N inputs, are not
applicable for estimating N2O emissions (e.g., clams, mussels, and oysters) and have not been included in the
analysis. The IPCC Tier 1 default emissions factor of 0.00169 kg N2O-N per kg offish/shellfish produced is applied to
the activity data to calculate total N2O emissions.
68 See https://www.fisheries.noaa.gov/resource/document/fisheries-united-states-2019-report; accessed August 2021.
6-114 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2020.
Uncertainty
Uncertainty estimates are based upon the Tier 1 default 95 percent confidence interval provided in Table 4.15,
chapter 4 of the Wetlands Supplement for N2O emissions and on expert judgment of the NOAA Fisheries of the
United States fisheries production data. Given the overestimate of fisheries production from coastal wetland areas
due to the inclusion offish production in non-coastal wetland areas, this is a reasonable initial first approximation
for an uncertainty range.
Uncertainty estimates for N2O emissions from aquaculture production are presented in Table 6-72 for N2O
emissions. The combined uncertainty is +/-H6 percent. In 2020, the total flux was 0.16 MMT CO2 Eq., with lower
and upper estimates of 0.00 and 0.34 MMT CO2 Eq.
Table 6-72: Approach 1 Quantitative Uncertainty Estimates for N2O Emissions from
Aquaculture Production in Coastal Wetlands in 2020 (MMT CO2 Eq. and Percent)
2020 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
0.16
0.00 0.34
-116% 116%
Wetlands
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
QA/QC and Verification
NOAA provided internal QA/QC review of reported fisheries data. The Coastal Wetlands Inventory team consulted
with the Coordinating Lead Authors of the Coastal Wetlands chapter of the Wetlands Supplement to assess which
fisheries production data to include in estimating emissions from aquaculture. It was concluded that N2O emissions
estimates should be applied to any fish production to which food supplement is supplied be they pond or coastal
open water and that salinity conditions were not a determining factor in production of N2O emissions.
Recalculations Discussion
A NOAA report was released in 2021 that contains updated fisheries data through 2018 and the 2017 production
estimate was revised from 283,808 to 286,287 MT, although it did not affect the resulting emissions (National
Marine Fisheries Service 2021). The updated production value was applied for 2017, and the 2018 value was
applied in 2019 and 2020. This resulted in an increase of N2O emissions by 0.02 MMT CO2 Eq. (0.04 kt N2O), a 7.7
percent increase, for 2018 and 2019 compared to the previous Inventory.
Flooded Land Remaining Flooded Land
Flooded lands are defined as water bodies where human activities have 1) caused changes in the amount of
surface area covered by water, typically through water level regulation (e.g., constructing a dam), 2) waterbodies
where human activities have changed the hydrology of existing natural waterbodies thereby altering water
residence times and/or sedimentation rates, in turn causing changes to the natural emission of greenhouse gases,
and 3) waterbodies that have been created by excavation, such as canals, ditches and ponds (IPCC 2019). Flooded
lands include waterbodies with seasonally variable degrees of inundation, but these waterbodies would be
expected to retain some inundated area throughout the year under normal conditions.
Land Use, Land-Use Change, and Forestry 6-115
-------
Flooded lands are broadly classified as "reservoirs" or "other constructed waterbodies" (IPCC 2019). Reservoirs are
defined as flooded land greater than 8 ha and includes the seasonally flooded land on the perimeter of
permanently flooded land (i.e., inundation areas). IPCC guidance (IPCC 2019) provides default emission factors for
reservoirs and several types of "other constructed waterbodies" including freshwater ponds and canals/ditches.
Land that has been flooded for greater than 20 years is defined as Flooded Land Remaining Flooded Land and land
flooded for 20 years or less is defined as Land Converted to Flooded Land. The distinction is based on literature
reports that CFU and CO2 emissions are high immediately following flooding as labile organic matter is rapidly
degraded but declines to a steady background level approximately 20 years after flooding. Emissions of CH4 are
estimated for Flooded Land Remaining Flooded Land, but CO2 emissions are not included as they are primarily the
result of decomposition of organic matter entering the waterbody from the catchment or contained in inundated
soils are included elsewhere in the IPCC guidelines (IPCC 2006).
Nitrous oxide emissions from flooded lands are largely related to input of organic or inorganic nitrogen from the
watershed. These inputs from runoff/leaching/deposition are largely driven by anthropogenic activities such as
land-use change, wastewater disposal or fertilizer application in the watershed or application of fertilizer or feed in
aquaculture. These emissions are not included here to avoid double-counting of N2O emissions which are captured
in other source categories, such as indirect N2O emissions from managed soils (Volume 4, Chapter 11, 2006 IPCC
Guidelines) and wastewater management (Volume 5, Chapter 6, 2006 IPCC Guidelines).
Emissions from Flooded Land Remaining Flooded Land-
Reservoirs
Reservoirs are designed to store water for a wide range of purposes including hydropower, flood control, drinking
water, and irrigation. The permanently wetted portion of reservoirs are typically surrounded by periodically
inundated land referred to as a "drawdown zone" or "inundation area." Greenhouse gas emissions from
inundation areas are considered significant and similar per unit area to the emissions from the water surface and
are therefore included in the total reservoir surface area when estimating greenhouse gas emissions from flooded
land. Lakes converted into reservoirs without substantial changes in water surface area or water residence times
are not considered to be managed flooded land (see Area Estimates below) (IPCC 2019).
In 2020, the conterminous United States hosted 7.2 million hectares of reservoir and associated inundation areas
in the Flooded Land Remaining Flooded Land category (see Methods below for calculation details). These
reservoirs are distributed across all six of the aggregated climate zones used to define flooded land emission
factors (Figure 6-10) (IPCC 2019). Alaska, Hawaii, and U.S. Territories are not included in this report due to a lack of
data (see Methodology).
6-116 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure 6-10: U.S. reservoirs (black polygons) in the Flooded Land Remaining Flooded Land
category in 2020. Colors represent climate zone used to derive IPCC default emission
factors.
boreal
cool temperate
tropical dry/montane
tropical moist/wet
warm temperate dry
warm temperate moist
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 surface of
reservoirs and inundation areas or downstream of the flooded land as CH4 enriched water passes through the dam
and the downstream river.
Table 6-73 and Table 6-74 below summarize nationally aggregated CH4 emissions from reservoirs and associated
inundation areas. 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
2016
2017
2018
2019
2020
Reservoirs
16.0
17.4
17.5
17.5
17.5
17.5
17.5
Surface Emission
14.7
16.0
16.1
16.1
16.1
16.1
16.1
Downstream Emission
1.3
1.4
1.4
1.4
1.4
1.4
1.4
Inundation Areas
1.2
1.3
1.3
1.3
1.3
1.3
1.3
Surface Emission
1.1
1.2
1.2
1.2
1.2
1.2
1.2
Downstream Emission
0,1
0.1
0.1
0.1
0.1
0.1
0.1
Total
17.2
18.7
18.8
18.8
18.8
18.8
18.8
Note: Alaska, Hawaii, and U.S. Territories not included.
Table 6-74: CH4 Emissions from Flooded Land Remaining Flooded Land—Reservoirs (kt Cl-U)
Source
1990
2005
2016
2017
2018
2019
2020
Reservoirs
640
696
700
700
700
700
700
Surface Emission
587
639
642
642
642
642
642
Downstream Emission
53
57
58
58
58
58
58
Inundation Areas
48
53
53
53
53
53
53
Surface Emission
44
48
49
49
49
49
49
Downstream Emission
4
4
4
4
4
4
4
Total
688
749
753
753
753
753
753
Note: Alaska, Hawaii, and U.S. Territories not included.
Land Use, Land-Use Change, and Forestry 6-117
-------
Methane emissions from reservoirs and inundation areas are greatest in Texas (Figure 6-11, Table 6-75) due to 1)
the large expanse of reservoirs and inundation areas in the state (Figure 6-10) and 2) the high CFU emission factor
for the tropical dry/montane climate zone which encompasses a majority of the flooded land area in the state
(Figure 6-10, Table 6-75). Florida has the second greatest CFU emission from reservoirs and inundation areas in the
United States, but the emissions are less than half of that from Texas. Louisiana and Georgia have the third and
fourth greatest CFU emission, respectively, in accordance with the relatively high extent of flooded lands in the
states and the high emission factor for CFU in the tropical moist/wet biome.
Twenty five percent of the increase in Cm emissions from 1990 to 2005 for this subcategory is due to the
transition of Lakes Sakakawea and Oahe in North Dakota and South Dakota to Flooded Land Remaining Flooded
Land between 2000 and 2003 (i.e., they were emitting Cm prior to 2000 and the emissions were included in the
Land Converted to Flooded Land category but these emissions are now included in Land Converted to Flooded
Land). Combined, these two large reservoirs have a surface area in excess of 0.25 million hectares.
Figure 6-11: Total ChU Emissions (Downstream + Surface) from Reservoirs and Associated
Inundation Areas in Flooded Land Remaining Flooded Land (kt ChU)
kt CH4 y~1
125
*
100
75
50
Table 6-75: Surface and Downstream ChU Emissions (kt ChU) from Reservoirs and
Associated Inundation Areas in Flooded Land Remaining Flooded Land in 2020
Reservoir Inundation area Total
State
Downstream
Ssurface
Downstream
Surface
Downstream
Surface
Alabama
2
19
+
+
2
19
Arizona
1
12
+
1
1
14
Arkansas
2
17
+
1
2
17
California
2
24
+
1
2
26
Colorado
+
5
+
+
+
6
Connecticut
+
2
+
+
+
2
Delaware
+
1
NO
NO
+
1
District of Columbia
+
+
NO
NO
+
+
Florida
4
47
NO
NO
4
47
Georgia
3
33
+
+
3
33
Idaho
1
8
+
+
1
8
Illinois
1
9
+
1
1
10
Indiana
+
3
+
1
+
3
Iowa
+
4
+
2
1
6
Kansas
1
6
+
3
1
9
Kentucky
1
10
+
1
1
11
6-118 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Louisiana
3
33
+
1
3
33
Maine
1
11
NO
NO
1
11
Maryland
1
11
+
+
1
11
Massachusetts
+
3
+
+
+
3
Michigan
1
6
+
+
1
6
Minnesota
2
19
NO
NO
2
19
Mississippi
1
12
+
3
1
15
Missouri
1
13
+
2
1
14
Montana
1
10
+
+
1
10
Nebraska
+
3
+
+
+
3
Nevada
2
18
+
+
2
18
New Hampshire
+
3
+
+
+
3
New Jersey
+
3
NO
NO
+
3
New Mexico
+
3
+
2
+
5
New York
1
10
+
+
1
11
North Carolina
2
19
+
1
2
20
North Dakota
1
10
+
+
1
10
Ohio
+
4
+
1
+
5
Oklahoma
2
19
+
5
2
24
Oregon
1
11
NO
NO
1
11
Pennsylvania
+
4
+
+
+
5
Rhode Island
+
1
NO
NO
+
1
South Carolina
2
23
NO
NO
2
23
South Dakota
1
8
+
+
1
8
Tennessee
1
16
+
1
1
16
Texas
9
99
2
20
11
119
Utah
2
18
+
+
2
18
Vermont
+
2
+
+
+
2
Virginia
2
18
+
+
2
19
Washington
2
18
+
+
2
18
West Virginia
+
2
+
+
+
2
Wisconsin
1
10
NO
NO
1
10
Wyoming
+
4
+
+
+
4
+ Indicates values less than 0.5 kt
NO (Not Occurring)-lndicates no reservoir or inundation area in the state.
Note: Alaska and Hawaii not included.
Methodology and Time-Series Consistency
Estimates of CFU emission for reservoirs and associated inundation areas 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 and inundation areas 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 6 climate zones using a regression tree approach. All six aggregated climate zone
are present in the conterminous United States.
Land Use, Land-Use Change, and Forestry 6-119
-------
Table 6-76: IPCC (2019) Default ChU Emission Factors for Surface Emission from Reservoirs
and Associated Inundation Areas 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
Area estimates
The Reservoirs in the conterminous United States were identified from the NHDArea and NHDWaterbody layers in
the National Hydrography Dataset Plus V2 (NHD),69 the National Lakes Assessment (NLA)70 data, the National
Inventory of Dams (NID),71 and the Navigable Waterways (NW)72 dataset. The NHD and NLA do not include Alaska,
Hawaii, or U.S. Territories, thus these areas are not included in the Inventory. Waterbodies in these data sets that
were greater than 20 years old, greater than 8 ha in surface area, and not identified as canal/ditch in NHD or NW
and met any of the following criteria were considered reservoirs in Flooded Land Remaining Flooded Land: 1) the
water body was classified "Reservoir" in the NHDWaterbody layer, 2) the water body name in the NHDWaterbody
layer included "reservoir", 3) the waterbody in the NHDWaterbody layer was located in close proximity to a dam in
the NID, 4) the water body was deemed "man-made" in the NLA, 5) the waterbody was included in NW, and 6)
inundation areas in the NHDArea layer that were associated with water bodies that met any of the above criteria
were assumed to represent drawdown zones and were included in the inventory of reservoirs.
The IPCC (2019) allows for the exclusion of reservoirs from the Inventory if the water surface area or residence
time was not substantially changed by the construction of the dam. The guidance does not quantify what
constitutes a "substantial" change, but here EPA excludes the U.S. Great Lakes from the Inventory based on expert
judgment that neither the surface area or water residence time was substantially altered by their associated dams.
EPA assumes that all other waterbodies identified through the six criteria described above were substantially
impacted by the construction of dams.
EPA assumes that all reservoirs included in the NW are subject to water-level management to maintain minimum
water depths required for navigation and are therefore included in the inventory. Reservoir age was determined
from the year the dam was completed as reported in the NID (available for 40,012 out of 54,670 reservoirs). When
dam completion year was not available, the reservoir was assumed to be greater than 20 years old. Reservoirs
were disaggregated by state (using boundaries from the 2016 U.S. Census Bureau73) and climate zone.
Downstream and surface emissions for cross-state reservoirs were allocated to states based on the surface area
that the reservoir occupied in each state. Only the U.S. portion of reservoirs that cross country borders were
included in the Inventory. Surface areas for reservoirs and associated inundation areas were taken from NHD or
the NW and the final inventory of reservoirs and associated inundation areas was screened to ensure no
waterbodies were duplicated.
Many reservoirs are surrounded by land that is dry for a portion of the year but is periodically flooded when water
inflows to the reservoir exceed outflows and the reservoir surface area expands into surrounding lands. This can
occur for a variety of reasons including high rates of water runoff from the watershed (i.e., snow melt, large
precipitation events), deliberate efforts to raise water levels for seasonal recreation or wildlife habitat, and
69 See https
//www.usgs.gov/core-science-svstems/ngp/national-hvdrographv.
70 See https
//www.epa.Rov/national-aquatic-resource-surveys/nla.
71 See https
//nid.sec.usace.armv.mil.
72 See https
//hif]d-geopiatform,opendata,arcgis,com/datasets/geoplatform::navigable-waterwav-network-lines-l/about.
73 See https
//www.census.gov/geographies/mapRing-files/time-series/geo/carto-boundarv-file.html.
6-120 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
management efforts to reduce inflows to downstream systems. These periodically flooded lands are represented
as "Inundation Areas" in the NHDArea layer (Figure 6-12). Inundation areas are considered equivalent to
"drawdown zones" in IPCC (2019) and CFU emissions from these lands are estimated using the same methodology
as for reservoirs.
The surface area of reservoirs and associated inundation areas in Flooded Land Remaining Flooded Land increased
by approximately 10 percent from 1990 to 2020 (Table 6-77) due to reservoirs matriculating into Flooded Land
Remaining Flooded Land when they reached 20 years of age.
Figure 6-12: Example of a Reservoir and Associated Inundation Area
Table 6-77: National Totals of Reservoirs and Associated Inundation Area Surface Area
(millions of ha) in Flooded Land Remaining Flooded Land
Surface Area (millions of ha)
1990 ¦ 2005 ¦ 2016
2017
2018
2019
2020
Reservoir
Inundation Area
6.05 6.70 6.76
0.39 0.43 0.44
6.76
0.44
6.76
0.44
6.76
0.44
6.76
0.44
Note: Alaska, Hawaii, and U.S. Territories not included.
Table 6-78: State breakdown of Reservoirs and Associated Inundation Area Surface Area
(millions of ha) in Flooded Land Remaining Flooded Land
State
1990
2005
2016
2017
2018
2019
2020
Alabama
0.17
0.21
0.21
0.21
0.21
0.21
0.21
Arizona
0.06
0.06
0.06
0.06
0.06
0.06
0.06
Arkansas
0.17
0.20
0.20
0.20
0.20
0.20
0.20
California
0.25
0.26
0.26
0.26
0.26
0.26
0.26
Colorado
0.07
0.07
0.07
0.07
0.07
0.07
0.07
Connecticut
0.03
0.03
0.03
0.03
0.03
0.03
0.03
Delaware
0.02
0.02
0.02
0.02
0.02
0.02
0.02
District of Columbia
0.00
0.00
0.00
0.00
0.00
0.00
0.00
Florida
0.32
0.33
0.33
0.33
0.33
0.33
0.33
Georgia
0.24
0.25
0.25
0.25
0.25
0.25
0.25
Idaho
0.12
0.13
0.13
0.13
0.13
0.13
0.13
Illinois
0.11
0.12
0.13
0.13
0.13
0.13
0.13
Indiana
0.03
0.04
0.04
0.04
0.04
0.04
0.04
Land Use, Land-Use Change, and Forestry 6-121
-------
Iowa
0.06
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
0.13
0.13
0.13
0.13
0.13
0.13
0.13
Louisiana
0.22
0.23
0.24
0.24
0.24
0.24
0.24
Maine
0.20
0.21
0.21
0.21
0.21
0.21
0.21
Maryland
0.14
0.14
0.14
0.14
0.14
0.14
0.14
Massachusetts
0.04
0.04
0.04
0.04
0.04
0.04
0.04
Michigan
0.11
0.12
0.12
0.12
0.12
0.12
0.12
Minnesota
0.33
0.34
0.34
0.35
0.35
0.35
0.35
Mississippi
0.14
0.15
0.15
0.15
0.15
0.15
0.15
Missouri
0.12
0.18
0.18
0.18
0.18
0.18
0.18
Montana
0.18
0.19
0.19
0.19
0.19
0.19
0.19
Nebraska
0.04
0.04
0.04
0.04
0.04
0.04
0.04
Nevada
0.10
0.10
0.10
0.10
0.10
0.10
0.10
New Hampshire
0.05
0.05
0.05
0.05
0.05
0.05
0.05
New Jersey
0.03
0.03
0.03
0.03
0.03
0.03
0.03
New Mexico
0.05
0.05
0.05
0.05
0.05
0.05
0.05
New York
0.18
0.18
0.18
0.18
0.18
0.18
0.18
North Carolina
0.23
0.25
0.25
0.25
0.25
0.25
0.25
North Dakota
0.03
0.19
0.19
0.19
0.19
0.19
0.19
Ohio
0.07
0.07
0.07
0.07
0.07
0.07
0.07
Oklahoma
0.25
0.27
0.27
0.27
0.27
0.27
0.27
Oregon
0.15
0.15
0.15
0.15
0.15
0.15
0.15
Pennsylvania
0.07
0.07
0.07
0.07
0.07
0.07
0.07
Rhode Island
0.01
0.01
0.01
0.01
0.01
0.01
0.01
South Carolina
0.19
0.20
0.20
0.20
0.20
0.20
0.20
South Dakota
0.05
0.15
0.15
0.15
0.15
0.16
0.16
Tennessee
0.20
0.20
0.20
0.20
0.20
0.20
0.20
Texas
0.57
0.62
0.62
0.62
0.62
0.62
0.62
Utah
0.13
0.14
0.17
0.17
0.17
0.17
0.17
Vermont
0.03
0.03
0.03
0.03
0.03
0.03
0.03
Virginia
0.23
0.23
0.23
0.23
0.23
0.23
0.23
Washington
0.19
0.19
0.19
0.19
0.19
0.19
0.19
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
0.07
0.07
0.08
0.08
0.08
0.08
0.08
Total
6.44
7.13
7.20
7.20
7.20
7.20
7.20
Note: Alaska, Hawaii, and U.S. Territories not included.
Uncertainty
Uncertainty in estimates of Cm emissions from reservoirs and associated inundation areas in Flooded Land
Remaining Flooded Land (Table 6-79) are developed using the IPCC Approach 2 and include uncertainty in the
default emission factors and land areas. Uncertainty ranges for the emission factors are provided in the 2019
Refinement to the 2006 IPCC Guidelines (IPCC 2019). Uncertainties in the spatial data include 1) uncertainty in area
estimates from the NHD and NW, and 2) uncertainty in the location of dams in the NID. Overall uncertainties in
these spatial datasets are unknown, but uncertainty for remote sensing products is assumed to be ±10 to 15
percent based on IPCC guidance (IPCC 2003). An uncertainty range of +/-15 percent for the reservoir and
associated inundation area estimates is assumed and is based on expert judgment.
Table 6-79: Approach 2 Quantitative Uncertainty Estimates for ChU Emissions from
Reservoirs and Associated Inundation Areas in Flooded Land Remaining Flooded Land
Source
Gas
2020 Emission Estimate
Uncertainty Range Relative to Emission
Estimate3
(MMT CO? Eq.)
(MMTCOzEq.) (%)
Lower Upper Lower Upper
Bound Bound Bound Bound
6-122 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Reservoir
Surface
ch4
16.1
15.7
16.4
-2%
2%
Downstream
ch4
1.5
1.4
1.8
-6.9%
22.4%
Inundation Area
Surface
ch4
1.2
1.2
1.2
-2.3%
2.5%
Downstream
ch4
0.1
0.1
0.1
-10.1%
17.5%
Total
ch4
18.8
18.5
19.3
-1.8%
2.6%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
The National Hydrography Data (NHD) is managed by the USGS in collaboration with many other federal, state, and
local entities. Extensive QA/QC procedures are incorporated into the curation of the NHD. The National Inventory
of Dams (NID) is maintained by the U.S. Army Corps of Engineers (USACE) in collaboration with the Federal
Emergency Management Agency (FEMA) and state regulatory offices. USACE resolves duplicative and conflicting
data from 68 data sources, which helps obtain the more complete, accurate, and updated NID. The National Lakes
Assessment is a survey of U.S. lakes and reservoirs conducted by the U.S. Environmental Protection Agency every 5
years. The program is subject to rigorous QA/QC as detailed in the Quality Assurance Project Plan.74 The Navigable
Waterways (NW) dataset is part of the U.S. Department of Transportation (USDOT)/Bureau of Transportation
Statistics's (BTS's) National Transportation Atlas Database (NTAD). The NW is a comprehensive network database
of the nation's navigable waterways updated on a continuing basis.
All calculations were executed independently in Excel and R. Ten percent of state and national totals were
randomly selected for comparison between the two approaches to ensure there were no computational errors.
Recalculations Discussion
This is a new category in the current Inventory.
Planned Improvements
The EPA is measuring greenhouse gas emissions from 108 flooded lands (reservoirs) in the conterminous United
States. The survey is expected to be complete by September 2023 and the data will be used to predict greenhouse
gas emission rates for all U.S. flooded lands. The Inventory will be updated at that time using these country-specific
emission factors developed through the measurement campaign.
Hawaii, Alaska, and U.S. Territories will be included in the next (i.e., 1990 through 2021) Inventory. Flooded lands
area data for these states and territories will be derived from the National Hydrography Dataset Plus High
Resolution (NHDPIus HighRes),75 an enhanced version of the NHD used in this Inventory.
To verify that waterbodies contained in NW are subject to water level management, EPA will overlay the NW with
other spatial datasets of water control structures including the inventory of U.S. Army Corps of Engineers locks for
water navigation76 and dams/weirs contained in the NHDPIus HighRes.
74 See https://www.epa.Eov/national-aauatic-resource-survevs/national-lakes-assessment-2017-qualitv-assurance-proiect-
plan.
75 See https://www.uses.eov/core-science-svstems/nep/national-hvdrographv/nhdplus-high-resolution.
76 See https://hifld-eeoplatform.opendata.arcgis.com/datasets/eeoplatform::locks/about.
Land Use, Land-Use Change, and Forestry 6-123
-------
Emissions from Flooded Land Remaining Flooded Land-Other
Constructed Waterbodies
The IPCC (IPCC 2019) provides emission factors for several types of "other constructed waterbodies" including
freshwater ponds and canals/ditches. IPCC (2019) describes ponds as waterbodies that are "...constructed by
excavation and/or construction of walls to hold water in the landscape for a range of uses, including agricultural
water storage, access to water for livestock, recreation, and aquaculture." Furthermore, the IPCC "Decision tree
for types of Flooded Land" (IPCC 2019, Fig. 7.2) defines a size threshold of 8 ha to distinguish reservoirs from
"other constructed waterbodies." For this Inventory, ponds are defined as managed flooded land that are 1) less
than 8 ha in surface area, and 2) not categorized as canals/ditches. IPCC (2019) further distinguishes saline versus
brackish ponds, with the former supporting lower CFU emissions than the latter. Activity data on pond salinity are
not uniformly available for the conterminous United States and all ponds in the inventory are assumed to be
freshwater. Ponds often receive high organic matter and nutrient loadings, may have low oxygen levels, and are
often sites of substantial CFU emissions from anaerobic sediments.
Canals and ditches (terms are used interchangeably) are linear water features constructed to transport water (i.e.
stormwater drainage, aqueduct), to irrigate or drain land, to connect two or more bodies of water, or to serve as a
waterway for watercraft. The geometry and construction of canals and ditches varies widely and includes narrow
earthen channels (
-------
Table 6-82: ChU Emissions (kt ChU) from Other Constructed Waterbodies in Flooded Land
Remaining Flooded Land in 2020
State
Canals and Ditches
Freshwater Ponds
Total
Alabama
+
0.6
0.6
Arizona
+
+
+
Arkansas
+
0.6
0.9
California
1.7
+
2.1
Colorado
+
0.6
0.6
Connecticut
+
+
+
Delaware
+
+
+
District of Columbia
+
+
+
Florida
1.9
+
2.0
Georgia
+
1.6
1.6
Idaho
+
+
+
Illinois
+
+
0.8
Indiana
+
+
+
Iowa
+
0.7
0.7
Kansas
+
1.0
1.0
Kentucky
+
+
+
Louisiana
12.5
+
12.7
Maine
+
+
+
Maryland
+
+
+
Massachusetts
+
+
+
Michigan
+
+
+
Minnesota
+
+
+
Mississippi
1.7
1.6
3.3
Missouri
+
1.4
1.5
Montana
+
1.3
1.4
Nebraska
+
+
0.5
Nevada
+
+
+
New Hampshire
+
+
+
New Jersey
+
+
+
New Mexico
+
+
+
New York
+
+
0.9
North Carolina
+
0.7
0.9
North Dakota
+
+
+
Ohio
+
+
+
Oklahoma
+
1.3
1.3
Oregon
+
+
+
Pennsylvania
+
+
+
Rhode Island
+
+
+
South Carolina
+
0.8
0.9
South Dakota
+
+
+
Tennessee
+
+
+
Texas
1.6
2.4
4.0
Utah
+
+
+
Vermont
+
+
+
Virginia
+
0.7
0.7
Washington
+
+
+
West Virginia
+
+
+
Wisconsin
+
+
+
Wyoming
+
+
+
Total
21.6
22.2
43.8
+ Indicates values less than 0.5 kt
Note: Alaska, Hawaii, and U.S. Territories not included.
Land Use, Land-Use Change, and Forestry 6-125
-------
Figure 6-13: ChU Emissions (kt ChU) from Other Constructed Waterbodies in Flooded Land
Remaining Flooded Land in 2020
kt CH4 y"1
12.0
_ 1.5
1.0
0.5
Methodology and Time-Series Consistency
Estimates of Cm emission 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).
Table 6-83: IPCC (2019) Default ChU Emission Factors for Surface Emissions from Other
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
Area estimates
Freshwater ponds in the conterminous United States were identified from the NHDArea and NHDWaterbody layers
in the National Hydrography Dataset Plus V2 (NHD),77 the National Lakes Assessment (NLA)78 data, the National
Inventory of Dams (NID),79 and the Navigable Waterways (NW)80 dataset. The NHD and NLA do not include Alaska,
Hawaii, or U.S. Territories, thus these areas are not included in the Inventory. Waterbodies in these data sets that
were greater than 20 years old, less than 8 ha in surface area, and not identified as canal/ditch in NHD or NW and
met any of the following criteria were considered ponds in Flooded Land Remaining Flooded Land: 1) the water
body was classified "Reservoir" in the NHDWaterbody layer, 2) the water body name in the NHDWaterbody layer
included "reservoir", 3) the water body in the NHDWaterbody layer was located in close proximity to a dam in the
77 See https://www.usgs.gov/core-science-svstems/ngp/national-hvdrographv.
78 See https://www.epa.gov/national-aauatic-resource-survevs/nla.
79 See https://nid.sec.usace.armv.mil.
80 See https://hifld-geoplatform.opendata.arcgis.com/datasets/geoplatform::navigable-waterwav-network-lines-l/about
6-126 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
NID, 4) the water body was deemed "man made" in the NLA, 5) the waterbody was included in NW, and 6)
inundation areas in the NHDArea layer that were associated with water bodies that met any of the above criteria
were assumed to represent drawdown zones and were included in the ponds inventory.
Surface areas for ponds and canals/ditches were taken from NHD or the NW. Waterbodies were disaggregated by
state (using boundaries from the 2016 U.S. Census Bureau81) and the final area inventory was screened to ensure
no waterbodies were duplicated. While the distribution of U.S. waterbodies <8 ha is well represented in NHD, it is
difficult to determine which of these waterbodies are subject to water level management. The presence or
absence of a flow control structure associated with these small waterbodies is typically not documented in NHD,
thus EPA used the NID for this purpose. The NID only includes dams that pose a hazard if they were to fail, equal or
exceed 25 feet in height and exceed 15 acre-feet in storage, or equal or exceed 50 acre-feet storage and exceed 6
feet in height.82 The extent to which these criteria fail to capture flow control structures associated with
freshwater ponds in the United States in unknown, but the freshwater pond area estimates presented here
certainly underestimates the surface area of U.S. freshwater ponds. There is a planned improvement to review
other data sources or approaches that could more fully capture all managed freshwater ponds in the United
States.
All waterbodies identified as "CANAL/DITCH" in the NHDArea layer of the NHD were classified as canals and
ditches , a subcategory of other constructed waterbodies (IPCC 2019), for this Inventory. This is an underestimate
of U.S. canals and ditches, however, because the majority of canal and ditch length is represented as one
dimensional flow lines in the NHDFIowline_Network layer of the NHD. While NHD flowlines can be used to
estimate length of ditches and canals, they are one-dimensional features and do not provide area estimates. There
is a planned improvement to review other data sources for approaches to better capture the surface area of
canals/ditches in the United States.
The age of freshwater ponds was derived from NID when available, otherwise they were assumed to be greater
than 20 years old throughout the time series. Age data were not available for canals and ditches and they were
assumed to be greater than 20 years old in 1990 and therefore included in Flooded Land Remaining Flooded Land
throughout the time series. For the year 2020, this Inventory contains 121,255 ha of freshwater ponds and 51,834
ha of canals and ditches in Flooded Land Remaining Flooded Land. The surface area of freshwater ponds increased
by 15 percent from 1990 to 2020 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.
Overall, the surface area of other constructed waterbodies increased 10 percent between 1990 and 2020, with
most of the increase occurring by 2005 (Table 6-84).
Table 6-84: National Surface Area (ha) Totals in Flooded Land Remaining Flooded Land -
Other Constructed Waterbodies
1990
2005
2016
2017
2018
2018
2020
Canals and ditches
Freshwater ponds
51,834
105,859
51,834
120,373
51,834
121,014
51,834
121,067
51,834
121,167
51,834
121,215
51,834
121,255
Total
157,693
172,207
172,848
172,901
173,001
173,049
173,089
Note: Alaska, Hawaii, and U.S. Territories not included.
Canals and ditches in the conterminous United States are most abundant in the Gulf Coast states and California
(Figure 6-14A). Louisiana contains over half of all U.S. canal and ditch surface area, most of which was created to
drain swamps and wetlands for agricultural production (Davis 1973). Freshwater ponds are more widely
distributed across the United States. (Figure 6-14B). Texas has the greatest surface area of freshwater ponds,
equivalent to 11 percent of all freshwater pond surface area in the United States. Texas also had the largest
increase in freshwater pond surface area during the time series.
81 See https://www.census.Eov/geographies/mapping-files/time-series/geo/carto-boundarv-file.html.
82 See https://nid.sec.usace.army.mil.
Land Use, Land-Use Change, and Forestry 6-127
-------
Figure 6-14: Surface Area (hectares) of Other Constructed Waterbodies in Flooded Land
Remaining Flooded Land in 2020
A. Area of Canals and Ditches B. Area of Freshwater Ponds
hectares
B 12,500
B 10,000
7,500
5,000
2,500
Table 6-85: State Totals of Surface Area (ha) in Flooded Land Remaining Flooded Land-
Other Constructed Waterbodies
State
1990
2005
2016
2017
2018
2019
2020
Alabama
3,007
3,244
3,252
3,252
3,255
3,255
3,255
Arizona
459
488
488
488
493
493
493
Arkansas
3,796
4,051
4,051
4,051
4,051
4,051
4,051
California
6,022
6,106
6,120
6,120
6,120
6,122
6,122
Colorado
3,227
3,367
3,381
3,381
3,390
3,390
3,390
Connecticut
1,180
1,226
1,226
1,226
1,226
1,226
1,226
Delaware
481
488
488
488
488
488
488
District of Columbia
6
6
6
6
6
6
6
Florida
5,169
5,214
5,225
5,225
5,225
5,228
5,228
Georgia
8,069
8,673
8,687
8,687
8,687
8,687
8,687
Idaho
389
433
433
433
433
433
433
Illinois
2,998
3,318
3,344
3,345
3,350
3,350
3,350
Indiana
756
836
843
843
843
843
843
Iowa
2,529
3,574
3,596
3,596
3,596
3,600
3,607
Kansas
3,724
5,357
5,381
5,388
5,404
5,411
5,411
Kentucky
1,143
1,327
1,327
1,330
1,330
1,330
1,330
Louisiana
30,900
30,991
30,995
30,995
30,995
30,995
30,995
Maine
224
243
247
247
247
247
247
Maryland
574
609
615
615
615
615
615
Massachusetts
1,813
1,871
1,897
1,902
1,908
1,912
1,919
Michigan
1,082
1,172
1,183
1,185
1,185
1,185
1,185
Minnesota
1,042
1,133
1,137
1,137
1,137
1,137
1,151
Mississippi
12,445
12,852
12,874
12,888
12,888
12,893
12,901
Missouri
5,312
7,684
7,700
7,700
7,700
7,700
7,700
Montana
7,113
7,411
7,411
7,411
7,411
7,416
7,416
Nebraska
1,844
2,590
2,605
2,605
2,630
2,630
2,630
Nevada
242
242
260
262
262
262
262
New Hampshire
451
497
517
517
517
517
517
New Jersey
1,381
1,396
1,396
1,396
1,399
1,399
1,399
New Mexico
444
453
453
453
453
453
453
New York
3,071
3,232
3,294
3,294
3,294
3,294
3,294
North Carolina
3,977
4,178
4,216
4,216
4,216
4,216
4,218
North Dakota
784
837
866
866
873
873
873
Ohio
2,008
2,201
2,229
2,238
2,238
2,238
2,238
Oklahoma
6,138
7,139
7,162
7,162
7,165
7,172
7,172
Oregon
1,448
1,520
1,533
1,533
1,536
1,536
1,536
Pennsylvania
1,483
1,610
1,630
1,630
1,631
1,631
1,631
6-128 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Rhode Island
257
258
258
258
258
258
258
South Carolina
4,172
4,617
4,638
4,640
4,640
4,640
4,642
South Dakota
2,403
2,500
2,537
2,539
2,539
2,548
2,548
Tennessee
1,372
1,562
1,569
1,569
1,576
1,579
1,579
Texas
14,634
17,052
17,082
17,086
17,090
17,091
17,091
Utah
784
829
829
829
829
829
829
Vermont
311
368
372
372
372
372
372
Virginia
3,549
3,721
3,721
3,721
3,721
3,721
3,721
Washington
653
727
759
759
761
761
761
West Virginia
124
124
124
124
124
124
124
Wisconsin
1,009
1,158
1,170
1,170
1,170
1,170
1,170
Wyoming
1,693
1,721
1,721
1,721
1,721
1,721
1,721
Total
157,693
172,207
172,848
172,901
173,001
173,049
173,089
Note: Alaska, Hawaii, and U.S. Territories not included.
Uncertainty
Uncertainty in estimates of Cm emissions from other constructed waterbodies (ponds, canals/ditches) in Flooded
Land Remaining Flooded Land (Table 6-86) are estimated using IPCC Approach 2 and include uncertainty in the
default emission factors and the flooded land area inventory. Uncertainty in emission factors is provided in the
2019 Refinement to the 2006 IPCC Guidelines (IPCC 2019). Uncertainties in the spatial data include 1) uncertainty in
area estimates from the NHD and NW, and 2) uncertainty in the location of dams in the NID. Overall uncertainties
in these spatial datasets are unknown, but uncertainty for remote sensing products is assumed to be ±10 to 15
percent based on IPCC guidance (IPCC 2003). An uncertainty range of +/-15 percent for the flooded land area
estimates is assumed and is based on expert judgment. These uncertainties do not include the underestimate of
pond and canal/ditch surface areas discussed above, see Planned Improvements for a discussion on steps being
taken to improve the area estimates.
Table 6-86: Approach 2 Quantitative Uncertainty Estimates for ChU Emissions from Other
Constructed Waterbodies in Flooded Land Remaining Flooded Land
Source
Gas
2020 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
0.54
0.45
0.62
-15.9%
15.8%
Freshwater pond
ch4
0.55
0.55
0.56
-0.3%
0.3%
Total
ch4
1.09
1.01
1.18
-8%
7.9%
aRange of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
QA/QC and Verification
The National Hydrography Data (NHD) is managed by the USGS in collaboration with many other federal, state, and
local entities. Extensive QA/QC procedures are incorporated into the curation of the NHD. The National Inventory
of Dams (NID) is maintained by the U.S. Army Corps of Engineers (USACE) in collaboration with the Federal
Emergency Management Agency (FEMA) and state regulatory offices. USACE resolves duplicative and conflicting
data from 68 data sources, which helps obtain the more complete, accurate, and updated NID. The National Lakes
Assessment is a survey of U.S. lakes and reservoirs conducted by the U.S. Environmental Protection Agency every 5
years. The program is subject to rigorous QA/QC as detailed in the Quality Assurance Project Plan.83 The Navigable
Waterways (NW) dataset is part of the U.S. Department of Transportation (USDOT)/Bureau of Transportation
83 See https://www.epa.gov/national-aquatic-resource-surveys/national-lakes-assessment-2017-qualitv-assurance-proiect-
plan.
Land Use, Land-Use Change, and Forestry 6-129
-------
Statistics's (BTS's) National Transportation Atlas Database (NTAD). The NW is a comprehensive network database
of the nation's navigable waterways updated on a continuing basis.
All calculations were executed independently in Excel and R. Ten percent of state and national totals were
randomly selected for comparison between the two approaches to ensure there were no computational errors.
Recalculations Discussion
This is a new category in the current Inventory.
Planned Improvements
EPA is unaware of ongoing work that could be used to develop country-specific emission factors for Other
Constructed Waterbodies but is working to improve land use representation of canals and ditches. Canals and
ditches are represented as both flow lines and area polygons in NHD. The area polygons are used in this Inventory,
but flow lines only contain the length of the feature and therefore cannot be directly used to calculate surface
area. EPA is researching methods for associating flow lines with a width, which would enable the area calculations
needed for the Inventory.
Canal and ditch surface area included here may overlap with ditches and canals included in Cm emission estimates
for ditches draining inland organic soils (IPCC 2013, section 2.2.2.1). EPA plans to reconcile ditch/canal surface
areas between the two managed land types (flooded land vs drained inland organic soils) in the next (i.e., 1990
through 2021) Inventory.
U.S. waterbodies less than 8 ha are well represented in NHD, but the presence or absence of water level control
structures associated with these small waterbodies is not well documented in national data sources. To improve
the representation of managed ponds in future Inventories, EPA will attempt to locate state or county records on
small dam construction permits and/or inspection records to supplement records in the NID. EPA will also use
surrounding land use as a proxy for management. For example, a pond surrounded by cultivated land is likely
subject to water level management and should be included in the inventory. Even if the pond were not subject to
water level management, greenhouse gas emissions from the system are likely enhanced by elevated nutrient and
sediment inputs from the surrounding managed lands, thus the emissions should be considered anthropogenic and
included in the inventory.
Hawaii, Alaska, and U.S. Territories will be included in the (i.e., 1990 through 2021) Inventory. Flooded lands area
data for these states and territories will be derived from the National Hydrography Dataset Plus High Resolution
(NHDPIus HighRes),84 an enhanced version of the NHD used in this Inventory.
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. All other land
84 See https://www.usgs.gov/core-science-svstems/ngp/national-hydrographv/nhdplus-high-resolution.
6-130 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
categories (i.e., Forest Land, Cropland, Grassland, Settlements and Other Lands) are identified as having some area
converting to Vegetated Coastal Wetlands. Between 1990 and 2020 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.85 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.86 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.
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).
In this Inventory, biomass, dead organic material (DOM) and soil C stock changes as well as CFUemissions are
quantified as a result of the land use conversion to coastal wetlands and the land is assumed to be held in this
category for up to 20 years after which it is classified as Coastal Wetlands Remaining Coastal Wetlands. Estimates
of emissions and removals are based on emission factor data that have been applied to assess changes in each
respective flux for Land Converted to Vegetated Coastal Wetlands. Following conversion to Vegetated Coastal
Wetlands, it is assumed there is a loss of biomass C stocks from the converted Forest Land, Cropland and Grassland
and as well as the loss of DOM C stocks from Forest Land. Converted lands are held in this land category for up to
20 years and the assumption is that the C stock losses from biomass and DOM all occur in the year of conversion.
There are no soil C losses from land use conversion. Carbon stock increases in coastal wetlands as a result of gains
in plant biomass and DOM on these converted lands are also included during the year of transition even though
the entire C stock accrual takes many years to occur. Soil C accumulation and Cm emissions are quantified using an
annual rate in this Inventory and thus are occurring over the period under which lands are held in this category;
therefore, the soil C removals and Cm emissions presented for a given year include the cumulative
removals/emissions for the new area that was converted during that year and the area held in this category for the
prior 19 years. At salinities less than half that of seawater, the transition from upland dry soils to wetland soils
results in CFU emissions. The United States calculates emissions and removals based upon stock change.
Conversion to coastal wetlands resulted in a biomass C stock loss of 0.1 MMT CO2 Eq. (0.03 MMT C) in 2020 (Table
6-87 and Table 6-88). 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
2020 (Table 6-87 and Table 6-88), 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. Once
Tier 1 or 2 DOM values are collated and accounted for in estuarine and palustrine scrub shrub coastal wetlands
and estuarine forested wetlands (in climates other than subtropical), the carbon emissions will decrease. Across all
time periods, soil C accumulation resulting from Lands Converted to Vegetated Coastal Wetlands is a carbon sink
and has ranged between -0.2 and -0.3 MMT CO2 Eq. (-0.04 and -0.07 MMT C; Table 6-87 and Table 6-88).
Conversion of lands to coastal wetlands resulted in CH4 emissions of 0.2 MMT CO2 Eq. (6.7 kt CH4) in 2020 (Table
6-89). 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 (0.2 MMT CO2 Eq., 10.0 kt CH4) and have continually
decreased to current levels.
85 Data from C-CAP; see https://coast.noaa.eov/dieitalcoast/tools/. Accessed August 2021.
86 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.
Land Use, Land-Use Change, and Forestry 6-131
-------
Table 6-87: Net CO2 Flux from C Stock Changes in Land Converted to Vegetated Coastal
Wetlands (MMT C02 Eq.)
Land Use/Carbon Pool
1990
2005
2016
2017
2018
2019
2020
Cropland Converted to Vegetated Coastal
Wetlands
<+)
(+)
(+)
(+)
(+)
(+)
(+)
Biomass C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Soil C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Forest Land Converted to Vegetated
Coastal Wetlands
0.49
0.50
(0.02)
(0.01)
+
0.01
0.02
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)
(0.18)
(0.17)
(0.16)
(0.15)
(0.14)
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)
(0.19)
(0.18)
(0.18)
(0.17)
(0.16)
Total Flux
0.46
0.47
(0.04)
(0.03)
(0.01)
(0.02)
(+)
+ Absolute value does not exceed 0.005 MMT C02 Eq.
Note: Totals may not sum due to independent rounding. Parentheses indicate net sequestration
Table 6-88: Net CO2 Flux from C Stock Changes in Land Converted to Vegetated Coastal
Wetlands (MMT C)
Land Use/Carbon Pool
1990
2005
2016
2017
2018
2019
2020
Cropland Converted to Vegetated Coastal
Wetlands
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Biomass C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Soil C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Forest Land Converted to Vegetated
Coastal Wetlands
0.13
0.14
(0.00)
(+)
+
+
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)
(0.05)
(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
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Soil C Stock
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Settlements Converted to Vegetated
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
6-132 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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) (0.05) (0.04)
Total Flux 0J3 0.13 (0.01) (0.01) (0.01) (+) (+)
+ Absolute value does not exceed 0.005 MMT C.
Note: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Table 6-89: ChU Emissions from Land Converted to Vegetated Coastal Wetlands (MMT CO2
Eq. and kt CH4)
Land Use/Carbon Pool
1990
2005
2016
2017
2018
2019
2020
Cropland Converted to Vegetated Coastal
Wetlands
CH4 Emissions (MMT C02 Eq.)
+
+
+
+
+
+
+
CH4 Emissions (kt CH4)
+
0.01
0.04
0.04
0.04
0.04
0.05
Forest Land Converted to Vegetated
Coastal Wetlands
CH4 Emissions (MMT C02 Eq.)
0.25
0.24
0.19
0.18
0.17
0.16
0.15
CH4 Emissions (kt CH4)
9.88
9.74
7.60
7.22
6.85
6.48
6.10
Grassland Converted to Vegetated Coastal
Wetlands
CH4 Emissions (MMT C02 Eq.)
+
+
+
+
+
+
+
CH4 Emissions (kt CH4)
0.01
0.01
0.05
0.06
0.07
0.07
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
0.37
0.40
0.43
0.47
0.50
Settlements Converted to Vegetated
Coastal Wetlands
CH4 Emissions (MMT C02 Eq.)
+
+
+
+
+
+
+
CH4 Emissions (kt CH4)
0.01
+
+
+
+
+
+
Total CH4 Emissions (MMT C02 Eq.)
0.25
0.25
0.20
0.19
0.18
0.18
0.17
Total CH4 Emissions (kt CH4)
9.98
9.91
8.05
7.72
7.39
7.06
6.73
+ Absolute value does not exceed 0.005 MMT C02 Eq. or 0.005 kt CH4.
Note: Totals may not sum due to independent rounding.
Methodology and Time-Series Consistency
The following section provides a description of the methodology used to estimate changes in biomass, dead
organic matter and soil C stocks and (Remissions for Land Converted to Vegetated Coastal Wetlands.
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2020.
Biomass Carbon Stock Changes
Biomass C stocks for Land Converted to Vegetated Coastal Wetlands are estimated for palustrine and estuarine
marshes for 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 within the U.S. Land Representation according to the national
LiDAR dataset, the national network of tide gauges and land use histories recorded in the 1996, 2001, 2005, 2011,
and 2016 NOAA C-CAP surveys (NOAA OCM 2020). Both federal and non-federal lands are represented.
Delineating Vegetated Coastal Wetlands from ephemerally flooded upland Grasslands represents a particular
challenge in remote sensing. Moreover, at the boundary between wetlands and uplands, which may be gradual on
low lying coastlines, the presence of wetlands may be ephemeral depending upon weather and climate cycles and
as such impacts on the emissions and removals will vary over these time frames. Trends in land cover change are
extrapolated to 1990 and 2020 from these datasets using the C-CAP change data closest in date to a given year.
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. Biomass is not sensitive to soil organic
Land Use, Land-Use Change, and Forestry 6-133
-------
content. Aboveground biomass carbon 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 carbon 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 (201787).
Root to shoot ratios from the Wetlands Supplement were used to account for belowground biomass, which were
multiplied by the aboveground carbon 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 carbon 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) carbon 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 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 2020 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.
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
87 See https://github.com/Smithsonian/Coastal-Wetland-NGGl-Data-Public; accessed October 2020.
6-134 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 carbon 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 2020 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 6-90 for each carbon pool and the Cm emissions. The combined
uncertainty is +/-A2.2 percent. In 2020, the total flux was 0.16 MMT CO2 Eq., with lower and upper estimates of
0.09 and 0.23 MMT C02 Eq.
Table 6-90: Approach 1 Quantitative Uncertainty Estimates for C Stock Changes occurring
within Land Converted to Vegetated Coastal Wetlands \n 2020 (MMT CO2 Eq. and Percent)
Source
2020 Estimate
Uncertainty Range Relative to Estimate3
(MMT CO? Eq.)
(MMT CO;
>Eq.)
(%)
Lower
Upper
Lower
Upper
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
(0.16)
(0.2)
(0.1)
-17.8%
17.8%
Methane Emissions
0.17
0.12
0.22
-29.9%
29.9%
Total Uncertainty
0.16
0.09
0.23
-42.2%
42.2%
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.
Land Use, Land-Use Change, and Forestry 6-135
-------
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. 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 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
No recalculations were needed for the current 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.88 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., 2023) Inventory submission.
Currently, the only coastal wetland conversion that is reported in the Inventory is Lands Converted to Vegetated
Coastal Wetlands. The next (2023) submission will include C stock change data for Lands Converted to Unvegetated
Open Water Coastal Wetlands.
Land Converted to Flooded Land
Flooded lands are defined as water bodies where human activities have 1) caused changes in the amount of
surface area covered by water, typically through water level regulation (e.g., constructing a dam), 2) waterbodies
where human activities have changed the hydrology of existing natural waterbodies thereby altering water
residence times and/or sedimentation rates, in turn causing changes to the natural production of greenhouse
gases, and 3) waterbodies that have been created by excavation, such as canals, ditches and ponds (IPCC 2019).
Flooded lands include waterbodies with seasonally variable degrees of inundation but would be expected to retain
some inundated area throughout the year under normal conditions.
Flooded lands are broadly classified as "reservoirs" or "other constructed waterbodies" (IPCC 2019). Reservoirs are
defined as flooded land greater than 8 ha and includes the seasonally flooded land on the perimeter of
permanently flooded land (i.e., inundation areas). IPCC guidance (IPCC 2019) provides default emission factors for
reservoirs and several types of "other constructed waterbodies" including freshwater ponds and canals/ditches.
Land that has been flooded for 20 years or greater is defined as Flooded Land Remaining Flooded Land and land
flooded for less than 20 years is defined as Land Converted to Flooded Land. The distinction is based on literature
reports that CO2 and CFU emissions are high immediately following flooding as labile organic matter is rapidly
degraded but decline to a steady background level approximately 20 years after flooding. Both CO2 and CFU
emissions are inventoried for Land Converted to Flooded Land.
88 See https://serc.si.edu/coastalcarbon; accessed August 2021.
6-136 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Nitrous oxide emissions from flooded lands are largely related to input of organic or inorganic nitrogen from the
watershed. These inputs from runoff/leaching/deposition are largely driven by anthropogenic activities such as
land-use change, wastewater disposal or fertilizer application in the watershed or application of fertilizer or feed in
aquaculture. These emissions are not inventoried here to avoid double-counting N2O emissions which are captured
in other source categories, such as indirect N2O emissions from managed soils (Volume 4, Chapter 11, 2006IPCC
Guidelines) and wastewater management (Volume 5, Chapter 6, 2006 IPCC Guidelines).
Emissions from Land Converted to Flooded Land-Reservoirs
Reservoirs are designed to store water for a wide range of purposes including hydropower, flood control, drinking
water, and irrigation. The permanently wetted portion of reservoirs are typically surrounded by periodically
inundated land referred to as a "drawdown zone" or "inundation area." Greenhouse gas emissions from
inundation areas are considered significant and similar per unit area to the emissions from the water surface and
are therefore included in the total reservoir surface area when estimating greenhouse gas emissions from flooded
land. Lakes converted into reservoirs without substantial changes in water surface area or water residence times
are not considered to be managed flooded land (see Area Estimates below) (IPCC 2019).
In 2020 the conterminous United States hosted 74,292 hectares of reservoir surface area in Land Converted to
Flooded Land (see Methodology and Time-Series Consistency below for calculation details) distributed across four
of the six aggregated climate zones used to define flooded land emission factors (Figure 6-15) (IPCC 2019). Alaska,
Hawaii, and U.S. Territories are not included in this report due to a lack of data (see the Methodology and Time-
Series Consistency section).
Figure 6-15: U.S. Reservoirs (black polygons) in the Land Converted to Flooded Land
Category in 2020
boreal
cool temperate
tropical dry/montane
tropical moist/wet
warm temperate dry
warm temperate moist
Note: Colors represent climate zone used to derive IPCC default emission factors.
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 6-91 and Table 6-92 below summarize nationally aggregated CFU and CO2
emissions from reservoirs and associated inundation areas in Land Converted to Flooded Land. The decrease in CO2
Land Use, Land-Use Change, and Forestry 6-137
-------
and Cm 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 15 years.
Table 6-91: ChU Emissions from Reservoirs and Inundation Areas in Land Converted to
Flooded Land (MMT CCh Eq.)
Source
1990
2005
2016
2017
2018
2019
2020
Reservoirs
2.3
0.2
0.2
0.2
0.2
0.2
0.2
Surface Emissions
2.1
0.2
0.2
0.2
0.2
0.2
0.2
Downstream Emissions
0.2
0.0
0.0
0.0
0.0
0.0
0.0
Inundation Areas
0.2
0.0
0.0
0.0
0.0
0.0
0.0
Surface Emissions
0.2
0.0
0.0
0.0
0.0
0.0
0.0
Downstream Emissions
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Total
2.5
0.2
0.2
0.2
0.2
0.2
0.2
Note: Alaska, Hawaii, and U.S. Territories are not included.
Table 6-92: ChU Emissions from Reservoirs and Inundation Areas in Land Converted to
Flooded Land (kt ChU)
Source
1990
2005
2016
2017
2018
2019
2020
Reservoirs
93
8
7
7
7
7
7
Surface Emissions
86
7
6
6
6
6
6
Downstream Emissions
8
1
1
1
1
1
1
Inundation Areas
7
0
0
0
0
0
0
Surface Emissions
6
0
0
0
0
0
0
Downstream Emissions
1
0
0
0
0
0
0
Total
100
8
7
7
7
7
7
Note: Alaska, Hawaii, and U.S. Territories are not included.
Table 6-93: CO2 Emissions from Reservoirs and Inundation Areas in Land Converted to
Flooded Land (MMT COz)
Source
1990
2005
2016
2017
2018
2019
2020
Reservoir
3.5
0.3
0.3
0.3
0.3
0.3
0.3
Inundation Area
0.3
+
+
+
+
+
+
Total
3.8
0.3
0.3
0.3
0.3
0.3
0.3
+lndicates values less than 0.05 MMT C02
Note: Alaska, Hawaii, and U.S. Territories are not included.
Table 6-94: CO2 Emissions from Reservoirs and Inundation Areas in Land Converted to
Flooded Land (MMT C)
Source
1990
2005
2016
2017
2018
2019
2020
Reservoir
1.0
0.1
0.1
0.1
0.1
0.1
0.1
Inundation Area
0.1
+
+
+
+
+
+
Total
1.0
0.1
0.1
0.1
0.1
0.1
0.1
+lndicates values less than 0.05 MMT C
Note: Alaska, Hawaii, and U.S. Territories are not included.
Methane and CO2 emissions from reservoirs and inundated areas in Minnesota were nearly 5 and 10-fold greater
than any other state, respectively (Figure 6-16 and Table 6-95). This is attributed to nine dams built in Minnesota
after 2001 which impound 58,875 ha of water, 88 percent of which is located in Mille Lacs lake. North Dakota is the
second largest source of CO2 and Cl-Ufrom reservoirs and inundated areas 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.
6-138 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure 6-16: 2020 A) ChU and B) CO2 Emissions from U.S. Reservoirs and Inundation Areas in
Land Converted to Flooded Land
Table 6-95: Methane and CO2 Emissions (kt) from Reservoirs and Associated Inundation
Areas in Land Converted to Flooded Land in 2020
State
Reservoir
Inundation Area
Total
CH
4
C02a
ch4
C02a
ch4
C02a
Downstream
Surface
Surface
Downstream Surface
Surface
Downstream
Surface
Surface
Alabama
0
0
0
0
0
0
0
0
0
Arizona
0
0
0
0
0
0
0
0
0
Arkansas
0
0
0
0
0
0
0
0
0
California
+
+
+
0
0
0
+
+
+
Colorado
+
+
+
0
0
0
+
+
+
Connecticut
+
+
+
0
0
0
+
+
+
Delaware
0
0
0
0
0
0
0
0
0
District of
0
0
0
0
0
0
0
0
0
Columbia
Florida
0
0
0
0
0
0
0
0
0
Georgia
0
0
0
0
0
0
0
0
0
Idaho
+
+
1
0
0
0
+
+
1
Illinois
+
+
6
0
0
0
+
+
6
Indiana
0
0
0
0
0
0
0
0
0
Iowa
+
+
8
0
0
0
+
+
8
Kansas
+
+
1
0
0
0
+
+
1
Kentucky
0
0
0
0
0
0
0
0
0
Louisiana
+
+
+
0
0
0
+
+
+
Maine
+
+
+
0
0
0
+
+
+
Maryland
0
0
0
0
0
0
0
0
0
Massachusetts
+
+
5
0
0
0
+
+
5
Michigan
+
+
+
0
0
0
+
+
+
Minnesota
+
5
215
+
+
5
+
5
220
Mississippi
+
+
+
0
0
0
+
+
+
Missouri
0
0
0
0
0
0
0
0
0
Montana
+
+
7
+
+
3
+
+
9
Nebraska
+
+
+
0
0
0
+
+
+
Nevada
+
+
+
0
0
0
+
+
+
New Hampshire
+
+
1
0
0
0
+
+
1
New Jersey
0
0
0
0
0
0
0
0
0
New Mexico
+
+
+
0
0
0
+
+
+
New York
+
+
+
0
0
0
+
+
+
North Carolina
+
+
1
0
0
0
+
+
1
North Dakota
+
1
23
0
0
0
+
1
23
Ohio
+
+
+
0
0
0
+
+
+
Land Use, Land-Use Change, and Forestry 6-139
-------
Oklahoma
+
+
+
0
0
0
+
+
+
Oregon
+
+
1
0
0
0
+
+
1
Pennsylvania
+
+
+
0
0
0
+
+
+
Rhode Island
0
0
0
0
0
0
0
0
0
South Carolina
0
0
0
0
0
0
0
0
0
South Dakota
+
+
+
0
0
0
+
+
+
Tennessee
+
+
3
0
0
0
+
+
3
Texas
+
+
+
0
0
0
+
+
+
Utah
+
+
+
0
0
0
+
+
+
Vermont
0
0
0
0
0
0
0
0
0
Virginia
+
+
+
0
0
0
+
+
+
Washington
+
+
+
0
0
0
+
+
+
West Virginia
0
0
0
0
0
0
0
0
0
Wisconsin
0
0
0
0
0
0
0
0
0
Wyoming
0
0
0
0
0
0
0
0
0
+ Indicates values less than 0.5 kt.
a C02: Only surface C02 emissions are included in the Inventory.
Note: Alaska, Hawaii, and U.S. Territories are not included.
Methodology and Time-Series Consistency
Estimates of Cm and CO2 emissions for reservoirs and associated inundation areas 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 6-96). Downstream Cm
emissions are calculated as 9 percent of the surface Cm 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 conterminous
United States.
Table 6-96: IPCC (2019) Default ChU and CO2 Emission Factors for Surface Emissions from
Reservoirs and Associated Inundation Areas 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
Earm Temperate Moist
0.1275
5.35
Tropical Dry/Montane
0.3923
10.82
Tropical Moist/Wet
0.2516
10.16
Area Estimates
Reservoirs in the conterminous United States were identified from the NHDArea and NHDWaterbody layers in the
National Hydrography Dataset Plus V2 (NHD),89 the National Lakes Assessment (NLA)90 data, the National
89 See https://www.uses.gov/core-science-svstems/nep/national-hvdroeraphv.
90 See https://www.epa.gov/national-aquatic-resource-surveys/nla.
6-140 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Inventory of Dams (NID),91 and the Navigable Waterways (NW)92 dataset. The NHD and NLA do not include Alaska,
Hawaii, or U.S. Territories, thus these areas are not included in the Inventory. Waterbodies less than 20 years old,
greater than 8 ha in surface area, and not identified as canal/ditch in NHD or NW and met any of the following
criteria were considered reservoirs in Land Converted to Flooded Land: 1) the water body was classified
"Reservoir" in the NHDWaterbody layer, 2) the water body name in the NHDWaterbody layer included "reservoir",
3) the water body in the NHDWaterbody layer was located in close proximity to a dam in the NID, 4) the water
body was deemed "man-made" in the NLA, 5) the waterbody was included in NW, and 6) inundation areas in the
NHDArea layer that were associated with water bodies that met any of the above criteria were assumed to
represent drawdown zones and were included in the flooded land inventory. Surface areas for identified flooded
lands were taken from NHD or the NW.
IPCC (2019) allows for the exclusion of reservoirs from the inventory if the water surface area or residence time
was not substantially changed by the construction of the dam. The guidance does not quantify what constitutes a
"substantial" change, but here EPA excludes the U.S. Great Lakes from the inventory based on expert judgment
that neither the surface area or water residence time was substantially altered by their associated dams. EPA
assumes that all other waterbodies identified through the six criteria described above were substantially impacted
by the construction of dams.
EPA assumes that all reservoirs included in the NW are subject to water-level management to maintain minimum
water depths required for navigation and are therefore managed flooded lands. Reservoir age was determined
from the year the dam was completed as reported in the NID (available for 40,012 out of 54,670 reservoirs). When
dam completion year was not available, the reservoir was assumed to be greater than 20 years old. Reservoirs
were disaggregated by state (using boundaries from the 2016 U.S. Census Bureau93) and climate zone.
Downstream and surface emissions for cross-state reservoirs were allocated to states based on the surface area
that the reservoir occupied in each state. Only the U.S. portion of reservoirs that cross country borders were
included in the inventory. Surface areas for reservoirs and associated inundation areas were taken from NHD or
the NW and the final inventory of reservoirs and associated inundation areas was screened to ensure no
waterbodies were duplicated.
Many reservoirs are surrounded by land that is dry for a portion of the year but is periodically flooded when water
inflows to the reservoir exceed outflows and the reservoir surface area expands into surrounding lands. This can
occur for a variety of reasons including high rates of water runoff from the watershed (i.e., snow melt, large
precipitation events), deliberate efforts to raise water levels for seasonal recreation or wildlife habitat, and
management efforts to reduce inflows to downstream systems. These periodically flooded lands are represented
as "Inundation Areas" in the NHDArea layer (Figure 6-17). Inundation areas are considered equivalent to
"drawdown zones" in IPCC (2019) and CO2 and CH4 emissions from these lands are estimated using the same
methodology as for reservoirs.
91 See https://nid.sec.usace.armv.mil.
92 See https://hifld-geoplatform.opendata.arcgis.com/datasets/eeoplatform::navieable-waterwav-network-lines-l/about.
93 See https://www.census.eov/eeoeraphies/mappine-files/time-series/eeo/carto-boundary-file.htrnl.
Land Use, Land-Use Change, and Forestry 6-141
-------
Figure 6-17: Example of a Reservoir and Associated Inundation Area in Land Converted to
Flooded Land
I Reservoir
I Inundation Areas Esri Canada, Esri, HERE, Garmin, SafeGraph, METI/NASA, USGS, Bureau of
Land Management EPA, NPS, USDA
The surface area of reservoirs and inundation areas in Land Converted to Flooded Land decreased by
approximately 90 percent from 1990 to 2020 (Table 6-97). This is due to reservoirs that were less than 20 years old
at beginning of time series entering the Flooded Land Remaining Flooded Land category when they reached 20
years of age. The rate at which flooded land has aged out of the Land Converted to Flooded Land category has
outpaced the rate of new dam construction. New dam construction has slowed considerably during the time series
with no new dams entering the inventory since 2017 (Figure 6-18).
Lakes Sakakawea and Oahe in North Dakota and South Dakota are notable examples of reservoirs that matriculated
out of Land Converted to Flooded Land during the time series. These Missouri River impoundments have a combined
surface area in excess of 0.25 million hectares and aged out of Land Converted to Flooded Land between 2000 and
2003, but in 2020 North Dakota still had the second largest expanse of surface area in this category due primarily to
a new dam on Devils Lake.
Table 6-97: National Totals of Reservoir and Associated Inundation Areas Surface Area
(thousands of ha) in Land Converted to Flooded Land
Surface Area (thousands of ha)
1990 2005 2016
2017
2018
2019
2020
Reservoir
699 1 75 I 74
73
73
72
72
Inundation Area
51 | 3 | 2
2
2
2
2
Note: Alaska, Hawaii, and U.S. Territories are not included.
6-142 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Figure 6-18: Number of dams built per year from 1990-2020
30-
if)
ro 20-
~o
5
CD
z
10-
o-
Table 6-98: State breakdown of Reservoirs and Associated Inundation Area Surface Area
(thousands of ha) in Land Converted to Flooded Land
State
1990
2005
2016
2017
2018
2019
2020
Alabama
32.3
0.1
0.0
0.0
0.0
0.0
0.0
Arizona
0.1
0.1
0.0
0.0
0.0
0.0
0.0
Arkansas
32.1
0.0
0.0
0.0
0.0
0.0
0.0
California
10.7
0.5
0.4
0.4
0.4
0.1
0.0
Colorado
5.8
0.1
0.1
0.1
0.1
0.1
0.1
Connecticut
0.1
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
10.4
0.2
0.0
0.0
0.0
0.0
0.0
Georgia
10.2
0.0
0.0
0.0
0.0
0.0
0.0
Idaho
10.1
0.8
0.4
0.4
0.4
0.4
0.4
Illinois
19.6
3.1
1.3
1.2
1.2
1.1
1.1
Indiana
6.2
2.3
0.0
0.0
0.0
0.0
0.0
Iowa
11.4
1.9
2.1
2.1
2.1
2.1
2.1
Kansas
26.7
0.4
0.3
0.3
0.2
0.2
0.2
Kentucky
0.6
0.0
0.0
0.0
0.0
0.0
0.0
Louisiana
16.9
3.0
0.0
0.0
0.0
0.0
0.0
Maine
13.2
3.4
0.1
0.1
0.1
0.1
0.1
Maryland
0.3
0.0
0.0
0.0
0.0
0.0
0.0
Massachusetts
1.0
0.6
1.3
1.3
1.3
1.2
1.2
Michigan
12.8
2.7
0.1
0.1
0.1
0.1
0.1
Minnesota
9.0
9.1
59.7
58.9
58.9
58.9
58.9
Mississippi
6.0
0.1
0.0
0.0
0.0
0.0
0.0
Missouri
56.7
0.0
0.0
0.0
0.0
0.0
0.0
Montana
8.5
3.8
2.5
2.5
2.5
2.5
2.5
Nebraska
5.5
1.3
0.2
0.2
0.2
0.2
0.1
Nevada
1.3
0.3
0.0
0.0
0.0
0.0
0.0
New Hampshire
0.4
0.2
0.2
0.2
0.2
0.2
0.2
New Jersey
0.3
0.0
0.0
0.0
0.0
0.0
0.0
New Mexico
0.2
0.0
0.0
0.0
0.0
0.0
0.0
New York
2.4
0.9
0.1
0.1
0.1
0.1
0.1
North Carolina
19.2
0.5
0.1
0.1
0.1
0.1
0.1
North Dakota
159.4
2.5
6.7
6.2
6.2
6.2
6.2
Land Use, Land-Use Change, and Forestry 6-143
-------
Ohio
3.1
0.0
0.0
0.0
0.0
0.0
0.0
Oklahoma
19.6
0.2
0.0
0.0
0.0
0.0
0.0
Oregon
3.9
0.2
0.2
0.2
0.1
0.1
0.1
Pennsylvania
3.2
0.3
0.3
0.3
0.1
0.1
0.1
Rhode Island
0.2
0.0
0.0
0.0
0.0
0.0
0.0
South Carolina
6.0
0.0
0.0
0.0
0.0
0.0
0.0
South Dakota
106.5
0.0
0.0
0.0
0.0
0.0
0.0
Tennessee
2.7
0.0
0.6
0.6
0.6
0.6
0.6
Texas
49.5
0.1
0.0
0.0
0.0
0.0
0.0
Utah
43.0
37.0
0.0
0.0
0.0
0.0
0.0
Vermont
0.1
0.1
0.0
0.0
0.0
0.0
0.0
Virginia
5.8
0.0
0.0
0.0
0.0
0.0
0.0
Washington
3.0
0.2
0.0
0.0
0.0
0.0
0.0
West Virginia
2.9
1.6
0.0
0.0
0.0
0.0
0.0
Wisconsin
1.9
0.4
0.0
0.0
0.0
0.0
0.0
Wyoming
9.1
0.3
0.0
0.0
0.0
0.0
0.0
Total
750.0
78.5
76.7
75.3
75.0
74.5
74.3
Note: Alaska, Hawaii,
and U.S. Territories are
not included.
Uncertainty
Uncertainty in estimates of CH4 and CO2 emissions from reservoirs on Land Converted to Flooded Land were
developed using IPCC Approach 2 and include uncertainty in the default emission factors and the flooded land area
inventory (Table 6-99). Uncertainty in emission factors is provided in the 2019 Refinement to the 2006 IPCC
Guidelines (IPCC 2019). Uncertainties in the spatial data include 1) uncertainty in area estimates from the NHD and
NW, and 2) uncertainty in the location of dams in the NID. Overall uncertainties in these spatial datasets are
unknown, but uncertainty for remote sensing products is assumed to be ±10 to 15 percent based on IPCC guidance
(IPCC 2003). An uncertainty range of ±15 percent for the flooded land area estimates is assumed and is based on
expert judgment.
Table 6-99: Approach 2 Quantitative Uncertainty Estimates for ChU and CO2 Emissions from
Reservoirs and Associated Inundation Areas in Land Converted to Flooded Land
2020 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
0.2
0.1
0.2
-12.7%
12.5%
Surface
C02
0.3
0.2
0.3
-13.4%
13.5%
Downstream
ch4
+
+
0.1
-57.1%
287.3
Inundation Area
Surface
ch4
+
+
+
-13.1%
12.9%
Surface
co2
+
+
+
-13.0%
14.1%
Downstream
ch4
+
+
+
-56.9%
290.7%
Total
0.5
0.4
0.5
-13.1%
16.0%
+ 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.
QA/QC and Verification
The National Hydrography Data (NHD) is managed by the USGS in collaboration many other federal, state, and
local entities. Extensive QA/QC procedures are incorporated into the curation of the NHD. The National Inventory
of Dams (NID) is maintained by the U.S. Army Corps of Engineers (USACE) in collaboration with the Federal
Emergency Management Agency (FEMA) and state regulatory offices. USACE resolves duplicative and conflicting
data from 68 data sources, which helps obtain the more complete, accurate, and updated NID. The National Lakes
6-144 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Assessment is a survey of U.S. lakes and reservoirs conducted by the U.S. Environmental Protection Agency every 5
years. The program is subject to rigorous QA/QC as detailed in the Quality Assurance Project Plan.94 The Navigable
Waterways (NW) dataset is part of the U.S. Department of Transportation (USDOT)/Bureau of Transportation
Statistics's (BTS's) National Transportation Atlas Database (NTAD). The NW is a comprehensive network database
of the nation's navigable waterways updated on a continuing basis.
All calculations were executed independently in Excel and R. Ten percent of state and national totals were
randomly selected for comparison between the two approaches to ensure there were no computational errors.
Recalculations Discussion
This is a new category in the current Inventory.
Planned Improvements
EPA is measuring greenhouse gas emissions from 108 flooded lands (reservoirs) in the conterminous United States.
The survey will be complete by September 2023 and the data will be used to predict greenhouse gas emission
rates for all U.S. flooded lands. The Inventory will be updated at that time using these country-specific emission
factors developed through the measurement campaign.
Hawaii, Alaska and U.S. Territories will be included in the next (i.e., 1990 through 2021) Inventory. Flooded lands
area data for these states and territories will be derived from the National Hydrography Dataset Plus High
Resolution (NHDPIus HighRes),95 an enhanced version of the NHD used in this Inventory.
To verify that waterbodies contained in NW are subject to water level management, EPA will overlay the NW with
other spatial datasets of water control structures including the inventory of U.S. Army Corps of Engineers locks for
water navigation96 and dams/weirs contained in the NHDPIus HighRes.
Emissions from Land Converted to Flooded Land-Other
Constructed Waterbodies
Freshwater ponds are the only type of flooded lands within the "other constructed waterbodies" subcategory of
Land Converted to Flooded Land that are included in this Inventory (see Methodology for details). IPCC (2019)
describes ponds as waterbodies that are "...constructed by excavation and/or construction of walls to hold water in
the landscape for a range of uses, including agricultural water storage, access to water for livestock, recreation,
and aquaculture." The IPCC "Decision tree for types of Flooded Land" (IPCC 2019, Fig. 7.2) elaborates on this
description by defining waterbodies less than 8 ha as a subset of "other constructed waterbodies." For this
inventory, ponds are defined as managed flooded land not flagged as "canal/ditch" in the NHD (see Methods
below) with surface area less than 8 ha. IPCC (2019) further distinguishes saline versus brackish ponds, with the
former supporting lower CH4 emission rates than the latter. Activity data on pond salinity is not uniformly available
for the conterminous United States and all ponds in Land Converted to Flooded Land are assumed to be
freshwater. Ponds often receive high organic matter and nutrient loadings, may have low oxygen levels, and are
sites of substantial CH4 and CO2 emissions from anaerobic sediments.
Methane and CO2 emissions from freshwater ponds decreased 98 percent from 1990 to 2020 due to flooded land
matriculating from Land Converted to Flooded Land to Flooded Land Remaining Flooded Land. Much of this decline
occurred by 2000, but declines have continued through 2020. In 2020, Massachusetts, Mississippi, and Kansas
94 See https://www.epa.gov/national-aquatic-resource-survevs/national-lakes-assessment-2017-qualitv-assurance-proiect-
plan.
95 See https://www.uses.eov/core-science-svstems/nep/national-hvdroeraphv/nhdplus-hieh-resolution.
96 See https://hifld-eeoplatform.opendata.arceis.com/datasets/eeoplatform::locks/about.
Land Use, Land-Use Change, and Forestry 6-145
-------
have the greatest CO2 and CH4 emissions for freshwater ponds in Land Converted to Flooded Land (Table 6-100
through Table 6-104, Figure 6-19).
Table 6-100: ChU Emissions from Other Constructed Waterbodies in Land Converted to
Flooded Land (MMT CCh Eq.)
Source
1990
2005
2016
2017
2018
2019
2020
Freshwater Ponds
0.1
+
+
+
+
+
+
+ Indicates values less than 0.05 MMT C02 Eq.
Note: Alaska, Hawaii, and U.S. Territories are not included.
Table 6-101: ChU Emissions from Other Constructed Waterbodies in Land Converted to
Flooded Land (kt Cm)
Source
1990
2005
2016
2017
2018
2019
2020
Freshwater Ponds
3
+
+
+
+
+
+
+ Indicates values less than 0.5 kt
Note: Alaska, Hawaii, and U.S. Territories are not included.
Table 6-102: CO2 Emissions from Other Constructed Waterbodies in Land Converted to
Flooded Land (MMT COz)
Source
1990
2005
2016
2017
2018
2019
2020
Freshwater Ponds
0.1
+
+
+
+
+
+
+ Indicates values less than 0.5 MMT C02 Eq.
Note: Alaska, Hawaii, and U.S. Territories are not included.
Table 6-103: CO2 Emissions from Other Constructed Waterbodies in Land Converted to
Flooded Land (MMT C)
Source
1990
2005
2016
2017
2018
2019
2020
Freshwater Ponds
0.02
+
+
+
+
+
+
+ Indicates values less than 0.005 MMT C
Note: Alaska, Hawaii, and U.S. Territories are not included.
Table 6-104: ChU and CO2 Emissions (MT CO2 Eq.) from Other Constructed Waterbodies in
Land Converted to Flooded Land in 2020
Freshwater Ponds
State
ch4
CO?
Total
Alabama
0
0
0
Arizona
0
0
0
Arkansas
0
0
0
California
18
25
43
Colorado
82
65
147
Connecticut
0
0
0
Delaware
0
0
0
District of Columbia
0
0
0
Florida
14
30
44
Georgia
43
96
139
Idaho
0
0
0
Illinois
26
31
57
Indiana
0
0
0
Iowa
8
10
18
Kansas
160
187
347
Kentucky
0
0
0
Louisiana
0
0
0
Maine
11
9
19
Maryland
26
31
57
6-146 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Massachusetts
229
239
468
Michigan
0
0
0
Minnesota
75
61
137
Mississippi
146
288
434
Missouri
33
38
71
Montana
37
30
68
Nebraska
43
39
82
Nevada
0
0
0
New Hampshire
41
34
75
New Jersey
0
0
0
New Mexico
29
24
53
New York
85
72
157
North Carolina
56
65
121
North Dakota
43
35
78
Ohio
14
16
30
Oklahoma
115
135
250
Oregon
9
7
16
Pennsylvania
21
20
40
Rhode Island
0
0
0
South Carolina
28
33
60
South Dakota
78
64
142
Tennessee
0
0
0
Texas
32
75
107
Utah
60
49
109
Vermont
0
0
0
Virginia
0
0
0
Washington
5
5
10
West Virginia
0
0
0
Wisconsin
30
25
55
Wyoming
24
19
43
Total
1,619
1,857
3,477
Note: Alaska, Hawaii, and U.S. Territories are not included.
Figure 6-19: ChU and CO2 Emissions (MT CO2 Eq.) from Other Constructed Waterbodies in
Land Converted to Flooded Land in 2020
A. C02 Emissions from Freshwater Ponds B. CH4 Emissions from Freshwater Ponds
Methodology and Time-Series Consistency
Estimates of CFU 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 6-105). 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
Land Use, Land-Use Change, and Forestry 6-147
-------
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 6-105: 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
Area estimates
Freshwater ponds in the conterminous United States were identified from the NHDArea and NHDWaterbody layers
in the National Hydrography Dataset Plus V2 (NHD),97 the National Lakes Assessment (NLA)98 data, the National
Inventory of Dams (NID),99 and the Navigable Waterways (NW)100 dataset. The NHD and NLA do not include
Alaska, Hawaii, or U.S. Territories, thus these areas are not included in the Inventory.
Waterbodies less than 20 years old, less than 8 Ha in surface area, and not identified as canal/ditch in NHD or NW
and met any of the following criteria were considered ponds in Land Converted to Flooded Land: 1) the water body
was classified "Reservoir" in the NHDWaterbody layer, 2) the water body name in the NHDWaterbody layer
included "reservoir", 3) the water body in the NHDWaterbody layer was located in close proximity to a dam in the
NID, 4) the water body was deemed "man-made" in the NLA, 5) the waterbody was included in NW, and 6)
inundation areas in the NHDArea layer that were associated with water bodies that met any of the above criteria
were assumed to represent drawdown zones and were included in the flooded land inventory. Flooded lands that
met any one of these criteria and 1) had a surface area less than 8 ha and 2) were not classified as
CANALS/DITCHES in the NHD, were classified as freshwater ponds, a subcategory of other constructed waterbodies
(IPCC 2019).
Surface areas for ponds were taken from NHD or the NW. Waterbodies were further disaggregated by state (using
boundaries from the 2016 U.S. Census Bureau101) and the final area inventory was screened to ensure no
waterbodies were duplicated.
While the distribution of U.S. waterbodies <8 ha is well represented in NHD, it is difficult to determine which of
these waterbodies are subject to water level management. The presence or absence of a flow control structure
associated with these small waterbodies is typically not documented in NHD, thus EPA used the NID for this
purpose. The NID only includes dams that pose a hazard if they were to fail, equal or exceed 25 feet in height and
exceed 15 acre-feet in storage, or equal or exceed 50 acre-feet storage and exceed 6 feet in height.102 The extent
97 See https://www.uses.gov/core-science-svstems/nep/national-hvdroeraphv.
98 See https://www.epa.gov/national-aquatic-resource-surveys/nla.
99 See https://nid.sec.usace.army.mil.
100 See https://hifld-eeoplatform.opendata.arceis.com/datasets/eeoplatform::navieable-waterwav-network-lines-l/about.
101 See https://www.census.gov/geographies/mapping-files/time-series/geo/carto-boundarv-file.htrnl.
102 See https://nid.sec.usace.army.mil.
6-148 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
to which these criteria fail to capture flow control structures associated with freshwater ponds in the United States
in unknown, but the freshwater pond area inventory presented here is an underestimate. There is a planned
improvement to review other data sources or approaches that could more fully capture all managed freshwater
ponds in the United States.
IPCC (2019) provides guidance for estimating Cm emissions from canals and ditches, a subcategory of other
constructed waterbodies, in Land Converted to Flooded Land. While U.S. canals and ditches can be identified in the
data sources described above, the age of these systems cannot. EPA assumes that all U.S. canals and ditches are
greater than 20 years old and therefore are not included in Land Converted to Flooded Land.
For the year 2020, this Inventory contains 354 ha of freshwater ponds in Land Converted to Flooded Land. The
surface area of freshwater ponds decreased by 98 percent from 1990 to 2020 due to flooded lands aging out of
Land Converted to Flooded Land more quickly than new flooded lands entered the category. The greatest
reduction in freshwater pond surface area occurred in Texas, Missouri, and Kansas (Table 6-107). Freshwater
ponds in the 2020 inventory are most abundant in Massachusetts, Kansas, and Mississippi, but show no
overarching geographical pattern (Figure 6-20).
Table 6-106: National Surface Area (ha) Totals of Other Constructed Waterbodies in Land
Converted to Flooded Land
Other Constructed Waterbody
1990 ¦
2005
2016
2017
2018
2019
2020
Freshwater Ponds
14,846
1015
590
542
442
393
354
Note: Alaska, Hawaii, and U.S. Territories are not included.
Figure 6-20: Surface Area (ha) of Other Constructed Waterbodies in Land Converted to
Flooded Land
Table 6-107: State Surface Area (ha) Totals of Other Constructed Waterbodies in Land
Converted to Flooded Land
State
1990
2005
2016
2017
2018
2019
2020
Alabama
240
11
3
3
0
0
0
Arizona
29
5
5
5
0
0
0
Arkansas
255
0
0
0
0
0
0
California
98
16
5
5
5
4
4
Colorado
140
27
27
27
18
18
18
Connecticut
46
0
0
0
0
0
0
Delaware
7
0
0
0
0
0
0
District of Columbia
0
0
0
0
0
0
0
Land Use, Land-Use Change, and Forestry 6-149
-------
Florida
46
16
6
6
6
3
3
Georgia
611
17
9
9
9
9
9
Idaho
44
0
0
0
0
0
0
Illinois
333
38
12
11
6
6
6
Indiana
81
7
0
0
0
0
0
Iowa
1,062
33
13
13
13
9
2
Kansas
1,636
74
65
58
42
35
35
Kentucky
184
2
2
0
0
0
0
Louisiana
95
5
0
0
0
0
0
Maine
23
4
2
2
2
2
2
Maryland
38
6
6
6
6
6
6
Massachusetts
70
65
72
67
61
57
50
Michigan
99
13
2
0
0
0
0
Minnesota
96
25
30
30
30
30
16
Mississippi
421
70
59
44
44
39
32
Missouri
2,375
23
7
7
7
7
7
Montana
298
5
13
13
13
8
8
Nebraska
751
44
35
35
9
9
9
Nevada
7
20
2
0
0
0
0
New Hampshire
62
20
9
9
9
9
9
New Jersey
15
3
3
3
0
0
0
New Mexico
9
0
6
6
6
6
6
New York
193
70
19
19
19
19
19
North Carolina
225
41
14
14
14
14
12
North Dakota
66
41
17
17
9
9
9
Ohio
202
38
13
3
3
3
3
Oklahoma
1,012
52
35
35
32
25
25
Oregon
81
18
5
5
2
2
2
Pennsylvania
137
20
5
5
5
5
5
Rhode Island
1
0
0
0
0
0
0
South Carolina
462
26
11
9
9
9
6
South Dakota
112
48
28
26
26
17
17
Tennessee
198
16
9
9
2
0
0
Texas
2,428
45
17
12
8
7
7
Utah
45
0
8
13
13
13
13
Vermont
61
4
0
0
0
0
0
Virginia
171
0
0
0
0
0
0
Washington
98
33
3
3
1
1
1
West Virginia
0
0
0
0
0
0
0
Wisconsin
154
11
7
7
7
7
7
Wyoming
28
3
5
5
5
5
5
TOTAL
14,846
1,015
590
542
442
393
354
Note: Alaska, Hawaii, and U.S. Territories are not included.
Uncertainty
Uncertainty in estimates of CO2 and Cmemissions from Land Converted to Flooded Land -Other Constructed Water
Bodies include uncertainty in the default emission factors and the flooded land area inventory. Uncertainty in
emission factors is provided in the 2019 Refinement to the 2006IPCC Guidelines (IPCC 2019). Uncertainties in the
spatial data include 1) uncertainty in area estimates from the NHD and NW, and 2) uncertainty in the location of
dams in the NID. Overall uncertainties in the NHD, NID, and NW are unknown, but uncertainty for remote sensing
products is ±10 to 15 percent (IPCC 2003). EPA assumes an uncertainty of ±15 percent for the flooded land area
inventory based on expert judgment. These uncertainties do not include the underestimate of pond surface area
discussed above.
6-150 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 6-108: Approach 2 Quantitative Uncertainty Estimates for ChU and CO2 Emissions from
Other Constructed Waterbodies in Land Converted to Flooded Land
Source
Gas
2020 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
2
2 2
-4.5%
5.0%
Freshwater ponds
C02
2
2 2
-4.1%
3.6%
Total
3
3 4
-3.6%
4.1%
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.
QA/QC and Verification
The National Hydrography Data (NHD) is managed by the USGS with collaboration from many other federal, state,
and local entities. Extensive QA/QC procedures are incorporated into the curation of the NHD. The National
Inventory of Dams (NID) is maintained by the U.S. Army Corps of Engineers (USACE) in collaboration with the
Federal Emergency Management Agency (FEMA) and state regulatory offices. USACE resolves duplicative and
conflicting data from 68 data sources, which helps obtain the more complete, accurate, and updated NID. The
National Lakes Assessment is a survey of U.S. lakes and reservoirs conducted by the U.S. Environmental Protection
Agency every 5 years. The program is subject to rigorous QA/QC as detailed in the Quality Assurance Project
Plan.103 The Navigable Waterways (NW) dataset is part of the U.S. Department of Transportation (USDOT)/Bureau
of Transportation Statistics1 (BTS's) National Transportation Atlas Database (NTAD). The NW is a comprehensive
network database of the nation's navigable waterways updated on a continuing basis.
All calculations were executed independently in Excel and R. Ten percent of state and national totals were
randomly selected for comparison between the two approaches to ensure there were no computational errors.
Recalculations Discussion
This is a new category in the current Inventory.
Planned Improvements
The distribution of U.S. waterbodies less than 8 ha is well known, but the presence or absence of water level
control structures associated with these small waterbodies is not well documented in national data sources. To
improve the representation of managed ponds in future inventories, EPA will attempt to locate state or county
records on small dam construction permits and/or inspection records to supplement records in the NID. EPA will
also use surrounding land use as a proxy for management. For example, a pond surrounded by cultivated land is
likely subject to water level management and should be included in the inventory. Even if the pond were not
subject to water level management, greenhouse gas emissions from the system are likely enhanced by elevated
nutrient and sediment inputs from the surrounding managed lands, thus the emissions should be considered
anthropogenic and included in the inventory.
Hawaii, Alaska, and U.S. Territories will be included in the next (i.e., 1990 through 2021) Inventory. Flooded lands
area data for these states and territories will be derived from the National Hydrography Dataset Plus High
Resolution (NHDPIus HighRes),104 an enhanced version of the NHD used in this Inventory.
103 See https://www.epa.gov/national-aquatic-resource-surveys/national-lakes-assessment-2017-qualitv-assurance-proiect-
plan.
104 See https://www.usgs.eov/core-science-svstems/ngp/national-hvdrographv/nhdplus-high-resolution.
Land Use, Land-Use Change, and Forestry 6-151
-------
6.10 Settlements Remaining Settlements
(CRF Category 4E1)
Soil Carbon Stock Changes (CRF Category 4E1)
Soil organic C stock changes for Settlements Remaining Settlements occur in both mineral and organic soils.
However, the United States does not estimate changes in soil organic C stocks for mineral soils in Settlements
Remaining Settlements. This approach is consistent with the assumption of the Tier 1 method in the 2006IPCC
Guidelines (IPCC 2006) that inputs equal outputs, and therefore the soil organic C stocks do not change. This
assumption may be re-evaluated in the future if funding and resources are available to conduct an analysis of soil
organic C stock changes for mineral soils in Settlements Remaining Settlements.
Drainage of organic soils is common when wetland areas have been developed for settlements. Organic soils, also
referred to as Histosols, include all soils with more than 12 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. Drainage of organic
soils leads to aeration of the soil that accelerates decomposition rate and CO2 emissions.105 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).
Settlements Remaining Settlements includes all areas that have been settlements for a continuous time period of
at least 20 years according to the 2015 United States Department of Agriculture (USDA) National Resources
Inventory (NRI) (USDA-NRCS 2018)106 or according to the National Land Cover Dataset (NLCD) for federal lands
(Yang et al. 2018; Fry et al. 2011; Homer et al. 2007, 2015). The Inventory includes settlements on privately-owned
lands in the conterminous United States and Hawaii. Alaska and the small amount of settlements on federal lands
are not included in this Inventory even though these areas are part of the U.S. managed land base. This leads to a
discrepancy with the total amount of managed area in Settlements Remaining Settlements (see Section 6.1
Representation of the U.S. Land Base) and the settlements area included in the Inventory analysis. 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.
CO2 emissions from drained organic soils in settlements are 15.9 MMT CO2 Eq. (4.3 MMT C) in 2020 (See Table
6-109 and Table 6-110). Although the flux is relatively small, the amount has increased by over 40 percent since
1990 due to an increase in area of drained organic soils in settlements.
Table 6-109: Net CO2 Flux from Soil C Stock Changes in Settlements Remaining Settlements
(MMT COz Eq.)
Soil Type 1990 2005 2016 2017 2018 2019 2020
Organic Soils 11.3 12.2 16.0 16.0 15.9 15.9 15.9
105 N20 emissions from soils are included in the N20 Emissions from Settlement Soils section.
106 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.
6-152 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 6-110: Net CO2 Flux from Soil C Stock Changes in Settlements Remaining Settlements
(MMT C)
Soil Type
1990
2005
2016
2017
2018
2019
2020
Organic Soils
3.1
3.3
4.4
4.4
4.3
4.3
4.3
Methodology and Time-Series Consistency
An IPCC Tier 2 method is used to estimate soil organic C stock changes for organic soils in Settlements Remaining
Settlements (IPCC 2006). Organic soils in Settlements Remaining Settlements are assumed to be losing C at a rate
similar to croplands due to deep drainage, and therefore emission rates are based on country-specific values for
cropland (Ogle et al. 2003).
The land area designated as settlements is based primarily on the 2018 NRI (USDA-NRCS 2018) with additional
information from the NLCD (Yang et al. 2018; Fry et al. 2011; Homer et al. 2007, 2015). It is assumed that all
settlement area on organic soils is drained, and those areas are provided in Table 6-111 (See Section 6.1,
Representation of the U.S. Land Base for more information). The area of drained organic soils is estimated from
the NRI spatial weights and aggregated to the country (Table 6-111). The area of land on organic soils in
Settlements Remaining Settlements has increased from 220 thousand hectares in 1990 to over 303 thousand
hectares in 2015. The area of land on organic soils are not currently available from NRI for Settlements Remaining
Settlements after 2015.
Table 6-111: Thousands of Hectares of Drained Organic Soils in Settlements Remaining
Settlements
Year
Area
(Thousand Hectares)
1990
220
2005
235
2014
291
2015
303
2016
ND
2017
ND
2018
ND
2019
ND
2020
ND
Note: No NRI data are available after 2015,
designated as ND (No data).
To estimate CO2 emissions from drained organic soils across the time series from 1990 to 2015, the total area of
organic soils in Settlements Remaining Settlements is multiplied by the country-specific emission factors for
Cropland Remaining Cropland under the assumption that there is deep drainage of the soils. The emission factors
are 11.2 MT C per ha in cool temperate 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 for more information).
In order to ensure time-series consistency, the same methods are applied from 1990 to 2015, and a linear
extrapolation method is used to approximate emissions for the remainder of the 2016 to 2020 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 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 in future Inventories to recalculate the estimates beyond 2015 as activity data become
available.
Land Use, Land-Use Change, and Forestry 6-153
-------
Uncertainty
Uncertainty for the Tier 2 approach is derived using a Monte Carlo approach, along with additional uncertainty
propagated through the Monte Carlo Analysis for 2016 to 2020 based on the linear time series model. The results
of the Approach 2 Monte Carlo uncertainty analysis are summarized in Table 6-112. Soil C losses from drained
organic soils in Settlements Remaining Settlements for 2020 are estimated to be between 7.4 and 24.4 MMT CO2
Eq. at a 95 percent confidence level. This indicates a range of 53 percent below and 53 percent above the 2020
emission estimate of 15.9 MMT CO2 Eq.
Table 6-112: Uncertainty Estimates for CO2 Emissions from Drained Organic Soils in
Settlements Remaining Settlements (MMT CO2 Eq. and Percent)
2020 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
7.4 24.4
-53% 53%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence
interval.
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. No errors were found in this Inventory.
Recalculations Discussion
There were no recalculations to the 1990 through 2019 time series in this Inventory.
Planned Improvements
This source will be updated to include CO2 emissions from drainage of organic soils in settlements of Alaska and
federal lands in order to provide a complete inventory of emissions for this category. See Table 6-113 for the
amount of managed land area in Settlements Remaining Settlements that is not included in the Inventory due to
these omissions. The managed settlements area that is not included in the Inventory is in the range of 150 to 160
thousand hectares each year. These improvements will be made as funding and resources are available to expand
the inventory for this source category.
Table 6-113: Area of Managed Land in Settlements Remaining Settlements that is not
included in the current Inventory (Thousand Hectares)
Area (Thousand Hectares)
Year
SRS Managed Land
Area (Section 6.1)
SRS Area Included
in Inventory
SRS Area Not
in Invenl
1990
30,585
30,425
159
1991
30,589
30,430
159
1992
30,593
30,434
159
1993
30,505
30,346
159
1994
30,423
30,264
159
1995
30,365
30,206
159
1996
30,316
30,157
158
1997
30,264
30,105
158
1998
30,200
30,041
159
6-154 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
30,144
30,101
30,041
30,034
30,530
31,011
31,522
31,964
32,469
33,074
33,646
34,221
34,814
35,367
36,308
37,281
38,210
ND
ND
ND
ND
ND
29,992
29,949
29,889
29,882
30,378
30,859
31,370
31,812
32,317
32,922
33,494
34,069
34,662
35,215
36,156
37,129
38,058
ND
ND
ND
ND
ND
152
152
152
152
152
152
152
152
152
152
152
152
152
152
152
152
152
ND
ND
ND
ND
ND
Note: NRI data are not available after 2015, and these years are designated as ND (No
data).
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. Trees in
settlement areas of the United States are estimated to account for an average annual net sequestration of 116.3
MMT CO2 Eq. (31.7 MMT C) over the period from 1990 through 2020. Net C sequestration from settlement trees in
2020 is estimated to be 129.8 MMT CO2 Eq. (35.4 MMT C) (Table 6-114). Dominant factors affecting carbon 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 35 percent between
1990 and 2020 due to increases in settlement area and changes in tree cover.
Trees in settlements often grow faster than forest trees because of their relatively open structure (Nowak and
Crane 2002). Because tree density in settlements is typically much lower than in forested areas, the C storage per
hectare of land is in fact smaller for settlement areas than for forest areas. Also, percent tree cover in settlement
areas are less than in forests and this tree cover varies significantly across the United States (e.g., Nowak and
Greenfield 2018a). To quantify the C stored in settlement trees, the methodology used here requires analysis per
unit area of tree cover, rather than per unit of total land area (as is done for Forest Lands).
Table 6-114: Net Flux from Trees in Settlements Remaining Settlements (WAT CO2 Eq. and
MMT C)a
Year
MMT C02 Eq.
MMT C
1990
(96.4)
(26.3)
2005
(117.4)
(32.0)
Land Use, Land-Use Change, and Forestry 6-155
-------
2016 (129.8) (35.4)
2017 (129.8) (35.4)
2018 (129.8) (35.4)
2019 (129.8) (35.4)
202 0 (129.8) (35.4)
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.
Methodology and Time-Series Consistency
To estimate net carbon sequestration in settlement areas, three types of data are required for each state:
1. Settlement area
2. Percent tree cover in settlement areas
3. Carbon sequestration density per unit of tree cover
Settlement Area
Settlements area is defined in Section 6.1 Representation of the U.S. Land Base as a land-use category representing
developed areas. The data used to estimate settlement area within Section 6.1 comes from the NRI as updated
through 2015 with the extension of the time series through 2018 based on assuming the settlements area is the
same as 2015, while harmonizing these data with the FIA dataset, which are available through 2018, and the NLCD
dataset, which is available through 2016. Settlement areas for 2020 are held constant with the 2018 values. This
process of combining the datasets extends the time series to ensure that there is a complete and consistent
representation of land use data for all source categories in the LULUCF sector. Annual estimates of CO2 flux (Table
6-114) were developed based on estimates of annual settlement area and tree cover derived from NLCD
developed lands. Developed land, which was used to estimate tree cover in settlement areas, is about six percent
higher than the area categorized as Settlements in the Representation of the U.S. Land Base developed for this
report.
Percent Tree Cover in Settlement Areas
Percent tree cover in settlement area by state is needed to convert settlement land area to settlement tree cover
area. Converting to tree cover area is essential as tree cover, and thus carbon estimates, can vary widely among
states in settlement areas due to variations in the amount of tree cover (e.g., Nowak and Greenfield 2018a).
However, since the specific geography of settlement area is unknown because they are based on NRI sampling
methods, NLCD developed land was used to estimate the percent tree cover to be used in settlement areas. NLCD
developed classes 21-24 (developed, open space (21), low intensity (22), medium intensity (23), and high intensity
(24)) were used to estimate percent tree cover in settlement area by state (U.S. Department of Interior 2018;
MRLC 2013).
a) "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.
b) "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
housing units." Plots designated as single family or low-density residential land were classified as
Developed, Low Intensity.
6-156 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 database107 and Forest Service urban forest inventory data
(e.g., Nowak et al. 2016, 2017) (Table 6-115). 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).
In each city, a random sample of plots were measured to assess tree stem diameter, tree height, crown height and
crown width, tree location, species, and canopy condition. The data for each tree were used to estimate total dry-
weight biomass using allometric models, a root-to-shoot ratio to convert aboveground biomass estimates to whole
tree biomass, and wood moisture content. Total dry weight biomass was converted to C by dividing by two (50
percent carbon content). An adjustment factor of 0.8 was used for open grown trees to account for settlement
107 See http://www.itreetools.org.
Land Use, Land-Use Change, and Forestry 6-157
-------
trees having less aboveground biomass for a given stem diameter than predicted by allometric models based on
forest trees (Nowak 1994). Carbon storage estimates for deciduous trees include only C stored in wood. Estimated
C storage was divided by tree cover in the area to estimate carbon storage per square meter of tree cover.
Table 6-115: 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)
City
Storage
SE
Gross
Seauestration
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
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
6-158 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
South Dakotab
Syracuse, NY
Tennessee15
Washington, DC
Woodbridge, NJ
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
based on expert judgment of the authors. Decomposition rates were based on literature estimates (Nowak et al.
2013).
Estimates of gross and net sequestration rates for each of the 50 states and the District of Columbia (Table 6-116)
were compiled in units of C sequestration per unit area of tree canopy cover. These rates were used in conjunction
with estimates of state settlement area and developed land percent tree cover data to calculate each state's
annual net C sequestration by urban trees. This method was described in Nowak et al. (2013) and has been
modified here to incorporate developed land percent tree cover data.
3.14
0.66
0.13
0.03
0.11
0.02
0.87
16.5
2.2
9.48
1.08
0.30
0.03
0.22
0.04
0.72
26.9
1.3
6.47
0.50
0.34
0.02
0.30
0.02
0.89
37.7
0.8
8.52
1.04
0.26
0.03
0.21
0.03
0.79
35.0
2.0
8.19
0.82
0.29
0.03
0.21
0.03
0.73
29.5
1.7
Land Use, Land-Use Change, and Forestry 6-159
-------
Net annual C sequestration estimates were obtained for all 50 states and the District of Columbia by multiplying
the gross annual emission estimates by 0.73, the average ratio for net/gross sequestration (Table 6-116). However,
state specific ratios were used where available.
State Carbon Sequestration Estimates
The gross and net annual C sequestration values for each state were multiplied by each state's settlement area of
tree cover, which was the product of the state's settlement area and the state's tree cover percentage based on
NLCD developed land. The model used to calculate the total carbon sequestration amounts for each state, can be
written as follows:
Equation 6-1: Net State Annual Carbon Sequestration
Net state annual C sequestration (t C/yr) = Gross state sequestration rate (t C/ha/yr) x Net to Gross state
sequestration ratio x state settlement Area (ha) x % state tree cover in settlement area
The results for all 50 states and the District of Columbia are given in Table 6-116. This approach is consistent with
the default IPCC Gain-Loss methodology in IPCC (2006), although sufficient field data are not yet available to
separately determine interannual gains and losses in C stocks in the living biomass of settlement trees. Instead, the
methodology applied here uses estimates of net C sequestration based on modeled estimates of decomposition,
as given by Nowak et al. (2013).
Table 6-116: Estimated Annual C Sequestration (Metric Tons C/Year), Tree Cover (Percent),
and Annual C Sequestration per Area of Tree Cover (kg C/m2/ year) for settlement areas in
United States by State and the District of Columbia (2020)
Gross Annual
Net Annual
Net: Gross
Sequestration
Sequestration
Annual
Gross Annual
Net Annual
Tree
per Area of
per Area of
Sequestration
State
Sequestration
Sequestration
Cover
Tree Cover
Tree Cover
Ratio
Alabama
2,060,001
1,501,070
53.5
0.376
0.274
0.73
Alaska
111,722
81,409
47.4
0.169
0.123
0.73
Arizona
172,750
125,878
4.6
0.388
0.283
0.73
Arkansas
1,266,164
922,622
48.9
0.362
0.264
0.73
California
2,007,869
1,463,083
16.9
0.426
0.311
0.73
Colorado
142,719
103,996
8.0
0.216
0.157
0.73
Connecticut
618,683
450,818
58.7
0.262
0.191
0.73
Delaware
97,533
71,070
24.4
0.366
0.267
0.73
DC
11,995
8,741
25.1
0.366
0.267
0.73
Florida
4,322,610
3,149,776
40.3
0.520
0.379
0.73
Georgia
3,411,478
2,485,857
56.3
0.387
0.282
0.73
Hawaii
285,700
208,182
41.7
0.637
0.464
0.73
Idaho
59,611
43,437
7.4
0.201
0.146
0.73
Illinois
662,891
483,032
15.5
0.310
0.226
0.73
Indiana
472,905
437,275
17.1
0.274
0.254
0.92
Iowa
177,692
129,480
8.6
0.263
0.191
0.73
Kansas
290,461
226,027
10.8
0.310
0.241
0.78
Kentucky
926,269
674,949
36.8
0.313
0.228
0.73
Louisiana
1,512,145
1,101,861
47.0
0.435
0.317
0.73
Maine
394,471
287,441
55.5
0.242
0.176
0.73
Maryland
818,044
596,088
40.1
0.353
0.257
0.73
Massachusetts
1,002,723
730,659
57.2
0.278
0.203
0.73
Michigan
1,343,325
978,847
34.7
0.241
0.175
0.73
Minnesota
313,364
228,340
13.1
0.251
0.183
0.73
Mississippi
1,518,448
1,106,454
57.3
0.377
0.275
0.73
Missouri
850,492
619,732
23.2
0.313
0.228
0.73
Montana
48,911
35,640
4.9
0.201
0.147
0.73
Nebraska
98,584
83,192
7.3
0.261
0.220
0.84
6-160 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Nevada
41,181
30,008
4.8
0.226
0.165
0.73
New Hampshire
363,989
265,229
59.3
0.238
0.174
0.73
New Jersey
904,868
659,355
40.7
0.321
0.234
0.73
New Mexico
177,561
129,384
10.2
0.288
0.210
0.73
New York
1,531,415
1,115,903
39.9
0.263
0.192
0.73
North Carolina
3,064,797
2,233,239
54.1
0.341
0.249
0.73
North Dakota
18,492
8,787
1.8
0.244
0.116
0.48
Ohio
1,248,841
909,999
28.2
0.271
0.198
0.73
Oklahoma
699,044
509,376
22.1
0.364
0.265
0.73
Oregon
682,468
497,297
39.9
0.265
0.193
0.73
Pennsylvania
1,794,939
1,307,927
40.2
0.267
0.195
0.73
Rhode Island
121,940
88,855
50.0
0.283
0.206
0.73
South Carolina
1,801,029
1,312,364
53.8
0.370
0.269
0.73
South Dakota
29,489
25,573
2.9
0.258
0.224
0.87
Tennessee
1,591,278
1,422,789
41.1
0.332
0.297
0.89
Texas
4,239,494
3,089,211
28.5
0.403
0.294
0.73
Utah
118,880
86,625
11.7
0.235
0.172
0.73
Vermont
176,564
128,658
50.6
0.234
0.170
0.73
Virginia
1,968,537
1,434,422
52.9
0.321
0.234
0.73
Washington
1,063,871
775,216
37.6
0.282
0.206
0.73
West Virginia
699,320
509,577
64.1
0.264
0.192
0.73
Wisconsin
697,863
508,515
25.9
0.246
0.180
0.73
Wyoming
29,984
21,849
4.7
0.199
0.145
0.73
Total
48,065,406
35,405,113
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 6-117). 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, urban soil flux is not quantified as part of this 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 2020. The results of this quantitative uncertainty analysis are summarized in Table
6-117. The change in C stocks in Settlement Trees in 2020 was estimated to be between -195.4 and -62.2 MMT CO2
Eq. at a 95 percent confidence level. This analysis indicates a range of 51 percent more sequestration to 52 percent
less sequestration than the 2020 flux estimate of -129.8 MMT CO2 Eq.
Land Use, Land-Use Change, and Forestry 6-161
-------
Table 6-117: Approach 2 Quantitative Uncertainty Estimates for Net CO2 Flux from Changes
in C Stocks in Settlement Trees (MMT CO2 Eq. and Percent)
Source Gas
2020 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
(129.8)
(195.42) (62.22)
-51% 52%
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
There were no recalculations to the 1990 through 2019 time series in this Inventory.
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. What needs to be determined is how much of this settlement area tree
cover might also be accounted for in "forest" area assessments as some of these forests may fall within 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
source category might also be counted in the urban areas. The potential overlap with settlement areas is unknown.
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."
N20 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
6-162 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 Settlements108 are 2.5 MMT CO2 Eq. (8 kt of N2O) in 2020.
There is an overall increase of 23 percent from 1990 to 2020 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 6-118.
Table 6-118: N2O Emissions from Soils in Settlements Remaining Settlements (MMT CO2 Eq.
and kt N2O)
1990
2005
2016
2017
2018
2019
2020
MMT CO? Eq.
Direct N20 Emissions from Soils
1.6
2.5
1.9
2.0
2.0
2.1
2.1
Synthetic Fertilizers
0.8
1.6
0.9
1.0
1.0
1.1
1.1
Biosolids
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Drained Organic Soils
0.6
0.7
0.8
0.8
0.8
0.8
0.8
Indirect N20 Emissions from Soils
0.4
0.6
0.3
0.4
0.4
0.4
0.4
Total
2.0
3.1
2.2
2.3
2.4
2.4
2.5
kt N20
Direct N20 Emissions from Soils
6
9
6
7
7
7
7
Synthetic Fertilizers
3
6
3
3
4
4
4
Biosolids
1
1
1
1
1
1
1
Drained Organic Soils
2
2
3
3
3
3
3
Indirect N20 Emissions from Soils
1
2
1
1
1
1
1
Total
7
10
8
8
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 7.1—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 2020 are based on 2012 values adjusted for annual total N fertilizer sales
in the United States 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 total amount of fertilizer N
108 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 6-163
-------
applied to settlements is multiplied by the IPCC default emission factor (1 percent) to estimate direct N2O
emissions (IPCC 2006) for 1990 to 2012.
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 2020.
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 2015 for fertilizer and from 1990
to 2020 for biosolids.
In order to ensure time-series consistency, the same methods are applied from 1990 to 2020 for biosolids. For
synthetic fertilizer and 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 2020 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 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 time series will be
recalculated for the years beyond 2015 in a future Inventory with the methods described above for 1990 to 2015.
This Inventory does incorporate updated activity data on biosolids application in settlements through 2020.
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 any
of these variables, except variation in the total amount of fertilizer N and biosolids application, which in turn, 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.109 Uncertainty in the area of
drained organic soils is based on the estimated variance from the NRI survey (USDA-NRCS 2018). For 2016 to 2020,
there is also additional uncertainty associated with the fit of the linear regression model for the data splicing
methods.
For biosolids, there is uncertainty in the amounts of biosolids applied to non-agricultural lands and used in surface
disposal. These uncertainties are derived from variability in several factors, including: (1) N content of biosolids; (2)
total sludge applied in 2000; (3) wastewater existing flow in 1996 and 2000; and (4) the biosolids disposal practice
109 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.
6-164 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
distributions to non-agricultural land application and surface disposal. In addition, there is uncertainty in the direct
and indirect emission factors that are provided by IPCC (2006).
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
6-119. Direct N2O emissions from soils in Settlements Remaining Settlements in 2020 are estimated to be between
1.3 and 2.5 MMT CO2 Eq. at a 95 percent confidence level. This indicates a range of 38 percent below to 22 percent
above the 2020 emission estimate of 2.1 MMT CO2 Eq. Indirect N2O emissions in 2020 are between 0.2 and 0.5
MMT CO2 Eq., ranging from 38 percent below to 38 percent above the estimate of 0.4 MMT CO2 Eq.
Table 6-119: Quantitative Uncertainty Estimates of N2O Emissions from Soils in Settlements
Remaining Settlements (MMT CO2 Eq. and Percent)
Source
Gas
2020 Emissions
(MMT C02 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT C02 Eq.) (%)
Settlements Remaining
Settlements
Lower
Bound
Upper
Bound
Lower Upper
Bound Bound
Direct N20 Emissions from Soils
N20
2.1
1.3
2.5
-38% 22%
Indirect N20 Emissions from Soils
n2o
0.4
0.2
0.5
-38% 38%
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
that was corrected.
Recalculations Discussion
Recalculations are associated with updated estimates for 2019 using the linear extrapolation method. As a result,
N2O Emissions from Settlement Soils has a smaller emission of 0.04 MMT CO2 Eq. in 2019. This represents less than
1 percent decrease in emissions compared to the previous Inventory.
Planned Improvements
This source will be extended to include soil N2O emissions from drainage of organic soils in settlements of Alaska
and federal lands in order to provide a complete inventory of emissions for this category. Data on fertilizer amount
and area of drained organic soils will be compiled to update emissions estimates from 2016 to 2020 in a future
Inventory.
Changes in Yard Trimmings and Food Scrap Carbon Stocks in
Landfills (CRF Category 4E1)
In the United States, yard trimmings (i.e., grass clippings, leaves, and branches) and food scraps account for a
significant portion of the municipal waste stream, and a large fraction of the collected yard trimmings and food
scraps are put in landfills. Carbon (C) contained in landfilled yard trimmings and food scraps can be stored for very
long periods.
Carbon storage estimates within the Inventory are associated with particular land uses. For example, harvested
wood products are reported under Forest Land Remaining Forest Land because these wood products originated
Land Use, Land-Use Change, and Forestry 6-165
-------
from the forest ecosystem. Similarly, C stock changes in yard trimmings and food scraps are reported under
Settlements Remaining Settlements because the bulk of the C, which comes from yard trimmings, originates from
settlement areas. While the majority of food scraps originate from cropland and grassland, in this Inventory they
are reported with the yard trimmings in the Settlements Remaining Settlements section. Additionally, landfills are
considered part of the managed land base under settlements (see Section 6.1 Representation of the U.S. Land
Base), and reporting these C stock changes that occur entirely within landfills fits most appropriately within the
Settlements Remaining Settlements section.
Both the estimated amount of yard trimmings collected annually and the fraction that is landfilled have been
declining. In 1990, over 53 million metric tons (wet weight) of yard trimmings and food scraps are estimated to
have been generated (i.e., put at the curb for collection to be taken to disposal sites or to composting facilities)
(EPA 2020). Since then, programs banning or discouraging yard trimmings disposal have led to an increase in
backyard composting and the use of mulching mowers, and a consequent estimated 1.1 percent increase between
1990 and 2020 in the tonnage of yard trimmings generated (i.e., collected for composting or disposal in landfills)
per year. At the same time, an increase in the number of municipal composting facilities has reduced the
proportion of collected yard trimmings that are discarded in landfills per year—from 72 percent in 1990 to 30
percent in 2020. The net effect of the slight increase in generation and the increase in composting is a 58 percent
decrease in the quantity of yard trimmings disposed of in landfills since 1990.
Food scrap generation has grown by an estimated 165 percent since 1990, and while the proportion of total food
scraps generated that are eventually discarded in landfills has decreased, from an estimated 82 percent in 1990 to
55 percent in 2020, the tonnage disposed of in landfills has increased considerably (by an estimated 78 percent)
due to the increase in food scrap generation. Although the total tonnage of food scraps disposed of in landfills has
increased from 1990 to 2020, the difference in the amount of food scraps added from one year to the next
generally decreased, and consequently the annual carbon stock net changes from food scraps have generally
decreased as well (as shown in Table 6-120 and Table 6-121). As described in the Methodology section, the carbon
stocks are modeled using data on the amount of food scraps landfilled since 1960. These food scraps decompose
over time, producing Cm and CO2. Decomposition happens at a higher rate initially, then decreases. As
decomposition decreases, the carbon stock becomes more stable. Because the cumulative carbon stock left in the
landfill from previous years is (1) not decomposing as much as the carbon introduced from food scraps in a single
more recent year; and (2) is much larger than the carbon introduced from food scraps in a single more recent year,
the total carbon stock in the landfill is primarily driven by the more stable "older" carbon stock, thus resulting in
less annual change in later years.
Overall, the decrease in the landfill disposal rate of yard trimmings has more than compensated for the increase in
food scrap disposal in landfills, and the net result is a decrease in annual net change landfill C storage from 24.5
MMT CO2 Eq. (6.7 MMT C) in 1990 to 12.2 MMT C02 Eq. (3.3 MMT C) in 2020 (Table 6-120 and Table 6-121), a
decrease of 50 percent over the time series.
Table 6-120: Net Changes in Yard Trimmings and Food Scrap Carbon Stocks in Landfills
(MMT COz Eq.)
Carbon Pool
1990
2005
2016
2017
2018
2019
2020
Yard Trimmings
(20.1)
(7.5)
(6.3)
(8.3)
(8.3)
(8.2)
(8.1)
Grass
(1.7)
(0.6)
(0.5)
(0.8)
(0.8)
(0.8)
(0.7)
Leaves
(8.7)
(3.4)
(3.0)
(3.8)
(3.8)
(3.8)
(3.7)
Branches
(9.8)
(3.4)
(2.8)
(3.7)
(3.7)
(3.7)
(3.6)
Food Scraps
(4.4)
(3.9)
(3.7)
(5.6)
(5.2)
(4.8)
(4.1)
Total Net Flux
(24.5)
(11.4)
(10.0)
(13.8)
(13.4)
(13.1)
(12.2)
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
6-166 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 6-121: Net Changes in Yard Trimmings and Food Scrap Carbon Stocks in Landfills
(MMT C)
Carbon Pool
1990
2005
2016
2017
2018
2019
2020
Yard Trimmings
(5.5)
(2.0)
(1.7)
(2.3)
(2.3)
(2.2)
(2.2)
Grass
(0.5)
(0.2)
(0.1)
(0.2)
(0.2)
(0.2)
(0.2)
Leaves
(2.4)
(0.9)
(0.8)
(1.0)
(1.0)
(1.0)
(1.0)
Branches
(2.7)
(0.9)
(0.8)
(1.0)
(1.0)
(1.0)
(1.0)
Food Scraps
(1.2)
(1.1)
(1.0)
(1.5)
(1.4)
(1.3)
(1.1)
Total Net Flux
(6.7)
(3.1)
(2.7)
(3.8)
(3.7)
(3.6)
(3.3)
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Methodology and Time-Series Consistency
When wastes of biogenic origin (such as yard trimmings and food scraps) are landfilled and do not completely
decompose, the C that remains is effectively removed from the C cycle. Empirical evidence indicates that yard
trimmings and food scraps do not completely decompose in landfills (Barlaz 1998, 2005, 2008; De la Cruz and
Barlaz 2010), and thus the stock of C in landfills can increase, with the net effect being a net atmospheric removal
of C. Estimates of net C flux resulting from landfilled yard trimmings and food scraps were developed by estimating
the change in landfilled C stocks between inventory years and are based on methodologies presented for the Land
Use, Land-Use Change, and Forestry sector in IPCC (2003) and the 2006IPCC Guidelines for National Greenhouse
Gas Inventories (IPCC 2006). Carbon stock estimates were calculated by determining the mass of landfilled C
resulting from yard trimmings and food scraps discarded in a given year; adding the accumulated landfilled C from
previous years; and subtracting the mass of C that was landfilled in previous years and has since decomposed and
been emitted as CO2 and Cm.
To determine the total landfilled C stocks for a given year, the following data and factors were assembled:
(1) The composition of the yard trimmings;
(2) The mass of yard trimmings and food scraps discarded in landfills;
(3) The C storage factor of the landfilled yard trimmings and food scraps; and
(4) The rate of decomposition of the degradable C.
The composition of yard trimmings was assumed to be 30 percent grass clippings, 40 percent leaves, and 30
percent branches on a wet weight basis (Oshins and Block 2000). The yard trimmings were subdivided, because
each component has its own unique adjusted C storage factor (i.e., moisture content and C content) and rate of
decomposition. The mass of yard trimmings and food scraps disposed of in landfills was estimated by multiplying
the quantity of yard trimmings and food scraps discarded by the proportion of discards managed in landfills. Data
on discards (i.e., the amount generated minus the amount diverted to centralized composting facilities) for both
yard trimmings and food scraps were taken primarily from Advancing Sustainable Materials Management: Facts
and Figures 2018 (EPA 2020), which provides data for 1960,1970,1980,1990, 2000, 2005, 2010, 2015, 2017 and
2018. To provide data for some of the missing years, detailed backup data were obtained from the 2012, 2013, and
2014, 2015, and 2017 versions of the Advancing Sustainable Materials Management: Facts and Figures reports
(EPA 2019), as well as historical data tables that EPA developed for 1960 through 2012 (EPA 2016). Remaining
years in the time series for which data were not provided were estimated using linear interpolation. Since the
Advancing Sustainable Materials Management: Facts and Figures reports for 2019 and 2020 were unavailable,
landfilled material generation, recovery, and disposal data for 2019 and 2020 were proxied equal to 2018 values.
The amount of C disposed of in landfills each year, starting in 1960, was estimated by converting the discarded
landfilled yard trimmings and food scraps from a wet weight to a dry weight basis, and then multiplying by the
initial (i.e., pre-decomposition) C content (as a fraction of dry weight). The dry weight of landfilled material was
calculated using dry weight to wet weight ratios (Tchobanoglous et al. 1993, cited by Barlaz 1998) and the initial C
contents and the C storage factors were determined by Barlaz (1998, 2005, 2008) (Table 6-122).
Land Use, Land-Use Change, and Forestry 6-167
-------
The amount of C remaining in the landfill for each subsequent year was tracked based on a simple model of C fate.
As demonstrated by Barlaz (1998, 2005, 2008), a portion of the initial C resists decomposition and is essentially
persistent in the landfill environment. Barlaz (1998, 2005, 2008) conducted a series of experiments designed to
measure biodegradation of yard trimmings, food scraps, and other materials, in conditions designed to promote
decomposition (i.e., by providing ample moisture and nutrients). After measuring the initial C content, the
materials were placed in sealed containers along with methanogenic microbes from a landfill. Once decomposition
was complete, the yard trimmings and food scraps were re-analyzed for C content; the C remaining in the solid
sample can be expressed as a proportion of the initial C (shown in the row labeled "C Storage Factor, Proportion of
Initial C Stored (%)" in Table 6-122).
The modeling approach applied to simulate U.S. landfill C flows builds on the findings of Barlaz (1998, 2005, 2008).
The proportion of C stored is assumed to persist in landfills. The remaining portion is assumed to degrade over
time, resulting in emissions of CFU and CO2. (The CFU emissions resulting from decomposition of yard trimmings
and food scraps are reported in the Waste chapter.) 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 6-122.
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 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 Section 7.1. The component-specific
decay rates are shown in Table 6-122.
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:
6-168 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Equation 6-2: Total C Stock for Yard Trimmings and Food Scraps in Landfills
f
LFCi.t = £ Win x (1 - MCi) x ICQx {[CSi~x ICQ + [(1 - (CS,x ICC)) x "J]}
n
where,
t = Year for which C stocks are being estimated (year),
/ = Waste type for which C stocks are being estimated (grass, leaves, branches, food
scraps),
LFQt = Stock of C in landfills in year t, for waste / (metric tons),
Wi,n = 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 < t),
MCi = Moisture content of waste / (percent of water),
CSi = Proportion of initial C that is stored for waste / (percent),
ICC, = Initial C content of waste / (percent),
e = Natural logarithm, and
k = First-order decay rate for waste /', (year-1).
For a given year t, the total stock of C in landfills (TLFCt) is the sum of stocks across all four materials (grass, leaves,
branches, food scraps). The annual flux of C in landfills (Ft) for year t is calculated in as the change in C stock
compared to the preceding year according to Equation 6-3:
Equation 6-3: C Stock Annual Flux for Yard Trimmings and Food Scraps in Landfills
Ft= TLFCt- TLFCt- u
Thus, as seen in Equation 1, the C placed in a landfill in year n is tracked for each year t through the end of the
inventory period. For example, disposal of food scraps in 1960 resulted in depositing about 1,135,000 metric tons
of C in landfills. Of this amount, 16 percent (179,000 metric tons) is persistent; the remaining 84 percent (956,000
metric tons) is degradable. By 1965, more than half of the degradable portion (507,000 metric tons) decomposes,
leaving a total of 628,000 metric tons (the persistent portion, plus the remainder of the degradable portion).
Continuing the example, by 2020, the total food scraps C originally disposed of in 1960 had declined to 179,000
metric tons (i.e., virtually all degradable C had decomposed). By summing the C remaining from 1960 with the C
remaining from food scraps disposed of in subsequent years (1961 through 2020), the total landfill C from food
scraps in 2020 was 49.6 million metric tons. This value is then added to the C stock from grass, leaves, and
branches to calculate the total landfill C stock in 2019, yielding a value of 285.7 million metric tons (as shown in
Table 6-123). In the same way total net flux is calculated for forest C and harvested wood products, the total net
flux of landfill C for yard trimmings and food scraps for a given year (Table 6-121) is the difference in the landfill C
stock for the following year (2021C stock was forecast using 1990 to 2020 C stocks) and the stock in the current
year. For example, the net change in 2020 shown in Table 6-121 (2.9 MMT C) is equal to the stock in 2021 (288.7
MMT C) minus the stock in 2020 (285.7 MMT C). The C stocks calculated through this procedure are shown in
Table 6-123.
Table 6-122: Moisture Contents, C Storage Factors (Proportions of Initial C Sequestered),
Initial C Contents, and Decay Rates for Yard Trimmings and Food Scraps in Landfills
Yard Trimmings
Variable ; Food Scraps
Grass Leaves Branches
Moisture Content (% H20)
70
30
10
70
C Storage Factor, Proportion of Initial C
Stored (%)
53
85
77
16
Initial C Content (%)
45
46
49
51
Decay Rate (year1)
0.313
0.179
0.015
0.151
Note: The decay rates are presented as weighted averages based on annual precipitation categories
and population residing in each precipitation category.
Land Use, Land-Use Change, and Forestry 6-169
-------
Table 6-123: C Stocks in Yard Trimmings and Food Scraps in Landfills (MMT C)
Carbon Pool
1990
2005
2016
2017
2018
2019
2020
2021a
Yard Trimmings
156.0
203.1
227.7
229.4
231.6
233.9
236.1
238.3
Branches
14.6
18.1
20.3
20.5
20.7
20.9
21.1
21.3
Leaves
66.7
87.4
98.6
99.4
100.4
101.5
102.5
103.5
Grass
74.7
97.7
108.7
109.5
110.5
111.5
112.5
113.5
Food Scraps
17.9
33.2
44.4
45.4
46.9
48.3
49.6
50.7
Total Carbon Stocks
173.9
236.3
272.0
274.8
278.5
282.2
285.7
289.1
a 2021 C stock estimate was forecasted using 1990 to 2020 data.
Note: Totals may not sum due to independent rounding.
To develop the 2021 C stock estimate, estimates of yard trimming and food scrap carbon stocks were forecasted
for 2021, based on data from 1990 through 2020. These forecasted values were used to calculate net changes in
carbon stocks for 2020. Excels FORECAST.ETS function was used to predict a 2021 value using historical data via an
algorithm called "Exponential Triple Smoothing." This method determined the overall trend and provided
appropriate carbon stock estimates for 2021.
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2020. The same data source was used for the analysis, when available. When data were unavailable,
missing values were estimated using linear interpolation or forecasting, as noted above.
Uncertainty
The uncertainty analysis for landfilled yard trimmings and food scraps includes an evaluation of the effects of
uncertainty for the following data and factors: disposal in landfills per year (tons of C), initial C content, moisture
content, decay rate, and proportion of C stored. The C storage landfill estimates are also a function of the
composition of the yard trimmings (i.e., the proportions of grass, leaves and branches in the yard trimmings
mixture). There are respective uncertainties associated with each of these factors.
A Monte Carlo (Approach 2) uncertainty analysis was applied to estimate the overall uncertainty of the
sequestration estimate for 2020. The results of the Approach 2 quantitative uncertainty analysis are summarized in
Table 6-124. Total yard trimmings and food scraps CO2 flux in 2020 was estimated to be between -20.5 and -5.4
MMT CO2 Eq. at a 95 percent confidence level. This indicates a range of 68 percent below to 56 percent above the
2020 flux estimate of -12.2 MMT CO2 Eq.
Table 6-124: Approach 2 Quantitative Uncertainty Estimates for CO2 Flux from Yard
Trimmings and Food Scraps in Landfills (MMT CO2 Eq. and Percent)
2020 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
(12.2)
(20.5)
(5.4)
-68% 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.
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 Landfilled Yard Trimmings and Food Scraps included checking that input data were properly
transposed within the spreadsheet, checking calculations were correct, and confirming that all activity data and
6-170 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
calculations documentation was complete and updated to ensure data were properly handled through the
inventory process.
Order of magnitude checks and checks of time-series consistency were performed to ensure data were updated
correctly and any changes in emissions estimates were reasonable and reflected changes in activity data. An
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.
Recalculations Discussion
The current Inventory has been revised to reflect updated data from the most recent Advancing Sustainable
Materials Management: Facts and Figures report. Recalculations based on these updates resulted in 2.8 percent
change in the annual carbon stocks and sequestration values as compared to the previous inventory values. The
largest changes occurred in the most recent years: a 3 percent increase in sequestration in 2016, a 40.3 percent
increase in sequestration in 2017, a 37.5 percent increase in sequestration in 2018, and a 28.6 percent increase in
sequestration in 2019. Large changes in yard trimmings can be attributed to updates to 2017 and 2018 yard
trimmings and food scraps landfilled values reported in Advancing Sustainable Materials Management: Facts and
Figures 2018 (EPA 2020). A large increase in sequestration in 2019 can be attributed to updated generation values
as well - 2019 landfill data were unavailable and were reported as 2018 values.
The bulk of the increase in sequestration is attributed to a change in food scrap measurement methodology in the
Advancing Sustainable Materials Management: Facts and Figures 2018 (EPA 2020). The revised methodology more
fully captures flows of recovery of excess food and food waste for 2018 data. The 2018 food scraps recovery
estimates include nine management pathways, three of which previously existed in the report (composting,
combustion with energy recovery, and landfilling). The six new management pathways are:
• Animal feed
• Bio-based materials/biochemical processing
• Land application
• Donation
• Codigestion/anaerobic digestion
• Sewer/wastewater treatment
Food scrap generation estimates increased over 50 percent between 2017 and 2018, from 40.7 million tons to 63.1
million tons, due to the change in food generation measurement. Food scrap recovery estimates increased by
nearly 800 percent (from 2,570 thousand tons in 2017 to 20,300 tons in 2018). Data on the six management
pathways from 1990 to 2017 were not available.
Planned Improvements
EPA plans to evaluated 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. Based upon this evaluation,
changes may be made to the default values.
EPA also plans 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 data changes over time. Currently the
inventory calculations use 2010 U.S. Census data.
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.
Land Use, Land-Use Change, and Forestry 6-171
-------
Six new food waste management pathways were introduced in the 2018 Advancing Sustainable Materials
Management report. Time series data all of these pathways are not provided prior to 2018 but EPA plans to
investigate potential data sources and/or methods to apply data for the remaining time series.
Finally, EPA plans to review available data to ensure all types of landfilled yard trimmings and food scraps are being
included in Inventory estimates, such as debris from road construction and commercial food waste not included in
other chapter estimates.
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).110 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). This Inventory includes all settlements in the conterminous
United States and Hawaii, but does not include settlements in Alaska. Areas of drained organic soils on settlements
in federal lands are also not included in this Inventory. Consequently, there is a discrepancy between the total
amount of managed area for Land Converted to Settlements (see Section 6.1 Representation of the U.S. Land Base)
and the settlements area included in the Inventory analysis.
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 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 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, but not for other land use conversions to
settlements.
Forest Land Converted to Settlements is the largest source of emissions from 1990 to 2020, accounting for
approximately 75 percent of the average total loss of C among all of the land use conversions in Land Converted to
Settlements. Losses of aboveground and belowground biomass, dead wood and litter C losses in 2020 are 36.7,
7.0, 6.4, and 9.3 MMT CO2 Eq., respectively (10.0,1.9,1.7, and 2.5 MMT C). Mineral and organic soils also lost 16.2
and 2.4 MMT C02 Eq. in 2020 (4.4 and 0.6 MMT C). The total net flux is 77.9 MMT C02 Eq. in 2020 (21.2 MMT C),
which is a 28 percent increase in CO2 emissions compared to the emissions in the initial reporting year of 1990
(Table 6-125 and Table 6-126). 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.
110 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.
6-172 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 6-125: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for
Land Converted to Settlements (MMT CO2 Eq.)
1990
2005
2016
2017
2018
2019
2020
Cropland Converted to
Settlements
3.4
9.8
6.0
6.0
5.9
5.9
5.9
Mineral Soils
2.8
8.4
5.2
5.2
5.2
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
52.6
57.7
61.3
61.5
61.6
61.6
61.5
Aboveground Live Biomass
31.7
34.2
36.5
36.6
36.7
36.7
36.7
Belowground Live Biomass
6.1
6.5
7.0
7.0
7.0
7.0
7.0
Dead Wood
5.5
5.9
6.4
6.4
6.4
6.4
6.4
Litter
8.0
8.7
9.3
9.3
9.3
9.3
9.3
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
5.2
16.3
11.3
11.3
11.3
11.3
11.2
Mineral Soils
4.6
14.9
10.4
10.4
10.4
10.4
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
0.4
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
31.7
34.2
36.5
36.6
36.7
36.7
36.7
Total Belowground Biomass Flux
6.1
6.5
7.0
7.0
7.0
7.0
7.0
Total Dead Wood Flux
5.5
5.9
6.4
6.4
6.4
6.4
6.4
Total Litter Flux
8.0
8.7
9.3
9.3
9.3
9.3
9.3
Total Mineral Soil Flux
8.1
23.8
16.3
16.2
16.2
16.2
16.2
Total Organic Soil Flux
1.4
3.6
2.4
2.4
2.4
2.4
2.4
Total Net Flux
60.8
82.8
77.8
77.9
78.0
77.9
77.9
+ 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.
Table 6-126: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for
Land Converted to Settlements (MMT C)
1990
2005
2016
2017
2018
2019
2020
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
14.3
15.7
16.7
16.8
16.8
16.8
16.8
Aboveground Live Biomass
8.6
9.3
10.0
10.0
10.0
10.0
10.0
Belowground Live Biomass
1.7
1.8
1.9
1.9
1.9
1.9
1.9
Dead Wood
1.5
1.6
1.7
1.7
1.7
1.7
1.7
Litter
2.2
2.4
2.5
2.5
2.5
2.5
2.5
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
1.4
4.4
3.1
3.1
3.1
3.1
3.1
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
Land Use, Land-Use Change, and Forestry 6-173
-------
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)
(0.4)
(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
8.6
9.3
10.0
10.0
10.0
10.0
10.0
Total Belowground Biomass
v
Flux
1.7
1.8
1.9
1.9
1.9
1.9
1.9
Total Dead Wood Flux
1.5
1.6
1.7
1.7
1.7
1.7
1.7
Total Litter Flux
2.2
2.4
2.5
2.5
2.5
2.5
2.5
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
0.7
0.6
0.6
0.6
Total Net Flux
16.6
22.6
21.2
21.3
21.3
21.3
21.2
+ 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.
Methodology and Time-Series Consistency
The following section includes a description of the methodology used to estimate C stock changes for Land
Converted to Settlements, including (1) loss of aboveground and belowground biomass, dead wood and litter C
with conversion of forest lands to settlements, as well as (2) the impact from all land use conversions to
settlements on soil organic C stocks in mineral and organic soils.
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 Settlements. 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 2020), however there is no country-specific data for settlements so the biomass, litter, and dead wood
carbon stocks on these converted lands were assumed to be zero. 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).
This inventory also includes estimates of change in dead organic matter for standing dead, deadwood and litter. 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 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).
6-174 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
In order to ensure time-series consistency, the same methods are applied from 1990 to 2020 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 are currently available 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.
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. Soil organic C stock changes from 2016 to 2020
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 2020 emissions in a future Inventory.
Organic Soil Carbon Stock Changes
Annual C emissions from drained organic soils in Land Converted to Settlements are estimated using the Tier 2
method provided in IPCC (2006). The Tier 2 method assumes that organic soils are losing C at a rate similar to
croplands, and therefore uses the country-specific values for cropland (Ogle et al. 2003). To estimate CO2
Land Use, Land-Use Change, and Forestry 6-175
-------
emissions from 1990 to 2015, the area of organic soils in Land Converted to Settlements is multiplied by the Tier 2
emission factor, which is 11.2 MT C per ha in cool temperate 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 for more information).
In order to ensure time-series consistency, the same methods are applied from 1990 to 2015, and a linear
extrapolation method is used to approximate emissions for the remainder of the 2016 to 2020 time series (See 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). Estimates
will be recalculated in future Inventories when new NRI data are available.
Uncertainty
The uncertainty analysis for C losses with Forest Land Converted to Settlements is 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), i.e., 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 analysis for mineral soil organic C
stock changes and annual C emission estimates from drained organic soils in Land Converted to Settlements is
estimated using a Monte Carlo approach, which is also described in the Cropland Remaining Cropland section.
Uncertainty estimates are presented in Table 6-127 for each subsource (i.e., biomass C, dead wood, litter, soil
organic C in mineral soil and organic 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), i.e., as described in the previous paragraph. There are also additional uncertainties
propagated through the analysis associated with the data splicing methods applied to estimate soil organic C stock
changes from 2016 to 2020. The combined uncertainty for total C stocks in Land Converted to Settlements ranges
from 34 percent below to 34 percent above the 2020 stock change estimate of 77.9 MMT CO2 Eq.
Table 6-127: 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)
2020 Flux Estimate
Uncertainty Range Relative to Flux Estimate3
Source
(MMT C02 Eq.)
(MMT C02
Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Cropland Converted to Settlements
5.9
2.1
9.8
-65%
65%
Mineral Soil C Stocks
5.1
1.3
8.9
-74%
74%
Organic Soil C Stocks
0.8
0.1
1.5
-86%
86%
Forest Land Converted to Settlements
61.5
37.3
85.8
-39%
39%
Aboveground Biomass C Stocks
36.7
13.9
59.5
-62%
62%
Belowground Biomass C Stocks
7.0
2.6
11.4
-62%
62%
Dead Wood
6.4
2.4
10.3
-62%
62%
Litter
9.3
3.5
15.1
-62%
62%
Mineral Soil C Stocks
1.9
1.3
2.5
-32%
32%
Organic Soil C Stocks
0.3
0.1
0.5
-72%
72%
Grassland Converted to Settlements
11.2
6.1
16.4
-46%
46%
Mineral Soil C Stocks
10.3
5.2
15.4
-49%
49%
Organic Soil C Stocks
0.9
0.1
1.7
-90%
90%
Other Lands Converted to Settlements
(1.2)
(2.0)
(0.4)
-68%
68%
Mineral Soil C Stocks
(1.3)
(2.0)
(0.5)
-61%
61%
Organic Soil C Stocks
0.1
(0.1)
0.3
-168%
168%
Wetlands Converted to Settlements
0.3
(0.2)
0.9
-150%
150%
Mineral Soil C Stocks
0.1
+
0.1
-103%
103%
Organic Soil C Stocks
0.3
(0.2)
0.8
-182%
182%
6-176 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Total: Land Converted to Settlements
77.9
51.4
104.4
-34%
34%
Aboveground Biomass C Stocks
36.7
13.9
59.5
-62%
62%
Belowground Biomass C Stocks
7.0
2.6
11.4
-62%
62%
Dead Wood
6.4
2.4
10.3
-62%
62%
Litter
9.3
3.5
15.1
-62%
62%
Mineral Soil C Stocks
16.2
9.7
22.6
-40%
40%
Organic Soil C Stocks
2.4
(6.2)
10.9
-361%
361%
+ 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.
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. No errors were found in this Inventory.
Recalculations Discussion
Recalculations are associated with new FIA data from 1990 to 2020 on biomass, dead wood and litter C stocks in
Forest Land Converted to Settlements, and updated estimates for mineral and organic soils from 2016 to 2020
using the linear extrapolation method. As a result, Land Converted to Settlements has an estimated smaller C loss
of 2.0 MMT CO2 Eq. on average over the time series. This represents a 2.6 percent decrease in C stock changes for
Land Converted to Settlements compared to the previous Inventory.
Planned Improvements
A planned improvement for the Land Converted to Settlements category is to develop an inventory of mineral soil
organic C stock changes in Alaska and losses of C from drained organic soils in federal lands. This includes C stock
changes for biomass, dead organic matter and soils. See Table 6-128 for the amount of managed land area in Land
Converted to Settlements that is not included in the Inventory due to these omissions. The managed area that is
not included in the Inventory ranges between 0 and about 600 thousand hectares depending on the year.
There are plans to improve classification of trees in settlements and to include transfer of biomass from forest land
to those areas in this category. There are also plans to extend the Inventory to included C losses associated with
drained organic soils in settlements occurring on federal lands.
New land representation data will also be compiled, and the emissions data will be recalculated for the latter years
in the time series that are estimated using data splicing methods in this Inventory. These improvements will be
made as funding and resources are available to expand the inventory for this source category.
Land Use, Land-Use Change, and Forestry 6-177
-------
Table 6-128: Area of Managed Land in Settlements Remaining Settlements that is not
included in 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
2,861
2,861
0
1991
3,238
3,238
0
1992
3,592
3,592
0
1993
4,178
4,107
72
1994
4,777
4,630
147
1995
5,384
5,161
223
1996
5,927
5,658
269
1997
6,520
6,174
346
1998
7,065
6,650
416
1999
7,577
7,116
461
2000
8,095
7,568
528
2001
8,544
7,947
597
2002
8,886
8,284
602
2003
8,941
8,335
606
2004
8,957
8,345
612
2005
8,947
8,341
606
2006
8,959
8,352
607
2007
8,902
8,295
607
2008
8,722
8,111
610
2009
8,541
7,930
611
2010
8,335
7,725
611
2011
8,108
7,498
611
2012
7,918
7,298
620
2013
7,504
6,932
572
2014
7,087
6,586
501
2015
6,589
6,165
424
2016
ND
ND
ND
2017
ND
ND
ND
2018
ND
ND
ND
2019
ND
ND
ND
2020
ND
ND
ND
Note: NRI data are not available after 2015, and these years are designated as ND (No data).
6.12 Other Land Remaining Other Land (CRF
Category 4F1)
Land use is constantly occurring, and areas under a number of differing land-use types remain in their respective
land-use type each year, just as other land can remain as other land. While the magnitude of Other Land
Remaining Other Land is known (see Table 6-4), research is ongoing to track C pools in this land use. Until such
time that reliable and comprehensive estimates of C for Other Land Remaining Other Land can be produced, it is
not possible to estimate CO2, Cm or N2O fluxes on Other Land Remaining Other Land at this time.
6-178 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
6.13 Land Converted to Other Land (CRF
Category 4F2)
Land-use change is constantly occurring, and areas under a number of differing land-use types are converted to
other land each year, just as other land is converted to other uses. While the magnitude of these area changes is
known (see Table 6-4), research is ongoing to track C across Other Land Remaining Other Land and Land Converted
to Other Land. Until such time that reliable and comprehensive estimates of C across these land-use and land-use
change categories can be produced, it is not possible to separate CO2, Cm or N2O fluxes on Land Converted to
Other Land from fluxes on Other Land Remaining Other Land at this time.
Land Use, Land-Use Change, and Forestry 6-179
-------
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 16.8 percent of total U.S. anthropogenic methane (Cm) emissions in
2020, 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
2.8 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
6.0 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
2020, waste activities generated emissions of 155.6 MMT CO2 Eq., or 2.6 percent of total U.S. greenhouse gas
emissions.
Figure 7-1: 2020 Waste Sector Greenhouse Gas Sources
Wastewater Treatment
Anaerobic Digestion at
Biogas Facilities
Composting
Landfills
0 10 20 30 40 50 60 70 80 90 100 110 120
MMT CO2 Eq.
Waste 7-1
-------
Figure 7-2: Trends in Waste Chapter Greenhouse Gas Sources
Table 7-1: Emissions from Waste (MMT CO2 Eq.)
Gas/Source
1990
2005
2016
2017
2018
2019
2020
ch4
197.3
153.6
129.1
130.3
132.4
134.1
130.0
Landfills
176.6
131.5
107.9
109.2
111.7
113.6
109.3
Wastewater T reatment
20.3
20.1
18.7
18.5
18.3
18.1
18.3
Composting
0.4
1.9
2.3
2.5
2.3
2.3
2.3
Anaerobic Digestion at Biogas
Facilities
+
+
0.2
0.2
0.2
0.2
0.2
n2o
16.9
22.0
24.8
25.4
25.5
25.4
25.6
Wastewater T reatment
16.6
20.3
22.8
23.2
23.5
23.4
23.5
Composting
0.3
1.7
2.0
2.2
2.0
2.0
2.0
Total
214.2
175.6
153.9
155.7
157.9
159.6
155.6
+ Does not exceed 0.05 MMT C02 Eq.
Note: Totals may not sum due to independent rounding.
Table 7-2: Emissions from Waste (kt)
Gas/Source
1990
2005
2016
2017
2018
2019
2020
ch4
7,890
6,144
5,164
5,212
5,296
5,365
5,201
Landfills
7,063
5,262
4,318
4,368
4,467
4,545
4,373
Wastewater Treatment
812
806
748
740
732
723
730
Composting
15
75
91
98
90
91
91
Anaerobic Digestion at Biogas
Facilities
1
2
7
6
6
6
6
n2o
57
74
83
85
86
85
86
Wastewater Treatment
56
68
76
78
79
79
79
Composting
1
6
7
7
7
7
7
Note: Totals may not sum due to independent rounding.
Carbon dioxide (CO2), Cm, 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
7-2 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 2020 resulted in 13.5 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 2019) to ensure that the trend
is accurate. EPA revised the factor used to estimate non-consumed protein in domestic wastewater treatment. EPA
also incorporated the Tier 2 methodology from 2019 Refinement (IPCC 2019) for Cm emissions from the discharge
of pulp and paper manufacturing wastewater to aquatic environments. EPA also revised the emissions factor for
centralized aerobic systems and updated the BOD concentration for wastewater entering constructed wetlands as
a tertiary treatment. Overall, these recalculations lead to an average decrease of 2.6 MMT CO2 Eq. (1.5 percent)
across the timeseries. 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, but not 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
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.
Waste 7-3
-------
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 Cm emissions from MSW landfills for the years 2010 to 2020 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).
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);
7-4 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
• 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.1 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,
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 2020, landfill CH4 emissions were approximately 109.3 MMT CO2 Eq. (4,373 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 86 percent of total landfill emissions (94.2 MMT CO2 Eq.), while
industrial waste landfills accounted for the remainder (15.1 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 CH4 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 2021a; EPA 2021b; EPA 2019; Waste Business
Journal [WBJ] 2016; WBJ 2010). The Environment Research & Education Foundation (EREF) conducted a
1 For more information regarding federal MSW landfill regulations, see
http://www.epa.gov/osw/nonhaz/municipal/landfill/msw regs.htm.
Waste 7-5
-------
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 2021b; 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 2021a; EREF 2016; BioCycle 2010). Larger landfills may have deeper cells where a
greater amount of area will be anaerobic (more Cm 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 6 percent to approximately 214 MMT in 2020 (see Annex 3.14, Table A-220). Emissions
decreased between 1990 to 2010 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, but the percentage of waste landfilled may decline due to increased recycling and composting
practices. The impacts of the coronavirus (COVID-19) pandemic with respect to landfilled waste cannot be
quantified until data sources such as the EPA's Advancing Sustainable Materials Management: Facts and Figures
report are published for 2019 and 2020. The quantities of waste landfilled for 2014 to 2020 (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 10.3 MMT in 2020 (see Annex
3.14, Table A-219). CH4 emissions from industrial waste landfills have also remained at similar levels recently,
ranging from 14.4 MMT CO2 Eq. in 2005 to 15.1 MMT CO2 Eq. in 2020 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
landfills and the GHGRP Subpart TT (Industrial Waste Landfills) dataset has confirmed C&D landfills, for example,
are insignificant CH4 generators.
EPA's Landfill Methane Outreach Program (LMOP) collects information on landfill gas energy projects currently
operational or under construction throughout the United States. LMOP's Landfill and Landfill Gas Energy Database
contains certain information on the gas collection and control systems in place at landfills provided by
organizations that are a part of the program, which can include the amount of landfill gas collected and flared. In
2020, LMOP identified 6 new landfill gas-to-energy (LFGE) projects (EPA 2021b) that began operation. While the
amount of landfill gas collected and combusted continues to increase, the rate of increase in collection and
combustion no longer exceeds the rate of additional CH4 generation from the amount of organic MSW landfilled as
the U.S. population grows (EPA 2021a).
Landfill gas collection and control is not accounted for at industrial waste landfills in this chapter (see the
Methodology discussion for more information).
7-6 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 7-3: ChU Emissions from Landfills (MMT CO2 Eq.)
Activity
1990
2005
2016
2017
2018
2019
2020
MSW CH4 Generation3
205.3
271.1
283.0
291.8
297.0
304.6
304.8
Industrial CH4 Generation
12.1
16.0
16.6
16.7
16.7
16.7
16.8
MSW CH4 Recovered3
(21.3)
(132.5)
(165.9)
(172.1)
(174.1)
(179.6)
(184.1)
MSW CH4 Oxidized3
(18.4)
(21.4)
(24.1)
(25.5)
(26.2)
(26.5)
(26.6)
Industrial CH4 Oxidized
(1.2)
(1.6)
(1.7)
(1.7)
(1.7)
(1.7)
(1.7)
MSW net CH4 Emissions
165.7
117.2
93.0
94.2
96.7
98.6
94.2
Industrial CH4 Emissions'5
10.9
14.4
15.0
15.0
15.0
15.1
15.1
Total
176.6
131.5
107.9
109.2
111.7
113.6
109.3
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 2020, 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 169 that report to Subpart TT (Industrial Waste Landfills) of the GHGRP had an active gas collection
and control system during the year 2020 (EPA 2021a).
Note: Totals may not sum due to independent rounding. Parentheses indicate negative values.
Table 7-4: ChU Emissions from Landfills (kt)
Activity
1990
2005
2016
2017
2018
2019
2020
MSW CH4 Generation3
8,214
10,845
11,321
11,672
11,878
12,186
12,193
Industrial CH4 Generation
484
638
666
667
668
669
672
MSW CH4 Recovered3
(851)
(5,301)
(6,637)
(6,884)
(6,965)
(7,182)
(7,362)
MSW CH4 Oxidized3
(736)
(856)
(965)
(1,020)
(1,046)
(1,061)
(1,063)
Industrial CH4 Oxidized
(48)
(64)
(67)
(67)
(67)
(67)
(67)
MSW net CH4 Emissions
6,627
4,687
3,719
3,768
3,867
3,943
3,768
Industrial net CH4 Emissions'5
436
575
599
600
601
602
605
Total
7,063
5,262
4,318
4,368
4,467
4,545
4,373
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 2020, 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 169 that report to Subpart TT (Industrial Waste Landfills) of the GHGRP had an active gas
collection and control system during the year 2020 (EPA 2021a).
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values.
Methodology and Time-Series Consistency
Methodology Applied for MSW Landfills
A combination of IPCC Tier 2 and 3 approaches (IPCC 2006) are used to calculate emissions from MSW Landfills.
Waste 7-7
-------
Methane emissions from landfills are estimated using two primary methods. The first method uses the first order
decay (FOD) model as described by the 2006IPCC Guidelines to estimate Cm generation. The amount of Cm
recovered and combusted from MSW landfills is subtracted from the Cm generation and is then adjusted with an
oxidation factor. The oxidation factor represents the amount of Cm in a landfill that is oxidized to CO2 as it passes
through the landfill cover (e.g., soil, clay, geomembrane). This method is presented below and is similar to
Equation HH-6 in 40 CFR Part 98.343 for MSW landfills, and Equation TT-6 in 40 CFR Part 98.463 for industrial
waste landfills.
Equation 7-1: Landfill Methane Generation
CH4,msw= (GCH4 - Rn) * (1 - OX)
where,
Cm,msw = Net CH4 emissions from solid waste
Gch4,msw = CH4 generation from MSW landfills, using emission factors for DOC, k, MCF, F from IPCC
(2006) and other peer-reviewed sources
R = CH4 recovered and combusted
Ox = CH4 oxidized from MSW landfills before release to the atmosphere, using Ox values from
IPCC (2006) and other peer-reviewed or scientifically validated literature (40 CFR Part 98)
The second method used to calculate CH4 emissions from landfills, also called the back-calculation method, is
based on directly measured amounts of recovered CH4 from the landfill gas and is expressed below and by
Equation HH-8 in 40 CFR Part 98.343. The two parts of the equation consider the portion of CFUin the landfill gas
that is not collected by the landfill gas collection system, and the portion that is collected. First, the recovered CH4
is adjusted with the collection efficiency of the gas collection and control system and the fraction of hours the
recovery system operated in the calendar year. This quantity represents the amount of CH4 in the landfill gas that is
not captured by the collection system; this amount is then adjusted for oxidation. The second portion of the
equation adjusts the portion of CH4 in the collected landfill gas with the efficiency of the destruction device(s), and
the fraction of hours the destruction device(s) operated during the year.
Equation 7-2: Net Methane Emissions from MSW Landfills
ch4 ,Solid Waste — \i~ " R \ X (1 - OX) + R x (l - (DE x fDest))]
\CE X f REC J v J
where,
CH4 ,solid waste — Net CH4 emissions from solid waste
R = Quantity of recovered CH4 from Equation HH-4 of EPA's GHGRP
CE = Collection efficiency estimated at the landfill, considering system coverage, operation, and
cover system materials from Table HH-3 of EPA's GHGRP. If area by soil cover type
information is not available, the default value of 0.75 should be used (percent)
fREc = fraction of hours the recovery system was operating (percent)
OX = oxidation factor (percent)
DE = destruction efficiency (percent)
fDest = fraction of hours the destruction device was operating (fraction)
The current Inventory uses both methods to estimate CH4 emissions across the time series within EPA's Waste
Model, as summarized in Figure 7-3 below. This chapter provides a summary of the methods, activity data, and
parameters used. Additional step-wise explanations to generate the net emissions are provided in Annex 3.14.
7-8 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
fourCH4 recovery
databases
• Back-calculated CH4
generation 3
• 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 present.
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 through 2020 Inventory estimates for MSW
landfills is as follows and is also illustrated in Annex Figure A-19:
• 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 CH4 generation, estimates for those years were
Waste 7-9
-------
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
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 2020: 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 current (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).
7-10 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Additional revisions made included incorporating facilities that have stopped reporting to the GHGRP,
new additions to the 2021 LMOP Database (EPA 2021b), corrections to the underlying database of non-
reporting landfills used to develop the 9 percent scale-up factor that were identified. For 2017 to 2020, a
scale-up factor of 11 percent is applied annually to the GHGRP net reported Cm 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 0.21 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.
The FOD equation from IPCC 2006 is used via the waste model to estimate methane emissions:
Equation 7-3: Net Methane Emissions from Industrial Waste Landfills
CH4.IND = (GCH4 - * (1 - OX)
where,
CH4 ,solid waste — Net CH4 emissions from solid waste
Waste 7-11
-------
GcH4,ind = Cm generation from industrial waste landfills, using production data multiplied by a
disposal factor and emission factors for DOC, k, MCF, F (IPCC 2006)
R = CH4 recovered and combusted (no recovery is assumed for industrial waste landfills)
OX = CH4 oxidized from industrial waste landfills before release to the atmosphere (using the
2006 IPCC Guidelines value for OX of 0.10)
The activity data used in the emission calculations are production data (e.g., the amount of meat, poultry,
vegetables processed; the amount of paper produced) versus disposal data. There are currently no facility-specific
data sources that track and report the amount and type of waste disposed of in the universe of industrial waste
landfills in the United States. EPA's GHGRP provides some insight into waste disposal in industrial waste landfills
but is not comprehensive. Data reported to the GHGRP on industrial waste landfills suggests that most of the
organic waste which would result in methane emissions is disposed at pulp and paper and food processing
facilities. Of the 168 facilities that reported to Subpart TT of the GHGRP in 2019, 92 (54 percent) are in the North
American Industrial Classification System (NAICS) for Pulp, Paper, and Wood Products (NAICS 321 and 322) and 12
(7 percent) are in Food Manufacturing (NAICS 311).
Based on this limited information, the Inventory methodology assumes most of the organic waste placed in
industrial waste landfills originates from the food processing (meat, vegetables, fruits) and pulp and paper sectors,
thus estimates of industrial landfill emissions focused on these two sectors. EPA validated this assumption through
an analysis of the Subpart TT of the GHGRP in the 2016 reporting year (RTI 2018b). The Subpart TT waste disposal
information for pulp and paper facilities correlates well with the activity data currently used to estimate Inventory
emissions; however, the waste disposal information in Subpart TT related to food and beverage facilities are
approximately an order of magnitude different than the Inventory disposal estimates for the entire time series.
EPA conducted a literature review in 2020 to investigate other sources of industrial food waste, which is briefly
described in the Planned Improvements section, and decided to maintain the currently used methodology for the
previous 1990 through 2019 Inventory due to questions around data availability across the 1990 to 2019 time
series, the completeness and representativeness of other estimates and methodologies, and the level of effort
required to reproduce and/or merge estimates across the 1990 to 2019 time series.
The composition of waste disposed of in industrial waste landfills is expected to be more consistent in terms of
composition and quantity than that disposed of in MSW landfills. The amount of waste landfilled is assumed to be
a fraction of production that is held constant over the time series as explained in Annex 3.14.
Landfill CH4 recovery is not accounted for in industrial waste landfills. Data collected through EPA's GHGRP for
industrial waste landfills (Subpart TT) show that only one of the 166 facilities, or 1 percent of facilities, have active
gas collection systems (EPA 2021a). However, because EPA's GHGRP is not a national database and comprehensive
data regarding gas collection systems have not been published for industrial waste landfills, assumptions regarding
a percentage of landfill gas collection systems, or a total annual amount of landfill gas collected for the non-
reporting industrial waste landfills have not been made for the Inventory methodology.
The amount of CH4 oxidized by the landfill cover at industrial waste landfills was assumed to be 10 percent of the
CH4 generated (IPCC 2006; Mancinelli and McKay 1985; Czepiel et al. 1996) for all years.
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 (and later land applied), 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,
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):
7-12 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
• 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 reports use a materials flow
methodology, commonly referred to as a 'top-down' methodology, which relies heavily on a mass balance
approach. Data are gathered from industry associations, key businesses, similar industry sources, and
government agencies (e.g., the Department of Commerce and the U.S. Census Bureau) and are used to estimate
tons of materials and products generated, recycled, combusted with energy recovery, other food management
pathways, or landfilled nationwide. The amount of MSW generated is estimated by estimating production and
then adjusting these values by addressing the imports and exports of produced materials to other countries.
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 (U.S. EPA 2020) uses a methodology that expanded the number of management pathways to
include: animal feed; bio-based materials/biochemical processing (i.e., rendering); codigestion/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
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 Cm
Waste 7-13
-------
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 2020, 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 2020. 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
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 CH4 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 CH4 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 CH4 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 CH4 gas recovery; however, the number of
unique landfills between the four databases does differ.
7-14 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 CFU 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
2020 Emission
Source
Gas
Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT CO? Eq.)
(MMT CO?
Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Total Landfills
ch4
109.3
84.7
133.3
-23%
22%
MSW
ch4
94.2
73.7
123.0
-22%
31%
Industrial
ch4
15.1
10.4
18.9
-31%
25%
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.
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 2006 IPCC 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;
Waste 7-15
-------
• 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.2 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.
For the current (1990 to 2020) Inventory, the scale-up factor methodology was revised from a total waste-in-place
comparison to a time-based threshold (50-years) for waste disposed between landfills that do and do not report to
Subpart HH of the GHGRP. An 11 percent scale-up factor resulted from the time-based threshold (between 1970 to
2020). This value is the same scale-up factor as that calculated when considering the total waste-in-place scale-up
factor incorporated into the previous 1990 through 2019 Inventory for 2017 to 2019. Additional details about the
scale-up factor are presented in Annex 3.14.
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 CH4
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
2015 to 2019 slightly increased or decreased total Subpart HH reported net emissions by ± 0.3 percent in the years
the Subpart HH data are applied (i.e., 2005 to 2019). No Subpart HH reports were resubmitted for the 2010 to
2014 reporting years that resulted in net emission changes. These changes resulted in annual increases of less than
2 See https://www.epa.gov/sites/production/files/2015-07/documents/eherp verification factsheet.pdf.
7-16 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
0.2 percent to the net Inventory emissions between 2005 to 2009, and annual decreases of less than 1 percent per
year to the net Inventory emissions between 2015 to 2019. A change in net Subpart HH reported emissions results
in the same percentage change in the Inventory emissions for that year.
The scale-up factor was also reassessed as a planned improvement for the current (1990 to 2020) Inventory.
Results from this effort did not change the 11 percent scale-up factor that was applied between 2017 to 2019 in
the current 1990 to 2020 inventory.
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 is
developing 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 future GHGRP rulemaking 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 2020, 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 next (1990 to 2021) inventory.
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). As was the case with this Inventory, if warranted, EPA will revise the scale-up factor to reflect
newly acquired information to ensure completeness of the Inventory. For this 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.
EPA is planning to account for unmanaged landfills in Puerto Rico and other U.S. Territories to the landfill
emissions estimates. Data limitations for the history of waste received at these sites make this challenging.
Presently emissions from managed sites in Puerto Rico and Guam are accounted for years 2005 to present as they
are reporters to GHGRP Subpart HH.
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
2021) Inventory cycle per the 2019 Refinement.
Waste 7-17
-------
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
MSW to WTE
12%
Other Food
Management
6%
Composted
8%
Landfilled
50%
Recycled
24%
Source: EPA (2020b)
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).
Figure 7-5: MSW Management Trends from 1990 to 2018
150
Landfilling
\
,/'\ —•-
* *'
140
120
100
80
50
40 _ Energy Recovery
20 — — — — —
_ — — — "" Comjjosting
Recycling
Combustion with
Other Food Management
(2018 only)
$ 4~ $ $ $ $ $ 4 $ 4> 4' & $ $ $ $ 4? 4^ 4'' 4* 4* 4^ 4~ 4? 4* 4 4*
Recycling — — Composting -Other Food Management Combustion with Energy Recovery — — — Landfiling
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,
7-18 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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.3
Note: 2018 is the latest year of available data. Only one year of data (2018) is available for the 'Other Food Management'
category.
Table 7-6 presents the average composition of waste disposed in a typical MSW landfill in the United States
overtime. It is important to note that the actual composition of waste entering each landfill will vary from
that presented in Table 7-6. Due to China's recent ban on accepting certain kinds of solid waste by the end of
2017 (WTO 2017), inclusive of some paper and paperboard waste, plastic waste, and other miscellaneous
inorganic wastes, there has been a slight increase in the disposal of paper and paperboard and plastic wastes
since 2017 (Table 7-6). Future impacts of China's recent waste ban to the composition of waste disposed in
U.S. landfills are unknown at this time.
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 CFU generation potential and CFU 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.
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)
3 See https://www.epa.gov/facts-and-figures-about-materials-waste-and-recvcling/advancing-sustainable-materials-
management.
Waste 7-19
-------
Figure 7-6: Percent of Degradable Materials Diverted from Landfills from 1990 to 2018
(Percent)
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 web site.4
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.
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.5 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 18
4 See https://www.epa.gov/facts-and-figures-about-materials-waste-and-recvcling/advancing-sustainable-materials-
management.
5 Throughout the Inventory, emissions from domestic wastewater also include any commercial and industrial wastewater
collected and co-treated with domestic wastewater.
7-20 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 it 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-
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 CH4 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 CH4 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 addition of organic matter or nitrogen from
wastewater discharges is generally expected to increase CH4 and N2O emissions from these 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.
Waste 7-21
-------
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.
In 2020, CH4 emissions from domestic wastewater treatment and discharge were estimated to be 10.0 MMT CO2
Eq. (400 kt CH4) and 1.8 MMT CO2 Eq. (73 kt CH4), respectively. 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 2020, CH4 emissions from industrial
wastewater treatment and discharge were estimated to be 5.9 MMT CO2 Eq. (238 kt CH4) and 0.5 MMT CO2 Eq. (20
kt CH4), respectively. Industrial emissions from wastewater treatment have generally increased across the time
series through 1999 and then fluctuated up and corresponding 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. However, in 2020, some industries (including starch-based ethanol production and petroleum refining),
experienced a more significant decrease in production (and, by extension, emissions) due to the COVID-
19 pandemic (see Table 7-20). Table 7-7 and Table 7-8 provide CH4 emission estimates from domestic and
industrial wastewater treatment.
With respect to N2O, emissions from domestic wastewater treatment and discharge in 2020 were estimated to be
18.1 MMT CO2 Eq. (61 kt N2O) and 4.9 MMT CO2 Eq. (16 kt N2O), respectively. Total N2O emissions from domestic
wastewater were estimated to be 23.0 MMT CO2 Eq. (77 kt N2O). Nitrous oxide emissions from wastewater
treatment processes gradually increased across the time series as a result of increasing U.S. population and protein
consumption. In 2020, N2O emissions from industrial wastewater treatment and discharge were estimated to be
0.4 MMT CO2 Eq. (1 kt N2O) and 0.1 MMT CO2 Eq. (0.3 kt N2O), respectively. 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,
though see the CH4 discussion above regarding 2020 industrial production. 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
2016
2017
2018
2019
2020
ch4
20.3
20.1
18.7
18.5
18.3
18.1
18.3
Domestic Treatment
13.5
12.9
10.8
10.5
10.2
9.8
10.0
Domestic Effluent
1.2
1.2
1.8
1.8
1.8
1.8
1.8
Industrial Treatment3
4.9
5.4
5.6
5.7
5.8
5.9
5.9
Industrial Effluent3
0.7
0.6
0.5
0.5
0.5
0.5
0.5
n2o
16.6
20.3
22.8
23.2
23.5
23.4
23.5
7-22 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Domestic Treatment
11.8
15.4
17.3
17.7
18.0
18.0
18.1
Domestic Effluent
4.4
4.4
5.0
5.0
5.0
4.9
4.9
Industrial Treatment15
0.3
0.4
0.4
0.4
0.4
0.5
0.4
Industrial Effluentb
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Total
36.9
40.5
41.5
41.7
41.8
41.5
41.8
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
2016
2017
2018
2019
2020
ch4
812
806
748
740
732
723
730
Domestic Treatment
540
518
434
420
407
393
400
Domestic Effluent
49
49
71
72
73
73
73
Industrial Treatment3
196
216
223
228
232
237
238
Industrial Effluent3
27
22
20
20
20
20
20
n2o
56
68
76
78
79
79
79
Domestic T reatment
40
52
58
59
60
60
61
Domestic Effluent
15
15
17
17
17
16
16
Industrial Treatment15
1
1
1
1
2
2
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 may not sum due to independent rounding.
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.6 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.
6IPCC (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.
Waste 7-23
-------
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2020. In the following cases, the source used to capture activity data changed over the time series. EPA
recognizes this may appear to lower data consistency; however, believes the result is more accurate. 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 2018-
2020. 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, 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 Cm 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);
• Centralized anaerobic systems (C);
• Anaerobic sludge digesters (D); and
• Centralized wastewater treatment effluent (E).
Methodological equations for each of these systems are presented in the subsequent subsections; total domestic
Cm emissions are estimated as follows:
Equation 7-4: Total Domestic ChU Emissions from Wastewater Treatment and Discharge
Total Domestic CH4 Emissions from Wastewater Treatment and Discharge (kt) = A+ B + C + D + E
Table 7-9 presents domestic wastewater CH4 emissions for both septic and centralized systems, including
anaerobic sludge digesters and emissions from centralized wastewater treatment effluent, in 2020.
Table 7-9: Domestic Wastewater ChU Emissions from Septic and Centralized Systems (2020,
kt, MMT CO2 Eq. and Percent)
CH4 Emissions
% of Domestic
CH4 Emissions (kt)
(MMT CO? Eq.)
Wastewater CH4
Septic Systems (A)
227
5.7
48.0
Centrally-Treated Aerobic Systems (B)
37
0.9
7.8
Centrally-Treated Anaerobic Systems (C)
127
3.2
26.9
Anaerobic Sludge Digesters (D)
8
0.2
1.7
Centrally-Treated Wastewater Effluent (E)
73
1.8
15.5
Total
473
11.8
100
7-24 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Emissions from Septic Systems:
Methane 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 2020; U.S. Census Bureau 2019) and an emission factor
and then converting the result to kt/year.
U.S. population data were taken from the U.S. Census Bureau International Database (U.S. Census Bureau 2021)
and include the populations of the United States, American Samoa, Guam, Northern Mariana Islands, Puerto Rico,
and the U.S. Virgin Islands. Table 7-12 presents U.S. population for 1990 through 2020. The fraction of the U.S.
population using septic systems or centralized treatment systems is based on data from the American Housing
Surveys (U.S. Census Bureau 2019). Methane emissions for septic systems are estimated as follows:
Equation 7-5: CH4 Emissions from Septic Systems
Emissions from Septic Systems (U.S. Specific) = A
= USpop X (Tseptic) X (EFseptic) X 1/109 X 365.25
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
1990-2020: U.S. Census
Bureau (2021)
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: Forecasted from the
rest of the time series
EFseptic
Methane emission factor - septic systems
(10.7)
g CH4/capita/day
1990-2020: Leverenz et al.
(2010)
1/109
Conversion factor
g to kt
Standard conversion
365.25
Conversion factor
Days in a year
Standard conversion
a Value of activity data varies over the Inventory time series.
Emissions from Centrally Treated Aerobic and Anaerobic Systems:
Methane emissions from POTWs depend on the total organics in wastewater. Table 7-12 presents total BODs
produced (also referred to as the total organically degradable material in wastewater or TOW) for 1990 through
2020. The BODs production rate was determined using BOD generation rates per capita weighted average both
with and without kitchen scraps as well as an estimated percent of housing units that utilize kitchen garbage
disposals. Households with garbage disposals (with kitchen scraps or ground up food scraps) typically have
wastewater with higher BOD than households without garbage disposals due to increased organic matter
contributions (ERG 2018a). The equations are as follows:
Equation 7-6: Total Wastewater BODs Produced per Capita (U.S.-Specific [ERG 2018a])
BODgenrate (kg/C<] p i t <1/ cl<]V') = BODwithout scrap X (1 - lX)(-1 iSpOS<11) -|- BODwith scraps X (lX)cliSp0S<11)
Equation 7-7: Total Organically Degradable Material in Domestic Wastewater (IPCC 2019
[Eq. 6.3])
TOW (Gg/year) = USpop X BODgenrate X 365.25
Waste 7-25
-------
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
1990-2020: Calculated
BODwithout scrap
Wastewater BOD produced per capita
without kitchen scraps3
kg/capita/day
1990-2003: Metcalf & Eddy
(2003)
2004-2013: Linear
interpolation
2014-2020: Metcalf & Eddy
(2014)
BODwith scraps
Wastewater BOD produced per capita
with kitchen scraps3
kg/capita/day
% disposal
Percent of housing units with kitchen
disposal3
%
1990-2013: U.S. Census
Bureau (2013)
2014-2020: Forecasted from
the rest of the time series
TOW
Total wastewater BOD Produced per
Capita3
Gg BOD/year
1990-2020: Calculated, ERG
(2018a)
USpop
U.S. population3
Persons
1990-2020: U.S. Census
Bureau (2021)
365.25
Conversion factor
Days in a year
Standard conversion
a Value of activity data varies over the Inventory time series.
Table 7-12: U.S. Population (Millions) and Domestic Wastewater BODs Produced (kt)
Activity
1990
2005
2016
2017
2018
2019
2020
Population
253
300
327
329
330
332
336
TOW
8,131
Vi 9,624
9,816
9,891
9,956
10,017
10,164
Sources: U.S. Census Bureau (2021); 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 2020), 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 factor7 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 sludge8 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
methodology (IPCC 2019) 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
judgment), primary treatment with anaerobic sludge digestion, or secondary treatment without primary
treatment. The organic component removed from anaerobic wastewater treatment and the amount of CH4
7 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)
and constructed wetlands (0.4) (IPCC 2014).
8 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.
7-26 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
recovered or flared from both aerobic and anaerobic wastewater treatment were set equal to the IPCC default of
zero.
The methodological equations for Cm emissions from aerobic and anaerobic systems are:
Equation 7-8: Total Domestic ChU 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
where,
Equation 7-9: Total Organics in Centralized Wastewater Treatment [IPCC 2019 (Eq. 6.3A)]
TOWcentralized (Gg BOD/year) = TOW X Tcentralized X Icollected
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)
TOWcentralized
Total organics in centralized
wastewater treatment3
Gg BOD/year
1990-2020: Calculated
TOW
Total wastewater BOD Produced per
Capita3
Gg BOD/year
1990-2020: Calculated, ERG (2018a)
Tcentralized
Percent collected3
%
1990-2019: U.S. Census Bureau (2019)
Data for intervening years obtained by
linear interpolation
2020: Forecasted from the rest of the
time series
'collected
Correction factor for additional
industrial BOD discharged (1.25)
No units
1990-2020: IPCC (2019)
a Value of this activity data varies over the time series.
Equation 7-10: Organic Component Removed from Aerobic Wastewater Treatment (IPCC
2019 [Eq. 6.3B])
Saerobic (Gg/year) = Smass x [(% aerobic w/primary x Krem,aer_Prim) + (% aerobic w/out primary x Krem,aer_noprim)
+ (%aerobiC+digeStion X Krem,aer_digest)] X 1000
Equation 7-11: Emissions from Centrally Treated Aerobic Systems (other than Constructed
Wetlands) (IPCC 2019 [Eq. 6.1])
Bl (ktCH4/year)
= [(TOWcentralized) X (% aerobicoTcw) - Saerobic] X EFaerobic - Raerobic
Table 7-14: Variables and Data Sources for ChU 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 CH4/year)
Saerobic
Organic component removed from
aerobic wastewater treatment3
Gg BOD/year
1990-2020: Calculated
Smass
Raw sludge removed from wastewater
treatment as dry mass3
Tg dry
weight/year
1988: EPA (1993c); EPA (1999)
1990-1995: Calculated based on
sewage sludge production change
per year EPA (1993c); EPA (1999);
Beecher et al. (2007)
Waste 7-27
-------
Variable
Variable Description
Units
Inventory Years: Source of Value
1996: EPA (1999)
2004: Beecher et al. (2007)
Data for intervening years obtained
by linear interpolation
2005-2020: Forecasted from the rest
of the time series
% 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.
2005-2020: 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
1990-2020: IPCC (2019)
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
1000
Conversion factor
metric tons to
kilograms
EF aerobic
Emission factor - aerobic systems
(0.018)
kg CH4/kg BOD
Raerobic
Amount CH4 recovered or flared from
aerobic wastewater treatment (0)
kg CH4/year
a Value of this activity data varies over the time series.
Constructed wetlands exhibit both aerobic and anaerobic treatment (partially anaerobic treatment) but are
referred to in this chapter as aerobic systems. Constructed wetlands may be used as the sole treatment unit at a
centralized wastewater treatment plant or may serve as tertiary treatment after simple settling and biological
treatment. Emissions from all constructed wetland systems were included in the estimates of emissions from
centralized wastewater treatment plant processes and effluent from these plants. Methane emissions equations
from constructed wetlands used as sole treatment were previously described. 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.
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.
7-28 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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
1990-2020: Calculated
% aerobiccw
Flow to aerobic systems,
constructed wetlands used as sole
treatment / total flow to POTWs.a
%
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.
2013-2020: Forecasted from the rest
of the time series
EFcw
Emission factor for constructed
wetlands (0.24)
kg CH4/kg
BOD
1990-2020: 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.
2013-2020: Forecasted from the rest
of the time series
BODcw.inf
BOD concentration in wastewater
entering the constructed wetland
(9.1)
mg/L
1990-2020: EPA (2013)
3.785
Conversion factor
liters to
gallons
Standard conversion
EFcw
Emission factor for constructed
wetlands (0.24)
kg CH4/kg
BOD
1990-2020: 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.
Data sources and methodologies for centrally treated anaerobic systems are similar to those described for aerobic
systems, other than constructed wetlands. See discussion above.
Equation 7-14: Emissions from Centrally Treated Anaerobic Systems [IPCC 2019 (Eq. 6.1)]
C (ktCH4/year)
— [(TOWcENTRALIZEd) X (% anaerobic) - Sanaerobic] X EFanaerobic - Ranaerobic
Waste 7-29
-------
Table 7-16: Variables and Data Sources for ChU Emissions from Centrally Treated Anaerobic
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
1990-2020: 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.
2005-2020: Forecasted from the
rest of the time series
Sanaerobic
Organic component removed from
anaerobic wastewater treatment (0)
Gg/year
1990-2020: IPCC(2019)
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
a Value of this activity data varies over the time series.
Emissions from Anaerobic Sludge Digesters:
Total Cm emissions from anaerobic sludge digesters were estimated by multiplying the wastewater influent flow
to POTWs with anaerobic sludge digesters, the cubic feet of digester gas generated per person per day divided by
the flow to POTWs, the fraction of CH4 in biogas, the density of CH4, one minus the destruction efficiency from
burning the biogas in an energy/thermal device and then converting the results to kt/year.
Equation 7-15: Emissions from Anaerobic Sludge Digesters (U.S. Specific)
D (ktCH4/year)
= [(POTW_flow_AD) x (biogas gen)/(100)] x 0.0283 x (FRAC_CH4) x 365.25 x (662) x (1-DE) x 1/109
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.
2005-2020: Forecasted from the
rest of the time series
biogas gen
Gas Generation Rate (1.0)
ft3/ca pita/day
1990-2020: Metcalf & Eddy
(2014)
100
Per Capita POTW Flow (100)
gal/capita/day
1990-2020: Ten-State Standards
(2004)
0.0283
Conversion factor
ft3 to m3
Standard Conversion
frac_ch4
Proportion of Methane in Biogas (0.65)
No units
1990-2020: Metcalf & Eddy
(2014)
365.25
Conversion factor
Days in a year
Standard conversion
662
Density of Methane (662)
g CH4/m3 CH4
1990-2020: EPA (1993a)
7-30 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Variable
Variable Description
Units
Inventory years: Source of
Value
DE
Destruction Efficiency (99% converted
to fraction)
No units
1990-2020: EPA (1998b); CAR
(2011); Sullivan (2007); Sullivan
(2010); and UNFCCC (2012)
1/109
Conversion factor
g to kt
Standard conversion
a Value of this activity data varies over the time series.
Emissions from Discharge of Centralized Treatment Effluent:
Methane emissions from the discharge of 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.
The BOD in treated effluent was determined by multiplying the total organics in centrally treated wastewater by
the percent of wastewater treated in primary, secondary, and tertiary treatment, and the fraction of organics
remaining after primary treatment (one minus the fraction of organics removed from primary treatment,
secondary treatment, and tertiary treatment).
Equation 7-16: Emissions from Centrally Treated Systems Discharge (U.S.-Specific)
E (kt CH4/year)
= (TOWrlE X EFrle) + (TOWother X EFother)
where,
Equation 7-17: Total Organics in Centralized Treatment Effluent (IPCC 2019 [Eq. 6.3D])
TOWEFFtreat.cENTRALizED (Gg BOD/year)
= [TOWcentralized X % primary X (l-TOWrem,PRiMARY)] + [TOWcentralized X % secondary X (1-
TOWrem,SECONDARY)] + [TOWCENTRALIZED X % tertiary X (l-TOWrem,TERTIARY)]
Equation 7-18: Total Organics in Effluent Discharged to Reservoirs, Lakes, or Estuaries (U.S.-
Specific)
TOWrle (Gg BOD/year)
= TOWEFFtreat.CENTRALIZED X PerCentRLE
Equation 7-19: Total Organics in Effluent Discharged to Other Waterbodies (U.S.-Specific)
TOWother (Gg BOD/year)
= TOWEFFtreat.CENTRALIZED X PerCentother
Table 7-18: Variables and Data Sources for ChU Emissions from Centrally Treated Systems
Discharge
Variable
Variable Description
Units
Source of Value
TOW EFFtreat.CENTRALIZED
Total organics in centralized treatment effluent3
Gg
BOD/year
1990-2020:
Calculated
TOWcentralized
Total organics in centralized wastewater treatment3
Gg
BOD/year
1990-2020:
Calculated
% 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
% secondary
Percent of secondary domestic centralized treatment3
%
% tertiary
Percent of tertiary domestic centralized treatment3
%
Waste 7-31
-------
Variable
Variable Description
Units
Source of Value
obtained by linear
interpolation.
2013-2020:
Forecasted from
the rest of the time
series
TOWrem.PRIMARY
Fraction of organics removed from primary domestic
centralized treatment (0.4)
No units
1990-2020: IPCC
(2019)
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
1990-2020:
Calculated
TOWother
Total organics in effluent discharge to other waterbodies3
Gg
BOD/year
EFrle
Emission factor (discharge to reservoirs/lakes/estuaries)
(0.021)
kg CH4/kg
BOD
1990-2020: IPCC
(2019)
EFother
Emission factor (discharge to other waterbodies) (0.114)
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)
2013-2020: Set
equal to 2012
Percentother
% discharged to other waterbodies3
%
a Value of this activity data varies over the time series.
Industrial Wastewater CH4 Emission Estimates
Industrial wastewater Cm emissions originate from on-site treatment systems, typically comprised of biological
treatment operations. The collection systems at an industrial plant are not as extensive as domestic wastewater
sewer systems; therefore, it is not expected that dissolved Cm will form during collection. However, some
treatment systems are designed to have anaerobic activity (e.g., anaerobic reactors or lagoons), or may
periodically have anaerobic conditions form (facultative lagoons or large stabilization basins). Emissions will also
result from discharge of treated effluent to waterbodies where carbon accumulates in sediments (typically slow-
moving systems, such as lakes, reservoirs, and estuaries).
Industry categories that are likely to produce significant Cm emissions from wastewater treatment were identified
and included in the Inventory. The main criteria used to identify U.S. industries likely to generate Cm are whether
they generate high volumes of wastewater, whether there is a high organic wastewater load, and whether the
wastewater is treated using methods that result in Cm emissions. The top six industries that meet these criteria
are pulp and paper manufacturing; meat and poultry processing; vegetables, fruits, and juices processing; starch-
based ethanol production; petroleum refining; and breweries. Wastewater treatment and discharge emissions for
these sectors for 2020 are displayed in Table 7-19 below. Further discussion of wastewater treatment for each
industry is included below.
7-32 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 7-19: Total Industrial Wastewater ChU Emissions by Sector (2020, MMT CO2 Eq. and
Percent)
CH4 Emissions
% of Industrial
Industry
(MMT CO? Eq.)
Wastewater CH4
Meat & Poultry
5.0
78.2
Pulp & Paper
0.8
12.2
Fruit & Vegetables
0.2
3.6
Ethanol Refineries
0.1
2.3
Breweries
0.1
2.1
Petroleum Refineries
0.1
1.5
Total
6.4
100
Note: Totals may not sum due to independent rounding.
Emissions from Industrial Wastewater Treatment Systems:
Equation 7-20 presents the general IPCC equation (Equation 6.4, IPCC 2019) to estimate methane emissions from
each type of treatment system used for each industrial category.
Equation 7-20: Total ChU Emissions from Industrial Wastewater
CH4 (industrial sector) = [(TOWi - Si) x EF -Ri]
where,
Cm (industrial sector) = Total CH4 emissions from industrial sector wastewater treatment (kg/year)
i = Industrial sector
TOWi = Total organics in wastewater for industrial sector / (kg COD/year)
Si = Organic component removed from aerobic wastewater treatment for industrial
sector / (kg COD/year)
EF = System-specific emission factor (kg Cm/kg COD)
Ri = Methane recovered for industrial sector I (kg Cl-U/year)
Equation 7-21 presents the general IPCC equation to estimate the total organics in wastewater (TOW) for each
industrial category.
Equation 7-21: TOW in Industry Wastewater Treatment Systems
TOWi = Pi x Wi x CODi
where,
TOWi = Total organically degradable material in wastewater for industry I (kg COD/yr)
i = Industrial sector
Pi = Total industrial product for industrial sector / (t/yr)
Wi = Wastewater generated (m3/t product)
CODi = Chemical oxygen demand (industrial degradable organic component in wastewater) (kg
COD/m3)
The annual industry production is shown in Table 7-20, and the average wastewater outflow and the organics
loading in the outflow is shown in Table 7-21. 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 CH4 emissions, wastewater
Waste 7-33
-------
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
where,
Spulp,asb
TOWpuip
% removal w/primary
Spuip.asb = TOWpuip x % removal w/primary
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
COD/yr)
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
where,
Saerobic
Smass
Krem
10-6
Saerobic — Smass X Krem XlO"1
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
Equation 7-24: Raw Sludge Removed from Wastewater Treatment as Dry Mass
where,
Smass
Sprim
Saer
P
w
Smass — (Sprim + Saer ) x PxW
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)
Default emission factors9 from IPCC (2019) were used. Information on methane recovery operations varied by
industry. See industry descriptions below.
9 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).
7-34 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 7-20: U.S. Pulp and Paper, Meat, Poultry, Vegetables, Fruits and Juices, Ethanol,
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
38.7
2.5
23.9
702.4
2005
92.4
31.4
25.1
42.9
11.7
23.1
818.6
2016
79.9
34.2
28.3
43.5
45.8
22.3
926.0
2017
80.3
35.4
28.9
42.9
47.2
21.8
933.5
2018
78.7
36.4
29.4
42.4
48.1
21.5
951.7
2019
76.3
37.4
30.1
43.5
47.1
21.1
940.0
2020
76.1
37.8
30.5
43.5
41.6
21.1
806.7
a Pulp and paper production is the sum of market pulp production plus paper and paperboard production.
Sources: Pulp and Paper - FAO (2021a) and FAO (2021b); Meat, Poultry, and Vegetables - USDA (2021a and 2021c);
Ethanol - Cooper (2018) and RFA (2021a and 2021b); Breweries - Beer Institute (2011) and TTB (2021); Petroleum
Refining - EIA (2021).
Table 7-21: U.S. Industrial Wastewater Characteristics Data (2020)
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).
Waste 7-35
-------
Table 7-22: U.S. Industrial Wastewater Treatment Activity Data
% Treated Aerobically
Industry
% Wastewater
% Treated
% Treated
% Treated in
Treated On Site
Anaerobically
Aerobically
% Treated in
ASBs
Other
Aerobic
Pulp and Paper
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
0
0
0
Ethanol Production - Dry Mill
75
75
0
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
a Wastewater is pretreated in anaerobic lagoons prior to aerobic treatment.
Note: Due to differences in data availability and methodology, zero values in the table are for calculation purposes only and
may indicate unavailable data.
Sources: ERG (2008b); ERG (2013a); ERG (2013b); ERG (2021a).
Table 7-23: Sludge Variables for Aerobic Treatment Systems
Variable
Pulp and
Paper
Industry
Fruit/Vegetable
Processing
Petroleum
Refining
Organic reduction associated with sludge removal (%)
Sludge Production (kg/m3)
Primary Sedimentation
Aerobic Treatment
Sludge Factor (kg BOD/kg dry mass sludge)
Aerobic Treatment w/Primary Sedimentation and No Anaerobic
Sludge Digestion
Aerobic Treatment w/out Primary Sedimentation
58
0.15
0.096
0.8
0.096
1.16
Sources: Organic reduction (pulp) - ERG (2008a); Sludge production - Metcalf & Eddy (2003); Sludge factors - IPCC (2019).
Emissions from Discharge of Industrial Wastewater Treatment Effluent:
Methane emissions from discharge of industrial wastewater treatment effluent are estimated via a Tier 1 method
for all industries except for pulp, paper, and paperboard. Emissions from discharge of pulp, paper, and paperboard
treatment effluent is estimated via a Tier 2 method and is described in the industry-specific data section. Tier 1
emissions from effluent are estimated by multiplying the total organic content of the discharged wastewater
effluent by an emission factor associated with the discharge:
Equation 7-25: ChU Emissions from Industrial Wastewater Treatment Discharge
CH4 EffluentiND = TOWeffluent,ind X EFeffluent
where,
Cm EffluentiND =
TOWeffluent.ind =
EFeffluent =
CH4 emissions from industrial wastewater discharge for inventory year (kg Cm/year)
Total organically degradable material in wastewater effluent from industry for inventory
year (kg COD/year or kg BOD/year)
Tier 1 emission factor for wastewater discharged to aquatic environments (0.028 kg
Cm/kg COD or 0.068 kg CH4/kg BOD) (IPCC 2019)
The COD or BOD in industrial treated effluent (TOWeffluent.ind) was determined by multiplying the total organics in
the industry's untreated wastewater that is treated on site by an industry-specific percent removal where available
7-36 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
or a more general percent removal based on biological treatment for other industries. Table 7-22 presents the
percent of wastewater treated onsite, while Table 7-24 presents the fraction of TOW removed during treatment.
Equation 7-26: TOW in Industrial Wastewater Effluent
TOWeffluent.ind = TOWind * %onsite * (1 - TOWrem)
where,
TOWeffluent.ind = Total organically degradable material in wastewater effluent from industry for inventory
year (kg COD/year or kg BOD/year)
TOWind = Total organics in untreated wastewater for industry (kg COD/year)
%onsite = Percent of industry wastewater treated on site (%)
TOWrem = Fraction of organics removed during treatment
Table 7-24: Fraction of TOW Removed During Treatment by Industry
Industry TOWrem Source
Pulp, Paper, and Paperboard 0.905
Red Meat and Poultry 0.85
Fruits and Vegetables 0.85
Ethanol Production
Biomethanator Treatment 0.90
Other Treatment 0.85
Petroleum Refining 0.93
Breweries 0.85
Discussion of Industry-Specific Data:
Pulp, Paper, and Paperboard manufacturing Wastewater Treatment. Wastewater treatment for the pulp, paper,
and paperboard manufacturing (hereinafter referred to as "pulp and paper") industry typically includes
neutralization, screening, sedimentation, and flotation/hydrocycloning to remove solids (World Bank 1999;
Nemerow and Dasgupta 1991). Secondary treatment (storage, settling, and biological treatment) mainly consists of
lagooning. About 60 percent of pulp and paper mills have on-site treatment with primary treatment and about half
of these also have secondary treatment (ERG 2008). In the United States, primary treatment is focused on solids
removal, equalization, neutralization, and color reduction (EPA 1993b). The vast majority of pulp and paper mills
with on-site treatment systems use mechanical clarifiers to remove suspended solids from the wastewater. About
10 percent of pulp and paper mills with treatment systems use settling ponds for primary treatment and these are
more likely to be located at mills that do not perform secondary treatment (EPA 1993b).
Approximately 42 percent of the BOD passes on to secondary treatment, which consists of activated sludge,
aerated stabilization basins, or non-aerated stabilization basins. Pulp and paper mill wastewater treated using
anaerobic ponds or lagoons or unaerated ponds were classified as anaerobic (with an MCF of 0.8). Wastewater
flow treated in systems with aerated stabilization basins or facultative lagoons was classified as partially anaerobic
(with an MCF of 0.2, which is the 2006IPCC Guidelines-suggested MCF for shallow lagoons). Wastewater flow
treated in systems with activated sludge systems or similarly aerated biological systems was classified as aerobic.
A time series of CFU emissions for 1990 through 2019 was developed based on paper and paperboard production
data and market pulp production data. Market pulp production values were available directly for 1998, 2000
through 2003, and 2010 through 2018. Where market pulp data were unavailable, a percent of woodpulp that is
market pulp was applied to woodpulp production values from FAOSTAT to estimate market pulp production (FAO
2020a). 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 2019 was forecasted from the rest of the time series. A time series of the overall wastewater
outflow is presented in Table 7-25. Data for 1990 through 1994 varies based on data outlined in ERG (2013a) to
Malmberg (2018)
IPCC (2019), Table 6.6b
IPCC (2019), Table 6.6b
ERG (2008a), ERG (2006b)
IPCC (2019), Table 6.6b
Kenari, Sarrafzadeh, and Tavakoli
(2010)
IPCC (2019), Table 6.6b
Waste 7-37
-------
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),
while data for 2018 were obtained from the 2020 AF&PA Sustainability Report (AF&PA 2020). Data for intervening
years were obtained by linear interpolation, while 2019 to 2020 were set equal to 2018. 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 2020 (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
Year Outflow (m3/ton)
1990 68
2005 43
2016 40
2017 39
2018 40
2019 40
2020 40
Sources: ERG (2013a), AF&PA (2014), AF&PA
(2016), AF&PA (2018), AF&PA (2020).
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
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)
7-38 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 Cm/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
Organic Content in Untreated
Commodity Wastewater Outflow (mB/ton) Wastewater (g BOD/L)
Vegetables
Potatoes 10.27 1.765
Other Vegetables 9.88 0.752
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).
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).
Waste 7-39
-------
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 Cm/kg COD for anaerobic treatment was used to estimate the
Cm 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 2011.10 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 Cm/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 2021). 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.
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
10 Available online at https://www.epa.gov/stationarv-sources-air-pollution/comprehensive-data-collected-petroleum-refining-
sector.
7-40 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 2020.
Table 7-27: Domestic Wastewater N2O Emissions from Septic and Centralized Systems
(2020, kt, MMT CO2 Eq. and Percent)
N20 Emissions (kt)
N20 Emissions
(MMT CO? Eq.)
% of Domestic
Wastewater N20
Septic Systems
3
0.9
3.8
Centrally-Treated Aerobic Systems
58
17.2
74.8
Centrally-Treated Anaerobic Systems
0
0.0
0
Centrally-Treated Wastewater Effluent
16
4.9
21.3
Total
77
23.0
100
Note: Totals may not sum due to independent rounding.
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)
Waste 7-41
-------
= 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
1990-2020: Calculated
PrOteinper capita
Daily per capita protein supply3
g/person/day
1990-2020: 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 (2021b)
2011-2018: FAO (2021c)
and scaling factor
2019, 2020: 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 2020; 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 the U.S. Census Bureau International Database (U.S. Census Bureau 2021)
and include the populations of the United States, American Samoa, Guam, Northern Mariana Islands, Puerto Rico,
and the U.S. Virgin Islands. 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. 10])
TNdom.septic (kg N/year)
= (USPOP X TsEPTIc) X Protein X FnPR X NhH X FNON-CON_septic X FlND-COM_septic
Equation 7-34: Emissions from Septic Systems (IPCC 2019 [Eq. 6.9])
A (kt N20/year)
= TNdom. SEPTIC X (EFseptic) X 44/28 X 1/106
Table 7-29: Variables and Data Sources for N2O Emissions from Septic System
Inventory Years: Source of
Variable
Variable Description
Units
Value
Emissions from Septic Systems
TNdom septic
Total nitrogen entering septic systems
kg N/year
1990-2020: Calculated
USpop
U.S. population3
Persons
1990-2020: U.S. Census
Bureau (2021)
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: Forecasted from the
rest of the time series
Fnpr
Fraction of nitrogen in protein (0.016)
kg N/kg protein
1990-2020: IPCC (2019)
7-42 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Variable
Variable Description
Units
Inventory Years: Source of
Value
Nhh
Additional nitrogen from household products (1.17)
No units
F|ND-COM_septic
Factor for Industrial and Commercial Co-Discharged
Protein, septic systems (1)
No units
FNON-CON_septic
Factor for Non-Consumed Protein Added to
Wastewater (1.13)
No units
EFseptic
Emission factor, septic systems (0.0045)
kg N20-N/kg N
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.
Emissions from Centrally Treated Aerobic and Anaerobic Systems:
Nitrous oxide emissions from POTWs depend on the total nitrogen entering centralized wastewater treatment. The
total nitrogen entering centralized wastewater treatment was estimated by multiplying the U.S. population by the
percent of wastewater collected for centralized treatment (about 83 percent in 2020), the consumed protein per
capita, the fraction of N in protein, the correction factor for additional N from household products, the factor for
industrial and commercial co-discharged protein into wastewater treatment, and the factor for non-consumed
protein added to wastewater.
Equation 7-35: Total Nitrogen Entering Centralized Systems (IPCC 2019 [Eq. 10])
TNdom.central (kg N/year)
= (USpop X Tcentralized) X Protein X Fnpr X Nhh X Fnon-con X Find-com
Table 7-30: Variables and Data Sources for Non-Consumed Protein and Nitrogen Entering
Centralized Systems
Variable
Variable Description
Units
Inventory Years: Source
of Value
USpop
U.S. population3
Persons
1990-2020: U.S. Census
Bureau (2021)
Tcentralized
Percent collected3
%
Odd years from 1989
through 2019: U.S.
Census Bureau (2019)
Data for intervening
years obtained by linear
interpolation
2020: Forecasted from
the rest of the time
series
Protein
Consumed protein per capita3
kg/person/year
1990-2020: Calculated
Fnpr
Fraction of nitrogen in protein (0.16)
kg N/kg protein
1990-2020: IPCC (2019)
Nhh
Factor for additional nitrogen from household
products (1.17)
No units
1990-2020: IPCC (2019)
Fnon-con
Factor for U.S. specific non-consumed protein
(1.13)
No units
Find-com
Factor for Industrial and Commercial Co-
Discharged Protein (1.25)
No units
a Value of this activity data varies over the Inventory time series.
Nitrous oxide emissions from POTWs were estimated by multiplying the total nitrogen entering centralized
wastewater treatment, the relative percentage of wastewater treated by aerobic systems (other than constructed
wetlands) and anaerobic systems, aerobic systems with constructed wetlands as the sole treatment, the emission
factor for aerobic systems and anaerobic systems, and the conversion from N2 to N2O.
Waste 7-43
-------
Table 7-34 presents the data for U.S. population, population served by centralized wastewater treatment plants,
available protein, and protein consumed. The methodological equations are:
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
1990-2020: 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.
2005-2020: Forecasted
from the rest of the time
series
EF aerobic
Emission factor - aerobic systems (0.015)
kg N20-N/kg N
1990-2020: ERG (2021b)
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
7-44 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 7-32: Variables and Data Sources for N2O Emissions from Centrally Treated Aerobic
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
1990-2020: 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.
2013-2020: Forecasted
from the rest of the time
series
EFcw
Emission factor for constructed wetlands
(0.0013)
kg N20-N/kg N
1990-2020: IPCC (2014)
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.
2013-2020: Forecasted
from the rest of the time
series
Ncw.inf
BOD concentration in wastewater entering the
constructed wetland (25)
mg/L
1990-2020: 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
1990-2020: IPCC (2014)
1/106
Conversion factor
mg to kg
Standard conversion
365.25
Conversion factor
Days in a year
Standard conversion
a Value of this activity data varies over the Inventory time series.
Data sources and methodologies are similar to those described for aerobic systems, other than constructed
wetlands. See discussion above.
Equation 7-40: Emissions from Centrally Treated Anaerobic Systems (IPCC 2019 [Eq. 6.9])
C (kt N20/year)
= [(TNdom.central) X (% anaerobic)] X EFanaerobic X 44/28 X 1/106
Table 7-33: Variables and Data Sources for N2O Emissions from Centrally Treated Anaerobic
Systems
Waste 7-45
-------
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
1990-2020: 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.
2005-2020: Forecasted from
the rest of the time series
EF anaerobic
Emission factor for anaerobic
reactors/deep lagoons (0)
kg N20-N/kg N
1990-2020: IPCC (2019)
44/28
Conversion factor
Molecular weight
ratio of N20 to N2
Standard conversion
1/106
Conversion factor
mg to kg
Standard conversion
a Value of this activity data varies over the Inventory time series.
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)
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
2016
327
81.1
44.5
34.3
2017
329
82.1
44.7
34.5
2018
330
82.9
44.9
34.7
2019
332
83.6
44.4
34.2
2020
336
82.7
44.4
34.2
Sources: Population - U.S. Census Bureau (2021); WWTP Population - U.S. Census
Bureau (2019); Available Protein - USDA (2021b); Protein Consumed - FAO (2021c).
Emissions from Discharge of Centralized Treatment Effluent:
Nitrous oxide emissions from the discharge of wastewater treatment effluent were estimated by multiplying the
total nitrogen in centrally treated wastewater effluent by the percent of wastewater treated in primary,
secondary, and tertiary treatment and the fraction of nitrogen remaining after primary, secondary, or tertiary
treatment and then multiplying by the percent of wastewater volume routed to waterbodies with nutrient-
impaired/eutrophic conditions and all other waterbodies (ERG 2021a) and emission factors for discharge to
impaired waterbodies and other waterbodies from IPCC (2019). The methodological equations are:
Equation 7-41: Emissions from Centrally Treated Systems Discharge (U.S.-Specific)
D (kt N20/year)
= [(Neffluent.imp X EFimp) + (Nefluent,nonimp X EFnonimp)] X 44/28 X 1/106
where,
Equation 7-42: Total Organics in Centralized Treatment Effluent (IPCC 2019 [Eq. 6.8])
Neffulent.dom (kg N/year)
7-46 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
= [TNdom.central11 X % primary X (l-Nrem,PRiMARY)] + [TNdom.central X % secondary X (l-Nrem,sEcoNDARY)] +
[TNdOM.CENTRAL X % tertiary X (l-Nrem,TERTIARY)]
Equation 7-43: Total Nitrogen in Effluent Discharged to Impaired Waterbodies (U.S.-
Specific)
Neffluent.imp (kg N/year)
= (Neffulent.dom X PercentiMp)/1000
Equation 7-44: Total Nitrogen in Effluent Discharged to Nonimpaired Waterbodies (U.S.-
Specific)
Neffluent.nonimp (kg N year)
= (Neffluent.dom X PercentNONiMp)/1000
Table 7-35: Variables and Data Sources for N2O Emissions from Centrally Treated Systems
Discharge
Variable
Variable Description
Units
Source of Value
Neffulent.dom
Total organics in centralized treatment effluent3
kg N/year
1990-2020: Calculated
Molecular
44/28
Conversion factor
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
1990-2020: Calculated
1000
Conversion factor
kg to kt
Standard Conversion
% primary
Percent of primary domestic centralized treatment3
%
1990,1991: Set equal to
% secondary
Percent of secondary domestic centralized treatment3
%
1992.
1992, 1996, 2000,
2004, 2008, 2012: EPA
(1992, 1996, 2000,
2004a, 2008, and
2012), respectively
% tertiary
Percent of tertiary domestic centralized treatment3
%
Data for intervening
years obtained by
linear interpolation.
2013-2020: Forecasted
from the rest of the
time series
Nrem.PRIMARY
Fraction of nitrogen removed from primary domestic
centralized treatment (0.1)
No units
Nrem.SECONDARY
Fraction of nitrogen removed from secondary domestic
centralized treatment (0.4)
No units
1990-2020: IPCC (2019)
Nrem.TERTIARY
Fraction of nitrogen removed from tertiary domestic
centralized treatment (0.9)
No units
Neffluent.imp
Total nitrogen in effluent discharged to impaired waterbodies
kg N/year
Neffluent.nonimp
Total nitrogen in effluent discharged to nonimpaired
waterbodies
kg N/year
1990-2020: Calculated
EFimp
Emission factor (discharge to reservoirs/lakes/estuaries)
(0.19)
kg N20-N/kg N
1990-2020: IPCC (2019)
EFlMONIMPr
Emissions factor (discharge to other waterbodies) (0.005)
kg N20-N/kg N
PercentiMP
Percent of wastewater discharged to impaired waterbodies3
%
1990-2010: Set equal to
PercentNONiMP
Percent of wastewater discharged to nonimpaired
waterbodies3
%
2010
2010: ERG (2021a)
11 See emissions from centrally treated aerobic and anaerobic systems for methodological equation calculating TNDom_central.
Waste 7-47
-------
Variable
Variable Description
Units
Source of Value
2011: Obtained by
linear interpolation
2012: ERG (2021a)
2013-2020: Set equal to
2012
a Value for this activity data varies over the Inventory time series.
Industrial Wastewater N2O Emission Estimates
Nitrous oxide emission estimates from industrial wastewater were added to the inventory for the first time and
developed according to the methodology described in the 2019 Refinement. U.S. industry categories that are likely
to produce significant N2O emissions from wastewater treatment were identified based on whether they generate
high volumes of wastewater, whether there is a high nitrogen wastewater load, and whether the wastewater is
treated using methods that result in N2O emissions. The top four industries that meet these criteria and were
added to the inventory are meat and poultry processing; petroleum refining; pulp and paper manufacturing; and
breweries (ERG 2021a). Wastewater treatment and discharge emissions for these sectors for 2020 are displayed in
Table 7-36 below. Table 7-20 contains production data for these industries.
Table 7-36: Total Industrial Wastewater N2O Emissions by Sector (2020, MMT CO2 Eq. and
Percent)
N20 Emissions
(MMT CO?
% of Industrial
Industry
Eq.)
Wastewater N20
Meat & Poultry
0.2
48.1
Petroleum Refineries
0.1
28.4
Pulp & Paper
0.1
22.7
Breweries
+
0.8
Total
0.5
100
+ Does not exceed 0.5 MMT CO2 Eq.
Note: Totals may not sum due to independent rounding.
Emissions from Industrial Wastewater Treatment Systems:
More recent research has revealed that emissions from nitrification or nitrification-denitrification processes at
wastewater treatment, previously judged to be a minor source, may in fact result in more substantial emissions
(IPCC 2019). N2O is generated as a by-product of nitrification, or as an intermediate product of denitrification.
Therefore, N2O emissions are primarily expected to occur from aerobic treatment systems. To estimate these
emissions, the total nitrogen entering aerobic wastewater treatment for each industry must be calculated. Then,
the emission factor provided by the 2019 Refinement is applied to the portion of wastewater that undergoes
aerobic treatment.
The total nitrogen that enters each industry's wastewater treatment system is a product of the total amount of
industrial product produced, the wastewater generated per unit of product, and the nitrogen expected to be
present in each meter cubed of wastewater (IPCC equation 6.13).
Equation 7-45: Total Nitrogen in Industrial Wastewater
TN,NDi = PtXWtX TNi
where,
TN iNDi — total nitrogen in wastewater for industry /' for inventory year, kg TN/year
/ = industrial sector
Pi = total industrial product for industrial sector / for inventory year, t/year
7-48 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Wi = wastewater generated per unit of production for industrial sector / for inventory year,
m3/t product
TNi = total nitrogen in untreated wastewater for industrial sector / for inventory year, kg TN/m3
For the four industries of interest, the total production and the total volume of wastewater generated has already
been calculated for Cm emissions. For these new N2O emission estimates, the total nitrogen in the untreated
wastewater was determined by multiplying the annual industry production, shown in Table 7-20, by the average
wastewater outflow, shown in Table 7-23, and the nitrogen loading in the outflow shown in Table 7-37.
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
a Units are kilograms N per air-dried metric ton of production.
Nitrous oxide emissions from industry wastewater treatment are calculated by applying an emission factor to the
percent of wastewater (and therefore nitrogen) that undergoes aerobic treatment (IPCC Equation 6.11).
Equation 7-46: N2O Emissions from Indsutrial Wastewater Treatment Plants
N20 PlantsIND = x EFi j x T/V/(VDl)] x ;
44
28
where,
N2O PlantsiND = N2O emissions from industrial wastewater treatment plants for inventory year, kg
INhO/year
TN iNDi — total nitrogen in wastewater from industry /' for inventory year, kg N/year
Ti,j = degree of utilization of treatment/discharge pathway or system j, for each industry / for
inventory year
/ = industrial sector
j = each treatment/discharge pathway or system
EFi.j = emission factor for treatment/discharge pathway or system j, kg INhO-N/kg N. 0.015 kg
INhO-N/kg N ERG (2021b)
44/28 = conversion of kg N2O-N into kg N2O
For each industry, the degree of utilization (Ti,j)—the percent of wastewater that undergoes each type of
treatment-was previously determined for CFU emissions and presented in Table 7-22.
Emissions from Industrial Wastewater Treatment Effluent:
Nitrous oxide emissions from industrial wastewater treatment effluent are estimated by multiplying the total
nitrogen content of the discharged wastewater effluent by an emission factor associated with the location of the
discharge. Where wastewater is discharged to aquatic environments with nutrient-impacted/eutrophic conditions
(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 N20/year)
Waste 7-49
-------
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
N Effluent|ND (kg N)
Industry-Specific
N Removal Factor
Meat & Poultry
8,773,308
0.082
Petroleum Refineries
1,698,953
0.045
Pulp & Paper
18,809,623
1.08
Breweries
1,069,919
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 2020 CH4 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
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 and Table 7-40. For
2020, methane emissions from wastewater treatment were estimated to be between 11.8 and 22.4 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 35 percent below to 23 percent above the 2020 emissions estimate of 18.3 MMT CO2 Eq. Nitrous
oxide emissions from wastewater treatment were estimated to be between 15.3 and 69.3 MMT CO2 Eq., which
indicates a range of approximately 35 percent below to 194 percent above the 2020 emissions estimate of 3.5
MMT C02Eq.
7-50 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
For 1990, methane emissions from wastewater treatment were estimated to be between 14.3 and 26.0 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 28 percent above the 1990 emissions estimate of 20.23 MMT CO2 Eq.
Nitrous oxide emissions from wastewater treatment were estimated to be between 11.8 and 50.8 MMT CO2 Eq.,
which indicates a range of approximately 29 percent below to 206 percent above the 1990 emissions estimate of
16.6 MMT CO2 Eq.
Table 7-39: Approach 2 Quantitative Uncertainty Estimates for 2020 Emissions from
Wastewater Treatment (MMT CO2 Eq. and Percent)
Source
Gas
2020 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
18.3
11.8
22.4
-35%
+23%
Domestic
ch4
11.8
6.5
14.9
-45%
+26%
Industrial
ch4
6.4
3.7
10.3
-42%
+60%
Wastewater T reatment
n2o
23.5
15.3
69.3
-35%
+194%
Domestic
n2o
23.0
14.4
68.2
-38%
+196%
Industrial
n2o
0.5
0.5
1.6
-1%
+207%
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 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.
EPA conducted engagement with stakeholders on development of a U.S.-specific emission factor for nitrous oxide
emissions from aerobic wastewater treatment systems. EPA received feedback on study data that reflect U.S.
operations and where data collection methods ensure the quality of the emissions data (ERG 2021b).
Waste 7-51
-------
Recalculations Discussion
Population data were updated to reflect revised U.S. Census Bureau datasets which resulted in changes to 2002
through 2019 values (U.S. Census Bureau 2021). The percent of onsite and collected wastewater were updated to
reflect revised American Housing Survey data which resulted in changes to the 2018 and 2019 values (U.S. Census
Bureau 2019). These changes affected the timeseries from 2002 through 2019. Protein data were updated to
reflect available protein values available for 2014 through 2018 (FAO 2021c). Pulp, paper, and paperboard
production data were updated to reflect revised values for 2017 through 2019 (FAO 2021a). Updated red meat
production values for 2019, as well as fruits and vegetables processing production values for 2016 through 2019,
were updated based on revised data (USDA 2021a; USDA 2021c). Updated ethanol production and petroleum
refining production values for 2018 and 2019 were based on revised data (RFA 2021a; RFA 2021b; EIA 2021).
The BOD concentration for wastewater entering constructed wetlands used as tertiary treatment was updated
from the secondary treatment standard (30 mg/L) to the median value (9.1 mg/L) provided (EPA 2013). Domestic
wastewater treatment and discharge CFU emissions decreased an average of 0.3 percent over the time series, with
the smallest decrease of 0.04 percent (0 MMT CO2 Eq.) in 1992 and largest decrease of 3.3 percent (0.1 MMT CO2
Eq.) in 2019.
EPA revised the domestic wastewater N2O methodology based on the 2019 Refinement (IPCC 2019) by updating
the factor for non-consumed protein (1.13). EPA also revised the emission factor for centralized aerobic systems
which affected 1990 through 2019 (ERG 2021b). All of these changes affected the time series from 1990 through
2019. Domestic wastewater treatment and discharge N2O emissions decreased an average 11.6 percent over the
time series, with the smallest decrease of 10.1 percent (2.6 MMT CO2 Eq.) in 2018 and largest decrease of 12.4
percent (3.1 MMT C02 Eq.) in 2015.
EPA revised the industrial wastewater CFU methodology based on the 2019 Refinement (IPCC 2019): updated
emission estimates from discharge of pulp and paper manufacturing wastewater to aquatic environments using a
Tier 2 methodology and default emission factor (ERG 2021b). All of these changes affected the time series from
1990 through 2019. Industrial wastewater treatment and discharge CH4 emissions increased an average of 1.6
percent over the time series, with the smallest increase of 0.9 percent (0.0 MMT CO2 Eq.) in 2017 and largest
increase of 2.6 percent (0.1 MMT CO2 Eq.) in 1990.
EPA revised the emissions factor for centralized aerobic systems which affected nitrous oxide emissions estimates
for 1990 through 2019 (ERG 2021b). Due to the updates to 2018 petroleum refining and pulp, paper, and
paperboard manufacturing production data, the total nitrogen entering treatment was updated for 1990 through
2019. All of these changes affected the time series emissions estimates from 1990 through 2019. Industrial
wastewater treatment and discharge N2O emissions decreased an average 5.2 percent over the time series, with
the smallest decrease of 5.1 percent (0.02 MMT CO2 Eq.) in 1995 and largest decrease of 6.0 percent (0.03 MMT
CO2 Eq.) in 2019.
The cumulative effect of these recalculations had a large impact on the overall wastewater treatment emission
estimates. Over the time series, the average total emissions decreased by 6 percent from the previous Inventory.
The changes ranged from the smallest decrease, 5.1 percent (2.0 MMT CO2 Eq.), in 1990, to the largest decrease,
7.4 percent (3.3 MMT C02 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:
• Evaluate a 2020 survey compiling 2018 data on biosolids from the North East Biosolids & Residuals
Association (NEBRA). NEBRA expects the final report to be available in September 2021. This report could
help refine total sludge data which have been forecasted since 2004.
7-52 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
• Evaluate the use of POTW BOD effluent discharge data from ICIS-NPDES.12 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.
• 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;
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 will continue to look for methods to improve the transparency of the fate of sludge produced in
wastewater treatment.
EPA notes the following improvements will continued to be investigated as time and resources allow, but there are
no immediate plans to implement 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.
• Review whether sufficient data exist to develop U.S.-specific CFU 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.
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.
12 ICIS-NPDES refers to EPA's Integrated Compliance Information System - National Pollutant Discharge Elimination System.
Waste 7-53
-------
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
microorganisms, which convert the CH4 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 2020, the amount of waste composted in the United States (see Table 7-42) increased from 3,810 kt
to 22,774 kt. 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 2020, a 24 percent increase in waste composted is
observed. Emissions of CH4 and N2O from composting from 2010 to 2020 have increased by the same percentage.
In 2020, CH4 emissions from composting (see Table 7-40 and Table 7-41) were 2.3 MMT CO2 Eq. (91 kt), and N2O
emissions from composting were 2.0 MMT CO2 Eq. (7 kt), representing consistent emissions trends when
compared to 2019. The wastes composted primarily include 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 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 then
land-apply the resultant compost to soils.
The growth in composting since the 1990s and specifically over the past decade is attributable to the following
factors: (1) the enactment of legislation by state and local governments that discouraged the disposal of yard
trimmings and food waste in landfills, (2) yard trimming collection and yard trimming drop off sites provided 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, 22 states representing about 44 percent of the nation's
population have enacted such legislation (NERC 2020). There are many more initiatives at the metro and municipal
level 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 April 2019, six states
(California, Connecticut, New York, Massachusetts, Rhode Island, Vermont) and seven municipalities (Austin, TX;
Boulder, CO; Hennepin County, MN; Metro, 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 (Harvard Law School and CET 2019). In most cases, organic waste reduction in
landfills is accomplished by following recycling guidelines, donating excess waste for human consumption, or by
sending waste to organics processing facilities (Harvard Law School and CET 2019). An example of an organic waste
7-54 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 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).
There are a growing number of initiatives to encourage households and businesses to compost or beneficially
reuse food waste.
Table 7-40: ChU and N2O Emissions from Composting (MMT CO2 Eq.)
Activity
1990
2005
2016
2017
2018
2019
2020
ch4
0.4
1.9
2.3
2.5
2.3
2.3
2.3
n2o
0.3
1.7
2.0
2.2
2.0
2.0
2.0
Total
0.7
3.5
4.3
4.6
4.3
4.3
4.3
Note: Totals may not sum due to independent rounding.
ble 7-41: CH4 and N2O Emissions from Composting (kt)
Activity
1990
2005
2016
2017
2018
2019
2020
ch4
15
75
91
98
90
91
91
n2o
1
6
7
7
7
7
7
Methodology
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
Ei= MX EFi
where,
Ei = CH4 or N2O emissions from composting, kt CH4or N2O
M = mass of organic waste composted in kt
EFi = emission factor for composting, 41 Cm/kt of waste treated (wet basis) and
0.31 INhO/kt of waste treated (wet basis) (IPCC 2006)
i = designates either CH4or N2O
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, 2005, and 2014 to 2015 were taken from EPA's Advancing
Sustainable Materials Management: Facts and Figures 2015 (EPA 2018); the estimates of the quantities composted
for 2016 to 2018 were taken from EPA's Advancing Sustainable Materials Management: 2018 Tables and Figures
(EPA 2020); the estimate of the quantity composted for 2019 and 2020 were extrapolated using the 2018 quantity
composted and a ratio of the U.S. population growth between 2018 to 2019, and 2019 to 2020, respectively (U.S.
Waste 7-55
-------
Census Bureau 2021). New to this inventory are estimates of waste composted by commercial facilities in Puerto
Rico provided by EPA Region 2 (Kijanka 2020). Only limited information across the time series was provided. This
inventory includes waste composted for 2017, 2018, and/or 2019 from three facilities in Puerto Rico, ranging from
1,200 tons to a high of 15,021 tons. The average waste composted for these years was used as the annual amount
composted for the respective facility for years the facility was operational, but for which annual waste composted
data are lacking. Additional efforts are being made to fill these historical data gaps.
Table 7-42: U.S. Waste Composted (kt)
Activity
1990
2005
2016
2017
2018
2019
2020
Waste Composted
3,810
18,655
22,795
24,501
22,594
22,709
22,774
Uncertainty
The major uncertainty drivers are the assumption that all composting emissions come from commercial windrow
facilities and the use of default emission factors (IPCC 2006) which are tied to a homogenous mixture of waste
processed across the country (largely yard trimmings). Data presented by BioCycle (BioCycle 2017) confirm most
composting operations use the windrow method and yard trimmings are the largest share of material composted
across the country, but there are other composting methods used and waste characteristics will vary at a facility
level. The material composted is estimated using a mass balance flows model with end results presented in the
EPA Sustainable Materials Management Reports. No primary data on actual annual quantities of material
composted are collected by the Inventory compilation team or the EPA Sustainable Materials Management Report
compilation team, thus, the activity data used to generate greenhouse gas emissions are also a source of
uncertainty. Additionally, there are composting operations in Puerto Rico and U.S. territories that are not explicitly
included in the national quantity of material composted as reported in the EPA Sustainable Materials Management
Reports because the methodological scope does not include Puerto Rico and U.S. territories. EPA took steps to
include emissions from Puerto Rico and U.S. Territories beginning in the 1990 to 2020 inventory and will continue
to seek out additional data in future Inventories.
The estimated uncertainty from the 2006 IPCC Guidelines is ±58 percent for the Approach 1 methodology. A ±30
percent uncertainty factor is applied to the annual amount of material composted (i.e., the activity data) and a ±50
percent uncertainty factor is applied to both the Cm and N2O emission factors. These uncertainty factors were
developed using expert judgment. Using the IPCC's error propagation equation (Equation 3.1 in IPCC 2006 Volume
1, Chapter 3), the combined uncertainty percentage is ±58 percent.
Emissions from composting in 2020 were estimated to range between 1.8 and 6.8 MMT CO2 Eq., which indicates a
range of 58 percent below to 58 percent above the 2020 emission estimate of each gas (see Table 7-43).
Table 7-43: Approach 1 Quantitative Uncertainty Estimates for ChU and N2O Emissions from
Composting (MMT CO2 Eq. and Percent)
Source
Gas
2020 Emission Estimate
Uncertainty Range Relative to Emission Estimate
(MMT CO? Eq.)
(MMT CO?
Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
ch4
2.3
0.9
3.6
-58%
+58%
Composting
n2o
2.0
0.8
3.2
-58%
+58%
Total
4.3
1.8
6.8
-58%
+58%
7-56 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
QA/QC and Verification
General QA/QC procedures were applied to data gathering and input, documentation, and calculations 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). No errors were found for the current Inventory.
Recalculations Discussion
The U.S. population estimate for 2019 was revised with current U.S. Census Bureau Data from Table 1. Annual
Estimates of the Resident Population for the United States, Regions, States, the District of Columbia, and Puerto
Rico: April 1, 2010 to July 1, 2019; April 1, 2020; and July 1, 2020 (U.S. Census Bureau, 2021). Because the 2019
composting estimates are extrapolated based on population growth, this recalculation also resulted in a minor
increase in the quantity of material composted. The quantity of material composted in 2019 (extrapolated based
on population growth) increased by 6 kt from the previous Inventory report. In addition to this recalculation, the
amount of waste composted in PR is added onto the 2019 recalculated value, resulting in 22,709 kt of waste
composted in 2019, a 0.09% increase from 22,687 kt in the previous Inventory report.
Efforts have been made to improve the completeness of the composting Inventory by incorporating composted
waste from U.S. Territories. In 2016, EPA conducted a desk-based investigation into industrial/commercial
composting facilities in the U.S. Territories and identified facilities in Puerto Rico. Historical data is generally lacking
for identified facilities in Puerto Rico and service disruptions have occurred in previous years due to weather-
related damage. Quantities of material composted at two facilities between 2017 to 2019 and for one facility in
2019 have been obtained and incorporated into the current 1990 to 2020 Inventory. These three facilities began
operating in 1998, 2007, and 2012, respectively. The average of waste composted for these years was used as a
general estimate of waste composted for each year these facilities have been operational in the time series.
Incorporating material composted from these three facilities resulted in a minimal, if any, increase of less than 0.5
percent in total material composted and emissions across the time series. Additional efforts are being made to
collect historical and current (i.e., 2020) estimates of the quantity of waste composted by these three facilities.
This data may be incorporated into future Inventories as a methodological improvement.
Planned Improvements
EPA 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 will also continue to seek out activity data including processing capacity and years of operation for commercial
composting facilities in U.S. Territories such as Guam for inclusion in a future national Inventory.
Waste 7-57
-------
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 soil amendments. 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 CH4 generated, IPCC 2006),
collected biogas that is not completely combusted, and entrained gas bubbles and residual gas potential in the
digested sludge. 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 CFU 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 manure and food waste from restaurants or food processing industry, or a combination of manure
and waste from energy crops or crop residues (EPA 2016). The moisture content of the feedstock (wet or dry)
impacts the amount of biogas generation. Wet anaerobic digesters process feedstock with a solids content 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, typically manage food waste from
different sources, including food and beverage processing industries. Some stand-alone digesters also co-
digest other organics such as yard waste.
• On-farm digesters that 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).
From 1990 to 2020, 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 in the early part of the time
7-58 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
series are general estimates, extrapolated from data collected later in the time series (i.e., 2015 and later). 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 2020, 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
1990-2020
Activity
1990
2005
2016
2017
2018
2019
2020
CH4 Generation
+
0.05
0.2
0.2
0.2
0.2
0.2
CH4 Recovered
(+)
(+)
(+)
(+)
(+)
(+)
(+)
CH4 Emissions
+
+
0.2
0.2
0.2
0.2
0.2
+ 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 1990-2020
Activity
1990
2005
2016
2017
2018
2019
2020
CH4 Generation
1
2
7
7
7
7
7
CH4 Recovered
(+)
(+)
(0.7)
(+)
(0.5)
(0.5)
(0.5)
CH4 Emissions
1
2
7
6
6
6
6
+ Does not exceed 0.5 kt.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values.
Methodology
m
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 x 10~3 — R
where,
Cm Emissions = total CH4 emissions in inventory year, Gg CH4
M, = mass of organic waste treated by biological treatment type /', Gg, see Table 7-46
EF = emission factor for treatment /', g Cl-U/kg waste treated, 0.8 Mg/Gg CH4
i = anaerobic digestion
R = total amount of CH4 recovered in inventory year, Gg CH4
Equation 7-50: Recovered Methane Estimation for Anaerobic Digestion
YftXYlXlteS 1
R = Biogas x 0.0283 x x Biogas CH4 Density x x Yq9 x ^ DE~)
where,
Biogas = the annual amount of biogas produced, standard cubic feet per minute (scfm)
Waste 7-59
-------
662
65%
1/109
0.99
0.0283
525,600
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 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 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 ^aC2015)
(Fac2 oi6 + Fac2 015)
where,
W
Fac
total waste processed in the respective survey year, food and non-food waste (short tons).
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
7-60 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
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 Table 7-46 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 and 2020 were extrapolated using the average of
the waste digested from the 2017 and 2018 survey data (EPA 2021) as a proxy. Data for these years will be
updated as future EPA survey reports are published.
Table 7-46: U.S. Waste Digested (kt) from 1990-2020
Activity
1990
2005
2016
2017
2018
2019
2020
Waste Digested3
786
2,357
9,305
8,206
8,320
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 2020.
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 2020.
Table 7-47: Estimated Number of Stand-Alone AD Facilities Operating3 from 1990-2020
Year
1990
2005
2016
2017
2018
2019
2020
Estimated Count of Operational
Facilities
4
12
62
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
Waste 7-61
-------
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 2006 IPCC 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 1990-2020"
Activity
1990
2005
2016
2017
2018
2019
2020
Total Biogas Produced (scfm)b
767
2,301
10,498
6,402
7,282
6,842
6,842
R, recovered CH4 from biogas (kt)c
(0.05)
(0.14)
(0.67)
(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 2015 to 2018. The quantities presented in this table are likely underestimates because not all
operational facilities provided a survey response to the EPA AD 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 2018, divided by the average number of survey responses to generate a weighted average
estimate of biogas collected per facility, which is then multiplied 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 (2019), 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 Cl-Uemissions.
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
to tons. This slightly underestimates the quantity of waste digested and Cl-Uemissions. EPA (2021) did
convert the liquid waste managed to tons for 2017 and 2018 using a general conversion factor.
3. The assumption required to estimate the activity data for 1990 to 2014 may overestimate the number of
facilities in operation because it assumes that each facility operates from its start year for the entire time
series (i.e. facility closures are not taken into account). This introduces a large amount of uncertainty in
the estimates compared to years where there is directly reported survey data. It is unclear whether this
under- or over-estimates the quantity of waste digested and Cm emissions.
The estimated uncertainty from the 2006 IPCC Guidelines is ±54 percent for the Approach 1 methodology.
Emissions from anaerobic digestion at biogas facilities in 2020 were estimated to be between 0.1 and 0.2 MMT CO2
Eq., which indicates a range of 54 percent below to 54 percent above the 2020 emission estimate of CH4 (see Table
7-49). A ±20 percent uncertainty factor is applied to the annual amount of material digested (i.e., the activity data),
which was developed with expert judgment. A ±50 percent default uncertainty factor is applied to the CH4
7-62 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
emission factor (IPCC 2006). Using the IPCC's error propagation equation (Equation 3.1 in IPCC 2006 Volume 1,
Chapter 3), the combined uncertainty percentage is ±54 percent.
Table 7-49: Approach 1 Quantitative Uncertainty Estimates for Emissions from Anaerobic
Digestion (MMT CO2 Eq. and Percent)
Source
Gas
2020 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 0.2
-54% +54%
QA/QC and Verification
General QA/QC procedures were applied to data gathering and input, documentation, and calculations consistent
with the U.S. Inventory QA/QC Plan, which is in accordance with Vol. 1 Chapter 6 of the 2006 IPCC Guidelines (see
Annex 8 for more details). No errors were found for the current Inventory.
Recalculations Discussion
EPA incorporated the EPA (2021) AD Data Collection Survey Data for 2017 and 2018 into the current (1990 to
2020) Inventory. Recalculations were made to the number of operating facilities compared to the 1990 to 2019
Inventory for the following years: 1992,1999, 2001, 2004, 2006, 2008, and 2010 to 2014. The number of
operational facilities was sourced from EPA (2019) in the 1990 to 2019 Inventory, and facilities counts for the
aforementioned years changed slightly (plus or minus one facility) in EPA (2021). Due to how the emissions are
estimated for 1990 to 2014, a change in the number of facilities in any of these years will impact the facility count
for all subsequent years in addition to the estimated annual quantities of waste digested and net CH4 emissions.
The recalculations resulted in changes in net CH4 emissions of less than 0.5 kt each year.
Recalculations were also made to the 2017 to 2019 emissions estimates with the inclusion of the EPA (2021)
survey data for 2017 and 2018. The estimated quantity of waste digested decreased by 1,428 kt, 1,314 kt, and
1,371 kt in 2017, 2018, and 2019, respectively. The decrease in annual waste digested resulted in a decrease in net
Cm emissions by approximately 0.9 kt each year from 2017 to 2019.
Planned Improvements
EPA will continue to incorporate survey data from future EPA AD Data Collection Surveys when the survey data are
published. These revisions will change the estimated emissions for later years in the time series (e.g., 2019, 2020).
EPA will also re-assess how best to estimate annual waste processed using proxy data for years between the EPA
AD Data Collection Survey reports as needed (e.g., for 2019, 2020). The methodology described here assumes the
same average amount of waste is processed each year for 2019 and 2020.
EPA will conduct additional desk-based research to improve the estimated number of operational facilities by year
prior to 2015 and how best to estimate the quantity of waste processed per year by these facilities with the goal of
better estimating the annual quantity of waste digested between 1990 to 2014. Available data will also be
compiled for facilities that did not directly respond to the EPA AD Data Collection surveys for completeness.
EPA will investigate the amount of recovered biogas for years prior to 2015 (i.e., the years prior to the EPA AD Data
Collection Surveys). Currently, partial data of recovered biogas are available between 2015 to 2018 from the EPA
AD Data Collection Surveys. The primary purpose of this improvement will be to understand whether the range of
recovered biogas from the survey data are reflective of earlier years in the time series.
Waste 7-63
-------
The uncertainty assessment will be further reviewed to confirm the appropriateness of the uncertainty factor(s) to
be applied.
7.5 Waste Incineration (CRF Source
Category 5C1)
As stated earlier in this chapter, carbon dioxide (CO2), nitrous oxide (N2O), and methane (CH4) 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 2020 resulted in 13.5 MMT CO2 Eq. of emissions.
For more details on emissions from the incineration of waste, see Section 3.3 of the Energy chapter.
Additional sources of emissions from waste incineration include non-hazardous industrial waste incineration and
medical waste incineration. As described in Annex 5 of this report, data are not readily available for these sources
and emission estimates are not provided.
An analysis of the likely level of medical waste incineration emissions was conducted based on a 2009 study of
hospital/ medical/ infectious waste incinerator (HMIWI) facilities in the United States (RTI 2009). Based on that
study's information of waste throughput and an analysis of the fossil-based composition of the waste, it was
determined that annual greenhouse gas emissions for medical waste incineration would be below 500 kt CO2 Eq.
per year and considered insignificant for the purposes of Inventory reporting under the UNFCCC. More information
on this analysis is provided in Annex 5.
Furthermore, an analysis was conducted on the likely level of sewage sludge incineration emissions based on the
total amount of sewage sludge generated and assumed percent incineration. Based on assumed amount of sludge
incinerated and non-CC>2 factors for solid biomass it was determined that annual greenhouse gas emissions for
sewage sludge incineration would be below 500 kt CO2 Eq. per year and considered insignificant for the purposes
of Inventory reporting under the UNFCCC.
7.6 Waste Sources of Precursor
Greenhouse Gases
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 UNFCCC13 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 2020 are provided in Table 7-50.
13 See http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
7-64 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Table 7-50: Emissions of NOx, CO, NMVOC, and SO2 from Waste (kt)
Gas/Source
1990
2005
2016
2017
2018
2019
2020
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
57
52
52
52
52
Wastewater Treatment
57
49
25
22
22
22
22
Miscellaneous3
557
43
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 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 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 (2022).14 NEI Tier 1 emission categories related to the waste sector categories in
this report include: Waste Disposal and Recycling (landfills; publicly owned treatment works; industrial
wastewater; treatment, storage, and disposal facilities; and other). As described in detail in the NEI Technical
Support Documentation (TSD) (EPA 2021b), 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 recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2020, which are described in detail in the NEI's TSD (EPA 2021b). No quantitative estimates of uncertainty
were calculated for this source category.
14 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, SO2,
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.
Waste 7-65
-------
8. Other
The United States does not report any greenhouse gas emissions under the Intergovernmental Panel on Climate
Change (IPCC) "Other" sector.
Other 8-1
-------
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 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; improved inventory methods become available; and/or for
correction of errors."
In general, when methodological changes have been implemented, the previous Inventory's time series (i.e., 1990
to 2019) will be recalculated to reflect the change, per guidance in IPCC (2006). Changes in historical data are
generally the result of changes in statistical data supplied by other agencies, and do not necessarily impact the
entire time series.
The results of all methodological changes and historical data updates made in the current Inventory are presented
in Figure 9-1, Table 9-1, and Table 9-2. Figure 9-1 presents the impact of recalculations by sector and on net total
emissions across the timeseries. Table 9-1 summarizes the quantitative effect of all changes on U.S. greenhouse
gas emissions by gas across the Energy, Industrial Processes and Product Use (IPPU), Agriculture, and Waste
sectors, while Table 9-2 summarizes the quantitative effect of changes on annual net fluxes from Land Use, Land-
Use Change, and Forestry (LULUCF). Both the figure and tables present results relative to the previously published
Inventory (i.e., the 1990 to 2019 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 Inventory includes new categories not included in the previous Inventory that improve completeness of the
national estimates. Specifically, the current report includes methane (CH4) emissions from post-meter uses (i.e.,
includes leak emissions from residential and commercial appliances, industrial facilities and power plants, and
natural gas fueled vehicles), fugitive CO2 emissions from coal mining, CO2 emissions from land converted to
flooded land, CH4 emissions from land remaining and land converted to flooded land (i.e., reservoirs and other
constructed waterbodies), and PFC (CF4) emissions from electrical transmission and distribution.
The following source and sink categories underwent the most significant methodological and historical data
changes, and all noted here are key categories except CO2 emissions from incineration of waste. A brief summary
of the recalculations and/or improvements undertaken are provided for these categories.
• Natural Gas Systems (CHa). EPA received information and data related to the Inventory emission
estimates through GHGRP reporting, the annual Inventory formal public notice periods, stakeholder
feedback on updates under consideration, and new studies. EPA thoroughly evaluated relevant
information available and made several updates to the Inventory, including incorporating post-meter
emissions, reassessing the Gas STAR reductions data and incorporating Methane Challenge data, adding
Recalculations and Improvements 9-1
-------
well blowout emissions, and using Pipeline and Hazardous Materials Safety Administration (PHMSA) data
to update underground storage well counts. The recalculations and inclusion of post-meter estimates
resulted in an average increase in Cm emission estimates across the 1990 through 2019 time series,
compared to the previous Inventory, of 13.2 MMT CO2 Eq„ or 8.1 percent.
• Land Converted to Grassland (CO2). Recalculations are associated with new FIA data from 1990 to 2020 on
biomass, dead wood and litter C stocks in Forest Land Converted to Grassland, and updated estimates for
mineral soils from 2016 to 2020 using the linear extrapolation method. As a result, Land Converted to
Grassland estimates increased an average of 2.9 MMT CO2 Eq. (14.9 percent decrease in carbon stock
change) on average over the time series.
• Wastewater Treatment (N2O). EPA revised the domestic wastewater N2O methodology based on the 2019
Refinement (IPCC 2019) by updating the factor for non-consumed protein (1.13). EPA also revised the
emission factor for centralized aerobic systems which affected 1990 through 2019 (ERG 2021a). All of
these changes affected the time series from 1990 through 2019. Nitrous oxide emissions from wastewater
treatment and discharge decreased an average 11.5 percent (2.6 MMT CO2 Eq.) over the time series, with
the smallest decrease of 10.0 percent (2.6 MMT CO2 Eq.) in 2018 and largest decrease of 12.4 percent (3.1
MMT CO2 Eq.) in 2015.
• Manure Management (CHa). The manure management emission estimates include a number of revisions
and methodology updates including: updated methodology for population distribution to states where
USDA population data are withheld due to disclosure concerns (ERG 2021b) as well as raw animal
population data (impacted poultry, sheep, and swine population estimates), revised MCF for pasture to
align with updated guidance from IPCC (2019), and updated Cattle Enteric Fermentation Model (CEFM)
output data. The cumulative effect of these recalculations led to an average decrease of 2.5 MMT CO2 Eq.
(5.0 percent) over the time series.
• Incineration of Waste (CCh). Waste incineration estimates in the current Inventory were derived following
a new methodology relying on different data sources than previously used. Specifically, waste tonnage
estimates for 2006 through 2019 relied on several new data sources. For 1990 through 2020, CO2
emissions were calculated with a new methodology using a carbon emission factor calculated from EPA's
GHGRP data. The previous methodology relied on generation, disposal, and incineration rates of synthetic
fibers, plastics, and synthetic rubber and the accompanying carbon contents to calculate CO2 emissions
for incineration of these materials. As a result of the changes in data and methodology, CO2 emissions
increased by an average of 2.1 MMT CO2 Eq. (20.0 percent) each year over the time series.
• Land Converted to Settlements (CO2). Recalculations are associated with new FIA data from 1990 to 2020
on biomass, dead wood and litter C stocks in Forest Land Converted to Settlements, and updated
estimates for mineral and organic soils from 2016 to 2020 using the linear extrapolation method. As a
result, emissions from Land Converted to Settlements decreased an average 2.0 MMT CO2 Eq. (2.6
percent) over the time series.
• Forest Land Remaining Forest Land (CCh). 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 2020 are consistent with those used in the previous (1990 through 2019) Inventory. New national
forest inventory (NFI) data in most states were incorporated in the latest Inventory which contributed to
increases in forest land area estimates and carbon stocks. Fire data sources were also updated for Alaska
through 2020 and this combined with the new NFI data for the years 2018 through 2020 resulted in
substantial changes in carbon stocks and stock changes. Soil carbon stocks increased in the latest
Inventory relative to the previous Inventory and this change can be attributed to refinements in the
Digital General Soil Map of the United States (STATSG02) dataset. This resulted in a structural change in
the soil organic carbon estimates for mineral and organic soils across the entire time series. Additionally,
recent land use change in AK (since 2015) also contributed to variability in soil carbon stocks and stock
changes in recent years in the time series which contributed to differences in historic estimates. New
HWP data suggest a continued decline in paper products in use over time due to changes in consumer
9-2 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
behavior (i.e., more use of electronic information sources) and a small drop in solid wood products in the
last year due to a downturn in the economy associated with the global pandemic. Overall, these revisions
lead to an average decrease of 2.0 MMT CO2 Eq. (0.3 percent) across the time series.
• Petroleum Systems (CHa). EPA received information and data related to the emission estimates through
GHGRP reporting, stakeholder feedback on updates under consideration, and new studies. EPA did not
make methodological updates for Petroleum Systems emission sources for this Inventory. However, for
certain sources, changes in emissions were driven mainly by GHGRP data submission revisions and
Enverus well count updates (Enverus 2021). The recalculations resulted in an average increase in CH4
emission estimates across the 1990 through 2019 time series, compared to the previous Inventory, of 1.5
MMT CO2 Eq. (3.6 percent).
• Grassland Remaining Grassland (CCh). Recalculations are associated with updated estimates for mineral
soils from 2016 to 2020 using the linear extrapolation method, in addition to a correction in the
estimation of biomass C. The correction is associated with foliage estimates for woodlands that had been
based on values for non-woodland foliage in the previous Inventory. The recalculations resulted in a
decrease in emission estimates 1.4 MMT CO2 Eq. (4.3 percent) on average over the time series compared
to the previous Inventory.
• Enteric Fermentation (CHa). In the previous Inventory, 1990 to 2017 estimates were retained from the
1990 through 2017 Inventory, and 2018 and 2019 estimates were based on a simplified approach that
used emission factors and extrapolated population estimates for all animals. For the current Inventory,
the Cattle Enteric Fermentation Model (CEFM) was used for cattle for all years, resulting in different
estimates for 2018 and 2019 than the prior Inventory. For non-cattle livestock in the current Inventory,
updated Tier 1 estimates were calculated for 2018 and 2019, yielding different results than the simplified
approach used for these years in the prior Inventory. In addition, there were changes to cattle-related
activity data including minor data revisions in the CEFM. Finally, non-cattle livestock emissions were
impacted as a result of revisions to the USDA animal population data and how "other" populations were
distributed to their respective states (ERGb 2021). The recalculations resulted in an average decrease in
CH4 emissions estimates across the 1990 through 2019 time series of 1.3 MMT CO2 Eq. (0.7 percent).
• Land Converted to Cropland (CO2). Recalculations are associated with new FIA data from 1990 to 2020 on
biomass, dead wood and litter C stocks in Forest Land Converted to Cropland, and updated estimates for
mineral soils from 2016 to 2020 using the linear extrapolation method. The recalculations resulted in an
estimated C loss of 0.1 MMT CO2 Eq. (1 percent) on average over the time series compared to the
previous Inventory.
Figure 9-1 presents the impact of recalculations by sector and on net total emissions across the timeseries.
Recalculations and Improvements 9-3
-------
Figure 9-1: Impacts from Recalculations to U.S. Greenhouse Gas Emissions by Sector
90 ¦ Change in Net Total Emissions
¦ Agriculture
80 ¦ Energy
¦ Industrial Processes and Product Use
70 ¦ LULUCF Sector Net Total
o^-HfNn^rLovor>.coa>0'-HfNro^rLovor^coa*Oi-ifNro'srLovors>.coa»
^^O^O^Cr*CT*CT>CT»^^OOOOOOOOOOi—it—i i—l l—I i—I i—It—It—ii—li—I
CT>a^a^cricriCT.cjioooooooooooooooooooo
HHririrtrtHHririfMfMfMNN(NfM(NfMOJfM(N(NfNlNlN(NfMfMfM
Table 9-1 and Table 9-2 present the quantitative impact of recalculations by gas for the Energy, IPPU, Agriculture,
and Waste sources and for the Land Use, Land-Use Change, and Forestry sources and sinks, respectively.
Table 9-1: Revisions to U.S. Greenhouse Gas Emissions (MMT CO2 Eq.)
Gas/Source
1990
2005
2016
2017
2018
2019
Average
Annual
Change
C02
9.0
3.1
3.7
3.2
1.2
3.3
5.5
Fossil Fuel Combustion
(0.3)
(1.5)
(1.9)
(1.2)
(2.1)
(4.4)
(0.7)
Electric Power Sector
NC
NC
+
+
+
0.1
+
Transportation
(0.2)
(0.1)
(2.2)
(2.5)
(3.8)
(3.5)
(0.6)
Industrial
(0.1)
(1.4)
0.2
0.3
0.5
(6.4)
(0.3)
Residential
NC
NC
+
+
0.2
4.6
0.2
Commercial
+
(+)
(+)
+
0.1
1.0
+
U.S. Territories
(+)
+
0.1
0.9
0.9
(0.3)
+
Non-Energy Use of Fuels
(0.6)
(0.2)
(0.3)
(0.9)
(0.9)
(2.0)
(0.6)
Natural Gas Systems
(0.1)
(0.2)
(0.3)
(0.1)
(1.5)
1.5
(0.1)
Cement Production
NC
NC
NC
NC
NC
NC
NC
Lime Production
NC
NC
NC
NC
NC
NC
NC
Other Process Uses of Carbonates
(0.1)
(0.2)
(0.2)
(0.1)
(0.1)
2.4
(0.1)
Glass Production
0.8
0.5
0.9
0.7
0.7
0.7
0.8
Soda Ash Production
NC
NC
NC
NC
NC
NC
NC
Carbon Dioxide Consumption
NC
NC
NC
NC
NC
NC
NC
Incineration of Waste
4.9
0.6
2.8
1.6
1.8
1.5
2.1
Titanium Dioxide Production
NC
NC
NC
NC
NC
NC
NC
Aluminum Production
NC
NC
NC
NC
NC
NC
NC
9-4 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Iron and Steel Production & Metallurgical
Coke Production
Ferroalloy Production
Ammonia Production
Urea Consumption for Non-Agricultural
Purposes
Phosphoric Acid Production
Petrochemical Production
Carbide Production and Consumption
Lead Production
Zinc Production
Petroleum Systems
Abandoned Oil and Gas Wells
Magnesium Production and Processing
Liming
Urea Fertilization
Coal Mining
International Bunker Fuelsa
Wood Biomass, Ethanol, and Biodiesel
Consumptionb
CH4c
Stationary Combustion
Mobile Combustion
Coal Mining
Abandoned Underground Coal Mines
Natural Gas Systems
Petroleum Systems
Abandoned Oil and Gas Wells
Petrochemical Production
Carbide Production and Consumption
Iron and Steel Production & Metallurgical
Coke Production
Ferroalloy Production
Enteric Fermentation
Manure Management
Rice Cultivation
Field Burning of Agricultural Residues
Landfills
Wastewater T reatment
Composting
Anaerobic Digestion at Biogas Facilities
Incineration of Waste
International Bunker Fuelsa
N2Oc
Stationary Combustion
Mobile Combustion
AdipicAcid Production
Nitric Acid Production
Manure Management
Agricultural Soil Management
Field Burning of Agricultural Residues
Wastewater T reatment
N20 from Product Uses
Caprolactam, Glyoxal, and Glyoxylic Acid
Production
Incineration of Waste
+
+
+
+
+
1.8
0.1
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
0.2
0.2
0.2
(0.2)
+
NC
NC
NC
(+)
(+)
+
+
NC
NC
(0.2)
(+)
NC
(0.1)
(+)
(0.1)
NC
NC
NC
NC
NC
(+)
NC
NC
NC
NC
NC
(+)
(+)
(+)
(+)
(+)
(+)
(+)
(+)
(+)
(0.1)
(0.1)
0.1
+
0.2
(0.6)
(+)
(+)
(+)
(+)
(+)
(+)
+
(+)
0.1
+
+
+
+
+
0.1
NC
NC
NC
NC
NC
(+)
(+)
NC
NC
(0.2)
(0.2)
(0.2)
(0.2)
(+)
4.6*
4.2*
2.8*
3.1*
3.1*
3.0*
4.0*
0.2
0.1
0.1
0.1
0.1
0.1
0.1
NC
NC
0.2
0.4
0.2
1.0
0.1
3.9
11.3
15.2
15.4
15.2
9.1
10.7
+
(+)
+
+
+
0.1
+
0.1
0.1
0.1
+
+
0.1
0.1
NC
NC
NC
NC
NC
NC
+
NC
NC
NC
NC
NC
NC
NC
*
LO
00
13.3*
17.9*
17.8*
19.2*
14.5*
13.2*
(1.1)
1.9
1.2
1.2
1.3
1.3
1.5
(0.3)
(0.4)
(0.5)
(0.3)
(0.3)
0.4
(0.3)
NC
NC
NC
NC
NC
(+)
(+)
(+)
NC
NC
NC
NC
NC
(+)
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
(1.2)
(1.3)
(0.9)
(0.9)
(2.3)
(2.4)
(1.3)
(2.3)
(2.6)
(2.5)
(2.4)
(2.3)
(3.7)
(2.5)
NC
NC
(+)
(+)
(+)
+
+
NC
NC
NC
NC
NC
NC
NC
NC
0.2
(0.1)
(0.2)
(0.4)
(0.9)
(+)
0.1
0.1
+
+
(0.1)
(0.3)
0.1
NC
+
+
+
+
+
+
+
(+)
NC
(+)
(+)
(+)
(+)
NC
NC
+
+
+
+
+
NC
NC
NC
NC
NC
NC
NC
(2.2)
(2.5)
(1.6)
(1.7)
(1.5)
(0.3)
(2.1)
(+)
(+)
+
+
+
+
+
(0.1)
(0.2)
0.3
0.3
0.3
2.0
0.1
NC
NC
0.1
0.1
0.2
(+)
+
NC
NC
NC
NC
NC
NC
NC
(0.1)
(0.1)
0.3
0.3
(0.2)
(0.1)
(+)
0.1
0.4
0.7
0.7
0.7
0.6
0.4
NC
NC
NC
NC
NC
NC
+
(2.1)
(2.7)
(3.1)
(3.2)
(2.6)
(3.0)
(2.6)
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
NC
0.1
0.1
0.1
0.1
+
Recalculations and Improvements 9-5
-------
Composting
NC
+
+
+
+
+
+
Electronics Industry
NC
+
+
(+)
+
+
+
Natural Gas Systems
(+)
(+)
(+)
(+)
(+)
+
(+)
Petroleum Systems
(+)
(+)
(+)
(+)
(+)
(+)
(+)
International Bunker Fuelsa
(+)
(+)
(+)
(+)
(+)
(+)
(+)
HFCs, PFCs, SF6and NF3
+
(0.1)
0.2
0.8
1.2
1.3
+
HFCs
+
(0.1)
0.2
0.8
1.3
1.3
+
Substitution of Ozone Depleting
Substancesd
NC
(0.1)
0.2
0.8
1.3
1.3
+
HCFC-22 Production
NC
NC
NC
NC
NC
NC
(+)
Electronics Industry
+
+
+
+
+
+
+
Magnesium Production and Processing
NC
NC
NC
+
NC
NC
+
PFCs
+
+
+
0.1
0.1
0.1
+
Aluminum Production
NC
NC
+
+
(+)
(+)
+
Electronics Industry
+
+
+
0.1
0.1
0.1
+
Substitution of Ozone Depleting
Substancesd
NC
NC
NC
NC
NC
NC
NC
Electrical Transmission and Distribution
NO*
+*
+*
+*
NO*
+*
+*
sf6
+
+
(+)
(+)
(+)
(+)
(+)
Electrical Transmission and Distribution
+
(+)
+
+
(0.1)
(+)
(+)
Electronics Industry
NC
+
(+)
(+)
+
(+)
+
Magnesium Production and Processing
NC
NC
(+)
NC
+
(+)
(+)
nf3
NC
NC
+
(+)
(+)
(+)
(+)
Electronics Industry
NC
NC
+
(+)
(+)
(+)
(+)
Change in Total Gross Emissions (Sources)
10.8
11.8
17.5
17.7
16.1
13.4
14.1
Percent Change in Total Gross Emissions
0.2%
0.2%
0.3%
0.3%
0.2%
0.2%
0.2%
NC (No Change)
NO (Not Occurring)
+ Absolute value does not exceed 0.05 MMT C02 Eq. or 0.05 percent.
* Indicates inclusion of a new source for the current Inventory year that were not estimated in previous inventory. Emissions from
new source categories and subcategories are captured in gross/net emissions and percent change totals.
a Emissions from International Bunker Fuels are not included in totals.
b Emissions from Wood Biomass, Ethanol, 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 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; Land Converted to Coastal Wetlands; Land Remaining Flooded Land; Land Converted to
Flooded Land; and N20 emissions from Forest Soils and Settlement Soils.
d Small amounts of PFC emissions also result from this source.
Notes: Net change in total emissions presented without LULUCF. Totals may not sum due to independent rounding.
Table 9-2: Revisions to U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land
Use, Land-Use Change, and Forestry (MMT CO2 Eq.)
Average
Annual
Land-Use Category
1990
2005
2016
2017
2018
2019
Change
Forest Land Remaining Forest Land
16.2
(21.2)
(1.6)
(29.2)
17.7
43.7
2.8
Changes in Forest Carbon Stocks3
13.6
(25.7)
(3.7)
(28.6)
21.5
57.0
2.0
Non-C02 Emissions from Forest Firesb
2.7
4.6
2.1
(0.6)
(3.8)
(13.3)
0.8
N20 Emissions from Forest Soilsc
NC
NC
NC
NC
NC
NC
NC
Non-C02 Emissions from Drained Organic
Soilsd
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Land Converted to Forest Land
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
9-6 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-
2020
-------
Changes in Forest Carbon Stocks6
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
Cropland Remaining Cropland
NC
NC
(+)
(+)
(+)
(+)
(+)
Changes in Mineral and Organic Soil Carbon
Stocks
NC
NC
(+)
(+)
(+)
(+)
(+)
Land Converted to Cropland
+
(0.1)
(0.3)
(0.3)
(0.3)
(0.3)
(0.1)
Changes in all Ecosystem Carbon Stocks'
+
(0.1)
(0.3)
(0.3)
(0.3)
(0.3)
(0.1)
Grassland Remaining Grassland
(1.4)
(1.3)
(1.8)
(2.0)
(2.0)
(2.1)
(1.4)
Changes in Mineral and Organic Soil Carbon
Stocks
(1.4)
(1.3)
(1.8)
(2.0)
(2.0)
(2.1)
(1.4)
Non-C02 Emissions from Grassland Fires8
NC
NC
NC
NC
NC
NC
NC
Land Converted to Grassland
3.1
3.1
1.5
1.7
1.8
1.7
2.9
Changes in all Ecosystem Carbon Stocks'
3.1
3.1
1.5
1.7
1.8
1.7
2.9
Wetlands Remaining Wetlands
18.2
19.8
19.9
19.9
19.9
19.9
19.5
Changes in Organic Soil Carbon Stocks in
Peatlands
NC
NC
NC
NC
(+)
(+)
(+)
Changes in Biomass, DOM, and Soil Carbon
Stocks in Coastal Wetlands
NC
NC
NC
NC
(+)
(+)
(+)
CH4 Emissions from Coastal Wetlands
Remaining Coastal Wetlands
+
(+)
NC
NC
NC
NC
(+)
N20 Emissions from Coastal Wetlands
Remaining Coastal Wetlands
NC
NC
NC
NC
NC
NC
NC
Non-C02 Emissions from Peatlands
Remaining Peatlands
NC
NC
NC
+
+
+
+
CH4 Emissions from Flooded Land Remaining
Flooded Land
18.2*
19.8*
19.9*
19.9*
19.9*
19.9*
19.9*
Land Converted to Wetlands
6.4
0.6
0.5
0.5
0.5
0.5
1.9
Changes in Biomass, DOM, and Soil Carbon
Stocks
(+)
+
+
+
+
+
+
CH4 Emissions from Land Converted to
Coastal Wetlands
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Changes in Land Converted to Flooded Land
3.9*
0.3*
0.3*
0.3*
0.3*
0.3*
0.3*
CH4 Emissions from Land Converted to
Flooded Land
2.6*
0.2*
0.2*
0.2*
0.2*
0.2*
0.2*
Settlements Remaining Settlements
+
(+)
(+)
(4.0)
(3.7)
(2.9)
(0.4)
Changes in Organic Soil Carbon Stocks
NC
NC
+
+
+
+
+
Changes in Settlement Tree Carbon Stocks
NC
NC
NC
NC
NC
NC
NC
Changes in Yard Trimming and Food Scrap
Carbon Stocks in Landfills
+
(+)
+
(+)
(+)
+
+
N20 Emissions from Settlement Soils'1
NC
NC
(+)
(4.0)
(3.7)
(2.9)
(0.4)
Land Converted to Settlements
(2.1)
(2.2)
(1.6)
(1.4)
(1.3)
(1.3)
(2.0)
Changes in all Ecosystem Carbon Stocks'
(2.1)
(2.2)
(1.6)
(1.4)
(1.3)
(1.3)
(2.0)
Change in LULUCF Total Net Flux'
16.7
(26.3)
(6.1)
(34.6)
15.9
51.9
1.8
Change in LULUCF Emissions1'
23.5
24.6
22.2
19.5
16.3
6.9
21.0
Change in LULUCF Sector Totalk
40.2
(1.8)
16.2
(15.1)
32.2
58.7
22.8
Percent Change in LULUCF Total Net Flux
4.5%
-0.2%
1.9%
4.0%
4.0%
7.4%
2.7%
NC (No Change)
+ Absolute value does not exceed 0.05 MMT C02 Eq. or 0.05 percent.
* Indicates inclusion of a new source category or subcategory for the current Inventory year that were not estimated in the
previous year. Emissions from new source categories and subcategories are captured in gross/net emissions and percent
change totals.
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.
Recalculations and Improvements 9-7
-------
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.
s Estimates include CH4 and N20 emissions from fires on both Grassland Remaining Grassland and Land Converted to
Grass/and.
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; Land Converted to Coastal Wetlands;
emissions from Forest Soils and Settlement Soils; Flooded Land Remaining Flooded Land; and Land Converted to Flooded
Land.
k The LULUCF Sector Net Total is the net sum of all LULUCF CH4 and N20 emissions to the atmosphere plus net carbon stock
changes.
Note: Totals may not sum due to independent rounding.
9-8 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
10. References and Abbreviations
Executive Summary
BEA (2022) 2021 Comprehensive Revision of the National Income and Product Accounts: Current-dollar and "real"
GDP, 1929-2021. Bureau of Economic Analysis (BEA), U.S. Department of Commerce, Washington, D.C. Available
online at: http://www.bea.gOv/national/index.htm#gdp.
Duffield, J. (2006) Personal communication. Jim Duffield, Office of Energy Policy and New Uses, U.S. Department of
Agriculture, and Lauren Flinn, ICF International. December 2006.
EIA (2022) Electricity Generation. Monthly Energy Review, February 2022. Energy Information Administration, U.S.
Department of Energy, Washington, D.C. DOE/EIA-0035(2019/11).
EIA (2021b) Electricity in the United States. Electricity Explained. Energy Information Administration, U.S.
Department of Energy, Washington, D.C. Available online at:
https://www.eia.gov/energyexplained/index.php?page=electricity in the united states.
EIA (2019) International Energy Statistics 1980-2019. Energy Information Administration, U.S. Department of
Energy. Washington, D.C. Available online at: https://www.eia.gov/beta/international/.
EPA (2021a) Acid Rain Program Dataset 1996-2020. Office of Air and Radiation, Office of Atmospheric Programs,
U.S. Environmental Protection Agency, Washington, D.C.
EPA (2021b) Greenhouse Gas Reporting Program (GHGRP). 2020 Envirofacts. Subpart HH: Municipal Solid Waste
Landfills and Subpart TT: Industrial Waste Landfills. Available online at: https://www.epa.gov/enviro/greenhouse-
gas-customized-search.
EPA (2021c) "1970 - 2020 Average annual emissions, all criteria pollutants in MS Excel." National Emissions
Inventory (NEI) Air Pollutant Emissions Trends Data. Office of Air Quality Planning and Standards, December 2021.
Available online at: https://www.epa.gov/air-emissions-inventories/air-pollutant-emissions-trends-data.
EPA (1997) Compilation of Air Pollutant Emission Factors, AP-42. Office of Air Quality Planning and Standards, U.S.
Environmental Protection Agency. Research Triangle Park, NC. October 1997.
FHWA (1996 through 2021) Highway Statistics. Federal Highway Administration, U.S. Department of
Transportation, Washington, D.C. Report FHWA-PL-96-024-annual. Available online at:
http://www.fhwa.dot.gov/policy/ohpi/hss/hsspubs.htm.
IEA (2021) CO2 Emissions from Fossil Fuel Combustion - Overview. International Energy Agency. Available online
at: https://www.iea.org/subscribe-to-data-services/co2-emissions-statistics.
References and Abbreviations 10-1
-------
IPCC (2021) Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth
Assessment Report of the Intergovernmental Panel on Climate Change. [Masson-Delmotte, V., P. Zhai, A. Pirani, S.
L. Connors, C. Pean, S. Berger, N. Caud, Y. Chen, L Goldfarb, M. I. Gomis, M. Huang, K. Leitzell, E. Lonnoy, J.B.R.
Matthews, T. K. Maycock, T. Waterfield, O. Yelekgi, R. Yu and B. Zhou (eds.)]. Cambridge University Press. In Press.
IPCC (2019) 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories, e [Buendia, E.,
Guendehou S., Limmeechokachai B., Pipatti R., Rojas Y., Sturgiss R., Tanabe K., Wirth T., (eds.)]. Cambridge
University Press. In Press.
IPCC (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. [Stocker, T.F., D. Qin, G.-K., Plattner, M.
Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom
996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
IPCC (1996) Climate Change 1995: The Science of Climate Change. Intergovernmental Panel on Climate Change.
[J.T. Houghton, LG. Meira Filho, B.A. Callander, N. Harris, A. Kattenberg, and K. Maskell (eds.)]. Cambridge
University Press. Cambridge, United Kingdom.
National Academies of Sciences, Engineering, and Medicine (2018) Improving Characterization of Anthropogenic
Methane Emissions in the United States. Washington, DC: The National Academies Press. Available online at:
https://doi.org/10.17226/2498?.
National Research Council (2010) Verifying Greenhouse Gas Emissions: Methods to Support International Climate
Agreements. Washington, DC: The National Academies Press. Available online at: https://doi.org/10.17226/12883.
NOAA/ESRL (2022a) Trends in Atmospheric Carbon Dioxide. Available online at:
http://www.esrl.noaa.gov/gmd/ccgg/trends/. 05 January 2022.
NOAA/ESRL (2022b) Trends in Atmospheric Methane. Available online at:
https://www.esrl.noaa.gov/gmd/ccgg/trends clW. 05 January 2022.
NOAA/ESRL (2022c) Nitrous Oxide (N2O) hemispheric and global monthly means from the NOAA/ESRL
Chromatograph for Atmospheric Trace Species data from baseline observatories (Barrow, Alaska; Summit,
Greenland; Niwot Ridge, Colorado; Mauna Loa, Hawaii; American Samoa; South Pole). Available online at:
https://www.esrl.noaa.gov/gmd/ccgg/trends n2o/. 05 January 2022.
UNFCCC (2014) Report of the Conference of the Parties on its Nineteenth Session, Held in Warsaw from 11 to 23
November 2013. (FCCC/CP/2013/10/Add.3). January 31, 2014. Available online at:
http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
U.S. Census Bureau (2021) U.S. Census Bureau International Database (IDB). Available online at:
https://www.census.gov/programs-survevs/intemational-programs.html.
Introduction
IPCC (2021) Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth
Assessment Report of the Intergovernmental Panel on Climate Change. [Masson-Delmotte, V., P. Zhai, A. Pirani, S.L
2 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Connors, C. Pean, S. Berger, N. Caud, Y. Chen, L. Goldfarb, M.I. Gomis, M. Huang, K. Leitzell, E. Lonnoy, J.B.R.
Matthews, T.K. Maycock, T. Waterfield, O. Yelekgi, R. Yu, and B. Zhou (eds.)]. Cambridge University Press. In Press
IPCC (2014) Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group III to the Fifth
Assessment Report of the Intergovernmental Panel on Climate Change [Edenhofer, O., R. Pichs-Madruga, Y.
Sokona, J. Minx, E. Farahani, S. Kadner, K. Seyboth, A. Adler, I. Baum, S. Brunner, P. Eickemeier, B. Kriemann, J.
Savolainen, S. Schlomer, C. von Stechow, and T. Zwickel (eds.)]. Cambridge University Press, Cambridge, United
Kingdom and New York, NY, USA, 1435 pp.
IPCC (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 [Stocker, T.F., D. Qin, G.-K. Plattner, M.
Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom
996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
IPCC (2001) Climate Change 2001: The Scientific Basis. Intergovernmental Panel on Climate Change. [J.T. Houghton,
Y. Ding, D.J. Griggs, M. Noguer, P.J. van der Linden, X. Dai, C.A. Johnson, and K. Maskell (eds.)]. Cambridge
University Press. Cambridge, United Kingdom.
IPCC/TEAP (2005) Special Report: Safeguarding the Ozone Layer and the Global Climate System, Chapter 4:
Refrigeration. 2005. Available online at: https://www.ipcc.ch/site/assets/uploads/2018/03/sroc04-l.pdf.
NOAA (2017) Vital Signs of the Planet. Available online at: http://climate.nasa.gov/causes/. Accessed on 9 January
2017.
NOAA/ESRL (2021a) Trends in Atmospheric Carbon Dioxide. Available online at:
https://gml.noaa.gov/ccgg/trends/gr.html. February 2, 2022.
NOAA/ESRL (2021b) Trends in Atmospheric Methane. Available online at: https://gml.noaa.gov/ccgg/trends ch4/.
February 2, 2022.
NOAA/ESRL (2021c) Trends in Atmospheric Nitrous Oxide. Available online at:
https://gml.noaa.gov/ccgg/trends n2o/. February 2, 2022.
NOAA/ESRL (2021d) Trends in Atmospheric Sulfur Hexafluoride. Available online at:
https://gml.noaa.gov/ccgg/trends sf6/. February 2, 2022.
UNEP/WMO (1999) Information Unit on Climate Change. Framework Convention on Climate Change. Available
online at: http://unfccc.int.
UNFCCC (2014) Report of the Conference of the Parties on its nineteenth session, held in Warsaw from 11 to 23
November 2013. (FCCC/CP/2013/10/Add.3). January 31, 2014. Available online at:
http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
USGCRP (2017) Climate Science Special Report: Fourth National Climate Assessment, Volume I. [Wuebbles, D.J.,
D.W. Fahey, K.A. Hibbard, D.J. Dokken, B.C. Stewart, and T.K. Maycock (eds.)]. U.S. Global Change Research
Program, Washington, DC, USA, 470 pp, doi: 10.7930/J0J964J6. Available online at:
https://science2017.globalchange.gov/.
WMO/UNEP (2018) Scientific Assessment of Ozone Depletion: 2018. Available online at:
https://csl.noaa.gov/assessments/ozone/2018.
References and Abbreviations 3
-------
Trends in Greenhouse Gas Emissions
BEA (2021) 2020 Comprehensive Revision of the National Income and Product Accounts: Current-dollar and "real"
GDP, 1929-2020. Bureau of Economic Analysis (BEA), U.S. Department of Commerce, Washington, D.C. Available
online at: http://www.bea.gOv/national/index.htm#gdp.
EIA (2022) Monthly Energy Review, February 2022. Energy Information Administration, U.S. Department of Energy,
Washington, D.C. DOE/EIA-0035(2022/02).
EIA (1991 through 2021) Fuel Oil and Kerosene Sales. Energy Information Administration, U.S. Department of
Energy. Washington, D.C. Available online at: http://www.eia.gov/petroleum/fueloilkerosene.
EIA (2018) "In 2017, U.S. electricity sales fell by the greatest amount since the recession" Available online at:
https://www. eia.gov/todayinenergy/deta il.php?id=35612.
EPA (2022) "Crosswalk of Precursor Gas Categories." U.S. Environmental Protection Agency. April 6, 2022.
EPA (2021a) Light-Duty Automotive Technology, Carbon Dioxide Emissions, and Fuel Economy Trends: 1975 - 2020.
Office of Transportation and Air Quality, U.S. Environmental Protection Agency. Available online at:
https://www.epa.gov/fuel-economv/trends-report.
EPA (2021b) 1970 - 2020 Average annual emissions, all criteria pollutants in MS Excel. National Emissions Inventory
(NEI) Air Pollutant Emissions Trends Data. Office of Air Quality Planning and Standards, May 2021. Available online
at: https://www.epa.gov/air-emissions-inventories/air-pollutant-emissions-trends-data.
IPCC (2021) Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth
Assessment Report of the Intergovernmental Panel on Climate Change. [Masson-Delmotte, V., P. Zhai, A. Pirani, S.
L. Connors, C. Pean, S. Berger, N. Caud, Y. Chen, L Goldfarb, M. I. Gomis, M. Huang, K. Leitzell, E. Lonnoy, J.B.R.
Matthews, T. K. Maycock, T. Waterfield, O. Yelekgi, R. Yu and B. Zhou (eds.)]. Cambridge University Press. In Press.
IPCC (2019) 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories. [Buendia, E.,
Guendehou S., Limmeechokachai B., Pipatti R., Rojas Y., Sturgiss R., Tanabe K., Wirth T., (eds.)]. Cambridge
University Press. In Press.
IPCC (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. [Stocker, T.F., D. Qin, G.-K., Plattner, M.
Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom
996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
U.S. Census Bureau (2021) U.S. Census Bureau International Database (IDB). Available online at:
https://www.census.gov/programs-survevs/international-programs.html.
U.S. Department of Agriculture, National Agricultural Statistics Service (2020) Farm Production Expenditures
Annual Summary. National Agricultural Statistics Service, U.S. Department of Agriculture, Washington DC. Available
online at: https://usda.librarv.cornell.edu/concern/publications/qz20ss48r?locale=en.
4 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Energy
EIA (2022) Monthly Energy Review, February 2022, Energy Information Administration, U.S. Department of Energy,
Washington, DC. DOE/EIA-0035(2022/02).
IEA (2020) Energy related CO2 emissions, 1990-2019, International Energy Agency, Paris. Available online at:
https://www.iea.org/data-and-statistics/charts/energv-related-co2-emissions-1990-2019.
Carbon Dioxide Emissions from Fossil Fuel Combustion
AAR (2008 through 2021) Railroad Facts. Policy and Economics Department, Association of American Railroads,
Washington, D.C. Private communication with Dan Keen.
AISI (2004 through 2020) Annual Statistical Report, American Iron and Steel Institute, Washington, D.C.
APTA (2007 through 2021) Public Transportation Fact Book. American Public Transportation Association,
Washington, D.C. Available online at: http://www.apta.com/resources/statistics/Pages/transitstats.aspx.
APTA (2006) Commuter Rail National Totals. American Public Transportation Association, Washington, D.C.
BEA (2022) Table 1.1.6. Real Gross Domestic Product, Chained 2012 Dollars. Bureau of Economic Analysis (BEA),
U.S. Department of Commerce, Washington, D.C. February 2021. Available online at:
https://apps.bea.gov/iTable/iTable.cfm?reqid=19&step=2#reqid=19&step=2&isuri=l&1921=survey.
BEA (1991 through 2015) Unpublished BE-36 survey data. Bureau of Economic Analysis, U.S. Department of
Commerce. Washington, D.C.
Benson, D. (2002 through 2004) Unpublished data. Upper Great Plains Transportation Institute, North Dakota State
University and American Short Line & Regional Railroad Association.
Browning, L (2020) GHG Inventory EF Development Using Certification Data. Technical Memo, September 2020.
Browning (2019) Updated On-highway Cm and N2O Emission Factors for GHG Inventory. Memorandum from ICF to
Sarah Roberts, Office of Transportation and Air Quality, U.S. Environmental Protection Agency. September 2019.
Browning, L (2018a) Updated Methodology for Estimating Electricity Use from Highway Plug-In Electric Vehicles.
Technical Memo, October 2018.
Browning, L. (2018b) Updated Non-Highway CH4 and N2O Emission Factors for U.S. GHG Inventory. Technical
Memo, November 2018.
Browning, L (2017) Updated Methodology for Estimating CH4 and N2O Emissions from Highway Vehicle Alternative
Fuel Vehicles. Technical Memo, October 2017.
Coffeyville Resources Nitrogen Fertilizers (2012) Nitrogen Fertilizer Operations. Available online at:
http://coffeyvillegroup.com/NitrogenFertilizerOperations/index.html.
Dakota Gasification Company (2006) CO2 Pipeline Route and Designation Information. Bismarck, ND.
DHS (2008) Email Communication. Elissa Kay, Department of Homeland Security and Joe Aamidor, ICF
International. January 11, 2008.
DLA Energy (2021) Unpublished data from the Fuels Automated System (FAS). Defense Logistics Agency Energy,
U.S. Department of Defense. Washington, D.C.
DOC (1991 through 2020) Unpublished Report of Bunker Fuel Oil Laden on Vessels Cleared for Foreign Countries.
Form-563. Foreign Trade Division, Bureau of the Census, U.S. Department of Commerce. Washington, D.C.
DOE (1991 through 2021) Transportation Energy Data Book. Edition 40. Office of Transportation Technologies,
Center for Transportation Analysis, Energy Division, Oak Ridge National Laboratory. ORNL-6978. Personal
References and Abbreviations 5
-------
Communication between Stacy Davis (DOE) and Deep Shah (ICF) for sharing selected tables from the pre-release
version.
DOE (2012) 2010 Worldwide Gasification Database. National Energy Technology Laboratory and Gasification
Technologies Council. Available online at:
http://www.netl.doe.gov/technologies/coalpower/gasification/worlddatabase/index.html. Accessed on 15 March
2012.
DOT (1991 through 2021) Airline Fuel Cost and Consumption. U.S. Department of Transportation, Bureau of
Transportation Statistics, Washington, D.C. DAI-10. Available online at: http://www.transtats.bts.gov/fuel.asp.
Eastman Gasification Services Company (2011) Project Data on Eastman Chemical Company's Chemicals-from-Coal
Complex in Kingsport, TN.
EIA (2022a) Monthly Energy Review, February 2022, Energy Information Administration, U.S. Department of
Energy, Washington, DC. DOE/EIA-0035(2022/02).
EIA (2022b) International Energy Statistics 1980-2020. Energy Information Administration, U.S. Department of
Energy. Washington, D.C. Available online at: https://www.eia.gov/beta/international/.
EIA (2021a) Quarterly Coal Report: July - September 2021. Energy Information Administration, U.S. Department of
Energy. Washington, D.C. DOE/EIA-0121.
EIA (2021b) Form EIA-923 detailed data with previous form data (EIA-906/920), Energy Information Administration,
U.S. Department of Energy. Washington, DC. DOE/EIA. September 2021.
EIA (2021c) Electric Power Annual 2020. Energy Information Administration, U.S. Department of Energy.
Washington, D.C. DOE/EIA-0348(20).
EIA (2021d) Natural Gas Annual 2020. Energy Information Administration, U.S. Department of Energy. Washington,
D.C. DOE/EIA-0131(20).
EIA (2021e) Annual Coal Report 2020. Energy Information Administration, U.S. Department of Energy. Washington,
D.C. DOE/EIA-0584.
EIA (2020b) "Natural gas prices, production, consumption, and exports increased in 2019." Today in Energy.
Available online at: https://www.eia.gov/todayinenergy/detail.php?id=3?892.
EIA (2018) "Both natural gas supply and demand have increased from year-ago levels." Today in Energy. Available
online at: https://www.eia.gov/todavinenergy/detail.php?id=37193.
EIA (1991 through 2020) Fuel Oil and Kerosene Sales. Energy Information Administration, U.S. Department of
Energy. Washington, D.C. Available online at: http://www.eia.gov/petroleum/fueloilkerosene.
EIA (2009a) Emissions of Greenhouse Gases in the United States 2008, Draft Report. Office of Integrated Analysis
and Forecasting, Energy Information Administration, U.S. Department of Energy. Washington, D.C. DOE-EIA-
0573(2009).
EIA (2009b) Manufacturing Consumption of Energy (MECS) 2006. U.S. Department of Energy, Energy Information
Administration, Washington, D.C. Released July 2009.
EIA (2008) Historical Natural Gas Annual, 1930 - 2008. Energy Information Administration, U.S. Department of
Energy. Washington, D.C.
EIA (2007) Personal Communication. Joel Lou, Energy Information Administration and Aaron Beaudette, ICF
International. Residual and Distillate Fuel Oil Consumption for Vessel Bunkering (Both International and Domestic)
for American Samoa, U.S. Pacific Islands, and Wake Island. October 24, 2007.
EIA (2003) Personal Communication. Kent Forsberg, Energy Information Administration and ICF International.
Distillate Fuel Oil Consumption.
6 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
EIA (2002) Alternative Fuels Data Tables. Energy Information Administration, U.S. Department of Energy.
Washington, D.C. Available online at: https://www.eia.gov/renewable/.
EIA (2001) U.S. Coal, Domestic and International Issues. Energy Information Administration, U.S. Department of
Energy. Washington, D.C. March 2001.
EIA (1990-2001) State Energy Data System. Energy Information Administration, U.S. Department of Energy.
Washington, D.C.
Environment and Climate Change Canada (2022) Personal Communication between Environment and Climate
Change Canada and Vincent Camobreco for imported CO2. March 2022.
EPA (2022) Acid Rain Program Dataset 1996-2020. Office of Air and Radiation, Office of Atmospheric Programs,
U.S. Environmental Protection Agency, Washington, D.C.
EPA (2020c) EPA Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2019: Updated Gasoline and Diesel
Fuel CO2 Emission Factors - Memo.
EPA (2021a) Light-Duty Automotive Technology, Carbon Dioxide Emissions, and Fuel Economy Trends: 1975 - 2020.
Office of Transportation and Air Quality, U.S. Environmental Protection Agency. Available online at:
https://www.epa.gov/fuel-economv/trends-report.
EPA (2019b) MOtor Vehicle Emissions Simulator (MOVES) 2014b. Office of Transportation and Air Quality, U.S.
Environmental Protection Agency, Washington, D.C. Available online at: https://www.epa.gov/moves.
EPA (2021b) MOtor Vehicle Emissions Simulator (MOVES3). Office of Transportation and Air Quality, U.S.
Environmental Protection Agency, Washington, D.C. Available online at: https://www.epa.gov/moves.
EPA (2021c) The Emissions & Generation Resource Integrated Database (eGRID) 2019 Technical Support
Document. Clean Air Markets Division, Office of Atmospheric Programs, U.S. Environmental Protection Agency,
Washington, D.C. Available Online at: https://www.epa.gov/sites/default/files/2021-
02/documents/egrid2019 technical guide.pdf
EPA (2010) Carbon Content Coefficients Developed for EPA's Mandatory Reporting Rule. Office of Air and
Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
Erickson, T. (2003) Plains CO2 Reduction (PCOR) Partnership. Presented at the Regional Carbon Sequestration
Partnership Meeting Pittsburgh, Pennsylvania, Energy and Environmental Research Center, University of North
Dakota. November 3, 2003.
FAA (2022) Personal Communication between FAA and John Steller, Mausami Desai, and Vincent Camobreco for
aviation emissions estimates from the Aviation Environmental Design Tool (AEDT). March 2022.
FHWA (1996 through 2021) Highway Statistics. Federal Highway Administration, U.S. Department of
Transportation, Washington, D.C. Report FHWA-PL-96-023-annual. Available online at:
http://www.fhwa.dot.gov/policy/ohpi/hss/hsspubs.htm.
FHWA (2015) Off-Highway and Public-Use Gasoline Consumption Estimation Models Used in the Federal Highway
Administration, Publication Number FHWA-PL-17-012. Available online at:
https://www.fhwa.dot.gov/policyinformation/pubs/pll7012.pdf.
Fitzpatrick, E. (2002) The Weyburn Project: A Model for International Collaboration.
FRB (2021) Industrial Production and Capacity Utilization. Federal Reserve Statistical Release, G.17, Federal
Reserve Board. Available online at: http://www.federalreserve.gov/releases/G17/tablel 2.htm.
Gaffney, J. (2007) Email Communication. John Gaffney, American Public Transportation Association and Joe
Aamidor, ICF International. December 17, 2007.
ICF (2020) Potential Improvements to Energy Sector Hydrocarbon Gas Liquid Carbon Content Coefficients.
Memorandum from ICF to Vincent Camobreco, U.S. Environmental Protection Agency. December 7, 2020.
References and Abbreviations 7
-------
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Marland, G. and A. Pippin (1990) "United States Emissions of Carbon Dioxide to the Earth's Atmosphere by
Economic Activity." Energy Systems and Policy, 14(4):323.
SAIC/EIA (2001) Monte Carlo Simulations of Uncertainty in U.S. Greenhouse Gas Emission Estimates. Final Report.
Prepared by Science Applications International Corporation (SAIC) for Office of Integrated Analysis and Forecasting,
Energy Information Administration, U.S. Department of Energy. Washington, D.C. June 22, 2001.
U.S. Aluminum Association (USAA) (2008 through 2020) U.S. Primary Aluminum Production. U.S. Aluminum
Association, Washington, D.C.
USAF (1998) Fuel Logistics Planning. U.S. Air Force: AFPAM23-221. May 1,1998.
U.S. Census Bureau (2001 through 2011) Current Industrial Reports Fertilizer Materials and Related Products:
Annual Summary. Available online at: https://www.census.gov/data/tables/time-series/econ/cir/mq325b.html.
United States Geological Survey (USGS) (2020a) 2020 Mineral Commodity Summaries: Aluminum. U.S. Geological
Survey, Reston, VA.
USGS (2021b) 2021 Mineral Commodity Summary: Titanium and Titanium Dioxide. U.S. Geological Survey, Reston,
VA.
USGS (2019) 2017Mineral Yearbook: Aluminum. U.S. Geological Survey, Reston, VA
USGS (2014 through 2021a) Mineral Industry Surveys: Silicon. U.S. Geological Survey, Reston, VA.
USGS (2014 through 2021b) Mineral Commodity Summary, Lead. U.S. Geological Survey, Reston, VA.
USGS (2014 through 2019) Minerals Yearbook: Nitrogen [Advance Release], Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/nitrogen/.
USGS (1991 through 2020) Minerals Yearbook - Iron and Steel Scrap. U.S. Geological Survey, Reston, VA.
USGS (1991 through 2015a) Minerals Yearbook: Manufactured Abrasives Annual Report. U.S. Geological Survey,
Reston, VA. Available online at: http://minerals.usgs.gov/minerals/pubs/commodity/abrasives/.
USGS (1991 through 2015b) Minerals Yearbook: Titanium. U.S. Geological Survey, Reston, VA.
USGS (1991 through 2015c) Minerals Yearbook: Silicon Annual Report. U.S. Geological Survey, Reston, VA. Available
online at: http://minerals.usgs.gov/minerals/pubs/commodity/silicon/.
USGS (1996 through 2013) Minerals Yearbook: Silicon. U.S. Geological Survey, Reston, VA.
USGS (1995 through 2013) Minerals Yearbook: Lead Annual Report. U.S. Geological Survey, Reston, VA.
USGS (1995,1998, 2000, 2001, 2002, 2007) Minerals Yearbook: Aluminum Annual Report. U.S. Geological Survey,
Reston, VA.
Stationary Combustion (excluding C02)
EIA (2022a) Monthly Energy Review, February 2022. Energy Information Administration, U.S. Department of
Energy. Washington, D.C. DOE/EIA-0035(2022/02).
EIA (2022b) International Energy Statistics 1980-2020. Energy Information Administration, U.S. Department of
Energy. Washington, D.C. Available online at: https://www.eia.gov/international/data/world.
EPA (2022) Acid Rain Program Dataset 1996-2020. Office of Air and Radiation, Office of Atmospheric Programs,
U.S. Environmental Protection Agency, Washington, D.C.
8 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
EPA (2021) MOtor Vehicle Emissions Simulator (MOVES3). Office of Transportation and Air Quality, U.S.
Environmental Protection Agency, Washington, D.C. Available online at: httpsi//www.epa.gov/moves.
EPA (1997) Compilation of Air Pollutant Emission Factors, AP-42. Office of Air Quality Planning and Standards, U.S.
Environmental Protection Agency. Research Triangle Park, NC. October 1997.
FHWA (1996 through 2021) Highway Statistics. Federal Highway Administration, U.S. Department of
Transportation, Washington, D.C. Report FHWA-PL-96-023-annual. Available online at:
http://www.fliwa.dot.gov/policy/ohpi/hss/hsspubs.htm.
ICF (2020) Potential Improvements to Energy Sector Hydrocarbon Gas Liquid Carbon Content Coefficients.
Memorandum from ICF to Vincent Camobreco, U.S. Environmental Protection Agency. December 7, 2020.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
SAIC/EIA (2001) Monte Carlo Simulations of Uncertainty in U.S. Greenhouse Gas Emission Estimates. Final Report.
Prepared by Science Applications International Corporation (SAIC) for Office of Integrated Analysis and Forecasting,
Energy Information Administration, U.S. Department of Energy. Washington, D.C. June 22, 2001.
Mobile Combustion (excluding C02)
AAR (2008 through 2021) Railroad Facts. Policy and Economics Department, Association of American Railroads,
Washington, D.C. Private communication with Dan Keen.
ANL (2021) The Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation Model (GREET2021).
Argonne National Laboratory. Available online at: https://greet.es.anl.gov.
ANL (2006) Argonne National Laboratory (2006) GREET model Version 1.7. June 2006.
APTA (2007 through 2021) Public Transportation Fact Book. American Public Transportation Association,
Washington, D.C. Available online at: http://www.apta.com/resources/statistics/Pages/transitstats.aspx.
APTA (2006) Commuter Rail National Totals. American Public Transportation Association, Washington, D.C.
Available online at: http://www.apta.com/research/stats/rail/crsum.cfm.
BEA (1991 through 2015) Unpublished BE-36 survey data. Bureau of Economic Analysis, U.S. Department of
Commerce. Washington, D.C.
Benson, D. (2002 through 2004) Personal communication. Unpublished data developed by the Upper Great Plains
Transportation Institute, North Dakota State University and American Short Line & Regional Railroad Association.
Browning (2020a) GHG Inventory EF Development Using Certification Data. Memorandum from ICF to Sarah
Roberts, Office of Transportation and Air Quality, U.S. Environmental Protection Agency. September 2020.
Browning, L. (2020b). Updated Methane and Nitrous Oxide Emission Factors for Non-Road Sources and On-road
Motorcycles. Technical Memorandum from ICF International to Sarah Roberts, Office of Transportation and Air
Quality, U.S. Environmental Protection Agency, September 2020.
Browning, L. (2019) Updated On-highway CH4 and N2O Emission Factors for GHG Inventory. Memorandum from ICF
to Sarah Roberts and Justine Geidosch, Office of Transportation and Air Quality, U.S. Environmental Protection
Agency. September 2019.
Browning, L. (2018a). Updated Methodology for Estimating Electricity Use from Highway Plug-In Electric Vehicles.
Technical Memorandum from ICF International to Sarah Roberts and Justine Geidosch, Office of Transportation
and Air Quality, U.S. Environmental Protection Agency. October 2018.
References and Abbreviations 9
-------
Browning, L (2018b) Updated Non-Highway Cm and N2O Emission Factors for U.S. GHG Inventory. Technical
Memorandum from ICF International to Sarah Roberts and Justine Geidosch, Office of Transportation and Air
Quality, U.S. Environmental Protection Agency. November 2018.
Browning, L (2017) Updated Methodology for Estimating CH4 and N2O Emissions from Highway Vehicle Alternative
Fuel Vehicles. Technical Memorandum from ICF International to Sarah Roberts and Justine Geidosch, Office of
Transportation and Air Quality, U.S. Environmental Protection Agency. October 2017.
Browning, L (2009) Personal communication with Lou Browning, "Suggested New Emission Factors for Marine
Vessels," ICF International.
Browning, L (2005) Personal communication with Lou Browning, "Emission control technologies for diesel highway
vehicles specialist," ICF International.
DHS (2008) Email Communication. Elissa Kay, Department of Homeland Security and Joe Aamidor, ICF
International. January 11, 2008.
DLA Energy (2021) Unpublished data from the Defense Fuels Automated Management System (DFAMS). Defense
Energy Support Center, Defense Logistics Agency, U.S. Department of Defense. Washington, D.C.
DOC (1991 through 2020) Unpublished Report of Bunker Fuel Oil Laden on Vessels Cleared for Foreign Countries.
Form-563. Foreign Trade Division, Bureau of the Census, U.S. Department of Commerce. Washington, D.C.
DOE (1993 through 2021) Transportation Energy Data Book Edition 40. Office of Transportation Technologies,
Center for Transportation Analysis, Energy Division, Oak Ridge National Laboratory. Personal Communication
between Stacy Davis (DOE) and Deep Shah (ICF) for sharing selected tables from the pre-release version.
DOT (1991 through 2021) Airline Fuel Cost and Consumption. U.S. Department of Transportation, Bureau of
Transportation Statistics, Washington, D.C. DAI-10. Available online at: http://www.transtats.bts.gov/fuel.asp.
EIA (2022) Monthly Energy Review, February 2022, Energy Information Administration, U.S. Department of Energy,
Washington, D.C. DOE/EIA-0035(2022/02).
EIA (2021f) Natural Gas Annual 2020. Energy Information Administration, U.S. Department of Energy, Washington,
D.C. DOE/EIA-0131(11).
EIA (1991 through 2020) Fuel Oil and Kerosene Sales. Energy Information Administration, U.S. Department of
Energy. Washington, D.C. Available online at: http://www.eia.gov/petroleum/fueloilkerosene.
EIA (2016) "Table 3.1: World Petroleum Supply and Disposition." International Energy Annual. Energy Information
Administration, U.S. Department of Energy. Washington, D.C. Available online at:
https://www.eia.gOv/cfapps/ipdbproiect/l EDlndex3.cfm?tid=5&pid=66&aid=13.
EIA (2011) Annual Energy Review 2010. Energy Information Administration, U.S. Department of Energy,
Washington, D.C. DOE/EIA-0384(2011). October 19, 2011.
EIA (2007) Personal Communication. Joel Lou, Energy Information Administration and Aaron Beaudette, ICF
International. Residual and Distillate Fuel Oil Consumption for Vessel Bunkering (Both International and Domestic)
for American Samoa, U.S. Pacific Islands, and Wake Island. October 24, 2007.
EIA (2002) Alternative Fuels Data Tables. Energy Information Administration, U.S. Department of Energy,
Washington, D.C. Available online at: http://www.eia.doe.gov/fuelrenewable.html.
EPA (2021a) Light-Duty Automotive Technology, Carbon Dioxide Emissions, and Fuel Economy Trends: 1975 - 2020.
Office of Transportation and Air Quality, U.S. Environmental Protection Agency. Available online at:
https://www.epa.gov/fuel-economv/trends-report.
EPA (2021b) Motor Vehicle Emissions Simulator (MOVES3). Office of Transportation and Air Quality, U.S.
Environmental Protection Agency. Available online at: https://www.epa.gov/moves.
EPA (2021c) Confidential Engine Family Sales Data Submitted to EPA by Manufacturers. Office of Transportation
and Air Quality, U.S. Environmental Protection Agency.
10 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
EPA (2021d) Annual Certification Test Results Report. Office of Transportation and Air Quality, U.S. Environmental
Protection Agency. Available online at: https://www.epa.gov/compliance-and-fuel-economy-data/annual-
certification-test-data-vehicles-and-engines.
EPA (2016g) "1970 - 2015 Average annual emissions, all criteria pollutants in MS Excel." National Emissions
Inventory (NEI)Air Pollutant Emissions Trends Data. Office of Air Quality Planning and Standards. Available online
at: https://www.epa.gov/air-emissions-inventories/air-pollutant-emissions-trends-data.
EPA (2004) Mobile6.2 Vehicle Emission Modeling Software. Office of Mobile Sources, U.S. Environmental
Protection Agency, Ann Arbor, Michigan.
EPA (1999a) Emission Facts: The History of Reducing Tailpipe Emissions. Office of Mobile Sources. May 1999. EPA
420-F-99-017. Available online at: https://www.epa.gov/nscep.
EPA (1999b) Regulatory Announcement: EPA's Program for Cleaner Vehicles and Cleaner Gasoline. Office of Mobile
Sources. December 1999. EPA420-F-99-051. Available online at:
https://nepis.epa.gov/Exe/ZvPDF.cgi/P100 lZ9W.PDF?Dockev=P1001Z9W.PDF.
EPA (1998) Emissions of Nitrous Oxide from Highway Mobile Sources: Comments on the Draft Inventory of U.S.
Greenhouse Gas Emissions and Sinks, 1990-1996. Office of Mobile Sources, Assessment and Modeling Division,
U.S. Environmental Protection Agency. August 1998. EPA420-R-98-009.
EPA (1994a) Automobile Emissions: An Overview. Office of Mobile Sources. August 1994. EPA 400-F-92-007.
Available online at: https://www.epa.gov/nscep.
EPA (1994b) Milestones in Auto Emissions Control. Office of Mobile Sources. August 1994. EPA 400-F-92-014.
Available online at: https://www.epa.gov/nscep.
EPA (1993) Automobiles and Carbon Monoxide. Office of Mobile Sources. January 1993. EPA 400-F-92-005.
Available online at: https://www.epa.gov/nscep.
Esser, C. (2003 through 2004) Personal Communication with Charles Esser, Residual and Distillate Fuel Oil
Consumption for Vessel Bunkering (Both International and Domestic) for American Samoa, U.S. Pacific Islands, and
Wake Island.
FAA (2022) Personal Communication between FAA and John Steller, Mausami Desai and Vincent Camobreco for
aviation emission estimates from the Aviation Environmental Design Tool (AEDT). March 2022.
FHWA (1996 through 2021) Highway Statistics. Federal Highway Administration, U.S. Department of
Transportation, Washington, D.C. Report FHWA-PL-96-023-annual. Available online at:
http://www.fhwa.dot.gov/policy/ohpi/hss/hsspubs.htm.
FHWA (2015) Off-Highway and Public-Use Gasoline Consumption Estimation Models Used in the Federal Highway
Administration, Publication Number FHWA-PL-17-012. Available online at:
https://www.fhwa.dot.gov/policyinformation/pubs/pll7012.pdf.
Gaffney, J. (2007) Email Communication. John Gaffney, American PublicTransportation Association and Joe
Aamidor, ICF International. December 17, 2007.
HybridCars.com (2019). Monthly Plug-In Electric Vehicle Sales Dashboard, 2010-2018. Available online at
https://www.hybridcars.com/december-2017-dashboard/.
ICF (2006a) Revised Gasoline Vehicle EFs for LEV and Tier 2 Emission Levels. Memorandum from ICF International to
John Davies, Office of Transportation and Air Quality, U.S. Environmental Protection Agency. November 2006.
ICF (2006b) Revisions to Alternative Fuel Vehicle (AFV) Emission Factors for the U.S. Greenhouse Gas Inventory.
Memorandum from ICF International to John Davies, Office of Transportation and Air Quality, U.S. Environmental
Protection Agency. November 2006.
ICF (2004) Update of Methane and Nitrous Oxide Emission Factors for On-Highway Vehicles. Final Report to U.S.
Environmental Protection Agency. February 2004.
References and Abbreviations 11
-------
ICF (2017b) Updated Non-Highway Cm and N2O Emission Factors for U.S. GHG Inventory. Memorandum from ICF
to Sarah Roberts and Justine Geidosch, Office of Transportation and Air Quality, U.S. Environmental Protection
Agency. October 2017.
Lipman, T. and M. Delucchi (2002) "Emissions of Nitrous Oxide and Methane from Conventional and Alternative
Fuel Motor Vehicles." Climate Change, 53:477-516.
SAE (2010) Utility Factor Definitions for Plug-In Hybrid Electric Vehicles Using Travel Survey Data. Society of
Automotive Engineers. Report J2841, Available online at: https://www.sae.org/standards/content/i2841 201009/.
Raillnc (2014 through 2021) Raillnc Short line and Regional Traffic Index. Carloads Originated Year-to-Date.
December 2021. Available online at: https://www.railinc.com/rportal/railinc-indexes.
Santoni, G., B. Lee, E. Wood, S. Herndon, R. Miake-Lye, S. Wofsy, J. McManus, D. Nelson, M. Zahniser (2011)
Aircraft emissions of methane and nitrous oxide during the alternative aviation fuel experiment. Environ Sci
Technol. 2011 Aug 15; 45(16):7075-82.
U.S. Census Bureau (2000) Vehicle Inventory and Use Survey. U.S. Census Bureau, Washington, D.C. Database CD-
EC97-VIUS.
Whorton, D. (2006 through 2014) Personal communication, Class II and III Rail energy consumption, American
Short Line and Regional Railroad Association.
Carbon Emitted from Non-Energy Uses of Fossil Fuels
ACC (2021a) "U.S. Resin Production & Sales 2020 vs. 2019." Available online at:
https://www.americanchemistry.com/chemistry-in-america/data-industry-statistics/statistics-on-the-plastic-
resins-industry/resources/u.s.-resin-production-sales-2020-vs.-2019.
ACC (2021b) "Guide to the Business of Chemistry, 2021," American Chemistry Council.
ACC (2020) "U.S. Resin Production & Sales 2019 vs. 2018." Available online at:
https://www.americanchemistrv.com/chemistrv-in-america/chemistrv-in-everyday-products/plastics.
ACC (2019) "U.S. Resin Production & Sales 2018 vs. 2017." Available online at:
https://www.americanchemistrv.com/chemistry-in-america/chemistrv-in-everyday-products/plasticsf.
ACC (2018) "U.S. Resin Production & Sales 2017 vs. 2016." Available online at:
https://www.americanchemistry.com/chemistrv-in-america/chemistry-in-everyday-products/plastics.
ACC (2017) "U.S. Resin Production & Sales 2016 vs. 2015."
ACC (2016) "U.S. Resin Production & Sales 2015 vs. 2014."
ACC (2015) "PIPS Year-End Resin Statistics for 2014 vs. 2013: Production, Sales and Captive Use." Available online
at: https://www.americanchemistry.com/chemistry-in-america/data-industry-statistics/statistics-on-the-plastic-
resins-industry/resin-report-subscr iptions.
ACC (2014) "U.S. Resin Production & Sales: 2013 vs. 2012," American Chemistry Council. Available online at:
http://www.americanchemistry.com/Jobs/EconomicStatistics/Plastics-Statistics/Production-and-Sales-Data-by-
Resin.pdf.
ACC (2013) "U.S. Resin Production & Sales: 2012 vs. 2011," American Chemistry Council. Available online at:
http://www.americanchemistry.com/Jobs/EconomicStatistics/Plastics-Statistics/Production-and-Sales-Data-by-
Resin.pdf.
ACC (2003-2011) "PIPS Year-End Resin Statistics for 2010: Production, Sales and Captive Use." Available online at:
http://www.americanchemistry.com/Jobs/EconomicStatistics/Plastics-Statistics/Production-and-Sales-Data-by-
Resin.pdf.
12 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Bank of Canada (2021) Financial Markets Department Year Average of Exchange Rates. Available online at:
https://www.bankofcanada.ca/rates/exchange/annual-average-exchangerates/#download.
Bank of Canada (2020) Financial Markets Department Year Average of Exchange Rates. Available online at:
https://www.bankofcanada.ca/rates/exchange/annual-average-exchange-rates/#download.
Bank of Canada (2019) Financial Markets Department Year Average of Exchange Rates. Available online at:
https://www.bankofcanada.ca/rates/exchange/annual-average-exchange-rates/#download.
Bank of Canada (2018) Financial Markets Department Year Average of Exchange Rates. Available online at:
https://www.bankofcanada.ca/rates/exchange/annual-average-exchange-rates/.
Bank of Canada (2017) Financial Markets Department Year Average of Exchange Rates. Available online at:
https://www.bankofcanada.ca/rates/exchange/legacy-noon-and-closing-rates/.
Bank of Canada (2016) Financial Markets Department Year Average of Exchange Rates. Available online at:
https://www.bankofcanada.ca/rates/exchange/legacv-noon-and-closing-rates/.
Bank of Canada (2014) Financial Markets Department Year Average of Exchange Rates. Available online at:
https://www.bankofcanada.ca/rates/exchange/legacv-noon-and-closing-rates/.
Bank of Canada (2013) Financial Markets Department Year Average of Exchange Rates. Available online at:
https://www.bankofcanada.ca/rates/exchange/legacv-noon-and-closing-rates/.
Bank of Canada (2012) Financial Markets Department Year Average of Exchange Rates. Available online at:
https://www.bankofcanada.ca/rates/exchange/legacv-noon-and-closing-rates/.
EIA (2021a) EIA Manufacturing Consumption of Energy (MECS) 2018. U.S. Department of Energy, Energy
Information Administration, Washington, D.C.
EIA (2021b) Monthly Energy Review, November 2021. Energy Information Administration, U.S. Department of
Energy, Washington, D.C. DOE/EIA-0035 (2021/11).
EIA (2020) Glossary. Energy Information Administration, U.S. Department of Energy, Washington, D.C. Available
online at: https://www. eia.gov/tools/glossary/index. php?id=N#nat Gas Liquids.
EIA (2019) Personal communication between EIA and ICF on November 11, 2019.
EIA (2017) EIA Manufacturing Consumption of Energy (MECS) 2014. U.S. Department of Energy, Energy Information
Administration, Washington, D.C.
EIA (2013) EIA Manufacturing Consumption of Energy (MECS) 2010. U.S. Department of Energy, Energy Information
Administration, Washington, D.C.
EIA (2010) EIA Manufacturing Consumption of Energy (MECS) 2006. U.S. Department of Energy, Energy Information
Administration, Washington, D.C.
EIA (2005) EIA Manufacturing Consumption of Energy (MECS) 2002. U.S. Department of Energy, Energy Information
Administration, Washington, D.C.
EIA (2001) EIA Manufacturing Consumption of Energy (MECS) 1998. U.S. Department of Energy, Energy Information
Administration, Washington, D.C.
EIA (1997) EIA Manufacturing Consumption of Energy (MECS) 1994. U.S. Department of Energy, Energy Information
Administration, Washington, D.C.
EIA (1994) EIA Manufacturing Consumption of Energy (MECS) 1991. U.S. Department of Energy, Energy Information
Administration, Washington, D.C.
EPA (2021a) Resource Conservation and Recovery Act (RCRA) Info, Biennial Report, GM Form (Section 2- Onsite
Management) and WR Form.
References and Abbreviations 13
-------
EPA (2021b) "Criteria pollutants National Tier 1 for 1970 - 2020." National Emissions Inventory (NEI) Air Pollutant
Emissions Trends Data. Office of Air Quality Planning and Standards, March 2021. Available online at:
https://www.epa.gov/air-emissions-inventories/air-pollutant-emissions-trends-data.
EPA (2019) Advancing Sustainable Materials Management: 2016 and 2017 Data Tables. Office of Land and
Emergency Management, U.S. Environmental Protection Agency. Washington, D.C. Available online at:
https://www.epa.gov/sites/production/files/2019-
ll/documents/2016 and 2017 facts and figures data tables O.pdf.
EPA (2018a) Advancing Sustainable Materials Management: Facts and Figures 2015, Assessing Trends in Material
Generation, Recycling and Disposal in the United States. Washington, D.C.
EPA (2018b) Resource Conservation and Recovery Act (RCRA) Info, Biennial Report, GM Form (Section 2- Onsite
Management) and WR Form.
EPA (2017) EPA's Pesticides Industry Sales and Usage, 2008 - 2012 Market Estimates. Available online at:
https://www.epa.gov/sites/production/files/2017-01/documents/pesticides-industry-sales-usage-2016 O.pdf.
Accessed September 2017.
EPA (2016a) Advancing Sustainable Materials Management: 2014 Facts and Figures Fact Sheet. Office of Solid
Waste and Emergency Response, U.S. Environmental Protection Agency, Washington, D.C. Available online at:
https://www.epa.gov/sites/production/files/2016-ll/documents/2014 smmfactsheet 508.pdf.
EPA (2016b) Resource Conservation and Recovery Act (RCRA) Info, Biennial Report, GM Form (Section 2- Onsite
Management) and WR Form.
EPA (2015) Resource Conservation and Recovery Act (RCRA) Info, Biennial Report, GM Form (Section 2- Onsite
Management) and WR Form.
EPA (2014a) Municipal Solid Waste in the United States: 2012 Facts and Figures. Office of Solid Waste and
Emergency Response, U.S. Environmental Protection Agency, Washington, D.C. Available online at:
https://www.epa.gov/sites/default/files/2015-09/documents/2012 msw dat tbls.pdf.
EPA (2014b) Chemical Data Access Tool (CDAT). U.S. Environmental Protection Agency, June 2014. Available online
at: https://edg.epa.gov/metadata/catalog/search/resource/details.page?uuid=%7B2D73C764-6919-4Q4D-8C9B-
61869B3330D6%7D. Accessed January 2015.
EPA (2013a) Municipal Solid Waste in the United States: 2011 Facts and Figures. Office of Solid Waste and
Emergency Response, U.S. Environmental Protection Agency, Washington, D.C. Available online at:
http://www.epa.gov/epaoswer/non-hw/muncpl/msw99.litm.
EPA (2013b) Resource Conservation and Recovery Act (RCRA) Info, Biennial Report, GM Form (Section 2- Onsite
Management) and WR Form.
EPA (2011) EPA's Pesticides Industry Sales and Usage, 2006 and 2007 Market Estimates. Available online at:
https://www.epa.gov/pesticides/pesticides-industry-sales-and-usage-2006-and-2007-market-estimates. Accessed
January 2012.
EPA (2009) Biennial Reporting System (BRS) Database. U.S. Environmental Protection Agency, Envirofacts
Warehouse. Washington, D.C. Available online at: https://www.epa.gov/enviro/br-search. Data for 2001-2007 are
current as of Sept. 9, 2009.
EPA (2004) EPA's Pesticides Industry Sales and Usage, 2000 and 2001 Market Estimates. Available online at:
https://nepis.epa.gov/Exe/Z.yPURL.cgi?Dockev=3000659P.TXT. Accessed September 2006.
EPA (2002) EPA's Pesticides Industry Sales and Usage, 1998 and 1999 Market Estimates, Table 3.6. Available online
at https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=200001G5.TXT. Accessed July 2003.
EPA (2001) AP 42, Volume I, Fifth Edition. Chapter 11: Mineral Products Industry. Available online at:
http://www.epa.gov/ttn/chief/ap42/chll/index.html.
14 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
EPA (2000a) Biennial Reporting System (BRS). U.S. Environmental Protection Agency, Envirofacts Warehouse.
Washington, D.C. Available online at: https://www.epa.gov/enviro/br search.
EPA (2000b) Toxics Release Inventory, 1998. U.S. Environmental Protection Agency, Office of Environmental
Information, Office of Information Analysis and Access, Washington, D.C. Available online at:
https://enviro.epa.gov/triexplorer/tri release.chemical.
EPA (1999) EPA's Pesticides Industry Sales and Usage, 1996-1997 Market Estimates. Available online at:
https://nepis.epa.gov/Exe/ZvPURL.cgi?Dockey=2000011LTXT.
EPA (1998) EPA's Pesticides Industry Sales and Usage, 1994-1995 Market Estimates. Available online at:
http://www.epa.gov/oppbeadl/pestsales/95pestsales/market estimatesl995.pdf.
FEB (2013) Fiber Economics Bureau, as cited in C&EN (2013) Lackluster Year for Chemical Output: Production
stayed flat or dipped in most world regions in 2012. Chemical &Engineering News, American Chemical Society, 1
July. Available online at: http://www.cen-online.org.
FEB (2012) Fiber Economics Bureau, as cited in C&EN (2012) Too Quiet After the Storm: After a rebound in 2010,
chemical production hardly grew in 2011. Chemical & Engineering News, American Chemical Society, 2 July.
Available online at: http://www.cen-online.org.
FEB (2011) Fiber Economics Bureau, as cited in C&EN (2011) Output Ramps up in all Regions. Chemical Engineering
News, American Chemical Society, 4 July. Available online at: http://www.cen-online.org.
FEB (2010) Fiber Economics Bureau, as cited in C&EN (2010) Output Declines in U.S., Europe. Chemical &
Engineering News, American Chemical Society, 6 July. Available online at: http://www.cen-online.org.
FEB (2009) Fiber Economics Bureau, as cited in C&EN (2009) Chemical Output Slipped In Most Regions Chemical &
Engineering News, American Chemical Society, 6 July. Available online at: http://www.cen-online.org.
FEB (2007) Fiber Economics Bureau, as cited in C&EN (2007) Gains in Chemical Output Continue. Chemical &
Engineering News, American Chemical Society. July 2, 2007. Available online at: http://www.cen-online.org.
FEB (2005) Fiber Economics Bureau, as cited in C&EN (2005) Production: Growth in Most Regions Chemical &
Engineering News, American Chemical Society, 11 July. Available online at: http://www.cen-online.org.
FEB (2003) Fiber Economics Bureau, as cited in C&EN (2003) Production Inches Up in Most Countries, Chemical &
Engineering News, American Chemical Society, 7 July. Available online at: http://www.cen-online.org.
FEB (2001) Fiber Economics Bureau, as cited in ACS (2001) Production: slow gains in output of chemicals and
products lagged behind U.S. economy as a whole Chemical & Engineering News, American Chemical Society, 25
June. Available online at: http://pubs.acs.org/cen.
Financial Planning Association (2006) Canada/US Cross-Border Tools: US/Canada Exchange Rates. Available online
at: http://www.fpanet.org/global/planners/US Canada ex rates.cfm. Accessed on August 16, 2006.
Gosselin, Smith, and Hodge (1984) "Clinical Toxicology of Commercial Products." Fifth Edition, Williams & Wilkins,
Baltimore.
ICIS (2016) "Production issues force US melamine plant down" Available online at:
https://www.icis.com/resources/news/2016/05/03/9994556/production-issues-force-us-melamine-plant-down/.
ICIS (2008) "Chemical profile: Melamine" Available online at:
https://www.icis.com/resources/news/2008/12/01/9174886/chemical-profile-melamine/. Accessed November,
2017.
IISRP (2003) "IISRP Forecasts Moderate Growth in North America to 2007" International Institute of Synthetic
Rubber Producers, Inc. New Release. Available online at: http://www.iisrp.com/press-releases/20Q3-Press-
Releases/IISRP-NA-Forecast-03-07.html.
IISRP (2000) "Synthetic Rubber Use Growth to Continue Through 2004, Says IISRP and RMA" International Institute
of Synthetic Rubber Producers press release.
References and Abbreviations 15
-------
INEGI (2006) Produccion bruta total de las unidades economicas manufactureras por Subsector, Rama, Subrama y
Clase de actividad. Available online at:
http://www.inegi.gob.mx/est/contenidos/espanol/provectos/censos/ce2004/tb manufacturas.asp. Accessed on
August 15, 2006.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Marland, G., and R.M. Rotty (1984) "Carbon dioxide emissions from fossil fuels: A procedure for estimation and
results for 1950-1982," Tellus 36b:232-261.
NPRA (2002) North American Wax - A Report Card. Available online at:
http://www.npra.org/members/publications/papers/lybes/LW-02-126.pdf.
RMA (2018) 2017 U.S. Scrap Tire Management Summary. Rubber Manufacturers Association, Washington, D.C. July
2018.
RMA (2016) 2015 U.S. Scrap Tire Management Summary. Rubber Manufacturers Association, Washington, D.C.
August 2016.
RMA (2014) 2013 U.S. Scrap Tire Management Summary. Rubber Manufacturers Association, Washington, D.C.
November 2014.
RMA (2011) U.S. Scrap Tire Management Summary: 2005-2009. Rubber Manufacturers Association, Washington,
D.C. October 2011, updated September 2013.
RMA (2009) "Scrap Tire Markets: Facts and Figures - Scrap Tire Characteristics." Rubber Manufacturers
Association., Washington D.C. Available online at:
http://www.rma.org/scrap tires/scrap tire markets/scrap tire characteristics/. Accessed on 17 September 2009.
U.S. Census Bureau (2021) 2017 Economic Census. Available online at:
https://www.census.gov/data/tables/2017/econ/economic-census/naics-sector-31-33.html. Accessed October
2021.
U.S. Census Bureau (2014) 2012 Economic Census. Available online at:
http://www.censys.gov/econ/census/schedule/whats been released.html. Accessed November 2014.
U.S. Census Bureau (2009) Soap and Other Detergent Manufacturing: 2007. Available online at:
http://smpbffl.dsd.census.gov/TheDataWeb HotReport/servlet/HotReportEngineServlet?emailname=vhg)boc&fil
ename=mfgl. hrml&20071204152004. Var. NAICS2002=325611&fo rward=20071204152004. Var.NAICS2002.
U.S. Census Bureau (2004) Soap and Other Detergent Manufacturing: 2002. Issued December 2004. EC02-31I-
325611 (RV). Available online at: http://www.census.gov/prod/ec02/ec0231i325611.pdf.
U.S. Census Bureau (1999) Soap and Other Detergent Manufacturing: 1997. Available online at:
http://www.census.gov/epcd/www/ec97stat.htm.
U.S. International Trade Commission (2021) "Interactive Tariff and Trade DataWeb: Quick Query." Available online
at: http://dataweb.usitc.gov/. Accessed September 2021.
Incineration of Waste
ArSova, Ljupka, Rob van Haaren, Nora Goldstein, Scott M. Kaufman, and Nickolas J. Themelis (2008) "16th Annual
BioCycle Nationwide Survey: The State of Garbage in America" BioCycle, JG Press, Emmaus, PA. December.
Bahor, B (2009) Covanta Energy's public review comments re: Draft Inventory of U.S. Greenhouse Gas Emissions
and Sinks: 1990-2007. Submitted via email on April 9, 2009 to Leif Hockstad, U.S. EPA.
16 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
De Soete, G.G. (1993) "Nitrous Oxide from Combustion and Industry: Chemistry, Emissions and Control." In A. R.
Van Amstel, (ed.) Proc. of the International Workshop Methane and Nitrous Oxide: Methods in National Emission
Inventories and Options for Control, Amersfoort, NL February 3-5,1993.
Energy Recovery Council (2018) Energy Recovery Council. 2018 Directory of Waste to Energy Facilities. Ted
Michaels and Karunya Krishnan. October 2018. Available online at: http://energyrecoverycouncil.org/wp-
content/uploads/2019/10/ERC-2018-directory.pdf.
Energy Recovery Council (2009) "2007 Directory of Waste-to-Energy Plants in the United States." Accessed on
September 29, 2009.
EIA (2017) MSW Incineration for Heating or Electrical Generation, December 2017, Energy Information
Administration, U.S. Department of Energy, Washington, DC. DOE/EIA-0035. Available online at:
https://www.eia.gov/opendata/?src=-f3.
EIA (2019) EIA St. Louis Federal Reserve's Economic Data (FRED) Consumer Price Index for All Urban Consumers:
Education and Communication (CPIEDUSL). Available online at: < https://www.eia.gov/opendata/excel/>
EPA (2020a) Advancing Sustainable Materials Management: 2018 Data Tables. Office of Land and Emergency
Management, U.S. Environmental Protection Agency. Washington, D.C. Available online at:
https://www.epa.gov/sites/production/files/2020-ll/documents/2018 ff fact sheet.pdf.
EPA (2020b). Greenhouse Gas Reporting Program (GHGRP). 2020 Envirofacts. Available online at:
https://ghgdata.epa.gov/ghgp/main.do.
EPA (2019) Advancing Sustainable Materials Management: 2016 and 2017 Data Tables. Office of Land and
Emergency Management, U.S. Environmental Protection Agency. Washington, D.C. Available online at:
https://www.epa.gov/sites/production/files/2019-
ll/documents/2016 and 2017 facts and figures data tables Q.pdf.
EPA (2018a) Advancing Sustainable Materials Management: 2015 Data Tables. Office of Land and Emergency
Management, U.S. Environmental Protection Agency. Washington, D.C. Available online at:
https://www.epa.gov/sites/production/files/2018-
07/documents/smm 2015 tables ; ~es 07252018 fnl 508 Q.pdf.
EPA (2018b). Greenhouse Gas Reporting Program Data. Washington, DC: U.S. Environmental Protection Agency.
Available online at: https://www.epa.gov/ghgreporting/ghg-reporting-program-data-sets.
EPA (2016) Advancing Sustainable Materials Management: 2014 Fact Sheet. Office of Land and Emergency
Management, U.S. Environmental Protection Agency. Washington, D.C. Available online at:
https://www.epa.gov/sites/production/files/2016-ll/documents/2014 smmfactsheet 508.pdf.
EPA (2015) Advancing Sustainable Materials Management: Facts and Figures 2013 - Assessing Trends in Material
Generation, Recycling and Disposal in the United States. Office of Solid Waste and Emergency Response, U.S.
Environmental Protection Agency. Washington, D.C. Available online at:
http://www3.epa.gov/epawaste/nonhaz/municipal/pubs/2013 advncng smm rpt.pdf.
EPA (2007, 2008, 2011, 2013, 2014) Municipal Solid Waste in the United States: Facts and Figures. Office of Solid
Waste and Emergency Response, U.S. Environmental Protection Agency. Washington, D.C. Available online at:
http://www.epa.gov/osw/nonhaz/municipal/msw99.htm.
EPA (2006) Solid Waste Management and Greenhouse Gases: A Life-Cycle Assessment of Emissions and Sinks.
Office of Solid Waste and Emergency Response, U.S. Environmental Protection Agency. Washington, D.C.
EPA (2000) Characterization of Municipal Solid Waste in the United States: Source Data on the 1999 Update. Office
of Solid Waste, U.S. Environmental Protection Agency. Washington, D.C. EPA530-F-00-024.
Goldstein, N. and C. Madtes (2001) "13th Annual BioCycle Nationwide Survey: The State of Garbage in America."
BioCycle, JG Press, Emmaus, PA. December 2001.
References and Abbreviations 17
-------
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Kaufman, et al. (2004) "14th Annual BioCycle Nationwide Survey: The State of Garbage in America 2004" Biocycle,
JG Press, Emmaus, PA. January 2004.
RMA (2020) "2019 U.S. Scrap Tire Management Summary." Rubber Manufacturers Association, Washington, DC.
October 2020. Available online at:
https://www.ustires.org/sites/default/files/2019%20USTMA%20Scrap%20Tire%20Management%20Summary%20
Report.pdf.
RMA (2018) "2017 U.S. Scrap Tire Management Summary." Rubber Manufacturers Association, Washington, DC.
July 2018. Available online at: https://www.ustires.org/system/files/USTMA scraptire summ 201? 072018.pdf.
RMA (2016) "2015 U.S. Scrap Tire Management Summary." Rubber Manufacturers Association. August 2016.
Available online at: https://rma.org/sites/default/files/RMA scraptire summ 2015.pdf.
RMA (2014) "2013 U.S. Scrap Tire Management Summary." Rubber Manufacturers Association. November 2014.
Available online at: https://www.ustires.org/sites/default/files/MAR 027 USTMA.pdf.
RMA (2013) "U.S. Scrap Tire Management Summary 2005-2009." Rubber Manufacturers Association. October
2011; Updated September 2013. Available online at:
https://www.ustires.org/sites/default/files/MAR 025 USTMA.pdf.
RMA (2012a) "Rubber FAQs." Rubber Manufacturers Association. Available online at: http://www.rma.org/about-
rma/rubber-faqs/. Accessed on 19 November 2014.
RMA (2012b) "Scrap Tire Markets: Facts and Figures - Scrap Tire Characteristics." Rubber Manufacturers
Association. Available online at: http://www.rma.org/scrap tires/scrap tire markets/scrap tire characteristics/.
Accessed 18 on January 2012.
Schneider, S. (2007) E-mail between Shelly Schneider of Franklin Associates (a division of ERG) and Sarah Shapiro of
ICF International, January 10, 2007.
Shin, D. (2014) Generation and Disposition of Municipal Solid Waste (MSW) in the United States-A National
Survey. Thesis. Columbia University, Department of Earth and Environmental Engineering, January 3, 2014.
Simmons, et al. (2006) "15th Nationwide Survey of Municipal Solid Waste Management in the United States: The
State of Garbage in America." BioCycle, JG Press, Emmaus, PA. April 2006.
van Haaren, Rob, Themelis, N., and Goldstein, N. (2010) "The State of Garbage in America." BioCycle, October
2010. Volume 51, Number 10, pg. 16-23.
Coal Mining
AAPG (1984) Coalbed Methane Resources of the United States. AAPG Studies in Geology Series #17.
Creedy, D.P. (1993) Methane Emissions from Coal Related Sources in Britain: Development of a Methodology.
Chemosphere, 26: 419-439.
DMME (2021) DGO Data Information System. Department of Mines, Minerals and Energy of Virginia. Available
online at https://www.dmme.virginia.gov/dgoinquiry/frmmain.aspx.
EIA (2021) Annual Coal Report 2020. Table 1. Energy Information Administration, U.S. Department of Energy.
Washington, D.C. DOE/EIA-0584.
El Paso (2009) Shoal Creek Mine Plan, El Paso Exploration & Production.
EPA (2021) Greenhouse Gas Reporting Program (GHGRP) 2020 Subpart FF: Underground Coal Mines.
18 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
EPA (2005) Surface Mines Emissions Assessment. Draft. U.S. Environmental Protection Agency.
EPA (1996) Evaluation and Analysis of Gas Content and Coal Properties of Major Coal Bearing Regions of the United
States. EPA/600/R-96-065. U.S. Environmental Protection Agency.
ERG (2021). Correspondence between ERG and Buchanan Mine.
Geological Survey of Alabama State Oil and Gas Board (GSA) (2021) Well Records Database. Available online at
http://www.gsa.state.al.us/ogb/database.aspx.
IEA (2021) Key World Energy Statistics. Coal Production, International Energy Agency.
IPCC (2019) 2019 Refinement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories. Calvo Buendia,
E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M., Ngarize S., Osako, A., Pyrozhenko, Y., Shermanau, P. and
Federici, S. (eds). Published: IPCC, Switzerland.
IPCC (2011) Use of Models and Facility-Level Data in Greenhouse Gas Inventories. Report of IPCC Expert Meeting on
Use of Models and Measurements in Greenhouse Gas Inventories 9-11 August 2010, Sydney, Australia. Eds:
Eggleston H.S., Srivastava N., Tanabe K., Baasansuren J., Fukuda M. IGES.
JWR (2010) No. 4 & 7 Mines General Area Maps. Walter Energy: Jim Walter Resources.
King, Brian (1994) Management of Methane Emissions from Coal Mines: Environmental, Engineering, Economic and
Institutional Implication of Options. Neil and Gunter Ltd.
McElroy OVS (2021) Marshall County VAM Abatement Project Offset Verification Statement submitted to
California Air Resources Board, August 2021.
MSHA (2021) Data Transparency at MSHA. Mine Safety and Health Administration. Available online at
http://www.msha.gov/.
Mutmansky, Jan M. and Yanbei Wang (2000) Analysis of Potential Errors in Determination of Coal Mine Annual
Methane Emissions. Mineral Resources Engineering, 9(4).
Saghafi, Abouna (2013) Estimation of Fugitive Emissions from Open Cut Coal Mining and Measurable Gas Content.
13th Coal Operators' Conference, University of Wollongong, The Australian Institute of Mining and Metallurgy &
Mine Managers Association of Australia. 306-313.
USBM (1986) Results of the Direct Method Determination of the Gas Contents of U.S. Coal Basins. Circular 9067.
U.S. Bureau of Mines.
West Virginia Geological & Economic Survey (WVGES) (2021) Oil & Gas Production Data. Available online at
http://www.wvgs.wvnet.edu/www/datastat/datastat.htm.
Abandoned Underground Coal Mines
CMM (2021) The International Coal Mine Methane Recovery and Utilization Project Database. Available online at:
https://www.globalmethane.org/resources/details.aspx?resourceid=1981
CMOP (2022) EPA's Coalbed Methane Outreach Program, Map of US Coal Mine Methane Current Projects and
Potential Opportunities. Available online at: https://www.epa.gov/cmop/map-us-coal-mine-methane-current-
projects-and-potential-opportunities
EPA (2004) Methane Emissions Estimates & Methodology for Abandoned Coal Mines in the U.S. Draft Final Report.
Washington, D.C. April 2004.
MSHA (2021) U.S. Department of Labor, Mine Health & Safety Administration, Mine Data Retrieval System.
Available online at: https://www.msha.gov/mine-data-retrieval-system.
References and Abbreviations 19
-------
Petroleum Systems
API (1992) Global Emissions of Methane from Petroleum Sources. American Petroleum Institute, Health and
Environmental Affairs Department, Report No. DR140, February 1992.
BOEM (2021a) BOEM Platform Structures Online Query. Available online at:
https://www.data.boem.gov/Platform/PlatformStructures/Default.aspx.
BOEM (2021b) BOEM Oil and Gas Operations Reports - Part A (OGOR-A). Production Data for 1947 to 2020.
Download "Production Data" online at: https://www.data.boem.gov/Main/RawData.aspx.
BOEM (2021c) BOEM Oil and Gas Operations Reports - Part A (OGOR-A). Production Data for 1996 to 2020.
Available online at: https://www.data.boem.gov/Main/OGOR-A.aspx.
BOEM (2021d) BOEM Oil and Gas Operations Reports - Part B (OGOR-B). Flaring volumes for 1996 to 2020.
Available online at: https://www.data.boem.gov/Main/OGOR-B.aspx.
EIA (2021) Crude Oil Production. Energy Information Administration.
Enverus (2021) August 2021 Download. Enverus, Inc.
EPA (1997) Compilation of Air Pollutant Emission Factors, AP-42. Office of Air Quality Planning and Standards, U.S.
Environmental Protection Agency. Research Triangle Park, NC. October 1997.
EPA (1999) Estimates of Methane Emissions from the U.S. Oil Industry (Draft Report). Prepared by ICF International.
Office of Air and Radiation, U.S. Environmental Protection Agency. October 1999.
EPA (2017) 2017 Nonpoint Oil and Gas Emission Estimation Tool, Version 1.2. Prepared for U.S. Environmental
Protection Agency by Eastern Research Group, Inc. (ERG). October 2019.
EPA (2021) Greenhouse Gas Reporting Program. U.S. Environmental Protection Agency. Data reported as of August
7, 2021.
EPA/GRI (1996) Methane Emissions from the Natural Gas Industry. Prepared by Radian. U.S. Environmental
Protection Agency. April 1996.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Natural Gas Systems
AHS (2021) U.S. Census Bureau's American Housing Survey (AHS). https://www.census.gov/programs-
surveys/ahs.html.
CBECS (2021) Energy Information Administration's Commercial Buildings Energy Consumption Survey (CBECS).
https://www.eia.gov/consumption/commercial.
CenSARA (2012) 2011 Oil and Gas Emission Inventory Enhancement Project for CenSARA States. Prepared by
ENVIRON International Corporation and Eastern Research Group, Inc. (ERG). Central States Air Resources Agencies
(CenSARA). December 2012.
Cusworth, D.H., Duren, R.M., Thorpe, A.K., Pandey S., Maasakkers, J.D., Aben, I., et al. (2021). Multisatellite
imaging of a gas well blowout enables quantification of total methane emissions. Geophysical Research Letters, 48,
e2020GL090864. https://doi.org/10.1029/2020GL090864
EIA (2021) Natural Gas Gross Withdrawals and Production. Energy Information Administration.
EIA (2021b) October 2021 Monthly Energy Review. Energy Information Administration.
https://www.eia.gov/totalenergv/data/monthly/archive/00352110.pdf.
20 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Enverus (2021) August 2021 Download. Enverus, Inc.
EPA (1977) Atmospheric Emissions from Offshore Oil and Gas Development and Production. Office of Air Quality
Planning and Standards, Research Triangle Park, NC. PB272268. June 1977.
EPA (2020) MOVES3. https://www.epa.gov/moves/latest-version-motor-vehicle-emission-simulator-moves.
EPA (2021a) Greenhouse Gas Reporting Program- Subpart W-Petroleum and Natural Gas Systems. Environmental
Protection Agency. Data reported as of August 7, 2021.
EPA (2021b) Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2020: Updates Under Consideration for
Anomalous Events Including Well Blowout and Well Release Emissions (Anomalous Events memo). U.S.
Environmental Protection Agency. September 2021. Available at: https://www.epa.gov/ghgemissions/natural-gas-
and-petroleum-systems.
EPA (2021c) Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2020: Updates Under Consideration for
Activity Data (Activity Data memo). U.S. Environmental Protection Agency. September 2021. Available at:
https://www.epa.gov/ghgemissions/natural-gas-and-petroleum-systems.
EPA (2021d) Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2020: Updates Under Consideration for
Gas STAR and Methane Challenge Reductions (Reductions memo). U.S. Environmental Protection Agency.
September 2021. Available at: https://www.epa.gov/ghgemissions/natural-gas-and-petroleum-systems.
EPA (2021e) Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2020: Updates Under Consideration for
Post-Meter Emissions (Post-Meter memo). U.S. Environmental Protection Agency. September 2021. Available at:
https://www.epa.gov/ghgemissions/natural-gas-and-petroleum-systems.
EPA (2021f) Nonpoint Oil & Gas Emission estimation Tool. Data received via email in November 2021.
Fischer et al. (2018) "An Estimate of Natural Gas Methane Emissions from California Homes." Environmental
Science & Technology 2018, 52 (17), 10205-10213. https://pubs.acs.org/doi/10.1021/acs.est.8b03217.
GRI/EPA (1996) Methane Emissions from the Natural Gas Industry. Prepared by Harrison, M., T. Shires, J. Wessels,
and R. Cowgill, eds., Radian International LLC for National Risk Management Research Laboratory, Air Pollution
Prevention and Control Division, Research Triangle Park, NC. EPA-600/R-96-080a.
GTI (2001) Gas Resource Database: Unconventional Natural Gas and Gas Composition Databases. Second Edition.
GRI-01/0136.
GTI (2019) Classification of Methane Emissions from Industrial Meters, Vintage vs Modern Plastic Pipe, and Plastic-
lined Steel and Cast-Iron Pipe. June 2019. Gas Technology Institute and U.S. Department of Energy GTI Project
Number 22070. DOE project Number ED-FE0029061.
Illinois Office of Oil and Gas Resource Management (2020) State-level natural gas production quantities.
Indiana Division of Oil & Gas (2020) State-level natural gas production quantities.
Kansas Department of Health and Environment (2020) County-level produced water quantities.
Lamb, et al. (2015) "Direct Measurements Show Decreasing Methane Emissions from Natural Gas Local
Distribution Systems in the United States." Environmental Science & Technology, Vol. 49 5161-5169.
Lavoie et al. (2017) "Assessing the Methane Emissions from Natural Gas-Fired Power Plants and Oil Refineries."
Environmental Science & Technology. 2017 Mar 21;51(6):3373-3381. doi: 10.1021/acs.est.6b05531.
Maasakkers, Joannes D., Mark Omara, Ritesh Gautam, Alba Lorente, Sudhanshu Pandey, Paul Tol, Tobias Borsdorff,
Sander Houweling, Use Aben (2022). Reconstructing and quantifying methane emissions from the full duration of a
38-day natural gas well blowout using space-based observations. Remote Sensing of Environment.
https://doi.Org/10.1016/j.rse.2021.112755.Ohio Environmental Protection Agency (2020) Well-level produced
water quantities.
Oklahoma Department of Environmental Quality (2020) Well-level produced water quantities.
References and Abbreviations 21
-------
Pandey, S., Gautam, R., Houweling, S., van der Gon, H. D., Sadavarte, P., Borsdorff, T., et al. (2019). Satellite
observations reveal extreme methane leakage from a natural gas well blowout. Proceedings of the National
Academy of Sciences, 116, 26376-26381. https://doi.org/10.1073/pnas.1908712116
PHMSA (2021a) Gas Distribution Annual Data. Pipeline and Hazardous Materials Safety Administration, U.S.
Department of Transportation, Washington, DC. Available online at: https://www.phmsa.dot.gov/data-and-
statistics/pipeline/annual-report-mileage-gas-distribution-systems.
PHMSA (2021b) Underground Natural Gas STAR, Part C. Pipeline and Hazardous Materials Safety Administration,
U.S. Department of Transportation, Washington, DC. https://www.phmsa.dot.gov/data-and-statistics/pipeline/gas-
distribution-gas-gathering-gas-transmission-hazardous-liquids.
West Virginia Department of Environmental Protection (2020) State-level natural gas production quantities.
Zimmerle et al. (2019) "Characterization of Methane Emissions from Gathering Compressor Stations." October
2019. Available at https://mountainscholar.org/handle/10217/195489.
Zimmerle, et al. (2015) "Methane Emissions from the Natural Gas Transmission and Storage System in the United
States." Environmental Science and Technology, Vol. 49 9374-9383.
Abandoned Oil and Gas Wells
Alaska Oil and Gas Conservation Commission, Available online at:
https://www.commerce.alaska.gov/web/aogcc/Data.aspx.
Arkansas Geological & Conservation Commission, "List of Oil & Gas Wells - Data From November 1,1936 to January
1,1955."
The Derrick's Handbook of Petroleum: A Complete Chronological and Statistical Review of Petroleum
Developments From 1859 to 1898 (V.l), (1898-1899) (V.2).
Enverus (2021) August 2021 Download. Enverus, Inc.
GRI/EPA (1996) Methane Emissions from the Natural Gas Industry. Prepared by Harrison, M., T. Shires, J. Wessels,
and R. Cowgill, eds., Radian International LLC for National Risk Management Research Laboratory, Air Pollution
Prevention and Control Division, Research Triangle Park, NC. EPA-600/R-96-080a.
Florida Department of Environmental Protection - Oil and Gas Program, Available online at:
https://floridadep.gov/water/oil-gas.
Geological Survey of Alabama, Oil & Gas Board, Available online at: https://www.gsa.state.al.us/ogb/.
GTI (2001) Gas Resource Database: Unconventional Natural Gas and Gas Composition Databases. Second Edition.
GRI-01/0136.
Interstate Oil and Gas Compact Commission (2021). IDLE AND ORPHAN OIL AND GAS WELLS: STATE AND
PROVINCIAL REGULATORY STRATEGIES 2021. Available online at:
https://iogcc.ok.gOv/sites/g/files/gmc836/f/iogcc idle and orphan wells 2021 final web.pdf.
Kang, et al. (2016) "Identification and characterization of high methane-emitting abandoned oil and gas wells."
PNAS, vol. 113 no. 48, 13636-13641, doi: 10.1073/pnas.l605913113.
Oklahoma Geological Survey. "Oklahoma Oil: Past, Present, and Future." Oklahoma Geology Notes, v. 62 no. 3,
2002 pp. 97-106.
Pennsylvania Department of Environmental Protection, Oil and Gas Reports - Oil and Gas Operator Well Inventory.
Available online at:
http://www.depreportingservices.state.pa.us/ReportServer/Pages/ReportViewer.aspx7/Oil Gas/OG Well Invento
ry.
22 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Texas Railroad Commission, Oil and Gas Division, "History of Texas Initial Crude Oil, Annual Production and
Producing Wells, Crude Oil Production and Well Counts (since 1935)."
Townsend-Small, et al. (2016) "Emissions of coalbed and natural gas methane from abandoned oil and gas wells in
the United States." Geophysical Research Letters, Vol. 43,1789-1792.
United States Department of Interior (2022). Overwhelming Interest in Orphan Well Infrastructure Investments.
Available online at: https://content.govdelivery.com/accounts/USDOI/bulletins/30416b5.
United States Geological Survey's (USGS) Mineral Resources of the United States Annual Yearbooks, available
online at: https://minerals.usgs.gov/minerals/pubs/usbmmyb.html.
Virginia Department of Mines Minerals and Energy, "Wells Drilled for Oil and Gas in Virginia prior to 1962.",
Virginia Division of Mineral Resources. Available online at:
https://www.dmme.virginia.gov/commercedocs/MRR 4.pdf.
International Bunker Fuels
Anderson, B.E., et al. (2011) Alternative Aviation Fuel Experiment (AAFEX), NASA Technical Memorandum, in press.
ASTM (1989) Military Specification for Turbine Fuels, Aviation, Kerosene Types, NATO F-34 (JP-8) and NATO F-35.
February 10,1989.
DHS (2008) Personal Communication with Elissa Kay, Residual and Distillate Fuel Oil Consumption (International
Bunker Fuels). Department of Homeland Security, Bunker Report. January 11, 2008.
DLA Energy (2021) Unpublished data from the Defense Fuels Automated Management System (DFAMS). Defense
Energy Support Center, Defense Logistics Agency, U.S. Department of Defense. Washington, D.C.
DOC (1991 through 2020) Unpublished Report of Bunker Fuel Oil Laden on Vessels Cleared for Foreign Countries.
Form-563. Foreign Trade Division, Bureau of the Census, U.S. Department of Commerce. Washington, D.C.
DOT (1991 through 2013) Fuel Cost and Consumption. Federal Aviation Administration, Bureau of Transportation
Statistics, U.S. Department of Transportation. Washington, D.C. DAI-10.
EIA (2022) Monthly Energy Review, February 2022, Energy Information Administration, U.S. Department of Energy,
Washington, D.C. DOE/EIA-0035(2022/02).
EPA (2020) EPA Inventory of U.S. Greenhouse Gas Emissions and Sinks 1990-2019: Updated Gasoline and Diesel
Fuel CO2 Emission Factors - Memo.
FAA (2022) Personal Communication between FAA and John Steller, Mausami Desai, and Vincent Camobreco for
aviation emissions estimates from the Aviation Environmental Design Tool (AEDT). March 2022.
IPCC/UNEP/OECD/IEA (1997) Revised 1996 IPCC Guidelines for National Greenhouse Gas Inventories. 31
Intergovernmental Panel on Climate Change, United Nations Environment Programme, Organization for Economic
32 Co-Operation and Development, International Energy Agency, Paris, France.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
USAF (1998) Fuel Logistics Planning. U.S. Air Force pamphlet AFPAM23-221, May 1,1998.
Wood Biomass and Biofuel Consumption
EIA (2022) Monthly Energy Review, February 2022. Energy Information Administration, U.S. Department of Energy.
Washington, D.C. DOE/EIA-0035(2022/02).
References and Abbreviations 23
-------
EIA (2020) Biofuels explained: Use of biomass-based diesel fuel. Energy Information Administration, U.S.
Department of Energy. Washington, D.C. Available online at: https://www.eia.gov/energyexplained/biofuels/use-
of-biodiesel.php.
EPA (2022) Acid Rain Program Dataset 1996-2020. Office of Air and Radiation, Office of Atmospheric Programs,
U.S. Environmental Protection Agency, Washington, D.C.
EPA (2010) Carbon Content Coefficients Developed for EPA's Mandatory Reporting Rule. Office of Air and
Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
Lindstrom, P. (2006) Personal Communication. Perry Lindstrom, Energy Information Administration and Jean Kim,
ICF International.
Energy Sources of Precursor Greenhouse Gases
EPA (2022) "Crosswalk of Precursor Gas Categories." U.S. Environmental Protection Agency. April 6, 2022.
EPA (2021a) "Criteria pollutants National Tier 1 for 1970 - 2020." National Emissions Inventory (NEI) Air Pollutant
Emissions Trends Data. Office of Air Quality Planning and Standards, March 2021. Available online at:
https://www.epa.gov/air-emissions-inventories/air-pollutant-emissions-trends-data.
EPA (2021b) "2017 National Emissions Inventory (NEI) Technical Support Document (TSD)." Office of Air Quality
Planning and Standards, April 2021. Available online at: https://www.epa.gov/air-emissions-inventories/2017-
national-emissions-inventory-nei-technical-support-document-tsd.
EPA (1997) Compilation of Air Pollutant Emission Factors, AP-42. Office of Air Quality Planning and Standards, U.S.
Environmental Protection Agency. Research Triangle Park, NC. October 1997.
Industrial Processes and Product Use
EPA (2014) 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.
EPA (2002) Quality Assurance/Quality Control and Uncertainty Management Plan for the U.S. Greenhouse Gas
Inventory: Procedures Manual for Quality Assurance/Quality Control and Uncertainty Analysis, U.S. Greenhouse
Gas Inventory Program, U.S. Environmental Protection Agency, Office of Atmospheric Programs, EPA430-R-02-
007B, June 2002.
IPCC (2011) Use of Models and Facility-Level Data in Greenhouse Gas Inventories (Report of IPCC Expert Meeting
on Use of Models and Measurements in Greenhouse Gas Inventories 9-11 August 2010, Sydney, Australia) eds.:
Eggleston H.S., Srivastava N., Tanabe K., Baasansuren J., Fukuda M., Pub. IGES, Japan 2011.
Cement Production
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
U.S. Bureau of Mines (1990 through 1993) Minerals Yearbook: Cement Annual Report. U.S. Department of the
Interior, Washington, D.C.
U.S. Environmental Protection Agency (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.
24 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
U.S. EPA (2021) Greenhouse Gas Reporting Program (GHGRP). Aggregation of Reported Facility Level Data under
Subpart H -National Level Clinker Production from Cement Production for Calendar Years 2014 through 2020.
Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington,
D.C.
United States Geological Survey (USGS) (2021) Mineral Commodity Summaries: Cement. U.S. Geological Survey,
Reston, VA. January 2021. Available at: https://pubs.usgs.gov/periodicals/mcs2021/mcs2021-cement.pdf.
USGS (1995 through 2014) Minerals Yearbook - Cement. U.S. Geological Survey, Reston, VA.
USGS (2020) 2017 Minerals Yearbook - Cement. U.S. Geological Survey, Reston, VA. August 2020.
Van Oss (2013a) 1990 through 2012 Clinker Production Data Provided by Hendrik van Oss (USGS) via email on
November 8, 2013.
Van Oss (2013b) Personal communication. Hendrik van Oss, Commodity Specialist of the U.S. Geological Survey
and Gopi Manne, Eastern Research Group, Inc. October 28, 2013.
Lime Production
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Males, E. (2003) Memorandum from Eric Males, National Lime Association to William N. Irving & Leif Hockstad,
Environmental Protection Agency. March 6, 2003.
Miner, R. and B. Upton (2002) Methods for estimating greenhouse gas emissions from lime kilns at kraft pulp mills.
Energy. Vol. 27 (2002), p. 729-738.
Seeger (2013) Memorandum from Arline M. Seeger, National Lime Association to Leif Hockstad, Environmental
Protection Agency. March 15, 2013.
U.S. Environmental Protection Agency (EPA (2021) Greenhouse Gas Reporting Program (GHGRP). Aggregation of
Reported Facility Level Data under Subpart S-National Lime Production for Calendar Years 2010 through
2020. Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency,
Washington, D.C.
United States Geological Survey (USGS) (2021a) 2021 Mineral Commodities Summary: Lime. U.S. Geological Survey,
Reston, VA (January 2021).
USGS (2021b) 2020 Minerals Yearbook Annual Tables: Lime. U.S. Geological Survey, Reston, VA (August 2021).
USGS (2020a) 2020 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA (January 2020).
USGS (2020b) (1992 through 2017) Minerals Yearbook: Lime. U.S. Geological Survey, Reston, VA (June 2020).
USGS (2020c) 2018 Minerals Yearbook Annual Tables: Lime. U.S. Geological Survey, Reston, VA (November 2020).
USGS (2020d) Personal communication. Lori E. Apodaca, U.S. Geological Survey and Amanda Chiu, U.S.
Environmental Protection Agency. December 17, 2020.
USGS (2019) 2016 Minerals Yearbook: Lime. U.S. Geological Survey, Reston, VA (August 2019).
USGS (2018a) 2018 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA (January 2018).
USGS (2018b) 2015 Minerals Yearbook: Lime. U.S. Geological Survey, Reston, VA (March 2018).
USGS (2012) 2012 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA (January 2012).
USGS (2011) 2011 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA (January 2011).
USGS (2010) 2010 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA (January 2010).
References and Abbreviations 25
-------
USGS (2008) 2008 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA (January 2008).
USGS (2007) 2007 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA (January 2007).
USGS (2002) 2002 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA (January 2002).
USGS (1996) 1996 Mineral Commodities Summary: Lime. U.S. Geological Survey, Reston, VA (January 1996).
USGS (1991) 1991 Minerals Yearbook: Lime. U.S. Geological Survey, Reston, VA (1991).
Glass Production
Federal Reserve (2021) Board of Governors of the Federal Reserve System (US), Industrial Production:
Manufacturing: Durable Goods: Glass and Glass Product (NAICS = 3272) [IPG3272N], retrieved from FRED, Federal
Reserve Bank of St. Louis. Available at: https://fred.stlouisfed.org/series/lPG3272N. Accessed on October 1, 2021.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
U.S. Bureau of Mines (1991 and 1993a) Minerals Yearbook: Crushed Stone Annual Report. U.S. Department of the
Interior. Washington, D.C.
U.S. Department of Energy (DOE) (2002) Glass Industry of the Future: Energy and Environmental Profile of the U.S.
Glass Industry. Office of Industrial Technologies, U.S. Department of Energy. Washington, D.C.
U.S. Environmental Protection Agency (EPA) (2009) Technical Support Document for the Glass Manufacturing
Sector: Proposed Rule for Mandatory Reporting of Greenhouse Gases. U.S. Environmental Protection Agency,
Washington, D.C.
U.S. EPA (2021) Greenhouse Gas Reporting Program (GHGRP). Aggregation of Reported Facility Level Data under
Subpart N -National Glass Production for Calendar Years 2010 through 2020. Office of Air and Radiation, Office of
Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
United States Geological Survey (USGS) (1995 through 2015b) Minerals Yearbook: Soda Ash Annual Report. U.S.
Geological Survey, Reston, VA.
USGS (2017) Minerals Industry Surveys: Soda Ash in January 2017. U.S. Geological Survey, Reston, VA. March 2017.
USGS (2018) Mineral Industry Surveys: Soda Ash in February 2018. U.S. Geological Survey, Reston, VA. 2018.
USGS (2019) Mineral Industry Surveys: Soda Ash in December 2018. U.S. Geological Survey, Reston, VA. March
2019.
USGS (2020) Mineral Industry Surveys: Soda Ash in April 2020. U.S. Geological Survey, Reston, VA. July 2020.
USGS (2021) Mineral Industry Surveys: Soda Ash in April 2021. U.S. Geological Survey, Reston, VA. July 2021.
Other Process Uses of Carbonates
AISI (2018 through 2020) Annual Statistical Report. American Iron and Steel Institute.
Kostick, D. S. (2012) Personal communication. Dennis S. Kostick, U.S. Geological Survey, Soda Ash Commodity
Specialist and Gopi Manne and Bryan Lange of Eastern Research Group, Inc. October 2012.
U.S. Bureau of Mines (1991 and 1993a) Minerals Yearbook: Crushed Stone Annual Report. U.S. Department of the
Interior. Washington, D.C.
U.S. Environmental Protection Agency (EPA) (2021). Greenhouse Gas Reporting Program (GHGRP). Dataset as of
August 7, 2021. Available online at: https://ghgdata.epa.gov/ghgp/
26 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
United States Geological Survey (USGS) (2017a) Mineral Industry Surveys: Soda Ash in January 2017. U.S.
Geological Survey, Reston, VA. March 2017.
USGS (2018) Mineral Industry Surveys: Soda Ash in February 2018. U.S. Geological Survey, Reston, VA. 2018.
USGS (2019) Mineral Industry Surveys: Soda Ash in April 2019. U.S. Geological Survey, Reston, VA. July 2019.
USGS (2020a) 2016 Minerals Yearbook: Stone, Crushed [Advanced Release]. U.S. Geological Survey, Reston, VA.
January 2020.
USGS (2020b) Mineral Industry Surveys: Soda Ash in April 2020. U.S. Geological Survey, Reston, VA. July 2020.
USGS (2020c) Minerals Yearbook 2017: Stone, Crushed [Advanced Data Release of the 2017 Annual Tables]. U.S.
Geological Survey, Reston, VA. August 2020.
USGS (2021a) 2017 Minerals Yearbook: Stone, Crushed [Advanced Release]. U.S. Geological Survey, Reston, VA.
June 2021.
USGS (2021b) 2020 Mineral Commodity Summaries: Stone (Crushed). U.S. Geological Survey, Reston, VA. January
2021.
USGS (2021c) Minerals Yearbook 2019: Soda Ash [Advanced Data Release of the 2019 Annual Tables]. U.S.
Geological Survey, Reston, VA. August 2021.
USGS (2021d) Mineral Industry Surveys: Soda Ash in April 2021. U.S. Geological Survey, Reston, VA. July 2021.
USGS (1995a through 2017) Minerals Yearbook: Crushed Stone Annual Report. U.S. Geological Survey, Reston, VA.
USGS (1994 through 2015b) Minerals Yearbook: Soda Ash Annual Report. U.S. Geological Survey, Reston, VA.
Willett (2017) Personal communication, Jason Christopher Willett, U.S. Geological Survey and Mausami Desai and
John Steller, U.S. Environmental Protection Agency. March 9, 2017.
Willett (2021) Personal communication, Jason Christopher Willett, U.S. Geological Survey and Amanda Chiu, U.S.
Environmental Protection Agency. August 31, 2021.
Ammonia Production
ACC (2020) Business of Chemistry (Annual Data). American Chemistry Council, Arlington, VA.
Bark (2004) Coffeyville Nitrogen Plant. December 15, 2004. Available online at:
http://www.gasification.org/uploads/downloads/Conferences/2003/07BARK.pdf.
Coffeyville Resources Nitrogen Fertilizers (2012) Nitrogen Fertilizer Operations. Available online at:
http://coffevvillegroup.com/NitrogenFertilizerOperations/index.html.
Coffeyville Resources Nitrogen Fertilizers (2011) Nitrogen Fertilizer Operations. Available online at:
http://coffevvillegroup.com/NitrogenFertilizerOperations/index.html.
Coffeyville Resources Nitrogen Fertilizers (2010) Nitrogen Fertilizer Operations. Available online at:
http://coffevvillegroup.com/NitrogenFertilizerOperations/index.html.
Coffeyville Resources Nitrogen Fertilizers (2009) Nitrogen Fertilizer Operations. Available online at:
http://coffevvillegroup.com/NitrogenFertilizerOperations/index.html.
Coffeyville Resources Nitrogen Fertilizers, LLC (2005 through 2007a) Business Data. Available online at:
http://www.coffeyvillegroup.com/businessSnapshot.asp.
Coffeyville Resources Nitrogen Fertilizers (2007b) Nitrogen Fertilizer Operations. Available online at:
http://coffeyvillegroup.com/nitrogenMain.aspx.
Coffeyville Resources Energy, Inc. (CVR) (2012) CVR Energy, Inc. 2012 Annual Report. Available online at:
http://cvrenergy.com.
References and Abbreviations 27
-------
CVR (2013) CVR Energy, Inc. 2013 Annual Report. Available online at: http://cvrenergy.com.
CVR (2014) CVR Energy, Inc. 2014 Annual Report. Available online at: http://cvrenergy.com.
CVR (2015) CVR Energy, Inc. 2015 Annual Report. Available online at: http://cvrenergy.com.
EFMA (2000a) Best Available Techniques for Pollution Prevention and Control in the European Fertilizer Industry.
Booklet No. 1 of 8: Production of Ammonium. Available online at:
http://fertilizerseurope.com/site/index.php7icb390.
EFMA (2000b) Best Available Techniques for Pollution Prevention and Control in the European Fertilizer Industry.
Booklet No. 5 of 8: Production of Urea and Urea Ammonium Nitrate. Available online at:
http://fertilizerseurope.com/site/index.php?id=390.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
United States Census Bureau (2011) Current Industrial Reports Fertilizer Materials and Related Products: 2010
Summary. Available online at: http://www.census.gov/manufacturing/cir/historical data/mq325b/index.html.
U.S. Census Bureau (2010) Current Industrial Reports Fertilizer Materials and Related Products: 2009 Summary.
Available online at: http://www.census.gov/manufacturing/cir/historical data/mq325b/index.html.
U.S. Census Bureau (2009) Current Industrial Reports Fertilizer Materials and Related Products: 2008 Summary.
Available online at: http://www.census.gov/manufacturing/cir/historical data/mq325b/index.html.
U.S. Census Bureau (2008) Current Industrial Reports Fertilizer Materials and Related Products: 2007Summary.
Available online at: http://www.census.gov/cir/www/325/mq325b/mq325b075.xls.
U.S. Census Bureau (2007) Current Industrial Reports Fertilizer Materials and Related Products: 2006 Summary.
Available online at: http://www.census.gOv/industry/l/mq325b065.pdf.
U.S. Census Bureau (2006) Current Industrial Reports Fertilizer Materials and Related Products: 2005 Summary.
Available online at: http://www.census.gov/cir/www/325/mq325b.html.
U.S. Census Bureau (2004, 2005) Current Industrial Reports Fertilizer Materials and Related Products: Fourth
Quarter Report Summary. Available online at: http://www.census.gov/cir/www/325/mq325b.html.
U.S. Census Bureau (1998 through 2003) Current Industrial Reports Fertilizer Materials and Related Products:
Annual Reports Summary. Available online at: http://www.census.gov/cir/www/325/mq325b.html.
U.S. Census Bureau (1991 through 1994) Current Industrial Reports Fertilizer Materials Annual Report. Report No.
MQ28B. U.S. Census Bureau, Washington, D.C.
United States Environmental Protection Agency (EPA) (2018) Greenhouse Gas Reporting Program. Aggregation of
Reported Facility Level Data under Subpart G -Annual Urea Production from Ammonia Manufacturing for Calendar
Years 2011-2016. Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection
Agency, Washington, D.C.
U.S. EPA (2021a) Greenhouse Gas Reporting Program. Aggregation of Reported Facility Level Data under Subpart G
-Annual Urea Production from Ammonia Manufacturing for Calendar Years 2017-2020. Office of Air and Radiation,
Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
U.S. EPA (2021b). Greenhouse Gas Reporting Program. Dataset as of August 7, 2021. Available online at:
https://ghgdata.epa.gov/ghgp/.
United States Geological Survey (USGS) (2021) 2021 Mineral Commodity Summaries: Nitrogen (Fixed) - Ammonia.
January 2021. Available online at: https://pubs.usgs.gov/periodicals/mcs2021/mcs2021-nitrogen.pdf.
USGS (1994 through 2009) Minerals Yearbook: Nitrogen. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/nitrogen/.
28 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Urea Consumption for Non-Agricultural Purposes
EFMA (2000) Best Available Techniques for Pollution Prevention and Control in the European Fertilizer Industry.
Booklet No. 5 of 8: Production of Urea and Urea Ammonium Nitrate.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
TFI (2002) U.S. Nitrogen Imports/Exports Table. The Fertilizer Institute. Available online at:
http://www.tfi.org/statistics/usnexim.asp. August 2002.
United States Census Bureau (2001 through 2011) Current Industrial Reports Fertilizer Materials and Related
Products: Annual Summary. Available online at:
http://www.census.gov/manufacturing/cir/historical data/index.html.
United States Department of Agriculture (2012) Economic Research Service Data Sets, Data Sets, U.S. Fertilizer
Imports/Exports: Standard Tables. Available online at: http://www.ers.usda.gov/data-products/fertilizer-
importsexports/standard-tables.aspx.
United States Environmental Protection Agency (EPA) (2018) Greenhouse Gas Reporting Program. Aggregation of
Reported Facility Level Data under Subpart G -Annual Urea Production from Ammonia Manufacturing for Calendar
Years 2011-2016. Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection
Agency, Washington, D.C.
U.S. EPA (2021a) Greenhouse Gas Reporting Program. Aggregation of Reported Facility Level Data under Subpart G
-Annual Urea Production from Ammonia Manufacturing for Calendar Years 2017-2020. Office of Air and Radiation,
Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
U.S. EPA (2021b). Greenhouse Gas Reporting Program. Dataset as of August 7, 2021. Available online at:
https://ghgdata.epa.gov/ghgp/.
United States International Trade Commission (ITC) (2002) United States International Trade Commission
Interactive Tariff and Trade DataWeb, Version 2.5.0. Available online at: http://dataweb.usitc.gov/. August 2002.
United States Geological Survey (USGS) (1994 through 2021a) Minerals Yearbook: Nitrogen. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/nitrogen/.
USGS (2021b) Minerals Commodity Summaries: Nitrogen (Fixed)-Ammonia. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/nitrogen/.
Nitric Acid Production
Climate Action Reserve (CAR) (2013) Project Report. Available online at:
https://thereserve2.apx.com/myModule/rpt/myrpt.asp?r=lll. Accessed on 18 January 2013.
Desai (2012) Personal communication. Mausami Desai, U.S. Environmental Protection Agency, January 25, 2012.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
United States Census Bureau (2010a) Current Industrial Reports. Fertilizers and Related Chemicals: 2009. 'Table 1:
Summary of Production of Principle Fertilizers and Related Chemicals: 2009 and 2008." June, 2010. MQ325B(08)-5.
Available online at: http://www.census.gov/manufacturing/cir/historical data/mq325b/index.html.
U.S. Census Bureau (2010b) Personal communication between Hilda Ward (of U.S. Census Bureau) and Caroline
Cochran (of ICF International). October 26, 2010 and November 5, 2010.
References and Abbreviations 29
-------
U.S. Census Bureau (2009) Current Industrial Reports. Fertilizers and Related Chemicals: 2008. 'Table 1: Shipments
and Production of Principal Fertilizers and Related Chemicals: 2004 to 2008." June, 2009. MQ325B(08)-5. Available
online at: http://www.census.gov/manufacturing/cir/historical data/mq325b/index.html.
U.S. Census Bureau (2008) Current Industrial Reports. Fertilizers and Related Chemicals: 2007. 'Table 1: Shipments
and Production of Principal Fertilizers and Related Chemicals: 2003 to 2007." June, 2008. MQ325B(07)-5. Available
online at: http://www.census.gov/manufacturing/cir/historical data/mq325b/index.html.
United States Environmental Protection Agency (EPA) (2021) Greenhouse Gas Reporting Program. Aggregation of
Reported Facility Level Data under Subpart V -National Nitric Acid Production for Calendar Years 2010 through
2020. Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency,
Washington, D.C.
U.S. 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.
U.S. EPA (2013) Draft Nitric Acid Database. U.S. Environmental Protection Agency, Office of Air and Radiation.
September 2010.
U.S. EPA (2012) Memorandum from Mausami Desai, U.S. EPA to Mr. Bill Herz, The Fertilizer Institute. November
26, 2012.
U.S. EPA (2010) Available and Emerging Technologies for Reducing Greenhouse Gas Emissions from the Nitric Acid
Production Industry. Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency. Research
Triangle Park, NC. December 2010. Available online at: http://www.epa.gov/nsr/ghgdocs/nitricacid.pdf.
U.S. EPA (1998) Compilation of Air Pollutant Emission Factors, AP-42. Office of Air Quality Planning and Standards,
U.S. Environmental Protection Agency. Research Triangle Park, NC. February 1998.
United States Geological Survey (USGS) (2021) 2021 Mineral Commodity Summaries: Nitrogen (Fixed) - Ammonia.
January 2021. Available online at: https://pubs.usgs.gov/periodicals/mcs2021/mcs2021-nitrogen.pdf.
Adipic Acid Production
ACC (2021) Business of Chemistry (Annual Data). American Chemistry Council, Arlington, VA.
C&EN (1995) "Production of Top 50 Chemicals Increased Substantially in 1994." Chemical & Engineering News,
73(15):17. April 10,1995.
C&EN (1994) 'Top 50 Chemicals Production Rose Modestly Last Year." Chemical & Engineering News, 72(15):13.
April 11,1994.
C&EN (1993) 'Top 50 Chemicals Production Recovered Last Year." Chemical & Engineering News, 71(15):11. April
12,1993.
C&EN (1992) "Production of Top 50 Chemicals Stagnates in 1991." Chemical & Engineering News, 70(15): 17. April
13,1992.
CMR (2001) "Chemical Profile: Adipic Acid." Chemical Market Reporter. July 16, 2001.
CMR (1998) "Chemical Profile: Adipic Acid." Chemical Market Reporter. June 15,1998.
CW (2005) "Product Focus: Adipic Acid." Chemical Week. May 4, 2005.
CW (1999) "Product Focus: Adipic Acid/Adiponitrile." Chemical Week, p. 31. March 10,1999.
Desai (2010, 2011) Personal communication. Mausami Desai, U.S. Environmental Protection Agency and Adipic
Acid Plant Engineers. 2010 and 2011.
ICIS (2007) "Adipic Acid." ICIS Chemical Business Americas. July 9, 2007.
30 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Reimer, R.A., Slaten, C.S., Seapan, M., Koch, T.A. and Triner, V.G. (1999) "Implementation of Technologies for
Abatement of N2O Emissions Associated with Adipic Acid Manufacture." Proceedings of the 2nd Symposium on
Non-CC>2 Greenhouse Gases (NCGG-2), Noordwijkerhout, The Netherlands, 8-10 Sept. 1999, Ed. J. van Ham et al.,
Kluwer Academic Publishers, Dordrecht, pp. 347-358.
Thiemens, M.H., and W.C. Trogler (1991) "Nylon production; an unknown source of atmospheric nitrous oxide."
Science 251:932-934.
United States Environmental Protection Agency (EPA) (2021) Greenhouse Gas Reporting Program. Subpart E Data.
Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington,
D.C. Available online at: https://www.epa.gov/ghgreporting/ghg-reporting-program-data-sets.
U.S. EPA (2019, 2020) Greenhouse Gas Reporting Program. Subpart E, S-CEMS, BB, CC, LL Data Set (XLSX) (Adipic
Acid Tab). Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency,
Washington, D.C. Available online at: https://www.epa.gov/ghgreporting/ghg-reporting-program-data-sets.
U.S. 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.
U.S. EPA (2014 through 2018) Greenhouse Gas Reporting Program. Subpart E, S-CEMS, BB, CC, LL Data Set (XLSX)
(Adipic Acid Tab). Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection
Agency, Washington, D.C. Available online at: http://www2.epa.gov/ghgreporting/ghg-reporting-program-data-
sets.
U.S. EPA (2010 through 2013) Analysis of Greenhouse Gas Reporting Program data - Subpart E (Adipic Acid), Office
of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
Caprolactam, Glyoxal and Glyoxylic Acid Production
American Chemistry Council (ACC) (2021) Business of Chemistry (Annual Data). American Chemistry Council,
Arlington, VA.
AdvanSix (2021) AdvanSix's Hopewell Facility Fact Sheet. Retrieved from:
https://www.advansix.com/hopewell/about-us/ on September 16, 2021.
BASF (2021) Welcome to BASF in Freeport Texas. Retrieved from https://www.basf.com/us/en/who-we-
are/organization/locations/featured-sites/Freeport.html on September 16, 2021.
Cline, D. (2019) Firm to Clean Up and Market Former Fibrant Site. The Augusta Chronicle. September 9, 2019.
Retrieved from https://www.augustachronicle.com.
ICIS (2004) Chemical Profile - Caprolactam. January 5, 2004. Available online at:
https://www.icis.com/explore/resources/news/2005/12/02/547244/chemical-profile-caprolactam/.
ICIS (2006) Chemical Profile - Caprolactam. October 15, 2006. Available online at:
https://www.icis.com/explore/resources/news/2006/10/18/2016832/chemical-profile-caprolactam/.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Shaw Industries Group, Inc. (Shaw) (2015) "Shaw Carpet Recycling Facility Successfully Processes Nylon and
Polyester". July 13, 2015. Available online at: https://shawinc.com/Newsroom/Press-Releases/Shaw-Carpet-
Recycling-Facility-Successfully-Proces/.
References and Abbreviations 31
-------
Textile World (2000) "Evergreen Makes Nylon Live Forever". Textile World. October 1, 2000. Available online at:
https://www.textileworld.com/textile-world/textile-news/2000/10/evergreen-makes-nylon-live-forever/.
Carbide Production and Consumption
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
UNFCCC GHG Data Interface (2021) Data accessed on 13 September 2021. Available online at:
https://di.unfccc.int/flex annexl.
United States Census Bureau (2005 through 2020) USITC Trade DataWeb. Available online at:
http://dataweb.usitc.gov/.
United States Geological Survey (USGS) (2020) 2017 Minerals Yearbook: Abrasives, Manufactured [Advance
Release], August 2020. U.S. Geological Survey, Reston, VA. Available online at:
https://www.usgs.gov/centers/nmic/manufactured-abrasives-statistics-and-information.
USGS (1991a through 2020) Minerals Yearbook: Manufactured Abrasives Annual Report. U.S. Geological Survey,
Reston, VA. Available online at: https://prd-wret.s3.us-west-
2. amazonaws.com/assets/palladium/production/atoms/files/mvbl-2017-abras. pdf.
USGS (1991b through 2020) Minerals Yearbook: Silicon Annual Report. U.S. Geological Survey, Reston, VA.
Available online at: http://minerals.usgs.gov/minerals/pubs/commodity/silicon/.
USGS (2021) Mineral Commodity Summaries: Abrasives (Manufactured), January 2021.Available online at:
https://pubs.usgs.gov/periodicals/mcs2021/mcs2021-abrasives.pdf.
Washington Mills (2021), North Grafton, MA. Available online at: https://www.washingtonmills.com/silicon-
carbide/sic-industries.
Titanium Dioxide Production
Gambogi, J. (2002) Telephone communication. Joseph Gambogi, Commodity Specialist, U.S. Geological Survey and
Philip Groth, ICF International. November 2002.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
United States Geological Survey (USGS) (2021a) Mineral Commodity Summaries: Titanium and Titanium Dioxide.
U.S. Geological Survey, Reston, Va. January 2021. Available online at:
https://pubs.usgs.gov/periodicals/mcs2021/mcs2021-titanium.pdf
USGS (2021b) Minerals Yearbook: Titanium, tables-only release of the 2019 tables. U.S. Geological Survey, Reston,
VA. September 2021
USGS (1991 through 2020) Minerals Yearbook: Titanium. U.S. Geological Survey, Reston, VA.
Soda Ash Production
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
32 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
United States Geological Survey (USGS) (2021a) Mineral Commodity Summary: Soda Ash. U.S. Geological Survey,
Reston, VA. Accessed September 2021.
USGS (2021b) Mineral Industry Surveys: Soda Ash in April 2021. U.S. Geological Survey, Reston, VA. Accessed
September 2021.
USGS (2020) Mineral Industry Surveys: Soda Ash in April 2020. U.S. Geological Survey, Reston, VA. Accessed
September 2020.
USGS (2019) Mineral Industry Surveys: Soda Ash in April 2019. U.S. Geological Survey, Reston, VA. Accessed August
2019.
USGS (2018a) Mineral Industry Surveys: Soda Ash in February 2018. U.S. Geological Survey, Reston, VA. Accessed
September 2018.
USGS (2017) Mineral Industry Surveys: Soda Ash in January 2017. U.S. Geological Survey, Reston, VA. March 2017.
USGS (2016) Mineral Industry Surveys: Soda Ash in November 2016. U.S. Geological Survey, Reston, VA. January
2017.
USGS (2015a) Mineral Industry Surveys: Soda Ash in July 2015. U.S. Geological Survey, Reston, VA. September
2015.
USGS (1994 through 2015b, 2018b) Minerals Yearbook: Soda Ash Annual Report. U.S. Geological Survey, Reston,
VA.
USGS (1995c) Trona Resources in the Green River Basin, Southwest Wyoming. U.S. Department of the Interior, U.S.
Geological Survey. Open-File Report 95-476. Wiig, Stephen, Grundy, W.D., Dyni, John R.
Petrochemical Production
ACC (2021) Business of Chemistry (Annual Data). American Chemistry Council, Arlington, VA.
AN (2014) About Acrylonitrile: Production. AN Group, Washington, D.C. Available online at:
http://www.angroup.org/about/production.cfm.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Johnson, G. L. (2005 through 2010) Personal communication. Greg Johnson of Liskow & Lewis, on behalf of the
International Carbon Black Association (ICBA) and Caroline Cochran, ICF International. September 2010.
Johnson, G. L. (2003) Personal communication. Greg Johnson of Liskow& Lewis, on behalf of the International
Carbon Black Association (ICBA) and Caren Mintz, ICF International. November 2003.
United States Environmental Protection Agency (EPA) (2021) Greenhouse Gas Reporting Program. Aggregation of
Reported Facility Level Data under Subpart X -National Petrochemical Production for Calendar Years 2010 through
2020. Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency,
Washington, D.C.
U.S. 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.
U.S. EPA (2008) Technical Support Document for the Petrochemical Production Sector: Proposed Rule for
Mandatory Reporting of Greenhouse Gases. U.S. Environmental Protection Agency. September 2008.
U.S. EPA (2000) Economic Impact Analysis for the Proposed Carbon Black Manufacturing NESHAP, U.S.
Environmental Protection Agency. Research Triangle Park, NC. EPA-452/D-00-003. May 2000.
References and Abbreviations 33
-------
HCFC-22 Production
ARAP (2010) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. September 10, 2010.
ARAP (2009) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. September 21, 2009.
ARAP (2008) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. October 17, 2008.
ARAP (2007) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. October 2, 2007.
ARAP (2006) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
Atmospheric Policy to Sally Rand of the U.S. Environmental Protection Agency. July 11, 2006.
ARAP (2005) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. August 9, 2005.
ARAP (2004) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. June 3, 2004.
ARAP (2003) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
Atmospheric Policy to Sally Rand of the U.S. Environmental Protection Agency. August 18, 2003.
ARAP (2002) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. August 7, 2002.
ARAP (2001) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. August 6, 2001.
ARAP (2000) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
Atmospheric Policy to Sally Rand of the U.S. Environmental Protection Agency. August 13, 2000.
ARAP (1999) Facsimile from Dave Stirpe, Executive Director, Alliance for Responsible Atmospheric Policy to
Deborah Ottinger Schaefer of the U.S. Environmental Protection Agency. September 23,1999.
ARAP (1997) Letter from Dave Stirpe, Director, Alliance for Responsible Atmospheric Policy to Elizabeth Dutrow of
the U.S. Environmental Protection Agency. December 23,1997.
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.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom
996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
RTI (2008) "Verification of Emission Estimates of HFC-23 from the Production of HCFC-22:Emissions from 1990
through 2006." Report prepared by RTI International for the Climate Change Division. March 2008.
RTI (1997) "Verification of Emission Estimates of HFC-23 from the Production of HCFC-22: Emissions from 1990
through 1996." Report prepared by Research Triangle Institute for the Cadmus Group. November 25,1997; revised
February 16,1998.
34 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
UNFCCC (2014) Report of the Conference of the Parties on its nineteenth session, held in Warsaw from 11 to 23
November 2013. United Nations Framework Convention on Climate Change, Warsaw. (FCCC/CP/2013/10/Add.3).
January 31, 2014. Available online at: http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
Carbon Dioxide Consumption
ARI (1990 through 2010) CO2 Use in Enhanced Oil Recovery. Deliverable to ICF International under Task Order 102,
July 15, 2011.
ARI (2007) CO2-EOR: An Enabling Bridge for the Oil Transition. Presented at "Modeling the Oil Transition—a
DOE/EPA Workshop on the Economic and Environmental Implications of Global Energy Transitions." Washington,
D.C. April 20-21, 2007.
ARI (2006) CO2-EOR: An Enabling Bridge for the Oil Transition. Presented at "Modeling the Oil Transition—a
DOE/EPA Workshop on the Economic and Environmental Implications of Global Energy Transitions." Washington,
D.C. April 20-21, 2006.
Broadhead (2003) Personal communication. Ron Broadhead, Principal Senior Petroleum Geologist and Adjunct
faculty, Earth and Environmental Sciences Department, New Mexico Bureau of Geology and Mineral Resources,
and Robin Petrusak, ICF International. September 5, 2003.
COGCC (2014) Monthly CO2 Produced by County (1999-2009). Available online at:
http://cogcc.state.co.us/COGCCReports/production.aspx?id=MonthlyCQ2ProdByCounty. Accessed October 2014.
Denbury Resources Inc. (2002 through 2010) Annual Report: 2001 through 2009, Form 10-K. Available online at:
http://www.denburv.com/investor-relations/SEC-Filings/SEC-Filings-Details/default.aspx?Filingld=9823015.
Accessed September 2014.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
New Mexico Bureau of Geology and Mineral Resources (2006) Natural Accumulations of Carbon Dioxide in New
Mexico and Adjacent Parts of Colorado and Arizona: Commercial Accumulation of CO2. Available online at:
http://geoinfo.nmt.edU/staff/broadhead/C02.html#commercial.
U.S. Environmental Protection Agency (EPA) (2021) Greenhouse Gas Reporting Program (GHGRP). Aggregation of
Reported Facility Level Data under Subpart PP -National Level CO2 Transferred for Food & Beverage Applications
for Calendar Years 2010 through 2020. Office of Air and Radiation, Office of Atmospheric Programs, U.S.
Environmental Protection Agency, Washington, D.C.
U.S. 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.
Phosphoric Acid Production
EFMA (2000) "Production of Phosphoric Acid." Best Available Techniques for Pollution Prevention and Control in
the European Fertilizer Industry. Booklet 4 of 8. European Fertilizer Manufacturers Association. Available online at:
http://www.efma.org/Publications/BAT%202000/Bat04/section04.asp.
Florida Institute of Phosphate Research (FIPR) (2003a) "Analyses of Some Phosphate Rocks." Facsimile Gary
Albarelli, Florida Institute of Phosphate Research, Bartow, Florida, to Robert Lanza, ICF International. July 29, 2003.
FIPR (2003b) Florida Institute of Phosphate Research. Personal communication. Mr. Michael Lloyd, Laboratory
Manager, FIPR, Bartow, Florida, to Mr. Robert Lanza, ICF International. August 2003.
Golder Associates and M3 Engineering, Bayovar 12 Phosphate Project: Nl 43-101 Updated Pre-Feasibility Study,
Issued June 28, 2016. Available at:
References and Abbreviations 35
-------
https://www.sec.gov/Archives/edgar/data/1471603/000121716016000634/focusiune2016bayovar techrep.htm.
Accessed on October 7, 2020.
NCDENR (2013) North Carolina Department of Environment and Natural Resources, Title V Air Permit Review for
PCS Phosphate Company, Inc. - Aurora. Available online at:
http://www.ncair.org/permits/permit reviews/PCS rev 08282012.pdf. Accessed on January 25, 2013.
United States Geological Survey (USGS) (2021a) Mineral Commodity Summaries: Phosphate Rock 2021. January
2021. U.S. Geological Survey, Reston, VA. Accessed August 2021. Available online at:
https://www.usgs.gov/centers/nmic/phosphate-rock-statistics-and-information.
USGS (2021b) Personal communication between Stephen Jasinski (USGS) and Amanda Chiu (EPA) on August 25,
2021.
USGS (2020) Mineral Commodity Summaries: Phosphate Rock 2020. January 2020. U.S. Geological Survey, Reston,
VA. Accessed September 2020. Available online at: https://www.usgs.gov/centers/nmic/phosphate-rock-statistics-
and-information.
USGS (2019) Mineral Commodity Summaries: Phosphate Rock 2019. February 2019. U.S. Geological Survey, Reston,
VA. Accessed August 2019. Available online at: https://www.usgs.gov/centers/nmic/phosphate-rock-statistics-and-
information.
USGS (2019b) Communication between Stephen Jasinski (USGS) and John Steller EPA on November 15, 2019.
USGS (2018) Mineral Commodity Summaries: Phosphate Rock 2018. January 2018. U.S. Geological Survey, Reston,
VA. Available online at: https://www.usgs.gov/centers/nmic/phosphate-rock-statistics-and-information.
USGS (2017) Mineral Commodity Summaries: Phosphate Rock 2017. January 2017. U.S. Geological Survey, Reston,
VA. Available online at: https://www.usgs.gov/centers/nmic/phosphate-rock-statistics-and-information.
USGS (2016) Mineral Commodity Summaries: Phosphate Rock 2016. January 2016. U.S. Geological Survey, Reston,
VA. Available online at: https://www.usgs.gov/centers/nmic/phosphate-rock-statistics-and-information.
USGS (1994 through 2015b) Minerals Yearbook. Phosphate Rock Annual Report. U.S. Geological Survey, Reston, VA.
USGS (2012) Personal communication between Stephen Jasinski (USGS) and Mausami Desai (EPA) on October 12,
2012.
Iron and Steel Production and Metallurgical Coke Production
American Coke and Coal Chemicals Institute (ACCCI) (2021) U.S. Coke Plants as of November 2021, ACCCI,
Washington, D.C. November 2021.
American Iron and Steel Institute (AISI) (2004 through 2021) Annual Statistical Report, American Iron and Steel
Institute, Washington, D.C.
Carroll (2016) Personal communication, Colin P. Carroll, Director of Environment, Health and Safety, American Iron
and Steel Institute and Mausami Desai, U.S. Environmental Protection Agency, December 2016.
Carroll (2017) Personal communication, Colin P. Carroll, Director of Environment, Health and Safety, American Iron
and Steel Institute and John Steller, U.S. Environmental Protection Agency, November 2017.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
IPCC/UNEP/OECD/IEA (1995) "Volume 3: Greenhouse Gas Inventory Reference Manual. Table 2-2." IPCC Guidelines
for National Greenhouse Gas Inventories. Intergovernmental Panel on Climate Change, United Nations
Environment Programme, Organization for Economic Co-Operation and Development, International Energy
Agency. IPCC WG1 Technical Support Unit, United Kingdom.
36 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Steiner (2008) Personal communication, Bruce Steiner, Technical Consultant with the American Iron and Steel
Institute and Mausami Desai, U.S. Environmental Protection Agency, October 2008.
Tuck (2020) Personal communication, Christopher Tuck, Commodity Specialist, U.S. Geological Survey and Amanda
Chiu, U.S. Environmental Protection Agency, November 2020.
United States Department of Energy (DOE) (2000) Energy and Environmental Profile of the U.S. Iron and Steel
Industry. Office of Industrial Technologies, U.S. Department of Energy. August 2000. DOE/EE-0229.EIA.
United States Energy Information Administration (EIA) (1998 through 2019) Quarterly Coal Report: October-
December, Energy Information Administration, U.S. Department of Energy, Washington, D.C.
U.S. EIA (2021) Quarterly Coal Report: January - March, Energy Information Administration, U.S. Department of
Energy. Washington, D.C.
U.S. EIA (2020) Natural Gas Annual 2019. Energy Information Administration, U.S. Department of Energy.
Washington, D.C. September 2020.
U.S. EIA (2017b) Monthly Energy Review, December 2017, Energy Information Administration, U.S. Department of
Energy, Washington, D.C. DOE/EIA-0035(2015/12).
U.S. EIA (1992) Coal and lignite production. EIA State Energy Data Report 1992, Energy Information Administration,
U.S. Department of Energy, Washington, D.C.
United States Environmental Protection Agency (EPA) (2010) Carbon Content Coefficients Developed for EPA's
Mandatory Reporting Rule. Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental
Protection Agency, Washington, D.C.
U.S. EPA (2021). Greenhouse Gas Reporting Program. Dataset as of August 7, 2021. Available online at:
https://ghgdata.epa.gov/ghgp/.
United States Geological Survey (USGS) (2021) 2021 Mineral Commodities Summaries: Iron and Steel. U.S.
Geological Survey, Reston, VA. January 2021.
USGS (2020) 2020 USGS Minerals Yearbook - Iron and Steel Scrap (unreleased tables). U.S. Geological Survey,
Reston, VA.
USGS (2019) 2019 USGS Minerals Yearbook - Iron and Steel Scrap (tables-only release). U.S. Geological Survey,
Reston, VA.
USGS (2018) 2018 USGS Minerals Yearbook - Iron and Steel Scrap (tables-only release). U.S. Geological Survey,
Reston, VA.
USGS (2017) 2017 USGS Minerals Yearbook - Iron and Steel. U.S. Geological Survey, Reston, VA.
USGS (1991 through 2017) USGS Minerals Yearbook - Iron and Steel Scrap. U.S. Geological Survey, Reston, VA.
Ferroalloy Production
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Onder, H., and E.A. Bagdoyan (1993) Everything You've Always Wanted to Know about Petroleum Coke. Allis
Mineral Systems.
United States Geological Survey (USGS) (2021a) 2020 Mineral Commodity Summaries: Silicon. U.S. Geological
Survey, Reston, VA. January 2021.
USGS (2021b) 2018 Minerals Yearbook: Ferroalloys (tables-only release). U.S. Geological Survey, Reston, VA. May
2021.
References and Abbreviations 37
-------
USGS (2021c) 2020 Mineral Yearbook: Silicon (tables-only release). U.S Geological Survey, Reston, VA. September
2021.
USGS (2020) 2016 Minerals Yearbook: Ferroalloys (Advanced Release). U.S. Geological Survey, Reston, VA. January
2020.
USGS (2019) Mineral Industry Surveys: Silicon in May 2019. U.S. Geological Survey, Reston, VA. August 2019.
USGS (2018a) 2015 Minerals Yearbook: Ferroalloys. U.S. Geological Survey, Reston, VA. May 2018.
USGS (2018b) Mineral Industry Surveys: Silicon in July 2018. U.S. Geological Survey, Reston, VA. September 2018.
USGS (2017) Mineral Industry Surveys: Silicon in April 2017. U.S. Geological Survey, Reston, VA. June 2017.
USGS (2016a) 2014 Minerals Yearbook: Ferroalloys. U.S. Geological Survey, Reston, VA. October 2016.
USGS (2016b) Mineral Industry Surveys: Silicon in December 2016. U.S. Geological Survey, Reston, VA. December
2016.
USGS (2015a) Mineral Industry Surveys: Silicon in June 2015. U.S. Geological Survey, Reston, VA. September 2015.
USGS (2014) Mineral Industry Surveys: Silicon in September 2014. U.S. Geological Survey, Reston, VA. December
2014.
USGS (2013) 2013 Minerals Yearbook: Chromium. U.S. Geological Survey, Reston, VA. March 2016.
USGS (1996 through 2013) Minerals Yearbook: Silicon. U.S. Geological Survey, Reston, VA.
Aluminum Production
EPA (2021) Greenhouse Gas Reporting Program (GHGRP). Envirofacts, Subpart: F Aluminum Production. Available
online at:
https://enviro.epa.gov/enviro/ef metadata html.ef metadata table?p table name=F SUBPA L INFORM
ATlON&p topic=GHG.
EPA (2020) Greenhouse Gas Reporting Program (GHGRP). Envirofacts, Subpart: F Aluminum Production. Available
online at: https://enviro.epa.gov/facts/ghg/search.html.
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.
IPPC (2019) 2019 Refinement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change. [Calvo Buendia, E.,
Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M., Ngarize, S., Osako, A., Pyrozhenko, Y., Shermanau, P. and
Federici, S. (eds.)]. Hayama, Kanagawa, Japan.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
USAA (2020) U.S. Primary Aluminum Production: Report for August 2020. U.S. Aluminum Association, Washington,
D.C. September 2020.
USAA (2019) U.S. Primary Aluminum Production: Report for August 2019. U.S. Aluminum Association, Washington,
D.C. September 2019.
USAA (2018) U.S. Primary Aluminum Production: Report for August 2018. U.S. Aluminum Association, Washington,
D.C. September 2018.
USAA (2017) U.S. Primary Aluminum Production: Report for September 2017. U.S. Aluminum Association,
Washington, D.C. October 2017.
38 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
USAA (2016a) U.S. Primary Aluminum Production: Report for February 2016. U.S. Aluminum Association,
Washington, D.C. March 2016.
USAA (2016b) U.S. Primary Aluminum Production: Report for August 2016. U.S. Aluminum Association,
Washington, D.C. August 2016.
USAA (2015) U.S. Primary Aluminum Production: Report for June 2015. U.S. Aluminum Association, Washington,
D.C.July 2015.
USAA (2014) U.S. Primary Aluminum Production 2013. U.S. Aluminum Association, Washington, D.C. October 2014.
USAA (2013) U.S. Primary Aluminum Production 2012. U.S. Aluminum Association, Washington, D.C. January 2013.
USAA (2012) U.S. Primary Aluminum Production 2011. U.S. Aluminum Association, Washington, D.C. January 2012.
USAA (2011) U.S. Primary Aluminum Production 2010. U.S. Aluminum Association, Washington, D.C.
USAA (2010) U.S. Primary Aluminum Production 2009. U.S. Aluminum Association, Washington, D.C.
USAA (2008, 2009) U.S. Primary Aluminum Production. U.S. Aluminum Association, Washington, D.C.
USAA (2004, 2005, 2006) Primary Aluminum Statistics. U.S. Aluminum Association, Washington, D.C.
USGS (2021) 2020 Mineral Commodity Summaries: Aluminum. U.S. Geological Survey, Reston, VA.
USGS (2020) Mineral Industry Surveys: Aluminum in December 2020. U.S. Geological Survey, Reston VA. December
2020
USGS (2020) 2019 Mineral Commodity Summaries: Aluminum. U.S. Geological Survey, Reston, VA.
USGS (2019) 2017 Mineral Yearbook: Aluminum. U.S. Geological Survey, Reston, VA.USGS (2007) 2006 Mineral
Yearbook: Aluminum. U.S. Geological Survey, Reston, VA.USGS (1995,1998, 2000, 2001, 2002) Minerals Yearbook:
Aluminum Annual Report. U.S. Geological Survey, Reston, VA.
Magnesium Production and Processing
ARB (2015) "Magnesium casters successfully retool for a cleaner future." California Air Resources Board News
Release. Release # 15-07. February 5, 2015. Accessed October 2017. Available online at:
https://www.arb.ca.gov/newsrel/newsrelease.php?id=704.
Bartos S., C. Laush, J. Scharfenberg, and R. Kantamaneni (2007) "Reducing greenhouse gas emissions from
magnesium die casting." Journal of Cleaner Production, 15: 979-987, March.
EPA (2020) Envirofacts. Greenhouse Gas Reporting Program (GHGRP), Subpart T: Magnesium Production and
Processing. Available online at: https://www.epa.gov/enviro/greenhouse-gas-customized-search. Accessed on
October 2020.
Gjestland, H. and D. Magers (1996) "Practical Usage of Sulphur [Sulfur] Hexafluoride for Melt Protection in the
Magnesium Die Casting Industry." #13,1996 Annual Conference Proceedings, International Magnesium
Association. Ube City, Japan.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
RAND (2002) RAND Environmental Science and Policy Center, "Production and Distribution of SF6 by End-Use
Applications" Katie D. Smythe. International Conference on SF6 and the Environment: Emission Reduction
Strategies. San Diego, CA. November 21-22, 2002.
USGS (1995 through 2020) Minerals Yearbook: Magnesium Annual Report. U.S. Geological Survey, Reston, VA.
Available online at: http://minerals.usgs.gOv/minerals/pubs/commoditv/magnesium/index.html#mis.
References and Abbreviations 39
-------
USGS (2010b) Mineral Commodity Summaries: Magnesium Metal. U.S. Geological Survey, Reston, VA. Available
online at: http://minerals.usgs.gov/minerals/pubs/commoditv/magnesium/mcs-2010-mgmet.pdf.
USGS (2005b) Personal Communication between Deborah Kramer of the USGS and Jeremy Scharfenberg of ICF
Consulting.
Lead Production
Dutrizac, J.E., V. Ramachandran, and J.A. Gonzalez (2000) Lead-Zinc 2000. The Minerals, Metals, and Materials
Society.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Morris, D., F.R. Steward, and P. Evans (1983) Energy Efficiency of a Lead Smelter. Energy 8(5):337-349.
Sjardin, M. (2003) CO2 Emission Factors for Non-Energy Use in the Non-Ferrous Metal, Ferroalloys and Inorganics
Industry. Copernicus Institute. Utrecht, the Netherlands.
Ullman (1997) Ullman's Encyclopedia of Industrial Chemistry: Fifth Edition. Volume A5. John Wiley and Sons.
United States Geological Survey (USGS) (2021) 2020 Mineral Commodity Summary, Lead. U.S. Geological Survey,
Reston, VA. February 2021.
USGS
(2020)
2019 Mineral
Commodity Summary,
Lead.
U.S.
Geological
Survey,
Reston,
VA.
February 2020.
USGS
(2019)
2018 Mineral
Commodity Summary,
Lead.
U.S.
Geological
Survey,
Reston,
VA.
February 2019.
USGS
(2018)
2017 Mineral
Commodity Summary,
Lead.
U.S.
Geological
Survey,
Reston,
VA.
January 2018.
USGS
(2017)
2016 Mineral
Commodity Summary,
Lead.
U.S.
Geological
Survey,
Reston,
VA.
January 2017.
USGS
(2016)
2015 Mineral
Commodity Summary,
Lead.
U.S.
Geological
Survey,
Reston,
VA.
January 2016.
USGS
(2015)
2014 Mineral
Commodity Summary,
Lead.
U.S.
Geological
Survey,
Reston,
VA.
January 2015.
USGS
(2014)
2013 Mineral
Commodity Summary,
Lead.
U.S.
Geological
Survey,
Reston,
VA.
February 2014.
USGS (1995 through 2013) Minerals Yearbook: Lead Annual Report. U.S. Geological Survey, Reston, VA.
Zinc Production
American Zinc Recycling (AZR) (2021) Summary of Company History. Available online at https://azr.com/our-
history/. Accessed on March 16, 2021.
AZR (2020) Personal communication. Erica Livingston, Environmental Affairs Manager, American Zinc Recycling and
Amanda Chiu, U.S. Environmental Protection Agency. October 29, 2020.
American Zinc Products (AZP) (2021) American Zinc Products Marks First Anniversary of Zinc Production. Available
online at https://americanzincproducts.com/american-zinc-products-marks-first-anniversary-of-zinc-production/.
Accessed on March 1, 2022.
Horsehead Corp. (2016) Form 10-k, Annual Report for the Fiscal Year Ended December 31, 2015. Available online
at: https://www.sec.gov/Archives/edgar/data/1385544/0Q0119312516725704/d236839dl0k.htm. Submitted on
January 25, 2017.
Horsehead Corp. (2015) Form 10-k, Annual Report for the Fiscal Year Ended December 31, 2014. Available online
at: http://www.sec.gov/Archives/edgar/data/1385544/000138554415000Q05/zinc-2014123110k.htm. Submitted
on March 2, 2015.
40 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Horsehead Corp. (2014) Form 10-k, Annual Report for the Fiscal Year Ended December 31, 2013. Available online
at: http://www.sec.gov/Archives/edgar/data/1385544/0001385544140000Q3/zinc-2013123110k.htm. Submitted
on March 13, 2014.
Horsehead Corp. (2013) Form 10-k, Annual Report for the Fiscal Year Ended December 31, 2012. Available online
at: http://www.sec.gov/Archives/edgar/data/1385544/000119312513110431/00Q1193125-13-110431-index.htm.
Submitted March 18, 2013.
Horsehead Corp. (2012a) Form 10-k, Annual Report for the Fiscal Year Ended December 31, 2011. Available online
at: http://www.sec.gov/Archives/edgar/data/1385544/00Q119312512107345/d293011dl0k.htm. Submitted on
March 9, 2012.
Horsehead Corp. (2012b) Horsehead's New Zinc Plant and its Impact on the Zinc Oxide Business. February 22, 2012.
Available online at: http://www.horsehead.net/downloadAttachmentNDO.php?ID=118. Accessed on September
10, 2015.
Horsehead Corp. (2011) 10-k Annual Report for the Fiscal Year Ended December 31, 2010. Available online at:
http://google.brand.edgar-online.com/default.aspx?sym=zinc. Submitted on March 16, 2011.
Horsehead Corp. (2010a) 10-k Annual Report for the Fiscal Year Ended December 31, 2009. Available online at:
http://google.brand.edgar-online.com/default.aspx?sym=zinc. Submitted on March 16, 2010.
Horsehead Corp. (2010b) Horsehead Holding Corp. Provides Update on Operations at its Monaco, PA Plant. July 28,
2010. Available online at: http://www.horsehead.net/pressreleases.php?showall=no&news=&lD=65.
Horsehead Corp (2009) 10-k Annual Report for the Fiscal Year Ended December 31, 2008. Available online at:
https://www.sec.gov/Archives/edgar/data/1385544/000095015209002674/l35087aelQvk.htm. Submitted on
March 16, 2009.
Horsehead Corp (2008) 10-k Annual Report for the Fiscal Year Ended December 31, 2007. Available online at:
http://google.brand.edgar-online.com/default.aspx?sym=zinc. Submitted on March 31, 2008.
Horsehead Corp (2007) Registration Statement (General Form) S-l. Available online at http://google.brand.edgar-
online.com/default.aspx?sym=zinc. Submitted on April 13, 2007.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Nyrstar (2017) 2016 Clarksville Fact Sheet. Available online at:
http://www.nyrstar.eom/~/media/Files/N/Nyrstar/operations/melting/fact-sheet-clarksville-en.pdf. Accessed on
September 27, 2017.
PIZO (2012) Available online at http://pizotech.com/index.html. Accessed on October 10, 2012.
PIZO (2021) Personal communication. Thomas Rheaume, Arkansas Department of Energy and Environment and
Amanda Chiu, U.S. Environmental Protection Agency. February 16, 2021.
Recycling Today (2020) "AZR to restart for zinc recycling plant in North Carolina." March 6, 2020.
https://www.recvclingtoday.com/article/american-zinc-recycling-restarting-north-carolina-plant-2020/. Accessed
October 10, 2020.
Steel Dust Recycling (SDR) (2021) Personal communication. Jeremy Whitten, EHS Manager, Steel Dust Recycling
LLC and Amanda Chiu, U.S. Environmental Protection Agency. January 8, 2021.
SDR (2018) Personal communication. Jeremy Whitten, EHS Manager, Steel Dust Recycling LLC and John Steller, U.S.
Environmental Protection Agency. October 25, 2018.
SDR (2017) Personal communication. Jeremy Whitten, EHS Manager, Steel Dust Recycling LLC and John Steller, U.S.
Environmental Protection Agency. January 26, 2017.
References and Abbreviations 41
-------
SDR (2015) Personal communication. Jeremy Whitten, EHS Manager, Steel Dust Recycling LLC and Gopi Manne,
Eastern Research Group, Inc. September 22, 2015.
SDR (2014) Personal communication. Art Rowland, Plant Manager, Steel Dust Recycling LLC and Gopi Manne,
Eastern Research Group, Inc. December 9, 2014.
SDR (2013) Available online at http://steeldust.com/home.htm. Accessed on October 29, 2013.
SDR (2012) Personal communication. Art Rowland, Plant Manager, Steel Dust Recycling LLC and Gopi Manne,
Eastern Research Group, Inc. October 5, 2012.
Sjardin (2003) CO2 Emission Factors for Non-Energy Use in the Non-Ferrous Metal, Ferroalloys and Inorganics
Industry. Copernicus Institute. Utrecht, the Netherlands.
United States Environmental Protection Agency (EPA) (1992) "Applications Analysis Report: Horsehead Resource
Development Company Inc., Flame Reactor Technology" EPA/540/A5-91/005. May 1992.
United States Geological Survey (USGS) (2020) 2020 Mineral Commodity Summary: Zinc. U.S. Geological Survey,
Reston, VA. February 2020. Available online at: https://pubs.usgs.gov/periodicals/mcs2020/mcs2020.pdf.
USGS (2021) 2021 Mineral Commodity Summary: Zinc. U.S. Geological Survey, Reston, VA. January 2021.
USGS (2020) 2020 Mineral Commodity Summary: Zinc. U.S. Geological Survey, Reston, VA. January 2020.
USGS (2019) 2019 Mineral Commodity Summary: Zinc. U.S. Geological Survey, Reston, VA. January 2019.
USGS (2018) 2018 Mineral Commodity Summary: Zinc. U.S. Geological Survey, Reston, VA. January 2018.
USGS (2017) 2017 Mineral Commodity Summary: Zinc. U.S. Geological Survey, Reston, VA. January 2017.
USGS (2016) 2016 Mineral Commodity Summary: Zinc. U.S. Geological Survey, Reston, VA. January 2016.
USGS (2015) 2015 Mineral Commodity Summary: Zinc. U.S. Geological Survey, Reston, VA. January 2015.
USGS (1995 through 2014) Minerals Yearbook: Zinc Annual Report. U.S. Geological Survey, Reston, VA.
Viklund-White (2000) The use ofLCAforthe environmental evaluation of the recycling of galvanized steel. ISIJ
International, Vol. 40. No. 3, pp 292-299.
Electronics Industry
Burton, C.S., and R. Beizaie (2001) "EPA's PFC Emissions Model (PEVM) v. 2.14: Description and Documentation"
prepared for Office of Global Programs, U. S. Environmental Protection Agency, Washington, DC. November 2001.
Citigroup Smith Barney (2005) Global Supply/Demand Model for Semiconductors. March 2005.
DisplaySearch (2010) DisplaySearch Q4'09 Quarterly FPD Supply/Demand and Capital Spending Report.
DisplaySearch, LLC.
Doering, R. and Nishi, Y (2000) "Handbook of Semiconductor Manufacturing Technology", Marcel Dekker, New
York, USA, 2000.
EPA (2006) Uses and Emissions of Liquid PFC Heat Transfer Fluids from the Electronics Sector. U.S. Environmental
Protection Agency, Washington, DC. EPA-430-R-06-901.
EPA (2010) Technical Support Document for Process Emissions from Electronics Manufacture (e.g., Micro-Electro-
Mechanical Systems, Liquid Crystal Displays, Photovoltaics, and Semiconductors). U.S. Environmental Protection
Agency, Washington, DC.
EPA Greenhouse Gas Reporting Program (GHGRP) Envirofacts. Subpart I: Electronics Manufacture. Available online
at: https://enviro.epa.gov/facts/ghg/search.html.
42 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
IPCC (2019) 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change. Calvo Buendia, E.,
Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M., Ngarize, S., Osako, A., Pyrozhenko, Y., Shermanau, P. and
Federici, S. (eds). Published: IPCC, Switzerland.
ITRS (2007, 2008, 2011, 2013) International Technology Roadmap for Semiconductors: 2006 Update, January 2007;
International Technology Roadmap for Semiconductors: 2007 Edition, January 2008; International Technology
Roadmap for Semiconductors: 2011, January 2012; Update, International Technology Roadmap for
Semiconductors: 2013 Edition, Available online at: https://www.semiconductors.org/resources/20Q7-intemational-
technology-roadmap-for-semiconductors-itrs/. These and earlier editions and updates are available online at:
https://www.semiconductors.org/resources/7fwp resource types=utilization-reports&fwp paged=2. Information
about the number of interconnect layers for years 1990-2010 is contained in Burton and Beizaie, 2001. PEVM is
updated using new editions and updates of the ITRS, which are published annually. SEMI - Semiconductor
Equipment and Materials Industry (2017) World Fab Forecast, August 2018 Edition.
Platzer, Michaela D. (2015) U.S. Solar Photovoltaic Manufacturing: Industry Trends, Global Competition, Federal
Support. Congressional Research Service. January 27, 2015. https://fas.org/sgp/crs/misc/R42509.pdf.
SEMI - Semiconductor Equipment and Materials Industry (2021) World Fab Forecast, June 2021 Edition.
SEMI - Semiconductor Equipment and Materials Industry (2018) World Fab Forecast, June 2018 Edition.
SEMI - Semiconductor Equipment and Materials Industry (2016) World Fab Forecast, May 2017 Edition.
SEMI - Semiconductor Equipment and Materials Industry (2013) World Fab Forecast, May 2013 Edition.
SEMI - Semiconductor Equipment and Materials Industry (2012) World Fab Forecast, August 2012 Edition.
Semiconductor Industry Association (SIA) (2009-2011) STATS: SICAS Capacity and Utilization Rates Q1-Q4 2008, Ql-
Q4 2009, Q1-Q4 2010. Available online at:
http://www.semiconductors.org/industry statistics/semiconductor capacity utilization sicas reports/.
United States Census Bureau (USCB) (2011, 2012, 2015, 2016, 2017, 2018, 2019, 2020) Historical Data: Quarterly
Survey of Plant Capacity Utilization. Available online at: https://www.census.gov/programs-surveys/qpc.html.
VLSI Research, Inc. (2012) Worldwide Silicon Demand. August 2012.
Substitution of Ozone Depleting Substances
Barwick, Vicki (1999) Sources of Uncertainty in Gas Chromatography and High Performance Liquid
Chromatography. Journal of Chromatography. 1999.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
EPA (2022a) Suppliers of Industrial GHGs and Products Containing GHGs. Greenhouse Gas Reporting Program.
Available online at: https://www.epa.gov/ghgreporting/suppliers-industrial-ghgs-and-products-containing-ghgs.
EPA (2022b) Summary of Updates to the Window Units End-use in the Vintaging Model. Prepared for U.S. EPA's
Stratospheric Protection Division by ICF under EPA Contract Number 68HERH19D0029.
EPA (2022c) Summary of Proposed Updates to the Metered Dose Inhaler End-use in the Vintaging Model. Prepared
for U.S. EPA's Stratospheric Protection Division by ICF under EPA Contract Number 68HERH19D0029.
References and Abbreviations 43
-------
EPA (2021a) GHGRP Data relevant to the AIM Act. Greenhouse Gas Reporting Program. Available online at:
https://www.epa.gov/ghgreporting/ghgrp-data-relevant-aim-act.
EPA (2021b) Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2019. U.S. Environmental Protection
Agency, April 2021. Available at: https://www.epa.gov/ghgemissions/inventory-us-greenhouse-gas-emissions-and-
sinks-1990-2019.
EPA (2021c) Summary of Research and Proposed Updates to Polyurethane Foam End-Uses in the Vintaging Model.
Prepared for U.S. EPA's Stratospheric Protection Division by ICF under EPA Contract Number 68HERH19D0029.
EPA (2018) EPA's Vintaging Model of ODS Substitutes: A Summary of the 2017 Peer Review. Office of Air and
Radiation. Document Number EPA-400-F-18-001. Available online at:
https://www.epa.gov/sites/production/files/2018-09/documents/epas-vintaging-model-of-ods-substitutes-peer-
review-factsheet.pdf.
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.
Hu et al. (2017) Considerable Contribution of the Montreal Protocol to Declining Greenhouse Gas Emissions from
the United States, Geophys. Res. Lett., 44, 8075-8083, doi:10.1002/2017GL074388, August 14, 2017.
NOAA/EPA (2020) Correspondence between Lei Hu (NOAA), Steve Montzka (NOAA) and Dave Godwin (U.S. EPA),
November 6-19, 2020.
U.S. Census (2021) Table A. Apportionment Population, Resident Population, and Overseas Population: 2020
Census and 2010 Census; downloaded 4/27/21 from
https://en.wikipedia.org/wiki/List of states and territories of the United States by populationtfcite note-
Census2020-8.
Electrical Transmission and Distribution
HIFLD (2019) Federal Energy Regulatory Commission. Homeland Infrastructure Foundation-Level Data (HIFLD).
2019. Accessed March 2021. Available online at: https://hifld-geoplatform.opendata.arcgis.com/datasets/electric-
power-transmission-lines.
HIFLD (2020) Federal Energy Regulatory Commission. Homeland Infrastructure Foundation-Level Data (HIFLD).
2020. Accessed October 2021. Available online at: https://hifld-
geoplatform.opendata.arcgis.com/datasets/electric-power-transmission-lines/explore?showTable=true.
Hu, L. et al. (2021) Atmosphere-based US emission estimates of SF6 for 2007 - 2018. Presented at General
Assembly 2021 (vEGU21: Gather Online).
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
Assessment Report of the Intergovernmental Panel on Climate Change. S. Solomon, D. Qin, M. Manning, Z. Chen,
M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.). Cambridge University Press. Cambridge, United Kingdom
996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
IPCC (1996) Climate Change 1995: The Science of Climate Change. Intergovernmental Panel on Climate Change, J.T.
Houghton, L.G. Meira Filho, B.A. Callander, N. Harris, A. Kattenberg, and K. Maskell (eds.). Cambridge University
Press. Cambridge, United Kingdom.
Levin et al. (2010) "The Global SF6Source Inferred from Long-term High Precision Atmospheric Measurements and
its Comparison with Emission Inventories." Atmospheric Chemistry and Physics, 10: 2655-2662.
44 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Middleton, B. (2000) Cold Weather Applications of Gas Mixture (SF6/N2, SF6/CF4) Circuit Breakers: A User Utility's
Perspective [Conference Presentation], The US Environmental Protection Agency's Conference on SF6 and the
Environment: Emission Reduction Strategies, San Diego, CA, United States. Available online at:
https://www.epa.gov/sites/default/files/2016-02/documents/conf00 roiddletort.pdf
O'Connell, P., F. Heil, J. Henriot, G. Mauthe, H. Morrison, L. Neimeyer, M. Pittroff, R. Probst, J.P. Tailebois (2002)
SFs in the Electric Industry, Status 2000, CIGRE. February 2002.
RAND (2004) 'Trends in SF6 Sales and End-Use Applications: 1961-2003," Katie D. Smythe. International Conference
on SFe and the Environment: Emission Reduction Strategies. RAND Environmental Science and Policy Center,
Scottsdale, AZ. December 1-3, 2004.
UDI (2017) 2017 UDI Directory of Electric Power Producers and Distributors, 125th Edition, Platts.
UDI (2013) 2013 UDI Directory of Electric Power Producers and Distributors,121st Edition, Platts.
UDI (2010) 2010 UDI Directory of Electric Power Producers and Distributors, 118th Edition, Platts.
UDI (2007) 2007 UDI Directory of Electric Power Producers and Distributors, 115th Edition, Platts.
UDI (2004) 2004 UDI Directory of Electric Power Producers and Distributors, 112th Edition, Platts.
UDI (2001) 2001 UDI Directory of Electric Power Producers and Distributors, 109th Edition, Platts.
UNFCCC (2014) Report of the Conference of the Parties on its nineteenth session, held in Warsaw from 11 to 23
November 2013. United Nations Framework Convention on Climate Change, Warsaw. (FCCC/CP/2013/10/Add.3).
January 31, 2014. Available online at: http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
Nitrous Oxide from Product Use
CGA (2003) "CGA Nitrous Oxide Abuse Hotline: CGA/NWSA Nitrous Oxide Fact Sheet." Compressed Gas
Association. November 3, 2003.
CGA (2002) "CGA/NWSA Nitrous Oxide Fact Sheet." Compressed Gas Association. March 25, 2002.
Heydorn, B. (1997) "Nitrous Oxide—North America." Chemical Economics Handbook, SRI Consulting. May 1997.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press. Cambridge, United Kingdom
996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Ottinger (2021) Personal communication. Deborah Ottinger, U.S. Environmental Protection Agency and Amanda
Chiu, U.S. Environmental Protection Agency. January 7, 2021.
Tupman, M. (2002) Personal communication. Martin Tupman, Airgas Nitrous Oxide and Laxmi Palreddy, ICF
International. July 3, 2002.
Industrial Processes and Product Use Sources of Precursor
Greenhouse Gases
EPA (2022) "Crosswalk of Precursor Gas Categories." U.S. Environmental Protection Agency. April 6, 2022.
References and Abbreviations 45
-------
EPA (2021a) "Criteria pollutants National Tier 1 for 1970 - 2020." National Emissions Inventory (NEI) Air Pollutant
Emissions Trends Data. Office of Air Quality Planning and Standards, March 2021. Available online at:
https://www.epa.gov/air-emissions-inventories/air-pollutant-emissions-trends-data.
EPA (2021b) "2017 National Emissions Inventory (NEI) Technical Support Document (TSD)." Office of Air Quality
Planning and Standards, April 2021. Available online at: https://www.epa.gov/air-emissions-inventories/2017-
national-emissions-inventory-nei-technical-support-document-tsd.
EPA (1997) Compilation of Air Pollutant Emission Factors, AP-42. Office of Air Quality Planning and Standards, U.S.
Environmental Protection Agency. Research Triangle Park, NC. October 1997.
Agriculture
Enteric Fermentation
Archibeque, S. (2011) Personal Communication. Shawn Archibeque, Colorado State University, Fort Collins,
Colorado and staff at ICF International.
Crutzen, P.J., I. Aselmann, and W. Seiler (1986) Methane Production by Domestic Animals, Wild Ruminants, Other
Herbivores, Fauna, and Humans. Tellus, 38B:271-284.
Donovan, K. (1999) Personal Communication. Kacey Donovan, University of California at Davis and staff at ICF
International.
Doren, P.E., J. F. Baker, C. R. Long and T. C. Cartwright (1989) Estimating Parameters of Growth Curves of Bulls, J
Animal Science 67:1432-1445.
Enns, M. (2008) Personal Communication. Dr. Mark Enns, Colorado State University and staff at ICF International.
EPA (2002) Quality Assurance/Quality Control and Uncertainty Management Plan for the U.S. Greenhouse Gas
Inventory: Procedures Manual for Quality Assurance/Quality Control and Uncertainty Analysis, U.S. Greenhouse
Gas Inventory Program, U.S. Environmental Protection Agency, Office of Atmospheric Programs, EPA430-R-02-
007B, June 2002.
ERG (2021) Updated Other Animal Population Distribution Methodology. ERG, Lexington, MA.
ERG (2016) Development of Methane Conversion Rate Scaling Factor and Diet-Related Inputs to the Cattle Enteric
Fermentation Model for Dairy Cows, Dairy Heifers, and Feedlot Animals. ERG, Lexington, MA. December 2016.
Galyean and Gleghorn (2001) Summary of the 2000 Texas Tech University Consulting Nutritionist Survey. Texas
Tech University. Available online at http://www.depts.ttu.edu/afs/burnett center/progress reports/bcl2.pdf.
June 2009.
Holstein Association (2010) History of the Holstein Breed (website). Available online at:
http://www.holsteinusa.com/holstein breed/breed history, html. Accessed September 2010.
ICF (2006) Cattle Enteric Fermentation Model: Model Documentation. Prepared by ICF International for the
Environmental Protection Agency. June 2006.
ICF (2003) Uncertainty Analysis of 2001 Inventory Estimates of Methane Emissions from Livestock Enteric
Fermentation in the U.S. Memorandum from ICF International to the Environmental Protection Agency. May 2003.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
Assessment Report of the Intergovernmental Panel on Climate Change. S. Solomon, D. Qin, M. Manning, Z. Chen,
M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.). Cambridge University Press. Cambridge, United Kingdom
996 pp.
46 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. H.S. Eggleston, L Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
IPCC (2019) 2019 Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The
Intergovernmental Panel on Climate Change. Calvo Buendia, E., Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M.,
Ngarize, S., Osako, A., Pyrozhenko, Y., Shermanau, P. and Federici, S. (eds). Hayama, Kanagawa, Japan.Johnson, D.
(2002) Personal Communication. Don Johnson, Colorado State University, Fort Collins, and ICF International.
Johnson, D. (1999) Personal Communication. Don Johnson, Colorado State University, Fort Collins, and David
Conneely, ICF International.
Kebreab E., K. A. Johnson, S. L Archibeque, D. Pape, and T. Wirth (2008) Model for estimating enteric methane
emissions from United States dairy and feedlot cattle. J. Anim. Sci. 86: 2738-2748.
Lippke, H., T. D. Forbes, and W. C. Ellis. (2000) Effect of supplements on growth and forage intake by stocker steers
grazing wheat pasture. J. Anim. Sci. 78:1625-1635.
National Bison Association (1999) Total Bison Population—1999. Report provided during personal email
communication with Dave Carter, Executive Director, National Bison Association, July 19, 2011.
Pinchak, W.E., D. R. Tolleson, M. McCloy, L. J. Hunt, R. J. Gill, R. J. Ansley, and S. J. Bevers (2004) Morbidity effects
on productivity and profitability of stocker cattle grazing in the southern plains. J. Anim. Sci. 82:2773-2779.
Platter, W. J., J. D. Tatum, K. E. Belk, J. A. Scanga, and G. C. Smith (2003) Effects of repetitive use of hormonal
implants on beef carcass quality, tenderness, and consumer ratings of beef palatability. J. Anim. Sci. 81:984-996.
Preston, R.L (2010) What's The Feed Composition Value of That Cattle Feed? Beef Magazine, March 1, 2010.
Available at: http://beefmagazine.com/nutrition/feed-composition-tables/feed-composition-value-cattle-0301.
Skogerboe, T. L, L Thompson, J. M. Cunningham, A. C. Brake, V. K. Karle (2000) The effectiveness of a single dose
of doramectin pour-on in the control of gastrointestinal nematodes in yearling stocker cattle. Vet. Parasitology
87:173-181.
Soliva, C.R. (2006) Report to the attention of IPCC about the data set and calculation method used to estimate
methane formation from enteric fermentation of agricultural livestock population and manure management in
Swiss agriculture. On behalf of the Federal Office for the Environment (FOEN), Berne, Switzerland.
U.S. Department of Agriculture (USDA) (2021a) Quick Stats: Agricultural Statistics Database. National Agriculture
Statistics Service, U.S. Department of Agriculture. Washington, D.C. Available online at
http://quickstats.nass.usda.gov/. Accessed May-June, 2021.
USDA (2021b) Economic Research Service Dairy Data. Available online at: https://www.ers.usda.gov/data-
products/dairy-data/. Accessed May 2021.
USDA (2019) 1987,1992,1997, 2002, 2007, 2012, and 2017 Census of Agriculture. National Agriculture Statistics
Service, U.S. Department of Agriculture. Washington, D.C. Available online at:
https://www.nass.usda.gov/AgCensus/index.php. May 2019.
USDA (1996) Beef Cow/Calf Health and Productivity Audit (CHAPA): Forage Analyses from Cow/Calf Herds in 18
States. National Agriculture Statistics Service, U.S. Department of Agriculture. Washington, D.C. Available online at
http://www.aphis.usda.gov/vs/ceah/cahm. March 1996.
USDA:APHIS:VS (2010) Beef 2007-08, Part V: Reference of Beef Cow-calf Management Practices in the United
States, 2007-08. USDA-APHIS-VS, CEAH. Fort Collins, CO.
USDA:APHIS:VS (2002) Reference of 2002 Dairy Management Practices. USDA-APHIS-VS, CEAH. Fort Collins, CO.
Available online at http://www.aphis.usda.gov/vs/ceah/cahm.
USDA:APHIS:VS (1998) Beef'97, Parts l-IV. USDA-APHIS-VS, CEAH. Fort Collins, CO. Available online at
http://www.aphis.usda.gov/animal health/nahms/beefcowcalf/index.shtml#beef97.
References and Abbreviations 47
-------
USDA:APHIS:VS (1996) Reference of 1996 Dairy Management Practices. USDA-APHIS-VS, CEAH. Fort Collins, CO.
Available online at http://www.aphis.usda.gov/vs/ceah/cahm.
USDA:APHIS:VS (1994) Beef Cow/Calf Health and Productivity Audit. USDA-APHIS-VS, CEAH. Fort Collins, CO.
Available online at http://www.aphis.usda.gov/vs/ceah/cahm.
USDA:APHIS:VS (1993) Beef Cow/Calf Health and Productivity Audit. USDA-APHIS-VS, CEAH. Fort Collins, CO.
August 1993. Available online at http://www.aphis.usda.gov/vs/ceah/cahm.
Vasconcelos and Galyean (2007) Nutritional recommendations of feedlot consulting nutritionists: The 2007 Texas
Tech University Study. J. Anim. Sci. 85:2772-2781.
Manure Management
ASAE (1998) ASAE Standards 1998, 45th Edition. American Society of Agricultural Engineers. St. Joseph, Ml.
Bryant, M.P., V.H. Varel, R.A. Frobish, and H.R. Isaacson (1976) In H.G. Schlegel (ed.)]; Seminar on Microbial Energy
Conversion. E. Goltz KG. Gottingen, Germany.
Bush, E. (1998) Personal communication with Eric Bush, Centers for Epidemiology and Animal Health, U.S.
Department of Agriculture regarding National Animal Health Monitoring System's (NAHMS) Swine '95 Study.
EPA (2021) AgSTAR Anaerobic Digester Database. Available online at: https://www.epa.gov/agstar/livestock-
ariaerobic-digester-database. Accessed September 2021.
EPA (2008) Climate Leaders Greenhouse Gas Inventory Protocol Offset Project Methodology for Project Type
Managing Manure with Biogas Recovery Systems.
EPA (2005) National Emission Inventory—Ammonia Emissions from Animal Agricultural Operations, Revised Draft
Report. U.S. Environmental Protection Agency. Washington, D.C. April 22, 2005.
EPA (2002a) Development Document for the Final Revisions to the National Pollutant Discharge Elimination System
(NPDES) Regulation and the Effluent Guidelines for Concentrated Animal Feeding Operations (CAFOS). U.S.
Environmental Protection Agency. EPA-821-R-03-001. December 2002.
EPA (2002b) Cost Methodology for the Final Revisions to the National Pollutant Discharge Elimination System
Regulation and the Effluent Guidelines for Concentrated Animal Feeding Operations. U.S. Environmental
Protection Agency. EPA-821-R-03-004. December 2002.
EPA (1992) Global Methane Emissions from Livestock and Poultry Manure, Office of Air and Radiation, U.S.
Environmental Protection Agency. February 1992.
ERG (2021) Updated Other Animal Population Distribution Methodology. Memorandum to EPA from ERG.
ERG (2019) "Incorporation of USDA 2016 ARMS Dairy Data into the Manure Management Greenhouse Gas
Inventory." Memorandum to USDA OCE and EPA from ERG, December 2019.
ERG (2018) "Incorporation of USDA 2009 ARMS Swine Data into the Manure Management Greenhouse Gas
Inventory." Memorandum to USDA OCE and EPA from ERG, November 2018.
ERG (2010a) "Typical Animal Mass Values for Inventory Swine Categories." Memorandum to EPA from ERG. July 19,
2010.
ERG (2010b) Telecon with William Boyd of USDA NRCS and Cortney Itle of ERG Concerning Updated VS and Nex
Rates. August 8, 2010.
ERG (2010c) "Updating Current Inventory Manure Characteristics new USDA Agricultural Waste Management Field
Handbook Values." Memorandum to EPA from ERG. August 13, 2010.
48 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
ERG (2008) "Methodology for Improving Methane Emissions Estimates and Emission Reductions from Anaerobic
Digestion System for the 1990-2007 Greenhouse Gas Inventory for Manure Management." Memorandum to EPA
from ERG. August 18, 2008.
ERG (2003a) "Methodology for Estimating Uncertainty for Manure Management Greenhouse Gas Inventory."
Contract No. GS-10F-0036, Task Order 005. Memorandum to EPA from ERG, Lexington, MA. September 26, 2003.
ERG (2003b) "Changes to Beef Calves and Beef Cows Typical Animal Mass in the Manure Management Greenhouse
Gas Inventory." Memorandum to EPA from ERG, October 7, 2003.
ERG (2001) Summary of development of MDP Factor for methane conversion factor calculations. ERG, Lexington,
MA. September 2001.
ERG (2000a) Calculations: Percent Distribution of Manure for Waste Management Systems. ERG, Lexington, MA.
August 2000.
ERG (2000b) Discussion of Methodology for Estimating Animal Waste Characteristics (Summary of Bo Literature
Review). ERG, Lexington, MA. June 2000.
Groffman, P.M., R. Brumme, K. Butterbach-Bahl, K.E. Dobbie, A.R. Mosier, D. Ojima, H. Papen, W.J. Parton, K.A.
Smith, and C. Wagner-Riddle (2000) "Evaluating annual nitrous oxide fluxes at the ecosystem scale." Global
Biogeochemical Cycles, 14(4): 1061-1070.
Hashimoto, A.G. (1984) "Methane from Swine Manure: Effect of Temperature and Influent Substrate Composition
on Kinetic Parameter (k)." Agricultural Wastes, 9:299-308.
Hashimoto, A.G., V.H. Varel, and Y.R. Chen (1981) "Ultimate Methane Yield from Beef Cattle Manure; Effect of
Temperature, Ration Constituents, Antibiotics and Manure Age." Agricultural Wastes, 3:241-256.
Hill, D.T. (1984) "Methane Productivity of the Major Animal Types." Transactions of the ASAE, 27(2):530-540.
Hill, D.T. (1982) "Design of Digestion Systems for Maximum Methane Production." Transactions of the ASAE,
25(l):226-230.
IPCC (2019) 2019 Refinement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change. [CalvoBuendia, E.,
Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M., Ngarize S., Osako, A., Pyrozhenko, Y., Shermanau, P. and
Federici, S. (eds)]. Switzerland.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Morris, G.R. (1976) Anaerobic Fermentation of Animal Wastes: A Kinetic and Empirical Design Fermentation. M.S.
Thesis. Cornell University.
National Bison Association (1999) Total Bison Population—1999. Report provided during personal email
communication with Dave Carter, Executive Director, National Bison Association July 19, 2011.
Ott, S.L (2000) Dairy '96 Study. Stephen L. Ott, Animal and Plant Health Inspection Service, U.S. Department of
Agriculture. June 19, 2000.
Robertson, G. P. and P. M. Groffman (2015) Nitrogen transformations. Soil Microbiology, Ecology, and
Biochemistry, pages 421-446. Academic Press, Burlington, Massachusetts, USA.
Safley, L.M., Jr. (2000) Personal Communication. Deb Bartram, ERG and LM. Safley, President, Agri-Waste
Technology. June and October 2000.
Sweeten, J. (2000) Personal Communication. John Sweeten, Texas A&M University and Indra Mitra, ERG. June
2000.
References and Abbreviations 49
-------
UEP (1999) Voluntary Survey Results—Estimated Percentage Participation/Activity. Caged Layer Environmental
Management Practices, Industry data submissions for EPA profile development, United Egg Producers and National
Chicken Council. Received from John Thorne, Capitolink. June 2000.
USDA (2021a) Quick Stats: Agricultural Statistics Database. National Agriculture Statistics Service, U.S. Department
of Agriculture. Washington, D.C. Available online at: http://quickstats.nass.usda.gov/.
USDA (2021b) Chicken and Eggs 2020 Summary. National Agriculture Statistics Service, U.S. Department of
Agriculture. Washington, D.C. February 2021. Available online at:
http://www.nass.usda.gov/Publications/index.asp.
USDA (2021c) Poultry - Production and Value 2020 Summary. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. April 2021. Available online at:
http://www.nass.usda.gov/Publications/index.asp.
USDA (2019b) Chicken and Eggs 2018 Summary. National Agriculture Statistics Service, U.S. Department of
Agriculture. Washington, D.C. February 2019. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2019b) Poultry - Production and Value 2018 Summary. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. April 2019. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2019c) Chicken and Eggs 2013-2017 Summary. National Agriculture Statistics Service, U.S. Department of
Agriculture. Washington, D.C. June 2019. Available online at: http://www.nass.usda.gov/Publications/index.php.
USDA (2019d) 1987,1992,1997, 2002, 2007, 2012, and 2017 Census of Agriculture. National Agriculture Statistics
Service, U.S. Department of Agriculture. Washington, D.C. Available online at:
https://www.nass.usda.gov/AgCensus/index.php. May 2019.
USDA (2018) Poultry - Production and Value 2017 Summary. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. April 2018. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2017) Poultry - Production and Value 2016 Summary. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. April 2017. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2016) Poultry - Production and Value 2015 Summary. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. April 2016. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2015) Poultry - Production and Value 2014 Summary. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. April 2015. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2014) Poultry - Production and Value 2013 Summary. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. April 2014. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2013a) Chicken and Eggs 2012 Summary. National Agriculture Statistics Service, U.S. Department of
Agriculture. Washington, D.C. February 2013. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2013b) Poultry - Production and Value 2012 Summary. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. April 2013. Available online at:
http://www.nass.usda.gov/Publications/index.php.
50 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
USDA (2012a) Chicken and Eggs 2011 Summary. National Agriculture Statistics Service, U.S. Department of
Agriculture. Washington, D.C. February 2012. Available online at:
http://www.nass. usda.gov/Publications/index.php.
USDA (2012b) Poultry - Production and Value 2011 Summary. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. April 2012. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2011a) Chicken and Eggs 2010 Summary. National Agriculture Statistics Service, U.S. Department of
Agriculture. Washington, D.C. February 2011. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2011b) Poultry - Production and Value 2010 Summary. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. April 2011. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2010a) Chicken and Eggs 2009 Summary. National Agriculture Statistics Service, U.S. Department of
Agriculture. Washington, D.C. February 2010. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2010b) Poultry - Production and Value 2009 Summary. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. April 2010. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2009a) Chicken and Eggs 2008 Summary. National Agriculture Statistics Service, U.S. Department of
Agriculture. Washington, D.C. February 2009. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2009b) Poultry - Production and Value 2008 Summary. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. April 2009. Available online at:
http://www.nass.usda.gov/Publications/index.php.
USDA (2009c) Chicken and Eggs - Final Estimates 2003-2007. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. March 2009. Available online at:
https://www.nass.usda.gov/Publications/index.php.
USDA (2009d) Poultry Production and Value—Final Estimates 2003-2007. National Agriculture Statistics Service,
U.S. Department of Agriculture. Washington, D.C. May 2009. Available online at:
https://www.nass.usda.gov/Publications/index.php.
USDA (2008) Agricultural Waste Management Field Handbook, National Engineering Handbook (NEH), Part 651.
Natural Resources Conservation Service, U.S. Department of Agriculture.
USDA (2004a) Chicken and Eggs—Final Estimates 1998-2003. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. April 2004. Available online at:
https://www.nass.usda.gov/Publications/index.php.
USDA (2004b) Poultry Production and Value—Final Estimates 1998-2002. National Agriculture Statistics Service,
U.S. Department of Agriculture. Washington, D.C. April 2004. Available online at:
https://www.nass.usda.gov/Publications/index.php.
USDA (1999) Poultry Production and Value—Final Estimates 1994-97. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. March 1999. Available online at:
https://www.nass.usda.gov/Publications/index.php.
USDA (1998) Chicken and Eggs—Final Estimates 1994-97. National Agriculture Statistics Service, U.S. Department
of Agriculture. Washington, D.C. December 1998. Available online at:
https://www.nass.usda.gov/Publications/index.php.
References and Abbreviations 51
-------
USDA (1996) Agricultural Waste Management Field Handbook, National Engineering Handbook (NEH), Part 651.
Natural Resources Conservation Service, U.S. Department of Agriculture. July 1996.
USDA (1995a) Poultry Production and Value—Final Estimates 1988-1993. National Agriculture Statistics Service,
U.S. Department of Agriculture. Washington, D.C. March 1995. Available online at:
https://www.nass.usda.gov/Publications/index.php.
USDA (1995b) Chicken and Eggs—Final Estimates 1988-1993. National Agriculture Statistics Service, U.S.
Department of Agriculture. Washington, D.C. December 1995. Available online at:
https://www.nass.usda.gov/Publications/index.php.
USDA (1994) Sheep and Goats—Final Estimates 1989-1993. National Agriculture Statistics Service, U.S. Department
of Agriculture. Washington, D.C. January 31,1994. Available online at:
https://www.nass.usda.gov/Publications/index.php.
USDA APHIS (2003) Sheep 2001, Part I: Reference of Sheep Management in the United States, 2001 and Part IV:
Baseline Reference of 2001 Sheep Feedlot Health and Management. USDA-APHIS-VS. Fort Collins, CO. #N356.0702.
Available online at http://www.aphis.usda.gov/animal health/nahrris/sheep/index.shtrol#sheep2001.
USDA APHIS (2000) Layers '99—Part II: References of 1999 Table Egg Layer Management in the U.S. USDA-APHIS-
VS. Fort Collins, CO. Available online at
http://www.aphis.usda.gov/animal health/nahms/poultry/downloads/layers99/Layers99 dr Partll.pdf.
USDA APHIS (1996) Swine '95: Grower/Finisher Part II: Reference of 1995 U.S. Grower/Finisher Health &
Management Practices. USDA-APHIS-VS. Fort Collins, CO. Available online at:
http://www.aphis.usda.gov/animal health/nahms/swine/downloads/swine95/Swine95 dr Partll.pdf.
Rice Cultivation
Baicich, P. (2013) The Birds and Rice Connection. Bird Watcher's Digest. Available online at:
http://www.usarice.com/doclib/194/6867.pdf.
Brockwell, P.J., and R.A. Davis (2016) Introduction to time series and forecasting. Springer.
Cantens, G. (2004 through 2005) Personal Communication. Janet Lewis, Assistant to Gaston Cantens, Vice
President of Corporate Relations, Florida Crystals Company and ICF International.
Cheng, K., S.M. Ogle, W.J. Parton, G. Pan. (2014) "Simulating greenhouse gas mitigation potentials for Chinese
croplands using the DAYCENT ecosystem model." Global Change Biology 20:948-962.
Cheng, K., S.M. Ogle, W.J. Parton and G. Pan. (2013) "Predicting methanogenesis from rice paddies using the
DAYCENT ecosystem model." Ecological Modelling 261-262:19-31.
Del Grosso, S.J., S.M. Ogle, W.J. Parton, and F.J. Breidt (2010) "Estimating Uncertainty in N2O Emissions from U.S.
Cropland Soils." Global Biogeochemical Cycles, 24, GB1009, doi:10.1029/2009GB003544.
Deren, C. (2002) Personal Communication and Dr. Chris Deren, Everglades Research and Education Centre at the
University of Florida and Caren Mintz, ICF International. August 15, 2002.
Fitzgerald, G.J., K. M. Scow & J. E. Hill (2000) "Fallow Season Straw and Rice Management Effects on Methane
Emissions in California Rice." Global biogeochemical cycles, 14 (3), 767-776.
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
the northern Central Valley of California determined by satellite imagery. California Fish and Game, 91: 207-215.
Gonzalez, R. (2007 through 2014) Email correspondence. Rene Gonzalez, Plant Manager, Sem-Chi Rice Company
and ICF International.
52 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Hardke, J.T. (2015) Trends in Arkansas rice production, 2014. B.R. Wells Arkansas Rice Research Studies 2014.
Norman, R.J. and Moldenhauer, K.A.K. (Eds.). Research Series 626, Arkansas Agricultural Experiment Station,
University of Arkansas.
Hardke, J. (2014) Personal Communication. Dr. Jarrod Hardke, Rice Extension Agronomist at the University of
Arkansas Rice Research and Extension Center and Kirsten Jaglo, ICF International. September 11, 2014.
Hardke, J. (2013) Email correspondence. Dr. Jarrod Hardke, Rice Extension Agronomist at the University of
Arkansas Rice Research and Extension Center and Cassandra Snow, ICF International. July 15, 2013.
Hardke, J.T., and Wilson, C.E. Jr., (2014) Trends in Arkansas rice production, 2013. B.R. Wells Arkansas Rice
Research Studies 2013. Norman, R.J., and Moldenhauer, K.A.K., (Eds.). Research Series 617, Arkansas Agricultural
Experiment Station, University of Arkansas.
Hardke, J.T., and Wilson, C.E. Jr., (2013) Trends in Arkansas rice production. B.R. Wells Arkansas Rice Research
Studies 2012. Norman, R.J., and Moldenhauer, K.A.K., (Eds.). Research Series 609, Arkansas Agricultural Experiment
Station, University of Arkansas.
Hollier, C. A. (ed), (1999) Louisiana rice production handbook. Louisiana State University Agricultural Center. LCES
Publication Number 2321. 116 pp.
Holzapfel-Pschorn, A., R. Conrad, and W. Seiler (1985) "Production, Oxidation, and Emissions of Methane in Rice
Paddies." FEMS Microbiology Ecology, 31:343-351.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Kirstein, A. (2003 through 2004, 2006) Personal Communication. Arthur Kirstein, Coordinator, Agricultural
Economic Development Program, Palm Beach County Cooperative Extension Service, FLand ICF International.
Klosterboer, A. (1997,1999 through 2003) Personal Communication. Arlen Klosterboer, retired Extension
Agronomist, Texas A&M University and ICF International. July 7, 2003.
Lindau, C.W. and P.K. Bollich (1993) "Methane Emissions from Louisiana First and Ratoon Crop Rice." Soil Science,
156:42-48.
Linquist, B.A., M.A. Adviento-Borbe, C.M. Pittelkow, C.v. Kessel, et al. (2012) Fertilizer management practices and
greenhouse gas emissions from rice systems: A quantitative review and analysis. Field Crops Research, 135:10-21.
Linscombe, S. (1999, 2001 through 2014) Email correspondence. Steve Linscombe, Professor with the Rice
Research Station at Louisiana State University Agriculture Center and ICF International.
LSU, (2015) Louisiana ratoon crop and conservation: Ratoon & Conservation Tillage Estimates. Louisiana State
University, College of Agriculture AgCenter. Online at: www.lsuagcenter.com.
Miller, M.R., Garr, J.D., and Coates, P.S., (2010) Changes in the status of harvested rice fields in the Sacramento
Valley, California: Implications for wintering waterfowl. Wetlands, 30: 939-947.
Neue, H.U., R. Wassmann, H.K. Kludze, W. Bujun, and R.S. Lantin (1997) "Factors and processes controlling
methane emissions from rice fields." Nutrient Cycling in Agroecosystems 49: 111-117.
Ogle, S.M., F.J. Breidt, M. Easter, S. Williams and K. Paustian. (2007) "An empirically based approach for estimating
uncertainty associated with modeling carbon sequestration in soils." Ecological Modelling 205:453-463.
Ogle, S.M., S. Spencer, M. Hartman, L. Buendia, L. Stevens, D. du Toit, J. Witi (2016) "Developing national baseline
GHG emissions and analyzing mitigation potentials for agriculture and forestry using an advanced national GHG
inventory software system." In Advances in Agricultural Systems Modeling 6, Synthesis and Modeling of
Greenhouse Gas Emissions and Carbon Storage in Agricultural and Forestry Systems to Guide Mitigation and
Adaptation, S. Del Grosso, LR. Ahuja and W.J. Parton (eds.), American Society of Agriculture, Crop Society of
America and Soil Science Society of America, pp. 129-148.
References and Abbreviations 53
-------
Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface Submodel: Description
and Testing". Glob. Planet. Chang. 19: 35-48.
Parton, W.J., D.S. Schimel, C.V. Cole, D.S. Ojima (1987) "Analysis of factors controlling soil organic matter levels in
Great Plains grasslands." Soil Science Society of America Journal 51:1173-1179.
Sass, R. L. (2001) Cm Emissions from Rice Agriculture. Good Practice Guidance and Uncertainty Management in
National Greenhouse Gas Inventories. 399-417. Available online at: http://www.ipcc-
nggip.iges.or.ip/public/gp/bgp/4 7 CH4 Rice Agriculture.pdf.
Sass, R.L, F.M. Fisher, P.A. Harcombe, and F.T. Turner (1990) "Methane Production and Emissions in a Texas Rice
Field." Global Biogeochemical Cycles, 4:47-68.
Sass, R.L, F.M. Fisher, S.T. Lewis, M.F. Jund, and F.T. Turner. (1994) "Methane emissions from rice fields: effect of
soil texture." Global Biogeochemical Cycles 8:135-140.
Schueneman, T. (1997,1999 through 2001) Personal Communication. Tom Schueneman, Agricultural Extension
Agent, Palm Beach County, FLand ICF International.
Slaton, N. (1999 through 2001) Personal Communication. Nathan Slaton, Extension Agronomist—Rice, University
of Arkansas Division of Agriculture Cooperative Extension Service and ICF International.
Stansel, J. (2004 through 2005) Email correspondence. Dr. Jim Stansel, Resident Director and Professor Emeritus,
Texas A&M University Agricultural Research and Extension Center and ICF International.
TAMU (2015) Texas Rice Crop Survey. Texas A&M AgriLIFE Research Center at Beaumont. Online at:
https://beaumont.tamu.edu/.
Texas Agricultural Experiment Station (2007 through 2014) Texas Rice Acreage by Variety. Agricultural Research
and Extension Center, Texas Agricultural Experiment Station, Texas A&M University System. Available online at:
http://beaumont.tamu.edu/CropSurvev/CropSurveyReport.aspx.
Texas Agricultural Experiment Station (2006) 2005 - Texas Rice Crop Statistics Report. Agricultural Research and
Extension Center, Texas Agricultural Experiment Station, Texas A&M University System, p. 8. Available online at:
http://beaumont.tamu.edu/eLibrary/TRRFReport default.htm.
University of California Cooperative Extension (UCCE) (2015) Rice Production Manual. Revised (2015) UCCE, Davis,
in collaboration with the California Rice Research Board.
USDA (2005 through 2015) Crop Production Summary. National Agricultural Statistics Service, Agricultural Statistics
Board, U.S. Department of Agriculture, Washington, D.C. Available online at: http://usda.mannlib.cornell.edu.
USDA (2012) Summary of USDA-ARS Research on the Interrelationship of Genetic and Cultural Management
Factors That Impact Grain Arsenic Accumulation in Rice. News and Events. Agricultural Research Service, U.S.
Department of Agriculture, Washington, D.C. Available online at:
http://www.ars.usda.gov/is/pr/2012/120919.htm. September 2013.
USDA (2003) Field Crops, Final Estimates 1997-2002. Statistical Bulletin No. 982. National Agricultural Statistics
Service, Agricultural Statistics Board, U.S. Department of Agriculture, Washington, D.C. Available online at:
http://usda.mannlib.cornell.edu/usda/reports/general/sb/. September 2005.
USDA (1998) Field Crops Final Estimates 1992-1997. Statistical Bulletin Number 947 a. National Agricultural
Statistics Service, Agricultural Statistics Board, U.S. Department of Agriculture, Washington, D.C. Available online
at: http://usda.mannlib.cornell.edu/. July 2001.
USDA (1994) Field Crops Final Estimates 1987-1992. Statistical Bulletin Number 896. National Agricultural Statistics
Service, Agricultural Statistics Board, U.S. Department of Agriculture, Washington, D.C. Available online at:
http://usda.mannlib.cornell.edu/. July 2001.
54 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory. Natural Resources Conservation Service,
Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd 1422028.pdf.
van Bodegom, P.M., R. Wassmann, T.M. Metra-Corton (2001) "A process based model for methane emission
predictions from flooded rice paddies." Global Biogeochemical Cycles 15: 247-263.
Wang, J.J., S.K. Dodla, S. Viator, M. Kongchum, S. Harrison, S. D. Mudi, S. Liu, Z. Tian (2013) Agriculture Field
Management Practices and Greenhouse Gas Emissions from Louisiana Soils. Louisiana Agriculture, Spring 2013: 8-
9. Available online at: http://www.lsuagcenter.com/NR/rdonlyres/78D8B61A-96A8-49El-B2EF-
BAlD4CE4E698/93016/v56no2Spring2013.pdf.
Wassmann, R. H.U. Neue, R.S. Lantin, K. Makarim, N. Chareonsil5, LV. Buendia, and H. Rennenberg (2000a)
Characterization of methane emissions from rice fields in Asia II. Differences among irrigated, rainfed, and
deepwater rice." Nutrient Cycling in Agroecosystems, 58(1): 13-22.
Wassmann, R., R.S. Lantin, H.U. Neue, LV. Buendia, et al. (2000b) "Characterization of Methane Emissions from
Rice Fields in Asia. III. Mitigation Options and Future Research Needs." Nutrient Cycling in Agroecosystems,
58(l):23-36.
Way, M.O., McCauley, G.M., Zhou, X.G., Wilson, L.T., and Morace, B. (Eds.), (2014) 2014 Texas Rice Production
Guidelines. Texas A&M AgriLIFE Research Center at Beaumont.
Wilson, C. (2002 through 2007, 2009 through 2012) Personal Communication. Dr. Chuck Wilson, Rice Specialist at
the University of Arkansas Cooperative Extension Service and ICF International.
Wilson, C.E. Jr., and Branson, J.W., (2006) Trends in Arkansas rice production. B.R. Wells Arkansas Rice Research
Studies 2005. Norman, R.J., Meullenet, J.-F., and Moldenhauer, K.A.K., (Eds.). Research Series 540, Arkansas
Agricultural Experiment Station, University of Arkansas.
Wilson, C.E. Jr., and Branson, J.W., (2005) Trends in Arkansas rice production. B.R. Wells Arkansas Rice Research
Studies 2004. Norman, R.J., Meullenet, J.-F., and Moldenhauer, K.A.K., (Eds.). Research Series 529, Arkansas
Agricultural Experiment Station, University of Arkansas.
Wilson, C.E. Jr., Runsick, S.K., and Mazzanti, R., (2010) Trends in Arkansas rice production. B.R. Wells Arkansas Rice
Research Studies 2009. Norman, R.J., and Moldenhauer, K.A.K., (Eds.). Research Series 581, Arkansas Agricultural
Experiment Station, University of Arkansas.
Wilson, C.E. Jr., Runsick, S.K., Mazzanti, R., (2009) Trends in Arkansas rice production. B.R. Wells Arkansas Rice
Research Studies (2008) Norman, R.J., Meullenet, J.-F., and Moldenhauer, K.A.K., (Eds.). Research Series 571,
Arkansas Agricultural Experiment Station, University of Arkansas.
Wilson, C.E. Jr., and Runsick, S.K., (2008) Trends in Arkansas rice production. B.R. Wells Arkansas Rice Research
Studies 2007. Norman, R.J., Meullenet, J.-F., and Moldenhauer, K.A.K., (Eds.). Research Series 560, Arkansas
Agricultural Experiment Station, University of Arkansas.
Wilson, C.E. Jr., and Runsick, S.K., (2007) Trends in Arkansas rice production. B.R. Wells Arkansas Rice Research
Studies 2006. Norman, R.J., Meullenet, J.-F., and Moldenhauer, K.A.K., (Eds.). Research Series 550, Arkansas
Agricultural Experiment Station, University of Arkansas.
Yan, X., H. Akiyana, K. Yagi, and H. Akimoto (2009) "Global estimations of the inventory and mitigation potential of
methane emissions from rice cultivation conducted using the 2006 Intergovernmental Panel on Climate Change
Guidelines." Global Biogeochemical Cycles, 23, DOI: 0.1029/2008GB003299.
Young, M. (2013) Rice and Ducks. Ducks Unlimited, Memphis, TN. Available online at:
http://www.ducks.org/conservation/farm-bill/rice-and-ducks—by-matt-young.
References and Abbreviations 55
-------
Agricultural Soil Management
AAPFCO (2008 through 2021) Commercial Fertilizers: 2008-2016. Association of American Plant Food Control
Officials. University of Missouri. Columbia, MO.
AAPFCO (1995 through 2000a, 2002 through 2007) Commercial Fertilizers: 1995-2007. Association of American
Plant Food Control Officials. University of Kentucky. Lexington, KY.
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
Cibrowski, P. (1996) Personal Communication. Peter Cibrowski, Minnesota Pollution Control Agency and Heike
Mainhardt, ICF Incorporated. July 29,1996.
Cheng, B., and D.M. Titterington (1994) "Neural networks: A review from a statistical perspective." Statistical
science 9: 2-30.
Claassen, R., M. Bowman, J. McFadden, D. Smith, and S. Wallander (2018) Tillage intensity and conservation
cropping in the United States, EIB 197. United States Department of Agriculture, Economic Research Service,
Washington, D.C.
CTIC (2004) 2004 Crop Residue Management Survey. Conservation Technology Information Center. Available at
http://www.ctic.purdue.edu/CRM/.
Del Grosso, S.J., A.R. Mosier, W.J. Parton, and D.S. Ojima (2005) "DAYCENT Model Analysis of Past and
Contemporary Soil N2O and Net Greenhouse Gas Flux for Major Crops in the USA." Soil Tillage and Research, 83: 9-
24. doi: 10.1016/j.still.2005.02.007.
Del Grosso, S.J., S.M. Ogle, W.J. Parton, and F.J. Breidt (2010) "Estimating Uncertainty in N2O Emissions from U.S.
Cropland Soils." Global Biogeochemical Cycles, 24, GB1009, doi:10.1029/2009GB003544.
Del Grosso, S.J., W.J. Parton, C.A. Keough, and M. Reyes-Fox. (2011) Special features of the DAYCENT modeling
package and additional procedures for parameterization, calibration, validation, and applications, in Methods of
Introducing System Models into Agricultural Research, L.R. Ahuja and Liwang Ma, editors, p. 155-176, American
Society of Agronomy, Crop Science Society of America, Soil Science Society of America, Madison, Wl. USA.
Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel (2001) "Simulated
Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the DAYCENT Model." In Schaffer, M., L. Ma,
S. Hansen, (eds.). Modeling Carbon and Nitrogen Dynamics for Soil Management. CRC Press. Boca Raton, Florida.
303-332.
Del Grosso, S.J., T. Wirth, S.M. Ogle, W.J. Parton (2008) Estimating agricultural nitrous oxide emissions. EOS 89,
529-530.
Delgado, J.A., S.J. Del Grosso, and S.M. Ogle (2009) "15N isotopic crop residue cycling studies and modeling suggest
that IPCC methodologies to assess residue contributions to N2O-N emissions should be reevaluated." Nutrient
Cycling in Agroecosystems, DOI 10.1007/sl0705-009-9300-9.
Edmonds, L, N. Gollehon, R.L. Kellogg, B. Kintzer, L. Knight, C. Lander, J. Lemunyon, D. Meyer, D.C. Moffitt, and J.
Schaeffer (2003) "Costs Associated with Development and Implementation of Comprehensive Nutrient
Management Plans." Part 1. Nutrient Management, Land Treatment, Manure and Wastewater Handling and
Storage, and Recordkeeping. Natural Resource Conservation Service, U.S. Department of Agriculture.
EPA (2003) Clean Watersheds Needs Survey 2000—Report to Congress, U.S. Environmental Protection Agency.
Washington, D.C. Available online at: http://www.epa.gov/owm/mtb/cwns/2000rtc/toc.htm.
EPA (1999) Biosolids Generation, Use and Disposal in the United States. Office of Solid Waste, U.S. Environmental
Protection Agency. Available online at: http://biosolids.policy.net/relatives/18941.PDF.
EPA (1993) Federal Register. Part II. Standards for the Use and Disposal of Sewage Sludge; Final Rules. U.S.
Environmental Protection Agency, 40 CFR Parts 257, 403, and 503.
56 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Firestone, M. K., and E.A. Davidson, Ed. (1989) Microbiological basis of NO and N2O production and consumption in
soil. Exchange of trace gases between terrestrial ecosystems and the atmosphere. New York, John Wiley & Sons.
ILENR (1993) Illinois Inventory of Greenhouse Gas Emissions and Sinks: 1990. Office of Research and Planning,
Illinois Department of Energy and Natural Resources. Springfield, IL.
IPCC (2013) 2013 Supplement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands. The
Intergovernmental Panel on Climate Change. [T, Hiraishi, T. Krug, K. Tanabe, N. Srivastava, B. Jamsranjav, M.
Fukuda and T. Troxler (eds.)]. Hayama, Kanagawa, Japan.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Little, R. (1988) "Missing-data adjustments in large surveys." Journal of Business and Economic Statistics 6: 287-
296.
McFarland, M.J. (2001) Biosolids Engineering, New York: McGraw-Hill, p. 2.12.
McGill, W.B., and C.V. Cole (1981) Comparative aspects of cycling of organic C, N, S and P through soil organic
matter. Geoderma 26:267-286.
Metherell, A.K., L.A. Harding, C.V. Cole, and W.J. Parton (1993) "CENTURYSoil Organic Matter Model
Environment." Agroecosystem version 4.0. Technical documentation, GPSRTech. Report No. 4, USDA/ARS, Ft.
Collins, CO.
NEBRA (2007) A National Biosolids Regulation, Quality, End Use & Disposal Survey. North East Biosolids and
Residuals Association, July 21, 2007.
Noller, J. (1996) Personal Communication. John Noller, Missouri Department of Natural Resources and Heike
Mainhardt, ICF Incorporated. July 30,1996.
Ogle, S.M., F.J. Breidt, M. Easter, S. Williams and K. Paustian (2007) "Empirically-Based Uncertainty Associated with
Modeling Carbon Sequestration Rates in Soils." Ecological Modeling 205:453-463.
Oregon Department of Energy (1995) Report on Reducing Oregon's Greenhouse Gas Emissions: Appendix D
Inventory and Technical Discussion. Oregon Department of Energy. Salem, OR.
Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface Submodel: Description
and Testing." Glob. Planet. Chang. 19: 35-48.
Potter, C., S. Klooster, A. Huete, and V. Genovese (2007) Terrestrial carbon sinks for the United States predicted
from MODIS satellite data and ecosystem modeling. Earth Interactions 11, Article No. 13, DOI 10.1175/EI228.1.
Potter, C. S., J.T. Randerson, C.B. Fields, P.A. Matson, P.M. Vitousek, H.A. Mooney, and S.A. Klooster (1993)
"Terrestrial ecosystem production: a process model based on global satellite and surface data." Global
Biogeochemical Cycles 7:811-841.
PRISM Climate Group (2018) PRISM Climate Data, Oregon State University, http://prism.oregonstate.edu,
downloaded 18 July 2018.
Pukelsheim, F. (1994) 'The 3-Sigma-Rule." American Statistician 48:88-91.
Ruddy B.C., D.L. Lorenz, and D.K. Mueller (2006) County-level estimates of nutrient inputs to the land surface of
the conterminous United States, 1982-2001. Scientific Investigations Report 2006-5012. U.S Department of the
Interior.
Scheer, C., S.J. Del Grosso, W.J. Parton, D.W. Rowlings, P.R. Grace (2013) Modeling Nitrous Oxide Emissions from
Irrigated Agriculture: Testing DAYCENT with High Frequency Measurements, Ecological Applications, in press.
Available online at: http://dx.doi.Org/10.1890/13-0570.l.
References and Abbreviations 57
-------
Soil Survey Staff (2019) Gridded Soil Survey Geographic (gSSURGO) Database for the Conterminous United States.
United States Department of Agriculture, Natural Resources Conservation Service. Available online at
https://gdg.sc.egov.usda.gov/. April, 2019 (FY2019 official release).
Towery, D. (2001) Personal Communication. Dan Towery regarding adjustments to the CTIC (1998) tillage data to
reflect long-term trends, Conservation Technology Information Center, West Lafayette, IN, and Marlen Eve,
National Resource Ecology Laboratory, Fort Collins, CO. February 2001.
TVA (1991 through 1992a, 1993 through 1994) Commercial Fertilizers. Tennessee Valley Authority, Muscle Shoals,
AL.
USDA-ERS (2018) Agricultural Resource Management Survey (ARMS) Farm Financial and Crop Production Practices:
Tailored Reports. Available online at: https://www.ers.usda.gov/data-products/arms-farm-financial-and-crop-
production-practices/.
USDA-ERS (1997) Cropping Practices Survey Data —1995. Economic Research Service, United States Department of
Agriculture. Available online at: http://www.ers.usda.gov/data/archive/93018/.
USDA-NASS (2019) Quick Stats. National Agricultural Statistics Service, United States Department of Agriculture,
Washington, D.C. http://quickstats.nass.usda.gov/.
USDA-NASS (2017) 2017 Census of Agriculture. USDA National Agricultural Statistics Service, Complete data
available at http://www.nass.usda.gov/AgCensus.
USDA-NASS (2012) 2012 Census of Agriculture. USDA National Agricultural Statistics Service, Complete data
available at http://www.nass.usda.gov/AgCensus.
USDA-NASS (2004) Agricultural Chemical Usage: 2003 Field Crops Summary. Report AgChl(04)a. National
Agricultural Statistics Service, U.S. Department of Agriculture, Washington, D.C. Available online at:
http://usda.mannlib.cornell.edu/reports/nassr/other/pcu-bb/agcs0504.pdfH.
USDA-NASS (1999) Agricultural Chemical Usage: 1998 Field Crops Summary. Report AgCHl(99). National
Agricultural Statistics Service, U.S. Department of Agriculture, Washington, DC. Available online at:
http://usda.mannlib.cornell.edu/reports/nassr/other/pcu-bb/agch0599.pdf.
USDA-NASS (1992) Agricultural Chemical Usage: 1991 Field Crops Summary. Report AgChl(92). National
Agricultural Statistics Service, U.S. Department of Agriculture, Washington, D.C. Available online at:
http://usda.mannlib.cornell.edu/reports/nassr/other/pcu-bb/agch0392.txtH.
USDA-NRCS (2012) Assessment of the Effects of Conservation Practices on Cultivated Cropland in the Upper
Mississippi River Basin. U.S. Department of Agriculture, Natural Resources Conservation Service,
https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/stelprdb 1042093.pdf.
USDA-NRCS (2018a) Summary Report: 2015 National Resources Inventory. Natural Resources Conservation Service,
Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
https://www.nrcs.usda.gov/lnternet/FSE DOCU ME NTS/nrcseprd 1422028.pdf.
USDA-NRCS (2018b) CEAP Cropland Farmer Surveys. USDA Natural Resources Conservation Service.
https://www.nrcs.usda.gov/wps/portal/nrcs/detail/national/technical/nra/ceap/na/?cid=nrcsl43 014163.
USFS (2019) Forest Inventory and Analysis Program. United States Department of Agriculture, U.S. Forest Service,
https://www.fia.fs.fed.us/tools-data/default.asp.
Van Buuren, S. (2012) "Flexible imputation of missing data." Chapman & Hall/CRC, Boca Raton, FL.
Wagner-Riddle, C., Congreves, K. A., Abalos, D., Berg, A. A., Brown, S. E., Ambadan, J. T., Gao, X. & Tenuta, M.
(2017) "Globally important nitrous oxide emissions from croplands induced by freeze-thaw cycles." Nature
Geosciences 10(4): 279-283.
Wisconsin Department of Natural Resources (1993) Wisconsin Greenhouse Gas Emissions: Estimates for 1990.
Bureau of Air Management, Wisconsin Department of Natural Resources, Madison, Wl.
58 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Yang, L, Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J., Fry, J., Funk, M.,
Granneman, B., Liknes, G. C., Rigge, M. & Xian, G. (2018) "A new generation of the United States National Land
Cover Database: Requirements, research priorities, design, and implementation strategies." ISPRS Journal of
Photogrammetry and Remote Sensing 146:108-123.
Liming
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Tepordei, V.V. (1997 through 2015) "Crushed Stone," In Minerals Yearbook. U.S. Department of the Interior/U.S.
Geological Survey. Washington, D.C. Available online at: http://minerals.usgs.gov/minerals/.
Tepordei, V.V. (2003b) Personal communication. Valentin Tepordei, U.S. Geological Survey and ICF Consulting,
August 18, 2003.
Tepordei, V.V. (1996) "Crushed Stone," In Minerals Yearbook 1994. U.S. Department of the Interior/Bureau of
Mines, Washington, D.C. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed August 2000.
Tepordei, V.V. (1995) "Crushed Stone," In Minerals Yearbook 1993. U.S. Department of the Interior/Bureau of
Mines, Washington, D.C. pp. 1107-1147.
Tepordei, V. V. (1994) "Crushed Stone," In Minerals Yearbook 1992. U.S. Department of the Interior/Bureau of
Mines, Washington, D.C. pp. 1279-1303.
USGS (2021) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the First Quarter of 2021, U.S.
Geological Survey, Reston, VA. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis.
USGS (2020) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the Fourth Quarter of 2020, U.S.
Geological Survey, Reston, VA. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis.
West, T.O., and A.C. McBride (2005) "The contribution of agricultural lime to carbon dioxide emissions in the
United States: dissolution, transport, and net emissions," Agricultural Ecosystems & Environment 108:145-154.
West, T.O. (2008) Email correspondence. Tristram West, Environmental Sciences Division, Oak Ridge National
Laboratory, U.S. Department of Energy and Nikhil Nadkarni, ICF International on suitability of liming emission
factor for the entire United States. June 9, 2008.
Willett, J.C. (2021b) Personal communication. Jason Willett. Preliminary data tables from "Crushed Stone," In 2020
Minerals Yearbook. U.S. Department of the Interior/U.S. Geological Survey. Washington, D.C. September, 2021.
Willett, J.C. (2021a) Personal communication. Jason Willett. Preliminary data tables from "Crushed Stone," In 2019
Minerals Yearbook. U.S. Department of the Interior/U.S. Geological Survey. Washington, D.C. September, 2021.
Willett, J.C. (2020b) Personal communication. Jason Willett. Preliminary data tables from "Crushed Stone," In 2018
Minerals Yearbook. U.S. Department of the Interior/U.S. Geological Survey. Washington, D.C. December 01, 2020.
Willett, J.C. (2019) Personal communication. Jason Willett. Preliminary data tables from "Crushed Stone," In 2017
Minerals Yearbook. U.S. Department of the Interior/U.S. Geological Survey. Washington, D.C. September 10, 2019.
Willett, J.C. (2020a) "Crushed Stone," In Minerals Yearbook 2016. U.S. Department of the Interior/U.S. Geological
Survey, Washington, D.C. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed November 2020.
References and Abbreviations 59
-------
Willett, J.C. (2017) "Crushed Stone," In Minerals Yearbook 2015. U.S. Department of the Interior/U.S. Geological
Survey, Washington, D.C. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed November 2017.
Willett, J.C. (2016) "Crushed Stone," In Minerals Yearbook 2014. U.S. Department of the Interior/U.S. Geological
Survey, Washington, D.C. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed September 2016.
Willett, J.C. (2015) "Crushed Stone," In Minerals Yearbook 2013. U.S. Department of the Interior/U.S. Geological
Survey, Washington, D.C. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed September 2015.
Willett, J.C. (2014) "Crushed Stone," In Minerals Yearbook 2012. U.S. Department of the Interior/U.S. Geological
Survey, Washington, D.C. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed September 2014.
Willett, J.C. (2013a) "Crushed Stone," In Minerals Yearbook 2011. U.S. Department of the Interior/U.S. Geological
Survey, Washington, D.C. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed May 2013.
Willett, J.C. (2013b) Personal Communication. Jason Willet, U.S. Geological Survey and ICF International.
September 9, 2013.
Willett, J.C. (2011a) "Crushed Stone," In Minerals Yearbook 2009. U.S. Department of the Interior/U.S. Geological
Survey, Washington, D.C. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed August 2011.
Willett, J.C. (2011b) "Crushed Stone," In Minerals Yearbook 2010. U.S. Department of the Interior/U.S. Geological
Survey, Washington, D.C. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed September 2012.
Willett, J.C. (2010) "Crushed Stone," In Minerals Yearbook 2008. U.S. Department of the Interior/U.S. Geological
Survey, Washington, D.C. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed August 2010.
Willett, J.C. (2009) "Crushed Stone," In Minerals Yearbook 2007. U.S. Department of the Interior/U.S. Geological
Survey, Washington, D.C. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed August 2009.
Willett, J.C. (2007a) "Crushed Stone," In Minerals Yearbook 2005. U.S. Department of the Interior/U.S. Geological
Survey, Washington, D.C. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed August 2007.
Willett, J.C. (2007b) "Crushed Stone," In Minerals Yearbook 2006. U.S. Department of the Interior/U.S. Geological
Survey, Washington, D.C. Available online at:
http://minerals.usgs.gov/minerals/pubs/commodity/stone crushed/index.html#mis. Accessed August 2008.
Urea Fertilization
AAPFCO (2008 through 2021) Commercial Fertilizers. Association of American Plant Food Control Officials.
University of Missouri. Columbia, MO.
AAPFCO (1995 through 2000a, 2002 through 2007) Commercial Fertilizers. Association of American Plant Food
Control Officials. University of Kentucky. Lexington, KY.
AAPFCO (2000b) 1999-2000 Commercial Fertilizers Data, ASCII files. Available from David Terry, Secretary, AAPFCO.
EPA (2000) Preliminary Data Summary: Airport Deicing Operations (Revised). EPA-821-R-00-016. August 2000.
60 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Itle, C. (2009) Email correspondence. Cortney Itle, ERG and Tom Wirth, U.S. Environmental Protection Agency on
the amount of urea used in aircraft deicing. January 7, 2009.
TVA (1991 through 1994) Commercial Fertilizers. Tennessee Valley Authority, Muscle Shoals, AL.
TVA (1992b) Fertilizer Summary Data 1992. Tennessee Valley Authority, Muscle Shoals, AL.
Field Burning of Agricultural Residues
Akintoye, H.A., Agbeyi, E.O., and Olaniyan, A.B. (2005) 'The effects of live mulches on tomato (Lycopersicon
esculentum) yield under tropical conditions." Journal of Sustainable Agriculture 26: 27-37.
Bange, M.P., Milroy, S.P., and Thongbai, P. (2004) "Growth and yield of cotton in response to waterlogging." Field
Crops Research 88:129-142.
Beyaert, R.P. (1996) The effect of cropping and tillage management on the dynamics of soil organic matter. PhD
Thesis. University ofGuelph.
Bouquet, D.J., and Breitenbeck, G.A. (2000) "Nitrogen rate effect on partitioning of nitrogen and dry matter by
cotton." Crop Science 40:1685-1693.
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.Cantens, G.
(2004 through 2005) Personal Communication. Janet Lewis, Assistant to Gaston Cantens, Vice President of
Corporate Relations, Florida Crystals Company and ICF International.
Brouder, S.M., and Cassman, K.G (1990) "Root development of two cotton cultivars in relation to potassium uptake
and plant growth in a vermiculitic soil." Field Crops Res. 23: 187-203.
Costa, L.D., and Gianquinto, G. (2002) "Water stress and watertable depth influence yield, water use efficiency,
and nitrogen recovery in bell pepper: lysimeter studies." Aust. J. Agric. Res. 53: 201-210.
Crafts-Brandner, S.J., Collins, M., Sutton, T.G., and Burton, H.R. (1994) "Effect of leaf maleic hydrazide
concentration on yield and dry matter partitioning in burley tobacco (Nicotiana tabacum L.)." Field Crops Research
37: 121-128.
De Pinheiro Henriques, A.R., and Marcelis, L.F.M. (2000) "Regulation of growth at steady-state nitrogen nutrition in
lettuce (Lactuca sativa L.): Interactive effects of nitrogen and irradiance." Annals of Botany 86: 1073-1080.0.
Diaz-Perez, J.C., Silvoy, J., Phatak,S.C., Ruberson, J., and Morse, R. (2008) Effect of winter cover crops and co-till on
the yield of organically-grown bell pepper (Capsicum annum L). Acta Hort. 767:243-247.
Dua, K.L., and Sharma, V.K. (1976) "Dry matter production and energy contents often varieties of sugarcane at
Muzaffarnagar (Western Uttar Pradesh)." Tropical Ecology 17: 45-49.
Fritschi, F.B., Roberts, B.A., Travis, R.L, Rains, D.W., and Hutmacher, R.B. (2003) "Seasonal nitrogen concentration,
uptake, and partitioning pattern of irrigated Acala and Pima cotton as influenced by nitrogen fertility level." Crop
Science 44:516-527.
Gerik, T.J., K.L Faver, P.M. Thaxton, and K.M. El-Zik. (1996) "Late season water stress in cotton: I. Plant growth,
water use, and yield." Crop Science 36: 914-921.
Gibberd, M.R., McKay, A.G., Calder, T.C., and Turner, N.C. (2003) "Limitations to carrot (Daucus carota L.)
productivity when grown with reduced rates of frequent irrigation on a free-draining, sandy soil ."Australian
Journal of Agricultural Research 54: 499-506.
References and Abbreviations 61
-------
Giglio, L., I. Csiszar, and C.O. Justice (2006) "Global distribution and seasonality of active fires as observed with the
Terra and Aqua Moderate Resolution Imaging Spectroradiometer (MODIS) sensors" J. Geophys. Res. Ill, G02016,
doi:10.1029/2005JG000142.
Halevy, J. (1976) "Growth rate and nutrient uptake of two cotton cultivars grown under irrigation." Agronomy
Journal 68: 701-705.
Halvorson, A.D., Follett, R.F., Bartolo, M.E., and Schweissing, F.C. (2002) "Nitrogen fertilizer use efficiency of
furrow-irrigated onion and corn." Agronomy Journal 94: 442-449.
Heitholt, J.J., Pettigrew, W.T., and Meredith, W.R. (1992) "Light interception and lint yield of narrow-row cotton."
Crop Science 32: 728-733.
Hollifield, C.D., Silvertooth, J.C., and Moser, H. (2000) "Comparison of obsolete and modern cotton cultivars for
irrigated production in Arizona." 2000 Arizona Cotton Report, University of Arizona College of Agriculture,
http://ag.arizona.edu/pubs/crops/azll70/.
Hopkinson, J.M. (1967) "Effects of night temperature on the growth of Nicotiana tabacum." Australian Journal of
Experimental Agriculture and Animal Husbandry 7: 78-82.
Huett, D.O., and Dettman, E.B. (1991) Effect of nitrogen on growth, quality and nutrient uptake of cabbages grown
in sand culture. Australian Journal of Experimental Agriculture 29: 875-81.
Huett, D.O., and Dettman, B. (1989) "Nitrogen response surface models of zucchini squash, head lettuce and
potato." Plant and Soil 134: 243-254.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
IPCC/UNEP/OECD/IEA (1997) Revised 1996 IPCC Guidelines for National Greenhouse Gas Inventories.
Intergovernmental Panel on Climate Change, United Nations Environment Programme, Organization for Economic
Co-Operation and Development, International Energy Agency, Paris, France.
Jacobs, J.L, Ward, G.N., and Kearney, G. (2004) "Effects of irrigation strategies and nitrogen fertilizer on turnip dry
matter yield, water use efficiency, nutritive characteristics and mineral content in western Victoria." Australian
Journal of Experimental Agriculture 44:13-26.
Jacobs, J.L, Ward, G.N., McDowell, A.M., and Kearney, G. (2002) "Effect of seedbed cultivation techniques, variety,
soil type and sowing time, on brassica dry matter yields, water use efficiency and crop nutritive characteristics in
western Victoria." Australian Journal of Experimental Agriculture 42: 945-952.
Jacobs, J.L, Ward, G.N., McDowell, A.M., and Kearney, G.A. (2001) "A survey on the effect of establishment
techniques, crop management, moisture availability and soil type on turnip dry matter yields and nutritive
characteristics in western Victoria." Australian Journal of Experimental Agriculture 41: 743-751.
Kage, H., Alt, C., and Stutzel, H. (2003) "Aspects of nitrogen use efficiency of cauliflower II. Productivity and
nitrogen partitioning as influenced by N supply." Journal of Agricultural Science 141: 17-29.
Kumar, A., Singh, D.P., and Singh, P. (1994) "Influence of water stress on photosynthesis, transpiration, water-use
efficiency and yield of Brassica juncea L" Field Crops Research 37: 95-101.
LANDFIRE (2008) Existing Vegetation Type Layer, LANDFIRE 1.1.0, U.S. Department of the Interior, Geological
Survey. Accessed 28 October 2010 at http://landfire.cr.usgs.gov/viewer/.
MacLeod, LB., Gupta, U.C., and Cutcliffe, J.A. (1971) "Effect of N, P, and K on root yield and nutrient levels in the
leaves and roots of rutabagas grown in a greenhouse." Plant and Soil 35: 281-288.
Mahrani, A., and Aharonov, B. (1964) "Rate of nitrogen absorption and dry matter production by upland cotton
grown under irrigation." Israel J. Agric. Res. 14: 3-9.
62 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Marcussi, F.F.N., Boas, R.L.V., de Godoy, L.J.G., and Goto, R. (2004) "Macronutrient accumulation and partitioning
in fertigated sweet pepper plants." 5c/. Agric. (Piracicaba, Braz.) 61: 62-68.
McCarty, J.L. (2011) "Remote Sensing-Based Estimates of Annual and Seasonal Emissions from Crop Residue
Burning in the Contiguous United States." Journal of the Air & Waste Management Association, 61:1, 22-34, DOI:
10.3155/1047-3289.61.1.22.
McCarty, J.L. (2010) Agricultural Residue Burning in the Contiguous United States by Crop Type and State.
Geographic Information Systems (GIS) Data provided to the EPA Climate Change Division by George Pouliot,
Atmospheric Modeling and Analysis Division, EPA. Dr. McCarty's research was supported by the NRI Air Quality
Program of the Cooperative State Research, Education, and Extension Service, USDA, under Agreement No.
20063511216669 and the NASA Earth System Science Fellowship.
McCarty, J.L. (2009) Seasonal and Interannual Variability of Emissions from Crop Residue Burning in the Contiguous
United States. Dissertation. University of Maryland, College Park.
McPharlin, I.R., Aylmore, P.M., and Jeffery, R.C. (1992) "Response of carrots (Daucus carota L.) to applied
phosphorus and phosphorus leaching on a Karrakatta sand, under two irrigation regimes." Australian Journal of
Experimental Agriculture 32:225-232.
Mondino, M.H., Peterlin, O.A., and Garay, F. (2004) "Response of late-planted cotton to the application of growth
regulator (chlorocholine chloride, CYCOCEL75)." Expl Agric. 40: 381-387.
Moustakas, N.K., and Ntzanis, H. (2005) "Dry matter accumulation and nutrient uptake in flue-cured tobacco
(Nicotiana tabacum L.)." Field Crops Research 94:1-13.
Peach, L., Benjamin, L.R., and Mead, A. (2000) "Effects on the growth of carrots (Daucus carota L), cabbage
(Brassica oleracea var. capitata L.) and onion (Allium cepa L.) of restricting the ability of the plants to intercept
resources." Journal of Experimental Botany 51: 605-615.
Pettigrew, W.T., and Meredith, W.R., Jr. (1997) "Dry matter production, nutrient uptake, and growth of cotton as
affected by potassium fertilization." / Plant Nutr. 20:531-548.
Pettigrew, W.T., Meredith, W.R., Jr., and Young, LD. (2005) "Potassium fertilization effects on cotton lint yield,
yield components, and reniform nematode populations." Agronomy Journal 97: 1245-1251.
PRISM Climate Group (2015) PRISM Climate Data. Oregon State University. July 24, 2015. Available online at:
http://prism.oregonstate.edu.
Reid, J.B., and English, J.M. (2000) "Potential yield in carrots (Daucus carota L): Theory, test, and an application."
Annals of Botany 85: 593-605.
Sadras, V.O., and Wilson, L.J. (1997) "Growth analysis of cotton crops infested with spider mites: II. Partitioning of
dry matter." Crop Science 37: 492-497.
Scholberg, J., McNeal, B.L., Jones, J.W., Boote, K.J., Stanley, C.D., and Obreza, T.A. (2000a) "Growth and canopy
characteristics of field-grown tomato." Agronomy Journal 92: 152-159.
Scholberg, J., McNeal, B.L., Boote, K.J., Jones, J.W., Locasio, S.J., and Olson, S.M. (2000b) "Nitrogen stress effects on
growth and nitrogen accumulation by field-grown tomato." Agronomy Journal 92:159-167.
Singels, A. and Bezuidenhout, C.N. (2002) "A new method of simulating dry matter partitioning in the Canegro
sugarcane model." Field Crops Research 78: 151 - 164.
Sitompul, S.M., Hairiah, K., Cadisch, G., and Van Noordwuk, M. (2000) "Dynamics of density fractions of macro-
organic matter after forest conversion to sugarcane and woodlots, accounted for in a modified Century model."
Netherlands Journal of Agricultural Science 48: 61-73.
Stirling, G.R., Blair, B.L, Whittle, P.J.L., and Garside, A.L. (1999) "Lesion nematode (Pratylenchus zeae) is a
component of the yield decline complex of sugarcane." In: Magarey, R.C. (Ed.), Proceedings of the First
Australasian Soilborne Disease Symposium. Bureau of Sugar Experiment Stations, Brisbane, pp. 15-16.
References and Abbreviations 63
-------
Tan, D.K.Y., Wearing, A.H., Rickert, K.G., and Birch, C.J. (1999) "Broccoli yield and quality can be determined by
cultivar and temperature but not photoperiod in south-east Queensland." Australian Journal of Experimental
Agriculture 39: 901-909.
Tadesse, T., Nichols, M.A., and Fisher, K.J. (1999) Nutrient conductivity effects on sweet pepper plants grown using
a nutrient film technique. 1. Yield and fruit quality. New Zealand Journal of Crop and Horticultural Science, 27: 229-
237.
Torbert, H.A., and Reeves, D.W. (1994) "Fertilizer nitrogen requirements for cotton production as affected by
tillage and traffic." Soil Sci. Soc. Am. J. 58:1416-1423.
USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory, Natural Resources Conservation Service,
Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd 1422028.pdf.
USDA (2019) Quick Stats: U.S. & All States Data; Crops; Production and Area Harvested; 1990 - 2018. National
Agricultural Statistics Service, U.S. Department of Agriculture. Washington, D.C. U.S. Department of Agriculture,
National Agricultural Statistics Service. Washington, D.C., Available online at: http://quickstats.nass.usda.gov/.
Valantin, M., Gary, C., Vaissiere, B.E., and Frossard, J.S. (1999) "Effect of fruit load on partitioning of dry matter and
energy in cantaloupe (Cucumis melo L)." Annals of Botany 84:173-181.
Wallach, D., Marani, A., and Kletter, E. (1978) 'The relation of cotton crop growth and development to final yield."
Field Crops Research 1: 283-294.
Wells, R., and Meredith, W.R., Jr. (1984) "Comparative growth of obsolete and modern cultivars. I. Vegetative dry
matter partitioning." Crop Science 24: 858-872.4.
Wiedenfels, R.P. (2000) "Effects of irrigation and N fertilizer application on sugarcane yield and quality." Field Crops
Research 43:101-108.
Land Use, Land-Use Change, and Forestry
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
UNFCCC (2014) Report of the Conference of the Parties on its nineteenth session, held in Warsaw from 11 to 23
November 2013. United Nations Framework Convention on Climate Change, Warsaw. (FCCC/CP/2013/10/Add.3).
January 31, 2014. Available online at: http://unfccc.int/resource/docs/2013/copl9/eng/10a03.pdf.
Representation of the U.S. Land Base
Alaska Department of Natural Resources (2006) Alaska Infrastructure 1:63,360. Available online at:
http://dnr.alaska.gov/SpatialUtility/SUC?cmd=extract&layerid=75.
Alaska Interagency Fire Management Council (1998) Alaska Interagency Wildland Fire Management Plan. Available
online at: http://agdc.usgs.gov/data/blm/fire/index.html.
Alaska Oil and Gas Conservation Commission (2009) Oil and Gas Information System. Available online at:
http://doa.alaska.gov/ogc/publicdb.html.
EIA (2011) Coal Production and Preparation Report Shapefile. Available online at: http://www.eia.gov/state/notes-
sources.cfmtfmaps.
ESRI (2008) ESRI Data & Maps. Redlands, CA: Environmental Systems Research Institute. [CD-ROM],
64 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L, Barnes, C., Herold, N., and J. Wickham. (2011) Completion of
the 2006 National Land Cover Database for the Conterminous United States, PE&RS, Vol. 77(9):858-864.
Homer, C., J. Dewitz, J. Fry, M. Coan, N. Hossain, C. Larson, N. Herold, A. McKerrow, J.N. VanDriel and J. Wickham.
(2007) Completion of the 2001 National Land Cover Database for the Conterminous United States,
Photogrammetric Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-
Representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, v.
81, no. 5, p. 345-354.
IPCC (2014) 2013 Supplement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands.
Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds.). Published: IPCC,
Switzerland.
IPCC (2010) Revisiting the use of managed land as a proxy for estimating national anthropogenic emissions and
removals. [Eggleston HS, Srivastava N, Tanabe K, Baasansuren J, (eds.)]. Institute for Global Environmental Studies,
Intergovernmental Panel on Climate Change, Hayama, Kanagawa, Japan.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Jin, S., L. Yang, P. Danielson, C. Homer, J. Fry, and G. Xian. (2013) A comprehensive change detection method for
updating the National Land Cover Database to circa 2011. Remote Sensing of Environment, 132: 159-175.
NOAA Coastal Change Analysis Program (C-CAP) Regional Land Cover Database. Data collected 1995-present
Charleston, SC: National Oceanic and Atmospheric Administration (NOAA) Coastal Services Center. Data accessed
at: www.csc.noaa.gov/landcover.
Nusser, S.M. and J.J. Goebel (1997) "The national resources inventory: a long-term multi-resource monitoring
programme." Environmental and Ecological Statistics 4:181-204.
Ogle, S.M., G. Domke, W.A. Zurz, M.T. Rocha, T. Huffman, A. Swan, J.E. Smith, C. Woodall, T. Krug (2018)
Delineating managed land for reporting greenhouse gas emissions and removals to the United Nations Framework
Convention on Climate Change. Carbon Balance and Management 13:9.
Smith, W.B., P.D. Miles, C.H. Perry, and S.A. Pugh (2009) Forest Resources of the United States, 2007. Gen. Tech.
Rep. WO-78. U.S. Department of Agriculture Forest Service, Washington, D.C.
U.S. Census Bureau (2010) Topological^ Integrated Geographic Encoding and Referencing (TIGER) system
shapefiles. U.S. Census Bureau, Washington, D.C. Available online at: http://www.census.gov/geo/www/tiger.
U.S. Department of Agriculture (2015) County Data - Livestock, 1990-2014. U.S. Department of Agriculture,
National Agriculture Statistics Service, Washington, D.C.
U.S. Department of Agriculture, Forest Service. (2012) Timber Product Output (TPO) Reports. Knoxville, TN: U.S.
Department of Agriculture Forest Service, Southern Research Station.
U.S. Department of Interior (2005) Federal Lands of the United States. National Atlas of the United States, U.S.
Department of the Interior, Washington D.C. Available online at:
http://nationalatlas.gov/atlasftp.html?openChapters=chpbound#chpbound.
United States Geological Survey (USGS), Gap Analysis Program (2012) Protected Areas Database of the United
States (PADUS), version 1.3 Combined Feature Class. November 2012.
USGS (2012) Alaska Resource Data File. Available online at: http://ardf.wr.usgs.gov/.
USGS (2005) Active Mines and Mineral Processing Plants in the United States in 2003. U.S. Geological Survey,
Reston, VA.
References and Abbreviations 65
-------
Yang, L, Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J., Fry, J., Funk, M.,
Granneman, B., Liknes, G. C., Rigge, M. & Xian, G. (2018) A new generation of the United States National Land
Cover Database: Requirements, research priorities, design, and implementation strategies. ISPRS Journal of
Photogrammetry and Remote Sensing 146:108-123.
Forest Land Remaining Forest Land: Changes in Forest Carbon
Stocks
AF&PA (2006a and earlier) Statistical roundup. (Monthly). Washington, D.C. American Forest & Paper Association.
AF&PA (2006b and earlier) Statistics of paper, paperboard and wood pulp. Washington, D.C. American Forest &
Paper Association.
AF&PA (2021) 2020 Statistics - Paper Industry - Annual Summary Data Through 2020. Washington, D.C.: American
Forest and Paper Association, 54 p.
Amichev, B.Y. and J.M. Galbraith (2004) "A Revised Methodology for Estimation of Forest Soil Carbon from Spatial
Soils and Forest Inventory Data Sets." Environmental Management 33(Suppl. 1):S74-S86.
Bechtold, W.A.; Patterson, P.L (2005) The enhanced forest inventory and analysis program—national sampling
design and estimation procedures. Gen. Tech. Rep. SRS-80. Asheville, NC: U.S. Department of Agriculture Forest
Service, Southern Research Station. 85 p.
Birdsey, R. (1996) "Carbon Storage for Major Forest Types and Regions in the Conterminous United States." In R.N.
Sampson and D. Hair, (eds.). Forest and Global Change, Volume 2: Forest Management Opportunities for
Mitigating Carbon Emissions. American Forests. Washington, D.C., 1-26 and 261-379 (appendices 262 and 263).
Coulston, J.W., Wear, D.N., and Vose, J.M. (2015) Complex forest dynamics indicate potential for slowing carbon
accumulation in the southeastern United States. Scientific Reports. 5: 8002.
Domke, G.M., J.E. Smith, and C.W. Woodall. (2011) Accounting for density reduction and structural loss in standing
dead trees: Implications for forest biomass and carbon stock estimates in the United States. Carbon Balance and
Management. 6:14.
Domke, G.M., Woodall, C.W., Smith, J.E., Westfall, J.A., McRoberts, R.E. (2012) Consequences of alternative tree-
level biomass estimation procedures on U.S. forest carbon stock estimates. Forest Ecology and Management. 270:
108-116.
Domke, G.M., Woodall, C.W., Walters, B.F., Smith, J.E. (2013) From models to measurements: comparing down
dead wood carbon stock estimates in the U.S. forest inventory. PLoS ONE 8(3): e59949.
Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., and Smith, J.E. (2016) A framework for estimating litter
carbon stocks in forests of the United States. Science of the Total Environment 557-558: 469-478.
Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., Nave, L., Swanston, C. (2017) Toward inventory-based
estimates of soil organic carbon in forests of the United States. Ecological Applications. 27(4), 1223-1235.
EPA (2006) Municipal solid waste in the United States: 2005 Facts and figures. Office of Solid Waste, U.S.
Environmental Protection Agency. Washington, D.C. (5306P) EPA530-R-06-011. Available online at:
http://www.epa.gov/msw/msw99.htm.
FAO (2021) Forest product statistics. Rome, Italy: FAO Forestry Division, fao.org/forestry/statistics/en. Accessed
August 16, 2021.
Frayer, W.E., and G.M. Furnival (1999) "Forest Survey Sampling Designs: A History." Journal of Forestry 97(12): 4-
10.
Freed, R. (2004) Open-dump and Landfill timeline spreadsheet (unpublished). ICF International. Washington, D.C.
66 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Hair, D. (1958) "Historical forestry statistics of the United States." Statistical Bull. 228. U.S. Department of
Agriculture Forest Service, Washington, D.C.
Hair. D. and A.H. Ulrich (1963) The Demand and price situation for forest products - 1963. U.S. Department of
Agriculture Forest Service, Misc Publication No. 953. Washington, D.C.
Harmon, M.E., C.W. Woodall, B. Fasth, J. Sexton, M. Yatkov. (2011) Differences between standing and downed
dead tree wood density reduction factors: A comparison across decay classes and tree species. Res. Paper. NRS-15.
Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 40 p.
Howard, J. L. and Liang, S. (2019) U.S. timber production, trade, consumption, and price statistics 1965 to 2017.
Res. Pap. FPL-RP-701. Madison, Wl: USDA, Forest Service, Forest Products Laboratory.
Howard, J. L. and Jones, K.C. (2016) U.S. timber production, trade, consumption, and price statistics 1965 to 2013.
Res. Pap. FPL-RP-679. Madison, Wl: USDA, Forest Service, Forest Products Laboratory.
Howard, J. L. (2007) U.S. timber production, trade, consumption, and price statistics 1965 to 2005. Res. Pap. FPL-
RP-637. Madison, Wl: USDA, Forest Service, Forest Products Laboratory.
Howard, J. L. (2003) U.S. timber production, trade, consumption, and price statistics 1965 to 2002. Res. Pap. FPL-
RP-615. Madison, Wl: USDA, Forest Service, Forest Products Laboratory. Available online at:
http://www.fpl.fs.fed.us/documnts/fplrp/fplrp615/fplrp615.pdf.
IPCC (2014) 2013 Supplement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands.
[Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M., and Troxler, T.G. (eds.)]. Switzerland.
IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M. Manning, Z. Chen,
M. Marquis, K.B. Averyt, M. Tignor and H.L Miller (eds.)]. Cambridge University Press, Cambridge, United Kingdom
and New York, NY, USA, 996 pp.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Jenkins, J.C., D.C. Chojnacky, L.S. Heath, and R.A. Birdsey (2003) "National-scale biomass estimators for United
States tree species." Forest Science 49(l):12-35.
Jandl, R., Rodeghiero, M., Martinez, C., Cotrufo, M. F., Bampa, F., van Wesemael, B., Harrison, R.B., Guerrini, I.A.,
deB Richter Jr., D., Rustad, L., Lorenz, K., Chabbi, A., Miglietta, F. (2014) Current status, uncertainty and future
needs in soil organic carbon monitoring. Science of the Total Environment, 468, 376-383.
Johnson, K. Domke, G.M., Russell, M.B., Walters, B.F., Horn, J., Peduzzi, A., Birdsey, R., Dolan, K., Huang, W. (2017)
Estimating aboveground live understory vegetation carbon in the United States. Environmental Research Letters.
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.,
Oswalt, S.N., Reeves, M.C. and Wear, D.N. (2020) Defining the United States land base: a technical document
supporting the USDA Forest Service 2020 RPA assessment. Gen. Tech. Rep. NRS-191., 191, pp.1-70.
Ogle, S.M., Woodall, C.W., Swan, A., Smith, J.E., Wirth. T. (In preparation) Determining the Managed Land Base for
Delineating Carbon Sources and Sinks in the United States. Environmental Science and Policy.
O'Neill, K.P., Amacher, M.C., Perry, C.H. (2005) Soils as an indicator of forest health: a guide to the collection,
analysis, and interpretation of soil indicator data in the Forest Inventory and Analysis program. Gen. Tech. Rep. NC-
258. St. Paul, MN: U.S. Department of Agriculture, Forest Service, North Central Research Station. 53 p.
Oswalt, S.N., Smith, W.B., Miles, P.D. and Pugh, S.A. (2019) Forest resources of the United States, 2017: A technical
document supporting the Forest Service 2020 RPA Assessment. Gen. Tech. Rep. WO-97. Washington, DC: U.S.
Department of Agriculture, Forest Service, Washington Office., 97.
References and Abbreviations 67
-------
Perry, C.H., C.W. Woodall, and M. Schoeneberger (2005) Inventorying trees in agricultural landscapes: towards an
accounting of "working trees". In: "Moving Agroforestry into the Mainstream." Proc. 9th N. Am. Agroforestry
Conf., Brooks, K.N. and Folliott, P.F. (eds.). 12-15 June 2005, Rochester, MN [CD-ROM], Dept. of Forest Resources,
Univ. Minnesota, St. Paul, MN, 12 p. Available online at: http://cinram.umn.edu/afta2005/. (verified 23 Sept 2006).
Russell, M.B.; D'Amato, A.W.; Schulz, B.K.; Woodall, C.W.; Domke, G.M.; Bradford, J.B. (2014) Quantifying
understory vegetation in the U.S. Lake States: a proposed framework to inform regional forest carbon stocks.
Forestry. 87: 629-638.
Russell, M.B.; Domke, G.M.; Woodall, C.W.; D'Amato, A.W. (2015) Comparisons of allometric and climate-derived
estimates of tree coarse root carbon in forests of the United States. Carbon Balance and Management. 10: 20.
Skog, K.E. (2008) Sequestration of carbon in harvested wood products for the United States. Forest Products
Journal 58:56-72.
Smith, J.E.; Heath, L.S.; Skog, K.E.; Birdsey, R.A. (2006) Methods for calculating forest ecosystem and harvested
carbon with standard estimates for forest types of the United States. Gen. Tech. Rep. NE-343. Newtown Square,
PA: U.S. Department of Agriculture, Forest Service, Northeastern Research Station. 216 p.
Smith, W. B., P. D. Miles, C. H. Perry, and S. A. Pugh (2009) Forest Resources of the United States, 2007. General
Technical Report WO-78, U.S. Department of Agriculture Forest Service, Washington Office.
Smith, J.E., L.S. Heath, and M.C. Nichols (2010) U.S. Forest Carbon Calculation Tool User's Guide: Forestland Carbon
Stocks and Net Annual Stock Change. General Technical Report NRS-13 revised, U.S. Department of Agriculture
Forest Service, Northern Research Station, 34 p.
Smith, J.E., Domke, G.M. and Woodall, C.W. (2022) Predicting downed woody material carbon stocks in forests of
the conterminous United States. Science of the Total Environment, 803, p.150061.
Steer, Henry B. (1948) Lumber production in the United States. Misc. Pub. 669, U.S. Department of Agriculture
Forest Service. Washington, D.C.
Ulrich, Alice (1985) U.S. Timber Production, Trade, Consumption, and Price Statistics 1950-1985. Misc. Pub. 1453,
U.S. Department of Agriculture Forest Service. Washington, D.C.
Ulrich, A.H. (1989) U.S. Timber Production, Trade, Consumption, and Price Statistics, 1950-1987. USDA
Miscellaneous Publication No. 1471, U.S. Department of Agriculture Forest Service. Washington, D.C., 77.
United Nations Framework Convention on Climate Change (2013) Report on the individual review of the inventory
submission of the United States of America submitted in 2012. FCCC/ARR/2012/USA. 42 p.
USDA Forest Service (2022a) Forest Inventory and Analysis National Program: Program Features. U.S. Department
of Agriculture Forest Service. Washington, D.C. Available online at: http://fia.fs.fed.us/program-features/.
Accessed 30 March 2022.
USDA Forest Service. (2022b) Forest Inventory and Analysis National Program: FIA Data Mart. U.S. Department of
Agriculture Forest Service. Washington, D.C. Available online at:
https://apps.fs.usda.gov/fia/datamart/datamart.html. Accessed on 30 March 2022.
USDA Forest Service. (2022c) Forest Inventory and Analysis National Program, FIA library: Field Guides, Methods
and Procedures. U.S. Department of Agriculture Forest Service. Washington, D.C. Available online at:
http://www.fia.fs.fed.us/library/field-guides-methods-proc/. Accessed on 30 March 2022.
USDA Forest Service (2022d) Forest Inventory and Analysis National Program, FIA library: Database
Documentation. U.S. Department of Agriculture, Forest Service, Washington Office. Available online at:
http://fia.fs.fed.us/library/database-documentation/. Accessed on 30 March 2022.
U.S. Census Bureau (1976) Historical Statistics of the United States, Colonial Times to 1970, Vol. 1. Washington,
D.C.
68 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Wear, D.N., Coulston, J.W. (2015) From sink to source: Regional variation in U.S. forest carbon futures. Scientific
Reports. 5:16518.
Westfall, J.A., Woodall, C.W., Hatfield, M.A. (2013) A statistical power analysis of woody carbon flux from forest
inventory data. Climatic Change. 118: 919-931.
Woodall, C.W., Coulston, J.W., Domke, G.M., Walters, B.F., Wear, D.N., Smith, J.E., Anderson, H.-E., Clough, B.J.,
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,
B.T. (2015a) The U.S. Forest Carbon Accounting Framework: Stocks and Stock change 1990-2016. Gen. Tech. Rep.
NRS-154. Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 49 pp.
Woodall, C.W., LS. Heath, G.M. Domke, and M.C. Nichols (2011a) Methods and equations for estimating
aboveground volume, biomass, and carbon for trees in the U.S. forest inventory, 2010. Gen. Tech. Rep. NRS-88.
Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 30 p.
Woodall, C.W., Amacher, M.C., Bechtold, W.A., Coulston, J.W., Jovan, S., Perry, C.H., Randolph, K.C., Schulz, B.K.,
Smith, G.C., Tkacz, B., Will-Wolf, S. (2011b) "Status and future of the forest health indicators program of the United
States." Environmental Monitoring and Assessment. 177: 419-436.
Woodall, C.W., and V.J. Monleon (2008) Sampling protocol, estimation, and analysis procedures for the down
woody materials indicator of the FIA program. Gen. Tech. Rep. NRS-22. Newtown Square, PA: U.S. Department of
Agriculture, Forest Service, Northern Research Station. 68 p.
Woodall, C.W., Walters, B.F., Oswalt, S.N., Domke, G.M., Toney, C., Gray, A.N. (2013) Biomass and carbon
attributes of downed woody materials in forests of the United States. Forest Ecology and Management 305: 48-59.
Woodall, C.W., Walters, B.F., Coulston, J.W., D'Amato, A.W., Domke, G.M., Russell, M.B., Sowers, P.A. (2015b)
Monitoring network confirms land use change is a substantial component of the forest carbon sink in the eastern
United States. Scientific Reports. 5:17028.
Zhu, Zhiliang, and McGuire, A.D., eds. (2016) Baseline and projected future carbon storage and greenhouse-gas
fluxes in ecosystems of Alaska: U.S. Geological Survey Professional Paper 1826,196 p., Available online at:
http://dx.doi.org/10.3133/ppl826.
Forest Land Remaining Forest Land: Non-C02 Emissions from
Forest Fires
Eidenshink, J., Schwind, B., Brewer, K., Zhu, Z.L, Quayle, B. and Howard, S. (2007) A project for monitoring trends
in burn severity. Fire ecology, 3(1), pp.3-21.
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.
Hyer, R. Keane, D. McKenzie, S.G. McNulty, B.E Law, R. Ottmar, D.R. Perez-Salicrup, J. Randerson, K.M. Robertson,
and M. Turetsky (2011) "Model comparisons for estimating carbon emissions from North American wildland fire."
Journal of Geophysical Research 116. 10.1029/2010JG001469
French, N.H.F., D. McKenzie, T. Erickson, B. Koziol, M. Billmire, K.A. Endsley, N.K.Y. Scheinerman, L Jenkins, M.E.
Miller, R. Ottmar, and S. Prichard (2014) "Modeling regional-scale fire emissions with the Wildland Fire Emissions
Information System." Earth Interactions 18, no. 16
Giglio, L., Boschetti, L., Roy, D. P., Humber, M. L, and Justice, C. O. (2018) The Collection 6 MODIS burned area
mapping algorithm and product. Remote Sensing of Environment, 217, 72-85.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Larkin, N. K., S. Raffuse, and T. T. Strand (2014) Wildland fire emissions, carbon, and climate: U.S. emissions
inventories. For. Ecol. Manage. 317:61-69. doi:10.1016/j.foreco.2013.09.012.
References and Abbreviations 69
-------
MTBS Data Access: Fire Level Geospatial Data (2021, April - revised) MTBS Project (USDA Forest Service/U.S.
Geological Survey). Available online at: http://mtbs.gov/direct-download. Accessed on 21 April 2021.
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.
(2018) Delineating managed land for reporting national greenhouse gas emissions and removals to the United
Nations framework convention on climate change. Carbon Balance and Management 13:9.
Forest Land Remaining Forest Land: N20 Emissions from Soils
Albaugh, T.J., Allen, H.L., Fox, T.R. (2007) Historical Patterns of Forest Fertilization in the Southeastern United
States from 1969 to 2004. Southern Journal of Applied Forestry, 31,129-137(9).
Binkley, D. (2004) Email correspondence regarding the 95 percent confidence interval for area estimates of
southern pine plantations receiving N fertilizer (±20%) and the rate applied for areas receiving N fertilizer (100 to
200 pounds/acre). Dan Binkley, Department of Forest, Rangeland, and Watershed Stewardship, Colorado State
University and Stephen Del Grosso, Natural Resource Ecology Laboratory, Colorado State University. September
19, 2004.
Binkley, D., R. Carter, and H.L Allen (1995) Nitrogen Fertilization Practices in Forestry. In: Nitrogen Fertilization in
the Environment, P.E. Bacon (ed.), Marcel Decker, Inc., New York.
Briggs, D. (2007) Management Practices on Pacific Northwest West-Side Industrial Forest Lands, 1991-2005: With
Projections to 2010. Stand Management Cooperative, SMC Working Paper Number 6, College of Forest Resources,
University of Washington, Seattle.
Fox, T.R., H. L. Allen, T.J. Albaugh, R. Rubilar, and C.A. Carlson (2007) Tree Nutrition and Forest Fertilization of Pine
Plantations in the Southern United States. Southern Journal of Applied Forestry, 31, 5-11.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
USDA Forest Service (2001) U.S. Forest Facts and Historical Trends. FS-696. U.S. Department of Agriculture Forest
Service, Washington, D.C. Available online at: http://www.fia.fs.fed.us/library/ForestFactslVletric.pdf.
Forest Land Remaining Forest Land: Drained Organic Soils
IPCC (2014) 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands,
Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds.). Published: IPCC,
Switzerland.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change, H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
STATSG02 (2016) Soil Survey Staff, Natural Resources Conservation Service, United States Department of
Agriculture. U.S. General Soil Map (STATSG02). Available online at https://sdmdataaccess.sc.egov.usda.gov.
Accessed 10 November 2016.
USDA Forest Service (2022b) Forest Inventory and Analysis National Program: FIA Data Mart. U.S. Department of
Agriculture Forest Service. Washington, DC; 2015. Available online at
https://apps.fs.usda.gov/fia/datamart/datamart.html. Accessed 30 March 2022.
70 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Land Converted to Forest Land
Birdsey, R. (1996) "Carbon Storage for Major Forest Types and Regions in the Conterminous United States." In R.N.
Sampson and D. Hair, (eds.). Forest and Global Change, Volume 2: Forest Management Opportunities for
Mitigating Carbon Emissions. American Forests. Washington, D.C., 1-26 and 261-379 (appendices 262 and 263).
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
Domke, G.M., J.E. Smith, and C.W. Woodall. (2011) Accounting for density reduction and structural loss in standing
dead trees: Implications for forest biomass and carbon stock estimates in the United States. Carbon Balance and
Management. 6:14.Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., Nave, L, Swanston, C. (2017) Toward
inventory-based estimates of soil organic carbon in forests of the United States. Ecological Applications. 27(4),
1223-1235.
Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., and Smith, J.E. (2016) A framework for estimating litter
carbon stocks in forests of the United States. Science of the Total Environment 557-558: 469-478.
Domke, G.M., Woodall, C.W., Walters, B.F., Smith, J.E. (2013) From models to measurements: comparing down
dead wood carbon stock estimates in the U.S. forest inventory. PLoS ONE 8(3): e59949.
Harmon, M.E., C.W. Woodall, B. Fasth, J. Sexton, M. Yatkov. (2011) Differences between standing and downed
dead tree wood density reduction factors: A comparison across decay classes and tree species. Res. Paper. NRS-15.
Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 40 p.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Jenkins, J.C., D.C. Chojnacky, L.S. Heath, and R.A. Birdsey (2003) "National-scale biomass estimators for United
States tree species." Forest Science 49(l):12-35.0gle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003)
"Uncertainty in estimating land use and management impacts on soil organic carbon storage for U.S.
agroecosystems between 1982 and 1997." Global Change Biology 9:1521-1542.
Ogle, S.M., F.J. Breidt, and K. Paustian. (2006) "Bias and variance in model results due to spatial scaling of
measurements for parameterization in regional assessments." Global Change Biology 12:516-523.
Smith, J.E.; Heath, L.S.; Skog, K.E.; Birdsey, R.A. (2006) Methods for calculating forest ecosystem and harvested
carbon with standard estimates for forest types of the United States. Gen. Tech. Rep. NE-343. Newtown Square,
PA: U.S. Department of Agriculture, Forest Service, Northeastern Research Station. 216 p.
USDA Forest Service (2022b) Forest Inventory and Analysis National Program: FIA Data Mart. U.S. Department of
Agriculture Forest Service. Washington, D.C. Available online at: https://apps.fs.usda.gov/fia/datamart/. Accessed
on 30 March 2022.
USDA Forest Service (2022c) Forest Inventory and Analysis National Program, FIA library: Field Guides, Methods
and Procedures. U.S. Department of Agriculture Forest Service. Washington, D.C. Available online at:
http://www.fia.fs.fed.us/library/field-guides-methods-proc/. Accessed on 30 March 2022.
USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory, Natural Resources Conservation Service,
Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd 1422028.pdf.
USDA-NRCS (1997) "National Soil Survey Laboratory Characterization Data," Digital Data, Natural Resources
Conservation Service, U.S. Department of Agriculture. Lincoln, NE.
Woodall, C.W., L.S. Heath, G.M. Domke, and M.C. Nichols (2011a) Methods and equations for estimating
aboveground volume, biomass, and carbon for trees in the U.S. forest inventory, 2010. Gen. Tech. Rep. NRS-88.
Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 30 p.
References and Abbreviations 71
-------
Woodall, C.W., and V.J. Monleon (2008) Sampling protocol, estimation, and analysis procedures for the down
woody materials indicator of the FIA program. Gen. Tech. Rep. NRS-22. Newtown Square, PA: U.S. Department of
Agriculture, Forest Service, Northern Research Station. 68 p.
Woodall, C.W., Walters, B.F., Coulston, J.W., D'Amato, A.W., Domke, G.M., Russell, M.B., Sowers, P.A. (2015b)
Monitoring network confirms land use change is a substantial component of the forest carbon sink in the eastern
United States. Scientific Reports. 5:17028.
Woodall, C.W., Walters, B.F., Oswalt, S.N., Domke, G.M., Toney, C., Gray, A.N. (2013) Biomass and carbon
attributes of downed woody materials in forests of the United States. Forest Ecology and Management 305: 48-59.
Yang, L, Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J., Fry, J., Funk, M.,
Granneman, B., Liknes, G. C., Rigge, M. & Xian, G. (2018) A new generation of the United States National Land
Cover Database: Requirements, research priorities, design, and implementation strategies. ISPRS Journal of
Photogrammetry and Remote Sensing 146:108-123.
Cropland Remaining Cropland
Armentano, T. V., and E.S. Menges (1986) Patterns of change in the carbon balance of organic soil-wetlands of the
temperate zone. Journal of Ecology 74: 755-774.
Brady, N.C. and R.R. Weil (1999) The Nature and Properties of Soils. Prentice Hall. Upper Saddle River, NJ, 881.
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
Cheng, B., and D.M. Titterington (1994) "Neural networks: A review from a statistical perspective." Statistical
science 9: 2-30.
Claassen, R., M. Bowman, J. McFadden, D. Smith, and S. Wallander (2018) Tillage intensity and conservation
cropping in the United States, EIB 197. United States Department of Agriculture, Economic Research Service,
Washington, D.C.
Conant, R. T., K. Paustian, and E.T. Elliott (2001) "Grassland management and conversion into grassland: effects on
soil carbon." Ecological Applications 11: 343-355.
CTIC (2004) National Crop Residue Management Survey: 1989-2004. Conservation Technology Information Center,
Purdue University, Available online at: http://www.ctic.purdue.edu/CRM/.
Daly, C., R.P. Neilson, and D.L Phillips (1994) "A Statistical-Topographic Model for Mapping Climatological
Precipitation Over Mountainous Terrain." Journal of Applied Meteorology 33:140-158.
Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel (2001) "Simulated
Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the DAYCENT Model." In Modeling Carbon
and Nitrogen Dynamics for Soil Management, Schaffer, M., L. Ma, S. Hansen, (eds.). CRC Press, Boca Raton, Florida,
pp. 303-332.
Del Grosso, S.J., S.M. Ogle, W.J. Parton (2011) Soil organic matter cycling and greenhouse gas accounting
methodologies, Chapter 1, pp 3-13 DOI: 10.1021/bk-2011-1072.ch001. In: Understanding Greenhouse Gas
Emissions from Agricultural Management, L. Guo, A. Gunasekara, L. McConnell (eds.). American Chemical Society,
Washington, D.C.
Edmonds, L, R. L. Kellogg, B. Kintzer, L. Knight, C. Lander, J. Lemunyon, D. Meyer, D.C. Moffitt, and J. Schaefer
(2003) "Costs associated with development and implementation of Comprehensive Nutrient Management Plans."
Part I—Nutrient management, land treatment, manure and wastewater handling and storage, and recordkeeping.
Natural Resources Conservation Service, U.S. Department of Agriculture. Available online at:
http://www.nrcs.usda.gov/technical/land/pubs/cnmpl.html.
72 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Euliss, N., and R. Gleason (2002) Personal communication regarding wetland restoration factor estimates and
restoration activity data. Ned Euliss and Robert Gleason of the U.S. Geological Survey, Jamestown, ND, to Stephen
Ogle of the National Resource Ecology Laboratory, Fort Collins, CO. August 2002.
Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and Wickham, J. (2011) Completion of the 2006
National Land Cover Database for the Conterminous United States, PE&RS, Vol. 77(9):858-864.
Griscom, B. W., Adams, J., Ellis, P. W., Houghton, R. A., Lomax, G., Miteva, D. A., Schlesinger, W. H., Shoch, D., Siikamaki, J.
V., Smith, P., Woodbury, P., Zganjar, C., Blackman, A., Campari, J., Conant, R. T., Delgado, C., Elias, P., Gopalakrishna, T.,
Hamsik, M. R., Herrero, M., Kiesecker, J., Landis, E., Laestadius, L., Leavitt, S. M., Minnemeyer, S., Polasky, S., Potapov, P.,
Putz, F. E., Sanderman, J., Silvius, M., Wollenberg, E. & Fargione, J. (2017) "Natural climate solutions." Proceedings of the
National Academy of Sciences of the United States of America 114(44): 11645-11650.
Hijmans, R.J., S.E. Cameron, J.L. Parra, P.G. Jones and A. Jarvis (2005) Very high resolution interpolated climate
surfaces for global land areas. International Journal of Climatology 25: 1965-1978.
Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A., VanDriel, J.N., and
Wickham, J. (2007) Completion of the 2001 National Land Cover Database for the Conterminous United States.
Photogrammetric Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-
Representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, v.
81, no. 5, p. 345-354.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
IPCC (2003) Good Practice Guidance for Land Use, Land-Use Change, and Forestry. The Intergovernmental Panel on
Climate Change, National Greenhouse Gas Inventories Programme, J. Penman, et al., eds. August 13, 2004.
Available online at: http://www.ipcc-nggip.iges.or.ip/public/gpglulucf/gpglulucf.htm.
Lai, R., Kimble, J. M., Follett, R. F. & Cole, C. V. (1998) The potential of U.S. cropland to sequester carbon and
mitigate the greenhouse effect. Chelsea, Ml: Sleeping Bear Press, Inc.
Little, R. (1988) "Missing-data adjustments in large surveys." Journal of Business and Economic Statistics 6: 287-
296.
McGill, W.B., and C.V. Cole (1981) Comparative aspects of cycling of organic C, N, S and P through soil organic
matter. Geoderma 26:267-286.
Metherell, A.K., L.A. Harding, C.V. Cole, and W.J. Parton (1993) "CENTURY Soil Organic Matter Model
Environment." Agroecosystem version 4.0. Technical documentation, GPSRTech. Report No. 4, USDA/ARS, Ft.
Collins, CO.
Mesinger, F., G. DiMego, E. Kalnay, K. Mitchell, P. C. Shafran, W. Ebisuzaki, D. Jovic, J. Woollen, E. Rogers, E. H.
Berbery, M. B. Ek, Y. Fan, R. Grumbine, W. Higgins, H. Li, Y. Lin, G. Manikin, D. Parrish, and W. Shi (2006) North
American regional reanalysis. Bulletin of the American Meteorological Society 87:343-360.
NRCS (1999) Soil Taxonomy: A basic system of soil classification for making and interpreting soil surveys, 2nd
Edition. Agricultural Handbook Number 436, Natural Resources Conservation Service, U.S. Department of
Agriculture, Washington, D.C.
NRCS (1997) "National Soil Survey Laboratory Characterization Data," Digital Data, Natural Resources Conservation
Service, U.S. Department of Agriculture. Lincoln, NE.
NRCS (1981) Land Resource Regions and Major Land Resource Areas of the United States, USDA Agriculture
Handbook 296, United States Department of Agriculture, Natural Resources Conservation Service, National Soil
Survey Cente., Lincoln, NE, pp. 156.
References and Abbreviations 73
-------
Ogle, S. M., Alsaker, C., Baldock, J., Bernoux, M., Breidt, F. J., McConkey, B., Regina, K. & Vazquez-Amabile, G. G.
(2019) "Climate and Soil Characteristics Determine Where No-Till Management Can Store Carbon in Soils and
Mitigate Greenhouse Gas Emissions." Scientific Reports 9(1): 11665.
Ogle, S.M., F.J. Breidt, M. Easter, S. Williams, K. Killian, and K. Paustian (2010) "Scale and uncertainty in modeled
soil organic carbon stock changes for U.S. croplands using a process-based model." Global Change Biology 16:810-
820.
Ogle, S.M., F.J. Breidt, M. Easter, S. Williams and K. Paustian (2007) "Empirically-Based Uncertainty Associated with
Modeling Carbon Sequestration Rates in Soils." Ecological Modeling 205:453-463.
Ogle, S.M., F.J. Breidt, and K. Paustian (2006) "Bias and variance in model results due to spatial scaling of
measurements for parameterization in regional assessments." Global Change Biology 12:516-523.
Ogle, S. M., et al. (2005) "Agricultural management impacts on soil organic carbon storage under moist and dry
climatic conditions of temperate and tropical regions." Biogeochemistry 72: 87-121.
Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and management
impacts on soil organic carbon storage for U.S. agroecosystems between 1982 and 1997." Global Change Biology
9:1521-1542.
Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface Submodel: Description
and Testing". Glob. Planet. Chang. 19: 35-48.
Parton, W.J., D.S. Ojima, C.V. Cole, and D.S. Schimel (1994) "A General Model for Soil Organic Matter Dynamics:
Sensitivity to litter chemistry, texture and management," in Quantitative Modeling of Soil Forming Processes.
Special Publication 39, Soil Science Society of America, Madison, Wl, 147-167.
Parton, W.J., D.S. Schimel, C.V. Cole, D.S. Ojima (1987) "Analysis of factors controlling soil organic matter levels in
Great Plains grasslands." Soil Science Society of America Journal 51:1173-1179.
Parton, W.J., J.W.B. Stewart, C.V. Cole. (1988) "Dynamics of C, N, P, and S in grassland soils: a model."
Biogeochemistry 5:109-131.
Paustian, K., et al. (1997a) "Agricultural soils as a sink to mitigate CO2 emissions." Soil Use and Management 13:
230-244.
Paustian, K., et al. (1997b) Management controls on soil carbon. In Soil organic matter in temperate
agroecosystems: long-term experiments in North America (Paul E.A., K. Paustian, and C.V. Cole, eds.). Boca Raton,
CRC Press, pp. 15-49.
Potter, C. S., J.T. Randerson, C.B. Fields, P.A. Matson, P.M. Vitousek, H.A. Mooney, and S.A. Klooster (1993)
"Terrestrial ecosystem production: a process model based on global satellite and surface data." Global
Biogeochemical Cycles 7:811-841.
Potter, C., S. Klooster, A. Huete, and V. Genovese (2007) Terrestrial carbon sinks for the United States predicted
from MODIS satellite data and ecosystem modeling. Earth Interactions 11, Article No. 13, DOI 10.1175/EI228.1.
PRISM Climate Group (2018) PRISM Climate Data, Oregon State University, http://prism.oregonstate.edu,
downloaded 18 July 2018.
Pukelsheim, F. (1994) 'The 3-Sigma-Rule." American Statistician 48:88-91
Soil Survey Staff (2016) State Soil Geographic (STATSGO) Database for State. Natural Resources Conservation
Service, United States Department of Agriculture. Available online at:
http://www.ncgc.nrcs.usda.gov/products/datasets/statsgo/index.html.
Spencer, S., S.M. Ogle, F.J. Breidt, J. Goebel, and K. Paustian. (2011) "Designing a national soil carbon monitoring
network to support climate change policy: a case example for US agricultural lands." Greenhouse Gas Management
& Measurement 1: 167-178.
74 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Towery, D. (2001) Personal Communication. Dan Towery regarding adjustments to the CTIC (1998) tillage data to
reflect long-term trends, Conservation Technology Information Center, West Lafayette, IN, and Marlen Eve,
National Resource Ecology Laboratory, Fort Collins, CO. February 2001.
USDA-ERS (2018) Agricultural Resource Management Survey (ARMS) Farm Financial and Crop Production Practices:
Tailored Reports. Available online at: https://www.ers.usda.gov/data-products/arms-farm-financial-and-crop-
production-practices/.
USDA-ERS (1997) Cropping Practices Survey Data —1995. Economic Research Service, United States Department of
Agriculture. Available online at: http://www.ers.usda.gov/data/archive/93018/.
USDA-FSA (2015) Conservation Reserve Program Monthly Summary-September 2015. U.S. Department of
Agriculture, Farm Service Agency, Washington, D.C. Available online at: https://www.fsa.usda.gov/Assets/USDA-
FSA-Public/usdafiles/Conservation/PDF/sep2015summary.pdf.
USDA-NASS (2017) 2017 Census of Agriculture. USDA National Agricultural Statistics Service, Complete data
available at http://www.nass.usda.gov/AgCensus.
USDA-NASS (2012) 2012 Census of Agriculture. USDA National Agricultural Statistics Service, Complete data
available at http://www.nass.usda.gov/AgCensus.
USDA-NASS (2004) Agricultural Chemical Usage: 2003 Field Crops Summary. Report AgChl(04)a. National
Agricultural Statistics Service, U.S. Department of Agriculture, Washington, D.C. Available online at:
http://usda.mannlib.cornell.edu/reports/nassr/other/pcu-bb/agcs0504.pdfH.
USDA-NASS (1999) Agricultural Chemical Usage: 1998 Field Crops Summary. Report AgCHl(99). National
Agricultural Statistics Service, U.S. Department of Agriculture, Washington, DC. Available online at:
http://usda.mannlib.cornell.edu/reports/nassr/other/pcu-bb/agch0599.pdf.
USDA-NASS (1992) Agricultural Chemical Usage: 1991 Field Crops Summary. Report AgChl(92). National
Agricultural Statistics Service, U.S. Department of Agriculture, Washington, D.C. Available online at:
http://usda.mannlib.cornell.edu/reports/nassr/other/pcu-bb/agch0392.txtH.
USDA-NRCS (2012) Assessment of the Effects of Conservation Practices on Cultivated Cropland in the Upper
Mississippi River Basin. U.S. Department of Agriculture, Natural Resources Conservation Service,
https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/stelprdb 1042093.pdf.
USDA-NRCS (2018a) Summary Report: 2015 National Resources Inventory. Natural Resources Conservation Service,
Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd 1422028.pdf.
USDA-NRCS (2018b) CEAP Cropland Farmer Surveys. USDA Natural Resources Conservation Service.
https://www.nrcs.usda.gov/wps/portal/nrcs/detail/national/technical/nra/ceap/na/?cid=nrcsl43 014163.
USDA-NRCS (2000) Digital Data and Summary Report: 1997 National Resources Inventory. Revised December 2000.
Resources Inventory Division, Natural Resources Conservation Service, United States Department of Agriculture,
Beltsville, MD.
Van Buuren, S. (2012) "Flexible imputation of missing data." Chapman & Hall/CRC, Boca Raton, FL.
Yang, L, Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J., Fry, J., Funk, M.,
Granneman, B., Liknes, G. C., Rigge, M. & Xian, G. (2018) "A new generation of the United States National Land
Cover Database: Requirements, research priorities, design, and implementation strategies." ISPRS Journal of
Photogrammetry and Remote Sensing 146:108-123.
Zomer RJ, Trabucco A, Bossio DA, van Straaten O, Verchot LV (2008) Climate Change Mitigation: A Spatial Analysis
of Global Land Suitability for Clean Development Mechanism Afforestation and Reforestation. Agric. Ecosystems
and Envir. 126: 67-80.
References and Abbreviations 75
-------
Zomer RJ, Bossio DA, Trabucco A, Yuanjie L, Gupta DC & Singh VP (2007) Trees and Water: Smallholder
Agroforestry on Irrigated Lands in Northern India. Colombo, Sri Lanka: International Water Management Institute,
pp 45. (IWMI Research Report 122).
Land Converted to Cropland
Sampson and D. Hair, (eds.). Forest and Global Change, Volume 2: Forest Management Opportunities for
Mitigating Carbon Emissions. American Forests. Washington, D.C., 1-26 and 261-379 (appendices 262 and 263).
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel (2001) "Simulated
Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the DAYCENT Model." In Modeling Carbon
and Nitrogen Dynamics for Soil Management, Schaffer, M., L. Ma, S. Hansen, (eds.). CRC Press, Boca Raton, Florida,
pp. 303-332.
Del Grosso, S.J., S.M. Ogle, W.J. Parton (2011) "Soil organic matter cycling and greenhouse gas accounting
methodologies." Chapter 1, pp 3-13 DOI: 10.1021/bk-2011-1072.ch001. In: Understanding Greenhouse Gas
Emissions from Agricultural Management (L. Guo, A. Gunasekara, L. McConnell. Eds.), American Chemical Society,
Washington, D.C.
Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel (2001) "Simulated
Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the DAYCENT Model." In Schaffer, M., L. Ma,
S. Hansen, (eds.); Modeling Carbon and Nitrogen Dynamics for Soil Management. CRC Press. Boca Raton, Florida.
303-332.
Domke, G.M., J.E. Smith, and C.W. Woodall. (2011) "Accounting for density reduction and structural loss in
standing dead trees: Implications for forest biomass and carbon stock estimates in the United States". Carbon
Balance and Management 6:14.
Domke, G.M., et al. (2013) "From models to measurements: comparing down dead wood carbon stock estimates in
the U.S. forest inventory." PLoS ONE 8(3): e59949.
Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., and Smith, J.E. (2016) "A framework for estimating litter
carbon stocks in forests of the United States." Science of the Total Environment 557-558: 469-478.
Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and Wickham, J. (2011) "Completion of the
2006 National Land Cover Database for the Conterminous United States." PE&RS, Vol. 77(9):858-864.
Harmon, M.E., C.W. Woodall, B. Fasth, J. Sexton, M. Yatkov. (2011) Differences between standing and downed
dead tree wood density reduction factors: A comparison across decay classes and tree species. Res. Paper. NRS-15.
Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 40 p.
Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A., VanDriel, J.N., and Wickham,
J. (2007) "Completion of the 2001 National Land Cover Database for the Conterminous United States."
Photogrammetric Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
Megown, K. (2015) "Completion of the 2011 National Land Cover Database for the conterminous United States-
Representing a decade of land cover change information." Photogrammetric Engineering and Remote Sensing 81:
345-354.
Houghton, R.A., et al. (1983) "Changes in the carbon content of terrestrial biota and soils between 1860 and 1980:
a net release of CO2 to the atmosphere." Ecological Monographs 53: 235-262.
Houghton, R. A. and Nassikas, A. A. (2017) "Global and regional fluxes of carbon from land use and land cover
change 1850-2015." Global Biogeochemical Cycles 31(3): 456-472.
76 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Jenkins, J.C., D.C. Chojnacky, L.S. Heath, and R.A. Birdsey (2003) "National-scale biomass estimators for United
States tree species." Forest Science 49(l):12-35.
Metherell, A.K., L.A. Harding, C.V. Cole, and W.J. Parton (1993) CENTURY Soil Organic Matter Model Environment.
Agroecosystem version 4.0. Technical documentation, GPSRTech. Report No. 4, USDA/ARS, Ft. Collins, CO.
Ogle, S.M., F.J. Breidt, M. Easter, S. Williams, K. Killian, and K. Paustian (2010) "Scale and uncertainty in modeled
soil organic carbon stock changes for U.S. croplands using a process-based model." Global Change Biology 16:810-
820.
Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and management
impacts on soil organic carbon storage for U.S. agroecosystems between 1982 and 1997." Global Change Biology
9:1521-1542.
Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface Submodel: Description
and Testing". Glob. Planet. Chang. 19: 35-48.
Parton, W.J., D.S. Ojima, C.V. Cole, and D.S. Schimel (1994) "A General Model for Soil Organic Matter Dynamics:
Sensitivity to litter chemistry, texture and management," in Quantitative Modeling of Soil Forming Processes.
Special Publication 39, Soil Science Society of America, Madison, Wl, 147-167.
Parton, W.J., D.S. Schimel, C.V. Cole, D.S. Ojima (1987) "Analysis of factors controlling soil organic matter levels in
Great Plains grasslands." Soil Science Society of America Journal 51:1173-1179.
Parton, W.J., J.W.B. Stewart, C.V. Cole. (1988) "Dynamics of C, N, P, and S in grassland soils: a model."
Biogeochemistry 5:109-131.
PRISM Climate Group (2018) PRISM Climate Data, Oregon State University, http://prism.oregonstate.edu,
downloaded 18 July 2018.
Smith, J.E.; Heath, L.S.; Skog, K.E.; Birdsey, R.A. (2006) Methods for calculating forest ecosystem and harvested
carbon with standard estimates for forest types of the United States. Gen. Tech. Rep. NE-343. Newtown Square,
PA: U.S. Department of Agriculture, Forest Service, Northeastern Research Station. 216 p.
Tubiello, F. N., et al. (2015) "The Contribution of Agriculture, Forestry and other Land Use activities to Global
Warming, 1990-2012." Global Change Biology 21:2655-2660.
USDA Forest Service (2020) Forest Inventory and Analysis National Program: FIA Data Mart. U.S. Department of
Agriculture Forest Service. Washington, D.C. Available online at: http://apps.fs.fed.us/fiadb-
downloads/datamart.html. Accessed on 10 October 2020.
USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory. Natural Resources Conservation Service,
Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd 1422028.pdf.
Woodall, C.W., and V.J. Monleon (2008) Sampling protocol, estimation, and analysis procedures for the down
woody materials indicator of the FIA program. Gen. Tech. Rep. NRS-22. Newtown Square, PA: U.S. Department of
Agriculture, Forest Service, Northern Research Station. 68 p.
Woodall, C.W., L.S. Heath, G.M. Domke, and M.C. Nichols (2011) Methods and equations for estimating
aboveground volume, biomass, and carbon for trees in the U.S. forest inventory, 2010. Gen. Tech. Rep. NRS-88.
Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 30 p.
Yang, L, Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J., Fry, J., Funk, M.,
Granneman, B., Liknes, G. C., Rigge, M. & Xian, G. (2018) "A new generation of the United States National Land
Cover Database: Requirements, research priorities, design, and implementation strategies." ISPRS Journal of
Photogrammetry and Remote Sensing 146:108-123.
References and Abbreviations 77
-------
Grassland Remaining Grassland: Soil Carbon Stock Changes
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
Del Grosso, S.J., S.M. Ogle, W.J. Parton (2011) Soil organic matter cycling and greenhouse gas accounting
methodologies, Chapter 1, pp 3-13 DOI: 10.1021/bk-2011-1072.ch001. In: Understanding Greenhouse Gas
Emissions from Agricultural Management (L Guo, A. Gunasekara, L McConnell. Eds.), American Chemical Society,
Washington, D.C.
Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel (2001) "Simulated
Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the DAYCENT Model." In Modeling Carbon
and Nitrogen Dynamics for Soil Management, Schaffer, M., L Ma, S. Hansen, (eds.). CRC Press, Boca Raton, Florida,
pp. 303-332.
Edmonds, L, R. L. Kellogg, B. Kintzer, L. Knight, C. Lander, J. Lemunyon, D. Meyer, D.C. Moffitt, and J. Schaefer
(2003) "Costs associated with development and implementation of Comprehensive Nutrient Management Plans."
Part I—Nutrient management, land treatment, manure and wastewater handling and storage, and recordkeeping.
Natural Resources Conservation Service, U.S. Department of Agriculture. Available online at:
http://www.nrcs.usda.gov/technical/land/pubs/cnmpl.html.
EPA (1999) Biosolids Generation, Use and Disposal in the United States. Office of Solid Waste, U.S. Environmental
Protection Agency. Available online at: http://biosolids.policy.net/relatives/18941.PDF.
Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L, Barnes, C., Herold, N., and Wickham, J. (2011) Completion of
the 2006 National Land Cover Database for the Conterminous United States, PE&RS, Vol. 77(9):858-864.
Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A., VanDriel, J.N., and
Wickham, J. (2007) Completion of the 2001 National Land Cover Database for the Conterminous United States.
Photogrammetric Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-
Representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, v.
81, no. 5, p. 345-354.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Kellogg, R.L., C.H. Lander, D.C. Moffitt, and N. Gollehon (2000) Manure Nutrients Relative to the Capacity of
Cropland and Pastureland to Assimilate Nutrients: Spatial and Temporal Trends for the United States. U.S.
Department of Agriculture, Washington, D.C. Publication number nps00-0579.
Metherell, A.K., L.A. Harding, C.V. Cole, and W.J. Parton (1993) "CENTURY Soil Organic Matter Model
Environment." Agroecosystem version 4.0. Technical documentation, GPSRTech. Report No. 4, USDA/ARS, Ft.
Collins, CO.
NEBRA (2007) A National Biosolids Regulation, Quality, End Use & Disposal Survey. North East Biosolids and
Residuals Association. July 21, 2007.
Nusser, S.M. and J.J. Goebel (1997) The national resources inventory: a long-term multi-resource monitoring
programme. Environmental and Ecological Statistics 4:181-204.
Ogle, S.M., F.J. Breidt, M. Easter, S. Williams, K. Killian, and K. Paustian (2010) "Scale and uncertainty in modeled
soil organic carbon stock changes for U.S. croplands using a process-based model." Global Change Biology 16:810-
820.
78 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and management
impacts on soil organic carbon storage for U.S. agroecosystems between 1982 and 1997." Global Change Biology
9:1521-1542.
Parton, W.J., D.S. Ojima, C.V. Cole, and D.S. Schimel (1994) "A General Model for Soil Organic Matter Dynamics:
Sensitivity to litter chemistry, texture and management," in Quantitative Modeling of Soil Forming Processes.
Special Publication 39, Soil Science Society of America, Madison, Wl, 147-167.
Parton, W.J., D.S. Schimel, C.V. Cole, D.S. Ojima (1987) "Analysis of factors controlling soil organic matter levels in
Great Plains grasslands." Soil Science Society of America Journal 51:1173-1179.
Parton, W.J., J.W.B. Stewart, C.V. Cole. (1988) "Dynamics of C, N, P, and S in grassland soils: a model."
Biogeochemistry 5:109-131.
Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface Submodel: Description
and Testing". Glob. Planet. Chang. 19: 35-48.PRISM Climate Group, Oregon State University,
http://prism.oregonstate.edu, created 24 July 2015.
PRISM Climate Group (2018) PRISM Climate Data, Oregon State University, http://prism.oregonstate.edu,
downloaded 18 July 2018.
United States Bureau of Land Management (BLM) (2014) Rangeland Inventory, Monitoring, and Evaluation
Reports. Bureau of Land Management. U.S. Department of the Interior. Available online at:
http://www.blm.gov/wo/st/en/prog/more/rangeland management/rangeland inventory.html.
USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory. Natural Resources Conservation Service,
Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd 1422028.pdf.
USDA Forest Service (2020) Forest Inventory and Analysis National Program: FIA Data Mart. U.S. Department of
Agriculture Forest Service. Washington, D.C. Available online at: http://apps.fs.fed.us/fiadb-
downloads/datamart.html. Accessed on 10 October 2020.
Yang, L, Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J., Fry, J., Funk, M.,
Granneman, B., Liknes, G. C., Rigge, M. & Xian, G. (2018) "A new generation of the United States National Land
Cover Database: Requirements, research priorities, design, and implementation strategies." ISPRS Journal of
Photogrammetry and Remote Sensing 146:108-123.
Grassland Remaining Grassland: Non-C02 Emissions from
Grassland Fires
Anderson, R.C. Evolution and origin of the Central Grassland of North America: climate, fire and mammalian
grazers. Journal of the Torrey Botanical Society 133: 626-647.
Andreae, M.O. and P. Merlet (2001) Emission of trace gases and aerosols from biomass burning. Global
Biogeochemical Cycles 15:955-966.
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
Chapin, F.S., S.F. Trainor, O. Huntington, A.L Lovecraft, E. Zavaleta, D.C. Natcher, A.D. McGuire, J.L. Nelson, L. Ray,
M. Calef, N. Fresco, H. Huntington, T.S. Rupp, L. DeWilde, and R.L. Naylor (2008) Increasing wildfires in Alaska's
Boreal Forest: Pathways to potential solutions of a wicked problem. Bioscience 58:531-540.
Daubenmire, R. (1968) Ecology of fire in grasslands. Advances in Ecological Research 5:209-266.
Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L, Barnes, C., Herold, N., and Wickham, J. (2011) Completion of
the 2006 National Land Cover Database for the Conterminous United States, PE&RS, Vol. 77(9):858-864.
References and Abbreviations 79
-------
Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A., VanDriel, J.N., and Wickham,
J. (2007) Completion of the 2001 National Land Cover Database for the Conterminous United States.
Photogrammetric Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
Homer, C.G., Dewitz, J.A., Yang, L, Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-
Representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, v. 81,
no. 5, p. 345-354.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Ogle, S.M., S. Spencer, M. Hartman, L. Buendia, L. Stevens, D. du Toit, J. Witi (2016) "Developing national baseline
GHG emissions and analyzing mitigation potentials for agriculture and forestry using an advanced national GHG
inventory software system." In Advances in Agricultural Systems Modeling 6, Synthesis and Modeling of
Greenhouse Gas Emissions and Carbon Storage in Agricultural and Forestry Systems to Guide Mitigation and
Adaptation, S. Del Grosso, LR. Ahuja and W.J. Parton (eds.), American Society of Agriculture, Crop Society of
America and Soil Science Society of America, pp. 129-148.
Nusser, S.M. and J.J. Goebel (1997) The national resources inventory: a long-term multi-resource monitoring
programme. Environmental and Ecological Statistics 4:181-204.
USDA-NRCS (2015) Summary Report: 2012 National Resources Inventory, Natural Resources Conservation Service,
Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa. Available
online at: http://www.nrcs.usda.gov/lnternet/FSE DOCU MENTSZnrcseprd396218.pdf.
land Converted to Grassland
Asner, G.P., Archer, S., Hughes, R.F., Ansley, R.J. and Wessman, C.A. (2003) "Net changes in regional woody
vegetation cover and carbon storage in Texas drylands, 1937-1999." Global Change Biology 9(3): 316-335.
Birdsey, R. (1996) "Carbon Storage for Major Forest Types and Regions in the Conterminous United States." In R.N.
Sampson and D. Hair, (eds.). Forest and Global Change, Volume 2: Forest Management Opportunities for
Mitigating Carbon Emissions. American Forests. Washington, D.C., 1-26 and 261-379 (appendices 262 and 263).
Breshears, D.D., Knapp, A.K., Law, D.J., Smith, M.D., Twidwell, D. and Wonkka, C.L, 2016. Rangeland Responses to
Predicted Increases in Drought Extremity. Rangelands, 38(4), pp. 191-196.
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
Del Grosso, S.J., S.M. Ogle, W.J. Parton. (2011) Soil organic matter cycling and greenhouse gas accounting
methodologies, Chapter 1, pp 3-13 DOI: 10.1021/bk-2011-1072.ch001. In: Understanding Greenhouse Gas
Emissions from Agricultural Management (L. Guo, A. Gunasekara, L. McConnell. Eds.), American Chemical Society,
Washington, D.C.
Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel (2001) "Simulated
Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the DAYCENT Model." In Modeling Carbon
and Nitrogen Dynamics for Soil Management (Schaffer, M., L. Ma, S. Hansen, (eds.). CRC Press, Boca Raton, Florida,
pp. 303-332.
Domke, G.M., J.E. Smith, and C.W. Woodall. (2011) Accounting for density reduction and structural loss in standing
dead trees: Implications for forest biomass and carbon stock estimates in the United States. Carbon Balance and
Management. 6:14.
Domke, G.M., et al. (2013) From models to measurements: comparing down dead wood carbon stock estimates in
the U.S. forest inventory. PLoS ONE 8(3): e59949.
80 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., and Smith, J.E. (2016) A framework for estimating litter
carbon stocks in forests of the United States. Science of the Total Environment 557-558: 469-478.
Epstein, H.E., Gill, R.A., Paruelo, J.M., Lauenroth, W.K., Jia, G.J. and Burke, I.C. (2002) The relative abundance of
three plant functional types in temperate grasslands and shrublands of North and South America: effects of
projected climate change. Journal of Biogeography, 29(7), pp.875-888.
Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L, Barnes, C., Herold, N., and Wickham, J. (2011) Completion of
the 2006 National Land Cover Database for the Conterminous United States, PE&RS, Vol. 77(9):858-864.
Harmon, M.E., C.W. Woodall, B. Fasth, J. Sexton, M. Yatkov. (2011) Differences between standing and downed
dead tree wood density reduction factors: A comparison across decay classes and tree species. Res. Paper. NRS-15.
Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 40 p.
Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A., VanDriel, J.N., and Wickham,
J. (2007) Completion of the 2001 National Land Cover Database for the Conterminous United States.
Photogrammetric Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
Homer, C.G., Dewitz, J.A., Yang, L, Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-
Representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, v. 81,
no. 5, p. 345-354.
Houghton, R.A., et al. (1983) "Changes in the carbon content of terrestrial biota and soils between 1860 and 1980:
a net release of CO2 to the atmosphere." Ecological Monographs 53: 235-262.
Houghton, R. A. and Nassikas, A. A. (2017) "Global and regional fluxes of carbon from land use and land cover
change 1850-2015." Global Biogeochemical Cycles 31(3): 456-472.
Huang, C.Y., Asner, G.P., Martin, R.E., Barger, N.N. and Neff, J.C. (2009) "Multiscale analysis of tree cover and
aboveground carbon stocks in pinyon-juniper woodlands." Ecological Applications 19(3): 668-681.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change, [H.S. Eggleston, L. Buendia, K. Miwa, T
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Jenkins, J.C., D.C. Chojnacky, L.S. Heath, and R.A. Birdsey (2003) "National-scale biomass estimators for United
States tree species." Forest Science 49(l):12-35.
Jurena, P.N. and Archer, S., (2003) Woody plant establishment and spatial heterogeneity in grasslands. Ecology,
84(4), pp.907-919.
Lenihan, J.M., Drapek, R., Bachelet, D. and Neilson, R.P., (2003) Climate change effects on vegetation distribution,
carbon, and fire in California. Ecological Applications, 13(6), pp.1667-1681.
Metherell, A.K., L.A. Harding, C.V. Cole, and W.J. Parton (1993) "CENTURYSoil Organic Matter Model
Environment." Agroecosystem version 4.0. Technical documentation, GPSRTech. Report No. 4, USDA/ARS, Ft.
Collins, CO.
Ogle, S.M., F.J. Breidt, M. Easter, S. Williams, K. Killian, and K. Paustian (2010) "Scale and uncertainty in modeled
soil organic carbon stock changes for U.S. croplands using a process-based model." Global Change Biology 16:810-
820.
Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and management
impacts on soil organic carbon storage for U.S. agroecosystems between 1982 and 1997." Global Change Biology
9:1521-1542.
Parton, W.J., D.S. Ojima, C.V. Cole, and D.S. Schimel (1994) "A General Model for Soil Organic Matter Dynamics:
Sensitivity to litter chemistry, texture and management," in Quantitative Modeling of Soil Forming Processes.
Special Publication 39, Soil Science Society of America, Madison, Wl, 147-167.
References and Abbreviations 81
-------
Parton, W.J., D.S. Schimel, C.V. Cole, D.S. Ojima (1987) "Analysis of factors controlling soil organic matter levels in
Great Plains grasslands." Soil Science Society of America Journal 51:1173-1179.
Parton, W.J., J.W.B. Stewart, C.V. Cole (1988) "Dynamics of C, N, P, and S in grassland soils: a model."
Biogeochemistry 5:109-131.
Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface Submodel: Description
and Testing". Glob. Planet. Chang. 19: 35-48.
PRISM Climate Group (2018) PRISM Climate Data, Oregon State University, http://prism.oregonstate.edu,
downloaded 18 July 2018.
Scholes, R.J. and Archer, S.R. (1997) Tree-grass interactions in savannas 1. Annual review of Ecology and
Systematics, 28(1), pp.517-544.
Sims, P.L, Singh, J.S. and Lauenroth, W.K. (1978) The structure and function often western North American
grasslands: I. Abiotic and vegetational characteristics. The Journal of Ecology, pp.251-285.
Smith, J.E.; Heath, L.S.; Skog, K.E.; Birdsey, R.A. (2006) Methods for calculating forest ecosystem and harvested
carbon with standard estimates for forest types of the United States. Gen. Tech. Rep. NE-343. Newtown Square,
PA: U.S. Department of Agriculture, Forest Service, Northeastern Research Station. 216 p.
Tarhouni, M., et al. (2016) Measurement of the aboveground biomass of some rangeland species using a digital
non-destructive technique. Botany Letters 163(3):281-287.
Tubiello, F. N., et al. (2015) "The Contribution of Agriculture, Forestry and other Land Use activities to Global
Warming, 1990-2012." Global Change Biology 21:2655-2660.
United States Bureau of Land Management (BLM) (2014) Rangeland Inventory, Monitoring, and Evaluation
Reports. Bureau of Land Management. U.S. Department of the Interior. Available online at:
http://www.blm.gov/wo/st/en/prog/more/rangeland management/rangeland inventory.html.
USDA Forest Service (2020) Forest Inventory and Analysis National Program: FIA Data Mart. U.S. Department of
Agriculture Forest Service. Washington, D.C. Available online at: http://apps.fs.fed.us/fiadb-
downloads/datamart.html. Accessed on 10 October 2020.
USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory. Natural Resources Conservation Service,
Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd 1422028.pdf.
Woodall, C.W., and V.J. Monleon (2008) Sampling protocol, estimation, and analysis procedures for the down
woody materials indicator of the FIA program. Gen. Tech. Rep. NRS-22. Newtown Square, PA: U.S. Department of
Agriculture, Forest Service, Northern Research Station. 68 p.
Woodall, C.W., LS. Heath, G.M. Domke, and M.C. Nichols. (2011) Methods and equations for estimating
aboveground volume, biomass, and carbon for trees in the U.S. forest inventory, 2010. Gen. Tech. Rep. NRS-88.
Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 30 p.
Yang, L, Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J., Fry, J., Funk, M.,
Granneman, B., Liknes, G. C., Rigge, M. & Xian, G. (2018) "A new generation of the United States National Land
Cover Database: Requirements, research priorities, design, and implementation strategies." ISPRS Journal of
Photogrammetry and Remote Sensing 146:108-123.
Wetlands Remaining Wetlands: C02, CH4, and N20 Emissions
from Peatlands Remaining Peatlands
Apodaca, L. (2011) Email correspondence. Lori Apodaca, Peat Commodity Specialist, USGS and Emily Rowan, ICF
International. November.
82 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Apodaca, L. (2008) E-mail correspondence. Lori Apodaca, Peat Commodity Specialist, USGS and Emily Rowan, ICF
International. October and November.
Brioche, A. (2021) E-mail correspondence. Amanda, Brioche, Peat Commodity Specialist, USGS and Spencer Sauter,
ERG. June.
Cleary, J., N. Roulet and T.R. Moore (2005) "Greenhouse gas emissions from Canadian peat extraction, 1990-2000:
A life-cycle analysis." Ambio 34:456-461.
Division of Geological & Geophysical Surveys (DGGS), Alaska Department of Natural Resources (1997-2015)
Alaska's Mineral Industry Report (1997-2014). Alaska Department of Natural Resources, Fairbanks, AK. Available
online at http://www.dggs.dnr.state.ak.us/pubs/pubs?reqtype=minerals.
IPCC (2013) 2013 Supplement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands.
Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds.). Published: IPCC,
Switzerland.
IPCC (2007) Climate Change 2007: Synthesis Report. Contribution of Working Groups I, II and III to the Fourth
Assessment Report (AR4) of the IPCC. The Intergovernmental Panel on Climate Change, R.K. Pachauri, A. Resinger
(eds.). Geneva, Switzerland.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
Szumigala, D.J. (2011) Phone conversation. Dr. David Szumigala, Division of Geological and Geophysical Surveys,
Alaska Department of Natural Resources and Emily Rowan, ICF International. January 18, 2011.
Szumigala, D.J. (2008) Phone conversation. Dr. David Szumigala, Division of Geological and Geophysical Surveys,
Alaska Department of Natural Resources and Emily Rowan, ICF International. October 17, 2008.
USGS (1991-2017) Minerals Yearbook: Peat (1994-2017). United States Geological Survey, Reston, VA. Available
online at http://minerals.usgs.gov/minerals/pubs/commodity/peat/index.html.
USGS (2021a) Minerals Yearbook: Peat (2018). United States Geological Survey, Reston, VA. Available online at
https://www.usgs.gov/centers/nmic/peat-statistics-and-information.
USGS (2021b) Mineral Commodity Summaries: Peat (1996-2021). United States Geological Survey, Reston, VA.
Available online at https://www.usgs.gov/centers/nmic/peat-statistics-and-information.
Wetlands Remaining Coastal Wetlands: Emissions and
Removals from Coastal Wetlands Remaining Coastal Wetlands
Bianchi, T. S., Allison, M. A., Zhao, J., Li, X., Comeaux, R. S., Feagin, R. A., & Kulawardhana, R. W. (2013) Historical
reconstruction of mangrove expansion in the Gulf of Mexico: linking climate change with carbon sequestration in
coastal wetlands. Estuarine, Coastal and Shelf Science 119: 7-16.
Byrd, K. B., Ballanti, L. R., Thomas, N. M., Nguyen, D. K., Holmquist, J. R., Simard, M., Windham-Myers, L., Schile, L.
M., Parker, V. T.,... and Castaneda-Moya, E. (2017) Biomass/Remote Sensing dataset: 30m resolution tidal marsh
biomass samples and remote sensing data for six regions in the conterminous United States: U.S. Geological Survey
data release, https://doi.org/10.5066/F77943K8.
Byrd, K. B., Ballanti, L., Thomas, N., Nguyen, D., Holmquist, J.R., Simard, M., and Windham-Myers, L. (2018) A
remote sensing-based model of tidal marsh aboveground carbon stocks for the conterminous United States. ISPRS
Journal of Photogrammetry and Remote Sensing 139: 255-271.
Byrd, K. B., Ballanti, L., Thomas, N., Nguyen, D., Holmquist, J.R., Simard, M., and Windham-Myers, L. (2020)
Corrigendum to "A remote sensing-based model of tidal marsh aboveground carbon stocks for the conterminous
United States". ISPRS Journal of Photogrammetry and Remote Sensing 166: 63-67.
References and Abbreviations 83
-------
Callaway, J. C., Borgnis, E. L., Turner, R. E. & Milan, C. S. (2012a) Carbon sequestration and sediment accretion in
San Francisco Bay tidal wetlands. Estuaries and Coasts 35(5): 1163-1181.
Callaway, J. C., Borgnis, E. L., Turner, R. E., Milan, C. S., Goodfriend, W., & Richmond, S. (2012b) "Wetland Sediment
Accumulation at Corte Madera Marsh and Muzzi Marsh". San Francisco Bay Conservation and Development
Commission.
Church, T. M., Sommerfield, C. K., Velinsky, D. J., Point, D., Benoit, C., Amouroux, D. & Donard, O. F. X. (2006)
Marsh sediments as records of sedimentation, eutrophication and metal pollution in the urban Delaware Estuary.
Marine Chemistry 102(1-2): 72-95.
Couvillion, B. R., Barras, J. A., Steyer, G. D., Sleavin, W., Fischer, M., Beck, H., & Heckman, D. (2011) Land area
change in coastal Louisiana (1932 to 2010) (pp. 1-12). U.S. Department of the Interior, U.S. Geological Survey.
Couvillion, B. R., Fischer, M. R., Beck, H. J. and Sleavin, W. J. (2016) Spatial Configuration Trends in Coastal
Louisiana from 1986 to 2010. Wetlands 1-13.
Craft, C. B., & Richardson, C. J. (1998) Recent and long-term organic soil accretion and nutrient accumulation in the
Everglades. Soil Science Society of America Journal 62(3): 834-843.
Crooks, S., Findsen, J., Igusky, K., Orr, M. K. and Brew, D. (2009) Greenhouse Gas Mitigation Typology Issues Paper:
Tidal Wetlands Restoration. Report by PWA and SAIC to the California Climate Action Reserve.
Crooks, S., Rybczyk, J., O'Connell, K., Devier, D. L, Poppe, K., Emmett-Mattox, S. (2014) Coastal Blue Carbon
Opportunity Assessment for the Snohomish Estuary: The Climate Benefits of Estuary Restoration. Report by
Environmental Science Associates, Western Washington University, EarthCorps, and Restore America's Estuaries.
DeLaune, R. D., & White, J. R. (2012) Will coastal wetlands continue to sequester carbon in response to an increase
in global sea level?: A case study of the rapidly subsiding Mississippi river deltaic plain. Climatic Change, 110(1),
297-314.
Holmquist, J. R., Windham-Myers, L, Bliss, N., Crooks, S., Morris, J. T., Megonigal, J. P. & Woodrey, M. (2018)
Accuracy and Precision of Tidal Wetland Soil Carbon Mapping in the Conterminous United States. Scientific reports
8(1): 9478.
Hu, Z., Lee, J. W., Chandran, K., Kim, S. and Khanal, S. K. (2012) N2O Emissions from Aquaculture: A Review.
Environmental Science & Technology 46(12): 6470-6480.
Hussein, A. H., Rabenhorst, M. C. & Tucker, M. L. (2004) Modeling of carbon sequestration in coastal marsh soils.
Soil Science Society of America Journal 68(5): 1786-1795.
IPCC (2000) Good Practice Guidance and Uncertainty Management in National Greenhouse Gas Inventories.
Quantifying Uncertainties in Practice, Chapter 6. Penman, J., Kruger, D., Galbally, I., Hiraishi, T., Nyenzi, B.,
Emmanuel, S., Buendia, L., Hoppaus, R., Martinsen, T., Meijer, J., Miwa, K. andTanabe, K. (eds). Institute of Global
Environmental Strategies (IGES), on behalf of the Intergovernmental Panel on Climate Change (IPCC): Hayama,
Japan.
IPCC (2003) Good Practice Guidance for Land Use, Land-Use Change and Forestry. LUCF Sector Good Practice
Guidance, Chapter 3. Penman, J., Gytarsky, M., Hiraishi, T., Krug, T., Kruger, D., Pipatti, R., Buendia, L, Miwa, K.,
Ngara, T., Tanabe, K. and Wagner, F. (eds). Institute of Global Environmental Strategies (IGES), on behalf of the
Intergovernmental Panel on Climate Change (IPCC): Hayama, Japan.
IPCC (2006) IPCC Guidelines for National Greenhouse Gas Inventories. Prepared by the National Greenhouse Gas
Inventories Programme, Eggleston H.S., Buendia L, Miwa K., NgaraT. andTanabe K. (eds). IGES, Japan.
IPCC (2014) 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands.
Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds.). Published: IPCC,
Switzerland.
Kearney, M. S. & Stevenson, J. C. (1991) Island land loss and marsh vertical accretion rate evidence for historical
sea-level changes in Chesapeake Bay. Journal of Coastal Research 7(2): 403-415.
84 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Koster, D., Lichter, J., Lea, P. D., & Nurse, A. (2007) Historical eutrophication in a river-estuary complex in mid-
coast Maine. Ecological Applications 17(3): 765-778.
Lu, M & Megonigal, J. P. (2017) Final Report for RAE Baseline Assessment Project. Memo to Silvestrum Climate
Associates by Smithsonian Environmental Research Center, Maryland.
Lynch, J. C. (1989) Sedimentation and nutrient accumulation in mangrove ecosystems of the Gulf of Mexico. M.S.
thesis, Univ. of Southwestern Louisiana, Lafayette, LA.
Marchio, D. A., Savarese, M., Bovard, B., & Mitsch, W. J. (2016) Carbon sequestration and sedimentation in
mangrove swamps influenced by hydrogeomorphic conditions and urbanization in Southwest Florida. Forests 7:
116-135.
McCombs, J. W., Herold, N. D., Burkhalter, S. G. and Robinson C. J. (2016) Accuracy Assessment of NOAA Coastal
Change Analysis Program 2006-2010 Land Cover and Land Cover Change Data. Photogrammetric Engineering &
Remote Sensing. 82:711-718.
Merrill, J. Z. (1999) Tidal Freshwater Marshes as Nutrient Sinks: particulate Nutrient Burial and Denitrification.
Ph.D. Dissertation, University of Maryland, College Park, MD, 342 pp.
National Marine Fisheries Service (2021) Fisheries of the United States, 2018. U.S. Department of Commerce,
NOAA Current Fishery Statistics No. 2018.
National Oceanic and Atmospheric Administration, Office for Coastal Management (2020) Coastal Change Analysis
Program (C-CAP) Regional Land Cover. Charleston, SC: NOAA Office for Coastal Management. Accessed October
2020 at .
Noe, G. B., Hupp, C. R., Bernhardt, C. E., & Krauss, K. W. (2016) Contemporary deposition and long-term
accumulation of sediment and nutrients by tidal freshwater forested wetlands impacted by sea level rise. Estuaries
and Coasts 39(4): 1006-1019.
Orson, R. A., Simpson, R. L., & Good, R. E. (1990) Rates of sediment accumulation in a tidal freshwater marsh.
Journal of Sedimentary Research 60(6): 859-869.
Orson, R., Warren, R. & Niering, W. (1998) Interpreting sea level rise and rates of vertical marsh accretion in a
southern New England tidal salt marsh. Estuarine, Coastal and Shelf Science 47(4): 419-429.
Roman, C., Peck, J., Allen, J., King, J. & Appleby, P. (1997) Accretion of a New England (USA) salt marsh in response
to inlet migration, storms, and sea-level rise. Estuarine, Coastal and Shelf Science 45(6): 717-727.
Villa, J. A. & Mitsch W. J. (2015) Carbon sequestration in different wetland plant communities of Southwest Florida.
International Journal for Biodiversity Science, Ecosystems Services and Management 11: 17-28
Weston, N. B., Neubauer, S. C., Velinsky, D. J., & Vile, M. A. (2014) Net ecosystem carbon exchange and the
greenhouse gas balance of tidal marshes along an estuarine salinity gradient. Biogeochemistry 120:163-189.
Wetlands Remaining Wetlands: Flooded Land Remaining
Flooded Land
Davis, D. W. (1973) Louisiana Canals and Their Influence on Wetland Development. Louisiana State University and
Agricultural & Mechanical College. LSU Historical Dissertations and Theses. 2386., Louisiana State University.
IPCC (2003) Good Practice Guidance for Land Use, Land-Use Change and Forestry. LUCF Sector Good Practice
Guidance, Chapter 3. Penman, J., Gytarsky, M., Hiraishi, T., Krug, T., Kruger, D., Pipatti, R., Buendia, L, Miwa, K.,
Ngara, T., Tanabe, K. and Wagner, F. (eds). Institute of Global Environmental Strategies (IGES), on behalf of the
Intergovernmental Panel on Climate Change (IPCC): Hayama, Japan.
IPCC (2006) IPCC Guidelines for National Greenhouse Gas Inventories. Prepared by the National Greenhouse Gas
Inventories Programme, Eggleston H.S., Buendia L, Miwa K., Ngara T. and Tanabe K. (eds). IGES, Japan.
References and Abbreviations 85
-------
IPCC. (2013) 2013 Supplement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands.
Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds). In: IPCC,
Switzerland.
IPCC (2019) 2019 Refinement to the 2006 Guidelines for National Greenhouse Gas Inventories. Wetlands, Chapter
7. Lovelock, C. E., Evans, C., Barros, N., Prairie, Y. T., Aim, J., Bastviken, D., Beaulieu, J. J., Garneau, M., Harby, A.,
Harrison, J. A., Pare, David, Raadal, Hanne Lerche, Sherman, B., Zhang, Chengyi, Ogle, S. M.
Lehner B, Reidy Liermann C, Revenga C, Vorosmarty C, Fekete B, Crouzet P, Doll P, et al. (2011b) Global Reservoir
and Dam Database, Version 1 (GRanDvl): Dams, Revision 01. In: Palisades, NY: NASA Socioeconomic Data and
Applications Center (SEDAC).
Prairie, Y. T., et al. (2017) The GHG ReservoirTool (G-res) User guide. UNESCO/IHA research project on the GHG
status of freshwater reservoirs. Joint publication of the UNESCO Chair in Global Environmental Change and the
International Hydropower Association: 38.
Land Converted to Wetlands: Emissions and Removals from
Land Converted to Vegetated Coastal Wetlands
Bianchi, T. S., Allison, M. A., Zhao, J., Li, X., Comeaux, R. S., Feagin, R. A., & Kulawardhana, R. W. (2013) Historical
reconstruction of mangrove expansion in the Gulf of Mexico: linking climate change with carbon sequestration in
coastal wetlands. Estuarine, Coastal and Shelf Science 119: 7-16.
Byrd, K. B., Ballanti, L. R., Thomas, N. M., Nguyen, D. K., Holmquist, J. R., Simard, M., Windham-Myers, L., Schile, L.
M., Parker, V. T.,... and Castaneda-Moya, E. (2017) Biomass/Remote Sensing dataset: 30m resolution tidal marsh
biomass samples and remote sensing data for six regions in the conterminous United States: U.S. Geological Survey
data release, https://doi.org/10.5066/F77943K8.
Byrd, K. B., Ballanti, L., Thomas, N., Nguyen, D., Holmquist, J.R., Simard, M., and Windham-Myers, L. (2018) A
remote sensing-based model of tidal marsh aboveground carbon stocks for the conterminous United States. ISPRS
Journal of Photogrammetry and Remote Sensing 139: 255-271.
Byrd, K. B., Ballanti, L., Thomas, N., Nguyen, D., Holmquist, J.R., Simard, M., and Windham-Myers, L. (2020)
Corrigendum to "A remote sensing-based model of tidal marsh aboveground carbon stocks for the conterminous
United States". ISPRS Journal of Photogrammetry and Remote Sensing 166: 63-67.
Callaway, J. C., Borgnis, E. L., Turner, R. E. & Milan, C. S. (2012a) Carbon sequestration and sediment accretion in
San Francisco Bay tidal wetlands. Estuaries and Coasts 35(5): 1163-1181.
Callaway, J. C., Borgnis, E. L, Turner, R. E., Milan, C. S., Goodfriend, W., & Richmond, S. (2012b) "Wetland Sediment
Accumulation at Corte Madera Marsh and Muzzi Marsh". San Francisco Bay Conservation and Development
Commission.
Church, T. M., Sommerfield, C. K., Velinsky, D. J., Point, D., Benoit, C., Amouroux, D. & Donard, O. F. X. (2006)
Marsh sediments as records of sedimentation, eutrophication and metal pollution in the urban Delaware Estuary.
Marine Chemistry 102(1-2): 72-95.
Craft, C. B., & Richardson, C. J. (1998) Recent and long-term organic soil accretion and nutrient accumulation in the
Everglades. Soil Science Society of America Journal 62(3): 834-843.
Crooks, S., Rybczyk, J., O'Connell, K., Devier, D.L., Poppe, K., Emmett-Mattox, S. (2014) Coastal Blue Carbon
Opportunity Assessment for the Snohomish Estuary: The Climate Benefits of Estuary Restoration. Report by
Environmental Science Associates, Western Washington University, EarthCorps, and Restore America's Estuaries.
Hussein, A. H., Rabenhorst, M. C. & Tucker, M. L. (2004) Modeling of carbon sequestration in coastal marsh soils.
Soil Science Society of America Journal 68(5): 1786-1795.
IPCC (2019) Refinement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories, Volume 4:
86 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Agriculture, Forestry, and Other Land Use. Calvo Buendia, E., Tanabe K., Kranjc, A., Baasansuren, J., Fukuda, M.,
Ngarize, S., Osako, A., Pyrozhenko, Y., Shermanau, P., & Federici, S. (eds). IPCC, Switzerland.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. Prepared by the National Greenhouse
Gas Inventories Programme, H.S.Eggleston, L Buendia, K. Miwa, T. Ngara & K. Tanabe (eds). IGES, Japan.
IPCC (2014) 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands.
Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds.). Published: IPCC,
Switzerland.
IPCC (2003) Good Practice Guidance for Land Use, Land-Use Change and Forestry. LUCF Sector Good Practice
Guidance, Chapter 3. Penman, J., Gytarsky, M., Hiraishi, T., Krug, T., Kruger, D., Pipatti, R., Buendia, L., Miwa, K.,
Ngara, T., Tanabe, K. & F. Wagner (eds). Institute of Global Environmental Strategies (IGES), on behalf of the
Intergovernmental Panel on Climate Change (IPCC): Hayama, Japan.
IPCC (2000) Good Practice Guidance and Uncertainty Management in National Greenhouse Gas Inventories.
Quantifying Uncertainties in Practice, Chapter 6. Penman, J and Kruger, D and Galbally, I and Hiraishi, T and Nyenzi,
B and Emmanuel, S and Buendia, L and Hoppaus, R and Martinsen, T and Meijer, J and Miwa, K and Tanabe, K
(eds). Institute of Global Environmental Strategies (IGES), on behalf of the Intergovernmental Panel on Climate
Change (IPCC): Hayama, Japan.
Kearney, M. S. & Stevenson, J. C. (1991) Island land loss and marsh vertical accretion rate evidence for historical
sea-level changes in Chesapeake Bay. Journal of Coastal Research 7(2): 403-415.
Koster, D., Lichter, J., Lea, P. D., & Nurse, A. (2007) Historical eutrophication in a river-estuary complex in mid-
coast Maine. Ecological Applications 17(3): 765-778.
Lu, M & Megonigal, J.P. (2017) Final Report for RAE Baseline Assessment Project. Memo to Silvestrum Climate
Associates by Smithsonian Environmental Research Center, Maryland.
Lynch, J. C., Sedimentation and nutrient accumulation in mangrove ecosystems of the Gulf of Mexico, M.S. thesis,
Univ. of Southwestern Louisiana, Lafayette, La., 1989.
Marchio, D.A., Savarese, M., Bovard, B., & Mitsch, W.J. (2016) Carbon sequestration and sedimentation in
mangrove swamps influenced by hydrogeomorphic conditions and urbanization in Southwest Florida. Forests 7:
116-135.
McCombs, J.W., Herold, N.D., Burkhalter, S.G. and Robinson C.J., (2016) Accuracy Assessment of NOAA Coastal
Change Analysis Program 2006-2010 Land Cover and Land Cover Change Data. Photogrammetric Engineering &
Remote Sensing. 82:711-718.
Merrill, J. Z. (1999) Tidal Freshwater Marshes as Nutrient Sinks: particulate Nutrient Burial and Denitrification.
Ph.D. Dissertation, University of Maryland, College Park, MD, 342pp.
National Oceanic and Atmospheric Administration, Office for Coastal Management (2020) Coastal Change Analysis
Program (C-CAP) Regional Land Cover. Charleston, SC: NOAA Office for Coastal Management. Accessed October
2020 at .
Noe, G. B., Hupp, C. R., Bernhardt, C. E., & Krauss, K. W. (2016) Contemporary deposition and long-term
accumulation of sediment and nutrients by tidal freshwater forested wetlands impacted by sea level rise. Estuaries
and Coasts 39(4): 1006-1019.
Orson, R. A., Simpson, R. L, & Good, R. E. (1990) Rates of sediment accumulation in a tidal freshwater marsh.
Journal of Sedimentary Research 60(6): 859-869.
Orson, R., Warren, R. & Niering, W. (1998) Interpreting sea level rise and rates of vertical marsh accretion in a
southern New England tidal salt marsh. Estuarine, Coastal and Shelf Science 47(4): 419-429.
Roman, C., Peck, J., Allen, J., King, J. & Appleby, P. (1997) Accretion of a New England (USA) salt marsh in response
to inlet migration, storms, and sea-level rise. Estuarine, Coastal and Shelf Science 45(6): 717-727.
References and Abbreviations 87
-------
Villa, J. A. & Mitsch W. J. (2015) "Carbon sequestration in different wetland plant communities of Southwest
Florida". International Journal for Biodiversity Science, Ecosystems Services and Management 11:17-28.
Weston, N. B., Neubauer, S. C., Velinsky, D. J., & Vile, M. A. (2014) Net ecosystem carbon exchange and the
greenhouse gas balance of tidal marshes along an estuarine salinity gradient. Biogeochemistry 120:163-189.
Land Converted to Wetlands: Land Converted to Flooded Land
IPCC (2003) Good Practice Guidance for Land Use, Land-Use Change and Forestry. LUCF Sector Good Practice
Guidance, Chapter 3. Penman, J., Gytarsky, M., Hiraishi, T., Krug, T., Kruger, D., Pipatti, R., Buendia, L., Miwa, K.,
Ngara, T., Tanabe, K. and Wagner, F. (eds). Institute of Global Environmental Strategies (IGES), on behalf of the
Intergovernmental Panel on Climate Change (IPCC): Hayama, Japan.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. Prepared by the National Greenhouse
Gas Inventories Programme, H.S.Eggleston, L Buendia, K. Miwa, T. Ngara & K. Tanabe (eds). IGES, Japan.
IPCC. (2013) 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands.
Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds). In: IPCC,
Switzerland.
IPCC (2019) 2019 Refinement to the 2006 Guidelines for National Greenhouse Gas Inventories. Wetlands, Chapter
7. Lovelock, C. E., Evans, C., Barros, N., Prairie, Y. T., Aim, J., Bastviken, D., Beaulieu, J. J., Garneau, M., Harby, A.,
Harrison, J. A., Pare, David, Raadal, Hanne Lerche, Sherman, B., Zhang, Chengyi, Ogle, S. M.
Lehner B, Reidy Liermann C, Revenga C, Vorosmarty C, Fekete B, Crouzet P, Doll P, et al. (2011b) Global Reservoir
and Dam Database, Version 1 (GRanDvl): Dams, Revision 01. In: Palisades, NY: NASA Socioeconomic Data and
Applications Center (SEDAC).
Prairie, Y. T., et al. (2017) The GHG ReservoirTool (G-res) User guide. UNESCO/IHA research project on the GHG
status of freshwater reservoirs. Joint publication of the UNESCO Chair in Global Environmental Change and the
International Hydropower Association: 38.
Settlements Remaining Settlements: Soil Carbon Stock
Changes
Armentano, T. V., and E.S. Menges (1986) Patterns of change in the carbon balance of organic soil-wetlands of the
temperate zone. Journal of Ecology 74: 755-774.
Brady, N.C. and R.R. Weil (1999) The Nature and Properties of Soils. Prentice Hall. Upper Saddle River, NJ, 881.
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L, Barnes, C., Herold, N., and J. Wickham. (2011) Completion of
the 2006 National Land Cover Database for the Conterminous United States, PE&RS 77(9):858-864.
Homer, C., J. Dewitz, J. Fry, M. Coan, N. Hossain, C. Larson, N. Herold, A. McKerrow, J.N. VanDriel and J. Wickham.
(2007) Completion of the 2001 National Land Cover Database for the Conterminous United States.
Photogrammetric Engineering and Remote Sensing 73(4): 337-341.
Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-
Representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing
81(5):345-354.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
88 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
NRCS (1999) Soil Taxonomy: A basic system of soil classification for making and interpreting soil surveys, 2nd
Edition. Agricultural Handbook Number 436, Natural Resources Conservation Service, U.S. Department of
Agriculture, Washington, D.C.
Nusser, S.M. and J.J. Goebel (1997) The national resources inventory: a long-term multi-resource monitoring
programme. Environmental and Ecological Statistics 4:181-204.
Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) Uncertainty in estimating land use and management
impacts on soil organic carbon storage for U.S. agroecosystems between 1982 and 1997. Global Change Biology
9:1521-1542.
Soil Survey Staff (2011) State Soil Geographic (STATSGO) Database for State. Natural Resources Conservation
Service, United States Department of Agriculture. Available online at:
http://www.ncgc.nrcs.usda.gov/products/datasets/statsgo/index.html.
USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory, Natural Resources Conservation Service,
Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd 1422028.pdf.
Yang, L, Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J., Fry, J., Funk, M.,
Granneman, B., Liknes, G. C., Rigge, M. & Xian, G. (2018) A new generation of the United States National Land
Cover Database: Requirements, research priorities, design, and implementation strategies. ISPRS Journal of
Photogrammetry and Remote Sensing 146:108-123.
Settlements Remaining Settlements: Changes in Carbon Stocks
in Settlement Trees
deVries, R.E. (1987) A Preliminary Investigation of the Growth and Longevity of Trees in Central Park. M.S. thesis,
Rutgers University, New Brunswick, NJ.
Fleming, L.E. (1988) Growth Estimation of Street Trees in Central New Jersey. M.S. thesis, Rutgers University, New
Brunswick, NJ.
Frelich, L.E. (1992) Predicting Dimensional Relationships for Twin Cities Shade Trees. University of Minnesota,
Department of Forest Resources, St. Paul, MN, p. 33.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
MRLC (2013) National Land Cover Database 2001 (NLCD2001). Available online at:
http://www.mrlc.gov/nlcd2001.php. Accessed August 2013.
Nowak, D.J. (1986) Silvics of an Urban Tree Species: Norway maple (Acer platanoides L.). M.S. thesis, College of
Environmental Science and Forestry, State University of New York, Syracuse, NY.
Nowak, D.J. (1994) Atmospheric carbon dioxide reduction by Chicago's urban forest. In: Chicago's Urban Forest
Ecosystem: Results of the Chicago Urban Forest Climate Project. E.G. McPherson, D.J. Nowak, and R.A. Rowntree
(eds.). General Technical Report NE-186. U.S. Department of Agriculture Forest Service, Radnor, PA. pp. 83-94.
Nowak, D.J. (2012) Contrasting natural regeneration and tree planting in 14 North American cities. Urban Forestry
and Urban Greening. 11: 374-382.
Nowak, D.J. and D.E. Crane (2002) Carbon storage and sequestration by urban trees in the United States.
Environmental Pollution 116(3):381-389.
Nowak, D.J. and E. Greenfield (2010) Evaluating the National Land Cover Database tree canopy and impervious
cover estimates across the conterminous United States: A comparison with photo-interpreted estimates.
Environmental Management. 46: 378-390.
References and Abbreviations 89
-------
Nowak, D.J. and E.J. Greenfield (2018a) U.S. urban forest statistics, values and projections. Journal of Forestry.
116(2): 164—177
Nowak, D.J. and E.J. Greenfield (2018b) Declining urban and community tree cover in the United States. Urban
Forestry and Urban Greening. 32:32-55.
Nowak, D.J., D.E. Crane, J.C. Stevens, and M. Ibarra (2002) Brooklyn's Urban Forest. General Technical Report NE-
290. U.S. Department of Agriculture Forest Service, Newtown Square, PA.
Nowak, D.J., R.E. Hoehn, D.E. Crane, J.C. Stevens, J.T. Walton, and J. Bond (2008) Aground-based method of
assessing urban forest structure and ecosystem services. Arboric. Urb. For. 34(6): 347-358.
Nowak, D.J., E.J. Greenfield, R.E. Hoehn, and E. Lapoint (2013) Carbon storage and sequestration by trees in urban
and community areas of the United States." Environmental Pollution 178: 229-236.
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,
2014. USDA Forest Service, Northern Research Station Resources Bulletin. NRS-100. Newtown Square, PA. 55 p.
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.
Brandeis, T.W. Lister (2017) Houston's Urban Forest, 2015. USDA Forest Service, Southern Research Station
Resources Bulletin. SRS-211. Newtown Square, PA. 91 p.
Smith, W.B. and S.R. Shifley (1984) Diameter Growth, Survival, and Volume Estimates for Trees in Indiana and
Illinois. Research Paper NC-257. North Central Forest Experiment Station, U.S. Department of Agriculture Forest
Service, St. Paul, MN.
U.S. Department of Interior (2018) National Land Cover Database 2011 (NLCD2011). Accessed online August 16,
2018. Available online at: https://www.mrlc.gov/nlcdll leg.php.
Settlements Remaining Settlements: N20 Emissions from Soils
Brakebill, J.W. and Gronberg, J.M. (2017) County-Level Estimates of Nitrogen and Phosphorus from Commercial
Fertilizer for the Conterminous United States, 1987-2012. U.S. Geological Survey,
https://doi.org/10.5066/F7H41PKX.
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
Soil Survey Staff (2016) State Soil Geographic (STATSGO) Database for State. Natural Resources Conservation
Service, United States Department of Agriculture. Available online at:
http://www.ncgc.nrcs.usda.gov/products/datasets/statsgo/index.html.
USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory, Natural Resources Conservation Service,
Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd 1422028.pdf.
Settlements Remaining Settlements: Changes in Yard
Trimmings and Food Scrap Carbon Stocks in Landfills
Barlaz, M.A. (2008) "Re: Corrections to Previously Published Carbon Storage Factors." Memorandum to Randall
Freed, ICF International. February 28, 2008.
Barlaz, M.A. (2005) "Decomposition of Leaves in Simulated Landfill." Letter report to Randall Freed, ICF Consulting.
June 29, 2005.
90 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Barlaz, M.A. (1998) "Carbon Storage during Biodegradation of Municipal Solid Waste Components in Laboratory-
Scale Landfills." Global Biogeochemical Cycles 12:373-380.
De la Cruz, F.B. and M.A. Barlaz (2010) "Estimation of Waste Component Specific Landfill Decay Rates Using
Laboratory-Scale Decomposition Data" Environmental Science & Technology 44:4722-4728.
Eleazer, W.E., W.S. Odle, Y. Wang, and M.A. Barlaz (1997) "Biodegradability of Municipal Solid Waste Components
in Laboratory-Scale Landfills." Environmental Science & Technology 31:911-917.
EPA (2020) Advancing Sustainable Materials Management: Facts and Figures 2018. U.S. Environmental Protection
Agency, Office of Solid Waste and Emergency Response, Washington, D.C. Available online at
https://www.epa.gov/smm/advancing-sustainable-materials-management-facts-and-figures-report.
EPA (2019) Advancing Sustainable Materials Management: Facts and Figures. U.S. Environmental Protection
Agency, Office of Solid Waste and Emergency Response, Washington, D.C. Available online at
https://www.epa.gov/smm/advancing-sustainable-materials-management-facts-and-figures-report.
EPA (2016) Municipal Solid Waste Generation, Recycling, and Disposal in the United States: Facts and Figures. U.S.
Environmental Protection Agency, Office of Solid Waste and Emergency Response, Washington, D.C. Available
online at https://archive.epa.gov/epawaste/nonhaz/municipal/web/html/msw99.html.
EPA (1995) Compilation of Air Pollutant Emission Factors. U.S. Environmental Protection Agency, Office of Air
Quality Planning and Standards, Research Triangle Park, NC. AP-42 Fifth Edition. Available online at
http://www3.epa.gov/ttnchiel/ap42/.
EPA (1991) Characterization of Municipal Solid Waste in the United States: 1990 Update. U.S. Environmental
Protection Agency, Office of Solid Waste and Emergency Response, Washington, D.C. EPA/530-SW-90-042.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
IPCC (2003) Good Practice Guidance for Land Use, Land-Use Change, and Forestry. The Intergovernmental Panel on
Climate Change, National Greenhouse Gas Inventories Programme, J. Penman et al. (eds.). Available online at
http://www.ipcc-nggip.iges.or.ip/public/gpglulucf/gpglulucf.htm.
Oshins, C. and D. Block (2000) "Feedstock Composition at Composting Sites." Biocycle 41(9):31-34.
Tchobanoglous, G., H. Theisen, and S.A. Vigil (1993) Integrated Solid Waste Management, 1st edition. McGraw-Hill,
NY. Cited by Barlaz (1998) "Carbon Storage during Biodegradation of Municipal Solid Waste Components in
Laboratory-Scale Landfills." Global Biogeochemical Cycles 12:373-380.
Land Converted to Settlements
Birdsey, R. (1996) "Carbon Storage for Major Forest Types and Regions in the Conterminous United States." In R.N.
Sampson and D. Hair, (eds.). Forest and Global Change, Volume 2: Forest Management Opportunities for
Mitigating Carbon Emissions. American Forests. Washington, D.C., 1-26 and 261-379 (appendices 262 and 263).
Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer. Domke, G.M.,
Perry, C.H., Walters, B.F., Woodall, C.W., and Smith, J.E. (2016) A framework for estimating litter carbon stocks in
forests of the United States. Science of the Total Environment 557-558: 469-478.
Domke, G.M., J.E. Smith, and C.W. Woodall. (2011) Accounting for density reduction and structural loss in standing
dead trees: Implications for forest biomass and carbon stock estimates in the United States. Carbon Balance and
Management. 6:14.
Domke, G.M., Woodall, C.W., Walters, B.F., Smith, J.E. (2013) From models to measurements: comparing down
dead wood carbon stock estimates in the U.S. forest inventory. PLoS ONE 8(3): e59949.
References and Abbreviations 91
-------
Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., and Smith, J.E. (2016) A framework for estimating litter
carbon stocks in forests of the United States. Science of the Total Environment 557-558: 469-478.
Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L, Barnes, C., Herold, N., and Wickham, J. (2011) Completion of
the 2006 National Land Cover Database for the Conterminous United States, PE&RS, Vol. 77(9):858-864.
Harmon, M.E., C.W. Woodall, B. Fasth, J. Sexton, M. Yatkov. (2011) Differences between standing and downed
dead tree wood density reduction factors: A comparison across decay classes and tree species. Res. Paper. NRS-15.
Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 40 p.
Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A., VanDriel, J.N., and
Wickham, J. (2007) Completion of the 2001 National Land Cover Database for the Conterminous United States.
Photogrammetric Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-
Representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, v.
81, no. 5, p. 345-354.
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change, H.S. Eggleston, L. Buendia, K. Miwa, T
Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
Jenkins, J.C., D.C. Chojnacky, L.S. Heath, and R.A. Birdsey (2003) "National-scale biomass estimators for United
States tree species." Forest Science 49(l):12-35.
Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and management
impacts on soil organic carbon storage for U.S. agroecosystems between 1982 and 1997." Global Change Biology
9:1521-1542.
Ogle, S.M., F.J. Breidt, and K. Paustian (2006) "Bias and variance in model results due to spatial scaling of
measurements for parameterization in regional assessments." Global Change Biology 12:516-523.
Schimel, D.S. (1995) "Terrestrial ecosystems and the carbon cycle." Global Change Biology 1: 77-91.
Smith, J.E.; Heath, L.S.; Skog, K.E.; Birdsey, R.A. (2006) Methods for calculating forest ecosystem and harvested
carbon with standard estimates for forest types of the United States. Gen. Tech. Rep. NE-343. Newtown Square,
PA: U.S. Department of Agriculture, Forest Service, Northeastern Research Station. 216 p.
Tubiello, F. N., et al. (2015) "The Contribution of Agriculture, Forestry and other Land Use activities to Global
Warming, 1990-2012." Global Change Biology 21:2655-2660.
USDA Forest Service (2020) Forest Inventory and Analysis National Program: FIA Data Mart. U.S. Department of
Agriculture Forest Service. Washington, D.C. Available online at: http://apps.fs.fed.us/fiadb-
downloads/datamart.html. Accessed on 10 October 2020.
USDA-NRCS (2018) Summary Report: 2015 National Resources Inventory, Natural Resources Conservation Service,
Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
https://www.nrcs.usda.gov/lnternet/FSE DOCUMENTS/nrcseprd 1422028.pdf.
USDA-NRCS (1997) "National Soil Survey Laboratory Characterization Data," Digital Data, Natural Resources
Conservation Service, U.S. Department of Agriculture. Lincoln, NE.
Woodall, C.W., L.S. Heath, G.M. Domke, and M.C. Nichols. (2011) Methods and equations for estimating
aboveground volume, biomass, and carbon for trees in the U.S. forest inventory, 2010. Gen. Tech. Rep. NRS-88.
Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 30 p.
Woodall, C.W., and V.J. Monleon (2008) Sampling protocol, estimation, and analysis procedures for the down
woody materials indicator of the FIA program. Gen. Tech. Rep. NRS-22. Newtown Square, PA: U.S. Department of
Agriculture, Forest Service, Northern Research Station. 68 p.
92 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Yang, L, Jin, S., Danielson, P., Homer, C., Gass, L., Bender, S. M., Case, A., Costello, C., Dewitz, J., Fry, J., Funk, M.,
Granneman, B., Liknes, G. C., Rigge, M. & Xian, G. (2018) A new generation of the United States National Land
Cover Database: Requirements, research priorities, design, and implementation strategies. ISPRS Journal of
Photogrammetry and Remote Sensing 146:108-123
Waste
Landfills
40 CFR Part 60, Subpart WWW (2005) Standards of Performance for Municipal Solid Waste Landfills, 60.750-
60.759, Code of Federal Regulations, Title 40. Available online at: https://www.ecfr.gov/current/title-40/chapter-
l/subchapter-C/part-60/subpart-WWW.
40 CFR Part 258, Subtitle D of RCRA (2012) Criteria for Municipal Solid Waste Landfills, 258.1—258.75, Code of
Federal Regulations, Title 40. Available online at: https://www.ecfr.gov/cgi-bin/text-idx?node=pt40.25.258.
ATSDR (2001). Chapter 2: Landfill Gas Basics. In Landfill Gas Primer - An Overview for Environmental Health
Professionals. Figure 2-1, pp. 5-6. https://www.atsdr.cdc.gov/HAC/landfill/PDFs/Landfill_2001_ch2mod.pdf.
BioCycle (2010) "The State of Garbage in America" By L. Arsova, R. Van Haaren, N. Goldstein, S. Kaufman, and N.
Themelis. BioCycle. December 2010. Available online at: https://www.biocycle.net/2010/10/26/the-state-of-
garbage-in america-4/.
BioCycle (2006) "The State of Garbage in America" By N. Goldstein, S. Kaufman, N. Themelis, and J. Thompson Jr.
BioCycle. April 2006. Available online at: https://www.biocycle.net/2006/04/21/the-state-of-garbage-in-america-
IL-
Bronstein, K., Coburn, J., and R. Schmeltz (2012) "Understanding the EPA's Inventory of U.S. Greenhouse Gas
Emissions and Sinks and Mandatory GHG Reporting Program for Landfills: Methodologies, Uncertainties,
Improvements and Deferrals." Prepared for the U.S. EPA International Emissions Inventory Conference, August
2012, Tampa, Florida. Available online at:
https://www3.epa.gov/ttnchiel/conference/ei20/session3/kbronstein.pdf.
Business for Social Responsibility (BSR) (2014). Analysis of U.S. Food Waste Among Food Manufacturers, Retailers,
and Restaurants. Available online at: http://www.foodwastealliance.org/wp-
content/uploads/2014/ll/FWRA BSR Tier3 FINAL.pdf.
BSR (2013) Analysis of U.S. Food Waste Among Food Manufacturers, Retailers, and Restaurants. Available online
at: http://www.foodwastealliance.org/wp-content/uploads/2013/06/FWRA BSR Tier2 FINAL.pdf.
Czepiel, P., B. Mosher, P. Crill, and R. Harriss (1996) "Quantifying the Effect of Oxidation on Landfill Methane
Emissions." Journal of Geophysical Research, 101(D11):16721-16730. Dou, Z.; Ferguson, J. D.; Galligan, D. T.; Kelly,
A. M.; Finn, S. T.; Giegengack, R. (2016) "Assessing U.S. food wastage and opportunities for reduction. Global Food
Security Volume 8, March 2016, Pages 19-26. https://doi.Org/10.1016/i.gfs.2016.02.001.
EIA (2007) Voluntary Greenhouse Gas Reports for EIA Form 1605B (Reporting Year 2006). Available online at:
https://www. eia.gov/environ ment/pdfpages/0608sl2009)index.php.
EPA (2021a) Greenhouse Gas Reporting Program (GHGRP). 2020 Amazon S3 Data. Subpart HH: Municipal Solid
Waste Landfills and Subpart TT: Industrial Waste Landfills. Accessed on September 19, 2021.
EPA (2021b) Landfill Methane Outreach Program (LMOP). 2021 Landfill and Project Level Data. March 2021.
Available online at: https://www.epa.gov/lmop/landfill-gas-energy-proiect-data.
References and Abbreviations 93
-------
EPA (2020a) Wasted Food Measurement Methodology Scoping Memo. July 2020. Available online at
https://www.epa.gov/sites/production/files/202Q-
06/documents/food measurement methodology scoping memo-6-18-2Q.pdf.
EPA (2020b) Advancing Sustainable Materials Management: Facts and Figures 2018. December 2020. Available
online at: https://www.epa.gov/sites/production/files/202Q-ll/documents/2018 tables and figures fnl 5Q8.pdf.
EPA (2019) Advancing Sustainable Materials Management: Facts and Figures 2016 and 2017. November 2019.
Available online at: https://www.epa.gov/sites/production/files/2019-
ll/documents/2016 and 2017 facts and figures data tables Q.pdf.
EPA (2018) Advancing Sustainable Materials Management: Facts and Figures 2015. July 2018. Available online at:
https://www.epa.gov/sites/production/files/2018-
07/documents/smm 2015 tables and figures 07252018 fnl 508 Q.pdf.
EPA (2016a) Industrial and Construction and Demolition Landfills. Available online at:
https://www.epa.gov/landfills/industrial-and-construction-and-demolition-cd-landfills.
EPA (2016b) Advancing Sustainable Materials Management: Facts and Figures 2014. December 2016. Available
online at: https://www.epa.gov/sites/production/files/2016-ll/documents/2014 smm tablesfigures 508.pdf.
EPA (2014) Advancing Sustainable Materials Management: Facts and Figures 2014. February 2014. Available online
at: https://www.epa.gov/sites/production/files/2015-09/documents/2012 msw dat tbls.pdf.
EPA (2008) Compilation of Air Pollution Emission Factors, Publication AP-42, Draft Section 2.4 Municipal Solid
Waste Landfills. October 2008.
EPA (1993) Anthropogenic Methane Emissions in the United States, Estimates for 1990: Report to Congress, U.S.
Environmental Protection Agency, Office of Air and Radiation. Washington, D.C. EPA/430-R-93-003. April 1993.
EPA (1988) National Survey of Solid Waste (Municipal) Landfill Facilities, U.S. Environmental Protection Agency.
Washington, D.C. EPA/530-SW-88-011. September 1988.
EREF (The Environmental Research & Education Foundation) (2016). Municipal Solid Waste Management in the
United States: 2010 & 2013.
ERG (2021) Production Data Supplied by ERG for 1990-2020 for Pulp and Paper, Fruits and Vegetables, and Meat.
June 29, 2021.
Food Waste Reduction Alliance (FWRA) (2016) Analysis of U.S. Food Waste Among Food Manufacturers, Retailers,
and Restaurants. Available online at: https://foodwastealliance.org/wp-content/uploads/2020/Q5/FWRA-Food-
Waste-Survey-2016-Report Final.pdf.
Intergovernmental Panel on Climate Change (IPCC) (2019) 2019 Refinement to the 2006 IPCC Guidelines for
National Greenhouse Gas Inventories. Available online at https://www.ipcc.ch/report/2019-refinement-to-the-
2006-ipcc-guidelines-for-national-greenhouse-gas-inventories/.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Mancinelli, R. and C. McKay (1985) "Methane-Oxidizing Bacteria in Sanitary Landfills." Proc. First Symposium on
Biotechnical Advances in Processing Municipal Wastes for Fuels and Chemicals, Minneapolis, MN, 437-450. August.
RTI (2018a) Methodological changes to the scale-up factor used to estimate emissions from municipal solid waste
landfills in the Inventory. Memorandum prepared by K. Bronstein and M. McGrath for R. Schmeltz (EPA). March 22,
2018.
RTI (2018b) Comparison of industrial waste data reported under Subpart TT and the Solid Waste chapter of the
GHG Inventory. Memorandum prepared by K. Bronstein, B. Jackson, and M. McGrath for R, Schmeltz (EPA).
October 12, 2018.
94 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
RTI (2017) Methodological changes to the methane emissions from municipal solid waste landfills as reflected in
the public review draft of the 1990-2015 Inventory. Memorandum prepared by K. Bronstein and M. McGrath for R.
Schmeltz (EPA). March 31, 2017.
RTI (2011) Updated Research on Methane Oxidation in Landfills. Memorandum prepared by K. Weitz (RTI) for R.
Schmeltz (EPA). January 14, 2011.
Waste Business Journal (WBJ) (2016) Directory of Waste Processing & Disposal Sites 2016.
WBJ (2010) Directory of Waste Processing & Disposal Sites 2010.
WTO (2017) "China's import ban on solid waste queried at import licensing meeting". World Trade Organization.
Published October 3, 2017. Available online at:
https://www.wto.org/english/news e/news!7 e/impl 03octl7 e.htro.
Wastewater Treatment and Discharge
AF&PA (2020) "2020 AF&PA Sustainability Report: Advancing the sustainability of an essential industry." American
Forest & Paper Association. Available online at: https://www.afandpa.org/sites/default/files/2021-07/2020 AF-
PA-Sustainability-Report.pdf. Access June 2021.
AF&PA (2018) "2018 AF&PA Sustainability Report: Advancing U.S. Paper and Wood Products Industry Sustainability
Performance." American Forest & Paper Association. Available online at: http://sustainability.afandpa.org/wp-
content/yploads/2013/06/2013SustainabilityReport PAGES.pdf. Accessed July 2019.
AF&PA (2016) "2016 AF&PA Sustainability Report: Advancing U.S. Paper and Wood Products Industry Sustainability
Performance." American Forest & Paper Association.
AF&PA (2014) "2014 AF&PA Sustainability Report." American Forest & Paper Association.
Beecher et al. (2007) "A National Biosolids Regulation, Quality, End Use & Disposal Survey, Preliminary Report."
Northeast Biosolids and Residuals Association, April 14, 2007. Available online at:
https://staticl.squarespace.eom/static/54806478e4b0dc44el698e88/t/5480c7a2e4b0787f2c73ad81/1417725858
575/NtlBiosldsRpt-AppD-FINAL.pdf. Accessed August 2021.
Beer Institute (2011) Brewers Almanac. Available online at: http://www.beerinstitute.org/multimedia/brewers-
almanac.
Benyahia, F., M. Abdulkarim, A. Embaby, and M. Rao. (2006) Refinery Wastewater Treatment: A true Technological
Challenge. Presented at the Seventh Annual U.A.E. University Research Conference.
BIER (2017) Beverage Industry Environmental Roundtable. 2016 Trends and Observations. Available online at:
https://www.bieroundtable.com/benchmarking-coeu. Accessed April 2018.
Brewers Association (2021) Statistics: Number of Breweries. Available online at:
https://www.brewersassociation.org/statistics-and-data/national-beer-stats/. Accessed August 2021.
Brewers Association (2017). 2016 Sustainability Benchmarking Update. Available online at:
https://www.brewersassociation.org/best-practices/sustainabilitv/sustainability-benchmarking-tools. Accessed
April 2018.
Brewers Association (2016a) 2015 Sustainability Benchmarking Report. Available online at:
https://www.brewersassociation.org/best-practices/sustainabilitv/sustainability-benchmarking-tools. Accessed
March 2018.
Brewers Association (2016b) Wastewater Management Guidance Manual. Available online at:
https://www.brewersassociation.org/educational-publications/wastewater-management-guidance-manual.
Accessed September 2017.
References and Abbreviations 95
-------
Cabrera (2017) "Pulp Mill Wastewater: Characteristics and Treatment." Biological Wastewater Treatment and
Resource Recovery. InTech. pp. 119-139.
CAST (1995) Council for Agricultural Science and Technology. Waste Management and Utilization in Food
Production and Processing. U.S.A. October 1995. ISBN 1-887383-02-6. Available online at: https://www.cast-
scierice.org/publicatiori/waste-management-arid-utilization-in food-production-arid-processing/.
Climate Action Reserve (CAR) (2011) Landfill Project Protocol V4.0, June 2011. Available online at:
http://www.climateactionreserve.org/how/protocols/us-landfill/.
Cooper (2018) Email correspondence. Geoff Cooper, Renewable Fuels Association to Kara Edquist, ERG. "Wet Mill
vs. Dry Mill Ethanol Production." May 18, 2018.
DOE (2013) U.S. Department of Energy Bioenergy Technologies Office. Biofuels Basics. Available online at:
http://energy.gov/eere/bioenergv/biofuels-basics. Accessed September 2013.
Donovan (1996) Siting an Ethanol Plant in the Northeast. C.T. Donovan Associates, Inc. Report presented to
Northeast Regional Biomass Program (NRBP). (April). Available online at:
https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.614.856&rep=repl&type=pdf. Accessed October
2006.
EIA (2021) Energy Information Administration. U.S. Refinery and Blender Net Production of Crude Oil and
Petroleum Products (Thousand Barrels). Available online at:
https://www.eia.gov/dnav/pet/pet pnp refp dc nus mbbl ro.htro. Accessed June 2021.
EPA (2019) Preliminary Effluent Guidelines Program Plan 14. EPA-821-R-19-005. Office of Water, U.S.
Environmental Protection Agency. Washington, DC. October 2019. Available online at:
https://www.epa.gov/sites/production/files/2019-10/documents/prelim-eg-pl * ' ; > ^ 319.pdf. Accessed July
2020.
EPA (2013) U.S. Environmental Protection Agency. Report on the Performance of Secondary Treatment
Technology. EPA-821-R-13-001. Office of Water, U.S. Environmental Protection Agency. Washington, D.C. March
2013. Available online at: https://www.epa.gov/sites/production/files/2015-
11/documents/npdes secondary treatment report march2013.pdf.
EPA (2012) U.S. Environmental Protection Agency. Clean Watersheds Needs Survey 2012 - Report to Congress. U.S.
Environmental Protection Agency, Office of Wastewater Management. Washington, D.C. Available online at:
https://www.epa.gOv/cwns/clean-watersheds-needs-survev-cwns-2012-report-and-data#access. Accessed
February 2016.
EPA (2008) U.S. Environmental Protection Agency. Clean Watersheds Needs Survey 2008 - Report to Congress. U.S.
Environmental Protection Agency, Office of Wastewater Management. Washington, D.C. Available online at:
https://www.epa.gov/cwns/clean-watersheds-needs-survev-cwns-2008-report-and-data. Accessed December
2015.
EPA (2004a) U.S. Environmental Protection Agency. Clean Watersheds Needs Survey 2004 - Report to Congress.
U.S. Environmental Protection Agency, Office of Wastewater Management. Washington, D.C. Available online at:
https://www.epa.gov/cwns/cleari-watersheds-rieeds-survev-cwris-report-corigress-2004.
EPA (2004b) Technical Development Document for the Final Effluent Limitations Guidelines and Standards for the
Meat and Poultry Products Point Source Category (40 CFR 432). Office of Water. EPA-821-R-04-011, Washington
DC, July.
EPA (2002) U.S. Environmental Protection Agency. Development Document for the Proposed Effluent Limitations
Guidelines and Standards for the Meat and Poultry Products Industry Point Source Category (40 CFR 432). EPA-
821-B-01-007. Office of Water, U.S. Environmental Protection Agency. Washington, D.C. January 2002.
EPA (2000) U.S. Environmental Protection Agency. Clean Watersheds Needs Survey 2000 - Report to Congress.
Office of Wastewater Management, U.S. Environmental Protection Agency. Washington, D.C. Available online at:
https://www.epa.gov/cwns/clean-watersheds-needs-survev-cwns-2000-report-and-data. Accessed July 2007.
96 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
EPA (1999) U.S. Environmental Protection Agency. Biosolids Generation, Use and Disposal in the United States.
Office of Solid Waste and Emergency Response, U.S. Environmental Protection Agency. Washington, D.C. EPA530-
R-99-009. September 1999.
EPA (1998a) U.S. Environment Protection Agency. How Wastewater Treatment Works...The Basics. EPA F98-002.
Office of Water. Washington D.C. May 1998. Available online at: https://www3.epa.gov/npdes/pubs/bastre.pdf.
EPA (1998b) U.S. Environmental Protection Agency. "AP-42 Compilation of Air Pollutant Emission Factors." Chapter
2.4, Table 2.4-3, page 2.4-13. Available online at: https://www.epa.gov/sites/default/files/2020-
10/documents/c02s04.pdf.
EPA (1997a) U.S. Environmental Protection Agency. Estimates of Global Greenhouse Gas Emissions from Industrial
and Domestic Wastewater Treatment. EPA-600/R-97-091. Office of Policy, Planning, and Evaluation, U.S.
Environmental Protection Agency. Washington, D.C. September 1997.
EPA (1997b) U.S. Environmental Protection Agency. Supplemental Technical Development Document for Effluent
Guidelines and Standards (Subparts B & E). EPA-821-R-97-011. Office of Water, U.S. Environmental Protection
Agency. Washington, D.C. October 1997.
EPA (1996) U.S. Environmental Protection Agency. 1996 Clean Water Needs Survey Report to Congress.
Assessment of Needs for Publicly Owned Wastewater Treatment Facilities, Correction of Combined Sewer
Overflows, and Management of Storm Water and Nonpoint Source Pollution in the United States. Office of
Wastewater Management, U.S. Environmental Protection Agency. Washington, D.C.
EPA (1993a) U.S. Environmental Protection Agency, "Anthropogenic Methane Emissions in the U.S.: Estimates for
1990, Report to Congress." Office of Air and Radiation, Washington, DC. April 1993.
EPA (1993b) U.S. Environmental Protection Agency. Development Document for the Proposed Effluent Limitations
Guidelines and Standards for the Pulp, Paper and Paperboard Point Source Category. EPA-821-R-93-019. Office of
Water, U.S. Environmental Protection Agency. Washington, D.C. October 1993.
EPA (1993c) Standards for the Use and Disposal of Sewage Sludge. 40 CFR Part 503.
EPA (1992) U.S. Environmental Protection Agency. Clean Watersheds Needs Survey 1992 - Report to Congress.
Office of Wastewater Management, U.S. Environmental Protection Agency. Washington, D.C.
EPA (1982) U.S. Environmental Protection Agency. Development Document for Effluent Limitations Guidelines and
standards for the Petroleum Refining. United States Environmental Protection Agency, Office of Water. EPA-440/1-
82-014. Washington D.C. October 1982.
EPA (1975) U.S. Environmental Protection Agency. Development Document for Interim Final and Proposed Effluent
Limitations Guidelines and New Source Performance Standards for the Fruits, Vegetables, and Specialties Segment
of the Canned and Preserved Fruits and Vegetables Point Source Category. United States Environmental Protection
Agency, Office of Water. EPA-440/1-75-046. Washington D.C. October 1975.
EPA (1974) U.S. Environmental Protection Agency. Development Document for Effluent Limitations Guidelines and
New Source Performance Standards for the Apple, Citrus, and Potato Processing Segment of the Canned and
Preserved Fruits and Vegetables Point Source Category. Office of Water, U.S. Environmental Protection Agency,
Washington, D.C. EPA-440/l-74-027-a. March 1974.
ERG (2021a) Revised Memorandum: Improvements to the 1990-2019 Wastewater Treatment and Discharge
Greenhouse Gas Inventory. March 2021.
ERG (2021b) Draft Memorandum: Improvements to the 1990-2020 Wastewater Treatment and Discharge
Greenhouse Gas Inventory. July 2021.
ERG (2019). Memorandum: Recommended Improvements to the 1990-2018 Wastewater Greenhouse Gas
Inventory. August 2019.
ERG (2018a) Memorandum: Updates to Domestic Wastewater BOD Generation per Capita. August 2018.
References and Abbreviations 97
-------
ERG (2018b) Memorandum: Inclusion of Wastewater Treatment Emissions from Breweries. July 2018.
ERG (2016) Revised Memorandum: Recommended Improvements to the 1990-2015 Wastewater Greenhouse Gas
Inventory. November 2016.
ERG (2013a) Memorandum: Revisions to Pulp and Paper Wastewater Inventory. October 2013.
ERG (2013b) Memorandum: Revisions to the Petroleum Refinery Wastewater Inventory. October 2013.
ERG (2008a) Memorandum: Planned Revisions of the Industrial Wastewater Inventory Emission Estimates for the
1990-2007 Inventory. 10 August 2008.
ERG (2008b) Memorandum: Estimation of Onsite Treatment at Industrial Facilities and Review of Wastewater
Characterization Data. 15 April 2008.
ERG (2006a) Memorandum: Recommended Improvements to EPA's Wastewater Inventory for Industrial
Wastewater. Prepared for Melissa Weitz, EPA. 11 August 2006.
ERG (2006b) Memorandum: Assessment of Greenhouse Gas Emissions from Wastewater Treatment of U.S. Ethanol
Production Wastewaters. Prepared for Melissa Weitz, EPA. 10 October 2006.
FAO (2021a) FAOSTAT-Forestry Database. Available online at: http://www.fao.Org/faostat/en/#data/FQ. Accessed
June 2021.
FAO (2021b) "Pulp and Paper Capacities Report." United States. From 1998 - 2003, 2000 - 2005, 2001 - 2006,
2002 - 2007, 2003 - 2008, 2010 - 2015, 2011 - 2016, 2012 - 2017, 2013 - 2018, 2014 - 2019, 2015 - 2020, 2016 -
2021, 2017 - 2022, 2018 - 2023, 2019 - 2024 reports. Available online at:
http://www.fao.org/forestry/statistics/30571/en/. Accessed June 2021.
FAO (2021c) FAOSTAT-Food Balance Sheets. Available online at: http://www.fao.Org/faostat/en/#data/FBS.
Accessed June 2021.
Foley et al. (2015) N2O and CHa Emission from Wastewater Collection and Treatment Systems: State of the Science
Report and Technical Report. GWRC Report Series. IWA Publishing, London, UK.
Great Lakes-Upper Mississippi River Board of State and Provincial Public Health and Environmental Managers.
(2004) Recommended Standards for Wastewater Facilities (Ten-State Standards).
Guisasola et al. (2008) Methane formation in sewer systems. Water Research 42(6-7): 1421-1430.
IPCC (2019) 2019 Refinement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change. [CalvoBuendia, E.,
Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M., Ngarize S., Osako, A., Pyrozhenko, Y., Shermanau, P. and
Federici, S. (eds)]. Switzerland.
IPCC (2014) 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands.
[Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds.)]. Published:
IPCC, Switzerland.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
Kenari et. al (2010) An Investigation on the Nitrogen Content of a Petroleum Refinery Wastewater and its Removal
by Biological Treatment. Journal of Environmental Health, Sciences, and Engineering. 7(1): 391-394.Leverenz, H.L,
G. Tchobanoglous, and J.L Darby (2010) "Evaluation of Greenhouse Gas Emissions from Septic Systems." Water
Environment Research Foundation. Alexandria, VA.
Lewis, A. (2019) Email correspondence. Ann Lewis, RFA to Kara Edquist, ERG. "Wet Mill vs Dry Mill Ethanol
Production." August 20, 2019.
98 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
Malmberg, B. (2018) Draft Pulp and Paper Information for Revision of EPA Inventory of U.S. Greenhouse Gas
Emissions and Sinks, Waste Chapter. National Council for Air and Stream Improvement, Inc. Prepared for Rachel
Schmeltz, EPA. June 13, 2018.
McFarland (2001) Biosolids Engineering, New York: McGraw-Hill, p. 2.12.
Merrick (1998) Wastewater Treatment Options for the Biomass-to-Ethanol Process. Report presented to National
Renewable Energy Laboratory (NREL). Merrick & Company. Subcontract No. AXE-8-18020-01. October 22,1998.
Metcalf & Eddy, Inc. (2014) Wastewater Engineering: Treatment and Resource Recovery, 5th ed. McGraw Hill
Publishing.
Metcalf & Eddy, Inc. (2003) Wastewater Engineering: Treatment, Disposal and Reuse, 4th ed. McGraw Hill
Publishing.
Nemerow, N.L and A. Dasgupta (1991) Industrial and Hazardous Waste Treatment. Van Nostrand Reinhold. NY.
ISBN 0-442-31934-7.
NRBP (2001) Northeast Regional Biomass Program. An Ethanol Production Guidebook for Northeast States.
Washington, D.C. (May 3).
Rendleman, C.M. and Shapouri, H. (2007) New Technologies in Ethanol Production. USDA Agricultural Economic
Report Number 842.
RFA (2021a) Renewable Fuels Association. Annual U.S. Fuel Ethanol Production. Available online at:
https://ethanolrfa.org/statistics/annual-ethanol-production. Accessed June 2021.
RFA (2021b) Renewable Fuels Association. Monthly Grain Use for U.S. Ethanol Production Report. Available online
at: https://ethanolrfa.org/statistics/feedstock-use-co-product-output. Accessed June 2021.
Ruocco (2006a) Email correspondence. Dr. Joe Ruocco, Phoenix Bio-Systems to Sarah Holman, ERG. "Capacity of
Bio-Methanators (Dry Milling)." October 6, 2006.
Ruocco (2006b) Email correspondence. Dr. Joe Ruocco, Phoenix Bio-Systems to Sarah Holman, ERG. "Capacity of
Bio-Methanators (Wet Milling)." October 16, 2006.
Short et al. (2017) Dissolved Methane in the Influent of Three Australian Wastewater Treatment Plants Fed by
Gravity Sewers. Sci Total Environ 599-600: 85-93.
Short et al. (2014) Municipal Gravity Sewers: an Unrecognised Source of Nitrous Oxide. Sci Total Environ 468-469:
211-218.
Stier, J. (2018) Personal communications between John Stier, Brewers Association Sustainability Mentor and Amie
Aguiar, ERG. Multiple dates.
Sullivan (SCS Engineers) (2010) The Importance of Landfill Gas Capture and Utilization in the U.S. Presented to
SWICS, April 6, 2010. Available online at: https://www.scsengineers.com/scs-white-papers/the-importance-of-
laridfill-gas-capture-arid-utilization-in-the-u-s/.
Sullivan (SCS Engineers) (2007) Current MSW Industry Position and State of the Practice on Methane Destruction
Efficiency in Flares, Turbines, and Engines. Presented to Solid Waste Industry for Climate Solutions (SWICS). July
2007. Available online at: https://www.scsengineers.com/wp-
content/uploads/2015/03/Sullivan LFG Destruction Efficiency White Paper.pdf.
TTB (2021) Alcohol and Tobacco Tax and Trade Bureau. Beer Statistics. Available online at:
https://www.ttb.gov/beer/beer-stats.shtml. Accessed June 2021.
UNFCCC (2012) CDM Methodological tool, Project emissions from flaring (Version 02.0.0). EB 68 Report. Annex 15.
Available online at: http://cdm.unfccc.int/methodologies/PAmethodologies/tools/am-tool-06-vl.pdf/history view.
U.S. Census Bureau (2021) International Database. Available online at: https://www.census.gov/data-
tools/demo/idb/#/trends?YR ANlM=2020&dashPages=DASH&FlPS SiNGLE=US£COUNTRY YEAR=2020&menu=tre
References and Abbreviations 99
-------
ndsViz&TREND RANGE=1990,2020&TREND STEP=1&TREND ADD YRS=&F1PS=AQ,GQ,CQ,RQ,US,VQ. Accessed
June 2021.
U.S. Census Bureau (2019) "American Housing Survey." Table 1A-4: Selected Equipment and Plumbing-All Housing
Units. From 1989,1991, 1993, 1995, 1997, 1999, 2001, 2003, 2005, 2007, and 2009 reports. Table C-04-AO
Plumbing, Water, and Sewage Disposal-All Occupied Units. From 2011, 2013, 2015, 2017, and 2019 reports.
Available online at http://www.census.gov/programs-surveys/ahs/data.html. Accessed August 2021.
U.S. Census Bureau (2013) "American Housing Survey." Table 1A-4: Selected Equipment and Plumbing-All Housing
Units. From 1989,1991, 1993, 1995, 1997, 1999, 2001, 2003, 2005, 2007, and 2009 reports. Table C-04-AO
Plumbing, Water, and Sewage Disposal-All Occupied Units. From 2011, and 2013 reports. Available online at
http://www.census.gov/programs-surveys/ahs/data.html. Accessed May 2020.
USDA (2021a) U.S. Department of Agriculture. National Agricultural Statistics Service. Washington, D.C. Available
online at: https://www.nass.usda.gov/Publications/ and https://quickstats.nass.usda.gov/. Accessed June 2021
USDA (2021b) U.S. Department of Agriculture. Economic Research Service. Nutrient Availability. Washington D.C.
Available online at: https://www.ers.usda.gov/data-products/food-availabilitv-per-capita-data-system/food-
availability-per-capita-data-system. Accessed June 2021.
USDA (2021c) U.S. Department of Agriculture. National Agricultural Statistics Service. Vegetables 2019 Summary.
Available online at: https://usda.library.cornell.edu/concern/publications/02870v86p?locale=en. Accessed June
2021.
U.S. Poultry (2006) Email correspondence. John Starkey, USPOULTRY to D. Bartram, ERG. 30 August 2006.
White and Johnson (2003) White, P.J. and Johnson, L.A. Editors. Corn: Chemistry and Technology. 2nd ed. AACC
Monograph Series. American Association of Cereal Chemists. St. Paul, MN.
World Bank (1999) Pollution Prevention and Abatement Handbook 1998, Toward Cleaner Production. The
International Bank for Reconstruction and Development/The WORLDBANK. 1818 H Street, N.W. Washington, DC.
20433, USA. ISBN 0-8213-3638-X.
Composting
BioCycle (2018a) Organic Waste Bans and Recycling Laws to Tackle Food Waste. Prepared by E. Broad Lieb, K.
Sandson, L. Macaluso, and C. Mansell. Available online at: https://www.biocycle.net/2018/09/ll/organic-waste-
bans-recycling-laws-tackle-food-waste/.
BioCycle (2018b). State Food Waste Recycling Data Collection, Reporting Analysis. Prepared by Nora Goldstein.
Available online at: http://compostcolab.wpengine.com/wp-content/uploads/2018/ll/State-Food-Waste-
Recycling-Data-Collection-Reporting-Analysis.pdf.
BioCycle (2010) The State of Garbage in America. Prepared by Rob van Haaren, Nickolas Themelis and Nora
Goldstein. Available online at http://www.biocvcle.net/images/art/1010/bcl01016 s.pdf.
BioCycle (2017) The State of Organics Recycling in the U.S. Prepared by Nora Goldstein. Available online at
http://www.biocycle.net/17 10 06 1/0001/BioCycle StateQfOrganicsUS.pdf.
Cornell Composting (1996). Monitoring Compost Moisture. Cornell Waste Management Institute. Available online
at: http://compost.css.cornell.edu/monitor/monitormoisture.html.
Cornell Waste Management Institute (2007) The Science of Composting. Available online at
http://cwmi.css.cornell.edu/chapterl.pdf.
EPA (2020) Advancing Sustainable Materials Management: 2018 Tables and Figures. Office of Solid Waste and
Emergency Response, U.S. Environmental Protection Agency, Washington, D.C. Available online at:
https://www.epa.gov/sites/default/files/2021-01/documents/2018 tables and figures dec 2020 fnl 508.pdf.
100 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
EPA (2018) Advancing Sustainable Materials Management: 2015 Tables and Figures. Office of Solid Waste and
Emergency Response, U.S. Environmental Protection Agency, Washington, D.C. Available online at
https://www.epa.gov/sites/production/files/2018-
07/documents/smm 2015 tables and figures 07252018 fnl 508 O.pdf.
EPA (2016) Advancing Sustainable Materials Management: Facts and Figures 2014. Office of Solid Waste and
Emergency Response, U.S. Environmental Protection Agency, Washington, D.C. Available online at
https://www.epa.gov/sites/production/files/2016-ll/documents/2014 smm tablesfigures 508.pdf.
EPA (2014) Municipal Solid Waste in the United States: 2012 Facts and Figures. Office of Solid Waste and
Emergency Response, U.S. Environmental Protection Agency, Washington, D.C. Available online at
https://www.epa.gov/sites/default/files/2015-09/documents/2012 msw fs.pdf.
Harvard Law School and Center for EcoTechnology (CET) (2019) Bans and Beyond: Designing and Implementing
Organic Waste Bans and Mandatory Organics Recycling Laws. Prepared by Katie Sandson and Emily Broad Leib,
Harvard Law School Food Law and Policy Clinic, with input from Lorenzo Macaluso and Coryanne Mansell, Center
for EcoTechnology (CET). Available online at https://wastedfood.cetonline.org/wp-
content/uploads/2019/07/Harvard-Law-School-FLPC-Center-for-EcoTechnology-CET-Organic-Waste-Bans~
Toolkit.pdf.
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. Volume 5: Waste, Chapter 4: Biological
Treatment of Solid Waste, Table 4.1. The National Greenhouse Gas Inventories Programme, The
Intergovernmental Panel on Climate Change, H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.).
Hayama, Kanagawa, Japan. Available online at https://www.ipcc-
nggip.iges.or.ip/public/2006gl/pdf/5 Volume5/V5 4 Ch4 Bio Treat.pdf.
Institute for Local Self-Reliance (ISLR) (2014). State of Composting in the US: What, Why, Where & How. Available
at http://ilsr.org/wp-content/uploads/2014/07/state-of-composting-in-us.pdf.
Kijanka (2020) Email correspondence. Kenin Kijanka, EPA Region 2 to Rachel Schmeltz, EPA HQ. "Puerto Rico
Composting Operations." November 13, 2020.
Northeast Recycling Council (NERC) (2020) Disposal Bans & Mandatory Recycling in the United States. Available
online at https://nerc.org/documents/disposal%20bans%20mandatorv%20recycling%20united%20states.pdf.
University of Maine (2016). Compost Report Interpretation Guide. Soil Testing Lab. Available online at:
https://umaine.edu/soiltestinglab/wp-content/uploads/sites/227/2016/07/Compost-Report-lnterpretation-
Guide.pdf.
U.S. Census Bureau (2021) Table 1. Annual Estimates of the Resident Population for the United States, Regions,
States, the District of Columbia, and Puerto Rico: April 1, 2010 to July 1, 2019; April 1, 2020; and July 1, 2020 (NST-
EST2020). Available online at https://www.census.gov/programs-surveys/popest/technical-
documentation/research/evaluation-estimates/2020-evaluation-estimates/2010s-totals-national.html.
U.S. Composting Council (2010) Yard Trimmings Bans: Impact and Support. Prepared by Stuart Buckner, Executive
Director, U.S., Composting Council. Available online at
https://cdn.ymaws.com/www.compostingcouncil.org/resource/resmgr/images/advocacy/Yard-Trimmings-Ban-
lmpacts-a.pdf.
Anaerobic Digestion at Biogas Facilities
IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. Volume 5: Waste, Chapter 4: Biological
Treatment of Solid Waste, Table 4.1. The National Greenhouse Gas Inventories Programme, The Intergovernmental
Panel on Climate Change, H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.). Hayama, Kanagawa,
Japan. Available online at https://www.ipcc-
nggip.iges.or.ip/public/2006gl/pdf/5 Volume5/V I Bio Treat.pdf.
References and Abbreviations 101
-------
EPA (2021) Anaerobic Digestion Facilities Processing Food Waste in the United States (2017 & 2018): Survey
Results. January 2021 EPA/903/S-21/001. Available online at https://www.epa.gov/sites/default/files/2021-
02/documents/2021 final ad report feb 2 with links.pdf.
EPA (2020) Types of Anaerobic Digesters: Common Ways to Describe Digesters. Available online at
https://www.epa.gov/anaerobic-digestion/types-anaerobic-digesters.
EPA (2019) Anaerobic Digestion Facilities Processing Food Waste in the United States in 2016: Survey Results.
September 2019 EPA/903/S-19/001. Available online at https://www.epa.gov/sites/production/files/2018-
08/documents/ad data report final 508 compliant no password.pdf.
EPA (2018) Anaerobic Digestion Facilities Processing Food Waste in the United States in 2015: Survey Results. May
2018 EPA/903/S-18/001. Available online at https://www.epa.gov/sites/production/files/2019-
09/documents/ad data report vlO - 508 comp vl.pdf.
EPA (2016) Frequently Asked Questions About Anaerobic Digestion. Available online at
https://www.epa.goV/anaerobic-digestion/frequent-questions-about-anaerobic-digestion#codigestion.
EPA (1993) Anthropogenic Methane Emissions in the U.S.: Estimates for 1990, Report to Congress. Office of Air and
Radiation, Washington, DC. April 1993.
Water Environment Federation (WEF) (2012) What Every Operator Should Know about Anaerobic Digestion.
Available online at https://www.wef.org/globalassets/assets-wef/direct-download-library/public/operator-
essentials/wet-operator-essentials—anaerobic-digestion—decl2.pdf.
Waste Incineration
RTI (2009) Updated Hospital/Medical/lnfectious Waste Incinerator (HMIWI) Inventory Database. Memorandum
dated July 6, 2009. Available online at: http://www.epa.gov/ttnatw01/129/hmiwi/hmiwi inventory.pdf.
Waste Sources of Precursor Greenhouse Gas Emissions
EPA (2022) "Crosswalk of Precursor Gas Categories." U.S. Environmental Protection Agency. April 6, 2022.
EPA (2021a) "Criteria pollutants National Tier 1 for 1970 - 2020." National Emissions Inventory (NEI) Air Pollutant
Emissions Trends Data. Office of Air Quality Planning and Standards, March 2021. Available online at:
https://www.epa.gov/air-emissions-inventories/air-pollutant-emissions-trends-data.
EPA (2021b) "2017 National Emissions Inventory (NEI) Technical Support Document (TSD)." Office of Air Quality
Planning and Standards, April 2021. Available online at: https://www.epa.gov/air-emissions-inventories/2017-
national-emissions-inventory-nei-technical-support-document-tsd.
Recalculations and Improvements
Enverus (2021) August 2021 Download. Enverus, Inc.
ERG (2021a) Draft Memorandum: Improvements to the 1990-2020 Wastewater Treatment and Discharge
Greenhouse Gas Inventory. July 2021.
ERG (2021b) Updated Other Animal Population Distribution Methodology. ERG, Lexington, MA.
IPCC (2019) 2019 Refinement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National
Greenhouse Gas Inventories Programme, The Intergovernmental Panel on Climate Change. [CalvoBuendia, E.,
Tanabe, K., Kranjc, A., Baasansuren, J., Fukuda, M., Ngarize S., Osako, A., Pyrozhenko, Y., Shermanau, P. and
Federici, S. (eds)]. Switzerland.
102 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
PHMSA (2021b) Underground Natural Gas STAR, Part C. Pipeline and Hazardous Materials Safety Administration,
U.S. Department of Transportation, Washington, DC. https://www.phmsa.dot.gov/data-and-statistics/pipeline/gas-
distribution-gas-gathering-gas-transmission-hazardous-liquids.
Shin, D. (2014) Generation and Disposition of Municipal Solid Waste (MSW) in the United States-A National
Survey. Thesis. Columbia University, Department of Earth and Environmental Engineering, January 3, 2014.
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
References and Abbreviations 103
-------
CKD
Cement kiln dust
FAA
CLE
Crown Light Exposure
FAO
CMA
Chemical Manufacturer's Association
FAOSTAT
CMM
Coal mine methane
FAS
CMOP
Coalbed Methane Outreach Program
FCCC
CMR
Chemical Market Reporter
FEB
CNG
Compressed natural gas
FEMA
CO
Carbon monoxide
FERC
C02
Carbon dioxide
FGD
COD
Chemical oxygen demand
FHWA
COGCC
Colorado Oil and Gas Conservation Commission
FIA
CONUS
Continental United States
FIADB
CRF
Common Reporting Format
FIPR
CRM
Component ratio method
FOD
CRP
Conservation Reserve Program
FOEN
CSRA
Carbon Sequestration Rural Appraisals
FOKS
CTIC
Conservation Technology Information Center
FQSV
CVD
Chemical vapor deposition
FSA
CWNS
Clean Watershed Needs Survey
FTP
d.b.h
Diameter breast height
g
DE
Digestible energy
G&B
DESC
Defense Energy Support Center-DoD's Defense
GaAs
Logistics Agency
GCV
DFAMS
Defense Fuels Automated Management System
GDP
DGGS
Division of Geological & Geophysical Surveys
GEI
DHS
Department of Homeland Security
GHG
DLA
DoD's Defense Logistics Agency
GHGRP
DM
Dry matter
GIS
DOC
Degradable organic carbon
GJ
DOC
U.S. Department of Commerce
GOADS
DoD
U.S. Department of Defense
GOM
DOE
U.S. Department of Energy
GPG
DOI
U.S. Department of the Interior
GRI
DOM
Dead organic matter
GSAM
DOT
U.S. Department of Transportation
GTI
DRE
Destruction or removal efficiencies
GWP
DRI
Direct Reduced Iron
ha
EAF
Electric arc furnace
HBFC
EDB
Aircraft Engine Emissions Databank
HC
EDF
Environmental Defense Fund
HCFC
EER
Energy economy ratio
HCFO
EF
Emission factor
HDDV
EFMA
European Fertilizer Manufacturers Association
HDGV
EJ
Exajoule
HDPE
EGR
Exhaust gas recirculation
HF
EGU
Electric generating unit
HFC
EIA
Energy Information Administration, U.S.
HFO
Department of Energy
HFE
El IP
Emissions Inventory Improvement Program
HHV
EOR
Enhanced oil recovery
HMA
EPA
U.S. Environmental Protection Agency
HMIWI
EREF
Environment Research & Education Foundation
HTF
ERS
Economic Research Service
HTS
ETMS
Enhanced Traffic Management System
HVAE
EV
Electric vehicle
HWP
EVI
Enhanced Vegetation Index
IBF
Federal Aviation Administration
Food and Agricultural Organization
Food and Agricultural Organization database
Fuels Automated System
Framework Convention on Climate Change
Fiber Economics Bureau
Federal Emergency Management Agency
Federal Energy Regulatory Commission
Flue gas desulfurization
Federal Highway Administration
Forest Inventory and Analysis
Forest Inventory and Analysis Database
Florida Institute of Phosphate Research
First order decay
Federal Office for the Environment
Fuel Oil and Kerosene Sales
First-quarter of silicon volume
Farm Service Agency
Federal Test Procedure
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
10-2 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
IC Integrated Circuit
ICAO International Civil Aviation Organization
ICBA International Carbon Black Association
ICE Internal combustion engine
ICR Information Collection Request
IEA International Energy Agency
IFO Intermediate Fuel Oil
IGES Institute of Global Environmental Strategies
IISRP International Institute of Synthetic Rubber
Products
ILENR Illinois Department of Energy and Natural
Resources
IMO International Maritime Organization
IPAA Independent Petroleum Association of America
IPCC Intergovernmental Panel on Climate Change
IPPU Industrial Processes and Product Use
ITC U.S. International Trade Commission
ITRS International Technology Roadmap for
Semiconductors
JWR Jim Walters Resources
KCA Key category analysis
kg Kilogram
kt Kiloton
kWh Kilowatt hour
LDPE Low density polyethylene
LDT Light-duty truck
LDV Light-duty vehicle
LEV Low emission vehicles
LFG Landfill gas
LFGTE Landfill gas-to-energy
LHV Lower Heating Value
LKD Lime kiln dust
LLDPE Linear low density polyethylene
LMOP EPA's Landfill Methane Outreach Program
LNG Liquefied natural gas
LPG Liquefied petroleum gas(es)
LTO Landing and take-off
LULUCF Land Use, Land-Use Change, and Forestry
LVAE Low Voltage Anode Effects
M&R Metering and regulating
MARPOL International Convention for the Prevention of
Pollution from Ships
MC Motorcycle
MCF Methane conversion factor
MCL Maximum Contaminant Levels
MCFD Thousand cubic feet per day
MDI Metered dose inhalers
MDP Management and design practices
MECS EIA Manufacturer's Energy Consumption Survey
MEMS Micro-electromechanical systems
MER Monthly Energy Review
MGO Marine gas oil
MgO Magnesium oxide
MJ Megajoule
MLRA Major Land Resource Area
mm Millimeter
MMBtu Million British thermal units
MMCF Million cubic feet
MMCFD Million cubic feet per day
MMS Minerals Management Service
MMT Million metric tons
MMTCE Million metric tons carbon equivalent
MMT C02 Million metric tons carbon dioxide equivalent
Eq.
MODIS Moderate Resolution Imaging
Spectroradiometer
MoU Memorandum of Understanding
MOVES U.S. EPA's Motor Vehicle Emission Simulator
model
MPG Miles per gallon
MRLC Multi-Resolution Land Characteristics
Consortium
MRV Monitoring, reporting, and verification
MSHA Mine Safety and Health Administration
MSW Municipal solid waste
MT Metric ton
MTBE Methyl Tertiary Butyl Ether
MTBS Monitoring Trends in Burn Severity
MVAC Motor vehicle air conditioning
MY Model year
N20 Nitrous oxide
NA Not applicable; Not available
NACWA National Association of Clean Water Agencies
NAHMS National Animal Health Monitoring System
NAICS North American Industry Classification System
NAPAP National Acid Precipitation and Assessment
Program
NARR North American Regional Reanalysis Product
NAS National Academies of Sciences, Engineering,
and Medicine
NASA National Aeronautics and Space Administration
NASF National Association of State Foresters
NASS USDA's National Agriculture Statistics Service
NC No change
NCASI National Council of Air and Stream
Improvement
NCV Net calorific value
ND No data
NE Not estimated
NEH National Engineering Handbook
NEI National Emissions Inventory
NEMA National Electrical Manufacturers Association
NEMS National Energy Modeling System
NESHAP National Emission Standards for Hazardous Air
Pollutants
NEU Non-Energy Use
NEV Neighborhood Electric Vehicle
NFs Nitrogen trifluoride
NFI National forest inventory
NGL Natural gas liquids
NID National inventory of Dams
NIR National Inventory Report
References 10-3
-------
NLA National Lime Association PLS
NLCD National Land Cover Dataset POTW
NMOC Non-methane organic compounds ppbv
NMVOC Non-methane volatile organic compound ppm
NMOG Non-methane organic gas ppmv
NO Not occurring pptv
N02 Nitrogen dioxide PRCI
NOx Nitrogen oxides PRP
NOAA National Oceanic and Atmospheric PS
Administration PSU
NOF Not on feed PU
NPDES National Pollutant Discharge Elimination System PVC
NPP Net primary productivity PV
NPRA National Petroleum and Refiners Association QA/QC
NRBP Northeast Regional Biomass Program QBtu
NRC National Research Council R&D
NRCS Natural Resources Conservation Service RECs
NREL National Renewable Energy Laboratory RCRA
NRI National Resources Inventory RFA
NSCEP National Service Center for Environmental RFS
Publications RMA
NSCR Non-selective catalytic reduction RPA
NSPS New source performance standards RTO
NWS National Weather Service SAE
OAG Official Airline Guide SAGE
OAP EPA Office of Atmospheric Programs SAIC
OAQPS EPA Office of Air Quality Planning and Standards SAN
ODP Ozone depleting potential SAR
ODS Ozone depleting substances SCR
OECD Organization of Economic Co-operation and SCSE
Development SDR
OEM Original equipment manufacturers SEC
OGJ Oil & Gas Journal SEMI
OGOR Oil and Gas Operations Reports
OH Hydroxyl radical SF6
OMS EPA Office of Mobile Sources SIA
ORNL Oak Ridge National Laboratory SiC
OSHA Occupational Safety and Health Administration SICAS
OTA Office of Technology Assessment SNAP
OTAQ EPA Office of Transportation and Air Quality SNG
OVS Offset verification statement S02
PADUS Protected Areas Database of the United States SOC
PAH Polycyclic aromatic hydrocarbons SOG
PCA Portland Cement Association SOHIO
PCC Precipitate calcium carbonate SSURGO
PDF Probability Density Function STMC
PECVD Plasma enhanced chemical vapor deposition SULEV
PET Polyethylene terephthalate SWANA
PET Potential evapotranspiration SWDS
PEVM PFC Emissions Vintage Model SWICS
PFC Perfluorocarbon TA
PFPE Perfluoropolyether TAM
PHEV Plug-in hybrid vehicles TAME
PHMSA Pipeline and Hazardous Materials Safety TAR
Administration TBtu
PI Productivity index TDN
Pregnant liquor solution
Publicly Owned Treatment Works
Parts per billion (109) by volume
Parts per million
Parts per million (106) by volume
Parts per trillion (1012) by volume
Pipeline Research Council International
Pasture/Range/Paddock
Polystyrene
Primary Sample Unit
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
10-4 Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2020
-------
TEDB Transportation Energy Data Book
TFI The Fertilizer Institute
TIGER Topological^ Integrated Geographic Encoding
and Referencing survey
TJ Terajoule
TLEV Traditional low emissions vehicle
TMLA Total Manufactured Layer Area
TOW Total organics in wastewater
TPO Timber Product Output
TRI Toxic Release Inventory
TSDF Hazardous waste treatment, storage, and
disposal facility
TTB Tax and Trade Bureau
TVA Tennessee Valley Authority
UAN Urea ammonium nitrate
UDI Utility Data Institute
UFORE U.S. Forest Service's Urban Forest Effects model
UG Underground (coal mining)
U.S. United States
U.S. ITC United States International Trade Commission
UEP United Egg Producers
ULEV Ultra low emission vehicle
UNEP United Nations Environmental Programme
UNFCCC United Nations Framework Convention on
Climate Change
USAA U.S. Aluminum Association
USAF United States Air Force
USDA United States Department of Agriculture
USFS United States Forest Service
USGS United States Geological Survey
USITC U.S. International Trade Commission
VAIP EPA's Voluntary Aluminum Industrial
Partnership
VAM Ventilation air methane
VKT Vehicle kilometers traveled
VMT Vehicle miles traveled
VOCs Volatile organic compounds
VS Volatile solids
WBJ Waste Business Journal
WEF Water Environment Federation
WERF Water Environment Research Federation
WFF World Fab Forecast (previously WFW, World
Fab Watch)
WGC World Gas Conference
WIP Waste-in-place
WMO World Meteorological Organization
WMS Waste management systems
WRRF Water resource recovery facilities
WTE Waste-to-energy
WW Wastewater
WWTP Wastewater treatment plant
ZEVs Zero emissions vehicles
References 10-5
------- |