EPA 430-P-18-001
DRAFT Inventory of U.S. Greenhouse Gas
Emissions and Sinks:
1990-2016
February 6,2018
U.S. Environmental Protection Agency
1200 Pennsylvania Ave., N.W.
Washington, DC 20460
U.S.A.
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
HOW TO OBTAIN COPIES
You can electronically download this document on the U.S. EPA's homepage at
.
All data tables of this document for the full time series 1990 through 2016, inclusive, will be made available for the
final report published on April 15, 2018 at the internet site mentioned above.
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
.
-------
1 Acknowledgments
2 The Environmental Protection Agency would like to acknowledge the many individual and organizational
3 contributors to this document, without whose efforts this report would not be complete. Although the complete list
4 of researchers, government employees, and consultants who have provided technical and editorial support is too
5 long to list here, EPA's Office of Atmospheric Programs would like to thank some key contributors and reviewers
6 whose work has significantly improved this year's report.
7 Work on emissions from fuel combustion was led by Vincent Camobreco. Sarah Roberts and Justine Geidosch
8 directed the work on mobile combustion and transportation. Work on fugitive methane emissions from the Energy
9 sector was directed by Melissa Weitz, Chris Sherry, and Cate Hight. Calculations for the Waste sector were led by
10 Rachel Schmeltz. Tom Wirth directed work on the Agriculture and the Land Use, Land-Use Change, and Forestry
11 chapters, with support from John Steller. Work on Industrial Processes and Product Use (IPPU) CO2, CH4, and N20
12 emissions was directed by John Steller. Work on emissions of HFCs, PFCs, SF6, and NF3 from the IPPU sector was
13 directed by Deborah Ottinger and Dave Godwin. Cross-cutting work was directed by Mausami Desai.
14 Within the EPA, other Offices also contributed data, analysis, and technical review for this report. The Office of
15 Transportation and Air Quality and the Office of Air Quality Planning and Standards provided analysis and review
16 for several of the source categories addressed in this report. The Office of Solid Waste and the Office of Research
17 and Development also contributed analysis and research.
18 The Energy Information Administration and the Department of Energy contributed invaluable data and analysis on
19 numerous energy-related topics. Other government agencies have contributed data as well, including the U.S.
20 Geological Survey, the Federal Highway Administration, the Department of Transportation, the Bureau of
21 Transportation Statistics, the Department of Commerce, the National Agricultural Statistics Service, the Federal
22 Aviation Administration, and the Department of Defense.
23 We thank the U.S. Department of Agriculture's Forest Service (Grant Domke, Brian Walters, Jim Smith, Mike
24 Nichols, and John Coulston) for compiling the inventories for carbon dioxide (CO2), methane (CH4), and nitrous
25 oxide (N20) fluxes associated with forest land.
26 We thank the Department of Agriculture's Agricultural Research Service (Stephen Del Grosso) and the Natural
27 Resource Ecology Laboratory at Colorado State University (Stephen Ogle, Keith Paustian, Bill Parton, F. Jay Breidt,
28 Shannon Spencer, Kendrick Killian, Ram Gurung, Ernie Marx, Stephen Williams, Cody Alsaker, Amy Swan, and
29 Chris Dorich) for compiling the inventories for CH4 emissions, N20 emissions, and CO2 fluxes associated with soils
30 in croplands, grasslands, and settlements.
31 We thank Silvestrum Climate Associates (Stephen Crooks, Lisa Schile Beers, Christine May), National Oceanic and
32 Atmospheric Administration (Nate Herold, Ariana Sutton-Grier, Meredith Muth), the Smithsonian Environmental
33 Research Center (J. Patrick Megonigal, Blanca Bernal, James Holmquist, Meng Lu) and Florida International
34 University (Tiffany Troxler) and members of the U.S. Coastal Wetland Carbon Working Group for compiling
35 inventories of land use change, soil carbon stocks and stock change, CH4 emissions, and N20 emissions from
36 aquaculture in coastal wetlands.
37 We would also like to thank Marian Martin Van Pelt, Leslie Chinery, Alexander Lataille, Sabrina Andrews and the
38 full Inventory team at ICF including Diana Pape, Robert Lanza, Lauren Marti, Mollie Averyt, Larry O'Rourke,
39 Deborah Harris, Tommy Hendrickson, Rebecca Ferenchiak, Kasey Knoell, Cory Jemison, Emily Kent, Rani Murali,
-------
1 Drew Stilson, Cara Blumenthal, Louise Huttinger, Helena Caswell, Charlotte Cherry, Katie O'Malley, Howard
2 Marano, and Neha Vaingankar for synthesizing this report and preparing many of the individual analyses.
3 We thank Eastern Research Group for their significant analytical support. Deborah Bartram, Kara Edquist, and
4 Amie Aguiar support the development of emissions estimates for wastewater. Cortney Itle, Amie Aguiar, Kara
5 Edquist, Amber Allen, and Spencer Sauter support the inventories for Manure Management, Enteric Fermentation,
6 Wetlands Remaining Wetlands, and Landfilled Yard Trimmings and Food Scraps (included in Settlements
7 Remaining Settlements). Casey Pickering, Brandon Long, Gopi Manne, and Aylin Sertkaya develop estimates for
8 Natural Gas and Petroleum Systems. Brian Guzzone supports the Coal Mining sector.
9 Finally, we thank the following teams for their significant analytical support: RTI International (Kate Bronstein,
10 Meaghan McGrath, Michael Laney, Carson Moss, David Randall, Gabrielle Raymond, Jason Goldsmith, Karen
11 Schaffner, Melissa Icenhour); Raven Ridge Resources, and Ruby Canyon Engineering Inc. (Michael Cote, Samantha
12 Phillips, and Phillip Cunningham).
13
-------
1 Preface
2 The United States Environmental Protection Agency (EPA) prepares the official U.S. Inventory of Greenhouse Gas
3 Emissions and Sinks to comply with existing commitments under the United Nations Framework Convention on
4 Climate Change (UNFCCC). Under decision 3/CP.5 of the UNFCCC Conference of the Parties, national inventories
5 for UNFCCC Annex I parties should be provided to the UNFCCC Secretariat each year by April 15.
6 In an effort to engage the public and researchers across the country, the EPA has instituted an annual public review
7 and comment process for this document. The availability of the draft document is announced via Federal Register
8 Notice and is posted on the EPA web site. Copies are also emailed upon request. The public comment period is
9 generally limited to 30 days; however, comments received after the closure of the public comment period are
10 accepted and considered for the next edition of this annual report. Public review of this report is occurring in
11 February 2018, and comments received will be posted to the EPA web site.
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Table of Contents
TABLE OF CONTENTS VI
LIST OF TABLES, FIGURES, AND BOXES IX
EXECUTIVE SUMMARY ES-1
ES. 1 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-18
ES.4 Other Information ES-23
1. INTRODUCTION 1-1
1.1 Background Information 1-3
1.2 National Inventory Arrangements 1-10
1.3 Methodology and Data Sources 1-15
1.4 Key Categories 1-16
1.5 Quality Assurance and Quality Control (QA/QC) 1-19
1.6 Uncertainty Analysis of Emission Estimates 1-21
1.7 Completeness 1-23
1.8 Organization of Report 1-23
2. TRENDS IN GREENHOUSE GAS EMISSIONS 2-1
2.1 Recent Trends in U.S. Greenhouse Gas Emissions and Sinks 2-1
2.2 Emissions by Economic Sector 2-23
2.3 Indirect Greenhouse Gas Emissions (CO, NOx, NMVOCs, and SO2) 2-34
3. ENERGY 3-1
3.1 Fossil Fuel Combustion (CRF Source Category 1A) 3-5
3.2 Carbon Emitted from Non-Energy Uses of Fossil Fuels (CRF Source Category 1A) 3-45
3.3 Incineration of Waste (CRF Source Category lAla) - TO BE UPDATED FOR FINAL INVENTORY
REPORT 3-51
3.4 Coal Mining (CRF Source Category lBla) 3-56
3.5 Abandoned Underground Coal Mines (CRF Source Category lBla) 3-61
3.6 Petroleum Systems (CRF Source Category lB2a) 3-65
3.7 Natural Gas Systems (CRF Source Category lB2b) 3-77
vi DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
3.8 Abandoned Oil and Gas Wells (CRF Source Categories lB2a and lB2b) 3-93
3.9 Energy Sources of Indirect Greenhouse Gas Emissions 3-96
3.10 International Bunker Fuels (CRF Source Category 1: Memo Items) 3-97
3.11 WoodBiomass andBiofuels Consumption (CRF Source Category 1A) 3-101
4. INDUSTRIAL PROCESSES AND PRODUCT USE 4-1
4.1 Cement Production (CRF Source Category 2A1) 4-8
4.2 Lime Production (CRF Source Category 2A2) - TO BE UPDATED FOR FINAL INVENTORY REPORT.
4-12
4.3 Glass Production (CRF Source Category 2A3) - TO BE UPDATED FOR FINAL INVENTORY REPORT
4-17
4.4 Other Process Uses of Carbonates (CRF Source Category 2A4) - TO BE UPDATED FOR FINAL
INVENTORY REPORT 4-20
4.5 Ammonia Production (CRF Source Category 2B1) 4-24
4.6 Urea Consumption for Non-Agricultural Purposes 4-28
4.7 Nitric Acid Production (CRF Source Category 2B2) 4-31
4.8 Adipic Acid Production (CRF Source Category 2B3) 4-34
4.9 Caprolactam, Glyoxal and Glyoxylic Acid Production (CRF Source Category 2B4) 4-38
4.10 Silicon Carbide Production and Consumption (CRF Source Category 2B5) 4-41
4.11 Titanium Dioxide Production (CRF Source Category 2B6) 4-44
4.12 Soda Ash Production (CRF Source Category 2B7) 4-47
4.13 Petrochemical Production (CRF Source Category 2B8) 4-50
4.14 HCFC-22 Production (CRF Source Category 2B9a) 4-56
4.15 Carbon Dioxide Consumption (CRF Source Category 2B10) 4-59
4.16 Phosphoric Acid Production (CRF Source Category 2B10) 4-62
4.17 Iron and Steel Production (CRF Source Category 2C1) and Metallurgical Coke Production 4-66
4.18 Ferroalloy Production (CRF Source Category 2C2) 4-76
4.19 Aluminum Production (CRF Source Category 2C3) 4-79
4.20 Magnesium Production and Processing (CRF Source Category 2C4) 4-84
4.21 Lead Production (CRF Source Category 2C5) 4-89
4.22 Zinc Production (CRF Source Category 2C6) 4-92
4.23 Semiconductor Manufacture (CRF Source Category 2E1) 4-97
4.24 Substitution of Ozone Depleting Substances (CRF Source Category 2F) 4-109
4.25 Electrical Transmission and Distribution (CRF Source Category 2G1) 4-117
4.26 Nitrous Oxide from Product Uses (CRF Source Category 2G3) 4-124
4.27 Industrial Processes and Product Use Sources of Indirect Greenhouse Gases 4-127
5. AGRICULTURE 5-1
5.1 Enteric Fermentation (CRF Source Category 3A) 5-3
5.2 Manure Management (CRF Source Category 3B) 5-9
vii
-------
1 5.3 Rice Cultivation (CRF Source Category 3C) 5-17
2 5.4 Agricultural Soil Management (CRF Source Category 3D) 5-23
3 5.5 Liming (CRF Source Category 3G) 5-42
4 5.6 Urea Fertilization (CRF Source Category 3H) 5-45
5 5.7 Field Burning of Agricultural Residues (CRF Source Category 3F) 5-47
6 6. LAND USE, LAND-USE CHANGE, AND FORESTRY 6-1
7 6.1 Representation of the U.S. Land Base 6-8
8 6.2 Forest Land Remaining Forest Land (CRF Category 4 Al) 6-22
9 6.3 Land Converted to Forest Land (CRF Category 4A2) 6-42
10 6.4 Cropland Remaining Cropland (CRF Category 4B1) 6-48
11 6.5 Land Converted to Cropland (CRF Category 4B2) 6-57
12 6.6 Grassland Remaining Grassland (CRF Category 4C1) 6-63
13 6.7 Land Converted to Grassland (CRF Category 4C2) 6-72
14 6.8 Wetlands Remaining Wetlands (CRF Category 4D1) 6-79
15 6.9 Land Converted to Wetlands (CRF Category 4D2) 6-96
16 6.10 Settlements Remaining Settlements (CRF Category 4E1) 6-99
17 6.11 Land Converted to Settlements (CRF Category 4E2) 6-116
18 6.12 Other Land Remaining Other Land (CRF Category 4F1) 6-121
19 6.13 Land Converted to Other Land (CRF Category 4F2) 6-122
20 7. WASTE 7-1
21 7.1 Landfills (CRF Source Category 5A1) 7-3
22 7.2 Wastewater Treatment (CRF Source Category 5D) 7-19
23 7.3 Composting (CRF Source Category 5B1) 7-33
24 7.4 Waste Incineration (CRF Source Category 5C1) - TO BE UPDATED FOR FINAL INVENTORY
25 REPORT 7-35
26 7.5 Waste Sources of Indirect Greenhouse Gases 7-36
27 8. OTHER 8-1
28 9. RECALCULATIONS AND IMPROVEMENTS 9-1
29 10. REFERENCES 10-1
30
31
32
viii DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
List of Tables, Figures, and Boxes
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-6
Table ES-3: CO2 Emissions from Fossil Fuel Combustion by End-Use Sector (MMT CO2 Eq.) ES-11
Table ES-4: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks by Chapter/IPCC Sector (MMT CO2 Eq.)
ES-18
Table ES-5: U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use, Land-Use Change, and
Forestry (MMT CO Eq.) ES-22
Table ES-6: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO2 Eq.) ES-24
Table ES-7: U.S. Greenhouse Gas Emissions by Economic Sector with Electricity-Related Emissions Distributed
(MMTCO2 Eq.) ES-25
Table ES-8: Recent Trends in Various U.S. Data (Index 1990 = 100) ES-26
Table 1-1: Global Atmospheric Concentration, Rate of Concentration Change, and Atmospheric Lifetime of
Selected Greenhouse Gases 1-4
Table 1-2: Global Warming Potentials and Atmospheric Lifetimes (Years) Used in this Report 1-9
Table 1-3: Comparison of 100-Year GWP values 1-10
Table 1-4: Key Categories for the United States (1990-2016) 1-16
Table 1-5: Estimated Overall Inventory Quantitative Uncertainty (MMT CO2 Eq. and Percent) - TO BE UPDATED
FOR FINAL INVENTORY REPORT 1-22
Table 1-6: IPCC Sector Descriptions 1-24
Table 1-7: List of Annexes 1-24
Table 2-1: Recent Trends 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 Chapter/IPCC Sector (MMT CO2 Eq.)....
2-7
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.) 2-12
Table 2-6: Emissions from Industrial Processes and Product Use (MMT CO2 Eq.) 2-16
Table 2-7: Emissions from Agriculture (MMT CO2 Eq.) 2-18
Table 2-8: U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use, Land-Use Change, and
Forestry (MMT CO2 Eq.) 2-20
U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO2 Eq. and Percent of Total in
2-24
Table 2-9: Emissions from Waste (MMT CO2 Eq.) 2-22
Table 2-10
2016)
Table 2-11: Electric Power-Related Greenhouse Gas Emissions (MMT CO2 Eq.) 2-26
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 2016 2-27
ix
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Table 2-13: Transportation-Related Greenhouse Gas Emissions (MMT CO2 Eq.) 2-30
Table 2-14: Recent Trends in Various U.S. Data (Index 1990 = 100) 2-33
Table 2-15: Emissions of NO,. CO, NMVOCs, and SO -(kt) 2-35
Table 3-1: CO2, CH4, and N20 Emissions from Energy (MMT CO2 Eq.) 3-2
Table 3-2: CO2, CH4, and N20 Emissions from Energy (kt) 3-3
Table 3-3: CO2, CH4, and N20 Emissions from Fossil Fuel Combustion (MMT CO2 Eq.) 3-5
Table 3-4: CO2, CH4, and N20 Emissions from Fossil Fuel Combustion (kt) 3-5
Table 3-5: CO2 Emissions from Fossil Fuel Combustion by Fuel Type and Sector (MMT CO2 Eq.) 3-6
Table 3-6: Annual Change in CO2 Emissions and Total 2016 Emissions fromFossil Fuel Combustion for Selected
Fuels and Sectors (MMT CO2 Eq. and Percent) 3-7
Table 3-7: CO2, CH4, andN20 Emissions fromFossil Fuel Combustion by Sector (MMT CO2 Eq.) 3-11
Table 3-8: CO2, CH4, andN20 Emissions fromFossil Fuel Combustion by End-Use Sector (MMT CO2 Eq.).... 3-12
Table 3-9: CO2 Emissions from Stationary Fossil Fuel Combustion (MMT CO2 Eq.) 3-13
Table 3-10: CH4 Emissions from Stationary Combustion (MMT CO2 Eq.) 3-13
Table 3-11: N20 Emissions from Stationary Combustion (MMT CO2 Eq.) 3-14
Table 3-12: CO2 Emissions from Fossil Fuel Combustion in Transportation End-Use Sector (MMT CO2 Eq.)... 3-24
Table 3-13: CH4 Emissions from Mobile Combustion (MMT CO2 Eq.) 3-26
Table 3-14: N2O Emissions from Mobile Combustion (MMT CO2 Eq.) 3-27
Table 3-15: Carbon Intensity from Direct Fossil Fuel Combustion by Sector (MMT CO2 Eq./QBtu) 3-32
Table 3-16: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Energy-Related Fossil Fuel
Combustion by Fuel Type and Sector (MMT CO2 Eq. and Percent) 3-35
Table 3-17: Approach 2 Quantitative Uncertainty Estimates for CH4 and N20 Emissions from Energy-Related
Stationary Combustion, Including Biomass (MMT CO2 Eq. and Percent) 3-39
Table 3-18: Approach 2 Quantitative Uncertainty Estimates for CH4 and N20 Emissions from Mobile Sources
(MMT CO2 Eq. and Percent) 3-42
Table 3-19: CO2 Emissions from Non-Energy Use Fossil Fuel Consumption (MMT CO2 Eq. and Percent) 3-45
Table 3-20: Adjusted Consumption of Fossil Fuels for Non-Energy Uses (TBtu) 3-46
Table 3-21: 2016 Adjusted Non-Energy Use Fossil Fuel Consumption, Storage, and Emissions 3-47
Table 3-22: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Non-Energy Uses of Fossil
Fuels (MMT CO2 Eq. and Percent) 3-48
Table 3-23: Approach 2 Quantitative Uncertainty Estimates for Storage Factors of Non-Energy Uses of Fossil Fuels
(Percent) 3-49
Table 3-24: CO2, CH4, andN20 Emissions from the Incineration of Waste (MMT CO2 Eq.) 3-52
Table 3-25: CO2, CH4, andN20 Emissions from the Incineration of Waste (kt) 3-52
Table 3-26: Municipal Solid Waste Generation (Metric Tons) and Percent Combusted (BioCycle dataset) 3-54
Table 3-27: Approach 2 Quantitative Uncertainty Estimates for CO2 andN20 from the Incineration of Waste (MMT
CO2 Eq. and Percent) 3-55
Table 3-28: Coal Production (kt) 3-57
Table 3-29: CH4 Emissions from Coal Mining (MMT CO2 Eq.) 3-57
x DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Table 3-30: CH4 Emissions from Coal Mining (kt) 3-57
Table 3-31: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from Coal Mining (MMT CO2 Eq.
and Percent) 3-61
Table 3-32: CH4 Emissions from Abandoned Coal Mines (MMT CO2 Eq.) 3-62
Table 3-33: CH4 Emissions from Abandoned Coal Mines (kt) 3-62
Table 3-34: Number of Gassy Abandoned Mines Present in U.S. Basins in 2016, grouped by Class according to
Post-Abandonment State 3-63
Table 3-35: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from Abandoned Underground Coal
Mines (MMT CO2 Eq. and Percent) 3-65
Table 3-36: CH4 Emissions from Petroleum Systems (MMT CO2 Eq.) 3-66
Table 3-37: CH4 Emissions from Petroleum Systems (kt) 3-66
Table 3-38: CO2 Emissions from Petroleum Systems (MMT CO2) 3-67
Table 3-39: CO2 Emissions from Petroleum Systems (kt) 3-67
Table 3-40: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from Petroleum Systems (MMT
CO2 Eq. and Percent) 3-69
Table 3-41: Oil Well Testing National CH4 Emissions (Metric Tons CH4) 3-71
Table 3-42: Oil Well Testing National CO2 Emissions (Metric Tons CO2) 3-71
Table 3-43: National Tank CO2 Emissions by Category and National Emissions (kt CO2) 3-72
Table 3-44: Associated Gas Venting and Flaring National CO2 Emissions (kt CO2) 3-73
Table 3-45: Associated Gas Venting and Flaring National CH4 Emissions (Metric Tons CH4) 3-73
Table 3-46: Miscellaneous Production Flaring National CO2 Emissions (kt CO2) 3-73
Table 3-47: Miscellaneous Production Flaring National CH4 Emissions (Metric Tons CH4) 3-74
Table 3-48: Producing Oil Well Count Data 3-74
Table 3-49: Quantity of CO2 Captured and Extracted for EOR Operations (MMT CO2) 3-77
Table 3-50: Quantity of CO2 Captured and Extracted for EOR Operations (kt) 3-77
Table 3-51: CH4 Emissions from Natural Gas Systems (MMT CO2 Eq.)a 3-79
Table 3-52: CH4 Emissions from Natural Gas Systems (kt)a 3-79
Table 3-53: Calculated Potential CH4 and Captured/Combusted CH4 from Natural Gas Systems (MMT CO2 Eq.)
3-80
Table 3-54: Non-combustion CO2 Emissions from Natural Gas Systems (MMT) 3-80
Table 3-55: Non-combustion CO2 Emissions from Natural Gas Systems (kt) 3-80
Table 3-56: Approach 2 Quantitative Uncertainty Estimates for CH4 and Non-energy CO2 Emissions from Natural
Gas Systems (MMT CO2 Eq. and Percent) 3-83
Table 3-57: Gas Well Testing National CH4 Emissions (Metric Tons CH4) 3-85
Table 3-58: Gas Well Testing National CO2 Emissions (Metric Tons CO2) 3-85
Table 3-59: Non-HF Gas Well Completions National CH4 Emissions (Metric Tons CH4) 3-86
Table 3-60: Non-HF Gas Well Completions National CO2 Emissions (Metric Tons CO2) 3-86
Table 3-61: HF Gas Well Completions National CO2 Emissions (kt CO2) 3-86
Table 3-62: Non-HF Gas Well Workovers National CH4 Emissions (Metric Tons CH4) 3-87
xi
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Table 3-63: Non-HF Gas Well Workovers National CO2 Emissions (kt CO2) 3-87
Table 3-64: Producing Gas Well Count Data 3-87
Table 3-65: Miscellaneous Production Flaring National CO2 Emissions (kt CO2) 3-88
Table 3-66: National Condensate Tank Emissions by Category and National Emissions (kt CO2) 3-88
Table 3-67: Processing CO2 Updates, National Emissions (kt CO2) 3-89
Table 3-68: Transmission and Storage CH4 Updates to Flaring, National Emissions (MT CH4) 3-89
Table 3-69: Transmission and Storage CO2 Updates, National Emissions (kt CO2) 3-90
Table 3-70: CH4 Emissions from Abandoned Oil and Gas Wells (MMT CO2 Eq.) 3-93
Table 3-71: CH4 Emissions from Abandoned Oil and Gas Wells (kt) 3-93
Table 3-72: Abandoned Oil Wells Activity Data and Methane Emissions (Metric Tons CH4) 3-94
Table 3-73: Abandoned Gas Wells Activity Data and Methane Emissions (Metric Tons CH4) 3-94
Table 3-74: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from Petroleum Systems (MMT
CO2 Eq. and Percent) 3-95
Table 3-75: NOx, CO, and NMVOC Emissions from Energy-Related Activities (kt) 3-96
Table 3-76: CO2, CH4, and N2O Emissions from International Bunker Fuels (MMT CO2 Eq.) 3-98
Table 3-77: CO2, CH4, and N20 Emissions from International Bunker Fuels (kt) 3-98
Table 3-78: Aviation Jet Fuel Consumption for International Transport (Million Gallons) 3-100
Table 3-79: Marine Fuel Consumption for International Transport (Million Gallons) 3-100
Table 3-80: CO2 Emissions from Wood Consumption by End-Use Sector (MMT CO2 Eq.) 3-102
Table 3-81: CO2 Emissions from Wood Consumption by End-Use Sector (kt) 3-102
Table 3-82: CO2 Emissions from Ethanol Consumption (MMT CO2 Eq.) 3-102
Table 3-83: CO2 Emissions from Ethanol Consumption (kt) 3-102
Table 3-84: CO2 Emissions from Biodiesel Consumption (MMT CO2 Eq.) 3-103
Table 3-85: CO2 Emissions from Biodiesel Consumption (kt) 3-103
Table 3-86: Woody Biomass Consumption by Sector (Trillion Btu) 3-103
Table 3-87: Ethanol Consumption by Sector (TrillionBtu) 3-104
Table 3-88: Biodiesel Consumption by Sector (TrillionBtu) 3-104
Table 4-1: Emissions from Industrial Processes and Product Use (MMT CO2 Eq.) 4-3
Table 4-2: Emissions from Industrial Processes and Product Use (kt) 4-4
Table 4-3: CO2 Emissions from Cement Production (MMT CO2 Eq. and kt) 4-9
Table 4-4: Clinker Production (kt) 4-10
Table 4-5: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Cement Production (MMT CO2
Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY REPORT 4-11
Table 4-6: CO2 Emissions from Lime Production (MMT CO2 Eq. and kt) 4-12
Table 4-7: Potential, Recovered, and Net CO2 Emissions from Lime Production (kt) 4-13
Table 4-8: High-Calcium- and Dolomitic-Quicklime, High-Calcium- and Dolomitic-Hydrated, and Dead-Burned-
Dolomite Lime Production (kt) 4-14
Table 4-9: Adjusted Lime Production (kt) 4-14
xii DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Table 4-10: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Lime Production (MMT CO2
2 Eq. and Percent) 4-15
3 Table 4-11: CO2 Emissions from Glass Production (MMT CO2 Eq. and kt) 4-17
4 Table 4-12: Limestone, Dolomite, and Soda Ash Consumption Used in Glass Production (kt) 4-19
5 Table 4-13: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Glass Production (MMT CO2
6 Eq. and Percent) 4-19
7 Table 4-14: CO2 Emissions from Other Process Uses of Carbonates (MMT CO2 Eq.) 4-21
8 Table 4-15: CO2 Emissions from Other Process Uses of Carbonates (kt) 4-21
9 Table 4-16: Limestone and Dolomite Consumption (kt) 4-22
10 Table 4-17: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Other Process Uses of
11 Carbonates (MMT CO2 Eq. and Percent) 4-23
12 Table 4-18: CO2 Emissions from Ammonia Production (MMT CO2 Eq.) 4-25
13 Table 4-19: CO2 Emissions from Ammonia Production (kt) 4-25
14 Table 4-20: Ammonia Production and Urea Production (kt) 4-26
15 Table 4-21: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Ammonia Production (MMT
16 C02 Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY REPORT 4-27
17 Table 4-22: CO2 Emissions from Urea Consumption for Non-Agricultural Purposes (MMT CO2 Eq.) 4-28
18 Table 4-23: CO2 Emissions from Urea Consumption for Non-Agricultural Purposes (kt) 4-29
19 Table 4-24: Urea Production, Urea Applied as Fertilizer, Urea Imports, and Urea Exports (kt) 4-30
20 Table 4-25: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Urea Consumption for Non-
21 Agricultural Purposes (MMT C02 Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY REPORT
22 4-30
23 Table 4-26: N2O Emissions from Nitric Acid Production (MMT CO2 Eq. and kt N2O) 4-31
24 Table 4-27: Nitric Acid Production (kt) 4-33
25 Table 4-28: Approach 2 Quantitative Uncertainty Estimates for N2O Emissions from Nitric Acid Production (MMT
26 C02 Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY REPORT 4-34
27 Table 4-29: N2O Emissions from Adipic Acid Production (MMT CO2 Eq. and kt N2O) 4-35
28 Table 4-30: Adipic Acid Production (kt) 4-37
29 Table 4-31: Approach 2 Quantitative Uncertainty Estimates for N20 Emissions from Adipic Acid Production
30 (MMT C02 Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY REPORT 4-37
31 Table 4-32: N20 Emissions from Caprolactam Production (MMT CO2 Eq. and kt N20) 4-39
32 Table 4-33: Caprolactam Production (kt) 4-40
33 Table 4-34: Approach 2 Quantitative Uncertainty Estimates for N20 Emissions from Caprolactam, Glyoxal and
34 Glyoxylic Acid Production (MMT C02 Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY
35 REPORT 4-40
36 Table 4-35: CO2 and CH4 Emissions from Silicon Carbide Production and Consumption (MMT CO2 Eq.) 4-42
37 Table 4-36: CO2 and CH4 Emissions from Silicon Carbide Production and Consumption (kt) 4-42
38 Table 4-37: Production and Consumption of Silicon Carbide (Metric Tons) 4-43
39 Table 4-38: Approach 2 Quantitative Uncertainty Estimates for CH4 and CO2 Emissions from Silicon Carbide
40 Production and Consumption (MMT CO2 Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY
41 REPORT 4-44
xiii
-------
1 Table 4-39: CO2 Emissions from Titanium Dioxide (MMT CO2 Eq. and kt) 4-45
2 Table 4-40: Titanium Dioxide Production (kt) 4-46
3 Table 4-41: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Titanium Dioxide Production
4 (MMT C02 Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY REPORT 4-46
5 Table 4-42: CO2 Emissions from Soda Ash Production (MMT CO2 Eq. and kt CO2) 4-48
6 Table 4-43: Soda Ash Production (kt) 4-49
7 Table 4-44: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Soda Ash Production (MMT
8 C02 Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY REPORT 4-49
9 Table 4-45: CO2 and CH4 Emissions from Petrochemical Production (MMT CO2 Eq.) 4-52
10 Table 4-46: CO2 and CH4 Emissions from Petrochemical Production (kt) 4-52
11 Table 4-47: Production of Selected Petrochemicals (kt) 4-54
12 Table 4-48: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from Petrochemical Production and
13 CO2 Emissions from Carbon Black Production (MMT CO2 Eq. and Percent) - TO BE UPDATED FOR FINAL
14 INVENTORY REPORT 4-55
15 Table 4-49: HFC-23 Emissions from HCFC-22 Production (MMT C02 Eq. and kt HFC-23) 4-57
16 Table 4-50: HCFC-22 Production (kt) 4-58
17 Table 4-51: Approach 2 Quantitative Uncertainty Estimates for HFC-23 Emissions from HCFC-22 Production
18 (MMT CO2 Eq. and Percent) 4-58
19 Table 4-52: CO2 Emissions from CO2 Consumption (MMT CO2 Eq. and kt) 4-59
20 Table 4-53: CO2 Production (kt CO2) and the Percent Used for Non-EOR Applications 4-61
21 Table 4-54: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from CO2 Consumption (MMT CO2
22 Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY REPORT 4-62
23 Table 4-55: CO2 Emissions from Phosphoric Acid Production (MMT CO2 Eq. and kt) 4-63
24 Table 4-56: Phosphate Rock Domestic Consumption, Exports, and Imports (kt) 4-64
25 Table 4-57: Chemical Composition of Phosphate Rock (Percent by Weight) 4-64
26 Table 4-58: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Phosphoric Acid Production
27 (MMT C02 Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY REPORT 4-65
28 Table 4-59: CO2 Emissions from Metallurgical Coke Production (MMT CO2 Eq.) 4-67
29 Table 4-60: CO2 Emissions from Metallurgical Coke Production (kt) 4-67
30 Table 4-61: CO2 Emissions from Iron and Steel Production (MMT CO2 Eq.) 4-67
31 Table 4-62: CO2 Emissions from Iron and Steel Production (kt) 4-68
32 Table 4-63: CH4 Emissions from Iron and Steel Production (MMT CO2 Eq.) 4-68
33 Table 4-64: CH4 Emissions from Iron and Steel Production (kt) 4-68
34 Table 4-65: Material Carbon Contents for Metallurgical Coke Production 4-69
35 Table 4-66: Production and Consumption Data for the Calculation of CO2 Emissions from Metallurgical Coke
36 Production (Thousand Metric Tons) 4-70
37 Table 4-67: Production and Consumption Data for the Calculation of CO2 Emissions from Metallurgical Coke
38 Production (Million ft3) 4-70
39 Table 4-68: Material Carbon Contents for Iron and Steel Production 4-71
40 Table 4-69: CH4 Emission Factors for Sinter and Pig Iron Production 4-71
xiv DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Table 4-70: CO2 Emission Factors for Sinter Production, Direct Reduced Iron Production and Pellet Production 4-72
Table 4-71: Production and Consumption Data for the Calculation of CO2 and CH4 Emissions from Iron and Steel
Production (Thousand Metric Tons) 4-73
Table 4-72: Production and Consumption Data for the Calculation of CO2 Emissions from Iron and Steel
Production (Million ft3 unless otherwise specified) 4-73
Table 4-73: Approach 2 Quantitative Uncertainty Estimates for CO2 and CH4 Emissions from Iron and Steel
Production and Metallurgical Coke Production (MMT CO2 Eq. and Percent) - TO BE UPDATED FOR FINAL
INVENTORY REPORT 4-74
Table 4-74: CO2 and CH4 Emissions from Ferroalloy Production (MMT CO2 Eq.) 4-76
Table 4-75: CO2 and CH4 Emissions from Ferroalloy Production (kt) 4-76
Table 4-76: Production of Ferroalloys (Metric Tons) 4-78
Table 4-77: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Ferroalloy Production (MMT
C02 Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY REPORT 4-79
Table 4-78: CO2 Emissions from Aluminum Production (MMT CO2 Eq. and kt) 4-80
Table 4-79: PFC Emissions from Aluminum Production (MMT CO2 Eq.) 4-80
Table 4-80: PFC Emissions from Aluminum Production (kt) 4-81
Table 4-81: Production of Primary Aluminum (kt) 4-83
Table 4-82: Approach 2 Quantitative Uncertainty Estimates for CO2 and PFC Emissions from Aluminum
Production (MMT CO2 Eq. and Percent) 4-84
Table 4-83: SF6, HFC-134a, FK 5-1-12 and CO2 Emissions from Magnesium Production and Processing (MMT
C02 Eq.) 4-85
Table 4-84: SF6, HFC-134a, FK 5-1-12 and CO2 Emissions from Magnesium Production and Processing (kt)... 4-85
Table 4-85: SF6 Emission Factors (kg SF6 per metric ton of magnesium) 4-87
Table 4-86: Approach 2 Quantitative Uncertainty Estimates for SF6, HFC-134a and CO2 Emissions from
Magnesium Production and Processing (MMT CO2 Eq. and Percent) 4-88
Table 4-87: CO2 Emissions from Lead Production (MMT CO2 Eq. and kt) 4-90
Table 4-88: Lead Production (Metric Tons) 4-91
Table 4-89: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Lead Production (MMT CO2
Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY REPORT 4-91
Table 4-90: Zinc Production (Metric Tons) 4-93
Table 4-91: CO2 Emissions from Zinc Production (MMT CO2 Eq. and kt) 4-94
Table 4-92: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Zinc Production (MMT CO2
Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY REPORT 4-96
Table 4-93: PFC, HFC, SF6, NF3, and N20 Emissions from Semiconductor Manufacture (MMT CO2 Eq.) 4-98
Table 4-94: PFC, HFC, SF6, NF3, and N20 Emissions from Semiconductor Manufacture (kt) 4-99
Table 4-95: F-HTF Emissions Based on GHGRP Reporting (MMT CO2 Eq.) 4-99
Table 4-96: F-HTF Compounds with Largest Emissions Based on GHGRP Reporting (tons of gas) 4-99
Table-4-97: Approach 2 Quantitative Uncertainty Estimates for HFC, PFC, SF6, NF3 and N20 Emissions from
Semiconductor Manufacture (MMT CO2 Eq. and Percent) 4-108
Table 4-98: Emissions of HFCs and PFCs from ODS Substitutes (MMT CO2 Eq.) 4-109
Table 4-99: Emissions of HFCs and PFCs from ODS Substitution (Metric Tons) 4-110
xv
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Table 4-100: Emissions of HFCs and PFCs from ODS Substitutes (MMT CO2 Eq.) by Sector 4-110
Table 4-101: Approach 2 Quantitative Uncertainty Estimates for HFC and PFC Emissions from ODS Substitutes
(MMT CO2 Eq. and Percent) 4-113
Table 4-102: U.S. HFC Consumption (MMT CO Eq.) 4-114
Table 4-103: Averaged U.S. HFC Demand (MMT CO Eq.) 4-116
Table 4-104: SF6 Emissions from Electric Power Systems and Electrical Equipment Manufacturers (MMT CO2 Eq.)
4-118
Table 4-105: SF6 Emissions from Electric Power Systems and Electrical Equipment Manufacturers (kt) 4-118
Table 4-106: Transmission Mile Coverage (kg) and Regression Coefficients (Percent) 4-121
Table 4-107: Approach 2 Quantitative Uncertainty Estimates for SF6 Emissions from Electrical Transmission and
Distribution (MMT CO2 Eq. and Percent) 4-123
Table 4-108: N2O Production (kt) 4-125
Table 4-109: N20 Emissions from N20 Product Usage (MMT CO2 Eq. and kt) 4-125
Table 4-110: Approach 2 Quantitative Uncertainty Estimates for N20 Emissions from N20 Product Usage (MMT
CO2 Eq. and Percent) 4-127
Table 4-111: NOx, CO, and NMVOC Emissions from Industrial Processes and Product Use (kt) 4-128
Table 5-1: Emissions from Agriculture (MMT CO2 Eq.) 5-2
Table 5-2: Emissions from Agriculture (kt) 5-2
Table 5-3: CH4 Emissions from Enteric Fermentation (MMT CO2 Eq.) 5-4
Table 5-4: CH4 Emissions from Enteric Fermentation (kt) 5-4
Table 5-5: Cattle Sub-Population Categories for 2016 Population Estimates 5-6
Table 5-6: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from Enteric Fermentation (MMT
CO2 Eq. and Percent) 5-8
Table 5-7: CH4 and N2O Emissions from Manure Management (MMT CO2 Eq.) 5-11
Table 5-8: CH4 and N20 Emissions from Manure Management (kt) 5-11
Table 5-9: Approach 2 Quantitative Uncertainty Estimates for CH4 and N20 (Direct and Indirect) Emissions from
Manure Management (MMT CO2 Eq. and Percent) 5-15
Table 5-10: IPCC (2006) Implied Emission Factor Default Values Compared with Calculated Values for CH4 from
Manure Management (kg/head/year) 5-16
Table 5-11: CH4 Emissions from Rice Cultivation (MMT CO2 Eq.) 5-18
Table 5-12: CH4 Emissions from Rice Cultivation (kt) 5-18
Table 5-13: Rice Area Harvested (1,000 Hectares) 5-20
Table 5-14: Average Ratooned Area as Percent of Primary Growth Area (Percent) 5-20
Table 5-15: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from Rice Cultivation (MMT CO2
Eq. and Percent) 5-22
Table 5-16: N20 Emissions from Agricultural Soils (MMT CO2 Eq.) 5-25
Table 5-17: N20 Emissions from Agricultural Soils (kt) 5-25
Table 5-18: Direct N20 Emissions from Agricultural Soils by Land Use Type and N Input Type (MMT CO2 Eq.)....
5-25
Table 5-19: Indirect N20 Emissions from Agricultural Soils (MMT CO2 Eq.) 5-26
xvi DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Table 5-20: Quantitative Uncertainty Estimates of N20 Emissions from Agricultural Soil Management in 2016
2 (MMT CO2 Eq. and Percent) 5-40
3 Table 5-21: Emissions from Liming (MMT CO2 Eq.) 5-42
4 Table 5-22: Emissions from Liming (MMT C) 5-42
5 Table 5-23: Applied Minerals (MMT) 5-44
6 Table 5-24: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Liming (MMT CO2 Eq. and
7 Percent) 5-44
8 Table 5-25: CO2 Emissions from Urea Fertilization (MMT CO2 Eq.) 5-45
9 Table 5-26: CO2 Emissions from Urea Fertilization (MMT C) 5-45
10 Table 5-27: Applied Urea (MMT) 5-46
11 Table 5-28: Quantitative Uncertainty Estimates for CO2 Emissions from Urea Fertilization (MMT CO2 Eq. and
12 Percent) 5-46
13 Table 5-29: CH4 and N20 Emissions from Field Burning of Agricultural Residues (MMT CO2 Eq.) 5-47
14 Table 5-30: CH4, N20, CO, and NOx Emissions from Field Burning of Agricultural Residues (kt) 5-48
15 Table 5-31: Agricultural Crop Production (kt of Product) 5-50
16 Table 5-32: U.S. Average Percent Crop Area Burned by Crop (Percent) 5-50
17 Table 5-33: Key Assumptions for Estimating Emissions from Field Burning of Agricultural Residues 5-51
18 Table 5-34: Greenhouse Gas Emission Ratios and Conversion Factors 5-51
19 Table 5-35: Approach 2 Quantitative Uncertainty Estimates for CH4 and N20 Emissions from Field Burning of
20 Agricultural Residues (MMT CO2 Eq. and Percent) 5-51
21 Table 6-1: Net CO2 Flux from Land Use, Land-Use Change, and Forestry (MMT CO2 Eq.) 6-2
22 Table 6-2: Emissions from Land Use, Land-Use Change, and Forestry by Gas (MMT CO2 Eq.) 6-3
23 Table 6-3: Emissions and Removals (Net Flux) from Land Use, Land-Use Change, and Forestry (MMT CO2 Eq.)....
24 6-4
25 Table 6-4: Emissions and Removals from Land Use, Land-Use Change, and Forestry (MMT CO2 Eq.) 6-5
26 Table 6-5: Emissions and Removals from Land Use, Land-Use Change, and Forestry (kt) 6-6
27 Table 6-6: Managed and Unmanaged Land Area by Land-Use Categories for All 50 States (Thousands of Hectares)
28 6-9
29 Table 6-7: Land Use and Land-Use Change for the U.S. Managed Land Base for All 50 States (Thousands of
30 Hectares) 6-10
31 Table 6-8: Data Sources Used to Determine Land Use and Land Area for the Conterminous United States, Hawaii,
32 and Alaska 6-16
33 Table 6-9: Total Land Area (Hectares) by Land-Use Category for U.S. Territories 6-22
34 Table 6-10: Net CO2 Flux from Forest Pools in Forest Land Remaining Forest Land and Harvested Wood Pools
35 (MMT C02 Eq.) 6-26
36 Table 6-11: Net C Flux from Forest Pools in Forest Land Remaining Forest Land and Harvested Wood Pools
37 (MMT C) 6-26
38 Table 6-12: Forest Area (1,000 ha) and C Stocks in Forest Land Remaining Forest Land and Harvested Wood
39 Pools (MMT C) 6-27
40 Table 6-13: Estimates of CO2 (MMT per Year) Emissions from Forest Fires in the Conterminous 48 States and
41 Alaska3 6-29
xvii
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Table 6-14: Quantitative Uncertainty Estimates for Net CO2 Flux from Forest Land Remaining Forest Land:
Changes in Forest C Stocks (MMT CO2 Eq. and Percent) 6-32
Table 6-15: Mean C Stocks, CO2 and CH4 Fluxes in Alaska between 2000 and 2009 6-35
Table 6-16: Non-CCh Emissions from Forest Fires (MMT CO2 Eq.)a 6-35
Table 6-17: Non-C02 Emissions from Forest Fires (kt)a 6-36
Table 6-18: Quantitative Uncertainty Estimates of Non-CCh Emissions from Forest Fires (MMT CO2 Eq. and
Percent)3 6-36
Table 6-19: N2O Fluxes from Soils in Forest Land Remaining Forest Land and Land Converted to Forest Land
(MMT C02 Eq. and kt N O) 6-37
Table 6-20: 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) 6-39
Table 6-21: Estimated CO2 and Non-CCh Emissions on Drained Organic Forest Soils3 (MMT CO2 Eq.) 6-40
Table 6-22: Estimated C (MMT C) and Non-C02 (kt) Emissions on Drained Organic Forest Soils3 6-40
Table 6-23: States identified as having Drained Organic Soils, Area of Forest on Drained Organic Soils, and
Sampling Error 6-41
Table 6-24: Quantitative Uncertainty Estimates for Annual CO2 and Non-C02 Emissions on Drained Organic Forest
Soils (MMT CO2 Eq. and Percent)3 6-42
Table 6-25: Net CO2 Flux from Forest C Pools in Land Converted to Forest Land by Land Use Change Category
(MMT C02 Eq.) 6-43
Table 6-26: Net C Flux from Forest C Pools in Land Converted to Forest Land by Land Use Change Category
(MMT C) 6-44
Table 6-27: Quantitative Uncertainty Estimates for Forest C Pool Stock Changes (MMT CO2 Eq. per Year) in 2016
fxomLand Converted to Forest Landby Land Use Change 6-46
Table 6-28: Net CO2 Flux from Soil C Stock Changes in Cropland Remaining Cropland (MMT CO2 Eq.) 6-49
Table 6-29: Net CO2 Flux from Soil C Stock Changes in Cropland Remaining Cropland (MMT C) 6-49
Table 6-30: Approach 2 Quantitative Uncertainty Estimates for Soil C Stock Changes occurring within Cropland
Remaining Cropland (MMT CO2 Eq. and Percent) 6-55
Table 6-31: 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.) 6-58
Table 6-32: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes in Land Converted to
Cropland (MMT C) 6-58
Table 6-33: 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) 6-62
Table 6-34: Net CO2 Flux from Soil C Stock Changes in Grassland Remaining Grassland (MMT CO2 Eq.) 6-64
Table 6-35: Net CO2 Flux from Soil C Stock Changes in Grassland Remaining Grassland (MMT C) 6-64
Table 6-36: Approach 2 Quantitative Uncertainty Estimates for C Stock Changes Occurring Within Grassland
Remaining Grassland (MMT CO2 Eq. and Percent) 6-68
Table 6-37: CH4 and N20 Emissions from Biomass Burning in Grassland (MMT CO2 Eq.) 6-69
Table 6-38: CH4, N20, CO, and NOx Emissions from Biomass Burning in Grassland (kt) 6-70
Table 6-39: Thousands of Grassland Hectares Burned Annually 6-70
Table 6-40: Uncertainty Estimates for Non-C02 Greenhouse Gas Emissions from Biomass Burning in Grassland
(MMT CO2 Eq. and Percent) 6-71
xviii DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Table 6-41: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for Land Converted to
Grassland (MMT CO2 Eq.) 6-73
Table 6-42: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for Land Converted to
Grassland (MMT C) 6-74
Table 6-43: 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) 6-77
Table 6-44: Emissions from PeatlandsRemaining Peatlands (MMT CO2 Eq.) 6-80
Table 6-45: Emissions from Peatlands Remaining Peatlands (kt) 6-80
Table 6-46: Peat Production of Lower 48 States (kt) 6-82
Table 6-47: Peat Production of Alaska (Thousand Cubic Meters) 6-82
Table 6-48: Approach 2 Quantitative Uncertainty Estimates for CO2, CH4, and N20 Emissions from Peatlands
Remaining Peatlands (MMT CO2 Eq. and Percent) 6-83
Table 6-49: Net CO2 Flux from Soil C Stock Changes in Vegetated Coastal Wetlands Remaining Vegetated Coastal
Wetlands (MMT CO2 Eq.) 6-86
Table 6-50: Net CO2 Flux from Soil C Stock Changes in Vegetated Coastal Wetlands Remaining Vegetated Coastal
Wetlands (MMT C) 6-86
Table 6-51: Net CH4 Flux from Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands (MMT CO2
Eq.) 6-86
Table 6-52: Net CH4 Flux from Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands (kt CH4). 6-86
Table 6-53: Approach 1 Quantitative Uncertainty Estimates for Emissions from C Stock Changes occurring within
Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands (MMT CO2 Eq. and Percent) 6-88
Table 6-54: Approach 1 Quantitative Uncertainty Estimates for CH4 Emissions occurring within Vegetated Coastal
Wetlands Remaining Vegetated Coastal Wetlands (MMT CO2 Eq. and Percent) 6-88
Table 6-55: Net CO2 Flux from Soil C Stock Changes in Vegetated Coastal Wetlands Converted to Unvegetated
Open Water Coastal Wetlands (MMT CO2 Eq.) 6-89
Table 6-56: Net CO2 Flux from Soil C Stock Changes in Vegetated Coastal Wetlands Converted to Unvegetated
Open Water Coastal Wetlands (MMT C) 6-89
Table 6-57: Approach 1 Quantitative Uncertainty Estimates for Net CO2 Flux Occurring within Vegetated Coastal
Wetlands Converted to Unvegetated Open Water Coastal Wetlands (MMT CO2 Eq. and Percent) 6-91
Table 6-58: Net CO2 Flux from Soil C Stock Changes from Unvegetated Open Water Coastal Wetlands Converted
to Vegetated Coastal Wetlands (MMT CO2 Eq.) 6-92
Table 6-59: Net CO2 Flux from Soil C Stock Changes from Unvegetated Open Water Coastal Wetlands Converted
to Vegetated Coastal Wetlands (MMT C) 6-92
Table 6-60: Approach 1 Quantitative Uncertainty Estimates for C Stock Changes Occurring within Unvegetated
Open Water Coastal Wetlands Converted to Vegetated Coastal Wetlands (MMT CO2 Eq. and Percent) 6-94
Table 6-61: Net N20 Emissions from Aquaculture in Coastal Wetlands (MMT CO2 Eq.) 6-95
Table 6-62: Net N20 Emissions from Aquaculture in Coastal Wetlands (kt N20) 6-95
Table 6-63: Approach 1 Quantitative Uncertainty Estimates for N20 Emissions for Aquaculture Production in
Coastal Wetlands (MMT CO2 Eq. and Percent) 6-96
Table 6-64: Net CO2 Flux from Soil C Stock Changes in Land Converted to Vegetated Coastal Wetlands (MMT
C02 Eq.) 6-97
Table 6-65: Net CO2 Flux from Soil C Stock Changes in Land Converted to Vegetated Coastal Wetlands (MMT C)
6-97
xix
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Table 6-66: Net CH4 Flux in Land Converted to Vegetated Coastal Wetlands (MMT CO2 Eq.) 6-97
Table 6-67: Net CH4 Flux from Soil C Stock Changes in Land Converted to Vegetated Coastal Wetlands (kt CH4)...
6-97
Table 6-68: Approach 1 Quantitative Uncertainty Estimates for Net CO2 Flux Changes occurring within Land
Converted to Vegetated Coastal Wetlands (MMT CO2 Eq. and Percent) 6-98
Table 6-69: Approach 1 Quantitative Uncertainty Estimates for CH4 Emissions occurring w ithin Land Converted to
Vegetated Coastal Wetlands (MMT CO2 Eq. and Percent) 6-99
Table 6-70: Net CO2 Flux from Soil C Stock Changes in Settlements Remaining Settlements (MMT CO2 Eq.)
6-100
Table 6-71: Net CO2 Flux from Soil C Stock Changes in Settlements Remaining Settlements (MMT C) 6-100
Table 6-72: Thousands of Hectares of Drained Organic Soils in Settlements Remaining Settlements 6-101
Table 6-73: Uncertainty Estimates for CO2 Emissions from Drained Organic Soils in Settlements Remaining
Settlements (MMT CO2 Eq. and Percent) 6-101
Table 6-74: Net C Flux from Urban Trees (MMT CO Eq. and MMT C) 6-103
Table 6-75: Annual C Sequestration (Metric Tons C7Year), Tree Cover (Percent), and Annual C Sequestration per
Area of Tree Cover (kg C/m2-yr) for 50 states plus the District of Columbia (2016) 6-105
Table 6-76: Approach 2 Quantitative Uncertainty Estimates for Net C Flux from Changes in C Stocks in Urban
Trees (MMT CO2 Eq. and Percent) 6-106
Table 6-77: N2O Emissions from Soils in Settlements Remaining Settlements (MMT CO2 Eq. and kt N2O) 6-108
Table 6-78: Quantitative Uncertainty Estimates of N2O Emissions from Soils in Settlements Remaining Settlements
(MMT CO2 Eq. and Percent) 6-110
Table 6-79: Net Changes in Yard Trimmings and Food Scrap Carbon Stocks in Landfills (MMT CO2 Eq.) 6-112
Table 6-80: Net Changes in Yard Trimmings and Food Scrap Carbon Stocks in Landfills (MMT C) 6-112
Table 6-81: Moisture Contents, C Storage Factors (Proportions of Initial C Sequestered), Initial C Contents, and
Decay Rates for Yard Trimmings and Food Scraps in Landfills 6-114
Table 6-82: C Stocks in Yard Trimmings and Food Scraps in Landfills (MMT C) 6-115
Table 6-83: Approach 2 Quantitative Uncertainty Estimates for CO2 Flux from Yard Trimmings and Food Scraps in
Landfills (MMT CO2 Eq. and Percent) 6-115
Table 6-84: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for Land Converted to
Settlements (MMT CO2 Eq.) 6-117
Table 6-85: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for Land Converted to
Settlements (MMT C) 6-117
Table 6-86: 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) 6-120
Table 7-1: Emissions from Waste (MMT CO2 Eq.) 7-1
Table 7-2: Emissions from Waste (kt) 7-2
Table 7-3: CH4 Emissions from Landfills (MMT CO2 Eq.) 7-4
Table 7-4: CH4 Emissions from Landfills (kt) 7-5
Table 7-5: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from Landfills (MMT CO2 Eq. and
Percent) 7-13
Table 7-6: Materials Discarded3 in the Municipal Waste Stream by Waste Type from 1990 to 2014 (Percent)b .. 7-17
Table 7-7: CH4 and N20 Emissions from Domestic and Industrial Wastewater Treatment (MMT CO2 Eq.) 7-20
xx DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Table 7-8: CH4 and N20 Emissions from Domestic and Industrial Wastewater Treatment (kt) 7-20
Table 7-9: U.S. Population (Millions) and Domestic Wastewater BOD5 Produced (kt) 7-22
Table 7-10: Domestic Wastewater CH4 Emissions from Septic and Centralized Systems (2016, MMT CO2 Eq. and
Percent) 7-23
Table 7-11: Industrial Wastewater CH4 Emissions by Sector (2016, MMT CO2 Eq. and Percent) 7-24
Table 7-12: U.S. Pulp and Paper, Meat, Poultry, Vegetables, Fruits and Juices, Ethanol, and Petroleum Refining
Production (MMT) 7-24
Table 7-13: Variables Used to Calculate Percent Wastewater Treated Anaerobically by Industry (Percent) 7-25
Table 7-14: Wastewater Flow (m3/ton) and BOD Production (g/L) for U.S. Vegetables, Fruits, and Juices
Production 7-27
Table 7-15: U.S. Population (Millions), Population Served by Biological Denitrification (Millions), Fraction of
Population Served by Wastewater Treatment (percent), Available Protein (kg/person-year), Protein Consumed
(kg/person-year), and Nitrogen Removed with Sludge (kt-N/year) 7-30
Table 7-16: Approach 2 Quantitative Uncertainty Estimates for CH4 Emissions from Wastewater Treatment (MMT
CO2 Eq. and Percent) 7-31
Table 7-17: CH4 and N20 Emissions from Composting (MMT CO2 Eq.) 7-34
Table 7-18: CH4 and N20 Emissions from Composting (kt) 7-34
Table 7-19: U.S. Waste Composted (kt) 7-34
Table 7-20: Approach 1 Quantitative Uncertainty Estimates for Emissions from Composting (MMT CO2 Eq. and
Percent) 7-35
Table 7-21: Emissions of NOx, CO, and NMVOC from Waste (kt) 7-36
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
Figure ES-1: Gross U.S. Greenhouse Gas Emissions by Gas (MMT CO2 Eq.) ES-4
Figure ES-2: Annual Percent Change in Gross U.S. Greenhouse Gas Emissions Relative to the Previous Year ..ES-5
Figure ES-3: Cumulative Change in Annual Gross U.S. Greenhouse Gas Emissions Relative to 1990 (1990=0,
MM I CO- Eq.) ES-5
Figure ES-4: 2016 U.S. Greenhouse Gas Emissions by Gas (Percentages based on MMT CO2 Eq.) ES-9
Figure ES-5: 2016 Sources of CO2 Emissions (MMT CO2 Eq.) ES-10
Figure ES-6: 2016 CO2 Emissions from Fossil Fuel Combustion by Sector and Fuel Type (MMT C02Eq.) ES-11
Figure ES-7: 2016 End-Use Sector Emissions of CO2 from Fossil Fuel Combustion (MMT CO2 Eq.) ES-11
Figure ES-8: Electric Power Generation (Billion kWh) and Emissions (MMT CO2 Eq.) ES-13
Figure ES-9: 2016 Sources of CH4 Emissions (MMT CO2 Eq.) ES-15
Figure ES-10: 2016 Sources of N20 Emissions (MMT CO2 Eq.) ES-16
Figure ES-11: 2016 Sources of HFCs, PFCs, SF6, and NF3 Emissions (MMT CO2 Eq.) ES-17
Figure ES-12: U.S. Greenhouse Gas Emissions and Sinks by Chapter/IPCC Sector (MMT CO2 Eq.) ES-18
Figure ES-13: 2016 U.S. Energy Consumption by Energy Source (Percent) ES-20
xxi
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Figure ES-14: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO2 Eq.) ES-24
Figure ES-15: U.S. Greenhouse Gas Emissions with Electricity-Related Emissions Distributed to Economic Sectors
(MMTCO2 Eq.) ES-26
Figure ES-16: U.S. Greenhouse Gas Emissions Per Capita and Per Dollar of Gross Domestic Product (GDP)
ES-27
Figure ES-17: 2016 Key Categories (MMT CO2 Eq.) ES-28
Figure 1-1: National Inventory Arrangements Diagram Inventory Process 1-12
Figure 1-2: U.S. QA/QC Plan Summary 1-21
Figure 2-1: Gross U.S. Greenhouse Gas Emissions by Gas (MMT CO2 Eq.) 2-1
Figure 2-2: Annual Percent Change in Gross U.S. Greenhouse Gas Emissions Relative to the Previous Year 2-2
Figure 2-3: Cumulative Change in Annual Gross U.S. Greenhouse Gas Emissions Relative to 1990 (1990=0, MMT
C02 Eq.) 2-2
Figure 2-4: U.S. Greenhouse Gas Emissions and Sinks by Chapter/IPCC Sector (MMT CO2 Eq.) 2-7
Figure 2-5: 2016 Energy Chapter Greenhouse Gas Sources (MMT CO2 Eq.) 2-9
Figure 2-6: 2016 U.S. Fossil Carbon Flows (MMT CO2 Eq.) 2-10
Figure 2-7: 2016 CO2 Emissions from Fossil Fuel Combustion by Sector and Fuel Type (MMT CO2 Eq.) 2-13
Figure 2-8: 2016 End-Use Sector Emissions of CO2 from Fossil Fuel Combustion (MMT CO2 Eq.) 2-13
Figure 2-9: Electric Power Generation (Billion kWh) and Emissions (MMT CO2 Eq.) 2-14
Figure 2-10: 2016 Industrial Processes and Product Use Chapter Greenhouse Gas Sources (MMT CO2 Eq.) 2-15
Figure 2-11: 2016 Agriculture Chapter Greenhouse Gas Sources (MMT CO2 Eq.) 2-18
Figure 2-12: 2016 LULUCF Chapter Greenhouse Gas Sources and Sinks (MMT CO2 Eq.) 2-20
Figure 2-13: 2016 Waste Chapter Greenhouse Gas Sources (MMT CO2 Eq.) 2-22
Figure 2-14: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO2 Eq.) 2-24
Figure 2-15: U.S. Greenhouse Gas Emissions with Electricity-Related Emissions Distributed to Economic Sectors
(MMT C02 Eq.) 2-27
Figure 2-16: U.S. Greenhouse Gas Emissions Per Capita and Per Dollar of Gross Domestic Product 2-34
Figure 3-1: 2016 Energy Chapter Greenhouse Gas Sources (MMT CO2 Eq.) 3-1
Figure 3-2: 2016 U.S. Fossil Carbon Flows (MMT CO2 Eq.) 3-2
Figure 3-3: 2016 U.S. Energy Consumption by Energy Source (Percent) 3-8
Figure 3-4: U.S. Energy Consumption (Quadrillion Btu) 3-8
Figure 3-5: 2016 CO2 Emissions from Fossil Fuel Combustion by Sector and Fuel Type (MMT CO2 Eq.) 3-9
Figure 3-6: Annual Deviations from Normal Heating Degree Days for the United States (1950-2016, Index Normal
= 100) 3-10
Figure 3-7: Annual Deviations from Normal Cooling Degree Days for the United States (1950-2016, Index Normal
= 100) 3-10
Figure 3-8: Fuels Used in Electric Power Generation (TBtu) and Total Electric Power Sector CO2 Emissions.... 3-16
Figure 3-9: Electric Power Retail Sales by End-Use Sector (Billion kWh) 3-16
Figure 3-10: Industrial Production Indices (Index 2012=100) 3-18
xxii DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Figure 3-11: Fuels Used in Residential and Commercial Sectors (TBtu), Heating Degree Days, and Total Sector
CO2 Emissions 3-19
Figure 3-12: Fuels Used in Transportation Sector (TBtu), Onroad VMT, and Total Sector CO2 Emissions 3-21
Figure 3-13: Sales-Weighted Fuel Economy of New Passenger Cars and Light-Duty Trucks, 1990-2016
(miles/gallon) 3-23
Figure 3-14: Sales of New Passenger Cars and Light-Duty Trucks, 1990-2016 (Percent) 3-23
Figure 3-15: Mobile Source CH4 and N20 Emissions (MMT CO2 Eq.) 3-26
Figure 3-16: U.S. Energy Consumption and Energy-Related CO2 Emissions Per Capita and Per Dollar GDP 3-33
Figure 4-1: 2016 Industrial Processes and Product Use Chapter Greenhouse Gas Sources (MMT CO2 Eq.) 4-2
Figure 4-2: U.S. HFC Consumption (MMT CO2 Eq.) 4-115
Figure 5-1: 2016 Agriculture Chapter Greenhouse Gas Emission Sources (MMT CO2 Eq.) 5-1
Figure 5-2: Annual CH4 Emissions from Rice Cultivation, 2012 (MMT CO2 Eq./Year)* 5-19
Figure 5-3: Sources and Pathways of N that Result in N20 Emissions from Agricultural Soil Management 5-24
Figure 5-4: Crops, 2012 Annual Direct N2O Emissions Estimated Using the Tier 3 DAYCENT Model (MMT CO2
Eq./year)* 5-27
Figure 5-5: Grasslands, 2012 Annual Direct N2O Emissions Estimated Using the Tier 3 DAYCENT Model (MMT
CO2 Eq./year)* 5-28
Figure 5-6: Crops, 2012 Annual Indirect N2O Emissions from Volatilization Using the Tier 3 DAYCENT Model
(MMT C02 Eq./year)* 5-29
Figure 5-7: Grasslands, 2012 Annual Indirect N2O Emissions from Volatilization Using the Tier 3 DAYCENT
Model (MMT CO Eq./year)* 5-30
Figure 5-8: Crops, 2012 Annual Indirect N2O Emissions from Leaching and Runoff Using the Tier 3 DAYCENT
Model (MMT CO Eq./year)* 5-31
Figure 5-9: Grasslands, 2012 Annual Indirect N20 Emissions from Leaching and Runoff Using the Tier 3
DAYCENT Model (MMT CO Eq./year)* 5-32
Figure 5-10: Comparison of Measured Emissions at Field Sites and Modeled Emissions Using the DAYCENT
Simulation Model and IPCC Tier 1 Approach (kg N20 per ha per year) 5-41
Figure 6-1: 2016 LULUCF Chapter Greenhouse Gas Sources and Sinks (MMT CO2 Eq.) 6-4
Figure 6-2: Percent of Total Land Area for Each State in the General Land-Use Categories for 2015a 6-12
Figure 6-3: Changes in Forest Area by Region for Forest Land Remaining Forest Land in the conterminous United
States and coastal Alaska (1990-2016, Million Hectares) 6-25
Figure 6-4: Estimated Net Annual Changes in C Stocks for All C Pools in Forest Land Remaining Forest Land in
the Conterminous U.S. and Coastal Alaska (1990-2016, MMT C per Year) 6-28
Figure 6-5: Total Net Annual Soil C Stock Changes for Mineral Soils under Agricultural Management within
States, 2012, Cropland Remaining Cropland * 6-50
Figure 6-6: Total Net Annual Soil C Stock Changes for Organic Soils under Agricultural Management within
States, 2012, Cropland Remaining Cropland* 6-51
Figure 6-7: Total Net Annual Soil C Stock Changes for Mineral Soils under Agricultural Management within
States, 2012, Grassland Remaining Grassland * 6-65
Figure 6-8: Total Net Annual Soil C Stock Changes for Organic Soils under Agricultural Management within
States, 2012, Grassland Remaining Grassland* 6-65
Figure 7-1: 2016 Waste Chapter Greenhouse Gas Sources (MMT CO2 Eq.) 7-1
xxiii
-------
1 Figure 7-2: Management of Municipal Solid Waste in the United States, 2014 7-16
2 Figure 7-3: MSW Management Trends from 1990 to 2014 7-16
3 Figure 7-4: Percent of Degradable Materials Diverted from Landfills from 1990 to 2014 (Percent) 7-18
4 Boxes
5 Box ES-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals ES-1
6 BoxES-2: EPA's Greenhouse Gas Reporting Program ES-2
7 Box ES-3: Improvements and Recalculations Relative to the Previous Inventory ES-5
8 Box ES-4: Use of Ambient Measurements Systems for Validation of Emission Inventories ES-14
9 Box ES-5: Recent Trends in Various U.S. Greenhouse Gas Emissions-Related Data ES-26
10 Box ES-6: Recalculations of Inventory Estimates ES-29
11 Box 1-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals 1-2
12 Box 1-2: The IPCC Fifth Assessment Report and Global Warming Potentials 1-9
13 Box 1-3: IPCC Reference Approach 1-15
14 Box 2-1: Methodology for Aggregating Emissions by Economic Sector 2-32
15 Box 2-2: Recent Trends in Various U.S. Greenhouse Gas Emissions-Related Data 2-33
16 Box 2-3: Sources and Effects of Sulfur Dioxide 2-36
17 Box 3-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals 3-4
18 Box 3-2: Energy Data from EPA's Greenhouse Gas Reporting Program 3-4
19 Box 3-3: Weather and Non-Fossil Energy Effects on CO2 from Fossil Fuel Combustion Trends 3-9
20 Box 3-4: Uses of Greenhouse Gas Reporting Program Data and Improvements in Reporting Emissions from
21 Industrial Sector Fossil Fuel Combustion 3-31
22 Box 3-5: Carbon Intensity of U.S. Energy Consumption 3-32
23 Box 3-6: Reporting of Lubricants, Waxes, and Asphalt and Road Oil Product Use in Energy Sector 3-51
24 Box 3-7: Carbon Dioxide Transport, Injection, and Geological Storage 3-76
25 Box 4-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals 4-6
26 Box 4-2: Industrial Processes Data from EPA's Greenhouse Gas Reporting Program 4-7
27 Box 5-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals 5-2
28 Box 5-2: Biennial Inventory Compilation 5-3
29 Box 5-3: Surrogate Data Method 5-21
30 Box 5-4: Tier 1 vs. Tier 3 Approach for Estimating N2O Emissions 5-33
31 Box 5-5: Surrogate Data Method 5-34
32 Box 5-6: Comparison of the Tier 2 U.S. Inventory Approach and IPCC (2006) Default Approach 5-43
33 Box 5-7: Comparison of Tier 2 U.S. Inventory Approach and IPCC (2006) Default Approach 5-49
34 Box 6-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals 6-7
35 Box 6-2: Biennial Inventory Compilation 6-8
36 Box 6-3: Preliminary Estimates of Land Use in U.S. Territories 6-21
xxiv DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Box 6-4: CO2 Emissions from Forest Fires 6-28
Box 6-5: Preliminary Estimates of Historical Carbon Stock Change and Methane Emissions from Managed Land in
Alaska (Represents Mean for Years 2000 to 2009) 6-34
Box 6-6: Surrogate Data Method 6-52
Box 6-7: Tier 3 Approach for Soil C Stocks Compared to Tier 1 or 2 Approaches 6-53
Box 6-8: Grassland Woody Biomass Analysis 6-69
Box 7-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals 7-2
Box 7-2: Waste Data from EPA's Greenhouse Gas Reporting Program 7-2
Box 7-3: Nationwide Municipal Solid Waste Data Sources 7-15
Box 7-4: Overview of the Waste Sector 7-16
Box 7-5: Description of a Modern, Managed Landfill 7-18
xxv
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Executive Summary
An emissions inventory that identifies and quantifies a country's primary 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 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
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.. ."3 The United States views this report as an opportunity
to fulfill these commitments.
This chapter summarizes the latest information on U.S. anthropogenic greenhouse gas emission trends from 1990
through 2016. To ensure that the U.S. emissions inventory is comparable to those of other UNFCCC Parties, the
estimates presented here were calculated using methodologies consistent with those recommended in the 2006
Intergovernmental Panel on Climate Change (IPCC) Guidelines for National Greenhouse Gas Inventories (IPCC
2006). The structure of this report is consistent with the UNFCCC guidelines for inventory reporting, as discussed in
BoxES-1.4
Box ES-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
1 Hie 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 UNEPAVMO Information Unit on Climate
Change. See .
3 Article 4(1 )(a) of the United Nations Framework Convention on Climate Change (also identified in Article 12). Subsequent
decisions by the Conference of the Parties elaborated the role of Annex I Parties in preparing national inventories. See
.
4 See .
Executive Summary ES-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
for National Greenhouse Gas Inventories (2006IPCC Guidelines). 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.
Box ES-2: EPA's Greenhouse Gas Reporting Program
On October 30, 2009, the U.S. Enviromnental Protection Agency (EPA) published 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). The rule
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.5 Annual reporting is at the facility level,
except for certain suppliers of fossil fuels and industrial greenhouse gases.
EPA's GHGRP dataset and the data presented in this Inventory report 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. 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 categories to improve the national estimates presented in this
Inventory consistent with IPCC guidance.6
ES.l Background Information
Greenhouse gases absorb infrared radiation, thereby trapping heat 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 other fluorine-containing halogenated substances. Although CO2, CH4, and N20 occur naturally
in the atmosphere, human activities have changed their atmospheric concentrations. From the pre-industrial era (i.e.,
ending about 1750) to 2016, concentrations of these greenhouse gases have increased globally by 44, 163, and 22
percent, respectively (IPCC 2013; NOAA/ESRL 2017a, 2017b, 2017c). This annual report estimates the total
national greenhouse gas emissions and removals associated with human activities across the United States.
Global Warming Potentials
Gases in the atmosphere can contribute to climate change both directly and indirectly. Direct effects occur when the
gas itself absorbs radiation. Indirect radiative forcing occurs when chemical transformations of the substance
produce other greenhouse gases, when a gas influences the atmospheric lifetimes of other gases, and/or when a gas
5 See and .
6 See
ES-2 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 affects atmospheric processes that alter the radiative balance of the earth (e.g., affect cloud formation or albedo).7
2 The IPCC developed the Global Warming Potential (GWP) concept to compare the ability of each greenhouse gas to
3 trap heat in the atmosphere relative to another gas.
4 The GWP of a greenhouse gas is defined as the ratio of the accumulated radiative forcing within a specific time
5 horizon caused by emitting 1 kilogram of the gas, relative to that of the reference gas CO2 (IPCC 2014). The
6 reference gas used is CO2, and therefore GWP-weighted emissions are measured in million metric tons of CO2
7 equivalent (MMT CO2 Eq.).8 9 All gases in this Executive Summary are presented in units of MMT CO2 Eq.
8 Emissions by gas in unweighted mass kilotons are provided in the Trends chapter of this report.
9 UNFCCC reporting guidelines for national inventories require the use of GWP values from the IPCC Fourth
10 Assessment Report (AR4) (IPCC 2007).10 All estimates are provided throughout the report in both CO2 equivalents
11 and unweighted units. A comparison of emission values using the AR4 GWP values versus the SAR (IPCC 1996),
12 and the IPCC Fifth Assessment Report (AR5) (IPCC 2013) GWP values can be found in Chapter 1 and, in more
13 detail, in Annex 6.1 of this report. The GWP values used in this report are listed below in Table ES-1.
14 Table ES-1: Global Warming Potentials (100-Year Time Horizon) Used in this Report
Gas
GWP
CO2
1
CH4a
25
N2O
298
HFC-23
14,800
HFC-32
675
HFC-125
3,500
HFC-134a
1,430
HFC-143a
4,470
HFC-152a
124
HFC-227ea
3,220
HFC-236fa
9,810
HFC-4310mee
1,640
CF4
7,390
c2f6
12,200
C4F10
8,860
C6Fl4
9,300
SFe
22,800
NF3
17,200
a The CH4 GWP 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 CO2 is not included.
Source: IPCC (2007)
15
7 Albedo is a measure of the Earth's reflectivity, and is defined as the fraction of the total solar radiation incident on a body that
is reflected by it.
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 .
Executive Summary ES-3
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
ES.2 Recent Trends in U.S. Greenhouse Gas
Emissions and Sinks
In 2016, total gross U.S. greenhouse gas emissions were 6,546.2 million metric tons (MMT) of CO2 Eq. Total U.S.
emissions have increased by 2.8 percent from 1990 to 2016, and emissions decreased from 2015 to 2016 by 2.0
percent (131.1 MMT CO2 Eq.). The decrease in total greenhouse gas emissions between 2015 and 2016 was driven
in large part by a decrease in CO2 emissions from fossil fuel combustion. The decrease in CO2 emissions from fossil
fuel combustion was a result of multiple factors, including:
(1) substitution from coal to natural gas and other sources in the electric power sector; and
(2) warmer winter conditions in 2016 resulting in a decreased demand for heating fuel in the residential and
commercial sectors.
Relative to 1990, the baseline fortius Inventory, gross emissions in 2016 are higher by 2.8 percent, down from a
high of 15.6 percent above 1990 levels in 2007. Overall, net emissions in 2016 were 11.6 percent below 2005 levels
as shown in Table ES-2. Figure ES-1 through Figure ES-3 illustrate the overall trends in total U.S. emissions by gas,
annual changes, and absolute change since 1990, and Table ES-2 provides a detailed summary of gross U.S.
greenhouse gas emissions and sinks for 1990 through 2016. Note, unless otherwise stated, all tables and figures
provide total emissions without LULUCF.
Figure ES-1: Gross U.S. Greenhouse Gas Emissions by Gas (MMT CO2 Eq.)
9,000
8,000
7,000
6,000
&
LU
0 5,000
u
1-
1 4,000
3,000
2,000
1,000
0
I HFCs, PFCs, SF, and NF. Subtotal
Nitrous Oxide
I Methane
I Carbon Dioxide
ES-4 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Figure ES-2: Annual Percent Change in Gross U.S. Greenhouse Gas Emissions Relative to the
Previous Year
Figure ES-3: Cumulative Change in Annual Gross U.S. Greenhouse Gas Emissions Relative to
1990 (1990=0, MMT COz Eq.)
1,200-
1,100-
Box ES-3: Improvements and Recalculations Relative to the Previous Inventory
Each year, some emission and sink estimates in the Inventory are recalculated with improved methods and/or data.
These improvements are also implemented consistently across the previous Inventory's time series (i.e., 1990 to
2015) to ensure that the trend is accurate (see also Box ES-6 on the recalculations approach). Below are categories
with recalculations resulting in an average change over the time series of greater than 10 MMT CO2 Eq. For more
information on specific improvements, please see the Recalculations and Improvements chapter (Chapter 9) and the
Energy chapter (Chapter 3).
• Fossil Fuel Combustion (CO2). Average increase of 14.0 MMT CO2 Eq. relative to the previous Inventory,
resulting primarily from incorporation of updated energy consumption statistics from EIA.
• Petroleum Systems (CO2). Average increase of 13.8 MMT CO2 Eq. relative to the previous Inventory,
resulting primarily from reallocation of CO2 from flaring to petroleum systems from natural gas systems.
Executive Summary ES-5
-------
1 • Petroleum Systems (CH4). Average decrease of 10.9 MMT CO2 Eq. relative to the previous Inventory,
2 resulting primarily from recalculation of associated gas venting and flaring emissions using a basin-level
3 approach.
4 • Natural Gas Systems (CO2). Average decrease of 10.3 MMT CO2 Eq. relative to the previous Inventory,
5 resulting primarily from reallocation of CO2 from flaring to petroleum systems from natural gas systems.
6 Other improvements of note include recalculations of CH4 estimates from Municipal Solid Waste (MSW) Landfills
7 (See Section 7.1 of the Waste chapter).
8
9 Table ES-2: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks (MMT CO2 Eq.)
Gas/Source
1990
2005
2012
2013
2014
2015
2016
CO2
5,136.8
6,150.8
5,383.7
5,541.7
5,590.5
5,449.5
5,333.4
Fossil Fuel Combustion
4,755.8
5,759.1
5,029.8
5,162.3
5,206.1
5,059.3
4,'976.7
Electric Power
1,820.8
2,400.9
2,022.2
2,038.1
2,038.0
1,900.7
1,808.8
Transportation
1,467.2
1,855.8
1,661.9
1,677.6
1,717.1
1,735.5
1,794.9
Industrial
874.5
867.8
818.4
848.7
830.8
819.3
807.6
Residential
338.3
357.8
282.5
329.7
345.3
316.8
296.2
Commercial
227.4
227.0
201.3
225.7
233.6
245.6
227.9
U.S. Territories
27.6
49.7
43.5
42.5
41.4
41.4
41.4
Non-Energy Use of Fuels
119.6
141.7
113.3
133.2
127.8
135.1
121.0
Iron and Steel Production &
Metallurgical Coke Production
101.5
68.0
55.4
53.3
58.2
47.7
42.2
Cement Production
33.5
46.2
35.3
36.4
39.4
39.9
39.4
Petrochemical Production
21.2
26.8
26.5
26.4
26.5
28.1
27.4
Natural Gas Systems
29.7
22.5
24.4
26.0
27.0
26.3
26.7
Petroleum Systems
9.4
17.0
25.6
29.7
32.9
38.0
25.5
Lime Production
11.7
14.6
13.8
14.0
14.2
13.3
13.3
Other Process Uses of Carbonates
4.9
6.3
8.0
10.4
11.8
11.2
11.2
Ammonia Production
13.0
9.2
9.4
10.0
9.6
10.6
11.2
Incineration of Waste
8.0
12.5
10.4
10.4
10.6
10.7
10.7
Urea Fertilization
2.4
3.5
4.3
4.4
4.5
4.9
5.1
Carbon Dioxide Consumption
1.5
1.4
4.0
4.2
4.5
4.5
4.5
Urea Consumption for Non-
Agricultural Purposes
3.8
3.7
4.4
4.1
1.5
4.2
4.0
Liming
4.7
4.3
6.0
3.9
3.6
3.8
3.9
Ferroalloy Production
2.2
1.4
1.9
1.8
1.9
2.0
1.8
Soda Ash Production
1.4
1.7
1.7
1.7
1.7
1.7
1.7
Titanium Dioxide Production
1.2
1.8
1.5
1.7
1.7
1.6
1.6
Aluminum Production
6.8
4.1
3.4
3.3
2.8
2.8
1.3
Glass Production
1.5
1.9
1.2
1.3
1.3
1.3
1.3
Phosphoric Acid Production
1.5
1.3
1.1
1.1
1.0
1.0
1.0
Zinc Production
0.6
1.0
1.5
1.4
1.0
0.9
0.9
Lead Production
0.5
0.6
0.5
0.5
0.5
0.5
0.5
Silicon Carbide Production and
Consumption
0.4
0.2
0.2
0.2
0.2
0.2
0.2
Magnesium Production and Processing
+
+
+
+
+
+
+
Wood Biomass, Ethanol, and Biodiesel
Consumption"
219.4
230.7
276.2
299.8
308.3
294.5
291.1
International Bunker Fuelsb
103.5
113.1
105.8
99.8
103.4
110.9
114.4
CH4c
778.1
679.3
661.3
659.6
665.3
664.0
655.8
Enteric Fermentation
164.2
168.9
166.7
165.5
164.2
166.5
170.1
Natural Gas Systems
193.7
160.0
156.8
159.6
164.2
164.4
162.1
ES-6 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
Landfills
179.6
132.7
117.0
113.3
112.7
111.7
107.7
Manure Management
37.2
56.3
65.6
63.3
62.9
66.3
67.7
Coal Mining
96.5
64.1
66.5
64.6
64.6
61.2
53.8
Petroleum Systems
42.3
34.7
35.4
38.8
41.0
39.4
39.3
Wastewater Treatment
15.7
15.8
15.1
14.9
15.0
15.1
14.8
Rice Cultivation
16.0
16.7
11.3
11.5
12.7
12.3
13.7
Stationary Combustion
8.6
7.9
7.3
8.7
8.8
7.8
7.2
Abandoned Oil and Gas Wells
6.5
6.9
7.0
7.0
7.1
7.2
7.1
Abandoned Underground Coal Mines
7.2
6.6
6.2
6.2
6.3
6.4
6.7
Mobile Combustion
9.8
6.6
4.0
3.7
3.4
3.1
3.0
Composting
0.4
1.9
1.9
2.0
2.1
2.1
2.1
Field Burning of Agricultural Residues
0.2
0.2
0.3
0.3
0.3
0.3
0.3
Petrochemical Production
0.2
0.1
0.1
0.1
0.1
0.2
0.2
Ferroalloy Production
+
+
+
+
+
+
+
Silicon Carbide Production and
Consumption
+
+
+
+
+
+
+
Iron and Steel Production &
Metallurgical Coke Production
+
+
+
+
+
+
+
Incineration of Waste
+
+
+
+
+
+
+
International Bunker Fuelsb
0.2
0.1
0.1
0.1
0.1
0.1
0.1
N2Oc
354.6
357.4
335.2
362.6
360.5
378.9
368.8
Agricultural Soil Management
250.5
253.5
247.9
276.6
274.0
295.0
283.6
Stationary Combustion
11.1
17.5
16.8
18.6
18.9
18.0
18.4
Manure Management
14.0
16.5
17.5
17.5
17.5
17.7
18.1
Mobile Combustion
41.5
38.4
23.8
22.0
20.2
18.8
17.8
Nitric Acid Production
12.1
11.3
10.5
10.7
10.9
11.6
10.2
Adipic Acid Production
15.2
7.1
5.5
3.9
5.4
4.3
7.0
Wastewater Treatment
3.4
4.4
4.6
4.7
4.8
4.8
5.0
NjO 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
2.0
2.0
2.0
2.0
2.0
Composting
0.3
1.7
1.7
1.8
1.9
1.9
1.9
Incineration of Waste
0.5
0.4
0.3
0.3
0.3
0.3
0.3
Semiconductor Manufacture
+
0.1
0.2
0.2
0.2
0.2
0.2
Field Burning of Agricultural Residues
0.1
0.1
0.1
0.1
0.1
0.1
0.1
International Bunker Fuelsb
0.9
1.0
0.9
0.9
0.9
0.9
1.0
HFCs
46.6
120.0
156.0
159.1
166.8
173.3
177.1
Substitution of Ozone Depleting
Substances'1
0.3
99.8
150.3
154.8
161.4
168.6
173.9
HCFC-22 Production
46.1
20.0
5.5
4.1
5.0
4.3
2.8
Semiconductor Manufacture
0.2
0.2
0.2
0.2
0.3
0.3
0.3
Magnesium Production and Processing
0.0
0.0
+
0.1
0.1
0.1
0.1
PFCs
24.3
6.7
5.9
5.8
5.6
5.1
4.3
Semiconductor Manufacture
2.8
3.3
3.0
2.8
3.1
3.1
3.0
Aluminum Production
21.5
3.4
2.9
3.0
2.5
2.0
1.4
Substitution of Ozone Depleting
Substances
0.0
+
+
+
+
+
+
SF«
28.8
11.7
6.6
6.3
6.3
5.9
6.2
Electrical Transmission and
Distribution
23.1
8.3
4.6
4.5
4.6
4.2
4.3
Magnesium Production and Processing
5.2
2.7
1.6
1.5
1.0
0.9
1.0
Semiconductor Manufacture
0.5
0.7
0.3
0.4
0.7
0.7
0.8
NF3
+
0.5
0.6
0.6
0.5
0.6
0.6
Semiconductor Manufacture
+
0.5
0.6
0.6
0.5
0.6
0.6
Executive Summary ES-7
-------
1
2
3
4
5
6
7
8
9
10
11
12
Total Emissions
6,369.2
7,326.4
6,549.4
6,735.6
6,795.6
6,677.3
6,546.2
LULUCF Emissions0
10.6
23.0
26.1
19.2
19.6
38.2
38.1
LULUCF CH4 Emissions
6.7
13.3
15.0
10.9
11.2
22.4
22.4
LULUCF N2O Emissions
3.9
H
11.1
8.3
8.4
15.8
15.7
LULUCF Carbon Stock Change6
(830.2)
(754.2)
(779.5)
(755.0)
(760.0)
(733.4)
(754.9)
LULUCF Sector Net Total'
(819.6)
(731.1)
(753.5)
(735.8)
(740.4)
(695.2)
(716.8)
Net Emissions (Sources and Sinks)
5,549.6
6,5^5.3
5,795.9
5,999.9
6,055.2
5,982.1
5,829.3
Notes: Total emissions presented without LULUCF. Net emissions presented with LULUCF.
+ Does not exceed 0.05 MMT CO2 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 Land Use, Land-Use Change,
and Forestry.
b Emissions from International Bunker Fuels are not included in totals.
0 LULUCF emissions of CH4 andN20 are reported separately from gross emissions totals. LULUCF emissions include the
CH4 and N2O 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 N2O
emissions from Forest Soils and Settlement Soils. Refer to Table ES-5 for a breakout of emissions and removals for Land
Use, Land-Use Change, and Forestry 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 ES-5 for a breakout of emissions and removals for
Land Use, Land-Use Change, and Forestry by gas and source category.
f The LULUCF Sector Net Total is the net sum of all CH4 and N2O emissions to the atmosphere plus net carbon stock changes.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or sequestration.
Figure ES-4 illustrates the relative contribution of the direct greenhouse gases to total U.S. emissions in 2016,
weighted by global warming potential. The primary greenhouse gas emitted by human activities in the United States
was CO2, representing approximately 81.5 percent of total greenhouse gas emissions. The largest source of CO2, and
of overall greenhouse gas emissions, was fossil fuel combustion. Methane emissions, which have decreased by 15.7
percent since 1990, resulted primarily from enteric fermentation associated with domestic livestock, natural gas
systems, and decomposition of wastes in landfills. Agricultural soil management, stationary fuel combustion,
manure management, and mobile source fuel combustion were the major sources of N20 emissions. Ozone depleting
substance substitute emissions and emissions of HFC-23 during the production of HCFC-22 were the primary
contributors to aggregate hydrofluorocarbon (HFC) emissions. Perfluorocarbon (PFC) emissions resulted from
semiconductor manufacturing and as a byproduct of primary aluminum production, electrical transmission and
distribution systems accounted for most sulfur hexafluoride (SF6) emissions, and semiconductor manufacturing is
the only source of nitrogen trifluoride (NF3) emissions.
ES-8 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Figure ES-4: 2016 U.S. Greenhouse Gas Emissions by Gas (Percentages based on MMT CO2
2 Eq.)
HFCs, PFCs, SF, and NF, Subtotal
2.9%
N20
5.6%
CH.
10.0%
CO,
81.5%
3
4
5 Overall, from 1990 to 2016, total emissions of CO2 increased by 196.5 MMT CO2 Eq. (3.8 percent), while total
6 emissions of CH4 decreased by 122.3 MMT CO: Eq. (15.7 percent), and N;0 emissions increased by 14.2 MMT
7 CO2 Eq. (4.0 percent). During the same period, aggregate weighted emissions of HFCs, PFCs, SF6 and NF;, rose by
8 88.6 MMT CO2 Eq. (88.8 percent). From 1990 to 2016, HFCs increased by 130.5 MMT CO2 Eq. (280.3 percent),
9 PFCs decreased by 19.9 MMT CO2 Eq. (82.1 percent), SF6 decreased by 22.6 MMT CO2 Eq. (78.5 percent), and
10 NF3 increased by 0.5 MMT CO2 Eq. (1,110.2 percent). Despite being emitted in smaller quantities relative to the
11 other principal greenhouse gases, emissions of HFCs, PFCs, SF6 and NF;, are significant because many of these
12 gases have extremely high global warming potentials and, in the cases of PFCs and SF6, long atmospheric lifetimes.
13 Conversely, U.S. greenhouse gas emissions were partly offset by carbon (C) sequestration in forests, trees in urban
14 areas, agricultural soils, landfilled yard trimmings and food scraps, and coastal wetlands, which, in aggregate, offset
15 11.5 percent of total emissions in 2016. The following sections describe each gas's contribution to total U.S.
16 greenhouse gas emissions in more detail.
17 Carbon Dioxide Emissions
18 The global carbon cycle is made up of large carbon flows and reservoirs. Billions of tons of carbon in the form of
19 CO2 are absorbed by oceans and living biomass (i.e., sinks) and are emitted to the atmosphere annually through
20 natural processes (i.e., sources). When in equilibrium, carbon fluxes among these various reservoirs are roughly
21 balanced.11
22 Since the Industrial Revolution (i.e., about 1750), global atmospheric concentrations of CO2 have risen
23 approximately 44 percent (IPCC 2013; NOAA/ESRL 2017a), principally due to the combustion of fossil fuels.
24 Globally, approximately 32,294 MMT of CO2 were added to the atmosphere through the combustion of fossil fuels
25 in 2015, of which the United States accounted for approximately 15 percent.12
26 Within the United States, fossil fuel combustion accounted for 93.3 percent of CO2 emissions in 2016. There are 24
27 additional sources of CO2 emissions included in the Inventory (see Figure ES-5). Although not illustrated in the
28 Figure ES-5, changes in land use and forestry practices can also lead to net CO2 emissions (e.g., through conversion
29 of forest land to agricultural or urban use) or to a net sink for CO2 (e.g., through net additions to forest biomass).
11 Hie term "flux" is used to describe the net emissions of greenhouse gases accounting for both the emissions of CO2 to and the
removals of CO2 from the atmosphere. Removal of CO2 from the atmosphere is also referred to as "carbon sequestration."
12 Global CO2 emissions from fossil fuel combustion were taken from International Energy Agency CO: Emissions from Fossil
Fuels Combustion -Highlights. IEA (2017). See . The publication has not yet been updated to include 2016 data.
Executive Summary ES-9
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Figure ES-5: 2016 Sources of CO2 Emissions (MMT CO2 Eq.)
Fossil Fuel Combustion
Non-Energy Use of Fuels
Iron and Steel Prod. & Metallurgical Coke Prod.
Cement Production
Petrochemical Production
Natural Gas Systems
Petroleum Systems
Lime Production
Other Process Uses of Carbonates
Ammonia Production
Incineration of Waste
Urea Fertilization
Carbon Dioxide Consumption
Urea Consumption for Non-Agricultural Purposes
Liming
Ferroalloy Production
Soda Ash Production
Titanium Dioxide Production
Aluminum Production
Glass Production
Phosphoric Acid Production
Zinc Production
Lead Production
Silicon Carbide Production and Consumption
Magnesium Production and Processing
| 4,977
< .05
< .05
< .05
CO2 as a Portion of all
Emissions
0
25
50
75 100
MMT CO= Eq.
125
150
As the largest source of U.S. greenhouse gas emissions, CO2 from fossil fuel combustion has accounted for
approximately 77 percent of GWP-weighted emissions since 1990. The fundamental factors influencing emissions
levels 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 most sectors of the economy.
Between 1990 and 2016, CO2 emissions from fossil fuel combustion increased from 4,755.8 MMT CO2 Eq. to
4,976.7 MMT CO2 Eq., a 4.6 percent total increase over the twenty-seven-year period. Conversely, CO2 emissions
from fossil fuel combustion decreased by 782.3 MMT CO2 Eq. from 2005 levels, a decrease of approximately 13.6
percent between 2005 and 2016. From 2015 to 2016, these emissions decreased by 82.6 MMT CO2 Eq. (1.6
percent).
Historically, changes in emissions from fossil fuel combustion have been the dominant factor affecting U.S.
emission trends. Changes in CO2 emissions from fossil fuel combustion are influenced by many long-term and
short-term factors. Long-term factors include population and economic trends, technological changes, 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.
The five major fuel consuming economic sectors contributing to CO2 emissions from fossil fuel combustion are
electric power, transportation, industrial, residential, and commercial. 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. For the discussion below, electric power emissions have been distributed to each end-use sector on the basis
of each sector's share of aggregate electricity use. 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.
ES-10 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Emissions from electric power are also addressed separately after the end-use sectors have been discussed. Note that
emissions from U.S. Territories are calculated separately due to a lack of specific consumption data for the
individual end-use sectors. Figure ES-6, Figure ES-7, and Table ES-3 summarize CO2 emissions from fossil fuel
combustion by end-use sector.
Figure ES-6: 2016 CO2 Emissions from Fossil Fuel Combustion by Sector and Fuel Type (MMT
COz Eq.)
2,500
2,000
£ 1,500
0
U
I-
1 1,000
500
Relative Contribution by Fuel Type
41
228
I Petroleum
I Coal
I Natural Gas
296
U.S. Territories Commercial
Residential
1,795
1,809
Industrial
Transportation Electric Power
Note on Figure ES-6: Fossil Fuel Combustion includes electric power, which also includes emissions of less than 0.5 MMT CO2
Eq. from geothermal-based generation.
Figure ES-7: 2016 End-Use Sector Emissions of CO2 from Fossil Fuel Combustion (MMT CO2
Eq.)
2,000
1,500
8 1,000-
b
s:
z
500-
I Direct Fossil Fuel Combustion
Indirect Fossil Fuel Combustion
1,798
1,314
U.S. Territories
Commercial
Residential
Industrial
Transportation
Table ES-3: CO2 Emissions from Fossil Fuel Combustion by End-Use Sector (MMT CO2 Eq.)
End-Use Sector
1990
2005
2012
2013
2014
2015
2016
Transportation
1,470.2
1,860.5
1,665.8
1,681.6
1,721.2
1,739.2
1,798.4
Combustion
1,467.2
1,855.8
1,661.9
1,677.6
1,717.1
1,735.5
1,794.9
Electricity
3.0
4.7
3.9
4.0
4.1
3.7
3.5
Industrial
1,561.3
1,604.4
1,411.2
1,443.4
1,424.0
1,368.8
1,313.8
Combustion
874.5
867.8
818.4
848.7
830.8
819.3
807.6
Electricity
686.7
736.6
592.8
594.7
593.2
549.6
506.2
Executive Summary ES-11
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Residential
931.4
1,214.1
1,007.8
1,064.6
1,080.0
1,001.1
957.0
Combustion
338.3
357.8
282.5
329.7
345.3
316.8
296.2
Electricity
593.0
856.3
725.3
734.9
734.7
684.3
660.7
Commercial
765.3
1,030.3
901.6
930.2
939.6
908.8
866.2
Combustion
227.4
227.0
201.3
225.7
233.6
245.6
227.9
Electricity
538.0
803.3
700.3
704.5
706.0
663.1
638.3
U.S. Territories3
27.6
49.7
43.5
42.5
41.4
41.4
41.4
Total
4,755.8
5,759.1
5,029.8
5,162.3
5,206.1
5,059.3
4,976.7
Electric Power
1,820.8
2,400.9
2,022.2
2,038.1
2,038.0
1,900.7
1,808.8
aFuel 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.
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.
Transportation End-Use Sector. When electricity-related emissions are distributed to economic end-use sectors,
transportation activities accounted for 36.1 percent of U.S. CO2 emissions from fossil fuel combustion in 2016. The
largest sources of transportation CO2 emissions in 2016 were passenger cars (42.2 percent), medium- and heavy-
duty trucks (23.3 percent), light-duty trucks, which include sport utility vehicles, pickup trucks, and minivans (17.0
percent), commercial aircraft (6.6 percent), rail (2.2 percent), other aircraft (2.8 percent), pipelines (2.2 percent), and
ships and boats (2.3 percent). Annex 3.2 presents the total emissions from all transportation and mobile sources,
including CO2, CH4, N20, and HFCs.
In terms of the overall trend, from 1990 to 2016, total transportation CO2 emissions increased due, in large part, to
increased demand for travel. The number of VMT by light-duty motor vehicles (i.e., passenger cars and light-duty
trucks) increased 43 percent from 1990 to 2016,13 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. 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.
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 26 percent of CO2 from fossil fuel
combustion in 2016. Approximately 61 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. In contrast to the other end-use sectors, emissions
from industry have declined 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.
Residential and Commercial End-Use Sectors. The residential and commercial end-use sectors accounted for 19 and
17 percent, respectively, of CO2 emissions from fossil fuel combustion in 2016. Both sectors relied heavily on
electricity for meeting energy demands, with 69 and 74 percent, respectively, of their emissions attributable to
electricity use for lighting, heating, cooling, and operating appliances. The remaining emissions were due to the
consumption of natural gas and petroleum for heating and cooking. Emissions from the residential and commercial
end-use sectors have increased by 3 percent and 13 percent since 1990, respectively.
Electric Power. The United States relies on electricity to meet a significant portion of its energy demands.
Electricity generators used 33 percent of U.S. energy from fossil fuels and emitted 36 percent of the CO2 from fossil
fuel combustion in 2016. The type of energy source used to generate electricity is the main factor influencing
13 VMT estimates are based on data from FHWA Highway Statistics Table VM-1 (FHWA 1996 through 2017). Table VM-1 data
for 2016 has not been published yet, therefore 2016 mileage data is estimated using the 1.7 percent increase in FHWA Traffic
Volume Trends from 2015 to 2016. 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 2016 time period. In absence of these method changes,
light-duty VMT growth between 1990 and 2016 would likely have been even higher.
ES-12 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
emissions. For example, some electricity is generated through non-fossil fuel options such as nuclear, hydroelectric,
wind, solar, or geothennal energy.
Including all electricity generation modes, electric power sector generators relied on coal for approximately 30
percent of their total energy requirements in 2016. In addition, the coal used by electricity generators accounted for
93 percent of all coal consumed for energy in the United States in 2016.14 Recently, a decrease in the carbon
intensity of the mix of fuels consumed to generate electricity has occurred due to a decrease in coal consumption,
increased natural gas consumption, and increased reliance on non-fossil generation sources. Including all electricity
generation modes, electric power sector generators used natural gas for approximately 34 percent of their total
energy requirements in 2016.
Across the time series, changes in electricity demand and the carbon intensity of fuels used for electric power have a
significant impact on CO2 emissions. While emissions from the electric power sector have decreased by
approximately 0.2 percent since 1990, the carbon intensity of the electric power sector, in terms of CO2 Eq. per
QBtu, input has significantly decreased—by 12 percent—during that same timeframe. This trend away from a direct
relationship between electric power and the resulting emissions is shown in Figure ES-8.
Figure ES-8: Electric Power Generation (Billion kWh) and Emissions (MMT CO2 Eq.)
5,000 ¦ Petroleum-based Generation (Billion kWh)
¦ Nuclear-based Generation (Billion kWh)
4 50O-I I Renewable-based Generation (Billion kWh)
¦ Natural Gas-based Generation (Billion kWh)
¦ Coal-based Generation (Billion kWh)
3,500
3,000
4,000-
Total Emissions (MMT COi Eq.) [Right Axis]
2,500
3,500-
3,000-
2,000
2,500-
1,500 .2
2,000-
1,500-
1,000
1,000-
Other significant CO2 trends included the following:
• Carbon dioxide emissions from non-energy use of fossil fuels increased by 1.5 MMT CO2 Eq. (1.2 percent)
from 1990 through 2016. Emissions from non-energy uses of fossil fuels were 121.0 MMT CO2 Eq. in
2016, which constituted 2.3 percent of total national CO2 emissions, approximately the same proportion as
in 1990.
14 See Table 6.2 Coal Consumption by Sector of EIA 2016.
Executive Summary ES-13
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
• Carbon dioxide emissions from iron and steel production and metallurgical coke production have decreased
by 59.3 MMT CO2 Eq. (58.4 percent) from 1990 through 2016, 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 approximately 9.1
percent between 1990 and 2016. 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.
Box ES-4: 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.15 Several recent studies have measured
emissions at the national or regional level with 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
greenhouse gas emissions could assist in improving the understanding of inventory estimates. In working with the
research community on ambient measurement and remote sensing techniques to improve national greenhouse gas
inventories, EPA follows guidance from the IPCC on the use of measurements and modeling to validate emission
inventories. 16An area of particular interest in EPA's outreach efforts is how ambient measurement data can be used
in a manner consistent with this Inventory report's transparency on 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.
In an effort 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 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.17
Methane Emissions
Methane (CH4) is 25 times as effective as CO2 at trapping heat in the atmosphere (IPCC 2007). Over the last two
hundred and fifty years, the concentration of CH4 in the atmosphere increased by 163 percent (IPCC 2013;
NOAA/ESRL 2017b). Anthropogenic sources of CH4 include natural gas and petroleum systems, agricultural
activities, LULUCF, landfills, coal mining, wastewater treatment, stationary and mobile combustion and certain
industrial processes (see Figure ES-9).
15 See .
16 See .
17 See .
ES-14 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Figure ES-9: 2016 Sources of ChU Emissions (MMT CO2 Eq.)
Enteric Fermentation
Natural Gas Systems
Landfills
Manure Management
Coal Mining
Petroleum Systems
Wastewater Treatment
Rice Cultivation
Stationary Combustion
Abandoned Oil and Gas Wells
Abandoned Underground Coal Mines
Mobile Combustion
Composting
Field Burning of Agricultural Residues
Petrochemical Production
Ferroalloy Production
Silicon Carbide Production and Consumption
Iron and Steel Production & Metallurgical Coke Production
Incineration of Waste
170
¦
¦
¦
I
I
< .05
< .05
< .05
< .05
< .05
< .05
CH. as a Portion
Emissions
of all
10.0%
25
50
75 100
MMT CO: Eq.
125
150
175
Note: LULUCF emissions are reported separately from gross emissions totals and are not included in Figure ES-9. Refer to Table
ES-5 for a breakout of LULUCF emissions by gas.
Significant trends for the largest sources of U.S. CH4 emissions include the following:
• Enteric fermentation is the largest anthropogenic source of CH4 emissions in the United States. In 2016,
enteric fermentation CH4 emissions were 170.1 MMT CO2 Eq. (25.9 percent of total CH4 emissions),
which represents an increase of 6.0 MMT CO2 Eq. (3.6 percent) since 1990. This increase in emissions
from 1990 to 2016 generally follows the increasing trends in cattle populations.
• Natural gas systems were the second largest anthropogenic source category of CH4 emissions in the United
States in 2016 with 162.1 MMT CO2 Eq. of CH4 emitted into the atmosphere. Those emissions have
decreased by 31.6 MMT CO2 Eq. (16.3 percent) since 1990. The decrease in CH4 emissions is largely due
to the decrease in emissions from transmission, storage, and distribution. The decrease in transmission and
storage emissions is largely due to reduced compressor station emissions (including emissions from
compressors and fugitives). The decrease in distribution emissions is largely attributed to increased use of
plastic piping, which has lower emissions than other pipe materials, and station upgrades at metering and
regulating (M&R) stations.
• Landfills are the third largest anthropogenic source of CH4 emissions in the United States (107.7 MMT
CO2 Eq.), accounting for 16.4 percent of total CH4 emissions in 2016. From 1990 to 2016, CH4 emissions
from landfills decreased by 71.9 MMT CO2 Eq. (40.0 percent), with small increases occurring in some
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 MSW landfills over the time series,18 which has more than offset the
18 Carbon dioxide emissions from landfills are not included specifically in summing waste sector totals. Net carbon fluxes from
changes in biogenic carbon reservoirs and disposed wood products are accounted for in the estimates for LULUCF.
Executive Summary ES-15
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
additional CH4 emissions that would have resulted from an increase in the amount of municipal solid waste
landfilled.
Nitrous Oxide Emissions
Nitrous oxide (N20) 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 N20
emissions are much lower than CO2 emissions, N20 is nearly 300 times more powerful than CO2 at trapping heat in
the atmosphere (IPCC 2007). Since 1750, the global atmospheric concentration of N20 lias risen by approximately
22 percent (IPCC 2013; NOAA/ESRL 2017c). The main anthropogenic activities producing N2O in the United
States are agricultural soil management, stationary fuel combustion, fuel combustion in motor vehicles, manure
management, and nitric acid production (see Figure ES-10).
Figure ES-10: 2016 Sources of N2O Emissions (MMT CO2 Eq.)
Agricultural Soil Management
Stationary Combustion
Manure Management
Mobile Combustion
Nitric Acid Production
Adipic Acid Production
Wastewater Treatment
NiO from Product Uses
Caprolactam, Glyoxal, and Glyoxylic Acid Production
Composting
Incineration of Waste
Semiconductor Manufacture
Field Burning of Agricultural Residues
Note: LULUCF emissions are reported separately from gross emissions totals and are not included in Figure ES-10. Refer to
Table ES-5 for a breakout of LULUCF emissions by gas.
Significant trends forthe largest sources of U.S. emissions of N20 include the following:
• Agricultural soils accounted for approximately 76.9 percent of N20 emissions and 4.3 percent of total
emissions in the United States in 2016. Estimated emissions from this source in 2016 were 283.6 MMT
CO2 Eq. Annual N20 emissions from agricultural soils fluctuated between 1990 and 2016, although overall
emissions were 13.2 percent higher in 2016 than in 1990. Year-to-year fluctuations are largely a reflection
of annual variation in weather patterns, synthetic fertilizer use, and crop production.
• Nitrous oxide emissions from stationary combustion increased 7.2 MMT CO2 Eq. (64.9 percent) from 1990
through 2016. Nitrous oxide emissions from this source increased primarily as a result of an increase in the
number of coal fluidized bed boilers in the electric power sector.
• In 2016, total N20 emissions from manure management were estimated to be 18.1 MMT CO2 Eq.;
emissions were 14.0 MMT CO2 Eq. in 1990. These values include both direct and indirect N20 emissions
from manure management. Nitrous oxide emissions have remained fairly steady 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 N20 emissions showed a 29.6 percent increase from 1990 to 2016 and a 2.4
percent increase from 2015 through 2016.
284
NjO as a Portion of all
Emissions
5.6%
< 0.5
< 0.5
< 0.5
10 15
MMT CO; Eq.
20
25
ES-16 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
HFC, PFC, SF6, and NF3 Emissions
Hydrofluorocarbons (HFCs) and perfluorocarbons (PFCs) are families of 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 and PFCs do not deplete the stratospheric ozone layer,
and are therefore acceptable alternatives under the Montreal Protocol on Substances that Deplete the Ozone Layer.
These compounds, however, along with SF6 and NF3, are potent greenhouse gases. In addition to having high global
warming potentials, SF6 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 lias evaluated (IPCC 2013).
Other emissive sources of these gases include HCFC-22 production electrical transmission and distribution systems,
semiconductor manufacturing, aluminum production and magnesium production and processing (see Figure ES-11).
Figure ES-11: 2016 Sources of HFCs, PFCs, SFe, and NF3 Emissions (MMT CO2 Eq.)
Substitution of Ozone Depleting Substances I
Semiconductor Manufacture
Electrical Transmission and Distribution
HCFC-22 Production
Aluminum Production
Magnesium Production and Processing
174
HFCs, PFCs, SFi, and NFs as a Portion of all Emissions
2.9%
10
MMT CO: Eq.
20
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 the substitution of ODS (e.g.,
chlorofluorocarbons [CFCs]) have been consistently increasing, from small amounts in 1990 to 173.9
MMT CO2 Eq. in 2016. This increase was in large part the result of efforts to phase out CFCs and other
ODS in the United States. In the short term, this trend is expected to continue, and will likely continue over
the next decade as hydrochlorofluorocarbons (HCFCs), which are interim substitutes in many applications,
are themselves phased out under the provisions of the Copenhagen Amendments to the Montreal Protocol.
• GWP-weighted PFC, HFC, SF6, and NF3 emissions from semiconductor manufacturing have increased by
32.8 percent from 1990 to 2016, due to competing factors 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, but have since declined
to 4.7 MMT CO2 Eq. in 2016 (a 47.6 percent decrease relative to 1999).
• Sulfur hexafluoride emissions from electric power transmission and distribution systems decreased by 81.3
percent (18.8 MMT CO2 Eq.) from 1990 to 2016. There are two potential causes for this decrease: (1) a
sharp increase in the price of SF6 during the 1990s and (2) a growing awareness of the enviromnental
impact of SF6 emissions through programs such as EPA's SF6 Emission Reduction Partnership for Electric
Power Systems.
Executive Summary ES-17
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
ES.3 Overview of Sector Emissions and Trends
In accordance with the UNFCCC decision to set the 2006IPCC Guidelines for National Greenhouse Gas
Inventories (IPCC 2006) as the standard for Annex I countries at the Nineteenth Conference of the Parties
(UNFCCC 2014), Figure ES-12 and Table ES-4 aggregate emissions and sinks by the sectors defined by those
guidelines. Ov er the twenty-seven-year period of 1990 to 2016, total emissions from the Energy, Industrial
Processes and Product Use. and Agriculture sectors grew by 136.2 MMT CO2 Eq. (2.6 percent), 35.2 MMT CO2 Eq.
(10.3 percent), and 73.4 MMT CO; Eq. (15.0 percent), respectively. Emissions from the Waste sector decreased by
67.9 MMT CO2 Eq. (34.1 percent). Over the same period, total C sequestration in the Land Use, Land-Use Change,
and Forestry (LULUCF) sector increased by 75.3 MMT CO2 (9.1 percent decrease in total C sequestration), and
emissions from the LULUCF sector increased by 27.4 MMT CO2 Eq. (258 percent).
Figure ES-12: U.S. Greenhouse Gas Emissions and Sinks by Chapter/IPCC Sector (MMT CO2
Eq.)
-7 ™ -1 Industrial Processes and Product Use
7,500- Waste \
7,000- \ ¦ V - M "*'(emissions)
6,500- ^Agriculture '*"
6,000-
5,500- I
5,000- I
cr 4,500- I
UJ
q 4,000- I
u 3,500- I
3,000- Energy
s 2,500- I
2,000- I
1,500- I
1,000- I
500- I
OH I
-500 Lanci Use' Land-Use Change and Forestry (LULUCF) (removals)
o*HrMcor^covor^coo^OTHr\iroTrinvo
c>c^(^(^crt(^o>o^c^(^ooooooooooTHHHHHHH
<7>a>cr>c^cr»cr»c^cn<^c^ooooooooooooooooo
¦r-* ¦»-< -»H -r-t r-i r-i *-4 *-4 •*-« T-i f\i CM fM fM CM Cvl fM CM C\1 Csl fNJ Csl CM fNJ OJ fNl Csl
Table ES-4: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks by Chapter/IPCC
Sector (MMT CO2 Eq.)
Chapter/IPCC Sector
1990
2005
2012
2013
2014
2015
2016
Energy
5,340.2
6,295.7
5,527.6
5,691.1
5,739.1
5,596.0
5,476.4
Fossil Fuel Combustion
4,755.8
5,759.1
5,029.8
5,162.3
5,206.1
5,059.3
4,976.7
Natural Gas Systems
223.4
182.5
181.2
185.6
191.2
190.8
188.8
Non-Energy Use of Fuels
119.6
141.7
113.3
133.2
127.8
135.1
121.0
Petroleum Systems
51.7
51.7
61.0
68.5
73.9
77.4
64.8
Coal Mining
96.5
64.1
66.5
64.6
64.6
61.2
53.8
Stationary Combustion
19.8
25.4
24.1
27.2
27.7
25.8
25.6
Mobile Combustion
51.3
45.0
27.8
25.7
23.6
21.9
20.8
Incineration of Waste
8.4
12.9
10.7
10.7
10.9
11.0
11.0
Abandoned Oil and Gas Wells
6.5
6.9
7.0
7.0
7.1
7.2
7.1
Abandoned Underground Coal Mines
7.2
6.6
6.2
6.2
6.3
6.4
6.7
Industrial Processes and Product Use
340.5
354.2
361.6
364.7
380.2
378.8
375.7
Substitution of Ozone Depleting
Substances
0.3
99.8
150.4
154.8
161.4
168.6
173.9
Iron and Steel Production &
Metallurgical Coke Production
101.5
68.1
55.5
53.4
58.2
47.7
42.2
ES-18 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
Cement Production
33.5
46.2
35.3
36.4
39.4
39.9
39.4
Petrochemical Production
21.4
26.9
26.6
26.5
26.6
28.2
27.6
Lime Production
11.7
14.6
13.8
14.0
14.2
13.3
13.3
Other Process Uses of Carbonates
4.9
6.3
8.0
10.4
11.8
11.2
11.2
Ammonia Production
13.0
9.2
9.4
10.0
9.6
10.6
11.2
Nitric Acid Production
12.1
11.3
10.5
10.7
10.9
11.6
10.2
Adipic Acid Production
15.2
7.1
5.5
3.9
5.4
4.3
7.0
Semiconductor Manufacture
3.6
4.7
4.4
4.0
4.9
5.0
5.0
Carbon Dioxide Consumption
1.5
1.4
4.0
4.2
4.5
4.5
4.5
Electrical Transmission and
Distribution
23.1
8.3
4.6
4.5
4.6
4.2
4.3
N2O from Product Uses
4.2
4.2
4.2
4.2
4.2
4.2
4.2
Urea Consumption for Non-
Agricultural Purposes
3.8
3.7
4.4
4.1
1.5
4.2
4.0
HCFC-22 Production
46.1
20.0
5.5
4.1
5.0
4.3
2.8
Aluminum Production
28.3
7.6
6.4
6.2
5.4
4.8
2.7
Caprolactam, Glyoxal, and Glyoxylic
Acid Production
1.7
2.1
2.0
2.0
2.0
2.0
2.0
Ferroalloy Production
2.2
1.4
1.9
1.8
1.9
2.0
1.8
Soda Ash Production
1.4
1.7
1.7
1.7
1.7
1.7
1.7
Titanium Dioxide Production
1.2
1.8
1.5
1.7
1.7
1.6
1.6
Glass Production
1.5
1.9
1.2
1.3
1.3
1.3
1.3
Magnesium Production and Processing
5.2
2.7
1.7
1.5
1.1
1.0
1.1
Phosphoric Acid Production
1.5
1.3
1.1
1.1
1.0
1.0
1.0
Zinc Production
0.6
1.0
1.5
1.4
1.0
0.9
0.9
Lead Production
0.5
0.6
0.5
0.5
0.5
0.5
0.5
Silicon Carbide Production and
Consumption
0.4
0.2
0.2
0.2
0.2
0.2
0.2
Agriculture
489.2
520.0
519.8
543.1
539.8
566.9
562.6
Agricultural Soil Management
250.5
253.5
247.9
276.6
274.0
295.0
283.6
Enteric Fermentation
164.2
168.9
166.7
165.5
164.2
166.5
170.1
Manure Management
51.1
72.9
83.2
80.8
80.4
84.0
85.9
Rice Cultivation
16.0
16.7
11.3
11.5
12.7
12.3
13.7
Urea Fertilization
2.4
3.5
4.3
4.4
4.5
4.9
5.1
Liming
4.7
4.3
6.0
3.9
3.6
3.8
3.9
Field Burning of Agricultural Residues
0.3
0.3
0.4
0.4
0.4
0.4
0.4
Waste
199.3
156.4
140.4
136.7
136.5
135.6
131.5
Landfills
179.6
132.7
117.0
113.3
112.7
111.7
107.7
Wastewater Treatment
19.1
20.2
19.7
19.6
19.8
20.0
19.8
Composting
0.7
3.5
3.7
3.9
4.0
4.0
4.0
Total Emissions3
6,369.2
7,326.4
6,549.4
6,735.6
6,795.6
6,677.3
6,546.2
Land Use, Land-Use Change, and
Forestry
(819.6)
(731.1)
(753.5)
(735.8)
(740.4)
(695.2)
(716.8)
Forest land
(784.3)
(730.0)
(723.3)
(733.3)
(731.7)
(709.9)
(714.2)
Cropland
2.4
(0.7)
1.3
11.9
11.2
16.8
13.8
Grassland
13.8
25.3
0.8
18.5
14.7
33.6
21.0
Wetlands
(4.0)
(5.3)
(4.1)
(4.1)
(4.1)
(4.1)
(4.2)
Settlements
(47.6)
(20.5)
(28.3)
(28.8)
(30.5)
(31.5)
(33.3)
Net Emission (Sources and Sinks)b
5,549.6
6,595.3
5,795.9
5,999.9
6,055.2
5,982.1
5,829.3
Notes: Total emissions presented without LULUCF. Net emissions presented with LULUCF.
a Total emissions without LULUCF.
b Total emissions with LULUCF.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or sequestration.
1 Energy
2 The Energy chapter contains emissions of all greenhouse gases resulting from stationary and mobile energy
3 activities including fuel combustion and fugitive fuel emissions, and the use of fossil fuels for non-energy purposes.
Executive Summary ES-19
-------
1 Energy-related activities, primarily fossil fuel combustion, accounted for the vast majority of U.S. CO2 emissions for
2 the period of 1990 through 2016.
3 In 2016, approximately 81 percent of the energy used in the United States (on a Btu basis) was produced through the
4 combustion of fossil fuels. The remaining 19 percent came from other energy sources such as hydropower, bio mass,
5 nuclear, wind, and solar energy (see Figure ES-13).
6 Energy-related activities are also responsible for CH4 and N20 emissions (43 percent and 10 percent of total U.S.
7 emissions of each gas, respectively). Overall, emission sources in the Energy chapter account for a combined 83.7
8 percent of total U.S. greenhouse gas emissions in 2016.
9 Figure ES-13: 2016 U.S. Energy Consumption by Energy Source (Percent)
12 The Industrial Processes and Product Use (IPPU) chapter includes greenhouse gas emissions occurring from
13 industrial processes and from the use of greenhouse gases in products.
14 In many cases, greenhouse gas emissions are produced as the byproducts of many non-energy-related industrial
15 activities. For example, industrial processes can chemically transform raw materials, which often release waste gases
16 such as CO2, CH4, N2O, and fluorinated gases (e.g., HFC-23). These processes include iron and steel production and
17 metallurgical coke production, cement production, lime production, other process uses of carbonates (e.g., flux
18 stone, flue gas desulfurization, and glass manufacturing), ammonia production and urea consumption, petrochemical
19 productioa aluminum production. HCFC-22 production, soda ash production and use, titanium dioxide production.
20 ferroalloy production, glass productioa zinc production, phosphoric acid production, lead productioa silicon
21 carbide production and consumption, nitric acid productioa adipic acid production, and caprolactam production.
22 Industrial manufacturing processes and use by end-consumers also release HFCs, PFCs, SFg, and NF3 and other
23 fluorinated compounds. In addition to the use of HFCs and some PFCs as ODS substitutes, HFCs, PFCs, SF(, NF3,
24 and other fluorinated compounds are employed and emitted by a number of other industrial sources in the United
25 States. These industries include semiconductor manufacture, electric power transmission and distributioa and
26 magnesium metal production and processing. In additioa !vO is used in and emitted by semiconductor
27 manufacturing and anesthetic and aerosol applications, and CO2 is consumed and emitted through various end-use
28 applications. Overall, emission sources in the Industrial Process and Product Use chapter account for 5.7 percent of
29 U.S. greenhouse gas emissions in 2016.
Nuclear Electric Power
8.6%
Renewable En
10.4%
Petroleum
36.9%
10
11 Industrial Processes and Product Use
ES-20 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Agriculture
The Agriculture chapter contains anthropogenic emissions from agricultural activities (except fuel combustion,
which is addressed in the Energy chapter, and some agricultural CO2 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: enteric fermentation in domestic livestock,
livestock manure management, rice cultivation, agricultural soil management, liming, urea fertilization, and field
burning of agricultural residues.
In 2016, agricultural activities were responsible for emissions of 562.6 MMT CO2 Eq., or 8.6 percent of total U.S.
greenhouse gas emissions. Methane, N20, and CO2 were the primary greenhouse gases emitted by agricultural
activities. Methane emissions from enteric fermentation and manure management represented approximately 25.9
percent and 10.3 percent of total CH4 emissions from anthropogenic activities, respectively, in 2016. Agricultural
soil management activities, such as application of synthetic and organic fertilizers, deposition of livestock manure,
and growing N-fixing plants, were the largest source of U.S. N20 emissions in 2016, accounting for 76.9 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 2-11 and Table 2-7
illustrate agricultural greenhouse gas emissions by source.
Land Use, Land-Use Change, and Forestry
The Land Use, Land-Use Change, and Forestry (LULUCF) chapter contains emissions of CH4 and N20, and
emissions and removals of CO2 from managed lands in the United States. Overall, managed land is a net sink for
CO2 (C sequestration) in the United States. The primary drivers of fluxes on managed lands include, for example,
forest management practices, tree planting in urban areas, the management of agricultural soils, 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, 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 and landfilled yard trimmings and food scraps, is a result of net tree growth and increased urban
forest size, as well as long-term accumulation of yard trimmings and food scraps carbon in landfills.
The LULUCF sector in 2016 resulted in a net increase in C stocks (i.e., net CO2 removals) of 754.9 MMT CO2 Eq.
(Table ES-5).19 This represents an offset of 11.5 percent of total (i.e., gross) greenhouse gas emissions in 2016.
Emissions of CH4 and N2O from LULUCF activities in 2016 are 38.1 MMT CO2 Eq. and represent 0.6 percent of
total greenhouse gas emissions.20 Between 1990 and 2016, total C sequestration in the LULUCF sector decreased by
9.1 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.
Forest fires were the largest source of CH4 emissions from LULUCF in 2016, totaling 18.5 MMT CO2 Eq. (740 kt of
CH4). Coastal Wetlands Remaining Coastal Wetlands resulted in CH4 emissions of 3.6 MMT CO2 Eq. (143 kt of
CH4). Grassland fires resulted in CH4 emissions of 0.3 MMT CO2 Eq. (11 kt of CH4). Peatlands Remaining
Peatlands, Land Converted to Wetlands, and Drained Organic Soils resulted in CH4 emissions of less than 0.05
MMT CO2 Eq. each.
Forest fires were also the largest source of N20 emissions from LULUCF in 2016, totaling 12.2 MMT CO2 Eq. (41
kt of N20). Nitrous oxide emissions from fertilizer application to settlement soils in 2016 totaled to 2.5 MMT CO2
Eq. (8 kt of N20). Additionally, the application of synthetic fertilizers to forest soils in 2016 resulted in N20
emissions of 0.5 MMT CO2 Eq. (2 kt of N20). Grassland fires resulted in N20 emissions of 0.3 MMT CO2 Eq. (1 kt
19 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.
20 LULUCF emissions include the CH4 and N2O 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 N2O emissions from Forest Soils and Settlement Soils.
Executive Summary ES-21
-------
1 of N2O). Coastal Wetlands Remaining Coastal Wetlands and Drained Organic Soils resulted in N2O emissions of
2 0.1 MMT CO2 Eq. each (less than 0.5 kt of N2O). Peatlands Remaining Peatlands resulted U1N2O emissions of less
3 than 0.05 MMT C02 Eq.
4 Carbon dioxide removals from C stock changes are presented in Table ES-5 along with CH4 and N20 emissions for
5 LULUCF source categories.
6 Table ES-5: U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use, Land-
7 Use Change, and Forestry (MMT CO2 Eq.)
Gas/Land-Use Category
1990
2005
2012
2013
2014
2015
2016
Carbon Stock Change3
(830.2)
(754.2)
(779.5)
(755.0)
(760.0)
(733.4)
(754.9)
Forest Land Remaining Forest Land
(697.7)
(664.6)
(666.9)
(670.9)
(669.3)
(666.2)
(670.5)
Land Converted to Forest Land
(92.0)
(81.6)
(74.9)
(74.9)
(75.0)
(75.0)
(75.0)
Cropland Remaining Cropland
(40.9)
(26.5)
(21.4)
(11.4)
(12.0)
(6.3)
(9.9)
Land Converted to Cropland
43.3
25.9
22.7
23.3
23.2
23.2
23.8
Grassland Remaining Grassland
(4.2)
5.5
(20.8)
(3.7)
(7.5)
9.6
(1.6)
Land Converted to Grassland
17.9
19.2
20.4
21.9
21.5
23.3
22.0
Wetlands Remaining Wetlands
(7.6)
(8.9)
(7.7)
(7.8)
(7.8)
(7.8)
(7.9)
Land Converted to Wetlands
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Settlements Remaining Settlements
(86.2)
(91.4)
(99.2)
(99.8)
(101.2)
(102.2)
(103.7)
Land Converted to Settlements
37.2
68.4
68.3
68.3
68.2
68.1
68.0
CH4
6.7
13.3
15.0
10.9
11.2
22.4
22.4
Forest Land Remaining Forest Land:
Forest Fires
3.2
9.4
10.8
7.2
7.2
18.5
18.5
Wetlands Remaining Wetlands: Coastal
Wetlands Remaining Coastal Wetlands
3.4
3.5
3.5
3.6
3.6
3.6
3.6
Grassland Remaining Grassland:
Grassland Fires
0.1
0.3
0.6
0.2
0.4
0.3
0.3
Forest Land Remaining Forest Land:
Drained Organic Soils
+
+
+
+
+
+
+
Land Converted to Wetlands: Land
Converted to Coastal Wetlands
+
+
+
+
+
+
+
Wetlands Remaining Wetlands: Peatlands
Remaining Peatlands
+
+
+
+
+
+
+
N2O
3.9
9.7
11.1
8.3
8.4
15.8
15.7
Forest Land Remaining Forest Land:
Forest Fires
2.1
6.2
7.1
4.8
4.7
12.2
12.2
Settlements Remaining Settlements:
Settlement Soilsb
1.4
2.5
2.7
2.6
2.6
2.5
2.5
Forest Land Remaining Forest Land:
Forest Soilsc
0.1
0.5
0.5
0.5
0.5
0.5
0.5
Grassland Remaining Grassland:
Grassland Fires
0.1
0.3
0.6
0.2
0.4
0.3
0.3
Wetlands Remaining Wetlands: Coastal
Wetlands Remaining Coastal Wetlands
0.1
0.2
0.1
0.1
0.1
0.1
0.1
Forest Land Remaining Forest Land:
Drained Organic Soils
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Wetlands Remaining Wetlands: Peatlands
Remaining Peatlands
+
+
+
+
+
+
+
LULUCF Emissions'1
10.6
23.0
26.1
19.2
19.6
38.2
38.1
LULUCF Carbon Stock Change3
(830.2)
(754.2)
(779.5)
(755.0)
(760.0)
(733.4)
(754.9)
LULUCF Sector Net Totale
(819.6)
(731.1)
(753.5)
(735.8)
(740.4)
(695.2)
(716.8)
+ Absolute value does not exceed 0.05 MMT CO2 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 emissions from N fertilizer additions on both Settlements Remaining Settlements and Land Converted to
Settlements.
c Estimates include emissions from N fertilizer additions on both Forest Land Remaining Forest Land and Land Converted to
Forest Land.
ES-22 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
d LULUCF emissions include the CH4 and N2O 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 N2O emissions from Forest Soils and Settlement Soils.
e The LULUCF Sector Net Total is the net sum of all CH4 and N2O emissions to the atmosphere plus net carbon stock
changes.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
1 Waste
2 The Waste chapter contains emissions from waste management activities (except incineration of waste, which is
3 addressed in the Energy chapter). Landfills were the largest source of anthropogenic greenhouse gas emissions in the
4 Waste chapter, accounting for 81.9 percent of this chapter's emissions, and 16.4 percent of total U.S. CH4
5 emissions.21 Additionally, wastewater treatment accounts for 15.1 percent of Waste emissions, 2.3 percent of U.S.
6 CH4 emissions, and 1.3 percent of U.S. N20 emissions. Emissions of CH4 and N20 from composting are also
7 accounted for in this chapter, generating emissions of 2.1 MMT CO2 Eq. and 1.9 MMT CO2 Eq., respectively.
8 Overall, emission sources accounted for in the Waste chapter generated 2.0 percent of total U.S. greenhouse gas
9 emissions in 2016.
10 ES.4 Other Information
11 Emissions by Economic Sector
12 Throughout the Inventory of U.S. Greenhouse Gas Emissions and Sinks report, emission estimates are grouped into
13 five sectors (i.e., chapters) defined by the IPCC: Energy; Industrial Processes and Product Use; Agriculture;
14 LULUCF; and Waste. While it is important to use this characterization for consistency with UNFCCC reporting
15 guidelines and to promote comparability across countries, it is also useful to characterize emissions according to
16 commonly used economic sector categories: residential, commercial, industry, transportation, electric power,
17 agriculture, and U.S. Territories.
18 Figure ES-14 shows the trend in emissions by economic sector from 1990 to 2016, and Table ES-6 summarizes
19 emissions from each of these economic sectors.
21 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.
Executive Summary ES-23
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Figure ES-14: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO2 Eq.)
2,500-
Electric Power Industry
2,000-
Transportation
iS" 1/500-
Industry
1,000-
Agriculture
Commercial (Red)
500-
Residential (Blue)
UD
cr
-------
1 The commercial and residential sectors accounted for 6.3 percent and 5.4 percent of emissions, respectively, and
2 U.S. Territories accounted for 0.7 percent of emissions; emissions from these sectors primarily consisted of CO2
3 emissions from fossil fuel combustion. CO2 was also emitted and sequestered by a variety of activities related to
4 forest management practices, tree planting in urban areas, the management of agricultural soils, landfilling of yard
5 trimmings, and changes in C stocks in coastal wetlands.
6 Electricity is ultimately used in the economic sectors described above. Table ES-7 presents greenhouse gas
7 emissions from economic sectors with emissions related to electric power distributed into end-use categories (i.e.,
8 emissions from electric power are allocated to the economic sectors in which the electricity is used). To distribute
9 electricity emissions among end-use sectors, emissions from the source categories assigned to electric power were
10 allocated to the residential, commercial, industry, transportation, and agriculture economic sectors according to retail
11 sales of electricity (EIA 2017 and Duffield 2006). These source categories include CO2 from fossil fuel combustion
12 and the use of limestone and dolomite for flue gas desulfurization, CO2 and N20 from incineration of waste, CH4
13 and N20 from stationary sources, and SF6 from electrical transmission and distribution systems.
14 When emissions from electricity use are distributed among these sectors, industrial activities and transportation
15 account for the largest shares of U.S. greenhouse gas emissions (28.9 percent and 28.5 percent, respectively) in
16 2016. The residential and commercial sectors contributed the next largest shares of total U.S. greenhouse gas
17 emissions in 2016. Emissions from these sectors increase substantially when emissions from electricity are included,
18 due to their relatively large share of electricity use (e.g., lighting, appliances). In all sectors except agriculture, CO2
19 accounts for at least 81 percent of greenhouse gas emissions, primarily from the combustion of fossil fuels.
20 Figure ES-15 shows the trend in these emissions by sector from 1990 to 2016.
21 Table ES-7: U.S. Greenhouse Gas Emissions by Economic Sector with Electricity-Related
22 Emissions Distributed (MMT CO2 Eq.)
Implied Sectors
I'm
2005
2012
2013
2014
2015
2016
Industry
2,321.9
2,225.4
1,977.8
2,040.5
2,033.7
1,985.1
1,888.8
Transportation
1,528.2
1,977.3
1,752.8
1,761.9
1,797.2
1,812.5
1,867.4
Commercial
978.2
1,218.7
1,099.9
1,128.2
1,139.8
1,108.9
1,063.1
Residential
951.2
1,240.4
1,055.7
1,120.6
1,142.2
1,067.3
1,028.4
Agriculture
606.5
614.8
636.4
636.1
656.8
651.9
U.S. Territories
58.1
48.5
48.1
46.6
46.6
46.6
Total Emissions
6,3(t').2
7,326.4
6,549.4
6,735.6
6,795.6
6,677.3
6,546.2
LULUCF Sector Net Total3
(8I'U.)
(731.1)
(753.5)
(735.8)
(740.4)
(695.2)
(716.8)
Net Emissions (Sources and Sinks)
5,549.6
6,595.3
5.795.9
5,999.9
6,055.2
5,982.1
5,829.3
a The LULUCF Sector Net Total is the net sum of all CH4 and N2O 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.
23
Executive Summary ES-25
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Figure ES-15: U.S. Greenhouse Gas Emissions with Electricity-Related Emissions Distributed
to Economic Sectors (MMT CO2 Eq.)
2,500
Industry
2,000
Transportation
lS"
~ 1,500
O
u
H
s:
Commercial (Red)
1,000
Residential (Blue)
Agriculture
500
o
o
©
s
©
s
©
c
-
CO
©
CTv
©
©
r-J
in
-
o
ro
in
1—1
CTv
Cv
-*—1
©
©
m
o
=
t <
o
0
o
©
o
o
©
Box ES-5: Recent Trends in Various U.S. Greenhouse Gas Emissions-Related Data
Total 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 electricity use, because the electric power
industry—utilities and non-utilities combined—was the second largest source of U.S. greenhouse gas emissions in
2016; (4) emissions per unit of total gross domestic product as a measure of national economic activity; and (5)
emissions per capita.
Table ES-8 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 grown at an average annual rate of 0.1 percent since 1990. This rate is slightly slower than
that for total energy use and fossil fuel consumption and much slower than that for electricity use, overall gross
domestic product (GDP), and national population (see Figure ES-16). These trends vary relative to 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 1.0 percent since 2005. Total energy use and fossil fuel
consumption have also decreased at slower rates than emissions since 2005, while electricity use, GDP, and national
population continued to increase.
Table ES-8: Recent Trends in Various U.S. Data (Index 1990 = 100)
Variable
1990
2005
2012
2013
2014
2015
2016
Avg. Annual
Change
since 1990
Avg. Annual
Change
since 2005a
Greenhouse Gas Emissions'5
100
115
103
106
107
105
103
0.1%
-1.0%
Energy Usec
100
118
112
115
117
115
116
0.6%
-0.2%
Fossil Fuel Consumption0
100
119
107
110
111
110
109
0.4%
-0.7%
Electricity Usec
100
134
135
136
138
137
136
1.2%
0.1%
GDPd
100
159
171
174
179
184
187
2.4%
1.5%
Population6
100
118
125
126
127
128
129
1.0%
0.8%
a Average annual growth rate
ES-26 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
b GWP-weighted values
c Energy content-weighted values (EIA 2017)
d Gross Domestic Product in chained 2009 dollars (BEA 2017)
e U.S. Census Bureau (2017)
Figure ES-16: U.S. Greenhouse Gas Emissions Per Capita and Per Dollar of Gross Domestic
Product (GDP)
200
190-
Real GDP
180-
170-
160-
— 15°"
§ 140-
II
8 13°"
Ol
CS 120-
85
~o
100—
Population
110-
Emissions per capita
90-
80-
70-
Emissions per $GDP
60-
hv co on o
§ o o o o
vO
ro
fM
ro
8 S
o
o
c
c
o
Source: BEA (2017), U.S. Census Bureau (2017), 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 lias 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."23 By
definition key categories are sources or sinks that have the greatest contribution to the absolute overall level of
national emissions in any of the years covered by the time series. In addition when an entire time series of emission
estimates is prepared, a thorough investigation of key categories must also account for the influence of trends of
individual source and sink categories. Finally, a qualitative evaluation of key categories should be performed, in
order to capture any key categories that were not identified in either of the quantitative analyses.
Figure ES-17 presents 2016 emission estimates for the key categories as defined by a level analysis including the
LULUCF sector (i.e., the absolute value of the contribution of each source or sink category to the total inventory
level). The UNFCCC reporting guidelines request that key category analyses be reported at an appropriate level of
disaggregation, which may lead to source and sink category names which differ from those used elsewhere in the
Inventory report. For more information regarding key categories, including a complete list of categories accounting
for the influence of trends of individual source and sink categories, see Section 1.4 - Key Categories and Annex 1.
23 See Chapter 4 "Methodological Choice and Identification of Key Categories" in IPCC (2006). See .
Executive Summary ES-27
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Figure ES-17: 2016 Key Categories (MMT CO2 Eq.)
COz Emissions from Mobile Combustion: Road
COz Emissions from Stationary Combustion - Coal - Electricity Generation
Net COz Emissions in Forest Land Remaining Forest Land"
COz Emissions from Stationary Combustion - Gas - Electricity Generation
CO* Emissions from Stationary Combustion - Gas - Industrial
COz Emissions from Stationary Combustion - Oil - Industrial
COz Emissions from Stationary Combustion - Gas - Residential
Direct NzO Emissions from Agricultural Soil Management
Emissions from Substitutes for Ozone Depleting Substances
COz Emissions from Stationary Combustion - Gas - Commercial
CH4 Emissions from Enteric Fermentation
COz Emissions from Mobile Combustion: Aviation
CH» Emissions from Natural Gas Systems
CO: Emissions from Non-Energy Use of Fuels
CH» Emissions from Landfills
Net COz Emissions in Settlements Remaining Settlements3
COz Emissions from Mobile Combustion: Other
Net COz Emissions in Land Converted to Forest Land3
Net COz Emissions in Land Converted to Settlements3
CH< Emissions from Manure Management
COz Emissions from Stationary Combustion - Coal - Industrial
COz Emissions from Stationary Combustion - Oil - Residential
COz Emissions from Stationaiy Combustion - Oil - Commercial
Fugitive Emissions from Coal Mining
Indirect NzO Emissions from Applied Nitrogen
COz Emissions from Iron and Steel Production & Metallurgical Coke Production
COz Emissions from Mobile Combustion: Marine
COz Emissions from Cement Production
CH< Emissions from Petroleum Systems
COz Emissions from Stationary Combustion - Oil - U.S. Territories
COz Emissions from Petrochemical Production
COz Emissions from Natural Gas Systems
COz Emissions from Petroleum Systems
Net COz Emissions in Land Converted to Cropland3
Net COz Emissions in Land Converted to Grassland3
CH« Emissions from Forest Firesb
NzO Emissions from Forest Fires11
Net COz Emissions in Cropland Remaining Cropland3
Net COz Emissions in Grassland Remaining Grassland3
Key Categories as a Portion of All
Emissions
0 200 400 600 800 1,000 1,200 1,400
MMT COz Eq.
a Hie absolute values of net CO2 emissions from LULUCF are presented in this figure but reported separately from gross
emissions totals. Refer to Table ES-5 for a breakout of emissions and removals for LULUCF by gas and source category.
b N011-CO2 emissions from Forest Fires are presented in this figure but reported separately from gross emissions totals. Refer to
Table ES-5 for a breakout of emissions and removals for LULUCF by gas and source category.
Note: For a complete discussion of the key category analysis, see Annex 1. Blue bars indicate either an Approach 1, or Approach
1 and Approach 2 level assessment key category. Gray bars indicate solely an Approach 2 level assessment key category.
Quality Assurance and Quality Control (QA/QC)
The United States seeks to continually improve the quality , transparency , and credibility 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
Quality Assurance/Quality Control and Uncertainty Management Plan (QA/QC Management Plan) for the
Inventory, and the UNFCCC reporting guidelines. The QA process includes expert and public reviews for both the
Inventory estimates and the Inventory report.
Uncertainty Analysis of Emission Estimates
Uncertainty estimates are an essential element of a complete inventory of greenhouse gas emissions and removals,
because they help to prioritize future work and improve overall quality. Some of the current estimates, such as those
for CO2 emissions from energy-related activities, are considered to have low uncertainties. This is because the
amount of CO2 emitted from energy-related 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 believed to be relatively small. For some other categories of emissions, however, a
lack of data or an incomplete understanding of how emissions are generated increases the uncertainty or systematic
ES-28 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
error associated with the estimates presented. 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 provided in accordance with UNFCCC reporting guidelines, a
qualitative discussion of uncertainty is presented for all source and sink categories. Within the discussion of each
emission source, specific factors affecting the uncertainty surrounding the estimates are discussed.
Box ES-6: Recalculations of Inventory Estimates
Each year, emission and sink estimates are recalculated and revised for all years in the Inventory of U.S. Greenhouse
Gas Emissions and Sinks, as attempts are made to improve both the analyses themselves, through the use of better
methods or data, and the 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 lias increased; new inventory methods become available; and for correction of
errors." 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.
In each Inventory report, the results of all methodology changes and historical data updates are presented in the
Recalculations and Improvements chapter of this report; detailed descriptions of each recalculation are contained
within each source's description contained in the report, if applicable. In general, when methodological changes
have been implemented, the entire time series (in the case of the most recent Inventory report, 1990 through 2014)
has been recalculated to reflect the change, per the 2006 IPCC Guidelines (IPCC 2006). Changes in historical data
are generally the result of changes in statistical data supplied by other agencies. References for the data are provided
for additional information. Significant changes made since the previous report are also noted in Box ES-3.
Executive Summary ES-29
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
i. Introduction
This report presents estimates by the United States government of U.S. anthropogenic greenhouse gas emissions and
sinks for the years 1990 through 2016. 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 estimates in these tables are
presented on both a full molecular 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
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.. ."4 The United States views this report as an opportunity
to fulfill these commitments under the UNFCCC.
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 2014). 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
1 More information provided in "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 .
4 Article 4(1)(a) of the United Nations Framework Convention on Climate Change (also identified in Article 12). Subsequent
decisions by the Conference of the Parties elaborated the role of Annex I Parties in preparing national inventories. See
.
Introduction 1-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
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 2006IPCC Guidelines as the standard methodological approach for Annex I
countries at the Nineteenth Conference of the Parties (Warsaw, November 11-23, 2013). This report presents
information in accordance with these guidelines.
Overall, this Inventory of anthropogenic greenhouse gas emissions and sinks provides a common and consistent
mechanism through which Parties to the UNFCCC can estimate emissions and compare the relative contribution of
individual sources, gases, and nations to climate change. The Inventory provides a national estimate of sources and
sinks for the United States, including all states and U.S. Territories.5 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.
Box 1-1: Methodological Approach for Estimating and Reporting U.S. Emissions and Removals
In following the UNFCCC requirement under Article 4.1 to develop and submit national greenhouse gas emissions
inventories, the gross emissions total presented in this report for the United States excludes emissions and removals
from LULUCF. The net emissions total presented in this report for the United States includes emissions and
removals from LULUCF. All emissions and removals are calculated using internationally-accepted methods
consistent with the IPCC Guidelines.6 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.7 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 report itself
follows this standardized format, and provides an explanation of the IPCC methods used to calculate emissions and
removals.
On October 30, 2009, the U.S. Enviromnental Protection Agency (EPA) published a rule for the mandatory
reporting of greenhouse gases from large greenhouse gas emissions sources in the United States. Implementation of
40 CFR Part 98 is referred to as the EPA's GHGRP. 40 CFR Part 98 applies to direct greenhouse gas emitters, fossil
fuel suppliers, industrial gas suppliers, and facilities that inject CO2 underground for sequestration or other reasons.8
Reporting is at the facility level, except for certain suppliers of fossil fuels and industrial greenhouse gases. The
GHGRP dataset and the data presented in this Inventory are complementary.
The EPA's 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 emissions 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. The GHGRP will continue to enhance QA/QC procedures and assessment of uncertainties.
EPA continues to analyze the data on an annual basis to improve the national estimates presented in this Inventory
and uses that data for a number of categories consistent with IPCC guidance.9 EPA has integrated GHGRP
information for several categories10 this year and also identifies other categories11 where EPA plans to integrate
5 U.S. Territories include American Samoa, Guam, Puerto Rico, U.S. Virgin Islands, Wake Island, and other U.S. Pacific Islands.
6 See .
7 See .
8 See .
9 See .
10 Energy Sector (Coal Mining, Stationary Combustion [Industrial Combustion Disaggregation], and Oil and Gas Systems);
Industrial Processes and Product Use (Adipic Acid Production, Aluminum Production, Carbon Dioxide Consumption, Electrical
Transmission and Distribution, HCFC-22 Production, Lime Production, Magnesium Production and Processing, ODS
Substitutes, Nitric Acid Production, Petrochemical Production, Semiconductor Manufacture); and Waste (Landfills).
11 Industrial Process and Product Use (Ammonia Production, Cement Production, and Other Fluorinated Gas Production)
1-2 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 additional GHGRP data in the next edition of this report (see those categories Planned Improvement sections for
2 details).
3
4 1.1 Background Information
5 Sc
6 For over the past 200 years, the burning of fossil fuels such as coal and oil, deforestation, land-use changes, and
7 other activities have caused the concentrations of heat-trapping "greenhouse gases" to increase significantly in our
8 atmosphere (NOAA 2017). These gases in the atmosphere absorb some of the energy being radiated from the
9 surface of the Earth that would otherwise be lost to space, essentially acting like a blanket that makes the Earth's
10 surface warmer than it would be otherwise.
11 Greenhouse gases are necessary to life as we know it. Without greenhouse gases to create the natural heat-trapping
12 properties of the atmosphere, the planet's surface would be about 60 degrees Fahrenheit cooler than present
13 (USGCRP 2017). Carbon dioxide is also necessary for plant growth. With emissions from biological and geological
14 sources, there is a natural level of greenhouse gases that is maintained in the atmosphere. Human emissions of
15 greenhouse gases and subsequent changes in atmospheric concentrations alters the balance of energy transfers
16 between space and the earth system (IPCC 2013). A gauge of these changes is called radiative forcing, which is a
17 measure of a substance's total net effect on the global energy balance for which a positive number represents a
18 warming effect and a negative number represents a cooling effect (IPCC 2013). IPCC concluded in its most recent
19 scientific assessment report that it is extremely likely that human influences have been the dominant cause of
20 warming since the mid-20lh century (IPCC 2013).
21 As concentrations of greenhouse gases continue to increase in from man-made sources, the Earth's temperature is
22 climbing above past levels. The Earth's average land and ocean surface temperature has increased by about 1.2 to
23 1.9 degrees Fahrenheit since 1880. The last three decades have each been the warmest decade successively at the
24 Earth's surface since 1850 (IPCC 2013). Other aspects of the climate are also changing, such as rainfall patterns,
25 snow and ice cover, and sea level. If greenhouse gas concentrations continue to increase, climate models predict that
26 the average temperature at the Earth's surface is likely to increase from 0.5 to 8.6 degrees Fahrenheit above 1986
27 through 2005 levels by the end of this century, depending on future emissions and the responsiveness of the climate
28 system (IPCC 2013).
29 For further information on greenhouse gases, radiative forcing, and implications for climate change, see the recent
30 scientific assessment reports from the IPCC,12 the U.S. Global Change Research Program (USGCRP),13 and the
31 National Academies of Sciences, Engineering, and Medicine (NAS).14
32 Greenhouse Gases
33 Although the Earth's atmosphere consists mainly of oxygen and nitrogen, neither plays a significant role in
34 enhancing the greenhouse effect because both are essentially transparent to terrestrial radiation. The greenhouse
35 effect is primarily a function of the concentration of water vapor, carbon dioxide (CO2), methane (CH4), nitrous
36 oxide (N20), and other trace gases in the atmosphere that absorb the terrestrial radiation leaving the surface of the
37 Earth (IPCC 2013).
38 Naturally occurring greenhouse gases include water vapor, CO2, CH4, N20, and ozone (O3). Several classes of
39 halogenated substances that contain fluorine, chlorine, or bromine are also greenhouse gases, but they are, for the
12 See .
13 See < https://science2017.globalchange.gov/>.
14 See .
Introduction 1-3
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
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.15 Some other fluorine-containing halogenated substances—hydrofluorocarbons (HFCs),
perfluorocarbons (PFCs), sulfur hexafluoride (SF6), 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. Comparatively, however, while the understanding of aerosols has increased in recent years, they still
account for the largest contribution to uncertainty estimates in global energy budgets (IPCC 2013).
Carbon dioxide, CH4, and N20 are continuously emitted to and removed from the atmosphere by natural processes
on Earth. Anthropogenic activities, 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. 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
CO2
CH4
N2O
SF«
CF4
Pre-industrial atmospheric concentration
Atmospheric concentration
Rate of concentration change
Atmospheric lifetime (years)
280 ppm
404 ppma
2.3 ppm/yrf
See footnote11
0.700 ppm
1.843 ppmb
7 ppb/yrf,B
12.4'
0.270 ppm
0.329 ppmc
0.8 ppb/yr®
121'
Oppt
8.9 pptd
0.27 ppt/yr®
3,200
40 ppt
79 ppte
0.7 ppt/yr®
50,000
a The atmospheric CO2 concentration is the 2016 annual average at the Mauna Loa, HI station (NOAA/ESRL 2017a). The
concentration in 2017 at Mauna Loa was 407 ppm. The global atmospheric CO2 concentration, computed using an average of
sampling sites across the world, was 403 ppm in 2016.
b The values presented are global 2016 annual average mole fractions (NOAA/ESRL 2017b).
c The values presented are global 2016 annual average mole fractions (NOAA/ESRL 2017c).
dThe values presented are global 2016 annual average mole fractions (NOAA/ESRL 2017d).
e The 2011 CF4 global mean atmospheric concentration is from the Advanced Global Atmospheric Gases Experiment (IPCC
2013).
f 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 ppb/year.
B The rate of concentration change for CO2 and CH4 is the average rate of change between 2007 and 2016 (NOAA/ESRL 2017a).
The rate of concentration change for N2O, SF6, and CF4 is the average rate of change between 2005 and 2011 (IPCC 2013).
h For a given amount of carbon dioxide emitted, some fraction of the atmospheric increase in concentration is quickly absorbed by
15 Emissions estimates of CFCs, HCFCs, halons and other ozone-depleting substances are included in this document for
informational purposes.
1-4 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
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.
1 This lifetime has been defined as an "adjustment time" that takes into account the indirect effect of the gas on its own residence
time.
Source: Pre-industrial atmospheric concentrations, atmospheric lifetime, and rate of concentration changes for CH4, N2O, SF6, and
CF4 are from IPCC (2013). The rate of concentration change for CO2 is an average of the rates from 2011 through 2016 and has
fluctuated between 1.9 to 3.0 ppm per year over this period (NOAA/ESRL 2017a).
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 (H20). 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, this has been determined to have a negligible effect on climate (IPCC 2013). 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 2013).
Carbon Dioxide (C02). 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
approximately 280 parts per million by volume (ppmv) in pre-industrial times to 404 ppmv in 2016 a 44 percent
increase (IPCC 2013; NOAA/ESRL 2017a).1617 The IPCC definitively states that "the increase of CO2 ... is caused
by anthropogenic emissions from the use of fossil fuel as a source of energy and from land use and land use
changes, in particular agriculture" (IPCC 2013). The predominant source of anthropogenic CO2 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 CO2. In its Fifth Assessment Report, the IPCC stated "it is
extremely likely that more than half of the observed increase in global average surface temperature from 1951 to
2010 was caused by the anthropogenic increase in greenhouse gas concentrations and other anthropogenic forcings
together," of which CC^is the most important (IPCC 2013).
Methane (CH4). 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. 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 163
percent since 1750, from a pre-industrial value of about 700 ppb to 1,843 ppb in 201618 although the rate of increase
decreased to near zero in the early 2000s, and has recently increased again to about 5 ppb/year. The IPCC has
estimated that slightly more than half of the current CH4 flux to the atmosphere is anthropogenic, from human
activities such as agriculture, fossil fuel use, and waste disposal (IPCC 2007).
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
16 The pre-industrial period is considered as the time preceding the year 1750 (IPCC 2013).
17 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).
18 This value is the global 2016 annual average mole fraction (NOAA/ESRL 2017b).
Introduction 1-5
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
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 2013). 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.
Nitrous Oxide (N20). 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 22
percent since 1750, from a pre-industrial value of about 270 ppb to 329 ppb in 2016,19 a concentration that has not
been exceeded during the last 800 thousand years. Nitrous oxide is primarily removed from the atmosphere by the
photolytic action of sunlight in the stratosphere (IPCC 2013).
Ozone (03). Ozone is present in both the upper stratosphere,20 where it shields the Earth from harmful levels of
ultraviolet radiation, and at lower concentrations in the troposphere,21 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 anthropogenic
emissions of chlorine and bromine compounds (IPCC 2013). The depletion of stratospheric ozone and its radiative
forcing was expected to reach a maximum in about 2000 before starting to recover.
The past increase in tropospheric ozone, which is also a greenhouse gas, is estimated to provide the fourth largest
increase in direct radiative forcing since the pre-industrial era, behind CO2, black carbon, and CH4. 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 2013).
Halocarbons, Sulfur Hexafluoride, and Nitrogen Trifluoride. 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 522 countries, including the U.S., 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.
19 This value is the global 2016 annual average (NOAA/ESRL 2017c).
20 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.
21 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.
22 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.
1-6 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
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). An amendment to the Montreal Protocol
was adopted in 2016 which includes obligations for Parties to phase down the production and consumption of HFCs.
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 2013).
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
stratospheric ozone when emitted from very high-altitude aircraft.23 Additionally, NOx emissions are also likely to
decrease CH4 concentrations, thus having a negative radiative forcing effect (IPCC 2013). 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 carbonaceous24 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 2013). 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 2013).
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, "despite the large uncertainty ranges on aerosol forcing, there is high confidence that aerosols have
offset a substantial portion of GHG forcing" (IPCC 20 13).25 Although because they remain in the atmosphere for
23 NOx emissions injected higher in the stratosphere, primarily from fuel combustion emissions from high altitude supersonic
aircraft, can lead to stratospheric ozone depletion.
24 Carbonaceous aerosols are aerosols that are comprised mainly of organic substances and forms of black carbon (or soot)
(IPCC 2013).
25 The IPCC (2013) defines high confidence as an indication of strong scientific evidence and agreement in this statement.
Introduction 1-7
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
only days to weeks, their concentrations respond rapidly to changes in emissions.26 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 2013).
Black carbon also influences cloud development, but the direction and magnitude of this forcing is an area of active
research.
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 2014). 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 gas
used is CO2, and therefore GWP-weighted emissions are measured in million metric tons of CO2 equivalent (MMT
CO2 Eq.).27 The relationship between kilotons (kt) of a gas and MMT CO2 Eq. can be expressed as follows:
MMT CO2 Eq. = Million metric tons of CO2 equivalent
kt = kilotons (equivalent to a thousand metric tons)
GWP = Global warming potential
MMT = 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 ±35 percent. Parties to the UNFCCC have also agreed to use GWPs
based upon a 100-year time horizon, 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...28
Greenhouse gases with relatively long atmospheric lifetimes (e.g., CO2, CH4, N20, 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.
26 Volcanic activity can inject significant quantities of aerosol producing sulfur dioxide and other sulfur compounds into the
stratosphere, which can result in a longer negative forcing effect (i.e., a few years) (IPCC 2013).
27 Carbon comprises 12/44ths of carbon dioxide by weight.
28 Framework Convention on Climate Change; Available online at: ;
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).
Global Warming Potentials
( MMT \
Eq. = (kt of gas) x (GWP) x oqq J
MMT CO-
where,
1-8 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Table 1-2: Global Warming Potentials and Atmospheric Lifetimes (Years) Used in this Report
Gas
Atmospheric Lifetime
GWP1
CO2
See footnoteb
1
CH4c
12
25
N2O
114
298
HFC-23
270
14,800
HFC-32
4.9
675
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
HFC-4310mee
15.9
1,640
CF4
50,000
7,390
C2F6
10,000
12,200
O
J?
O
2,600
8,860
CdFl4
3,200
9,300
SFo
3,200
22,800
NF3
740
17,200
a 100-year time horizon.
b For a given amount of carbon dioxide 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 CO2 is not included.
Source: (IPCC 2007)
Box 1-2: The IPCC Fifth Assessment Report and Global Warming Potentials
In 2014, the IPCC published its Fifth Assessment Report (AR5), which updated its comprehensive scientific
assessment of climate change. Within the AR5 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), and the IPCC Fourth Assessment Report (AR4) (IPCC 2007). 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 AR5, the IPCC applied an improved calculation of CO2 radiative forcing and an improved CO2 response
function in presenting updated GWP values. Additionally, the atmospheric lifetimes of some gases have been
recalculated, and updated background concentrations were used. In addition, the values for radiative forcing and
lifetimes have been recalculated for a variety of halocarbons, and the indirect effects of methane on ozone have been
adjusted to match more recent science. Table 1-3 presents the new GWP values, relative to those presented in the
AR4 and 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 GWP values, as required by the 2013 revision to the UNFCCC reporting guidelines
for national inventories.29 All estimates provided throughout this report are also presented in unweighted units. For
29 See .
Introduction 1-9
-------
1 informational purposes, emission estimates that use GWPs from other IPCC Assessment Reports are presented in
2 detail in Annex 6.1 of this report.
3 Table 1-3: Comparison of 100-Year GWP values
AR5 with
Gas
SAR
AR4
AR5
feedbacks6
Comparison to AR4
SAR
AR5
AR5 with
feedbacksb
CO2
1
1
1
1
NC
NC
NC
CH4c
21
25
28
34
(4)
3
9
N2O
310
298
265
298
12
(33)
0
HFC-23
11,700
14,800
12,400
13,856
(3,100)
(2,400)
(944)
HFC-32
650
675
677
817
(25)
2
142
HFC-125
2,800
3,500
3,170
3,691
(700)
(330)
191
HFC-134a
1,300
1,430
1,300
1,549
(130)
(130)
119
HFC-143a
3,800
4,470
4,800
5,508
(670)
330
1,038
HFC-152a
140
124
138
167
16
14
43
HFC-227ea
2,900
3,220
3,350
3,860
(320)
130
640
HFC-236fa
6,300
9,810
8,060
8,998
(3,510)
(1,750)
(812)
HFC-4310mee
1,300
1,640
1,650
1,952
(340)
10
312
CF4
6,500
7,390
6,630
7,349
(890)
(760)
(41)
C2F6
9,200
12,200
11,100
12,340
(3,000)
(1,100)
140
C4F10
7,000
8,860
9,200
10,213
(1,860)
340
1,353
C6Fl4
7,400
9,300
7,910
8,780
(1,900)
(1,390)
(520)
SF«
23,900
22,800
23,500
26,087
1,100
700
3,287
NF3
NA
17,200
16,100
17,885
NA
(1,100)
685
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-
CO2 gases in order to be consistent with the approach used in calculating the CO2 lifetime.
Additionally, the AR5 reported separate values for fossil versus biogenic methane in order to account
for the CO2 oxidation product.
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 CO2 is
only included in the value from AR5 that includes climate-carbon feedbacks.
Note: Parentheses indicate negative values.
Source: (IPCC 2013, IPCC 2007, IPCC 2001, IPCC 1996).
4
5 1.2 National Inventory Arrangements
6 The U.S. Environmental Protection Agency (EPA), in cooperation with other U.S. government agencies, prepares
7 the Inventory of U.S. Greenhouse Gas Emissions and Sinks. A wide range of agencies and individuals are involved
8 in supplying data to, planning methodological approaches and improvements, reviewing, or preparing portions of the
9 U.S. Inventory—including federal and state government authorities, research and academic institutions, industry
10 associations, and private consultants.
11 Within EPA, the Office of Atmospheric Programs (OAP) is the lead office responsible for the emission calculations
12 provided in the Inventory, as well as the completion of the National Inventory Report and the Common Reporting
13 Format (CRF) tables. EPA's Office of Transportation and Air Quality (OTAQ) is also involved in calculating
14 emissions for the Inventory. While the U.S. Department of State officially submits the annual Inventory to the
1-10 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
UNFCCC, EPA's OAP serves as the National Inventory Focal Point for technical questions and comments on the
U.S. Inventory. The staff of EPA coordinate the annual methodological choice, activity data collection, and emission
calculations at the individual source 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 collection and
consistency of cross-cutting issues in the Inventory.
Several other government agencies contribute to the collection and analysis of the underlying activity data used in
the Inventory calculations. Formal and informal relationships exist between EPA and other U.S. agencies that
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 military fuel consumption
and bunker fuels. Informal relationships also exist with other U.S. agencies to provide activity data for use in EPA's
emission calculations. These 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 diagrams the National Inventory Arrangements.
Introduction 1-11
-------
Figure 1-1: National Inventory Arrangements Diagram Inventory Process
United States National Inventory Arrangements
United Nations
Framework Convention
on Climate Change
Inventory Submission
Inventory Compilation
U.S. Environmental
Protection Agency
Inventory Compiler
Emission Calculations
U.S. Environmental
Protection Agency
Other U.S.
Government Agencies
USDA Forest Service,
USDA Agricultural Research
Service, NOAA, DOD, FAA
Data Collection
Energy
• U.S. Department of Energy and its National Laboratories
• Energy Information Administration
• U.S. Department of Transportation
• Bureau of Transportation Statistics
• Federal Highway Administration
• Federal Aviation Administration
• U.S. Department of Defense - Defense Logistics Agency
• U.S. Department of Commerce - Bureau of the Census
• U.S. Department of Homeland Security
• U.S. Department of Labor's Mine Safety and Health Administration
• EPA Office of Transportation and Air Quality MOVES Model
• EPA Greenhouse Gas Reporting Program (GHGRP) and Acid Rain Program
• American Association of Railroads
• American Public Transportation Association
• U.S. Department of Labor - Mine Safety and Health Administration
• Data from research studies, trade publications, and —=
industry associations
A /
Agriculture/LULUCF
• U.S. Department of Agriculture (USDA) National Agricultural Statistics
Service
• USDA Natural Resources Conservation Service
• USDA Economic Research Service
• USDA Farm Service Agency
• USDA Animal Plant Health Inspection Service
• Conservation Technology Information Service
• U.S. Geological Survey
¦ USDA Forest Service
• National Oceanic and Atmospheric Administration (NOAA)
• U.S. Department of the Interior Bureau of Land Management
• EPA Office of Solid Waste
• U.S. Census Bureau
• Alaska Department of Natural Resources
• American Society of Agricultural Engineers
• Association of American Plant Food Control Officials
• Tennessee Valley Authority
• Data from research studies, trade publications, LggflH
and industry associations
~\ /
Industrial Processes and Product Use
• U.S. Geological Survey National Minerals Information
Center
• EPA GHGRP
• U.S. Department of Commerce
• American Iron and Steel Institute (AISI)
• American Chemistry Council (ACC)
• U.S. Aluminum Association
• Air-Conditioning, Heating, and Refrigeration Institute
• Data from research studies, trade publications, and
industry associations
Waste
• EPA GHGRP
• EPA Office of Land and Emergency
Management
• Data from research studies, trade
publications, and industry associations
1-12 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
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 data/document manager is responsible for compiling all emission estimates and ensuring
consistency and quality throughout the NIR and CRF tables. Emission calculations for individual sources are the
responsibility of individual source leads, who are most familiar with each source category and the unique
characteristics of its emissions profile. The individual source leads determine the most appropriate methodology and
collect the best activity data to use in the emission calculations, based upon their expertise in the source category, as
well as coordinating with researchers and contractors familiar with the sources. A multi-stage process for collecting
information from the individual source leads and producing the Inventory is undertaken annually to compile all
information and data.
Methodology Development, Data Collection, and Emissions
and Sink Estimation
Source leads at EPA collect input data and, as necessary, evaluate or develop the estimation methodology for the
individual source categories. Because EPA has been preparing the Inventory for many years, for most source
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 category are being developed for the first time, or if the methodology is changing for an existing source
category (e.g., the United States is implementing a higher Tiered approach for that source category), then the source
category lead will develop a new methodology, gather the most appropriate activity data and emission factors (or in
some cases direct emission measurements) for the entire time series, and conduct a special source-specific review
process involving relevant experts from industry, government, and universities (see Box ES-6 on approach to
recalculations).
Once the methodology is in place and the data are collected, the individual source leads calculate emissions and sink
estimates. The source leads then update or create the relevant text and accompanying annexes for the Inventory.
Source leads are also responsible for completing the relevant sectoral background tables of the CRF, conducting
quality assurance and quality control (QA/QC) checks, and uncertainty analyses.
The treatment of confidential business information (CBI) in the Inventory is based on EPA internal guidelines, as
well as regulations30 applicable to the data used. EPA has specific procedures in place to safeguard CBI during the
inventory compilation process. When information derived from CBI data is used for development of inventory
calculations, EPA procedures ensure that these confidential data are sufficiently aggregated to protect confidentiality
while still providing useful information for analysis. For example, within the Energy and Industrial Process and
Product Use (IPPU) sectors, EPA has used aggregated facility-level data from the Greenhous Gas Reporting
Program (GHGRP) to develop, inform, and/or quality-assure U.S. emissions estimates. In 2014, the EPA's GHGRP,
with industry engagement, compiled criteria that would be used for aggregating its confidential data to shield the
underlying CBI from public disclosure.31 In the Inventory, EPA is publishing only data values that meet the
GHGRP aggregation criteria.32 Specific uses of aggregated facility-level data are described in the respective
methodological sections within those chapters. In addition, EPA also 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 confidential business information.
30 40 CFR part 2, Subpart B titled "Confidentiality of Business Information" which is the regulation establishing rules governing
handling of data entitled to confidentiality treatment. See .
31 Federal Register Notice on "Greenhouse Gas Reporting Program: Publication of Aggregated Greenhouse Gas Data." See pp,
79 and 110 of notice at .
32 U.S. EPA Greenhouse Gas Reporting Program. Developments on Publication of Aggregated Greenhouse Gas Data, November
25, 2014. See .
Introduction 1-13
-------
1 Summary Data Compilation and Storage
2 The inventory coordinator at EPA with support from the data/document manager collect the source and sink
3 categories' descriptive text and Annexes, and also aggregates the emission estimates into a summary spreadsheet
4 that links the individual source category spreadsheets together. This summary sheet contains all of the essential data
5 in one central location, in formats commonly used in the Inventory document. In addition to the data from each
6 source category, national trend and related data are also gathered in the summary sheet for use in the Executive
7 Summary, Introduction, and Recent Trends sections of the Inventory report. Electronic copies of each year's
8 summary spreadsheet, which contains all the emission and sink estimates for the United States, are kept on a central
9 server at EPA under the jurisdiction of the inventory coordinator.
10 National Inventory Report Preparation
11 The NIR is compiled from the sections developed by each individual source or sink lead. In addition, the inventory
12 coordinator prepares a brief overview of each chapter that summarizes the emissions from all sources discussed in
13 the chapters. The inventory coordinator then carries out a key category analysis for the Inventory, consistent with the
14 2006IPCC Guidelines for National Greenhouse Gas Inventories, and in accordance with the reporting requirements
15 of the UNFCCC. Also at this time, the Introduction, Executive Summary, and Recent Trends sections are drafted, to
16 reflect the trends for the most recent year of the current Inventory. The analysis of trends necessitates gathering
17 supplemental data, including weather and temperature conditions, economic activity and gross domestic product,
18 population, atmospheric conditions, and the annual consumption of electricity, energy, and fossil fuels. Changes in
19 these data are used to explain the trends observed in greenhouse gas emissions in the United States. Furthermore,
20 specific factors that affect individual sectors are researched and discussed. Many of the factors that affect emissions
21 are included in the Inventory document as separate analyses or side discussions in boxes within the text. Text boxes
22 are also created to examine the data aggregated in different ways than in the remainder of the document, such as a
23 focus on transportation activities or emissions from electricity generation. The document is prepared to match the
24 specification of the UNFCCC reporting guidelines for National Inventory Reports.
25 Common Reporting Format Table Compilation
26 The CRF tables are compiled from individual tables completed by each individual source or sink lead, which contain
27 source emissions and activity data. The inventory coordinator integrates the source data into the UNFCCC's "CRF
28 Reporter" for the United States, assuring consistency across all sectoral tables. The summary reports for emissions,
29 methods, and emission factors used, the overview tables for completeness and quality of estimates, the recalculation
30 tables, the notation key completion tables, and the emission trends tables are then completed by the inventory
31 coordinator. Internal automated quality checks on the CRF Reporter, as well as reviews by the source leads, are
32 completed for the entire time series of CRF tables before submission.
33 Ci\h\€ jn4
34 QA/QC and uncertainty analyses are supervised by the QA/QC and uncertainty coordinators, who have general
35 oversight over the implementation of the QA/QC plan and the overall uncertainty analysis for the Inventory (see
36 sections on QA/QC and Uncertainty, below). These coordinators work closely with the source leads to ensure that a
37 consistent QA/QC plan and uncertainty analysis is implemented across all inventory sources. The inventory QA/QC
38 plan, detailed in a following section, is consistent with the quality assurance procedures outlined by EPA and IPCC.
39 The QA/QC and uncertainty findings also inform overall improvement planning, and specific improvements are
40 noted in the Planned Improvements sections of respective categories. QA processes are outlined below.
41 Expert, Public, and UNFCCC Review Periods
42 During the 30-day Expert Review period, a first draft of the document is sent to a select list of technical experts
43 outside of EPA who are not directly involved in preparing estimates. The purpose of the Expert Review is to provide
44 an objective review, encourage feedback on the methodological and data sources used in the current Inventory,
45 especially for sources which have experienced any changes since the previous Inventory.
1-14 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Once comments are received and addressed, a second draft of the document is released for public review by
publishing a notice in the U.S. Federal Register and posting the document on the EPA Web site. 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 categories Planned Improvement sections. See those sections for specific
details. EPA publishes comments received with publication of the 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.33
Feedback from these review processes all contribute to improving inventory quality over time.
Final Submittal to UNFCCC and Document Printing
After the final revisions to incorporate any comments from the Expert Review and Public Review periods, EPA
prepares the final National Inventory Report and the accompanying Common Reporting Format Reporter database.
EPA as the National Inventory focal point with the U.S. Department of State sends the official submission of the
U.S. Inventory to the UNFCCC. The document is then formatted and posted online, available for the public.34
1.3 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). 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 data sources. In addition to official statistics, the
report utilizes findings from academic studies, trade association surveys and statistical reports, along with expert
judgement, consistent with 2006 IPCC Guidelines.
The IPCC 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. 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. Complete documentation is provided in the annexes on the detailed methodologies and data sources utilized
in the calculation of each source category.
Box 1-3: IPCC Reference Approach
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. 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.
33 See .
34 See .
Introduction 1-15
-------
1 and therefore the carbon in them is oxidized and released into the atmosphere. Accounting for actual consumption of
2 fuels at the sectoral or sub-national level is not required.
3
4 1.4 Key Categories
5 The 2006IPCC Guidelines (IPCC 2006) defines a key category as a "[category] that is prioritized within the
6 national inventory system because its estimate lias a significant influence on a country's total inventory of
7 greenhouse gases in terms of the absolute level, the trend, or the uncertainty in emissions and removals."35 By
8 definition key categories include those categories that have the greatest contribution to the absolute level of national
9 emissions. In addition when an entire time series of emission and removal estimates is prepared, a thorough
10 investigation of key categories must also account for the influence of trends and uncertainties of individual source
11 and sink categories. This analysis can identify source and sink categories that diverge from the overall trend in
12 national emissions. Finally, a qualitative evaluation of key categories is performed to capture any categories that
13 were not identified in any of the quantitative analyses.
14 Approach 1, as defined in the 2006 IPCC Guidelines (IPCC 2006), was implemented to identify the key categories
15 for the United States. This analysis was performed twice; one analysis included sources and sinks from the Land
16 Use, Land-Use Change, and Forestry (LULUCF) sector, the other analysis did not include the LULUCF categories.
17 Following Approach 1, Approach 2, as defined in the 2006 IPCC Guidelines (IPCC 2006), was then implemented to
18 identify any additional key categories not already identified in Approach 1 assessment. This analysis, which includes
19 each source category's uncertainty assessments (or proxies) in its calculations, was also performed twice to include
20 or exclude LULUCF categories.
21 In addition to conducting Approach 1 and 2 level and trend assessments, a qualitative assessment of the source
22 categories, as described in the 2006 IPCC Guidelines (IPCC 2006), was conducted to capture any key categories that
23 were not identified by either quantitative method. For this inventory, no additional categories were identified using
24 criteria recommend by IPCC, but EPA continues to update its qualitative assessment on an annual basis.
25 Table 1-4: Key Categories for the United States (1990-2016)
CRF Source Categories
Gas
Approach 1
Approach 2
Quala
2016
Emissions
(MMT
CO2 Eq.)
Level Trend Level Trend
Without Without With With
LULUCF LULUCF LULUCF LULUCF
Level Trend Level Trend
Without Without With With
LULUCF LULUCF LULUCF LULUCF
Energy
CO2 Emissions from
Mobile Combustion:
Road
CO2
.
.
1,504.0
CO2 Emissions from
Stationary Combustion
- Coal - Electricity
Generation
CO2
.
.
1,241.3
CO2 Emissions from
Stationary Combustion
- Gas - Electricity
Generation
CO2
.
.
545.9
CO2 Emissions from
Stationary Combustion
- Gas - Industrial
CO2
.
.
478.8
35 See Chapter 4 Volume 1, "Methodological Choice and Identification of Key Categories" in IPCC (2006). See
.
1-16 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
CO2 Emissions from
Stationary Combustion
- Oil - Industrial
CO2
.
.
269.7
CO2 Emissions from
Stationary Combustion
- Gas - Residential
CO2
.
•
238.3
CO2 Emissions from
Stationary Combustion
- Gas - Commercial
CO2
.
.
170.3
CO2 Emissions from
Mobile Combustion:
Aviation
CO2
.
.
169.6
CO2 Emissions from
Non-Energy Use of
Fuels
CO2
•
121.0
CO2 Emissions from
Mobile Combustion:
Other
CO2
80.1
CO2 Emissions from
Stationary Combustion
- Coal - Industrial
CO2
.
.
59.0
CO2 Emissions from
Stationary Combustion
- Oil - Residential
CO2
.
•
58.0
CO2 Emissions from
Stationary Combustion
- Oil - Commercial
CO2
.
•
55.3
CO2 Emissions from
Mobile Combustion:
Marine
CO2
41.1
CO2 Emissions from
Stationary Combustion
- Oil - U.S. Territories
CO2
.
34.3
CO2 Emissions from
Natural Gas Systems
CO2
26.7
CO2 Emissions from
Petroleum Systems
CO2
.
.
25.5
CO2 Emissions from
Stationary Combustion
- Oil - Electricity
Generation
CO2
.
.
21.2
CO2 Emissions from
Stationary Combustion
- Gas - U.S. Territories
CO2
•
3.0
CO2 Emissions from
Stationary Combustion
- Coal - Commercial
CO2
•
2.3
CH4 Emissions from
Natural Gas Systems
CH4
.
.
162.1
Fugitive Emissions
from Coal Mining
ch4
.
.
53.8
CH4 Emissions from
Petroleum Systems
ch4
•
•
39.3
N011-CO2 Emissions
from Stationary
Combustion -
Residential
ch4
•
3.4
N011-CO2 Emissions
from Stationary
Combustion -
Electricity Generation
N2O
•
•
14.9
Introduction 1-17
-------
N2O Emissions from
Mobile Combustion:
Road
N2O
.
.
13.1
Non- CO2 Emissions
from Stationary
Combustion - Industrial
N2O
•
2.4
International Bunker
Fuelsb
Several
•
115.5
Industrial Processes and Product Use
CO2 Emissions from
Iron and Steel
Production &
Metallurgical Coke
Production
CO2
.
.
42.2
CO2 Emissions from
Cement Production
CO2
•
39.4
CO2 Emissions from
Petrochemical
Production
CO2
•
27.4
CO2 Emissions from
Other Process Uses of
Carbonates
CO2
11.2
N2O Emissions from
Adipic Acid Production
N2O
7.0
Emissions from
Substitutes for Ozone
Depleting Substances
HiGWP
.
.
173.9
SFo Emissions from
Electrical Transmission
and Distribution
HiGWP
•
4.3
HFC-23 Emissions
from HCFC-22
Production
HiGWP
.
•
2.8
PFC Emissions from
Aluminum Production
HiGWP
•
1.4
Agriculture
CO2 Emissions from
Liming
CO2
•
3.9
CH4 Emissions from
Enteric Fermentation
CH4
•
•
170.1
CH4 Emissions from
Manure Management
ch4
.
.
67.7
CH4 Emissions from
Rice Cultivation
ch4
•
13.7
Direct N2O Emissions
from Agricultural Soil
Management
N2O
.
.
237.6
Indirect N2O
Emissions from
Applied Nitrogen
N2O
.
.
45.9
Waste
CH4 Emissions from
Landfills
CH4
.
.
107.7
Land Use, Land Use Change, and Forestry
Net CO2 Emissions
from Land Converted to
Settlements
CO2
•
•
68.0
Net CO2 Emissions
from Land Converted to
Cropland
CO2
•
•
23.8
1-18 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
Net CO2 Emissions
from Land Converted to
Grassland
CO2
•
22.0
Net CO2 Emissions
from Grassland
Remaining Grassland
CO2
•
(1.6)
Net CO2 Emissions
from Cropland
Remaining Cropland
CO2
•
•
(9.9)
Net CO2 Emissions
from Land Converted to
Forest Land
CO2
•
(75.0)
Net CO2 Emissions
from Settlements
Remaining Settlements
CO2
•
•
(103.7)
Net CO2 Emissions
from Forest Land
Remaining Forest Land
CO2
•
•
(670.5)
CH4 Emissions from
Forest Fires
CH4
•
18.5
N2O Emissions from
Forest Fires
N2O
•
12.2
Subtotal Without LULUCF
6,390.8
Total Emissions Without LULUCF
6,546.2
Percent of Total Without LULUCF
98%
Subtotal With LULUCF
5,651.5
Total Emissions With LULUCF
5,829.3
Percent of Total With LULUCF
97%
a Qualitative criteria.
b Emissions from this source not included in totals.
Note: Parentheses indicate negative values (or sequestration).
1
2 1.5 Quality Assurance and Quality Control
3 (QA/QC)
4 As part of efforts to achieve its stated goals for inventory quality, transparency, and credibility, the United States has
5 developed a quality assurance and quality control plan designed to check, document and improve the quality of its
6 inventory over time. QA/QC activities on the Inventory are undertaken within the framework of the U.S. Quality
1 Assurance/Quality Control and Uncertainty Management Plan (QA/QC plan) for the U.S. Greenhouse Gas
8 Inventory: Procedures Manual for OA/OC and Uncertainty Analysis.
9 Key attributes of the QA/QC plan are summarized in Figure 1-2. These attributes include:
10 • Procedures and Forms: detailed and specific systems that serve to standardize the process of documenting
11 and archiving information as well as to guide the implementation of QA/QC and the analysis of
12 uncertainty
13 • Implementation of Procedures: application of QA/QC procedures throughout the whole inventory
14 development process from initial data collection through preparation of the emission estimates, to
15 publication of the Inventory
16 • Quality Assurance: expert and public reviews for both the Inventory estimates and the Inventory report
Introduction 1-19
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
(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
2006IPCC Guidelines (IPCC 2006)
• Quality Control, consideration of secondary data and category-specific checks (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.
In addition, based on the national QA/QC plan for the Inventory, source-specific QA/QC plans have been developed
for a number of sources. These plans follow the procedures outlined in the national QA/QC plan, tailoring the
procedures to the specific text and spreadsheets of the individual sources. For each greenhouse gas emissions source
or sink included in this Inventory, a minimum of general or Tier 1 QA/QC analysis has been undertaken. Where
QA/QC activities for a particular source go beyond the minimum Tier 1 level, and include category-specific checks
(Tier 2) further explanation is provided within the respective source 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 category sections in each chapter.
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 estimates. The QA/QC procedures also include feedback loops
and provide for corrective actions that are designed to improve the inventory estimates over time.
1-20 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Figure 1-2: U.S. QA/QC Plan Summary
_>*
03
C
<
>
CI
Obtain data in electronic
format (if possible)
Review spreadsheet
construction
Avoid hardwiring
• Use data validation
Protect cells
Develop automatic
checkers for:
Outliers, negative
values, or missing
data
• Variabletypes
match values
• Time series
consistency
Maintain tracking tab for
status of gathering
efforts
Check input data for
transcription errors
Inspect automatic
checkers
Identify spreadsheet
modifications that could
provide additional
QA/QC checks
Contact reports for non-
electroniccommunications
Provide cell references for
primary data elements
Obtain copies of all data
sources
~stand location of any
working/external
spreadsheets
Document assumptions
Check citations in
spreadsheet and text for
accuracy and style
Check reference docket for
new citations
Review documentation for
any data/ methodology
changes
Clearly label parameters,
units, and conversion
factors
Review spreadsheet
integrity
¦ Equations
• Units
¦ Inputs and output
Develop automated
checkers for:
¦ Input ranges
¦ Calculations
• Emission aggregation
Reproduce calculations
Reviewtime series
consistency
Review changes in
data/consistency with IPCC
methodology
Common starting
versions for each
inventory year
Utilize unalterable
summarytab foreach
source spreadsheetfor
linkingtoa master
summary spreadsheet
Follow strict version
control procedures
Document QA/QC
procedures
Data Gathering
Data Documentation Calculating Emissions
Cross-Cutting
Coordination
1.6 Uncertainty Analysis of Emission Estimates
Uncertainty estimates are an essential element of a complete and transparent emissions inventory. Uncertainty
information is not intended to dispute the validity of the Inventory estimates, but to help prioritize efforts to improve
the accuracy of future Inventories and guide future decisions on methodological choice. While the U.S. Inventory
calculates its emission estimates with the highest possible accuracy, uncertainties are associated to a varying degree
with the development of emission estimates for any inventory. Some of the current estimates, such as those for
carbon dioxide (CO2) emissions from energy-related activities, are considered to have minimal uncertainty
associated with them. For some other limited categories of emissions, however, a lack of data or an incomplete
understanding of how emissions are generated increases the uncertainty or systematic error associated with the
estimates presented. The UNFCCC reporting guidelines follow the recommendation in the 2006 IPCC Guidelines
(IPCC 2006) and require that countries provide single point estimates for each gas and emission or removal source
category. Within the discussion of each emission source, specific factors affecting the uncertainty associated with
the estimates are discussed.
Additional research in the following areas could help reduce uncertainty in the U.S. Inventory:
• Incorporating excluded emission sources. Quantitative estimates for some of the sources and sinks of
greenhouse gas emissions are not available at this time. In particular, emissions from some land-use
activities (e.g., emissions and removals from interior Alaska) and industrial processes are not included in
the inventory either because data are incomplete or because methodologies do not exist for estimating
Introduction 1-21
-------
1 emissions from these source categories. See Annex 5 of this report for a discussion of the sources of
2 greenhouse gas emissions and sinks excluded from this report.
3 • Improving the accuracy of emission factors. Further research is needed in some cases to improve the
4 accuracy of emission factors used to calculate emissions from a variety of sources. For example, the
5 accuracy of current emission factors applied to CH4 and N20 emissions from stationary and mobile
6 combustion is highly uncertain.
7 • Collecting detailed activity data. Although methodologies exist for estimating emissions for some sources,
8 problems arise in obtaining activity data at a level of detail where more technology or process-specific
9 emission factors can be applied.
10 The overall uncertainty estimate for total U.S. greenhouse gas emissions was developed using the IPCC Approach 2
11 uncertainty estimation methodology. Estimates of quantitative uncertainty for the total U.S. greenhouse gas
12 emissions are shown below, in Table 1-5.
13 The IPCC provides good practice guidance on two approaches—Approach 1 and Approach 2—to estimating
14 uncertainty for individual source categories. Approach 2 uncertainty analysis, employing the Monte Carlo Stochastic
15 Simulation technique, was applied wherever data and resources permitted; further explanation is provided within the
16 respective source category text and in Annex 7. Consistent with the 2006 IPCC Guidelines (IPCC 2006), over a
17 multi-year timeframe, the United States expects to continue to improve the uncertainty estimates presented in this
18 report.
19 Table 1-5: Estimated Overall Inventory Quantitative Uncertainty (MMT CO2 Eq. and Percent)
20 - TO BE UPDATED FOR FINAL INVENTORY REPORT
2015 Emission
Uncertainty Range Relative to Emission
Standard
Estimate3
Estimateb
Mean0
Deviation0
Gas
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
(MMT CO2 Eq.)
Lower
Upper
Lower
Upper
Boundd
Boundd
Bound
Bound
CO2
5,411.0
5,305.4
5,652.4
-2%
4%
5,474.3
90.2
CH4e
655.7
599.9
779.2
-9%
19%
681.8
45.3
N2Oe
334.8
302.5
424.6
-10%
27%
357.0
30.7
PFC,HFC,SF6,andNF3e
184.7
183.1
204.4
-1%
11%
193.4
5.5
Total
6,586.2
6,505.0
6,919.9
-1%
5%
6,706.6
106.0
LULUCF Emissions'
19.7
14.6
38.2
-26%
94%
23.3
6.3
LULUCF Total Net Flux®
(778.7)
(993.1)
(620.7)
-20%
28%
(808.4)
94.7
LULUCF Sector Total"
(758.9)
(969.7)
(597.9)
-21%
28%
(785.1)
94.8
Net Emissions (Sources and
Sinks)
5,827.3
5,643.8
6,207.4
-3%
7%
5,921.5
142.8
Notes: Total emissions (excluding emissions for which uncertainty was not quantified) is presented without LULUCF. Net
emissions is presented with LULUCF.
a Emission estimates reported in this table correspond to emissions from only those source categories for which quantitative
uncertainty was performed this year. Thus the totals reported in this table exclude approximately 0.4 MMT CO2 Eq. of
emissions for which quantitative uncertainty was not assessed. Hence, these emission estimates do not match the filial total U.S.
greenhouse gas emission estimates presented in this Inventory.
b 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.
c 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.
d 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.
e The overall uncertainty estimates did not take into account the uncertainty in the GWP values for CH4, N2O and high GWP
gases used in the Inventory emission calculations for 2015.
f LULUCF emissions include the CH4 and N2O emissions reported for N011-CO2 Emissions from Forest Fires, Emissions from
Drained Organic Soils, N2O Fluxes from Forest Soils, N011-CO2 Emissions from Grassland Fires, N2O Fluxes from Settlement
1-22 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
Soils. Coastal Wetlands Remaining Coastal Wetlands, Peatlands Remaining Peatlands, and CI 11 Emissions from Land
Converted to Coastal Wetlands..
•" Net C( h llu\ is the net C stock change from the follow ing categories: I'oresl I.and Remaining I'oresl I,and, hind ('onverlcd lo
I'oresl Land. ( ropland Remaining ( ropland. Land ( (inverted lo ( ropland, (irassland Remaining (irassland. I,and ( (inverted
10 ('wasshmd. Changes in ()rganic Soils Carbon Stocks. Changes in Urban Tree Carbon Slocks. Changes in Yard Trimmings and
Food Scrap Carbon Stocks in 1 .andlills, I.and ('(inverted lo Settlements. II ellands Remaining Wetlands, and {.and ('(inverted lo
11 ellands.
11 The H JI.UCF Sector Total is the net sum of all emissions (i.e., sources) of greenhouse gases to the atmosphere plus removals
of C(): (i.e.. sinks or negative emissions) from the atmosphere.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
1 Emissions calculated for the U.S. Inventory reflect current best estimates; in some cases, however, estimates are
2 based on approximate methodologies, assumptions, and incomplete data. As new information becomes available in
3 the future, the United States will continue to improve and revise its emission estimates. See Annex 7 of this report
4 for further details on the U. S. process for estimating uncertainty associated with the emission estimates and for a
5 more detailed discussion of the limitations of the current analysis and plans for improvement. Annex 7 also includes
6 details on the uncertainty analysis performed for selected source categories.
7 1.7 Completeness
8 This report, along with its accompanying CRF tables, serves as a thorough assessment of the anthropogenic sources
9 and sinks of greenhouse gas emissions for the United States for the time series 1990 through 2015. This report is
10 intended to be comprehensive and includes the vast majority of emissions and removals identified as anthropogenic,
11 consistent with IPCC and UNFCCC guidelines. In general, sources not accounted for in this Inventory are excluded
12 because they are not occurring in the U.S., or because data are unavailable to develop an estimate and/or the sources
13 were determined to be insignificant36 in terms of overall national emissions per UNFCCC reporting guidelines.
14 The United States is continually working to improve upon the understanding of such sources and seeking to find the
15 data required to estimate related emissions. As such improvements are implemented, new emission sources are
16 quantified and included in the Inventory. For a list of sources not included and more information, see Annex 5 and
17 the respective source category sections in each chapter of this report.
is 1.8 Organization of Report
19 In accordance with the revision of the UNFCCC reporting guidelines agreed to at the nineteenth Conference of the
20 Parties (UNFCCC 2014), this Inventory of U.S. Greenhouse Gas Emissions and Sinks is segregated into five sector-
21 specific chapters consistent with the UN Common Reporting Framework (CRF), listed below in Table 1-6. In
22 addition, chapters on Trends in Greenhouse Gas Emissions and Other information to be considered as part of the
23 U.S. Inventory submission are included.
36 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 CO2 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."
Introduction 1-23
-------
1 Table 1-6: IPCC Sector Descriptions
Chapter/IPCC Sector
Activities Included
Energy
Emissions of all greenhouse gases resulting from stationary and mobile energy
activities including fuel combustion and fugitive fuel emissions, and non-
energy use of fossil fuels.
Industrial Processes and
Emissions resulting from industrial processes and product use of greenhouse
Product Use
gases.
Agriculture
Emissions from agricultural activities except fuel combustion, which is
addressed under Energy.
Land Use, Land-Use
Emissions and removals of CO2, and emissions of CH4, and N2O from land use,
Change, and Forestry
land-use change and forestry.
Waste
Emissions from waste management activities.
2 Within each chapter, emissions are identified by the anthropogenic activity that is the source or sink of the
3 greenhouse gas emissions being estimated (e.g., coal mining). Overall, the following organizational structure is
4 consistently applied throughout this report:
5 Chapter/IPCC Sector: Overview of emission trends for each IPCC defined sector.
6 CRF Source or category: Description of source pathway and emission/removal trends based on IPCC
7 methodologies, consistent with UNFCCC reporting guidelines.
8 Methodology: Description of analytical methods (e.g. 2006 IPCC Guidelines) employed to produce emission
9 estimates and identification of data references, primarily for activity data and emission factors.
10 Uncertainty and Time Series Consistency: A discussion and quantification of the uncertainty in emission
11 estimates and a discussion of time-series consistency.
12 QA/QC and Verification: A discussion on steps taken to QA/QC and verily the emission estimates, consistent with
13 the U.S. QA/QC plan, and any key findings.
14 Recalculations: A discussion of any data or methodological changes that necessitate a recalculation of previous
15 years' emission estimates, and the impact of the recalculation on the emission estimates, if applicable.
16 Planned Improvements: A discussion on any category-specific planned improvements, if applicable.
17 Special attention is given to CO2 from fossil fuel combustion relative to other sources because of its share of
18 emissions and its dominant influence on emission trends. For example, each energy consuming end-use sector (i.e.,
19 residential, commercial, industrial, and transportation), as well as the electricity generation sector, is described
20 individually. Additional information for certain source categories and other topics is also provided in several
21 Annexes listed in Table 1-7.
22 Table 1-7: List of Annexes
ANNEX 1 Key Category Analysis
ANNEX 2 Methodology and Data for Estimating CO2 Emissions from Fossil Fuel Combustion
2.1. Methodology for Estimating Emissions of CO2 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, N2O, and Indirect Greenhouse Gases from Stationary
Combustion
3.2. Methodology for Estimating Emissions of CH4, N2O, 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 CO2 Emissions from Petroleum Systems
3.6. Methodology for Estimating CH4 Emissions from Natural Gas Systems
3.7. Methodology for Estimating CO2 and N2O 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
1-24 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
3.10. Methodology for Estimating CH4 Emissions from Enteric Fermentation
3.11. Methodology for Estimating CH4 and N2O Emissions from Manure Management
3.12. Methodology for Estimating N2O 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 CO2 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. Sulfur Dioxide Emissions
6.4. Complete List of Source Categories
6.5. Constants, Units, and Conversions
6.6. Abbreviations
6.7. Chemical Formulas
ANNEX 7 Uncertainty
7.1. Overview
7.2. Methodology and Results
7.3. Reducing Uncertainty
7.4. Planned Improvements
7.5. Additional Information on Uncertainty Analyses by Source
ANNEX 8 QA/QC Procedures
8.1. Background
8.2. Purpose
8.3. Assessment F actors
8.4 . Responses During the Review Process
Introduction 1-25
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
2. Trends in Greenhouse Gas Emissions
2.1 Recent Trends in U.S. Greenhouse Gas
Emissions and Sinks
In 2016, total gross U.S. greenhouse gas emissions were 6,546.2 MMT, or million metric tons, carbon dioxide (CO2)
Eq. Total U.S. emissions have increased by 2.8 percent from 1990 to 2016, and emissions decreased from 2015 to
2016 by 2.0 percent (131.1 MMT CO2 Eq.). The decrease in total greenhouse gas emissions between 2015 and 2016
was driven in large part by a decrease in CO2 emissions from fossil fuel combustion. The decrease in CO2 emissions
from fossil fuel combustion was a result of multiple factors, including:
(1) substitution from coal to natural gas and other sources in the electric power sector; and
(2) warmer winter conditions in 2016 resulting in a decreased demand for heating fuel in the residential and
commercial sectors.
Since 1990, U.S. emissions have increased at an average annual rate of 0.1 percent. Figure 2-1 through Figure 2-3
illustrate the overall trend in total U.S. emissions by gas, annual changes, and absolute changes since 1990. Overall,
net emissions in 2016 were 11.6 percent below 2005 levels as shown in Table 2-1.
Figure 2-1: Gross U.S. Greenhouse Gas Emissions by Gas (MMT CO2 Eq.)
9,000
8,000
7,000
6,000
o 5,000
U
4,000
3,000
2,000
1,000
0
I HFCs, PFCs, SFf
Nitrous Oxide
I Methane
I Carbon Dioxide
and NF3 Subtotal
Trends 2-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Figure 2-2: Annual Percent Change in Gross U.S. Greenhouse Gas Emissions Relative to the
Previous Year
4%
2%
0%
-2%
-4%
-6%
-8%
1,8% 1.7%
H M 1A% 1.3%
1
0.8% Mjj 0 6%
2.2%
0.5% 0-6%
1.9%
0.1%
1
3.5%
12.8%
I 0.9%
1 II || "l|
-2.8% I
^HfNro5'U">i£>fsvCOcrvO'-irNjro^u->ioi-vCOvo
On Ov On On On On On O O O O u O u O o o iH
OnOnOnOnOnOnOnOnOnCjOOOOCsOOOOOOOOOOO
Figure 2-3: Cumulative Change in Annual Gross U.S. Greenhouse Gas Emissions Relative to
1990 (1990=0, MMT COz Eq.)
1,200-
Overall, from 1990 to 2016, total emissions of CO2 increased by 196.5 MMT CO2 Eq. (3.8 percent), while total
emissions of methane (CH4) decreased by 122.3 MMT CO: Eq. (15.7 percent), and total emissions of nitrous oxide
(N2O) increased by 14.2 MMT CO2 Eq. (4.0 percent). During the same period, aggregate weighted emissions of
hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), sulfur hexafluoride (SF6), and nitrogen trifluoride (NF3) rose
by 88.6 MMT CO2 Eq. (88.8 percent). Despite being emitted in smaller quantities relative to the other principal
greenhouse gases, emissions of HFCs, PFCs, SF6, and NF3 are significant because many of them have extremely
high global wanning 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
11.5 percent of total emissions in 2016.
Table 2-1 summarizes 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.
2-2 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Table 2-1: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks (MMT CO2 Eq.)
Gas/Source
1990
2005
2012
2013
2014
2015
2016
CO2
5,136.8
6,150.8
5,383.7
5,541.7
5,590.5
5,449.5
5,333.4
Fossil Fuel Combustion
4,755.8
5,759.1
5,029.8
5,162.3
5,206.1
5,059.3
4,976.7
Electric Power
1,820.8
2,400.9
2,022.2
2,038.1
2,038.0
1,900.7
1,808.8
Treimportation
1,467.2
1,855.8
1,661.9
1,677.6
1,717.1
1,735.5
1,794.9
Industrial
874.5
867.8
818.4
848.7
830.8
819.3
807.6
Residential
338.3
357.8
282.5
329.7
345.3
316.8
296.2
Commercial
227.4
227.0
201.3
225.7
233.6
245.6
227.9
U.S. Territories
27.6
49.7
43.5
42.5
41.4
41.4
41.4
Non-Energy Use of Fuels
119.6
141.7
113.3
133.2
127.8
135.1
121.0
Iron and Steel Production &
Metallurgical Coke Production
101.5
68.0
55.4
53.3
58.2
47.7
42.2
Cement Production
33.5
46.2
35.3
36.4
39.4
39.9
39.4
Petrochemical Production
21.2
26.8
26.5
26.4
26.5
28.1
27.4
Natural Gas Systems
29.7
22.5
24.4
26.0
27.0
26.3
26.7
Petroleum Systems
9.4
17.0
25.6
29.7
32.9
38.0
25.5
Lime Production
11.7
14.6
13.8
14.0
14.2
13.3
13.3
Other Process Uses of Carbonates
4.9
6.3
8.0
10.4
11.8
11.2
11.2
Ammonia Production
13.0
9.2
9.4
10.0
9.6
10.6
11.2
Incineration of Waste
8.0
12.5
10.4
10.4
10.6
10.7
10.7
Urea Fertilization
2.4
3.5
4.3
4.4
4.5
4.9
5.1
Carbon Dioxide Consumption
1.5
1.4
4.0
4.2
4.5
4.5
4.5
Urea Consumption for Non-
Agricultural Purposes
3.8
3.7
4.4
4.1
1.5
4.2
4.0
Liming
4.7
4.3
6.0
3.9
3.6
3.8
3.9
Ferroalloy Production
2.2
1.4
1.9
1.8
1.9
2.0
1.8
Soda Ash Production
1.4
1.7
1.7
1.7
1.7
1.7
1.7
Titanium Dioxide Production
1.2
1.8
1.5
1.7
1.7
1.6
1.6
Aluminum Production
6.8
4.1
3.4
3.3
2.8
2.8
1.3
Glass Production
1.5
1.9
1.2
1.3
1.3
1.3
1.3
Phosphoric Acid Production
1.5
1.3
1.1
1.1
1.0
1.0
1.0
Zinc Production
0.6
1.0
1.5
1.4
1.0
0.9
0.9
Lead Production
0.5
0.6
0.5
0.5
0.5
0.5
0.5
Silicon Carbide Production and
Consumption
0.4
0.2
0.2
0.2
0.2
0.2
0.2
Magnesium Production and
Processing
+
+
+
+
+
+
+
Wood Biomass, Ethanol, and
Biodiesel Consumption"
219.4
230.7
276.2
299.8
308.3
294.5
291.1
International Bunker Fuelsb
103.5
113.1
105.8
99.8
103.4
110.9
114.4
CH4c
778.1
679.3
661.3
659.6
665.3
664.0
655.8
Enteric Fermentation
164.2
168.9
166.7
165.5
164.2
166.5
170.1
Natural Gas Systems
193.7
160.0
156.8
159.6
164.2
164.4
162.1
Landfills
179.6
132.7
117.0
113.3
112.7
111.7
107.7
Manure Management
37.2
56.3
65.6
63.3
62.9
66.3
67.7
Coal Mining
96.5
64.1
66.5
64.6
64.6
61.2
53.8
Petroleum Systems
42.3
34.7
35.4
38.8
41.0
39.4
39.3
Wastewater Treatment
15.7
15.8
15.1
14.9
15.0
15.1
14.8
Rice Cultivation
16.0
16.7
11.3
11.5
12.7
12.3
13.7
Stationary Combustion
8.6
7.9
7.3
8.7
OO
OO
7.8
7.2
Abandoned Oil and Gas Wells
6.5
6.9
7.0
7.0
7.1
7.2
7.1
Abandoned Underground Coal
Mines
7.2
6.6
6.2
6.2
6.3
6.4
6.7
Mobile Combustion
9.8
6.6
4.0
3.7
3.4
3.1
3.0
Composting
0.4
1.9
1.9
2.0
2.1
2.1
2.1
Field Burning of Agricultural
Residues
0.2
0.2
0.3
0.3
0.3
0.3
0.3
Trends 2-3
-------
Petrochemical Production
0.2
0.1
0.1
0.1
0.1
0.2
0.2
Ferroalloy Production
+
+
+
+
+
+
+
Silicon Carbide Production and
Consumption
+
+
+
+
+
+
+
Iron and Steel Production &
Metallurgical Coke Production
+
+
+
+
+
+
+
Incineration of Waste
+
+
+
+
+
+
+
International Bunker Fuelsb
0.2
0.1
0.1
0.1
0.1
0.1
0.1
N2Oc
354.6
357.4
335.2
362.6
360.5
378.9
368.8
Agricultural Soil Management
250.5
253.5
247.9
276.6
274.0
295.0
283.6
Stationary Combustion
11.1
17.5
16.8
18.6
18.9
18.0
18.4
Manure Management
14.0
16.5
17.5
17.5
17.5
17.7
18.1
Mobile Combustion
41.5
38.4
23.8
22.0
20.2
18.8
17.8
Nitric Acid Production
12.1
11.3
10.5
10.7
10.9
11.6
10.2
Adipic Acid Production
15.2
7.1
5.5
3.9
5.4
4.3
7.0
Wastewater Treatment
3.4
4.4
4.6
4.7
4.8
4.8
5.0
NjO 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
2.0
2.0
2.0
2.0
2.0
Composting
0.3
1.7
1.7
1.8
1.9
1.9
1.9
Incineration of Waste
0.5
0.4
0.3
0.3
0.3
0.3
0.3
Semiconductor Manufacture
+
0.1
0.2
0.2
0.2
0.2
0.2
Field Burning of Agricultural
Residues
0.1
0.1
0.1
0.1
0.1
0.1
0.1
International Bunker Fuelsb
0.9
1.0
0.9
0.9
0.9
0.9
1.0
HFCs
46.6
120.0
156.0
159.1
166.8
173.3
177.1
Substitution of Ozone Depleting
Substances'1
0.3
99.8
150.3
154.8
161.4
168.6
173.9
HCFC-22 Production
46.1
20.0
5.5
4.1
5.0
4.3
2.8
Semiconductor Manufacture
0.2
0.2
0.2
0.2
0.3
0.3
0.3
Magnesium Production and
Processing
0.0
0.0
+
0.1
0.1
0.1
0.1
PFCs
24.3
6.7
5.9
5.8
5.6
5.1
4.3
Semiconductor Manufacture
2.8
3.3
3.0
2.8
3.1
3.1
3.0
Aluminum Production
21.5
3.4
2.9
3.0
2.5
2.0
1.4
Substitution of Ozone Depleting
Substances
0.0
+
+
+
+
+
+
SF«
28.8
11.7
6.6
6.3
6.3
5.9
6.2
Electrical Transmission and
Distribution
23.1
8.3
4.6
4.5
4.6
4.2
4.3
Magnesium Production and
Processing
5.2
2.7
1.6
1.5
1.0
0.9
1.0
Semiconductor Manufacture
0.5
0.7
0.3
0.4
0.7
0.7
0.8
NF3
+
0.5
0.6
0.6
0.5
0.6
0.6
Semiconductor Manufacture
+
0.5
0.6
0.6
0.5
0.6
0.6
Total Emissions
6,369.2
7,326.4
6,549.4
6,735.6
6,795.6
6,677.3
6,546.2
LULUCF Emissions0
10.6
23.0
26.1
19.2
19.6
38.2
38.1
LULUCF CH4 Emissions
6.7
13.3
15.0
10.9
11.2
22.4
22.4
LULUCF N2O Emissions
3.9
9.7
11.1
8.3
8.4
15.8
15.7
LULUCF Carbon Stock Change'
(830.2)
(754.2)
(779.5)
(755.0)
(760.0)
(733.4)
(754.9)
LULUCF Sector Net Total'
(819.6)
(731.1)
(753.5)
(735.8)
(740.4)
(695.2)
(716.8)
Net Emissions (Sources and Sinks)
5,549.6
6,595.3
5,795.9
5,999.9
6,055.2
5,982.1
5,829.3
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.
+ Does not exceed 0.05 MMT CO2 Eq.
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.
2-4 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
c LULUCF emissions of CH4 and N2O are reported separately from gross emissions totals. LULUCF emissions include
the CH4 and N2O 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 N2O 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 CH4 and N2O emissions to the atmosphere plus net carbon stock
changes.
1 Table 2-2: Recent Trends in U.S. Greenhouse Gas Emissions and Sinks (kt)
Gas/Source
1W0
2005
2012
2013
2014
2015
2016
CO2
5,136,|SI4
6,150,751
5,383,670
5,541,745
5,590,474
5,449,474
5,333,352
Fossil Fuel Combustion
4,755,819
5,759.056
5,029,830
5,162,315
5,206,135
5,059,288
4,976,737
Electric Power
1,820,SIS
2,400,874
2,022,181
2,038,122
2,038,018
1,900,673
1,808,797
Transportation
1,467,193
1,855.751
1,661,895
1,677,593
1,717,132
1,735,503
1,794,886
Industrial
874,5-IS
867,833
818,402
848,669
830,766
819,273
807,565
Residential
338,347
357,834
282,501
329,742
345,296
316,822
296,238
Commercial
227,358
227,041
201,325
225,722
233,557
245,637
227,871
U.S. Territories
27,555
49,723
43,527
42,467
41,365
41,380
41,380
Non-Energy Use of Fuels
119,588
141.669
113,275
133,176
127,'778
135,106
121,049
Iron and Steel Production &
Metallurgical Coke
Production
101,487
68.047
55,449
53,348
58,234
47,718
42,219
Cement Production
33,484
46.194
35,270
36,369
39,439
39,907
39,439
Petrochemical Production
21,203
26.794
26,501
26,395
26,496
28,062
27,411
Natural Gas Systems
29,708
22.529
24,398
26,004
27,004
26,329
26,739
Petroleum Systems
9,384
17.004
25,629
29,695
32,895
37,971
25,543
Lime Production
11,700
14.552
13,785
14,028
14,210
13,342
13,342
Other Process Uses of
Carbonates
4,907
6.339
8,022
10,414
11,811
11,237
11,237
Ammonia Production
13,047
9,196
9,377
9,962
9,619
10,571
11,234
Incineration of Waste
7,950
12.469
10,392
10,363
10,608
10,676
10,676
Urea Fertilization
2,417
3,504
4,282
4,443
4,538
4,888
5,098
Carbon Dioxide Consumption
1,472
1.375
4,019
4,188
4,471
4,471
4,471
Urea Consumption for Non-
Agricultural Purposes
3,784
3.653
4,392
4,074
1,541
4,169
3,959
Liming
4,667
4,349
5,978
3,907
3,609
3,777
3,863
Ferroalloy Production
2,152
1.392
1,903
1,785
1,914
1,960
1,796
Soda Ash Production
1,431
1,655
1,665
1,694
1,685
1,714
1,723
Titanium Dioxide Production
1,195
1.755
1,528
1,715
1,688
1,635
1,608
Aluminum Production
6,831
4,142
3,439
3,255
2,833
2,767
1,334
Glass Production
1,535
1.928
1,248
1,317
1,336
1,299
1,299
Phosphoric Acid Production
1,529
1,342
1,118
1,149
1,038
999
992
Zinc Production
632
1.030
1,486
1,429
956
933
925
Lead Production
516
527
546
459
473
482
Silicon Carbide Production and
Consumption
375
219
158
169
173
180
174
Magnesium Production and
Processing
1
2
2
2
3
3
Wood Biomass, Ethanol, and
Biodiesel Consumption"
219,413
230,700
276,201
299,785
308,346
294,469
291,069
International Bunker Fuelsh
103,463
113,139
105,805
99,763
103,400
110,887
114,394
Trends 2-5
-------
CH4c
31,125
27,170
26,451
26,383
26,613
26,561
26,232
Enteric Fermentation
6,566
6,755
6,670
6,619
6,567
6,661
6,805
Natural Gas Systems
7,748
6,399
6,273
6,385
6,568
6,578
6,483
Landfills
7,182
5,310
4,680
4,531
4,509
4,467
4,306
Manure Management
1,486
2,254
2,625
2,530
2,514
2,651
2,709
Coal Mining
3,860
2,56.5
2,658
2,584
2,583
2,449
2,153
Petroleum Systems
1,693
1,386
1,415
1,553
1,639
1,576
1,571
Wastewater Treatment
627
631
604
596
598
605
593
Rice Cultivation
641
66"
453
462
510
493
549
Stationary Combustion
346
314
292
346
352
313
288
Abandoned Oil and Gas Wells
260
275
279
280
282
286
284
Abandoned Underground Coal
Mines
288
264
249
249
253
256
268
Mobile Combustion
393
26o
160
149
136
123
119
Composting
15
7.5
77
81
84
84
85
Field Burning of Agricultural
Residues
9
8
11
11
11
11
11
Petrochemical Production
9
3
3
5
7
7
Ferroalloy Production
1
-
1
+
1
1
1
Silicon Carbide Production and
Consumption
1
+ /
+
+
+
+
+
Iron and Steel Production &
Metallurgical Coke
Production
1
1
+
+
+
+
+
Incineration of Waste
+
+ /
+
+
+
+
+
International Bunker Fuelsb
7
.5
4
3
3
3
4
N2Oc
1,190
1,199
1,125
1,217
1,210
1,272
1,238
Agricultural Soil Management
840
851
832
928
920
990
952
Stationary Combustion
37
59
56
62
63
60
62
Manure Management
47
5.5
59
59
59
59
61
Mobile Combustion
139
129
80
74
68
63
60
Nitric Acid Production
41
38
35
36
37
39
34
Adipic Acid Production
51
24
19
13
18
14
23
Wastewater Treatment
11
1.5
16
16
16
16
17
N20 from Product Uses
14
14
14
14
14
14
14
Caprolactam, Glyoxal, and
Glyoxylic Acid Production
6
7
7
7
7
7
Composting
1
6
6
6
6
6
6
Incineration of Waste
2
1
1
1
1
1
1
Semiconductor Manufacture
+
-
1
1
1
1
1
Field Burning of Agricultural
Residues
+
-
+
+
+
+
+
International Bunker Fuelsb
3
3
3
3
3
3
3
HFCs
M
IV1
M
M
M
M
M
Substitution of Ozone
Depleting Substances'1
M
M
M
M
M
M
M
HCFC-22 Production
3
1
+
+
+
+
+
Semiconductor Manufacture
M
M
M
M
M
M
M
Magnesium Production and
Processing
0
0
+
+
+
+
+
PFCs
M
IV1
M
M
M
M
M
Semiconductor Manufacture
M
M
M
M
M
M
M
Aluminum Production
M
M
M
M
M
M
M
Substitution of Ozone
Depleting Substances
0
-
+
+
+
+
+
SF«
1
1
+
+
+
+
+
Electrical Transmission and
Distribution
1
-
+
+
+
+
+
2-6 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Magnesium Production and
Processing +
+
+ + + + +
Semiconductor Manufacture +
+
+ + + + +
NF3 +
+
+ + + + +
Semiconductor Manufacture +
+
+ + + + +
+ Does not exceed 0.5 kt.
M - Mixture of multiple gases
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 CUt and N2O are reported separately from gross emissions totals. Refer to Lable 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: Lotals may not sum due to independent rounding. Parentheses indicate negative values or sequestration.
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 2-4 and Table 2-3 illustrate that over the twenty-seven-
year period of 1990 to 2016, total emissions from the Energy, Industrial Processes and Product Use, and Agriculture
sectors grew by 136.2 MMT CO2 Eq. (2.6 percent), 35.2 MMT CO2 Eq. (10.3 percent), and 73.4 MMT CO2 Eq.
(15.0 percent), respectively. Emissions from the Waste sector decreased by 67.9 MMT CO2 Eq. (34.1 percent). Over
the same period, total C sequestration in the Land Use. Land-Use Change, and Forestry (LULUCF) sector increased
by 75.3 MMT CO2 (9.1 percent decrease in total C sequestration), and emissions from the LULUCF sector increased
by 27.4 MMT CO2 Eq. (258 percent).
Figure 2-4; U.S. Greenhouse Gas Emissions and Sinks by Chapter/IPCC Sector (MMT CO2
Eq.)
o
u
7,500-
7,000-
6,500-
6,000-
5,500-
5,000-
4,500-
4,000-
3,500-
3,000-
2,500-
2,000-
1,500-
1,000-
500-
0-
-500-
Waste
Industrial Processes and Product Use
—
Agriculture
LULUCF (emissions)
Energy
Land Use, Land-Use Change and Forestry (LULUCF) (removals)
o^rMroo»HrMro^rmvo
^0>O^ON<^0>(TtO^<^C^OOOOOOOOOO»-lrHHHH»-IH
cr>cr>cncr>(j>cr»cr*cncr>cr»ooooooooooooooooo
H »—I t-i r-t tH i-i y-t t—I fM fN| fM (M CM
-------
Petroleum Systems
51.7
51."
61.0
68.5
73.9
77.4
64.8
Coal Mining
96.5
64.1
66.5
64.6
64.6
61.2
53.8
Stationary Combustion
19.8
25.4
24.1
27.2
27.7
25.8
25.6
Mobile Combustion
51.3
45.0
27.8
25.7
23.6
21.9
20.8
Incineration of Waste
8.4
12.9
10.7
10.7
10.9
11.0
11.0
Abandoned Oil and Gas Wells
6.5
6.9
7.0
7.0
7.1
7.2
7.1
Abandoned Underground Coal Mines
7.2
6.6
6.2
6.2
6.3
6.4
6.7
Industrial Processes and Product Use
340.5
354.2
361.6
364.7
380.2
378.8
375.7
Substitution of Ozone Depleting
Substances
0.3
99.8
150.4
154.8
161.4
168.6
173.9
Iron and Steel Production &
Metallurgical Coke Production
101.5
68.1
55.5
53.4
58.2
47.7
42.2
Cement Production
33.5
46.2
35.3
36.4
39.4
39.9
39.4
Petrochemical Production
21.4
26.9
26.6
26.5
26.6
28.2
27.6
Lime Production
11.7
14.6
13.8
14.0
14.2
13.3
13.3
Other Process Uses of Carbonates
4.9
6.3
8.0
10.4
11.8
11.2
11.2
Ammonia Production
13.0
9.2
9.4
10.0
9.6
10.6
11.2
Nitric Acid Production
12.1
11.3
10.5
10.7
10.9
11.6
10.2
Adipic Acid Production
15.2
7.1
5.5
3.9
5.4
4.3
7.0
Semiconductor Manufacture
3.6
4."
4.4
4.0
4.9
5.0
5.0
Carbon Dioxide Consumption
1.5
1.4
4.0
4.2
4.5
4.5
4.5
Electrical Transmission and
Distribution
23.1
8.3
4.6
4.5
4.6
4.2
4.3
NjO from Product Uses
4.2
4.2
4.2
4.2
4.2
4.2
4.2
Urea Consumption for Non-
Agricultural Purposes
3.8
3.7
4.4
4.1
1.5
4.2
4.0
HCFC-22 Production
46.1
20.0
5.5
4.1
5.0
4.3
2.8
Aluminum Production
28.3
7.6
6.4
6.2
5.4
4.8
2.7
Caprolactam, Glyoxal, and Glyoxylic
Acid Production
1.7
2.1
2.0
2.0
2.0
2.0
2.0
Ferroalloy Production
2.2
1.4
1.9
1.8
1.9
2.0
1.8
Soda Ash Production
1.4
1."
1.7
1.7
1.7
1.7
1.7
Titanium Dioxide Production
1.2
1.8
1.5
1.7
1.7
1.6
1.6
Glass Production
1.5
1.9
1.2
1.3
1.3
1.3
1.3
Magnesium Production and
Processing
5.2
2."
1.7
1.5
1.1
1.0
1.1
Phosphoric Acid Production
1.5
1.3
1.1
1.1
1.0
1.0
1.0
Zinc Production
0.6
1.0
1.5
1.4
1.0
0.9
0.9
Lead Production
0.5
0.6
0.5
0.5
0.5
0.5
0.5
Silicon Carbide Production and
Consumption
0.4
0.2
0.2
0.2
0.2
0.2
0.2
Agriculture
489.2
520.0
519.8
543.1
539.8
566.9
562.6
Agricultural Soil Management
250.5
253.5
247.9
276.6
274.0
295.0
283.6
Enteric Fermentation
164.2
168.9
166.7
165.5
164.2
166.5
170.1
Manure Management
51.1
72.9
83.2
80.8
80.4
84.0
85.9
Rice Cultivation
16.0
16.7
11.3
11.5
12.7
12.3
13.7
Urea Fertilization
2.4
3.5
4.3
4.4
4.5
4.9
5.1
Liming
4.7
4.3
6.0
3.9
3.6
3.8
3.9
Field Burning of Agricultural
Residues
0.3
0.3
0.4
0.4
0.4
0.4
0.4
Waste
199.3
156.4
140.4
136.7
136.5
135.6
131.5
Landfills
179.6
132.7
117.0
113.3
112.7
111.7
107.7
Wastewater Treatment
19.1
20.2
19.7
19.6
19.8
20.0
19.8
Composting
0.7
3.5
3.7
3.9
4.0
4.0
4.0
Total Emissions3
6,369.2
7,326.4
6,549.4
6,735.6
6,795.6
6,677.3
6,546.2
Land Use, Land-Use Change, and
Forestry
(819.6)
(731.1)
(753.5)
(735.8)
(740.4)
(695.2)
(716.8)
Forest land
(784.3)
(730.0)
(723.3)
(733.3)
(731.7)
(709.9)
(714.2)
2-8 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Cropland
2.4
(0.7)
1.3
11.9
11.2
16.8
13.8
Grassland
13.8
25.3
0.8
18.5
14.7
33.6
21.0
Wetlands
(4.0)
(5.3)
(4.1)
(4.1)
(4.1)
(4.1)
(4.2)
Settlements
(47.6)
(20.5)
(28.3)
(28.8)
(30.5)
(31.5)
(33.3)
Net Emission (Sources and Sinks)b
5,549.6
6,595.3
5,795.9
5,999.9
6,055.2
5,982.1
5,829.3
Notes: Total emissions presented without LULUCF. Net emissions presented with LULUCF.
a Total emissions without LULUCF.
b Net emissions with LULUCF.
Notes: 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 2016. Fossil fuel combustion is the largest source of energy-related emissions, with CO2
being the primary gas emitted (see Figure 2-5). Due to their relative importance, fossil fuel combustion-related CO2
emissions are considered in detail in the Energy chapter (see Figure 2-6).
In 2016, approximately 81 percent of the energy consumed in the United States (on a Btu basis) was produced
through the combustion of fossil fuels. The remaining 19 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 in the Energy chapter. Energy-related activities are
also responsible for CH4 and N20 emissions (43 percent and 10 percent of total U.S. emissions of each gas,
respectively). Table 2-4 presents greenhouse gas emissions from the Energy chapter, by source and gas.
Figure 2-5: 2016 Energy Chapter Greenhouse Gas Sources (MMT CO2 Eq.)
COi Emissions from Fossil Fuel Combustion
Natural Gas Systems
Non-Energy Use of Fuels
Petroleum Systems
Coal Mining
Non-COs Emissions from Stationary
Combustion
Non-CO: Emissions from Mobile Combustion
Incineration of Waste
Abandoned Oil and Gas Wells
Abandoned Underground Coal Mines
0 50 100 150 200 250 300
MMT CO: Eq.
4,977
Energy as a Portion of
all Emissions
83.7%
Trends 2-9
-------
1 Figure 2-6: 2016 U.S. Fossil Carbon Flows (MMT CO2 Eq.)
2
4 Table 2-4: Emissions from Energy (MMT CO2 Eq.)
Gas/Source
1990
2005
2012
2013
2014
2015
2016
CO2
4,922.4
5,952.7
5,203.5
5,361.6
5,404.4
5,269.4
5,160.7
Fossil Fuel Combustion
4,755.8
5,759.1
5,029.8
5,162.3
5,206.1
5,059.3
4,976.7
Electric Power
1,820.8
2,400.9
2,022.2
2,038.1
2,038.0
1,900.7
1,808.8
Transportation
1,467.2
1,855.8
1,661.9
1,677.6
1,717.1
1,735.5
1,794.9
Industrial
874.5
867.8
818.4
848.7
830.8
819.3
807.6
Residential
338.3
357.8
282.5
329.7
345.3
316.8
296.2
Commercial
227.4
227.0
201.3
225.7
233.6
245.6
227.9
U.S. Territories
27.6
49.7
43.5
42.5
41.4
41.4
41.4
Non-Energy Use of Fuels
119.6
141.7
113.3
133.2
127.8
135.1
121.0
Natural Gas Systems
29.7
22.5
24.4
26.0
27.0
26.3
26.7
Petroleum Systems
9.4
17.0
25.6
29.7
32.9
38.0
25.5
Incineration of Waste
8.0
12.5
10.4
10.4
10.6
10.7
10.7
Biomass- Wood"
215.2
206.9
194.9
211.6
218.9
201.5
190.2
International Bunker Fuelsb
103.5
113.1
105.8
99.8
103.4
110.9
114.4
Biofuels-Ethanol"
4.2
22.9
72.8
74.7
76.1
78.9
81.2
Biofuels-Biodiesel"
0.0
0.9
8.5
13.5
13.3
14.1
19.6
CH4
364.7
286.7
283.1
288.7
295.4
289.5
279.2
Natural Gas Systems
193.7
160.0
156.8
159.6
164.2
164.4
162.1
Coal Mining
96.5
64.1
66.5
64.6
64.6
61.2
53.8
Petroleum Systems
42.3
34.7
35.4
38.8
41.0
39.4
39.3
Stationary Combustion
8.6
7.9
7.3
8.7
00
OO
7.8
7.2
Abandoned Oil and Gas Wells
6.5
6.9
7.0
7.0
7.1
7.2
7.1
Abandoned Underground Coal
Mines
7.2
6.6
6.2
6.2
6.3
6.4
6.7
Mobile Combustion
9.8
6.6
4.0
3.7
3.4
3.1
3.0
Incineration of Waste
+
+
+
+
+
+
+
International Bunker Fuelsb
0.2
0.1
0.1
0.1
0.1
0.1
0.1
N2O
53.1
56.4
40.9
40.9
39.4
37.1
36.5
Stationary Combustion
11.1
17.5
16.8
18.6
18.9
18.0
18.4
Mobile Combustion
41.5
38.4
23.8
22.0
20.2
18.8
17.8
Incineration of Waste
0.5
0.4
0.3
0.3
0.3
0.3
0.3
International Bunker Fuelsb
0.9
1.0
0.9
0.9
0.9
0.9
1.0
Total
5,340.2
6,295.7
5,527.6
5,691.1
5,739.1
5,596.0
5,476.4
+ Does not exceed 0.05 MMT CO2 Eq.
Coal Emissions
1,316
NEU Exports
146
Combustion
Emissions
1,307
NEU Emissions 5
Combustion
Emissions 1,477
NEU Emissions
Atmospherii
Emissions
5,292
Domestic
Fossil Fuel
Production
4,495
Apparent
Consumption
5,390
Petroleum
Emissions
2,299
Combustion
Emissions
2,193
Petroleum
1,367
- Natural Gas Liquids,
Liquefied Refinery Gas,
& Other Liquids
295
Petroleum
1,383 „
Fossil Fuel
Energy
Imports
1,796
Balancing
Item
(110)
NEU U.S.
Territories
Stock
Changes
Fossil Fuel
Energy Exports
Non-Energy Use
Carbon Sequestered
Note: Totals may not sum due to independent rounding.
The "Balancing Item" above accounts for the statistical imbalances
and unknowns in the reported data sets combined here.
NEU = Non-Energy Use
2-10 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
a 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 77 percent of GWP-weighted emissions for the entire time series since 1990. Emissions from this
source category grew by 4.6 percent (220.9 MMT CO2 Eq.) from 1990 to 2016 and were responsible for most of the
increase in national emissions during this period. Conversely, CO2 emissions from fossil fuel combustion decreased
from 2005 levels by 782.3 MMT CO2 Eq., a decrease of approximately 13.6 percent between 2005 and 2016. From
2015 to 2016, these emissions decreased by 1.6 percent (82.6 MMT CO2 Eq.). Historically, changes in emissions
from fossil fuel combustion have been the dominant factor affecting U.S. emission trends.
Changes in CO2 emissions from fossil fuel combustion are influenced by many long-term and short-term factors,
including population and economic growth, energy price fluctuations and market trends, technological changes,
energy fuel choices, and seasonal temperatures. 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 will
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 sector, which leads to a decrease in emissions from reduced fuel consumption.
Energy-related CO2 emissions also depend on the type of fuel consumed or energy used and its C intensity.
Producing a unit of heat or electricity using natural gas instead of coal, for example, can reduce the CO2 emissions
because of the lower C content of natural gas (see Table A-3 9 in Annex 2.1 for more detail on the C Content
Coefficient of different fossil fuels).
Trends in CO2 emissions from fossil fuel combustion over the past decade have been in large part driven by the
electric power sector, which historically has accounted for the majority of emissions from this source (see Figure
2-7). In recent years, the types of fuel consumed to produce electricity have changed. Carbon dioxide emissions
from coal consumption for electric power generation decreased by 36.7 percent since 2008, and there has been a
shift to the use of less-CCh-intensive natural gas to supply electricity. There has also been a rapid increase in
renewable energy capacity additions in the electric power sector in recent years. In 2016, renewable energy sources
accounted for 63 percent of capacity additions with natural gas accounting for the majority of the remaining
additions. The share of renewable energy capacity additions has grown significantly since 2010, when renewable
energy sources accounted for only 28 percent of total capacity additions (EIA 2017d). The decrease in coal-powered
electricity generation and increase in renewable energy capacity have contributed to a 4.8 percent decrease in
emissions from electric power generation from 2015 to 2016 (see Figure 2-9), and lower CO2 emissions from fossil
fuel combustion over the time series (i.e., 1990 through 2016).
Total petroleum use is another major driver of CO2 emissions from fossil fuel combustion, particularly in the
transportation sector, which represents the second largest source of CO2 emissions from fossil fuel combustion.
Emissions from petroleum consumption for transportation have increased by 22.6 percent since 1990, which can be
primarily attributed to a 46.8 percent increase in vehicle miles traveled (VMT) over the time series. 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. Since 2005, average new vehicle fuel economy has increased while the
market share of light-duty trucks has decreased. Total transportation sector CO2 emissions have increased by 5.9
percent since 2010.
The overall trends in CO2 emissions from fossil fuel combustion in the residential and commercial sectors closely
align with heating degree days. Emissions from the residential and commercial sectors decreased by 6.5 percent and
7.2 percent from 2015 to 2016, respectively. This trend can be largely attributed to a 5 percent decrease in heating
degree days which led to a decreased demand for heating fuel and electricity for heat in the residential and
commercial sectors. In addition, an increase in energy efficiency standards and the use of energy efficient products
Trends 2-11
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
in residential and commercial buildings has resulted in an overall reduction in energy use, contributing to a decrease
in emissions in both of these sectors (EIA 2017b). Combined residential and commercial sector emissions have
decreased by 6.2 percent since 2010.
The increase in transportation sector petroleum CO2 emissions from 2015 to 2016 offset emission reductions from
decreased coal use in the electric power sector and decreased demand for heating fuel in the residential and
commercial sectors. Although emissions from the transportation sector have increased, emissions from all other
sectors and U.S. Territories have decreased in recent years, contributing to a 1.6 percent decrease in total CO2
emissions from fossil fuel combustion from 2015 to 2016 and a 7.3 percent reduction since 2010.
Carbon dioxide emissions from fossil fuel combustion are presented in Table 2-5 based on the underlying U.S.
energy consumer data collected by the U.S. Energy Information Administration (EIA). 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.)
• Electric Power. EIA's fuel consumption data for the electric power sector are comprised of electricity-only
and combined-heat-and-power (CHP) plants within the North American Industry Classification System
(NAICS) 22 category whose primary business is to sell electricity, or electricity and heat, to the public.
(Non-utility power producers are included in this sector as long as they meet the electric power sector
definition.)
• Industry. EIA statistics for the industrial sector include fossil fuel consumption that occurs in the fields of
manufacturing, agriculture, mining, and construction. EIA'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.)
• Transportation. EIA'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.
• Residential. EIA's fuel consumption data for the residential sector consist of living quarters for private
households.
• Commercial. EIA'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 and Figure 2-7 summarize CO2 emissions from fossil fuel combustion by end-use sector. Figure 2-8
further describes the total emissions from fossil fuel combustion, separated by end-use sector, including CH4 and
N20 in addition to CO2.
Table 2-5: CO2 Emissions from Fossil Fuel Combustion by End-Use Sector (MMT CO2 Eq.)
End-Use Sector
1990
2005
2012
2013
2014
2015
2016
Transportation
1,470.2
1,860.5
1,665.8
1,681.6
1,721.2
1,739.2
1,798.4
Combustion
1,467.2
1,855.8
1,661.9
1,677.6
1,717.1
1,735.5
1,794.9
Electricity
3.0
4.7
3.9
4.0
4.1
3.7
3.5
Industrial
1,561.3
1,604.4
1,411.2
1,443.4
1,424.0
1,368.8
1,313.8
Combustion
874.5
867.8
818.4
848.7
830.8
819.3
807.6
Electricity
686.7
736.6
592.8
594.7
593.2
549.6
506.2
Residential
931.4
1,214.1
1,007.8
1,064.6
1,080.0
1,001.1
957.0
Combustion
338.3
357.8
282.5
329.7
345.3
316.8
296.2
Electricity
593.0
856.3
725.3
734.9
734.7
684.3
660.7
Commercial
765.3
1,030.3
901.6
930.2
939.6
908.8
866.2
Combustion
227.4
227.0
201.3
225.7
233.6
245.6
227.9
Electricity
538.0
803.3
700.3
704.5
706.0
663.1
638.3
U.S. Territories3
27.6
49.7
43.5
42.5
41.4
41.4
41.4
Total
4,755.8
5,759.1
5,029.8
5,162.3
5,206.1
5,059.3
4,976.7
2-12 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Electric Power
1,820.8
2,400.9
2,022.2 2,038.1 2,038.0 1,900.7 1,808.8
a 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.
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.
Figure 2-7: 2016 CO2 Emissions from Fossil Fuel Combustion by Sector and Fuel Type (MMT
COz Eq.)
2,500
2,000
S 1,500
0
u
1-
1 1,000
500
Relative Contribution by Fuel Type
I Petroleum
Coal
I Natural Gas
1,795
1,809
228
41
U.S. Territories Commercial
Residential
Industrial
Transportation Electric Power
Note on Figure 2-7: Fossil Fuel Combustion includes electric power, which also includes emissions of less than 0.5 MMT CO2
Eq. from geothermal-based generation.
Figure 2-8: 2016 End-Use Sector Emissions of CO2 from Fossil Fuel Combustion (MMT CO2
Eq.)
2,000
1,500-
8 1,000-
500-
I Direct Fossil Fuel Combustion
Indirect Fossil Fuel Combustion
1,798
U.S. Territories
Commercial
Residential
Industrial
Transportation
The main driver of emissions in the Energy sector is CO2 from fossil fuel combustion. Electric power is the largest
emitter of CO2, and electricity generators used 33 percent of U.S. energy from fossil fuels and emitted 36 percent of
the CO2 from fossil fuel combustion in 2016. Changes in electricity demand and the carbon intensity of fuels used
for electric power have a significant impact on CO2 emissions. Emissions from the electric power sector have
decreased by approximately 0.2 percent since 1990, and the carbon intensity of the electric power sector, in terms of
CO2 Eq. per QBtu input has significantly decreased by 12 percent during that same timeframe. This decoupling of
electric power and the resulting emissions is shown below in Figure 2-9.
Trends 2-13
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Figure 2
-9: Electric Power Generation (Billion kWh) and Emissions (MMT CO2 Eq.)
I Petroleum-based Generation (Billion kWh)
I Nuclear-based Generation (Billion kWh)
Renewable-based Generation (Billion kWh)
I Natural Gas-based Generation (Billion kWh)
Coal-based Generation (Billion kWh)
I Total Emissions (MMT CO; Eq.) [Right Axis]
3,500
3,000
Electric power emissions can also be allocated to the end-use sectors that are using that electricity, as presented in
Table 2-5. The transportation end-use sector accounted for 1,798.4 MMT CO2 Eq. in 2016 or approximately 36
percent of total CO2 emissions from fossil fuel combustion. The industrial end-use sector accounted for 26 percent
of CO2 emissions from fossil fuel combustion. The residential and commercial end-use sectors accounted for 19 and
17 percent, respectively, of CO2 emissions from fossil fuel combustion. 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 69 and 74 percent of emissions from the residential and commercial end-use
sectors, respectively. Significant trends in emissions from energy source categories over the twenty-seven-year
period from 1990 through 2016 included the following:
• Total CO2 emissions from fossil fuel combustion increased from 4,755.8 MMT CO2 Eq. in 1990 to 4,976.7
MMT CO2 Eq. in 2016—a 4.6 percent total increase over the twenty-seven-year period. From 2015 to
2016, these emissions decreased by 82.6 MMT CO2 Eq. (1.6 percent).
• Methane emissions from natural gas systems and petroleum systems (combined here) decreased from 236.0
MMT CO2 Eq. in 1990 to 201.4 MMT CO2 Eq. in 2016 (34.7 MMT CO2 Eq. or 14.7 percent decrease from
1990 to 2016). Natural gas systems CH4 emissions decreased by 31.6 MMT CO2 Eq. (16.3 percent) since
1990, largely due to a decrease in emissions from distribution, transmission and storage, processing, and
exploration. The decrease in distribution emissions is largely attributed to increased use of plastic piping,
which lias lower emissions than other pipe materials, and station upgrades at metering and regulating
(M&R) stations. The decrease in transmission and storage emissions is largely due to reduced compressor
station emissions (including emissions from compressors and leaks). Petroleum systems CH4 emissions
decreased by 3.0 MMT CO2 Eq. (or 7.2 percent) since 1990. This decrease is due primarily to decreases in
tank emissions and associated gas venting. Carbon dioxide emissions from natural gas and petroleum
systems increased by 34% from 1990 to 2016, due to increases in flaring emissions.
• Carbon dioxide emissions from non-energy uses of fossil fuels increased by 1.5 MMT CO2 Eq. (1.2
percent) from 1990 through 2016. Emissions from non-energy uses of fossil fuels were 121.0 MMT CO2
Eq. in 2016, which constituted 2.3 percent of total national CO2 emissions, approximately the same
proportion as in 1990.
2-14 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
• Nitrous oxide emissions from stationary combustion increased by 7.2 MMT CO2 Eq. (64.9 percent) from
1990 through 2016. Nitrous oxide emissions from this source increased primarily as a result of an increase
in the number of coal fluidized bed boilers in the electric power sector.
• Nitrous oxide emissions from mobile combustion decreased by 23.6 MMT CO2 Eq. (57.0 percent) from
1990 through 2016, primarily as a result of N20 national emission control standards and emission control
technologies for on-road vehicles.
• Carbon dioxide emissions from incineration of waste (10.7 MMT CO2 Eq. in 2016) increased by 2.7 MMT
CO2 Eq. (34.3 percent) from 1990 through 2016, as the volume of plastics and other fossil carbon-
containing materials in municipal solid waste grew.
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.
In many cases, greenhouse gas emissions are produced as the byproducts of many non-energy-related industrial
activities. For example, industrial processes can chemically transform raw materials, which often release waste gases
such as CO2, CH4, N2O, and fluorinated gases (e.g., HFC-23). These processes are shown in Figure 2-10. 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 employed and emitted by a number of other
industrial sources in the United States. These industries include semiconductor manufacture, electric power
transmission and distribution and magnesium metal production and processing. In addition, N20 is used in and
emitted by semiconductor manufacturing and anesthetic and aerosol applications. Table 2-6 presents greenhouse gas
emissions from industrial processes by source category.
Figure 2-10: 2016 Industrial Processes and Product Use Chapter Greenhouse Gas Sources
(MMT COz Eq.)
Substitution of Ozorie Depleting Substances
Iron and Steel Production &. Metallurgical Coke Production
Cement Production
Petrochemical Production
Lime Production
Other Process Uses of Carbonates
Ammonia Production
Nitric Acid Production
Adipic Acid Production
Semiconductor Manufacture
Carbon Dioxide Consumption
Electrical Transmission and Distribution
NsO from Product Uses
Urea Consumption for Non-Agricultural Purposes
HCFC-22 Production
Aluminum Production
Caprolactam, Glyoxal, and Glyoxylic Acid Production
Ferroalloy Production
Soda Ash Production
Titanium Dioxide Production
Glass Production
Magnesium Production and Processing
Phosphoric Acid Production
Zinc Production
Lead Production
Silicon Carbide Production and Consumption
174
¦
¦
¦
I
I
I
I
I
I
I
< 0.5
Industrial Processes and Product
Use as a Portion of all Emissions
5.7%
0
10
20
30
40
50
60
70
MMT CO, Eg.
Trends 2-15
-------
1 Table 2-6: Emissions from Industrial Processes and Product Use (MMT CO2 Eq.)
Gas/Source
1990
2005
2012
2013
2014
2015
2016
CO2
207.3
190.2
169.9
171.8
177.9
171.4
163.6
Iron and Steel Production & Metallurgical Coke
Production
101.5
68.0
55.4
53.3
58.2
47.7
42.2
Iron and Steel Production
99.0
66.0
54.9
51.5
56.2
44.9
40.9
Metallurgical Coke Production
2.5
2.0
0.5
1.8
2.0
2.8
1.3
Cement Production
33.5
46.2
35.3
36.4
39.4
39.9
39.4
Petrochemical Production
21.2
26.8
26.5
26.4
26.5
28.1
27.4
Lime Production
11.7
14.6
13.8
14.0
14.2
13.3
13.3
Other Process Uses of Carbonates
4.9
6.3
8.0
10.4
11.8
11.2
11.2
Ammonia Production
13.0
9.2
9.4
10.0
9.6
10.6
11.2
Carbon Dioxide Consumption
1.5
1.4
4.0
4.2
4.5
4.5
4.5
Urea Consumption for Non-Agricultural
Purposes
3.8
3.7
4.4
4.1
1.5
4.2
4.0
Ferroalloy Production
2.2
1.4
1.9
1.8
1.9
2.0
1.8
Soda Ash Production
1.4
1.7
1.7
1.7
1.7
1.7
1.7
Titanium Dioxide Production
1.2
1.8
1.5
1.7
1.7
1.6
1.6
Aluminum Production
6.8
4.1
3.4
3.3
2.8
2.8
1.3
Glass Production
1.5
1.9
1.2
1.3
1.3
1.3
1.3
Phosphoric Acid Production
1.5
1.3
1.1
1.1
1.0
1.0
1.0
Zinc Production
0.6
1.0
1.5
1.4
1.0
0.9
0.9
Lead Production
0.5
0.6
0.5
0.5
0.5
0.5
0.5
Silicon Carbide Production and Consumption
0.4
0.2
0.2
0.2
0.2
0.2
0.2
Magnesium Production and Processing
+
+
+
+
+
+
+
CH4
0.3
0.1
0.1
0.1
0.2
0.2
0.2
Petrochemical Production
0.2
0.1
0.1
0.1
0.1
0.2
0.2
Ferroalloy Production
+
+
+
+
+
+
+
Silicon Carbide Production and Consumption
+
+
+
+
+
+
+
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
22.4
21.0
22.8
22.3
23.7
Nitric Acid Production
12.1
11.3
10.5
10.7
10.9
11.6
10.2
Adipic Acid Production
15.2
7.1
5.5
3.9
5.4
4.3
7.0
N2O from Product Uses
4.2
4.2
4.2
4.2
4.2
4.2
4.2
Caprolactam, Glyoxal, and Glyoxylic Acid
1.7
2.1
2.0
2.0
2.0
2.0
2.0
Semiconductor Manufacturing
+
0.1
0.2
0.2
0.2
0.2
0.2
HFCs
46.6
120.0
156.0
159.1
166.8
173.3
177.1
Substitution of Ozone Depleting Substances3
0.3
99.8
150.3
154.8
161.4
168.6
173.9
HCFC-22 Production
46.1
20.0
5.5
4.1
5.0
4.3
2.8
Semiconductor Manufacturing
0.2
0.2
0.2
0.2
0.3
0.3
0.3
Magnesium Production and Processing
0.0
0.0
+
0.1
0.1
0.1
0.1
PFCs
24.3
6.7
5.9
5.8
5.6
5.1
4.3
Semiconductor Manufacturing
2.8
3.3
3.0
2.8
3.1
3.1
3.0
Aluminum Production
21.5
3.4
2.9
3.0
2.5
2.0
1.4
Substitution of Ozone Depleting Substances
0.0
+
+
+
+
+
+
SF«
28.8
11.7
6.6
6.3
6.3
5.9
6.2
Electrical Transmission and Distribution
23.1
8.3
4.6
4.5
4.6
4.2
4.3
Magnesium Production and Processing
5.2
2.7
1.6
1.5
1.0
0.9
1.0
Semiconductor Manufacturing
0.5
0.7
0.3
0.4
0.7
0.7
0.8
NF3
+
0.5
0.6
0.6
0.5
0.6
0.6
Semiconductor Manufacturing
+
0.5
0.6
0.6
0.5
0.6
0.6
Total
340.5
354.2
361.6
364.7
380.2
378.8
375.7
+ Does not exceed 0.05 MMT CO2 Eq.
2-16 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
a Small amounts of PFC emissions also result from this source.
Note: Totals may not sum due to independent rounding.
Overall, emissions from the IPPU sector increased by 10.3 percent from 1990 to 2016. Significant trends in
emissions from IPPU source categories over the twenty-seven-year period from 1990 through 2016 included the
following:
• Hydrofluorocarbon and perfluorocarbon emissions from ODS substitutes have been increasing from small
amounts in 1990 to 173.9 MMT CO2 Eq. in 2016. This increase was in large part the result of efforts to
phase out chlorofluorocarbons (CFCs) and other ODSs in the United States. In the short term, this trend is
expected to continue, and will likely continue over the next decade as hydrochlorofluorocarbons (HCFCs),
which are interim substitutes in many applications, are themselves phased-out under the provisions of the
Copenhagen Amendments to the Montreal Protocol.
• Combined CO2 and CH4 emissions from iron and steel production and metallurgical coke production
decreased by 11.5 percent to 42.2 MMT CO2 Eq. from 2015 to 2016, and have declined overall by 59.3
MMT CO2 Eq. (58.4 percent) from 1990 through 2016, due to restructuring of the industry, technological
improvements, and increased scrap steel utilization.
• Carbon dioxide emissions from ammonia production (11.2 MMT CO2 Eq. in 2016) decreased by 1.8 MMT
CO2 Eq. (13.9 percent) 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.
• Nitrous oxide emissions from adipic acid production were 7.0 MMT CO2 Eq. in 2016, and have decreased
significantly since 1990 due to both the widespread installation of pollution control measures in the late
1990s and plant idling in the late 2000s. Emissions from adipic acid production have decreased by 53.9
percent since 1990 and by 58.5 percent since a peak in 1995.
• PFC emissions from aluminum production decreased by 93.7 percent (20.1 MMT CO2 Eq.) from 1990 to
2016, 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, N20, and CO2 were the primary greenhouse gases emitted by agricultural activities.
In 2016, agricultural activities were responsible for emissions of 562.6 MMT CO2 Eq., or 8.6 percent of total U.S.
greenhouse gas emissions. Methane emissions from enteric fermentation and manure management represented
approximately 25.9 percent and 10.3 percent of total CH4 emissions from anthropogenic activities, respectively, in
2016. Agricultural soil management activities, such as application of synthetic and organic fertilizers, deposition of
livestock manure, and growing N-ftxing plants, were the largest source of U.S. N20 emissions in 2016, accounting
for 76.9 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 2-11 and
Table 2-7 illustrate agricultural greenhouse gas emissions by source.
Trends 2-17
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Figure 2-11: 2016 Agriculture Chapter Greenhouse Gas Sources (MMT CO2 Eq.)
Agricultural Soil Management
Enteric Fermentation
Manure Management
Rice Cultivation
Urea Fertilization
Liming
Field Burning of Agricultural Residues
284
¦
Agriculture as a Portion of all
Emissions
8.6%
< 0.5
0 20 40 60 80 100 120 140 160 180
MMT COi Eg.
Table 2-7: Emissions from Agriculture (MMT CO2 Eq.)
Gas/Source
1990
2005
2012
2013
2014
2015
2016
CO2
7.1
7.9
10.3
8.4
8.1
8.7
9.0
Urea Fertilization
2.4
3.5
4.3
4.4
4.5
4.9
5.1
Liming
4.7
4.3
6.0
3.9
3.6
3.8
3.9
CH4
217.6
242.1
244.0
240.6
240.1
245.4
251.8
Enteric Fermentation
164.2
168.9
166.7
165.5
164.2
166.5
170.1
Manure Management
37.2
56.3
65.6
63.3
62.9
66.3
67.7
Rice Cultivation
16.0
16.7
11.3
11.5
12.7
12.3
13.7
Field Burning of Agricultural
Residues
0.2
0.2
0.3
0.3
0.3
0.3
0.3
N2O
264.5
270.1
265.5
294.2
291.6
312.8
301.8
Agricultural Soil Management
250.5
253.5
247.9
276.6
274.0
295.0
283.6
Manure Management
14.0
16.5
17.5
17.5
17.5
17.7
18.1
Field Burning of Agricultural
Residues
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Total
489.2
520.0
519.8
543.1
539.8
566.9
562.6
Note: Totals may not sum due to independent rounding.
Some significant trends in U.S. emissions from Agriculture source categories include the following:
• Agricultural soils is the largest anthropogenic source of N20 emissions in the United States, accounting for
approximately 76.9 percent of N2O emissions in 2016. Estimated emissions from this source in 2016 were
283.6 MMT CO2 Eq. Annual N20 emissions from agricultural soils fluctuated between 1990 and 2016,
although overall emissions were 13.2 percent higher in 2016 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 2016,
enteric fermentation CH4 emissions were 170.1 MMT CO2 Eq. (25.9 percent of total CH4 emissions),
which represents an increase of 6.0 MMT CO2 Eq. (3.6 percent) since 1990. This increase in emissions
from 1990 to 2016 in enteric fermentation generally follows the increasing trends in cattle populations.
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
2-18 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
as beef cattle populations again decreased. Emissions increased from 2014 to 2016, consistent with an
increase in beef cattle population over those same years.
• Liming and urea fertilization are the only source of CO2 emissions reported in the Agriculture sector.
Estimated emissions from these sources were 3.9 and 5.1 MMT CO2 Eq., respectively. Liming and urea
fertilization emissions increased by 2.3 percent and 4.3 percent, respectively, relative to 2015, and
decreased by 17.2 percent and increased by 110.9 percent, respectively since 1990.
• Overall, emissions from manure management increased 67.8 percent between 1990 and 2016. This
encompassed an increase of 82.2 percent for CH4, from 37.2 MMT CO2 Eq. in 1990 to 67.7 MMT CO2 Eq.
in 2016; and an increase of 29.6 percent for N2O, from 14.0 MMT CO2 Eq. in 1990 to 18.1 MMT CO2 Eq.
in 2016. The majority of the increase observed in CH4 resulted from swine and dairy cattle manure, where
emissions increased 63 and 140 percent, respectively, from 1990 to 2016. From 2015 to 2016, there was a
2.2 percent increase 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.
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 CH4 and N20. Overall,
managed land is a net sink for CO2 (C sequestration) in the United States. The drivers of fluxes on managed lands
include, for example, forest management practices, tree planting in urban areas, the management of agricultural
soils, the landfilling of yard trimmings and food scraps, and activities that cause changes in C stocks in coastal
wetlands. 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 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 2016 resulted in a net increase in C stocks (i.e., net CO2 removals) of 754.9 MMT CO2 Eq.
(Table 2-8).1 This represents an offset of approximately 11.5 percent of total (i.e., gross) greenhouse gas emissions
in 2016. Emissions of CHi and N2O from LULUCF activities in 2016 were 38.1 MMT CO2 Eq. and represent 0.6
percent of total greenhouse gas emissions.2 Between 1990 and 2016, total C sequestration in the LULUCF sector
decreased by 9.1 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.
Forest fires were the largest source of CHi emissions from LULUCF in 2016, totaling 18.5 MMT CO2 Eq. (740 kt of
CH4). Coastal Wetlands Remaining Coastal Wetlands resulted in CH4 emissions of 3.6 MMT CO2 Eq. (143 kt of
CH4). Grassland fires resulted in CH4 emissions of 0.3 MMT CO2 Eq. (11 kt of CH4). Peatlands Remaining
Peatlands, Land Converted to Wetlands, and Drained Organic Soils resulted in CH4 emissions of less than 0.05
MMT CO2 Eq. each.
Forest fires were also the largest source of N20 emissions from LULUCF in 2016, totaling 12.2 MMT CO2 Eq. (41
kt of N20). Nitrous oxide emissions from fertilizer application to settlement soils in 2016 totaled to 2.5 MMT CO2
Eq. (8 kt of N20). Additionally, the application of synthetic fertilizers to forest soils in 2016 resulted in N20
emissions of 0.5 MMT CO2 Eq. (2 kt of N20). Grassland fires resulted in N20 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
0.1 MMT CO2 Eq. each (less than 0.5 kt of N2O). Peatlands Remaining Peatlands resulted U1N2O emissions of less
than 0.05 MMT C02 Eq.
1 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.
2 LULUCF emissions include the CH4 and N2O 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 N2O emissions from Forest Soils and Settlement Soils.
Trends 2-19
-------
1 Carbon dioxide removals from C stock changes are presented in Figure 2-12 and Table 2-8 along with CH4 and N20
2 emissions for LULUCF source categories.
3 Figure 2-12: 2016 LULUCF Chapter Greenhouse Gas Sources and Sinks (MMT CO2 Eq.)
Forest Land Remaining Forest Land -670.5 |
Settlements Remaining Settlements
Land Converted to Forest Land
Cropland Remaining Cropland
Wetlands Remaining Wetlands
Grassland Remaining Grassland
Land Converted to Wetlands |< 0.5|
Non-COi Emissions from Peatlands Remaining Peatlands | |< 0.5|
CHi Emissions from Land Converted to Coastal Wetlands I |< 0.5|
Non-COi Emissions from Drained Organic Soils | |< 0.5|
NiO Emissions from Forest Soils | | < 0.51
Non-COi Emissions from Grassland Fires
N:0 Emissions from Settlement Soils
Non-COz Emissions from Coastal Wetlands Remaining Coastal Wetlands
Land Converted to Grassland
Land Converted to Cropland g Carbon Stock Change
Non-COi Emissions from Forest Fires | Emissions
Land Converted to Settlements
-300 -250 -200 -150 -100 -50 0 50 100
MMT CO* Eg.
4
5 Table 2-8: U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land Use, Land-
6 Use Change, and Forestry (MMT CO2 Eq.)
Gas/Land-Use Category
1990
2005
2012
2013
2014
2015
2016
Carbon Stock Change3
(830.2)
(754.2)
(779.5)
(755.0)
(760.0)
(733.4)
(754.9)
Forest Land Remaining Forest Land
(697.7)
(664.6)
(666.9)
(670.9)
(669.3)
(666.2)
(670.5)
Land Converted to Forest Land
(92.0)
(81.6)
(74.9)
(74.9)
(75.0)
(75.0)
(75.0)
Cropland Remaining Cropland
(40.9)
(26.5)
(21.4)
(11.4)
(12.0)
(6.3)
(9.9)
Land Converted to Cropland
43.3
25.9
22.7
23.3
23.2
23.2
23.8
Grassland Remaining Grassland
(4.2)
5.5
(20.8)
(3.7)
(7.5)
9.6
(1.6)
Land Converted to Grassland
17.9
19.2
20.4
21.9
21.5
23.3
22.0
Wetlands Remaining Wetlands
(7.6)
(8.9)
(7.7)
(7.8)
(7.8)
(7.8)
(7.9)
Land Converted to Wetlands
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Settlements Remaining Settlements
(86.2)
(91.4)
(99.2)
(99.8)
(101.2)
(102.2)
(103.7)
Land Converted to Settlements
37.2
68.4
68.3
68.3
68.2
68.1
68.0
CH4
6.7
13.3
15.0
10.9
11.2
22.4
22.4
Forest Land Remaining Forest Land:
Forest Fires
3.2
9.4
10.8
7.2
7.2
18.5
18.5
Wetlands Remaining Wetlands: Coastal
Wetlands Remaining Coastal Wetlands
3.4
3.5
3.5
3.6
3.6
3.6
3.6
Grassland Remaining Grassland:
Grassland Fires
0.1
0.3
0.6
0.2
0.4
0.3
0.3
Forest Land Remaining Forest Land:
Drained Organic Soils
+
+
+
+
+
+
+
Land Converted to Wetlands: Land
Converted to Coastal Wetlands
+
+
+
+
+
+
+
Wetlands Remaining Wetlands:
Peatlands Remaining Peatlands
+
+
+
+
+
+
+
N2O
3.9
9.7
11.1
8.3
8.4
15.8
15.7
Forest Land Remaining Forest Land:
Forest Fires
2.1
6.2
7.1
4.8
4.7
12.2
12.2
Settlements Remaining Settlements:
Settlement Soilsb
1.4
2.5
2.7
2.6
2.6
2.5
2.5
2-20 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Forest Land Remaining Forest Land:
Forest Soilsc
0.1
0.5
0.5
0.5
0.5
0.5
0.5
Grassland Remaining Grassland:
Grassland Fires
0.1
0.3
0.6
0.2
0.4
0.3
0.3
Wetlands Remaining Wetlands: Coastal
Wetlands Remaining Coastal Wetlands
0.1
0.2
0.1
0.1
0.1
0.1
0.1
Forest Land Remaining Forest Land:
Drained Organic Soils
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Wetlands Remaining Wetlands:
Peatlands Remaining Peatlands
+
+
+
+
+
+
+
LULUCF Emissions'1
10.6
23.0
26.1
19.2
19.6
38.2
38.1
LULUCF Carbon Stock Change3
(830.2)
(754.2)
(779.5)
(755.0)
(760.0)
(733.4)
(754.9)
LULUCF Sector Net Total'
(819.6)
(731.1)
(753.5)
(735.8)
(740.4)
(695.2)
(716.8)
+ Absolute value does not exceed 0.05 MMT CO2 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 emissions from N fertilizer additions on both Settlements Remaining Settlements and Land Converted to
Settlements.
c Estimates include emissions from N fertilizer additions on both Forest Land Remaining Forest Land and Land Converted
to Forest Land.
d LULUCF emissions include the CH4 and N2O 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 N2O emissions from Forest Soils and Settlement Soils.
e Hie LULUCF Sector Net Lotal is the net sum of all CH4 and N2O emissions to the atmosphere plus net carbon stock
changes.
Notes: Lotals may not sum due to independent rounding. Parentheses indicate net sequestration.
Other significant trends from 1990 to 2016 in emissions from LULUCF categories include:
• Annual C sequestration by forest land (i.e., annual C stock accumulation in the five C pools and harvested
wood products for Forest Land Remaining Forest Land and Land Converted to Forest Land) lias decreased
by approximately 5.6 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, urban trees,
and landfilled yard trimmings and food scraps) has increased by 20.3 percent over the period from 1990 to
2016. This is primarily due to an increase in urbanized land area in the United States.
• Annual emissions from Land Converted to Grassland increased by approximately 23.3 percent from 1990
to 2016 due to losses in aboveground biomass, belowground biomass, dead wood, and litter C stocks from
Forest Land Converted to Grassland.
• Annual emissions from Land Converted to Settlements increased by approximately 82.6 percent from 1990
to 2016 due to losses in aboveground biomass C stocks from Forest Land Converted to Settlements and
mineral soils C stocks from Grassland Converted to Settlements.
• Nitrous oxide emissions from fertilizer application to settlement soils in 2016 totaled to 2.5 MMT CO2 Eq.
(8 kt of N2O). This represents an increase of 74.6 percent since 1990. Additionally, the application of
synthetic fertilizers to forest soils in 2016 resulted in N20 emissions of 0.5 MMT CO2 Eq. (2 kt of N20).
Nitrous oxide emissions from fertilizer application to forest soils have increased by 455 percent since 1990,
but still account for a relatively small portion of overall emissions.
Waste
Waste management and treatment activities are sources of greenhouse gas emissions (see Figure 2-13). In 2016,
landfills were the third-largest source of U.S. anthropogenic CH4 emissions, accounting for 16.4 percent of total
Trends 2-21
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
U.S. CH4 emissions.3 Additionally, wastewater treatment accounts for 15.1 percent of Waste emissions, 2.3 percent
of U.S. CH4 emissions, and 1.3 percent of N20 emissions. Emissions of CH4 and N20 from composting grew from
1990 to 2016, and resulted in emissions of 4.0 MMT CO2 Eq. in 2016. A summary of greenhouse gas emissions
from the Waste chapter is presented in Table 2-9.
Figure 2-13: 2016 Waste Chapter Greenhouse Gas Sources (MMT CO2 Eq.)
Landfills
Wastewater
Treatment
Composting
108
Waste as a Portion of all Emissions
2.0%
20 40 60 80
MMT COi Eq.
100 120
Overall, in 2016, waste activities generated emissions of 131.5 MMT CO2 Eq., or 2.0 percent of total U.S.
greenhouse gas emissions.
Table 2-9: Emissions from Waste (MMT CO2 Eq.)
Gas/Source
1990
2005
2012
2013
2014
2015
2016
CH4
195.6
150.4
134.0
130.2
129.8
128.9
124.6
Landfills
179.6
132.7
117.0
113.3
112.7
111.7
107.7
Wastewater Treatment
15.7
15.8
15.1
14.9
15.0
15.1
14.8
Composting
0.4
1.9
1.9
2.0
2.1
2.1
2.1
N2O
3.7
6.1
6.4
6.5
6.7
6.7
6.8
Wastewater Treatment
3.4
4.4
4.6
4.7
4.8
4.8
5.0
Composting
0.3
1.7
1.7
1.8
1.9
1.9
1.9
Total
199.3
156.4
140.4
136.7
136.5
135.6
131.5
Note: Totals may not sum due to independent rounding.
Some significant trends in U.S. emissions from waste source categories include the following:
• From 1990 to 2016, net CHi emissions from landfills decreased by 71.9 MMT CO2 Eq. (40.0 percent), with
small increases occurring in interim years. This downward trend in emissions coincided with increased
landfill gas collection and control systems, and a reduction of decomposable materials (i.e., paper and
paperboard, food scraps, and yard trimmings) discarded in MSW landfills over the time series.
• Combined CHi and N2O emissions from composting have generally increased since 1990, from 0.7 MMT
CO2 Eq. to 4.0 MMT CO2 Eq. in 2016, which represents slightly less than a five-fold increase over the time
series. The growth in composting since the 1990s is attributable to primarily two factors: (1) steady growth
in population and residential housing, and (2) the enactment of legislation by state and local governments
that discouraged the disposal of yard trimmings in landfills.
• From 1990 to 2016, CH4 and N20 emissions from wastewater treatment decreased by 0.9 MMT CO2 Eq.
(5.5 percent) and increased by 1.6 MMT CO2 Eq. (46.5 percent), respectively. Methane emissions from
3 Landfills also store carbon, due to incomplete degradation of organic materials such as wood products and yard trimmings, as
described in the Land Use, Land-Use Change, and Forestry chapter.
2-22 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
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.
5 2.2 Emissions by Economic Sector
6 Throughout this report, emission estimates are grouped into five sectors (i.e., chapters) defined by the IPCC and
7 detailed above: Energy; Industrial Processes and Product Use; Agriculture; LULUCF; and Waste. While it is
8 important to use this characterization for consistency with UNFCCC reporting guidelines and to promote
9 comparability across countries, it is also useful to characterize emissions according to commonly used economic
10 sector categories: residential, commercial, industry, transportation, electric power, and agriculture, as well as U.S.
11 Territories.
12 Using this categorization, transportation activities, in aggregate, accounted for the largest portion (28.5 percent) of
13 total U.S. greenhouse gas emissions in 2016. Emissions from electric power, in aggregate, accounted for the second
14 largest portion (28.2 percent). Emissions from industry accounted for about 22 percent of total U.S. greenhouse gas
15 emissions in 2016. Emissions from industry have in general declined over the past decade due to a number of
16 factors, including structural changes in the U.S. economy (i.e., shifts from a manufacturing-based to a service-based
17 economy), fuel switching, and efficiency improvements.
18 The remaining 22 percent of U.S. greenhouse gas emissions were contributed by the residential, agriculture, and
19 commercial sectors, plus emissions from U.S. Territories. The residential sector accounted for 5 percent, and
20 primarily consisted of CO2 emissions from fossil fuel combustion. Activities related to agriculture accounted for
21 roughly 9 percent of U.S. emissions; unlike other economic sectors, agricultural sector emissions were dominated by
22 N20 emissions from agricultural soil management and CH4 emissions from enteric fermentation, rather than CO2
23 from fossil fuel combustion. The commercial sector accounted for roughly 6 percent of emissions, while U.S.
24 Territories accounted for less than 1 percent. Carbon dioxide was also emitted and sequestered (in the form of C) by
25 a variety of activities related to forest management practices, tree planting in urban areas, the management of
26 agricultural soils, landfilling of yard trimmings, and changes in C stocks in coastal wetlands.
27 Table 2-10 presents a detailed breakdown of emissions from each of these economic sectors by source category, as
28 they are defined in this report. Figure 2-14 shows the trend in emissions by sector from 1990 to 2016.
Trends 2-23
-------
1 Figure 2-14: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO2 Eq.)
2,500-
Electric Power Industry
2,000-
Transportation
1,500-
Industry
1,000-
Agriculture
Commercial (Red)
500-
Residential (Blue)
o^rMro5,ir>^orvcoo>0'-Jm^u->v£>r^coo%o--JC>^-mio
o!flISSo\oIoIoIoIo!o 00000 00 000000000
3
4
5 Table 2-10: U.S. Greenhouse Gas Emissions Allocated to Economic Sectors (MMT CO2 Eq. and
6 Percent of Total in 2016)
Sector/Source
1990
2005
2012
2013
2014
2015
2016a
Percent3
Transportation
1,525.1
1,972.5
1,748.9
1,757.8
1,793.1
1,808.8
1,863.8
28.5%
CO2 from Fossil Fuel Combustion
1,467.2
1,855.8
1,661.9
1,677.6
1,717.1
1,735.5
1,794.9
27.4%
Substitution of Ozone Depleting
Substances
+
67.1
55.1
49.8
47.2
45.1
42.4
0.6%
Mobile Combustion
46.1
39.5
23.6
21.6
19.6
18.2
17.1
0.3%
Non-Energy Use of Fuels
11.8
10.2
8.3
8.8
9.1
10.0
9.5
0.1%
Electric Power Industry
1,861.7
2,439.3
2,055.7
2,074.1
2,075.5
1,936.9
1,845.7
28.2%
CO2 from Fossil Fuel Combustion
1,820.8
2,400.9
2,022.2
2,038.1
2,038.0
1,900.7
1,808.8
27.6%
Stationary Combustion
6.9
14.0
14.2
15.6
16.0
15.4
16.0
0.2%
Incineration of Waste
8.4
12.9
10.7
10.7
10.9
11.0
11.0
0.2%
Other Process Uses of Carbonates
2.5
3.2
4.0
5.2
5.9
5.6
5.6
0.1%
Electrical Transmission and
Distribution
23.1
8.3
4.6
4.5
4.6
4.2
4.3
0.1%
Industry
1,654.6
1,515.1
1,418.2
1,477.6
1,474.2
1,466.7
1,412.4
21.6%
CO2 from Fossil Fuel Combustion
843.1
820.5
767.3
798.6
780.0
771.8
759.3
11.6%
Natural Gas Systems
223.4
182.5
181.2
185.6
191.2
190.8
188.8
2.9%
Non-Energy Use of Fuels
102.1
123.4
100.2
119.0
113.5
120.0
106.5
1.6%
Petroleum Systems
51.7
51.7
61.0
68.5
73.9
77.4
64.8
1.0%
Coal Mining
96.5
64.1
66.5
64.6
64.6
61.2
53.8
0.8%
Iron and Steel Production
101.5
68.1
55.5
53.4
58.2
47.7
42.2
0.6%
Cement Production
33.5
46.2
35.3
36.4
39.4
39.9
39.4
0.6%
Petrochemical Production
21.4
26.9
26.6
26.5
26.6
28.2
27.6
0.4%
Substitution of Ozone Depleting
Substances
+
7.4
18.8
20.4
22.3
24.7
26.9
0.4%
Lime Production
11.7
14.6
13.8
14.0
14.2
13.3
13.3
0.2%
Ammonia Production
13.0
9.2
9.4
10.0
9.6
10.6
11.2
0.2%
Nitric Acid Production
12.1
11.3
10.5
10.7
10.9
11.6
10.2
0.2%
Abandoned Oil and Gas Wells
6.5
6.9
7.0
7.0
7.1
7.2
7.1
0.1%
2-24 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
Adipic Acid Production
15.2
7.1
5.5
3.9
5.4
4.3
7.0
0.1%
Abandoned Underground Coal
Mines
7.2
6.6
6.2
6.2
6.3
6.4
6.7
0.1%
Other Process Uses of Carbonates
2.5
3.2
4.0
5.2
5.9
5.6
5.6
0.1%
Semiconductor Manufacture
3.6
4.7
4.4
4.0
4.9
5.0
5.0
0.1%
Carbon Dioxide Consumption
1.5
1.4
4.0
4.2
4.5
4.5
4.5
0.1%
N2O from Product Uses
4.2
4.2
4.2
4.2
4.2
4.2
4.2
0.1%
Urea Consumption for Non-
Agricultural Purposes
3.8
3.7
4.4
4.1
1.5
4.2
4.0
0.1%
Stationary Combustion
5.0
4.8
4.0
4.1
4.0
3.9
3.7
0.1%
Mobile Combustion
4.4
4.5
3.3
3.3
3.1
3.0
3.0
+%
HCFC-22 Production
46.1
20.0
5.5
4.1
5.0
4.3
2.8
+%
Aluminum Production
28.3
7.6
6.4
6.2
5.4
4.8
2.7
+%
Caprolactam, Glyoxal, and
Glyoxylic Acid Production
1.7
2.1
2.0
2.0
2.0
2.0
2.0
+%
Ferroalloy Production
2.2
1.4
1.9
1.8
1.9
2.0
1.8
+%
Soda Ash Production
1.4
1.7
1.7
1.7
1.7
1.7
1.7
+%
Titanium Dioxide Production
1.2
1.8
1.5
1.7
1.7
1.6
1.6
+%
Glass Production
1.5
1.9
1.2
1.3
1.3
1.3
1.3
+%
Magnesium Production and
Processing
5.2
2.7
1.7
1.5
1.1
1.0
1.1
+%
Phosphoric Acid Production
1.5
1.3
1.1
1.1
1.0
1.0
1.0
+%
Zinc Production
0.6
1.0
1.5
1.4
1.0
0.9
0.9
+%
Tead Production
0.5
0.6
0.5
0.5
0.5
0.5
0.5
+%
Silicon Carbide Production and
Consumption
0.4
0.2
0.2
0.2
0.2
0.2
0.2
+%
Agriculture
521.5
568.5
571.8
594.0
591.5
615.1
611.7
9.3%
N2O from Agricultural Soil
Management
250.5
253.5
247.9
276.6
274.0
295.0
283.6
4.3%
Enteric Fermentation
164.2
168.9
166.7
165.5
164.2
166.5
170.1
2.6%
Manure Management
51.1
72.9
83.2
80.8
80.4
84.0
85.9
1.3%
CO2 from Fossil Fuel Combustion
31.4
47.4
51.1
50.0
50.8
47.5
48.3
0.7%
Rice Cultivation
16.0
16.7
11.3
11.5
12.7
12.3
13.7
0.2%
Urea Fertilization
2.4
3.5
4.3
4.4
4.5
4.9
5.1
0.1%
Timing
4.7
4.3
6.0
3.9
3.6
3.8
3.9
0.1%
Mobile Combustion
0.8
1.0
0.9
0.8
0.8
0.7
0.7
+%
Field Burning of Agricultural
Residues
0.3
0.3
0.4
0.4
0.4
0.4
0.4
+%
Stationary Combustion
+
+
+
0.1
0.1
0.1
0.1
+%
Commercial
428.2
402.6
388.0
411.3
420.8
433.2
411.7
6.3%
CO2 from Fossil Fuel Combustion
227.4
227.0
201.3
225.7
233.6
245.6
227.9
3.5%
Tandfills
179.6
132.7
117.0
113.3
112.7
111.7
107.7
1.6%
Substitution of Ozone Depleting
Substances
+
17.6
45.1
47.5
49.3
50.3
50.9
0.8%
Wastewater Treatment
15.7
15.8
15.1
14.9
15.0
15.1
14.8
0.2%
Human Sewage
3.4
4.4
4.6
4.7
4.8
4.8
5.0
0.1%
Composting
0.7
3.5
3.7
3.9
4.0
4.0
4.0
0.1%
Stationary Combustion
1.5
1.4
1.2
1.4
1.4
1.6
1.5
+%
Residential
344.9
370.4
318.4
372.7
393.9
370.0
354.1
5.4%
CO2 from Fossil Fuel Combustion
338.3
357.8
282.5
329.7
345.3
316.8
296.2
4.5%
Substitution of Ozone Depleting
Substances
0.3
7.7
31.4
37.0
42.6
48.4
53.8
0.8%
Stationary Combustion
6.3
4.9
4.5
5.9
6.1
4.7
4.1
0.1%
U.S. Territories
33.3
58.1
48.5
48.1
46.6
46.6
46.6
0.7%
CO2 from Fossil Fuel Combustion
27.6
49.7
43.5
42.5
41.4
41.4
41.4
0.6%
Non-Energy Use of Fuels
5.7
8.1
4.8
5.4
5.1
5.1
5.1
0.1%
Stationary Combustion
0.1
0.2
0.2
0.2
0.2
0.2
0.2
+%
Total Emissions
6,369.2
7,326.4
6,549.4
6,735.6
6,795.6
6,677.3
6,546.2
100.0%
Trends 2-25
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
LULUCF Sector Net Total"
(819.6)
(731.1)
(753.5)
(735.8)
(740.4)
(695.2)
(716.8)
(11.0%)
Net Emissions (Sources and
Sinks)
5,549.6
6,595.3
5,795.9
5,999.9
6,055.2
5,982.1
5,829.3
89.0%
Notes: Total emissions presented without LULUCF. Total net emissions presented with LULUCF.
+ Does not exceed 0.05 MMT CO2 Eq. or 0.05 percent.
3 Percent of total (gross) emissions excluding emissions from LULUCF for 2016.
b The LULUCF Sector Net Total is the net sum of all CH4 and N2O emissions to the atmosphere plus net carbon stock
changes.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or sequestration.
Emissions with Electricity Distributed to Economic Sectors
It can also be 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, which is the second
largest economic sector in the United States, accounted for 28 percent of total U.S. greenhouse gas emissions in
2016. Electric power-related emissions decreased by 1 percent since 1990 and by 4.7 percent from 2015 to 2016,
primarily due to decreased CO2 emissions from fossil fuel combustion due to increased natural gas consumption,
decreased coal consumption.
Overall, between 2015 and 2016, the amount of electricity generated (in kWh) increased by less than 0.1 percent.
However, total emissions from the electric power sector decreased by 4.7 percent from 2015 to 2016 due to changes
in the consumption of coal and natural gas for electric power, which were driven by changes in their relative prices.
Coal consumption decreased by 8.1 percent, while natural gas consumption increased by 3.8 percent. The
consumption of petroleum for electric power decreased by 12.9 percent in 2016 relative to 2015.
Electricity sales to the residential and commercial end-use sectors increased by 0.2 percent and decreased by 0.1
percent, respectively, from 2015 to 2016. The sales trend in the residential sector can largely be attributed to an
increase in the number of households in the United States. The sales trend in the commercial sector can largely be
attributed to warmer, less energy-intensive winter conditions compared to 2015. Electricity sales to the industrial
sector from 2015 to 2016 decreased by approximately 5.1 percent.
Table 2-11 provides a detailed summary of emissions from electric power-related activities.
Table 2-11: Electric Power-Related Greenhouse Gas Emissions (MMT CO2 Eq.)
Gas/Fuel Type or Source
1990
2005
2012
2013
2014
2015
2016
CO2
1,831.2
2,416.5
2,036.6
2,053.7
2,054.5
1,917.0
1,825.1
Fossil Fuel Combustion
1,820.8
2,400.9
2,022.2
2,038.1
2,038.0
1,900.7
1,808.8
Coal
1,547.6
1,983.8
1,511.2
1,571.3
1,569.1
1,350.5
1,241.3
Natural Gas
175.3
318.8
492.2
444.0
443.2
526.1
545.9
Petroleum
97.5
97.9
18.3
22.4
25.3
23.7
21.2
Geothermal
0.4
0.4
0.4
0.4
0.4
0.4
0.4
Incineration of Waste
8.0
12.5
10.4
10.4
10.6
10.7
10.7
Other Process Uses of
Carbonates
2.5
3.2
4.0
5.2
5.9
5.6
5.6
CH4
0.4
0.9
1.1
1.0
1.0
1.1
1.1
Stationary Sources3
0.4
0.9
1.1
1.0
1.0
1.1
1.1
Incineration of Waste
+
+
+
+
+
+
+
N2O
6.9
13.6
13.4
14.9
15.3
14.6
15.2
Stationary Sources3
6.5
13.2
13.1
14.6
15.0
14.3
14.9
Incineration of Waste
0.5
0.4
0.3
0.3
0.3
0.3
0.3
SF«
23.1
8.3
4.6
4.5
4.6
4.2
4.3
Electrical Transmission and
Distribution
23.1
8.3
4.6
4.5
4.6
4.2
4.3
Total
1,861.7
2,439.3
2,055.7
2,074.1
2,075.5
1,936.9
1,845.7
2-26 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
+ Does not exceed 0.05 MMT CO2 Eq.
a 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 2017a; Duffield 2006).
These source categories include CO2 from Fossil Fuel Combustion, CH4 and N20 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.4
When emissions from electricity use are distributed among these sectors, industrial activities account for the largest
share of total U.S. greenhouse gas emissions (28.9 percent), followed closely by emissions from transportation (28.5
percent). Emissions from the residential and commercial sectors also increase substantially when emissions from
electricity are included. In all sectors except agriculture, CO2 accounts for more than 81 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 2-15 shows the trend in these emissions by sector from 1990 to 2016.
Figure 2-15: U.S. Greenhouse Gas Emissions with Electricity-Related Emissions Distributed
to Economic Sectors (MMT CO2 Eq.)
2,500
Industry
2,000
Transportation
iff
.. 1,500
o
u
Commercial (Red)
1,000
Residential (Blue)
Agriculture
500
Q) Q) 0"» ffi Ov O"* 01 m O
rN
in
=
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 2016
Sector/Gas
1990
2005
2012
2013
2014
2015
2016
Percent3
Industry
2,321.9
2,225.4
1,977.8
2,040.5
2,033.7
1,985.1
1,888.8
28.9%
Direct Emissions
CO2
1,654.6
1,189.1
1,515.1
1,170.4
1,418.2
1,083.5
1,477.6
1,139.9
1,474.2
1,125.4
1,466.7
1,122.0
1,412.4
1,076.1
21.6%
16.4%
4 Emissions were not distributed to U.S. Territories, since the electric power sector only includes emissions related to the
generation of electricity in the 50 states and the District of Columbia.
Trends 2-27
-------
ch4
351.9
277.2
275.2
279.6
286.3
281.6
271.9
4.2%
n2o
37.3
29.3
26.5
25.2
27.0
26.4
27.7
0.4%
HFCs, PFCs, SFs.andNFs
76.3
38.2
33.0
32.9
35.5
36.7
36.9
0.6%
Electricity-Related
667.3
710.3
559.6
562.9
559.5
518.4
476.4
7.3%
CO2
656.4
703.7
554.4
557.4
553.9
513.1
471.1
7.2%
CH4
0.2
0.3
0.3
0.3
0.3
0.3
0.3
+%
N2O
2.5
3.9
3.7
4.1
4.1
3.9
3.9
0.1%
SFe
8.3
2.4
1.3
1.2
1.2
1.1
1.1
+%
Transportation
1,528.2
; 1,977.3
1,752.8
1,761.9
1,797.2
1,812.5
1,867.4
28.5%
Direct Emissions
1,525.1
1,972.5
1,748.9
1,757.8
1,793.1
1,808.8
1,863.8
28.5%
CO2
1,479.0
1,865.9
1,670.2
1,686.4
1,726.3
1,745.5
1,804.3
27.6%
CH4
5.8
3.1
2.0
1.9
1.7
1.7
1.6
+%
N2O
40.2
36.5
21.6
19.7
17.8
16.5
15.5
0.2%
HFCsb
+
67.1
55.1
49.8
47.2
45.1
42.4
0.6%
Electricity-Related
3.1
4.8
3.9
4.1
4.1
3.8
3.6
0.1%
CO2
3.1
4.8
3.9
4.0
4.1
3.8
3.6
0.1%
CH4
+
+
+
+
+
+
+
+%
N2O
+
+
+
+
+
+
+
+%
SFe
+
+
+
+
+
+
+
+%
Commercial
978.2
1,218.7
1,099.9
1,128.2
1,139.8
1,108.9
1,063.1
16.2%
Direct Emissions
428.2
402.6
388.0
411.3
420.8
433.2
411.7
6.3%
CO2
227.4
227.0
201.3
225.7
233.6
245.6
227.9
3.5%
CH4
196.7
151.5
135.0
131.3
130.9
130.1
125.8
1.9%
N2O
4.1
6.4
6.6
6.8
7.0
7.1
7.2
0.1%
HFCs
+
17.6
45.1
47.5
49.3
50.3
50.9
0.8%
Electricity-Related
550.1
816.1
711.9
716.9
719.0
675.8
651.3
10.0%
CO2
541.1
808.5
705.3
709.9
711.7
668.8
644.1
9.8%
CH4
0.1
0.3
0.4
0.4
0.4
0.4
0.4
+%
N2O
2.0
4.5
4.6
5.2
5.3
5.1
5.4
0.1%
SFe
6.8
2.8
1.6
1.5
1.6
1.5
1.5
+%
Residential
951.2
1,240.4
1,055.7
1,120.6
1,142.2
1,067.3
1,028.4
15.7%
Direct Emissions
344.9
370.4
318.4
372.7
393.9
370.0
354.1
5.4%
CO2
338.3
357.8
282.5
329.7
345.3
316.8
296.2
4.5%
CH4
5.2
4.1
3.7
5.0
5.1
3.9
3.4
0.1%
N2O
1.0
0.9
0.7
1.0
1.0
0.8
0.7
+%
HFCs
0.3
7.7
31.4
37.0
42.6
48.4
53.8
0.8%
Electricity-Related
606.3
870.0
737.3
747.9
748.2
697.3
674.2
10.3%
CO2
596.4
861.9
730.5
740.5
740.7
690.1
666.7
10.2%
CH4
0.1
0.3
0.4
0.4
0.4
0.4
0.4
+%
N2O
2.3
4.8
4.8
5.4
5.5
5.2
5.5
0.1%
SFe
7.5
3.0
1.7
1.6
1.7
1.5
1.6
+%
Agriculture
556.3
606.5
614.8
636.4
636.1
656.8
651.9
10.0%
Direct Emissions
521.5
568.5
571.8
594.0
591.5
615.1
611.7
9.3%
CO2
38.5
55.2
61.3
58.4
58.9
56.1
57.3
0.9%
CH4
218.0
242.6
244.3
240.8
240.3
245.5
251.9
3.8%
N2O
264.9
270.7
266.2
294.9
292.3
313.5
302.5
4.6%
Electricity-Related
34.8
38.0
43.0
42.3
44.6
41.6
40.2
0.6%
CO2
34.2
37.7
42.6
41.9
44.1
41.2
39.7
0.6%
CH4
+
+
+
+
+
+
+
+%
N2O
0.1
0.2
0.3
0.3
0.3
0.3
0.3
+%
SFe
0.4
0.1
0.1
0.1
0.1
0.1
0.1
+%
U.S. Territories
33.3
58.1
48.5
48.1
46.6
46.6
46.6
0.7%
Total Emissions
6,369.2
7,326.4
6,549.4
6,735.6
6,795.6
6,677.3
6,546.2
100.0%
LULUCF Sector Net Totalc
(819.6)
(731.1)
(753.5)
(735.8)
(740.4)
(695.2)
(716.8)
(11.0%)
Net Emissions (Sources and
Sinks)
5,549.6
6,595.3
5,795.9
5,999.9
6,055.2
5,982.1
5,829.3
89.0%
Notes: Total emissions presented without LULUCF. Net emissions presented with LULUCF.
+ Does not exceed 0.05 MMT CO2 Eq. or 0.05 percent.
a Percent of total (gross) emissions excluding emissions from LULUCF for year 2016.
2-28 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
b Includes primarily HFC-134a.
c The LULUCF Sector Net Total is the net sum of all CH4 and N2O 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.
Industry
The industry end-use sector includes CO2 emissions from fossil fuel combustion from all manufacturing facilities, in
aggregate. 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 CH4
emissions from petroleum and natural gas systems, fugitive CH4 emissions from coal mining, byproduct CO2
emissions from cement manufacture, and HFC, PFC, SF6, and NF3 byproduct emissions from semiconductor
manufacture, to name a few.
Since 1990, industrial sector emissions have declined. 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.
Transportation
When electricity-related emissions are distributed to economic end-use sectors, transportation activities accounted
for 28.5 percent of U.S. greenhouse gas emissions in 2016. The largest sources of transportation greenhouse gases in
2016 were passenger cars (41.8 percent), freight trucks (22.7 percent), light-duty trucks, which include sport utility
vehicles, pickup trucks, and minivans (17.7 percent), commercial aircraft (6.4 percent), other aircraft (2.7 percent),
ships and boats (2.4 percent), rail (2.2 percent), and pipelines (2.1 percent). These figures include direct CO2, CH4,
and N20 emissions from fossil fuel combustion used in transportation 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.
In terms of the overall trend, from 1990 to 2016, total transportation emissions increased due, in large part, to
increased demand for travel. The number of vehicle miles traveled (VMT) by light-duty motor vehicles (passenger
cars and light-duty trucks) increased 43 percent from 1990 to 2016,5 as a result of a confluence of factors including
population growth, economic growth, urban sprawl, and periods of low fuel prices.
The decline in new light-duty vehicle fuel economy between 1990 and 2004 reflected the increasing market share of
light-duty trucks, which grew from 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. Light-duty VMT grew by less than one percent or declined each year between 2005 and 20146 and has
since grown at a faster rate (1.2 percent from 2014 to 2015, and 2.6 percent from 2015 to 2016). 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 percent and 43 percent. Light-duty truck share
is about 43 percent of new vehicles in model year 2016 (EPA 2016a).
5 VMT estimates are based on data from FHWA Highway Statistics Table VM-1 (FHWA 1996 through 2017). Table VM-1 data
for 2016 has not been published yet, therefore 2016 mileage data is estimated using the 1.7 percent increase in FHWA Traffic
Volume Trends from 2015 to 2016. 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 2016 time period. In absence of these method changes,
light-duty VMT growth between 1990 and 2016 would likely have been even higher.
6 In 2007 and 2008 light-duty VMT decreased 3 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 absence of these method
changes, light-duty VMT growth between 2006 and 2007 would likely have been higher. See previous footnote.
Trends 2-29
-------
1 Table 2-13 provides a detailed summary of greenhouse gas emissions from transportation-related activities with
2 electricity-related emissions included in the totals.
3 Almost all of the energy consumed for transportation was supplied by petroleum-based products, with more than
4 half being related to gasoline consumption in automobiles and other highway vehicles. Other fuel uses, especially
5 diesel fuel for freight trucks and jet fuel for aircraft, accounted for the remainder. The primary driver of
6 transportation-related emissions was CO2 from fossil fuel combustion, which increased by 22 percent from 1990 to
7 2016.7 This rise in CO2 emissions, combined with an increase in HFCs from close to zero emissions in 1990 to 42.4
8 MMT CO2 Eq. in 2016, led to an increase in overall emissions from transportation activities of 22 percent.8
9 Table 2-13: Transportation-Related Greenhouse Gas Emissions (MMT CO2 Eq.)
Gas/Vehicle
1990
2005
2012
2013
2014
2015
2016
Passenger Cars
639.6
693.1
745.9
740.9
756.8
761.1
781.3
CO2
612.3
642.6
711.3
710.9
729.6
735.8
758.4
CH4
3.2
1.3
0.9
0.8
0.7
0.6
0.6
N2O
24.1
17.6
13.1
11.8
10.5
9.7
8.9
HFCs
0.0
31.7
20.6
17.3
16.0
14.9
13.4
Light-Duty Trucks
326.8
539.7
316.2
313.2
334.3
324.9
331.4
CO2
312.3
490.6
281.3
281.5
304.7
297.5
306.3
CH4
1.7
0.8
0.3
0.3
0.3
0.2
0.2
N2O
12.8
15.0
5.3
4.7
4.4
3.8
3.4
HFCs
0.0
33.3
29.3
26.7
25.0
23.4
21.5
Medium- and Heavy-Duty
Trucks
230.3
397.8
387.3
394.3
405.6
413.9
424.0
CO2
229.3
395.4
383.6
390.3
401.5
409.5
419.5
CH4
0.3
0.1
0.1
0.1
0.1
0.1
0.1
N2O
0.7
1.2
1.0
0.9
0.9
0.8
0.8
HFCs
0.0
1.1
2.6
2.9
3.2
3.4
3.7
Buses
8.5
12.2
18.0
18.2
19.5
20.0
20.5
CO2
8.4
11.6
17.2
17.5
18.8
19.3
19.8
CH4
+
0.2
0.3
0.2
0.2
0.2
0.2
N2O
+
+
+
+
+
+
+
HFCs
0.0
0.3
0.4
0.4
0.4
0.4
0.4
Motorcycles
1.7
1.6
4.0
3.8
3.8
3.7
3.8
CO2"
1.7
1.6
3.9
3.7
3.7
3.7
3.8
CH4
+
+
+
+
+
+
+
N2O
+
+
+
+
+
+
+
Commercial Aircraft3
110.9
134.0
114.3
115.4
116.3
120.1
120.1
CO2
109.9
132.7
113.3
114.3
115.2
119.0
119.0
CH4
0.0
0.0
0.0
0.0
0.0
0.0
0.0
N2O
1.0
1.2
1.0
1.1
1.1
1.1
1.1
Other Aircraftb
78.3
59.7
32.1
34.7
35.0
40.4
51.1
CO2
77.5
59.1
31.8
34.4
34.7
40.0
50.6
CH4
0.1
0.1
+
+
+
+
+
N2O
0.7
0.5
0.3
0.3
0.3
0.4
0.5
Ships and Boats0
45.3
45.8
41.9
41.5
31.0
35.7
44.9
CO2
44.3
44.3
39.3
38.6
28.0
32.3
41.1
CH4
0.5
0.5
0.4
0.4
0.3
0.3
0.3
N2O
0.6
0.6
0.5
0.5
0.3
0.4
0.5
HFCs
0.0
0.5
1.7
2.0
2.3
2.6
2.9
Rail
38.9
51.0
44.1
45.0
46.4
44.3
41.1
CO2
38.5
50.3
43.4
44.2
45.6
43.5
40.3
7 See previous footnote.
8 See previous footnote.
2-30 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
ch4
O.I
0.1
0.1
0.1
0.1
0.1
0.1
N20
0.3
0.4
0.3
0.3
0.3
0.3
0.3
HFCs
0.0
0.2
0.3
0.3
0.3
0.4
0.4
Other Emissions from
Electric Powerd
O.I
+
+
+
+
+
+
Pipelines®
36.0
32.4
40.5
46.2
39.4
38.5
39.6
CO2
36.0
32.4
40.5
46.2
39.4
38.5
39.6
Lubricants
ll.N
10.2
8.3
8.8
9.1
10.0
9.5
CO2
II.8
10.2
8.3
OO
OO
9.1
10.0
9.5
Total Transportation
1,52N.2
1,977.3
1,752.8
1,761.9
1,797.2
1,812.5
1,867.4
International Bunker Fuel/
104.5
114.2
106.8
100.7
104.4
111.9
115.5
Ethanol CO2s
4.1
22.4
71.5
73.4
74.9
75.9
78.2
Biodiesel CO2s
0.0
0.9
8.5
13.5
13.3
14.1
19.6
+ Does not exceed 0.05 MMT CO2 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 CO2 estimates reflect natural gas used to power pipelines, but not electricity. While the operation of pipelines
produces CH4 and N2O, 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.
B Ethanol and biodiesel CO2 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.
Commercial
The commercial sector 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 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, a reduction in heating degree days, and increases in energy efficiency.
Landfills and wastewater treatment are included in the commercial sector, with landfill emissions decreasing since
1990 and wastewater treatment emissions decreasing slightly.
Residential
The residential sector is heavily reliant on electricity for meeting energy needs, with electricity consumption 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 are often correlated with short-term fluctuations in
energy consumption 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 2017b).
Trends 2-31
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Agriculture
The agriculture end-use sector includes a variety of processes, including enteric fermentation in domestic livestock,
livestock manure management, and agricultural soil management. In 2016, agricultural soil management was the
largest source of N20 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 like tractors.
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 emissions from the combustion of fossil fuels that are included in
the EIA electric utility fuel-consuming sector. Stationary combustion emissions of CH4 and N20 are also based on
the EIA electric power sector. Additional sources include CO2, CHi and N20 from waste incineration, as the
majority of municipal solid waste is combusted in "trash-to-steam" electric power plants. 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 N20 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 N20
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.
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 included in
supplementary sources of agriculture fuel use, because EIA does not include an agriculture fuel-consuming sector.
Agriculture equipment is included in the EIA industrial fuel-consuming sector. Agriculture fuel use estimates are
obtained from U.S. Department of Agriculture survey data, in combination with separate EIA fuel sales reports
(USDA 2016; EIA 2017c). 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 CH4 and N20 emissions from stationary
and mobile combustion are then apportioned to the Agriculture economic sector based on agricultural fuel use.
The other emission sources included in the Agriculture economic sector are intuitive for the agriculture sectors, such
as N2O emissions from Agricultural Soils, CH4 from Enteric Fermentation, CH4 and N20 from Manure
2-32 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Management, CH4 from Rice Cultivation. CO2 emissions from Liming and Urea Application, and CH4 and N20
from Field Burning of Agricultural Residues.
The Residential economic sector includes CO2 emissions from the combustion of fossil fuels that are included in the
EIA residential fuel-consuming sector. Stationary combustion emissions of CH4 and N20 are also based on the EIA
residential fuel-consuming sector. Substitution of Ozone Depleting Substances are apportioned to the Residential
economic sector based on emissions from residential air-conditioning systems. Nitrous oxide emissions from the
application of fertilizers to developed land (termed "settlements" by the IPCC) are also included in the Residential
economic sector.
The Commercial economic sector includes CO2 emissions from the combustion of fossil fuels that are included in
the EIA commercial fuel-consuming sector. Emissions of CH4 and N20 from Mobile Combustion are also
apportioned to the Commercial economic sector based on the EIA commercial fuel-consuming sector. Substitution
of Ozone Depleting Substances emissions are apportioned to the Commercial economic sector based on emissions
from commercial refrigeration/air-conditioning systems. Public works sources including direct CH4 from Landfills,
CH4 and N20 from Wastewater Treatment, and Composting are also included in the Commercial economic sector.
Box 2-2: Recent Trends in Various U.S. Greenhouse Gas Emissions-Related Data
Total 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 electricity use, because the electric power
industry—utilities and non-utilities combined—was the second largest source of U.S. greenhouse gas emissions in
2016; (4) emissions per unit of total gross domestic product as a measure of national economic activity; or (5)
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 grown at an average annual rate of 0.1 percent since 1990. This rate is slightly slower than
that for total energy use and fossil fuel consumption, and much slower than that for electricity use, overall gross
domestic product (GDP) and national population (see Table 2-14 and Figure 2-16). These trends vary relative to
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 1.0 percent since 2005. Total energy use
and fossil fuel consumption have also decreased at slower rates than emissions since 2005, while electricity use,
GDP, and national population continued to increase.
Table 2-14: Recent Trends in Various U.S. Data (Index 1990 = 100)
Variable
1990
2005
2012
2013
2014
2015
2016
Avg. Annual Avg. Annual
Change Change
since 1990a since 2005a
Greenhouse Gas Emissions'5
100
115
103
106
107
105
103
0.1%
-1.0%
Energy Usec
100
118
112
115
117
115
116
0.6%
-0.2%
Fossil Fuel Consumption0
100
119
107
110
111
110
109
0.4%
-0.7%
Electricity Usec
100
134
135
136
138
137
136
1.2%
0.1%
GDPd
100
159
171
174
179
184
187
2.4%
1.5%
Population6
100
118
125
126
127
128
129
1.0%
0.8%
a Average annual growth rate
b GWP-weighted values
c Energy-content-weighted values (EIA 2017a)
d Gross Domestic Product in chained 2009 dollars (BEA 2017)
e U.S. Census Bureau (2017)
Trends 2-33
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Figure 2-16: U.S. Greenhouse Gas Emissions Per Capita and Per Dollar of Gross Domestic
Product
200
190-
180-
170-
160-
150-
©
2 140-
II
o 130-
U 120-
110-
100--
90-
80-
70-
60-
50
5S
Real GDP
Population
Emissions per capita
Emissions per $GDP
Source: BEA (2017), U.S. Census Bureau (2017), and emission estimates in this report.
2.3 Indirect Greenhouse Gas Emissions (CO,
NOx, NMVOCs, and S02)
The reporting requirements of the UNFCCC9 request that information be provided on indirect greenhouse gases,
which include CO, NOx, NMVOCs, and SO2. These gases do not have a direct global warming effect, but indirectly
affect terrestrial radiation absorption by influencing the formation and destruction of tropospheric and stratospheric
ozone, or, in the case of SO2, by affecting the absorptive characteristics of the atmosphere. Additionally, some of
these gases may react with other chemical compounds in the atmosphere to form compounds that are greenhouse
gases. Carbon monoxide is produced when carbon-containing fuels are combusted incompletely. Nitrogen oxides
(i.e., NO and NO2) are created by lightning, fires, fossil fuel combustion, and in the stratosphere from N20. Non-
methane volatile organic compounds—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, 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.
Sulfur-containing compounds emitted into the atmosphere tend to exert a negative radiative forcing (i.e., cooling)
and therefore are discussed separately.
One important indirect climate change effect of NMVOCs and NOx is their role as precursors for tropospheric ozone
formation. They can also alter the atmospheric lifetimes of other greenhouse gases. Another example of indirect
greenhouse gas formation into greenhouse gases is the interaction of CO with the hydroxyl radical—the major
9 See .
2-34 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 atmospheric sink for CH4 emissions—to form CO2. Therefore, increased atmospheric concentrations of CO limit the
2 number of hydroxyl molecules (OH) available to destroy CH4.
3 Since 1970, the United States has published estimates of emissions of CO, NOx, NMVOCs, and SO2 (EPA 2016b),10
4 which are regulated under the Clean Air Act. Table 2-15 shows that fuel combustion accounts for the majority of
5 emissions of these indirect greenhouse gases. Industrial processes—such as the manufacture of chemical and allied
6 products, metals processing, and industrial uses of solvents—are also significant sources of CO, NOx, and
7 NMVOCs.
8 Table 2-15: Emissions of NOx, CO, NMVOCs, and SO2 (kt)
Gas/Activity
1990
2005
2012
2013
2014
2015
2016
NOx
21,791
17,443
12,038
11,388
10,807
10,252
9,278
Mobile Fossil Fuel Combustion
10,862
10,295
6,871
6,448
6,024
5,417
4,814
Stationary Fossil Fuel Combustion
10,02:'
5,858
3,655
3,504
3,291
3,061
2,692
Oil and Gas Activities
139
321
663
704
745
745
745
Forest Fires
81
239
276
185
185
474
474
Industrial Processes and Product Use
592
572
443
434
424
424
424
Waste Combustion
82
128
82
91
100
100
100
Grassland Fires
5
21
39
13
27
21
19
Agricultural Burning
6
6
7
7
8
8
7
Waste
-
2
2
2
2
2
2
CO
132,926
75,569
54,109
48,589
46,875
54,977
52,990
Mobile Fossil Fuel Combustion
119,360
58,615
36,153
34,000
31,848
29,881
27,934
Forest Fires
2,880
8,484
9,804
6,624
6,595
16,752
16,752
Stationary Fossil Fuel Combustion
5,000
4,648
4,027
3,884
3,741
3,741
3,741
Waste Combustion
978
1,403
1,318
1,632
1,947
1,947
1,947
Industrial Processes and Product Use
4,129
1,557
1,246
1,262
1,273
1,273
1,273
Oil and Gas Activities
302
318
666
723
780
780
780
Grassland Fires
84
358
657
217
442
356
324
Agricultural Burning
191
178
232
239
240
239
230
Waste
1
7
6
8
9
9
9
NMVOCs
20,930
13,154
11,464
11,202
10,935
10,647
10,362
Industrial Processes and Product Use
7,638
5,849
3,861
3,793
3,723
3,723
3,723
Mobile Fossil Fuel Combustion
10,932
5,724
4,243
3,924
3,605
3,318
3,032
Oil and Gas Activities
554
510
2,651
2,786
2,921
2,921
2,921
Stationary Fossil Fuel Combustion
912
716
569
539
507
507
507
Waste Combustion
222
241
94
108
121
121
121
Waste
673
114
45
51
57
57
57
Agricultural Burning
NA
NA
NA
NA
NA
NA
NA
SO2
20,935
13,196
5,876
5,874
4,357
3,448
2,457
Stationary Fossil Fuel Combustion
18,40"
11,541
5,006
5,005
3,640
2,756
1,790
Industrial Processes and Product Use
1,30"
831
604
604
496
496
496
Mobile Fossil Fuel Combustion
390
180
108
108
93
93
93
Oil and Gas Activities
79'
619
142
142
95
70
44
Waste Combustion
38
25
15
15
32
32
32
Waste
+
1
+
+
1
1
1
Agricultural Burning
NA
NA
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 2016b) except for estimates from Field Burning of Agricultural Residues.
10 NOx and CO emission estimates from Field Burning of Agricultural Residues were estimated separately, and therefore not
taken from EPA (2016b).
Trends 2-35
-------
1
Box 2-3: Sources and Effects of Sulfur Dioxide
2 Sulfur dioxide (SO2) emitted into the atmosphere through natural and anthropogenic processes affects the earth's
3 radiative budget through its photochemical transformation into sulfate aerosols that can:
8 The indirect effect of sulfur-derived aerosols on radiative forcing can be considered in two parts. The first indirect
9 effect is the aerosols' tendency to decrease water droplet size and increase water droplet concentration in the
10 atmosphere. The second indirect effect is the tendency of the reduction in cloud droplet size to affect precipitation
11 by increasing cloud lifetime and thickness. Although still highly uncertain, the radiative forcing estimates from both
12 the first and the second indirect effect are believed to be negative, as is the combined radiative forcing of the two
13 (IPCC 2013).
14 Sulfur dioxide is also a major contributor to the formation of regional haze, which can cause significant increases in
15 acute and chronic respiratory diseases. Once SO2 is emitted, it is chemically transformed in the atmosphere and
16 returns to the earth as the primary source of acid rain. Because of these harmful effects, the United States has
17 regulated SO2 emissions in the Clean Air Act.
18 Electric power is the largest anthropogenic source of SO2 emissions in the United States, accounting for 43.8 percent
19 in 2016. Coal combustion contributes nearly all of those emissions (approximately 92 percent). Sulfur dioxide
20 emissions have decreased in recent years, primarily as a result of electric power generators switching from liigh-
21 sulfur to low-sulfur coal and installing flue gas desulfurization equipment.
4
5
6
7
(1) scatter radiation from the sun back to space, thereby reducing the radiation reaching the earth's surface;
(2) affect cloud formation; and
(3) affect atmospheric chemical composition (e.g., by providing surfaces for heterogeneous chemical
reactions).
22
23
2-36 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
3. Energy
Energy-related activities were the primary sources of U.S. anthropogenic greenhouse gas emissions, accounting for
83.7 percent of total greenhouse gas emissions on a carbon dioxide (CO2) equivalent basis in 2016.1 This included
97, 43, and 10 percent of the nation's CO2, methane (CH4), and nitrous oxide (N20) emissions, respectively. Energy-
related CO2 emissions alone constituted 78.8 percent of national emissions from all sources on a CO2 equivalent
basis, while the non-C02 emissions from energy-related activities represented a much smaller portion of total
national emissions (4.8 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). Globally, approximately 32,294 million metric tons (MMT) of CO2 were
added to the atmosphere through the combustion of fossil fuels in 2015, of which the United States accounted for
approximately 15 percent.2 Due to their relative importance, fossil fuel combustion-related CO2 emissions are
considered separately, and in more detail than other energy-related emissions (see Figure 3-2). Fossil fuel
combustion also emits CH4 and N20. Stationary combustion of fossil fuels was the second largest source of N20
emissions in the United States and mobile fossil fuel combustion was the fourth largest source.
Figure 3-1: 2016 Energy Chapter Greenhouse Gas Sources (MMT CO2 Eq.)
CO: Emissions from Fossil Fuel Combustion
Natural Gas Systems
Non-Energy Use of Fuels
Petroleum Systems
Coal Mining
Non-CO: Emissions from Stationary
Combustion
Non-CO? Emissions from Mobile Combustion
Incineration of Waste
Abandoned Oil and Gas Wells
Abandoned Underground Coal Mines _
0 50 100 150 200 250 300
MMT CO* Eq.
4,977
Energy as a Portion of
all Emissions
83.7%
1 Estimates are presented in units of million metric tons of carbon dioxide equivalent (MMT CO2 Eq.), which weight each gas by
its global wanning potential, or GWP, value. See section on global wanning potentials in the Executive Summary.
2 Global CO2 emissions from fossil fuel combustion were taken from International Energy Agency CO: Emissions from Fossil
Fuels Combustion - Highlights IEA (2017).
Energy 3-1
-------
1
Figure 3-2: 2016 U.S. Fossil Carbon Flows (MMT CO2 Eq.)
International.
Bunkers
Industrial
Processes
Fossil Fuel
Energy Exports
Coal Emissions
1,316
NEU Exports
Combustion
Emissions
1,307
Natural Gas Emissions
1,482
Coal
1,377
Combustion
Emissions 1,477
NEU Emissions 107
Atmospheric
Emissions
5,292
Domestic
Fossil Fuel
Production
4,495
Apparent
Consumption
5,390
Petroleum
Emissions
2,299
Natural Gas
1,455
Combustion
Emissions
2,193
Petroleum
Natural Gas Liquids,
Liquefied Refinery Gas,
& Other Liquids
Non-Energy Use
Carbon Sequestered
Fossil Fuel
Energy
Imports
Petroleum
1,383 ,
Balancing
Item
(110)
NEU U.S.
Territories
Note: Totals may not sum due to independent rounding.
The "Balancing Item" above accounts for the statistical imbalances
and unknowns in the reported data sets combined here.
NEU = Non-Energy Use
Fossil Fuel
Consumption
U.S.
Territories
Stock
Changes
(68)
Natural Gas 163
Coal 23'
NEU Imports
Other 226
2
3 Energy-related activities other than fuel combustion such as the production, transmission storage, and distribution
4 of fossil fuels, also emit greenhouse gases. These emissions consist primarily of fugitive CH4 from natural gas
5 systems, petroleum systems, and coal mining. Table 3-1 summarizes emissions from the Energy sector in units of
6 MMT CO2 Eq., while unweighted gas emissions in kilotons (kt) are provided in Table 3-2. Overall, emissions due to
7 energy-related activities were 5,476.4 MMT CO2 Eq. in 2016,3 an increase of 2.6 percent since 1990 and a decrease
8 of 2.1 percent since 2015.
9 Table 3-1: CO2, ChU, and N2O Emissions from Energy (MMT CO2 Eq.)
Gas/Source
1990
2005
2012
2013
2014
2015
2016
CO2
4,922.4
5,952.7
5,203.5
5,361.6
5,404.4
5,269.4
5,160.7
Fossil Fuel Combustion
4,755.8
5,759.1
5,029.8
5,162.3
5,206.1
5,059.3
4,916.1
Electric Power
1,820.8
2,400.9
2,022.2
2,038.1
2,038.0
1,900.7
1,808.8
Transportation
1,467.2
1,855.8
1,661.9
1,677.6
1,717.1
1,735.5
1,794.9
Industrial
874.5
867.8
818.4
848.7
830.8
819.3
807.6
Residential
338.3
357.8
282.5
329.7
345.3
316.8
296.2
Commercial
227.4
227.0
201.3
225.7
233.6
245.6
227.9
U.S. Territories
27.6
49.7
43.5
42.5
41.4
41.4
41.4
Non-Energy Use of Fuels
119.6
141.7
113.3
133.2
127.8
135.1
121.0
Natural Gas Systems
29.7
22.5
24.4
26.0
27.0
26.3
26.7
Petroleum Systems
9.4
17.0
25.6
29.7
32.9
38.0
25.5
Incineration of Waste
8.0
12.5
10.4
10.4
10.6
10.7
10.7
Biomass-Wood"
215.2
206.9
194.9
211.6
218.9
201.5
190.2
International Bunker Fuelsb
103.5
113.1
105.8
99.8
103.4
110.9
114.4
Biofuels-Ethanol"
4.2
22.9
72.8
74.7
76.1
78.9
81.2
Biofuels-Biodiesel"
0.0
0.9
8.5
13.5
13.3
14.1
19.6
CH4
364.7
286.7
283.1
288.7
295.4
289.5
279.2
Natural Gas Systems
193.7
160.0
156.8
159.6
164.2
164.4
162.1
3 Following the revised reporting requirements under the UNFCCC, this Inventory report presents CO2 equivalent values based
on the IPCC Fourth Assessment Report (AR4) GWP values. See the Introduction chapter for more information.
3-2 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
Coal Mining
96.5
64.1
66.5
64.6
64.6
61.2
53.8
Petroleum Systems
42.3
34.7
35.4
38.8
41.0
39.4
39.3
Stationary Combustion
8.6
7.9
7.3
8.7
8.8
7.8
7.2
Abandoned Oil and Gas
Wells
6.5
6.9
7.0
7.0
7.1
7.2
7.1
Abandoned Underground
Coal Mines
7.2
6.6
6.2
6.2
6.3
6.4
6.7
Mobile Combustion
9.8
6.6
4.0
3.7
3.4
3.1
3.0
Incineration of Waste
+
+
+
+
+
+
+
International Bunker Fuelsb
0.2
0.1
0.1
0.1
0.1
0.1
0.1
N2O
53.1
56.4
40.9
40.9
39.4
37.1
36.5
Stationary Combustion
11.1
17.5
16.8
18.6
18.9
18.0
18.4
Mobile Combustion
41.5
38.4
23.8
22.0
20.2
18.8
17.8
Incineration of Waste
0.5
0.4
0.3
0.3
0.3
0.3
0.3
International Bunker Fuelsb
0.9
1.0
0.9
0.9
0.9
0.9
1.0
Total
5,340.2
6,295.7
5,527.6
5,691.1
5,739.1
5,596.0
5,476.4
+ Does not exceed 0.05 MMT CO2 Eq.
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 2006IPCC Guidelines and UNFCCC reporting obligations.
Note: Totals may not sum due to independent rounding.
1 Table 3-2: CO2, ChU, and N2O Emissions from Energy (kt)
Gas/Source
1990
2005
2012
2013
2014
2015
2016
CO2
4,922,449
5,952,727
5,203,523
5,361,554
5,404,420
5,269,370
5,160,744
Fossil Fuel Combustion
4,755,819
5,759,056
5,029,830
5,162,315
5,206,135
5,059,288
4,976,737
Non-Energy Use of Fuels
119,588
141,669
113,275
133,176
127,778
135,106
121,049
Natural Gas Systems
29,708
22,529
24,398
26,004
27,004
26,329
26,739
Petroleum Systems
9,384
17,004
25,629
29,695
32,895
37,971
25,543
Incineration of Waste
7,950
12,469
10,392
10,363
10,608
10,676
10,676
Biomass-Wood"
215,186
206,901
194,903
211,581
218,922
201,457
190,171
International Bunker Fuelsb
103,463
113,139
105,805
99,763
103,400
110,887
114,394
Biofuels-Ethanol"
4,227
22,943
72,827
74,743
76,075
78,934
81,250
Biofuels-Biodiesel"
0
856
8,470
13,462
13,349
14,077
19,648
cm
14,587
11,467
11,326
11,548
11,814
11,581
11,166
Natural Gas Systems
7,748
6,399
6,273
6,385
6,568
6,578
6,483
Coal Mining
3,860
2,565
2,658
2,584
2,583
2,449
2,153
Petroleum Systems
1,693
1,386
1,415
1,553
1,639
1,576
1,571
Stationary Combustion
346
314
292
346
352
313
288
Abandoned Oil and Gas
Wells
260
275
279
280
282
286
284
Abandoned Underground
Coal Mines
288
264
249
249
253
256
268
Mobile Combustion
393
263
160
149
136
123
119
Incineration of Waste
+
+
+
+
+
+
+
International Bunker Fuelsb
7
5
4
3
3
3
4
N2O
178
189
137
137
132
125
122
Stationary Combustion
37
59
56
62
63
60
62
Mobile Combustion
139
129
80
74
68
63
60
Incineration of Waste
2
1
1
1
1
1
1
International Bunker Fuelsb
3
3
3
3
3
3
3
+ 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.
Energy 3-3
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Box 3-1: 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 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 do 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.
Box 3-2: Energy Data from EPA's Greenhouse Gas Reporting Program
On October 30, 2009, the U.S. Enviromnental 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). The rule 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. Data reporting by affected facilities includes the reporting of emissions from fuel combustion at
that affected facility. In general, the threshold for reporting is 25,000 metric tons or more of CO2 Eq. per year.
EPA's GHGRP dataset and the data presented in this Inventory 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. 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 categories to improve the national estimates presented in this
Inventory consistent with IPCC guidelines (see, also. Box 3-4).4 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. Most methodologies used in EPA's
GHGRP are consistent with IPCC, 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. 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. Further information on the reporting categorizations in EPA's GHGRP and specific data
caveats associated with monitoring methods in EPA's GHGRP lias been provided on the GHGRP website.5
EPA presents the data collected by its GHGRP through a data publication tool that allows data to be viewed in
several formats including maps, tables, charts and graphs for individual facilities or groups of facilities.6
4 See .
5 See
.
6 See .
3-4 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
In addition to using GHGRP data to estimate emissions, 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. The industrial
end-use sector activity data collected for the Inventory (EIA 2017) 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, the GHGRP data are used to provide a more detailed breakout of total
emissions in the industrial end-use sector within that source category.
3.1 Fossil Fuel Combustion (CRF Source
Category 1A)
Emissions from the combustion of fossil fuels for energy include the gases CO2, CH4, and N20. 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. Following that is a discussion
of emissions of all three gases from fossil fuel combustion presented by sectoral breakdowns. Methodologies for
estimating CO2 from fossil fuel combustion also differ from the estimation of CH4 and N20 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 N20
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
loon
2005
2012
2013
2014
2015
2016
CO2
4,755.8
5,759.1
5,029.8
5,162.3
5,206.1
5,059.3
4,976.7
CH4
18.5
14.4
11.3
12.4
12.2
10.9
10.2
N2O
52.6
56.0
40.6
40.6
39.0
36.8
36.2
Total
4,826.')
5,829.5
5,081.7
5,215.3
5,257.4
5,107.0
5,023.1
Note: Totals may not sum due to independent rounding
able 3-4: CO2, ChU, and N2O Emissions from Fossil Fuel Combustion (kt)
Gas
1990
2005
2012
2013
2014
2015
2016
CO2
4,755,819
5,759,056
5.029,830
5,162,315
5,206,135
5,059,288
4,976,737
CH4
739
578
452
495
488
436
407
N2O
177
188
136
136
131
124
121
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
2016, CO2 emissions from fossil fuel combustion decreased by 1.6 percent relative to the previous year. The
decrease in CO2 emissions from fossil fuel combustion was a result of multiple factors, including: (1) substitution
from coal to natural gas and other sources in the electric power sector; and (2) warmer winter conditions in 2016
resulting in a decreased demand for heating fuel in the residential and commercial sectors. In 2016, CO2 emissions
from fossil fuel combustion were 4,976.7 MMT CO2 Eq., or 4.6 percent above emissions in 1990 (see Table 3-5).7
7 An additional discussion of fossil fuel emission trends is presented in the Trends in U.S. Greenhouse Gas Emissions chapter.
Energy 3-5
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Table 3-5: CO2 Emissions from Fossil Fuel Combustion by Fuel Type and Sector (MMT CO2
Eq.)
Fuel/Sector
1990
2005
2012
2013
2014
2015
2016
Coal
1,718.4
2,112.3
1,592.8
1,653.8
1,652.6
1,423.3
1,306.6
Residential
3.0
0.8
NO
NO
NO
NO
0.0
Commercial
12.0
9.3
4.1
3.9
3.8
2.9
2.3
Industrial
155.3
115.3
74.1
75.7
75.6
65.9
59.0
Transportation
NE
NE
NE
NE
NE
NE
NE
Electric Power
1,547.6
1,983.8
1,511.2
1,571.3
1,569.1
1,350.5
1,241.3
U.S. Territories
0.6
3.0
3.4
2.8
4.0
4.0
4.0
Natural Gas
1,000.3
1,166.7
1,352.6
1,391.2
1,422.0
1,464.2
1,477.0
Residential
238.0
262.2
224.8
266.2
277.9
253.2
238.3
Commercial
142.1
162.9
156.9
179.1
189.3
175.7
170.3
Industrial
408.9
388.5
434.8
451.9
468.4
466.7
478.8
Transportation
36.0
33.1
41.3
47.0
40.3
39.5
40.6
Electric Power
175.3
318.8
492.2
444.0
443.2
526.1
545.9
U.S. Territories
NO
1.3
2.6
3.0
3.0
3.0
3.0
Petroleum
2,036.6
2,479.7
2,084.0
2,116.9
2,131.1
2,171.3
2,192.7
Residential
97.4
94.9
57.7
63.5
67.4
63.6
58.0
Commercial
73.3
54.9
40.4
42.7
40.4
67.0
55.3
Industrial
310.4
364.0
309.6
321.1
286.8
286.7
269.7
Transportation
1,431.2
1,822.7
1,620.6
1,630.6
1,676.9
1,696.0
1,754.2
Electric Power
97.5
97.9
18.3
22.4
25.3
23.7
21.2
U.S. Territories
26.9
45.4
37.5
36.6
34.3
34.3
34.3
Geothermal3
0.4
0.4
0.4
0.4
0.4
0.4
0.4
Total
4,755.8
5,759.1
5,029.8
5,162.3
5,206.1
5,059.3
4,976.7
NE (Not Estimated)
NO (Not Occurring)
a Although not technically a fossil fuel, geothermal energy-related CO2 emissions are included for reporting
purposes.
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.
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, power plants, steel mills, and light bulbs), and social planning
and consumer behavior (e.g., walking, bicycling, or telecommuting to work instead of driving).
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.8
Table 3-6 shows annual changes in emissions during the last five years for coal, petroleum, and natural gas in
selected sectors.
8 Based on national aggregate carbon content of all coal, natural gas, and petroleum fuels combusted in the United States.
3-6 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Table 3-6: Annual Change in CO2 Emissions and Total 2016 Emissions from Fossil Fuel
Combustion for Selected Fuels and Sectors (MMT CO2 Eq. and Percent)
Sector
Fuel Type
2012 to 2013
2013 to 2014
2014 to 2015
2015 to 2016
Total 2016
Electric Power
Coal
60.1
4.0%
-2.2
-0.1%
-218.7
-13.9%
-109.2
-8.1%
1,241.3
Electric Power
Natural Gas
-48.3
-9.8%
-0.8
-0.2%
82.9
18.7%
19.8
3.8%
545.9
Electric Power
Petroleum
4.1
22.3%
2.9
12.8%
-1.6
-6.4%
-2.5
-10.4%
21.2
Transportation
Petroleum
10.0
0.6%
46.3
2.8%
19.2
1.1%
58.2
3.4%
1,754.2
Residential
Natural Gas
41.4
18.4%
11.6
4.4%
-24.7
-8.9%
-14.9
-5.9%
238.3
Commercial
Natural Gas
22.3
14.2%
10.2
5.7%
-13.6
-7.2%
-5.4
-3.1%
170.3
Industrial
Coal
1.7
2.3%
-0.1
-0.1%
-9.8
-12.9%
-6.8
-10.4%
59.0
Industrial
Natural Gas
17.1
3.9%
16.5
3.7%
-1.7
-0.4%
12.2
2.6%
478.8
All Sectors3
All Fuels3
132.5
2.6%
43.8
0.8%
-146.8
-2.8%
-82.6
-1.6%
4,976.7
a Includes sector and fuel combinations not shown in this table.
Note: Totals may not sum due to independent rounding.
As shown in Table 3-6, recent trends in CO2 emissions from fossil fuel combustion show a 2.6 percent increase from
2012 to 2013, then a 0.8 percent increase from 2013 to 2014, then a 2.8 percent decrease from 2014 to 2015, and a
1.6 percent decrease from 2015 to 2016. Total electric power generation remained relatively flat over that time
period but emission trends generally mirror the trends in the amount of coal used to generate electricity. The
consumption of coal used to generate electricity increased by roughly 4 percent from 2012 to 2013, stayed relatively
flat from 2013 to 2014, decreased by 14 percent from 2014 to 2015, and decreased by 8 percent from 2015 to 2016.
The overall CO2 emission trends from fossil fuel combustion also follow closely changes in heating degree days
over that time period. Heating degree days increased by 18 percent from 2012 to 2013, increased by 2 percent from
2013 to 2014, decreased by 10 percent from 2014 to 2015 and decreased by 5 percent from 2015 to 2016. A
decrease in heating degree days leads to decreased demand for heating fuel and electricity for heat in the residential
and commercial sector, primarily in winter months. The overall CO2 emission trends from fossil fuel combustion
also generally follow changes in overall petroleum use and emissions. Carbon dioxide emissions from all petroleum
increased by 1.6 percent from 2012 to 2013, increased by 0.7 percent from 2013 to 2014, increased by 1.9 percent
from 2014 to 2015, and increased by 1.0 percent from 2015 to 2016. The increase in petroleum CO2 emissions from
2015 to 2016 somewhat offsets emission reductions from other sources like decreased coal use in the electricity
sector.
In the United States, 81 percent of the energy used in 2016 was produced through the combustion of fossil fuels such
as coal, natural gas, and petroleum (see Figure 3-3 and Figure 3-4). The remaining portion was supplied by nuclear
electric power (9 percent) and by a variety of renewable energy sources (10 percent), primarily hydroelectric power,
wind energy and biofuels (EIA 2017a).9 Specifically, petroleum supplied the largest share of domestic energy
demands, accounting for 37 percent of total U.S. energy used in 2016. Natural gas and coal followed in order of
energy demand importance, accounting for approximately 29 percent and 15 percent of total U.S. energy used,
respectively. Petroleum was consumed primarily in the transportation end-use sector and the vast majority of coal
was used in the electric power end-use sector. Natural gas was broadly consumed in all end-use sectors except
transportation (see Figure 3-5) (EIA 2017a).
9 Renewable energy, as defined in EIA's energy statistics, includes the following energy sources: hydroelectric power,
geothermal energy, biofuels, solar energy, and wind energy.
Energy 3-7
-------
1
Figure 3-3:
2016 U.S. Energy Consumption by Energy Source (Percent)
Nuclear Electric Power
8,6%
Renewable Energy
10.4%
Petroleum
36.9%
Coal
14.8%
Natural Gas
29.2%
Figure 3-4: U.S. Energy Consumption (Quadrillion Btu)
120-
100-
CD
cy
Cl
£
3
<3
>.
as
80-
60-
40
20-
Total Energy
Fossil Fuels
Renewable & Nuclear
o
CM
ro
LT>
UD
r**
€0
ON
o
ro
in
VD
fv
00
Ch
o
V i
rsj
ro
*r
U~>
a
CTv
Ch
o\
&¦
<3%
a\
G\
CT'
Q
O
o
o
o
o
CJ
CJ
a
d:
¦H
'—
•*—
¦*—
-•—
tH
¦»—
o\
c\
o\
&
O
O
©
o
c
o
o
o
o
o
o
c
o
O
o
o
O
11
* i
v 4
V 1
V 1
* i
tH
i—<
fM
Psl
rsi
rsj
rsj
rsj
rsi
rsj
OJ
rsj
f*Nj
rsj
rsi
fM
04
fN
3-8 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Figure 3-5: 2016 CO2 Emissions from Fossil Fuel Combustion by Sector and Fuel Type (MMT
COz Eq.)
2,500
2,000
S 1,500
o
u
i-
11,000
500
Relative Contribution by Fuel Type
I Petroleum
Coal
I Natural Gas
228
41
1,795
1,809
U.S. Territories Commercial Residential Industrial Transportation Electric Power
Note on Figure 3-5: Fossil Fuel Combustion includes electric power, which also includes emissions of less than 0.5 MMT CO2
Eq. from geothermal-based generation.
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, CO, and NMVOCs.10 These other C-containing non-CO; gases are emitted as a byproduct of
incomplete fuel combustion but are, for the most part, eventually oxidized to CO2 in the atmosphere. Therefore, it is
assumed all of the C in fossil fuels used to produce energy is eventually converted to atmospheric CO2.
Box 3-3: Weather and Non-Fossil Energy Effects on CO2 from Fossil Fuel Combustion Trends
In 2016, weather conditions, and a warm first and fourth quarter of the year in particular, caused a significant
decrease in demand for heating fuels and is reflected in the decreased residential emissions from 2015 to 2016. The
United States in 2016 also experienced a wanner winter overall compared to 2015, as heating degree days decreased
(5.1 percent). Wanner winter conditions compared to 2015 resulted in a decrease in the amount of energy required
for heating, and heating degree days in the United States were 14.2 percent below nonnal (see Figure 3-6). Cooling
degree days increased, by 4.6 percent, and increased demand for air conditioning in the residential and commercial
sector, this led in part to an overall residential electricity demand increase of 0.2 percent. Summer conditions were
significantly wanner in 2016 compared to 2015, with cooling degree days 28.0 percent above nonnal (see Figure
3-7) (EIA 2017a).11
10 See the sections entitled Stationary Combustion and Mobile Combustion in this chapter for information on non-CCb gas
emissions from fossil fuel combustion.
11 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 1981 through 2010. The variation in these
normals during this time period was +12 percent and +19 percent for heating and cooling degree days, respectively (99 percent
confidence interval).
Energy 3-9
-------
1 Figure 3-6: Annual Deviations from Normal Heating Degree Days for the United States
2 (1950-2016, Index Normal = 100)
10-
-20-
Normal
(4,524 Heating Degree Days)
99% Confidence
Note: Ciimatological normal data are highlighted. Statistical confidence interval for "normal" climatology period of 1981 through 2010.
orsi*r vDooorsi*Tv£>co o oj t ud co o cm ud co OfMTu^coooj t y) co o r-j t ud
Lnir>i/7t/>u->u3vp^0<:0vorvi>»rvr-»rH»aotx>oococoONgNONONONOQOOQ-'-<^-«'-«-r-i
on on o>i on on on On on on on on on on on on on on On on on On On on On on oo ooo oooo
*-4 -r-i T-) w —< . *-t *-* w ^ tH —• — — tH ^ t-( H N (N Osl CM rM IN fM M M
Figure 3-7: Annual Deviations from Normal Cooling Degree Days for the United States
(1950-2016, Index Normal = 100)
Normal
(1,216 cooling degree days)
99% Confidence
_20- Note: Ciimatological normal data are highlighted. Statistical confidence interval for "normal" climatology period of 1981 through 2010.
ocvj<3-i£»GOOfNjTry>coof"sj
unLnLOLOLnuDUDUDU3'UDr*vfN>. . „ _
On On On On On On On On On On On On On On On On On On On On On
UDCOOrsJTUDCOOr-JTT
O) CO CO CO CO ON ^ ON
i£>COOrvJ
-------
1 2016, nuclear power represented 20 percent of total electricity production. In recent years, the wind and solar power
2 sectors have been showing strong growth, such that, on the margin, they are becoming relatively important
3 electricity sources. Between 1990 and 2016, renewable energy generation (in kWh) from solar and wind energy
4 have increased from 0.1 percent in 1990 to 7 percent in 2016, which helped drive the decreases in the carbon
5 intensity of the electricity supply in the United States.
6
7 Fossil Fuel Combustion Emissions by Sector
8 In addition to the CO2 emitted from fossil fuel combustion, CH4 and N20 are emitted from stationary and mobile
9 combustion as well. Table 3-7 provides an overview of the CO2, CH4, and N20 emissions from fossil fuel
10 combustion by sector.
11 Table 3-7: CO2, ChU, and N2O Emissions from Fossil Fuel Combustion by Sector (MMT CO2
12 Eq.)
End-Use Sector
1990
2005
2012
2013
2014
2015
2016
Electric Power
1,827.7
2,414.9
2,036.3
2,053.8
2,054.0
1,916.1
1,824.8
CO2
1,820.8
2,400.9
2,022.2
2,038.1
2,038.0
1,900.7
1,808.8
CH4
0.4
0.9
1.1
1.0
1.0
1.1
1.1
N2O
6.-
13.2
13.1
14.6
15.0
14.3
14.9
Transportation
1,518.5
1,900.8
1,689.7
1,703.3
1,740.7
1,757.4
1,815.7
CO2
1,467.2
1,855.8
1,661.9
1,677.6
1,717.1
1,735.5
1,794.9
CH4
9.8
6.6
4.0
3.7
3.4
3.1
3.0
N2O
41.5
38.4
23.8
22.0
20.2
18.8
17.8
Industrial
879.fi
872.6
822.5
852.8
834.8
823.2
811.4
CO2
874 -
867.8
818.4
848.7
830.8
819.3
807.6
CH4
1.9
1.8
1.5
1.5
1.5
1.5
1.4
N2O
3.2
3.0
2.6
2.6
2.5
2.5
2.4
Residential
344.fi
362.8
287.0
335.7
351.4
321.6
300.3
CO2
338. ^
357.8
282.5
329.7
345.3
316.8
296.2
CH4
5.2
4.1
3.7
5.0
5.1
3.9
3.4
N2O
1.0
0.9
0.7
1.0
1.0
0.8
0.7
Commercial
228. S
228.5
202.5
227.1
235.0
247.2
229.4
CO2
227.4
227.0
201.3
225.7
233.6
245.6
227.9
CH4
1.1
1.1
0.9
1.1
1.1
1.2
1.2
N2O
0.4
0.3
0.3
0.3
0.3
0.4
0.3
U.S. Territories3
27.7
49.9
43.7
42.6
41.5
41.5
41.5
Total
4,826.'J
5,829.5
5,081.7
5,215.3
5,257.4
5,107.0
5,023.1
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 consumption by each end-use sector.
13 Other than CO2, gases emitted from stationary combustion include the greenhouse gases CH4 and N20 and the
14 indirect greenhouse gases NOx, CO, and NMVOCs.13 Methane and N20 emissions from stationary combustion
15 sources depend upon fuel characteristics, size and vintage, along with combustion technology, pollution control
16 equipment, ambient environmental conditions, and operation and maintenance practices. Nitrous oxide emissions
17 from stationary combustion are closely related to air-fuel mixes and combustion temperatures, as well as the
18 characteristics of any pollution control equipment that is employed. Methane emissions from stationary combustion
19 are primarily a function of the CH4 content of the fuel and combustion efficiency.
13 Sulfur dioxide (SO2) emissions from stationary combustion are addressed in Annex 6.3.
Energy 3-11
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Mobile combustion produces greenhouse gases other than CO2, including CH4, N20, and indirect greenhouse gases
including NOx, CO, and NMVOCs. As with stationary combustion, N20 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 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 were defined: industrial, transportation, residential, and
commercial. In the table 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 N20 from transportation.14 Emissions
from U.S. Territories are also calculated separately due to a lack of end-use-specific consumption data.15 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 3-8.
Table 3-8: CO2, ChU, and N2O Emissions from Fossil Fuel Combustion by End-Use Sector
(MMT COz Eq.)
End-Use Sector
1990
2005
2012
2013
2014
2015
2016
Transportation
1,521.5
1,905.5
1,693.6
1,707.3
1,744.7
1,761.1
1,819.2
CO2
1,470.2
1,860.5
1,665.8
1,681.6
1,721.2
1,739.2
1,798.4
CH4
9.8
6.6
4.0
3.7
3.4
3.1
3.0
N2O
41.5
38.4
23.8
22.0
20.2
18.8
17.8
Industrial
1,568.9
1,613.5
1,419.4
1,452.1
1,432.7
1,377.2
1,322.1
CO2
1,561.3
1,604.4
1,411.2
1,443.4
1,424.0
1,368.8
1,313.8
CH4
2.0
2.0
1.8
1.8
1.8
1.8
1.8
N2O
5.6
7.1
6.4
6.8
6.9
6.6
6.5
Residential
939.9
1,224.1
1,017.3
1,076.2
1,091.9
1,011.4
966.9
CO2
931.4
1,214.1
1,007.8
1,064.6
1,080.0
1,001.1
957.0
CH4
5.4
4.4
4.1
5.3
5.4
4.3
3.8
N2O
3.2
5.6
5.5
6.2
6.4
5.9
6.1
Commercial
768.8
1,036.5
907.7
937.0
946.5
915.7
873.3
CO2
765.3
1,030.3
901.6
930.2
939.6
908.8
866.2
CH4
1.2
1.4
1.3
1.4
1.5
1.6
1.6
N2O
2.3
4.8
4.8
5.4
5.5
5.4
5.6
U.S. Territories3
27.7
49.9
43.7
42.6
41.5
41.5
41.5
Total
4,826.9
5,829.5
5,081.7
5,215.3
5,257.4
5,107.0
5,023.1
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.
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 3-9 presents CO2 emissions from fossil
14 Separate calculations were performed for transportation-related CH4 and N2O. The methodology used to calculate these
emissions are discussed in the mobile combustion section.
15 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.
3-12 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 fuel combustion by stationary sources. The CO2 emitted is closely linked to the type of fuel being combusted in each
2 sector (see Methodology section of CO2 from Fossil Fuel Combustion). Other than CO2, gases emitted from
3 stationary combustion include the greenhouse gases CH4 and N20. Table 3-10 and Table 3-11 present CH4 and N20
4 emissions from the combustion of fuels in stationary sources. The CH4 and N20 emission estimation methodology
5 utilizes facility-specific technology and fuel use data reported to EPA's Acid Rain Program (EPA 2017a) (see
6 Methodology section for CH4 and N20 from Stationary Combustion). Table 3-7 presents the corresponding direct
7 C02, CH4, and N20 emissions from all sources of fuel combustion, without allocating emissions from electricity use
8 to the end-use sectors.
9 Table 3-9: CO2 Emissions from Stationary Fossil Fuel Combustion (MMT CO2 Eq.)
Sector/Fuel Type
1990
2005
2012
2013
2014
2015
2016
Electric Power
1,820.8
2,400.9
2,022.2
2,038.1
2,038.0
1,900.7
1,808.8
Coal
1,547.6
1,983.8
1,511.2
1,571.3
1,569.1
1,350.5
1,241.3
Natural Gas
175.3
318.8
492.2
444.0
443.2
526.1
545.9
Fuel Oil
97.5
97.9
18.3
22.4
25.3
23.7
21.2
Geo thermal
0.4
0.4
0.4
0.4
0.4
0.4
0.4
Industrial
874.5
867.8
818.4
848.7
830.8
819.3
807.6
Coal
155.3
115.3
74.1
75.7
75.6
65.9
59.0
Natural Gas
408.9
388.5
434.8
451.9
468.4
466.7
478.8
Fuel Oil
310.4
364.0
309.6
321.1
286.8
286.7
269.7
Commercial
227.4
227.0
201.3
225.7
233.6
245.6
227.9
Coal
12.0
9.3
4.1
3.9
3.8
2.9
2.3
Natural Gas
142.1
162.9
156.9
179.1
189.3
175.7
170.3
Fuel Oil
73.3
54.9
40.4
42.7
40.4
67.0
55.3
Residential
338.3
357.8
282.5
329.7
345.3
316.8
296.2
Coal
3.0
0.8
NO
NO
NO
NO
NO
Natural Gas
238.0
262.2
224.8
266.2
277.9
253.2
238.3
Fuel Oil
97.4
94.9
57.7
63.5
67.4
63.6
58.0
U.S. Territories
27.6
49.7
43.5
42.5
41.4
41.4
41.4
Coal
0.6
3.0
3.4
2.8
4.0
4.0
4.0
Natural Gas
NO
1.3
2.6
3.0
3.0
3.0
3.0
Fuel Oil
26.9
45.4
37.5
36.6
34.3
34.3
34.3
Total
3,288.6
3,903.3
3,367.9
3,484.7
3,489.0
3,323.8
3,181.9
NO (Not Occurring)
Note: Totals may not sum due to independent rounding.
10 Table 3-10: ChU Emissions from Stationary Combustion (MMT CO2 Eq.)
Sector/Fuel Type
1990
2005
2012
2013
2014
2015
2016
Electric Power
0.4
0.9
1.1
1.0
1.0
1.1
1.1
Coal
0.3
0.4
0.3
0.3
0.3
0.3
0.2
Fuel Oil
+
+
+
+
+
+
+
Natural gas
0.1
0.5
0.8
0.7
0.7
0.9
0.9
Wood
+
+
+
+
+
+
+
Industrial
1.9
1.8
1.5
1.5
1.5
1.5
1.4
Coal
0.4
0.3
0.2
0.2
0.2
0.2
0.2
Fuel Oil
0.2
0.3
0.2
0.2
0.2
0.2
0.2
Natural gas
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Wood
1.0
1.0
1.0
0.9
0.9
0.9
0.9
Commercial
1.1
1.1
0.9
1.1
1.1
1.2
1.2
Coal
+
+
+
+
+
+
+
Fuel Oil
0.3
0.2
0.1
0.2
0.1
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
0.4
0.5
0.5
0.6
0.6
Energy 3-13
-------
Residential
5.2
4.1
3.7
5.0
5.1
3.9
3.4
Coal
0.2
0.1
NO
NO
NO
NO
NO
Fuel Oil
0.3
0.3
0.2
0.2
0.2
0.2
0.2
Natural Gas
0.5
0.6
0.5
0.6
0.6
0.6
0.5
Wood
4.1
3.1
3.0
4.1
4.2
3.1
2.7
U.S. Territories
+
0.1
0.1
0.1
0.1
0.1
0.1
Coal
+
+
+
+
+
+
+
Fuel Oil
+
0.1
0.1
0.1
0.1
0.1
0.1
Natural Gas
NO
+
+
+
+
+
+
Wood
NO
NO
NO
NO
NO
NO
NO
Total
8.6
7.9
7.3
8.7
8.8
7.8
7.2
+ Does not exceed 0.05 MMT CO2 Eq.
NO (Not Occurring)
Note: Totals may not sum due to independent rounding.
1 Table 3-11: N2O Emissions from Stationary Combustion (MMT CO2 Eq.)
Sector/Fuel Type
1990
2005
2012
2013
2014
2015
2016
Electric Power
6.5
13.2
13.1
14.6
15.0
14.3
14.9
Coal
6.1
11.1
9.8
11.6
11.9
10.6
11.1
Fuel Oil
0.1
0.1
+
+
+
+
+
Natural Gas
0.3
1.9
3.2
3.0
3.1
3.6
3.7
Wood
+
+
+
+
+
+
+
Industrial
3.2
3.0
2.6
2.6
2.5
2.5
2.4
Coal
0.7
0.5
0.4
0.4
0.4
0.3
0.3
Fuel Oil
0.6
0.6
0.5
0.5
0.4
0.4
0.4
Natural Gas
0.2
0.2
0.2
0.2
0.3
0.2
0.3
Wood
1.6
1.6
1.5
1.5
1.5
1.5
1.5
Commercial
0.4
0.3
0.3
0.3
0.3
0.4
0.3
Coal
0.1
+
+
+
+
+
+
Fuel Oil
0.2
0.1
0.1
0.1
0.1
0.2
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
0.7
1.0
1.0
0.8
0.7
Coal
+
+
NO
NO
NO
NO
NO
Fuel Oil
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Natural Gas
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Wood
0.7
0.5
0.5
0.7
0.7
0.5
0.4
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
0.1
0.1
0.1
0.1
Natural Gas
NO
+
+
+
+
+
+
Wood
NO
NO
NO
NO
NO
NO
NO
Total
11.1
17.5
16.8
18.6
18.9
18.0
18.4
+ Does not exceed 0.05 MMT CO2 Eq.
NO (Not Occurring)
Note: Totals may not sum due to independent rounding.
2 Electric Power Sector
3 The process of generating electricity is the single largest source of CO2 emissions in the United States, representing
4 34 percent of total CO2 emissions from all CO2 emissions sources across the United States. Methane and N20
5 accounted for a small portion of total greenhouse gas emissions from electric power, representing 0.1 percent and
6 0.8 percent, respectively. Electric power also accounted for the largest share of CO2 emissions from fossil fuel
7 combustion, approximately 36.3 percent in 2016. Methane and N2O from electric power represented 11.2 and 41.1
8 percent of total CH4 and N20 emissions from fossil fuel combustion in 2016, respectively.
3-14 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
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 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). Electric generation is reported as
occurring in other sectors where the producer of the power indicates that its primary business is something other
than the production of electricity.16
Emissions from the electric power sector have decreased by 0.2 percent since 1990. The carbon intensity of the
electric power sector, in terms of CO2 Eq. per QBtu, input has significantly decreased by 12 percent during that
same timeframe with the majority of the emissions and carbon intensity decreases coming in the past decade as
shown below in Figure 3-8. This recent decarbonization of the electric power sector is a result of several key drivers.
Coal-fired electric power (in kilowatt-hours [kWh]) decreased from almost 54 percent of generation in 1990 to 32
percent in 2016.17 This generation corresponded with an increase in natural gas 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 27-year period to represent 33 percent of electric power sector generation
in 2016.
In 2016, CO2 emissions from the electric power sector decreased by 4.8 percent relative to 2015. This decrease in
CO2 emissions was a result of changes in the types of fuel consumed to produce electricity in the electric power
sector in recent years. The shift from coal to less-CCh-intensive natural gas to supply electricity has accelerated in
recent years. Consumption of coal for electric power decreased by 8.1 percent from 2015 to 2016, while
consumption of natural gas increased by 3.8 percent. There has also been a rapid increase in renewable energy
capacity additions in the electric power sector in recent years. In 2016, renewable energy sources accounted for 63
percent of capacity additions, with natural gas accounting for the remaining additions. The share of renewable
energy capacity additions has grown significantly since 2010, when renewable energy sources accounted for only 28
percent of total capacity additions (EIA 2017e). The decrease in coal-powered electricity generation and increase in
renewable energy capacity contributed to a decrease in emissions from electric power generation over the time series
(see Figure 3-8).
Decreases in natural gas costs and the associated increase in natural gas generation, particularly between 2005 and
2016, was one of the main driver 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 57 percent while the cost of coal (in
$/MMBtu) increased by 83 percent (EIA 2017a). Also, between 1990 and 2016, renewable energy generation (in
kWh) from wind and solar energy have increased from 0.1 percent in 1990 to 7 percent in 2016, 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 3,711 billion kWh in 2016.
16 Utilities primarily generate power for the U.S. electric grid for sale to retail customers. Non-utilities produce electricity for
their own use, to sell to large consumers, or to sell on the wholesale electricity market (e.g., to utilities for distribution and resale
to customers).
17 Values represent electricity net generation from the electric power sector (EIA 2017a).
Energy 3-15
-------
1 Figure 3-8: Fuels Used in Electric Power Generation (TBtu) and Total Electric Power Sector
2 CO2 Emissions
I Petroleum (TBtu)
I Nuclear (TBtu)
Renewable Energy Sources (TBtu)
I Natural Gas (TBtu)
Coal (TBtu)
Net Generation (Index vs. 1990) [Right Axis]
I Sector COs Emissions (Index vs. 1990) [Right Axis]
30,000
25,000
' 20,000
15,000
10,000
5,000
_ ,
On On On On On On On On On G* C5 CD • i ~—i ~w—4 <
ononononononononononSoooocjcjooooooo coo
Electricity was consumed primarily in the residential, commercial, and industrial end-use sectors for lighting,
heating, electric motors, appliances, electronics, and air conditioning (see Figure 3-9).
Figure 3-9: Electric Power Retail Sales by End-Use Sector (Billion kWh)
160
140
120
100
80
60
40
20
u
T3
1,500
1,400-
1,300-
$ 1,200
1,000-
900-
800
Residential
Commercia
Industrial
O rH fN ro IA VD
ON On ON On On On On
On On On On On On On On
oo On
ON On gN
O -h
o o
OJ PnJ
T id VD N
o o o c
o o o o
N (N fN fN
ON O
C -H
a o
r-j fNj
8 The industrial, residential, and commercial end-use sectors, as presented in Table 3-8, were reliant on electricity for
9 meeting energy' needs. The residential and commercial end-use sectors are especially reliant on electricity use for
10 lighting, heating, air conditioning, and operating appliances. In 2016. electricity sales to the residential end-use
11 sector increased by 0.2 percent and sales to the commercial end-use sector decreased by 0.1 percent, respectively.
12 Electricity sales to the industrial sector in 2016 decreased approximately 5.1 percent. Overall, in 2016, the amount of
13 electricity retail sales (in kWh) decreased by 1.2 percent.
3-16 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Industrial Sector
2 Industrial sector CO2, CH4, and N20, emissions accounted for 16, 14, and 7 percent of CO2, CH4, and N20,
3 emissions from fossil fuel combustion, respectively. Carbon dioxide, CH4, and N20 emissions resulted from the
4 direct consumption of fossil fuels for steam and process heat production.
5 The industrial end-use sector, per the underlying energy use data from EIA, includes activities such as
6 manufacturing, construction, mining, and agriculture. The largest of these activities in terms of energy use is
7 manufacturing, of which six industries—Petroleum Refineries, Chemicals, Paper, Primary Metals, Food, and
8 Nonmetallic Mineral Products—represent the vast majority of the energy use (EIA 2017a; EIA 2009b).
9 There are many dynamics that impact emissions from the industrial sector including economic activity, changes in
10 the make-up of the industrial sector, changes in the emissions intensity of industrial processes, and weather impacts
11 on heating of industrial buildings.18 Structural changes within the U.S. economy that lead to shifts in industrial
12 output away from energy-intensive manufacturing products to less energy-intensive products (e.g., from steel to
13 computer equipment) have had a significant effect on industrial emissions.
14 From 2015 to 2016, total industrial production and manufacturing output decreased by 1.2 percent (FRB 2017).
15 Over this period, output increased across production indices for Food, Petroleum Refineries, Chemicals, and
16 Nonmetallic Mineral Products, and decreased slightly for Primary Metals and Paper (see Figure 3-10). Through
17 EPA's Greenhouse Gas Reporting Program (GHGRP), specific industrial sector trends can be discerned from the
18 overall total EIA industrial fuel consumption data used for these calculations.
19 For example, from 2015 to 2016, the underlying EIA data showed decreased consumption of coal, and relatively flat
20 use of natural gas in the industrial sector. The GHGRP data highlights that several industries contributed to these
21 trends, including chemical manufacturing; pulp, paper and print; and food processing, beverages and tobacco.19
18 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.
19 Further details on industrial sector combustion emissions are provided by EPA's GHGRP. See
.
Energy 3-17
-------
1 Figure 3-10: Industrial Production Indices (Index 2012=100)
140-
120-
Total excluding Computers, Communications Equipment, and Semiconductors
100-
Total Industrial
80-
Stone, Clay, & Glass Products
Chemicals
140-
120-
Primary Metals
100J
80-
Petroleum Refineries
3 Despite the growth in industrial output (60 percent) and the overall U.S. economy (87 percent) from 1990 to 2016,
4 CO?, emissions from fossil fuel combustion in the industrial sector decreased by 7.7 percent over the same time
5 series. A number of factors are believed to have caused this disparity between growth in industrial output and
6 decrease in industrial emissions, including: (1) more rapid growth in output from less energy-intensive industries
7 relative to traditional manufacturing industries, and (2) energy-intensive industries such as steel are employing new
8 methods, such as electric arc furnaces, that are less carbon intensive than the older methods. In 2016, CO?, CH4. and
9 N2O emissions from fossil fuel combustion and electricity use within the industrial end-use sector totaled 1,322.1
10 MMT CO2 Eq., a 4.0 percent decrease from 2015 emissions.
11 Residential and Commercial Sectors
12 Emissions from the residential and commercial sectors have increased since 1990, and are often correlated with
13 short-term fluctuations in energy use caused by weather conditions, rather than prevailing economic conditions.
14 More significant changes in emissions from the residential and commercial sectors in recent years can be largely
3-18 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
attributed to an overall reduction in energy use, a reduction in heating degree days, and increases in energy
efficiency (see Figure 3-11).
In 2016 the residential and commercial sectors accounted for 6 and 5 percent of CO2 emissions from fossil fuel
combustion, 33 and 11 percent of CH4 emissions from fossil fuel combustion, and 2 and 1 percent of N20 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 both of these end-use sectors. In 2016, total emissions (CO2, CH4, and N20) from fossil
fuel combustion and electricity use within the residential and commercial end-use sectors were 966.9 MMT CO2 Eq.
and 873.3 MMT CO2 Eq., respectively. Total CO2, CH4, and N20 emissions from fossil fuel combustion and
electricity use within the residential and commercial end-use sectors decreased by 4.4 and 4.6 percent from 2015 to
2016, respectively, and heating degree days decreased by 5 percent over the same time period. A decrease in heating
degree days led to a decreased demand for heating fuel and electricity for heat in the residential and commercial
sectors. In addition, a shift toward energy efficient products and more stringent energy efficiency standards for
household equipment lias also contributed to a decrease in energy demand in households (EIA 2017f), resulting in a
decrease in energy-related emissions. In the long term, the residential sector is also affected by population growth,
migration trends toward wanner areas, and changes in housing and building attributes (e.g., larger sizes and
improved insulation).
In 2016, combustion emissions from natural gas consumption represented 80 and 75 percent of the direct fossil fuel
CO2 emissions from the residential and commercial sectors, respectively. Natural gas combustion CO2 emissions
from the residential and commercial sectors in 2016 decreased by 5.9 percent and 3.1 percent from 2015 levels,
respectively.
Figure 3-11: Fuels Used in Residential and Commercial Sectors (TBtu), Heating Degree Days,
and Total Sector CO2 Emissions
Coal (TBtu) | Heating Degree Days (Index vs. 1990) [Right Axis]
Renewable Energy Sources (TBtu) ¦ Sector CO: Emissions (Index vs. 1990) [Right Axis] 140
¦ Petroleum (TBtu)
¦ Natural Gas (TBtu)
Electricity Use (TBtu)
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 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, CH4, and N20 emissions are not
presented for U.S. Territories in the tables above by sector, though the emissions will include some transportation
and mobile combustion sources.
Energy 3-19
-------
1 Transportation Sector and Mobile Combustion
2 This discussion of transportation emissions follows the alternative method of presenting combustion emissions by
3 allocating emissions associated with electricity generation to the transportation end-use sector, as presented in Table
4 3-8. Table 3-7 presents direct CO2, CH4, and N20 emissions from all transportation sources (i.e., excluding
5 emissions allocated to electricity consumption in the transportation end-use sector).
6 The transportation end-use sector and other mobile combustion accounted for 1,819.2 MMT CO2 Eq. in 2016, which
7 represented 36 percent of CO2 emissions, 29 percent of CH4 emissions, and 49 percent of N20 emissions from fossil
8 fuel combustion, respectively.20 Fuel purchased in the United States for international aircraft and marine travel
9 accounted for an additional 115.5 MMT CO2 Eq. in 2016; these emissions are recorded as international bunkers and
10 are not included in U.S. totals according to UNFCCC reporting protocols.
11 Transportation End-Use Sector
12 From 1990 to 2016, transportation emissions from fossil fuel combustion rose by 20 percent due, in large part, to
13 increased demand for travel (see Figure 3-12). The number of vehicle miles traveled (VMT) by light-duty motor
14 vehicles (passenger cars and light-duty trucks) increased 43 percent from 1990 to 2016,21 as a result of a confluence
15 of factors including population growth, economic growth, urban sprawl, and periods of low fuel prices.
16 From 2015 to 2016, CO2 emissions from the transportation end-use sector increased by 3.4 percent. The increase in
17 emissions can largely be attributed to increased VMT and motor gasoline consumption by light duty vehicles, as
18 well as diesel consumption by medium-and heavy-duty vehicles. From 2015 to 2016, there were also increases in
19 residual fuel oil consumption by ships and boats and jet fuel use in general aviation aircraft.
20 Commercial aircraft emissions were similar between 2015 and 2016, but have decreased 15 percent since 2007
21 (FAA 2017).22 Decreases in jet fuel emissions (excluding bunkers) since 2007 are due in part to improved
22 operational efficiency that results in more direct flight routing, improvements in aircraft and engine technologies to
23 reduce fuel burn and emissions, and the accelerated retirement of older, less fuel efficient aircraft.
24 Almost all of the energy consumed for transportation was supplied by petroleum-based products, with more than
25 half being related to gasoline consumption in automobiles and other highway vehicles. Other fuel uses, especially
26 diesel fuel for freight trucks and jet fuel for aircraft, accounted for the remainder. The primary driver of
27 transportation-related emissions was CO2 from fossil fuel combustion. Annex 3.2 presents the total emissions from
28 all transportation and mobile sources, including CO2, N20, CH4, and HFCs.
20 Note that these totals include CO2, CH4 and N2O emissions from some sources in the U.S. Territories (ships and boats,
recreational boats, non-transportation mobile sources) and CH4 and N2O emissions from transportation rail electricity.
21 VMT estimates are based on data from FHWA Highway Statistics Table VM-1 (FHWA 1996 through 2017). Table VM-1 data
for 2016 has not been published yet, therefore 2016 mileage data is estimated using the 1.7 percent increase in FHWA Traffic
Volume Trends from 2015 to 2016. 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 2016 time period. In absence of these method changes,
light-duty VMT growth between 1990 and 2016 would likely have been even higher.
22 Commercial aircraft, as modeled in FAA's AEDT (FAA 2017), consists of passenger aircraft, cargo, and other chartered
flights.
3-20 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Figure 3-12: Fuels Used in Transportation Sector (TBtu), Onroad VMT, and Total Sector CO2
Emissions
¦ Aviation Gasoline (TBtu)
30,000 ¦ LPG (TBtu)
¦ Residual Fuel (TBtu)
I Natural Gas (TBtu)
Renewable Energy Sources (TBtu)
Jet Fuel
Distillate Fuel (TBtu)
I Motor Gasoline (TBtu)
Onroad VMT (Index vs. 1990) [Right Axis]
I Sector CO; Emissions (Index vs. 1990) [Right Axis]
160
140
co eh o h f\i n t
o o I -
Transportation Fossil Fuel Combustion CO 2 Emissions
Domestic transportation CO2 emissions increased by 22 percent (328.2 MMT CO2) between 1990 and 2016, an
annualized increase of 0.8 percent. Among domestic transportation sources in 2016, light-duty vehicles (including
passenger cars and light-duty trucks) represented 59 percent of CO2 emissions from fossil fuel combustion, medium-
and heavy-duty trucks and buses 24 percent, commercial aircraft 7 percent, and other sources 10 percent. See Table
3-12 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 from the transportation
sector lias increased from 0.7 billion gallons in 1990 to 13.5 billion gallons in 2016, while biodiesel consumption
has increased from 0.01 billion gallons in 2001 to 2.1 billion gallons in 2016. For further information, see Section
3.11 on biofuel consumption at the end of this chapter and Table A-96 in Annex 3.2.
Carbon dioxide emissions from passenger cars and light-duty trucks totaled 1,064.7 MMT CO2 in 2016. This is an
increase of 15 percent (140.2 MMT CO2) from 1990 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 2016). Carbon dioxide emissions from passenger
cars and light-duty trucks peaked at 1,150.6 MMT CO2 in 2004, and since then have declined about 7 percent. The
decline in new light-duty vehicle fuel economy between 1990 and 2004 (Figure 3-13) 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
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 .
Energy 3-21
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
modestly for much of the period. Light-duty VMT grew by less than one percent or declined each year between
2005 and 201324 and has since grown at a faster rate (2.6 percent from 2014 to 2015, and 1.7 percent from 2015 to
2016). 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 43 percent.
Light-duty truck share is about 38 percent of new vehicles in model year 2016 (EPA 2016a). See also Annex 3.2 for
data by vehicle mode and information on VMT and the share of new vehicles (in VMT).
Medium- and heavy-duty truck CO2 emissions increased by 83 percent from 1990 to 2016. This increase was largely
due to a substantial growth in medium- and heavy-duty truck VMT, which increased by 98 percent between 1990
and 20 1 6.25 Carbon dioxide from the domestic operation of commercial aircraft increased by 8 percent (9.1 MMT
CO2) from 1990 to 2016.26 Across all categories of aviation, excluding international bunkers, CO2 emissions
decreased by 9 percent (17.7 MMT CO2) between 1990 and 2016 27 This includes a 65 percent (22.8 MMT CO2)
decrease in CO2 emissions from domestic military operations.
Transportation sources also produce CH4 and N20; these emissions are included in Table 3-13 and Table 3-14 and in
the CH4 and N20 from Mobile Combustion section. Annex 3.2 presents total emissions from all transportation and
mobile sources, including CO2, CH4, N20, and HFCs.
24 VMT estimates are based on data from FHWA Highway Statistics Table VM-1 (FHWA 1996 through 2017). Table VM-1
data for 2016 has not been published yet, therefore 2016 mileage data is estimated using the 1.7 percent increase in FHWA
Traffic Volume Trends from 2015 to 2016. 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 2016 time period. In absence of these method changes, light-duty VMT
growth between 2006 and 2007 would likely have been higher.
25 While FHWA data shows consistent growth in medium- and heavy-duty truck VMT over the 1990 to 2016 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 2016 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 Commercial aircraft, as modeled in FAA's AEDT, consists of passenger aircraft, cargo, and other chartered flights.
27 Includes consumption of jet fuel and aviation gasoline. Does not include aircraft bunkers, which are not included in national
emission totals, in line with IPCC methodological guidance and UNFCCC reporting obligations.
3-22 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Figure 3-13: Sales-Weighted Fuel Economy of New Passenger Cars and Light-Duty Trucks,
2 1990-2016 (miles/gallon)
30.0.
29.0-
28.0-
27.0-
26,0-
c 25.0-
o 24.0-
23.0-
22.0-
21.0-
20.0-
19.0-
18.0-
17.0"
o-^rMro^-mvDr^oochO-rHfNm'g-cnuDf,^oo\D
4 Source: EPA (2016a)
5
6 Figure 3-14: Sales of New Passenger Cars and Light-Duty Trucks, 1990-2016 (Percent)
100%
75%-
% Passenger
° 50%-
% Light-Duty Trucks
25%-
0%
8 Source: EPA (2016a)
Energy 3-23
-------
1
2 Table 3-12: CO2 Emissions from Fossil Fuel Combustion in Transportation End-Use Sector
3 (MMT COz Eq.)
Fuel/Vehicle Type
1990
2005
2012a
2013a
2014a
2015a
2016a
Gasolineb
956.9
1,152.4
1,029.8
1,030.2
1,072.0
1,070.5
1,102.7
Passenger Cars
604.4
638.3
707.2
706.9
725.4
731.3
753.6
Light-Duty Trucks
300.6
464.4
268.2
268.3
290.2
283.2
291.8
Medium- and Heavy-Duty
Trucksc
37."
33.9
37.4
38.2
39.5
39.3
40.5
Buses
0.3
0.4
0.8
0.8
0.9
0.9
0.9
Motorcycles
ir
1.6
3.9
3.7
3.7
3.7
3.8
Recreational Boats'1
12.3
13.8
12.3
12.2
12.2
12.2
12.1
Distillate Fuel Oil (Diesel)b
262.9
457.5
427.5
433.9
446.3
459.8
466.1
Passenger Cars
7.9
4.2
4.1
4.1
4.1
4.3
4.4
Light-Duty Trucks
11.5
25.8
12.9
12.9
13.8
13.9
14.2
Medium- and Heavy-Duty
Trucksc
190.5
360.2
344.4
350.0
360.0
368.6
377.4
Buses
8.0
10.6
15.4
15.5
16.8
17.3
17.7
Rail
35.5
45.5
39.5
40.1
41.5
39.8
36.7
Recreational Boats
2.0
3.2
3.7
3.7
3.8
3.9
4.0
Ships and Non-Recreational
Boats6
7.5
8.0
7.5
7.5
6.1
12.0
11.6
International Bunker Fuel/
11.'
9.4
6.8
5.6
6.1
8.4
8.7
Jet Fuel
184.2
189.3
143.4
147.1
148.4
157.6
168.2
Commercial Aircraft8
109.9
132.7
113.3
114.3
115.2
119.0
119.0
Military Aircraft
35.0
19.4
12.1
11.0
14.0
13.5
12.2
General Aviation Aircraft
39.4
37.3
18.0
21.8
19.2
25.1
37.0
International Bunker Fuel/
38.0
60.1
64.5
65.7
69.6
71.9
71.9
International Bunker Fuels
from Commercial Aviation
30.0
55.6
61.4
62.8
66.3
68.6
68.6
Aviation Gasoline
3.1
2.4
1.7
1.5
1.5
1.5
1.4
General Aviation Aircraft
3.1
2.4
1.7
1.5
1.5
1.5
1.4
Residual Fuel Oil
22.6
19.3
15.8
15.1
5.8
4.2
13.4
Ships and Boats6
22.6
19.3
15.8
15.1
5.8
4.2
13.4
International Bunker Fuel/
53.'
43.6
34.5
28.5
27.7
30.6
33.8
Natural Gas J
36.0
33.1
41.3
47.0
40.3
39.5
40.6
Passenger Cars
+ /J
+
+
+
+
+
+
Light-Duty Trucks
+
+
+
+
+
+
+
Medium- and Heavy-Duty
Trucks
+
+
+
+
+
+
+
Buses
+ :i:;
0.6
0.8
0.8
0.8
0.9
1.0
Pipeline11
36.0
32.4
40.5
46.2
39.4
38.5
39.6
LPGJ
1.4
1.7
2.3
2.7
2.9
2.5
2.5
Passenger Cars
+
+
+
+
+
0.2
0.4
Light-Duty Trucks
0.2
0.3
0.2
0.3
0.6
0.4
0.2
Medium- and Heavy-Duty
Trucks6
1.1
1.3
1.8
2.1
1.9
1.6
1.6
Buses
0.1
0.1
0.3
0.4
0.3
0.3
0.2
Electricity
3.0
4.7
3.9
4.0
4.1
3.7
3.5
Rail
3.0
4.7
3.9
4.0
4.1
3.7
3.5
Totalk
1,470.2
1,860.5
1,665.8
1,681.6
1,721.2
1,739.2
1,798.4
Total (Including Bunkers)'
1,573.7
1,973.6
1,771.6
1,781.4
1,824.6
1,850.1
1,912.8
Biofuels-Ethanol'
4.1
22.4
71.5
73.4
74.9
75.9
78.2
Biofuels-Biodiesel'
0
0.9
8.5
13.5
13.3
14.1
19.6
4 + Does not exceed 0.05 MMT CO2 Eq.
3-24 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
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 2016 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-21, MF-27, and VM-1 (FHWA 1996 through 2017). Table VM-1 data for 2016 has not been published yet, therefore
2016 mileage data is estimated using the 1.7 percent increase in FHWA Traffic Volume Trends from 2015 to 2016. 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. 1 through A.6
(DOE 1993 through 2016). TEDB data for 2015 and 2016 has not been published yet, therefore 2014 data is used as a proxy.
c Includes medium- and heavy-duty trucks over 8,500 lbs.
d In 2014, EPA incorporated the NONROAD2008 model into MOVES2014. The current Inventory uses the NONROAD
component ofMOVES2014a foryears 1999 through 2016.
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.
B Commercial aircraft, as modeled in FAA's AEDT, consists of passenger aircraft, cargo, and other chartered flights.
h Pipelines reflect CO2 emissions from natural gas powered pipelines transporting natural gas.
'Ethanol andbiodiesel estimates are presented for informational purposes only. See Section 3.11 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.
J Transportation sector natural gas and LPG consumption are based on data fromEIA (2017). Prior to the previous (i.e., 1990
through 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 previous Inventory and apply to the
1990 to 2016 time period.
k Includes emissions from rail electricity.
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 CO2 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 andN2O Emissions
Mobile combustion includes emissions of CH4 and N20 from all transportation sources identified in the U.S.
Inventory with the exception of pipelines and electric locomotives;28 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.). 29 Annex 3.2 includes a summary of all emissions from both transportation
28 Emissions of CH4 from natural gas systems are reported separately. More information on the methodology used to calculate
these emissions are included in this chapter and Annex 3.4.
29 See the methodology sub-sections of the CO2 from Fossil Fuel Combustion and CH4 and N2O from Mobile Combustion
sections of this chapter. Note that N2O and CH4 emissions are reported using different categories than CO2. CO2 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. CO2 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 CO2 emissions
are included in the Industrial end-use sector instead of the Transportation end-use sector). CH4 and N2O emissions are reported
using the "Mobile Combustion" category, which includes non-transportation mobile sources. CH4 and N2O 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, CO2 emissions from non-
transportation mobile sources are presented separately from their overall end-use sector in Annex 3.2.
Energy 3-25
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
and mobile sources. Table 3-13 and Table 3-14 provide mobile fossil fuel CH4 and N20 emission estimates in MMT
C02 Eq.30
Mobile combustion was responsible for a small portion of national CH4 emissions (0.5 percent) but was the fourth
largest source of U.S. N20 emissions (4.8 percent). From 1990 to 2016, mobile source CH4 emissions declined by
70 percent, to 3.0 MMT CO: Eq. (119 kt CH4), due largely to control technologies employed in on-road vehicles
since the mid-1990s to reduce CO, NOx, NMVOC, and CH4 emissions. Mobile source emissions of N20 decreased
by 57 percent, to 17.8 MMT C02 Eq. (60 kt N20). Earlier generation control technologies initially resulted in higher
N20 emissions, causing a 30 percent increase in N20 emissions from mobile sources between 1990 and 1997.
Improvements in later-generation emission control technologies have reduced N20 output, resulting in a 67 percent
decrease in mobile source N20 emissions from 1997 to 2016 (Figure 3-15). Overall, CH4 and N20 emissions were
predominantly from gasoline-fueled passenger cars and light-duty trucks. See also Annex 3.2 for data by vehicle
mode and information on VMT and the share of new vehicles (in VMT).
Figure 3-15: Mobile Source ChU and N2O Emissions (MMT CO2 Eq.)
50-
40-
O"
LU
8 30-
H
2:
X
20-
10-
Table 3-13: ChU Emissions from Mobile Combustion (MMT CO2 Eq.)
Fuel Type/Vehicle Type3
1990
2005
2012
2013
2014
2015
2016
Gasoline On-Roadb
5.2
2.2
1.3
1.1
1.0
0.9
0.8
Passenger Cars
3.2
1.3
0.9
0.8
0.7
0.6
0.6
Light-Duty Trucks
1.7
0.8
0.3
0.3
0.2
0.2
0.2
Medium- and Heavy-Duty
Trucks and Buses
0.3
0.1
0.1
0.1
+
+
+
Motorcycles
+
+
+
+
+
+
+
Diesel On-Roadb
+
+
+
+
+
+
+
Passenger Cars
+
+
+
+
+
+
+
Light-Duty Trucks
+
+
+
+
+
+
+
Medium- and Heavy-Duty
Trucks and Buses
+
+
+
+
+
+
+
Alternative Fuel On-Road
+
0.2
0.3
0.3
0.2
0.3
0.3
Non-Roadc
4.6
4.2
2.4
2.3
2.1
1.9
1.8
Ships and Boats
0.5
0.5
0.4
0.4
0.3
0.3
0.3
30 See Annex 3.2 for a complete time series of emission estimates for 1990 through 2016.
3-26 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
Rail
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Aircraft
0.1
0.1
+
+
+
+
+
Agricultural Equipment"1
0.4
0.4
0.2
0.2
0.2
0.1
0.1
Construction/Mining
Equipment6
0.4
0.3
0.2
0.3
0.2
0.1
0.1
Otherf
3.1
2.8
1.5
1.4
1.3
1.2
1.2
Total
9.8
6.6
4.0
3.7
3.4
3.1
3.0
+ Does not exceed 0.05 MMT CO2 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 (FHWA 1996 through 2017). Table VM-1 data for 2016 has not been published yet, therefore 2016
mileage data is estimated using the 1.7 percent increase in FHWA Traffic Volume Trends from 2015 to 2016.
These mileage 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 2016). TEDB data for 2015 and 2016 has not been published yet,
therefore 2014 data is used as a proxy for the Public Review draft.
c Rail emissions do not include emissions from electric powered locomotives. Class II, Class III, commuter, and
intercity rail diesel consumption data for 2014 to 2016 are not available yet, therefore 2013 data is used as a
proxy for the Public Review draft.
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.
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 2016 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.
1 Table 3-14: N2O Emissions from Mobile Combustion (MMT CO2 Eq.)
Fuel Type/Vehicle Type3
19911
2005
2012
2013
2014
2015
2016
Gasoline On-Roadb
37.5
33.5
19.1
17.2
15.4
14.0
12.7
Passenger Cars
24.1
17.5
13.1
11.8
10.5
9.7
8.9
Light-Duty Trucks
12.8
15.0
5.3
4.7
4.4
3.8
3.4
Medium- and Heavy-Duty
Trucks and Buses
0.5
0.9
0.7
0.6
0.5
0.5
0.4
Motorcycles
-
+
+
+
+
+
Diesel On-Roadb
0.2
0.3
0.4
0.4
0.4
0.4
0.4
Passenger Cars
-
-
+
+
+
+
+
Light-Duty Trucks
-
+
+
+
+
+
Medium- and Heavy-Duty
Trucks and Buses
0.2
0.3
0.4
0.4
0.4
0.4
0.4
Alternative Fuel On-Road
+
+
+
+
+
+
+
Non-Road
3.S
4.6
4.4
4.5
4.3
4.5
4.7
Ships and Boats
0.6
0.6
0.5
0.5
0.3
0.4
0.5
Rail0
0.3
0.3
0.3
0.3
0.3
0.3
0.3
Aircraft
1."
1.8
1.3
1.4
1.4
1.5
1.6
Agricultural Equipment"1
0.4
0.6
0.7
0.7
0.7
0.7
0.7
Construction/Mining
Equipment6
0.5
0.8
0.9
0.9
0.9
0.9
0.9
Otherf
0.4
0.6
0.7
0.7
0.7
0.7
0.7
Total
41.5
38.4
23.8
22.0
20.2
18.8
17.8
Energy 3-27
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
+ Does not exceed 0.05 MMT CO2 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 (FHWA 1996 through 2017). Table VM-1 data for 2016 has not been published yet, therefore 2016 mileage data is
estimated using the 1.7 percent increase in FHWA Traffic Volume Trends from 2015 to 2016. These mileage
estimates are combined with estimates of fuel shares by vehicle type from DOE's TEDB Annex Tables A. 1 through
A.6 (DOE 1993 through 2016). TEDB data for 2015 and 2016 has not been published yet, therefore 2014 data is used
as a proxy for the Public Review draft.
c Rail emissions do not include emissions from electric powered locomotives. Class II, Class III, commuter, and
intercity rail diesel consumption data for 2014 to 2016 are not available yet, therefore 2013 data is used as a proxy for
the Public Review draft.
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.
Note: 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 2016 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.
C02 from Fossil Fuel Combustion
Methodology
CO2 emissions from fossil fuel combustion are estimated in line with a Tier 2 method described by the IPCC in the
2006 IPCC Guidelines for National Greenhouse Gas Inventories (IPCC 2006) with some exceptions as discussed
below.31 A detailed description of the U.S. methodology is presented in Annex 2.1, and is characterized by the
following steps:
1. Determine total fuel consumption by fuel type and sector. Total fossil fuel consumption for each year is
estimated by aggregating consumption data by end-use sector (e.g., commercial, industrial, etc.), primary
fuel type (e.g., coal, petroleum, gas), and secondary fuel category (e.g., motor gasoline, distillate fuel oil,
etc.). 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 2017a). The EIA does not
include territories in its national energy statistics, so fuel consumption data for territories were collected
separately fromEIA's International Energy Statistics (EIA 2017b).32
For consistency of reporting, the IPCC has recommended that countries report energy data using the
International Energy Agency (IEA) reporting convention and/or IEA data. Data in the IEA format are
presented "top down"—that is, energy consumption for fuel types and categories are estimated from energy
production data (accounting for imports, exports, stock changes, and losses). The resulting quantities are
referred to as "apparent consumption." The data collected in the United States by EIA on an annual basis
and used in this Inventory are predominantly from mid-stream or conversion energy consumers such as
refiners and electric power generators. These annual surveys are supplemented with end-use energy
consumption surveys, such as the Manufacturing Energy Consumption Survey, that are conducted on a
31 The IPCC Tier 3B methodology is used for estimating emissions from commercial aircraft.
32 Fuel consumption by U.S. Territories (i.e., American Samoa, Guam, Puerto Rico, U.S. Virgin Islands, Wake Island, and other
U.S. Pacific Islands) is included in this report and contributed total emissions of 41.4 MMT CO2 Eq. in 2016.
3-28 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
periodic basis (every four years). These consumption data sets help inform the annual surveys to arrive at
the national total and sectoral breakdowns for that total.33
Also, note that U.S. fossil fuel energy statistics are generally presented using gross calorific values (GCV)
(i.e., higher heating values). Fuel consumption activity data presented here have not been adjusted to
correspond to international standards, which are to report energy statistics in terms of net calorific values
(NCV) (i.e., lower heating values).34
2. 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 2016), Coffeyville (2012), U.S.
Census Bureau (2001 through 2011), EIA (2017a, 2017c, 2017d), USAA (2008 through 2017), USGS
(1991 through 2015a), (USGS 2016a), USGS (2014 through 2016a), USGS (2014 through 2016b), USGS
(1995 through 2013), USGS (1995, 1998, 2000, 2001), USGS (2017), USGS (1991 through 2013), USGS
(2016d), USGS (2015b), USGS (2014), USGS (1996 through 2013), USGS (1991 through 2015b), USGS
(2015 and 2016), USGS (1991 through 2015c).35
3. Adjust for conversion offuels and exports of CO 2. Fossil fuel consumption estimates are adjusted
downward to exclude fuels created from other fossil fuels and exports of CO2.36 Synthetic natural gas is
created from industrial coal, and is currently included in EIA statistics for both coal and natural gas.
Therefore, synthetic natural gas is subtracted from energy consumption statistics.37 Since October 2000, the
Dakota Gasification Plant has been exporting CO2 to Canada by pipeline. Since this CO2 is not emitted to
the atmosphere in the United States, the associated fossil fuel burned to create the exported CO2 is
subtracted from fossil fuel consumption statistics. 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, additional data for ethanol and biodiesel
were collected from EIA (2017a), data for synthetic natural gas were collected from EIA (2017d), and data
for CO2 exports were collected from the Eastman Gasification Services Company (2011), Dakota
Gasification Company (2006), Fitzpatrick (2002), Erickson (2003), EIA (2008) and DOE (2012).
4. Adjust Sectoral Allocation of Distillate Fuel Oil and Motor Gasoline. EPA had 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 was 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 2017), Benson
33 See IPCC Reference Approach for Estimating CO2 Emissions from Fossil Fuel Combustion in Annex 4 for a comparison of
U.S. estimates using top-down and bottom-up approaches.
34 A crude convention to convert between gross and net calorific values is to multiply the heat content of solid and liquid fossil
fuels by 0.95 and gaseous fuels by 0.9 to account for the water content of the fuels. Biomass-based fuels in U.S. energy statistics,
however, are generally presented using net calorific values.
35 See sections on Iron and Steel Production and Metallurgical Coke Production, Ammonia Production and Urea Consumption,
Petrochemical Production, Titanium Dioxide Production, Ferroalloy Production, Aluminum Production, and Silicon Carbide
Production and Consumption in the Industrial Processes and Product Use chapter.
36 Energy statistics from EIA (2017a) are already adjusted downward to account for ethanol added to motor gasoline, biodiesel
added to diesel fuel, and biogas in natural gas.
37 These adjustments are explained in greater detail in Annex 2.1.
Energy 3-29
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
(2002 through 2004), DOE (1993 through 2016), EIA (2007), EIA (1991 through 2016), EPA (2017b), and
FHWA (1996 through 2017).38
5. Adjust for fuels consumed for non-energy uses. U.S. aggregate energy statistics include consumption of
fossil fuels for non-energy purposes. These are fossil fuels that are manufactured into plastics, asphalt,
lubricants, or other products. Depending on the end-use, this can result in storage of some or all of the C
contained in the fuel for a period of time. As the emission pathways of C used for non-energy purposes are
vastly different than fuel combustion (since the C in these fuels ends up in products instead of being
combusted), these emissions are estimated separately in Section 3.2 - Carbon Emitted and Stored in
Products from Non-Energy Uses of Fossil Fuels. Therefore, the amount of fuels used for non-energy
purposes was subtracted from total fuel consumption. Data on non-fuel consumption was provided by EIA
(2017a).
6. Subtract consumption of international bunker fuels. According to the UNFCCC reporting guidelines
emissions from international transport activities, or bunker fuels, should not be included in national totals.
U.S. energy consumption statistics include these bunker fuels (e.g., distillate fuel oil, residual fuel oil, and
jet fuel) as part of consumption by the transportation end-use sector, however, so emissions from
international transport activities were calculated separately following the same procedures used for
emissions from consumption of all fossil fuels (i.e., estimation of consumption, and determination of C
content).39 The Office of the Under Secretary of Defense (Installations and Environment) and the Defense
Logistics Agency Energy (DLA Energy) of the U.S. Department of Defense (DoD) (DLA Energy 2017)
supplied data on military jet fuel and marine fuel use. Commercial jet fuel use was obtained from FAA
(2017); residual and distillate fuel use for civilian marine bunkers was obtained from DOC (1991 through
2017) for 1990 through 2001 and 2007 through 2014, and DHS (2008) for 2003 through 2006.
Consumption of these fuels was subtracted from the corresponding fuels in the transportation end-use
sector. Estimates of international bunker fuel emissions for the United States are discussed in detail in
Section 3.10 - International Bunker Fuels.
7. Determine the total C 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. The C
content coefficients used by the United States were obtained from EIA's Emissions of Greenhouse Gases in
the United States 2008 (EIA 2009a), and an EPA analysis of C content coefficients developed for the
GHGRP (EPA 2010). A discussion of the methodology used to develop the C content coefficients are
presented in Annexes 2.1 and 2.2.
8. Estimate C02 Emissions. Total CO2 emissions are the product of the adjusted energy consumption (from
the previous methodology steps 1 through 6), the C 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).
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 (2017a) and USAF (1998).40
• For on-road vehicles, annual estimates of combined motor gasoline and diesel fuel consumption by
vehicle category were obtained fromFHWA (1996 through 2017); for each vehicle category, the
38 Bottom-up gasoline and diesel highway vehicle fuel consumption estimates are based on data from FHWA Highway Statistics
Table MF-21, MF-27, and VM-1 (FHWA 1996 through 2017).
39 See International Bunker Fuels section in this chapter for a more detailed discussion.
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 CO2) and International Bunker Fuels sections of the Energy chapter, Annex 3.2, and Annex 3.8,
respectively.
3-30 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
percent gasoline, diesel, and other (e.g., CNG, LPG) fuel consumption are estimated using data from
DOE (1993 through 2016).4142
• For non-road vehicles, activity data were obtained from AAR (2008 through 2017), APTA (2007
through 2016), APTA (2006)' BEA (2016), Benson (2002 through 2004), DOE (1993 through 2016),
DLA Energy (2017), DOC (1991 through 2017), DOT (1991 through 2016), EIA (2009a), EIA
(2017a), EIA (2017), EIA (1991 through 2016), EPA (2017b),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.
Box 3-4: Uses of Greenhouse Gas Reporting Program Data and Improvements in Reporting Emissions from
Industrial Sector Fossil Fuel Combustion
As described in the calculation methodology, total fossil fuel consumption for each year is based on aggregated end-
use sector consumption published by the EIA. The availability of facility-level combustion emissions through EPA's
GHGRP lias provided an opportunity to better characterize the industrial sector's energy consumption and emissions
in the United States, through a disaggregation of EIA's industrial sector fuel consumption data from select
industries.
For GHGRP 2010 through 2016 reporting years, facility-level fossil fuel combustion emissions reported through
EPA's GHGRP were categorized and distributed to specific industry types by utilizing facility-reported NAICS
codes (as published by the U.S. Census Bureau). As noted previously in this report, the definitions and provisions
for reporting fuel types in EPA's GHGRP include some differences from the Inventory's use of EIA national fuel
statistics to meet the UNFCCC reporting guidelines. The IPCC has provided guidance on aligning facility-level
reported fuels and fuel types published in national energy statistics, which guided this exercise.44
As with previous Inventory reports, the current effort represents an attempt to align, reconcile, and coordinate the
facility-level reporting of fossil fuel combustion emissions under EPA's GHGRP with the national-level approach
presented in this report. Consistent with recommendations for reporting the Inventory to the UNFCCC, progress was
made on certain fuel types for specific industries and has been included in the CRF tables that are submitted to the
41 Data from FHWA's Table VM-1 is used to estimate the share of fuel consumption between each on-road vehicle class. Since
VM-1 data for 2016 has not been published yet, fuel consumption shares from 2015 are used as a proxy for the current Inventory.
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 2016). TEDB data for 2016 has not been published yet, therefore 2015 data is used as a
proxy. 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 2015 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.
42 Transportation sector natural gas and LPG consumption are based on data from EIA (2017a). In previous Inventory years,
data from DOE 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 current Inventory and apply to the 1990 to 2015 time period.
43 In 2014, EPA incorporated the NONROAD2008 model into MOVES2014. The current Inventory uses the NONROAD
component of MOVES2014a for years 1999 through 2016.
44 See Section 4 "Use of Facility-Level Data in Good Practice National Greenhouse Gas Inventories" of the IPCC meeting report,
and specifically the section on using facility-level data in conjunction with energy data, at .
Energy 3-31
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
UNFCCC along with this report.45 The efforts in reconciling fuels focus on standard, common fuel types (e.g.,
natural gas, distillate fuel oil, etc.) where the fuels inEIA's national statistics aligned well with facility-level
GHGRP data. For these reasons, the current information presented in the 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 2016 time series 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.
Box 3-5: Carbon Intensity of U.S. Energy Consumption
The amount of C emitted from the combustion of fossil fuels is dependent upon the C content of the fuel and the
fraction of that C that is oxidized. Fossil fuels vary in their average C 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.46 In general, the C
content per unit of energy of fossil fuels is the highest for coal products, followed by petroleum, and then natural
gas. The overall C intensity of the U.S. economy is thus dependent upon the quantity and combination of fuels and
other energy sources employed to meet demand.
Table 3-15 provides a time series of the C 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 C
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 3-15 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 C intensity, which is related to the large percentage of its energy
derived from natural gas for heating. The C intensity of the commercial sector lias 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 C intensity of the transportation sector was closely related to the C content of petroleum products (e.g.,
motor gasoline and jet fuel, both around 70 MMT CO2 Eq./EJ), which were the primary sources of energy. Lastly,
the electric power sector had the highest C intensity due to its heavy reliance on coal for generating electricity.
Table 3-15: Carbon Intensity from Direct Fossil Fuel Combustion by Sector (MMT CO2
Eq./QBtu)
Sector
1990
2005
2012
2013
2014
2015
2016
Residential3
57.4
56.6
55.5
55.3
55.4
55.5
55.4
Commercial3
59.6
57.7
56.3
56.1
55.8
57.2
56.7
Industrial3
64.4
64.5
62.3
62.1
61.6
61.2
60.7
Transportation3
71.1
71.4
71.5
71.4
71.5
71.5
71.5
Electric Powerb
87.3
85.8
79.9
81.3
81.2
78.1
76.9
U.S. Territories0
73.0
73.5
72.2
71.9
72.3
72.3
72.3
All Sectors0
73.0
73.5
70.9
70.9
70.7
69.7
69.2
3 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.
45 See .
46 One exajoule (E.T) is equal to 1018 joules or 0.9478 QBtu.
3-32 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
For the time period of 1990 through about 2008. the C 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 C intensity lias 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 in 2016 was approximately
10.5 percent below levels in 1990 (see Figure 3-16). To differentiate these estimates from those of Table 3-15, the C
intensity trend shown in Figure 3-16 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 2017).
Figure 3-16: U.S. Energy Consumption and Energy-Related CO2 Emissions Per Capita and Per
Dollar GDP
110
100
8 90
S 80
rH
X
-8
£ 70
60
50
CO:/Energy Consumption
Energy Consumption/capita
CO./capita
Energy Consumption/$GDP
COj/$ GDP
o-«-»fNro^*u^vor^oo^O'^r«jm^-mvor>»ooo>0'^-'fNro^ri/^vo
G^O^ONO^G^C^O^O^O^O^OOOO COC OO O »—« 4
0^0>(^a>^O^OGNO^(^COOCOOCOOOOCOOOOO
f>4 fNJ OJ Oi CM fN fN CM CM CM CM CM CM CM CM CM CM
C intensity estimates were developed using nuclear and renewable energy data from EIA (2017a), EPA (2010), and
fossil fuel consumption data as discussed above and presented in Annex 2.1.
Uncertainty and Time-Series Consistency
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).
Although 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
Energy 3-33
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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 problems 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.
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 3.10 -
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 120 input variables were modeled for CO2 from energy-related Fossil Fuel Combustion (including about 10
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.47 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.48
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).49 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 3-16. Fossil fuel
combustion CO2 emissions in 2016 were estimated to be between 4,868.8 and 5,202.9 MMT CO2 Eq. at a 95 percent
47 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.
48 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.
49 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.
3-34 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 confidence level. This indicates a range of 2 percent below to 5 percent above the 2016 emission estimate of 4,976.7
2 MMT C02 Eq.
3 Table 3-16: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Energy-
4 Related Fossil Fuel Combustion by Fuel Type and Sector (MMT CO2 Eq. and Percent)
2016 Emission Estimate Uncertainty Range Relative to Emission Estimate3
Fuel/Sector (MMT CO2 Eq.) (MMT CO2 Eq.) (%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Coal"
1,306.6
1,261.2
1,430.4
-3%
9%
Residential
NE
NE
NE
NE
NE
Commercial
2.3
2.2
2.7
-5%
15%
Industrial
59.0
56.1
68.3
-5%
16%
Transportation
NE
NE
NE
NE
NE
Electric Power
1,241.3
1,192.8
1,361.2
-4%
10%
U.S. Territories
4.0
3.5
4.8
-12%
19%
Natural Gasb
1,477.0
1,460.2
1,545.5
-1%
5%
Residential
238.3
231.5
255.0
-3%
7%
Commercial
170.3
165.5
182.3
-3%
7%
Industrial
478.8
464.4
513.3
-3%
7%
Transportation
40.6
39.5
43.5
-3%
7%
Electric Power
545.9
530.1
574.0
-3%
5%
U.S. Territories
3.0
2.6
3.5
-13%
17%
Petroleumb
2,192.7
2,056.6
2,320.3
-6%
6%
Residential
58.0
54.8
61.0
-5%
5%
Commercial
55.3
52.1
58.1
-6%
5%
Industrial
269.7
212.6
322.4
-21%
20%
Transportation
1,754.2
1,639.4
1,866.5
-7%
6%
Electric Power
21.2
20.0
23.2
-6%
9%
U.S. Territories
34.3
31.7
38.2
-8%
11%
Total (excluding Geothermal)1
" 4,976.3
4,868.4
5,202.4
-2%
5%
Geothermal
0.4
NE
NE
NE
NE
Total (including Geothermal)b'c 4,976.7
4,868.8
5,202.9
-2%
5%
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 CO2 emissions
from geothermal production.
5 Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
6 through 2016. Details on the emission trends through time are described in more detail in the Methodology section,
7 above.
8 QA/QC and Verification
9 A source-specific QA/QC plan for CO2 from fossil fuel combustion was developed and implemented consistent with
10 the 2006IPCC Guidelines and the Quality Assurance/Quality Control and Uncertainty Management Plan (QA/QC
11 Management Plan) referenced in this report and described further in Annex 8. This effort included a general (Tier 1)
12 analysis, as well as portions of a category-specific (Tier 2) analysis. The Tier 2 procedures that were implemented
13 involved checks specifically focusing on the activity data and methodology used for estimating CO2 emissions from
14 fossil fuel combustion in the United States. Emission totals for the different sectors and fuels were compared and
Energy 3-35
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
trends were investigated to determine whether any corrective actions were needed. Minor corrective actions were
taken.
Recalculations Discussion
The Energy Information Administration (EIA 2017a) updated energy consumption statistics across the time series
relative to the previous Inventory. EIA revised ethylene consumption in the industrial sector for the years 1990
through 2015, which had a significant impact on emissions from Liquefied Petroleum Gas (LPG) across the time
series (i.e., 1990 through 2015). EIA revised LPG consumption in the residential sector for the years 2010 through
2015, and in the commercial and transportation sectors for the years 2011 through 2015. EIA also revised 2014 and
2015 distillate fuel consumption in the residential, commercial, industrial, and transportation sectors, and 2015
natural gas consumption in the residential, commercial, transportation, and electric power sectors. Revisions to LPG
and distillate fuel consumption resulted in an average annual increase of 14.0 MMT CO2 Eq. (0,6 percent) in CO2
emissions from petroleum. Revisions to natural gas consumption resulted in an average annual increase of less than
0.5 MMT CO2 Eq. (less than 0.05 percent) in CO2 emissions from natural gas. Overall, these changes resulted in an
average annual increase of 14.0 MMT CO2 Eq. (0.3 percent) in CO2 emissions from fossil fuel combustion for the
period 1990 through 2015, relative to the previous Inventory.
In addition, changes were made to the historic allocation of gasoline to on-road and non-road applications. In 2016,
the Federal Highway Administration (FHWA) changed its methods for estimating the share of gasoline used in on-
road and non-road applications. Among other updates, FHWA included lawn and garden equipment as well as off-
road recreational equipment in its estimates of non-road gasoline consumption for the first time. This change created
a time-series inconsistency between the data reported for years 2015 and 2016 and previous years. To create a more
consistent time series of motor gasoline consumption and emissions data for the current Inventory, the historical
time series was modified. Specifically, the lawn, garden, and recreational vehicle gasoline consumption from EPA's
NONROAD model is subtracted from the highway motor gasoline consumption from FHWA Table MF-21 when
determining the total highway motor gasoline consumption foryears 1990 through 2014.
Planned Improvements
To reduce uncertainty of CO2 from fossil fuel combustion estimates for U.S. Territories, efforts will be made to
improve the quality of the U.S. Territories data, including through work with EIA and other agencies. This
improvement is part of an ongoing analysis and efforts to continually improve the CO2 from fossil fuel combustion
estimates. In addition, further expert elicitation may be conducted to better quantify the total uncertainty associated
with emissions from this source.
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, 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.50 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, efforts will be taken to incorporate updated industrial fuel consumption data from
50 See .
3-36 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 EIA's Manufacturing Energy Consumption Survey (MECS), with updated data for 2014. Additional work will look
2 at CO2 emissions from biomass to ensure they are separated in the facility-level reported data, and maintaining
3 consistency with national energy statistics provided by EIA. In implementing improvements and integration of data
4 from EPA's GHGRP, the latest guidance from the IPCC on the use of facility-level data in national inventories will
5 continue to be relied upon.51
6 An ongoing planned improvement is to develop improved estimates of domestic waterborne fuel consumption. The
7 Inventory estimates for residual and distillate fuel used by ships and boats is based in part on data on bunker fuel use
8 from the U.S. Department of Commerce. Domestic fuel consumption is estimated by subtracting fuel sold for
9 international use from the total sold in the United States. It may be possible to more accurately estimate domestic
10 fuel use and emissions by using detailed data on marine ship activity. The feasibility of using domestic marine
11 activity data to improve the estimates will continue to be investigated.
12 EPA received a comment from FHWA that the trend of decreasing electricity use in the transportation sector does
13 not align with increased sales of electric and plug-in hybrid vehicles. Electricity data is allocated between economic
14 sectors based on electricity sales data provided by the industry through EIA reports. The data for electricity used in
15 transportation only includes electricity used for railroads and railways. Electricity used to charge electric vehicles
16 would fall under other sectors like residential and commercial use associated with home and public charging
17 stations. As a planned improvement, EPA will look into the possibility of breaking out electricity used to charge
18 electric vehicles and report that electricity use under the transportation sector.
19 EPA will evaluate and potentially update methods for allocating motor gasoline consumption to the transportation,
20 industrial, and commercial sectors. In 2016, FHWA changed its methods for estimating the share of gasoline used in
21 on-road and non-road applications, creating a time-series inconsistency in the current Inventory between 2015 and
22 previous years.52 EPA will continue to explore approaches to address this inconsistency, including using MOVES
23 on-road fuel consumption output to define the percentage of the FHWA consumption totals (from MF -21) that are
24 attributable to "transportation", and applying that percentage to the EIA total. This would define gasoline
25 consumption from "transportation," such that the remainder would be defined as consumption by the industrial and
26 commercial sectors.
27 CH4 and N20 from Stationary Combustion
28 Methodology
29 Methane and N20 emissions from stationary combustion were estimated by multiplying fossil fuel and wood
30 consumption data by emission factors (by sector and fuel type for industrial, residential, commercial, and U.S.
31 Territories; and by fuel and technology type for the electric power sector). The electric power sector utilizes a Tier 2
32 methodology, whereas all other sectors utilize a Tier 1 methodology. The activity data and emission factors used are
33 described in the following subsections.
34 Industrial, Residential, Commercial, and U.S. Territories
35 National coal, natural gas, fuel oil, and wood consumption data were grouped by sector: industrial, commercial,
36 residential, and U.S. Territories. For the CH4 and N20 estimates, consumption data for each fuel were obtained from
37 EIA's Monthly Energy Review (EIA 2017a). Because the United States does not include territories in its national
38 energy statistics, fuel consumption data for territories were provided separately by EIA's International Energy
51 See .
52 The previous and new FHWA methodologies for estimating non-road gasoline are described in Off-Highway and Public-Use
Gasoline Consumption Estimation Models Used in the Federal Highway Administration, Publication Number FHWA-PL-17-012.
Energy 3-37
-------
1 Statistics (EIA 2017b).53 Fuel consumption for the industrial sector was adjusted to subtract out construction and
2 agricultural use, which is reported under mobile sources.54 Construction and agricultural fuel use was obtained from
3 EPA (2017b) and FHWA (1996 through 2016). Estimates for wood biomass consumption for fuel combustion do
4 not include wood wastes, liquors, municipal solid waste, tires, etc., that are reported as biomass by EIA. Tier 1
5 default emission factors for these three end-use sectors were provided by the 2006IPCC Guidelines for National
6 Greenhouse Gas Inventories (IPCC 2006). U.S. Territories' emission factors were estimated using the U.S. emission
7 factors for the primary sector in which each fuel was combusted.
8 Electric Power Sector
9 The electric power sector uses a Tier 2 emission estimation methodology as fuel consumption for the electric power
10 sector by control-technology type was obtained from EPA's Acid Rain Program Dataset (EPA 2017a). These
11 combustion technology- and fuel- use data were available by facility from 1996 to 2016. The Tier 2 emission factors
12 used are based in part on emission factors published by EPA, and EPA's Compilation of Air Pollutant Emission
13 Factors, AP-42 (EPA 1997) for combined cycle natural gas units.55
14 Since there was a difference between the EPA (2017a) and EIA (2017a) total fuel consumption estimates, the
15 remaining consumption from EIA (2017a) was apportioned to each combustion technology type and fuel
16 combination using a ratio of fuel consumption by technology type from 1996 to 2016.
17 Fuel consumption estimates were not available from 1990 to 1995 in the EPA (2017a) dataset, and as a result,
18 consumption was calculated using total electric power production from EIA (2017a) and the ratio of combustion
19 technology and fuel types from EPA (2017a). The consumption estimates from 1990 to 1995 were estimated by
20 applying the 1996 consumption ratio by combustion technology type to the total EIA consumption for each year
21 from 1990 to 1995. Emissions were estimated by multiplying fossil fuel and wood consumption by technology - and
22 fuel-specific Tier 2 country specific emission factors.
23 Lastly, there were significant differences between wood biomass consumption in the electric power sector between
24 the EPA (2017a) and EIA (2017a) datasets. The higher wood biomass consumption from EIA (2017a) in the electric
25 power sector was distributed to the residential, commercial, and industrial sectors according to their percent share of
26 wood biomass energy consumption calculated from EIA (2017a).
27 More detailed information on the methodology for calculating emissions from stationary combustion, including
28 emission factors and activity data, is provided in Annex 3.1.
29 Uncertainty and Time-Series Consistency
30 Methane emission estimates from stationary sources exhibit high uncertainty, primarily due to difficulties in
31 calculating emissions from wood combustion (i.e., fireplaces and wood stoves). The estimates of CH4 and N20
32 emissions presented are based on broad indicators of emissions (i.e., fuel use multiplied by an aggregate emission
33 factor for different sectors), rather than specific emission processes (i.e., by combustion technology and type of
34 emission control).
35 An uncertainty analysis was performed by primary fuel type for each end-use sector, using the IPCC-recommended
36 Approach 2 uncertainty estimation methodology, Monte Carlo Stochastic Simulation technique, with @RISK
37 software.
38 The uncertainty estimation model for this source category was developed by integrating the CH4 and N20 stationary
39 source inventory estimation models with the model for CO2 from fossil fuel combustion to realistically characterize
53 U.S. Territories data also include combustion from mobile activities because data to allocate territories' energy use were
unavailable. For this reason, CH4 and N2O emissions from combustion by U.S. Territories are only included in the stationary
combustion totals.
54 Though emissions from construction and farm use occur due to both stationary and mobile sources, detailed data was not
available to determine the magnitude from each. Currently, these emissions are assumed to be predominantly from mobile
sources.
55 Several of the U.S. Tier 2 emission factors were used in IPCC 2006 as Tier 1 emission factors.
3-38 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
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 N20 emission factors, based on the SAIC/EIA (2001) report.56 For these variables, the uncertainty
ranges were assigned to the input variables based on the data reported in SAIC/EIA (2001).57 However, the CH4
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 3-17. Stationary
combustion CH4 emissions in 2016 (including biomass) were estimated to be between 5.1 and 15.5 MMT CO2 Eq. at
a 95 percent confidence level. This indicates a range of 29 percent below to 115 percent above the 2016 emission
estimate of 7.2 MMT CO2 Eq.58 Stationary combustion N20 emissions in 2016 (including biomass) were estimated
to be between 14.2 and 27.5 MMT CO2 Eq. at a 95 percent confidence level. This indicates a range of 23 percent
below to 50 percent above the 2016 emission estimate of 18.4 MMT CO2 Eq.
Table 3-17: Approach 2 Quantitative Uncertainty Estimates for ChU and N2O Emissions from
Energy-Related Stationary Combustion, Including Biomass (MMT CO2 Eq. and Percent)
Source
Gas
2016 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Upper
Lower Upper
Bound
Bound
Bound Bound
Stationary Combustion
CH4
7.2
5.1
15.5
-29% +115%
Stationary Combustion
N2O
18.4
14.2
27.5
-23% +50%
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 CH4 and N20 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 CH4 and N20 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.
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2016 as discussed below. Details on the emission trends through time are described in more detail in the
Methodology section, above.
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 QA/QC and Verification Procedures section in the introduction of the
IPPU Chapter.
56 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.
57 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.
58 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 3-39
-------
1
QA/QC and Verification
2 A source-specific QA/QC plan for stationary combustion was developed and implemented consistent with the 2006
3 IPCC Guidelines and the QA/QC Management Plan referenced in this report and described further in Annex 8. This
4 effort included a general (Tier 1) analysis, as well as portions of a category-specific (Tier 2) analysis. The Tier 2
5 procedures that were implemented involved checks specifically focusing on the activity data and emission factor
6 sources and methodology used for estimating CH4, N20, and the indirect greenhouse gases from stationary
7 combustion in the United States. Emission totals for the different sectors and fuels were compared and trends were
8 investigated.
9 Recalculations Discussion
10 Methane and N20 emissions from stationary sources (excluding CO2) across the entire time series were revised due
11 to revised data from EIA (2017a), EIA (2017b), and EPA (2017a) relative to the previous Inventory. Methane and
12 N20 emission factors for combined cycle natural gas units were updated to be consistent with EPA's Compilation of
13 Air Pollutant Emission Factors, AP-42 (EPA 1997). In addition, the GWPs for CH4 and N2O for the Acid Rain
14 Program Dataset (EPA 2017a) were updated to be consistent with the IPCC Fourth Assessment Report (AIM)
15 values. The historical data changes resulted in an average annual increase 0.4 MMT CO2 Eq. (5.2 percent) in CH4
16 emissions, and an average annual decrease 2.3 MMT CO2 Eq. (12 percent) in N20 emissions from stationary
17 combustion for the 1990 through 2015 period.
18 Planned Improvements
19 Several items are being evaluated to improve the CH4 and N20 emission estimates from stationary combustion and
20 to reduce uncertainty for U.S. Territories. Efforts will be taken to work with EIA and other agencies to improve the
21 quality of the U.S. Territories data. Because these data are not broken out by stationary and mobile uses, further
22 research will be aimed at trying to allocate consumption appropriately. In addition, the uncertainty of bio mass
23 emissions will be further investigated since it was expected that the exclusion of biomass from the estimates would
24 reduce the uncertainty; and in actuality the exclusion of biomass increases the uncertainty. These improvements are
25 not all-inclusive, but are part of an ongoing analysis and efforts to continually improve these stationary combustion
26 estimates from U.S. Territories.
27 Fuel use was adjusted for the industrial sector to subtract out construction and agricultural use, which is reported
28 under mobile sources. Mobile source CH4 and N20 also include emissions from sources that may be captured as part
29 of the commercial sector. Future research will look into the need to adjust commercial sector fuel consumption to
30 account for sources included elsewhere.
31 CH4 and N20 from Mobile Combustion
32 Methodology
33 Estimates of CH4 and N20 emissions from mobile combustion were calculated by multiplying emission factors by
34 measures of activity for each fuel and vehicle type (e.g., light-duty gasoline trucks). Activity data included vehicle
35 miles traveled (VMT) for on-road vehicles and fuel consumption for non-road mobile sources. The activity data and
36 emission factors used are described in the subsections that follow. A complete discussion of the methodology used to
37 estimate CH4 and N20 emissions from mobile combustion and the emission factors used in the calculations is provided
38 in Annex 3.2.
3-40 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
On-Road Vehicles
Estimates of CH4 and N20 emissions from gasoline and diesel on-road vehicles are based on VMT and emission
factors 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 by vehicle and fuel type.59
Emissions factors for N20 from newer on-road gasoline vehicles were calculated based upon a regression analysis
done by EPA (Browning 2017). Methane emission factors were calculated based on the ratio of NMOG emission
standards for newer vehicles. Older gasoline vehicles on-road emissions factors were developed by ICF (2004).
These factors were derived from EPA, California Air Resources Board (CARB) and Environment Canada laboratory
test results of different vehicle and control technology types. The EPA, CARB and Environment Canada tests were
designed following the Federal Test Procedure (FTP), which 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 segment 2 were used to define running
emissions, and subtracted from the total FTP emissions to determine start emissions. These were then recombined
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,
to approximate average driving characteristics.60 Diesel on-road vehicle emission factors were developed by ICF
(2006b).
CH4 and N20 emission factors for AFVs were developed based on the 2016 GREET model. 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 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 operation only (tank-to-
wheels); well-to-tank emissions are calculated elsewhere in the Inventory.
Annual VMT data for 1990 through 2015 were obtained from the Federal Highway Administration's (FHWA)
Highway Performance Monitoring System database as reported in Highway Statistics (FHWA 1996 through
2016).61 VMT data in the VM-1 table for 2016 has not been published yet; therefore 2016 mileage data is estimated
using the 1.7 percent increase in FHWA Traffic Volume Trends from 2015 to 2016. 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 2017) and information on total motor
vehicle fuel consumption by fuel type from FHWA (1996 through 2017). VMT for AFVs were estimated based on
Browning (2017). The age distributions of the U.S. vehicle fleet were obtained from EPA (2017b, 2000), and the
average annual age-specific vehicle mileage accumulation of U.S. vehicles were obtained from EPA (2017b).
59 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.
60 Additional information regarding the MOBILE model can be found online at .
61 The source of VMT 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 2016 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.
Energy 3-41
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Control technology and standards data for on-road vehicles were obtained from EPA's Office of Transportation and
Air Quality (EPA 2007a, 2007b, 2000, 1998, and 1997) 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).
Non-Road Mobile Sources
To estimate 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 N20 and CH4 per kilogram of fuel
consumed).62 Activity data were obtained from AAR (2008 through 2016), APTA (2007 through 2016), APTA
(2006), BEA (1991 through 2015), Benson (2002 through 2004), DHS (2008), DLA Energy (2015), DOC (1991
through 2015), DOE (1993 through 2016), DOT (1991 through 2017), EIA (2002, 2007, 2016a), EIA (2007 through
2016), EIA (1991 through 2017), EPA (2017b), Esser (2003 through 2004), FAA (2017), FHWA (1996 through
20 1 7),63 Gaffney (2007), and Whorton (2006 through 2014). Emission factors for non-road modes were taken from
IPCC (2006) and Browning (2017).
Uncertainty and Time-Series Consistency-TO BE UPDATED FOR FINAL
INVENTORY REPORT
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 VvRISK
software. The uncertainty analysis was performed on 2015 estimates of CH4 and N20 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 and (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 since 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
CH4 and N20 emission factors due to limited emission test data, and because, unlike CO: emissions, the emission
pathways of CH4 and N20 are highly complex.
Mobile combustion CH4 emissions from all mobile sources in 2015 were estimated to be between 1.6 and 2.5 MMT
C02 Eq. at a 95 percent confidence level. This indicates a range of 18 percent below to 27 percent above the
corresponding 2015 emission estimate of 2.0 MMT C02 Eq. Also at a 95 percent confidence level, mobile
combustion N20 emissions from mobile sources in 2015 were estimated to be between 13.2 and 17.9 MMT C02
Eq., indicating a range of 13 percent below to 19 percent above the corresponding 2015 emission estimate of 15.1
MMT C02 Eq.
Table 3-18: Approach 2 Quantitative Uncertainty Estimates for ChU and N2O Emissions from
Mobile Sources (MMT CO2 Eq. and Percent)
Source Gas 2015 Emission Estimate3 Uncertainty Range Relative to Emission Estimate3
62 Hie consumption of international bunker fuels is not included in these activity data, but is estimated separately under the
International Bunker Fuels source category.
63 This Inventory uses FHWA's Agriculture, Construction, and Commercial/Industrial MF-24 fuel volumes along with the
MOVES NONROAD model gasoline volumes to estimate non-road mobile source CH4 and N2O emissions for these categories.
For agriculture, the MF-24 gasoline volume is used directly because it includes both off-road trucks and equipment. For
construction and commercial/industrial gasoline estimates, the 2014 and older MF-24 volumes represented off-road trucks only;
therefore, the MOVES NONROAD gasoline volumes for construction and commercial/industrial 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
NONROAD equipment gasoline volumes in the construction and commercial/industrial categories.
3-42 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Mobile Sources
CH4 2.0
1.6
2.5
-18%
+27%
Mobile Sources
N2O 15.1
13.2
17.9
-13%
+19%
a Range of emission
estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
1 This uncertainty analysis is a continuation of a multi-year process for developing quantitative uncertainty estimates
2 for this source category using the IPCC Approach 2 uncertainty analysis. As a result, as new information becomes
3 available, uncertainty characterization of input variables may be improved and revised. For additional information
4 regarding uncertainty in emission estimates for CH4 and N20 please refer to the Uncertainty Annex.
5 QA/QC and Verification
6 A source-specific Quality Assurance/Quality Control plan for mobile combustion was developed and implemented.
7 This plan is based on the IPCC-recommended QA/QC Plan. The specific plan used for mobile combustion was
8 updated prior to collection and analysis of this current year of data. This effort included a general (Tier 1) analysis,
9 as well as portions of a category-specific (Tier 2) analysis. The Tier 2 procedures focused on the emission factor and
10 activity data sources, as well as the methodology used for estimating emissions. These procedures included a
11 qualitative assessment of the emission estimates to determine whether they appear consistent with the most recent
12 activity data and emission factors available. A comparison of historical emissions between the current Inventory and
13 the previous Inventory was also conducted to ensure that the changes in estimates were consistent with the changes
14 in activity data and emission factors.
15 Recalculations Discussion
16 Updates were made to the on-road, non-road and alternative fuel CH4 and N20 emissions calculations this year
17 resulting in both increases and decreases to different source categories. Decreases in on-road gasoline emissions
18 were offset by large increases in alternative fuel and non-road emissions. The collective result of all of these changes
19 was a net increase in CH4 and N20 emissions from mobile combustion relative to the previous Inventory. CH4
20 emissions increased by 52.7 percent. N20 emissions increased by 24.5 percent. Each of these changes is described
21 below.
22 New emissions factors for N20 emissions were developed for on-road vehicles based on an EPA regression analysis
23 of the relationship between NOx and N20. New CH4 emission factors were calculated based on the ratio of NMOG
24 emission standards. These new emission factors allowed the inclusion of additional emissions standards, including
25 Federal Tier 3 emission standards and two levels of California emission standards (LEV II and LEV III) to the
26 control technology breakouts.
27 In addition new non-road emissions factors were developed. Previously emission factors were taken from the 1996
28 IPCC Guidelines and represented the IPCC Tier 1 factors. This year new emission factors were calculated using the
29 updated 2006 IPCC Tier 3 guidance and EPA's MOVES2014a model. CH4 emission factors were calculated directly
30 from MOVES. N20 emission factors were calculated using NONROAD activity and emission factors by fuel type
31 from the European Enviromnent Agency. Gasoline engines were broken out by 2- and 4-stroke engine types.
32 Equipment using liquefied petroleum gas (LPG) and compressed natural gas (CNG) were included.
33 New emission factors for alternative fuel vehicles were estimated using GREET 2016. The updated emission factors
34 have been generated for CH4 and N20. For light-duty trucks, EPA used a curve fit of 1999 through 2011 travel
35 fractions for LDT1 and LDT2 (MOVES Source Type 31 for LDT1 and MOVES Source Type 32 for LDT2). For
36 medium duty vehicles, EPA used emission factors for Light Heavy-Duty Vocational Trucks. For heavy-duty
37 vehicles, EPA used emission factors for Long Haul Combination Trucks. For Buses, EPA used emission factors for
38 Transit Buses. The emissions factors developed represent vehicle operation only (tank-to-wheels).
39 In addition changes were made to the historic allocation of gasoline to on-road and non-road applications. In 2016,
40 the Federal Highway Administration (FHWA) changed its methods for estimating the share of gasoline used in on-
41 road and non-road applications. Among other updates, FHWA included lawn and garden equipment as well as off-
Energy 3-43
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
road recreational equipment in its estimates of non-road gasoline consumption for the first time. This change created
a time-series inconsistency between the data reported for years 2015 and 2016 and previous years. To create a more
consistent time series of motor gasoline consumption and emissions data for the current Inventory, the historical
time series was modified. Specifically, the lawn, garden, and recreational vehicle gasoline consumption from EPA's
NONROAD model is subtracted from the highway motor gasoline consumption from FHWA Table MF-21 when
determining the total highway motor gasoline consumption for years 1990 through 2014.
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2016 with one recent notable exception. An update by FHWA to the method for estimating on-road VMT
created an inconsistency in on-road CH4 and N20 for the time periods 1990 to 2006 and 2007 to 2016. 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 have been
identified that could potentially be improved in the near term given available resources.
• Evaluate and potentially update EPA's method for estimating motor gasoline consumption for non-road
mobile sources to improve accuracy and create a more consistent time series. As discussed in the
Methodology section above and in Annex 3.2, CH4 and N20 estimates for gasoline-powered non-road
sources in this Inventory are based on a variety of inputs, including FHWA Highway Statistics Table MF-
24. In 2016, FHWA changed its methods for estimating the share of gasoline used in on-road and non-road
applications.64 These method changes created a time-series inconsistency in the current Inventory between
2015 and previous years in CH4 and N20 estimates for agricultural, construction, commercial, and
industrial non-road mobile sources. In the current Inventory EPA has implemented one approach to address
this inconsistency. EPA will test other approaches including using MOVES on-road fuel consumption
output to define the percentage of the FHWA consumption totals (from MF-21) that are attributable to
"transportation." This percentage would then be applied to the EIA total, thereby defining gasoline
consumption from "transportation," such that the remainder would be defined as consumption by the
industrial and commercial sectors.
• Explore updates to on-road diesel emissions factors for CH4 and N20 to incorporate diesel after treatment
technology for light-duty vehicles.
• 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. 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 continues to be
investigated. Additionally, the feasibility of including data from a broader range of domestic and
international sources for domestic bunker fuels, including data from studies such as the Third 1MO GHG
Study 2014, continues to be explored.
64 The previous and new FHWA methodologies for estimating non-road gasoline are described in Off-Highway and Public-Use
Gasoline Consumption Estimation Models Used in the Federal Highway Administration, Publication Number FHWA-PL-17-012.
3-44 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 3.2 Carbon Emitted from Non-Energy Uses of
2 Fossil Fuels (CRF Source Category 1A)
3 In addition to being combusted for energy, fossil fuels are also consumed for non-energy uses (NEU) in the United
4 States. The fuels used for these purposes are diverse, including natural gas, liquefied petroleum gases (LPG), asphalt
5 (a viscous liquid mixture of heavy crude oil distillates), petroleum coke (manufactured from heavy oil), and coal
6 (metallurgical) coke (manufactured from coking coal). The non-energy applications of these fuels are equally
7 diverse, including feedstocks for the manufacture of plastics, rubber, synthetic fibers and other materials; reducing
8 agents for the production of various metals and inorganic products; and non-energy products such as lubricants,
9 waxes, and asphalt (IPCC 2006). Emissions from non-energy uses of fossil fuels are reported in the Energy sector,
10 as opposed to the IPPU sector, to reflect national circumstances in its choice of methodology and to increase
11 transparency of this source category's unique country-specific data sources and methodology (see Box 3-6).
12 Carbon dioxide emissions arise from non-energy uses via several pathways. Emissions may occur during the
13 manufacture of a product, as is the case in producing plastics or rubber from fuel-derived feedstocks. Additionally,
14 emissions may occur during the product's lifetime, such as during solvent use. Overall, throughout the time series
15 and across all uses, about 61 percent of the total C consumed for non-energy purposes was stored in products, and
16 not released to the atmosphere; the remaining 39 percent was emitted.
17 There are several areas in which non-energy uses of fossil fuels are closely related to other parts of this Inventory.
18 For example, some of the NEU products release CO2 at the end of their commercial life when they are combusted
19 after disposal; these emissions are reported separately within the Energy chapter in the Incineration of Waste source
20 category. In addition, there is some overlap between fossil fuels consumed for non-energy uses and the fossil-
21 derived CO2 emissions accounted for in the Industrial Processes and Product Use chapter, especially for fuels used
22 as reducing agents. To avoid double-counting, the "raw" non-energy fuel consumption data reported by EIA are
23 modified to account for these overlaps. There are also net exports of petrochemicals that are not completely
24 accounted for in the EIA data, and the Inventory calculations adjust for the effect of net exports on the mass of C in
25 non-energy applications.
26 As shown in Table 3-19, fossil fuel emissions in 2016 from the non-energy uses of fossil fuels were 121.0 MMT
27 CO2 Eq., which constituted approximately 2 percent of overall fossil fuel emissions. In 2016, the consumption of
28 fuels for non-energy uses (after the adjustments described above) was 4,844.4 TBtu (see Table 3-20). A portion of
29 the C in the 4,844.4 TBtu of fuels was stored (207.7 MMT CO2 Eq.), while the remaining portion was emitted
30 (121.0 MMT C02Eq.).
31 Table 3-19: CO2 Emissions from Non-Energy Use Fossil Fuel Consumption (MMT CO2 Eq. and
32 Percent)
Year
1990
2005
2012
2013
2014
2015
2016
Potential Emissions
312.1
377.5
312.6
329.3
323.8
339.6
328.7
C Stored
192.5
235.9
199.4
196.2
196.0
204.5
207.7
Emissions as a % of Potential
38%
38%
36%
40%
39%
40%
37%
Emissions
119.6
141.7
113.3
133.2
127.8
135.1
121.0
33 Methodology
34 The first step in estimating C stored in products was to determine the aggregate quantity of fossil fuels consumed for
35 non-energy uses. The C content of these feedstock fuels is equivalent to potential emissions, or the product of
36 consumption and the fuel-specific C content values. Both the non-energy fuel consumption and C content data were
3 7 supplied by the EIA (2013, 2016) (see Annex 2.1). Consumption of natural gas, LPG, pentanes plus, naphthas, other
38 oils, and special naphtha were adjusted to subtract out net exports of these products that are not reflected in the raw
39 data from EIA. Consumption values for industrial coking coal, petroleum coke, other oils, and natural gas in Table
40 3-20 and Table 3-21 have been adjusted to subtract non-energy uses that are included in the source categories of the
Energy 3-45
-------
1 Industrial Processes and Product Use chapter.65-66 Consumption values were also adjusted to subtract net exports of
2 intermediary chemicals.
3 For the remaining non-energy uses, the quantity of C stored was estimated by multiplying the potential emissions by
4 a storage factor.
5 • For several fuel types—petrochemical feedstocks (including natural gas for non-fertilizer uses, LPG,
6 pentanes plus, naphthas, other oils, still gas, special naphtha, and industrial other coal), asphalt and road oil,
7 lubricants, and waxes—U.S. data on C stocks and flows were used to develop C storage factors, calculated
8 as the ratio of (a) the C stored by the fuel's non-energy products to (b) the total C content of the fuel
9 consumed. A lifecycle approach was used in the development of these factors in order to account for losses
10 in the production process and during use. Because losses associated with municipal solid waste
11 management are handled separately in the Energy sector under the Incineration of Waste source category,
12 the storage factors do not account for losses at the disposal end of the life cycle.
13 • For industrial coking coal and distillate fuel oil, storage factors were taken from IPCC (2006), which in turn
14 draws from Marland and Rotty (1984).
15 • For the remaining fuel types (petroleum coke, miscellaneous products, and other petroleum), IPCC does not
16 provide guidance on storage factors, and assumptions were made based on the potential fate of C in the
17 respective NEU products.
18 Table 3-20: Adjusted Consumption of Fossil Fuels for Non-Energy Uses (TBtu)
Year
1990
2005
2012
2013
2014
2015
2016
Industry
4,215.!S
5,110.7
4,373.3
4,627.5
4,591.4
4,762.0
4,626.6
Industrial Coking Coal
0.0
80.4
132.5
119.3
48.8
121.8
88.6
Industrial Other Coal
8.2
11.9
10.3
10.3
10.3
10.3
10.3
Natural Gas to Chemical Plants
281.6
260.9
292.7
297.1
305.1
302.2
289.5
Asphalt & Road Oil
1,170.2
1,323.2
826.7
783.3
792.6
831.7
853.4
LPG
1,120.5
1,610.0
1,883.4
2,069.2
2,103.4
2,160.0
2,117.6
Lubricants
186.:-
160.2
130.5
138.1
144.0
156.8
148.9
Pentanes Plus
117.6
95.5
40.3
45.4
43.5
78.4
53.0
Naphtha (<401 °F)
326.}
679.5
432.2
498.8
435.2
417.8
396.6
Other Oil (>401 °F)
662.1
499.4
267.4
209.1
236.2
216.8
203.8
Still Gas
36."
67.7
160.6
166.7
164.5
162.2
166.1
Petroleum Coke
27.2
105.2
0.0
0.0
0.0
0.0
0.0
Special Naphtha
100.9
60.9
14.1
96.6
104.4
97.0
88.7
Distillate Fuel Oil
7.0
11.7
5.8
5.8
5.8
5.8
5.8
Waxes
33.3
31.4
15.3
16.5
14.8
12.4
12.9
Miscellaneous Products
137.8
112.8
161.6
171.2
182.7
188.9
191.3
Transportation
176.0
151.3
123.2
130.4
136.0
148.1
140.6
Lubricants
176.0
151.3
123.2
130.4
136.0
148.1
140.6
U.S. Territories
85/.
123.2
72.0
82.4
77.3
77.3
77.3
Lubricants
0."
4.6
1.0
1.0
1.0
1.0
1.0
Other Petroleum (Misc. Prod.)
84.9
118.6
71.0
81.4
76.2
76.2
76.2
Total
4,477.4
5,385.2
4,568.5
4,840.3
4,804.7
4,987.4
4,844.5
65 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.
66 Some degree of double counting may occur between these estimates of non-energy use of fuels and process emissions from
petrochemical production presented in the Industrial Processes and Produce Use sector. Data integration is not feasible at this
time as feedstock data from EIA used to estimate non-energy uses of fuels are aggregated by fuel type, rather than disaggregated
by both fuel type and particular industries (e.g., petrochemical production) as currently collected through EPA's GHGRP and
used for the petrochemical production category.
3-46 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Table 3-21: 2016 Adjusted Non-Energy Use Fossil Fuel Consumption, Storage, and Emissions
Adjusted
Carbon
Non-Energy
Content
Potential
Storage
Carbon
Carbon
Carbon
Use3
Coefficient
(MMT
Carbon
Factor
Stored
Emissions
Emissions
(MMT
Sector/Fuel Type
(TBtu)
C/QBtu)
(MMT C)
(MMT C)
(MMT C)
CO2 Eq.)
Industry
4,626.6
NA
85.3
NA
56.2
29.0
106.5
Industrial Coking Coal
88.6
31.00
2.7
0.10
0.3
2.5
9.1
Industrial Other Coal
10.3
25.82
0.3
0.66
0.2
0.1
0.3
Natural Gas to
Chemical Plants
289.5
14.47
4.2
0.66
2.8
1.4
5.2
Asphalt & Road Oil
853.4
20.55
17.5
1.00
17.5
0.1
0.3
LPG
2,117.6
17.06
36.1
0.66
23.8
12.3
45.2
Lubricants
148.9
20.20
3.0
0.09
0.3
2.7
10.0
Pentanes Plus
53.0
19.10
1.0
0.66
0.7
0.3
1.3
Naphtha (<401° F)
396.6
18.55
7.4
0.66
4.8
2.5
9.2
Other Oil (>401° F)
203.8
20.17
4.1
0.66
2.7
1.4
5.1
Still Gas
166.1
17.51
2.9
0.66
1.9
1.0
3.6
Petroleum Coke
0.0
27.85
0.0
0.30
0.0
0.0
0.0
Special Naphtha
88.7
19.74
1.8
0.66
1.2
0.6
2.2
Distillate Fuel Oil
5.8
20.17
0.1
0.50
0.1
0.1
0.2
Waxes
12.9
19.80
0.3
0.58
0.1
0.1
0.4
Miscellaneous Products
191.3
20.31
3.9
0.00
0.0
3.9
14.2
Transportation
140.6
NA
2.8
NA
0.3
2.6
9.5
Lubricants
140.6
20.20
2.8
0.09
0.3
2.6
9.5
U.S. Territories
77.3
NA
1.5
NA
0.2
1.4
5.1
Lubricants
1.0
20.20
0.0
0.09
0.0
0.0
0.1
Other Petroleum (Misc.
Prod.)
76.2
20.00
1.5
0.10
0.2
1.4
5.0
Total
4,844.5
89.7
56.6
33.0
121.0
+ Does not exceed 0.05 TBtu, MMT C, MMT CO2 Eq.
NA (Not Applicable)
aTo avoid double counting, net exports have been deducted.
Note: Totals may not sum due to independent rounding.
Lastly, emissions were estimated by subtracting the C stored from the potential emissions (see Table 3-19). 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, 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 2016a), Toxics Release
Inventory, 1998 (EPA 2000b), Biennial Reporting System (EPA 2000a, 2009), Resource Conservation and Recovery
Act Information System (EPA 2013b, 2015b, 2016c), pesticide sales and use estimates (EPA 1998, 1999, 2002,
2004, 2011, 2017), and the Chemical Data Access Tool (EPA 2012); the EIA Manufacturer's Energy Consumption
Survey (MECS) (EIA 1994, 1997, 2001, 2005, 2010, 2013, 2017a); the National Petrochemical & Refiners
Association (NPRA 2002); the U.S. Census Bureau (1999, 2004, 2009, 2014); Bank of Canada (2012, 2013, 2014,
2016, 2017); Financial Planning Association (2006); INEGI (2006); the United States International Trade
Energy 3-47
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Commission (1990 through 2016); Gosselin, Smith, and Hodge (1984); EPA's Municipal Solid Waste (MSW) Facts
and Figures (EPA 2013a, 2014a, 2016b); the Rubber Manufacturers' Association (RMA 2009, 2011, 2014, 2016);
the International Institute of Synthetic Rubber Products (IISRP 2000, 2003); the Fiber Economics Bureau (FEB
2001, 2003, 2005, 2007, 2009, 2010, 2011, 2012, 2013, 2017); the EPA Chemical Data Access Tool (CDAT) (EPA
2014b); the American Chemistry Council (ACC 2003 through 2011, 2013, 2014, 2015a, 2016b, 2017); and the
Guide to the Business of Chemistry (ACC 2012, 2015b, 2016a). Specific data sources are listed in full detail in
Annex 2.3.
Uncertainty and lime-Serfi insistency
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.
As noted above, the non-energy use analysis is based on U.S.-specific storage factors for (1) feedstock materials
(natural gas, LPG, pentanes plus, 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 3-20 and Table
3-21), 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 3-22 (emissions) and Table
3-23 (storage factors). Carbon emitted from non-energy uses of fossil fuels in 2016 was estimated to be between
93.5 and 166.5 MMT CO2 Eq. at a 95 percent confidence level. This indicates a range of 23 percent below to 38
percent above the 2016 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 3-22: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Non-
Energy Uses of Fossil Fuels (MMT CO2 Eq. and Percent)
Source
Gas
2016 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Feedstocks
CO2
72.3
51.0
121.3
-29%
68%
Asphalt
CO2
0.3
0.1
0.6
-58%
120%
Lubricants
CO2
19.5
16.1
22.8
-18%
16%
Waxes
CO2
0.4
0.3
0.7
-23%
77%
Other
CO2
28.5
17.3
31.1
-39%
9%
Total
CO2
121.0
93.5
166.5
-23%
38%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence
interval.
Note: Totals may not sum due to independent rounding.
3-48 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Table 3-23: Approach 2 Quantitative Uncertainty Estimates for Storage Factors of Non-
Energy Uses of Fossil Fuels (Percent)
Source
Gas
2016 Storage Factor
(%)
Uncertainty Range Relative to Emission Estimate3
(%) (%, Relative)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Feedstocks
CO2
65.9%
50%
70%
-19%
5%
Asphalt
CO2
99.6%
99%
100%
-0.5%
0.3%
Lubricants
CO2
9.2%
4%
17%
-58%
91%
Waxes
CO2
57.8%
48%
67%
-17%
17%
Other
CO2
6.4%
6%
43%
-4%
569%
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 3-23, 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 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.
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2016 as discussed below. Details on the emission trends through time are described in more detail in the
Methodology section, above.
QA/QC and Verification
A source-specific QA/QC plan for non-energy uses of fossil fuels was developed and implemented. 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 category-specific 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 2015 totals as well as their trends
across the time series.
Petrochemical input data reported by EIA will continue to be investigated in an attempt to address an input/output
discrepancy in the NEU model. Prior to 2001, the C balance inputs exceed outputs, then starting in 2001 through
2009, outputs exceeded inputs. In 2010 through 2016, inputs exceeded outputs. A portion of this discrepancy has
been reduced and two strategies have been developed to address the remaining portion (see the Planned
Improvements section, below).
Energy 3-49
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
Recalculations Discussion
A number of updates to historical production values were included in the most recent Monthly Energy Review; these
have been populated throughout the Inventory.
Pesticide production data for 2007 through 2015 were updated using EPA's Pesticides Industry Sales and Usage
2008 - 2012 Market Estimates (EPA 2017). This resulted in a slight increase in emissions from pesticides compared
to previous estimates for 2007 through 2015. Pesticide production data for 1990 through 2015 were updated by
correcting rounding errors and molecular weights and chemical formulas for certain pesticides.
The calculated ratio of urea production to melamine production from 2001 to 2015 was updated to approximately
95/5 based on ICIS (2016) and ICIS (2008), rather tlian an even 50/50 split as previously estimated.
Planned Improvements
There are several future improvements planned:
• Analyzing the fuel and feedstock data from EPA's GHGRP subpart X (Petrochemical Production) to better
disaggregate CO2 emissions in NEU model and CO2 process emissions from petrochemical production.
• More accurate accounting of C in petrochemical feedstocks. EPA has worked with EI A 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. 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
(miscellaneous products). A better understanding of these trends will be pursued to identify any
mischaracterized or misreported fuel consumption for non-energy uses. For example, "miscellaneous
products" category includes miscellaneous products that are not reported elsewhere in the EIA data set. The
EIA does not have firm data concerning the amounts of various products that are being reported in the
"miscellaneous products" category; however, EIA has indicated that recovered sulfur from petroleum and
natural gas processing, and potentially also C black feedstock could be reported in this category. Recovered
sulfur would not be reported in the NEU calculation or elsewhere in the Inventory.
• 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
3-50 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
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.
Box 3-6: 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 Industrial Processes and Product Use (IPPU) sector. 67 In this Inventory, C storage
and C emissions from product use of lubricants, waxes, and asphalt and road oil are reported under the Energy sector
in the Carbon Emitted from Non-Energy Uses of Fossil Fuels source category (CRF Source Category 1A).68
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. 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
Section 3.2, Table 3-21). 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.69
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. 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. These artificial adjustments would also result in the C emissions for lubricants, waxes,
and asphalt and road oil being reported under IPPU, while the C storage for lubricants, waxes, and asphalt and road
oil would be reported under Energy. 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, 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 IPPU chapter, as
they were consumed during non-energy related industrial activity. Emissions from 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.
3.3 Incineration of Waste (CRF Source
67 See Volume 3: Industrial Processes and Product Use, Chapter 5: Non-Energy Products from Fuels and Solvent Use of the
2006 IPCC Guidelines for National Greenhouse Gas Inventories (IPCC 2006).
68 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.
69 Data and calculations for lubricants and waxes and asphalt and road oil are in Annex 2.3 - Methodology and Data for
Estimating CO2 Emissions from Fossil Fuel Combustion.
Energy 3-51
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Category lAla) - TO BE UPDATED FOR
FINAL INVENTORY REPORT
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; Goldstein
andMadtes 2001; Kaufman et al. 2004; Simmons et al. 2006; van Haarenet 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 IPCC guidelines, when the CO2 emitted is of fossil origin,
it is counted as a net anthropogenic emission of CO2 to the atmosphere. Thus, the emissions from waste 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 municipal solid wastes 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—plastics, synthetic rubber, synthetic fibers, and carbon black in scrap tires—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
municipal solid wastes 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 municipal solid waste. 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 30.1 million metric tons of MSW were incinerated in the United States in 2014 (EPA 2016). Data for
the amount of MSW incinerated in 2015 were not available, so data for 2015 was assumed to be equal to data for
2014. CO2 emissions from incineration of waste rose 34 percent since 1990, to an estimated 10.7 MMT CO2 Eq.
(10,676 kt) in 2015, as the volume of scrap tires and other fossil C-containing materials in waste increased (see
Table 3-24 and Table 3-25). Waste incineration is also a source of CH4 and N20 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 0.5 kt CH4) in 2015, and have decreased by 32 percent since 1990. Nitrous oxide emissions from the
incineration of waste were estimated to be 0.3 MMT CO2 Eq. (1 kt N20) in 2015, and have decreased by 32 percent
since 1990.
Table 3-24: CO2, ChU, and N2O Emissions from the Incineration of Waste (MMT CO2 Eq.)
Gas/Waste Product
1990
2005
2011
2012
2013
2014
2015
CO2
8.0
12.5
10.6
10.4
10.4
10.6
10.7
Plastics
5.6
6.9
5.8
5.7
5.8
5.9
5.9
Synthetic Rubber in Tires
0.3
1.6
1.4
1.3
1.2
1.2
1.2
Carbon Black in Tires
0.4
2.0
1.7
1.5
1.4
1.4
1.5
Synthetic Rubber in
MSW
0.9
0.8
0.7
0.7
0.7
0.7
0.7
Synthetic Fibers
0.8
1.2
1.1
1.1
1.3
1.3
1.3
CH4
+
+
+
+
+
+
+
N2O
0.5
0.4
0.3
0.3
0.3
0.3
0.3
Total
8.4
12.8
10.9
10.7
10.7
10.9
11.0
+ Does not exceed 0.05 MMT CO2 Eq.
Table 3-25: CO2, ChU, and N2O Emissions from the Incineration of Waste (kt)
Gas/Waste Product 1990 2005 2011 2012 2013 2014 2015
3-52 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
C02
7,950
12,469
10,564
10,379
10,398
10,608
10,676
Plastics
5,588
6,919
5,757
5,709
5,815
5,928
5,928
Synthetic Rubber in Tires
308
1,599
1,363
1,261
1,158
1,189
1,220
Carbon Black in Tires
385
1,958
1,663
1,537
1,412
1,449
1,487
Synthetic Rubber in
MSW
854
766
712
706
729
729
729
Synthetic Fibers
816
1,227
1,070
1,166
1,284
1,313
1,313
ch4
+
+
+
+
+
+
+
n2o
2
1
1
1
1
1
1
+ Does not exceed 0.5 kt
Methodology
Emissions of CO2 from the incineration of waste include CO2 generated by the incineration of plastics, synthetic
fibers, and synthetic rubber in MSW, as well as the incineration of synthetic rubber and carbon black in scrap tires.
The emission estimates are calculated for all four sources on a mass-basis based on the data available. These
emissions were estimated by multiplying the mass of each material incinerated by the C content of the material and
the fraction oxidized (98 percent). Plastics incinerated in municipal solid wastes were categorized into seven plastic
resin types, each material having a discrete C content. Similarly, synthetic rubber is categorized into three product
types, and synthetic fibers were categorized into four product types, each having a discrete C content. 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. 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. More detail on the
methodology for calculating emissions from each of these waste incineration sources is provided in Annex 3.7.
For each of the methods used to calculate CO2 emissions from the incineration of waste, data on the quantity of
product combusted and the C content of the product are needed. For plastics, synthetic rubber, and synthetic fibers in
MSW, the amount of specific materials discarded as municipal solid waste (i.e., the quantity generated minus the
quantity recycled) was taken from Municipal Solid H aste Generation, Recycling, and Disposal in the United States:
Facts and Figures (EPA 2000 through 2003, 2005 through 2014), and Advancing Sustainable Materials
Management: Facts and Figures: Assessing Trends in Material Generation, Recycling and Disposal in the United
States (EPA 2015, 2016) and detailed unpublished backup data for some years not shown in the reports (Schneider
2007). For 2015, the amount of MSW incinerated was assumed to be equal to that in 2014, due to the lack of
available data. The proportion of total waste discarded that is incinerated was derived from Shin (2014). Data on
total waste incinerated was not available in detail for 2012 through 2015, so these values were assumed to equal to
the 2011 value (Shin 2014). For synthetic rubber and carbon black in scrap tires, information was obtained from
U.S. Scrap Tire Management Summary for 2005 through 2015 data (RMA 2016). Average C contents for the
"Other" plastics category and synthetic rubber in municipal solid wastes were calculated from 1998 and 2002
production statistics: C content for 1990 through 1998 is based on the 1998 value; C content for 1999 through 2001
is the average of 1998 and 2002 values; and C content for 2002 to date is based on the 2002 value. Carbon content
for synthetic fibers was calculated from a weighted average of production statistics from 1990 to date. Information
about scrap tire composition was taken from the Rubber Manufacturers' Association internet site (RMA 2012a). The
mass of incinerated material is multiplied by its C content to calculate the total amount of carbon stored.
The assumption that 98 percent of organic C is oxidized (which applies to all waste incineration categories for CO2
emissions) was reported in EPA's life cycle analysis of greenhouse gas emissions and sinks from management of
solid waste (EPA 2006). This percentage is multiplied by the carbon stored to estimate the amount of carbon
emitted.
Incineration of waste, including MSW, also results in emissions of CH4 and N20. These emissions were calculated
as a function of the total estimated mass of waste incinerated and emission factors. As noted above, CH4 and N20
emissions are a function of total waste incinerated in each year; for 1990 through 2008, these data were derived from
the information published in BioCvcle (van Haaren et al. 2010). Data for 2009 and 2010 were interpolated between
2008 and 2011 values. Data for 2011 were derived from Shin (2014). Data on total waste incinerated was not
Energy 3-53
-------
1 available in the BioCvcle data set for 2012 through 2015, so these values were assumed to equal the 2011 BioCvcle
2 data set value.
3 Table 3-26 provides data on municipal solid waste discarded and percentage combusted for the total waste stream.
4 The emission factors of N20 and CH4 emissions per quantity of municipal solid waste combusted are default
5 emission factors for the default continuously-fed stoker unit MSW incineration technology type and were taken from
6 IPCC (2006).
7 Table 3-26: Municipal Solid Waste Generation (Metric Tons) and Percent Combusted
8 (BioCycle dataset)
Incinerated (% of
Year Waste Discarded Waste Incinerated Discards)
1990 235,733,657 30,632,057 13.0%
2005 259,559,787 25,973,520 10.0%
2011 273,116,704 20,756,870 7.6%
2012 273,116,704a 20,756,870 7.6%
2013 273,116,704a 20,756,870 7.6%
2014 273,116,704a 20,756,870 7.6%
201 5 273,116,704a 20,756,870 7.6%
a Assumed equal to 2011 value.
Source: van Haaren et al. (2010)
9 Uncertainty and Time-Series Consistency
10 An Approach 2 Monte Carlo analysis was performed to determine the level of uncertainty surrounding the estimates
11 of CO2 emissions and N20 emissions from the incineration of waste (given the very low emissions for CH4, no
12 uncertainty estimate was derived). IPCC Approach 2 analysis allows the specification of probability density
13 functions for key variables within a computational structure that mirrors the calculation of the Inventory estimate.
14 Uncertainty estimates and distributions for waste generation variables (i.e., plastics, synthetic rubber, and textiles
15 generation) were obtained through a conversation with one of the authors of the Municipal Solid Waste in the
16 United States reports. Statistical analyses or expert judgments of uncertainty were not available directly from the
17 information sources for the other variables; thus, uncertainty estimates for these variables were determined using
18 assumptions based on source category knowledge and the known uncertainty estimates for the waste generation
19 variables.
20 The uncertainties in the waste incineration emission estimates arise from both the assumptions applied to the data
21 and from the quality of the data. Key factors include MSW incineration rate; fraction oxidized; missing data on
22 waste composition; average C content of waste components; assumptions on the synthetic/biogenic C ratio; and
23 combustion conditions affecting N20 emissions. The highest levels of uncertainty surround the variables that are
24 based on assumptions (e.g., percent of clothing and footwear composed of synthetic rubber); the lowest levels of
25 uncertainty surround variables that were determined by quantitative measurements (e.g., combustion efficiency, C
26 content of C black).
27 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3-27. Waste incineration
28 CO2 emissions in 2015 were estimated to be between 9.6 and 12.1 MMT CO2 Eq. at a 95 percent confidence level.
29 This indicates a range of 10 percent below to 13 percent above the 2015 emission estimate of 10.7 MMT CO2 Eq.
30 Also at a 95 percent confidence level, waste incineration N20 emissions in 2015 were estimated to be between 0.2
31 and 1.3 MMT CO2 Eq. This indicates a range of 51 percent below to 3 3 0 percent above the 2015 emission estimate
32 of 0.3 MMT C02 Eq.
3-54 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
Table 3-27: Approach 2 Quantitative Uncertainty Estimates for CO2 and N2O from the
Incineration of Waste (MMT CO2 Eq. and Percent)
2015 Emission Estimate Uncertainty Range Relative to Emission Estimate3
Source Gas (MMT CP2 Eq.) (MMT CO2 Eq.) [%)
Lower Upper Lower Upper
Bound Bound Bound Bound
Incineration of Waste CO2 10.7 9.6 12.1 -10% +13%
Incineration of Waste N2O (13 02 L3 -51% +330%
a Range of emission estimates predicted by Monte Carlo Simulation for a 95 percent confidence interval.
3 Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
4 through 2015 as discussed below. Details on the emission trends through time are described in more detail in the
5 Methodology section above.
6 QA/QC and Verification
7 A source-specific Quality Assurance/Quality Control plan was implemented for incineration of waste. This effort
8 included a general (Tier 1) analysis, as well as portions of a category-specific (Tier 2) analysis. The Tier 2
9 procedures that were implemented involved checks specifically focusing on the activity data and specifically
10 focused on the emission factor and activity data sources and methodology used for estimating emissions from
11 incineration of waste. Trends across the time series were analyzed to determine whether any corrective actions were
12 needed. Actions were taken to streamline the activity data throughout the calculations on incineration of waste.
13 Recalculations Discussion
14 For the current Inventory, emission estimates for 2014 have been updated based on Advancing Sustainable
15 Materials Management: 2014 Fact Sheet (EPA 2016). The data used to calculate the percent incineration was not
16 updated in the current Inventory. BioCvcle lias not released a new State of Garbage in America Report since 2010
17 (with 2008 data), which used to be a semi-annual publication which publishes the results of the nation-wide MSW
18 survey. The results of the survey have been published in Shin (2014). This provided updated incineration data for
19 2011, so the generation and incineration data for 2012 through 2015 are assumed equivalent to the 2011 values. The
20 data for 2009 and 2010 were based on interpolations between 2008 and 2011.
21 A transcription error in 2013 plastics production data from EPA's Advancing Sustainable Materials Management:
22 Facts and Figures 2013: Assessing Trends in Material Generation, Recycling and Disposal in the United States
23 (EPA 2015) was identified and corrected. The amount of HDPE discarded in 2013 was misreported and the value
24 has been updated. This update results in updated emission estimate for the CO2 from Plastics for 2013.
25 Previously, the carbon content for synthetic fiber was assumed to be the weighted average of carbon contents of four
26 fiber types (polyester, nylon olefin and acrylic) based on 1999 fiber production data. This methodology lias been
27 updated. A weighted average for the carbon content of synthetic fibers based on production data from 1990 through
28 2015 was developed for each year based on the amount of fiber produced. For each year, the weighted average
29 carbon content was used to develop the amount of carbon emitted. This methodology update affects the synthetic
30 fiber CO2 estimates.
31 Planned Improvements
32 The availability of facility-level waste incineration data through EPA's Greenhouse Gas Reporting Program
33 (GHGRP) will be examined to help better characterize waste incineration operations in the United States. This
34 characterization could include future improvements as to the operations involved in waste incineration for energy,
35 whether in the power generation sector or the industrial sector. Additional examinations will be necessary as, unlike
Energy 3-55
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
the reporting requirements for this chapter under the UNFCCC reporting guidelines,70 some facility-level waste
incineration emissions reported under EPA's GHGRP may also include industrial process emissions. In line with
UNFCCC reporting guidelines, emissions for waste incineration with energy recovery 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 waste
incineration 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 ensuring CO2 emissions from the biomass component of waste are separated in
the facility-level reported data, and on maintaining consistency with national waste generation and fate statistics
currently used to estimate total, national U.S. greenhouse gas emissions. 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.71 GHGRP data is available for MSW combustors, which contains
information on the CO2, CH4, and N20 emissions from MSW combustion plus the fraction of the emissions that are
biogenic. To calculate biogenic versus total CO2 emissions, a default biogenic fraction of 0.6 is used. The biogenic
fraction will be calculated using the current input data and assumptions to verily the current MSW emission
estimates.
If GHGRP data would not provide a more accurate estimate of the amount of solid waste combusted, new data
sources for the total MSW generated will be explored given that the data previously published semi-annually in
BioCvcIe (van Haaren et al. 2010) has ceased to be published, according to the authors. Equivalent data was derived
from Shin (2014) for 2011. A new methodology would be developed considering the available data within the
annual update of EPA's Ach'ancing Sustainable Materials Management: Facts and Figures 2014: Assessing Trends
in Material Generation, Recycling and Disposal in the United States (EPA 2016) and a report from the
Enviromnental Research & Education Foundation (2016), MSW Management in the U.S.: 2010 & 2013, that has
data for 2010 and 2013. In developing the new methodology, appropriate assumptions would need to be made to
ensure that the MSW figures include the same boundaries. Consideration would also be made to be consistent with
calculations in other waste categories including landfilling and composting.
Additional improvements will be conducted to improve the transparency in the current reporting of waste
incineration. Currently, hazardous industrial waste incineration is included within the overall calculations for the
Carbon Emitted from Non-Energy Uses of Fossil Fuels source category. Waste incineration activities that do not
include energy recovery will be examined. Synthetic fibers within scrap tires are not included in this analysis and
will be explored for future Inventories. The carbon content of fibers within scrap tires would be used to calculate the
associated incineration emissions. Updated fiber content data from the Fiber Economics Bureau will also be
explored.
3.4 Coal Mining (CRF Source Category lBla)
Three types of coal mining-related activities release CH4 to the atmosphere: underground mining, surface mining,
and post-mining (i.e., coal-handling) activities. While surface mines account for the majority of U.S. coal
production underground coal mines contribute the largest share of CH4 emissions (see Table 3-29 and Table 3-30)
due to the higher CH4 content of coal in the deeper underground coal seams. In 2016, 251 underground coal mines
and 439 surface mines were operating in the United States. In recent years the total number of active coal mines in
the United States lias declined. In 2016, the United States was the third largest coal producer in the world (660
MMT), after China (3,242 MMT) and India (708 MMT) (IEA 2017).
70 See .
71 See .
3-56 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
Table 3-28: 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
334,398
789
691.448
1.398
1,025.846
2012
488
310,608
719
610,307
1,207
920,915
2013
395
309,546
637
581,270
1,032
890,815
2014
345
321,783
613
583,974
958
905,757
2015
305
278,342
529
534,127
834
812,469
2016
251
228,403
439
431,485
690
659,888
2 Underground mines liberate CH4 from ventilation systems and from degasification systems. Ventilation systems
3 pump air through the mine workings to dilute noxious gases and ensure worker safety; these systems can exhaust
4 significant amounts of CH4 to the atmosphere in low concentrations. Degasification systems are wells drilled from
5 the surface or boreholes drilled inside the mine that remove large, often highly concentrated volumes of CH4 before,
6 during, or after mining. Some mines recover and use CH4 generated from ventilation and degasification systems,
7 thereby reducing emissions to the atmosphere.
8 Surface coal mines liberate CH4 as the overburden is removed and the coal is exposed to the atmosphere. CH4
9 emissions are normally a function of coal rank (a classification related to the percentage of carbon in the coal) and
10 depth. Surface coal mines typically produce lower-rank coals and remove less than 250 feet of overburden, so their
11 level of emissions is much lower than from underground mines.
12 In addition, CH4 is released during post-mining activities, as the coal is processed, transported, and stored for use.
13 Total CH4 emissions in 2016 were estimated to be 2,153 kt (53.8 MMT CO2 Eq.), a decline of 44 percent since 1990
14 (see Table 3-29 and Table 3-30). Of these total emissions, underground mines accounted for approximately 76
15 percent, surface mines accounted for 13 percent, and post-mining activities accounted for 12 percent.
16 Table 3-29: ChU Emissions from Coal Mining (MMT CO2 Eq.)
Activity
1990
2005
2012
2013
2014
2015
2016
Underground (UG) Mining
74.2
42.0
47.3
46.2
46.1
44.9
40.7
Liberated
80.8
59.7
65.8
64.5
63.1
61.2
57.1
Recovered & Used
(6.6)
(17.7)
(18.5)
(18.3)
(17.0)
(16.4)
(16.3)
Surface Mining
10.8
11.9
10.3
9.7
9.6
8.7
6.8
Post-Mining (UG)
9.2
7.6
6.7
6.6
6.7
5.8
4.8
Post-Mining (Surface)
2.3
2.6
2.2
2.1
2.1
1.9
1.5
Total
96.5
64.1
66.5
64.6
64.6
61.2
53.8
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values.
17 Table 3-30: ChU Emissions from Coal Mining (kt)
Activity
1990
2005
2012
2013
2014
2015
2016
UG Mining
2,968
1,682
1,891
1,849
1,844
1,796
1,629
Liberated
3,234
2,390
2,631
2,580
2,524
2,450
2,282
Recovered & Used
(266)
(708)
(740)
(730)
(680)
(654)
(654)
Surface Mining
430
475
410
388
386
347
273
Post-Mining (UG)
368
306
268
263
270
231
192
Post-Mining (Surface)
93
103
89
84
84
75
59
Total
3,860
2,565
2,658
2,584
2,583
2,449
2,153
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values.
Energy 3-57
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Methodology
The methodology for estimating CH4 emissions from coal mining consists of two steps:
• Estimate 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 CH4
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 from degasification systems. Some mines recover
and use the generated CH4, thereby reducing emissions to the atmosphere. Total CH4 emitted from underground
mines equals the CH4 liberated from ventilation systems, plus the CH4 liberated from degasification systems, minus
the CH4 recovered and used.
Step 1.1: Estimate CH4 Liberatedfrom Ventilation Systems
To estimate CH4 liberated from ventilation systems, EPA uses data collected through its Greenhouse Gas Reporting
Program (GHGRP)72 (subpart FF, "Underground Coal Mines"), data provided by the U.S. Mine Safety and Health
Administration (MSHA), 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 20 1 6).73 Mines that report to EPA's GHGRP must report quarterly measurements of CH4 emissions
from ventilation systems to EPA; they have the option of recording 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.74
Since 2013, ventilation emission estimates have been calculated based on both GHGRP data submitted by
underground mines, and on quarterly measurement data obtained directly from MSHA for the remaining mines. The
quarterly measurements are used to determine the average daily emissions rate for the reporting year quarter.
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.
Step 1.2: Estimate CH4 Liberatedfrom 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-five
mines used degasification systems in 2016, and the CH4 removed through these systems was reported to EPA's
GHGRP under subpart FF (EPA 2017). Based on the weekly measurements reported to EPA's GHGRP,
degasification data summaries for each mine were added to estimate the CH4 liberated from degasification systems.
Fifteen of the 25 mines with degasification systems had operational CH4 recovery and use projects (see step 1.3
72 In implementing improvements and integrating data from EPA's GHGRP, the EPA followed the latest guidance from the
IPCC on the use of facility-level data in national inventories (IPCC 2011).
73 Underground coal mines report to EPA under Subpart FF of the GHGRP. In 2016, 90 underground coal mines reported to the
program.
74 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.
3-58 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
below), and EPA's GHGRP reports show the remaining ten mines vented CH4from degasification systems to the
atmosphere.75
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 to
estimate CH4 liberated from degasification systems at 20 of the 25 mines that used degasification systems in 2016.
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.76 EPA's GHGRP does not require
gas production from virgin coal seams (coalbed methane) to be reported by coal mines under subpart FF.77 Most
pre-mining wells drilled from the surface are considered coalbed methane wells prior to mine-through and
associated CH4 emissions are reported under another subpart of the program (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. For five mines with degasification systems that include pre-mining wells that
were mined through in 2016, GHGRP information was supplemented with historical data from state gas well
production databases (DMME 2017; GSA 2017; WVGES 2017), 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).
EPA's GHGRP reports with CH4 liberated from degasification systems are reviewed for errors in reporting. For one
of the 25 mines, due to a lack of mine-provided information used in prior years and a GHGRP reporting
discrepancy, the CH4 liberated was based on both an estimate from historical mine-provided CH4 recovery and use
rates and state gas sales records (DMME 2017).
Step 1.3: Estimate CH4 Recovered from Ventilation and Degasification Systems, and Utilized or
Destroyed (Emissions Avoided)
Fifteen mines had CH4 recovery and use projects in place in 2016. Fourteen of these mines sold the recovered CH4
to a pipeline, including one that also used CH4 to fuel a thermal coal dryer. In addition, one mine used recovered
CH4to heat mine ventilation air.
EPA's GHGRP data was exclusively used to estimate the CH4 recovered and used from ten of the 15 mines that
deployed degasification systems in 2016. Based on weekly measurements, the GHGRP degasification destruction
data summaries for each mine were added together to estimate the CH4 recovered and used from degasification
systems.
All 10 mines with degasification systems used pre-mining wells as part of those systems, but only four of the mines
intersected pre-mining wells in 2016. EPA's GHGRP and supplemental data were used to estimate CH4 recovered
and used at two of these four mines; supplemental data alone (GSA 2017) was used to estimate CH4 recovered and
used at the other two mines. Supplemental information was used for these four 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. The supplemental data came from state gas
production databases as well as mine-specific information on the timing of mined-through pre-mining wells.
EPA's GHGRP information was not used to estimate CH4 recovered and used at two mines. At one of these mines, a
portion of reported CH4 vented was applied to an ongoing mine air heating project. Because of a lack of mine-
provided information used in prior years and a GHGRP reporting discrepancy, the 2016 CH4 recovered and used
from pre-mining wells at the other mine was based on an estimate from historical mine-provided CH4 recovery and
use rates. Emissions recovered and used from the active mine degasification system were estimated based on a state
gas production data information system.
75 Several of the mines venting CH4 from degasification systems use a small portion the gas to fuel gob well blowers in remote
locations where electricity is not available. However, this CH4use is not considered to be a formal recovery and use project.
76 A well is "mined through" when coal mining development or the working face intersects the borehole or well.
77 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.
Energy 3-59
-------
1 In 2016, one mine destroyed a portion of its CH4 emissions from ventilation systems using thermal oxidation
2 technology. The amount of CH4 recovered and destroyed by the project was determined through publicly-available
3 emission reduction project information (ACR 2017).
4 Step 2: Estimate CH4 Emitted from Surface Mines and Post-Mining Activities
5 Mine-specific data are not available for estimating CH4 emissions from surface coal mines or for post-mining
6 activities. For surface mines, basin-specific coal production obtained from the Energy Information Administration's
7 Annual Coal Report (EIA 2017) was multiplied by basin-specific CHi contents (EPA 1996, 2005) and a 150 percent
8 emission factor (to account for CH4from over- and under-burden) to estimate CH4 emissions (King 1994; Saghafi
9 2013). For post-mining activities, basin-specific coal production was multiplied by basin-specific gas contents and a
10 mid-range 32.5 percent emission factor for CH4 desorption during coal transportation and storage (Creedy 1993).
11 Basin-specific in situ gas content data were compiled from AAPG (1984) and USBM (1986).
12 Uncertainty and Time-Series Consistency
13 A quantitative uncertainty analysis was conducted for the coal mining source category using the IPCC-
14 recommended Approach 2 uncertainty estimation methodology. Because emission estimates from underground
15 ventilation systems were based on actual measurement data from EPA's GHGRP or from MSHA, uncertainty is
16 relatively low. A degree of imprecision was introduced because the ventilation air measurements used were not
17 continuous but rather quarterly instantaneous readings that were used to determine the average daily emissions rate
18 for the quarter. Additionally, the measurement equipment used can be expected to have resulted in an average of 10
19 percent overestimation of annual CH4 emissions (Mutmansky & Wang 2000). GHGRP data were used for a
20 significant number of the mines that reported their own measurements to the program beginning in 2013; however,
21 the equipment uncertainty is applied to both GHGRP and MSHA data.
22 Estimates of CH4 recovered by degasification systems are relatively certain for utilized CH4 because of the
23 availability of EPA's GHGRP data and gas sales information. Many of the recovery estimates use data on wells
24 within 100 feet of a mined area. However, uncertainty exists concerning the radius of influence of each well. The
25 number of wells counted, and thus the avoided emissions, may vary if the drainage area is found to be larger or
26 smaller than estimated.
27 EPA's GHGRP requires weekly CH4 monitoring of mines that report degasification systems, and continuous CH4
28 monitoring is required for utilized CH4 on- or off-site. Since 2012, GHGRP data have been used to estimate CH4
29 emissions from vented degasification wells, reducing the uncertainty associated with prior MSHA estimates used for
30 this subsource. Beginning in 2013, GHGRP data were also used for determining CH4 recovery and use at mines
31 without publicly available gas usage or sales records, which has reduced the uncertainty from previous estimation
32 methods that were based on information from coal industry contacts.
33 In 2015 and 2016, a small level of uncertainty was introduced with using estimated rather than measured values of
34 recovered methane from two of the mines with degasification systems. An increased level of uncertainty was applied
35 to these two subsources, but the change had little impact on the overall uncertainty.
36 Surface mining and post-mining emissions are associated with considerably more uncertainty than underground
37 mines, because of the difficulty in developing accurate emission factors from field measurements. However, since
38 underground emissions constitute the majority of total coal mining emissions, the uncertainty associated with
39 underground emissions is the primary factor that determines overall uncertainty.
40 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3-31. Coal mining CH4
41 emissions in 2016 were estimated to be between 47.5 and 61.7 MMT CO2 Eq. at a 95 percent confidence level. This
42 indicates a range of 11.8 percent below to 14.6 percent above the 2016 emission estimate of 53.8 MMT CO2 Eq.
3-60 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Table 3-31: Approach 2 Quantitative Uncertainty Estimates for ChU Emissions from Coal
Mining (MMT CO2 Eq. and Percent)
Source
Gas
2016 Emission Estimate
(MMT CO2 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT CO2 Eq.) (%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Coal mining
CH4
53.8
47.5 61.7
-11.8% +14.6%
a Range of emission estimates predicted by Monte Carlo stochastic simulation for a 95 percent confidence interval.
Methodological recalculations were applied to the entire time series to ensure consistency from 1990 through 2016.
Details on the emission trends through time are described in more detail in the methodology section.
Recalculations Discussion
For the current Inventory, revisions were made to the 2014 and 2015 underground liberated and recovered
emissions. The EPA's GHGRP data that was used to calculate the emissions liberated and destroyed in 2014 and
2015 from a mine with a ventilation air methane (VAM) project was incorrect. The GHGRP spreadsheet for Subpart
FF reporting does not accommodate methane destruction from VAM, and therefore the emissions avoided are
reported as degasification. In 2016, the VAM project's verified emission reductions registered with the California
Air Resources Board were deducted from the total reported destroyed methane; and the remaining emissions
destroyed were applied to the mine's degasification emissions recovered and destroyed total. The revised
methodology was used to recalculate and update the emissions avoided in 2014 and 2015.
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 CH4 that may find its
way to surface structures through overburden fractures or via ground water aquifers. 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 through 2016,
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
Energy 3-61
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
of gassy mine78 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. There were five gassy mine closures in 2016. In 2016, gross
abandoned mine emissions increased slightly from 9.0 to 9.5 MMT CO2 Eq. (see Table 3-32 and Table 3-33). Gross
emissions are reduced by CH4 recovered and used at 45 mines, resulting in net emissions in 2016 of 6.7 MMT CO2
Eq.
Table 3-32: ChU Emissions from Abandoned Coal Mines (MMT CO2 Eq.)
Activity
1990
2005
2012
2013
2014
2015
2016
Abandoned Underground Mines
7.2
8.4
8.9
8.8
8.7
9.0
9.5
Recovered & Used
+
1.8
2.7
2.6
2.4
2.6
2.8
Total
7.2
6.6
6.2
6.2
6.3
6.4
6.7
+ Does not exceed 0.05 MMT CO2 Eq.
Note: Totals may not sum due to independent rounding.
Table 3-33: ChU Emissions from Abandoned Coal Mines (kt)
Activity
1990
2005
2012
2013
2014
2015
2016
Abandoned Underground Mines
288
334
OO
353
350
359
380
Recovered & Used
+
70
109
104
97
102
112
Total
288
264
249
249
253
256
268
+ Does not exceed 0.5 kt
Note: Totals may not sum due to independent rounding.
Methodology
Estimating CH4 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 and method 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 which 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.
In order to estimate CH4 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. Permeability and isotherm data were gathered from each coal basin and histograms were
generated. The low, mid and high values of each parameter were combined in nine separate flow simulations for
each coal basin using a computational fluid dynamics simulation model used in the oil and gas industry, which
generated individual decline curves. These decline curves fit a hyperbolic equation commonly used in the oil and
78 A mine is considered a "gassy" mine if it emits more than 100 thousand cubic feet of CH4 per day (100 mcfd).
3-62 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
gas industry for forecasting gas well production. A rate-time equation can be generated that can be used to predict
future emissions. This equation is expressed as:
q = (1 + &A0("1/fc)
where,
q
b
Dj
t
Gas flow rate at time t in million cubic feet per day (mmcfd)
Initial gas flow rate at time zero (tQ), mmcfd
The hyperbolic exponent, dimensionless
Initial decline rate, 1/year
Elapsed time from tQ (years)
This equation is applied to mines of various initial emission rates that have similar initial pressures, permeability and
adsorption isotherms relative to their coal basin (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 CH4 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. There was not enough data to establish basin-specific equations as was done with the vented,
non-flooding mines (EPA 2004).
Seals have an inhibiting effect on the flow rate of CH4 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) account for about 98
percent of all CH4 emissions. This same relationship is assumed for abandoned mines. It was determined that the
531 abandoned mines closed after 1972 produced emissions greater than 100 mcfd when active. Further, the status
of 304 of the 531 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 CH4 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 3-34: Number of Gassy Abandoned Mines Present in U.S. Basins in 2016, grouped by
Class according to Post-Abandonment State
q = q^
where,
q
q>
D
t
Gas flow rate at time t in mmcfd
Initial gas flow rate at time zero (tQ), mmcfd
Decline rate, 1/year
Elapsed time from tQ (years)
Basin
Sealed
Vented
Flooded
Total
Known
Unknown Total Mines
Central Appl.
Illinois
Northern Appl.
40
34
47
26
3
22
52
14
16
118
51
85
147
31
39
265
82
124
Energy 3-63
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
Warrior Basin
Western Basins
0
28
0
4
16
2
16
34
0
10
16
44
Total
149
55
100
304
227
531
Inputs to the decline equation require the average emission rate and the date of abandonment. Generally, this data is
available for mines abandoned after 1971; however, such data are largely unknown for mines closed before 1972,
which marks the beginning of comprehensive methane emissions data by the Bureau of Mines. 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 CH4 emissions from coal mining came from seventeen counties in seven states. In
addition, mine closure dates were obtained for two states, Colorado and Illinois, for the hundred-year period
extending from 1900 through 1999. The data were used to establish a frequency of mine closure histogram (by
decade) and applied to the other five states with gassy mine closures. As a result, basin-specific decline curve
equations were applied to the 145 gassy coal mines estimated to have closed between 1920 and 1971 in the United
States, representing 78 percent of the emissions. State-specific, initial emission rates were used based on average
coal mine CH4 emissions rates during the 1970s (EPA 2004) and closure dates were summarized by decade.
Emissions from pre-1972 mines represent approximately 17 percent of total abandoned mine methane emissions.
Abandoned mine emission estimates are based on all closed mines known to have active mine CH4 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 2016). Coal mine degasification data are not
available for years prior to 1990, thus the initial emission rates used reflect ventilation emissions only for pre-1990
closures. CH4 degasification amounts were added to the quantity of CH4 vented to determine the total CH4 liberation
rate for all mines that closed between 1992 and 2016. Since the sample of gassy mines 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 to account for all U.S. abandoned mine emissions.
From 1993 through 2016, emission totals were reduced by subtracting abandoned mine CH4 emissions avoided. The
Inventory totals were not adjusted for abandoned mine reductions from 1990 through 1992 because no data was
reported for abandoned coal mining CH4 recovery projects during that time.
A quantitative uncertainty analysis was conducted to estimate the uncertainty surrounding the estimates of emissions
from abandoned underground coal mines. The uncertainty analysis described below provides for the specification of
probability density functions for key variables within a computational structure that mirrors the calculation of the
inventory estimate. The results provide the range within which, with 95 percent certainty, emissions from this source
category are likely to fall.
As discussed above, the low, mid and high model generated decline curves for each basin were fitted to a hyperbolic
decline curve. The decline curve parameters, Di and b, for the low, mid and high decline curves were then used to
define a triangular distribution and together with the initial rate value of a mine's emissions and time from
abandonment, a probability density function for each mine in the coal basin was generated. These density functions
were then summed together using Monte Carlo simulation software to produce the AMM inventory which would be
expressed in terms of a 95 percent confidence interval.
The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 3-35. Annual abandoned
coal mine CH4 emissions in 2016 were estimated to be between 5.5 and 8.2 MMT CO2 Eq. at a 95 percent
confidence level. This indicates a range of 18 percent below to 22 percent above the 2016 emission estimate of 6.7
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 after 1972. Emissions from mines closed prior to 1972 have the largest degree of
uncertainty because no mine-specific CH4 liberation rates exist. Pre-1972 mines represent 17 percent of the total
abandoned mine inventory.
Uncertainty and Time-Series Consistency
3-64 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Table 3-35: Approach 2 Quantitative Uncertainty Estimates for ChU Emissions from
2 Abandoned Underground Coal Mines (MMT CO2 Eq. and Percent)
Source
Gas
2016 Emission Estimate
(MMT CO2 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT CO2 Eq.) (%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Abandoned Underground
Coal Mines
CH4
6.7
5.5 8.2
-18% +22%
a Range of emission estimates predicted by Monte Carlo Simulation for a 95 percent confidence interval.
3
4 Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
5 through 2016. Details on the emission trends through time are described in more detail in the Methodology section,
6 above.
7 3.6 Petroleum Systems (CRF Source Category
s lB2a)
9 Methane emissions from petroleum systems are primarily associated with onshore and offshore crude oil production,
10 transportation, and refining operations. During these activities, CH4 is released to the atmosphere as leak emissions,
11 vented emissions (including emissions from operational upsets) and emissions from fuel combustion. Leak and
12 vented CO2 emissions from petroleum systems are primarily associated with crude oil production and refining
13 operations but are negligible in transportation operations. Total CH4 emissions from petroleum systems in 2016
14 were 39.3 MMT CO2 Eq. (1,571 kt), a decrease of 7 percent from 1990. Total CO2 emissions from petroleum
15 systems in 2016 were 25.5 MMT CO2 Eq. (25,543 kt), an increase of a factor of 1.7 from 1990.
16 Exploration. Exploration includes well drilling, testing, and completions. Exploration accounts for approximately 5
17 percent of total CH4 emissions from petroleum systems. The predominant sources of emissions from exploration are
18 hydraulically fractured oil well completions and well testing. Other sources include well completions without
19 hydraulic fracturing and well drilling. Since 1990, exploration CH4 emissions have increased 168 percent due to
20 increases in the number of wells completed. Emissions of CH4 from exploration decreased 7 percent from 2015 to
21 2016. Exploration accounts for less than 1 percent of total CO2 emissions from petroleum systems. Emissions of
22 CO2 from exploration in 2016 decreased by 84 percent from 1990, and 85 percent from 2015, due to a decrease in
23 well testing flaring CO2 emissions.
24 Production. Production accounts for approximately 92 percent of total CH4 emissions from petroleum systems. The
25 predominant sources of emissions from production field operations are pneumatic controllers, offshore oil platforms,
26 oil tanks, gas engines, chemical injection pumps, associated gas venting and flaring, and leaks from oil wellheads.
27 Since 1990, CH4 emissions from production have decreased by 12 percent, due to decreases in tank emissions and in
28 associated gas venting. Overall, production segment methane emissions decreased by less than 1 percent from 2015
29 levels, although emissions from tanks increased by 53 percent, emissions from associated gas venting and flaring
30 decreased by 40 percent, and emissions from miscellaneous production flaring decreased by 45 percent in 2016
31 compared to 2015. The change in CH4 emissions from 2015 to 2016 for tanks, associated gas venting and flaring,
32 and miscellaneous production flaring reflects differences in reported GHGRP subpart W emissions levels for
33 reporting year (RY) 2016 as compared to RY2015. Production field operations account for approximately 85 percent
34 of the total CO2 emissions from petroleum systems. The principal sources of CO2 emissions are associated gas
35 flaring, oil tanks with flares, and miscellaneous production flaring. These three sources together account for over 99
36 percent of the CO2 emissions from production.
37 Crude Oil Transportation. Crude oil transportation activities account for less than 1 percent of total CH4 emissions
38 from the oil industry. Emissions from tanks, truck loading, rail loading, and marine vessel loading operations
39 account for 88 percent of CH4 emissions from crude oil transportation. Leak emissions, almost entirely from floating
Energy 3-65
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
roof tanks, account for approximately 12 percent of CH4 emissions from crude oil transportation. Since 1990, CH4
emissions from transportation have increased by 27 percent. However, because emissions from crude oil
transportation account for such a small percentage of the total emissions from the petroleum industry, this has had
little impact on the overall emissions. Methane emissions from transportation in 2016 decreased by less than 1
percent from 2015 levels.
Crude Oil Refining. Crude oil refining processes and systems account for approximately 2 percent of total CH4
emissions from the oil industry. 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 an insignificant amount of CH4 in all refined products. Within
refineries, incomplete combustion accounts for 38 percent of the CH4 emissions, while vented and leak emissions
account for approximately 52 and 10 percent, respectively. Flaring accounts for 82 percent of combustion CH4
emissions. Refinery system blowdowns for maintenance and process vents are the primary venting contributors (97
percent). Most of the leak CH4 emissions from refineries are from equipment leaks and storage tanks (85 percent).
Methane emissions from refining of crude oil have increased by approximately 51 percent since 1990; however,
similar to the transportation subcategory, this increase has had little effect on the overall emissions of CH4. From
1990 to 2015, CH4 emissions from crude oil refining fluctuated between 24 and 28 kt; in 2016, emissions increased
to 37 kt as process vent emissions increased. Crude oil refining processes and systems account for approximately 15
percent of total CO2 emissions from the oil industry. Almost all (97 percent) of the CO2 from refining is from
flaring. Refinery CO2 emissions increased by approximately 13 percent from 1990 to 2016.
Table 3-36: ChU Emissions from Petroleum Systems (MMT CO2 Eq.)
Activity
1990
2005
2012
2013
2014
2015
2016
Exploration3
0.8
1.0
2.8
3.0
3.3
2.2
2.1
Production (Total)
40.8
32.8
31.7
35.0
36.8
36.3
36.1
Pneumatic controller venting
19.1
16.6
14.3
17.2
17.9
18.0
18.5
Offshore platforms
5.3
4.6
4.7
4.7
4.7
4.7
4.7
Associated gas venting and
flaring
3.5
3.0
3.4
3.2
3.6
2.7
1.6
Tanks
6.4
2.0
1.4
1.6
1.9
2.1
3.2
Gas engines
2.1
1.8
2.1
2.2
2.3
2.3
2.2
Chemical injection pumps
1.2
1.7
2.0
2.1
2.1
2.1
2.0
Other Sources
3.0
3.1
3.9
4.1
4.3
4.4
3.9
Crude Oil Transportation
0.2
0.1
0.2
0.2
0.2
0.2
0.2
Refining
0.6
0.7
0.7
0.7
0.7
0.7
0.9
Total
42.3
34.7
35.4
38.8
41.0
39.4
39.3
a Exploration includes well drilling, testing, and completions.
Note: Totals may not sum due to independent rounding.
Table 3-37: ChU Emissions from Petroleum Systems (kt)
Activity
1990
2005
; 2012
2013
2014
2015
2016
Exploration3
31
39
113
120
131
89
82
Production (Total)
1,631
1,314
1,269
1,399
1,474
1,451
1,443
Pneumatic controller venting
766
663
570
687
716
721
739
Offshore platforms
211
185
188
188
188
188
188
Associated gas venting and
flaring
140
120
136
127
145
106
64
Tanks
258
84
57
65
77
82
127
Gas Engines
85
70
83
87
92
93
89
Chemical injection pumps
49
67
80
82
85
85
81
Other Sources
122
125
155
163
172
175
156
Crude Oil Transportation
7
5
6
7
8
8
8
Refining
24
28
27
27
26
28
37
Total
1,693
1,386
1,415
1,553
1,639
1,576
1,571
a Exploration includes well drilling, testing, and completions.
Note: Totals may not sum due to independent rounding.
3-66 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
Table 3-38: CO2 Emissions from Petroleum Systems (MMT CO2)
Activity
1990
2005
2012
2013
2014
2015
2016
Exploration
0.2
0.2
0.2
0.3
0.3
0.3
0.0
Production
5.9
13.1
22.0
25.8
29.2
33.7
21.8
Crude Refining
3.3
3.7
3.4
3.6
3.4
4.0
3.7
Total
9.4
17.0
25.6
29.7
32.9
38.0
25.5
Note: Totals may not sum due to independent rounding.
ible 3-39: CO2 Emissions from Petroleum Systems (kt)
Activity
1990
2005
2012
2013
2014
2015
2016
Exploration
243
207
247
255
264
262
39
Production
5,859
13,071
21,957
25,835
29,217
33,695
21,794
Crude Refining
3,282
3,726
3,425
3,605
3,414
4,014
3,710
Total
9,384
17,004
25,629
29,695
32,895
37,971
25,543
Note: Totals may not sum due to independent rounding.
Methodology
See Annex 3.5 for the Ml time series of emissions data, activity data, and emission factors, and additional
information on methods and data sources.
The estimates of CH4 emissions from petroleum systems are largely based on RY2010 through RY2016 GHGRP
data, Drillinglnfo, EPA/GRI 1996, and EPA 1999. Petroleum systems includes emission estimates for activities
occurring in petroleum systems from the oil wellhead through crude oil refining, including activities for crude oil
production field operations, crude oil transportation activities, and refining operations. Annex 3.5 provides further
detail on the emission estimates for these activities, including year-specific emission factor and activity data
information. 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).
References for emission factors include 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),
Drillinglnfo (2017), consensus of industry peer review panels, Bureau of Ocean Energy Management (BOEM)
reports and analysis of GHGRP data.
The emission factors for pneumatic controllers and chemical injection pumps were developed using GHGRP data
for reporting year 2014. The emission factors for tanks, well testing, associated gas venting and flaring, and
miscellaneous production flaring were developed using GHGRP data for reporting year 2015 and 2016. Emission
factors for hydraulically fractured (HF) oil well completions (controlled and uncontrolled) were developed using
Drillinglnfo data analyzed for the 2015 NSPS OOOOa proposal. For offshore oil production, two emission factors
were calculated using data collected for all federal offshore platforms; one for oil platforms in shallow water, and
one for oil platforms in deep water. For most sources, emission factors were held constant for the period 1990
through 2016, and trends in emissions reflect changes in activity levels. For tanks, well testing, and associated gas
venting and flaring, year-specific emission factors were developed for 2015 and 2016 and the 2015 emission factors
were applied back to 1990. Basin-specific emission factors for associated gas venting and flaring were developed
and applied for the four basins that in any year from 2011 through 2016 contributed at least 10 percent of total
emissions (on a CO2 Eq. basis) from associated gas venting and flaring in the GHGRP: Williston, Permian, Gulf
Coast, and Anadarko basins. Associated gas venting and flaring data in all other basins were combined, and
emission factors and activity factors developed for the other basins as a single group. For more information, see
Recalculations Discussion below. For miscellaneous production flaring, year-specific emission factors were
developed for 2015 and 2016, an emission factor of 0 was assumed for 1990 through 1992, and linear interpolation
was applied to develop emission factors for 1993 through 2014. Emission factors from EPA 1999 are used for all
other production and transportation activities.
Energy 3-67
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
References for activity data include Drillinglnfo (2017), 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 the GHGRP (RY2010 through RY2016).
For many sources, complete activity data were not available for all years of the time series. In such cases, one of
three approaches was employed to estimate values, consistent with IPCC good practice. Where appropriate, the
activity data were calculated from related statistics using ratios developed based on EPA/GRI 1996 and/or GHGRP
data. For floating roof tanks, the activity data were held constant from 1990 through 2016 based on EPA 1999. 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, the previous year's data were used for domestic barges and
tankers as current year were not yet available. For offshore production, the number of platforms in shallow water
and the number of platforms in deep water are used as activity data and are taken from BOEM datasets.
For the production segment, in general, CO2 emissions for each source are estimated with GHGRP data or by
multiplying CO2 emission factors by the corresponding CH4 data, as the CO2 content of gas relates to the CH4
content of gas. Sources with CO2 emissions calculated from GHGRP data are associated gas venting and flaring,
tanks, well testing, pneumatic controllers, chemical injection pumps, and miscellaneous production flaring. For these
sources, CO2 was calculated using the same methods as used for CH4. Emission factors for offshore oil production
(shallow and deep water) were derived using data from BOEM. 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 petroleum refining activities, 2010 to 2016 emissions were directly obtained from EPA's GHGRP. All U.S.
refineries have been required to report CH4 and CO2 emissions for all major activities starting with emissions that
occurred in 2010. However, GHGRP does have provisions that refineries are not required to report to the GHGRP if
their emissions fall below certain thresholds (see Planned Improvements for additional discussion). The reported
total of CH4 and CO2 emissions for each activity was used for the 2010 to 2016 emissions. 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), which were then applied to the annual
refinery feed to estimate CH4 and CO2 emissions for 1990 to 2009.
A complete list of references for emission factors and activity data by emission source is provided in Annex 3.5.
Through EPA's stakeholder process on oil and gas in the Inventory, EPA received initial stakeholder feedback on
updates under consideration for the Inventory. Stakeholder feedback is noted below in Uncertainty and Time -Series
Consistency, Recalculations Discussion, and Planned Improvements.
Uncertainty and Time-Series Consistency
In recent years, EPA has made significant revisions to the Inventory methodology to use updated activity and
emissions data. To update its characterization of uncertainty, EPA has conducted a draft quantitative uncertainty
analysis using the IPCC Approach 2 methodology (Monte Carlo Simulation technique). The 95 percent confidence
intervals presented here are based on 2015 data from the previous (i.e., 1990 through 2015) Inventory. EPA is still
seeking comment on the approach to calculate uncertainty and may update its approach in the final version of the
current Inventory. Initial stakeholder feedback on the uncertainty analysis included support for annual updates to the
uncertainty assessment, so that the uncertainty ranges will continue to reflect new data as they become available.
Stakeholders supported the approach of calculating uncertainty for the top emitters. For more information, please
see the Planned Improvements section, and the memorandum Inventory of U.S. Greenhouse Gas Emissions and
Sinks 1990-2016: Updates Under Consideration for Natural Gas and Petroleum Systems Uncertainty Estimates
(Draft 2018 Uncertainty Memo)?9
79 See
3-68 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
To develop the values in Table 3-40 below, EPA applied the uncertainty bounds calculated for the 2015 emission
estimates presented in the previous Inventory. To develop the uncertainty bounds, EPA used the IPCC Approach 2
methodology (Monte Carlo Simulation technique). Microsoft Excel's @RISK add-in tool was used to estimate the
95 percent confidence bound around methane emissions from petroleum systems. For the analysis, EPA focused on
the five highest methane-emitting sources for the year 2015, which together emitted 79 percent of methane from
petroleum systems in 2015, 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.
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 2016, using the recommended IPCC methodology. The results of the
Approach 2 uncertainty analysis are summarized in Table 3-40. Petroleum systems CH4 emissions in 2016 were
estimated to be between 26.7 and 53.4 MMT CO2 Eq., while CO2 emissions were estimated to be between 17.3 and
34.7 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.
Table 3-40: Approach 2 Quantitative Uncertainty Estimates for ChU Emissions from
Petroleum Systems (MMT CO2 Eq. and Percent)
Source
Gas
2016 Emission Estimate
(MMT CO2 Eq.)b
Uncertainty Range Relative to Emission Estimate3
(MMT CO2 Eq.) (%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Petroleum Systems
CH4
39.3
26.7 53.4
-32% +36%
Petroleum Systems
CO2
25.5
17.3 34.7
-32% +36%
a Range of emission estimates estimated by applying the 95 percent confidence intervals obtained from the Monte Carlo
Simulation analysis conducted for the year 2015.
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.
c An uncertainty analysis for the non-energy CO2 emissions was not performed. The relative uncertainty estimated (expressed
as a percent) from the CH4 uncertainty analysis was applied to the point estimate of non-energy CO2 emissions.
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 1990 through 2016, 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 2014 by interpolating activity data or emission factors or both between 1992 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.
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
Energy 3-69
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
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.80
As in previous years, EPA conducted early engagement and communication with stakeholders on updates prior to
public review. EPA held stakeholder workshops on greenhouse gas data for oil and gas in June and October of 2017,
and held webinars in April and August of 2017. In advance of each workshop, EPA released memos detailing
updates under consideration and requesting stakeholder feedback. Stakeholder feedback received through these
processes is discussed in the Recalculations Discussion and Planned Improvements sections below.
In recent years, several studies have measured emissions at the source level and at the national or regional level and
calculated emission estimates that may differ from the Inventory. There are a variety of potential uses of data from
new studies, including replacing a previous estimate or factor, verifying or QA of an existing estimate or factor, and
identifying areas for updates. In general, there are two major types of studies related to oil and gas greenhouse gas
data: studies that focus on measurement or quantification of emissions from specific activities, processes, and
equipment, and studies that use tools such as inverse modeling to estimate the level of overall emissions needed to
account for measured atmospheric concentrations of greenhouse gases at various scales. The first type of study can
lead to direct improvements to or verification of Inventory estimates. In the past few years, EPA has reviewed and in
many cases, incorporated data from these data sources. The second type of study can provide general indications on
potential over- and under-estimates. A key challenge in using these types of studies to assess Inventory results is
having a relevant basis for comparison (i.e., the independent study should assess data from the Inventory and not
another data set, such as EDGAR). In an effort to improve the ability to compare the national-level 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 degree x 0.1 degree spatial
resolution, monthly temporal resolution, and detailed scale-dependent error characterization.81 The gridded methane
inventory is designed to be consistent with the U.S. EPA Inventory of U.S. Greenhouse Gas Emissions and Sinks
(1990-2014) estimates for the year 2012, which presents national totals.82
Recalculations Discussion
The 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
June and October 2017, the EPA released draft memoranda: Inventory of U.S. Greenhouse Gas Emissions and Sinks
1990-2016: Revisions Under Consideration for CO2 Emissions {Draft 2018 CO 2 Memo),^ and Inventory of U.S.
Greenhouse Gas Emissions and Sinks 1990-2016: Additional Revisions Under Consideration {Draft 2018 Other
Updates Memo).84 The memos discussed changes under consideration, and requested stakeholder feedback on those
changes.
The EPA thoroughly evaluated relevant information available, and made updates to exploration and production
segment methodologies for the Inventory, including to define an exploration segment separate from production (not
a methodological change, but a change in presentation of information), revising activity and CH4 and CO2 emissions
data for associated gas venting and flaring, miscellaneous production flaring, and well testing. Production segment
CO2 emissions data were also revised for oil tanks, pneumatic controllers, and chemical injection pumps.
80 See
81 See
82 See
83 See < https://www. epa.gov/sites/production/files/2017-10/documents/2018_ghgi_co2_revisions_under_consideration_2017-
10-25_to_post.pdf>
84 See < https://www.epa.gov/sites/production/files/2017-10/documents/2018_ghgi_ng-
petro_revisions_under_consideration_2017- 10-26_pdf_to_post.pdf
3-70 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
The combined impact of revisions to 2015 petroleum systems CH4 emissions, compared to the previous Inventory, is
a decrease from 39.9 to 39.4 MMT CO2 Eq. (0.5 MMT CO2 Eq., or 1 percent). The recalculations resulted in an
average decrease in CH4 emission estimates across the 1990 through 2015 time series, compared to the previous
Inventory, of 11 MMT CO2 Eq, or 22 percent. The CH4 emissions estimate decrease was primarily due to
recalculations related to associated gas venting and flaring which were updated to use a basin-level approach, and
has the largest impact on years prior to 2013.
The combined impact of revisions to 2015 petroleum systems CO2 emissions, compared to the previous Inventory, is
an increase from 3.6 to 38.0 MMT CO2 (34.4 MMT CO2, or by a factor of 9). The recalculations resulted in an
average increase in emission estimates across the 1990 through 2015 time series, compared to the previous
Inventory, of 13.8 MMT CO2 Eq, or 360 percent. The CO2 emissions estimate increase was primarily due to
recalculations related to the reallocation of CO2 from flaring to petroleum systems from natural gas systems.
Previously, data were not available to disaggregate flared emissions between natural gas systems and petroleum
systems. The largest sources of CO2 from flaring are associated gas flaring, tanks with flares, and miscellaneous
production flaring.
Exploration
Petroleum systems was reorganized for the current Inventory to include an exploration segment to improve
conformance with the IPCC guidelines. Exploration activities were previously included under the production
segment. The activities included under exploration are hydraulically fractured oil well completions, oil well
completions without hydraulic fracturing, well drilling, and well testing. Of these activities, well testing was the
only source with a new methodology, which is discussed below.
Well Testing
EPA developed a new estimate for oil well testing (during non-completion events) using GHGRP data. In previous
Inventories, only well testing conducted as part of a completion event was included. CH4 and CO2 emission factors
were developed, on a per-event basis, for vented and flared oil well testing events using RY2015 and RY2016 data.
EPA developed activity factors (i.e., number of events per oil well) to determine the number of well testing events in
a year, also using RY2015 and RY2016 data. GHGRP RY2015 activity and emission factors are applied to all prior
years of the time series. Methane emissions from well testing averaged 8.1 kt (or 0.2 MMT CO2 Eq.) over the time
series. There was a large decrease in emissions from oil well testing from 2015 to 2016 as observed in reported
GHGRP data. Carbon dioxide emissions from well testing averaged 216 kt (0.2 MMT CO2) over the time series. See
the Draft 2018 Other Updates Memo for additional discussion.
Table 3-41: Oil Well Testing National ChU Emissions (Metric Tons ChU)
Source
1'WO
2005
2012
2013
2014
2015
2016
Non-Completion Well Testing -
Vented
8.043
6,819
8,022
8,272
8,559
8,567
2,811
Non-Completion Well Testing -
Flared
%l
815
959
989
1,023
1,024
157
Table 3-42: Oil Well Testing National CO2 Emissions (Metric Tons CO2)
Source I'J'JO 2005 2012 2013 2014 2015 2016~
Non-Completion Well Testing- ^ 3Q8 363 ^ ^ ^ ^
Vented
Non-Completion Well Testing- 241.362 204,643 240,754 248,234 256,853 257,101 34,481
r lared
Production
In addition to the memos discussed above, this section references the memorandum, Inventory of U.S. Greenhouse
Gas Emissions and Sinks 1990-2015: Revisions for Natural Gas and Petroleum Systems Production Emissions
Energy 3-71
-------
1 (2017 Production Memo).85 The 2017 Production Memo contains further details and documentation of
2 recalculations.
3 CO2 Updates
4 EPA updated CO2 emissions for a number of sources in the Inventory. See the Draft 2018 CO2 Memo for more
5 details. The overall impact was an average increase of 13.8 MMT CO2 (or 364 percent) over the time series in
6 petroleum systems, which is primarily due to the reallocation of flaring CO2 emissions from natural gas systems to
7 petroleum systems, which was not possible in the past because the previous data source aggregated venting and
8 flaring activity data from both petroleum and natural gas systems, but is now possible through use of the GHGRP
9 data. A stakeholder noted that the update uses the best available data for this source.
10 Sources with the largest impacts include tanks with flares, associated gas flaring, and miscellaneous production
11 flaring. These sources are discussed in detail below. Other sources (i.e., pneumatic controllers and chemical
12 injection pumps) had increases or decreases of less than 1 MMT CO2.
13 Tanks
14 EPA developed CO2 emissions estimates for oil tanks using GHGRP data and a throughput-based approach. This
15 approach is identical to the methodology to calculate CH4 emissions; for more information, please see the 2017
16 Production Memo. The overall impact of the change is an increase in calculated CO2 emissions by a factor of nine
17 on average over the time series.
18 Table 3-43: National Tank CO2 Emissions by Category and National Emissions (kt CO2)
CO2 Emissions
mo
2005
2012
2013
2014
2015
2016
Large Tanks w/ Flares
0
3,407
5,978
6,870
8,054
8,657
7,282
Large Tanks w/ VRU
0
6
11
13
15
16
11
Large Tanks w/o Control
25
"
4
5
6
6
9
Small Tanks w/ Flares
0
4
7
8
9
10
21
Small Tanks w/o Flares
9
6
7
8
8
7
Malfunctioning Dump Valves
20
14
17
20
23
25
22
Total Emissions
53
3,444
6,023
6,922
8,115
8,722
7,351
Previous Estimated Emissions
329
247
366
433
520
520
NA
NA (Not Applicable)
19 Associated Gas Venting and Flaring
20 EPA developed a new estimate for CO2 from associated gas venting and flaring. EPA's considerations for this
21 source are documented in the Draft 2018 C02 Memo. As noted above in the Methodology section, EPA used a
22 basin-level and well-based approach to calculate emissions from this source. In the Draft 2018 C02 Memo, a NEMS
23 Region-level and well-based approach was presented, however, stakeholder feedback on the Draft 2018 C02 Memo
24 supported the use of GHGRP data to calculate emissions from this source at a basin-level. EPA evaluated basin-
25 level associated gas venting and flaring data reported to GHGRP from 2011 to 2016 and developed the estimates
26 below with that approach. If a basin contributed at least 10 percent of total annual emissions (on a CO2 Eq. basis)
27 from associated gas venting and flaring in any year, then basin-specific emission factors and activity factors were
28 developed. Four basins met this criteria: Williston, Permian, Gulf Coast, and Anadarko. Associated gas venting and
29 flaring data in all other basins were combined, and emission factors and activity factors developed for the other
30 basins as a single group. For each basin or group, emission factors were calculated for 2015 and 2016; the 2015
31 emission factors were applied to all prior years. Two activity factors were also calculated for each basin or group:
32 the percent of oil wells that either flare or vent associated gas and, for those wells in that category (those that vent or
33 flare associated gas), the fraction that vents and the fraction that flares. The percent of oil wells that flare or vent
34 associated gas was calculated for 2015 and 2016, and the 2015 activity factors applied to all prioryears. The specific
35 fractions that vent and flare associated gas were developed for 2011 through 2016, and the 2011 fractions applied to
85 Available at
3-72 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
all prior years. Stakeholders also suggested a production-based approach for the basin-level analysis, and EPA will
continue evaluating this approach for the final version of the current Inventory (as noted in the Planned
Improvements section). Stakeholders have noted that past (e.g., 1990 through 2010) associated gas venting and
flaring likely varied significantly from year to year and from region to region. However, data are not presently
available to take variation prior to 2011 into account.
Table 3-44: Associated Gas Venting and Flaring National CO2 Emissions (kt CO2)
Source
Associated Gas Well Venting ^ ^ ^
Emissions
Associated Gas Well Flaring .
Emissions "'
Previous Estimated emissions j
from stripper wells
2005
2012
2013
2014
2015
2016
95
89
75
83
41
19
5,102
8,820
11,582
13,510
17,414
10,137
I
1
1
1
1
NA
NA (Not Applicable)
The CH4 methodology was previously developed for the previous Inventory and used a national-level approach.
EPA updated its CH4 calculations for associated gas venting and flaring to be consistent with the basin-level
approach to calculate CO2 emissions from this source. Overall, the change decreased calculated CH4 emissions over
the time series by around 70 percent for both associated gas venting and associated gas flaring, with the largest
decreases occurring early in the time series.
Table 3-45: Associated Gas Venting and Flaring National ChU Emissions (Metric Tons ChU)
Source Two 2005 2012 2013 2014 2015 2016
Associated Gas Well Venting
Emissions
Associated Gas Well Flaring
Emissions
Previous Associated Gas Well
Venting Emissions
Previous Associated Gas Well
Flaring Emissions
119.697 101,157 105,312 87,013 97,477 44,636 31,040
20.769 18,380 31,024 40,203 47,101 61,576 33,015
608,758 511,701 482,816 214,665 89,333 42,518 NA
76,176 64,031 104,513 146,292 149,694 105,706 NA
NA (Not Applicable)
Miscellaneous Production Flaring
The EPA developed new estimates for CO2 and CH4 emissions from miscellaneous production flaring using
GHGRP subpart W data. Along with other updates to flaring emissions in both oil and gas production, this replaces
the estimate for flaring that was previously reported in the natural gas systems emissions totals. EPA developed
emission factors from 2015 and 2016 GHGRP data; the 2015 emission factor is applied to all prior years. The
emission factors are on a per-well basis and were applied to all oil wells in each year. Details are provided in the
Draft 2018 CO2 Memo. Initial stakeholder feedback on this update suggested use of production-based emission
factors as opposed to well-based emission factors.
Table 3-46: Miscellaneous Production Flaring National CO2 Emissions (kt CO2)
Source 19911 2005 2012 2013 2014 2015 2016
Miscellaneous Production
Flaring
Previous Estimated emissions
from flaring (natural gas and
petroleum)"
0
4,349
6,944
7,160
7,409
7,416
4,183
9,093
7,193
12,704
15,684
17,629
17,629
NA
a The previous estimated emissions from flaring were reported under Natural Gas Systems and included
emissions from multiple sources, including associated gas, and natural gas systems, but is provided for reference.
NA (Not Applicable)
Energy 3-73
-------
1 Table 3-47: Miscellaneous Production Flaring National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2012
2013
2014
2015
2016
Miscellaneous Production
Flaring
Previous Estimated emissions
from flaring"
I)
14,889
23,773
24,511
25,362
25,387
13,844
0
0
0
0
0
0
NA
a Prior Inventories did not estimate methane emissions from a source similar to miscellaneous production flaring.
NA (Not Applicable)
2 Activity Data Updates
3 Well Counts
4 EPA has used a more recent version of the Drillinglnfo data set to update well counts data in the Inventory. There
5 are not methodological changes to this source in the current Inventory or major changes to the activity data, but
6 because this is a key input, results are highlighted here.
7 Table 3-48: Producing Oil Well Count Data
Oil Well Count
1990
2005
2012
2013
2014
2015
2016
Number of Oil Wells
Previous Estimate
553,899
572,639
469,632
481,340
552,504
564,348
569,670
580,960
589,450
598,627
590,017
586,896
561,964
NA
NA (Not Applicable)
8 In December 2017, EIA released a 2000 through 2016 time series of national oil and gas well counts. EIA total (oil
9 and gas) well counts for 2016 were 1,010,441. EPA's total well counts were 978,845. Over the 2000 to 2016 time
10 series, EPA's well counts were on average 2 percent lower than EIA's. EIA's well counts include side tracks,
11 completions, and recompletions, and therefore are expected to be higher than EPA's which include only producing
12 wells. EPA and EIA use a different threshold for distinguishing between oil versus gas (EIA uses 6 mcf/bbl, while
13 EPA uses 100 mcf/bbl), which results in EIA having a lower fraction of oil wells and a higher fraction of gas wells
14 than EPA. Across the 2000 through 2016 EIA time series, EIA estimates on average 111,420 (or 20 percent) fewer
15 oil wells in each year than EPA.
16 Equipment Counts
17 EPA recalculated activity factors of equipment per well using the latest GHGRP RY2015 data, which included some
18 resubmissions. This resulted in minor changes across the time series. For example, the number of heater/treaters per
19 well decreased by 9 percent over the time series, the number of separators and headers per well decreased by 4
20 percent and 3 percent, respectively, while chemical injection pumps and pneumatic controllers per well increased by
21 4 percent and less than 1 percent, respectively. The impact of the changes in equipment counts per well along with
22 changes in well counts resulted in minor changes in methane emissions across the time series for heater/treaters (-12
23 percent), separators (17 percent), headers (-5 percent), pneumatic controllers (-2 percent), and chemical injection
24 pumps (2 percent).
25 Transportation
26 Recalculations due to updated activity data for quantity of petroleum transported by barge or tanker in the
27 transportation segment have resulted in an average decrease in calculated emissions over the time series from this
28 segment of less than 0.01 percent.
29 Refining
30 Recalculations due to resubmitted GHGRP data-in particular from flaring in the refining segment have resulted in an
31 average increase in calculated CH4 emissions over the time series from this segment of 3 percent and an average
32 increase in calculated CO2 emissions over the time series of 6 percent.
3-74 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Planned improvements
2 Plans for Final 2018 Inventory (1990 through 2016) and Future Inventories
3 Associated Gas Venting and Flaring
4 For the final version of the current Inventory, EPA is considering applying a production-based approach at the
5 basin-level, as opposed to a well-based approach at the basin-level, to calculate emissions for this source based on
6 stakeholder feedback. Preliminary analysis of emissions calculated using a basin-level, production-based approach
7 indicates that the emission estimates will be lower than those calculated with the basin-level, well-based approach.
8 EPA applied a basin-level approach to avoid potential overestimates of venting and flaring activity that might occur
9 by applying data from GHGRP at a national level without basin-specific adjustments. EPA seeks feedback on this
10 approach and how other approaches might avoid over- or underestimating emissions from this source.
11 Uncertainty
12 The uncertainty analysis results presented for this public review Inventory were based on the top five methane-
13 emitting sources for 2015 from the previous Inventory. EPA will re-evaluate the highest emitting sources, based on
14 the final version of the current Inventory, and update the uncertainty analysis to reflect these sources and their
15 methodology, as necessary. EPA will also consider further stakeholder feedback on the Draft 2018 Uncertainty
16 Memo.
17 Miscellaneous Production Flaring
18 Miscellaneous production flaring emission factors are currently applied on a well-basis at the national-level. EPA is
19 considering two additional options for the final version of the current Inventory, based on stakeholder feedback: a
20 production-based approach and developing factors at a basin-level. Each of these options are being considered to
21 avoid over- or underestimating emissions from this source.
22 Refineries
23 The GHGRP includes provisions at 40 CFR 98.2(i) that allows facilities to discontinue complying with the GHGRP
24 if their emissions fall below certain thresholds. EPA is assessing to what extent this provision has affected the
25 subpart Y reported emissions. If certain refineries are not reporting emissions to the GHGRP, options to address this
26 will be considered.
27 Offshore Platforms
28 EPA is considering updates to the offshore platform emissions calculation methodology, as discussed in the Draft
29 2018 Other Updates Memo. The current emission factors were based on data from the 2011 DOI/Bureau of Ocean
30 Energy Management's (BOEM) Gulf Offshore Activity Data System (GOADS), and 2014 GOADS data is
31 available. A different source for platform counts is also being considered.
32 N2() Emissions
33 N20 emissions are currently not included in petroleum systems estimates, but EPA is considering developing a
34 methodology to estimate N2O emissions. The Draft 2018 Other Updates Memo provides discussion on this topic.
35 EPA will consider options such as using GHGRP data for sources that already rely on GHGRP data for CH4 or CO2
36 estimates. GHGRP RY2015 reported N20 flaring emissions specific to petroleum systems were 124 metric tons (or
37 0.04 MMT CO2 Eq. In addition, 36 metric tons N2O (or 0.01 MMT CO2 Eq.) flaring emissions were reported for
38 GHGRP RY2015 for sources that fall within both natural gas and petroleum systems.
39 Well-Related Activity Data
Energy 3-75
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
As described in the Recalculations Discussion EPA lias updated the emission factors for certain well-related
emission sources, including well testing. EPA will continue to assess available data, including data from the
GHGRP and stakeholder feedback on considerations, to improve activity estimates for these types of sources.
Upcoming Data, and Additional Data that Could Inform the Inventory
EPA will continue to review data available from the GHGRP, in particular new data on hydraulically fractured oil
well completions and workovers and new well-specific information, available in 2017 for the first time (for
RY2016). EPA will consider revising its methods to take into account the new GHGRP data.
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, such as an upcoming field
study by API on pneumatic controllers. 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, stakeholder comments have highlighted areas where additional data that could inform the Inventory are
currently limited or unavailable:
• Tank malfunction and control efficiency data. See Tanks in Recalculations Discussion.
• 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. See Associated Gas Venting
and Flaring in Recalculations Discussion.
• Refineries emissions data. One stakeholder noted a recent study (Lavoie et al. 2017) that measured three
refineries and found higher average emissions than in the Inventory, and the stakeholder suggested that
EPA evaluate the study and any additional information available on this source.
One stakeholder suggested that the Inventory should be updated with site-level and basin-level data, noting the EPA
could first use basin-level data to assess the inventory, and that future research could focus on collecting data in
basins with the largest discrepancies.
EPA will continue to seek available data on these and other sources as part of the process to update the Inventory.
Box 3-7: 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. 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 from naturally-produced CO2 are estimated based on the specific application.
In the Inventory, CO2 that 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. The naturally-occurring CO2 used in EOR operations is
assumed to be fully sequestered. Additionally, all anthropogenic CO2 emitted from natural gas processing and
ammonia plants is assumed to be emitted to the atmosphere, regardless of whether the CO2 is captured or not. These
emissions are currently included in the Natural Gas Systems and the Ammonia Production sections of the Inventory
report, respectively.
IPCC includes methodological guidance to estimate emissions from the capture, transport, injection, and geological
storage of CO2. The methodology is based on the principle that the carbon capture and storage system should be
handled in a complete and consistent manner across the entire Energy sector. The approach accounts for CO2
captured at natural and industrial sites as well as emissions from capture, transport, and use. For storage specifically,
a Tier 3 methodology is outlined for estimating and reporting emissions based on site-specific evaluations. However,
IPCC (IPCC 2006) notes that if a national regulatory process exists, emissions information available through that
process may support development of CO2 emission estimates for geologic storage.
3-76 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 In the United States, facilities that produce CO2 for various end-use applications (including capture facilities such as
2 acid gas removal plants and ammonia plants), importers of CO2, exporters of CO2, facilities that conduct geologic
3 sequestration of CO2, and facilities that inject CO2 underground, are required to report greenhouse gas data annually
4 to EPA through its GHGRP. Facilities conducting geologic sequestration of CO2 are required to develop and
5 implement an EPA-approved site-specific monitoring, reporting and verification plan, and to report the amount of
6 CO2 sequestered using a mass balance approach.
7 GHGRP data relevant for this inventory estimate consists of national-level annual quantities of CO2 captured and
8 extracted for EOR applications for 2010 to 2016. However, for 2015 and 2016, data from EPA's GHGRP (Subpart
9 PP) were unavailable for use in the current Inventory report due to data confidentiality reasons. The estimate for
10 2014 was held constant here to estimate 2015 and 2016 emissions. EPA will continue to evaluate the availability of
11 additional GHGRP data and other opportunities for improving the emission estimates. For reporting year 2016, one
12 facility reported data to the GHGRP under subpart RR (Geologic Sequestration of Carbon Dioxide). This facility
13 reported 3.1 MMT of CO2 sequestered in subsurface geological formations and 56 metric tons of CO2 emitted from
14 surface equipment leaks and vents.
15 These estimates indicate that the amount of CO2 captured and extracted from natural and industrial sites for EOR
16 applications in 2016 is 59.3 MMT CO2 Eq. (59,318 kt) (see Table 3-49 and Table 3-50). Site-specific monitoring
17 and reporting data for CO2 injection sites (i.e., EOR operations) were not readily available, therefore, the quantity of
18 CO2 captured and extracted is noted here for information purposes only; CO2 captured and extracted from industrial
19 and commercial processes is assumed to be emitted and included in emissions totals from those processes.
20 Table 3-49: Quantity of CO2 Captured and Extracted for EOR Operations (MMT CO2)
Stage
1990
2005
2012
2013
2014
2015
2016
Capture Facilities
4.8
6
9.3
12.2
13.1
13.1
13.1
Extraction Facilities
20.8
28.3
48.9
47.0
46.2
46.2
46.2
Total
25.6
34.7
58.1
59.2
59.3
59.3
59.3
Note: Totals may not sum due to independent rounding.
21 Table 3-50: Quantity of CO2 Captured and Extracted for EOR Operations (kt)
Stage 1990 2005 2012 2013 2014 2015 2016
Capture Facilities
4,832
6,475
9,267
12,205
13,093
13,093
13,093
Extraction Facilities
20,811
28,267
48,869
46,984
46,225
46,225
46,225
Total
25,643
34,742
58,136
59,189
59,318
59,318
59,318
Note: Totals may not sum due to independent rounding.
22
23 3.7 Natural Gas Systems (CRF Source Category
24 lB2b)
25 The U.S. natural gas system encompasses hundreds of thousands of wells, hundreds of processing facilities, and
26 over a million miles of transmission and distribution pipelines. Overall, natural gas systems emitted 162.1 MMT
27 CO2 Eq. (6,483 kt) of CH4 in 2016, a 16 percent decrease compared to 1990 emissions, and a 1.4 percent decrease
28 compared to 2015 emissions (see Table 3-51, Table 3-52, and Table 3-53) and 26.7 MMT CO2 Eq. (26,739 kt) of
29 non-combustion CO2 in 2016, a 10 percent decrease compared to 1990 emissions.
30 The 1990 to 2016 trend in CH4 is not consistent across segments. Overall, the 1990 to 2016 decrease in CH4
31 emissions is due primarily to the decrease in emissions from distribution (75 percent decrease), transmission and
32 storage (44 percent decrease), processing (48 percent decrease), and exploration (81 percent decrease) segments.
33 Over the same time period, the production segments saw increased methane emissions of 60 percent (with onshore
Energy 3-77
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
production emissions increasing 30 percent, offshore production emissions increasing 7 percent, and gathering and
boosting emissions increasing 103 percent). The 1990 to 2016 decrease in CO2 is due primarily to decreases in acid
gas removal emissions in the processing segment, where acid gas removal emissions per plant have decreased over
time.
Methane and non-combustion 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, bleed and discharge emissions from pneumatic controllers, and fugitive 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.
Below is a characterization of the five major stages of the natural gas system. Each of the stages is described and the
different factors affecting CH4 and non-combustion CO2 emissions are discussed.
Exploration. Exploration includes well drilling, testing, and completions. Emissions from exploration account for
less than 1 percent of CH4 emissions and 2 percent of non-combustion CO2 emissions from natural gas systems in
2016. Well completions account for most of the CH4 emissions in 2016, with well testing and drilling also
contributing emissions. Flaring emissions account for most of the non-combustion CO2 emissions. Methane
emissions from exploration decreased by 81 percent from 1990 to 2016, with the largest decreases coming from
hydraulically fractured gas well completions without reduced emissions completions (RECs) or flaring. Carbon
dioxide emissions from exploration increased by 74 percent from 1990 to 2016 due to increases in flaring.
Production (including gathering and boosting). In the production stage, wells are used to withdraw raw gas from
underground formations. Emissions arise from the wells themselves, and well-site gas treatment equipment such as
dehydrators and separators. 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). Emissions from production (including gathering and boosting) account
for 66 percent of CH4 emissions and 15 percent of non-combustion CO2 emissions from natural gas systems in 2016.
Emissions from gathering stations, pneumatic controllers, gas engines, liquids unloading, and offshore platforms
account for most of the CH4 emissions in 2016. Flaring emissions account for most of the non-combustion CO2
emissions with the highest emissions coming from miscellaneous flaring, flaring from tanks, offshore flaring, and
flaring at workovers. Methane emissions from production increased by 60 percent from 1990 to 2016, due primarily
to increases in emissions from gathering and boosting stations (driven by an increase in gas production), increases in
emissions from pneumatic controllers (due to an increase in the number of controllers, particularly in the number of
intermittent bleed controllers), and gas engines. Carbon dioxide emissions from production increased by a factor of
4.9 from 1990 to 2016 due to increases in flaring.
Processing. In this stage, 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. Fugitive CH4 emissions from compressors,
including compressor seals, are the primary emission source from this stage. Most of the non-combustion CO2
emissions come from acid gas removal (AGR) units, which are designed to remove CO2 from natural gas.
Processing plants account for 7 percent of CH4 emissions and 82 percent of non-combustion CO2 emissions from
natural gas systems. Methane emissions from processing decreased by 48 percent from 1990 to 2016 as emissions
from compressors (leaks and venting) and equipment leaks decreased. Carbon dioxide emissions from processing
decreased by 22 percent from 1990 to 2016, due to a decrease in acid gas removal emissions.
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 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). In 2016, emissions from the final months of
the Aliso Canyon leak event in Southern California contributed 0.5 MMT CO2 Eq. to transmission and storage
emissions, around 2 percent of total emissions for this segment. Compressors and dehydrators are the primary
contributors to emissions from storage. Methane emissions from the transmission and storage sector account for
3-78 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
approximately 20 percent of emissions from natural gas systems, while CO2 emissions from transmission and
storage account for less than 1 percent of the non-combustion CO2 emissions from natural gas systems. CH4
emissions from this source decreased by 44 percent from 1990 to 2016 due to reduced compressor station emissions
(including emissions from compressors and leaks). CO2 emissions from transmission and storage have decreased by
14 percent from 1990 to 2016, also due to reduced compressor station emissions.
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 1,284,241 miles of distribution mains in 2016, an increase of over 340,000 miles since 1990
(PHMSA 2017a; PHMSA 2017b). Distribution system emissions, which account for 7 percent of CH4 emissions
from natural gas systems and less than 1 percent of non-combustion 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 CH4 emissions in 2016 were 75 percent lower than 1990 levels
(changed from 43.5 MMT CO2 Eq. to 11.0 MMT CO2 Eq.), while distribution CO2emissions in 2016 were 72
percent lower than 1990 levels (CO2 emission from this segment are less than 0.1 MMT CO2 Eq. across the time
series).
Total CH4 emissions for the five major stages of natural gas systems are shown in MMT CO2 Eq. (Table 3-51) and
kt (Table 3-52). Table 3-53 provides additional information on how the estimates in Table 3-49 were calculated.
With recent updates to the Inventory, most emissions are calculated using a net emission approach. However,
certain sources are still calculated with a potential emission approach. Table 3-53 shows the calculated potential
CH4 release (i.e., potential emissions before any controls are applied) from each stage, and the amount of CH4 that is
estimated to have been flared, captured, or otherwise controlled, and therefore not emitted to the atmosphere.
Subtracting the value for CH4 that is controlled, from the value for calculated potential release of CH4, results in the
total net emissions values. More disaggregated information on potential emissions and emissions is available in
Annex 3.6. See Methodology for Estimating CH4 and CO2 Emissions from Natural Gas Systems.
Table 3-51: ChU Emissions from Natural Gas Systems (MMT CO2 Eq.)a
Stage
1990
2005
2012
2013
2014
2015
2016
Exploration6
3.5
9.7
3.5
2.6
2.6
1.0
0.7
Production
66.7
85.9
104.1
104.2
107.2
107.4
106.6
Onshore Production
34.8
47.7
49.6
48.8
47.8
45.6
45.2
Offshore Production
3.5
4.3
3.8
3.8
3.8
3.8
3.8
Gathering and Boosting0
28.4
33.8
50.7
51.6
55.6
58.1
57.7
Processing
21.3
11.6
10.0
10.8
11.0
11.0
11.2
Transmission and Storage
58.6
30.8
27.9
30.8
32.2
34.0
32.7
Distribution
43.5
22.1
11.3
11.2
11.2
11.0
11.0
Total
193.7
160.0
156.8
159.6
164.2
164.4
162.1
a These values represent CH4 emitted to the atmosphere. CH4 that is captured, flared, or otherwise controlled
(and not emitted to the atmosphere) has been calculated and removed from emission totals.
b Exploration includes well drilling, testing, and completions.
c Gathering and boosting includes gathering and boosting stations, gathering pipeline leaks, and gathering and
boosting station episodic events.
Note: Totals may not sum due to independent rounding.
Table 3-52: ChU Emissions from Natural Gas Systems (kt)a
Stage
1990
2005
2012
2013
2014
2015
2016
Explorationb
142
387
138
105
104
39
26
Production
2,669
3,435
4,165
4,169
4,288
4,296
4,264
Onshore Production
1,392
1,907
1,985
1,954
1,912
1,822
1,806
Offshore Production
141
173
151
151
151
151
151
Gathering and Boosting0
1,136
1,354
2,029
2,064
2,226
2,324
2,307
Processing
853
463
401
430
441
441
448
Transmission and Storage
2,343
1,230
1,117
1,232
1,287
1,360
1,306
Distribution
1,741
884
451
450
447
441
439
Total
7,748
6,399
6,273
6,385
6,568
6,578
6,483
Energy 3-79
-------
a These values represent CH4 emitted to the atmosphere. CH4 that is captured, flared, or otherwise controlled (and
not emitted to the atmosphere) has been calculated and removed from emission totals.
b Exploration includes well drilling, testing, and completions.
c Gathering and boosting includes gathering and boosting stations, gathering pipeline leaks, and gathering and
boosting station episodic events.
Note: Totals may not sum due to independent rounding.
1 Table 3-53: Calculated Potential CH4 and Captu red/Com busted Cm from Natural Gas
2 Systems (MMT CO2 Eq.)
1990
2005
2012
2013
2014
2015
2016
Calculated Potential3
193.7
179.8
176.5
178.1
182.6
182.9
180.3
Exploration
3.5
9.7
3.5
2.6
2.6
1.0
0.7
Production
66.7
92.1
113.3
113.3
116.3
116.5
115.5
Processing
21.3
11.6
10.0
10.8
11.0
11.0
11.2
Transmission and Storage
58.6
43.2
37.3
39.1
40.5
42.3
40.9
Distribution
43.5 |
23.3
12.4
12.3
12.2
12.0
12.0
Captured/Combusted
NA i
19.8
19.6
18.4
18.4
18.4
18.4
Exploration
NA i
NA
NA
NA
NA
NA
NA
Production
NA
6.2
9.1
9.1
9.1
9.1
8.9
Processing
NA
NA
NA
NA
NA
NA
NA
Transmission and Storage
NA
12.4
9.4
8.3
8.3
8.3
8.3
Distribution
NA
1.2
1.1
1.0
1.0
1.0
1.0
Net Emissions
193.7
160.0
156.8
159.6
164.2
164.4
162.1
Exploration
3.5
9.7
3.5
2.6
2.6
1.0
0.7
Production
66.7
85.9
104.1
104.2
107.2
107.4
106.6
Processing
21.3
11.6
10.0
10.8
11.0
11.0
11.2
Transmission and Storage
58.6
30.8
27.9
30.8
32.2
34.0
32.7
Distribution
43.5
22.1
11.3
11.2
11.2
11.0
11.0
a In this context, "potential" means the total emissions calculated before voluntary reductions and regulatory
controls are applied.
NA (Not Applicable)
Note: Totals may not sum due to independent rounding.
3 Table 3-54: Non-combustion CO2 Emissions from Natural Gas Systems (MMT)
Stage
1990
2005
2012
2013
2014
2015
2016
Exploration
0.3
1.4
1.6
1.5
1.9
1.1
0.6
Production
0.8
2.1
3.6
3.8
3.9
4.0
4.0
Processing
28.3
18.9
19.1
20.5
21.0
21.0
22.0
Transmission and Storage
0.2
0.1
0.1
0.1
0.1
0.1
0.1
Distribution
0.1
+
+
+
+
+
+
Total
29.7
22.5
24.4
26.0
27.0
26.3
26.7
+ Does not exceed 0.1 MMT CO2 Eq.
Note: Totals may not sum due to independent rounding.
4 Table 3-55: Non-combustion CO2 Emissions from Natural Gas Systems (kt)
Stage
1990
2005
2012
2013
2014
2015
2016
Exploration
328
1,406
1,568
1,517
1,874
1,101
571
Production
825
2,082
3,560
3,822
3,924
4,023
4,002
Processing
28,338
18,875
19,120
20,508
21,044
21,044
22,009
Transmission and Storage
166
140
135
142
148
147
143
Distribution
51
27
15
14
14
14
14
Total
29,708
22,529
24.398
26,004
27,004
26,329
26,739
Note: Totals may not sum due to independent rounding.
3-80 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Methodology
2 See Annex 3.6 for the Ml time series of emissions data, activity data, and emission factors, and additional
3 information on methods and data sources.
4 The methodology for natural gas emission estimates in the Inventory involves the calculation of CH4 and CO2
5 emissions for over 100 emissions sources, and then the summation of emissions for each natural gas segment.
6 The approach for calculating emissions for natural gas systems generally involves the application of emission factors
7 to activity data. For most sources, the approach uses technology-specific emission factors or emission factors that
8 vary over time and take into account changes to technologies and practices, which are used to calculate net
9 emissions directly. For others, the approach uses what are considered "potential methane factors" and reduction data
10 to calculate net emissions.
11 Emission Factors. Key references for emission factors for CH4 and non-combustion-related CO2 emissions from the
12 U.S. natural gas industry include a 1996 study published by the Gas Research Institute (GRI) and EPA (GRI/EPA
13 1996), the EPA's Greenhouse Gas Reporting Program (GHGRP 2017), and others.
14 The EPA/GRI study developed over 80 CH4 emission factors to characterize emissions from the various components
15 within the operating stages of the U.S. natural gas system. The EPA/GRI study was based on a combination of
16 process engineering studies, collection of activity data, and measurements at representative gas facilities conducted
17 in the early 1990s. Year-specific natural gas CH4 compositions are calculated using U.S. Department of Energy's
18 Energy Information Administration (EIA) annual gross production for National Energy Modeling System (NEMS)
19 oil and gas supply module regions in conjunction with data from the Gas Technology Institute (GTI, formerly GRI)
20 Unconventional Natural Gas and Gas Composition Databases (GTI 2001). These year-specific CH4 compositions are
21 applied to emission factors, which therefore may vary from year to year due to slight changes in the CH4
22 composition for each NEMS region.
23 GHGRP Subpart W data were used to develop both CH4 and CO2 emission factors for several sources in the
24 Inventory. In the onshore production segment, GHGRP data were used to develop emission factors used for all time
25 series years for well testing, gas well completions and workovers with and without hydraulic fracturing, pneumatic
26 controllers and chemical injection pumps, condensate tanks, liquids unloading, and miscellaneous flaring. In the
27 processing segment, for recent years of the times series, GHGRP data were used to develop emission factors for
28 fugitives, compressors, flares, dehydrators, and blowdowns/venting. In the transmission and storage segment, for
29 recent years of the times series, GHGRP data were used to develop factors for pneumatic controllers.
30 Other data sources used for CH4 emission factors include Marchese et al. (2015) for gathering stations, Zimmerle et
31 al. (2015) for transmission and storage station fugitives and compressors, and Lamb et al. (2015) for recent years for
32 distribution pipelines and meter/regulator stations.
33 For sources in the exploration, production and processing segments that use emission factors not directly calculated
34 from GHGRP data, data from the 1996 GRI/EPA study and a 2001 GTI publication were used to adapt the CH4
35 emission factors into non-combustion related CO2 emission factors. For sources in the transmission and storage
36 segment that use emission factors not directly calculated from GHGRP data, and for sources in the distribution
37 segment, data from the 1996 GRI/EPA study and a 1993 GTI publication were used to adapt the CH4 emission
38 factors into non-combustion related CO2 emission factors. See Annex 3.6 for more detailed information on the
39 methodology and data used to calculate CH4 and non-combustion CO2 emissions from natural gas systems.
40 Activity Data. Activity data were taken from various published data sets, as detailed in Annex 3.6. Key activity data
41 sources include data sets developed and maintained by EPA's GHGRP; Drillinglnfo, Inc.; U.S. Department of the
42 Interior's Bureau of Ocean Energy Management, Regulation and Enforcement (BOEMRE, previously Minerals and
43 Management Service); Federal Energy Regulatory Commission (FERC); EIA; the Natural Gas STAR Program
44 annual emissions savings data; Oil and Gas Journal; PHMSA; the Wyoming Conservation Commission; and the
45 Alabama State Oil and Gas Board.
46 For a few sources, recent direct activity data are not available. For these sources, either 2015 data were used as a
47 proxy for 2016 data, or a set of industry activity data drivers was developed and used to calculate activity data over
48 the time series. Drivers include statistics on gas production, number of wells, system throughput, miles of various
Energy 3-81
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
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 certain sectors, some sources are calculated using
potential emission factors, and the step of deducting CH4 that is not emitted from the total CH4 potential estimates to
develop net CH4 emissions is applied. 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 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 for certain sources in the production, transmission, and
distribution segments.
In fall of 2015, a well in a California storage field began leaking methane at an initial average rate of around 50
metric tons (MT) of methane (CH4) an hour, and continued leaking until it was permanently sealed in February of
20 1 6.86 An emission estimate from the leak event was included for 2015 and 2016, using the estimate of the leak
published by the California Air Resources Board (99,638 MT CH4 for the duration of the leak). The 2015 and 2016
emission estimates of 78,350 MT CH4 and 21,288 MT CH4, respectively, were added to the 2015 and 2016
estimates of fugitive emissions from storage wells. For more information, please see Inventory of U.S. Greenhouse
Gas Emissions and Sinks 1990-2015: Update for Storage Segment Emissions,87
Through EPA's stakeholder process on oil and gas in the Inventory, EPA received initial stakeholder feedback on
updates under consideration for the Inventory. Stakeholder feedback is noted below in Uncertainty and Time -Series
Consistency, Recalculations Discussion, and Planned Improvements.
Uncertainty and Time-Series Consistency
In recent years, EPA has made significant revisions to the Inventory methodology to use updated activity and
emissions data. To update its characterization of uncertainty, EPA has conducted a draft quantitative uncertainty
analysis using the IPCC Approach 2 methodology (Monte Carlo Simulation technique). The 95 percent confidence
intervals presented here are based on 2015 data from the previous (i.e., 1990 through 2015) Inventory. EPA is still
seeking comment on the approach to calculate uncertainty and may update its approach in the current Inventory.
Initial stakeholder feedback on the uncertainty analysis included support for annual updates to the uncertainty
assessment, so that the uncertainty ranges will continue to reflect new data as they become available. Similarly, a
stakeholder cautioned against not updating the uncertainty range to reflect updated data, in particular for
transmission and storage, in future Inventories. Stakeholders supported the approach of calculating uncertainty for
the top emitters. For more information, please see the Planned Improvements section, and the Draft 2018
Uncertainty Memo88.
To develop the values in Table 3-56 below, EPA has applied the uncertainty bounds calculated for the 2015
emission estimates presented in the previous Inventory. To develop the uncertainty bounds, EPA used the IPCC
Approach 2 methodology (Monte Carlo Simulation technique). Microsoft Excel's @RISK add-in tool was used to
estimate the 95 percent confidence bound around methane emissions from natural gas systems. For the analysis,
EPA focused on the 14 highest-emitting sources for the year 2015, which together emitted 77 percent of methane
from natural gas systems in 2015, and extrapolated the estimated uncertainty for the remaining sources. The @RISK
86 For more information on the Aliso Canyon event, and the measurements conducted of the leak, please see Ensuring Safe and
Reliable Underground Natural Gas Storage, Final Report of the Interagency Task Force on Natural Gas Storage Safety, available
at .
87 -
88 See < https://www.epa.gov/sites/production/files/2017-
10/documents/revision_under_consideration_for_ghgi_ng_and_petro_uncertainty_2017-10-25_to_post.pdf>
3-82 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
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 2016, using the IPCC methodology. The results of the Approach 2
uncertainty analysis are summarized in Table 3-56. Natural gas systems CH4 emissions in 2016 were estimated to be
between 137.9 and 190.4 MMT CO2 Eq. at a 95 percent confidence level. Natural gas systems non-energy CO2
emissions in 2016 were estimated to be between 22.8 and 31.4 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.
Table 3-56: Approach 2 Quantitative Uncertainty Estimates for CH4 and Non-energy CO2
Emissions from Natural Gas Systems (MMT CO2 Eq. and Percent)
Source
Gas
2016 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT CO2 Eq.)b
(MMT CO2 Eq.)
(%)
Lower Upper
Boundb Boundb
Lower Upper
Boundb Boundb
Natural Gas Systems
CH4
162.1
137.9 190.4
-15% +17%
Natural Gas Systems0
CO2
26.7
22.8 31.4
-15% +17%
a Range of emission estimates estimated by applying the 95 percent confidence intervals obtained from the Monte Carlo
Simulation analysis conducted for the year 2015.
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 3-51 and Table 3-52.
c An uncertainty analysis for the non-energy CO2 emissions was not performed. The relative uncertainty estimated (expressed
as a percent) from the CH4 uncertainty analysis was applied to the point estimate of non-energy CO2 emissions.
GHGRP data available (starting in 2011) and other recent data sources have improved estimates of emissions from
natural gas systems. To develop a consistent time series for 1990 through 2016, 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.
QA/QC and Verification Discussion
The natural gas 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
Energy 3-83
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
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.89
As in previous years, EPA conducted early engagement and communication with stakeholders on updates prior to
public review. EPA held stakeholder workshops on greenhouse gas data for oil and gas in June and October of
2017, and held webinars in April and August of 2017. In advance of each workshop, EPA released memos detailing
updates under consideration and requesting stakeholder feedback. Stakeholder feedback received through these
processes is discussed in the Recalculations Discussion and Planned Improvements sections below.
In recent years, several studies have measured emissions at the source level and at the national or regional level and
calculated emission estimates that may differ from the Inventory. There are a variety of potential uses of data from
new studies, including replacing a previous estimate or factor, verifying or QA of an existing estimate or factor, and
identifying areas for updates. In general, there are two major types of studies related to oil and gas greenhouse gas
data: studies that focus on measurement or quantification of emissions from specific activities, processes and
equipment, and studies that use tools such as inverse modeling to estimate the level of overall emissions needed to
account for measured atmospheric concentrations of greenhouse gases at various scales. The first type of study can
lead to direct improvements to or verification of Inventory estimates. In the past few years, EPA has reviewed and in
many cases, incorporated data from these data sources. The second type of study can provide general indications on
potential over- and under-estimates. A key challenge in using these types of studies to assess Inventory results is
having a relevant basis for comparison (i.e., the independent study should assess data from the Inventory and not
another data set, such as EDGAR.). In an effort to improve the ability to compare the national-level 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.90 The gridded methane inventory is
designed to be consistent with the 2016 Inventory of U.S. Greenhouse Gas Emissions and Sinks (1990-2014)
estimates for the year 2012, which presents national totals.91
Recalculations Discussion
The 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
June and October 2017, the EPA released draft memoranda, Inventory of U.S. Greenhouse Gas Emissions and Sinks
1990-2016: Revisions Under Consideration for CO2 Emissions {Draft 2018 CO2 Memo),92 Inventory of U.S.
Greenhouse Gas Emissions and Sinks 1990-2016: Updates Under Consideration for Natural Gas and Petroleum
Systems Uncertainty Estimates {Draft 2018 Uncertainty Memo) ,93 and Inventory of U.S. Greenhouse Gas Emissions
and Sinks 1990-2016: Additional Revisions Under Consideration {Draft 2018 Other Updates Memo),94. The
memos discussed changes under consideration, and requested stakeholder feedback on those changes.
The EPA thoroughly evaluated relevant information available, and made several updates to the Inventory, including
to define an exploration segment separate from production (not a methodological change, but a change in
presentation of information), calculate activity and emission factors for well testing and non-hydraulically fractured
completions from GHGRP data, recalculate production segment major equipment activity factors using updated
GHGRP data, and calculate new CO2 emission factors for several sources throughout all segments directly from
GHGRP data.
89 See .
90 See .
91 See .
92 See
93 See < https://www.epa.gov/sites/production/files/2017-
10/documents/revision_under_consideration_for_ghgi_ng_and_petro_uncertainty_2017-10-25_to_post.pdf>
94
3-84 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 The combined impact of revisions to 2015 natural gas sector CH4 emissions, compared to the previous Inventory, is
2 an increase from 162.4 to 164.4 MMT CO2 Eq. (2.0 MMT CO2 Eq., or 1.2 percent). The recalculations resulted in
3 an average increase in CH4 emission estimates across the 1990 through 2015 time series, compared to the previous
4 Inventory, of 0.1 MMT CO2 Eq, or 0.1 percent.
5 The combined impact of revisions to 2015 natural gas sector CO2 emissions, compared to the previous Inventory, is
6 a decrease from 42.4 to 26.3 MMT CO2 (16.0 MMT CO2, or 38 percent). The recalculations resulted in an average
7 decrease in emission estimates across the 1990 through 2015 time series, compared to the previous Inventory, of
8 10.3 MMT CO2 Eq, or 29 percent. The decreased estimate results primarily from recalculations related to the
9 reallocation of CO2 from flaring to petroleum systems from natural gas systems. Previously, data were not available
10 to disaggregate flared emissions between natural gas and petroleum.
11 Exploration
12 The natural gas system segments were reorganized for the current Inventory and now include a specific exploration
13 segment to improve conformance with the IPCC guidelines. Exploration activities were previously included under
14 the production segment. The activities included under exploration are hydraulically fractured (HF) gas well
15 completions, gas well completions without HF, well drilling, and well testing. EPA developed a new methodology
16 to estimate emissions from well testing (not during completions) using GHGRP data, revised the methodology for
17 non-HF gas well completions to use GHGRP data, and updated the HF gas well completions methodology for CO2
18 emissions. These recalculations are discussed below.
19 Well Testing
20 EPA developed a new estimate for gas well testing (during non-completion events) using GHGRP data. In previous
21 Inventories, only well testing conducted as part of a completion event was included. CH4 and CO2 emission factors
22 were developed, on a per-event basis, for vented and flared gas well testing events using RY2015 and RY2016 data.
23 EPA developed activity factors (i.e., number of events per gas well) to determine the number of well testing events
24 in a year, also using RY2015 and RY2016 data. GHGRP RY2015 activity and emission factors are applied to all
25 prior years of the time series. Methane emissions from well testing averaged 1.5 kt (or less than 0.05 MMT CO2 Eq.)
26 over the time series. There was a large decrease in methane emissions from gas well testing from 2015 to 2016 as
27 observed in reported GHGRP data. Carbon dioxide emission from well testing averaged 3.1 kt (or less than 0.05
28 MMT CO2) over the time series. See the Draft 2018 Other Updates Memo for additional discussion.
29 Table 3-57: Gas Well Testing National ChU Emissions (Metric Tons ChU)
Source
1'WO
2005
2012
2013
2014
2015
2016
Non-Completion Well Testing -
Vented
949
1,673
2,080
2,054
2,071
2,043
614
Non-Completion Well Testing -
Flared
13
23
29
29
29
29
2
30 Table 3-58: Gas Well Testing National CO2 Emissions (Metric Tons CO2)
Source
1'WO
2005
2012
2013
2014
2015
2016
Non-Completion Well Testing -
Vented
Non-Completion Well Testing -
Flared
30
1.914
53
3,375
66
4,198
65
4,144
65
4,179
64
4,123
39
323
31 Non-HF Gas Well Completions
32 EPA developed new emission factors for controlled and uncontrolled non-HF gas well completions using GHGRP
33 data, and applied the new factors over all time series years. The emission factor for non-HF gas well completions in
34 the Inventory was previously derived from the GRI 1996 study which defines the factor as covering both gas well
Energy 3-85
-------
1 completions and well flow testing, and based on the assumption that all gas is flared. CH4 and CO2 emission factors
2 were developed, on a per-event basis, for vented and flared gas well non-HF completion events using RY2015 and
3 RY2016 GHGRP data. EPA did not revise the overall counts of non-HF gas well completions. For the split between
4 vented and flared events, EPA used GHGRP data for year 2011 forward, and 2011 data (which show 3 percent of
5 events flared) as a proxy for all earlier years. Methane emissions from non-HF completions averaged 8.6 kt CH4 (or
6 0.2 MMT CO2 Eq.) over the time series. The previous estimate was an average of 0.01 kt CH4 over the time series.
7 Carbon dioxide emission from non-HF completions averaged 7.3 kt (or less than 0.05 MMT CO2) over the time
8 series. The previous estimate was an average of 0.001 kt CH4 over the time series. See the Draft 2018 Other
9 Updates Memo for additional discussion.
10 Table 3-59: Non-HF Gas Well Completions National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2012
2013
2014
2015
2016
Non-HF Completions - Vented
5,713
10,074
11,009
5,890
1,404
13,680
8,065
Non-HF Completions - Flared
20
35
2
39
12
36
89
11 Table 3-60: Non-HF Gas Well Completions National CO2 Emissions (Metric Tons CO2)
Source 1990 2005 2012 2013 2014 2015 2016
Non-HF Completions - Vented
216
381
101
182
72
172
829
Non-HF Completions - Flared
4,643
8,187
565
6,695
2,683
5,909
16,407
12 CO2 Updates
13 EPA developed new CO2 emission factors for the four control categories of HF gas well completions using the same
14 GHGRP data sets and methodology as established for CH4. EPA did not change the activity data methodology for
15 this source, other than to break out HF completions and workovers as separate line items (where completions are
16 included in Exploration and workovers remain within the Production segment). As noted in Planned Improvements,
17 EPA is considering year-specific GHGRP-based emission factor for this source. See the Draft 2018 CO2 Memo for
18 additional discussion.
19 Table 3-61: HF Gas Well Completions National CO2 Emissions (kt CO2)
Source
1990
2005
2012
2013
2014
2015
2016
HF Completions - Non-REC with
Venting
HF Completions - Non-REC with
Flaring
HF Completions - REC with
Venting
HF Completions - REC with
Flaring
Total Emissions
10
311
26
1,062
8
759
6
587
6
613
+
273
+
114
0
0
2
304
2
794
3
911
2
1,246
3
814
2
437
321
1,394
1,563
1,506
1,867
1,091
553
Previous Estimated Emissions
74
305
99
75
66
66
NA
NA (Not Applicable)
20 + Does not exceed 0.5 kt CO2.
21 Production
22 In addition to the memos discussed above, this section references the memorandum, Inventory of U.S. Greenhouse
23 Gas Emissions and Sinks 1990-2015: Revisions for Natural Gas and Petroleum Systems Production Emissions
24 (2017 Production Memo).95
95 See .
3-86 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Non-HF Gas Well Workovers
2 EPA developed new emission factors for controlled and uncontrolled non-HF gas well workovers using GHGRP
3 data, and applied the new factors over all time series years. The emission factor for non-HF gas well workovers in
4 the Inventory was previously derived from the GRI 1996 study. Methane and CO2 emission factors were developed,
5 on a per-event basis, for vented and flared gas well non-HF workover events using RY2015 andRY2016 data. EPA
6 did not revise the overall counts of non-HF gas well workovers. For the split between vented and flared events, EPA
7 used GHGRP data for year 2011 forward, and interpolated to 100 percent vented and 0 percent flared in year 1992
8 (GRI basis). Methane emissions from non-HF workovers averaged 0.6 kt CH4 (or 0.02 MMT CO2 Eq.) over the time
9 series. The previous estimate was an average of 0.4 kt CH4 over the time series. Carbon dioxide emission from non-
10 HF workovers averaged 2.2 kt (or less than 0.05 MMT CO2) over the time series. The previous estimate was an
11 average of 0.03 kt CH4 over the time series. See the Draft 2018 Other Updates Memo for additional discussion.
12 Table 3-62: Non-HF Gas Well Workovers National ChU Emissions (Metric Tons ChU)
Source
1990
2005
2012
2013
2014
2015
2016
Non-HF Workovers - Vented
Non-HF Workovers - Flared
509
0
631
19
1,486
0
429
6
441
2
525
26
517
1
13
14 Table 3-63: Non-HF Gas Well Workovers National CO2 Emissions (kt CO2)
~~Source 1990 2005 2012 2013 2014 2015 2016
Non-HF Workovers - Vented 30 38 92 24 28 45 25
Non-HF Workovers - Flared 0 3,164 97 942 548 3,192 5,836
15 Activity data updates
16 Well Counts
17 EPA has used a more recent version of the Drillinglnfo data set to update well counts data in the Inventory. There
18 are not methodological changes to this source in the 2018 Inventory or major changes to the activity data, but
19 because this is a key input, results are highlighted here.
20 Table 3-64: Producing Gas Well Count Data
Gas Well Count
1990
2005
2012
2013
2014
2015
2016
Number of Gas Wells
Previous Estimate
197,626
202,628
348,470
355,234
433,390
438,672
427,828
431,926
431,446
433,941
425,651
421,893
416,881
NA
NA (Not Applicable)
21 In December 2017, EIA released a 2000 through 2016 time series of national oil and gas well counts. EIA total (oil
22 and gas) well counts for 2016 were 1,010,441. EPA's total well counts were 978,845. Over the 2000 through 2016
23 time series, EPA's well counts were on average 2 percent lower than EIA's. EIA's well counts include side tracks,
24 completions, and recompletions, and therefore are expected to be higher than EPA's which include only producing
25 wells. EPA and EIA use different thresholds for distinguishing between oil and gas (EIA uses 6 mcf/bbl, while EPA
26 uses 100 mcf/bbl), which results in EIA having a lower fraction of oil wells and a higher fraction of gas wells than
27 EPA. Across the 2000 through 2016 EIA time series, EIA estimates (which include multiple well categories, as
28 noted above) on average 128,335 (or 31 percent) more gas wells in each year than EPA's gas well counts (which
29 include only producing wells).
30 Equipment Counts
31 EPA recalculated activity factors of equipment per well using the latest GHGRP RY2015 data. This resulted in
32 changes across the time series. For example, the number of heaters per well decreased by 20 percent over the time
33 series, the number of chemical injection pumps per well decreased by 4 percent, and the number of dehydrators per
34 well increased by 5 percent. The impact of the changes in equipment counts per well along with changes in well
Energy 3-87
-------
1 counts resulted in changes in methane emissions across the time series for heaters (-21 percent), chemical injection
2 pumps (-6 percent), and dehydrators (+3 percent).
3 CO2 Updates
4 EPA updated CO2 emissions for a number of sources in the production segment. See the Draft 2018 C02 Memo for
5 more details. The overall impact was an average decrease of 9.6 MMT CO2 (or 81 percent) over the time series,
6 which is partially due to the reallocation of CO2 emissions from associated gas and miscellaneous onshore
7 production flaring from Natural Gas Systems to Petroleum Systems, which was not possible in the past because the
8 previous data source aggregated venting and flaring activity data from both petroleum and natural gas systems, but
9 is now possible because through use of the GHGRP data.
10 Sources with the largest impacts include flaring (decrease of 9.9 MMT CO2 on average over the time series), and
11 tanks (increase of 0.5 MMT CO2 over the time series). These sources are discussed in detail below. Other sources
12 recalculated had increases or decreases of less than 0.5 MMT CO2.
13 Miscellaneous Production Flaring
14 The EPA developed new estimates for CO2 and CH4 emissions from miscellaneous production flaring using
15 GHGRP subpart W data. Along with other updates to flaring emissions in both oil and gas production, this replaces
16 the estimate for onshore flaring that was previously reported in the natural gas systems CO2 emissions totals. EPA
17 developed emission factors from 2015 and 2016 GHGRP data; the 2015 emission factor is applied to all prior years.
18 The emission factors are on a per-well basis and were applied to all gas wells in each year. Details are provided in
19 the Draft 2018 C02 Memo. Initial stakeholder feedback on this update suggested use of production-based emission
20 factors as opposed to well-based emission factors.
21 Table 3-65: Miscellaneous Production Flaring National CO2 Emissions (kt CO2)
Source
1WO
2005
2012
2013
2014
2015
2016
Miscellaneous Production
Flaring
Previous Estimated emissions
from flaring (natural gas and
petroleum)"
0
1,009
1,834
1,810
1,826
1,801
1,445
9,093
7,193
12,704
15,684
17,629
17,629
NA
a The previous estimated emissions from ilaring included emissions from multiple sources in the production and
processing segments, and also included petroleum systems flaring emissions.
NA (Not Applicable)
22 Tanks
23 EPA developed CO2 emissions estimates for condensate tanks using GHGRP data and a throughput-based approach.
24 This approach is identical to the methodology to calculate CH4 emissions; for more information, please see the 2017
25 Production Memo. The overall impact of the change is an increase in calculated CO2 emissions by 0.5 MMT CO2
26 over the time series.
27 Table 3-66: National Condensate Tank Emissions by Category and National Emissions (kt
28 COz)
CO2 Emissions
1990
2005
2012
2013
2014
2015
2016
Large Tanks w/ Flares
28"
363
819
985
1,030
1,044
1,398
Large Tanks w/ VRU
0
1
2
3
3
3
3
Large Tanks w/o Control
1
+
+
1
1
1
1
Small Tanks w/ Flares
0
9
27
33
35
35
42
Small Tanks w/o Flares
6
4
8
9
10
10
15
Malfunctioning Dump Valves
Total Emissions
+
294
+ /
378
+
857
+
1,030
+
1,078
+
1,093
+
1,460
Previous Estimated Emissions
296
383
870
1,045
1,093
1,108
NA
NA (Not Applicable)
+ Does not exceed 0.5 kt CO2.
3-88 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Processing
2 There were no updates to the CH4 emissions estimation methodology for the processing segment. Updates to activity
3 data resulted in a minor decrease (less than 0.1 MMT CO2 Eq., or 0.5 percent) in CH4 emissions estimates for this
4 segment across the time series. EPA updated CO2 emissions for a number of sources in the processing segment to
5 use emission factors directly calculated from subpart W data. See the Draft 2018 C02 Memo for more details. The
6 overall impact was an average decrease of 1.7 MMT CO2 (or 8 percent) over the time series, which is primarily due
7 to the incorporation of GHGRP data for acid gas removal vents. Acid gas removal CO2 emissions decreased by an
8 average of 4.7 MMT CO2, or 21 percent over the time series. Incorporation of GHGRP data for flaring in
9 processing increased emissions by 3.0 MMT CO2.
10 Table 3-67: Processing CO2 Updates, National Emissions (kt CO2)
Source
1990
2005
2012
2013
2014
2015
2016
Acid Gas Removal
28,282
15,320
13,579
14,565
14,946
14,946
16,565
Previous Acid Gas Removal
27,708
21,694
21,404
21,690
23,643
23,643
NA
Processing-flaring
0
3,516
5,502
5,902
6,056
6,056
5,404
Previous processing flaring a
NA
NA
NA
NA
NA
NA
NA
NA (Not Applicable)
a The previous estimated emissions from flaring included emissions from multiple sources in the natural gas and
petroleum production segments, and natural gas processing segment. The previous estimate was presented as a
single emission source in the natural gas systems production segment.
11 Transmission and Storage
12 Changes in the estimates for CH4 from transmission and storage include the addition of flaring emissions and
13 recalculations due to updated data (e.g., GHGRP station counts, the GHGRP split between dry and wet seal
14 centrifugal compressors, and GHGRP pneumatic controller data). Stakeholder feedback (one stakeholder) expressed
15 support for this approach. These changes resulted in an average increase in calculated emissions over the time series
16 from this segment of less than 0.1 MMT CO2 Eq., or 0.1 percent.
17 Additional information on inclusion of the Aliso Canyon emissions can be found in the Methodology section above
18 and in the 2017 Transmission and Storage Memo96 and not in the Recalculation Discussion section as it did not
19 involve recalculation of a previous year of the Inventory.
20 Table 3-68: Transmission and Storage ChU Updates to Flaring, National Emissions (MT ChU)
Source
1990
2005
2012
2013
2014
2015
2016
Transmission-flaring *
307
276
281
303
326
326
395
Storage-flaring*
235
223
231
232
232
227
198
Previous flaring (transmission
NA
NA ••
NA
NA
NA
NA
NA
and storage)
NA (Not Applicable)
*Estimates are developed from GHGRP data, wherein compressor stations that service underground storage
fields might be classified as transmission compression as the primary function. A significant fraction of the
transmission station flaring emissions presented in this table likely occurs at stations that service storage
facilities; such stations typically require flares, compared to a typical transmission compressor station used solely
for mainline compression that does not require liquids separation, dehydration, and flaring.
21 EPA updated CO2 emissions for pneumatic controllers and flares in the transmission and storage segment. See the
22 Draft 2018 CO2 Memo for more details. The overall impact was an average increase of 0.1 MMT CO2 (or by a
23 factor of 3) over the time series. The updated CO2 data for pneumatic controllers increased estimated emissions
24 from pneumatic controllers by less than 0.1 MMT CO2, or 53 percent over the time series. The addition of an
25 estimate for flares increased CO2 emissions from transmission and storage by 2.5 MMT CO2 over the time series.
96 See .
Energy 3-89
-------
1
Table 3-69: Transmission and Storage CO2 Updates, National Emissions (kt CO2)
Source
mo
2005
2012
2013
2014
2015
2016
Transmission-pneumatic
controllers
6.3
4.0
2.9
3.0
0.9
0.8
0.7
Previous transmission pneumatic
controllers
6.1
2.1
0.6
0.8
0.8
0.8
NA
Storage-pneumatic controllers
Previous Storage pneumatic
controllers
1.3
1.3
1.2
1.0
O
00 0
1.0
0.9
0.9
0.9
0.6
0.7
1.0
NA
Transmission-flaring *
78.8
71.0
72.2
78.0
83.7
83.9
88.4
Storage-flaring*
Previous flaring (transmission
and storage)
24.5
a :i
23.2
V,l
24.0
NA
24.1
NA
24.1
NA
23.6
NA
15.3
NA
NA (Not Applicable)
* Estimates are developed from GHGRP data, wherein compressor stations that service underground storage
fields might be classified as transmission compression as the primary function. A significant fraction of the
transmission station flaring emissions presented in this table likely occurs at stations that service storage
facilities; such stations typically require flares, compared to a typical transmission compressor station used solely
for mainline compression that does not require liquids separation, dehydration, and flaring.
2 Distribution
3 Although there were no methodological updates to the distribution segment, recalculations due to updated data (e.g.,
4 GHGRP M&R station counts) resulted in an average increase in calculated emissions over the time series from this
5 segment of less than 0.01 MMT CO2 Eq. CH4 (or less than 0.1 percent) and less than 0.01 MMT CO2 (or 1.9
6 percent).
7 Planned Improvements
8 Plans for 2018 Inventory (1990 through 2016) and Future Inventories
9 EPA seeks stakeholder feedback on the improvements noted below for the final version of the current Inventory and
10 future Inventories.
11 Uncertainty
12 The uncertainty analysis results presented for this public review Inventory were based on the top methane-emitting
13 sources for 2015 from the previous Inventory. EPA will re-evaluate the highest emitting sources for 2016, based on
14 the final version of the current Inventory, and update the uncertainty analysis to reflect these sources and their
15 methodology, as necessary. EPA will also consider further stakeholder feedback on the Draft 2018 Uncertainty
16 Memo.
17 Miscellaneous Production Flaring
18 Miscellaneous production flaring emission factors are currently applied on a well-basis at the national-level. EPA is
19 considering two additional options for the final version of the current Inventory, based on stakeholder feedback: a
20 production-based approach and developing factors at a basin-level. Each of these options is being considered to
21 avoid over- or underestimating emissions from this source.
22 Gas STAR Reductions
23 As detailed in the Draft 2018 Other Updates Memo, EPA is continuing to evaluate sources that currently use
24 voluntary reduction data in calculating emissions to identify instances where an emission source's calculation
25 methodology could be updated to calculate net emissions or instances where the current methodology could be
26 simplified to acknowledge sources that likely no longer necessitate consideration of Gas STAR reductions. EPA has
27 assessed Gas STAR reduction data and is considering several updates for the final version of the current Inventory.
3-90 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
In the Production Segment, a spreadsheet error in the public review draft resulted in a miscalculation of the scaling
factor for the "other reductions" value. Correcting this error results in a decrease in the Gas STAR reductions for
this segment of around 1 MMT CO2 Eq. (from 9 MMT CO2 Eq. to 8 MMT CO2 Eq.). In the transmission and
storage segment, Gas STAR reductions associated with transmission station fugitives were not removed from the
Gas STAR data when the emissions were recalculated with a "net approach." Correcting the estimates to remove
those reductions results in a decrease in the Gas STAR reductions of 0.1 MMT CO2 Eq. (from 8.3 MMT CO2 Eq. to
8.2 MMT CO2 Eq.). In the distribution segment, Gas STAR reductions associated with mishaps (dig-ins) and
pipeline blowdowns might be removed, which would result in a small decrease in the Gas STAR reductions of less
than 0.1 MMT CO2 Eq. (from 1.0 MMT CO2 Eq. to 0.99 MMT CO2 Eq.). See Draft 2018 Other Updates Memo for
more information.
EPA continues to review unassigned Gas STAR reductions in the transmission and storage segment and distribution
segment (currently grouped and identified as "other" reductions).
Liquids Unloading
EPA is considering several updates to liquids unloading for the final version of the current Inventory. Based on
stakeholder feedback, EPA is considering developing region-specific liquids unloading emissions and activity
factors, rather than national-level. Preliminary analysis of emissions calculated using a basin-level approach
indicates that the emission estimates will be lower than those calculated with the national-level approach.
Additionally, the emissions and activity data methodology used in the current Inventory rely exclusively on recently
collected data (from 2011 or later). The EPA is evaluating the liquids unloading data collected for the 1996
GRI/EPA study to determine if it better represents early time series years.
Year-Specific Emission Factors
EPA is considering the development of year-specific emission factors, using GHGRP data, for a number of sources
with annual emissions currently calculated with data from one year or an average of a several years. For example,
for hydraulically fractured gas well completions and workovers, while changes in practices over time are currently
reflected in the Inventory due to annual practice-specific activity data, changes in emissions within each practice-
specific category are not currently reflected. Preliminary analysis of emissions calculated using a year-specific
emission factor approach for hydraulically fractured gas well completions and workovers indicates that the emission
estimates will be lower than those calculated with the average emission factor approach. EPA is considering
updating emission factors for this and potentially other sources for the final version of the current Inventory.
Well-Related Activity Data
As described in the Recalculations Discussion, EPA has updated the emission factors for several well-related
emission sources, including testing, completions, and workovers. EPA will continue to assess available data that and
stakeholder feedback on considerations to improve activity estimates for these types of sources. For example, the
current Inventory assumes that 1 percent of HF gas wells and 4.35 percent of non-HF gas wells undergo workovers
each year based on historical assumptions; EPA will review available data including from the GHGRP to consider
updating the activity data methodology.
LNG Segment Emissions
The current Inventory estimates emissions from LNG storage stations and LNG import terminals in the transmission
and storage segment of natural gas systems. The emission factors are based on the 1996 GRI/EPA study, which
developed emission factors using underground natural gas storage and transmission compressor station data; specific
emissions data for LNG storage stations and LNG import terminals were not available in the GRI/EPA study. EPA's
GHGRP subpart W collects data from LNG storage and LNG import and export facilities that meet a reporting
threshold of 25,000 metric tons of CO2 equivalent (MT CO2 Eq.) emissions. EPA is considering approaches and
seeking stakeholder feedback on incorporating GHGRP data to improve LNG emissions estimates in the Inventory.
Refer to the Draft 2018 Other Updates Memo for additional details. Incorporating GHGRP data would likely
decrease emissions from this segment.
Energy 3-91
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
N2() Emissions
N20 emissions are currently not included in petroleum systems estimates, but EPA is considering developing a
methodology to estimate N2O emissions. The Draft 2018 Other Updates Memo provides discussion on this topic.
EPA will consider options such as using GHGRP data directly, for sources that already rely on GHGRP data for
CH4 or CO2 estimates. GHGRP RY2015 reported N20 flaring emissions specific to natural gas systems were 26
metric tons (or less than 0.01 MMT CO2 Eq.). In addition, 36 metric tons N2O (or 0.01 MMT CO2 Eq.) flaring
emissions were reported for GHGRP RY2015 for sources that fall within both natural gas and petroleum systems.
Offshore Platforms
EPA is considering updates to the offshore platform emissions calculation methodology, as discussed in the Draft
2018 Other Updates Memo. The current emission factors were based on data from the 2011 DOI/Bureau of Ocean
Energy Management's (BOEM) Gulf Offshore Activity Data System (GOADS), and 2014 GOADS data is
available. A different source for platform counts is also being considered.
Upcoming Data, and Additional Data that Could Inform the Inventory
EPA will continue to review data available from its GHGRP, in particular new data on gathering and boosting
stations, gathering pipelines, and transmission pipeline blowdowns and new well-specific information, available in
2017 (for reporting year 2016) for the first time. EPA will consider revising its methods to take into account the new
GHGRP data.
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. Key studies in progress
include DOE-funded work on the following sources: vintage and new plastic pipelines (distribution segment),
industrial meters (distribution segment), and sources within the gathering and storage segments97; an API field study
on pneumatic controllers; and a Pipeline Research Council International (PRCI) project in which researchers are
gathering and analyzing subpart W data on transmission compressor stations and underground storage facilities.
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, stakeholder comments have highlighted areas where additional data that could inform the Inventory are
currently limited or unavailable:
• Tank malfunction and control efficiency data.
• Consider updating engine emission factors, including using subpart W data to the extent possible, and
considering whether and how to represent differences between rich- and lean-burn engines.
• Activity data and emissions data for production facilities that do not report to GHGRP.
• Natural gas leaks at point of use estimates. A recent study (Lavoie et al. 2017) measured three natural gas
power plants and found them to be large sources of natural gas leak emissions, and the stakeholder
suggested that EPA evaluate the study and any additional information available on this source. At least one
country, the United Kingdom, includes an emission estimate for residential and commercial customer
natural gas use leaks (e.g., domestic heating boiler cycling and pre-ignition losses from domestic and
commercial gas appliances) in its national greenhouse gas emissions inventory; the EPA seeks available
data to estimate emissions from this source in the U.S. Stakeholder feedback (one stakeholder) supports
use of data from Lavoie et al. or use of the U.K. approach to calculate emissions from this source.
One stakeholder suggested that the Inventory should be updated with site-level and basin-level data, noting the EPA
could first use basin-level data to assess the Inventory, and that future research could focus on collecting data in
basins with the largest discrepancies.
97 See .
3-92 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
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" encompasses various types of 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 well is around 3 million (with around 2.6 million abandoned oil wells and 0.6
million abandoned gas wells). Wells that are plugged have much lower methane emissions than wells that are
unplugged (less than 1 kg CH4 per well per year, versus over 100 kg CH4 per well per year). Around 30 percent of
the abandoned well population in the U.S. is plugged. This fraction has increased over the time series (from around
19 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 230 kt CH4 in 2016. Emissions increased by 3 percent from 1990,
as the total population of abandoned oil wells increased 25 percent. Emissions decreased by 1 percent between 2015
and 2016 as a result of well plugging activities.
Abandoned gas wells. Abandoned gas wells emitted 54 kt CH4 in 2016. Emissions increased by 51 percent from
1990, as the total population of abandoned gas wells increased 73 percent. Emissions decreased by 1 percent
between 2015 and 2016 as a result of well plugging activities.
Table 3-70: ChU Emissions from Abandoned Oil and Gas Wells (MMT CO2 Eq.)
Activity 1990
2005
2012
2013
2014
2015
2016
Abandoned Oil Wells 5.6
5.8
5.8
5.8
5.8
5.8
5.8
Abandoned Gas Wells 0.9
I.I
1.2
1.2
1.3
1.4
1.4
Total 6.5
7.0
7.0
7.1
7.2
7.1
Note: Totals may not sum due to independent rounding.
able 3-71: ChU Emissions from Abandoned Oil and Gas Wells (kt)
Activity 1990
2005
2012
2013
2014
2015
2016
Abandoned Oil Wells 224
] 233
231
230
230
232
230
Abandoned Gas Wells 36
! 42
48
50
52
55
54
Total 260
275
279
280
282
286
284
Note: Totals may not sum due to independent rounding.
Methodology
EPA developed abandoned well emission factors using data from Kang et al. (2016) and Townsend-Small et al.
(2016). Plugged and unplugged abandoned well emission factors were developed at the national-level (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.
Energy 3-93
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
The total population of abandoned wells over the time series was estimated using historical data and Drillinglnfo
data. The abandoned well population was then split into plugged and unplugged wells by assuming that all
abandoned wells were unplugged in 1950, 31 percent of abandoned wells were plugged in 2016 (based on an
analysis of Drillinglnfo data), and applying linear interpolation for intermediate years. See the Draft 2018
Abandoned Wells Memo for details.98
Abandoned Oil Wells
Table 3-72: Abandoned Oil Wells Activity Data and Methane Emissions (Metric Tons ChU)
Source
Plugged abandoned oil
wells
Unplugged abandoned oil
wells
Total Abandoned Oil Wells
Abandoned oil wells in
Appalachia
Abandoned oil wells
outside of Appalachia
Methane from plugged
abandoned oil wells (MT)
Methane from unplugged
abandoned oil wells (MT)
Total Methane from
Abandoned oil wells (MT)
1990
2005
2012
2013
2014
2015
2016
382,446
610,884
719,901
736,830
754,118
776,450
788,396
1,666,399
1,769,214
1,768,266
1,769,425
1,770,862
1,783,308
1,771,362
2,048,846
2,380,098
2,488,167
2,506,255
2,524,980
2,559,758
2,559,758
26%
24%
24%
23%
23%
23%
23%
74%
76%
76%
77%
77%
77%
77%
314
471
539
549
560
574
582
223,780
232,546
230,070
229,885
229,735
231,011
229,464
224,094
233,017
230,609
230,434
230,295
231,585
230,046
Abandoned Gas Wells
Table 3-73: Abandoned Gas Wells Activity Data and Methane Emissions (Metric Tons ChU)
Source
1990
2005
2012
2013
2014
2015
2016
Plugged abandoned gas wells
59,480
103,379
138,537
146,187
156,144
167,011
169,580
Unplugged abandoned gas
wells
259,166
299,402
340,284
351,053
366,666
383,580
381,011
Total Abandoned Gas Wells
318,645
402,781
478,821
497,239
522,810
550,591
550,591
Abandoned gas wells in
Appalachia
28%
29%
30%
30%
30%
30%
30%
Abandoned gas wells outside
S ' /
of Appalachia
72%
71%
70%
70%
70%
70%
70%
Methane from plugged
abandoned gas wells (MT)
53
96
131
139
149
159
162
Methane from unplugged
abandoned gas wells (MT)
35,810
42,064
48,176
49,754
52,024
54,483
54,118
Total Methane from
Abandoned gas wells (MT)
35,863
42,160
18,307
49,893
52,173
54,643
54,280
Uncertainty and Time-Series Consistency
An uncertainty analysis for abandoned well emissions was not performed. To develop the values in Table 3-74
below, EPA has applied the uncertainty bounds calculated for the 2015 emission estimates presented in the previous
(i.e., 1990 through 2015) Inventory for Petroleum Systems and Natural Gas Systems. EPA is still seeking comment
98 https://www.epa.gOv/sites/production/files/2017-10/documents/2018_ghgi_draft_revision_-_abandoned_wells_2017-10-
25_to_post.pdf
3-94 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 on the approach to calculate uncertainty and may update its approach in the final version of the current Inventory,
2 such as by incorporating uncertainty information from Townsend-Small et al. and Kang et al. For more information,
3 please see the Planned Improvements sections for Petroleum Systems and Natural Gas Systems, and the 2018
4 Uncertainty Memo."
5 Table 3-74: Approach 2 Quantitative Uncertainty Estimates for ChU Emissions from
6 Petroleum Systems (MMT CO2 Eq. and Percent)
Source
Gas
2016 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT CO2 Eq.)b
(MMT CO2 Eq.)
(%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Abandoned Oil Wells
CH4
5.8
3.9 7.9
-32% +36%
Abandoned Gas Wells
ch4
1.4
1.2 1.6
-15% +17%
a Range of emission estimates estimated by applying the 95 percent confidence intervals obtained from the Monte Carlo
Simulation analysis conducted for the year 2015 for natural gas and petroleum systems.
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.
7 To calculate a time series of emissions for abandoned wells, EPA developed annual activity data for 1990 through
8 2016 by summing an estimate of total abandoned wells not included in recent databases, to an annual estimate of
9 abandoned wells in the Drillinglnfo data set. As discussed above, the abandoned well population was split into
10 plugged and unplugged wells by assuming that all abandoned wells were unplugged in 1950, 31 percent of
11 abandoned wells were plugged in 2016 (based on an analysis of Drillinglnfo data), and applying linear interpolation
12 for intermediate years. The same emission factors were applied to the corresponding categories for each year of the
13 time series.
14 QA/QC and Verific cussion
15 The emission estimates in the Inventory are continually being reviewed and assessed to determine whether emission
16 factors and activity factors accurately reflect current industry practices. A QA/QC analysis was performed for data
17 gathering and input, documentation, and calculation. QA/QC checks are consistently conducted to minimize human
18 error in the model calculations.
19 As in previous years, EPA conducted early engagement and communication with stakeholders on updates prior to
20 public review. EPA held stakeholder workshops on greenhouse gas data for oil and gas in June and October of
21 2017, and held webinars in April and August of 2017. In advance of each workshop, EPA released memos detailing
22 updates under consideration and requesting stakeholder feedback. Stakeholder feedback received through these
23 processes is discussed in the Planned Improvements sections below.
24 Planned Improvements
25 Through EPA's stakeholder process on oil and gas in the Inventory, EPA received initial stakeholder feedback on
26 the abandoned wells update to the Inventory. Stakeholders noted varying definitions regarding abandoned well
27 populations and subpopulations and plugging status, and noted varying degrees of plugging, due to state-level
28 programs to plug abandoned wells. A stakeholder noted limited coverage of abandoned wells studies in the U.S., and
29 cautioned that it may be premature to develop national level estimates for this source, while another stakeholder
30 supported the inclusion of this emission sources and noted that the update uses the best available data for this source.
31 EPA will also continue to assess new data and stakeholder feedback on considerations (such as the disaggregation of
32 the well population into Appalachia and other regions) to improve the abandoned well count estimates and emission
99 https://www.epa.gov/sites/production/files/2017-
10/documents/revision_under_consideration_for_ghgi_ng_and_petro_uncertainty_2017-10-25_to_post.pdf
Energy 3-95
-------
1 factors. In addition, the studies used to develop CH4 emission factors did not provide CO2 data. EPA will consider
2 developing a methodology to estimate CO2 emissions from abandoned wells, such as applying a ratio of CO2 to CH4
3 content in gas.
4 EPA will assess updates to the Uncertainty Analysis, such as incorporating data from Townsend-Small et al. and
5 Kang et al. to improve the uncertainty estimates.
6 3.9 Energy Sources of Indirect Greenhouse Gas
7 Emissions
8 In addition to the main greenhouse gases addressed above, many energy-related activities generate emissions of
9 indirect greenhouse gases. Total emissions of nitrogen oxides (NOx), carbon monoxide (CO), and non-CH4 volatile
10 organic compounds (NMVOCs) from energy-related activities from 1990 to 2016 are reported in Table 3-75.
11 Table 3-75: NOx, CO, and NMVOC Emissions from Energy-Related Activities (kt)
Gas/Activity
1990
2005
2012
2013
2014
2015
2016
NOx
21,106
16.602
11,271
10,747
10,161
9,323
8,352
Mobile Fossil Fuel Combustion
10,862
10.295
6,871
6,448
6,024
5,417
4,814
Stationary Fossil Fuel Combustion
10,023
5.858
3,655
3,504
3,291
3,061
2,692
Oil and Gas Activities
139
321
663
704
745
745
745
Waste Combustion
82
128
82
91
100
100
100
International Bunker Fuels"
1,956
1,70-1
1,398
1,139
1,139
1,226
1,313
CO
125,640
64,'M5
42,164
40,239
38,315
36,348
34,401
Mobile Fossil Fuel Combustion
119,360
58.615
36,153
34,000
31,848
29,881
27,934
Stationary Fossil Fuel Combustion
5,000
4.648
4,027
3,884
3,741
3,741
3,741
Waste Combustion
978
1.403
1,318
1,632
1,947
1,947
1,947
Oil and Gas Activities
302
318
666
723
780
780
780
International Bunker Fuels"
103
133
133
129
135
141
143
NMVOCs
12,620
7,m
7,558
7,357
7,154
6,867
6,581
Mobile Fossil Fuel Combustion
10,932
5.724
4,243
3,924
3,605
3,318
3,032
Oil and Gas Activities
554
510
2,651
2,786
2,921
2,921
2,921
Stationary Fossil Fuel Combustion
912
716
569
539
507
507
507
Waste Combustion
222
241
94
108
121
121
121
International Bunker Fuels"
57
5-1
46
41
42
47
49
a These values are presented for informational purposes only and are not included in totals.
Note: Totals may not sum due to independent rounding.
12 Methodology
13 Emission estimates for 1990 through 2016 were obtained from data published on the National Emission Inventory
14 (NEI) Air Pollutant Emission Trends web site (EPA 2016), and disaggregated based on EPA (2003). Emission
15 estimates for 2012 and 2013 for non-electric generating units (EGU) were updated to the most recent available data
16 in EPA (2016). Emission estimates for 2012 and 2013 for non-mobile sources are recalculated emissions by
17 interpolation from 2016 in EPA (2016). Emissions were calculated either for individual categories or for many
18 categories combined, using basic activity data (e.g., the amount of raw material processed) as an indicator of
19 emissions. National activity data were collected for individual applications from various agencies.
20 Activity data were used in conjunction with emission factors, which together relate the quantity of emissions to the
21 activity. Emission factors are generally available from the EPA's Compilation of Air Pollutant Emission Factors,
22 AP-42 (EPA 1997). The EPA currently derives the overall emission control efficiency of a source category from a
23 variety of information sources, including published reports, the 1985 National Acid Precipitation and Assessment
24 Program emissions inventory, and other EPA databases.
3-96 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Uncertainty and Time-Series Consistency
2 Uncertainties in these estimates are partly due to the accuracy of the emission factors used and accurate estimates of
3 activity data. A quantitative uncertainty analysis was not performed.
4 Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
5 through 2016. Details on the emission trends through time are described in more detail in the Methodology section,
6 above.
7 3.10 International Bunker Fuels (CRF Source
8 Category 1: Memo Items)
9 Emissions resulting from the combustion of fuels used for international transport activities, termed international
10 bunker fuels under the UNFCCC, are not included in national emission totals, but are reported separately based upon
11 location of fuel sales. The decision to report emissions from international bunker fuels separately, instead of
12 allocating them to a particular country, was made by the Intergovernmental Negotiating Committee in establishing
13 the Framework Convention on Climate Change.100 These decisions are reflected in the IPCC methodological
14 guidance, including IPCC (2006), in which countries are requested to report emissions from ships or aircraft that
15 depart from their ports with fuel purchased within national boundaries and are engaged in international transport
16 separately from national totals (IPCC 2006).101
17 Two transport modes are addressed under the IPCC definition of international bunker fuels: aviation and marine.102
18 Greenhouse gases emitted from the combustion of international bunker fuels, like other fossil fuels, include CO2,
19 CH4 and N20 for marine transport modes, and CO2 and N20 for aviation transport modes. Emissions from ground
20 transport activities—by road vehicles and trains—even when crossing international borders are allocated to the
21 country where the fuel was loaded into the vehicle and, therefore, are not counted as bunker fuel emissions.
22 The 2006 IPCC Guidelines distinguish between different modes of air traffic. Civil aviation comprises aircraft used
23 for the commercial transport of passengers and freight, military aviation comprises aircraft under the control of
24 national armed forces, and general aviation applies to recreational and small corporate aircraft. The 2006 IPCC
25 Guidelines further define international bunker fuel use from civil aviation as the fuel combusted for civil (e.g.,
26 commercial) aviation purposes by aircraft arriving or departing on international flight segments. However, as
27 mentioned above, and in keeping with the 2006 IPCC Guidelines, only the fuel purchased in the United States and
28 used by aircraft taking-off (i.e., departing) from the United States are reported here. The standard fuel used for civil
29 aviation is kerosene-type jet fuel, while the typical fuel used for general aviation is aviation gasoline.103
30 Emissions of CO2 from aircraft are essentially a function of fuel use. Nitrous oxide emissions also depend upon
31 engine characteristics, flight conditions, and flight phase (i.e., take-off, climb, cruise, decent, and landing). Recent
32 data suggest that little or no CH4 is emitted by modern engines (Anderson et al. 2011), and as a result, CH4
33 emissions from this category are considered zero. Injet engines, N20 is primarily produced by the oxidation of
34 atmospheric nitrogen, and the majority of emissions occur during the cruise phase. International marine bunkers
35 comprise emissions from fuels burned by ocean-going ships of all flags that are engaged in international transport.
1°° 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).
101 Note that the definition of international bunker fuels used by the UNFCCC differs from that used by the International Civil
Aviation Organization.
102 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).
103 Naphtha-type jet fuel was used in the past by the military in turbojet and turboprop aircraft engines.
Energy 3-97
-------
1 Ocean-going ships are generally classified as cargo and passenger carrying, military (i.e., U.S. Navy), fishing, and
2 miscellaneous support ships (e.g., tugboats). For the purpose of estimating greenhouse gas emissions, international
3 bunker fuels are solely related to cargo and passenger carrying vessels, which is the largest of the four categories,
4 and military vessels. Two main types of fuels are used on sea-going vessels: distillate diesel fuel and residual fuel
5 oil. Carbon dioxide is the primary greenhouse gas emitted from marine shipping.
6 Overall, aggregate greenhouse gas emissions in 2016 from the combustion of international bunker fuels from both
7 aviation and marine activities were 115.5 MMT CO2 Eq., or 10.5 percent above emissions in 1990 (see Table 3-76
8 and Table 3-77). Emissions from international flights and international shipping voyages departing from the United
9 States have increased by 88.9 percent and decreased by 35.1 percent, respectively, since 1990. The majority of these
10 emissions were in the form of CO2; however, small amounts of CH4 (from marine transport modes) and N20 were
11 also emitted.
12 Table 3-76: CO2, ChU, and N2O Emissions from International Bunker Fuels (MMT CO2 Eq.)
Gas/Mode
1990
2005
2012
2013
2014
2015
2016
CO2
103.5
113.1
105.8
99.8
103.4
110.9
114.4
Aviation
38.0
60.1
64.5
65.7
69.6
71.9
71.9
Commercial
30.0
55.6
61.4
62.8
66.3
68.6
68.6
Military
8.1
4.5
3.1
2.9
3.3
3.3
3.3
Marine
65.4
53.0
41.3
34.1
33.8
38.9
42.5
CH4
0.2
0.1
0.1
0.1
0.1
0.1
0.1
Aviation8
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Marine
0.2
0.1
0.1
0.1
0.1
0.1
0.1
N2O
0.9
1.0
0.9
0.9
0.9
0.9
1.0
Aviation
0.4
0.6
0.6
0.6
0.7
0.7
0.7
Marine
0.5
0.4
0.3
0.2
0.2
0.3
0.3
Total
104.5
114.2
106.8
100.7
104.4
111.9
115.5
+ Does not exceed 0.05 MMT CO2 Eq.
a CH4 emissions from aviation are estimated to be zero.
Notes: Totals may not sum due to independent rounding. Includes aircraft cruise altitude emissions.
13 Table 3-77: CO2, ChU, and N2O Emissions from International Bunker Fuels (kt)
Gas/Mode
1990
2005
2012
2013
2014
2015
2016
CO2
103,463
113,139
105,805
99,763
103,400
110,887
114,394
Aviation
38,034
60,125
64,524
65,664
69,609
71,942
71,859
Marine
65,429
53,014
41,281
34,099
33,791
38,946
42,535
CH4
7
5
4
3
3
3
4
Aviation3
0
0
0
0
0
0
0
Marine
7
5
4
3
3
3
4
N2O
3
3
3
3
3
3
3
Aviation
1
2
2
2
2
2
2
Marine
2
1
1
1
1
1
1
aCH4 emissions from aviation are estimated to be zero.
Notes: Totals may not sum due to independent rounding. Includes aircraft cruise altitude emissions.
14 Methodology
15 Emissions of CO2 were estimated by applying C content and fraction oxidized factors to fuel consumption activity
16 data. This approach is analogous to that described under Section 3.1- CO2 from Fossil Fuel Combustion. Carbon
17 content and fraction oxidized factors for jet fuel, distillate fuel oil, and residual fuel oil were taken directly from EIA
18 and are presented in Annex 2.1, Annex 2.2, and Annex 3.8 of this Inventory. Density conversions were taken from
19 Chevron (2000), ASTM (1989), and USAF (1998). Heat content for distillate fuel oil and residual fuel oil were
20 taken from EIA (2017) and USAF (1998), and heat content for jet fuel was taken from EIA (2017). A complete
21 description of the methodology and a listing of the various factors employed can be found in Annex 2.1. See Annex
3-98 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
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 N20 were calculated by multiplying emission factors by measures of fuel
consumption by fuel type and mode. Emission factors used in the calculations of CH4 and N20 emissions were
obtained from the 2006IPCC 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 N20 (IPCC 2006). For marine vessels
consuming either distillate diesel or residual fuel oil the following values (g/MJ), were employed: 0.32 for CH4 and
0.08 for N:0. 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, 2000 through 2016 as modeled with the Aviation Enviromnental 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 CO: estimates for 1990 and 2000 through 2016 are obtained from FAA's AEDT model (FAA
2017). The radar-informed method that was used to estimate CO: emissions for commercial aircraft for 1990, and
2000 through 2016 is not possible for 1991 through 1999 because the radar data set is not available for years prior to
2000. FAA developed OAG schedule-informed inventories modeled with AEDT and great circle trajectories for
1990, 2000 and 2010. Because fuel consumption and CO: 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.
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 Enviromnent), 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 2017). 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 3-78. See Annex 3.8 for additional discussion of military data.
Activity data on distillate diesel and residual fuel oil consumption by cargo or passenger carrying marine vessels
departing from U.S. ports were taken from unpublished data collected by the Foreign Trade Division of the U.S.
Department of Commerce's Bureau of the Census (DOC 2017) for 1990 through 2001, 2007 through 2016, 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 (2017). 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 3-79.
Energy 3-99
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Table 3-78: Aviation Jet Fuel Consumption for International Transport (Million Gallons)
Nationality
1990
2005
2012
2013
2014
2015
2016
U.S. and Foreign Carriers
3,222
5,983
6,604
6,748
7,126
7,383
7,383
U.S. Military
862
462
321
294
339
341
333
Total
4,084
6,445
6,925
7,042
7,465
7,725
7,716
Table 3-79: Marine Fuel Consumption for International Transport (Million Gallons)
Fuel Type
1990
2005
2012
2013
2014
2015
2016
Residual Fuel Oil
4,781
3,881
3,069
2,537
2,466
2,718
3,011
Distillate Diesel Fuel & Other
617
444
280
235
261
492
534
U.S. Military Naval Fuels
522
471
381
308
331
326
314
Total
5,920
4,796
3,730
3,081
3,058
3,536
3,858
Note: Totals may not sum due to independent rounding.
Uncertainty and Time-Series Consistency
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.104 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, and in the activity data on military
operations and training that were used to estimate percentages of total fuel use reported as bunker fuel emissions.
Total aircraft and ship fuel use estimates were developed from DoD records, which document fuel sold to the Navy
and Air Force from the Defense Logistics Agency. These data may slightly over or under estimate actual total fuel
use in aircraft and ships because each Service may have procured fuel from, and/or may have sold to, traded with,
and/or given fuel to other ships, aircraft, governments, or other entities. There are uncertainties in 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 2001. 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, Department and military Service data, and expert judgments. The magnitude of the potential errors
104 See uncertainty discussions under Carbon Dioxide Emissions from Fossil Fuel Combustion.
3-100 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 related to the various uncertainties has not been calculated, but is believed to be small. The uncertainties associated
2 with future military bunker fuel emission estimates could be reduced through additional data collection.
3 Although aggregate fuel consumption data have been used to estimate emissions from aviation, the recommended
4 method for estimating emissions of gases other than CO2 in the 2006IPCC Guidelines (IPCC 2006) is to use data by
5 specific aircraft type, number of individual flights and, ideally, movement data to better differentiate between
6 domestic and international aviation and to facilitate estimating the effects of changes in technologies. The IPCC also
7 recommends that cruise altitude emissions be estimated separately using fuel consumption data, while landing and
8 take-off (LTO) cycle data be used to estimate near-ground level emissions of gases other than CO2.105
9 There is also concern regarding the reliability of the existing DOC (2017) data on marine vessel fuel consumption
10 reported at U.S. customs stations due to the significant degree of inter-annual variation.
11 Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
12 through 2016. Details on the emission trends through time are described in more detail in the Methodology section,
13 above.
14 QA/QC and Verification
15 A source-specific QA/QC plan for international bunker fuels was developed and implemented. This effort included a
16 general analysis, as well as portions of a category specific analysis. The category specific procedures that were
17 implemented involved checks specifically focusing on the activity data and emission factor sources and
18 methodology used for estimating CO2, CH4, and N20 from international bunker fuels in the United States. Emission
19 totals for the different sectors and fuels were compared and trends were investigated. No corrective actions were
20 necessary.
21 Planned Improvements
22 The feasibility of including data from a broader range of domestic and international sources for bunker fuels,
23 including data from studies such as the Third IMO GHG Study 2014 (IMO 2014), is being considered.
24 3.11 Wood Biomass and Biofuefs
25 Consumption (CRF Source Category 1A)
26 The combustion of biomass fuels such as wood, charcoal, and wood waste and biomass-based fuels such as ethanol,
27 biogas, and biodiesel generates CO2 in addition to CH4 and N20 already covered in this chapter. In line with the
28 reporting requirements for inventories submitted under the UNFCCC, CO2 emissions from biomass combustion
29 have been estimated separately from fossil fuel CO2 emissions and are not directly included in the energy sector
30 contributions to U.S. totals. In accordance with IPCC methodological guidelines, any such emissions are calculated
31 by accounting for net carbon (C) fluxes from changes in biogenic C reservoirs in wooded or crop lands. For a more
32 complete description of this methodological approach, see the Land Use, Land-Use Change, and Forestry chapter
105 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. The estimates in Mobile Combustion are also likely to include emissions from
ocean-going vessels departing from U.S. ports on international voyages.
Energy 3-101
-------
1 (Chapter 6), which accounts for the contribution of any resulting CO2 emissions to U.S. totals within the Land Use,
2 Land-Use Change, and Forestry sector's approach.
3 Therefore, CO2 emissions from wood biomass and biofuel consumption are not included specifically in summing
4 energy sector totals and are instead included in net carbon fluxes from changes in biogenic carbon reservoirs in the
5 estimates for Land Use, Land-Use Change, and Forestry. However, they are presented here for informational
6 purposes and to provide detail on wood biomass and biofuels consumption.
7 In 2016, total CO2 emissions from the burning of woody biomass in the industrial, residential, commercial, and
8 electric power sectors were approximately 190.2 MMT CO2 Eq. (190,171 kt) (see Table 3-80 and Table 3-81). As
9 the largest consumer of woody biomass, the industrial sector was responsible for 63.3 percent of the CO2 emissions
10 from this source. The residential sector was the second largest emitter, constituting 20.2 percent of the total, while
11 the commercial and electric power sectors accounted for the remainder.
12 Table 3-80: CO2 Emissions from Wood Consumption by End-Use Sector (MMT CO2 Eq.)
End-Use Sector
1990
2005
2012
2013
2014
2015
2016
Industrial
135.3
136.3
125.7
123.1
124.4
122.6
120.4
Residential
59.8
44.3
43.3
59.8
60.9
45.4
38.4
Commercial
6.8
7.2
6.3
7.2
7.8
8.4
8.5
Electric Power
13.3
19.1
19.6
21.4
25.9
25.1
22.9
Total
215.2
206.9
194.9
211.6
218.9
201.5
190.2
Note: Totals may not sum due to independent rounding.
13 Table 3-81: CO2 Emissions from Wood Consumption by End-Use Sector (kt)
End-Use Sector
1990
2005
2012
2013
2014
2015
2016
Industrial
135,348
136,269
125,724
123,149
124,369
122,575
120,417
Residential
59,808
44,340
43,309
59,808
60,884
45,359
38,419
Commercial
6,779
7,218
6,257
7,235
7,760
8,377
8,457
Electric Power
13,252
19,074
19,612
21,389
25,908
25,146
22,878
Total
215,186
206,901
I'M,903
211,581
218,922
201,457
190,171
Note: Totals may not sum due to independent rounding.
14 The transportation sector is responsible for most of the fuel ethanol consumption in the United States. Ethanol used
15 for fuel is currently produced primarily from corn grown in the Midwest, but it can be produced from a variety of
16 biomass feedstocks. Most ethanol for transportation use is blended with gasoline to create a 90 percent gasoline, 10
17 percent by volume ethanol blend known as E-10 or gasohol.
18 In 2016, the United States transportation sector consumed an estimated 1,186.9 trillion Btu of ethanol, and as a
19 result, produced approximately 81.2 MMT CO2 Eq. (81,250 kt) (see Table 3-82 and Table 3-83) of CO2 emissions.
20 Ethanol fuel production and consumption has grown significantly since 1990 due to the favorable economics of
21 blending ethanol into gasoline and federal policies that have encouraged use of renewable fuels.
22 Table 3-82: CO2 Emissions from Ethanol Consumption (MMT CO2 Eq.)
End-Use Sector
1990
2005
2012
2013
2014
2015
2016
Transportation3
4.1
22.4
71.5
73.4
74.9
75.9
78.2
Industrial
0.1
0.5
1.1
1.2
1.0
1.2
1.2
Commercial
0.0
0.1
0.2
0.2
0.2
1.8
1.8
Total
4.2
22.9
72.8
74.7
76.1
78.9
81.2
+ Does not exceed 0.05 MMT CO2 Eq.
a See Annex 3.2, Table A-95 for additional information on transportation consumption of these fuels.
Note: Totals may not sum due to independent rounding.
23 Table 3-83: CO2 Emissions from Ethanol Consumption (kt)
End-Use Sector 1990 2005 2012 2013 2014 2015 2016
Transportation3 4,136 22,414 4 71,510 73,359 74,857 75,946 78,174
3-102 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Industrial 56 468 1,142 1,202 970 1,203 1,238
Commercial 34 60 175 183 249 1,785 1,838
Total 4,227 22,943 72,827 74,743 76,075 78,934 81,250
a See Annex 3.2, Table A-95 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
2017a). 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 2017b).
In 2016, the United States consumed an estimated 266.1 trillion Btu of biodiesel, and as a result, produced
approximately 19.6 MMT CO2 Eq. (19,648 kt) (see Table 3-84 and Table 3-85) 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 2017b). There was no
measured biodiesel consumption prior to 2001 EIA (2017a).
Table 3-84: CO2 Emissions from Biodiesel Consumption (MMT CO2 Eq.)
End-Use Sector
1990
2005
2012
2013
2014
2015
2016
Transportation3
0.0
0.9
8.5
13.5
13.3
14.1
19.6
Total
0.0
0.9
8.5
13.5
13.3
14.1
19.6
+ Does not exceed 0.05 MMT CO2 Eq.
a See Annex 3.2, Table A-95 for additional information on transportation consumption of these fuels.
Note: Totals may not sum due to independent rounding.
Table 3-85: CO2 Emissions from Biodiesel Consumption (kt)
End-Use Sector
1990
2005
2012
2013
2014
2015
2016
Transportation3
0
856
8,470
13,462
13,349
14,077
19,648
Total
0
856
8,470
13,462
13,349
14,077
19,648
a See Annex 3.2, Table A-95 for additional information on transportation consumption of these fuels.
Note: Totals may not sum due to independent rounding.
Methodology
Woody biomass emissions were estimated by applying two EIA gross heat contents (Lindstrom 2006) to U.S.
consumption data (EIA 2017a) (see Table 3-86), provided in energy units for the industrial, residential, commercial,
and electric generation 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. It was assumed that
the woody biomass contains black liquor and other wood wastes, has a moisture content of 12 percent, and is
converted into CO2 with 100 percent efficiency. The emissions from ethanol consumption were calculated by
applying an emission factor of 18.7 MMT C/QBtu (EPA 2010) to U.S. ethanol consumption estimates that were
provided in energy units (EIA 2017a) (see Table 3-87). 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 2017a) (see Table 3-88).
Table 3-86: Woody Biomass Consumption by Sector (Trillion Btu)
End-Use Sector 1990 2005 2012 2013 2014 2015 2016
Industrial 1,441.9 1,451.7 1,339.4 1,312.0 1,325.0 1,305.8 1,282.9
Residential 580.0 430.0 420.0 580.0 590.4 439.9 372.6
Commercial 65.7 70.0 60.7 70.2 75.3 81.2 82.0
Energy 3-103
-------
Electric Power
128.5
185.0
190.2 207.4 251.3 243.9 221.9
Total 2,216.2 2,136.7 2,010.3 2,169.5 2,241.9 2,070.8 1,959.3
Note: Totals may not sum due to independent rounding.
1 Table 3-87: Ethanol Consumption by Sector (Trillion Btu)
End-Use Sector
1990
2005
2012
2013
2014
2015
2016
Transportation
60.4
327.4
1.044.6
1,071.6
1,093.5
1,109.4
1,142.0
Industrial
0.8
6.8
16.7
17.6
14.2
17.6
18.1
Commercial
0.5
0.9
2.6
2.7
3.6
26.1
26.8
Total
61.7
335.1
I.I 163.8
1,091.8
1,111.3
1,153.1
1,186.9
Note: Totals may not sum due to independent rounding.
2 Table 3-88: Biodiesel Consumption by Sector (Trillion Btu)
End-Use Sector
1990
2005
2012
2013
2014
2015
2016
Transportation
0.0
11.6 1
114.7
182.3
180.8
190.6
266.1
Total
0.0
11.6
114.7
182.3
180.8
190.6
266.1
Note: Totals may not sum due to independent rounding.
3 Uncertainty and Time-Series Consistency
4 It is assumed that the combustion efficiency for woody biomass is 100 percent, which is believed to be an
5 overestimate of the efficiency of wood combustion processes in the United States. Decreasing the combustion
6 efficiency would decrease emission estimates for CO2. Additionally, the heat content applied to the consumption of
7 woody biomass in the residential, commercial, and electric power sectors is unlikely to be a completely accurate
8 representation of the heat content for all the different types of woody biomass consumed within these sectors.
9 Emission estimates from ethanol and biodiesel production are more certain than estimates from woody biomass
10 consumption due to better activity data collection methods and uniform combustion techniques.
11 Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
12 through 2016. Details on the emission trends through time are described in more detail in the Methodology section,
13 above.
14 Recalculations Discussion
15 Ethanol and biodiesel values for 1990 through 2015 were not revised relative to the previous Inventory, as there
16 were no historical revisions fromEIA's Monthly Energy Review (EIA 2017a).
17 Planned Improvements
18 Future research will look into the availability of data on woody biomass heat contents and carbon emission factors
19 the see if there are newer, improved data sources available for these factors.
20 The availability of facility-level combustion emissions through EPA's GHGRP will be examined to help better
21 characterize the industrial sector's energy consumption in the United States, and further classify woody biomass
22 consumption by business establishments according to industrial economic activity type. Most methodologies used in
23 EPA's GHGRP are consistent with IPCC, though for EPA's GHGRP, facilities collect detailed information specific
24 to their operations according to detailed measurement standards, which may differ with the more aggregated data
25 collected for the Inventory to estimate total, national U.S. emissions. In addition, and unlike the reporting
26 requirements for this chapter under the UNFCCC reporting guidelines, some facility-level fuel combustion
3-104 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
emissions reported under EPA's GHGRP may also include industrial process emissions.106 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.107
Currently emission estimates from biomass and biomass-based fuels included in this inventory are limited to woody
biomass, ethanol, and biodiesel. Other forms of biomass-based fuel consumption include biogas. An effort will be
made to examine sources of data for biogas including data from EIA for possible inclusion. EIA (2017a) 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.
As per discussion in Section 3.1, an additional planned improvement is to evaluate and potentially update EPA's
method for allocating motor gasoline consumption across the Transportation, Industrial and Commercial sectors to
improve accuracy and create a more consistent time series. Further research will be conducted to determine if
changes also need to be made to ethanol allocation between these sectors to match gasoline's sectoral distribution.
106 See .
107 See .
Energy 3-105
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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. 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.
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 (N20), and fluorinated GHGs (e.g., HFC-23). The GHG
byproduct generating processes included in this chapter include iron and steel production and metallurgical coke
production, cement production, lime production, other process uses of carbonates (e.g., flux stone, flue gas
desulfurization, and glass manufacturing), ammonia production and urea consumption, petrochemical production,
aluminum production, HCFC-22 production, soda ash production and use, titanium dioxide production, ferroalloy
production, glass production, zinc production, phosphoric acid production, lead production, silicon carbide
production and consumption, nitric acid production, adipic acid production, and caprolactam production.
Greenhouse gases that are used in manufacturing processes or by end-consumers include man-made compounds
such as hydrofluorocarbons (HFCs), perfluorocarbons (PFCs), sulfur hexafluoride (SF6), 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 accumulate in the atmosphere as long as emissions continue. 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 semiconductor manufacture, electric power transmission and distribution, 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 semiconductor manufacturing and anesthetic and
aerosol applications.
In 2016, IPPU generated emissions of 375.7 million metric tons of CO2 equivalent (MMT CO2 Eq.), or 5.7 percent
of total U.S. greenhouse gas emissions. Carbon dioxide emissions from all industrial processes were 163.6 MMT
CO2 Eq. (163,647 kt CO2) in 2016, or 3.1 percent of total U.S. CO2 emissions. Methane emissions from industrial
processes resulted in emissions of approximately 0.2 MMT CO2 Eq. (8 kt CH4) in 2016, which was less than 1
percent of U.S. CH4 emissions. Nitrous oxide emissions from IPPU were 23.7 MMT CO2 Eq. (79 kt N2O) in 2016,
or 6.4 percent of total U.S. N20 emissions. In 2016 combined emissions of HFCs, PFCs, SF6, and NF3 totaled 188.2
MMT CO2 Eq. Total emissions from IPPU in 2016 were 10.3 percent more than 1990 emissions. Indirect
greenhouse gas emissions also result from IPPU, and are presented in Table 4-111 in kilotons (kt).
Industrial Processes and Product Use 4-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Figure 4-1: 2016 Industrial Processes and
(MMT COz Eq.)
Product Use Chapter Greenhouse Gas Sources
Substitution of Ozone Depleting Substances
Iron and Steel Production & Metallurgical Coke Production
Cement Production
Petrochemical Production
Lime Production
Other Process Uses of Carbonates
Ammonia Production
Nitric Acid Production
Adipic Acid Production
Semiconductor Manufacture
Carbon Dioxide Consumption
Electrical Transmission and Distribution
NiO from Product Uses
Urea Consumption for Non-Agricultural Purposes
HCFC-22 Production
Aluminum Production
Caprolactam, Glyoxal, and Glyoxylic Acid Production
Ferroalloy Production
Soda Ash Production
Titanium Dioxide Production
Glass Production
Magnesium Production and Processing
Phosphoric Acid Production
Zinc Production
Lead Production
Silicon Carbide Production and Consumption
174
¦
¦
¦
I
I
I
I
I
< 0.5
Industrial Processes and Product
Use as a Portion of all Emissions
5.7%
10
20
30
40
50
60
70
MMT CO, Eq.
The increase in overall IPPU emissions since 1990 reflects a range of emission trends among the emission sources.
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). Similarly,
CO2 and CH4 emissions from many chemical production sources have either decreased or not changed significantly
since 1990, with the exception of petrochemical production which lias steadily increased. Emissions from mineral
sources have either increased (e.g., cement manufacturing) or not changed significantly (e.g., glass and lime
manufacturing) since 1990 but largely follow economic cycles. Hydrofluorocarbon emissions from the substitution
of ODS have increased drastically since 1990, while the emissions of HFCs, PFCs, SF6, and NF3 from other sources
have generally declined. Nitrous oxide emissions from the production of adipic and nitric acid have decreased, while
N2O emissions from product uses have remained nearly constant over time. Trends are explained further within each
emission source category throughout the chapter.
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 revised United Nations Framework Convention on Climate
Change (UNFCCC) reporting guidelines for national inventories (IPCC 2007).1 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.
1 See .
4-2 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Table 4-1: Emissions from Industrial Processes and Product Use (MMT CO2 Eq.)
Gas/Source 1990 2005 2012 2013 2014 2015 2016
co2
207.3
190.2
169.9
171.8
177.9
171.4
163.6
Iron and Steel Production &
Metallurgical Coke Production
101.5
68.0
55.4
53.3
58.2
47.7
42.2
Iron and Steel Production
99.0
66.0
54.9
51.5
56.2
44.9
40.9
Metallurgical Coke Production
2.5
2.0
0.5
1.8
2.0
2.8
1.3
Cement Production
33.5
46.2
35.3
36.4
39.4
39.9
39.4
Petrochemical Production
21.2
26.8
26.5
26.4
26.5
28.1
27.4
Lime Production
11.7
14.6
13.8
14.0
14.2
13.3
13.3
Other Process Uses of Carbonates
4.9
6.3
8.0
10.4
11.8
11.2
11.2
Ammonia Production
13.0
9.2
9.4
10.0
9.6
10.6
11.2
Carbon Dioxide Consumption
1.5
1.4
4.0
4.2
4.5
4.5
4.5
Urea Consumption for Non-
Agricultural Purposes
3.8
3.7
4.4
4.1
1.5
4.2
4.0
Ferroalloy Production
2.2
1.4
1.9
1.8
1.9
2.0
1.8
Soda Ash Production
1.4
1.7
1.7
1.7
1.7
1.7
1.7
Titanium Dioxide Production
1.2
1.8
1.5
1.7
1.7
1.6
1.6
Aluminum Production
6.8
4.1
3.4
3.3
2.8
2.8
1.3
Glass Production
1.5
1.9
1.2
1.3
1.3
1.3
1.3
Phosphoric Acid Production
1.5
1.3
1.1
1.1
1.0
1.0
1.0
Zinc Production
0.6
1.0
1.5
1.4
1.0
0.9
0.9
Lead Production
0.5
0.6
0.5
0.5
0.5
0.5
0.5
Silicon Carbide Production and
Consumption
0.4
0.2
0.2
0.2
0.2
0.2
0.2
Magnesium Production and
Processing
+
+
+
+
+
+
+
CH4
0.3
0.1
0.1
0.1
0.2
0.2
0.2
Petrochemical Production
0.2
0.1
0.1
0.1
0.1
0.2
0.2
Ferroalloy Production
+
+
+
+
+
+
+
Silicon Carbide Production and
Consumption
+
+
+
+
+
+
+
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
22.4
21.0
22.8
22.3
23.7
Nitric Acid Production
12.1
11.3
10.5
10.7
10.9
11.6
10.2
Adipic Acid Production
15.2
7.1
5.5
3.9
5.4
4.3
7.0
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
2.0
2.0
2.0
2.0
2.0
Semiconductor Manufacturing
+
0.1
0.2
0.2
0.2
0.2
0.2
HFCs
46.6
120.0
156.0
159.1
166.8
173.3
177.1
Substitution of Ozone Depleting
Substances3
0.3
99.8
150.3
154.8
161.4
168.6
173.9
HCFC-22 Production
46.1
20.0
5.5
4.1
5.0
4.3
2.8
Semiconductor Manufacturing
0.2
0.2
0.2
0.2
0.3
0.3
0.3
Magnesium Production and
Processing
0.0
0.0
+
0.1
0.1
0.1
0.1
PFCs
24.3
6.7
5.9
5.8
5.6
5.1
4.3
Semiconductor Manufacturing
2.8
3.3
3.0
2.8
3.1
3.1
3.0
Aluminum Production
21.5
3.4
2.9
3.0
2.5
2.0
1.4
Substitution of Ozone Depleting
Substances3
0.0
+
+
+
+
+
+
SF«
28.8
11.7
6.6
6.3
6.3
5.9
6.2
Electrical Transmission and
Distribution
23.1
8.3
4.6
4.5
4.6
4.2
4.3
Industrial Processes and Product Use 4-3
-------
Magnesium Production and
Processing 5.2 2.7
Semiconductor Manufacturing 0.5 0.7
NFs + 0.5
Semiconductor Manufacturing + 0.5
1.6
0.3
0.6
0.6
1.5
0.4
0.6
0.6
1.0
0.7
0.5
0.5
0.9
0.7
0.6
0.6
1.0
0.8
0.6
0.6
Total
340.5
354.2
361.6
364.7
380.2
378.8
375.7
+ Does not exceed 0.05 MMT CO2 Eq.
a Small amounts of PFC emissions also result from this source.
Note: Totals may not sum due to independent rounding.
1 Table 4-2: Emissions from Industrial Processes and Product Use (kt)
Gas/Source
1990
2005
2012
2013
2014
2015
2016
CO2
207,281
190,171
169,888
171,841
177,906
171,439
163,647
Iron and Steel Production &
Metallurgical Coke Production
101,48"
68,047
55,449
53,348
58,234
47,718
42,219
Iron and Steel Production
98,984
66,003
54,906
51,525
56,220
44,879
40,896
Metallurgical Coke Production
2,503
2,044
543
1,824
2,014
2,839
1,323
Cement Production
33,484
46,194
35,270
36,369
39,439
39,907
39,439
Petrochemical Production
21,20 ^
26,794
26,501
26,395
26,496
28,062
27,411
Lime Production
11,700
14,552
13,785
14,028
14,210
13,342
13,342
Other Process Uses of Carbonates
4,90"
6,339
8,022
10,414
11,811
11,237
11,237
Ammonia Production
13,04"
9,196
9,377
9,962
9,619
10,571
11,234
Carbon Dioxide Consumption
1,472
1,375
4,019
4,188
4,471
4,471
4,471
Urea Consumption for Non-
Agricultural Purposes
3,784
3,653
4,392
4,074
1,541
4,169
3,959
Ferroalloy Production
2,152
1,392
1,903
1,785
1,914
1,960
1,796
Soda Ash Production
1,431
1,655
1,665
1,694
1,685
1,714
1,723
Titanium Dioxide Production
1,195
1,755
1,528
1,715
1,688
1,635
1,608
Aluminum Production
6,831
4,142
3,439
3,255
2,833
2,767
1,334
Glass Production
1,535
1,928
1,248
1,317
1,336
1,299
1,299
Phosphoric Acid Production
1,529
1,342
1,118
1,149
1,038
999
992
Zinc Production
632
1,030
1,486
1,429
956
933
925
Lead Production
516
553
527
546
459
473
482
Silicon Carbide Production and
Consumption
375
219
158
169
173
180
174
Magnesium Production and
Processing
1
3
2
2
2
3
3
CH4
12
4
4
4
6
9
8
Petrochemical Production
9
3
3
3
5
7
7
Ferroalloy Production
1
+
1
+
1
1
1
Silicon Carbide Production and
Consumption
1
+
+
+
+
+
+
Iron and Steel Production &
Metallurgical Coke Production
1
1
+
+
+
+
+
Iron and Steel Production
/
1
+
+
+
+
+
Metallurgical Coke Production
0
0
0
0
0
0
0
N2O
112
84
75
71
77
75
79
Nitric Acid Production
41
38
35
36
37
39
34
Adipic Acid Production
51
24
19
13
18
14
23
NjO from Product Uses
14
14
14
14
14
14
14
Caprolactam, Glyoxal, and
Glyoxylic Acid Production
6
7
7
7
7
7
7
Semiconductor Manufacturing
+
+
1
1
1
1
1
HFCs
1V1
M
M
M
M
M
M
Substitution of Ozone Depleting
Substances3
M
M
M
M
M
M
M
4-4 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
HCFC-22 Production
1
+
+
+
+
+
Semiconductor Manufacturing
M
M
M
M
M
M
M
Magnesium Production and
Processing
0
0 !
+
+
+
+
+
PFCs
M
M
M
M
M
M
M
Semiconductor Manufacturing
M
M j
M
M
M
M
M
Aluminum Production
M
M i
M
M
M
M
M
Substitution of Ozone Depleting
Substances3
0
+
+
+
+
+
+
SF«
1
1
+
+
+
+
+
Electrical Transmission and
Distribution
1
+
+
+
+
+
+
Magnesium Production and
Processing
+
+
+
+
+
+
+
Semiconductor Manufacturing
+
+
+
+
+
+
+
NF3
+
+
+
+
+
+
+
Semiconductor Manufacturing
-
+
+
+
+
+
+
+ Does not exceed 0.5 kt.
M (Mixture of gases)
a Small amounts of PFC emissions also result from this source.
Note: Totals may not sum due to independent rounding.
The UNFCCC incorporated the 2006IPCC Guidelines for National Greenhouse Gas Inventories (2006IPCC
Guidelines) as the standard for Annex I countries at the Nineteenth Conference of the Parties (Warsaw, November
11-23, 2013). This chapter presents emission estimates calculated in accordance with the methodological guidance
provided in these guidelines. For additional detail on IPPU sources that are not estimated in this Inventory report,
please review Annex 5, Assessment of the Sources and Sinks of Greenhouse Gas Emissions Not Included. These
sources are not estimated due to various national circumstances, such as emissions from a source may not be
currently occurring in the United States, data are not currently available for those emission sources (e.g., ceramics,
non-metallurgical magnesium production), emissions are included elsewhere within the Inventory report, or also that
data suggest that emissions are not significant. Information on planned improvements for specific IPPU source
categories can be found in the Planned Improvements section of the individual source category.
Finally, as mentioned in the Energy chapter of this report (Box 3-6), 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
fossil fuels are to be reported under IPPU, rather than Energy; however, due to national circumstances regarding the
allocation of energy statistics and carbon (C) balance data, the United States reports non-energy uses in the Energy
chapter of this Inventory. Reporting these non-energy use emissions under IPPU would involve making artificial
adjustments to the non-energy use C balance. These artificial adjustments would also result in the C emissions for
lubricants, waxes, and asphalt and road oil being reported under IPPU, while the C storage for lubricants, waxes, and
asphalt and road oil would be reported under Energy. To avoid presenting an incomplete C balance, double-
counting, and adopting a less transparent approach, the entire calculation of C storage and C emissions is therefore
conducted in the Non-Energy Uses of Fossil Fuels category calculation methodology and reported under the Energy
sector. For more information, see the Methodology section for CO2 from Fossil Fuel Combustion and Section 3.2,
Carbon Emitted from Non-Energy Uses of Fossil Fuels.
In addition, 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 and zinc production) are reported in the IPPU chapter, unless otherwise noted due to specific national
circumstances. 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 1 A)) and Annex 2.1, Methodology for Estimating Emissions of CO2 from Fossil Fuel Combustion.
Industrial Processes and Product Use 4-5
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
QA/QC and Verification Procedures
For IPPU sources, a detailed QA/QC plan was developed and implemented for specific categories. This plan was
based on the overall Quality Assurance/Quality Control and Uncertainty Management Plan for the U.S. Greenhouse
Gas Inventory (QA/QC Management Plan), but was tailored to include specific procedures recommended for these
sources. 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.
General or tier 1 QA/QC procedures 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 ODS) 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, and planned improvement sections.
For most IPPU categories, activity data are obtained via aggregation of facility-level data from EPA's GHGRP,
national commodity surveys conducted by U.S. Geologic Survey National Minerals Information Center, U.S.
Department of Energy (DOE), U.S. Census Bureau, 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.
The uncertainty analysis performed to quantify uncertainties associated with the 2016 emission estimates from IPPU
continues a multi-year process for developing credible quantitative uncertainty estimates for these source categories
using the IPCC Tier 2 approach. As the process continues, the type and the characteristics of the actual probability
density functions underlying the input variables are identified and better characterized (resulting in development of
more reliable inputs for the model, including accurate characterization of correlation between variables), based
primarily on expert judgment. Accordingly, the quantitative uncertainty estimates reported in this section should be
considered illustrative and as iterations of ongoing efforts to produce accurate uncertainty estimates. The correlation
among data used for estimating emissions for different sources can influence the uncertainty analysis of each
individual source. While the uncertainty analysis recognizes very significant connections among sources, a more
compiclicnsn c ippio i :h that accounts for all linkages will be identified as the uncertainty analysis moves forward.
cal Approach for Estimating and Reporting U.S. Emissions and
I
In following ilic U nncd 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). 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
4-6 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
these reports are comparable. The presentation of emissions and removals provided in this Inventory do 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.
Box 4-2: Industrial Processes Data from EPA's Greenhouse Gas Reporting Program
On October 30, 2009, the U.S. 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). The rule 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 ("Subparts"). 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 reported for facilities
subject to 40 CFR Part 98, though some source categories first reported data for calendar year 2011.
EPA's GHGRP dataset and the data presented in this Inventory 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. 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 categories to
improve the national estimates presented in this Inventory consistent with IPCC guidelines. While many
methodologies used in EPA's GHGRP are consistent with IPCC, it should be noted that the definitions for source
categories in EPA's GHGRP may differ from those used in this Inventory in meeting the UNFCCC reporting
guidelines (IPCC 2011). 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. Further information on
the reporting categorizations in EPA's GHGRP and specific data caveats associated with monitoring methods in
EPA's GHGRP lias been provided on the GHGRP website.2
For certain source categories in this Inventory (e.g., nitric acid production, cement production and petrochemical
production), EPA has also 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.3 Specific uses of aggregated facility-level data are
described in the respective methodological sections. For other source categories in this chapter, as indicated in the
respective planned improvements sections, 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. As stated previously in the Introduction chapter, this year EPA has integrated
GHGRP information for various Industrial Processes and Product Use categories4 and also identified places where
EPA plans to integrate additional GHGRP data in additional categories5 (see those categories' Planned Improvement
sections for details). EPA lias paid particular attention to ensuring time-series consistency for major recalculations
that have occurred from the incorporation of GHGRP data into these categories, consistent with 2006 IPCC
2 See .
3 U.S. EPA Greenhouse Gas Reporting Program. Developments on Publication of Aggregated Greenhouse Gas Data, November
25, 2014. See .
4 Adipic Acid Production, Aluminum Production, Carbon Dioxide Consumption, Cement Production, Electrical Transmission
and Distribution, HCFC-22 Production, Lime Production, Magnesium Production and Processing, Substitution of ODS, Nitric
Acid Production, Petrochemical Production, and Semiconductor Manufacture.
5 Ammonia Production, Glass Production and Other fluorinated gas production.
Industrial Processes and Product Use 4-7
-------
1 Guidelines and IPCC Technical Bulletin on Use of Facility-Specific Data in National GHG Inventories.6 EPA
2 verifies annual facility-level reports through a multi-step process to identify potential errors and ensure that data
3 submitted to EPA are accurate, complete, and consistent.7 The GHGRP dataset is a particularly important annual
4 resource and will continue to be important for improving emissions estimates from Industrial Process and Product
5 Use in future Inventory reports. Additionally, EPA's GHGRP has and will continue to enhance QA/QC procedures
6 and assessment of uncertainties within the IPPU categories (see those categories for specific QA/QC details
7 regarding the use of GHGRP data).
8
9 4.1 Cement Production (CRF Source Category
10 2A1)
11 Cement production is an energy- and raw material-intensive process that results in the generation of carbon dioxide
12 (CO2) from both the energy consumed in making the cement and the chemical process itself. Emissions from fuels
13 consumed for energy purposes during the production of cement are accounted for in the Energy chapter.
14 During the cement production process, calcium carbonate (CaCCh) is heated in a cement kiln at a temperature range
15 of about 700 to 1000 degrees Celsius (1,292 to 1,832 degrees Fahrenheit) to form lime (i.e., calcium oxide or CaO)
16 and CO2 in a process known as calcination or calcining. The quantity of CO2 emitted during cement production is
17 directly proportional to the lime content of the clinker. During calcination, each mole of limestone (CaCCh) heated
18 in the clinker kiln forms one mole of lime (CaO) and one mole of CO2:
19 CaC03 + heat -» CaO + C02
20 Next, the lime is combined with silica-containing materials to produce clinker (an intermediate product), with the
21 earlier byproduct CO2 being released to the atmosphere. The clinker is then rapidly cooled to maintain quality,
22 mixed with a small amount of gypsum and potentially other materials (e.g., slag, etc.), and used to make Portland
23 cement.8
24 Carbon dioxide emitted from the chemical process of cement production is the second largest source of industrial
25 CO2 emissions in the United States. Cement is produced in 34 states and Puerto Rico. Texas, California, Missouri,
26 Florida, and Alabama were the leading cement-producing states in 2016 and accounted for almost 50 percent of total
27 U.S. production (USGS 2017). Clinker production in 2016 decreased approximately 1 percent from 2015 levels as
28 cement sales increased significantly in 2016, with much of the increase accounted for by imports. In 2016, U.S.
29 clinker production totaled 75,800 kilotons (EPA 2017). The resulting CO2 emissions were estimated to be 39.4
30 MMT C02 Eq. (39,439 kt) (see Table 4-3).
6 See .
7 See .
8 Approximately three percent of total clinker production is used to produce masonry cement, which is produced using
plasticizers (e.g., ground limestone, lime, etc.) and Portland cement (USGS 2011). Carbon dioxide emissions that result from the
production of lime used to create masonry cement are included in the Lime Manufacture source category.
4-8 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Table 4-3: CO2 Emissions from Cement Production (MMT CO2 Eq. and kt)
Year
MMT CO2 Eq.
kt
1000
33.5
33,484
2005
46.2
46,104
2012
35.3
35,270
2013
36.4
36,360
2014
30.4
30,430
2015
30.0
30,007
2016
30.4
30,430
2 Greenhouse gas emissions from cement production increased every year from 1991 through 2006 (with the
3 exception of a slight decrease in 1997), but decreased in the following years until 2009. Emissions from cement
4 production were at their lowest levels in 2009 (2009 emissions are approximately 28 percent lower than 2008
5 emissions and 12 percent lower than 1990). Since 2010, emissions have increased by roughly 25 percent. In 2016,
6 emissions from cement production decreased by 1 percent from 2015 levels.
7 Emissions since 1990 have increased by 18 percent. Emissions decreased significantly between 2008 and 2009, due
8 to the economic recession and associated decrease in demand for construction materials. Emissions increased
9 slightly from 2009 levels in 2010, and continued to gradually increase during the 2011 through 2015 time period due
10 to increasing consumption. Emissions in 2016 decreased slightly from 2015 levels. Cement continues to be a critical
11 component of the construction industry; therefore, the availability of public and private construction funding, as well
12 as overall economic conditions, have considerable impact on the level of cement production.
13 Methodology
14 Carbon dioxide emissions were estimated using the Tier 2 methodology from the 2006IPCC Guidelines. The Tier 2
15 methodology was used because detailed and complete data (including weights and composition) for carbonate(s)
16 consumed in clinker production are not available, and thus a rigorous Tier 3 approach is impractical. Tier 2 specifies
17 the use of aggregated plant or national clinker production data and an emission factor, which is the product of the
18 average lime fraction for clinker of 65 percent and a constant reflecting the mass of CO2 released per unit of lime.
19 The U.S. Geological Survey (USGS) mineral commodity expert for cement has confirmed that this is a reasonable
20 assumption for the United States (VanOss 2013a). This calculation yields an emission factor of 0.51 tons of CO2 per
21 ton of clinker produced, which was determined as follows:
22 EFciinker = 0.650 CaO X [(44.01 g/mole CO2) (56.08 g/mole CaO)] = 0.510 tons CCh/ton clinker
23 During clinker production, some of the clinker precursor materials remain in the kiln as non-calcinated, partially
24 calcinated, or fully calcinated cement kiln dust (CKD). The emissions attributable to the calcinated portion of the
25 CKD are not accounted for by the clinker emission factor. The IPCC recommends that these additional CKD CO2
26 emissions should be estimated as two percent of the CO2 emissions calculated from clinker production (when data
27 on CKD generation are not available). Total cement production emissions were calculated by adding the emissions
28 from clinker production to the emissions assigned to CKD (IPCC 2006).
29 Furthermore, small amounts of impurities (i.e., not calcium carbonate) may exist in the raw limestone used to
30 produce clinker. The proportion of these impurities is generally minimal, although a small amount (1 to 2 percent) of
31 magnesium oxide (MgO) may be desirable as a flux. Per the IPCC Tier 2 methodology, a correction for MgO is not
32 used, since the amount of MgO from carbonate is likely very small and the assumption of a 100 percent carbonate
33 source of CaO already yields an overestimation of emissions (IPCC 2006).
34 The 1990 through 2012 activity data for clinker production (see Table 4-4) were obtained from USGS (Van Oss
35 2013a, Van Oss 2013b). Clinker production data for 2013 were also obtained from USGS (USGS 2014). The data
36 were compiled by USGS (to the nearest ton) through questionnaires sent to domestic clinker and cement
37 manufacturing plants, including the facilities in Puerto Rico. During the 1990 through 2015 Inventory report cycle,
Industrial Processes and Product Use 4-9
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
EPA began incorporating clinker production data from EPA's GHGRP to estimate emissions in these respective
years. Clinker production values in the current Inventory report utilize GHGRP data for the years 2014, 2015 and
2016 (EPA 2017). More details on how this change compares to USGS reported data can be found in the section on
Uncertainty and Time-Series Consistency.
Table 4-4: Clinker Production (kt)
Year Clinker
1000 64.355
2005 88.783
2012 67,788
2013 60,000
2014 75,800
2015 76,700
201 6 75,800
Notes: Clinker production from 1000 through 2016
includes Puerto Rico.
Uncertainty and Time-Series Consistency
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 CaCCb, 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). CKD loss can range from 1.5 to 8 percent depending upon plant
specifications. 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 surface area. Because the amount of CO2 reabsorbed is thought to be minimal, it was
not estimated. However, see Planned Improvements described below to reassess this assumption by conducting a
review to identify recent studies that may provide information or data on reabsorption rates of cement products.
Total U.S. clinker production is assumed to have low uncertainty. USGS takes a number of manual steps to review
clinker production reported through their voluntary surveys. EPA continues to assess the accuracy of reported
clinker production data required by GHGRP Subpart H facilities for current and future Inventory reports. EPA
verifies annual facility-level reports through a multi-step process (e.g., combination of electronic checks and manual
reviews by staff) to identify potential errors and ensure that data submitted to EPA are accurate, complete, and
consistent. Based on the results of the verification process, the EPA follows up with facilities to resolve mistakes
that may have occurred.9 Facilities are also required to monitor and maintain records of monthly clinker production.
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 GHGRP with existing data from USGS. This
was to ensure time-series consistency of the emission estimates presented in the Inventory. For the year 2014, USGS
and GHGRP clinker production data showed a difference of approximately 2 percent, while in 2015 and in 2016 that
difference decreased to less than 1 percent between the two sets of activity data. This difference resulted in an
increase of emissions compared to USGS data (USGS 2016a) by 0.7 MMT CO2 Eq. in 2014 and 0.0 MMT CO2 Eq.
in 2015 and in 2016.
The results of 1 he \pproach 2 quaulitali\ e uuccriaiiits ; 111; 11\ sis are suniniari/cd 111 Table 4-5 IJased 011 ihc
uncertainties associated u 11 li total I S clinker production. I lie ('() emission factor for clinker production, and the
emission factor lor additional CO emissions from CKI). ) I(> CO emissions from cement production were
estimated to he hem con "— 0 ;nn.l 41 X MMT CO l!q at the l>5 percent confidence le\ el I his confidence le\ el
9 See .
4-10 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 indicalcsa laimc <> pcivciil lvlo\\ and (> pcicciil ;ihn\c I lie emission esiimale of 'lM \1\1T('()
2 I t|
3 Table 4-5: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Cement
4 Production (MMT CO2 Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY REPORT
Souriv
(¦;is
21116 remission 1'sii 111:1 lc-
(MM'I'CO: i:|KT I.OXUT I |)|KT
1 Sound Bound Bound Bound
Cement Prodn
ction
t ( )
30.4
37.0 41.8 -6% +6%
Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
5 Methodological approaches were applied to the entire time series to ensure time-series consistency from 1990
6 through 2016. Details on the emission trends through time are described in more detail in the Methodology section,
7 above. More information on the consistency in clinker production data and emissions across the time series with the
8 use of GHGRP clinker data for 2014 through 2016 can be found in the Uncertainty and Time-Series Consistency
9 section.
10 For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
11 Chapter 6 of the 2006IPCC Guidelines, see QA/QC and Verification Procedures section in the introduction of the
12 IPPU chapter.
13 Planned Improvements
14 In response to comments from the Portland Cement Association (PCA) and UNFCCC expert technical reviews, EPA
15 is continuing to evaluate and analyze data reported under EPA's GHGRP that would be useful to improve the
16 emission estimates for the Cement Production source category. EPA held a technical meeting with PCA in August
17 2016 to review Inventory methods and available data from the GHGRP data set. Most cement production facilities
18 reporting under EPA's GHGRP use Continuous Emission Monitoring Systems (CEMS) to monitor and report CO2
19 emissions, thus reporting combined process and combustion emissions from kilns. In implementing further
20 improvements and integration of data from EPA's GHGRP, the latest guidance from the IPCC on the use of facility -
21 level data in national inventories will be relied upon, in addition to category specific QC methods recommended by
22 2006 IPCC Guidelines,10 EPA's long-term improvement plan includes continued assessment of the feasibility of
23 using additional GHGRP information, in particular disaggregating aggregated GHGRP emissions consistent with
24 IPCC and UNFCCC guidelines to present both national process and combustion emissions streams. This long-term
25 planned analysis is still in development and has not been updated for this current Inventory.
26 Finally, in response to feedback from PCA during the public review of the draft Inventory in March 2017, EPA
27 plans to meet with PCA to discuss additional long-term improvements to review methods and data used to estimate
28 CO2 emissions from cement production to account for both organic material and magnesium carbonate in the raw
29 material, and to discuss the carbonation that occurs later in the cement product lifecycle. EPA will work with PCA
30 to identify data and studies on the average MgO content of clinker produced in the United States, the average carbon
31 content for organic materials in kiln feed in the United States, and CO2 reabsorption rates via carbonation for various
32 cement products.
10 See .
Industrial Processes and Product Use 4-11
-------
1
2
3
4 I.inic is ;in iniporimil ni;iiiiil';icliircd product wiili iii;in\ iiidusiri;il. chcniic;il. ;uid cu\ iroiinicui;il ;ipphc;ilious l.unc
5 production iu\ol\cs lliree 111:1111 processes' sioue prcp;ir;ilioii. ji 11:11 k>ii. ;iud hulr;ilioii Ciirbou dio\idc (CO ) is
6 ucucriilcd diiriuu ilie c;ilciu;iliou si;iuc. w lieu limestone niosiK :isictl :il liiuli
7 icnipcmiiircs 111 ;i kiln lo produce ji 11111 o\ide (( :K )) ;nid C() I lie CO is m\cu i»IT;is ;i u;is ;md is ik>i'iii:iIl\
8 eniilled In I he ;iiniosphcrc
9 C.uCO ¦ — CuO + CO.
10 Sonic ii|" 1 lie ('() uc 1 ici": 11ctl diiiinu 1 lie production process. Iriwe\cr. is rcco\ crcd ;il some f;ieililies for use 111 su&ir
11 rcfiiiiim ;md preeipil;iled e;ileiiim c;irbou;ilc (l'C(') prodiielion 11 IEmissions from fuels consumed lor ciicrus
12 purposes durum ilie prodiielion ol" lime :ire ;ieeouiiied lor 1111 lie l!ueru\ eli;ipier
13 I"nr I S. operations. ilie lerm "lime " ;ielu;ill\ refers 10 :i \:irielv of elieniienl eonipouuds These include C;iO. or
14 hmh-c;ilcuini quicklime. c;ilciuni h\dro\ide (( ;i(OI 11 1. or h\dr;iled lime, dolomiiic (|iucklime (|C;i( )»\1u()|i. ;md
15 dolomiiic 11> tlr: 11e (|( ;i(OI I) •\1uO| or |( :i(OI 11 'MuiOl I) 11
16 The currciii lime nuirkel is ;ippro\ini;iicl> disirihuied ;icioss l'i\e cud-use ciilcuories ;is follows nicl;illurmc;il uses.
17 perccui. eu\ iroiinicui;il uses. "51 pcrcciii. chcniic;il ;uid uidiisin;il uses. 22 percciii. coiisirucliou uses. l) pereeui.
18 ;uid relr;ielor\ doloniiie. I pereeui (I S( iS 2d I (ih 1 .The ni;ijor uses :ire 111 si eel ni;ikiim. Hue u:is desiill'uri/;iiioii
19 svsienis ;ii eo;il-l'ired eleeirie power phnils. coiisirucliou. ;iud w;iler ire;iinieul. ;is well ;is uses 111 milium, pulp ;iud
20 p;iper;uid preeipil;iled ciilciuni e;irhou;ile iii;iiiiil';ieluriim l.inie is ;ilso used ;is ;i CO scrubber. ;iud llierc h;is hcen
21 c\pcriniciii;iliou 011 ilie use of lime lo c:ipiure ('() from elcciric power pl;uils
22 l.inie prodiielion 111 ilie I niled Si:iics iucludiim I'ucrio kico w;is reported lo be IX.2~lH\iloloiis 111 2D 15
23 (Coniihers 2(>l~i. \i \c: 1 r-e 1 it! 2<)I5. llierc were opcriiiuii: pruii;ir\ lime phnils 1111 lie I lined Si;iics. iiicludiim
24 I'ucrio kico 1 - Prineip:il lime producing suites ;irc Missouri. \l:ih:i 111:1. Kentucky. Ohio. Tc\:is (I S(iS 2ul<>;n
25 I S. lime production resulted 111 estimated ucl CO emissions of H ' MMICO Lq (I V ^42 kn (see I :ible 4-(> ;md
26 T;ihle 4-~i The ireuds 111 C() emissions from lime prodiielion ;ire dircclK proporiiou;il lo ireuds 111 production.
27 w liieli ;ire described below
28 Table 4-6: CO2 Emissions from Lime Production (MMT CO2 Eq. and kt)
^ i ;i r
,MM"I" CO; i:<|.
kl
I 990
I I 700
201 |
2012
14.6
14.0
13.8
14,552
13,982
13,785
2013
2014
I4.0
I4.2
I"?
14.028
14.210
1"? ,4?
11 PCC is obtained from the reaction of CO2 with calcium hydroxide. It is used as a filler and/or coating in the paper, food, and
plastic industries.
12 In 2015, 74 operating primary lime facilities in the United States reported to the EPA Greenhouse Gas Reporting Program.
4-12 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
Table 4-7: Potential, Recovered, and Net CO2 Emissions from Lime Production (kt)
U-;ir
Puk-nlhil
RlinWIVll 1
Nil 1! miss in ns
I 990
I 1.959
259
1 1 700
2005
15.074
522
14.552
201 I
14,389
407
13.982
20I2
14.258
473
13,785
2013
14,495
467
14.028
20I4
14.715
S()S
14,210
2015
13.764
422
13.342
" for s
ugar rclining ant:
[ l'CC production.
Note:
Totals may nol si
mi due to indepen
dent rounding.
2 In 2<>15. lime prndiielinn decreased cnnip;ired In 2(>I4 levels (decrease nl'jihmil (> percciili ;il IX.2~9 kilnlnns. nwinu
3 in decreased cniisnnipiinii hv ilie I S iiniil'errniis nicl;illiirme;il indiisiries (prini;irilv enpperi ;md si eel industries
4 (( nnithers 2o I I S( iS 2(> I(<;ii
5 Methodology
6 Ti< c;ilcul;itc ennssiniis. the miinuuis nf limh-c;ilcuuii ;md dnlnniitic lime produced were multiplied In ilieir
7 respective eniissinn liictnrs iisiiiu ilie Tier 2 ;ipprn;ich Irnni llie 2nnf- ll'< '<' < ini,!i. /iih > The emission liictnr is ilie
8 prnduct ill ilie stoichiometric r;itio between ('() ;md (;i(). ;md 1 lie ;iver;me ( ;i() ;md \lu() content for lime The (;i()
9 ;md \lu() content for lime is ;issiimed In he 95 pereenl for both limh-c;ilciuni ;md doloniitic lime (IIH ( 2()()<¦) I lie
10 emission factors were e;ileul;iled ;is follows'
11 I'nr limh-c;ilcuuii lime
12 |( 14.01 ti/niok' CO ¦) -h (.Ki.OH n/niok' CnO)| x (()/).">()() (I;i()/1i 111c) = 0.7-l-.">."> n CO ¦/}• lime
13 I'nrdnlnmilie lime'
14 |(tm.02 o/mnk-CO ) h- t>/mc»k* C:i())| x (().').">()() CnO/limt') = O.H(i7.~> ?¦ CO ¦/}• lime
15 Production w;is ;id|iistcd In remove I he iikiss of chcniic;illv enmhined w;iler found in lmlr;iled lime, delermined
16 ;ieenrdinu In I he innleeiil;ir weiulil nil ins nf II () In (( ;ii( )l 11 ;ind |( ;ii( )l h • \ 1 u< <) 11) 11 < ll'('(' 2()()(ii These liictors
17 sel lhe chcniic;illv enmhined w;iler ennleiil In 24 ' pereenl for hmh-c;ilciuni hvdmtcd lime. ;md 2" 2 pereenl Inr
18 dnlnmilie hulniled lime
19 The 2nnf- ll'< '<' i ini,L/iih> ( Tier 2 nielhnd) ;ilsn reeniiiniends ;iccoiiiilum I'nr ennssiniis from lime kiln dnsi i l.kDi
20 ihrnimh ;ipphe;ilinn of ;i enrreelinn liictor l.kl) is ;i hv product of the lime iii;iiiiif;icluriim prneess tvpic;illv nnl
21 reeve led h;iek In kilns I ,kl) is ;i v erv I'liie-unined ni;ileri;il ;md is cspccmllv useful Inr ;ipplie;ilinns rci|iiiriim v erv
22 siikiII p;iriiele si/e Mnsi enmmnii l.kl) ;ipphe;iliniis include soil reel;im;ilinii mid ;mriciilliirc. ( iirrenllv. d;il;i on
23 ;iiiiiii;iI I .kl) prndiielinn is nnl rendilv ;iv ;nl;ihle In dev elnp ;i enimirv specific enrreelinn l';ielnr I .line eniissinn
24 esiiiiKiles were multiplied hv ;i l';iclnrnf I ()2 In ;iccninil I'nr eniissimis from l.kl) (ll'( ( 2()(K>i See I he Hmnicd
25 I niprnv enienis seclinn ;issnci;iled w nil elTnrls in iniprnv e iiiiceri;niilv ;m;ilv sis ;md eniissinn es|ini;iles ;issnci;iled w nil
26 l.kl)
27 I .line eniissinn esiini;iles were liiriher ;id| listed In ;iccniiiil I'nr I he ;inininil nf ('() c;ipiiircd Inr use in nn-siie
28 prncesses \ll ilie dnniesiic lime l;icililies ;ire rei|inred In repnri iliese d;il;i In I P \ under iIs (il l( ikl' The lnl;il
29 n;ilinii;il-lcv el ;iiimi;il ;imniinl nf ('() c;ipiiircd Inr nn-siie prneess use w;is nhliiined I rnni IP Vs (il l(d{lJ (IP \
30 2o l(>) h;ised nil repnried liieihlv lev el d;il;i Inr vein's 2o lo ihrnimh 2d 15 I lie ;ininiiiil nf ('() e;ipiiired reenv ered Inr
31 nn-siie prneess use is dedneled Irnni I he lnl;il pnlenli;il ennssiniis n e . Irnni lime prndiielinn ;md l.kl)) Ilie nel lime
32 ennssiniis ;ire presenled in T;ihlc 4-<> ;md T;ihlc 4-~ (il l( ikf d;il;i on ( () reninv;iIs (i e . CO e;ipiiired reenv ered)
33 w;is ;iv ;nl;ihle oulv I'nr 2o In llirniiuh 2o 15 Smee (il l( ikl'd;il;i ;ire nnl ;iv ;nl;ihle I'nr I990 ihrnimh 2<>(>l>. II'CC
34 "splieinu"' leehini|iies were used ;is per I he 2m if' ll'< '<'' :ni
-------
1 I.line prndiiclimi d;il;i (In l\pe. hmh-c;ilcinni- ;md dnlnnnlic-t|iiickhnic. hmh-c;ilciiini- ;ind dnlnniilic-hulr;ilcd. ;ind
2 dc;id-hiiriicd dnlnnnici Inr Iihrnimh 2ul5 (sec T;ihlc 4-Si were nhijiincd Irnni llic I S (icnlnmc;il Siir\c>
3 (I S(iS)il S( iS 2(> | I ~i ;imiii;il rcpnris ;md ;irc cnmpilcd In I S( iS in llic iicnrcsi inn Vilnrcil
4 lis d mil lie I line, w Inch is produced I'rnni ( ;i() ;md hulmiihc c;ilci n in silic;iles. is iu(>(>) Since d;il;i Inr
8 llic iiidi\ idinl lime l\pes ihiuh cnlcinni ;md dnlnmilio were ik ;il c;ipii\c lime prndnclinii r;icihlics As nnled ;ihn\c. lime h;is ni;in\ different chcniic;il.
21 iiiclnsiri;il. cn\ irniimeiil;il. ;md cniisiriiclinn ;ipplic;ilinns. In iii;in\ processes. ('() rc;icls w illi llic lime In crc;ilc
22 c;ilcinni c;irhnii;ilc (c.u . w;ilcr snl'lciiiimi (';irhnn din\idc rcnhsnrpiinn r;11cs \;ny Imwe\er. dcpciidiim nil llic
23 ;ipphc;ilinn I'nr cviniplc. I nil percent nf llic lime used In prndncc prccipil;ilcd c;ilcinni c;irhnn;ilc re;icls Willi ('() .
24 w licrcns mnsi oil lie lime used in siccl ninkiim rc;icls w illi i mpiiril ics such ;is silic;i. siill'iir. ;md ;il n mi nil in
4-14 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
compounds Oiinulil\ inu I lie nniouul of ('() lh;il is renhsorhed would require ;i dclmlcd nccouuliim of lime use in ilie
I lined Sink's ;iud nddilionnl iiiforninliou nhoul ilie nssocinlcd processes w here holli I he lime ;md h\ product ('() nrc
"reused" nrc required lo quniiiil\ ilie ;iiiku2i lu nccordnuce Willi ll'CC niclhodolomcnl
uuiclelilies. nu\ such emissions nrc cnlculnled In nccouuiiuu lor uel ( Hiincs from chnuucs mi hiouemc C reser\oirs
mi wooded or crop Inuds (see I he l.nud I sc. I.nud-l se Clinuue. nud l-'oresir\ chnpien.
In i he ense of wnler ireninieul plnuis. lime is used iu i he sofiemim process Some In rue wnler ireninieui plnuis nin>
rcco\er ilieir wnsie cnlciuni cnrhounlc nud cnlcuie il iulo c|iii pcrccul flies nlso uoic llinl nddilionnl emissions (nppro\ininlel> 2 perceui i nin> nlso he ueuernled lliroimh
prodiicliou of oilier In products wnsies (off-spec lime llinl is noi recvcled. scruhher sluduc) nl lime plnnis i Seeuer
2o| m I'uhlicK n\nilnhleou I.KI) ueuernliou mies. loinl qunuiiiics uoi used iiicemeui prodiicliou. nud i\pes of oilier
h\ products wnsies produced nl lime fnciluics is linuied IT \ niiiinled n dinlouue willi \l. \ lo discuss dnln needs io
ueuernle n coiuiirs -specific I.KI) I'nclor nud is rc\ lew nm ihe iiilorninlioii pro\ ided In \l. \ \l. \ compiled nnd
slinred hisioricnl eniissious iiiforninliou nud qunuiilies for some wnsie products reported In nieniher fncililies
nssocinled w illi ueuernliou of loinl cnlciued In products nud I .Kl). ns well ns nielhodolous nud cnlciilnliou
worksheets thnl nieniher fncililies complete w lieu reportiim There is iiuceriniuis reunrdum the n\ nilnhilits of dnln
ncross ilie lime series needed lo ueuernle n represeuinine couuirv-specific I.KI) I'nclor I iiceriniuis of ihe ncli\ n\
dnln is nlso n fuucliou of ilie relinhiln\ nud conipleleiiess of \ oluuinriK reported plnui-lc\cl production dnln I 'uriher
resenrch nud dnln is needed lo impro\ e iiiidersinudiim of nddilionnl cnlciunliou emissions io consider rc\ ismu I lie
curreiii nssumpiioiis Mini nrc hnsed on ll'CC guidelines More iiiforninliou cnu he found mi I he Plnuued
I nipro\ enieuis section helow
The results of the \ppronch 2 qunutiinli\ e uuceriniui\ minis sis nrc sunininri/.ed iu I'nhle 4-In I .line CO emissions
for 2o 15 were esiininled lo he helweeu I2l> nud I ^ - \1\1l(() l!q nl I he l>5 perceui confidence lc\ el This
confidence lc\cl nidicnlesn rnuue ofnppro\ininlel> ' perceui helow nud ' perceui nho\e ilie emission esiimnle of
I ^ ^ \l\11 CO l!q
Table 4-10: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Lime
Production (MMT CO2 Eq. and Percent)
2015 l.missiim l!siim;iu- I iui-rl;iiiil> K;iii
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
I.IHUT
lillllll(l
I |)|KT
lilllllHl
I.IHUT
lilllllHl
I |)|KT
liuLMUl
I.ime Production
C( h I
12.9
13.7
-V'/n
+ i%
Range of emission estimates predicted by Monle Carlo Stochastic Simulation lor a 95 percent confidence interval.
\1clhodolomc;il ;ippro;ichcs were ;ipphcd u» I lie enure lime series io ensure cousisteucv in emissions from I'WD
lIiixmiuIi 2d 15 I )el:iils on ilie emission trends throimh lime ;ire described in more del;nl in ilie Melhodolouv seeliou.
;iho\ e
I'or more iuforni;iiiou on ilie ueiler;iI o \ OC proeess applied lo ilns source c;ilcuor\. eousisieni with N'olunie I.
( kipicr<> of ilie li'< '<'' iititU-linc*. see o \ OC ;md \ erifie;iliou I'roeedures seeliou mi ilie iiiirodiieliou of ilie
IH'l ( h;ipier
I pd;iled d;il;i from I.is;i ('outliers (I S (ieolomc;il Sur\e\ i (( 'omihers 2u I ~i resulted in I lmli-( ;ileuini Oiiicklinic
prodiielion d;il;i ekiuues for 2d 14 ;iud I )olonnlic Oiiicklinic production d;il;i ch;iimcs for 2d I i ;md 2d 14. ;is show u mi
Tnhlc 4-X
kcco\ ered emissions show u mi T;iblc 4-~ were upd;iled usuiu ;murcu;ilcd Cil l( ¦ l esiini;iie eniissious from prodiielion of
l.kl) lu response lo Mils iechuic;il nieeinm. mi .l;iiiu;ir\ ;uid l'ehru;ir> 2d|(i. \|. \ compiled ;uid sh;ired lnsioric;il
emissions MiforiiKiiiou reported In nieniher f:icililies on ;iii ;iiiiiu;il h;isis under \oluui;ir\ reporiiuu unti;iii\ es o\er
2uo2 throimh 2d I I ;issoci;iled w ilh ueuer;iliou of tol;il c;ilciued In products ;iud I .Kl) (I .Kl) reporinm oul\
dilfereiili;iled s|;iriiuu mi 2d Id i This emissions iiiform;iliou w;is reported oil ;i \ oluul;ir\ h;isis consisieiil willi \ l. Vs
l;icilit> -lc\ el reportum protocol ;ilso receuil\ pro\ ided IP \ needs ;iddiliou;il lime lo rc\ lew the iiiforni;iiiou
pro\ ided In \l. \ ;uid pl;nis to work w ilh llieni lo ;iddress needs for I P X's ;iu;il\ sis. ;is there is hniiled iiiforni;iliou
;icross the lime series I)uc lo hniiled resources ;uid need lor ;iddiliou;il o \ of uiforni;iiiou. this pkuiued
iniproN enieui is still mi process ;iud h;is not heeu iiicorpomtcd into this curreui lu\eiitors report As ;ui imenm step.
IPX h;is updated the c|ii;ilil;ili\ e description of uuccri;iiiil\ lo reflecl the iufornuitiou pro\ ided In \l. X
In ;iddilion. I !l* X pkius lo ;ilso re\ lew (il l( ikl' emissions ;iud ;icli\ il\ d;il;i reported lo I PA under Suhp;iri S. mi
p;irticiil;ir. rc\ lew of ;mureu;iled ;icli\ il\ d;il;i on lime prodiielion In t\pe IJ:irticuhir ;ilteuliou w ill he m;idc lo ;ilso
eusiiriiiu iinie-series coiisisieucs of ihe emissions esiinuites presented mi fiiiure lu\ euior\ reports, eousisieni w ilh
IK ( ;iud I \l'('('(' uiiidehues This is required ;is ilie l;icihi\-lexel rcporlumd;it;i from l!P X's (il ICiKP. with the
prour;ini's iuiiiiil requirements for rcporiiuu of emissions in c;ilcud;ir\e;ir 2d Id. ;ire not ;i\;ul;ihle for;ill iu\ciiior\
\ c;irs 11 e . I iwu ihrouuh 2dd>Ji ;is required for this 11 in cuiors In i mple nieiiii uu impro\ cine ills ;md iuieur;iiiou of d;it;i
from I P.X's (il l( ikl\ ihe kilest miid;iucc from ilie IK ( on ihe use of l;icihl \-lexel d;il;i mi ii;iiiou;il inxeniories will
he relied upon
15 See.
4-16 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
(il;iss product ion is ;in cuerus ;iud mw -ni;iieri;il iuieusn e process lh;ii results in the ucueniliou of ('() from hoili i lie
cuerus consumed in ni;ikiuu ul;iss;uid ilie d;iss process itself Emissions from fuels consumed forciierus purposes
durum ilie prodiielion of ul;iss ;ire ;ieeouiiied lor in ilie I ineruv seelor
(ihiss producliou eniploss ;i \;iriel> of r;iw m;iieri;ils iii;i ul;iss-h;iieli These include lornieiY Hu\es. si;ihih/ers. ;uid
sonielinies color;iuis The nuijor r;n\ ni;iieri;ils 11 e.. rinses ;uid si;ihih/ersi w Inch emu process-rekiled c;irhou
dioxide (CO i emissions durum ilie ul;iss nielliuu process ;ue liniesioiie. dolomite. ;uid sod;i ;ish The ni;iiu rornier mi
;ill i\ pes of d;iss is sihc;i (Si() i Oilier imjor rorniers in ul;iss include feldspar ;iud hone ;icid lie. hor;i\) I'Iiincs
;ire ;idded lo lower I he leniper;ilure ;il w Inch I he h;ilch mells Most coniniouK used llu\ iu;ileri;i Is ;ire sod;i ;ish
(soduini c;irhon;ile. \;i CO i ;uid pol;ish (pol;issuini c;irhou;ile. K Oi Si;ihih/ers ;ire used lo ni;ike ul;iss more
cheniic;ill\ s|;ihle ;uid lo keep I he finished ukiss Irom dissol\ niu ;uid or l;il h uu ;ip;irt. ('onimouK used s|;ibili/um
;meuls in ul;iss producliou ;ire liniesioiie (C;iCO i. doloniile (C;iCO \lu('() i. ;ilunnii;i I \l O i. ni;muesi;i i\1u()i.
h;iriiini c;irhou;ile (I5;i( () i. siroiiliuni c;irhou;ile (SrCO i. Ill hiuiu c;irhon;ile 11 .i CO i. ;iud /ircoum i/.ri) i (Oil
2(i(>21 (il;iss milkers ;ilsi> use ;i cerl;nii ;uuouui of recvcled scr;ip ul;iss icullel i. w Inch conies from iii-house reluru of
ul;issw;ire broken mi ilie process or oilier ul;iss spilkme or releuiiou such ;is recschuu or cullel broker ser\ ices
The mw ni;iieri;ils (pnni;iril> liniesioiie. doloniile ;uid sod;i ;isln release ('() emissions in ;i coniple\ liiuh-
leniperniiire cheniic;il re;icliou duriim I lie d;iss niellum process This process is noi direclK conip;ir;ihle lo l lie
c;ilciu;iliou process used mi lime ni;iiiuf;icluriim. cenieul ni;iiiiif;icliirum. ;iud process uses of c;irhou;iles (i.e..
liniesioiie doloniile usei. hul h;is ilie s;une uel effect mi lernis of ( () emissions i ll'( ( ' 2uu<>) I lie I S uhiss industry
c;iii he di\ ided nilo four ni;iiu c;ileuorics coiil;iiiicrs. fl;ii i\\ mdow i ul;iss. fiher ul;iss. ;iud speci;ill\ uhiss. The
ni;i|orii\ of coniniercuil ul;iss produced is coiii;uuer ;iud fl;il ukiss i IT \ 2(>ui>). The I in led Si;iies is one of I lie nuijor
uloh;il exporters of uhiss l)oniesiic;ill>. denuuid conies ni;uul> from I lie coiisirucliou. ;iuio. hoillum. ;uid couijiiiier
iiidiisiries There ;ire o\er l.5oo companies ih;ii ni;uiuf;iclure ukiss in I lie I mied Si;iies. with I lie kiruest heiuu
Corinim. (iii;irdi;iu Iiidiisiries. Owens-Illinois. ;uid lJKi Iiidiisiries
lu 2d 15. kilolous of liniesioiie ;uid 2. kilolous of sod:i ;ish were consumed lor ul;iss producliou (I S( iS
2d 15c. Willell 2d I ~) I )olomilc cousiinipiioii d;il;i for ukiss ni;iiiiif;icluriim w;is reporled lo he /ero for 2o 15 I se of
liniesioiie ;uid sod;i ;ish mi ul;iss producliou resulted in ;murcu;ilc ('() emissions oi l ' \I\1T CO I !i|. (1,2l>l> kl) (see
I;ihle 4-1 I) ()\er;ill. emissions h;i\e decreased 15 perceui from I wo ihrouuh 2ol5
Imissions in 2o 15 decreased ;ippro\ini;iiel\ ' perceui from 2o 14 le\els w Inle. mi uenenil. emissions from ul;iss
producliou h;i\e reni;uiied rcl;ili\el> coustjiui o\er I he lime series with some I1iicIii;iiioiis since ll^<> lu ueuenil.
lliese fliiclii;iiious were rekiled lo l lie heh;i\ lor of I lie e\pori ni;irkel ;uid I lie I S ecouoim. Speci fic;i 11\. ilie extended
dowiiiuni mi resideuii;il ;uid coninierci;il coiisirucliou ;uid ;iuiomoii\c iiidiisiries helweeu 2()(>X ;ind 2<>l<) resulied in
reduced coiisumpiiou of ukiss producls. c;iusiim ;i drop in uloh;il demmid for limestone doloniile ;iud sod;i ;ish. ;uid ;i
correspoudiim decrease mi emissions I 'uriherniore. ilie ukiss coui;iiuer sector is one of the le;idiuu sod;i ;ish
coiisunnim sectors in the I mied Suites Some coninierci;il food ;uid hc\ crime p;ick;me ni;iiiiif;iclurers ;ire sin ft i uii
from ul;iss coiit;uuers towards lmhler;uid more cost effecti\e pol\elh\ lene lerephlh;il;ite (HT) h;ised coiit;uuers.
pulliim dow uw;ird pressure oil domestic consumption of sod;i ;ish (I S(iS Iiw5 throimh 2o15c).
Table 4-11: CO2 Emissions from Glass Production (MMT CO2 Eq. and kt)
U-:ir MM l ( (): l'.(|. kl_
1990 I I ^^5
16 Excerpt from Glass & Glass Product Manufacturing Industry Profile, First Research. Available online at:
.
Industrial Processes and Product Use 4-17
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
2005 1.9 1,928
2011 1.3 1.299
2012 1.2 1.248
2013 1.3 1,317
2014 1.3 1.336
2015 1.3 1,299
Nolo: Totals may not sum due to
independent rounding
C;irhou dio\ide emissions were e;ileul;iled h;ised on i lie yitit- li'( V 1 (iuiiLiiiuw Tier ' method In niullipK niu I lie
c | m ; 111111\ of inpiil e;irhon;iles (liniesioiie. dolomite. ;ind sod;i ;isln In I lie e;irhou;ile-h;ised emission f;ielor (in nieli'ie
UHis ('() nieli'ie Ion e;irhon;ile) liniesioiie. n4'l)"l. dolomite. "4 ^2. ;ind sod;i ;ish. D.414^2
( ousu mpi ion d;il;i lor I 'J'Jii lliroimh 2t> I 5 of limesioiie. dolomile. ;md sod;i ;isli used lor ul;iss ni;iiiiif;ieluriim were
ohl;nued from llie I S (ieolome;il Sur\e> il S(iSi \/iihr,i/.\ )l-,ir/'uull < rn>lhi! Sinih-. iimn.il lu /'ori (|iw5 lliroimli
2ol5hi. 2 <) 15 prel i mi ikiia d;il;i Irom i lie I S( iS (rushed Sione ( oniniodils l!\peri (Willell 2< > I-1. i lie ' V/.s
\liihr,il.s )ciirhi»iL Su./n . i.sh . imin.il AV/"*/-/ i |'W5 lliroimh 2t) 151 < I S(iS l'W5 lliroimh 2t>l5e). I S( iS \liner;il
liidiisir\ Siiia e\ s lor Sod;i \sli in .l;iiiu;ir\ 2t)l5(l S( iS 2d 15;n ;ind llie I S I5ure;iu of Mines i liw| ;ind I'N'in.
w lneli ;ire reporied lo llie ne;iresi ion Durum ivwo ;iud ll^2.ihel S( iS did noi eondiiel ;i detailed sur\ e\ of
limesioiie ;md dolomile eoiisiimpiioii In end-use ('oiisiinipiiou lor Iiwt) w;is esiini;iied h\ ;ippl\ iiiu ilie II
pereeuiiiues of iol;il liniesioiie ;iud dolomile use eousiiiuied In llie indix idu;il limesioiie ;iud dolomile uses lo I'wo
loi;il use Siniil;irl\. llie I ^>2 eoiisiinipiiou fmures were ;ippro\ini;iied h\ ;ippl> iiiu ;iu ;i\ er;me of llie I w I ;iud I '
percentiles of ioi;il limesioiie ;iud dolomile use eousiiiuied h\ llie iudi\ idu;il limesioiie ;iud dolomile uses lo I lie
ly): loini
\ddilioii;ill\. e;ich \e;ir llie I S( iS w iililiolds d;it;i oil eei1;nu limesioiie ;iud dok proprietors d;il;i I'orilie purposes of iliis ;iu;il\sis. eniissn e end-uses iluii eoui;iiued
w illilield d;il;i were esinu;iied usiuu one of i lie follow iiiu ieeliuic|iies 111 llie \ ;ilue for ;ill I lie w illilield d;il;i pouils lor
liniesioiie or dolomile use w;is disirihuied c\ eiiI\ lo ;ill w illilield end-uses, or 121 llie ;i\ crime pereeui of ioi;il
limesioiie or diilonnie I'orilie w illilield end-use in i lie preceding ;iud sueeeediuu\e;irs.
I here is ;i l;irue t|ii;iuiii\ of liniesioiie ;iud dolomile reporied lo llie I S( iS under I lie e;ileuories "unspecified
reporied" ;iud "unspecified esiini;iied " \ poriiou of iliis eoiisiinipiiou is lvlie\ ed lo he liniesioiie or dolomile used
for uhiss ni;iuuf;ieluriim I lie i|ii;iulilies lisied under llie "iinspeeified" e;ilei:ories were, llierelore. ;illoe;iled lo uhiss
iii;iiiuf;icluriim ;iccordiuu lo llie pereeui limesioiie or dolomite eoiisiinipiiou for uhiss iii;iiiuf;icluriim end use for lh;il
>e;ir1
I5;ised oil llie 2t) I 5 reporied d;il;i. llie es|ini;iled disirihuliou of sod;i ;ish eoiisiinipiiou lor uhiss prodiieliou compared
lo lol;il donieslie sod;i ;ish eoiisiinipiiou is 4X pereeui (I S( iS I lliroimh 2d 15ei
17 This approach was recommended by USGS.
4-18 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Table 4-12: Limestone, Dolomite, and Soda Ash Consumption Used in Glass Production (kt)
.Uli\ il\
mil
2005
2011
2012
2013
2014
2015
Limestone
430
920
614
> > >
693
765
699
Dolomite
59
541
0
0
0
0
0
Soda Ash
3.177
3.050
2.480
2.420
2.440
2.410
2.390
Tuliil
4.511
3.0'M
2,975
3.133
3,175
3,0N'J
2 unceridiruy anu i imeoeries Lonsisiency
3 The uiiccri;iiul\ lc\cls presented in lliis section ;irisc in p;iri due lo \ ;iri;ilious in llie chcniic;il composition oT
4 limestone used in uhiss production In ;iddiliou lo c;ilciuni c;irbou;ilc. limestone ni;i\ coiilmu siii;illcr miiouuls oT
5 ni;mucsi;i. sihc;i. ;ind sulfur. jinioim oilier niiiicr;ils i pol;issiuni c;irbou;ilc. siroiiliuni c;irbou;ilc ;ind bmiiini c;irbou;ilc.
6 ;ind dc;id burned dolomite) SiniikirK. I lie i|ii;ilil\ ol" llie limestone (;ind mix oTc;irbou;ilcsi used lor ul;iss
7 11i;iiiiil;icl11riiiu w ill depend on ilie i\ pe ol' ul;iss benm nimiiiTiiclurcd
8 The esinmies below ;ilso ;iccoiiul Tor uuccri;iiiil\ ;issoci;ilcd w illi ;icli\ its d;il;i I .;iruc lluclimlioiis in reporied
9 consumption exist. rcUccliim >e;ir-io-\e;ireli;iimes in ilie number oTsur\e\ respouders The iiuccri;iiui\ resiiliinu
10 I'roni ;i shilling sur\e\ population is e\;ieerh;iled In llie mips in ilie lime series of reports The ;iccur;ic> of
11 distribution In end use is ;ilso iiuccrimu lve;nise llns \;ilne is reporied In ilie iii;iinir;ieliirer of llie inpiii e;irhon;iies
12 i limesioiie. dolomiie ;md sod;i ;isln mid noi llie end user Tor 2d 15. ihere li;is heen no reporied eoiisuinpiioii of
13 dolomiie lor ul;iss iii;iiiuT;icluriim These d;il;i h;i\ e been reporied in I S(iS b\ dolomiie iii;iinir;ieliirers ;md nol end-
14 users ii.e . ul;iss inmiiiliieliirersi There is ;i hiuli uiiccri;uui\ ;issoei;iled w illi lliis esiiin;iie. ;is dolomiie is ;i nuijor r;iw
15 m;ileri;i I eonsiimed in uhiss prodiielion \ddiliou;ill\. I lie re is siumfiemil inhereiil uuccri;iiiil\ ;issoei;iled w illi
16 esiiiiKiiinu w iihheld d;il;i ponils Tor specific end uses oT IniiesUiiie ;md dolomite The iiiiccri;uul\ oT llie es|iin;iies Tor
17 limestone mid dolomiie used mi ul;iss ni;ikum is cspcci;ill> lnuli l.;istl\. niueli oT llie limestone consumed 11i llie
18 I lined Suites is reporied ;is "oilier uuspeeil'ied uses;" therelore. it is diTTieuli to ;iccur;iicl> ;illoe;ite this unspecified
19 i|ii;iuiii\ lo llie correct cud-uses T'uriher research is needed mio ;ilteru;iie ;md more complete sources ol'd;il;i ou
20 c;irboii;ilc-b;iscd r;iw ni;iteri;il eoiisuinpiioii b\ llie ul;iss mdiistn "This \e;ir. I !l* \ reiuiti;iled dmlouuc w illi the I S(iS
21 Vitioiinl \ 11ner;iIs luTorm;iiiou Center Crushed Stone coniniodits e\peri lo ;issess the current uuccri;iiiii\ rmmes
22 ;issoci;iled willi c| ii;i lit i lies oT c;irbou;ilcs coiisunied Tor ul;iss production compiled ;uid puhhshed ml S( iS reports
23 The results ol" the \ppro;ich 2 i|ii;inlil;ili\ e iuiccri;iiiil\ ;ui;il> sis ;ire sunini;iri/ed in T;iblc 4-1 ' lu 2d 15. ul;iss
24 prodiielion C() emissions were estinuiied lo he helweeu I 2 ;iud I 4 \ 1 \ 1"T ("<) lx| ;il the '->5 percent confidence
25 lc\ el This iudic;iles ;i r;umc ol";ippro\ini;iiel> 4 perceui helow mid 5 pcrccul ;iho\e llie emission esiini;iie ol" I '
26 \l\TTCO I iq
27 Table 4-13: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Glass
28 Production (MMT CO2 Eq. and Percent)
Si ill I'l l"
Ci;is
2015 llmissiiin Ksiim.iU-
I iui'il;iinl\ Ki-I.iMm- In 1". miss in 11 r!sliiii;ik"'
(MMT CO: Kii.)
(MM 1 ( (): Kd.) ("..)
I.I HUT I |)|KT I.I HUT I |>|KT
liiiund 1 {iiiiihI 1 {iiiiihI ISiimihI
Class Production
I ( )
1.3
1.2 1.4 -4% +5%
Range ol"emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
29 \1clhodolomc;il ;ippro;ichcs were ;ipphed lo llie eulire lime series to ensure coiisisiencv 111 emissions I'roni llwo
30 ihroimh 2d 15 I )el;nls 011 llie emission I rends ihroimh lime ;ire described iu more del;iil 1111 he Melhodolouv secliou.
31 ;ibo\c
32 I'or more luTornuiiiou 011 the ue 1 ler; 11 o \ OC process applied lo llns source c;ileuor\. coiisisieut with Volume I.
33 Ch;ipier (> ol" llie Jwir. ll'< '< 11 ;///,/, 7/'//r.v see o \ n(' mid VcriTic;iliou I'rocedures secliou 111 llie iiiiroducliou ol" ihe
34 ll'l'l Ckipier
Industrial Processes and Product Use 4-19
-------
1
2 I .Milestone ;iud dolomite consumption d:il:i for 2d 14 were rc\ iscd rcl;ili\ c lo I lie prc\ ions 11 in cnu>r\ bused on i lie
3 pivlimiikiia d;il;i obtained dircclK from ilie I S(iS ( rushed Stone ( oniniodils expert. Tisou \\ iNell (\\ illell
4 In llie pie\ kmis In\eiiun"\ n e . Il>l><> throimh 141. prelimiikiia d;il;i were used I'oi'2d|4. which were upd;iled for
5 I lie ciirreiil lii\cutor\ I lie published lime series w;is re\ lewed lo ensure lime-series coiisisiciic> I his upd;ile e;iused
6 ;i dee reuse in 2d 14 emissions of less ih;iu I pereeni compared lo 2d 14 emissions preseuied mi ilie prc\ kmis 11 in ciiioia
7 ne. I wd throimh 2Dl4i
8 riannea improvemenis
9 \s noted mi I lie prc\ kmis re pons, currcul puhl iel\ ;i\;ul;iblc ;icli\ n\ d;il;i shows eoiisunipiioii ol'oiik hniesione ;iud
10 sod;i ;ish lor ul;iss ni;iiiiif;icliirum While hniesione ;iud sod:i ;ish ;ire I lie predominant e;irhou;iles used in ul;iss
11 iii;iiiiif;icliirnm. I lie re ;ire oilier e;irhou;iles lli;il ;ire ;ilso eousunied lor ul;iss ni;iiiuf;icluriim. ;ill IkmiuIi hi siii;i I ler
12 (|ii;iulilies. I :P \ h;is niilKiled re\ lew of ;i\;ul;ihlc ;icli\ its d;il;i oil c;irhou;ile eoiisunipiioii In |\pe mi I he d;iss iudusir\
13 from IPX's (iieeiihouse (us keportnm I'rournni i( il l( ikl'i reported ;imiii;ill\ siuee 2d Id. ;is well ;is I S(iS
14 piihhe;ilious
15 I P \ h;is Minuted rc\ iew of I Ins ;icli\ it\ d;it;i ;iud ;uiticip;ilcs lo fui;ili/e ;issessmeui lor fiiiure uitcurnliou of d;il;i
16 reported under I P. Vs (il l( ¦ |< I* mi the sprum of 2d I" in inipro\ e I he eonipleleuess of emission esiini;iles ;iud
17 f;ieilil;ile c;ilcuor\ -speeifie n( per Volume I of I lie 2nnf- ll'< '<' < iniiL/iih* for I lie (ikiss Production source c;itcuor\.
18 I P Vs (il l( ikl' h;is; 11 i emission ihreshold for rcporiiui:. so i lie ;isscssniciii w ill consider I lie eonipleleuess of
19 c;irhou;ilc eoiisunipiioii d;il;i for uhiss production mi 1 he I nil eel Slnlcs knrticiikir ;illeuiiou will ;ilso he nindc lo ;ilso
20 eiisuriim lime-series coiisisieucs of I lie emissions csiini;iics preseuied mi fiiiure ln\ euior\ reporis. cousisieui w nil
21 I lJ("(' ;iirI I Nl'('('(' uiudehiies I'liis is required ;is i lie l;icihi\ -lex el report um d;il;i from IP Vs (il l( ¦ k I*. w illi I lie
22 prouriun's uiiiiiil re(|iiirenieuis lor reporiiuu of emissions in c;ilcud;ir xe;ir 2d Id. ;ire uoi ;i\;ul;ihlc fornll iii\euior\
23 \e;irs 11 e . Iwo ihrouuh 2dd^i ;is required for ilns 11inciiioia In implement inu inipro\enieuis ;uid uitcurniiou of
24 d;il;i from IP Vs (il l( ikl'. I he l;iles| miidnucc from the ll'('(' on the use of f;icilil> -lex el d;il;i in u;iliou;il i nx eulories
25 will he relied upon |s These pliiuued iniproxenienis;ireouuoiim;uid IP \ m;ix ;ilso 111111;11e research mio oilier
26 sources of ;iclix nx d;il;i for c;irhou;ile eoiisunipiioii h\ I he ul;iss industry
27
28
29
30 I .Milestone <( ;i( () i. dolonule (( ;i('() \lu('() i.1'1 ;ind oilier c;nliou;iles such ;is sod;i ;ish. ninmicsitc. ;uid siderile ;ire
31 h;isic ni;ileri;ils used In ;i wide x;inelx of industries, iiicludiim construction. nuriculturc. chcniic;il. mel;illurux. ulnss
32 production. ;iud eiin iroiimeui;il pollution control This section addresses oiils hniesione ;uid dolomite use for
33 iikIiisIri;iI ;ipphc;ilious. c;irbou;itcs such ;is limesioue ;uid dolomite ;ire honied siifficieutIx eiiouuh lo cnlciuc the
34 ni;iteri;il ;iud uciicmtc ( () ;is ;i In product
35 CuCO, — CuO + CO.
36 MyCO-. - Mi]() + CO.
18 See .
19 Limestone and dolomite are collectively referred to as limestone by the industry, and intermediate varieties are seldom
distinguished.
4-20 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
I A;imples of such ;ipphc;ilious include limestone used ;is ;i l ln\ or purifier in nicl;illurmc;il lurii;iccs. ;is ;i sorhcni in
Hue u;is dcsiill'iiri/;iliou (l '(il )i s\ stems lor ulilils ;iud indnsiri;il pkiuls. ;ind ;is ;i r;i\\ ni;ilcri;il lor I lie production of
ul;iss. 11inc. ;md ccnicul I Emissions from limestone ;ind dolomite used in oilier process seclors such ;is cement. lime.
ul;iss prodiielion. ;md iron ;md slccl. ;ire c\cludcd from llns section ;ind reported under ilieir respccli\c source
c;ilcuorics ic u . Section 4 V (ikiss I'roduclioui I Emission from sod;i ;isli consumption is rcporicd under respecli\ c
c;ilcuorics (c u . (ihiss \1;iiiiil';icliirum (( kl' Source ( ;ileuor\ 2 \') ;iud Sod;i \sli I'roduclioii ;ind ('ousiinipiioii (( kl'
Source C;ilcuoi'\ 215-)) I Emissions from fuels coiisunied forcucrus purposes durum iliese processes ;ire ;iccoiiuied lor
in llie I !nerii\ cluipicr.
I.iniesioiie is w idel> disirihuied iliroimlioui ilie world in deposiis til'\;ir\ nm si/cs ;iud decrees of purii\ I.;irue
deposiis ol' limesione occur m ue;irl\ c\ er\ sinle iu llie I lined Sinles. ;iud siumric;uil t|iinuiiiics ;irc cMrnclcd lor
iudiisiri;il ;ipplic;ilious lu 2d 14. ilie Icndiuu liniesioiie prodiicum sinles ;ire I e\;is. Missouri. I'londn. ()lno. ;uid
kciilucks. w Inch couirihulc 4^ percent of the lotnl I S ouipiilll S(iS l'W5n lliroimh 2d| 5) Siniilnrk. dolomite
deposiis ;ire nlso w idesprend ihroimhoui ihe world I)oloniilc deposiis ;irc found iu the I uiied Sinles. (nundn. \le\ico.
I mi lope. Allien. ;uid Urn/il lu llic I lined Sinles. llic lendum dolonnle prodiiciuu s|;iies ;irc Illinois. IVuuss l\ ;ini;i. ;ind
New York, w Inch contribute 55 perccul of the lotnl 2d 14 I S ouipiilll S(iS l'W5n lliroimh 2dI5i
lu 2dI5. 2^.251 kl ol'liniesioiie ;iud \244 kl ol'doloniiie were coiisunied lor iliese ennssi\e ;ipphc;ilious. c\cludiim
ul;iss niniiiifncluriim i Wi licit 2d I "hi I snuc ol' liniesioiie ;iud dolonnle resulied iu nuurcunlc ('() emissions ol' I 1.2
\1\11 ( () I !c| (I 1.2 '(i kl i (see Tnhle 4-14 ;iud Tnhle 4-15) While 2d 15 emissions h;i\c decreased 5 perceui
compared lo 2d|4. o\er;ill eniissious h;i\e incre;ised I2'J perceni Ironi I1>1>d ihrouuh 2d|5
Table 4-14: CO2 Emissions from Other Process Uses of Carbonates (MMT CO2 Eq.)
W;ir l"lu\ Slum-
k;d
M;ir acid water
treatment. acid neutralization. and sugar
Note: Totals may not sum due lo independent round
Table 4-15: CO2 Emissions from Other Process Uses of Carbonates (kt)
^ e;ir
l"lu\ Slime
ix;d
Mii^iU'siinii
I'l'iiduiiiiiii
Oilier
Misielhiiieiiiis
I SI S 1
Tiil.il
1990
2.592
1,432
64
819
4.907
2005
2.649
2,973
0
718
6,339
201 1
1.467
5.420
0
2.449
9.335
2012
1.077
5.797
0
1.148
8.022
2013
2.307
6.309
0
1.798
10.414
2014
2.91 1
7.1 1 1
0
1.790
1 1.81 1
2015
3,031
7.335
0
871
11.236
;i "()ther miscellaneous uses" include chemical stone, mine dusting or acid water
treatment, acid neutralization, and sugar refining.
Note: Totals may not sum due to independent rounding.
Industrial Processes and Product Use 4-21
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Cnrhou dii<\idc emissions were cnlculnlcd hnsed on the ym/- H'( '<' (iuiiLiiiuw Tier 2 method In niullipK inu I lie
1111;1111il\ of limestone iir dolomite consumed In the emission I'nctor lor limestone or dolomite cnlciiinliou.
rcspccti\cl>. Inblc 2 I limestone o4^~l metric inn ('() metric iiincnrbounlc. ;ind dolomite.<) 4 ^2 metric ton
('() metric Iiin cnrhounlc This niclhodolous wns used liir flu\ sioue. Hue uns dcsiilliiri/nliou s\ skins. j;iiiI use ol' limestone ;ind dolonnic llinl
produced CO emissions Ai ilie end of 2<>t)|. ilie sole ninmicsiiini production plnui opernliuu mi ilie I lined Sinles
llinl produced iiinuuesiuni nielnl usinun dolonuiic process Mini resulted iu llic relense of CO eniissious censed its
opernlious 11 S(iS llJ'J5h Miroimh 2<>I2. I S(iS2<)Pi
('oiisiinipliou dnln for Iiwt) lliroimh 2t> 15 of limestone nud doloniiie used for flu\ sioue. flue uns desiilfiiri/nliou
s\ sienis. c I ic 111 i en I sioue. mine diisiiuu orncid wnler irenimeui. nciil iieuirnli/nliou. nud suunr reluiiiiu (see Tnhle
4-l(>i were ohinuied from die I S. (icolomcnl Sur\c> (I S(iSi \liihru/.\ < rn>lhJ Siniw.iiiiiihi/ lu/'uri
i ll>l>5n Miroimh 2ul5i. preliniiunr\ dnln l'or2n|5 from I S(iS ( rushed Sionc Coniniodils l\peri iWillell 2u|~hi.
\mericnii Iron nud Sieel Iusimile Iimesioiie nud doloniiie consumption dnln i MSI 2t> l<>). nud llic I S. IJurenii of
Mines i llw| nud ll->l^m. which nre reporied lo llic uenresi ion The production cnpncits dnln for I'Wt) Miroimh 2<>I5
of dolonuiic ninuiiesiuni nielnl nlso enme I'roni the I S(iS i l'W5h Miroimh 2ul2. I S(iS 2(>| i) nud the I S IJurenu of
Mines 11lwo throimh IWhi Durum I'wnniul ll>l>2. the I SCSdid not conduct n detniled sur\ e\ of liniestoue nud
doloniiie consumption In end-use ('oiisiinipliou for I'J'Jt) wns estininled In nppKiuuihc II pcrcciilnucs of lotnl
Iimesioiie nud dolomite use constituted h\ the i ml i\ idunl Iimesioiie nud dolomite uses to I lwo totnl use Sum In rl\.
the I ^>2 eoiisiinipiioii I'iuures were nppro.Mninled In nppl> nm nil n\ ernue of the I I nud I W perceutnues of totnl
limestone nud dolomite use constituted In the iudi\ idunl Iimesioiie nud dolomite uses to the Il)l>2 totnl
\ddiliounll>. encli \enr the I S( iS w iihhokls dnln on certniu Iimesioiie nud dolomite end-uses due lo coufidciiiinhis
nureenients reunrdnm conipnm proprielnr\ dntn I 'or llic purposes of this niinls sis. eniissi\ c end-uses Mint coiiiniued
w itliliekl dnln were esiimnled usiuu one of the follow nm iechuii|iies. (I) the \ nine for nil the w itliliekl dnln points for
limestone or doloniiie use wns distributed c\cnl\ lo nil withheld end-uses. (2) the n\crnuc percent oftoinl liniestoue
or doloniiie for the w ilhheld end-use iu llic prcccdum nud succccdiim \ en in. or i ' i llic n\ ernue Irnclioii of lotnl
limestone or doloniiie for llic cud-use o\ cr llic entire time period
There is n In rue i|iinuiii\ of crushed sioue reporied lo llic I S( iS under llic cnlcuorv "unspecified uses " \ poriiou of
lliis eoiisiinipiioii is helie\ ed to he limestone or doloniiie used for ciiiissin c cud uses The i|iiniiiu\ listed for
"unspecified uses" wns. therefore, nllocnled lo encli reporied end-use nccordnm lo encli cud-use's I'rnction of lotnl
eoiisiinipiioii mi Mint \enr
Table 4-16: Limestone and Dolomite Consumption (kt)
Aili\ il\
I'WO
2005
2011
2012
2013
2014
2015
l-'lux Stone
C) 7^7
7.022
4.396
1 (S()(S
6.345
7 *^90
7.834
Limestone
5.804
3.165
2.531
3.108
4.380
4.243
4 ^90
Dolomite
3.857
1.865
S50
1.965
3.356
3 244
l'Cil)
3.258
6.761
12.326
13.185
14.347
16.171
16.680
Oilier Miscellaneon;
; Uses
1.835
1,632
5.548
2,610
i 973
4 069
1.980
lnhil
11.S30
15,415
22,270
24,
-------
1 uncertainty ana i imeoeries Lonsisiency
2 The iiuccn;iiul\ lc\ els presented in this section iiccouul for iiiiccri;iuily ;issoci;ilcd w ilh ;icli\ ily d;il;i l);il;ion
3 linicsiouc ;uid dolomite consumption ;ire collected b\ I S(iS lliroimh \ oluul;ir\ u;iliou;il siiiac> s I S( iS contacts the
4 mimics (i c . producers ii|" \ ;irious |\ pes of crushed stone) for ;iiiiim;iI s;ilcs d;il;i. I );ii;i on other c;irbou;ilc consumption
5 ;ire iml rc;idil> ;i\;iil;iblc The producers report I lie ;iiiiim;iI (|ii;nilil\ sold lo \;irious end-users ;md industrs l\pes.
6 I S( iS estimi;iics ilie historical response r;iie lor ilie erushed stone snr\ e\ lo he ;ippro\im;iiel> pereeiil. ;ind llie resi
7 is esiimined In I S(iS I.;iruc I'luelikiikiiix mi reporied consumption e\isi. rcflccliuu \e;ir-ui-\e;ir ckinucs in ihe
8 number of sur\ c> responders I'lie iiiieei"l;iiiil\ resiilnim from ;i shiftiim siiia e\ population is e\;icerb;iled In llie iz;ips
9 mi ihe lime series of reporis. The ;iccur;ic> of distribution In end use is ;ilso iiiiccrl;iiii hee;iuse lliis \ ;ilue is reporied
10 h\ ihe producer mines ;md noi ihe end user \ddilion;ill\. I lie re is sium fie;i nl inhereiil inieeri;ii ni\ ;issoci;ilcd w ilh
11 esiiiikiiiii'-i w uhheld d;il;i pomis lor speeil ie end uses of hmesioiie ;iud dolomiie I .;istl\. much of ihe hmesioiie
12 eoiisunied iu ihe I lined Scues is reporied ;is "oilier uuspeeified uses." iherefore. il is diffieuli lo ;iccur;itcl> ;illoe;ile
13 llns uuspeeified i|ii;iulil> lo ihe eorreel end-uses This \e;ir. I P \ reuiili;iled dmlouuc w ilh ihe I S( iS \;iiiou;il
14 Minerals luforni;iliou Ceuler ( rushed Sioue eoniniodils e\peri lo ;issess ihe curreiil iiuccri;iiul\ r;umcs ;issoei;iled
15 w ilh ihe hmesioiie ;iud dolomiie consumption d;il;i compiled ;iud published b\ I S( iS I )uriuu llns discussion, ihe
16 e\peri confirnied lluii lil' Vs r;nmc of iiiiccri;uui\ w;is still rc;isou;iblc (W illed 2t)|~;n
17 I iiccri;iint\ mi ihe cs|im;ilcs ;ilso ;irises iu p;ul due lo \ ;iri;ilious mi ihe chemic;il coniposiiKiu of hmesioiie lu
18 ;iddiliou lo c;ilcium c;irhou;ile. Iimesioue ni;i> coul;uu sm;iller ;imouiils of ni;imicsi;i. sihc;i. ;md sulfur. ;inioim oilier
19 niineriils The evict specil ic;ilious for hniesioiie or dolomiie used ;is llu\ sioue \ ;ir\ w ilh ihe p\ ronicl;illuruic;il
20 process ;uid ihe kind of lire processed
21 The results of ihe \ppro;ich 2 i|ii;nilil;ili\ e iiiiccri;iiul> ;111;11\ sis ;ue sunini;ui/ed in T;iblc 4-1 ~ ( ';irboii dio\ide
22 emissions from oilier process uses of c;irhoii;iics iu 2<) 15 were esiini;iied lo he hem ecu ;uid I ' 2 \1\1T ('() I !i|.
23 ;ii I he l>5 pereeiil confidence lex el This indicates ;i nuiue of ;ippro\ini;iiel> I ' perceui below ;ind l(> perceni ;ibo\e
24 ihe emission cs|ini;ilc of I 1.2 \1\1T ( '() I !i|
25 Table 4-17: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Other
26 Process Uses of Carbonates (MMT CO2 Eq. and Percent)
2015 I'.mission
Si ill I'l l'
(iilS
I'.siiniiiU'
I nii'i'hiiim Riinm'Kikilivi'In I'missiun I'slimiili''
I .MM'I'CO: Kii.)
(MM 1 CO:
i«i»
1.1 HUT
l |>|KT
I.I HUT I |>|KT
lii >11 ml
lililllld
Bound Bound
()ther Process 1 Jses
1 1 2
1 i 2
-1 i% +16%
of Carbonates
Range of emission estimates predicted by Monte Curio Stochastic Simulation for a 95 percent confidence interval.
27 \1clhodolomc;il ;ippro;ichcs were applied lo 1 lie eulire nine series ui ensure coiisisiencs 111 eniissioiis from I'Wt)
28 ihroimh 2d 15 I )el;nls 0111 lie emission ireuds ihroimh lime ;ire described 111 more dcl;iil 111 ihe Methodology seclion.
29 ;ibo\e
30 for more iuforni;iiiou 011 ihe ucucnil o \ OC process applied lo llns source c;ilcuor\. cousisieui Willi Volume I.
31 ( h;ipler (> of 1 lie 2mif- li'< '<' < inith liih .s. see OA 0(' ;ind Verification I'rocediires seel 10111111 lie introduction of the
32 ll'l'l ( kiptcr
33 Recalculations Discussion
34 I .Milestone ;iud dokiniite coiisunipiiou d;il;i by cud-use for 2<> 14 were updated rcl;iti\ e lo the pre\ ions 11 in cuiory
35 b;ised 011 the prchnnii;iry d;il;i pro\ ided by I S(iS ( rush Stone ( oniniodily expert. l;isou \\ illell lu ihe pre\ ions
36 Iuxeiiuirx lie.. I'J'Ji) lliroimh 2<) 14). prchniiii;iry d;il;i were used for 2<> 14 w Inch were upd;ilcd for ihe currcui
37 lii\ciitory The published lime series w;is rc\ icwed lo ensure linie-senes consistency This updiilc c;iused ;i decrease
38 iuioi;il limestone ;md dolomite consumption for cnnssi\c cud uses 111 2ul4 by ;ipproxini;itcly 2 perceui. rcl;ili\clo
39 the prc\ ious report
Industrial Processes and Product Use 4-23
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
I'eiidiiiu ;i\ ailahle lesnuiees. llns seelimi \x ill iiilemale and present emissmiis Irnm snda ;ish eniisumpiinii Inr niher
chemical uses ( nmi-ulass prndiiclimi) ( uiTeiilh . in llns dneiimeiil. I hose esii males are presented almm w nil
emissions Irmii snda ;ish prndiiclimi (ll'CC Caleunrs 2I>~) This imprn\email is planned ;md will he implemented
iiilii ihe ne\i In\enU'i"\ repnride. Ii u» 2<>I( 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 [CO(NH2)2], which has a variety of agricultural and industrial applications.
The chemical reaction that produces urea is:
2NH3+ C02 -> NH2COONH4 -> CO(NH2)2 +h2o
4-24 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Only the CO2 emitted directly to the atmosphere from the synthetic ammonia production process is accounted for in
2 determining emissions from ammonia production. The CO2 that is captured during the ammonia production process
3 and used to produce urea does not contribute to the CO2 emission estimates for ammonia production presented in
4 this section. Instead, CO2 emissions resulting from the consumption of urea are attributed to the urea consumption or
5 urea application source category (under the assumption that the carbon stored in the urea during its manufacture is
6 released into the environment during its consumption or application). Emissions of CO2 resulting from agricultural
7 applications of urea are accounted for in the Agriculture chapter. Previously, these emission estimates from the
8 agricultural application of urea were accounted for in the Cropland Remaining Cropland section of the Land Use,
9 Land Use Change, and Forestry chapter. Emissions of CO2 resulting from non-agricultural applications of urea (e.g.,
10 use as a feedstock in chemical production processes) are accounted for in the Urea Consumption for Non-
11 Agricultural Purposes section of this chapter.
12 Total emissions of CO2 from ammonia production in 2016 were 11.2 MMT CO2 Eq. (11,234 kt), and are
13 summarized in Table 4-18 and Table 4-19. Ammonia production relies on natural gas as both a feedstock and a fuel,
14 and as such, market fluctuations and volatility in natural gas prices affect the production of ammonia. Since 1990,
15 emissions from ammonia production have decreased by 14 percent. Emissions in 2016 have increased by
16 approximately 6 percent from the 2015 levels.
17 Table 4-18: CO2 Emissions from Ammonia Production (MMT CO2 Eq.)
Source
1990
2005
2012
2013
2014
2015
2016
Ammonia Production
13.0
9.2 .
9.4
10.0
9.6
10.6
11.2
Total 13.0 9.2 9.4 10.0 9.6 10.6 11.2
18 Table 4-19: CO2 Emissions from Ammonia Production (kt)
Source
1990
2005
2012
2013
2014
2015
2016
Ammonia Production
13,047
9,196
9,377
9,962
9,619
10,571
11,234
Total
13,047
9,196
9,377
9,962
9,619
10,571
11,234
19 Methodology
20 For the United States Inventory, carbon dioxide emissions from production of synthetic ammonia from natural gas
21 feedstock are estimated using a country-specific approach modified from the 2006IPCC Guidelines (IPCC 2006)
22 Tier 1 and 2 methods. In the country-specific approach, emissions are not based on total fuel requirement per the
23 2006 IPCC Guidelines due to data disaggregation limitations of energy statistics provided by the Energy
24 Information Administration (EIA). A country-specific emission factor is developed and applied to national ammonia
25 production to estimate emissions. The method uses a CO2 emission factor published by the European Fertilizer
26 Manufacturers Association (EFMA) that is based on natural gas-based ammonia production technologies that are
27 similar to those employed in the United States. This CO2 emission factor of 1.2 metric tons CCVmetric ton NH3
28 (EFMA 2000a) is applied to the percent of total annual domestic ammonia production from natural gas feedstock.
29 Emissions of CO2 from ammonia production are then adjusted to account for the use of some of the CO2 produced
30 from ammonia production as a raw material in the production of urea. The CO2 emissions reported for ammonia
31 production are reduced by a factor of 0.733 multiplied by total annual domestic urea production. This corresponds to
32 a stoichiometric CCVurea factor of 44/60, assuming complete conversion of ammonia (NH3) and CO2 to urea (IPCC
33 2006; EFMA 2000b).
34 All synthetic ammonia production and subsequent urea production are assumed to be from the same process—
35 conventional catalytic reforming of natural gas feedstock, with the exception of ammonia production from
36 petroleum coke feedstock at one plant located in Kansas. Annual ammonia and urea production are shown in Table
37 4-20. The CO2 emission factor for production of ammonia from petroleum coke is based on plant-specific data,
38 wherein all carbon contained in the petroleum coke feedstock that is not used for urea production is assumed to be
39 emitted to the atmosphere as CO2 (Bark 2004). Ammonia and urea are assumed to be manufactured in the same
40 manufacturing complex, as both the raw materials needed for urea production are produced by the ammonia
Industrial Processes and Product Use 4-25
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
production process. The CO2 emission factor of 3.57 metric tons CCh/metric ton NH3 for the petroleum coke
feedstock process (Bark 2004) is applied to the percent of total annual domestic ammonia production from
petroleum coke feedstock.
The emission factor of 1.2 metric ton CCVmetric ton NH3 for production of ammonia from natural gas feedstock
was taken from the EFMA Best Available Techniques publication, Production of Ammonia (EFMA 2000a). The
EFMA reported an emission factor range of 1.15 to 1.30 metric ton C02/metric ton NH3, with 1.2 metric ton
CCVmetric 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.
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 CO2 from 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.
The total ammonia production data for 2011 through 2015 were obtained from American Chemistry Council (2016).
ACC ammonia production data for 2016 was not yet available and so 2015 data were used as a proxy. For years
before 2011, ammonia production data (see Table 4-20) 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 years 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, 2014, 2015, 2016, and
2017) for 2012 through 2016. 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, and urea production data for 2011 through 2015 were obtained from the Minerals Yearbook:
Nitrogen (USGS 2015, 2016, 2017). USGS urea production data for 2016 was not yet published and so 2015 data
were used as a proxy.
Table 4-20: Ammonia Production and Urea Production (kt)
Year
Ammonia
Production
Urea
Production
1990
15,425
7,450
2005
10.143
5.270
5,220
5,480
5,230
5,540
5,540
2012 10,305
2013 10,930
2014 10,515
2015 11,505
2016 11,505
Uncertainty and Time-Sern insistency
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 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 collocated plants from the same natural gas raw
4-26 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 material. The uncertainty of the total urea production activity data, based on USGS Minerals Yearbook: Nitrogen
2 data, is a function of the reliability of reported production data and is influenced by the completeness of the survey
3 responses. In addition, due to the fact that 2016 nitrogen data has yet to be published, 2015 is used as a proxy which
4 may result in greater uncertainty.
5 Recovery of CO2 from ammonia production plants for purposes other than urea production (e.g., commercial sale,
6 etc.) has not been considered in estimating the CO2 emissions from ammonia production, as data concerning the
7 disposition of recovered CO2 are not available. Such recovery may or may not affect the overall estimate of CO2
8 emissions depending upon the end use to which the recovered CO2 is applied. Further research is required to
9 determine whether byproduct CO2 is being recovered from other ammonia production plants for application to end
10 uses that are not accounted for elsewhere.
11 The icsnlis nf ilie \ppmach 2 qiianinali\e iiiiceriaiiiis aiials sis are snniniai'i/ed 111 I able 4-2 I ( aihuii dio\ide
12 emissions I'min ammonia pmdnclkiii 111 2t> I<¦ were esiinialed lo be heluceii Id ' and 12 I \1\1TCO l!q al ilie l>5
13 pei'ceni confidence le\el This mdieales a ranue of appiiv\inialel> N peiveni helou and X pereeni aho\ e llie emission
14 esiimale of I I 2 \1\11 (() l!q
15 Table 4-21: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from
16 Ammonia Production (MMT CO2 Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY
17 REPORT
Si ill l\v
(¦;is
2016 Emission Ksiim;iu-
IMMK O: i:<|.)
I iHiTl.iiim Ki-hiliu- In Kmissimi I'slimiik-'
(MM 1 ('(): Ku.)
I.I HUT I |)|KT
1$111111(1 1$111111(1
I.I HUT I |>|KT
Bound liiniiid
Ammonia
Prodnc
lion
t ( )
1 1.2
10.3 12.1
-8% +8%
Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
18 Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
19 through 2016. Details on the emission trends through time are described in more detail in the Methodology section,
20 above.
21 For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
22 Chapter 6 of the 2006IPCC Guidelines, see QA/QC and Verification Procedures section in the introduction of the
23 IPPU Chapter.
24 Recalculations Discussion
25 For the previous version of the Inventory (i.e., 1990 through 2015), 2015 urea production data were not published
26 and so 2014 activity data was used as a proxy. Production estimates for urea production for 2015 were updated
27 relative to the previous Inventory using information obtained from the recent 2015 Minerals Yearbook: Nitrogen
28 (USGS 2017). This update resulted in a slight increase of emissions by approximately 6 percent for 2015 relative to
29 the previous Inventory.
30 Planned Improvements
31 Future improvements involve continuing to evaluate and analyze data reported under EPA's GHGRP to improve the
32 emission estimates for the Ammonia Production source category, in particular new data from updated reporting
33 requirements finalized in October of 2014 (79 FR 63750) and December 2016 (81 FR 89188),22 that include facility-
34 level ammonia production data and feedstock consumption. This data will first be reported by facilities in 2018 and
35 available post-verification to assess in early 2019 for use in future reports (e.g. 2020 Inventory report) if the data
36 meets GHGRP CBI aggregation criteria. Particular attention will be made to ensure time-series consistency of the
37 emission estimates presented in future Inventory reports, along with application of appropriate category-specific QC
38 procedures consistent with IPCC and UNFCCC guidelines. For example, data reported in 2018 will reflect activity
22 See .
Industrial Processes and Product Use 4-27
-------
1 in 2017 and may not be representative of activity in prior years of the time series. This assessment is required as the
2 new facility-level reporting data from EPA's GHGRP associated with new requirements is only applicable starting
3 with reporting of emissions in calendar year 2017, and thus is not available for all inventory years (i.e., 1990 through
4 2016) as required for this Inventory.
5 In implementing improvements and integration of data from EPA's GHGRP, the latest guidance from the IPCC on
6 the use of facility-level data in national inventories will be relied upon.23 Specifically, the planned improvements
7 include assessing the anticipated new data to update the emission factors to include both fuel and feedstock CO2
8 emissions to improve consistency with 2006 IPCC Guidelines, in addition to reflecting CO2 capture and storage
9 practices (beyond use of CO2 for urea production). Methodologies will also be updated if additional ammonia
10 production plants are found to use hydrocarbons other than natural gas for ammonia production. Due to limited
11 resources and ongoing data collection effort, this planned improvement is still in development and so is not
12 incorporated into this Inventory.
13 4.6 Urea Consumption for Non-Agricultural
14 Purposes
15 Urea is produced using ammonia and carbon dioxide (CO2) as raw materials. All urea produced in the United States
16 is assumed to be produced at ammonia production facilities where both ammonia and CO2 are generated. There were
17 31 plants producing ammonia in the United States during 2016, with two additional plants sitting idle for the entire
18 year (USGS 2017b).
19 The chemical reaction that produces urea is:
20 2NH3+ C02 -> NH2COONH4 -> CO(NH2)2 + H20
21 This section accounts for CO2 emissions associated with urea consumed exclusively for non-agricultural purposes.
22 Carbon dioxide emissions associated with urea consumed for fertilizer are accounted for in the Agriculture chapter.
23 Urea is used as a nitrogenous fertilizer for agricultural applications and also in a variety of industrial applications.
24 The industrial applications of urea include its use in adhesives, binders, sealants, resins, fillers, analytical reagents,
25 catalysts, intermediates, solvents, dyestuffs, fragrances, deodorizers, flavoring agents, humectants and dehydrating
26 agents, formulation components, monomers, paint and coating additives, photosensitive agents, and surface
27 treatments agents. In addition, urea is used for abating nitrogen oxide (NOx) emissions from coal-fired power plants
28 and diesel transportation motors.
29 Emissions of CO2 from urea consumed for non-agricultural purposes in 2016 were estimated to be 4.0 MMT CO2
30 Eq. (3,959 kt), and are summarized in Table 4-22 and Table 4-23. Net CO2 emissions from urea consumption for
31 non-agricultural purposes in 2016 have increased by approximately 5 percent from 1990 and decreased by
32 approximately 5 percent from 2015. The significant decrease in emissions during 2014 can be attributed to a
33 decrease in the amount of urea imported by the United States during that year.
34 Table 4-22: CO2 Emissions from Urea Consumption for Non-Agricultural Purposes (MMT CO2
35 Eq.)
Source
1990
2005
2012
2013
2014
2015
2016
Urea Consumption
3.8
3.7
4.4
4.1
1.5
4.2
4.0
Total 3.8 3.7 4.4 4.1 1.5 4.2 4.0
23 See .
4-28 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Table 4-23: CO2 Emissions from Urea Consumption for Non-Agricultural Purposes (kt)
Source
1990
2005
2012
2013
2014
2015
2016
Urea Consumption
3,784
C 3,653
4,392
4,074
1,541
4,169
3,959
Total
3,784
3,653
4,392
4,074
1,541
4,169
3,959
Methodology
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, and consistent with the 2006IPCC Guidelines.
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) and is reported in Table 4-24, from the total domestic supply of urea. In previous Inventory reports, the
quantity of urea fertilizer applied to agricultural lands was obtained directly from the Cropland Remaining Cropland
section of the Land Use, Land Use Change, and Forestry chapter. The domestic supply of urea is estimated based on
the amount of urea produced plus the sum of net urea imports and 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 producing urea from ammonia and CO2. This corresponds to a
stoichiometric CCh/urea factor of 44/60, assuming complete conversion of NH3 and CChto urea (IPCC 2006; EFMA
2000).
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 (2011). The
U.S. Census Bureau ceased collection of urea production statistics in 2011, therefore, urea production data from
2011 to 2015 were obtained from the Minerals Yearbook: Nitrogen (USGS 2014 through 2017a). Urea production
data for 2016 are not yet publicly available and so 2015 data (ACC 2015) have been used as proxy.
Urea import data for 2016 are not yet publicly available and so 2015 data have been used as proxy. Urea import data
for 2013 to 2015 were obtained from the Minerals Yearbook: Nitrogen (USGS 2016; USGS 2017a). 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 (TFI2002) 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 4-24).
Urea export data for 2016 are not yet publicly available and so 2015 data have been used as proxy. Urea export data
for 2013 to 2015 were obtained from the Minerals Yearbook: Nitrogen (USGS 2016; USGS 2017a). Urea 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.
Industrial Processes and Product Use 4-29
-------
1 Table 4-24: Urea Production, Urea Applied as Fertilizer, Urea Imports, and Urea Exports (kt)
Year
Urea
Production
Urea Applied
as Fertilizer
Urea
Imports
Urea
Exports
1990
7,450
3,296
1,860
854
2005
5,270
4.779
5.026
536
2012
5,220
5,838
6,944
336
2013
5,480
6,059
6,470
335
2014
5,230
6,188
3,510
451
2015
5,540
6,665
7,190
380
2016
5,540
6,952
7,190
380
2 Uncertainty and Time-Series Consistency
3 There is limited publicly-available data on the quantities of urea produced and consumed for non-agricultural
4 purposes. Therefore, the amount of urea used for non-agricultural purposes is estimated based on a balance that
5 relies on estimates of urea production, urea imports, urea exports, and the amount of urea used as fertilizer. The
6 primary uncertainties associated with this source category are associated with the accuracy of these estimates as well
7 as the fact that each estimate is obtained from a different data source. Because urea production estimates are no
8 longer available from the USGS, there is additional uncertainty associated with urea produced beginning in 2011.
9 There is also uncertainty associated with the assumption that all of the carbon in urea is released into the
10 environment as CO2 during use.
11 The icmiIis of 1 he \ppmadi 2 i|ii;inlil;ili\e iiiiceriaiiiis aiials sis are siininian/ed 111 I able 4-25 ( aihun dio\ide
12 emissions associated u illi urea aiiisiinipiiiin IV»r iioii-auricnliiiral purposes were esiimaled lo he heluceii v5 and 4 4
13 \1\1I 'CO I !c| ;illhc l>5 perceiii confidence le\ el This mdieales a rauue nf apprn\inialcl> 12 pereeui helow and 12
14 pereeiii alxne llie emission esiimale ol'4 0 \l\ITCO l!i|.
15 Table 4-25: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Urea
16 Consumption for Non-Agricultural Purposes (MMT CO2 Eq. and Percent) - TO BE UPDATED
17 FOR FINAL INVENTORY REPORT
Si hi I'l l'
(¦as
2016 llmission I'isiimaU'
(MMT CO: l.(|.)
I ni'iTlaiim kan^i' Ki-laliu- In 1". miss ion Ksiimak"1
(MM 1 ( (): l.ii.) ("..)
1 .ON l'l* I ppiT
1 {oiiiid 1 Sound
I.I HUT I ppi'l'
Bound liound
I Jrea Consiimp
for Non-Agrii
Purposes
lion
uiltural
t ( )
4.0
3.5 4.4
-12% +12%
Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
18 Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
19 through 2016 Methodological approaches were applied to the entire time series to ensure consistency above.
20 For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
21 Chapter 6 of the 2006IPCC Guidelines, see QA/QC and Verification Procedures section in the introduction of the
22 IPPU Chapter.
23 Recalculations Discussion
24 The amount of urea consumed for agricultural purposes (used for calculating urea consumption for non-agricultural
25 purposes) in the United States for the years 1990 through 2016 was revised based on updated urea application
26 estimates obtained from the Agriculture chapter (see Table 5-25). These updates resulted in the following changes to
27 the emission estimates relative to the previous Inventory report: a decrease of less than 1 percent in 2012, an
4-30 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 increase of 1.5 percent in 2013, an increase of 12 percent in 2014, and an increase of 270 percent in 2015. As stated
2 previously in the Methodology section, in previous Inventory reports the quantity of urea fertilizer applied to
3 agricultural lands was obtained directly from the Cropland Remaining Cropland section of the Land Use, Land-Use
4 Change, and Forestry chapter; urea consumption is reported in the Agriculture chapter for the current report.
5 4.7 Nitric Acid Production (CRF Source
6 Category 2B2)
7 Nitrous oxide (N20) is emitted during the production of nitric acid (HNO3), an inorganic compound used primarily
8 to make synthetic commercial fertilizers. It is also a major component in the production of adipic acid—a feedstock
9 for nylon—and explosives. Virtually all of the nitric acid produced in the United States is manufactured by the high-
10 temperature catalytic oxidation of ammonia (EPA 1998). There are two different nitric acid production methods:
11 weak nitric acid and high-strength nitric acid. The first method utilizes oxidation, condensation, and absorption to
12 produce nitric acid at concentrations between 30 and 70 percent nitric acid. High-strength acid (90 percent or greater
13 nitric acid) can be produced from dehydrating, bleaching, condensing, and absorption of the weak nitric acid. The
14 basic process technology for producing nitric acid has not changed significantly over time. Most U.S. plants were
15 built between 1960 and 2000. As of 2016, there were 35 active weak nitric acid production plants, including one
16 high-strength nitric acid production plant in the United States (EPA 2010; EPA 2017).
17 During this reaction, N20 is formed as a byproduct and is released from reactor vents into the atmosphere.
18 Emissions from fuels consumed for energy purposes during the production of nitric acid are accounted for in the
19 Energy chapter.
20 Nitric acid is made from the reaction of ammonia (NH3) with oxygen (O2) in two stages. The overall reaction is:
21 4NH3 + 802 -> 4HNO:i +4H20
22 Currently, the nitric acid industry controls emissions of NO and NO2 (i.e., NOx). As such, the industry in the United
23 States uses a combination of non-selective catalytic reduction (NSCR) and selective catalytic reduction (SCR)
24 technologies. In the process of destroying NOx, NSCR systems are also very effective at destroying N20. However,
25 NSCR units are generally not preferred in modern plants because of high energy costs and associated high gas
26 temperatures. NSCR systems were installed in nitric plants built between 1971 and 1977 with NSCRs installed at
27 approximately one-third of the weak acid production plants. U.S. facilities are using both tertiary (i.e., NSCR) and
28 secondary controls (i.e., alternate catalysts).
29 Nitrous oxide emissions from this source were estimated to be 10.2 MMT CO2 Eq. (34 kt of N20) in 2016 (see
30 Table 4-26). Emissions from nitric acid production have decreased by 16 percent since 1990, with the trend in the
31 time series closely tracking the changes in production. Emissions have decreased by 30 percent since 1997, the
32 highest year of production in the time series.
33 Table 4-26: N2O Emissions from Nitric Acid Production (MMT CO2 Eq. and kt N2O)
Year MMT CO2 Eq. kt N2O
1990 12.1 41
2005 11.3 38
35
36
37
39
34
2012 10.5
2013 10.7
2014 10.9
2015 11.6
2016 10.2
Industrial Processes and Product Use 4-31
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Methodology
Emissions of N20 were calculated using the estimation methods provided by the 2006IPCC Guidelines and country
specific methods from EPA's GHGRP. The 2006IPCC Guidelines Tier 2 method was used to estimate 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 N20 emissions for 2010 through 2016.
2010 through 2016
Process N20 emissions and nitric acid production data were obtained directly from EPA's GHGRP for 2010 through
2016 by aggregating reported facility-level data (EPA 2017). In the United States, all nitric acid facilities producing
weak nitric acid (30 to 70 percent in strength) are required to report annual greenhouse gas emissions data to EPA as
per the requirements of its GHGRP. As of 2016, there were 35 facilities that reported to EPA, including the known
single high-strength nitric acid production facility in the United States (EPA 2017). All nitric acid (weak acid)
facilities are required to calculate process emissions using a site-specific emission factor developed through annual
performance testing under typical operating conditions or by directly measuring N20 emissions using monitoring
equipment.24 The high-strength nitric acid facility also reports N20 emissions associated with weak acid production
and this may capture all relevant emissions, pending additional further EPA research. More details on the
calculation, monitoring and QA/QC methods applicable to nitric acid facilities can be found under Subpart V: Nitric
Acid Production of the regulation, Part 98.25 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. Based on the results of the verification process, the EPA
follows up with facilities to resolve mistakes that may have occurred.26
To calculate emissions from 2010 through 2016, the GHGRP nitric acid production data are utilized to develop
weighted country specific emission factors used to calculate emissions estimates. Based on aggregated nitric acid
production data by abatement type (i.e., with, without) provided by EPA's GHGRP, the percent of production values
and associated emissions of nitric acid with and without abatement technologies are calculated. These percentages
are the basis for developing the country specific weighted emission factors which vary from year to year based on
the amount of nitric acid production with and without abatement technologies.
1990 through 2009
Using GHGRP data for 2010,27 country-specific N20 emission factors were calculated for nitric acid production
with abatement and without abatement (i.e., controlled and uncontrolled emission factors), as previously stated. The
following 2010 emission factors were derived for production with abatement and without abatement: 3.3 kg
N20/metric ton HNO3 produced at plants using abatement technologies (e.g., tertiary systems such as NSCR
systems) and 5.99 kg N20/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 N20 emissions from nitric acid production for years prior to the availability of GHGRP data
(i.e., 1990 through 2008 and 2009). A separate weighted 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
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.
24 Facilities must use standard methods, either EPA Method 320 or ASTM D6348-03 and must follow associated QA/QC
procedures consistent during these performance test consistent with category-specific QC of direct emission measurements.
25 See .
26 See .
27 National N2O 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 2016 (i.e., percent production with and
without abatement).
4-32 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
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 also year over year due to changes in
application of facility-level abatement technologies, maintenance of abatement technologies, and also due to plant
closures and start-ups (EPA 2012, 2013; Desai 2012; CAR 2013). The installation dates of N20 abatement
technologies are not known at most facilities, but it is assumed that facilities reporting abatement technology use
have had this technology installed and operational for the duration of the time series considered in this report
(especially NSCRs).
The country-specific weighted N20 emission factors were used in conjunction with annual production to estimate
N20 emissions for 1990 through 2009, using the following equations:
Ei — Pi X EFwelg)lt:ecl l
where,
EFWeighted,i = [(°/oPc,i X EFc) + (%Punc,L X EFunc)\
E = Annual N20 Emissions for year i (kg/yr)
Pi = Annual nitric acid production for year i (metric tons HNO3)
EF weighted,i = Weighted N20 emission factor for year i (kg N20/metric ton HNO3)
%Pc,i = Percent national production of HNO3 with N20 abatement technology (%)
EFC = N20 emission factor, with abatement technology (kg N20/metric ton HNO3)
%Punc,i = Percent national production of HNO3 without N20 abatement technology (%)
EFunc = N20 emission factor, without abatement technology (kg N20/metric ton HNO3)
i = year from 1990 through 2009
For 2009: Weighted N20 emission factor = 5.46 kg N20/metric ton HNO3.
For 1990 through 2008: Weighted N20 emission factor = 5.66 kg N20/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 4-27). 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). EPA has previously conducted a review of operating
permits to obtain more current information due to the lack of publicly-available data on use of abatement
technologies for 1990 through 2007, as stated previously; therefore, the share of national production with and
without abatement for 2008 was assumed to be constant for 1990 through 2007.
Table 4-27: Nitric Acid Production (kt)
Year kt
1990 7,200
2005 6.710
2012 7,460
2013 7,580
2014 7,660
2015 7,210
2016 7,810
Uncertainty and Time-Series Consistency
Uncertainty associated with the parameters used to estimate N20 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
Industrial Processes and Product Use 4-33
-------
1 (especially prior to 2010) for a variety of facility level variables, including plant specific production levels, plant
2 production technology (e.g., low, high pressure, etc.), and abatement technology type, installation date of abatement
3 technology, and accurate destruction and removal efficiency rates. Production data prior to 2010 were obtained from
4 National Census Bureau, which does not provide uncertainty estimates with their data. Facilities reporting to EPA's
5 GHGRP must measure production using equipment and practices used for accounting purposes. At this time EPA
6 does not estimate uncertainty of the aggregated facility-level information. As noted in the Methodology section,
7 EPA verifies annual facility-level reports through a multi-step process (e.g., combination of electronic checks and
8 manual reviews by staff) to identify potential errors and ensure that data submitted to EPA are accurate, complete,
9 and consistent. The annual production reported by each nitric acid facility under EPA's GHGRP and then
10 aggregated to estimate national N20 emissions is assumed to have low uncertainty.
11 The results olThis \pproach 2 quauiiiali\ e uuceriaiuis aiials sis are sununari/ed in I able 4-2S \iirous o\ide
12 emissions limn uiiiicacid producnm! were esiimaled In he helueeu ') "and Ins \l\l I (() l!q al I lie l>5 perceui
13 confidence le\el This mdicales a rauue of approMinaleK 5 peiveui below lo <> peiveui aho\e llie )I(> emissions
14 esiiniale of Id 2 \1\1T CO l!q.
15 Table 4-28: Approach 2 Quantitative Uncertainty Estimates for N2O Emissions from Nitric
16 Acid Production (MMT CO2 Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY
17 REPORT
Si ill I'l l"
2016 l.missiim l'.slim;ili'
(MM 1 ( (): Kii.)
I iiii'i'hiiiih l{:iii|;r Riliiliu- In Kmissimi l;.sliiii;iu-'
(MM 1 ( (): l u.) ("..)
I.I HUT I |1|KT I.IHUT I |1|KT
lilllllld 1 $111111(1 1 $111111(1 1 $111111(1
Nitric Acid l'n
)duction N:() 10.2
9.7 10.8 -5% +6%
Range of emission estimates predicted by Monte Carlo Stochastic Simulation lor a l)5 percent confidence interval.
18 Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
19 through 2016. To maintain consistency across the time series and with the rounding approaches taken by other data
20 sets, a new rounding approach was performed for the GHGRP Subpart V: Nitric Acid data. This resulted in
21 production data changes across the time series of 2010 to 2016, in which EPA's GHGRP data have been utilized.
22 The results of this update have had an insignificant impact on the emission estimates across the 2010 to 2016 time
23 series. Details on the emission trends through time are described in more detail in the Methodology section, above.
24 For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
25 Chapter 6 of the 2006IPCC Guidelines, see QA/QC and Verification Procedures section in the introduction of the
26 IPPU Chapter.
27 Planned Improvements
28 Pending resources, EPA is considering both near-term and long-term improvement to estimates and associated
29 characterization of uncertainty. In the short-term, with 7 years of EPA's GHGRP data, EPA anticipates completing
30 updates of category-specific QC procedures to potentially also improve both qualitative and quantitative uncertainty
31 estimates. Longer term, in 2020, EPA anticipates having information from EPA's GHGRP facilities on the
32 installation date of any N20 abatement equipment, per recent revisions finalized in December 2016 to EPA's
33 GHGRP. This information will enable more accurate estimation of N20 emissions from nitric acid production over
34 the time series.
35 4.8 Adipic Acid Production (CRF Source
36 Category 2B3)
37 Adipic acid is produced through a two-stage process during which nitrous oxide (N20) is generated in the second
38 stage. Emissions from fuels consumed for energy purposes during the production of adipic acid are accounted for in
4-34 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 the Energy chapter. The first stage of manufacturing usually involves the oxidation of cyclohexane to form a
2 cyclohexanone/cyclohexanol mixture. The second stage involves oxidizing this mixture with nitric acid to produce
3 adipic acid. Nitrous oxide is generated as a byproduct of the nitric acid oxidation stage and is emitted in the waste
4 gas stream (Thiemens and Trogler 1991). The second stage is represented by the following chemical reaction:
5 ('CH2)5CO(cyclohexanone) + (CH2)zCHOH (cyclohexanol) + wHN03
6 -» HOOC(CH2)4COOH(adipic acid) + xN20 + yH20
7 Process emissions from the production of adipic acid vary with the types of technologies and level of emission
8 controls employed by a facility. In 1990, two major adipic acid-producing plants had N20 abatement technologies in
9 place and, as of 1998, three major adipic acid production facilities had control systems in place (Reimer et al. 1999).
10 In 2016, catalytic reduction, non-selective catalytic reduction (NSCR) and thermal reduction abatement technologies
11 were applied as N20 abatement measures at adipic acid facilities (EPA 2017).
12 Worldwide, only a few adipic acid plants exist. The United States, Europe, and China are the major producers, with
13 the United States accounting for the largest share of global adipic acid production capacity in recent years. In 2016,
14 the United States had two companies with a total of two adipic acid production facilities (one in Texas and one in
15 Florida) following the ceased operations of a third major production facility at the end of 2015 (EPA 2017).
16 Adipic acid is a white crystalline solid used in the manufacture of synthetic fibers, plastics, coatings, urethane
17 foams, elastomers, and synthetic lubricants. Commercially, it is the most important of the aliphatic dicarboxylic
18 acids, which are used to manufacture polyesters. Eighty-four percent of all adipic acid produced in the United States
19 is used in the production of nylon 6,6; 9 percent is used in the production of polyester polyols; 4 percent is used in
20 the production of plasticizers; and the remaining 4 percent is accounted for by other uses, including unsaturated
21 polyester resins and food applications (ICIS 2007). Food grade adipic acid is used to provide some foods with a
22 "tangy" flavor (Thiemens and Trogler 1991).
23 National adipic acid production has increased by approximately 40 percent over the period of 1990 through 2016, to
24 approximately 1,055,000 metric tons (ACC 2016). Nitrous oxide emissions from adipic acid production were
25 estimated to be 7.0 MMT CO2 Eq. (23 kt N2O) in 2016 (see Table 4-29). Over the period 1990 through 2016,
26 emissions have been reduced by 54 percent due to both the widespread installation of pollution control measures in
27 the late 1990s and plant idling in the late 2000s. Very little information on annual trends in the activity data exist for
28 adipic acid.
29 Table 4-29: N2O Emissions from Adipic Acid Production (MMT CO2 Eq. and kt N2O)
Year
MMT CO2 Eq.
kt N2O
1990
15.2
51
2005
7.1
24
2012
5.5
19
2013
3.9
13
2014
5.4
18
2015
4.3
14
2016
7.0
23
30 Methodology
31 Emissions are estimated using both Tier 2 and Tier 3 methods consistent with the 2006IPCC Guidelines. Due to
32 confidential business information, plant names are not provided in this section. Therefore, the four adipic acid-
33 producing facilities that have operated over the time series will be referred to as Plants 1 through 4. Overall, as noted
34 above, the two currently operating facilities use catalytic reduction, NSCR and thermal reduction abatement
35 technologies.
Industrial Processes and Product Use 4-35
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
2010 through 2016
All emission estimates for 2010 through 2016 were obtained through analysis of GHGRP data (EPA 2014 through
2017), which is consistent with the 2006IPCC Guidelines Tier 3 method. Facility-level greenhouse gas emissions
data were obtained from EPA's GHGRP for the years 2010 through 2016 (EPA 2014 through 2017) and aggregated
to national N20 emissions. Consistent with IPCC Tier 3 methods, all adipic acid production facilities are required to
calculate emissions using a facility-specific emission factor developed through annual performance testing under
typical operating conditions or by directly measuring N20 emissions using monitoring equipment.28 More
information on the calculation, monitoring and QA/QC methods for process N20 emissions applicable to adipic acid
production facilities under Subpart E can be found in the electronic code of federal regulations.29 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.30
1990 through 2009
For years prior to EPA's GHGRP reporting, for both Plants 1 and 2, 1990 to 2009 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 confidential business information 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 N20 emissions were estimated using the following Tier 2 equation from the 2006
IPCC Guidelines'.
Eaa = Qaa x EFaa X (1 - [DF X UF])
where,
Eaa
N20 emissions from adipic acid production, metric tons
Qaa
Quantity of adipic acid produced, metric tons
EFaa
Emission factor, metric ton N20/metric ton adipic acid produced
DF
N20 destruction factor
UF
Abatement system utility factor
The adipic acid production is multiplied by an emission factor (i.e., N20 emitted per unit of adipic acid produced),
which has been estimated, based on experiments that the reaction stoichiometry for N20 production in the
preparation of adipic acid, to be approximately 0.3 metric tons of N20 per metric ton of product (IPCC 2006). The
"N20 destruction factor" in the equation represents the percentage of N20 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
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 2016; 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
28 Facilities must use standard methods, either EPA Method 320 or ASTM D6348-03, and must follow associated QA/QC
procedures during these performance tests consistent with category-specific QC of direct emission measurements.
29 See .
30 See .
4-36 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 same as 1994 data. The 1997 plant capacities were taken from Chemical Market Reporter, "Chemical Profile:
2 Adipic Acid" (CMR 1998). The 1998 plant capacities for all four plants and 1999 plant capacities for three of the
3 plants were obtained from Chemical Week, Product Focus: Adipic Acid/Adiponitrile (CW 1999). Plant capacities for
4 2000 for three of the plants were updated using Chemical Market Reporter, "Chemical Profile: Adipic Acid" (CMR
5 2001). For 2001 through 2003, the plant capacities for three plants were held constant at year 2000 capacities. Plant
6 capacity for 1999 to 2003 for the one remaining plant was kept the same as 1998.
7 National adipic acid production data (see Table 4-30) from 1990 through 2015 were obtained from the American
8 Chemistry Council (ACC 2016). Updated ACC data for 2016 was not currently available and 2015 data were used
9 as a proxy.
10 Table 4-30: Adipic Acid Production (kt)
Year
kt
1990
755
2005
865
2012
950
2013
980
2014
1,025
2015
1,055
2016
1,055
11 Uncertainty and Time-Series Consistency
12 Uncertainty associated with N20 emission estimates includes the methods used by companies to monitor and
13 estimate emissions. While some information has been obtained through outreach with facilities, limited information
14 is available over the time series on these methods, abatement technology destruction and removal efficiency rates
15 and plant specific production levels.
16 The ivsiilis olTlns \ppronch 2 i|ii;inlil;ili\c iiiiceilami\ ;111;11\ sis are snnininii/cd in Table 4-' I \iirons o\idc
17 emissions from adipie acid production I'm' 2u I(> were esiimaled In he heUxeen <•." and ~ ' \1\1l (() I a| al llic l>5
18 pci'cciil confidence lc\cl. These \nines indicnlc n innue of ;ippi'o\ininlcl> 4 pcrccni heknx In 4 peiceni nho\c 1 he
19 2dl
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
4.9 Caprolactam, Glyoxal and Glyoxylic Acid
Production (CRF Source Category 2B4)
Caprolactam
Caprolactam (CV,Hi iNO) is a colorless monomer produced for nylon-6 fibers and plastics, with a substantial
proportion of the fiber used in carpet manufacturing. Commercial processes for the manufacture of caprolactam are
based on either toluene or benzene. The production of caprolactam can give rise to emissions of nitrous oxide (N20).
During the production of caprolactam, emissions of N20 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 (Reimschuessel 1977).
Where caprolactam is produced from benzene, the main process, the benzene is hydrogenated to cyclohexane
which is then oxidized to produce cyclohexanone (CY,Hh,0). The classical route (Raschig process) and basic
reaction equations for production from cyclohexanone are (Reimschuessel 1977; Lowenheim and Moran 1975;
IPCC 2006):
Oxidation of NH3 to NO/N02
I
NH3 reacted with C02/H20 to yield ammonium carbonate (NH4)2C03
I
(NH4)2C03 reacted with NO/N02 (from NH3 oxidation) to yield ammonium nitrite (NH4N02)
I
NH3 reacted with S02/H20 to yield ammonium bisulphite (NH4HS03)
I
NH4N02 and (NH4HS03) reacted to yield hydroxylamine disulphonate (NOH(S03NH4)2)
I
(NOH(S03NH4)2) hydrolised to yield hydroxylamine sulphate ({NH2OH)2. H2S04) and
ammonium sulphate ((NH4)2S04)
I
Cylohexanone reaction-.
1
C6H10O + ~(NH2OH)2.H2S04(+NH3 and H2S04) -> C6H10NOH + (NH4)2S04 + H20
I
Beckmann rearrangement:
C6H10NOH (+H2S04 and S02) -> C6HuNO.H2S04 (+4NH3 and H20) -> C6HuNO + 2(NH4)2S04
In 1999, there were four caprolactam production facilities in the United States. As of 2016, the United States had 3
companies with a total of 3 caprolactam production facilities: AdvanSix in Virginia (AdvanSix 2017), BASF in
Texas (BASF 2017), and Fibrant LLC in Georgia (Fibrant 2017) (TechSci n.d. 2017).
Nitrous oxide emissions from caprolactam production in the United States were estimated to be 2.0 MMT CO2 Eq.
(7 kt N20) in 2016 (see Table 4-32). National caprolactam production has increased by approximately 21 percent
over the period of 1990 through 2016, to approximately 760 thousand metric tons (ACC 2015). Very little
information on annual trends in the activity data exist for caprolactam.
4-38 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Table 4-32: N2O Emissions from Caprolactam Production (MMT CO2 Eq. and kt N2O)
Year
MMT CO2 Eq.
kt N2O
1990
1.7
6
2005
2.1
"
2012
2.0
7
2013
2.0
7
2014
2.0
7
2015
2.0
7
2016
2.0
7
2
3 Glyoxal
4 Glyoxal is mainly used as a crosslinking agent for acrylic resins, disinfectant, gelatin hardening agent, and textile
5 finishing agent etc. It's produced from oxidation of acetaldehyde with concentrated nitric acid, or from the catalytic
6 oxidation of ethylene glycol, and N20 is emitted in the process of oxidation of acetaldehyde.
7
8 Glyoxal (ethanedial) (C2H2O2) is produced from oxidation of acetaldehyde (ethanal) (C2H4O) with concentrated
9 nitric acid (HNO3). Glyoxal can also be produced from catalytic oxidation of ethylene glycol (ethanediol)
10 (CH2OHCH2OH). Glyoxal is used as a crosslinking agent for vinyl acetate/acrylic resins, disinfectant, gelatin
11 hardening agent, textile finishing agent (permanent-press cotton, rayon fabrics), wet-resistance additive (paper
12 coatings) (Ashford 1994; IPCC 2006).
13
14 Glyoxylic Acid
15 Glyoxylic acid is produced by nitric acid oxidation of glyoxal. Glyoxylic acid is used for the production of synthetic
16 aromas, agrochemicals and pharmaceutical intermediates (Babusiaux 2005).
17 The EPA does not currently estimate the emissions associated with the production of Glyoxal and Glyoxylic Acid
18 due to data availability and a lack of publicly available information on the industry in the United States.
19 Methodology
20 Emissions of N2O were calculated using the estimation methods provided by the 2006 IPCC Guidelines. The 2006
21 IPCC Guidelines Tier 1 method was used to estimate emissions from caprolactam production for 1990 through
22 2016, as shown in this formula:
23 EN20 = EF x CP
24 where,
25 En2o — Annual N2O Emissions (kg)
26 EF = N20 emission factor (default) (kg N20/metric ton caprolactam produced)
27 CP = Caprolactam production (metric tons)
28 During the caprolactam production process, nitrous oxide is generated as a byproduct of the high temperature
29 catalytic oxidation of ammonia (NH3), which is the first reaction in the series of reactions to produce caprolactam.
30 The amount of nitrous oxide emissions can be estimated based on the chemical reaction shown above. Based on this
31 formula, which is consistent with an IPCC Tier 1 approach, approximately 111.1 metric tons of caprolactam are
32 required to generate one metric ton of N20, or an emission factor of 9.0 kg N20 per metric ton of caprolactam (IPCC
33 2006). When applying the Tier 1 method, the 2006IPCC Guidelines state that it is good practice to assume that
34 there is no abatement of N20 emissions and to use the highest default emission factor available in the guidelines. In
35 addition, EPA did not find support for the use of secondary catalysts to reduce N20 emissions, like those employed
36 at nitric acid plants. Thus, the 760 thousand metric tons (kt) of caprolactam produced in 2016 (ACC 2015) resulted
37 in N2O emissions of approximately 2.0 MMT CO2 Eq. (7 kt).
Industrial Processes and Product Use 4-39
-------
1 The activity data for caprolactam production (see Table 4-33) from 1990 to 2015 were obtained from the ACC
2 Guide to the Business of Chemistry report (ACC 2015). For 2016, caprolactam production data was not available
3 and EPA used 2015 production data as proxy. EPA will continue to analyze and assess alternative sources of
4 production data as a quality control measure.
5 Table 4-33: Caprolactam Production (kt)
Year
kt
1990
626
2005
795
2012
750
2013
750
2014
755
2015
760
2016
760
6
7 Carbon dioxide and methane emissions may also occur from the production of caprolactam but currently the IPCC
8 does not have methodologies for calculating these emissions associated with caprolactam production.
9 Uncertainty and Time-Series Consistency
10 Estimation of emissions of N20 from caprolactam production can be treated as analogous to estimation of
11 emissions of N20 from nitric acid production. Both production processes involve an initial step of NH3 oxidation
12 which is the source of N20 formation and emissions (IPCC 2006). Therefore, uncertainties for the default values in
13 the 2006IPCC Guidelines is an estimate based on default values for nitric acid plants. In general, default emission
14 factors for gaseous substances have higher uncertainties because mass values for gaseous substances are influenced
15 by temperature and pressure variations and gases are more easily lost through process leaks. The default values for
16 caprolactam production have a relatively high level of uncertainty due to the limited information available (IPCC
17 2006).
18 The rcsnlis olThis \pproach 2 i|ii;inlil;ili\ c iiiiccriainis aiials sis are summari/cd in I able 4-^4 \iirons o\ide
19 emissions from (aprolaclam. (il\o\al and (il>i'\\ lie \cid hodiiclion lor 2<> l<> were csiinialcd lo he helween I 2
20 and 2 S \I\1T CO I al llie l>5 pereeni confidence lex el. These \ allies mdicalc a mime of appro\imalel> 4 emission esiiniale of 2 u \I\1T ('() I a|
22 Table 4-34: Approach 2 Quantitative Uncertainty Estimates for N2O Emissions from
23 Caprolactam, Glyoxal and Glyoxylic Acid Production (MMT CO2 Eq. and Percent) - TO BE
24 UPDATED FOR FINAL INVENTORY REPORT
Siuinv
(¦us
2016 riiiiissiiui I'.sliniiik'
(MM 1 CO: l.i|.)
I luvriiiiim in I", miss in 11 r'.siiinuk-1
(MM 1 ( (): l.i|.) ("..)
I.I HUT I |1|KT I.IIIUT I |1|KT
liiillllil Bound liiillllil liiillllil
Caprolacta
111 Production
\ ( 1
2.0
1.2 2.8 -40% +40%
Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
25 Details on the emission trends through time are described in more detail in the Methodology section, above.
26 Caprolactam was not reported as a source category in previous Inventory reports (previously reported as "NE") and
27 EPA has taken measures to ensure emission estimates are consistent with 2006 IPCC Guidelines and good practice,
28 ensuring time-series consistency.
4-40 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
2 Chapter 6 of the 2006IPCC Guidelines, see QA/QC and Verification Procedures section in the introduction of the
3 IPPU Chapter.
4 Recalculations Discussion
5 With the addition of caprolactam production emissions reported within the current Inventory report, recalculations
6 have occurred to the IPPU chapter aggregated total emissions estimate. Across the 1990 to 2016 time series, the
7 addition of caprolactam production has added emissions ranging between 1.6 and 2.2 MMT CO2 Eq.
8 Planned Improvements
9 Pending resources, EPA will research other available datasets for caprolactam production and industry trends,
10 including facility-level data. EPA will also research the production process and emissions associated with the
11 production of glyoxal and glyoxylic acid. During the Expert Review comment period for the current Inventory
12 report, EPA sought expert solicitation on data available for these emissions source categories. EPA did not receive
13 information regarding these industries during Expert Review but will continue to research alternative datasets.
14 4.10 Silicon Carbide Production and
15 Consumption (CRF Source Category 2B5)
16 Carbon dioxide (CO2) and methane (CH4) are emitted from the production of silicon carbide (SiC), a material used
17 as an industrial abrasive. Silicon carbide is produced for abrasive, metallurgical, and other non-abrasive applications
18 in the United States. Production for metallurgical and other non-abrasive applications is not available and therefore
19 both CO2 and CH4 estimates are based solely upon production estimates of silicon carbide for abrasive applications.
20 Emissions from fuels consumed for energy purposes during the production of silicon carbide are accounted for in the
21 Energy chapter.
22 Carbon dioxide and CH4 are also emitted during the production of calcium carbide, a chemical used to produce
23 acetylene. Carbon dioxide is implicitly accounted for in the storage factor calculation for the non-energy use of
24 petroleum coke in the Energy chapter. However, data are currently not available to estimate CH4 emissions from this
25 source. See Annex 5 for additional information on sources and sinks not included in this report.
26 To produce SiC, silica sand or quartz (SiCh) is reacted with C in the form of petroleum coke. A portion (about 35
27 percent) of the carbon contained in the petroleum coke is retained in the SiC. The remaining C is emitted as CO2,
28 CH4, or carbon monoxide (CO). The overall reaction is shown below (but in practice it does not proceed according
29 to stoichiometry):
30 Si02 + 3C -> SiC + 2CO (+ 02 -> 2C02)
31 Carbon dioxide is also emitted from the consumption of SiC for metallurgical and other non-abrasive applications.
32 Markets for manufactured abrasives, including SiC, are heavily influenced by activity in the U.S. manufacturing
33 sector, especially in the aerospace, automotive, furniture, housing, and steel manufacturing sectors. The U.S.
34 Geological Survey (USGS) reports that a portion (approximately 50 percent) of SiC is used in metallurgical and
35 other non-abrasive applications, primarily in iron and steel production (USGS 2006a). As a result of the economic
36 downturn in 2008 and 2009, demand for SiC decreased in those years. Low cost imports, particularly from China,
37 combined with high relative operating costs for domestic producers, continue to put downward pressure on the
38 production of SiC in the United States. However, demand for SiC consumption in the United States has recovered
39 somewhat from its low in 2009 (USGS 2012a). Abrasive-grade silicon carbide was manufactured at one facility in
40 2015 in the United States (USGS 2017a).
41 Carbon dioxide emissions from SiC production and consumption in 2016 were 0.2 MMT CO2 Eq. (174 kt CO2) (see
42 Table 4-35 and Table 4-36). Approximately 51 percent of these emissions resulted from SiC production while the
Industrial Processes and Product Use 4-41
-------
1 remainder resulted from SiC consumption. Methane emissions from SiC production in 2016 were 0.01 MMT CO2
2 Eq. (0.4 kt CH4) (see Table 4-35 and Table 4-36). Emissions have not fluctuated greatly in recent years, but 2016
3 emissions are about 52 percent lower than emissions in 1990.
4 Table 4-35: CO2 and ChU Emissions from Silicon Carbide Production and Consumption (MMT
5 COz Eq.)
Year
1990
2005
2012
2013
2014
2015
2016
CO2
CH4
0.4
+
0.2
+
0.2
+
0.2
+
0.2
+
0.2
+
0.2
+
Total
0.4
0.2
0.2
0.2
0.2
0.2
0.2
+ Does not exceed 0.05 MMT CO2 Eq.
6 Table 4-36: CO2 and ChU Emissions from Silicon Carbide Production and Consumption (kt)
Year
1990
2005
2012
2013
2014
2015
2016
CO2
CH4
375
1
219
+
158
+
169
+
173
+
180
+
174
+
+ Does not exceed 0.5 kt.
7 Methodology
8 Emissions of CO2 and CH4 from the production of SiC were calculated31 using the Tier 1 method provided by the
9 2006IPCC Guidelines. Annual estimates of SiC production were multiplied by the appropriate emission factor, as
10 shown below:
11 ESc,C02 = EFsc,co2 X Qsc
/I metric ton\
12 Esc'CH4 = EFsc'CHA X Qsc X { 1000 kg )
13 where,
14 Esc,co2
15 EF sc,co2
16 QSc
17 ESCjch4
18 EFsc,ch4
19
20 Emission factors were taken from the 2006 IPCC Guidelines:
21 • 2.62 metric tons CCVmetric ton SiC
22 • 11.6 kg CH i/mctric ton SiC
23 Emissions of CO2 from silicon carbide consumption for metallurgical uses were calculated by multiplying the
24 annual utilization of SiC for metallurgical uses (reported annually in the USGS Minerals Yearbook: Silicon) by the
25 carbon content of SiC (31.5 percent), which was determined according to the molecular weight ratio of SiC.
26 Emissions of CO2 from silicon carbide consumption for other non-abrasive uses were calculated by multiplying the
27 annual SiC consumption for non-abrasive uses by the carbon content of SiC (31.5 percent). The annual SiC
28 consumption for non-abrasive uses was calculated by multiplying the annual SiC consumption (production plus net
CO2 emissions from production of SiC, metric tons
Emission factor for production of SiC, metric ton CCh/metric ton SiC
Quantity of SiC produced, metric tons
CH4 emissions from production of SiC, metric tons
Emission factor for production of SiC, kilogram CH4/metric ton SiC
31 EPA has not integrated aggregated facility-level GHGRP information to inform these estimates. 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.
4-42 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 imports) by the percent used in metallurgical and other non-abrasive uses (50 percent) (USGS 2006a) and then
2 subtracting the SiC consumption for metallurgical use.
3 The petroleum coke portion of the total CO2 process emissions from silicon carbide production is adjusted for within
4 the Energy chapter, as these fuels were consumed during non-energy related activities. Additional information on
5 the adjustments made within the Energy sector for Non-Energy Use of Fuels is described in both the Methodology
6 section of CO2 from Fossil Fuel Combustion (3.1 Fossil Fuel Combustion (CRF Source Category 1 A)) and Annex
7 2.1, Methodology for Estimating Emissions of CO2 from Fossil Fuel Combustion.
8 Production data for 1990 through 2013 were obtained from the Minerals Yearbook: Manufactured Abrasives (USGS
9 1991a through 2015). Production data for 2014 and 2015 were obtained from the Minerals Industry Surveys:
10 Abrasives (Manufactured) (USGS 2016). Production data for 2016 were obtained from the Mineral Industry
11 Surveys: Abrasives (Manufactured) (USGS 2017b). Silicon carbide production data obtained through the USGS
12 National Minerals Information Center has been previously been rounded to the nearest 5,000 metric tons to avoid
13 disclosing company proprietary data. Silicon carbide consumption by major end use for 1990 through 2015 were
14 obtained from the Minerals Yearbook: Silicon (USGS 1991b through 2017c) (see Table 4-37). 2016 silicon carbide
15 consumption was not yet published by the USGS; therefore, 2015 data are used as a proxy for 2016. Net imports and
16 exports for the entire time series were obtained from the U.S. International Trade Commission (USITC) database
17 updated from data provided by the U.S. Census Bureau (2005 through 2017).
18 Table 4-37: Production and Consumption of Silicon Carbide (Metric Tons)
Year Production Consumption
1000 105.000 172.465
2005 35.000 220.140
2012
35,000
114,265
2013
35,000
134,055
2014
35,000
140,733
2015
35,000
153,475
2016
35,000
142,104
19 Uncertainty and Time-Series Consistency
20 There is uncertainty associated with the emission factors used because they are based on stoichiometry as opposed to
21 monitoring of actual SiC production plants. An alternative would be to calculate emissions based on the quantity of
22 petroleum coke used during the production process rather than on the amount of silicon carbide produced. However,
23 these data were not available. For CH4, there is also uncertainty associated with the hydrogen-containing volatile
24 compounds in the petroleum coke (IPCC 2006). There is also uncertainty associated with the use or destruction of
25 methane generated from the process in addition to uncertainty associated with levels of production, net imports,
26 consumption levels, and the percent of total consumption that is attributed to metallurgical and other non-abrasive
27 uses.
28 The ivsiilis of ilic \pproach 2 (|iianiiiali\ e uiiceriaiiiis ; 111; 11\ sis are suniniaii/ed 111 I able 4-^X. Silicon carhide
29 produclioii and consunipiioii ('() emissions fmm 2d I(> were esiimaled lo he bclweeii l) peiceiii helow and peiceiii
30 alxn e llie emission esiiniale oft) I~ \1\1T( O I !t| al I lie l)5 peiceiii confidence le\ el. Silicon carhide produclion
31 CI I emissions were esiimaled lo he hclweeii l) perceni below and It) perceni aho\e 1 lie emission esiiniale oft) t)l
32 M\11 'CO I !c| al I lie l)5 perceni confidence le\ el
Industrial Processes and Product Use 4-43
-------
1 Table 4-38: Approach 2 Quantitative Uncertainty Estimates for ChU and CO2 Emissions from
2 Silicon Carbide Production and Consumption (MMT CO2 Eq. and Percent) - TO BE UPDATED
3 FOR FINAL INVENTORY REPORT
Si HI I'l l'
C;is
2016 I'lniissiiiii l.slim;ili'
(MMT CO: i:<|.)
I iHiTi.iiim R;mm'Ri-hiiiw in Kmissimi l.siim;iii:'
(MM 1 C (): i:(|.) ("i.)
Li HUT
liiiund
I ppi'i'
Bmind
I.I HUT
1 $111111(1
I ppi'i'
liiilllHl
Silicon Carbide Production
and Consumption
t ( )
0.I7
0.15
0.19
-9%
+9%
Silicon Carbide Production
CI 11
+
+
+
-9%
+ 10%
+ I )oes not exceed 0.05 MMT C(): ] Cq.
11 Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
4 Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
5 through 2016. Details on the emission trends through time are described in more detail in the Methodology section,
6 above.
7 For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
8 Chapter 6 of the 2006IPCC Guidelines, see QA/QC and Verification Procedures section in the introduction of the
9 IPPU Chapter.
10 Recalculations Discussion
11 In the previous Inventory report (i.e., 1990 through 2015), 2015 silicon carbide consumption data by end-use was
12 not available which resulted in the use of 2014 data as proxy. In the current Inventory, advance release data were
13 available for 2015 and the value was updated (USGS 2017c). This recalculation resulted in an insignificant change
14 to the total silicon carbide emissions estimate for the year 2015 compared to the previous Inventory report.
15 4.11 Titanium Dioxide Production (CRF Source
is Category 2B6)
17 Titanium dioxide (TiCh) is manufactured using one of two processes: the chloride process and the sulfate process.
18 The chloride process uses petroleum coke and chlorine as raw materials and emits process-related carbon dioxide
19 (CO2). Emissions from fuels consumed for energy purposes during the production of titanium dioxide are accounted
20 for in the Energy chapter. The chloride process is based on the following chemical reactions:
21 2FeTi03 +7Cl2 +3C -> 2TiCl4 +2FeCl3 +3C02
22 2TiCl4 + 202 -> 2Ti02 +4Cl2
23 The sulfate process does not use petroleum coke or other forms of carbon as a raw material and does not emit CO2.
24 The C in the first chemical reaction is provided by petroleum coke, which is oxidized in the presence of the chlorine
25 and FeTiCb (rutile ore) to form CO2. Since 2004, all TiC>2 produced in the United States has been produced using the
26 chloride process, and a special grade of "calcined" petroleum coke is manufactured specifically for this purpose.
27 The principal use of TiC>2 is as a pigment in white paint, lacquers, and varnishes; it is also used as a pigment in the
28 manufacture of plastics, paper, and other products. In 2016, U.S. TiC>2 production totaled 1,200,000 metric tons
29 (USGS 2017). There were a total five plants producing TiC>2 in the United States in 2016.
4-44 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Emissions of CO2 from titanium dioxide production in 2016 were estimated to be 1.6 MMT CO2 Eq. (1,608 kt CO2),
2 which represents an increase of 35 percent since 1990 (see Table 4-39). Compared to 2015, emissions from titanium
3 dioxide production decreased by 2 percent in 2016 due to a 2 percent decrease in production.
4 Table 4-39: CO2 Emissions from Titanium Dioxide (MMT CO2 Eq. and kt)
Year
MMT CO2 Eq.
kt
1990
1.2
1,195
2005
1.8
1.755
2012
1.5
1,528
2013
1.7
1,715
2014
1.7
1,688
2015
1.6
1,635
2016
1.6
1,608
5 Methodology
6 Emissions of CO2 from TiCh production were calculated by multiplying annual national TiCh production by chloride
7 process-specific emission factors using a Tier 1 approach provided in 2006IPCC Guidelines. The Tier 1 equation is
8 as follows:
9 Etd = EFtd X Qtd
10 where,
11 Etd CO2 emissions from Ti02 production, metric tons
12 EFtd = Emission factor (chloride process), metric ton C02/metric ton Ti02
13 Qtd Quantity of Ti02 produced
14 The petroleum coke portion of the total CO2 process emissions from TiC>2 production is adjusted for within the
15 Energy chapter as these fuels were consumed during non-energy related activities. Additional information on the
16 adjustments made within the Energy sector for Non-Energy Use of Fuels is described in both the Methodology
17 section of CO2 from Fossil Fuel Combustion (3.1 Fossil Fuel Combustion (CRF Source Category 1A)) and Annex
18 2.1, Methodology for Estimating Emissions of CO2 from Fossil Fuel Combustion.
19 Data were obtained for the total amount of TiC>2 produced each year. For years prior to 2004, it was assumed that
20 TiC>2 was produced using the chloride process and the sulfate process in the same ratio as the ratio of the total U.S.
21 production capacity for each process. As of 2004, the last remaining sulfate process plant in the United States
22 closed; therefore, 100 percent of post-2004 production uses the chloride process (USGS 2005b). The percentage of
23 production from the chloride process is estimated at 100 percent since 2004. An emission factor of 1.34 metric tons
24 CCVmetric ton TiC>2 was applied to the estimated chloride-process production (IPCC 2006). It was assumed that all
25 TiC>2 produced using the chloride process was produced using petroleum coke, although some TiC>2 may have been
26 produced with graphite or other carbon inputs.
27 The emission factor for the TiC>2 chloride process was taken from the 2006 IPCC Guidelines. Titanium dioxide
28 production data and the percentage of total TiC>2 production capacity that is chloride process for 1990 through 2013
29 (see Table 4-40) were obtained through the Minerals Yearbook: Titanium Annual Report (USGS 1991 through
30 2015). Production data for 2014 through 2016 were obtained from the Minerals Commodity Summary: Titanium and
31 Titanium Dioxide (USGS 2016; USGS 2017).32 Data on the percentage of total TiC>2 production capacity that is
32 chloride process were not available for 1990 through 1993, so data from the 1994 USGS Minerals Yearbook were
32 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 4-45
-------
1 used for these years. Because a sulfate process plant closed in September 2001, the chloride process percentage for
2 2001 was estimated based on a discussion with Joseph Gambogi (2002). By 2002, only one sulfate process plant
3 remained online in the United States and this plant closed in 2004 (USGS 2005b).
4 Table 4-40: Titanium Dioxide Production (kt)
Year
kt
1990
979
2005
1,310
2012
1,140
2013
1,280
2014
1,260
2015
1,220
2016
1,200
5 Uncertainty and Time-Series Consistency
6 Each year, the U.S. Geological Survey (USGS) collects titanium industry data for titanium mineral and pigment
7 production operations. If TiCh pigment plants do not respond, production from the operations is estimated on the
8 basis of prior year production levels and industry trends. Variability in response rates varies from 67 to 100 percent
9 of Ti02 pigment plants over the time series.
10 Although some TiCh may be produced using graphite or other carbon inputs, information and data regarding these
11 practices were not available. Titanium dioxide produced using graphite inputs, for example, may generate differing
12 amounts of CO2 per unit of TiCh produced as compared to that generated through the use of petroleum coke in
13 production. While the most accurate method to estimate emissions would be to base calculations on the amount of
14 reducing agent used in each process rather than on the amount of TiCh produced, sufficient data were not available
15 to do so.
16 As of 2004, the last remaining sulfate-process plant in the United States closed. Since annual TiCh production was
17 not reported by USGS by the type of production process used (chloride or sulfate) prior to 2004 and only the
18 percentage of total production capacity by process was reported, the percent of total TiCh production capacity that
19 was attributed to the chloride process was multiplied by total TiCh production to estimate the amount of TiCh
20 produced using the chloride process. Finally, the emission factor was applied uniformly to all chloride-process
21 production, and no data were available to account for differences in production efficiency among chloride-process
22 plants. In calculating the amount of petroleum coke consumed in chloride-process TiCh production, literature data
23 were used for petroleum coke composition. Certain grades of petroleum coke are manufactured specifically for use
24 in the TiCh chloride process; however, this composition information was not available.
25 The results oil lie \pproach 2 t|ii;inlil;ili\ e iiuccriniuis ;in;il\sis are suniniari/cd in Table 4-41 Tiinmum dioxide
26 consumption ( () emissions limn 2u I (> were csiinialed In he between I 4and I X \l\l'l CO Lq al llie lJ5 perccui
27 confidence le\el This indicates a rnime of approximate^ 12 percent below and I ' percent abo\e llie emission
28 esiimale of I <> \1\1T CO \ x\
29 Table 4-41: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Titanium
30 Dioxide Production (MMT CO2 Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY
31 REPORT
Si HI I'l l'
(¦as
2016 1"missiiin KslimaU-
IMMK O: i:c|.)
I iHi-iiaiim Kan^c Ri hilhi-1<> l.missiim
(MM 1 ( (): l.t|.) ("¦
llsiinialr1
i>)
I.I HUT I |)|KT 1.1 HUT
lilllllld 1 $111111(1 1 $111111(1
I ppi-r
liiillllll
Titanium Dioxide Prod
uclion
t ( )
1.6
1.4 1.8 -12%
+ 13%
Range of emission estimates predicted by Monte Carlo Stochastic Simulation lor a 95 percent confidence interval.
4-46 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
2 through 2016. Details on the emission trends through time are described in more detail in the Methodology section,
3 above.
4 For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
5 Chapter 6 of the 2006IPCC Guidelines, see QA/QC and Verification Procedures section in the introduction of the
6 IPPU Chapter.
7 Planned Improvements
8 Planned improvements include researching the significance of titanium-slag production in electric furnaces and
9 synthetic-rutile production using the Becher process in the United States. Significant use of these production
10 processes will be included in future Inventory reports. Due to resource constraints, this planned improvement is still
11 in development by EPA and is not included in this report. EPA continues to assess the potential of integrating
12 aggregated facility-level GHGRP information for titanium dioxide production facilities based on criteria to shield
13 underlying CBI from public disclosure. Pending available resources, EPA will also evaluate use of GHGRP data to
14 improve category-specific QC consistent with both Volume 1, Chapter 6 of 2006 IPCC Guidelines and the latest
15 IPCC guidance on the use of facility-level data in national inventories.33
16 4.12 Soda Ash Production (CRF Source
n Category 2B7)
18 Carbon dioxide (CO2) is generated as a byproduct of calcining trona ore to produce soda ash, and is eventually
19 emitted into the atmosphere. In addition, CO2 may also be released when soda ash is consumed. Emissions from
20 soda ash consumption in chemical production processes are reported under Section 4.4 Other Process Uses of
21 Carbonates (IPCC Category 2A4) and emissions from fuels consumed for energy purposes during the production
22 and consumption of soda ash are accounted for in the Energy sector.
23 Calcining involves placing crushed trona ore into a kiln to convert sodium bicarbonate into crude sodium carbonate
24 that will later be filtered into pure soda ash. The emission of CO2 during trona-based production is based on the
25 following reaction:
26 2Na2C03 • NaHC03 • 2H20(Trona) -» 3Na2C03(Soda Ash) + 5H20 + C02
27 Soda ash (sodium carbonate, Na2CC>3) is a white crystalline solid that is readily soluble in water and strongly
28 alkaline. Commercial soda ash is used as a raw material in a variety of industrial processes and in many familiar
29 consumer products such as glass, soap and detergents, paper, textiles, and food. Emissions from soda ash used in
30 glass production are reported under Section 4.3, Glass Production (CRF Source Category 2A3). Glass production is
31 its own source category and historical soda ash consumption figures have been adjusted to reflect this change. After
32 glass manufacturing, soda ash is used primarily to manufacture many sodium-based inorganic chemicals, including
33 sodium bicarbonate, sodium chromates, sodium phosphates, and sodium silicates (USGS 2015b). Internationally,
34 two types of soda ash are produced, natural and synthetic. The United States produces only natural soda ash and is
35 second only to China in total soda ash production. Trona is the principal ore from which natural soda ash is made.
36 The United States represents about one-fourth of total world soda ash output. Only two states produce natural soda
37 ash: Wyoming and California. Of these two states, only net emissions of CO2 from Wyoming were calculated due to
38 specifics regarding the production processes employed in the state.34 Based on 2016 reported data, the estimated
33 See .
34 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 CO2 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
Industrial Processes and Product Use 4-47
-------
1 distribution of soda ash by end-use in 2016 (excluding glass production) was chemical production, 57 percent; soap
2 and detergent manufacturing, 12 percent; distributors, 11 percent; flue gas desulfurization, 8 percent; other uses, 7
3 percent; pulp and paper production, 3 percent, and water treatment, 2 percent (USGS 2017).35
4 U.S. natural soda ash is competitive in world markets because the majority of the world output of soda ash is made
5 synthetically. Although the United States continues to be a major supplier of world soda ash, China, which
6 surpassed the United States in soda ash production in 2003, is the world's leading producer.
7 In 2016, CO2 emissions from the production of soda ash from trona were approximately 1.7 MMT CO2 Eq. (1,723 kt
8 CO2) (see Table 4-42). Total emissions from soda ash production in 2016 increased by approximately 1 percent
9 from emissions in 2015, and have increased by approximately 20 percent from 1990 levels.
10 Emissions have remained relatively constant over the time series with some fluctuations since 1990. In general,
11 these fluctuations were related to the behavior of the export market and the U.S. economy. The U.S. soda ash
12 industry continued a trend of increased production and value in 2016 since experiencing a decline in domestic and
13 export sales caused by adverse global economic conditions in 2009.
14 Table 4-42: CO2 Emissions from Soda Ash Production (MMT CO2 Eq. and kt CO2)
Year MMT CO2 Eq. kt CO2
1990 1.4 1,431
2005 1.7 1.655
2012 1.7 1,665
2013 1.7 1,694
2014 1.7 1,685
2015 1.7 1,714
2016 1.7 1,723
Note: Totals may not sum due to independent rounding.
15 Methodology
16 During the production process, trona ore is calcined in a rotary kiln and chemically transformed into a crude soda
17 ash that requires further processing. Carbon dioxide and water are generated as byproducts of the calcination
18 process. Carbon dioxide emissions from the calcination of trona can be estimated based on the chemical reaction
19 shown above. Based on this formula, which is consistent with an IPCC Tier 1 approach, approximately 10.27 metric
20 tons of trona are required to generate one metric ton of CO2, or an emission factor of 0.0974 metric tons CO2 per
21 metric ton trona (IPCC 2006). Thus, the 17.7 million metric tons of trona mined in 2016 for soda ash production
22 (USGS 2017) resulted in CO2 emissions of approximately 1.7 MMT CO2 Eq. (1,723 kt).
23 Once produced, most soda ash is consumed in chemical production, with minor amounts in soap production, pulp
24 and paper, flue gas desulfurization, and water treatment (excluding soda ash consumption for glass manufacturing).
25 As soda ash is consumed for these purposes, additional CO2 is usually emitted. Consistent with the 2006 IPCC
26 Guidelines for National Greenhouse Gas Inventories, emissions from soda ash consumption in chemical production
27 processes are reported under Section 4.4 Other Process Uses of Carbonates (IPCC Category 2A4).
sodium carbonate. Although CO2 is generated as a byproduct, the CO2 is recovered and recycled for use in the carbonation stage
and is not emitted. A third state, Colorado, produced soda ash until the plant was idled in 2004. The lone producer of sodium
bicarbonate no longer mines trona 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 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.
35 Percentages may not add up to 100 percent due to independent rounding.
4-48 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 The activity data for trona production (see Table 4-43) for 1990 to 2016 were obtained from the U.S. Geological
2 Survey (USGS) Minerals Yearbook for Soda Ash (1994 through 2015b) and USGS Mineral Industry Surveys for
3 Soda Ash (USGS 2017). Soda ash production36 data were collected by the USGS from voluntary surveys of the U.S.
4 soda ash industry. EPA will continue to analyze and assess opportunities to use facility-level data from EPA's
5 GHGRP to improve the emission estimates for Soda Ash Production source category consistent with IPCC37 and
6 UNFCCC guidelines.
7 Table 4-43: Soda Ash Production (kt)
Year
Production3
1990
14,700
2005
17.000
2012
17,100
2013
17,400
2014
17,300
2015
17,600
2016
17,700
a Soda ash produced from trona ore only.
8 Uncertainty and Time-Series Consistency
9 Emission estimates from soda ash production have relatively low associated uncertainty levels in that reliable and
10 accurate data sources are available for the emission factor and activity data for trona-based soda ash production.
11 EPA plans to work with other entities to reassess the uncertainty of these emission factors and activity data based on
12 the most recent information and data. Through EPA's GHGRP, EPA is aware of one facility producing soda ash
13 from a liquid alkaline feedstock process. Soda ash production data was collected by the USGS from voluntary
14 surveys. A survey request was sent to each of the five soda ash producers, all of which responded, representing 100
15 percent of the total production data (USGS 2016). One source of uncertainty is the purity of the trona ore used for
16 manufacturing soda ash. The emission factor used for this estimate assumes the ore is 100 percent pure, and likely
17 overestimates the emissions from soda ash manufacture. The average water-soluble sodium carbonate-bicarbonate
18 content for ore mined in Wyoming ranges from 85.5 to 93.8 percent (USGS 1995).
19 The results nl" I lie \pproaeli 2 qiiaiitiiali\ e iiiiccriaiiits ;iii;iI\ sis are siininiari/ed in Table 4-44 Soda Ash hodiiclion
20 CO emissions I'm- 2<) I (> were estimated In he between I (> and I.XMMTCO liq alllie l>5 pereenl coiilideiice le\ el
21 This indicates a ranue ol appi'o\inialel> ~ pereenl below and <> percent aho\e the emission estimate of I ~ MMT
22 CO l!q
23 Table 4-44: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Soda Ash
24 Production (MMT CO2 Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY REPORT
Sou I'll'
(¦as
2016 rimissiiiii l.sliniiili'
I MMT CO: l!i|.)
I niirlaiim kan^i'ki'laliu'In llniissiiui 1'siimale-1
(MM 1 ( (): i:t|.) C'i.)
I.I HUT I |1|KT I.IHUT I |1|KT
Bound Bound Bound Bound
Soda Ash Production CO: 1.7 1.6 1.8
36 EPA has assessed feasibility of using emissions information (including activity data) from EPA's GHGRP program; however,
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.
37 See .
Industrial Processes and Product Use 4-49
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
;i Range of emission estimates predicted by Monte Curio Stochastic Simulation for a 95 percent confidence interval.
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2016.
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 QA/QC and Verification Procedures section in the introduction of the
IPPU Chapter.
Recalculations Discussion
Emissions from soda ash consumption in chemical production processes were removed from this section and
reported under Section 4.4 Other Process Uses of Carbonates (IPCC Category 2A4) to be consistent with 2006 IPCC
Guidelines. This revision resulted in a decrease in emissions associated with the soda ash production source category
ranging from approximately 39 percent to 49 percent across the time series of 1990 through 2015 compared to the
previous Inventory report.
Planned Improvements
EPA plans to use GHGRP data for conducting category-specific QC of emission estimates consistent with both
Volume 1, Chapter 6 of 2006 IPCC Guidelines and the latest IPCC guidance on the use of facility-level data in
national inventories.38 This planned improvement is ongoing and has not been incorporated into this Inventory
report.
4.13 Petrochemical Production (CRF Source
Category 2B8)
The production of some petrochemicals results in the release of small amounts 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 CRF 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 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 2006
IPCC 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
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 also produces byproduct CO2, carbon monoxide (CO), and water
38 See .
4-50 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
from the direct oxidation of the propylene feedstock, and produces other hydrocarbons from side reactions in the
ammoxidation process.
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 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. Carbon black is also produced in the United States by the thermal cracking of
acetylene-containing feedstocks (i.e., acetylene black process), by the thermal cracking of other hydrocarbons (i.e.,
thermal black process), and by the open burning of carbon black feedstock (i.e., lamp black process); each of these
processes is used at only one U.S. plant (EPA 2000).
Ethylene (C2H4) is consumed in the production processes of the plastics industry including polymers such as high,
low, and linear low density polyethylene (HDPE, LDPE, LLDPE); polyvinyl chloride (PVC); ethylene dichloride;
ethylene oxide; and ethylbenzene. Virtually all ethylene is produced from steam cracking of ethane, propane, butane,
naphtha, gas oil, and other feedstocks. The representative chemical equation for steam cracking of ethane to ethylene
is shown below:
C2H6 -» C2H4 + H2
Small amounts of 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 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 + -02 + 2HCI -» C2H4Cl2 + 2H20 (oxychlorination)
C2H4 + 3 02 —> 2C02 + 2 H20 (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
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, etc.) 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.
Industrial Processes and Product Use 4-51
-------
1 Emissions of CO2 and CH4 from petrochemical production in 2016 were 27.4 MMT CChEq. (27,411 kt CO2) and
2 0.2 MMT CO2 Eq. (7 kt CH4), respectively (see Table 4-45 and Table 4-46). Since 1990, total CO2 emissions from
3 petrochemical production increased by 29 percent. Methane emissions from petrochemical (methanol and
4 acrylonitrile) production have decreased by approximately 18 percent since 1990, given declining production;
5 however, CH4 emissions have been increasing since 2011 due to a rebound in methanol production.
6 Table 4-45: CO2 and ChU Emissions from Petrochemical Production (MMT CO2 Eq.)
Year
1990
2005
2012
2013
2014
2015
2016
CO2
21.2
26.8
26.5
26.4
26.5
28.1
27.4
CH4
0.2
0.1
0.1
0.1
0.1
0.2
0.2
Total
21.4
26.9
26.6
26.5
26.6
28.2
27.6
+ Does not exceed 0.05 MMT CO2 Eq.
Note: Totals may not sum due to independent rounding.
7 Table 4-46: CO2 and ChU Emissions from Petrochemical Production (kt)
Year
1990
2005
2012
2013
2014
2015
2016
CO2
CH4
21,203
9
26,794
3
26,501
3
26,395
3
26,496
5
28,062
7
27,411
7
8 Methodology
9 Emissions of CO2 and CH4 were calculated using the estimation methods provided by the 2006IPCC Guidelines
10 and country-specific methods from EPA's GHGRP. The 2006 IPCC Guidelines Tier 1 method was used to estimate
11 CO2 and CH4 emissions from production of acrylonitrile and methanol,39 and a country-specific approach similar to
12 the IPCC Tier 2 method was used to estimate CO2 emissions from carbon black, ethylene, ethylene oxide, and
13 ethylene dichloride. The Tier 2 method for petrochemicals is a total feedstock C mass balance method used to
14 estimate total CO2 emissions, but is not applicable for estimating CH4 emissions. As noted in the 2006 IPCC
15 Guidelines, the total feedstock C mass balance method (Tier 2) is based on the assumption that all of the C input to
16 the process is converted either into primary and secondary products or into CO2. Further, the guideline states that
17 while the total C mass balance method estimates total C emissions from the process but does not directly provide an
18 estimate of the amount of the total C emissions emitted as CO2, CH4, or non-CH4 volatile organic compounds
19 (NMVOCs). This method accounts for all the C as CO2, including CH4. Note, a subset of facilities reporting under
20 EPA's GHGRP use alternate methods to the C balance approach (e.g., Continuous Emission Monitoring Systems
21 (CEMS) or other engineering approaches) to monitor CO2 emissions and these facilities are required to also report
22 CH4 and N20 emissions from combustion of process off-gas. Preliminary analysis of aggregated annual reports
23 shows that these emissions are less than 500 kt/year. EPA's GHGRP is still reviewing this data across reported years
24 to facilitate update of category-specific QC documentation and EPA plans to address this more completely in future
25 reports.
26 Carbon Black, Ethylene, Ethylene Dichloride and Ethylene Oxide
27 2010 through 2016
28 Carbon dioxide emissions and national production were aggregated directly from EPA's GHGRP dataset for 2010
29 through 2016 (EPA 2017). In 2016, GHGRP data reported CO2 emissions of 3,160,000 metric tons from carbon
30 black production; 19,600,000 metric tons of CChfrom ethylene production; 447,000 metric tons of CChfrom
31 ethylene dichloride production; and 1,100,000 metric tons of CO2 from ethylene oxide production. These emissions
39 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.
4-52 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
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
facilities40 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. More details on the greenhouse gas
calculation, monitoring and QA/QC methods applicable to petrochemical facilities can be found under Subpart X
(Petrochemical Production) of the regulation (40 CFR Part 98).41 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.42
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 CO2 from Fossil Fuel Combustion (3.1 Fossil Fuel Combustion (CRF Source
Category 1 A)) 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 2010 through 2016 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. Carbon dioxide emission factors were derived from EPA's GHGRP data by dividing
annual CO2 emissions for petrochemical type "i" with annual production for petrochemical type "i" and then
averaging the derived emission factors obtained for each calendar year 2010 through 2016. 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 2010 through 2016 GHGRP data
are as follows:
• 2.62 metric tons CCVmetric ton carbon black produced
• 0.77 metric tons CCVmetric ton ethylene produced
• 0.040 metric tons CCVmetric ton ethylene dichloride produced
40 A few facilities producing ethylene dichloride used CO2 CEMS, those CO2 emissions have been included in the aggregated
GHGRP emissions presented here. For ethylene production processes, nearly all process emissions are from the combustion of
process off-gas. Under EPA's GHGRP, Subpart X, ethylene facilities can report CO2 emissions from burning of process gases
using the optional combustion methodology for ethylene production processes, which is requires estimating emissions based on
fuel quantity and carbon contents of the fuel. This is consistent with the 2006 IPCC Guidelines (p. 3.57) which recommends
including combustion emissions from fuels obtained from feedstocks (e.g., off-gases) in petrochemical production under in the
IPPU sector. In 2014, for example, this methodology was used by more than 20 of the 65 reporting facilities. In addition to CO2,
these facilities are required to report emissions of CH4 and N2O from combustion of ethylene process off-gas in 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, as applicable. Preliminary analysis of the aggregated reported CH4 and N2O emissions
from facilities using CEMS and the optional combustion methodology suggests that these annual emissions are less than 500
kt/yr so not significant enough to prioritize for inclusion in the report at this time. Pending resources and significance, EPA may
include these emissions in future reports to enhance completeness.
41 See .
42 See .
Industrial Processes and Product Use 4-53
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
• 0.43 metric tons COVmctric 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 and ethylene dichloride for
1990 through 2009 were obtained from the American Chemistry Council's (ACC's) Guide to the Business of
Chemistry (ACC 2002, 2003, 2005 through 2011). Annual production data for ethylene oxide were obtained from
ACC's U.S. Chemical Industry Statistical Handbook for 2003 through 2009 (ACC 2014a) and from ACC's Business
of Chemistry for 1990 through 2002 (ACC 2014b). As noted above, annual 2010 through 2016 production data for
carbon black, ethylene, ethylene dichloride, and ethylene oxide, were obtained fromEPA's GHGRP.
Acrylonitrile
Carbon dioxide and methane emissions from acrylonitrile production were estimated using the Tier 1 method in the
2006IPCC Guidelines. Annual acrylonitrile production data were used with IPCC default Tier 1 CO2 and CH4
emission factors to estimate emissions for 1990 through 2016. Emission factors used to estimate acrylonitrile
production emissions are as follows:
• 0.18 kg CH4/metric ton acrylonitrile produced
• 1.00 metric tons COVmctric ton acrylonitrile produced
Annual acrylonitrile production data for 1990 through 2015 were obtained from ACC's Business of Chemistry (ACC
2016). In this current report, 2016 ACC production data was not yet available and 2015 data was used as proxy.
Methanol
Carbon dioxide and methane emissions from methanol production were estimated using Tier 1 method in the 2006
IPCC Guidelines. Annual methanol production data were used with IPCC default Tier 1 CO2 and CH4 emission
factors to estimate emissions for 1990 through 2016. Emission factors used to estimate methanol production
emissions are as follows:
• 2.3 kg CH4/metric ton methanol produced
• 0.67 metric tons COVmctric ton methanol produced
Annual methanol production data for 1990 through 2015 were obtained from the ACC's Business of Chemistry
(ACC 2016). As mentioned previously, 2016 ACC production data was not yet available and 2015 data was used as
proxy.
Table 4-47: Production of Selected Petrochemicals (kt)
Chemical
1990
2005
2012
2013
2014
2015
2016
Carbon Black
1,307
1.651
1,280
1,230
1,210
1,220
1,190
Ethylene
16,542
23.975
24,800
25,300
25,500
26,900
26,600
Ethylene Dichloride
6,283
11.260
11,300
11,500
11,300
11,300
11,700
Ethylene Oxide
2,429
3.220
3,110
3,150
3,140
3,240
3,210
Acrylonitrile
1,214
1.325
1,220
1,075
1,095
1,050
1,050
Methanol
3,750
1,225
995
1,235
2,105
3,065
3,065
As noted earlier in the introduction section of the Petrochemical Production chapter, 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 reporting year 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
4-54 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 (81 FR 89260).43 Unfortunately, both of these data elements have not passed GHGRP's CBI aggregation criteria and
2 are thus not available for use in the Inventory at this time. As a result, the United States is currently unable to report
3 non-energy fuel use from petrochemical production under the IPPU chapter. Therefore, consistent with 2006IPCC
4 Guidelines, fuel consumption data reported by EIA are modified to account for these overlaps to avoid double-
5 counting. More information on the non-energy use of fossil fuel feedstocks for petrochemical production can be
6 found in Annex 2.3.
7 Uncertainty and Time-Series Consistency
8 The CH4 and CO2 emission factors used for acrylonitrile and methanol production are based on a limited number of
9 studies. Using plant-specific factors instead of default or average factors could increase the accuracy of the emission
10 estimates; however, such data were not available for the current Inventory report.
11 The results of the quantitative uncertainty analysis for the CO2 emissions from carbon black production, ethylene,
12 ethylene dichloride, and ethylene oxide are based on reported GHGRP data. Refer to the Methodology section for
13 more details on how these emissions were calculated and reported to EPA's GHGRP. There is some uncertainty in
14 the applicability of the average emission factors for each petrochemical type across all prior years. While
15 petrochemical production processes in the United States have not changed significantly since 1990, some
16 operational efficiencies have been implemented at facilities over the time series.
17 The results nl" llie \pproach 2 quaulilali\e iiuccriaiulv aiials sis ;iiv summarized in Table 4-4S. IVtrochcniical
18 production ('() emissions I'roni 2<) l<> were esiimuled in he between 2<> <> mid 2X S \ 1 \1T (() l\q. ill ilie l>5 percent
19 confidence le\el This indicates a rauue ol appro\inialcl> 5 percent below lo 5 pereeui aho\c llie emission esiiniale
20 of 2" 4 \1\1T ('() I x|. IVirochcniical production CM emissions IV0111 2t> l<> were est 1 mated lo he between t> <><1 and
21 11.22 \I\IT ('() I a| at llie l>5 pereeui confidence le\ el This indicates a rauuc ol appio\inialel> 5" percent helow to
22 4(> pereeui aho\e the emission esiiniale ol'u 2 \l\1TCO l!q.
23 Table 4-48: Approach 2 Quantitative Uncertainty Estimates for ChU Emissions from
24 Petrochemical Production and CO2 Emissions from Carbon Black Production (MMT CO2 Eq.
25 and Percent) - TO BE UPDATED FOR FINAL INVENTORY REPORT
Si ill I'l l'
(¦;is
2016 liiiiissiiin
r. si i m:i u-
(MM"I" CO: i:<|. 1
I iiiirl;iiiil\ Ki-hiliu- In 1". miss inn
(MM 1 ( (): i:t|.l
r.slilll;lli''
I.I HUT I |1|KT I.IHUT
lillllllll Bound 1 $111111(1
l |1|KT
lillllllll
Petrochemical
Production
Petrochemical
Production
t ( 1
CI Ii
27.4
0.2
26.0 28.8 -5%
0 Of) 0 22 -^7.
Industrial Processes and Product Use 4-55
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
this category included the QA/QC requirements and verification procedures of EPA's GHGRP. Future QC efforts to
validate the use of Tier 1 default EFs and report on the comparison of Tier 1 emissions estimates and GHGRP data
are described below in the Planned Improvements section.
Planned Improvements
Improvements include completing category-specific QC of activity data and EFs, along with further assessment of
CH4 and N20 emissions to enhance completeness in reporting of emissions from U.S. petrochemical production,
pending resources, significance and time series consistency considerations. As of this current report, timing and
resources have not allowed EPA to complete this analysis of activity data and EFs 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. Some degree of double
counting may occur between CO2 estimates of non-energy use of fuels in the energy sector and CO2 process
emissions from petrochemical production in this sector. As noted previously in the methodology section, data
integration is not feasible at this time as 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. As described in the
methodology section of this source category, EPA is currently unable to use GHGRP reported data on quantities of
fuel consumed as feedstocks by petrochemical producers, only feedstock type, due to the data failing GHGRP CBI
aggregation criteria. Incorporating this data into future inventories will allow for easier data integration between the
non-energy uses of fuels category and the petrochemicals category presented in this chapter. This planned
improvement is still under development and has not been completed to report on progress in this current Inventory.
4.14 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 is scheduled to be phased out by 2020 under the U.S. Clean Air Act.44 Feedstock production, however, is
permitted to continue indefinitely.
HCFC-22 is produced by the reaction of chloroform (CHCI3) 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 (CHCI2F), HCFC-22
(CHCIF2), HFC-23 (CHF3), HC1, 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, HC1 and residual HF. The
HC1 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.
44 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]
4-56 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Two facilities produced HCFC-22 in the United States in 2016. Emissions of HFC-23 from this activity in 2016
were estimated to be 2.8 MMT CO2 Eq. (0.19 kt) (see Table 4-49). This quantity represents a 34 percent decrease
from 2015 emissions and a 94 percent decrease from 1990 emissions. The decrease from 2015 emissions and the
decrease from 1990 emissions were caused primarily by changes in the HFC-23 emission rate (kg HFC-23
emitted/kg 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 factor was higher than that of the other two plants, ceased
production of HCFC-22 in 2013.
Table 4-49: HFC-23 Emissions from HCFC-22 Production (MMT COz Eq. and kt HFC-23)
Year
MMT CChEq.
kt HFC-23
1990
46.1
3
2005
20.0
1
2012
5.5
0.4
2013
4.1
0.3
2014
5.0
0.3
2015
4.3
0.3
2016
2.8
0.2
Methodology
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. Emissions
for 2010 through 2016 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 to estimate their emissions of HFC-23. Plants using thermal oxidation to abate their HFC-
23 emissions monitor the performance of their oxidizers to verily that the HFC-23 is almost completely destroyed.
Plants that release (or historically have released) some of their byproduct HFC-23 periodically measure HFC-23
concentrations in the output stream 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 2016 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 (RTI 1997; RTI2008). 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 4-50.
Industrial Processes and Product Use 4-57
-------
1
Table 4-50: HCFC-22 Production (kt)
Year
kt
1990
139
2005
156
2012
96
2013
C
2014
C
2015
C
2016
C
C (CBI)
Note: HCFC-22 production in 2012 through
2016 is considered Confidential Business
Information (CBI) as there were only two
producers of HCFC-22 in those years.
2 Uncertainty and Time-Series Consistency
3 The uncertainty analysis presented in this section was based on a plant-level Monte Carlo Stochastic Simulation for
4 2006. The Monte Carlo analysis used estimates of the uncertainties in the individual variables in each plant's
5 estimating procedure. This analysis was based on the generation of 10,000 random samples of model inputs from the
6 probability density functions for each input. A normal probability density function was assumed for all
7 measurements and biases except the equipment leak estimates for one plant; a log-normal probability density
8 function was used for this plant's equipment leak estimates. The simulation for 2006 yielded a 95-percent
9 confidence interval for U.S. emissions of 6.8 percent below to 9.6 percent above the reported total.
10 The relative errors yielded by the Monte Carlo Stochastic Simulation for 2006 were applied to the U.S. emission
11 estimate for 2016. The resulting estimates of absolute uncertainty are likely to be reasonably accurate because (1)
12 the methods used by the two remaining plants to estimate their emissions are not believed to have changed
13 significantly since 2006, and (2) although the distribution of emissions among the plants has changed between 2006
14 and 2016 (because one plant has closed), the plant that currently accounts for most emissions had a relative
15 uncertainty in its 2006 (as well as 2005) emissions estimate that was similar to the relative uncertainty for total U.S.
16 emissions. Thus, the closure of one plant is not likely to have a large impact on the uncertainty of the national
17 emission estimate.
18 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-51. HFC-23 emissions
19 from HCFC-22 production were estimated to be between 2.6 and 3.1 MMT CO2 Eq. at the 95 percent confidence
20 level. This indicates a range of approximately 7 percent below and 10 percent above the emission estimate of 2.8
21 MMT C02 Eq.
22 Table 4-51: Approach 2 Quantitative Uncertainty Estimates for HFC-23 Emissions from
23 HCFC-22 Production (MMT CO2 Eq. and Percent)
Source
_ 2016 Emission Estimate
(MMT CO2 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT CO2 Eq.) (%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
HCFC-22 Production
HFC-23 2.8
2.6 3.1
-7% +10%
a Range of emissions reflects a 95 percent confidence interval.
24 QA/QC and Verification
25 Tier 1 and Tier 2 QA/QC activities were conducted consistent with the U.S. QA/QC plan. Source-specific quality
26 control measures for the HCFC-22 Production category included the QA/QC requirements and verification
4-58 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 procedures of EPA's GHGRP. Under EPA's GHGRP, HCFC-22 producers are required to (1) measure
2 concentrations of HFC-23 and HCFC-22 in the product stream at least weekly using equipment and methods (e.g.,
3 gas chromatography) with an accuracy and precision of 5 percent or better at the concentrations of the process
4 samples, (2) measure mass flows of HFC-23 and HCFC-22 at least weekly using measurement devices (e.g.,
5 flowmeters) with an accuracy and precision of 1 percent of full scale or better, (3) calibrate mass measurement
6 devices at the frequency recommended by the manufacturer using traceable standards and suitable methods
7 published by a consensus standards organization, (4) calibrate gas chromatographs at least monthly through analysis
8 of certified standards, and (5) document these calibrations.
9 EPA verifies annual facility-level reports from HCFC-22 producers through a multi-step process (e.g., a
10 combination of electronic checks and manual reviews by staff) to identify potential errors and ensure that data
11 submitted to EPA are accurate, complete, and consistent. Based on the results of the verification process, EPA
12 follows up with facilities to resolve mistakes that may have occurred.45
n 4.15 Carbon Dioxide Consumption (CRF Source
14 Category 2610}
15 Carbon dioxide (CO2) is used for a variety of commercial applications, including food processing, chemical
16 production, carbonated beverage production, and refrigeration, and is also used in petroleum production for
17 enhanced oil recovery (EOR). Carbon dioxide used for EOR is injected underground to enable additional petroleum
18 to be produced. For the purposes of this analysis, CO2 used in commercial applications other than EOR is assumed
19 to be emitted to the atmosphere. Carbon dioxide used in EOR applications is discussed in the Energy chapter under
20 "Carbon Capture and Storage, including Enhanced Oil Recovery" and is not discussed in this section.
21 Carbon dioxide is produced from naturally-occurring CO2 reservoirs, as a byproduct from the energy and industrial
22 production processes (e.g., ammonia production, fossil fuel combustion, ethanol production), and as a byproduct
23 from the production of crude oil and natural gas, which contain naturally occurring CO2 as a component. Only CO2
24 produced from naturally occurring CO2 reservoirs, and as a byproduct from energy and industrial processes, and
25 used in industrial applications other than EOR is included in this analysis. Carbon dioxide captured from biogenic
26 sources (e.g., ethanol production plants) is not included in the Inventory. Carbon dioxide captured from crude oil
27 and gas production is used in EOR applications and is therefore reported in the Energy chapter.
28 Carbon dioxide is produced as a byproduct of crude oil and natural gas production. This CO2 is separated from the
29 crude oil and natural gas using gas processing equipment, and may be emitted directly to the atmosphere, or
30 captured and reinjected into underground formations, used for EOR, or sold for other commercial uses. A further
31 discussion of CO2 used in EOR is described in the Energy chapter in Box 3-7 titled "Carbon Dioxide Transport,
32 Injection, and Geological Storage."
33 In 2016, the amount of CO2 produced and captured for commercial applications and subsequently emitted to the
34 atmosphere was 4.5 MMT C02Eq. (4,471 kt) (see Table 4-52). This is consistent with 2014 and 2015 levels and is
35 an increase of approximately 204 percent since 1990.
36 Table 4-52: CO2 Emissions from CO2 Consumption (MMT CO2 Eq. and kt)
Year MMT CO2 Eq.
kt
1990
1.5
1,472
2005
1.4
1.37 5
2012
2013
4.0
4.2
4,019
4,188
45 See .
Industrial Processes and Product Use 4-59
-------
Year MMT CO2 Eg.
kt
2014
2015
2016
4.5
4.5
4.5
4,471
4,471
4,471
1 Methodology
2 Carbon dioxide emission estimates for 1990 through 2016 were based on the quantity of CO2 extracted and
3 transferred for industrial applications (i.e., non-EOR end-uses). Some of the CO2 produced by these facilities is used
4 for EOR and some is used in other commercial applications (e.g., chemical manufacturing, food production). It is
5 assumed that 100 percent of the CO2 production used in commercial applications other than EOR is eventually
6 released into the atmosphere.
8 For 2010 through 2014, data from EPA's GHGRP (Subpart PP) were aggregated from facility-level reports to
9 develop a national-level estimate for use in the Inventory (EPA 2016). Facilities report CO2 extracted or produced
10 from natural reservoirs and industrial sites, and CO2 captured from energy and industrial processes and transferred to
11 various end-use applications to EPA's GHGRP. This analysis includes only reported CO2 transferred to food and
12 beverage end-uses. EPA is continuing to analyze and assess integration of CO2 transferred to other end-uses to
13 enhance the completeness of estimates under this source category. Other end-uses include industrial applications,
14 such as metal fabrication. EPA is analyzing the information reported to ensure that other end-use data excludes non-
15 emissive applications and publication will not reveal confidential business information (CBI). Reporters subject to
16 EPA's GHGRP Subpart PP are also required to report the quantity of CO2 that is imported and/or exported.
17 Currently, these data are not publicly available through the GHGRP due to data confidentiality reasons and hence
18 are excluded from this analysis.
19 Facilities subject to Subpart PP of EPA's GHGRP are required to measure CO2 extracted or produced. More details
20 on the calculation and monitoring methods applicable to extraction and production facilities can be found under
21 Subpart PP: Suppliers of Carbon Dioxide of the regulation, Part 98.46 The number of facilities that reported data to
22 EPA's GHGRP Subpart PP (Suppliers of Carbon Dioxide) for 2010 through 2014 is much higher (ranging from 44
23 to 48) than the number of facilities included in the Inventory for the 1990 to 2009 time period prior to the
24 availability of GHGRP data (4 facilities). The difference is largely due to the fact the 1990 to 2009 data includes
25 only CO2 transferred to end-use applications from naturally occurring CO2 reservoirs and excludes industrial sites.
26 Starting in 2015, data from EPA's GHGRP (Subpart PP) was unavailable for use in the current Inventory report due
27 to data confidentiality reasons. As a result, the emissions estimates for 2015 and 2016 have been held constant from
28 2014 levels to avoid disclosure of proprietary information. EPA will continue to evaluate options for utilizing
29 GHGRP data to update these values in future inventories.
31 For 1990 through 2009, data from EPA's GHGRP are not available. For this time period, CO2 production data from
32 four naturally-occurring CO2 reservoirs were used to estimate annual CO2 emissions. These facilities were Jackson
33 Dome in Mississippi, Brave and West Bravo Domes in New Mexico, and McCallum Dome in Colorado. The
34 facilities in Mississippi and New Mexico produced CO2 for use in both EOR and in other commercial applications
35 (e.g., chemical manufacturing, food production). The fourth facility in Colorado (McCallum Dome) produced CO2
36 for commercial applications only (New Mexico Bureau of Geology and Mineral Resources 2006).
37 Carbon dioxide production data and the percentage of production that was used for non-EOR applications for the
38 Jackson Dome, Mississippi facility were obtained from Advanced Resources International (ARI 2006, 2007) for
46 See .
7 2010 through 2016
30 1990 through 2009
4-60 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
1990 to 2000, and from the Annual Reports of Denbury Resources (Denbury Resources 2002 through 2010) for
2001 to 2009 (see Table 4-53). 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 (ARI 1990 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 4-53: CO2 Production (kt CO2) and the Percent Used for Non-EOR Applications
Year
Jackson Dome,
Bravo Dome,
West Bravo
McCallum
Total CO2
%
MS
NM
Dome, NM CO2
Dome, CO
Production
Non-
CO2 Production
CO2 Production
Production
CO2 Production
from Extraction
EOR3
(kt) (% Non-
(kt) (% Non-
(kt) (% Non-
(kt) (% Non-
and Capture
EOR)
EOR)
EOR)
EOR)
Facilities (kt)
1990
1,344(100%)
63 (1%)
+
65 (100%)
NA
NA
2005
1.254 (27%)
58(1%)
63 (100%)
NA
NA
2012
NA
NA
NA
NA
66,326
6%
2013
NA
NA
NA
NA
68,435
6%
2014
NA
NA
NA
NA
72,000
6%
2015
NA
NA
NA
NA
72,000
6%
2016
NA
NA
NA
NA
72,000
6%
+ Does not exceed 0.5 percent.
NA (Not available) - For 2010 through 2014, the publicly available GHGRP data were aggregated at the national level. For
2015 and 2016, values were held constant with those from 2014. Facility-level data are not publicly available from EPA's
GHGRP.
a Includes only food & beverage applications.
Uncertainty and Time-Series Consistency
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 a threshold for reporting
under GHGRP applicable to those reporting under Subpart PP, in addition to the exclusion of the amount of CO2
transferred to all other end-use categories. This latter category might include CO2 quantities that are being used for
non-EOR industrial applications such as firefighting. Second, uncertainty is associated with the exclusion of
imports/exports data for CO2 suppliers. Currently these data are not publicly available through EPA's GHGRP and
hence are excluded from this analysis. EPA verifies annual facility-level reports through a multi-step process (e.g.,
combination of electronic checks and manual reviews by staff) to identify potential errors and ensure that data
submitted to EPA are accurate, complete, and consistent. Based on the results of the verification process, the EPA
follows up with facilities to resolve mistakes that may have occurred.47
The resiilis nl" 1 lie \pprnaeh 2 qiianliiali\ e iiiiceriaiiiis ; 111; 11\ sis are suiiimari/ed in Table 4-54 ( ailxin dioside
anMimpikni CO emissions fur 2<>l<> wereesimialed in he helueen 4. ^ and 4 " \l\ITCO l!q al llie l>5 pereeni
47 See .
Industrial Processes and Product Use 4-61
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
ainl'idcMcc le\el. This indicates ;i mime iifapprii\iiiialcl> 5 pcrcciil hckm in 5 percent :ihi»\e I lie emission esiiniale
ol'4.5 \1\1T CO \ x\
Table 4-54: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from CO2
Consumption (MMT CO2 Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY
REPORT
Sounv
21116 rimissiiin l!siiin;iU'
(MMT CO: l'.(|.)
I iuvrl;iinl\ K;iii|KT I.I HUT
lilllllld 1 $111111(1 1 $111111(1
I |)|KT
lion 11(1
C(): Consumption C'():
4.5
4.3 4.7 -5%
Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a l)5 percent confidence interval.
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2016.
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 QA/QC and Verification Procedures section in the introduction of the
IPPU chapter.
Planned Improvements
EPA will continue to evaluate the potential to include additional GHGRP data on other emissive end-uses to
improve 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.48 These improvements, in addition to updating the time series when new data is available, are still in
process and will be incorporated into future Inventory reports.
4.16 Phosphoric Acid Production (CRF Source
0)
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, as well as in Idaho and Utah and is used primarily as a raw material for wet-process phosphoric acid
production (USGS 2017). The composition of natural phosphate rock varies depending upon 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 also may contain organic carbon. The calcium carbonate component of the
phosphate rock is integral to the phosphate rock chemistry. Phosphate rock can also contain organic carbon that is
physically incorporated into the mined rock but is not an integral component of the phosphate rock chemistry.
48 See .
4-62 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
The phosphoric acid production process involves chemical reaction of the calcium phosphate (Ca3(P04)2)
component of the phosphate rock with sulfuric acid (H2SO4) and recirculated phosphoric acid (H3PO4) (EFMA
2000). However, the generation of CO2 is due to the associated limestone-sulfuric acid reaction, as shown below:
CaCO3 + //2SO4 + H20 —* CaS04 • 2H20 + C02
Total U.S. phosphate rock production sold or used in 2016 was 26.5 million metric tons (USGS 2017). Total imports
of phosphate rock to the United States in 2016 were approximately 1.6 million metric tons (USGS 2017). Between
2012 and 2015, most of the imported phosphate rock (58 percent) came from Peru, with the remaining 42 percent
being from Morocco (USGS 2017). All phosphate rock mining companies are vertically integrated with fertilizer
plants that produce phosphoric acid located near the mines. Some additional phosphoric acid production facilities
are located in Texas, Louisiana, and Mississippi that used imported phosphate rock.
Over the 1990 to 2016 period, domestic production has decreased by nearly 47 percent. Total CO2 emissions from
phosphoric acid production were 1.0 MMT CO2 Eq. (992 kt CO2) in 2016 (see Table 4-55). Domestic consumption
of phosphate rock in 2016 was estimated to have gone unchanged over 2015 levels (USGS 2017).
Table 4-55: CO2 Emissions from Phosphoric Acid Production (MMT CO2 Eq. and kt)
Year
MMT CO2 Eq.
kt
1990
1.5
1,529
2005
1.3
1.342
2012
1.1
1,118
2013
1.1
1,149
2014
1.0
1,038
2015
1.0
999
2016
1.0
992
Methodology
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:
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 is based on the assumption 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. The United States
uses a country-specific methodology to calculate emissions from production of phosphoric acid from phosphate
rock.
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 4-56).
For the years 1990 through 1992, and 2005 through 2016, 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 average share of U.S. production in those states
Industrial Processes and Product Use 4-63
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
from 1993 to 2004 data. For the years 2005 through 2016, the same approximation method is used, but the share of
U.S. production in those states data were obtained from the USGS commodity specialist for phosphate rock (USGS
2012). 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 2016 were obtained from
USGS Minerals Yearbook: Phosphate Rock (USGS 1994 through 2015b), and from USGS Minerals Commodity
Summaries: Phosphate Rock (USGS 2016, 2017). From 2004 through 2016, the USGS reported no exports of
phosphate rock from U.S. producers (USGS 2005 through 2015b).
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
(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 carbon. Calcined phosphate rock mined in
North Carolina and Idaho contains approximately 0.41 percent and 0.27 percent inorganic C, respectively (see Table
4-57).
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 (80 percent of domestic production) and
carbonate content data for phosphate rock mined in Morocco are used to calculate CO2 emissions from consumption
of imported phosphate rock. The CO2 emissions calculation is based on the assumption that all of the 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 4-56: Phosphate Rock Domestic Consumption, Exports, and Imports (kt)
Location/Year
1990
2005
2012
2013
2014
2015
2016
U.S. Domestic Consumption
49,800
35,200
27,300
28,800
26,700
26,200
26,500
FLandNC
42,494
28,160
21,840
23,040
21,360
20,960
21,200
ID and UT
7,306
7,040
5,460
5,760
5,340
5,240
5,300
Exports—FL and NC
6,240
0
0
0
0
0
0
Imports
451
2,630
3,570
3,170
2,390
1,960
1,600
Total U.S. Consumption
44,011
37,830
30,870
31,970
29,090
28,160
28,100
Table 4-57: Chemical Composition of Phosphate Rock (Percent by Weight)
Central
North
North Carolina
Idaho
Composition
Florida
Florida
(calcined)
(calcined)
Morocco
Total Carbon (as C)
1.60
1.76
0.76
0.60
1.56
Inorganic Carbon (as C)
1.00
0.93
0.41
0.27
1.46
Organic Carbon (as C)
0.60
0.83
0.35
0.00
0.10
Inorganic Carbon (as CO2)
3.67
3.43
1.50
1.00
5.00
Source: FIPR (2003a).
Uncertainty and Time-Series Consistency
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 2016. For previous years in the time series,
USGS provided the data disaggregated regionally; however, beginning in 2006, only total U.S. phosphate rock
production was reported. Regional production for 2016 was estimated based on regional production data from
previous years and multiplied by regionally-specific emission factors. There is uncertainty associated with the
degree to which the estimated 2016 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 calculation are reported 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.
4-64 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 An additional source of uncertainty in the calculation of CO2 emissions from phosphoric acid production is the
2 carbonate composition of phosphate rock; the composition of phosphate rock varies depending upon where the
3 material is mined, and may also vary over time. The Inventory relies on one study (FIPR 2003a) of chemical
4 composition of the phosphate rock; limited data are available beyond this study. Another source of uncertainty is the
5 disposition of the organic carbon content of the phosphate rock. A representative of the Florida Institute of
6 Phosphate Research (FIPR) indicated that in the phosphoric acid production process, the organic C content of the
7 mined phosphate rock generally remains in the phosphoric acid product, which is what produces the color of the
8 phosphoric acid product (FIPR 2003b). Organic carbon is therefore not included in the calculation of CO2 emissions
9 from phosphoric acid production.
10 A third source of uncertainty is the assumption that all domestically-produced phosphate rock is used in phosphoric
11 acid production and used without first being calcined. Calcination of the phosphate rock would result in conversion
12 of some of the organic C in the phosphate rock into CO2. However, according to air permit information available to
13 the public, at least one facility has calcining units permitted for operation (NCDENR 2013).
14 Finally, USGS indicated that approximately 7 percent of domestically-produced phosphate rock is used to
15 manufacture elemental phosphorus and other phosphorus-based chemicals, rather than phosphoric acid (USGS
16 2006). According to USGS, there is only one domestic producer of elemental phosphorus, in Idaho, and no data
17 were available concerning the annual production of this single producer. Elemental phosphorus is produced by
18 reducing phosphate rock with coal coke, and it is therefore assumed that 100 percent of the carbonate content of the
19 phosphate rock will be converted to CO2 in the elemental phosphorus production process. The calculation for CO2
20 emissions is based on the assumption that phosphate rock consumption, for purposes other than phosphoric acid
21 production, results in CO2 emissions from 100 percent of the inorganic carbon content in phosphate rock, but none
22 from the organic carbon content.
23 The ivsiilis of ihe \pproach 2 (|ii;inlil;ili\e iiiiceriaiiils aiials sis are summari/ed 111 I ahle 4-5X 2d l(> phosphoric acid
24 produclioii (() emissions were csiimaled lo he helweeii 0 X and I 2\l\l l ( () lx| al 1 lie l>5 percciil confidence
25 le\ el This mdicales a mime of appi'o\inialel\ 11> percenl below and 2<> percciil aho\ e I he emission esiimale of I 0
26 \1\1T'CO \x\
27 Table 4-58: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from
28 Phosphoric Acid Production (MMT CO2 Eq. and Percent) - TO BE UPDATED FOR FINAL
29 INVENTORY REPORT
Shu ni-
(¦iis
20H> rimissiiin Kslim;iU-
1 MMT CO: Kil l
I iHiTl;iinl> ki'liiliu'In I'.iiiissiiui Kslimuk"'
(MM I' CO: Km.) ("..)
I.I HUT I |)|KT I.I HUT I |)|KT
liiillllll Bound liiillllil liiillllil
Phosphoric Acid l'roi
luction
t ()
1.0
0.8 1.2 -19% +20%
Range of emission estimates predicted by Monte Carlo Stochastic Simulation lor a 95 percenl confidence interval.
30 Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
31 through 2016. Details on the emission trends through time are described in more detail in the Methodology section,
32 above.
33 For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
34 Chapter 6 of the 2006IPCC Guidelines, see QA/QC and Verification Procedures section in the introduction of the
35 IPPU Chapter.
36 Planned Improvements
37 EPA continues to evaluate potential improvements to the Inventory estimates for this source category, which include
38 direct integration of EPA's GHGRP data for 2010 through 2016 and the use of reported GHGRP data to update the
39 inorganic C content of phosphate rock for prior years. Confidentiality of CBI continues to be assessed, in addition to
40 the applicability of GHGRP data for the averaged inorganic C content data (by region) from 2010 through 2016 to
41 inform estimates in prior years in the required time series (i.e., 1990 through 2009). In implementing improvements
42 and integration of data from EPA's GHGRP, the latest guidance from the IPCC on the use of facility-level data in
Industrial Processes and Product Use 4-65
-------
1 national inventories will be relied upon.49 This planned improvement is still in development by EPA and have not
2 been implemented into the current Inventory report.
3 4.17 Iron and Steel Production (CRF Source
4 Category 2C1) and Metallurgical Coke
5 Production
6 Iron and steel production is a multi-step process that generates process-related emissions of carbon dioxide (CO2)
7 and methane (CH4) as raw materials are refined into iron and then transformed into crude steel. Emissions from
8 conventional fuels (e.g., natural gas, fuel oil) consumed for energy purposes during the production of iron and steel
9 are accounted for in the Energy chapter.
10 Iron and steel production includes six distinct production processes: coke production, sinter production, direct
11 reduced iron (DRI) production, pig iron50 production, electric arc furnace (EAF) steel production, and basic oxygen
12 furnace (BOF) steel production. The number of production processes at a particular plant is dependent upon the
13 specific plant configuration. Most process CO2 generated from the iron and steel industry is a result of the
14 production of crude iron.
15 In addition to the production processes mentioned above, CO2 is also generated at iron and steel mills through the
16 consumption of process byproducts (e.g., blast furnace gas, coke oven gas) used for various purposes including
17 heating, annealing, and electricity generation. Process byproducts sold for use as synthetic natural gas are deducted
18 and reported in the Energy chapter. In general, CO2 emissions are generated in these production processes through
19 the reduction and consumption of various carbon-containing inputs (e.g., ore, scrap, flux, coke byproducts). In
20 addition, fugitive CH4 emissions can also be generated from these processes, as well as from sinter, direct iron and
21 pellet production.
22 Currently, there are approximately 11 integrated iron and steel steelmaking facilities that utilize BOFs to refine and
23 produce steel from iron. These facilities have 21 active blast furnaces between them as of 2015. More than 100
24 steelmaking facilities utilize EAFs to produce steel primarily from recycled ferrous scrap (USGS 2017). The trend in
25 the United States for integrated facilities has been a shift towards fewer BOFs and more EAFs. EAFs use scrap steel
26 as their main input and use significantly less energy than BOFs. In addition, there are 16 cokemaking facilities, of
27 which 6 facilities are co-located with integrated iron and steel facilities (ACCCI2016). In the United States, raw
28 steel is produced in 37 states, but six states - Alabama, Arkansas, Indiana, Kentucky, Mississippi, and Tennessee -
29 count for roughly 50 percent of total production (AISI2017).
30 Total annual production of crude steel in the United States was fairly constant between 2000 and 2008 ranged from a
31 low of 99,320,000 tons to a high of 109,880,000 tons (2001 and 2004, respectively). Due to the decrease in demand
32 caused by the global economic downturn (particularly from the automotive industry), crude steel production in the
33 United States sharply decreased to 65,459,000 tons in 2009. In 2010, crude steel production rebounded to
34 88,731,000 tons as economic conditions improved and then continued to increase to 95,237,000 tons in 2011 and
35 97,769,000 tons in 2012; crude steel production slightly decreased to 95,766,000 tons in 2013 and then slightly
36 increased to 97,195,000 tons in 2014 (AISI 2017); crude steel production decreased to 86,912,000 tons in 2015 and
37 decreased again slightly in 2016 to 86,504,000 tons, a decrease of roughly 11 percent from 2014 levels. The United
38 States was the fourth largest producer of raw steel in the world, behind China, Japan and India, accounting for
39 approximately 4.8 percent of world production in 2016 (AISI 2017).
49 See .
50 pjg jj.on js 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.
4-66 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 The majority of CO2 emissions from the iron and steel production process come from the use of coke in the
2 production of pig iron and from the consumption of other process byproducts, with lesser amounts emitted from the
3 use of flux and from the removal of carbon from pig iron used to produce steel.
4 According to the 2006IPCC Guidelines, the production of metallurgical coke from coking coal is considered to be
5 an energy use of fossil fuel and the use of coke in iron and steel production is considered to be an industrial process
6 source. Therefore, the 2006 IPCC Guidelines suggest that emissions from the production of metallurgical coke
7 should be reported separately in the Energy sector, while emissions from coke consumption in iron and steel
8 production should be reported in the Industrial Processes and Product Use sector. However, the approaches and
9 emission estimates for both metallurgical coke production and iron and steel production are presented here because
10 much of the relevant activity data is used to estimate emissions from both metallurgical coke production and iron
11 and steel production. For example, some byproducts (e.g., coke oven gas) of the metallurgical coke production
12 process are consumed during iron and steel production, and some byproducts of the iron and steel production
13 process (e.g., blast furnace gas) are consumed during metallurgical coke production. Emissions associated with the
14 consumption of these byproducts are attributed at the point of consumption. Emissions associated with the use of
15 conventional fuels (e.g., natural gas, fuel oil) for electricity generation, heating and annealing, or other
16 miscellaneous purposes downstream of the iron and steelmaking furnaces are reported in the Energy chapter.
17 Metallurgical Coke Production
18 Emissions of CO2 from metallurgical coke production in 2016 were 1.3 MMT CO2 Eq. (1,323 kt CO2) (see Table
19 4-59 and Table 4-60). Emissions decreased significantly in 2016 by 54 percent from 2015 levels and have decreased
20 by 47 percent (1.2 MMT CO2 Eq.) since 1990. Coke production in 2016 was 43 percent lower than in 2000 and 57
21 percent below 1990.
22 Table 4-59: CO2 Emissions from Metallurgical Coke Production (MMT CO2 Eq.)
Gas 1990 2005 2012 2013 2014 2015 2016
CO2 2.5 : 2.0 0.5 1.8 2.0 2.8 1.3
Total 2.5 2.0 0.5 1.8 2.0 2.8 1.3
23 Table 4-60: CO2 Emissions from Metallurgical Coke Production (kt)
Gas 1990 2005 2012 2013 2014 2015 M16
CO2 2,503 2,044 543 1,824 2,014 2,839 1,323
Total 2,503 2,044 543 1,824 2,014 2,839 1,323
25
26 Iron and Steel Production
27 Emissions of CO2 and CH4 from iron and steel production in 2016 were 40.9 MMT CO2 Eq. (40,896 kt) and 0.0074
28 MMT CO2 Eq. (0.3 kt CH4), respectively (see Table 4-61 through Table 4-64), totaling approximately 40.9 MMT
29 CO2 Eq. Emissions decreased in 2016 from 2015 and have decreased overall since 1990 due to restructuring of the
30 industry, technological improvements, and increased scrap steel utilization. Carbon dioxide emission estimates
31 include emissions from the consumption of carbonaceous materials in the blast furnace, EAF, and BOF, as well as
32 blast furnace gas and coke oven gas consumption for other activities at the steel mill.
33 In 2016, domestic production of pig iron decreased by 12 percent from 2015 levels. Overall, domestic pig iron
34 production has declined since the 1990s. Pig iron production in 2016 was 53 percent lower than in 2000 and 55
35 percent below 1990. Carbon dioxide emissions from iron production have decreased by 78 percent since 1990.
36 Carbon dioxide emissions from steel production have decreased by 14 percent (1.1 MMT CO2 Eq.) since 1990,
37 while overall CO2 emissions from iron and steel production have declined by 59 percent (58.1 MMT CO2 Eq.) from
38 1990 to 2016.
39 Table 4-61: CO2 Emissions from Iron and Steel Production (MMT CO2 Eq.)
Source/Activity Data 1990 2005 2012 2013 2014 2015 2016"
Sinter Production 2.4 1.7 1.2 1.1 1.1 1.0 0.9
Industrial Processes and Product Use 4-67
-------
Iron Production
45.6
17.5
10.9
11.9
18.6
11.7
9.9
Pellet Production
1.8
1.5
1.2
1.2
1.1
1.0
0.9
Steel Production
7.9
9.4
9.9
8.6
7.5
6.9
6.8
Other Activities3
41.2
35.9
31.7
28.7
27.9
24.3
22.4
Total
99.0
66.0
54.9
51.5
56.2
44.9
40.9
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.
1 Table 4-62: CO2 Emissions from Iron and Steel Production (kt)
Source/Activity Data
1990
2005
2012
2013
2014
2015
2016
Sinter Production
2,448
1,663
1,159
1,117
1,104
1,016
877
Iron Production
45,592
17,545
10,918
11,935
18,629
11,696
9,853
Pellet Production
1,817
1,503
1,219
1,146
1,126
964
869
Steel Production
7,933
9,356
9,860
8,617
7,450
6,924
6,850
Other Activitiesa
41,193
35,934
31,750
28,709
27,911
24,280
22,448
Total
98,984
66,003
54,906
51,525
56,220
44,879
40,896
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.
2 Table 4-63: Cm Emissions from Iron and Steel Production (MMT CO2 Eq.)
Source/Activity Data
1990
2005
2012
2013
2014
2015
2016
Sinter Production + +; + + + + +
Total
+
+
+
+
+
+
+
+ Does not exceed 0.05 MMT CO2 Eq.
3 Table 4-64: Cm Emissions from Iron and Steel Production (kt)
Source/Activity Data
1990
2005
2012
2013
2014
2015
2016
Sinter Production
0.9
0.6 5
0.4
0.4
0.4
0.3
0.3
Total 0.9 0.6 0.4 0.4 0.4 0.3 0.3
Methodology
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Emission estimates presented in this chapter utilize a country-specific approach based on Tier 2 methodologies
provided by the 2006IPCC Guidelines. These Tier 2 methodologies call for a mass balance accounting of the
carbonaceous inputs and outputs during the iron and steel production process and the metallurgical coke production
process. Tier 1 methods are used for certain iron and steel production processes (i.e., sinter production, pellet
production and DRI production) for which available data are insufficient to apply a Tier 2 method.
The Tier 2 methodology equation is as follows:
Eco2 ~
X Ca) ~ X Q)
44
12
where,
Ec02
a
b
Qa
ca
Qb
Emissions from coke, pig iron, EAF steel, or BOF steel production, metric tons
Input material a
Output material b
Quantity of input material a, metric tons
Carbon content of input material a, metric tons C/metric ton material
Quantity of output material b, metric tons
4-68 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
Cb
44/12
Carbon content of output material b, metric tons C/metric ton material
Stoichiometric ratio of CO2 to C
4 The Tier 1 methodology equations are as follows:
5
Es,p = Qs x EFSiP
6
Ed,C02 — Qd X EFd£Q2
1
Ep,co2 — Qp x EFp C02
8 where,
9
10
11
12
13
14
15
16
17
EFp,co2
Ed,C02
Qd
EFd,co2
QP
Emissions from sinter production process for pollutant p (CO2 or CH4), metric ton
Quantity of sinter produced, metric tons
Emission factor for pollutant p (CO2 or CH4), metric ton /Vmctric ton sinter
Emissions from DRI production process for CO2, metric ton
Quantity of DRI produced, metric tons
Emission factor for CO2, metric ton CCh/metric ton DRI
Quantity of pellets produced, metric tons
Emission factor for CO2, metric ton CCh/metric ton pellets produced
is Metallurgical Coke Production
19 Coking coal is used to manufacture metallurgical coke that is used primarily as a reducing agent in the production of
20 iron and steel, but is also used in the production of other metals including zinc and lead (see Zinc Production and
21 Lead Production sections of this chapter). Emissions associated with producing metallurgical coke from coking coal
22 are estimated and reported separately from emissions that result from the iron and steel production process. To
23 estimate emissions from metallurgical coke production, a Tier 2 method provided by the 2006IPCC Guidelines was
24 utilized. The amount of carbon contained in materials produced during the metallurgical coke production process
25 (i.e., coke, coke breeze and coke oven gas) is deducted from the amount of carbon contained in materials consumed
26 during the metallurgical coke production process (i.e., natural gas, blast furnace gas, and coking coal). Light oil,
27 which is produced during the metallurgical coke production process, is excluded from the deductions due to data
28 limitations. The amount of carbon contained in these materials is calculated by multiplying the material-specific
29 carbon content by the amount of material consumed or produced (see Table 4-65). The amount of coal tar produced
30 was approximated using a production factor of 0.03 tons of coal tar per ton of coking coal consumed. The amount of
31 coke breeze produced was approximated using a production factor of 0.075 tons of coke breeze per ton of coking
32 coal consumed (AISI 2008; DOE 2000). Data on the consumption of carbonaceous materials (other than coking
33 coal) as well as coke oven gas production were available for integrated steel mills only (i.e., steel mills with co-
34 located coke plants). Therefore, carbonaceous material (other than coking coal) consumption and coke oven gas
35 production were excluded from emission estimates for merchant coke plants. Carbon contained in coke oven gas
36 used for coke-oven underfiring was not included in the deductions to avoid double-counting.
37 Table 4-65: Material Carbon Contents for Metallurgical Coke Production
Material
kgC/kg
Coal Tar
0.62
Coke
0.83
Coke Breeze
0.83
Coking Coal
0.73
Material
kgC/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.
38 Although the 2006IPCC Guidelines provide a Tier 1 CH4 emission factor for metallurgical coke production (i.e.,
39 0.1 g CH4 per metric ton of coke production), it is not appropriate to use because CO2 emissions were estimated
Industrial Processes and Product Use 4-69
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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
reports 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 2017) (see Table 4-66). 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
(AISI2004 through 2017) and through personal communications with AISI (AISI 2008) (see Table 4-67). The factor
for the quantity of coal tar produced per ton of coking coal consumed was provided by AISI (AISI 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 coking coal, metallurgical coke, coal tar, coke oven gas, and blast
furnace gas were provided by the 2006IPCC Guidelines. The C content for coke breeze was assumed to equal the C
content of coke.
Table 4-66: Production and Consumption Data for the Calculation of CO2 Emissions from
Metallurgical Coke Production (Thousand Metric Tons)
Source/Activity Data
1990
2005
2012
2013
2014
2015
2016
Metallurgical Coke Production
Coking Coal Consumption at Coke Plants
35,269
21,259
18,825
19,481
19,321
17,879
14,955
Coke Production at Coke Plants
25,054
15,167
13,764
13,898
13,748
12,479
10,755
Coal Breeze Production
2,645
1,594
1,412
1,461
1,449
1,341
1,122
Coal Tar Production
1,058
638
565
584
580
536
449
Table 4-67: Production and Consumption Data for the Calculation of CO2 Emissions from
Metallurgical Coke Production (Million ft3)
Source/Activity Data
1990
2005
2012
2013
2014
2015
2016
Metallurgical Coke Production
Coke Oven Gas Production
250,767
114,213
113,772
108,162
102,899
84,336
74,807
Natural Gas Consumption
599
2,996
3,267
3,247
3,039
2,338
2,077
Blast Furnace Gas Consumption
24,602
4,460
4,351
4,255
4,346
4,185
3,741
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). The carbon contained in the pig iron, blast furnace gas, and blast furnace inputs was estimated by
multiplying the material-specific C content by each material type (see Table 4-68). 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 C contained in inputs such as natural gas, coke oven gas, fluxes (e.g.
burnt lime or dolomite), and pig iron. In each case, the carbon was calculated by multiplying material-specific
carbon contents by each material type (see Table 4-68). For EAFs, the amount of EAF anode consumed was
4-70 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 approximated by multiplying total EAF steel production by the amount of EAF anode consumed per metric ton of
2 steel produced (0.002 metric tons EAF anode per metric ton steel produced [AISI 2008]). The amount of flux (e.g.,
3 burnt lime or dolomite) used in pig iron production was deducted from the "Other Process Uses of Carbonates"
4 source category (CRF Source Category 2A4) to avoid double-counting.
5 Carbon dioxide emissions from the consumption of blast furnace gas and coke oven gas for other activities occurring
6 at the steel mill were estimated by multiplying the amount of these materials consumed for these purposes by the
7 material-specific carbon content (see Table 4-68).
8 Carbon dioxide emissions associated with the sinter production, direct reduced iron production, pig iron production,
9 steel production, and other steel mill activities were summed to calculate the total CO2 emissions from iron and steel
10 production (see Table 4-61 and Table 4-62).
11 Table 4-68: Material Carbon Contents for Iron and Steel Production
Material
kgC/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
kgC/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.
12 The production process for sinter results in fugitive emissions of CH4, which are emitted via leaks in the production
13 equipment, rather than through the emission stacks or vents of the production plants. The fugitive emissions were
14 calculated by applying Tier 1 emission factors taken from the 2006 IPCC Guidelines for sinter production (see Table
15 4-69). Although the 1995 IPCC Guidelines (IPCC/UNEP/OECD/IEA 1995) provide a Tier 1 CH4 emission factor
16 for pig iron production, it is not appropriate to use because CO2 emissions were estimated using the Tier 2 mass
17 balance methodology. The mass balance methodology makes a basic assumption that all carbon that enters the pig
18 iron production process either exits the process as part of a carbon-containing output or as CO2 emissions; the
19 estimation of CH4 emissions is precluded. A preliminary analysis of facility-level emissions reported during iron
20 production further supports this assumption and indicates that CH4 emissions are below 500 kt CO2 Eq. and well
21 below 0.05 percent of total national emissions. The production of direct reduced iron also results in emissions of
22 CH4 through the consumption of fossil fuels (e.g., natural gas, etc.); however, these emission estimates are excluded
23 due to data limitations. Pending further analysis and resources, EPA may include these emissions in future reports to
24 enhance completeness. EPA is still assessing the possibility of including these emissions in future reports and have
25 not included this data in the current report.
26 Table 4-69: ChU Emission Factors for Sinter and Pig Iron Production
Material Produced
Factor
Unit
Sinter
0.07
kg CEU/metric ton
Source: IPCC (2006), Table 4.2.
27 Emissions of CO2 from sinter production, direct reduced iron production and pellet production were estimated by
28 multiplying total national sinter production and the total national direct reduced iron production by Tier 1 CO2
29 emission factors (see Table 4-70). Because estimates of sinter production, direct reduced iron production and pellet
30 production were not available, production was assumed to equal consumption.
Industrial Processes and Product Use 4-71
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Table 4-70: CO2 Emission Factors for Sinter Production, Direct Reduced Iron Production and
Pellet Production
Metric Ton
Material Produced
CCh/Metric 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 2016 were obtained from AISI's Annual
Statistical Report (AISI2004 through 2017) and through personal communications with AISI (AISI2008) (see
Table 4-71). 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 2016) and personal communication with
the USGS Iron and Steel Commodity Specialist (Fenton 2015). However, data for DRI consumed inEAFs 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 2017) and through personal communications with AISI (AISI
2008) (see Table 4-71 and Table 4-72).
Data for EAF steel production, flux, EAF charge carbon, and natural gas consumption were obtained from AISI's
Annual Statistical Report (AISI 2004 through 2017) and through personal communications with AISI (AISI 2006
through 2016 and AISI 2008). The factor for the quantity of EAF anode consumed per ton of EAF steel produced
was provided by AISI (AISI 2008). Data for BOF steel production, flux, natural gas, natural ore, pellet, sinter
consumption as well as BOF steel production were obtained from AISI's Annua! Statistical Report (AISI 2004
through 2017) and through personal communications with AISI (AISI 2008). Data for EAF and BOF scrap steel, pig
iron, andDRI consumption were obtained from the USGS Minerals Yearbook-Iron and Steel Scrap (USGS 1991
through 2016). 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 .innual Statistical Report (AISI 2004 through 2017) and
through personal communications with AISI (AISI 2008).
Data on blast furnace gas and coke oven gas sold for use as synthetic natural gas were obtained from EIA's Natural
Gas Annual (EIA 2016b). Carbon contents for direct reduced iron, EAF carbon electrodes, EAF charge carbon,
limestone, dolomite, pig iron, and steel were provided by the 2006 IPCC Guidelines. The carbon contents for natural
gas, fuel oil, and direct injection coal were obtained from EIA (EIA 2016c) 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 . Innual Statistical Report (AISI 2004 through 2017). Heat contents for coke oven gas and
blast furnace gas were provided in Table 37 of AISI's Annua! Statistical Report (AISI 2004 through 2017) and
confirmed by AISI staff (Carroll 2016).
4-72 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Table 4-71: Production and Consumption Data for the Calculation of CO2 and ChU Emissions
2 from Iron and Steel Production (Thousand Metric Tons)
Source/Activity Data
1990
2005
2012
2013
2014
2015
2016
Sinter Production
Sinter Production
12,239
8,315
5,795
5,583
5,521
5,079
4,385
Direct Reduced Iron
Production
Direct Reduced Iron
Production
516
1,303
3,530
3,350
4,790
4,790
4,777
Pellet Production
Pellet Production
60,563
50,096
40,622
38,198
37,538
32,146
28,967
Pig Iron Production
Coke Consumption
24,946
13,832
9,571
9,308
11,136
7,969
7,124
Pig Iron Production
49,669
37,222
32,063
30,309
29,375
25,436
22,293
Direct Injection Coal
Consumption
1,485
2,5"'
2,802
2,675
2,425
2,275
1,935
EAF Steel Production
EAF Anode and Charge
Carbon Consumption
67
1,127
1,318
1,122
1,062
1,072
1,120
Scrap Steel
Consumption
42,691
46,600
50,900
47,300
48,873
44,000
43,211
Flux Consumption
319
695
748
771
771
998
998
EAF Steel Production
33,511
52,194
52,415
52,641
55,174
49,451
52,589
BOF Steel Production
Pig Iron Consumption
47,307
34,400
31,500
29,600
23,755
20,349
18,620
Scrap Steel
Consumption
14,713
11,400
8,350
7,890
5,917
4,526
4,573
Flux Consumption
576
582
476
454
454
454
408
BOF Steel Production
43,9"'
42,705
36,282
34,238
33,000
29,396
25,888
3 Table 4-72: Production and Consumption Data for the Calculation of CO2 Emissions from
4 Iron and Steel Production (Million ft3 unless otherwise specified)
Source/Activity Data
1990
2005
2012
2013
2014
2015
2016
Pig Iron Production
Natural Gas
Consumption
56,273
59,844
62,469
48,812
47,734
43,294
38,396
Fuel Oil Consumption
(thousand gallons)
163,397
16,170
19,240
17,468
16,674
9,326
6,124
Coke Oven Gas
Consumption
22,033
16,557
18,608
17,710
16,896
13,921
12,404
Blast Furnace Gas
Production
1,439,380
1,299,980
1,139,578
1,026,973
1,000,536
874,670
811,005
EAF Steel Production
Natural Gas
Consumption
15,905
19,985
11,145
10,514
9,622
8,751
3,915
BOF Steel Production
Coke Oven Gas
Consumption
3,851
524
568
568
524
386
367
Other Activities
Coke Oven Gas
Consumption
224,883
97,132
94,596
89,884
85,479
70,029
62,036
Blast Furnace Gas
Consumption
1,414,778
1,295,520
1,135,227
1,022,718
996,190
870,485
807,264
Industrial Processes and Product Use 4-73
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Uncertainty and Time-Series Consistency
The estimates of CO2 emissions from metallurgical coke production are based on 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 C 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; some of this coal may be used for electricity generation.
There is also uncertainty associated with the C contents for pellets, sinter, and natural ore, which are assumed to
equal the C contents of direct reduced iron, when consumed in the blast furnace. 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 this Inventory report, EPA initiated conversation with AISI to update the qualitative and quantitative uncertainty
metrics associated with AISI data elements. EPA has yet to incorporate these changes into this current Public
Review draft but will include them in the final Inventory report published in April 2018.
The lesiills of I he \pproach 2 t|ii;inlil;ili\ e iiiiccriaiuls ; 111; 11\ sis ;iiv suniniai'i/ed 111 I ahle 4-~' fur niclallurmcal coke
production and iron and sieel production Tolal CO emissions Iroin niclalhumcal coke production and iron ;ind steel
production for 2d lii were estimated lo he between "5 t> and 4l>.4 \1\11 ( () I !q ;iiihe l>5 percent confidence le\ el
This uidic;iles a raime of approximate^ I ~ perceul helow and I ~ perceui aho\ e llie emission esiiniale of 42.2 \1\1T
CO I !q Tolal CI I emissions from niclalhumcal coke producliou and iron and sieel production for2<>|()<¦ and 0 no1; \|\| I CO I !q al llie l>5 perceul confidence lex el This indicates a ranue of
approximate^ ll> perceul helow and I'J perceul aho\e llie emission esiiniale of 0 00~ \|\IT CO I !q
Table 4-73: Approach 2 Quantitative Uncertainty Estimates for CO2 and CH4 Emissions from
Iron and Steel Production and Metallurgical Coke Production (MMT CO2 Eq. and Percent) -
TO BE UPDATED FOR FINAL INVENTORY REPORT
Sou I'll'
.. 21116 riiiiissiiui I'sliniiili'
" (MM 1 ( (): l.(|.)
I iKi'iliiinlN kiin^i' Ri'l;ili\i' in I'.iiiissiiui IMimuli'1
(MM 1 ( (): i.tl.) C'i.)
I.I HUT I |)|KT 1.1 HUT I |)|KT
Bound Bound Bound Bound
Metallu
and St
rgical Cc:
eel Prodi
ike & Iron
iction
CO: 42.2
35.0 49.4 -17% +17%
Metallu
rgical Cc:
ike & Iron
CI It +
+ + -19% +19%
and St
eel Prod 1
.iction
Does not exceed 0.05 MMT CO: Kq.
Range of emission estimates predicted by Monte Curio Stochastic Simulation lor a 95 percent confidence interval.
4-74 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
2 through 2016.
3 For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
4 Chapter 6 of the 2006IPCC Guidelines, see QA/QC and Verification Procedures section in the introduction of the
5 IPPU Chapter.
6 Recalculations Discussion
7 Updated USGS consumption data was provided for iron and steel manufacturing facilities for years 2014 to 2016.
8 The 2014 consumption from BOF facilities was updated from 2,680 kt to 0 kt, and the 2015 consumption from BOF
9 facilities was updated from a proxy value of 2,680 kt to 992 kt. These updates had measurable impacts on CO2
10 emission calculations. The total CO2 emissions from iron and steel production dropped 0.2 MMT CO2 Eq. (less than
11 one percent decrease) in 2014 and 1.0 MMT CO2 Eq. (2 percent decrease) in 2015 to 56.4 and 45.0 MMT CO2 Eq.
12 respectively (see Table 4-61 and Table 4-62).
13 The 2015 EAF dolomite consumption value has been updated from 250 thousand tons, used in the previous
14 Inventory report, to 550 thousand tons. This change had an insignificant impact on the emission estimates compared
15 to the previous Inventory.
16 The 2014 and 2015 EAF charge carbon consumption values have been updated from a proxy of 2013 values in the
17 previous report, to updated values reported by AISI (Carroll 2017). These changes had insignificant impacts on the
18 emissions from steel production in Table 4-61, and on the EAF anode and charge carbon consumption values in
19 Table 4-71.
20 Planned Improvements
21 Future improvements involve improving activity data and emission factor sources for estimating CO2 and CH4
22 emissions from pellet production. EPA will also evaluate and analyze data reported under EPA's GHGRP to
23 improve the emission estimates for this and other Iron and Steel Production process categories. Particular attention
24 will be made to ensure time-series consistency of the emissions estimates presented in future Inventory reports,
25 consistent with IPCC and UNFCCC guidelines. This is required as the facility-level reporting data from EPA's
26 GHGRP, with the program's initial requirements for reporting of emissions in calendar year 2010, are not available
27 for all inventory years (i.e., 1990 through 2009) as required for this Inventory. In implementing improvements and
28 integration of data from EPA's GHGRP, the latest guidance from the IPCC on the use of facility-level data in
29 national inventories will be relied upon.51
30 Additional improvements include accounting for emission estimates for the production of metallurgical coke to the
31 Energy chapter as well as identifying the amount of carbonaceous materials, other than coking coal, consumed at
32 merchant coke plants. Other potential improvements include identifying the amount of coal used for direct injection
33 and the amount of coke breeze, coal tar, and light oil produced during coke production. Efforts will also be made to
34 identify information to better characterize emissions from the use of process gases and fuels within the Energy and
35 Industrial Processes and Product Use chapters. This planned improvement is still in development and is not included
36 in this current Inventory report.
37 EPA also received comments during this Inventory's Expert Review cycle on recommendations to improve the
38 description of the iron and steel industry and emissive processes in the Inventory. EPA has begun to incorporate
39 some of these recommendations into this current Public Review draft and will require some additional time to
40 implement other substantive changes.
41 As noted in the Uncertainty and Time-Series Consistency section, EPA has initiated dialogue with AISI to better
42 characterize the uncertainties associated with AISI data elements used in the Inventory report. EPA plans to update
43 the qualitative uncertainty description and the quantitative uncertainty ranges of AISI data elements. This
51 See .
Industrial Processes and Product Use 4-75
-------
1
2
improvement is not yet incorporated and is planned to be included in the final Inventory report published in April
2018.
3 4.18 Ferroalloy Production (CRF Source
4 Category 2C2)
5 Carbon dioxide (CO2) and methane (CH4) are emitted from the production of several ferroalloys. Ferroalloys are
6 composites of iron (Fe) and other elements such as silicon (Si), manganese (Mn), and chromium (Cr). Emissions
7 from fuels consumed for energy purposes during the production of ferroalloys are accounted for in the Energy
8 chapter. Emissions from the production of two types of ferrosilicon (25 to 55 percent and 56 to 95 percent silicon),
9 silicon metal (96 to 99 percent silicon), and miscellaneous alloys (32 to 65 percent silicon) have been calculated.
10 Emissions from the production of ferrochromium and ferromanganese are not included here because of the small
11 number of manufacturers of these materials in the United States, and therefore, government information disclosure
12 rules prevent the publication of production data for these production facilities.
13 Similar to emissions from the production of iron and steel, CO2 is emitted when metallurgical coke is oxidized
14 during a high-temperature reaction with iron and the selected alloying element. Due to the strong reducing
15 environment, CO is initially produced, and eventually oxidized to CO2. A representative reaction equation for the
16 production of 50 percent ferrosilicon (FeSi) is given below:
17 Fe203 + 2Si02 + 7C -> 2FeSi + 7CO
18 While most of the carbon contained in the process materials is released to the atmosphere as CO2, a percentage is
19 also released as CH4 and other volatiles. The amount of CH4 that is released is dependent on furnace efficiency,
20 operation technique, and control technology.
21 When incorporated in alloy steels, ferroalloys are used to alter the material properties of the steel. Ferroalloys are
22 used primarily by the iron and steel industry, and production trends closely follow that of the iron and steel industry.
23 As of 2014, twelve companies in the United States produce ferroalloys (USGS 2016a).
24 Emissions of CO2 from ferroalloy production in 2016 were 1.8 MMT CO2 Eq. (1,796 kt CO2) (see Table 4-74 and
25 Table 4-75), which is a 17 percent reduction since 1990. Emissions of CHi from ferroalloy production in 2016 were
26 0.01 MMT CO2 Eq. (0.5 kt CH4), which is a 26 percent decrease since 1990.
27 Table 4-74: CO2 and ChU Emissions from Ferroalloy Production (MMT CO2 Eq.)
Gas
1990
2005
2012
2013
2014
2015
2016
CO2
2.2
1.4
1.9
1.8
1.9
2.0
1.8
CH4
+
+
+
+
+
+
+
Total
2.2
1.4
1.9
1.8
1.9
2.0
1.8
+ Does not exceed 0.05 MMT CO2 Eq.
28 Table 4-75: CO2 and ChU Emissions from Ferroalloy Production (kt)
Gas
1990
2005
2012
2013
2014
2015
2016
CO2
CH4
2,152
1
1,392
+
1,903
1
1,785
+
1,914
1
1,960
1
1,796
1
+ Does not exceed 0.5 kt.
4-76 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Methodology
Emissions of CO2 and CH4 from ferroalloy production were calculated52 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:
EC02 = Yj
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
• Silicon Metal: Annual production data were available from 1990 through 2005. The production data for
2005 were used as proxy for 2006 through 2010.
• 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 2016 (USGS 2013, 2014, 2015b, 2016b, 2017).
Table 4-76: 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
2012 175,108 154,507 169,385 NA
2013 164,229 144,908 158,862 NA
2014 176,161 155,436 170,404 NA
2015 180,372 159,151 174,477 NA
201 6 165,282 145,837 159,881 NA
NA - Not Available for product type, aggregated along with ferrosilicon (25-55% Si)
Uncertainty and Time-Series Consistency
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 proxy for 2005 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.
Also, 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.53 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 CH4 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 will
reduce CH4 emissions; however, specific furnace information was not available or included in the CH4 emission
estimates.
53 Emissions and sinks of biogenic carbon are accounted for in the Land Use, Land-Use Change, and Forestry chapter.
4-78 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
The lesiilis nl" llie \ppruadi 2 qiiaiililali\e iiiiceriaiiils ;in;il\ sis are siiiiimari/ed in Tahle 4- I'erRialkn prudiicliiiii
CO emissions finiii 2o l(i were esiimaled In he helueen I.Xand22 \I\1I 'CO l!q alllie l>5 perceni coiilideiice
le\el l liis Midlines ;i raime ill" ;ippi'ii\iiii;ik'l> 12 perceiil helnu ;md 12 percenl alxne llie emission esiiniale of 2 D
MM'I'CO I !q I'erroalkn prodiiclkiii CI I emissions were esiimaled lo he helueen ;i r:i nize ill" ;ippixiNi m;iiel\ 12
peieeiil hekm ;md 12 percent ;ihn\e llie einisskm esiiniale nl'<> n I MM'I'CO l!q
Table 4-77: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from
Ferroalloy Production (MMT CO2 Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY
REPORT
Shu nv
(¦;is
2016 Emission r.slim.iU-
(MM"I" CO: Kd.)
I ni'iThiiim Ki'kiliM' In riniissiiiii I'slimiik-'
(MM 1 ( O: Ktl.)
I.I HUT I |1|KT I.IHUT I |1|KT
lilllllld lillllllll lillllllll lillllllll
Ferroalloy 1
'reduction
t ( 1
2.0
1.8 2.2 -12% +12%
Ferroallov 1
'reduction
CIIi
+ + -12% +12%
Docs not exceed 0.05 MM'I'CO: Fq.
Range of emission estimates predicted by Monte Carlo Stochastic Simulation l"or a 95 percent confidence interval.
Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
through 2016. Details on the emission trends through time are described in more detail in the Methodology section,
above.
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 QA/QC and Verification Procedures section in the introduction of the
IPPU Chapter.
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, particular attention will be made to ensure time-series consistency of the emissions estimates presented in
future Inventory reports, consistent with IPCC and UNFCCC guidelines. This is required as the facility-level
reporting data from EPA's GHGRP, with the program's initial requirements for reporting of emissions in calendar
year 2010, are not available for all inventory years (i.e., 1990 through 2009) as required for this Inventory. In
implementing improvements and integration of data from EPA's GHGRP, the latest guidance from the IPCC on the
use of facility-level data in national inventories will be relied upon.54 EPA is still assessing the possibility of
incorporating this planned improvement into the national Inventory report and has not included these data sets into
the current Inventory report.
4.19 Aluminum Production (CRF Source
Category 2C3)
Aluminum is a light-weight, 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 ninth
largest producer of primary aluminum, with approximately 1 percent of the world total production (USGS 2017).
The United States was also a major importer of primary aluminum. The production of primary aluminum—in
54 See .
Industrial Processes and Product Use 4-79
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
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 (Na3AlF6). 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 1.3 MMT CO2 Eq. (1,334 kt) in 2016
(see Table 4-78). 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
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 4-78: CO2 Emissions from Aluminum Production (MMT CO2 Eq. and kt)
Year
MMT CO2 Eq.
kt
1990
6.8
6,831
2005
4.1
4.142
2012
3.4
3,439
2013
3.3
3,255
2014
2.8
2,833
2015
2.8
2,767
2016
1.3
1,334
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 "anode effects." These anode effects 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.
Since 1990, emissions of CF4 and C2F6 have declined by 94 percent and 87 percent, respectively, to 0.9 MMT CO2
Eq. of CF4 (0.1 kt) and 0.4 MMT CO2 Eq. of C2F6 (0.04 kt) in 2016, as shown in Table 4-79 and Table 4-80. 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 80 percent, while the combined CF4 and C2F6 emission rate (per metric
ton of aluminum produced) has been reduced by 69 percent. Emissions decreased by approximately 32 percent
between 2015 and 2016 due to decreases in aluminum production. CF4 and C2F6 emissions per metric ton of
aluminum produced increased between 2015 and 2016, in part because production decreased at low PFC emitting
facilities and stayed relatively stable at high PFC emitting facilities.
Table 4-79: 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
2012
2.3
0.7
2.9
2013
2.3
0.7
3.0
2014
1.9
0.6
2.5
2015
1.5
0.5
2.0
4-80 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
2016 0.9 0.4 1.4
Note: Totals may not sum due to
independent rounding.
Table 4-80: PFC Emissions from Aluminum Production (kt)
Year CF4 C2F6
1990 2.4 0.3
2005 0.4
+
2012 0.3
2013 0.3
2014 0.3
2015 0.2
2016 0.1
0.1
0.1
0.1
+
+
+ Does not exceed 0.05 kt.
In 2016, U.S. primary aluminum production totaled approximately 0.8 million metric tons, a 48 percent decrease
from 2015 production levels (USAA 2017). In 2016, three companies managed production at eight operational
primary aluminum smelters. One smelter remained on standby throughout the year, two smelters were temporarily
idled, and one non-operating smelters were permanently shut down during 2016 (USGS 2017). During 2016,
monthly U.S. primary aluminum production was lower for every month in 2016, when compared to the
corresponding months in 2015 (USAA 2017; USAA 2016b).
For 2017, total production for the January to September period was approximately 0.55 million metric tons
compared to 0.63 million metric tons for the same period in 2016, a 12 percent decrease (USAA 2017). Based on the
decrease in production, process CO2 and PFC emissions are likely to be lower in 2017 compared to 2016 if there are
no significant changes in process controls at operational facilities.
Process CO2 and PFC (i.e., CF4 and C2F6) emission estimates from primary aluminum production for 2010 through
2016 are available fromEPA's GHGRP—Subpart F (Aluminum Production) (EPA 2017). Under EPA's GHGRP,
facilities began reporting primary aluminum production process emissions (for 2010) in 2011; as a result, GHGRP
data (for 2010 through 2016) are available to be incorporated into the Inventory. EPA included reported emissions
for one facility that reported late to EPA's GHGRP in the previous Inventory estimate, so the total emissions
reported in the Inventory are not exactly equal to the emissions reported in the publically available GHGRP data.
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 Soderberg electrolysis cells, CO2 emissions from anode
consumption during electrolysis in all prebake and Soderberg 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).55 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
55 Code of Federal Regulations, Title 40: Protection of Environment, Part 98: Mandatory Greenhouse Gas Reporting, Subpart
F—Aluminum Production. See .
Methodology
Industrial Processes and Product Use 4-81
-------
1 Fuel Combustion (3.1 Fossil Fuel Combustion (CRF Source Category 1A)) and Annex 2.1, Methodology for
2 Estimating Emissions of CO2 from Fossil Fuel Combustion.
3 Process CO2 Emissions from Anode Consumption and Anode Baking
4 Carbon dioxide emission estimates for the years prior to the introduction of EPA's GHGRP in 2010 were estimated
5 2006IPCC Guidelines methods, but individual facility reported data were combined with process-specific emissions
6 modeling. These estimates were based on information previously gathered from EPA's Voluntary Aluminum
7 Industrial Partnership (VAIP) program, U.S. Geological Survey (USGS) Mineral Commodity reviews, and The
8 Aluminum Association (USAA) statistics, among other sources. Since pre- and post-GHGRP estimates use the same
9 methodology, emission estimates are comparable across the time series.
10 Most of the CO2 emissions released during aluminum production occur during the electrolysis reaction of the C
11 anode, as described by the following reaction:
12 2AI2O3 + 3C -> 4A1 + 3C02
13 For prebake smelter technologies, CO2 is also emitted during the anode baking process. These emissions can
14 account for approximately 10 percent of total process CO2 emissions from prebake smelters.
15 Depending on the availability of smelter-specific data, the CO2 emitted from electrolysis at each smelter was
16 estimated from: (1) the smelter's annual anode consumption, (2) the smelter's annual aluminum production and rate
17 of anode consumption (per ton of aluminum produced) for previous and/or following years, or (3) the smelter's
18 annual aluminum production and IPCC default CO2 emission factors. The first approach tracks the consumption and
19 carbon content of the anode, assuming that all C in the anode is converted to CO2. Sulfur, ash, and other impurities
20 in the anode are subtracted from the anode consumption to arrive at a C consumption figure. This approach
21 corresponds to either the IPCC Tier 2 or Tier 3 method, depending on whether smelter-specific data on anode
22 impurities are used. The second approach interpolates smelter-specific anode consumption rates to estimate
23 emissions during years for which anode consumption data are not available. This approach avoids substantial errors
24 and discontinuities that could be introduced by reverting to Tier 1 methods for those years. The last approach
25 corresponds to the IPCC Tier 1 method (IPCC 2006), and is used in the absence of present or historic anode
26 consumption data.
27 The equations used to estimate CO2 emissions in the Tier 2 and 3 methods vary depending on smelter type (IPCC
28 2006). For Prebake cells, the process formula accounts for various parameters, including net anode consumption,
29 and the sulfur, ash, and impurity content of the baked anode. For anode baking emissions, the formula accounts for
30 packing coke consumption, the sulfur and ash content of the packing coke, as well as the pitch content and weight of
31 baked anodes produced. For Soderberg cells, the process formula accounts for the weight of paste consumed per
32 metric ton of aluminum produced, and pitch properties, including sulfur, hydrogen, and ash content.
33 Through the VAIP, anode consumption (and some anode impurity) data have been reported for 1990, 2000, 2003,
34 2004, 2005, 2006, 2007, 2008, and 2009. Where available, smelter-specific process data reported under the VAIP
35 were used; however, if the data were incomplete or unavailable, information was supplemented using industry
36 average values recommended by IPCC (2006). Smelter-specific CO2 process data were provided by 18 of the 23
37 operating smelters in 1990 and 2000, by 14 out of 16 operating smelters in 2003 and 2004, 14 out of 15 operating
38 smelters in 2005, 13 out of 14 operating smelters in 2006, 5 out of 14 operating smelters in 2007 and 2008, and 3 out
39 of 13 operating smelters in 2009. For years where CO2 emissions data or CO2 process data were not reported by
40 these companies, estimates were developed through linear interpolation, and/or assuming representative (e.g.,
41 previously reported or industry default) values.
42 In the absence of any previous historical smelter specific process data (i.e., 1 out of 13 smelters in 2009; 1 out of 14
43 smelters in 2006, 2007, and 2008; 1 out of 15 smelters in 2005; and 5 out of 23 smelters between 1990 and 2003),
44 CO2 emission estimates were estimated using Tier 1 Soderberg and/or Prebake emission factors (metric ton of CO2
45 per metric ton of aluminum produced) from IPCC (2006).
46 Process PFC Emissions from Anode Effects
47 Smelter-specific PFC emissions from aluminum production for 2010 through 2016 were reported to EPA under its
48 GHGRP. To estimate their PFC emissions and report them under EPA's GHGRP, smelters use an approach identical
4-82 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
to the Tier 3 approach in the 2006IPCC Guidelines (IPCC 2006). Specifically, they use a smelter-specific slope
coefficient as well as smelter-specific operating data to estimate an emission factor using the following equation:
3
PFC = SxAE
4
5 where,
AE = F XD
6
7
PFC
CF4 or C2F6, kg/MT aluminum
Slope coefficient, PFC/AE
9
10
11
12
8
S
AE
F
D
Anode effect, minutes/cell-day
Anode effect frequency per cell-day
Anode effect duration, minutes
13 They then multiply this emission factor by aluminum production to estimate PFC emissions. All U.S. aluminum
14 smelters are required to report their emissions under EPA's GHGRP.
15 Perfluorocarbon emissions for the years prior to 2010 were estimated using the same equation, but the slope-factor
16 used for some smelters was technology-specific rather than smelter-specific, making the method a Tier 2 rather than
17 a Tier 3 approach for those smelters. Emissions and background data were reported to EPA under the VAIP. For
18 1990 through 2009, smelter-specific slope coefficients were available and were used for smelters representing
19 between 30 and 94 percent of U.S. primary aluminum production. The percentage changed from year to year as
20 some smelters closed or changed hands and as the production at remaining smelters fluctuated. For smelters that did
21 not report smelter-specific slope coefficients, IPCC technology-specific slope coefficients were applied (IPCC
22 2006). The slope coefficients were combined with smelter-specific anode effect data collected by aluminum
23 companies and reported under the VAIP to estimate emission factors over time. For 1990 through 2009, smelter-
24 specific anode effect data were available for smelters representing between 80 and 100 percent of U.S. primary
25 aluminum production. Where smelter-specific anode effect data were not available, representative values (e.g.,
26 previously reported or industry averages) were used.
27 For all smelters, emission factors were multiplied by annual production to estimate annual emissions at the smelter
28 level. For 1990 through 2009, smelter-specific production data were available for smelters representing between 30
29 and 100 percent of U.S. primary aluminum production. (For the years after 2000, this percentage was near the high
30 end of the range.) Production at non-reporting smelters was estimated by calculating the difference between the
31 production reported under VAIP and the total U.S. production supplied by USGS or USAA, and then allocating this
32 difference to non-reporting smelters in proportion to their production capacity. Emissions were then aggregated
33 across smelters to estimate national emissions.
34 Between 1990 and 2009, production data were provided under the VAIP by 21 of the 23 U.S. smelters that operated
35 during at least part of that period. For the non-reporting smelters, production was estimated based on the difference
36 between reporting smelters and national aluminum production levels (USGS and USAA 1990 through 2009), with
37 allocation to specific smelters based on reported production capacities (USGS 1990 through 2009).
38 National primary aluminum production data for 2016 were obtained via USAA (USAA 2017). For 1990 through
39 2001, and 2006 (see Table 4-81) data were obtained from USGS Mineral Industry Surveys: Aluminum Annual
40 Report (USGS 1995, 1998, 2000, 2001, 2002, 2007). For 2002 through 2005, and 2007 through 2015, national
41 aluminum production data were obtained from the US AA's Primary Aluminum Statistics (US AA 2004 through
42 2006, 2008 through 2016b).
43 Table 4-81: Production of Primary Aluminum (kt)
Year kt
1990 4,048
2005 2.478
2012 2,070
2013 1,948
Industrial Processes and Product Use 4-83
-------
2014 1,710
2015 1,587
2016 818
1 Uncertainty and Time-Series Consistency
2 Uncertainty was assigned to the CO2, CF4, and C2F6 emission values reported by each individual facility to EPA's
3 GHGRP. As previously mentioned, the methods for estimating emissions for EPA's GHGRP and this report are the
4 same, and follow the 2006IPCC Guidelines methodology. As a result, it was possible to assign uncertainty bounds
5 (and distributions) based on an analysis of the uncertainty associated with the facility-specific emissions estimated
6 for previous Inventory years. Uncertainty surrounding the reported CO2, CF4, and C2F6 emission values were
7 determined to have a normal distribution with uncertainty ranges of ±6, ±16, and ±20 percent, respectively. A Monte
8 Carlo analysis was applied to estimate the overall uncertainty of the CO2, CF4, and C2F6 emission estimates for the
9 U.S. aluminum industry as a whole, and the results are provided below.
10 The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 4-82. Aluminum
11 production-related CO2 emissions were estimated to be between 1.3 and 1.4 MMT CO2 Eq. at the 95 percent
12 confidence level. This indicates a range of approximately 3 percent below to 2 percent above the emission estimate
13 of 1.3 MMT CO2 Eq. Also, production-related CF4 emissions were estimated to be between 0.8 and 1.0 MMT CO2
14 Eq. at the 95 percent confidence level. This indicates a range of approximately 10 percent below to 10 percent above
15 the emission estimate of 0.9 MMT CO2 Eq. Finally, aluminum production-related C2F6 emissions were estimated to
16 be between 0.4 and 0.5 MMT CO2 Eq. at the 95 percent confidence level. This indicates a range of approximately 17
17 percent below to 17 percent above the emission estimate of 0.4 MMT CO2 Eq.
18 Table 4-82: Approach 2 Quantitative Uncertainty Estimates for CO2 and PFC Emissions from
19 Aluminum Production (MMT CO2 Eq. and Percent)
Source
Gas
2016 Emission Estimate
(MMT CO2 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT CO2 Eq.) (%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Aluminum Production
CO2
1.3
1.3
1.4
-3%
+2%
Aluminum Production
CF4
0.9
0.8
1.0
-10%
+10%
Aluminum Production
C2F6
0.4
0.4
0.5
-17%
+17%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
20
21 Methodological approaches were applied to the entire time-series to ensure time-series consistency from 1990
22 through 2016. Details on the emission trends through time are described in more detail in the Methodology section
23 above.
24 For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
25 Chapter 6 of the 2006 IPCC Guidelines, see QA/QC and Verification Procedures section in the introduction of the
26 IPPU Chapter.
27 4.20 Magnesium Production and Processing
28 (CRF Source Category 2C4)
29 The magnesium metal production and casting industry uses sulfur hexafluoride (SF6) as a cover gas to prevent the
30 rapid oxidation of molten magnesium in the presence of air. Sulfur hexafluoride lias been used in this application
31 around the world for more than thirty years. A dilute gaseous mixture of SF6 with dry air and/or carbon dioxide
32 (CO2) is blown over molten magnesium metal to induce and stabilize the formation of a protective crust. A small
33 portion of the SF6 reacts with the magnesium to form a thin molecular film of mostly magnesium oxide and
4-84 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
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.
The magnesium industry emitted 1.0 MMT CO2 Eq. (0.04 kt) of SFg, 0.1 MMT CO2 Eq. (0.07 kt) of HFC-134a, and
0.003 MMT CO2 Eq. (2.7 kt) of CO2 in 2016. This represents an increase of approximately 12 percent from total
2015 emissions (see Table 4-83). The increase can be attributed to an increase in secondary and die casting SF6
emissions between 2015 and 2016 as reported through EPA's GHGRP. In 2016, SF6 emissions increased by 12
percent. The increase in SF6 emissions is likely due in part to increased secondary production from reporting
facilities in 2016. In 2016, total HFC-134a emissions increased from 0.09 MMT CO2 Eq. to 0.10 MMT CO2 Eq., or
a 5 percent increase as compared to 2015 emissions. This is mainly attributable to the increased use of this
alternative for primary production. FK 5-1-12 emissions did not change substantially from 2015 levels. The
emissions of the carrier gas, CO2, increased from 2.6 kt in 2015 to 2.7 kt in 2016.
Table 4-83: SF6, HFC-134a, FK 5-1-12 and CO2 Emissions from Magnesium Production and
Processing (MMT CO2 Eq.)
Year
1990
2005
2012
2013
2014
2015
2016
SFe
5.2
2.7
1.6
1.5
1.0
0.9
1.0
HFC-134a
0.0
0.0
+
0.1
0.1
0.1
0.1
CO2
+
+
+
+
+
+
+
FK 5-1-12"
0.0
0.0
+
+
+
+
+
Total
5.2
2.7
1.7
1.5
1.1
1.0
1.1
+ Does not exceed 0.05 MMTCO2 Eq.
a Emissions of FK 5-1-12 are not included in totals.
Table 4-84: SF6, HFC-134a, FK 5-1-12 and CO2 Emissions from Magnesium Production and
Processing (kt)
Year
19')0
2005
2012
2013
2014
2015
2016
SFe
0.2
0.1
0.1
0.1
+
+
+
HFC-134a
0.0
0.0
+
0.1
0.1
0.1
0.1
CO2
1.4
2.9
2.3
2.1
2.3
2.6
2.7
FK 5-1-12"
0.0
0.0
+
+
+
+
+
+ Does not exceed 0.5 kt.
a Emissions of FK 5-1-12 are not included in totals.
Methodology
Emission estimates for the magnesium industry incorporate information provided by industry participants in EPA's
SF6 Emission Reduction Partnership for the Magnesium Industry as well as emissions data reported through subpart
T (Magnesium Production and Processing) of the 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 also reported their HFC-134a and FK 5-1-12 usage, which is also
assumed to be equal to emissions. The last reporting year was 2010 under the Partnership. Emissions data for 2011
through 2016 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 2016 (EPA GHGRP). The
methodologies described below also make use of magnesium production data published by the U.S. Geological
Survey (USGS).
Industrial Processes and Product Use 4-85
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
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 SF6 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 is reported in the RAND survey of major SF6 manufacturers, which shows
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 4-83. These emission factors for the other processes (i.e., permanent mold, wrought, and anode
casting) were based on discussions with industry representatives.
The quantities of CO2 carrier gas used for each production type have been estimated using the 1999 estimated CO2
emissions data and the annual calculated rate of change of SF6 use in the 1990 through 1999 time period. For each
year and production type, the rate of change of SF6 use between the current year and the subsequent year was first
estimated. This rate of change is then applied to the CO2 emissions of the subsequent year to determine the CO2
emission of the current year. The emissions of carrier gases for permanent mold, wrought and anode processes are
not estimated in this Inventory.
1999 through 2010
The 1999 through 2010 emissions from primary and secondary production are 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 are 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 and from 2008 through 2010, Partners did not account for all die casting tracked by USGS,
and, therefore, it was necessary to estimate the emissions of die casters who were not Partners. 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. For 2008 through 2010, the characteristics of the die casters who were not
Partners were not well known, and therefore the emission factor for these die casters was set equal to 3.0 kg SF6 per
metric ton of magnesium, the average of the emission factors reported over the same period by the die casters who
were Partners.
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 1999-2001, the sand
casting emission factor was held constant at the 2002 Partner-reported level. For 2007 through 2010, the sand
4-86 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 casting Partner did not report and the reported emission factor from 2005 was applied to the Partner and to all other
2 sand casters. Activity data for 2005 was obtained from USGS (USGS 2005b).
3 The emission factors for primary production, secondary production and sand casting for the 1999 to 2010 are not
4 published to protect company-specific production information. However, the emission factor for primary production
5 has not risen above the average 1995 Partner value of 1.1 kg SF6 per metric ton. The emission factors for the other
6 industry sectors (i.e., permanent mold, wrought, and anode casting) were based on discussions with industry
7 representatives. The emission factors for casting activities are provided below in Table 4-85.
8 The emissions of HFC-134a and FK-5-1-12 were included in the estimates for only instances where Partners
9 reported that information to the Partnership. Emissions of these alternative cover gases were not estimated for
10 instances where emissions were not reported.
11 Carbon dioxide carrier gas emissions were estimated using the emission factors developed based on GHGRP -
12 reported carrier gas and cover gas data, by production type. It was assumed that the use of carrier gas, by production
13 type, is proportional to the use of cover gases. Therefore, an emission factor, in kg CO2 per kg cover gas and
14 weighted by the cover gases used, was developed for each of the production types. GHGRP data on which these
15 emissions factors are based was available for primary, secondary, die casting and sand casting. The emission factors
16 were applied to the total quantity of all cover gases used (SF6, HFC-134a, and FK-5-1-12) by production type in this
17 time period. Carrier gas emissions for the 1999 through 2010 time period were only estimated for those Partner
18 companies that reported using CO2 as a carrier gas through the GHGRP. Using this approach helped ensure time-
19 series consistency. The emissions of carrier gases for permanent mold, wrought and anode processes are not
20 estimated in this Inventory.
21 Table 4-85: SF6 Emission Factors (kg SF6 per metric ton of magnesium)
Year Die Casting" 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
2.30
2
1 1
2010
2.94
2
1 1
a Weighted average includes all die casters, Partners and non-Partners. For
the majority of the time series (2000-2007), 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.
22 2011 through 2016
23 For 2011 through 2016, for the primary and secondary producers, GHGRP-reported cover and carrier gases
24 emissions data were used. For die and sand casting, some emissions data was obtained through EPA's GHGRP.
25 The balance of the emissions for these industry segments were estimated based on previous Partner reporting (i.e.,
26 for Partners that did not report emissions through EPA's GHGRP) or were estimated by multiplying emission
27 factors by the amount of metal produced or consumed. Partners who did not report through EPA's GHGRP were
28 assumed to have continued to emit SF6 at the last reported level, which was from 2010 in most cases, unless
29 publically available sources indicated that these facilities have closed or otherwise eliminated SF6 emissions from
30 magnesium production (ARB 2015). All Partners were assumed to have continued to consume magnesium at the
31 last reported level. Where the total metal consumption estimated for the Partners fell below the U.S. total reported
32 by USGS, the difference was multiplied by the emission factors discussed in the section above, i.e. non-partner
33 emission factors. For the other types of production and processing (i.e., permanent mold, wrought, and anode
Industrial Processes and Product Use 4-87
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
casting), emissions were estimated by multiplying the industry emission factors with the metal production or
consumption statistics obtained from USGS (USGS 2016). USGS data for 2016 was not yet available at the time of
the analysis, so the 2015 values were held constant through 2016 as a proxy.
Uncertainty and Time-Serii insistency
Uncertainty surrounding the total estimated emissions in 2016 is attributed to the uncertainties around SF6, HFC-
134a, and CO2 emission estimates. To estimate the uncertainty surrounding the estimated 2016 SF6 emissions from
magnesium production and processing, the uncertainties associated with three variables were estimated: (1)
emissions reported by magnesium producers and processors for 2016 through EPA's GHGRP, (2) emissions
estimated for magnesium producers and processors that reported via the Partnership in prior years but did not report
2016 emissions through EPA's GHGRP, and (3) emissions estimated for magnesium producers and processors that
did not participate in the Partnership or report through EPA's GHGRP. An uncertainty of 5 percent was assigned to
the emissions (usage) data reported by each GHGRP reporter for all the cover and carrier gases (per the 2006IPCC
Guidelines). If facilities did not report emissions data during the current reporting year through EPA's GHGRP
reporting program, SF6 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. Alternate cover gas and carrier gases data was set equal to zero if the facilities did not report via the
GHGRP program. 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 4-86). 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 4-86. Total emissions
associated with magnesium production and processing were estimated to be between 1.1 and 1.2 MMT CO2 Eq. at
the 95 percent confidence level. This indicates a range of approximately 5 percent below to 5 percent above the
2016 emission estimate of 1.1 MMT CO2 Eq. The uncertainty estimates for 2016 are smaller relative to the
uncertainty reported for 2015 in the previous Inventory. In the previous Inventory, the emissions factor of sand
casting had a significant impact on the uncertainty because of relatively high uncertainty from the facility that was
not reporting under EPA's GHGRP. This year, the sand casting facility was confirmed to have ceased emissions of
SF6, lowering the uncertainty bounds on the total emission estimate.
Table 4-86: Approach 2 Quantitative Uncertainty Estimates for SFe, HFC-134a and CO2
Emissions from Magnesium Production and Processing (MMT CO2 Eq. and Percent)
Source
Gas
2016 Emission Estimate
(MMT CO2 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT CO2 Eq.) (%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Magnesium
Production
SFe, HFC-
134a, CO2
1.1
1.1 1.2
-5% +5%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
Methodological approaches were applied to the entire time series to ensure time-series consistency from 1990
through 2016. Details on the emission trends through time are described in more detail in the Methodology section,
above.
4-88 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
2 Chapter 6 of the 2006IPCC Guidelines, see QA/QC and Verification Procedures section in the introduction of the
3 IPPU Chapter.
4 Recalculations Discussion
5 Estimates for SF6 emissions from secondary production for 2014 and 2015 were slightly adjusted to correct for a
6 mathematical error. This resulted in a slight increase in SF6 emissions for both 2014 and 2015.
7 Research on former partner facilities showed that one facility whose SF6 emissions were previously held constant at
8 2010 levels stopped emitting SF6 at the beginning of 2015. This facility's 2015 SF6 emissions estimates were
9 adjusted to correct for this change, slightly decreasing overall SF6 emissions for 2015.
10 Planned Improvements
11 Cover gas research conducted over the last decade has found that SF6 used for magnesium melt protection can have
12 degradation rates on the order of 20 percent in die casting applications (Bartos et al. 2007). Current emission
13 estimates assume (per the 2006IPCC Guidelines) that all SF6 utilized is emitted to the atmosphere. Additional
14 research may lead to a revision of the 2006 IPCC Guidelines to reflect this phenomenon and until such time,
15 developments in this sector will be monitored for possible application to the Inventory methodology.
16 Usage and emission details of carrier gases in permanent mold, wrought and anode processes will be researched as
17 part of a future inventory. Based on this research, it will be determined if CO2 carrier gas emissions are to be
18 estimated.
19 Additional emissions are generated as byproducts from the use of alternate cover gases, which are not currently
20 accounted for. Research on this topic is developing, and as reliable emission factors become available, these
21 emissions will be incorporated into the Inventory.
22 4.21 Lead Production (CRF Source Category
23 2C5)
24 In 2016, lead was produced in the United States only using secondary production processes. Until 2014, both lead
25 production in the United States involved both primary and secondary processes—both of which emit carbon dioxide
26 (CO2) (Sjardin 2003). Emissions from fuels consumed for energy purposes during the production of lead are
27 accounted for in the Energy chapter.
28 Primary production of lead through the direct smelting of lead concentrate produces CO2 emissions as the lead
29 concentrates are reduced in a furnace using metallurgical coke (Sjardin 2003). Primary lead production, in the form
30 of direct smelting, previously occurred at a single smelter in Missouri. This primary lead smelter was closed at the
31 end of 2013. In 2014, the smelter processed a small amount of residual lead during demolition of the site (USGS
32 2015) and in 2016 the smelter processed no lead (USGS 2016, 2017).
33 Similar to primary lead production, CO2 emissions from secondary lead production result when a reducing agent,
34 usually metallurgical coke, is added to the smelter to aid in the reduction process. Carbon dioxide emissions from
35 secondary production also occur through the treatment of secondary raw materials (Sjardin 2003). Secondary
36 production primarily involves the recycling of lead acid batteries and post-consumer scrap at secondary smelters. Of
37 all the domestic secondary smelters operational in 2016, 11 smelters had capacities of 30,000 tons or more and were
38 collectively responsible for more than 95 percent of secondary lead production in 2016 (USGS 2017). Secondary
39 lead production has increased in the United States over the past decade while primary lead production has decreased
40 to production levels of zero. In 2016, secondary lead production accounted for 100 percent of total lead production.
41 The lead-acid battery industry accounted for about 85 percent of the reported U.S. lead consumption in 2016 (USGS
42 2017).
Industrial Processes and Product Use 4-89
-------
1 In 2016, total secondary lead production in the United States was slightly higher than that in 2015. A new secondary
2 lead refinery, located in Nevada, was completed in 2016 and production was expected to begin by the end of the
3 year. The plant was expected to produce about 80 tons per day of high-purity refined lead for use in advanced lead-
4 acid batteries using an electromechanical battery recycling technology system. The United States has become more
5 reliant on imported refined lead in recent years owing to the closure of the last primary lead smelter in 2013, and to
6 an increase in exports of spent starting-lighting-ignition lead-acid batteries that reduced the availability of scrap for
7 secondary smelters (USGS 2017).
8 As in 2015, U.S. primary lead production remained at production levels of zero for 2016, and has also decreased by
9 100 percent since 1990. This is due to the closure of the only domestic primary lead smelter in 2013 (year-end), as
10 stated previously. In 2016, U.S. secondary lead production increased from 2015 levels (increase of 2 percent), and
11 has increased by 16 percent since 1990 (USGS 1995 through 2017).
12 In 2016, U.S. primary and secondary lead production totaled 1,070,000 metric tons (USGS 2017). The resulting
13 emissions of CO2 from 2016 lead production were estimated to be 0.5 MMT CO2 Eq. (482 kt) (see Table 4-87). All
14 2016 lead production is from secondary processes, which accounted for 100 percent of total 2016 CO2 emissions
15 from lead production. At last reporting, the United States was the third largest mine producer of lead in the world,
16 behind China and Australia, accounting for approximately 7 percent of world production in 2016 (USGS 2017).
17 Table 4-87: CO2 Emissions from Lead Production (MMT CO2 Eq. and kt)
Year
MMT CO2 Eq.
kt
1990
0.5
516
2005
0.6
2012 0.5 527
2013 0.5 546
2014 0.5 459
2015 0.5 473
2016 0.5 482
18 After a steady increase in total emissions from 1995 to 2000, total emissions have gradually decreased since 2000
19 and are currently 7 percent lower than 1990 levels.
20 Methodology
21 The methods used to estimate emissions for lead production56 are based on Sjardin's work (Sjardin 2003) for lead
22 production emissions and Tier 1 methods from the 2006IPCC Guidelines. The Tier 1 equation is as follows:
23 C02 Emissions = (DS x EFDS) + (5 x EFS)
24 where,
25 DS = Lead produced by direct smelting, metric ton
26 S Lead produced from secondary materials
27 EFds = Emission factor for direct Smelting, metric tons CCh/metric ton lead product
28 EFS = Emission factor for secondary materials, metric tons CCh/metric ton lead product
29 For primary lead production using direct smelting, Sjardin (2003) and the IPCC (2006) provide an emission factor of
30 0.25 metric tons CCVmetric ton lead. For secondary lead production, Sjardin (2003) and IPCC (2006) provide an
31 emission factor of 0.25 metric tons CCh/metric ton lead for direct smelting, as well as an emission factor of 0.2
32 metric tons CCh/metric ton lead produced for the treatment of secondary raw materials (i.e., pretreatment of lead
56 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.
4-90 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 acid batteries). Since the secondary production of lead involves both the use of the direct smelting process and the
2 treatment of secondary raw materials, Sjardin recommends an additive emission factor to be used in conjunction
3 with the secondary lead production quantity. The direct smelting factor (0.25) and the sum of the direct smelting and
4 pretreatment emission factors (0.45) are multiplied by total U.S. primary and secondary lead production,
5 respectively, to estimate CO2 emissions.
6 The production and use of coking coal for lead production is adjusted for within the Energy chapter as this fuel was
7 consumed during non-energy related activities. Additional information on the adjustments made within the Energy
8 sector for Non-Energy Use of Fuels is described in both the Methodology section of CO2 from Fossil Fuel
9 Combustion (3.1 Fossil Fuel Combustion (CRF Source Category 1A)) and Annex 2.1, Methodology for Estimating
10 Emissions of CO2 from Fossil Fuel Combustion.
11 The 1990 through 2016 activity data for primary and secondary lead production (see Table 4-88) were obtained from
12 the U.S. Geological Survey (USGS 1995 through 2017).
13 Table 4-88: Lead Production (Metric Tons)
Year
Primary
Secondary
1990
404,000
922,000
2005
143,000
1,150.000
2012
111,000
1,110,000
2013
114,000
1,150,000
2014
1,000
1,020,000
2015
0
1,050,000
2016
0
1,070,000
14 Uncertainty and Time-Series Consistency
15 Uncertainty associated with lead production relates to the emission factors and activity data used. The direct
16 smelting emission factor used in primary production is taken from Sjardin (2003) who averaged the values provided
17 by three other studies (Dutrizac et al. 2000; Morris et al. 1983; Ullman 1997). For secondary production, Sjardin
18 (2003) added a CO2 emission factor associated with battery treatment. The applicability of these emission factors to
19 plants in the United States is uncertain. There is also a smaller level of uncertainty associated with the accuracy of
20 primary and secondary production data provided by the USGS which is collected via voluntary surveys; the
21 uncertainty of the activity data is a function of the reliability of reported plant-level production data and the
22 completeness of the survey response.
23 The results of the \pproach 2 i|iianiiiali\ e iiiiccriniiits aiials sis are snnininii/cd 111 I able 4-XlJ I .cad production ('()
24 emissions in 2<> I (> were esimialed in he between 0 4 and <> <> \1\11 ('() I al I lie l>5 percent confidence le\el This
25 indicates a raimc of appio\iniatcl\ 15 percent hclou and l<> percent aho\c the emission estimate ofn 5 \ 1 \ 1T CO
26 I :i|
27 Table 4-89: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Lead
28 Production (MMT CO2 Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY REPORT
Si nine
(¦as
2016 1"missiiiii r.siimale
(MMT CO: !¦:«.)
I iHiTlaiim kan^e kelaliu'In Kmissiun l!slimale:l
(MMT CO: ("..)
1.1 HUT I |)|KT I.I HUT I |>|KT
liiiund Bound limind limind
Lead l'roduc
lion CO:
()
0 4 0 6 -1^% + 1 6%
Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
Industrial Processes and Product Use 4-91
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Methodological approaches discussed below were applied to applicable years to ensure time-series consistency in
emissions from 1990 through 2016. Details on the emission trends through time are described in more detail in the
Methodology section, above.
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 QA/QC and Verification Procedures section in the introduction of the
IPPU chapter.
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.57 EPA is still
reviewing available GHGRP data and assessing the possibility of including this planned improvement in future
Inventory reports.
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 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.
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 metallurgical, hydrometallurgical,
and pyrometallurgical processes. Worldwide primary zinc production also employs a pyrometallurgical process
using the Imperial Smelting Furnace process; however, this process is not used in the United States (Sjardin 2003).
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)
57 See .
4-92 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
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).
The only companies in the United States that use emissive technology to produce secondary zinc products are
American Zinc Recycling (AZR) (formerly "Horsehead Corporation"), PIZO, and Steel Dust Recycling (SDR). For
AZR, EAF dust is recycled in Waelz kilns at their Calumet, IL; Palmerton, PA; Rockwood, TN; and Barnwell, SC
facilities. These Waelz kiln 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 shut down their Monaca smelter. This was replaced by
their new facility in Mooresboro, NC. The new Mooresboro facility uses a hydro metallurgical process (i.e., solvent
extraction with electrowinning technology) to produce zinc products. The current capacity of the new facility is
155,000 short tons, with plans to expand to 170,000 short tons per year. Direct consumption of coal, coke, and
natural gas have been replaced with electricity consumption at the new Mooresboro facility. The new facility is
reported to have a 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 up on these surfaces, the cathodes are periodically harvested in
order to strip the zinc from their surfaces (Horsehead 2015). Hydrometallurgical production processes are assumed
to be non-emissive since no carbon is used in these processes (Sjardin 2003).
PIZO and SDR recycle EAF dust into intermediate zinc products using Waelz kilns, and then sell the intermediate
products to companies who smelt it into refined products.
In 2016, United States primary and secondary refined zinc production were estimated to total 140,000 metric tons
(USGS 2017) (see Table 4-90). Domestic zinc mine production decreased by 5 percent in 2016, owing mostly to a
decrease in production in Tennessee; in December 2015, the Middle Tennessee Mines (50,000-ton-per-year
capacity) were closed in response to low zinc prices at the time. Refined zinc production decreased by 19 percent as
a result of a decline in secondary zinc production; in January, the zinc recycling facility in Mooresboro, NC
(140,000-ton-per-year capacity) closed as a result of low zinc prices and ongoing equipment and technical issues
(USGS 2017). Primary zinc production (primary slab zinc) decreased by 11 percent in 2016, while secondary zinc
production in 2016 decreased by 42 percent relative to 2015.
Emissions of CO2 from zinc production in 2016 were estimated to be 0.9 MMT CO2 Eq. (925 kt CO2) (see Table
4-91). All 2016 CO2 emissions resulted from secondary zinc production processes. Emissions from zinc production
in the United States have increased overall since 1990 due to a gradual shift from non-emissive primary production
to emissive secondary production. In 2016, emissions were estimated to be 46 percent higher than they were in
1990.
Table 4-90: Zinc Production (Metric Tons)
Year
Primary
Secondary
Total
1990
262,704
95,708
358,412
2005
191.120
156.000
347,120
2012
114,000
147,000
261,000
2013
106,000
127,000
233,000
2014
110,000
70,000
180,000
Industrial Processes and Product Use 4-93
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
2015 125,000 50,000 172,000
2016 111,000 29,000 140,000
Table 4-91: CO2 Emissions from Zinc Production (MMT CO2 Eq. and kt)
Year
MMT CO2 Eq.
kt
1000
0.6
632
2005
1.0
1.030
2012
1.5
1,486
2013
1.4
1,420
2014
1.0
056
2015
0.0
033
2016
0.0
025
Methodology
The methods used to estimate non-energy CO2 emissions from zinc production58 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:
Eco2 ~ Zn x EFdefaldt
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 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:
1.19 metric tons coke 0.85 metric tons C 3.67 metric tons C02 3.70 metric tons C02
/v / iiir 11(^1? KLIti * * ^ T ^ , , ,
metric tons zinc metric tons coke metric tons C metric tons zinc
58 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.
4-94 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
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:
0.4 metric tons coke
0.85 metric tons C 3.67 metric tons C02
1.24 metric tons C02
EFEAF Dust —
metric tons EAF Dust
x
metric tons coke metric tons C
metric tons EAF Dust
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). Total EAF dust consumed by AZR at their Waelz kilns was not available for
2016 so 2015 data was used as proxy. Consumption levels for 1990 through 2005 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). 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 2016 (SDR 2012, 2014, 2015, and 2017). SDR's facility in Alabama underwent
expansion in 2011 to include a second unit (operational since early- to mid-2012). SDR's facility has been
operational since 2008. 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).
PIZO Technologies Worldwide LLC's facility in Arkansas has been operational since 2009. The amount of EAF
dust consumed by PIZO's facility for 2009 through 2016 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 2016 were estimated by applying the average annual capacity utilization rates for AZR and
SDR (Grupo PROMAX) to PIZO's annual capacity (Horsehead 2012, 2013, 2014, 2015, and 2016; SDR 2012, 2014
and 2017; PIZO 2012, 2014 and 2017). The 1.24 metric tons CO^metric ton EAF dust consumed emission factor
was then applied to PIZO's and SDR's estimated EAF dust consumption to develop CO2 emission estimates for
those Waelz kiln facilities.
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 was replaced by AZR's new facility in Mooresboro, NC. The new
facility uses hydrometallurgical process to produce refined zinc products. This process is assumed to be non-
emissive. Production levels for 1990 through 2004 were extrapolated using the percentage changes in annual refined
zinc production at secondary smelters in the United States as provided by USGS Minerals Yearbook: Zinc (USGS
1995 through 2005). The 3.70 metric tons CO^metric ton zinc emission factor was then applied to the Monaca
facility's production levels to estimate CO2 emissions for the facility. The Waelz kiln production emission factor
was applied in this case rather than the EAF dust consumption emission factor since AZR's Monaca facility did not
consume EAF dust.
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 judgement from the USGS mineral commodity expert to assess approaches for splitting total production into
primary and secondary values. During 2016, only one facility produced primary zinc. Primary zinc produced from
Industrial Processes and Product Use 4-95
-------
1 this facility was subtracted from the USGS 2016 total zinc production statistic to estimate secondary zinc production
2 for 2016.
3 Uncertainty and Time-Series Consistency
4 The uncertainty associated with these estimates is two-fold, relating to activity data and emission factors used.
5 First, there is uncertainty associated with the amount of EAF dust consumed in the United States to produce
6 secondary zinc using emission-intensive Waelz kilns. The estimate for the total amount of EAF dust consumed in
7 Waelz kilns is based on (1) an EAF dust consumption value reported annually by AZR/Horsehead Corporation as
8 part of its financial reporting to the Securities and Exchange Commission (SEC), and (2) an EAF dust consumption
9 value obtained from the Waelz kiln facility operated in Alabama by Steel Dust Recycling LLC. Since actual EAF
10 dust consumption information is not available for PIZO's facility (2009 through 2010) and SDR's facility (2008
11 through 2010), the amount is estimated by multiplying the EAF dust recycling capacity of the facility (available
12 from the company's website) by the capacity utilization factor for AZR (which is available from Horsehead
13 Corporation financial reports). Also, the EAF dust consumption for PIZO's facility for 2011 through 2016 was
14 estimated by multiplying the average capacity utilization factor developed from AZR and SDR's annual capacity
15 utilization rates by PIZO's EAF dust recycling capacity. Therefore, there is uncertainty associated with the
16 assumption used to estimate PIZO and SDR's annual EAF dust consumption values (except SDR's EAF dust
17 consumption for 2011 through 2016, which were obtained from SDR's recycling facility in Alabama).
18 Second, there is uncertainty associated with the emission factors used to estimate CO2 emissions from secondary
19 zinc production processes. The Waelz kiln emission factors are based on materials balances for metallurgical coke
20 and EAF dust consumed as provided by Viklund-White (2000). Therefore, the accuracy of these emission factors
21 depend upon the accuracy of these materials balances. Data limitations prevented the development of emission
22 factors for the electrothermic process. Therefore, emission factors for the Waelz kiln process were applied to both
23 electrothermic and Waelz kiln production processes.
24 The resnlis of Hie \pproach 2 i|ii;inlil;ili\e iiiiceriainis aiials sis are snniniari/ed 111 I able 4-92. Zinc prodiiclioii ('()
25 emissions limn 2<>I(> were esiimaled In he helueen 0 ~ and I I \l\1TCO Ia| al llie 95 perceni confidence le\el
26 This mdicalesa ranue of appro\inialel\ 19 percent helou and 21 perceiii aho\e llie emission esiiniale of t> 9 \1\IT
27 CO \x\
28 Table 4-92: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Zinc
29 Production (MMT CO2 Eq. and Percent) - TO BE UPDATED FOR FINAL INVENTORY REPORT
20Ki riiiiissiiui
Siuinv
(¦;is l'.sliin;ik'
I iuvrl;iini\ Ki-I;iiiw- in lliiiissiiui l.siim;ik''
(MMT CO: l.(|.)
(MM 1 ( (): Ktl.) ("..)
I.I HUT I |)|KT I.I HUT I |)|KT
Bound Bound 1 {iiuiiil Bound
Zinc l'rodn
ction CO: 0.9
0.7 1.1 -19% +21%
Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent conlklence interval.
30 Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
31 through 2016. Details on the emission trends through time are described in more detail in the Methodology section,
32 above.
33 For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
34 Chapter 6 of the 2006IPCC Guidelines, see QA/QC and Verification Procedures section in the introduction of the
35 IPPU chapter.
36 Planned Improvements
37 Pending resources and prioritization of improvements for more significant sources, EPA will continue to evaluate
38 and analyze data reported under EPA's GHGRP that would be useful to improve the emission estimates and
39 category specific QC for the Zinc Production source category, in particular considering completeness of reported
4-96 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 zinc production given the reporting threshold. Given the small number of facilities in the US, particular attention
2 will be made to risks for disclosing CBI and ensuring time series consistency of the emissions estimates presented in
3 future Inventory reports, consistent with IPCC and UNFCCC guidelines. This is required as the facility-level
4 reporting data from EPA's GHGRP, with the program's initial requirements for reporting of emissions in calendar
5 year 2010, are not available for all inventory years (i.e., 1990 through 2009) as required for this Inventory. In
6 implementing improvements and integration of data from EPA's GHGRP, the latest guidance from the IPCC on the
7 use of facility-level data in national inventories will be relied upon.59 EPA is still assessing the possibility of
8 including this planned improvement in future Inventory reports.
9 4.23 Semiconductor Manufacture (CRF Source
10 Category 2E1)
11 The semiconductor industry uses multiple greenhouse gases (GHGs) in its manufacturing processes. These include
12 long-lived fluorinated greenhouse gases used for plasma etching and chamber cleaning, fluorinated heat transfer
13 fluids used for temperature control and other applications, and nitrous oxide (N20) used to produce thin films
14 through chemical vapor deposition.
15 The gases most commonly employed in plasma etching and chamber cleaning are trifluoromethane (HFC-23 or
16 CHF3), perfluoromethane (CF4), perfluoroethane (C2F6), nitrogen trifluoride (NF3), and sulfur hexafluoride (SF6),
17 although other fluorinated compounds such as perfluoropropane (C3F8) and perfluorocyclobutane (c-CiFg) are also
18 used. The exact combination of compounds is specific to the process employed.
19 A single 300 mm silicon wafer that yields between 400 to 600 semiconductor products (devices or chips) may
20 require more than 100 distinct fluorinated-gas-using process steps, principally to deposit and pattern dielectric films.
21 Plasma etching (or patterning) of dielectric films, such as silicon dioxide and silicon nitride, is performed to provide
22 pathways for conducting material to connect individual circuit components in each device. The patterning process
23 uses plasma-generated fluorine atoms, which chemically react with exposed dielectric film to selectively remove the
24 desired portions of the film. The material removed as well as undissociated fluorinated gases flow into waste
25 streams and, unless emission abatement systems are employed, into the atmosphere. Plasma enhanced chemical
26 vapor deposition (PECVD) chambers, used for depositing dielectric films, are cleaned periodically using fluorinated
27 and other gases. During the cleaning cycle the gas is converted to fluorine atoms in plasma, which etches away
28 residual material from chamber walls, electrodes, and chamber hardware. Undissociated fluorinated gases and other
29 products pass from the chamber to waste streams and, unless abatement systems are employed, into the atmosphere.
30 In addition to emissions of unreacted gases, some fluorinated compounds can also be transformed in the plasma
31 processes into different fluorinated compounds which are then exhausted, unless abated, into the atmosphere. For
32 example, when C2F6 is used in cleaning or etching, CF4 is generated and emitted as a process byproduct. In some
33 cases, emissions of the byproduct gas can rival or even exceed emissions of the input gas, as is the case for NF3 used
34 in remote plasma chamber cleaning, which generates CF4 as a byproduct.
35 Besides dielectric film etching and PECVD chamber cleaning, much smaller quantities of fluorinated gases are used
36 to etch polysilicon films and refractory metal films like tungsten.
37 N20 is used in manufacturing semiconductor devices to produce thin films by CVD and nitridation processes as well
38 as for N-doping of compound semiconductors and reaction chamber conditioning (Doering 2000).
39 Liquid perfluorinated compounds are also used as heat transfer fluids (F-HTFs) for temperature control, device
40 testing, cleaning substrate surfaces and other parts, and soldering in certain types of semiconductor manufacturing
41 production processes. Leakage and evaporation of these fluids during use is a source of fluorinated gas emissions
42 (U.S. EPA 2006). Unweighted F-HTF emissions consist primarily of perfluorinated amines, hydrofluoroethers,
59 See .
Industrial Processes and Product Use 4-97
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
perfluoropolyethers, and perfluoroalkylmorpholines. One percent or less consist of HFCs or PFCs (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 GWPs near 10,000.60
For 2016, total GWP-weighted emissions of all fluorinated greenhouse gases and nitrous oxide from deposition,
etching, and chamber cleaning processes in the U.S. semiconductor industry were estimated to be 5.0 MMT CO2 Eq.
These emissions are presented in Table 4-93 and Table 4-94 below for the years 1990, 2005, and the period 2011 to
2016. (Emissions of F-HTFs that are HFCs or PFCs are presented in Table 4-93 and Table 4-94. Emissions of F-
HTFs that are not HFCs or PFCs are presented in Table 4-94, Table 4-95, and Table 4-96 but are not included in
Inventory totals.) The rapid growth of this industry and the increasing complexity (growing number of layers)61 of
semiconductor products led to an increase in emissions of 153 percent between 1990 and 1999, when emissions
peaked at 9.1 MMTCO2 Eq. Emissions began to decline after 1999, reaching a low point in 2009 before rebounding
slightly and plateauing at the current level, which represents a 46 percent decline from 1999 levels. 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 38 percent between 1990 and 2016.
Total emissions from semiconductor manufacture in 2016 were similar to 2015 emissions, increasing less than 1
percent.
Only F-HTF emissions that consist of HFCs or PFCs are included in the Inventory totals; emissions of other F-
HTFs, which account for the vast majority of F-HTF emissions, are provided for informational purposes and are not
included in the Inventory totals. Since reporting of F-HTF emissions began under the GHGRP in 2011, F-HTF
emissions (reported and estimated non-reported) have fluctuated between 0.7 MMT CO2 Eq. and 1.0 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 12 percent and 18 percent of total annual emissions (F-GHG, N20 and F-HTFs) from
semiconductor manufacturing.62 Table 4-96 shows the emissions of the F-HTF compounds with the highest
emissions in tons based on reporting to EPA's GHGRP during years 2011 through 2016.
Table 4-93: PFC, HFC, SFe, NF3, and N2O Emissions from Semiconductor Manufacture63 (MMT
COz Eq.)
Year
19')0
2005
2011
2012
2013
2014
2015
2016
CF4
0.8
1.1
1.4
1.2
1.2
1.5
1.5
1.5
C2F6
2.0
2.0
1.8
1.5
1.4
1.4
1.4
1.2
C3F8
+ /Jl
0.1
0.2
0.1
0.1
0.1
0.1
0.1
C4Fs
0.0
0.1
0.1
0.1
0.1
0.1
0.1
0.1
HFC-23
0.2
0.2
0.2
0.2
0.2
0.3
0.3
0.3
SFe
0.5
0.7
0.4
0.3
0.4
0.7
0.7
0.8
NF3
0.5
0.6
0.6
0.6
0.5
0.6
0.6
Total F-GHGs
3.6
4.6
4.6
4.2
3.9
4.6
4.7
4.7
N2O
0.1
0.2
0.2
0.2
0.2
0.2
0.2
HFC and PFC F-
HTFs
0.0
+
+
+
+
+
+
+
60 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.
61 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.
62 Emissions data for HTFs (in tons of gas) from the semiconductor industry from 2011 through 2015 were obtained from the
EPA GHGRP annual facility emissions reports.
63 An extremely small portion of emissions from Semiconductor Manufacture are from the manufacturing of MEMs and
photovoltaic cells.
4-98 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
Total 3.6 4.7 4.8 4.4 4.0 4.9 4.9 5.0
+ Does not exceed 0.05 MMT CO2 Eq.
Note: Totals may not sum due to independent rounding.
1 Table 4-94: PFC, HFC, SFe, NF3, and N2O Emissions from Semiconductor Manufacture (kt)
Year
1990
2005
2011
2012
2013
2014
2015
2016
CF4
0.11
0.15
0.19
0.17
0.16
0.20
0.21
0.21
C2F6
0.16
0.16
0.14
0.13
0.12
0.11
0.11
0.10
C3F8
+ s?>;
+
+
+
+
+
+
+
C4Fs
0.0
+
+
+
+
+
+
+
HFC-23
+ •.
+
+
+
+
+
+
+
SFe
+ ft?
+
+
+
+
+
+
+
NF3
+
+
+
+
+
+
+
N2O
0.12
0.41
0.80
0.67
0.62
0.73
0.78
0.79
HFC and PFC F-
0.00
+
+
+
+
+
+
+
HTFs
Total
0.43
0.81
1.23
1.05
0.97
1.15
1.21
1.21
+ Does not exceed 0.05 kt.
2 Table 4-95: F-HTF Emissions Based on GHGRP Reporting (MMT CO2 Eq.)
Year
2011
2012
2013
2014
2015
2016
HFCs
PFCs
Other F-HTFs
0.001
0.008
0.984
0.001
0.010
0.945
0.001
0.004
0.668
0.001
0.002
0.797
0.002
0.002
0.763
0.002
0.002
0.651
Total F-HTFs
0.992
0.956
0.673
0.800
0.766
0.655
3 Table 4-96: F-HTF Compounds with Largest GWP-Weighted Emissions Based on GHGRP
4 Reporting (tons of gas)
Fluorinated Heat Transfer
GHGRP-Reported Emissions (tons of gas)
Fluid
GWP
2011
2012
2013
2014
2015
2016
Perfluorotripropylamine
(3M™ FC-3283/FC-8270)
10,000
27.14
38.46
24.95
45.31
37.95
35.76
Perfluoroisopropylmorpholine
(3M™ FC-770)
10,000
12.27
10.07
10.09
11.82
7.94
7.41
PFPMIE fraction, BP 200 °C
(Solvay Galden™ HT-200)
10,000
5.81
6.04
9.49
3.03
7.14
7.08
PFPMIE fraction, BP 165 °C
(Solvay Galden™ D02-TS)
10,000
2.61
2.45
4.89
4.00
3.35
2.58
PFPMIE fraction, BP 170 °C
(Solvay Galden™ HT-170)
10,000
3.37
6.93
0.57
0.90
3.59
2.22
Perfluorotributylamine
(PTBA, 3M™ FC-40/FC-
43)
10,000
10.52
3.40
1.49
2.09
7.50
1.99
PFPMIE fraction, BP 110 °C
(Solvay Galden™ HT-110)
10,000
1.90
1.53
0.87
3.08
1.11
1.43
PFPMIE fraction, BP 135 °C
(Galden™ HT-135)
10,000
1.08
1.23
2.14
1.29
1.43
1.33
PFPMIE fraction, BP 81 °C
(Galden™ DET)
10,000
0.00
0.00
0.46
1.79
1.55
1.09
PFPMIE fraction, BP 203 °C
(Solvay Galden™ D03)
10,000
0.00
0.00
0.00
1.02
0.26
0.55
Industrial Processes and Product Use 4-99
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Additional Emissions from MEMS and PV
Similar to semiconductor manufacturing, the manufacturing of MEMs and photovoltaic cells requires the use of
multiple long-lived fluorinated GHGs for various processes. GHGRP-reported emissions from the manufacturing of
MEMs and photovoltaic cells are available for the years 2011 to 2016. They are not included in the semiconductor
manufacturing totals reported above. The emissions reported by facilities manufacturing MEMs included emissions
of C2F6, C3F8, C4F8, CF4, HFC-23, NF3, and SF6, and were equivalent to only 0.10 percent to 0.42 percent of the
total reported emissions from semiconductor manufacturing in 2011 to 2016. These emissions ranged from 0.0045 to
0.0185 MMT CO2 Eq. from 2011 to 2016. Similarly, emissions from manufacturing of photovoltaic cells were
equivalent to only 0.23 percent and 0.15 percent of the total reported emissions from semiconductor manufacturing
in 2015 and 2016 respectively. Reported emissions from photovoltaic cell manufacturing consisted of CF4, C2F6,
C4F8, and CHF3.
Based upon information in the WFF, it appears that some GHGRP reporters that manufacture both semiconductors
and MEMS are only reporting semiconductor emissions.
Methodology
Emissions are based on data reported through Subpart I, Electronics Manufacture, of EPA's GHGRP, Partner
reported emissions data received through the EPA's PFC64 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),65 and estimates of industry activity (i.e.,
total manufactured layer area). The availability and applicability of reported emissions data from the EPA
Partnership and EPA's GHGRP, and activity data differs across the 1990 through 2016 time series. Consequently,
fluorinated greenhouse gas (F-GHG) emissions from etching and chamber cleaning processes 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 and2016. 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 and2016.
Facility emissions of F-HTFs from semiconductor manufacturing are reported to EPA under its GHGRP, and are
available for the years 2011 through 2016. EPA estimates the emissions of F-HTFs from non-reporting 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 electronics manufacturing (U.S. EPA 2006). For time-series consistency, EPA
interpolated the share of F-HTF emissions to F-GHG emissions between 2000 (at 0 percent) and 2011 (at 22
percent) and applied these shares to the unadjusted F-GHG emissions during those years to estimate the fluorinated
HTF emissions.
1990 through 1994
From 1990 through 1994, Partnership data were unavailable and emissions were modeled using PEVM (Burton and
Beizaie 2001).66 The 1990 to 1994 emissions are assumed to be uncontrolled, since reduction strategies such as
chemical substitution and abatement were yet to be developed.
64 In the context of the EPA Partnership and PEVM, PFC refers to perfluorocompounds, not perfluorocarbons.
65 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.
66 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.
4-100 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 PEVM is based on the recognition that fluorinated greenhouse gas emissions from semiconductor manufacturing
2 vary with: (1) the number of layers that comprise different kinds of semiconductor devices, including both silicon
3 wafer and metal interconnect layers, and (2) silicon consumption (i.e., the area of semiconductors produced) for
4 each kind of device. The product of these two quantities, Total Manufactured Layer Area (TMLA), constitutes the
5 activity data for semiconductor manufacturing. PEVM also incorporates an emission factor that expresses emissions
6 per unit of manufactured layer-area. Emissions are estimated by multiplying TMLA by this emission factor.
7 PEVM incorporates information on the two attributes of semiconductor devices that affect the number of layers: (1)
8 linewidth technology (the smallest manufactured feature size),67 and (2) product type (discrete, memory or logic).68
9 For each linewidth technology, a weighted average number of layers is estimated using VLSI product-specific
10 worldwide silicon demand data in conjunction with complexity factors (i.e., the number of layers per Integrated
11 Circuit (IC) specific to product type (Burton and Beizaie 2001; ITRS 2007). PEVM derives historical consumption
12 of silicon (i.e., square inches) by linewidth technology from published data on annual wafer starts and average wafer
13 size (VLSI Research, Inc. 2012).
14 The emission factor in PEVM is the average of four historical emission factors, each derived by dividing the total
15 annual emissions reported by the Partners for each of the four years between 1996 and 1999 by the total TMLA
16 estimated for the Partners in each of those years. Over this period, the emission factors varied relatively little (i.e.,
17 the relative standard deviation for the average was 5 percent). Since Partners are believed not to have applied
18 significant emission reduction measures before 2000, the resulting average emission factor reflects uncontrolled
19 emissions. The emission factor is used to estimate world uncontrolled emissions using publicly-available data on
20 world silicon consumption.
21 As it was assumed for this time period that there was no consequential adoption of fluorinated-gas-reducing
22 measures, a fixed distribution of fluorinated-gas use was assumed to apply to the entire U.S. industry to estimate
23 gas-specific emissions. This distribution was based upon the average fluorinated-gas purchases made by
24 semiconductor manufacturers during this period and the application of IPCC default emission factors for each gas
25 (Burton and Beizaie 2001).
26 To estimate N20 emissions, it is assumed the proportion of N20 emissions estimated for 1995 (discussed below)
27 remained constant for the period of 1990 through 1994.
28 1995 through 1999
29 For 1995 through 1999, total U.S. emissions were extrapolated from the total annual emissions reported by the
30 Partners (1995 through 1999). Partner-reported emissions are considered more representative (e.g., in terms of
31 capacity utilization in a given year) than PEVM-estimated emissions, and are used to generate total U.S. emissions
32 when applicable. The emissions reported by the Partners were divided by the ratio of the total capacity of the plants
33 operated by the Partners and the total capacity of all of the semiconductor plants in the United States; this ratio
34 represents the share of capacity attributable to the Partnership. This method assumes that Partners and non-Partners
35 have identical capacity utilizations and distributions of manufacturing technologies. Plant capacity data is contained
36 in the World Fab Forecast (WFF) database and its predecessors, which is updated quarterly. Gas-specific emissions
37 were estimated using the same method as for 1990 through 1994.
38 For this time period, the N20 emissions were estimated using an emission factor that was applied to the annual, total
39 U.S. TMLA manufactured. The emission factor was developed using a regression-through-the-origin (RTO) model:
40 GHGRP reported N20 emissions were regressed against the corresponding TMLA of facilities that reported no use
67 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).
68 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 4-101
-------
1 of abatement systems. Details on EPA's GHGRP reported emissions and development of emission factor using the
2 RTO model are presented in the 2011 through 2012 section. The total U.S. TMLA was estimated using PEVM.
3 2000 through 2006
4 Emissions for the years 2000 through 2006—the period during which Partners began the consequential application
5 of fluorinated greenhouse gas-reduction measures—were estimated using a combination of Partner-reported
6 emissions and adjusted PEVM modeled emissions. The emissions reported by Partners for each year were accepted
7 as the quantity emitted from the share of the industry represented by those Partners. Remaining emissions, those
8 from non-Partners, were estimated using PEVM, with one change. To ensure time-series consistency and to reflect
9 the increasing use of remote clean technology (which increases the efficiency of the production process while
10 lowering emissions of fluorinated greenhouse gases), the average non-Partner emission factor (PEVM emission
11 factor) was assumed to begin declining gradually during this period. Specifically, the non-Partner emission factor
12 for each year was determined by linear interpolation, using the end points of 1999 (the original PEVM emission
13 factor) and 2011 (a new emission factor determined for the non-Partner population based on GHGRP-reported data,
14 described below).
15 The portion of the U.S. total emissions attributed to non-Partners is obtained by multiplying PEVM's total U.S.
16 emissions figure by the non-Partner share of U.S. total silicon capacity for each year as described above.69 Gas-
17 specific emissions from non-Partners were estimated using linear interpolation of gas-specific emission distribution
18 of 1999 (assumed same as total U.S. Industry in 1994) and 2011 (calculated from a subset of non-Partner facilities
19 from GHGRP reported emissions data). Annual updates to PEVM reflect published figures for actual silicon
20 consumption from VLSI Research, Inc., revisions and additions to the world population of semiconductor
21 manufacturing plants, and changes in IC fabrication practices within the semiconductor industry (see ITRS 2008 and
22 Semiconductor Equipment and Materials Industry 2011).70>71>72
23 N20 emissions were estimated using the same methodology as the 1995 through 1999 methodology.
24 2007 through 2010
25 For the years 2007 through 2010, emissions were also estimated using a combination of Partner reported emissions
26 and adjusted PEVM modeled emissions to provide estimates for non-Partners; however, two improvements were
27 made to the estimation method employed for the previous years in the time series. First, the 2007 through 2010
69 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.
70 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 2-3 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.
71 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.
72 Two versions of PEVM are used to model non-Partner emissions during this period. For the years 2000 to 2003 PEVM
v3.2.0506.0507 was used to estimate non-Partner emissions. During this time, discrete devices did not use PFCs during
manufacturing and therefore only memory and logic devices were modeled in the PEVM v3.2.0506.0507. From 2004 onwards,
discrete device fabrication started to use PFCs, hence PEVM v4.0.0701.0701, the first version of PEVM to account for PFC
emissions from discrete devices, was used to estimate non-Partner emissions for this time period.
4-102 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
emission estimates 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.73 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 actual utilization data was used to account for differences in fab utilization for
manufacturers of discrete and IC products for 2010 emissions for non-Partners. The Semiconductor Capacity
Utilization (SICAS) Reports from SIA provides the global semiconductor industry capacity and utilization,
differentiated by discrete and IC products (SIA, 2009 through 2011). PEVM estimates were adjusted using
technology-weighted capacity shares that reflect the relative influence of different utilization. Gas-specific emissions
for non-Partners were estimated using the same method as for 2000 through 2006.
N20 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 GaAs technology in addition to Si technology.74 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 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 N20 emissions from CVD and other processes. The F-GHGs and N20 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 PFCs and HFCs
are included in the total emission estimates from semiconductor manufacturing, and these GHGRP-reported
emissions have been compiled and presented in Table 4-93. F-HTF emissions resulting from other types of gases
(e.g., HFEs) are not presented in semiconductor manufacturing totals in Table 4-93 and Table 4-94 but are shown in
Table 4-95 and Table 4-96 for informational purposes.
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 N20 and estimates of manufacturing activity. The new emission factors (in units of
mass of CO2 Eq./TMLA [MSI]) are based on the emissions reported under EPA's GHGRP by facilities without
73 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.
74 GaAs and Si technologies refer to the wafer on which devices are manufactured, which use the same PFCs but in different
ways.
Industrial Processes and Product Use 4-103
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
abatement and on the TMLA estimates for these facilities based on the WFF (SEMI 2012; SEMI 20 13).75 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 or less
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, C4F8,
CHF3, SF6 and NF3)76 were regressed against the corresponding TMLA to estimate an aggregate F-GHG emissions
factor (CO2 Eq./MSI TMLA), and facility-reported N20 emissions were regressed against the corresponding TMLA
to estimate a N20 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 similar subpopulations by wafer size 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
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 N20 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 N20 for 2011, 2012, and
2015 and 2016, resulting in one set of proportions for F-GHGs and one set for N20, 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.
2015 through 2016
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 and 2016, 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
75 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 POU abatement. These fabs were therefore excluded from the regression analysis. (They are still included
in the national totals.)
76 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.
4-104 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
previous years, EPA was able to develop new annual emission factors for 2015 and 2016 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 destruction and removal efficiencies (DREs) were
available. Fab-wide (DREs) represent total fab CO2 Eq.-weighted controlled F-GHG and N20 emissions (emissions
after the use of abatement) divided by total fab CO2 Eq.-weighted uncontrolled F-GHG and N20 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
calculated uncontrolled emissions (each total F-GHG and N20) 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 N20), and wafer size (200 mm 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 N20 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 2b Method in the 2006IPCC Guidelines, this method
uses different emission and byproduct generation factors for different F-GHGs and process types, but it goes beyond
the Tier 2b Method by requiring use of updated factors for different wafer sizes (i.e., 300mm vs. 150 and 200mm)
and CVD clean subtypes (in situ thermal, in situ thermal, 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 (40 CFR Part 98). For the years 2011 through 2013 reported emissions, semiconductor
manufacturers used older emission factors to estimate their F-GHG Emissions (Federal Register / Vol. 75, No. 230
/December 1, 2010, 74829). Subpart I emission factors were updated for 2014 by EPA as a result of a larger set of
emission factor data becoming available as part of the Subpart I petition process, which took place from 2011
through 2013.
Historically, partners estimated and reported their emissions using a range of methods and uneven documentation. It
is assumed that most Partners used a method at least as accurate as the IPCC's Tier 2a Methodology, recommended
in the 2006 IPCC Guidelines. Partners are estimated to have accounted for between 56 and 79 percent of F-GHG
emissions from U.S. semiconductor manufacturing between 1995 and 2010, with the percentage declining in recent
years as Partners increasingly implemented abatement measures.
Estimates of operating plant capacities and characteristics for Partners and non-Partners were derived from the
Semiconductor Equipment and Materials Industry (SEMI) WFF (formerly World Fab Watch) database (1996
through 2012 and 2015) (e.g., Semiconductor Materials and Equipment Industry, 2017). 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 and 2015 were obtained from the
U.S. Census Bureau's Historical Data Quarterly Survey of Plant Capacity Utilization (USCB 2011; 2012; 2015,
2016).
\ qiianiilaliNe iiiicci'iaiun auaKsis ol'this stunce caleuois was performed usnm 1 lie I IH'('-recommended Approach 2
iiiiccriaiun estimation nielhodolous. the Monte ( arlo Stochastic Simulation technique The equation used lo
estimate iiiiccriaiuh is
Industrial Processes and Product Use 4-105
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Tol:il Emissions (lli) = (ilKiKI' l\cp«rU'd f-GIK! Emissions (IT.i .,ii.,) + Non-Ucpoi'lcrs' fslimnU'd I'-CIIK!
I -missions (f-.i-.i i.mi.) + (II KiKI' kcporU'd N ¦() I-missions (i!i-.-. i.) + Non-Kcporlcrs' fslimnU-d N ¦() I-missions
w here IMini I! denote loi;ils lor ihe iiidie;iled snhe;ileuones of emissions lor I"-(¦ 11C ¦;iiid \ (). respeeli\ el>
I lie niiccri;iiiii\ in Ipresented in T;ible-4-l)~ below resiilis from I he eon\ ohiiioii of lour distributions of emissions.
e;ieh relleelnm sep;ir;ile es|ini;iles of possible \ ;ilnes of I! .1! .1! . ;md f I lie ;ippro;ieh ;md
melhods foresiini;iiiim e;ieh disirihiiiion ;md conibiinim I hem lo ;irri\e ;il I he reported l)5 pereeni eonfideiiee mier\;il
(CI) ;ire described in ihe reni;iinder of iliis seelion
I lie niiccri;iiiii\ esiini;iie of I. or (il l( ikl'-reporied f-( il l( i emissions, is de\ eloped bused on u;is-speeil ie
iiiiccrt;iiiii\ esiinuiies of emissions lor mo industry seunienis. one processus 2(>o mm w;ifers ;md one proeessiim
'iio mm \x;ifers I iicertmiiiics in emissions lor e;ieh u;is ;md nidiisir\ seunieni were de\ eloped durum ihe
;issessiiienl of emission es|ini;ilion melhods for ihe snbp;iri I (il l( iRlJ rnlcni;ikiim in 2o 12 (see /r> huii.// Support t<>r
Modifications lo the Fluorinated (ireenhou.se das Emission Estimation Method Option for Semiconductor Facilities
under Subpart I. doekel ff \ IK.) ()\k 2o I I oo2Xi I'lie 2o 12 ;in;il\sis did nol i;ike mlo ;ieeoiml ihe use of
;ib;iienieni for ihe indiisirs seunieni lluil processed 2oo mm w;ifers. esiini;iles of iiiieeri;niiiies ;ii ;i l)5 pereeni ( I
mimed from ±2l) pereeni I'orC I-', lo ±|o pereeni I'orCf for the eorrespoiidiim '()() mm indiisirs seunieni.
esiini;iies of ihe l)5 pereeni CI r;inued from ±'<> pereeni for C f ¦ lo ± l<> pereeni for Cf These u;is ;md w;ifer-
speeilie iiiiceri;iiiil\ es|im;iles ;ire applied lo ihe lol;il emissions of ihe liicililies lluil did nol ;ib;ile emissions ;is
reported under I Vs (il l( ikP
I-'or l hose l;ieililies report iiiu ;ib;iienieni of emissions under f f Vs (il KikP. esiini;iies of iiiieeri;iniiies lor ihe no
;ib;iienieni iiidnsirs segments ;ire modified lo relleel ihe use of full ;ib;iienieni (;ib;iienieni of .///¦:; ises from.///
elemiinu ;md elehinu equipment I ;md p;irti;il ;ib;ilenienl These ;issiimpiions used lo de\ elop iiiiccriniiilics for ihe
p;irii;il ;md lull ;ib;ilcnieiil liieililies ;ire identical for 2oo mm ;md '<>o mm w;ifer proeessinu facilities for;ill
l;ieililies reportiiiu u;is ;ib;ilenienl. ;i iri;nmiil;ir disiribiilion of desirnelioii or renio\ ;il elTicieiicv is ;issnnied for e;ieh
u;is. The iri;iimnl;ir disiribiilions r;iimc from ;m ;is\ mmciric ;md limliK iiiiccrtnin disiribiilion of /ero pereeni
mi in iiiii in lo l)o pereeni m;i\imiim with ~o pereeni mosi I ikelv \;ihielorCf lo ;i s\ ninieirie ;md less iiiiccri;iiii
disiribiilion of S5 pereeni minimum lo l)5 pereeni m;i\imiim w illi')() pereeni mosi likcls \;ilne for (' I-'.. \f . mid
Sf for f;ieililies reporimu p;irii;il ;ib;ilenienl. ihe disiribiilion of fr;ielion of ihe u;is fed lliroiiuh ihe ;ib;ilenienl
dc\ lee. lore;ieh u;is. is ;issiinied lo be iri;iiimil;irl\ disiribnied ;is well II is ;issiinied lh;ii no more llimi 50 pereeni of
ihe u;ises ;ire ;ib;iled (i e . ihe m;i\imiim \ ;ilnei ;md lh;il 5o pereeni is ihe mosi 11 ke I \ \ ;ilne ;ind ihe mini iiiii m is zero
pereeni Coiisider;ilion of ;ib;ilenieni llien resnlled in four ;iddilion;il indiisir\ segments. iwo 2oo-nini w;ifcr-
proeessiiii: segments (one lulls ;md one p;irii:ill\ ;ib;ilnm e;ieh u;isi ;md Iwo '00-nini w;ifer-proeessiim seunienl (one
lulls ;md ihe oilier p;irii;ill\ ;ib;ilnm e;ieh u;isi (i;is-speeilie emission iiiieeri;inilies were es|ini;iled In eon\ ol\ iiiu ihe
disiribiilions of iin;ib;iled emissions w illi ihe ;ippropri;ilc disiribiilion of ;ib;ilenienl elTicieiicv lor lulls ;md p;irii;ill\
;ib;iled facilities iisinu ;i Monlel ( ;irlo simiikilion
I lie niiccri;iiiil\ in f is obl;iined In ;illoc;itiim ihe es|im;iles of miccrimiilics lo ihe lol;il (il l( ikf-reporied
emissions from c;ich of ihe si\ inclnsir\ segments. ;md ilien riiminiu ;i Monie (;irlo simiikiiion w Inch resiilis mi ihe l)5
pereeni CI for emissions from (il l( ikf reporiinu f;ieihiies if i
I lie niiccri;iiiil\ in f is obcimed In iissiimniu lh;il ihe iiiiccri;niil\ in ihe emissions reported b\ e;ieh of the
(ilKikf reporting f;ieihiies resiilis from ihe iiiiccrt;iiiii\ iiH|ii;iiilil\ of \ Oeoiisnmed ;ind ihe \ () emission lliclor
(or nliliAilioii) Siniikir lo ;in;il> ses eonipleled forsiibpnri I (see /r> hnii nl Support tur \ A<,//'//> utiniis in the
77 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.
4-106 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
/¦'htorinalcd (ireenhousv (ias amission Ustimalion leihod Option for Semiconductor I'aeililies under Subpart I.
docket I !l* \ I lo () \k 2d I I i)()2S). ilie unccrtainis of \ () consumed was assumed li» he 2d pereeul
(\>iisiimpikiii nl" \ O l\.ii" (111( ikl' report um facilities was estimated h\ hack - ca leu latum from emissions reported
and assuming iki abatement I lie quantus of N () uiili/ed (the complement of the emission ladon was assumed to
lia\ e a triaimular disiribuiKiu w uli a luiuimum \alue of zero pereeul. mode of 2d pereeul and iii;i\iiiiiini \ alue of 84
pereeul. The minimum was seleeled based mi phssical limitations. ilie mode was sel cquisaleni to ilie subpart I
default \ () in 11 i/al kiii rale lor eliemieal s apor deposition. and I lie ma\imum was set equal lo llie ma\imum
ulili/alioii rale found iu I SMI \uals sis of Niirous (Kide Sur\ cs I )ata (I SMI. 2doi)i The uipuis were used lo
simulate emissions lor eaeli of ilie (il l( iklJ reporium. \ ()-emiiliuu facilities The unccrtainis lor llie loial reported
\ () emissions was ilieu estimated b\ eonibiiinm the uiieeriaiuiies of each of llie laeililies reported emissions usinu
Mouie ( 'arlo simiilaliou.
llie estimate of unccrtainis iu I! and I! entailed des elopum estimales of uiieeriaiuiies for I lie emissions
laclors foreaeh iiou-reporiiuu suh-calcuors and llie corrcspoudiim estimates iif TMI. \
I lie unccrtainis mi l\ 11. \ depends on llie unccrtainis (if two \ anables au esiimale of I lie unccrtainis iu I lie as eraue
annual capacils ulili/alioii foreaeh lescl of produeliou of labs ie u.. full seale or k»V:l) prodiieliom and a
eorrespoudiuu estimate (if llie unccrtainis mi I lie number (if lasers ma uu I'ael 11 red lor hoili \ anables. llie
disirihiiti(ius of capacils uuli/alidiis and number (if maiiiifaelured lasers are assumed iriauuular lor all categories (if
iiou-rcporliim labs. for pr(idueli(iu labs and for I'aeililies dial niaiiufaelure discrete des ices, ilie most probable
111111/aIKi11 is assumed lo be 82 percent, wiili ilie liiuliesi and lowcsi ulili/alioii assumed l(i be 8l) pereeul. and ~d
pereeul. rcspccli\els I lie most pi\ibable \allies lor iiiili/alKin for k«V:l) I'aeililies are assumed lo be 2d pereeul. with
llie lnuliesi iiliIi/aIi(i11 al 'D percent, and llie lowest niili/aliou al Id pereeul for the triangular distributions thai
uos ern the number (if possible lasers manufactured, it is assumed llie most pixihahle \ alue is one laser less than
reported iu llie I I'kS. the smallest number saned hs tcchuolous ucucraliou between one and iwo lasers less than
uiseu iu ilie 11 kSaud laruest number (if lasers corresponded lo the liuiire mseu mi i lie 11 kS
I lie unccrtainis hounds for the as eraue capacils ulili/alioii and the iiumher (if lasers maiiiifaelured are used as
inputs mi a separate Mouie ('arlo simulation to estimate the unccrtainis around the TMI. \ (if hoili tudis idual
facilities as well as ilie u>ial uou-rcporiiim TMI. \ (if each suh-populalioii
I lie unccrtainis around the emission laclors lor each uoii-rcporiiim catcuors (if facilities is dependent on the
unccrtainis (if the total emissions i MM'I' ('() I a| units) and I lie I Ml. \ of each report nm laciliis mi I hat calcuors.
for each subpopulalioii (if reporiiuu facilities, total emissions were repressed on TMI. \ is\ uli au intercept forced lo
zero) for Id.odd emission and Id.odd TMI. \ sallies iu a Mouie Carlo simulation, which results iu Id.odd total
regression coefficients (emission factors) The 2 5ih and the l>~ 5ih percentile (if ihese emission factors are
delermiiied and the hounds are assigned as i lie percent difference from I lie estimated emission factor
I-'or siniphcils. the results of the Mouie ('arlo simulations on the hounds (if the uas- and wafer size-specific
emissions as well as ilie TMI. \ and emission factors are assumed lo he uornialls distributed and the unccrtainis
hounds are assmucd al I standard des lalious around the estimated mean l lie departures from iiormaliis were
(ihsers ed t(i he small.
I lie final step iu esiimaliuu the iiuceriaiuis iu emissions of iioii-rcporiuiu facilities is cons (ils iuu llie disirihiiii(iu (if
emission factors w uli the distribution ol TMI. \ usiuu Monte Carlo simulation.
The results of I lie \pproach 2 c|iiauiiial is e unccrtainis auals sis are suniniari/ed iu Tahle-4-lJ~. x\ liicli is als(i
(ihiaiued bs c(ius(i|s iiiu usinu Mouie Carlo simulation the distributions (if emissions lor each rcportum and nou-
rcportum I'acihis The emissions estimate lor total I S I"-C¦ 11(¦ aikI \ () eniissious from semiconductor
ninuiifacliiriim were estimated lo he helweeu 4 8 and 5 ' MMT ('() I x|. al a l>5 percent c(iiil'ideuce les el. This rauue
represeuis 5 percent below lo 5 percent ahos e llie 2d 15 emission estimate of 5 I MMT CO I This rauue and the
associated percentages appls t(i llie estimate of total eniissious rather than 11 Rise (if mdis idual uases I uccriaiuiics
associated s\ uli mdis idual uases s\ ill he soniew hat liiuher iliau the auureuale. but were ikii c\plicills modeled
Industrial Processes and Product Use 4-107
-------
1 Table-4-97: Approach 2 Quantitative Uncertainty Estimates for HFC, PFC, SFg, NF3 and N2O
2 Emissions from Semiconductor Manufacture (MMT CO2 Eq. and Percent)
Smiriv
Ci;is
21115 lliiiissiiiii
l!slim;iU-
(MM'I'CO: l.(|.)
I ni'iThiiim ki'liiliM'in rimissiiiii
(MM 1 ( (): l!(|.) ("
l!slini;iU"'
I.I HUT I |>|KT I.IHUT
liiiund1' 1 {iiiiihI1' liiiund
I |)|KT
liiiund
Semiconductor
Manufacture
III C. PI C. Sl '„,
Nl;:v and N:()
S ()
4.8 5.3 -5%
Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent conlidence interval.
Absolute lower and upper bounds were calculated using the corresponding lower and upper bounds in percentages.
3 The emissions reported nuclei" IT Vs (il l( ikP fur 2<> 14.20 15 and 2(> l<>. w Inch arc included in I he n\erall eniissimis
4 esi i iiKiies. were h;ised mi ;iii iipd;iled sel nf default cnnssmu faclnrs I his m;i\ h;i\e ;iffeeled I he irend seen hclweeu
5 2<> I ' ;ind 2d 14 (;i 24-perceul mere;ise). w Inch re\ersed 1 he I rend seen helween 2 nf i he li'< '<' > ini
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
In addition, EPA's GHGRP requires the reporting of emissions from other types of electronics manufacturing,
including micro-electro-mechanical systems (MEMs), flat panel displays, and photovoltaic cells. There currently
are seven MEMs manufacturers (most of which report emissions for semiconductor and MEMs manufacturing
separately), one photovoltaic cell manufacturer, and no flat panel displays manufacturing facilities reporting to
EPA's GHGRP. 78 Emissions from MEMs and photovoltaic cell manufacturing could be included in totals in future
Inventory reports - currently they are not represented in Inventory emissions totals for electronics manufacturing.
These emissions could be estimated for the full time series (including prior to the GHGRP) and for MEMS and
photovoltaic cell manufacturers that are not reporting to the GHGRP; however, at this time the contribution to total
emissions is not significant enough to warrant the development of the methodologies that would be necessary to
backcast these emissions to 1990 and estimate emissions for non-reporters for 2011 through 2016.
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.
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 1990.79 Ozone depleting substances—chlorofluorocarbons (CFCs), halons, carbon
tetrachloride, methyl chloroform, and hydrochlorofluorocarbons (HCFCs)—are used in a variety of industrial
applications including refrigeration and air conditioning equipment, solvent cleaning, foam production, sterilization,
fire extinguishing, and aerosols. Although HFCs and PFCs are not harmful to the stratospheric ozone layer, they are
potent greenhouse gases. Emission estimates for HFCs and PFCs used as substitutes for ODSs are provided in Table
4-98 and Table 4-99.
Table 4-98: Emissions of HFCs and PFCs from ODS Substitutes (MMT CO2 Eq.)
Gas 1990 2005 2012 2013 2014 2015 2016
HFC-23
+
+
+
+
+
+
+
HFC-32
+
0.3
4.4
5.4
6.4
7.5
8.5
HFC-125
+
9.5
43.6
49.9
55.9
61.9
67.7
HFC-134a
+
73.4
67.8
62.8
60.8
59.0
55.8
HFC-143a
+
9.4
24.4
26.0
27.3
28.1
28.8
HFC-236fa
+
1.2
1.5
1.5
1.4
1.3
1.2
CF4
+
+
+
+
+
+
+
Others3
0.3
5.9
8.6
9.0
9.5
10.7
11.9
78 Based upon information in the WFF, it appears that a small portion of GHGRP semiconductor reporters are manufacturing
both semiconductors and MEMS; however, these reporters are only reporting semiconductor emissions.
79 [42 U.S.C § 7671, CAA Title VI]
Industrial Processes and Product Use 4-109
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Total 03 99,8 150.4 154.8 161.4 168.6 173.9
+ Does not exceed 0.05 MMT CO2 Eq.
a Others include HFC-152a, HFC-227ea, HFC-245fa, HFC-43-10mee, HCFO-1233zd(E), HFO-1234yf,
HFO-1234ze(E), HFO-1336mzz(Z), C4F10, 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 C6F14.
Note: Totals may not sum due to independent rounding.
Table 4-99: Emissions of HFCs and PFCs from ODS Substitution (Metric Tons)
Gas
1990
2005
2012
2013
2014
2015
2016
HFC-23
+
1
2
2
3
3
3
HFC-32
+
511
6,479
7,985
9,475
11,052
12,623
HFC-125
+
2,703
12,452
14,266
15,981
17,693
19,330
HFC-134a
+
51,314
47,439
43,946
42,532
41,272
39,004
HFC-143a
+
2,110
5,460
5,821
6,096
6,281
6,433
HFC-236fa
+
125
148
151
148
135
127
CF4
+
2
4
4
4
4
4
Others3
M
M
M
M
M
M
M
+ Does not exceed 0.5 MT.
M (Mixture of Gases)
a Others include HFC-152a, HFC-227ea, HFC-245fa, HFC-43-10mee, HCFO-1233zd(E), HFO-1234yf,
HFO-1234ze(E), HFO-1336mzz(Z), C4F10, 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—and HFC-134a in refrigeration end-uses. 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.80 In 1993, the use of HFCs
in foam production began, and in 1994 ODS substitutes for halons entered widespread use in the United States as
halon production was phased-out. In 1995, these compounds also found applications as solvents.
The use and subsequent emissions of HFCs and PFCs as ODS substitutes has been increasing from small amounts in
1990 to 173.9 MMT CO2 Eq. emitted in 2016 This increase was in large part the result of efforts to phase out CFCs
and other ODSs in the United States. In the short term, this trend is expected to continue, and will likely continue
over the next decade as HCFCs, which are interim substitutes in many applications, are themselves phased-out under
the provisions of the Copenhagen Amendments to the Montreal Protocol. Improvements in the technologies
associated with the use of these gases and the introduction of alternative gases and technologies, however, may help
to offset this anticipated increase in emissions.
Table 4-100 presents emissions of HFCs and PFCs as ODS substitutes by end-use sector for 1990 through 2016. The
end-use sectors that contributed the most toward emissions of HFCs and PFCs as ODS substitutes in 2016 include
refrigeration and air-conditioning (149.6 MMT CO2 Eq., or approximately 86 percent), aerosols (10.7 MMT CO2
Eq., or approximately 6 percent), and foams (10.3 MMT CO2 Eq., or approximately 6 percent). Within the
refrigeration and air-conditioning end-use sector, motor vehicle air-conditioning was the highest emitting end-use
(34.9 MMT CO2 Eq.), followed by refrigerated retail food and refrigerated transport. Each of the end-use sectors is
described in more detail below.
Table 4-100: Emissions of HFCs and PFCs from ODS Substitutes (MMT CO2 Eq.) by Sector
Sector
1990
2005
2012
2013
2014
2015
2016
Refrigeration/Air
Conditioning
+
87.8
130.1
133.7
139.4
145.0
149.6
Aerosols
0.3
7.6
10.3
10.5
10.8
11.0
10.7
Foams
+
2.1
6.9
7.5
8.0
9.3
10.3
80 R-404A contains HFC-125, HFC-143a, andHFC-134a.
4-110 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Solvents
Fire Protection
+
+
1.7
0.7
1.7 1.8 1.8 1.8 1.9
1.3 1.3 1.4 1.5 1.5
Total
0.3
99.8
150.4 154.8 161.4 168.6 173.9
+ Does not exceed 0.05 MMT CO2 Eq.
Note: Totals may not sum due to independent rounding.
Refrigeration/Air Conditioning
The refrigeration and air-conditioning sector includes a wide variety of equipment types that have historically used
CFCs or HCFCs. End-uses within this sector include motor vehicle air-conditioning, retail food refrigeration,
refrigerated transport (e.g., ship holds, truck trailers, railway freight cars), household refrigeration, residential and
small commercial air-conditioning and heat pumps, chillers (large comfort cooling), cold storage facilities, and
industrial process refrigeration (e.g., systems used in food processing, chemical, petrochemical, pharmaceutical, oil
and gas, and metallurgical industries). As the ODS phaseout has taken effect, most equipment has been retrofitted
or replaced to use HFC-based substitutes. Common HFCs in use today in refrigeration/air-conditioning equipment
are HFC-134a, R-410A,81 R-404A, and R-507A.82 Lower -GWP options such as 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. These refrigerants are emitted to the atmosphere
during equipment manufacture and operation (as a result of component failure, leaks, and purges), as well as at
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). Many 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 has started to use 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 ODS 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.
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.
81 R-410A contains HFC-32 and HFC-125.
82 R-507A, also called R-507, contains HFC-125 and HFC-143a.
Foams
Industrial Processes and Product Use 4-111
-------
i Solvents
2 Chlorofluorocarbons, methyl chloroform (1,1,1 -trichloroethane or TCA), and to a lesser extent carbon tetrachloride
3 (CCI4) were historically used as solvents in a wide range of cleaning applications, including precision, electronics,
4 and metal cleaning. Since their phaseout, metal cleaning end-use applications have primarily transitioned to non-
5 fluorocarbon solvents and not-in-kind processes. The precision and electronics cleaning end-uses have transitioned
6 in part to high-GWP gases, due to their high reliability, excellent compatibility, good stability, low toxicity, and
7 selective solvency. These applications rely on HFC-43-10mee, HFC-365mfc, HFC-245fa, and to a lesser extent,
8 PFCs. Electronics cleaning involves removing flux residue that remains after a soldering operation for printed
9 circuit boards and other contamination-sensitive electronics applications. Precision cleaning may apply to either
10 electronic components or to metal surfaces, and is characterized by products, such as disk drives, gyroscopes, and
11 optical components, that require a high level of cleanliness and generally have complex shapes, small clearances,
12 and other cleaning challenges. The use of solvents yields fugitive emissions of these HFCs and PFCs.
13 Fire Protection
14 Fire protection applications include portable fire extinguishers ("streaming" applications) that originally used halon
15 1211, and total flooding applications that originally used halon 1301, as well as some halon 2402. Since the
16 production and import of virgin halons were banned in the United States in 1994, the halon replacement agent of
17 choice in the streaming sector has been dry chemical, although HFC-236fa is also used to a limited extent. In the
18 total flooding sector, HFC-227ea has emerged as the primary replacement for halon 1301 in applications that require
19 clean agents. Other HFCs, such as HFC-23 and HFC-125, are used in smaller amounts. The majority of HFC-227ea
20 in total flooding systems is used to protect essential electronics, as well as in civil aviation, military mobile weapons
21 systems, oil/gas/other process industries, and merchant shipping. Fluoroketone FK-5-1-12 is also used as low-GWP
22 option and 2-BTP is being considered. As fire protection equipment is tested or deployed, emissions of these HFCs
23 occur.
24 Methodology
25 A detailed Vintaging Model of ODS-containing equipment and products was used to estimate the actual—versus
26 potential—emissions of various ODS substitutes, including HFCs and PFCs. The name of the model refers to the
27 fact that it tracks the use and emissions of various compounds for the annual "vintages" of new equipment that enter
28 service in each end-use. The Vintaging Model predicts ODS and ODS substitute use in the United States based on
29 modeled estimates of the quantity of equipment or products sold each year containing these chemicals and the
30 amount of the chemical required to manufacture and/or maintain equipment and products over time. Emissions for
31 each end-use were estimated by applying annual leak rates and release profiles, which account for the lag in
32 emissions from equipment as they leak over time. By aggregating the data for 66 different end-uses, the model
33 produces estimates of annual use and emissions of each compound. Further information on the Vintaging Model is
3 4 contained in Annex 3.9.
35 Uncertainty and Time-Series Consistency — TO BE UPDATED
36 FOR FINAL INVENTORY REPORT
37 (1 in en I li;il emissions of ()l )S subsi nines occur from ihoiisands of different kinds of equipment and lioni 111111 ions of
38 poiui and mobile sources throimhoiii ilie I mied Siales. emission esiiniales ninsi he made usinu aiials Ileal lools such
39 as ilie \ iiiiaunm Model or ilie nielhods 0111 lined 111 IIH'(' 12t)u sis quauiifies 1 lie le\el of iiiiccriaiuis associated Willi ilie auureuale emissions across ihc<><>
44 end-uses 111 ilie Yiuiamuu Model lu order lo calculate iiiiccriaiuis. functional forms were de\ eloped lo simplify
45 some of 1 he coniple\ "\ iiiiamuu" aspects of some end-use sectors. especialK w illi respect to rel rmeratioii and air-
46 coiidiliouiiii:. and lo a lesser decree. lire cMiimuishiim These sectors calculate emissions based 0111 lie entire
4-112 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
111'cl i mc of equipment. ik(• "¦> ;iud ISS 4 \I\1T CO I
;il llie K>5 percent cuiifidcucc lex el I Ins indicates ;i niime nl" ;ipprnNiiii;iielx n.'J" perceni helnw in I I ~K> perceni
;ihnxe llie emission cs|ini;ilc nl' I(>S 5 \1\11 ( '() I x|
Table 4-101: Approach 2 Quantitative Uncertainty Estimates for HFC and PFC Emissions
from ODS Substitutes (MMT CO2 Eq. and Percent)
Sim riv
(iilSl'S
21116 I'.niissiiui
l'.slilll;ili-
(MMT CO: I'.ii.)
I niiThiiim k;ui^i' Ki-hiiiu- in Kmissiiui I'siimiiii-'
(MM 1 ( (): Ku.) ("..)
I.I HUT I |)|)IT I.IIXUT I |)|)IT
Bound liiiuiid liiiiiml 1 {iiuihI
11 Range of emission estimates predicted by Monte Carlo Stochastic Simulation lor a 95 percent confidence interval.
\lelliiidiilnmc;il ;ipprii;iclics were applied lo llie enure lime series 10 ensure lime-series consistency from I'wn
llirouuli 2n|(i I)ei;nls 011 llie eniission irends ilirnuuli lime ;ire described 111 more del;nl 111 ihe Melliodoloux secliou.
;ihox e
for more iiifornijiiiou 0111 lie ueueiiil (,).\ (,)(' process applied lo lliis source c;ileuor\. cousisiciii w 11I1 N'olunie I.
Clinpier (> of llie 2unt- ll'i V 11 see o \ (.H' ;iud \'eril'ic;iliou I'rocedures seclion 1111 lie iiiiroducliou of 1 lie
IH'l ( h;ipier.
Comparison of Reported Consumption to Modeled Consumption of HFCs
Data from EPA's Greenhouse Gas Reporting Program (GHGRP) was also used to perform quality control as a
reference scenario check on the modeled emissions from this source category as specified in 2006IPCC Guidelines
for National Greenhouse Gas Inventories. To do so, consumption patterns demonstrated through data reported
under GHGRP Subpart 00—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 (excluding HFC-23) 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 chemicals), destroyed, or exported from the
United States.83 This allows for a quality control check on emissions from this source because the Vintaging Model
uses modeled demand for new chemical as a proxy for total amount supplied, which is similar to net supply, as an
input to the emission calculations in the model.
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
83 Chemical that is exported, transformed, or destroyed—unless otherwise imported back to the United States—will never be
emitted in the United States.
Industrial Processes and Product Use 4-113
-------
1 to EPA in 2012 (for supply that occurred in 2011). Beginning in 2015, bulk consumption data for aggregated HFCs
2 reported under Subpart 00 were made publicly available under EPA's GHGRP. Data include all saturated HFCs
3 (except HFC-23) reported to EPA across the GHGRP-reporting time series (2010 through 2016). The data include
4 all 26 such saturated HFCs listed in Table A-l of 40 CFR Part 98, where regulations for EPA's GHGRP are
5 promulgated, though not all species were reported in each reporting year. For the first time in 2016, net imports of
6 HFCs contained in pre-charged equipment or closed-cell foams reported under Subpart QQ were made publicly
7 available under EPA's GHGRP.
8 Modeled Consumption (VintagingModel Bottom-Up Estimate)
9 The Vintaging Model, used to estimate emissions from this source category, calculates chemical demand based on
10 the quantity of equipment and products sold, serviced and retired each year, and the amount of the chemical required
11 to manufacture and/or maintain the equipment and products.84 It is assumed that the total demand equals the amount
12 supplied by either new production, chemical import, or quantities recovered (usually reclaimed) and placed back on
13 the market. In the Vintaging Model, demand for new chemical, as a proxy for consumption, is calculated as any
14 chemical demand (either for new equipment or for servicing existing equipment) that cannot be met through
15 recycled or recovered material. No distinction is made in the Vintaging Model between whether that need is met
16 through domestic production or imports. To calculate emissions, the Vintaging Model estimates the quantity
17 released from equipment over time. Thus, verifying the Vintaging Model's calculated consumption against GHGRP
18 reported data is one way to check the Vintaging Model's emission estimates.
19 There are ten saturated HFC species modeled in the Vintaging Model: HFC-23, HFC-32, HFC-125, HFC-134a,
20 HFC-143a, HFC-152a, HFC-227ea, HFC-236fa, HFC-245fa, and HFC-43-10mee. For the purposes of this
21 comparison, only nine HFC species are included (HFC-23 is excluded), to more closely align with the aggregated
22 total reported under EPA's GHGRP. While some amounts of less-used saturated HFCs, including isomers of those
23 included in the Vintaging Model, are reportable under EPA's GHGRP, the data are believed to represent an amount
24 comparable to the modeled estimates as a quality control check.
25 Comparison Results and Discussion
26 Comparing the estimates of consumption from these two approaches (i.e., reported and modeled) ultimately supports
27 and improves estimates of emissions, as noted in the 2006IPCC Guidelines (which refer to fluorinated greenhouse
28 gas consumption based on supplies as "potential emissions"):
29 [W]hen considered along with estimates of actual emissions, the potential emissions approach can assist in
30 validation of completeness of sources covered and as a QC check by comparing total domestic
31 consumption as calculated in this 'potential emissions approach' per compound with the sum of all activity
32 data of the various uses (IPCC 2006).
33 Table 4-102 and Figure 4-2 compare the net supply of saturated HFCs (excluding HFC-23) in MMT CO2 Eq. as
34 determined from Subpart OO (industrial greenhouse gas suppliers) and Subpart QQ (supply of HFCs in products) of
35 EPA's GHGRP for the years 2010 through 2016 and the chemical demand as calculated by the Vintaging Model for
36 the same time series. 2016 GHGRP values are not yet publically available and are proxied to 2015 estimates.
37 Table 4-102: U.S. HFC Consumption (MMT COz Eq.)
2010
2011
2012
2013
2014
2015
2016
Reported Net Supply (GHGRP)
235
249
245
295
279
290
290
Industrial GHG Suppliers
235
241
227
278
254
264
264
Imports of HFCs in Products
NA
7
18
17
25
26
26
Modeled Supply (Vintaging Model)
256
256
273
278
282
284
281
Percent Difference
9%
3%
11%
-6%
1%
-2%
-3%
38 NA (Not Available)
39 a Importers and exporters of fluorinated gases in products were not required to report until 2011.
84 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.
4-114 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Figure 4-2: U.S. HFC Consumption (MMT CO2 Eq.)
Modeled Consumption
¦ Reported Imports in Products
¦ Reported Bulk Supply
2010 2011 2012 2013 2014 2015 2016
350
300
250
S 200
o
u
150
100
50
0
As show n, the estimates from the Vintaging Model are generally higher than the GHGRP estimates by an average of
2 percent across the time series (i.e., 2010 through 2016). Potential reasons for these differences include:
• The Vintaging Model includes fewer HFCs than are reported to EPA's GHGRP. 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. Because the GHGRP data generally increases over time (although some year-to-year
variations exist) and the Vintaging Model estimates also increase, EPA would expect the modeled estimates
to be slightly lower than the corresponding GHGRP data due to this temporal effect.
• 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 may not be accounted for in the
GHGRP data. On the other hand, some exports might also not be accounted for in this data.
• In some years, imports and exports may be greater than consumption because the excess is being used to
increase chemical or equipment stockpiles; in other years, the opposite may hold true. Similarly, relocation
of manufacturing facilities or recovery from the recession could contribute to variability in imports or
exports. Averaging imports and exports over multiple years can minimize the impact of such fluctuations.
Industrial Processes and Product Use 4-115
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
For example, when the 2012 and 2013 net additions to the supply are averaged, as shown in Table 4-103,
the percent difference between the consumption estimates decreases compared to the 2012-only estimates.
Table 4-103: Averaged U.S. HFC Demand (MMT COz Eq.)
2010-2011
2011-2012
2012-2013
2013-2014
2014-2015
2015-2016
Avg.
Avg.
Avg.
Avg.
Avg.
Avg.
Reported Net Supply
(GHGRP)
Modeled Demand
(Vintaging Model)
242
256
247
264
270
275
287
280
284
283
290
284
Percent Difference
6%
7%
2%
-2%
0%
-2%
• 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 a slowly increasing overall demand,
actual consumption for specific chemicals or equipment may vary over time and could even switch from
positive to negative (indicating more chemical exported, transformed, or destroyed than produced or
imported in a given year). Furthermore, consumption as calculated in the Vintaging Model is a function of
demand not met by disposal recovery. 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 and recovery, actual supply would decrease, and hence
consumption would increase to meet that demand not satisfied by recovered quantities, increasing the
GHGRP data and bringing those totals closer to the Vintaging Model estimates.
• The Vintaging Model is used to estimate the emissions that occur in the United States. As such, all
equipment or products that contain ODS 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 OO (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 some 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 most years higher than GHGRP data:
• Saturated HFCs are also known to be used as a cover gas in the production of magnesium. The Vintaging
Model estimates here do not include the amount of HFCs for this use, but rather only the amount for uses
that traditionally were served by ODS. Nonetheless, EPA expects that this supply not included in the
Vintaging Model estimates to be very small compared to the ODS substitute use for the years analyzed. An
indication of the different magnitudes of these categories is seen in the fact that the 2016 emissions from
that non-modeled source (0.1 MMT CO2 Eq.) are much smaller than those for the ODS substitute sector
(173.9 MMT C02 Eq.).
Using a Tier 2 bottom-up modeling methodology to estimate emissions requires assumptions and expert judgement.
Comparing the Vintaging Model's estimates to GHGRP-reported estimates, particularly for more widely used
chemicals, can help validate the model but it is expected that the model will have limitations. This comparison
shows that Vintaging Model consumption estimates are well within the same order of magnitude as the actual
consumption data as reported to EPA's GHGRP although the differences in reported net supply and modeled
demand are still significant in some of the years. Although it can be difficult to capture the observed market
variability, the Vintaging Model is periodically reviewed and updated to ensure that the model reflects the current
and future trajectory of ODS and ODS substitutes across all end-uses and the Vintaging Model will continue to be
4-116 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
compared to available top-down estimates in order to ensure the model accurately estimates HFC consumption and
emissions.
Recalculations Discussion
For the current Inventory, a refrigerated food processing and dispensing equipment end-use was added to the
Vintaging Model. Methodological recalculations were applied to the entire time-series to ensure time-series
consistency from 1990 through 2016. The refrigerated food processing and dispensing equipment end-use was added
based on a review of technical reports and sales data. Additionally, the Industrial Process Refrigeration (IPR) end-
use was updated to transition fromHCFC-123 to HCFO-1233zd(E) in response to upcoming HCFC-123 phaseout
requirements under the Montreal Protocol, and the vending machine end-use was updated to transition CO2 systems
to R-290 beginning in 2018 to reflect manufacturer forecasting. Together, these updates increased greenhouse gas
emissions on average by 0.03 percent between 1990 and 2016.
Planned Improvements
Future improvements to the Vintaging Model are planned for the Refrigeration/Air Conditioning and Fire Protection
sectors. End-uses representing medium-duty and heavy-duty vehicle and truck air conditioners may be added to the
refrigeration and air-conditioning sector. In addition, updates are expected in response to a peer review conducted on
end-uses within the Refrigeration/Air Conditioning and Fire Protection sectors.
Finally, EPA received comments from 3M, the Halon Alternatives Research Corporation, and Trakref on the public
review draft of the Substitution of Ozone Depleting Substances category of the previous Inventory. These
comments, concerning the Fire Protection and Refrigeration/Air Conditioning applications, will be considered while
updating the model used to develop the emission estimates from these applications.
4.25 Electrical Transmission and Distribution
(CRF Source Category 2G1)
The largest use of sulfur hexafluoride (SF6), 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.
Fugitive emissions of SF6 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 from equipment manufacturing and from electrical transmission and distribution systems were
estimated to be 4.3 MMT CO2 Eq. (0.2 kt) in 2016. This quantity represents an 81 percent decrease from the
estimate for 1990 (see Table 4-104 and Table 4-105). There are two potential causes for this decrease: a sharp
increase in the price of SF6 during the 1990s and a growing awareness of the magnitude and 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. Utilities participating in the Partnership have lowered their emission
factor (kg SF6 emitted per kg of nameplate capacity) by more than 85 percent since the Partnership began in 1999. 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 26 percent from 2011 to 20 1 6,85 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
85 Analysis of emission trends from the GHGRP is imperfect due to an inconsistent group of reporters year to year.
Industrial Processes and Product Use 4-117
-------
1 hanging fruit," such as replacing major leaking circuit breakers) that Partners have already taken advantage of under
2 the voluntary program (Ottinger et al. 2014).
3 Table 4-104: SF6 Emissions from Electric Power Systems and Electrical Equipment
4 Manufacturers (MMT CO2 Eq.)
Electrical
Electric Power
Equipment
Year
Systems
Manufacturers
Total
1990
22.8
0.3
23.1
2005
0.6
8.3
2011
5.3
0.7
6.0
2012
4.4
0.3
4.6
2013
4.1
0.4
4.5
2014
4.3
0.3
4.6
2015
4.0
0.3
4.2
2016
4.1
0.2
4.3
Note: Totals may not sum due to independent rounding.
5 Table 4-105: SF6 Emissions from Electric Power Systems and Electrical Equipment
6 Manufacturers (kt)
Year Emissions
1990 1.0
2005 0.4
2011 0.3
2012 0.2
2013 0.2
2014 0.2
2015 0.2
2016 0.2
7 Methodology
8 The estimates of emissions from Electrical Transmission and Distribution are comprised of emissions from electric
9 power systems and emissions from the manufacture of electrical equipment. The methodologies for estimating both
10 sets of emissions are described below.
11 1990 through 1998 Emissions from Electric Power Systems
12 Emissions from electric power systems from 1990 through 1998 were estimated based on (1) the emissions
13 estimated for this source category in 1999, which, as discussed in the next section, were based on the emissions
14 reported during the first year of EPA's SF6 Emission Reduction Partnership for Electric Power Systems
15 (Partnership), and (2) the RAND survey of global SF6 emissions. Because most utilities participating in the
16 Partnership reported emissions only for 1999 through 2011, modeling was used to estimate SF6 emissions from
17 electric power systems for the years 1990 through 1998. To perform this modeling, U.S. emissions were assumed to
18 follow the same trajectory as global emissions from this source during the 1990 to 1999 period. To estimate global
19 emissions, the RAND survey of global SF6 sales were used, together with the following equation for estimating
20 emissions, which is derived from the mass-balance equation for chemical emissions (Volume 3, Equation 7.3) in the
4-118 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
2006IPCC Guidelines,86 (Although Equation 7.3 of the 2006IPCC 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.)
Emissions (kilograms SF6) = SF6 purchased to refill existing equipment (kilograms) + nameplate capacity of retiring
equipment (kilograms) 87
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 SF6
emission rates for Europe and Japan for 1995 (IPCC 2006). The 40-year lifetime for electrical equipment is also
based on IPCC (2006). The results of the two components of the above equation were then summed to yield
estimates of global SF6 emissions from 1990 through 1999.
U.S. emissions between 1990 and 1999 are assumed to follow the same trajectory as global emissions during this
period. To estimate U.S. emissions, global emissions for each year from 1990 through 1998 were divided by the
estimated global emissions from 1999. The result was a time series of factors that express each year's global
emissions as a multiple of 1999 global emissions. Historical U.S. emissions were estimated by multiplying the factor
for each respective year by the estimated U.S. emissions of SF6 from electric power systems in 1999 (estimated to be
14.3 MMT C02 Eq.).
Two factors may affect the relationship between the RAND sales trends and actual global emission trends. One is
utilities' inventories of SF6 in storage containers. When SF6 prices rise, utilities are likely to deplete internal
inventories before purchasing new SF6 at the higher price, in which case SF6 sales will fall more quickly than
emissions. On the other hand, when SF6 prices fall, utilities are likely to purchase more SF6 to rebuild inventories, in
which case sales will rise more quickly than emissions. This effect was accounted for by applying 3 -year smoothing
to utility SF6 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 2016 Emissions from Electric Power Systems
Emissions from electric power systems from 1999 to 2016 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 2017 Utility Data Institute (UDI) Directories of Electric Power
Producers and Distributors (UDI 2001, 2004, 2007, 2010, 2013, 2017), which was applied to the electric power
systems that do not report to EPA (Non-Reporters). (Transmission miles are defined as the miles of lines carrying
voltages above 34.5 kV).
Partners
86 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.
87 Nameplate capacity is defined as the amount of SFe within fully charged electrical equipment.
Industrial Processes and Product Use 4-119
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Over the period from 1999 to 2016, Partner utilities, which for inventory purposes are defined as utilities that either
currently are or previously have been part of the Partnership,88 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 lateryears) through EPA's GHGRP (discussed
further below) rather than through the Partnership. In 2016, approximately 0.2 percent of the total emissions
attributed to Partner utilities were reported through Partnership reports. Approximately 92 percent of the total
emissions attributed to Partner utilities were reported and verified through EPA's GHGRP. Partners without verified
2016 data accounted for approximately 8 percent of the total emissions attributed to Partner utilities.89
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 CO2 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 18 percent of U.S. transmission miles and 21
percent of estimated U.S. emissions from electric power system in 2016.90
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.91 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
88 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.
89 It should be noted that data reported through EPA's GHGRP must go through a verification process; only data verified as of
September 1, 2017 could be used in the emission estimates for the prior year of reporting. For Partners whose GHGRP data was
not yet verified, emissions were extrapolated based upon historical Partner-specific transmission mile growth rates, and those
Partners are included in the 'non-reporting Partners' category.
For electric power systems, verification involved a series of electronic range, completeness, and algorithm checks for each report
submitted. In addition, EPA manually reviewed the reported data and compared each facility's reported transmission miles with
the corresponding quantity in the UDI 2017 database (UDI 2017). In the first year of GHGRP reporting, EPA followed up with
reporters where the discrepancy between the reported miles and the miles published by UDI was greater than 10 percent, with a
goal to improve data quality. Only GHGRP data verified as of September 1, 2017 was included in the emission estimates for
2011,2012, 2013,2014, 2015, and 2016.
90 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.
91 In the United States, SF6 is contained primarily in transmission equipment rated above 34.5 kV.
4-120 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
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: It was determined that there continued to be no statistically significant
difference between the emission rates reported by Partners and by GHGRP-Only Reporters. Therefore, the
emissions data from both groups were combined to develop regression equations for 2012. This was
repeated for 2013, 2014, 2015, and 2016 using Partner and GHGRP-Only Reporter data for each year.
o The 2016 regression equation for reporters was developed based on the emissions reported by a subset
of Partner utilities and GHGRP-Only utilities (representing approximately 67 percent of total U.S.
transmission miles). The regression equation for 2016 is:
Emissions (kg) = 0.213 x Transmission Miles
Table 4-106 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 increased between 2015 and 2016.
Table 4-106: Transmission Mile Coverage (kg) and Regression Coefficients (Percent)
1999
2011
2012
2013
2014
2015
2016
Percentage of Miles Covered by Reporters
50
70
73
73
73
70
67
Regression Coefficient3
0.71 /
0.28
0.24
0.23
0.23
0.21
0.21
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. In 2016, one reporter
was removed as a result of the outlier test.
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, 2017). 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. This growth rate grew to 1.7 percent from 2006 to 2009 as transmission miles increased by more than 33,000
miles. The growth rate for 2009 through 2012 was calculated to be 1.2 percent as transmission miles grew by
approximately 24,000 during this time period. Starting in 2012, two growth rates were calculated to differentiate
between the growth of transmission miles reported by GHGRP reporters and the growth of transmission miles
observed for non-reporters based on the transmission miles included in the UDI database. The annual growth rate for
2012 through 2016 for non-reporters was calculated to be -0.2 percent whereas the growth rate for GHGRP reporters
was calculated to be 1.2 percent. Total transmission miles for both groups grew by approximately 22,000 over this
time period.
Total Industry Emissions
As a final step, total electric power system emissions from 1999 through 2016 were determined for each year by
summing the Partner reported and estimated emissions (reported data was available through the EPA's SF6 Emission
Industrial Processes and Product Use 4-121
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Reduction Partnership for Electric Power Systems), the GHGRP-Only reported emissions, and the non-reporting
utilities' emissions (determined using the regression equations).
1990 through 2016 Emissions from Manufacture of Electrical Equipment
Three different methods were used to estimate 1990 to 2016 emissions from original electrical equipment
manufacturers (OEMs).
• OEM 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 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 (152 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 emissions from 2011 through 2016 were estimated using the SF6 emissions from OEMs reporting to
the GHGRP, and an assumption that these reported emissions account for a conservative estimate of 50
percent of the total emissions from all U.S. OEMs.
Uncertainty and Time-Series Consistency
To estimate the uncertainty associated with emissions of SF6 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 5.4 percent. The uncertainty associated with extrapolated or
interpolated emissions from non-reporting Partners was assumed to be 20 percent.
For GHGRP-Only Reporters, reported SF6 data was assumed to have an uncertainty of 20 percent.92 Based on a
Monte Carlo analysis, the cumulative uncertainty of all GHGRP-Only reported data was estimated to be 9.7 percent.
There are two sources of uncertainty associated with the regression equations used to estimate emissions in 2016
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 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.
92 Uncertainty is assumed to be higher for the GHGRP-Only category, because 2011 is the first year that those utilities have
reported to EPA.
4-122 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 4-107. Electrical
2 Transmission and Distribution SF6 emissions were estimated to be between 3.9 and 4.8 MMT CO2 Eq. at the 95
3 percent confidence level. This indicates a range of approximately 9 percent below and 10 percent above the
4 emission estimate of 4.3 MMT CO2 Eq.
5 Table 4-107: Approach 2 Quantitative Uncertainty Estimates for SF6 Emissions from
6 Electrical Transmission and Distribution (MMT CO2 Eq. and Percent)
2016 Emission
Source
Gas
Estimate
Uncertainty Range Relative to 2016 Emission Estimate3
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Bound
Upper
Bound
Lower Upper
Bound Bound
Electrical Transmission
SFe
4.3
3.9
-9% +10%
and Distribution
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
7 In addition to the uncertainty quantified above, there is uncertainty associated with using global SF6 sales data to
8 estimate U.S. emission trends from 1990 through 1999. However, the trend in global emissions implied by sales of
9 SF6 appears to reflect the trend in global emissions implied by changing SF6 concentrations in the atmosphere. That
10 is, emissions based on global sales declined by 29 percent between 1995 and 1998 (RAND 2004), and emissions
11 based on atmospheric measurements declined by 17 percent over the same period (Levin et al. 2010).
12 Several pieces of evidence indicate that U.S. SF6 emissions were reduced as global emissions were reduced. First,
13 the decreases in sales and emissions coincided with a sharp increase in the price of SF6 that occurred in the mid-
14 1990s and that affected the United States as well as the rest of the world. A representative from DILO, a major
15 manufacturer of SF6 recycling equipment, stated that most U.S. utilities began recycling rather than venting SF6
16 within two years of the price rise. Finally, the emissions reported by the one U.S. utility that reported its emissions
17 for all the years from 1990 through 1999 under the Partnership showed a downward trend beginning in the mid-
18 1990s.
19 Methodological approaches were applied to the entire time series to ensure time-series consistency from 1990
20 through 2016. Details on the emission trends through time are described in more detail in the Methodology section,
21 above.
22 For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
23 Chapter 6 of the 2006IPCC Guidelines, see QA/QC and Verification Procedures section in the introduction of the
24 IPPU Chapter.
25 Recalculations Discussion
26 The historical emissions estimated for this source category have undergone some revisions. SF6 emission estimates
27 for the period 1990 through 2015 were updated relative to the previous report based on revisions to interpolated and
28 extrapolated non-reported Partner data.93 For the current Inventory, historical estimates for the period 2011 through
29 2015 were also updated relative to the previous report based on revisions to reported historical data in EPA's
30 GHGRP. One correction was made to a 2015 transmission mile value reported by one GHGRP partner, who
31 reported a value roughly 25 times higher than was reported in 2011 through 2014 and 2016. As a result of this
32 correction, the emissions regression coefficient for 2015 increased compared to the previous coefficient.
33 For the 1999 through 2016 inventory, UDI Platts data was purchased to obtain transmission mile estimates for 2016
34 estimates. The U.S. transmission mile growth rate used for years 2013 through 2015 was updated to an interpolation
35 between the 2012 and 2016 data, changing the growth rate that was previously used for extrapolating 2013 through
36 2015 estimates. Additionally, a methodology was implemented this year to differentiate between a transmission mile
93 The earlier year estimates within the time series (i.e., 1990 through 1998) were updated based on revisions to the 1999 U.S.
emission estimate because emissions for 1990 through 1998 are estimated by multiplying a series of annual factors by the
estimated U.S. emissions of SF6 from electric power systems in 1999 (see Methodology section).
Industrial Processes and Product Use 4-123
-------
1 growth rate for GHGRP reporters (GHGRP growth rate) and a transmission mile growth rate for non-reporters based
2 on UDI data for those utilities (UDI growth rate). This was done to derive growth rates that were based on roughly
3 the same populations during the three years. From 2013 to 2015, the UDI growth rate is applied only to utilities that
4 do not report to the GHGRP; GHGRP reporters provide transmission mile estimates during these years, which is a
5 better indication of the actual transmission miles for this population. Additionally, the transmission mile growth
6 rates calculated for GHGRP reporters and non-reporters were found to be significantly different. The impact of this
7 methodology improvement is a more accurate representation of transmission miles between 2013 and 2015. This
8 resulted in a decrease in the overall estimate of transmission miles from 2013 and 2014, and an increase in the
9 estimate of transmission miles in 2015.
10 The methodology for calculating uncertainty was updated this year due to an error that was found from the prior
11 year. For the 2015 inventory, a methodology change was implemented to incorporate data reported from Subpart SS
12 of the GHGRP. This methodology change was not previously reflected in the uncertainty methodology. This year,
13 two inputs were modified to reflect the new methodology: the total emissions from Subpart SS GHGRP reporters,
14 and the assumption that 50 percent of the emissions reported to Subpart SS of the GHGRP represented the market
15 share for the entire United States.
16 Finally, a correction was made to the calculation of total U.S. nameplate capacity. For the 1999 through 2015
17 inventory, an additional year (2006) was included in the calculation of emission regression coefficients via
18 extrapolation, though this update was inadvertently not carried to the leak rate calculation, which is used for
19 nameplate capacity. This issue was corrected for the 1999 through 2016 inventory, which smoothed the nameplate
20 capacity trend between 1999 and 2011.
21 As a result of the recalculations, SF6 emissions from electrical transmission and distribution increased by 2.0 percent
22 for 2015 relative to the previous report. On average, the change in SF6 emission estimates for the entire time series is
23 approximately -0.3 percent per year.
24 Planned Improvements
25 EPA is continuing research to improve the methodology for estimating non-reporter nameplate capacity. The current
26 methodology uses Beginning of Year Nameplate Capacity and the Net Increase in Nameplate Capacity for the
27 GHGRP reporters, which aggregates a small portion of hermetically sealed equipment and high-voltage equipment.
28 More research is needed to determine the impact of removing the Net Increase in Nameplate Capacity. Additionally,
29 the current methodology relies on country-wide emission estimates to calculate Total Nameplate Capacity. This
30 reliance on calculated emission values to estimate nameplate capacity often results in similar trends between the
31 values. EPA is planning to revisit the reasoning behind the current methodology, and to evaluate whether a change
32 should be made to relate nameplate capacity directly to transmission miles.
33 Due to the GHGRP policy that allows reporters to "off-ramp" from the reporting program when their emissions
34 remain below certain levels for certain periods of time (e.g., below 25,000 MT CO2 Eq. for five years), the number
35 of electric power systems whose reports are used to develop regression coefficients and country-wide emissions
36 estimates is decreasing. While EPA continues to account for emissions from these electric power systems using the
37 estimation method for non-reporters, it is possible that their cessation of reporting could influence the value and/or
38 stability of the emission factors (per transmission mile) that are applied to non-reporters. EPA is planning to explore
39 whether this is the case. If so, EPA is planning to evaluate whether the current methodology for scaling emissions is
40 the best option.
41 4.26 Nitrous Oxide from Product Uses (CRF
42 Source Category 2G3)
43 Nitrous oxide (N20) is a clear, colorless, oxidizing liquefied gas, with a slightly sweet odor which is used in a wide
44 variety of specialized product uses and applications. The amount of N20 that is actually emitted depends upon the
45 specific product use or application.
4-124 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 There are a total of three N20 production facilities currently operating in the United States (Ottinger 2014). Nitrous
2 oxide is primarily used in carrier gases with oxygen to administer more potent inhalation anesthetics for general
3 anesthesia, and as an anesthetic in various dental and veterinary applications. The second main use of N20 is as a
4 propellant in pressure and aerosol products, the largest application being pressure-packaged whipped cream. Small
5 quantities of N20 also are used in the following applications:
6 • Oxidizing agent and etchant used in semiconductor manufacturing;
7 • Oxidizing agent used, with acetylene, in atomic absorption spectrometry;
8 • Production of sodium azide, which is used to inflate airbags;
9 • Fuel oxidant in auto racing; and
10 • Oxidizing agent in blowtorches used by jewelers and others (Heydorn 1997).
11 Production of N20 in 2016 was approximately 15 kt (see Table 4-108).
12 Table 4-108: N2O Production (kt)
Year
kt
1000
16
2005
15
2012 15
2013 15
2014 15
2015 15
2016 15
13 Nitrous oxide emissions were 4.2 MMT C02 Eq. (14 kt N20) in 2016 (see Table 4-109). Production of N20
14 stabilized during the 1990s because medical markets had found other substitutes for anesthetics, and more medical
15 procedures were being performed on an outpatient basis using local anesthetics that do not require N20. The use of
16 N20 as a propellant for whipped cream has also stabilized due to the increased popularity of cream products
17 packaged in reusable plastic tubs (Heydorn 1997).
18 Table 4-109: N2O Emissions from N2O Product Usage (MMT CO2 Eq. and kt)
Year MMT CO2 Eq. kt
1000 4.2 14
2005 4.2 14
2012 4.2 14
2013 4.2 14
2014 4.2 14
2015 4.2 14
2016 4.2 14
19 Methodology
20 Emissions from N20 product uses were estimated using the following equation:
21 Epu = ^\p x 5a x ERa)
a
22 where,
23 Epu = N20 emissions from product uses, metric tons
24 P Total U.S. production of N20, metric tons
Industrial Processes and Product Use 4-125
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
a = specific application
Sa = Share of N20 usage by application a
ERa = Emission rate for application a, percent
The share of total quantity of N20 usage by end-use represents the share of national N20 produced that is used by
the specific subcategory (e.g., anesthesia, food processing). In 2016, the medical/dental industry used an estimated
86.5 percent of total N20 produced, followed by food processing propellants at 6.5 percent. All other categories
combined used the remainder of the N20 produced. This subcategory breakdown has changed only slightly over the
past decade. For instance, the small share of N20 usage in the production of sodium azide has 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 N20 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). The N20 was allocated across the following categories: medical applications, food processing propellant, and
sodium azide production (pre-1996). A usage emissions rate was then applied for each sector to estimate the amount
of N20 emitted.
Only the medical/dental and food propellant subcategories were estimated to release emissions into the atmosphere,
and therefore these subcategories were the only usage subcategories with emission rates. For the medical/dental
subcategory, due to the poor solubility of N20 in blood and other tissues, none of the N20 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 N20 used as a propellant in pressurized and aerosol food
products, none of the N20 is reacted during the process and all of the N20 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 N20
is consumed/reacted during the process, and therefore the emission rate was considered to be zero percent (Tupman
2002).
The 1990 through 1992 N20 production data were obtained from SRI Consulting's Nitrous Oxide, North America
report (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 N20 production to range between 13.6 and 18.1 thousand metric tons. Tupman (2003) 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 (2003) data are considered more industry-specific and current. Therefore, the midpoint of
the narrower production range was used to estimate N20 emissions for years 1993 through 2001 (Tupman 2003).
The 2002 and 2003 N20 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 N20 production to range between 13.6 and 15.9 thousand metric tons. Due to the
unavailability of data, production estimates for years 2004 through 2016 were held constant at the 2003 value.
The 1996 share of the total quantity of N20 used by each subcategory was obtained from SRI Consulting's Nitrous
Oxide, North America report (Heydorn 1997). The 1990 through 1995 share of total quantity of N20 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 N20 usage by sector was obtained from communication with a N20 industry expert (Tupman 2002).
The 2002 and 2003 share of total quantity of N20 usage by sector was obtained from CGA (2002, 2003). Due to the
unavailability of data, the share of total quantity of N20 usage data for years 2004 through 2016 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 report (Heydorn 1997), and confirmed by a N20 industry expert
(Tupman 2002). The emissions rate for all other subcategories was obtained from communication with a N20
industry expert (Tupman 2002). The emissions rate for the medical/dental subcategory was obtained from the 2006
IPCC Guidelines.
Uncertainty and Time-Series Consistency
The overall uncertainty associated with the 2016 N20 emission estimate from N20 product usage was calculated
using the 2006 IPCC Guidelines (2006) Approach 2 methodology. Uncertainty associated with the parameters used
to estimate N20 emissions include production data, total market share of each end use, and the emission factors
applied to each end use, respectively.
4-126 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 4-110. Nitrous oxide
2 emissions from N20 product usage were estimated to be between 3.2 and 5.2 MMT CO2 Eq. at the 95 percent
3 confidence level. This indicates a range of approximately 24 percent below to 24 percent above the emission
4 estimate of 4.2 MMT CO2 Eq.
5 Table 4-110: Approach 2 Quantitative Uncertainty Estimates for N2O Emissions from N2O
6 Product Usage (MMT CO2 Eq. and Percent)
Source
Gas
2016 Emission Estimate
(MMT CO2 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT CO2 Eq.) (%)
Lower Upper Lower Upper
Bound Bound Bound Bound
N2O from Product Uses
N2O
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.
7 Methodological approaches were applied to the entire time series to ensure consistency in emissions from 1990
8 through 2016. Details on the emission trends through time are described in more detail in the Methodology section,
9 above.
10 For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
11 Chapter 6 of the 2006IPCC Guidelines, see QA/QC and Verification Procedures section in the introduction of the
12 IPPU Chapter.
13 Planned Improvements
14 EPA has recently initiated an evaluation of alternative production statistics for cross-verification and updating time-
15 series activity data, emission factors, assumptions, etc., and a reassessment of N20 product use subcategories that
16 accurately represent trends. This evaluation includes conducting a literature review of publications and research that
17 may provide additional details on the industry. This work is currently ongoing and thus the results have not been
18 incorporated into the current Inventory report.
19 Pending additional resources and planned improvement prioritization, EPA may also evaluate production and use
20 cycles, and the potential need to incorporate a time lag between production and ultimate product use and resulting
21 release of N20. Additionally, planned improvements include considering imports and exports of N20 for product
22 uses.
23 Finally, for future inventories, EPA will examine data from EPA's GHGRP to improve the emission estimates for
24 the N20 product use subcategory. Particular attention will be made to ensure aggregated information can be
25 published without disclosing CBI and time-series consistency, as the facility-level reporting data from EPA's
26 GHGRP are not available for all inventory years as required in this Inventory. EPA is still assessing the possibility
27 of incorporating aggregated GHGRP CBI data to estimate emissions; therefore, this planned improvement is still in
28 development and not incorporated in the current Inventory report.
29 4.27 Industrial Processes and Product Use
30 Sources of Indirect Greenhouse Gases
31 In addition to the main greenhouse gases addressed above, many industrial processes can result in emissions of
32 various ozone precursors (i.e., indirect greenhouse gases). As some of industrial applications also employ thermal
33 incineration as a control technology, combustion byproducts, such as carbon monoxide (CO) and nitrogen oxides
34 (NOx), are also reported with this source category. Non-CH4 volatile organic compounds (NMVOCs), commonly
35 referred to as "hydrocarbons," are the primary gases emitted from most processes employing organic or petroleum
36 based products, and can also result from the product storage and handling. Accidental releases of greenhouse gases
37 associated with product use and handling can constitute major emissions in this category. In the United States,
Industrial Processes and Product Use 4-127
-------
1 emissions from product use are primarily the result of solvent evaporation, whereby the lighter hydrocarbon
2 molecules in the solvents escape into the atmosphere. The major categories of product uses include: degreasing,
3 graphic arts, surface coating, other industrial uses of solvents (e.g., electronics), dry cleaning, and non-industrial
4 uses (e.g., uses of paint thinner). Product usage in the United States also results in the emission of small amounts of
5 hydrofluorocarbons (HFCs) and hydrofluoroethers (HFEs), which are included under Substitution of Ozone
6 Depleting Substances in this chapter.
7 Total emissions of NOx, CO, and NMVOCs from non-energy industrial processes and product use from 1990 to
8 2016 are reported in Table 4-111.
9 Table 4-111: NOx, CO, and NMVOC Emissions from Industrial Processes and Product Use (kt)
Gas/Source
1990
2005
2012
2013
2014
2015
2016
NOx
592
572
443
434
424
424
424
Industrial Processes
Other Industrial Processes3
343
43"
315
311
306
306
306
Metals Processing
88
60
65
66
68
68
68
Chemical and Allied Product
Manufacturing
152
55
45
44
42
42
42
Storage and Transport
3
15
14
10
5
5
5
Miscellaneousb
5
2
3
3
2
2
2
Product Uses
Surface Coating
1
1
1
1
1
1
Graphic Arts
+
0
0
0
0
0
0
Degreasing
+
0
0
0
0
0
0
Dry Cleaning
+
0
0
0
0
0
0
Other Industrial Processes3
+
0
0
0
0
0
0
Non-Industrial Processes0
+
0
0
0
0
0
0
Other
NA
0
0
0
0
0
0
CO
4,129
1,557
1,246
1,262
1,273
1,273
1,273
Industrial Processes
Metals Processing
2,395
752
647
600
553
553
553
Other Industrial Processes3
487
484
388
470
551
551
551
Chemical and Allied Product
Manufacturing
1,073
189
140
128
117
117
117
Miscellaneous13
101
32
49
48
42
42
42
Storage and Transport
69
9"
19
14
9
9
9
Product Uses
Surface Coating
+
2
2
2
1
1
1
Other Industrial Processes3
4
0
0
0
0
0
0
Dry Cleaning
+
0
0
0
0
0
0
Degreasing
+
0
0
0
0
0
0
Graphic Arts
+
0
0
0
0
0
0
Non-Industrial Processes0
+
0
0
0
0
0
0
Other
NA
0
0
0
0
0
0
NMVOCs
7,638
5,849
3,861
3,793
3,723
3,723
3,723
Industrial Processes
Storage and Transport
1,352
1,308
837
727
618
618
618
Other Industrial Processes3
364
414
306
313
320
320
320
Chemical and Allied Product
Manufacturing
575
213
73
70
68
68
68
Metals Processing
111
45
29
28
26
26
26
Miscellaneous15
20
r
27
26
23
23
23
Product Uses
Surface Coating
2,289
1,578
1,061
1,077
1,093
1,093
1,093
Non-Industrial Processes0
1,724
1,446
972
987
1,002
1,002
1,002
Degreasing
675
280
189
191
194
194
194
Dry Cleaning
195
230
155
157
160
160
160
Graphic Arts
249
194
130
132
134
134
134
Other Industrial Processes3
85
88
59
60
61
61
61
4-128 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
Other + 36 24 24 25 25 25
+ Does not exceed 0.5 kt
NA (Not Available)
a Includes rubber and plastics manufacturing, and other miscellaneous applications.
b Miscellaneous includes the following categories: catastrophic/accidental release, other combustion, health
services, cooling towers, 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.
c Includes cutback asphalt, pesticide application adhesives, consumer solvents, and other miscellaneous
applications.
Note: Totals may not sum due to independent rounding.
1 Methodology
2 Emission estimates for 1990 through 2016 were obtained from data published on the National Emission Inventory
3 (NEI) Air Pollutant Emission Trends web site (EPA 2016), and disaggregated based on EPA (2003). Data were
4 collected for emissions of CO, NOx, volatile organic compounds (VOCs), and sulfur dioxide (SO2) from metals
5 processing, chemical manufacturing, other industrial processes, transport and storage, and miscellaneous sources.
6 Emission estimates for 2012 and 2013 for non-electric generating units (EGU) were updated to the most recent
7 available data in EPA (2016). Emission estimates for 2012 and 2013 for non-mobile sources are recalculated
8 emissions by interpolation from 2016 in EPA (2016). Emissions were calculated either for individual source
9 categories or for many categories combined, using basic activity data (e.g., the amount of raw material processed or
10 the amount of solvent purchased) as an indicator of emissions. National activity data were collected for individual
11 categories from various agencies. Depending on the category, these basic activity data may include data on
12 production, fuel deliveries, raw material processed, etc.
13 Emissions for product use were calculated by aggregating product use data based on information relating to product
14 uses from different applications such as degreasing, graphic arts, etc. Emission factors for each consumption
15 category were then applied to the data to estimate emissions. For example, emissions from surface coatings were
16 mostly due to solvent evaporation as the coatings solidify. By applying the appropriate product-specific emission
17 factors to the amount of products used for surface coatings, an estimate of NMVOC emissions was obtained.
18 Emissions of CO and NOx under product use result primarily from thermal and catalytic incineration of solvent-
19 laden gas streams from painting booths, printing operations, and oven exhaust.
20 Activity data were used in conjunction with emission factors, which together relate the quantity of emissions to the
21 activity. Emission factors are generally available from the EPA's Compilation of Air Pollutant Emission Factors,
22 AP-42 (EPA 1997). The EPA currently derives the overall emission control efficiency of a source category from a
23 variety of information sources, including published reports, the 1985 National Acid Precipitation and Assessment
24 Program emissions inventory, and other EPA databases.
25 Uncertainty and Time-Series Consistency
26 Uncertainties in these estimates are partly due to the accuracy of the emission factors and activity data used. A
27 quantitative uncertainty analysis was not performed.
28 Methodological approaches were applied to the entire time series to ensure time-series consistency from 1990
29 through 2016. Details on the emission trends through time are described in more detail in the Methodology section,
30 above.
31 For more information on the general QA/QC process applied to this source category, consistent with Volume 1,
32 Chapter 6 of the 2006IPCC Guidelines, see QA/QC and Verification Procedures section in the introduction of the
33 IPPU Chapter.
Industrial Processes and Product Use 4-129
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
5. Agriculture
Agricultural activities contribute directly to emissions of greenhouse gases through a variety of processes. This
chapter provides an assessment of methane (CH4) and nitrous oxide (N20) emissions from enteric fermentation in
domestic livestock, livestock manure management, rice cultivation, agricultural soil 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 N20 fluxes from agriculture-related land-use and land-use conversion activities, such
as cultivation of cropland, grassland fires and conversion of forest land to cropland, are presented in the Land Use,
Land-Use Change, and Forestry (LULUCF) chapter. Carbon dioxide emissions from on-fann energy use are
reported in the Energy chapter.
Figure 5-1: 2016 Agriculture Chapter Greenhouse Gas Emission Sources (MMT CO2 Eq.)
Agricultural Soil Management
Enteric Fermentation
Manure Management
Rice Cultivation
Urea Fertilization I
Liming |
Field Burning of Agricultural Residues < 0.5
0 20 40 60 80 100 120 140 160 180
MMT CO: Eq.
In 2016, the Agriculture sector was responsible for emissions of 562.6 MMT CO2 Eq.,1 or 8.6 percent of total U.S.
greenhouse gas emissions. Methane emissions from enteric fermentation and manure management represent 25.9
percent and 10.3 percent of total CH4 emissions from anthropogenic activities, respectively. Of all domestic animal
types, beef and dairy cattle were by far the largest emitters of CH4. Rice cultivation and field burning of agricultural
residues were minor sources of CH4. Emissions of N20 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. N20 emissions, accounting for 76.9 percent. Manure management and field burning of agricultural
residues were also small sources of N20 emissions. Urea fertilization and liming each accounted for 0.1 percent of
total CO2 emissions from anthropogenic activities.
1 Following the current reporting requirements under the United Nations Framework Convention on Climate Change (UNFCCC),
this Inventory report presents CO2 equivalent values based on the IPCC Fourth Assessment Report (AR4) GWP values. See the
Introduction chapter for more information.
Agriculture as a Portion of all Emissions
"A
Agriculture 5-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Table 5-1 and Table 5-2 present emission estimates for the Agriculture sector. Between 1990 and 2016, CO2 and
CH4 emissions from agricultural activities increased by 26.5 percent and 15.8 percent, respectively, while N20
emissions from agricultural activities fluctuated from year to year, but increased by 14.1 percent overall.
Table 5-1: Emissions from Agriculture (MMT CO2 Eq.)
Gas/Source
1990
2005
2012
2013
2014
2015
2016
CO2
7.1
7.9
10.3
8.4
8.1
8.7
9.0
Urea Fertilization
2.4
3.5
4.3
4.4
4.5
4.9
5.1
Liming
4.7
4.3
6.0
3.9
3.6
3.8
3.9
CH4
217.6
242.1
244.0
240.6
240.1
245.4
251.8
Enteric Fermentation
164.2
168.9
166.7
165.5
164.2
166.5
170.1
Manure Management
37.2
56.3
65.6
63.3
62.9
66.3
67.7
Rice Cultivation
16.0
16.7
11.3
11.5
12.7
12.3
13.7
Field Burning of Agricultural Residues
0.2
0.2
0.3
0.3
0.3
0.3
0.3
N2O
264.5
270.1
265.5
294.2
291.6
312.8
301.8
Agricultural Soil Management
250.5
253.5
247.9
276.6
274.0
295.0
283.6
Manure Management
14.0
16.5
17.5
17.5
17.5
17.7
18.1
Field Burning of Agricultural Residues
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Total
489.2
520.0
519.8
543.1
539.8
566.9
562.6
Note: Totals may not sum due to independent rounding.
Table 5-2: Emissions from Agriculture (kt)
Gas/Source
1990
2005
2012
2013
2014
2015
2016
CO2
7,084
7,854
10,259
8,350
8,147
8,665
8,961
Urea Fertilization
2,417
3,504
4,282
4,443
4,538
4,888
5,098
Liming
4,667
4,349
5,978
3,907
3,609
3,777
3,863
CH4
8,702
9,684
9,760
9,623
9,602
9,816
10,073
Enteric Fermentation
6,566
6,755
6,670
6,619
6,567
6,661
6,805
Manure Management
1,486
2,254
2,625
2,530
2,514
2,651
2,709
Rice Cultivation
641
667
453
462
510
493
549
Field Burning of Agricultural Residues
9
8
11
11
11
11
11
N2O
888
906
891
987
979
1,050
1,013
Agricultural Soil Management
840
851
832
928
920
990
952
Manure Management
47
55
59
59
59
59
61
Field Burning of Agricultural Residues
+
+
+
+
+
+
+
+ Does not exceed 0.5 kt.
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 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 do 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.
5-2 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Box 5-2: Biennial Inventory Compilation
For the current Inventory (i.e., 1990 through 2016 report), a biennial inventory compilation process lias been
implemented for the Agriculture and LULUCF chapters. As part of this biennial compilation process, during
alternating years, modified approaches will be applied to extend the emissions/removals time series of some
Agriculture and LULUCF source and sink categories rather than implementing a full inventory compilation (i.e.,
updating activity data and running models). In the current Inventory, for each category where these modified
approaches for extending the time series have been utilized, the alternative methods have been transparently
documented in their respective Methodology sections of the chapter. This biennial compilation schedule has been
adopted for the Agriculture and LULUCF chapters in order to conserve and efficiently utilize resources that are
needed to implement key improvements. Over the next four to six years, this process will result in more rapid
improvements to the Agriculture and LULUCF chapters. The next Inventory report (i.e., 1990 through 2017 report)
will include a full compilation of the Agriculture and LULUCF chapters along with a number of key improvements.
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 CH4 as a byproduct, which can be exhaled or eructated by the animal. The amount of
CH4 produced and emitted by an individual animal depends primarily upon the animal's digestive system, and the
amount and type of feed it consumes.
Ruminant animals (e.g., cattle, buffalo, sheep, goats, and camels) are the major emitters of CH4 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 CH4 emissions per unit of body mass among all animal types.
Non-ruminant animals (e.g., swine, horses, and mules and asses) also produce CH4 emissions through enteric
fermentation although this microbial fermentation occurs in the large intestine. These non-ruminants emit
significantly less CH4 on a per-animal-mass basis than ruminants because the capacity of the large intestine to
produce CH4 is lower.
In addition to the type of digestive system, an animal's feed quality and feed intake also affect CH4 emissions. In
general, lower feed quality and/or higher feed intake leads to higher CH4 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 CH4
emissions in 2016 were 170.1 MMT CO: Eq. (6,805 kt). Beef cattle remain the largest contributor of CH4 emissions
from enteric fermentation accounting for 71 percent in 2016. Emissions from dairy cattle in 2016 accounted for 25
percent, and the remaining emissions were from horses, sheep, swine, goats, American bison, mules and asses.
Agriculture 5-3
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Table 5-3: ChU Emissions from Enteric Fermentation (MMT CO2 Eq.)
Livestock Type
1990
2005
2012
2013
2014
2015
2016
Beef Cattle
119.1
125.2
119.1
118.0
116.5
118.1
121.3
Dairy Cattle
39.4
37.6
41.7
41.6
42.0
42.6
42.9
Swine
2.0
2.3
2.5
2.5
2.4
2.6
2.6
Horses
1.0
1.7
1.6
1.6
1.6
1.5
1.5
Sheep
2.3
1.2
1.1
1.1
1.0
1.1
1.1
Goats
0.3
0.4
0.3
0.3
0.3
0.3
0.3
American Bison
0.1
0.4
0.3
0.3
0.3
0.3
0.3
Mules and Asses
+
0.1
0.1
0.1
0.1
0.1
0.1
Total
164.2
168.9
166.7
165.5
164.2
166.5
170.1
+ Does not exceed 0.05 MMT CO2 Eq.
Note: Totals may not sum due to independent rounding.
Table 5-4: ChU Emissions from Enteric Fermentation (kt)
Livestock Type
1990
2005
2012
2013
2014
2015
2016
Beef Cattle
4,763
5,007
4,763
4,722
4,660
4,724
4,853
Dairy Cattle
1,574
1,503
1,670
1,664
1,679
1,706
1,715
Swine
81
92
100
98
96
102
105
Horses
40
70
65
64
62
61
62
Sheep
91
49
43
43
42
42
42
Goats
13
14
13
13
12
12
13
American Bison
4
17
13
13
12
12
11
Mules and Asses
1
2
3
3
3
3
3
Total 6,566 6,755 6,670 6,619 6,567 6,661 6,805
Note: Totals may not sum due to independent rounding.
From 1990 to 2016, emissions from enteric fermentation have increased by 3.6 percent. While emissions generally
follow trends in cattle populations, over the long term there are exceptions as population decreases have been
coupled with both production increases and minor decreases. For example, beef cattle emissions increased 1.9
percent from 1990 to 2016, while beef cattle populations actually declined by 3.7 percent and beef production
increased (USDA 2016), and while dairy cattle emissions increased 8.9 percent over the entire time series, the
population has declined by 2.1 percent and milk production increased 54 percent (USDA 2016). This trend indicates
that while emission factors per head are increasing, emission factors per unit of product are going down.
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 feedlot cattle.
Beef cattle emissions generally increased from 2004 to 2007, as beef cattle populations underwent increases 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 in 2015 and 2016,
consistent with another increase in population over those same years. 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). Dairy cattle emissions have continued to trend
upward since 2007, in line with dairy cattle population increases. Regarding trends in other animals, populations of
sheep have steadily declined, with an overall decrease of 53 percent since 1990. Horse populations are 56 percent
greater than they were in 1990, but their numbers have been declining by about 2 percent annually since 2007. Goat
populations increased by about 20 percent through 2007, steadily decreased through 2015, then jumped by 13
percent from 2015 to 2016. Swine populations have trended upward through most of the time series, increasing 19
percent from 1990 to 2016. The population of American bison nearly tripled over the 1990 to 2016 time period,
while mules and asses have more than quadrupled.
5-4 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Methodology
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 CH4 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 handled using a
less detailed approach (i.e., IPCC Tier 1).
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
CH4 emissions from enteric fermentation, incorporates this information and other analyses of livestock population,
feeding practices, and production characteristics. For the current Inventory, CEFM results for 1990 through 2015
were carried over from the 1990 to 2015 Inventory report (i.e., 2017 Inventory submission), and a simplified
approach was used to estimate 2016 enteric emissions from cattle.
1990 to 2015 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 Stackers
o Animals inFeedlots (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 2016).
Diet characteristics were estimated by region for dairy, grazing beef, and feedlot beef cattle. These diet
characteristics were used to calculate digestible energy (DE) values (expressed as the percent of gross energy intake
digested by the animal) and CH4 conversion rates (Ym) (expressed as the fraction of gross energy converted to CH4)
for each regional population category. The IPCC recommends Ym ranges of 3.0±1.0 percent for feedlot cattle and
6.5±1.0 percent for other well-fed cattle consuming temperate-climate feed types (IPCC 2006). Given the
availability of detailed diet information for different regions and animal types in the United States, DE and Ym
values unique to the United States were developed. The diet characterizations and estimation of DE and Ym values
were based on information from state agricultural extension specialists, a review of published forage quality studies
and scientific literature, expert opinion, and modeling of animal physiology.
The diet characteristics for dairy cattle were based on Donovan (1999) and an extensive review of nearly 20 years of
literature from 1990 through 2009. Estimates of DE were national averages based on the feed components of the
diets observed in the literature for the following year groupings: 1990 through 1993, 1994 through 1998, 1999
Agriculture 5-5
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
through 2003, 2004 through 2006, 2007, and 2008 onward.2 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 CH4 emissions, the population was divided into state, age, sub-type (i.e., dairy cows and replacements,
beef cows and replacements, heifer and steer stackers, 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 CH4 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
CH4 emission factors for the following cattle types: dairy cows, beef cows, dairy replacements, beef replacements,
steer stackers, heifer stackers, 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.
2016 Inventory Methodology for Cattle
As noted above, a simplified approach for cattle enteric emissions was used in lieu of the CEFM for 2016. First,
2016 populations for each of the CEFM cattle sub-populations were estimated, then these populations were
multiplied by the corresponding implied emission factors developed from the CEFM for the previous Inventory
year. Dairy cow, beef cow, and bull populations for 2016 were based on data directly from the USDA-NASS
QuickStats database (USDA 2017). Because the remaining CEFM cattle sub-population categories do not
correspond exactly to the remaining QuickStats cattle categories, 2016 populations for these categories were
estimated by extrapolating the 2015 populations based on percent changes from 2015 to 2016 in similar QuickStats
categories, consistent with Volume 1, Chapter 5 of the 2006 IPCC Guidelines on time-series consistency. Table 5-5
lists the QuickStats categories used to estimate the percent change in population for each of the CEFM categories.
Table 5-5: Cattle Sub-Population Categories for 2016 Population Estimates
CEFM Cattle Category
USDA-NASS Quickstats Cattle Category
Dairy Calves
Dairy Cows
Dairy Replacements 7-11 months
Cattle, Calves
Cattle, Cows, Milk
Cattle, Heifers, GE 500 lbs, Milk Replacement
2 Due to inconsistencies in the 2003 literature values, the 2002 values were used for 2003, as well.
5-6 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Dairy Replacements 12-23 months
Bulls
Beef Calves
Beef Cows
Beef Replacements 7-11 months
Beef Replacements 12-23 months
Steer Stackers
Heifer Stackers
Steer Feedlot
Heifer Feedlot
Cattle, Heifers, GE 500 lbs, Milk Replacement
Cattle, Bulls, GE 500 lbs
Cattle, Calves
Cattle, Cows, Beef
Cattle, Heifers, GE 500 lbs, Beef Replacement
Cattle, Heifers, GE 500 lbs, Beef Replacement
Cattle, Steers, GE 500 lbs
Cattle, Heifers, GE 500 lbs, (Excl. Replacement)
Cattle, On Feed
Cattle, On Feed
Non-Cattle Livestock
Emission estimates for other animal types were based on average emission factors representative of entire
populations of each animal type. Methane emissions from these animals accounted for a minor portion of total CH4
emissions from livestock in the United States from 1990 through 2016. 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 2015 for sheep; swine; goats; horses; mules and asses; and American
bison were obtained for available years from USDA-NASS (USDA 2016). Horse, goat and mule and ass population
data were available for 1987, 1992, 1997, 2002, 2007, and 2012 (USDA 1992, 1997, 2016); the remaining years
between 1990 and 2015 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, and 2012 (USDA 2016) 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. For the 2016
Inventory simplified approach, values for 1990 through 2015 remained the same as the 1990 through 2015
Inventory. For 2016 populations, sheep and swine population data were obtained from USD A-NASS (USDA 2017).
The 2016 populations for the other animal groups were extrapolated based on 1990 through 2015 values.
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).
See Annex 3.10 for more detailed information on the methodology and data used to calculate CH4 emissions from
enteric fermentation.
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 report (i.e., 2003
submission to the UNFCCC). There have been no significant changes to the methodology since that time;
consequently, these uncertainty estimates were directly applied to the 2016 emission estimates in this Inventory
report.
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
Uncertainty and Time-Series Consistency
Agriculture 5-7
-------
1 exogenous correlation coefficients between the probability distributions of selected activity-related variables were
2 developed through expert judgment.
3 The uncertainty ranges associated with the activity data-related input variables were plus or minus 10 percent or
4 lower. However, for many emission factor-related input variables, the lower- and/or the upper-bound uncertainty
5 estimates were over 20 percent. The results of the quantitative uncertainty analysis are summarized in Table 5-6.
6 Based on this analysis, enteric fermentation CH4 emissions in2016 were estimated to be between 151.4 and 200.7
7 MMT CO2 Eq. at a 95 percent confidence level, which indicates a range of 11 percent below to 18 percent above the
8 2016 emission estimate of 170.1 MMT CO2 Eq. Among the individual cattle sub-source categories, beef cattle
9 account for the largest amount of CH4 emissions, as well as the largest degree of uncertainty in the emission
10 estimates—due mainly to the difficulty in estimating the diet characteristics for grazing members of this animal
11 group. Among non-cattle, horses represent the largest percent of uncertainty in the previous uncertainty analysis
12 because the Food and Agricultural Organization of the United Nations (FAO) population estimates used for horses
13 at that time had a higher degree of uncertainty than for the USD A population estimates used for swine, goats, and
14 sheep. The horse populations are now from the same USDA source as the other animal types, and therefore the
15 uncertainty range around horses is likely overestimated. Cattle calves, American bison, mules and asses were
16 excluded from the initial uncertainty estimate because they were not included in emission estimates at that time.
17 Table 5-6: Approach 2 Quantitative Uncertainty Estimates for ChU Emissions from Enteric
18 Fermentation (MMT CO2 Eq. and Percent)
2016 Emission
Source
Gas
Estimate
Uncertainty Range Relative to Emission Estimate3'b'c
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower Upper
Lower Upper
Bound Bound
Bound Bound
Enteric Fermentation
CH4
170.1
151.4 200.7
-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 2016 estimates.
c The overall uncertainty calculated in 2003, and applied to the 2016 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.
19 Details on the emission trends through time are described in more detail in the Methodology section.
20 QA/QC arid Verification
21 In order to ensure the quality of the emission estimates from enteric fermentation, the IPCC Tier 1 and Tier 2
22 Quality Assurance/Quality Control (QA/QC) procedures were implemented consistent with the U.S. QA/QC plan
23 (EPA 2002). Tier 2 QA procedures included independent review of emission estimate methodologies from previous
24 inventories. Over the past few years, particular importance has been placed on harmonizing the data exchange
25 between the enteric fermentation and manure management source categories. The current Inventory now utilizes the
26 transition matrix from the CEFM for estimating cattle populations and weights for both source categories, and the
27 CEFM is used to output volatile solids and nitrogen excretion estimates using the diet assumptions in the model in
28 conjunction with the energy balance equations from the IPCC (2006). This approach facilitates the QA/QC process
29 for both of these source categories.
30 Recalculations Discussion
31 No recalculations were performed for the 1990 to 2015 estimates. The 2016 estimates were developed using a
32 simplified approach, as noted earlier in the chapter.
5-8 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Planned Improvements
Continued research and regular updates are necessary to maintain an emissions inventory that reflects the current
base of knowledge. Depending upon the outcome of ongoing investigations, future improvements for enteric
fermentation could include some of the following options:
• 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 and beef cow lactation
rates;
• Investigation of 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;
• 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;
• Potentially updating to a Tier 2 methodology for other animal types (i.e., sheep, swine, goats, horses);
• Investigation of methodologies and emission factors for including enteric fermentation emission estimates from
poultry;
• Comparison of the current CEFM processing of animal population data to estimates developed using annual
average populations to determine if the model could be simplified to use annual population data; 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.
5.2 Manure Management (CRF Source
Category 3B)
The treatment, storage, and transportation of livestock manure can produce anthropogenic CH4 and N20 emissions.
Methane is produced by the anaerobic decomposition of manure and 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 or poultry manure are 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 CH4. 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 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 type, including the animal's digestive system, 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.
Agriculture 5-9
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
As previously stated, nitrous oxide emissions are produced through both direct and indirect pathways. Direct N20
emissions are produced as part of the nitrogen (N) cycle through the nitrification and denitrification of the organic N
in livestock dung and urine.3 There are two pathways for indirect N20 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 N03~) onto soils and the surface of lakes and other waters. The second pathway is the runoff and leaching
of N from manure to the groundwater below, in riparian zones receiving drain or runoff water, or in the ditches,
streams, rivers, and estuaries into which the land drainage water eventually flows.
The production of direct N20 emissions from livestock manure depends on the composition of the manure and urine,
the type of bacteria involved in the process, and the amount of oxygen and liquid in the manure system. For direct
N20 emissions to occur, the manure must first be handled aerobically where ammonia (NH3) or organic N is
converted to nitrates and nitrites (nitrification), and then handled anaerobically where the nitrates and nitrites are
reduced to dinitrogen gas (N2), with intermediate production of N20 and nitric oxide (NO) (denitrification)
(Groffman et al. 2000). 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 N20 in the waste management system (WMS). Indirect N20 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 CH4 emissions from manure management in 2016 were 67.7 MMT C02 Eq. (2,709 kt); in 1990,
emissions were 37.2 MMT C02 Eq. (1,486 kt). This represents an 82 percent increase in emissions from 1990.
Emissions increased on average by 1.1 MMT C02 Eq. (3.0 percent) annually over this period. The majority of this
increase is due to swine and dairy cow manure, where emissions increased 63 and 140 percent, respectively. From
2015 to 2016, there was a 2.2 percent increase in total CH4 emissions from manure management, due to an increase
in animal populations.
Although the majority 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 generally been decreasing 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 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
3 Direct and indirect N2O emissions from dung and urine spread onto fields either directly as daily spread or after it is removed
from manure management systems (i.e., lagoon, pit, etc.) and from livestock dung and urine deposited on pasture, range, or
paddock lands are accounted for and discussed in the Agricultural Soil Management source category within the Agriculture
sector.
5-10 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 (MCF) values in combination with the 1992, 1997, 2002, 2007 and 2012 farm-size distribution data reported in the
2 U.S. Department of Agriculture (USDA) Census of Agriculture (USDA 2016d).
3 In 2016, total N2O emissions from manure management were estimated to be 18.1 MMT CO2 Eq. (61 kt); in 1990,
4 emissions were 14.0 MMT CO2 Eq. (47 kt). These values include both direct and indirect N20 emissions from
5 manure management. Nitrous oxide emissions have remained fairly steady since 1990. Small changes in N20
6 emissions from individual animal groups exhibit the same trends as the animal group populations, with the overall
7 net effect that N20 emissions showed a 30 percent increase from 1990 to 2016 and a 1.1 percent increase from 2015
8 through 2016. Overall shifts toward liquid systems have driven down the emissions per unit of nitrogen excreted as
9 dry manure handling systems have greater aerobic conditions that promote N20 emissions.
10 Table 5-7 and Table 5-8 provide estimates of CH4 and N20 emissions from manure management by animal
11 category.
12 Table 5-7: ChU and N2O Emissions from Manure Management (MMT CO2 Eq.)
Gas/Animal Type
1990
2005
2012
2013
2014
2015
2016
CH4a
37.2
56.3
65.6
63.3
62.9
66.3
67.7
Dairy Cattle
14.7
26.4
34.3
33.4
34.0
34.8
35.3
Beef Cattle
3.1
3.3
3.2
3.0
3.0
3.1
3.2
Swine
15.6
22.9
24.5
23.2
22.2
24.6
25.4
Sheep
0.2
0.1
0.1
0.1
0.1
0.1
0.1
Goats
+
+
+
+
+
+
+
Poultry
3.3
3.2
3.2
3.2
3.3
3.4
3.5
Horses
0.2
0.3
0.2
0.2
0.2
0.2
0.2
American Bison
+
+
+
+
+
+
+
Mules and Asses
+
+
+
+
+
+
+
N2Ob
14.0
16.5
17.5
17.5
17.5
17.7
18.1
Dairy Cattle
5.3
5.6
5.9
5.9
5.9
6.1
6.1
Beef Cattle
5.9
7.2
7.7
7.7
7.8
7.7
7.9
Swine
1.2
1.7
1.9
1.9
1.8
2.0
2.0
Sheep
0.1
0.3
0.3
0.3
0.3
0.3
0.3
Goats
+
+
+
+
+
+
+
Poultry
1.4
1.6
1.6
1.6
1.6
1.6
1.6
Horses
0.1
0.1
0.1
0.1
0.1
0.1
0.1
American Bison0
NA
NA
NA
NA
NA
NA
NA
Mules and Asses
+
+
+
+
+
+
+
Total
51.1
72.9
83.2
80.8
80.4
84.0
85.9
+ Does not exceed 0.05 MMT CO2 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 N2O emissions.
c There are no American bison N2O emissions from managed systems.
Notes: Totals may not sum due to independent rounding. American bison are maintained entirely
on unmanaged WMS.
13 Table 5-8: ChU and N2O Emissions from Manure Management (kt)
Gas/Animal Type
1990
2005
2012
2013
2014
2015
2016
CH4a
1,486
2,254
2,625
2,530
2,514
2,651
2,709
Dairy Cattle
590
1,057
1,373
1,338
1,361
1,391
1,413
Beef Cattle
126
133
128
121
120
126
130
Swine
622
916
982
930
890
985
1,014
Sheep
7
3
3
3
3
3
3
Goats
1
1
1
1
1
1
1
Poultry
131
129
128
128
131
135
138
Horses
9
12
10
9
9
9
9
American Bison
+
+
+
+
+
+
+
Agriculture 5-11
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Mules and Asses
+
+
+
+
+
+
+
N2Ob
47
55
59
59
59
59
61
Dairy Cattle
18
19
20
20
20
20
21
Beef Cattle
20
24
26
26
26
26
27
Swine
4
6
6
6
6
7
7
Sheep
+
1
1
1
1
1
1
Goats
+
+
+
+
+
+
+
Poultry
5
5
5
5
5
5
5
Horses
+
+
+
+
+
+
+
American Bison0
NA
NA
NA
NA
NA
NA
NA
Mules and Asses
+
+
+
+
+
+
+
+ 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 N2O emissions.
c There are no American bison N2O emissions from managed systems.
Notes: Totals may not sum due to independent rounding. American bison are maintained entirely
on unmanaged WMS.
Methodology
The methodologies presented in IPCC (2006) form the basis of the CH4 and N20 emission estimates for each animal
type. This section presents a summary of the methodologies used to estimate CH4 and N20 emissions from manure
management. For the current Inventory, results for 1990 through 2015 were carried over from the 1990 to 2015
Inventory (i.e., 2017 submission) and a simplified approach was used to estimate manure management emissions for
2016. See Annex 3.11 for more detailed information on the methodology and data used to calculate CH4 and N20
emissions from manure management.
Methane Calculation Methods
The following inputs were used in the calculation of manure management CH4 emissions for 1990 through 2015:
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;
Volatile solids (VS) production rate (by animal type and state or United States);
Methane producing potential (B0) 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 2015 for all livestock types, except goats, horses, mules and
asses, and American bison were obtained from the USDA-NASS. For cattle, the USDA populations were
utilized in conjunction with birth rates, detailed feedlot placement information, and slaughter weight data to
create the transition matrix in the Cattle Enteric Fermentation Model (CEFM) that models cohorts of individual
animal types and their specific emission profiles. The key variables tracked for each of the cattle population
categories are described in Section 5.1 and in more detail in Annex 3.10. Goat population data for 1992, 1997,
2002, 2007, and 2012; horse and mule and ass population data for 1987, 1992, 1997, 2002, 2007, and 2012; and
American bison population for 2002, 2007 and 2012 were obtained from the Census of Agriculture (USDA
2014a). 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
5-12 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
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 2016d) and EPA (ERG 2000a; EPA 2002a
and 2002b). 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.
• The maximum CH4-producing capacity of the VS (B0) 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).
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 presented in the AgSTAR Digest (EPA 2000, 2003, 2006) and the
AgSTAR project database (EPA 2016). 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 maximum CH4 producing capacity of the
VS (B0) (m3 CH4 per kg VS), the MCF for that WMS (percent), and the density of CH4 (kg CH4per m3 CH4). The
CH4 emissions for each WMS, state, and animal type were summed to determine the total U.S. CH4 emissions.
The following approach was used in the calculation of manure management CH4 emissions for 2016:
National-level population data for cattle, poultry, and swine animal types were downloaded from USDA-NASS
Quickstats. National-level population data for goats, horses, bison, mules, and asses were extrapolated based on the
1990 through 2015 population values. The national populations were then multiplied by the animal-specific 2015
implied emission factors4 for CH4 to calculate national-level 2016 CH4 emissions estimates. These methods were
utilized in order to maintain time-series consistency as referenced in Volume 1, Chapter 5 of the 2006 IPCC
Guidelines.
4 An implied emission factor is defined as emissions divided by the relevant measure of activity: The implied emission factors is
equal to emissions per activity data unit. For source/sink categories that are composed of several subcategories, the emissions and
activity data are summed up across all subcategories. Hence, the implied emission factors are generally not equivalent to the
emission factors used to calculate emission estimates, but are average values that could be used, with caution, in data
comparisons (UNFCCC 2017).
Agriculture 5-13
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Nitrous Oxide Calculation Methods
The following inputs were used in the calculation of direct and indirect manure management N20 emissions for
1990 through 2015:
• 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 N20 emission factor (EFwms);
• Indirect N20 emission factor for volatilization (EFvoiatuization);
• Indirect N20 emission factor for runoff and leaching (EFnlM0rrk,,ch):
• Fraction of N loss from volatilization of NH3 and NOx (Fracgas); and
• Fraction of N loss from runoff and leaching (FraCnmoff/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 rates 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 USD A'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 rates
were assumed to be the same as NOF bulls.5
• All N20 emission factors (direct and indirect) were taken from IPCC (2006). These data are appropriate because
they were developed using U.S. data.
• Country-specific estimates for the fraction of N loss from volatilization (Fracgas) and runoff and leaching
(FraCrunoff/ieach) were developed. Fracg(ls values were based on WMS-specific volatilization values as estimated
from EPA's National Emission Inventory - Ammonia Emissions from Animal Agriculture Operations (EPA
2005). FraCrunoff/ieaching values were based on regional cattle runoff data from EPA's Office of Water (EPA
2002b; see Annex 3.11).
To estimate N20 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 N20 emissions were calculated by multiplying the amount of N excreted (kg per year) in each WMS by the
N20 direct emission factor for that WMS (EFwms, in kg N20-N per kg N) and the conversion factor of N20-N to
N20. These emissions were summed over state, animal, and WMS to determine the total direct N20 emissions (kg of
N20 per year).
Next, indirect N20 emissions from volatilization (kg N20 per year) were calculated by multiplying the amount of N
excreted (kg per year) in each WMS by the fraction of N lost through volatilization (Fractas) divided by 100, the
emission factor for volatilization (EFvoiatiiization, in kg N20 per kg N), and the conversion factor of N20-N to N20.
Indirect N20 emissions from runoff and leaching (kg N20 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 (FraCnmoff/ieach)
divided by 100, and the emission factor for runoff and leaching (EFrunoff/ieach, in kg N20 per kg N), and the
5 The N2O emissions from N excreted (Nex) by 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 N2O emissions from unmanaged WMS are not included in the
Manure Management category, there are no N2O emissions from American bison included in the Manure Management category.
5-14 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 conversion factor of N20-N to N20. The indirect N20 emissions from volatilization and runoff and leaching were
2 summed to determine the total indirect N20 emissions.
3 Following these steps, direct and indirect N20 emissions were summed to determine total N20 emissions (kg N20
4 per year) for the years 1990 to 2015.
5 The following approach was used in the calculation of manure management N2O emissions for 2016:
6 National-level population data for cattle, poultry, and swine animal types were downloaded from USDA-NASS
7 Quickstats. National-level population data for goats, horses, bison, mules, and asses were extrapolated based on the
8 1990 through 2015 population values. The national populations were then multiplied by the animal-specific 2015
9 implied emission factors for N20 (combines both direct and indirect N20) to calculate national-level 2016 N20
10 emissions estimates. These methods were utilized in order to maintain time-series consistency as referenced in
11 Volume 1, Chapter 5 of the 2006IPCC Guidelines.
12 Uncertainty and Time-Series Consistency
13 An analysis (ERG 2003a) was conducted for the manure management emission estimates presented in the 1990
14 through 2001 Inventory report (i.e., 2003 submission to the UNFCCC) to determine the uncertainty associated with
15 estimating CH4 and N20 emissions from livestock manure management. The quantitative uncertainty analysis for
16 this source category was performed in 2002 through the IPCC-recommended Approach 2 uncertainty estimation
17 methodology, the Monte Carlo Stochastic Simulation technique. The uncertainty analysis was developed based on
18 the methods used to estimate CH4 and N20 emissions from manure management systems. A normal probability
19 distribution was assumed for each source data category. The series of equations used were condensed into a single
20 equation for each animal type and state. The equations for each animal group contained four to five variables around
21 which the uncertainty analysis was performed for each state. These uncertainty estimates were directly applied to the
22 2016 emission estimates as there have not been significant changes in the methodology since that time.
23 The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 5-9. Manure management
24 CH4 emissions in 2016 were estimated to be between 55.5 and 81.3 MMT C02 Eq. at a 95 percent confidence level,
25 which indicates a range of 18 percent below to 20 percent above the actual 2016 emission estimate of 67.7 MMT
26 C02 Eq. At the 95 percent confidence level, N20 emissions were estimated to be between 15.2 and 22.5 MMT C02
27 Eq. (or approximately 16 percent below and 24 percent above the actual 2016 emission estimate of 18.1 MMT C02
28 Eq.).
29 Table 5-9: Approach 2 Quantitative Uncertainty Estimates for CH4 and N2O (Direct and
30 Indirect) Emissions from Manure Management (MMT CO2 Eq. and Percent)
2016 Emission
Source
Gas
Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Manure Management
CH4
67.7
55.5 81.3
-18% 20%
Manure Management
N2O
18.1
15.2 22.5
-16% 24%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
31 QA/QC and Verification
32 Tier 1 and Tier 2 QA/QC activities were conducted consistent with the U.S. QA/QC plan. Tier 2 activities focused
33 on comparing estimates for the previous and current Inventories for N20 emissions from managed systems and CH4
34 emissions from livestock manure. All errors identified were corrected. Order of magnitude checks were also
35 conducted, and corrections made where needed.
36 Time-series data, including population, are validated by experts to ensure they are representative of the best
37 available U.S.-specific data. The U.S.-specific values for TAM, Nex, VS, B0, and MCF were also compared to the
38 IPCC default values and validated by experts. Although significant differences exist in some instances, these
39 differences are due to the use of U.S.-specific data and the differences in U.S. agriculture as compared to other
Agriculture 5-15
-------
1 countries. The U.S. manure management emission estimates use the most reliable country-specific data, which are
2 more representative of U.S. animals and systems than the IPCC (2006) default values.
3 For additional verification of the 1990 to 2015 estimates, the implied CH4 emission factors for manure management
4 (kg of CH4 per head per year) were compared against the default IPCC (2006) values.6 Table 5-10 presents the
5 implied emission factors of kg of CH4 per head per year used for the manure management emission estimates as well
6 as the IPCC (2006) default emission factors. The 2015 U.S. implied emission factors fall within the range of the
7 IPCC (2006) default values, except in the case of sheep, goats, and some years for horses and dairy cattle. The U.S.
8 implied emission factors are greater than the IPCC (2006) default value for those animals due to the use of U.S.-
9 specific data for typical animal mass and VS excretion. There is an increase in implied emission factors for dairy
10 cattle and swine across the time series. This increase reflects the dairy cattle and swine industry trend towards larger
11 farm sizes; large farms are more likely to manage manure as a liquid and therefore produce more CH4 emissions.
12 Table 5-10: IPCC (2006) Implied Emission Factor Default Values Compared with Calculated
13 Values for ChU from Manure Management (kg/head/year)
IPCC Default
CH4 Emission Implied CH4 Emission Factors (kg/head/year)
(kg/head/year)
1990
2005
2011
2012
2013
2014
2015
Dairy Cattle
48-112
30.2
59.4
70.3
73.9
72.3
73.4
74.0
Beef Cattle
1-2
1.5
1.6
1.7
1.7
1.6
1.6
1.7
Swine
10-45
11.5
15.0
14.5
14.8
14.2
13.8
14.5
Sheep
0.19-0.37
0.6
0.6
0.5
0.5
0.5
0.5
0.5
Goats
0.13-0.26
0.4
0.3
0.3
0.3
0.3
0.3
0.3
Poultry
0.02-1.4
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Horses
1.56-3.13
4.3
3.1
2.6
2.7
2.5
2.5
2.6
American Bison
NA
1.8
2.0
2.1
2.1
2.0
2.0
2.1
Mules and Asses
0.76-1.14
0.9
1.0* ,*
1.0
1.0
0.9
0.9
1.0
Note: CH4 implied emission factors were not calculated for 2016 due to the simplified emissions estimation
approach used to estimate emissions for that year.
14 In addition, default IPCC (2006) emission factors for N20 were compared to the U.S. Inventory implied N20
15 emission factors. Default N2O emission factors from the 2006 IPCC Guidelines were used to estimate N2O emission
16 from each WMS in conjunction with U.S.-specific Nex values. The implied emission factors differed from the U.S.
17 Inventory values due to the use of U. S.-specific Nex values and differences in populations present in each WMS
18 throughout the time series.
19 Recalculations Discussion
20 No recalculations were performed for the 1990 to 2015 estimates. The 2016 estimates were developed using a
21 simplified approach, as noted in the Methodology section of this chapter.
22 Planned Improvements
23 Potential data sources (such as the USD A Agricultural Resource Management Survey) for updated WMS
24 distribution estimates have been obtained and discussed with USD A. EPA is working with USD A to review these
25 data sources for potential implementation in future Inventory reports.
26 In addition, EPA may pursue the following improvements in future Inventory years:
27 • Implement a methodology to calculate monthly emissions estimates to present data that show seasonal changes
28 in emissions from each WMS.
6 CH4 implied emission factors were not calculated for 2016 due to the simplified emissions estimation approach used to estimate
emissions for that year.
5-16 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
• Revise 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).
• Update the B0 data used in the Inventory, which are dated.
• Compare CH4 and N20 emission estimates with estimates from other models and more recent studies and
compare the results to the Inventory, such as USDA's Dairy Gas Emissions Model.
• Compare manure management emission estimates with on-farm WMS measurement data to identify
opportunities for improved estimates.
• Improve 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 they depend on each other's outputs to properly calculation emissions.
• Changes that have been implemented to the CH4 and N20 estimates warrant an assessment of the current
uncertainty analysis; therefore, a revision of the quantitative uncertainty surrounding emission estimates from
this source will be initiated.
5.3 Rice Cultivation (CRF Source Category 3C)
Most of the world's rice is grown on flooded fields (Baicich 2013), and flooding creates anaerobic conditions that
foster CH4 production through a process known as methanogenesis. Approximately 60 to 90 percent of the CH4
produced by methanogenie bacteria is oxidized in the soil and converted to C02by 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).
Water management is arguably the most important factor affecting CH4 emissions, 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 CH4, but large amounts of CH4canbe emitted in continuously irrigated fields, which is the most
common practices 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 N20 emissions. Deepwater rice fields (i.e., fields with flooding depths greater than one meter, such as
natural wetlands) tend to have less 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 exudates7 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 influences CH4 emissions, particularly the use of fertilizers with sulfate (Wassmann et al. 2000b;
Linquist et al. 2012). Other environmental variables also impact the methanogenesis process such as soil
temperature and soil type. Soil temperature is an important factor regulating 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 twelve states, including Arkansas, California, Florida, Illinois, Kentucky, Louisiana,
Minnesota, Mississippi, Missouri, New York, South Carolina, Tennessee and Texas. Soil types, rice varieties, and
cultivation practices vary across the United States, but most farmers apply fertilizers and do not harvest crop
residues. In addition, a second, ratoon rice crop is sometimes grown in the Southeast. 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
7 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 5-17
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
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). While a combination of Tier 1 and 3 methods are used to estimate CH4
emissions from rice cultivation across most of the time series, a surrogate data method has been applied to estimate
national emissions in the last few years of this Inventory. National emission estimates based on surrogate data will
be recalculated in a future Inventory using the Tier 1 and 3 methods.
Overall, rice cultivation is a minor source of CH4 emissions in the United States relative to other source categories
(see Table 5-11, Table 5-12, and Figure 5-2). The majority of emission occur in Arkansas, California, Louisiana and
Texas. In 2016, CH4 emissions from rice cultivation were 13.7 MMT CO2 Eq. (549 kt). Annual emissions fluctuate
between 1990 and 2016, and emissions in 2016 represented a 14 percent decrease compared to 1990. Variation in
emissions is largely due to differences in the amount of rice harvested areas over time, which has been decreasing
over the past two decades.
Table 5-11: ChU Emissions from Rice Cultivation (MMT CO2 Eq.)
State
1990
2005
2012
2013
2014
2015
2016
Arkansas
3.3
4.7
3.8
NE
NE
NE
NE
California
2.0
2.1
2.0
NE
NE
NE
NE
Florida
+
0.1
+
NE
NE
NE
NE
Illinois
+
+
+
NE
NE
NE
NE
Kentucky
+
+
+
NE
NE
NE
NE
Louisiana
6.1
6.5
3.9
NE
NE
NE
NE
Minnesota
+
+
+
NE
NE
NE
NE
Mississippi
0.6
0.6
0.5
NE
NE
NE
NE
Missouri
0.3
0.6
0.3
NE
NE
NE
NE
New York
+
+
+
NE
NE
NE
NE
South Carolina
+
+ /'i
+
NE
NE
NE
NE
Tennessee
+
+ '/'**
+
NE
NE
NE
NE
Texas
3.7
2.1
0.9
NE
NE
NE
NE
Total
16.0
16.7
11.3
11.5
12.7
12.3
13.7
+ Does not exceed 0.05 MMT CO2 Eq.
NE (Not Estimated). State-level emissions are not estimated for 2013 through 2016 Inventory, and national
emissions are determined using a surrogate data method.
Note: Totals may not sum due to independent rounding.
Table 5-12: ChU Emissions from Rice Cultivation (kt)
State
1990
2005
2012
2013
2014
2015
2016
Arkansas
132
187
151
NE
NE
NE
NE
California
81
82
81
NE
NE
NE
NE
Florida
+
3
+
NE
NE
NE
NE
Illinois
+
+
+
NE
NE
NE
NE
Kentucky
+
+
+
NE
NE
NE
NE
Louisiana
246
261
156
NE
NE
NE
NE
Minnesota
1
2
1
NE
NE
NE
NE
Mississippi
23
23
19
NE
NE
NE
NE
Missouri
12
22
12
NE
NE
NE
NE
New York
+
+
+
NE
NE
NE
NE
South Carolina
+
+
+
NE
NE
NE
NE
Tennessee
+
+
+
NE
NE
NE
NE
Texas
146
86
34
NE
NE
NE
NE
Total
641
667
453
462
510
493
549
+ Does not exceed 0.5 kt.
NE (Not Estimated). State-level emissions are not estimated for 2013 through 2016 Inventory, and national
emissions are determined using a surrogate data method.
Note: Totals may not sum due to independent rounding.
5-18 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
Figure 5-2: Annual ChU Emissions from Rice Cultivation, 2012 (MMT CO2 Eq./Year)*
MT C02 Eq. ha1 yr
~ < 5
~ 5 to 10
¦ 10 to 15
¦ 15 to 20
¦ > 20
2
3 * Only national-scale emissions are estimated for 2013 through 2016 in this Inventory using the surrogate data method
4 described in the Methodology section, therefore the fine-scale emission patterns in this map are based on the previous
5 Inventory.
6 Methodology
7 The methodology used to estimate CH4 emissions from rice cultivation is based on a combination of IPCC Tier 1
8 and 3 approaches. The Tier 3 method utilizes a process-based model (DAYCENT) to estimate CH4 emissions from
9 rice cultivation (Cheng et al. 2013), and has been tested in the United States (see Annex 3.12) and Asia (Cheng et al.
10 2013, 2014). The model simulates hydrological conditions and thermal regimes, organic matter decomposition, root
11 exudation, rice plant growth and its influence on oxidation of CH4, as well as CH4 transport through the plant and
12 via ebullition (Cheng et al. 2013). The method simulates the influence of organic amendments and rice straw
13 management on methanogenesis in the flooded soils. In addition to CH4 emissions, DAYCENT simulates soil C
14 stock changes and N20 emissions (Parton et al. 1987 and 1998; Del Grosso et al. 2010), and allows for a seamless
15 set of simulations for crop rotations that include both rice and non-rice crops.
16 The Tier 1 method is applied to estimate CH4 emissions from rice when grown in rotation with crops that are not
17 simulated by DAYCENT, such as vegetables and perennial/horticultural crops. The Tier 1 method is also used for
18 areas converted between agriculture (i.e., cropland and grassland) and other land uses, such as forest land, wetland,
19 and settlements. In addition, the Tier 1 method is used to estimate CH4 emissions from organic soils (i.e., Histosols)
20 and from areas with very gravelly, cobbly, or shaley soils (greater than 35 percent by volume). The Tier 3 method
21 using DAY CENT lias not been fully tested for estimating emissions associated with these crops and rotations, land
22 uses, as well as organic soils or cobbly, gravelly, and shaley mineral soils.
23 The Tier 1 method for estimating CH4 emissions from rice production utilizes a default base emission rate and
24 scaling factors (IPCC 2006). The base emission factor represents emissions for continuously flooded fields with no
25 organic amendments. Scaling factors are used to adjust for water management and organic amendments that differ
26 from continuous flooding with no organic amendments. The method accounts for pre-season and growing season
27 flooding; types and amounts of organic amendments; and the number of rice production seasons within a single year
Agriculture 5-19
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
(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).8
Rice cultivation areas are based on cropping and land use histories recorded in the USD A National Resources
Inventory (NRI) survey (USDA-NRCS 2015). The NRI is a statistically-based sample of all non-federal land, and
includes 380,956 survey points of which 1,588 are in locations with rice cultivation at the end of the NRI time
series. The Tier 3 method is used to estimate CH4 emissions from 1,393 of the NRI survey locations, and the
remaining 195 survey locations are estimated with the Tier 1 method. Each NRI survey point is associated with an
"expansion factor" that allows scaling of CH4 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 sample point). 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 2012 (USDA-NRCS 2015). The current Inventory only uses NRI data through
2012 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 5-13.
Table 5-13: Rice Area Harvested (1,000 Hectares)
State/Crop
1990
2005
2012
2013
2014
2015
2016
Arkansas
599
796
613
NE
NE
NE
NE
California
248
24"
244
NE
NE
NE
NE
Florida
0
11
0
NE
NE
NE
NE
Illinois
0
0
0
NE
NE
NE
NE
Kentucky
0
0
0
NE
NE
NE
NE
Louisiana
380
402
226
NE
NE
NE
NE
Minnesota
4
10
6
NE
NE
NE
NE
Mississippi
119
115
92
NE
NE
NE
NE
Missouri
47
93
46
NE
NE
NE
NE
New York
1
0
0
NE
NE
NE
NE
South Carolina
0
0
0
NE
NE
NE
NE
Tennessee
0
1
0
NE
NE
NE
NE
Texas
300
150
66
NE
NE
NE
NE
Total
1,698
1,82ft
1,292
NE
NE
NE
NE
NE (Not Estimated).
Notes: Totals may not sum due to independent rounding. States are included if NRI reports an area of rice
production in the state at any time between 1990 and 2012. Rice harvested area data have not been compiled
for 2013 to 2016.
The Southeastern states have sufficient growing periods for a ratoon crop in some years. For example, in Arkansas,
the length of growing season is occa sionally sufficient for ratoon crops on an average of 1 percent of the rice fields.
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 foryears 1993 through 2014), 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. Ratooned crop area as a percent of primary crop area is presented in
Table 5-14.
Table 5-14: Average Ratooned Area as Percent of Primary Growth Area (Percent)
State 1990-2012
Arkansas8 1%
California 0%
Floridab 49%
Louisiana0 32%
Mississippi3 1%
8 See .
5-20 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Missouri3 0%
Texas'1 45%
a Arkansas: 1990-2000 (Slaton 1999 through 2001); 2001-2011 (Wilson 2002 through 2007,2009 through 2012); 2012-2013
(Hardke 2013, 2014).
b Florida - Ratoon: 1990-2000 (Schueneman 1997, 1999 through 2001); 2001 (Deren2002); 2002-2003 (Kirstein 2003 through
2004,2006); 2004 (Cantens 2004 through 2005); 2005-2013 (Gonzalez 2007 through 2014)
cLouisiana: 1990-2013 (Linscombe 1999,2001 through2014).
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 CH4 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 the winter
flood. 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 period.
A surrogate data method is used to estimate emissions for the land area associated with the Tier 3 method for 2013
to 2016, and for the land areas associated with the Tier 1 method for 2016. Specifically, a linear regression model
with autoregressive moving-average (ARMA) errors was used to estimate the relationship between the surrogate
data and the 1990-2012 emissions data that was derived using the Tier 1 and 3 methods (Brockwell and Davis
2016). Surrogate data for this model are based on rice commodity statistics from USDA-NASS
(https://quickstats.nass.usda.gov/). See Box 5-3 for more information about the surrogate data method.
Box 5-3: Surrogate Data Method
An approach to extend the time series is needed to estimate emissions from Rice cultivation because the Inventory is
only fully re-compiled every two years for the Agriculture, Forestry, and Other Land Use (AFOLU) sector as part of
the biennial update reporting process, and even in years that the Inventory is fully re-compiled, there are typically
gaps 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 lias 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 2012 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+ e,
where Y is the response variable (e.g., soil organic carbon), X(3 contains specific surrogate data depending on the
response variable, and e is the remaining unexplained error. We tested models with a variety of surrogate data,
including commodity statistics, weather data, or other relevant information. Parameters are estimated from the
observed data for 1990 to 2012 using standard statistical techniques, and these estimates are used to predict the
missing emissions data for 2013 to 2016.
A critical issue in using splicing methods in general, is to adequately account for the additional uncertainty
introduced by predicting emissions with related information without compiling the full inventory. For example,
predicting CH4 emissions will increase the total variation in the emission estimates for these specific years.
Agriculture 5-21
-------
1 compared to those years in which the full inventory is compiled. This added uncertainty is quantified within the
2 model framework using a Monte Carlo approach. The approach requires estimating parameters for results in each
3 Monte Carlo simulation for the full inventory (i.e., the surrogate data model is refit with the emissions estimated in
4 each Monte Carlo iteration from the full inventory analysis with data from 1990 to 2012).
5
6 Uncertainty and Time-Series Consistency
7 Sources of uncertainty in the Tier 3 method include management practices, uncertainties in model structure (i.e.,
8 algorithms and parameterization), and variance associated with the NRI sample. Sources of uncertainty in the IPCC
9 (2006) Tier 1 method include the emission factors, management practices, and variance associated with the NRI
10 sample. A Monte Carlo analysis was used to propagate uncertainties in the Tier 1 and 3 methods. For 2013 to 2016,
11 there is additional uncertainty propagated through the Monte Carlo Analysis associated with the surrogate data
12 method. (See Box 5-3 for information about propagating uncertainty with the surrogate data method.) The
13 uncertainties from the Tier 1 and 3 approaches are combined to produce the final CH4 emissions estimate using
14 simple error propagation (IPCC 2006). Additional details on the uncertainty methods are provided in Annex 3.12.
15 Rice cultivation CH4 emissions in 2016 were estimated to be between 9.3 and 22.5 MMT CO2 Eq. at a 95 percent
16 confidence level, which indicates a range of 32 percent below to 64 percent above the actual 2016 emission estimate
17 of 13.7 MMT C02 Eq. (see Table 5-15).
18 Table 5-15: Approach 2 Quantitative Uncertainty Estimates for ChU Emissions from Rice
19 Cultivation (MMT CO2 Eq. and Percent)
Source
Inventory
Method
Gas
2016 Emission
Estimate
Uncertainty Range Relative to Emission
Estimate3
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Bound
Upper
Bound
Lower Upper
Bound Bound
Rice Cultivation
Tier 3
CH4
11.9
7.7
16.2
-36% +36%
Rice Cultivation
Tier 1
ch4
1.8
0.8
2.8
-55% +55%
Rice Cultivation
Total
ch4
13.7
9.3
22.5
-32% +64%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
20 QA/QC and Verification
21 Quality control measures include checking input data, model scripts, and results to ensure data are properly handled
22 throughout the inventory process. Inventory reporting forms and text are reviewed and revised as needed to correct
23 transcription errors. No errors were found in the reporting forms and text.
24 Model results are compared to field measurements to verily if results adequately represent CH4 emissions. The
25 comparisons included over 15 long-term experiments, representing about 80 combinations of management
26 treatments across all of the sites. A statistical relationship was developed to assess uncertainties in the model
27 structure, adjusting the estimates for model bias and assessing precision in the resulting estimates (methods are
28 described in Ogle et al. 2007). See Annex 3.12 for more information.
29 Recalculations Discussion
30 The rice CH4 emissions data in this Inventory were not recalculated from the previous Inventory with the exception
31 of 2013 through 2015, which were estimated using the surrogate data method (Box 5-3). This change resulted in an
32 increase in emissions of less than 1 percent on average across the time series relative to the previous Inventory.
5-22 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Planned Improvements
New land representation data and rice cultivation data were not compiled for the current Inventory. A surrogate data
method has been 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 a future Inventory will be to update the time
series for CH4 emissions from rice cultivation by compiling the latest land use data and related management
statistics.
In addition, a major improvement is underway to update the time series of management data with information from
the USDA-NRCS Conservation Effects Assessment Program (CEAP). This improvement will fill several gaps in the
management data including more specific data on fertilizer rates, updated tillage practices, water management,
organic amendments and more information on planting and harvesting dates. This improvement is expected to be
completed for the 1990 through 2017 Inventory (i.e., 2019 submission). However, the timeline may be extended if
there are insufficient resources to fund this improvement.
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).9 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
atmosphere.10 A number of agricultural activities increase mineral N availability in soils that lead to direct N20
emissions from nitrification and denitrification at the site of a management activity (see Figure 5-3) (Mosier et al.
1998), including N fertilization; application of managed livestock manure and other organic materials such as
biosolids (i.e., sewage sludge); deposition of manure on soils by domesticated animals in pastures, rangelands, and
paddocks (PRP) (i.e., by grazing animals and other animals whose manure is not managed); production of N-fixing
crops and forages; retention of crop residues; and drainage of organic soils (i.e., soils with a high organic matter
content, otherwise known as Histosols11) (IPCC 2006). Additionally, agricultural soil management activities,
including irrigation, drainage, tillage practices, and fallowing of land, can influence N mineralization from soil
organic matter and levels of asymbiotic N fixation by impacting moisture and temperature regimes in soils. Indirect
emissions of N20 occur when N is transported from a site and is subsequently converted to N20; there are two
pathways for indirect emissions: (1) volatilization and subsequent atmospheric deposition of applied/mineralized N,
and (2) surface runoff and leaching of applied/mineralized N into groundwater and surface water.12 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; N20 emissions from Forest Land and Settlements soils are found
in Sections 6.2 and 6.10, respectively).
9 Nitrification and denitrification are driven by the activity of microorganisms in soils. Nitrification is the aerobic microbial
oxidation of ammonium (NH4+) to nitrate (NO3"), and denitrification is the anaerobic microbial reduction of nitrate to N2. Nitrous
oxide is a gaseous intermediate product in the reaction sequence of denitrification, which leaks from microbial cells into the soil
and then into the atmosphere. Nitrous oxide is also produced during nitrification, although by a less well-understood mechanism
(Nevison 2000).
10 Asymbiotic N fixation is the fixation of atmospheric N2 by bacteria living in soils that do not have a direct relationship with
plants.
11 Drainage of organic soils in former wetlands enhances mineralization of N-rich organic matter, thereby increasing N2O
emissions from these soils.
12 These processes entail volatilization of applied or mineralized N as NH3 and NOx, transformation of these gases within the
atmosphere (or upon deposition), and deposition of the N primarily in the form of particulate NH4+, nitric acid (HNO3), and NOx,
in addition to leaching and runoff of NO3" that is converted to N2O in aquatic systems.
Agriculture 5-23
-------
Figure 5-3: Sources and Pathways of N that Result in N2O Emissions from Agricultural Soil
Management
Sources and Pathways of N that Result in N2O Emissions from Agricultural Soil Management
N Volatilization
RERniTZEB
SyntheticN Fertilizers
Synthetic N fertilizer appl ied to soil
Organic
Amendments
Indudesboth commercial and
non-co,m mercislfertilizers (i.e.,
animal manure compost,
sewage sludge, tankage, etc.)
Urine and Dung from
Grazing Animals
Manure deposited on pasture range,
and paddock
Crop Residues
Includes above-and belowground
residues for all crops (non-N and In-
fixing (and from perennial forage
crops and pastures following renewal
Mineralization of
Soil Organic Matter
•
Includes N converted to mineral form
upon decomposition of soil organic
matter
Asymbiotic Fixation
Fixation of atm ospheric N2 by bacteria
living in soils that do not have a direct
relationship with plants
N Flows:
0
N Inputs to
Managed Soils
Direct N20
Emissions
N Volatilization
and Deposition
0
Indirect N20
Emissions
" Histosol
Cultivation
This graphic illustrates the sources and pathways of nitrogen that result
in direct and indirect N,0 emissions from soils using the methodologies
described in this Inventory. Emission pathways are shown with arrows.
On the lower right-hand side is a cut-away view of a representative
section of a managed soil: histosol cultivation is represented here.
:^g r
Surface
W
Groundwater
5-24 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Agricultural soils produce the majority of N20 emissions in the United States. Estimated emissions from this source
2 in 2016 are 283.6 MMT CO2 Eq. (952 kt) (see Table 5-16 and Table 5-17). Annual N2O emissions from agricultural
3 soils are 13.2 percent greater in the 2016 compared to 1990, but emissions fluctuated between 1990 and 2016 due to
4 inter-annual variability largely associated with weather patterns, synthetic fertilizer use, and crop production. From
5 1990 to 2016, on average, cropland accounted for approximately 70 percent of total direct emissions, while
6 grassland accounted for approximately 30 percent. On average, approximately 81 percent of indirect emissions are
7 from croplands and 19 percent from grasslands. Estimated direct and indirect N20 emissions by sub-source category
8 are shown in Table 5-18 and Table 5-19.
9 Table 5-16: N2O Emissions from Agricultural Soils (MMT CO2 Eq.)
Activity
1990
2005
2012
2013
2014
2015
2016
Direct
212.0
218.5
215.6
233.3
231.4
244.5
237.6
Cropland
147.5
153.9
156.7
165.5
165.1
169.3
168.0
Grassland
64.5
64.6
59.0
67.9
66.3
75.3
69.9
Indirect
38.5
35.0
32.3
43.3
42.6
50.5
45.9
Cropland
30.9
28.0
25.4
35.5
34.9
41.9
37.9
Grassland
7.6
7.0
6.9
7.8
7.7
8.6
8.1
Total
250.5
253.5
247.9
276.6
274.0
295.0
283.6
Notes: Estimates after 2012 are based on a data splicing method (See Methodology section). Totals
may not sum due to independent rounding.
10 Table 5-17: N2O Emissions from Agricultural Soils (kt)
Activity
1990
2005
2012
2013
2014
2015
2016
Direct
711
733
724
783
777
821
797
Cropland
495
516
526
555
554
568
564
Grassland
217
217
198
228
223
253
234
Indirect
129
118
108
145
143
169
154
Cropland
104
94
85
119
117
140
127
Grassland
26
23
23
26
26
29
27
Total
840
851
832
928
920
990
952
Notes: Estimates after 2012 are based on a data splicing method (See Methodology section). Totals
may not sum due to independent rounding.
11 Table 5-18: Direct N2O Emissions from Agricultural Soils by Land Use Type and N Input Type
12 (MMT CO2 Eq.)
Activity
1990
2005
2012
2013
2014
2015
2016
Cropland
147.5
153.9
156.7
165.5
165.1
169.3
168.0
Mineral Soils
144.1
150.6
153.5
161.9
161.6
165.8
164.6
Synthetic Fertilizer
53.6
54.6
60.4
63.3
62.4
64.1
63.6
Organic Amendment3
10.0
10.9
11.3
10.8
10.4
10.7
10.6
Residue Nb
22.1
22.9
23.5
25.3
25.7
26.4
26.3
Mineralization and
Asymbiotic Fixation
58.4
62.2
58.2
62.6
63.1
64.6
64.1
Drained Organic Soils
3.3
3.3
3.2
3.6
3.5
3.5
3.5
Grassland
64.5
64.6
59.0
67.9
66.3
75.3
69.6
Mineral Soils
61.3
61.1
55.7
64.3
62.7
71.7
66.0
Synthetic Fertilizer
0.9
0.8
0.7
0.9
0.8
1.0
0.9
PRP Manure
16.1
13.8
13.3
15.2
14.8
16.0
15.2
Managed Manure0
0.9
U
1.1
1.2
1.2
1.4
1.3
Biosolids (i.e., Sewage
Sludge)
0.2
0.5
0.6
0.6
0.6
0.6
0.6
Residue Nd
14.4
15.8
14.2
16.5
16.1
18.7
17.0
Mineralization and
Asymbiotic Fixation
28.5
29.2
25.8
29.9
29.3
34.0
31.0
Drained Organic Soils
3.3
3.5
3.3
3.6
3.6
3.6
3.6
Total
212.0
218.5
215.6
233.3
231.4
244.5
237.6
Agriculture 5-25
-------
a Organic amendment inputs include managed manure, daily spread manure, and commercial organic
fertilizers (i.e., dried blood, dried manure, tankage, compost, and other).
b Cropland residue N inputs include N in unharvested 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.
d Grassland residue N inputs include N in ungrazed legumes as well as ungrazed grass residue N.
Notes: Estimates after 2012 are based on a data splicing method (See Methodology section). Totals may not
sum due to independent rounding.
1 Table 5-19: Indirect N2O Emissions from Agricultural Soils (MMT CO2 Eq.)
Activity
1990
2005
2012
2013
2014
2015
2016
Cropland
30.9
28.0
25.4
35.5
34.9
41.9
37.9
Volatilization & Atm.
Deposition
5.9
6.6
6.5
7.0
7.0
7.1
6.9
Surface Leaching & Run-Off
25.0
21.4
18.9
28.5
27.9
34.8
30.9
Grassland
7.6
7.0
6.9
7.8
7.7
8.6
8.1
Volatilization & Atm.
Deposition
4.4
4.5
4.2
4.6
4.6
5.1
4.8
Surface Leaching & Run-Off
3.2
2.5
2.6
3.1
3.1
3.5
3.3
Total
38.5
35.0
32.3
43.3
42.6
50.5
45.9
Notes: Estimates after 2012 are based on a data splicing method (See Methodology section). Totals may not
sum due to independent rounding.
2 Figure 5-4 and Figure 5-5 show regional patterns for direct N20 emissions, Figure 5-6 and Figure 5-7 show indirect
3 N20 emissions from volatilization, and Figure 5-8 and Figure 5-9 show the indirect N20 emissions from leaching
4 and runoff in croplands and grasslands, respectively. Annual emissions in 201213 are shown for the Tier 3 Approach
5 only.
13 Only national-scale emissions are estimated for 2013 to 2016 in the current Inventory using the splicing method, and therefore
the fine-scale emission patterns in these maps are based on Inventory data from 2012.
5-26 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Figure 5-4: Crops, 2012 Annual Direct N2O Emissions Estimated Using the Tier 3 DAYCENT
2 Model (MMT CO2 Eq./year)*
3
4 * Only national-scale emissions are estimated for 2013 to 2016 using a splicing method, and therefore the fine-scale
5 emission patterns in this map are based on Inventory data from 2012.
6 Direct N;0 emissions from croplands occur throughout all of the cropland regions but tend to be high in the
7 Midwestern Com Belt Region (Illinois, Iowa, Indiana, Ohio, southern Mirmesota and Wisconsin, and eastern
8 Nebraska), where a large portion of the land is used for growing highly fertilized corn and N-fixing soybean crops
9 (see Figure 5-4). Emissions are also high in the Lower Mississippi River Basin from Missouri to Louisiana, and
10 highly productive irrigated areas, such as Platte River, which flows from Colorado through Nebraska. Snake River
11 Valley in Idaho and the Central Valley in California. Direct emissions are low in many parts of the eastern United
12 States because only a small portion of land is cultivated as well as in many western states where rainfall and access
13 to irrigation water are limited.
14 Direct emissions from grasslands are highest in the southeast, particularly Kentucky and Tennessee, in addition to
15 areas in east Texas and Iowa, where there tends to be higher rates of manure amendments on a relatively small
16 amount of pasture, compared to other regions of the United States. However, total emissions from grasslands tend to
17 be higher in the Great Plains and western United States (see Figure 5-5) where a high proportion of the land is
18 dominated by grasslands and used for cattle grazing.
Agriculture 5-27
-------
1
2
3
4
5
6
7
8
9
10
11
12
Figure 5-5: Grasslands, 2012 Annual Direct N2O Emissions Estimated Using the Tier 3
DAYCENT Model (MMT CO2 Eq./year)*
* Only national-scale emissions are estimated for 2013 to 2016 using a splicing method, and therefore the fine-scale
emission patterns in this map are based on Inventory data from 2012.
Indirect N<) emissions from volatilization in croplands have a similar pattern as the direct N:0 emissions with high
emissions in the Midwestern Corn Belt and Lower Mississippi River Basin. Indirect N20 emissions from
volatilization in grasslands are higher in the Southeastern United States than in other regions. The higher emissions
in this region are mainly due to highly productive pastures that support intensive grazing, which in turn, stimulates
NH3 volatilization. Indirect N20 emissions from surface runoff and leaching of applied/mineralized N is highest in
the Eastern United States for both croplands and grasslands. This region has greater precipitation and higher levels
of leaching and runoff compared to the more arid region in the Western United States.
5-28 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
Figure 5-6: Crops, 2012 Annual Indirect N2O Emissions from Volatilization Using the Tier 3
DAYCENT Model (MMT CO2 Eq./year)*
:
MT C02 Eq. ha-1 yr
~ 0.1 -0.25 ¦ 1.5-2
¦ 0.25 - 0.5 ¦ > 2
* Only national-scale emissions are estimated for 2013 to 2016 using a splicing method, and therefore the fine-scale
emission patterns in this map are based on Inventory data from 2012.
Agriculture 5-29
-------
1 Figure 5-7: Grasslands, 2012 Annual Indirect N2O Emissions from Volatilization Using the
2 Tier 3 DAYCENT Model (MMT CO2 Eq./year)*
MT C02 Eq. ha1 yr1
~ <0.01 ¦ 0.25 to 0.5
~ 0.01 to 0.05 ¦ 0.5 to 1
¦ 0.1 to 0.25 m>2
4 * Only national-scale emissions are estimated for 2013 to 2016 using a splicing method, and therefore the tine-scale
5 emission patterns in this map are based on Inventory data from 2012.
5-30 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Figure 5-8: Crops, 2012 Annual Indirect N2O Emissions from Leaching and Runoff Using the
2 Tier 3 DAYCENT Model (MMT CO2 Eq./year)*
4 * Only national-scale emissions are estimated for 2013 to 2016 using a splicing method, and therefore the fine-scale
5 emission patterns in this map are based on Inventory data from 2012.
Agriculture 5-31
-------
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Figure 5-9: Grasslands, 2012 Annual Indirect N2O Emissions from Leaching and Runoff
Using the Tier 3 DAYCENT Model (MMT CO2 Eq./year)*
* Only national-scale emissions are estimated for 2013 to 2016 using a splicing method, and therefore the fine-scale
emission patterns in this map are based on Inventory data from 2012.
Methodology
The 2006 IPCC 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., sewage sludge) applications, crop residues, organic amendments, and biological
N fixation associated with planting of legumes on cropland and grassland soils; (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 N;0.
In this source category, the United States reports on all croplands, as well as all "managed" grasslands, whereby
management of a land use implies there are anthropogenic impacts on greenhouse gas emissions (IPCC 2006),
including direct and indirect N20 emissions from asymbiotic fixation14 and mineralization of soil organic matter and
litter. One recommendation from IPCC (2006) that lias 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 N;0 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
14 N inputs from asymbiotic N fixation are not directly addressed in 2006 IPCC Guidelines, but are a component of the total
emissions from managed lands and are included in the Tier 3 approach developed for this source.
5-32 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
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 with the exception of biosolids (i.e., sewage sludge)
amendments (Del Grosso et al. 2010). The Tier 3 approach has been specifically designed and tested to estimate
N20 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 5-4 for further elaboration). Moreover, the Tier 3 approach
allows for the Inventory to address direct N20 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 N20) in a single inventory analysis ensures that there is consistent activity data and treatment
of the processes, and interactions are taken into account between C and N cycling in soils.
The Tier 3 approach is based on the cropping and land use histories recorded in the USDA National Resources
Inventory (NRI) (USDA-NRCS 2015). The NRI is a statistically-based sample of all non-federal land,15 and
includes 363,286 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 the remaining 205,487 in the NRI survey that
are designated as cropland or grassland (discussed later in this section). Each point is associated with an "expansion
factor" that allows scaling of N20 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 sample point). Each NRI point 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, and the annual data are currently available through 2012 (USDA-NRCS
2015).
Box 5-4: Tier 1 vs. Tier 3 Approach for Estimating N2O Emission
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 N20 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 fortius Inventory employs 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.
Consequently, the Tier 3 approach produces more accurate estimates; it accounts more comprehensively for land-use
and management impacts and their interaction with environmental factors (i.e., weather patterns and soil
characteristics), which will enhance or dampen anthropogenic influences. However, the Tier 3 approach requires
more detailed activity data (e.g., crop-specific N amendment rates), additional data inputs (i.e., daily weather, soil
types, etc.), 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 N20 emissions only during that year and cannot be stored in
soils and contribute to N20 emissions in subsequent years. This is a simplifying assumption that is likely to create
bias in estimated N20 emissions for a specific year. In contrast, the process-based model used 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 N20 during subsequent years.
DAYCENT is used to estimate N20 emissions associated with production of alfalfa hay, barley, corn, cotton, dry
beans, grass hay, grass-clover hay, lentils, oats, onions, peanuts, peas, potatoes, rice, sorghum, soybeans, sugar
beets, sunflowers, tobacco, tomatoes, and wheat, but is not applied to estimate N20 emissions from other crops or
15 The NRI survey does include sample points on federal lands, but the program does not collect data from those sample
locations.
Agriculture 5-33
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
rotations with other crops,16 such as sugarcane, some vegetables, tobacco, 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 N20 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, the Tier 1IPCC (2006) methodology is used to estimate (1) direct emissions from crops on mineral
soils that are not simulated by DAYCENT; (2) direct emissions from PRP on federal grasslands; and (3) direct
emissions from drained organic soils in croplands and grasslands.
A splicing method is used to estimate soil N20 emissions from 2013 to 2016 at the national scale as an alternative to
the Tier 1 and Tier 3 methods because new NRI activity data are not available 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 2012 emissions that are derived using the Tier 3
methods. Surrogate data for these regression models include corn and soybean yields from USDA-NASS statistics
(https://quickstats.nass.usda.gov/), and weather data from the PRISM Climate Group (PRISM 2015). For the Tier 1
method, a linear-time series model is used to estimate emissions from 2013 to 2016 without surrogate data. See Box
5-5 for more information about the splicing method. Emission estimates for 2013 to 2016 will be recalculated in
future Inventory reports when new NRI data are available.
Box 5-5: Surrogate Data Method
J!
An approach to extend the time series is needed for Agricultural Soil Management because the Inventory is only
fully re-compiled every two years for many categories in the AFOLU sector as part of the biennial update reporting
process, and even in years that the Inventory is re-compiled fully with the Tier 1 and 3 methods, there are typically
gaps at the end of the time series. This is mainly because the National Resources Inventory (NRI), 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 2012 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 + e,
where Y is the response variable (e.g., soil organic carbon), xp for the Tier 3 data contains specific surrogate data
depending on the response variable, and e is the remaining unexplained error. Models with a variety of surrogate
data were tested, including commodity statistics, weather data, or other relevant information, xp for the Tier 1 data
only contains year as a predictor of emission patterns over the time series, and therefore, is a linear time series
model with no surrogate data. Parameters are estimated from the emissions data for 1990 to 2012 using standard
statistical techniques, and these estimates are used in the model described above to predict the missing emissions
data for 2013 to 2016.
A critical issue when applying splicing methods is to account for the additional uncertainty introduced by predicting
emissions with related information without compiling the full inventory. Specifically, uncertainty will increase for
years with imputed estimates based on the splicing methods, compared to those years in which the full inventory is
compiled. This added uncertainty is quantified within the model framework using a Monte Carlo approach. We
combine the uncertainty from the original inventory data produced with the Tier 1 and 3 methods, with the
uncertainty in the parameters from the linear regression model. Specifically, the original inventory data are derived
through a series of random draws from probability distribution functions that produce multiple results (e.g., 100
results are produced with the DAYCENT simulations for the Tier 3 method). In order to propagate the uncertainty
16 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.
5-34 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
from the original Monte Carlo analysis, a separate linear regression model is derived for each result from the Monte
Carlo Analysis (i.e., 100 linear regression models are produced for the Tier 3 method). For each linear regression
model, we randomly select parameters and apply the model to estimate emissions, and in turn, have multiple
estimates of N20 emissions for 2013 to 2016 associated with each of the Monte Carlo results produced for the 1990
to 2012 time series.
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 N20 emissions from mineral cropland soils that are managed for production of a wide variety of crops (see list
in previous paragraph) based on the cropping histories in the 2012 NRI (USDA-NRCS 2015). Crops simulated by
DAYCENT are grown on approximately 91 percent of total cropland area in the United States. For agricultural
systems in the central region of the United States, crop production for key crops (i.e., corn, soybeans, sorghum,
cotton, and wheat) is simulated in DAYCENT with a NASA-CASA production algorithm (Potter et al. 1993; Potter
et al. 2007) using the Moderate Resolution Imaging Spectroradiometer (MODIS) Enhanced Vegetation Index (EVI)
products, MOD13Q1 and MYD13Q1, with a pixel resolution of 250m.17
DAYCENT is used to estimate direct N20 emissions due to mineral N available from the following sources: (1) the
application of synthetic fertilizers; (2) the application of livestock manure; (3) the retention of crop residues 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 soil organic matter; and (5) asymbiotic fixation. Note
that commercial organic fertilizers (TVA 1991 through 1994; AAPFCO 1995 through 2016) are addressed with the
Tier 1 method because county-level application data would be needed to simulate applications in DAYCENT, and
currently data are only available at the national scale. The third and fourth sources are generated internally by the
DAYCENT model.
Synthetic fertilizer data are based on fertilizer use and rates by crop type for different regions of the United States
and are obtained primarily from the USD A Economic Research Service. The data collection program was known as
the Cropping Practices Surveys through 1995 (USDA-ERS 1997), and then became the Agricultural Resource
Management Surveys (ARMS) (USDA-ERS 2015). Additional data are compiled through other sources particularly
the National Agricultural Statistics Service (NASS 1992, 1999, 2004). Frequency and rates of livestock manure
application to cropland during 1997 are estimated from data compiled by the USDA Natural Resources
Conservation Service (Edmonds et al. 2003), and then adjusted using county-level estimates of manure available for
application in other years. The adjustments are based on county-scale ratios of manure available for application to
soils in other years relative to 1997 (see Annex 3.12 for further details). Greater availability of managed manure N
relative to 1997 is assumed to increase the area amended with manure, while reduced availability of manure N
relative to 1997 is assumed to reduce the amended area. Data on the county-level N available for application is
estimated for managed manure systems based on the total amount of N excreted in manure minus N losses during
storage and transport, and including the addition of N from bedding materials. Nitrogen losses include direct N20
emissions, volatilization of ammonia and NOx, runoff and leaching, and poultry manure used as a feed supplement.
For unmanaged manure systems, it is assumed that no N losses or additions occur prior to the application of manure
to the soil. More information on livestock manure production is available in Section 5.2 Manure Management and
Annex 3.11.
The IPCC approach 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
N20 emissions, but these are not model inputs. The N20 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-C02 greenhouse gas emissions from agricultural residue burning. The estimate of residue
17 See .
Agriculture 5-35
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
burning is based on state inventory data (ILENR 1993; Oregon Department of Energy 1995; Noller 1996; Wisconsin
Department of Natural Resources 1993; Cibrowski 1996).
Additional sources of data are used to supplement the mineral N (USDA-ERS 1997, 2011), livestock manure
(Edmonds et al. 2003), and land-use information (USDA-NRCS 2015). The Conservation Technology Information
Center (CTIC 2004) provides annual data on tillage activity with adjustments for long-term adoption of no-till
agriculture (Towery 2001). Tillage has an influence on soil organic matter decomposition and subsequent soil N20
emissions. The time series of tillage data from CTIC began in 1989 and ended in 2004, so further changes in tillage
practices since 2004 are not currently captured in the Inventory and practices used in 2004 are assumed to apply for
subsequent years. Daily weather data are used as an input in the model simulations, based on gridded weather data at
a 32 km scale from the North America Regional Reanalysis Product (NARR) (Mesinger et al. 2006). Soil attributes
are obtained from the Soil Survey Geographic Database (SSURGO) (Soil Survey Staff 2011).
Each NRI point is run 100 times as part of the uncertainty assessment, yielding a total of over 18 million simulations
for the analysis. Soil N20 emission estimates from DAYCENT are adjusted using a structural uncertainty estimator
to account for uncertainty in model algorithms and parameter values (Del Grosso et al. 2010). Soil N20 emissions
and associated 95 percent confidence intervals are estimated for each year between 1990 and 2012, but emissions
from 2013 to 2016 are estimated using a splicing method that accounts for uncertainty in the original inventory data
and the splicing method (See Box 5-5). Annual data are currently available through 2012 (USDA-NRCS 2015), and
the Inventory time series will be updated in the future when new NRI data are released.
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 N20 production (nitrification and denitrification). It is not
possible to partition N20 emissions into each anthropogenic activity directly from model outputs due to the
complexity of the interactions (e.g., N20 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 N20 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 N20, 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 N20 emissions with individual sources of N.
Tier 1 Approach for Mineral Cropland Soils
The IPCC (2006) Tier 1 methodology is used to estimate direct N20 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 Approach, estimates of direct N20 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.18 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:
18 Commercial organic fertilizers include dried blood, tankage, compost, and other, but the dried manure and biosolids (i.e.,
sewage sludge) is removed from the dataset in order to avoid double counting with other datasets that are used for manure N and
biosolids.
5-36 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
• A process-of-elimination approach is used to estimate synthetic N fertilizer additions for crop areas not
simulated by DAYCENT. The total amount of fertilizer used on farms has been estimated at the county- level
by the USGS from sales records (Ruddy et al. 2006), and these data are aggregated to obtain state-level N
additions to farms. For 2002 through 2012, state-level fertilizer for on-farm use is adjusted based on annual
fluctuations in total U.S. fertilizer sales (AAPFCO 1995 through 2007, 2008 through 2012). After subtracting
the portion of fertilizer applied to crops and grasslands simulated by DAYCENT (see Tier 3 Approach for
Mineral Cropland Soils and Direct N20 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 amount of manure N applied in the Tier 3 approach to crops and grasslands is
subtracted from total manure N available for land application (see Tier 3 Approach for Mineral Cropland Soils
and Direct N20 Emissions from Grassland Soils sections for information on data sources), and 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 to units of N using average organic fertilizer N content (TVA 1991 through 1994; AAPFCO 1995
through 2012). Commercial fertilizers do include some manure and biosolids (i.e., sewage sludge), but the
amounts are removed from the commercial fertilizer data to avoid double counting with the manure N dataset
described above and the biosolids (i.e., 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 2013), crop production yield statistics (USDA-NASS 2015), 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 increase in soil mineral N from applied fertilizers and crop residues is multiplied by the IPCC (2006)
default emission factor to derive an estimate of direct N20 emissions using the Tier 1 Approach.
Tier 1 soil N20 emissions from 2013 to 2016 are estimated using a splicing method that is described in Box 5-5. As
with the Tier 3 method, the time series will be recalculated in future Inventory reports when new activity data are
compiled (see Planned Improvements section).
Tier 1 Approach for Drainage of Organic Soils in Croplands and Grasslands
The IPCC (2006) Tier 1 methods are used to estimate direct N20 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 2012 NRI (USDA-NRCS 2015) using soils data from the Soil Survey Geographic Database (SSURGO)
(Soil Survey Staff 2011). Temperature data from Daly et al. (1994 and 1998) are used to subdivide areas into
temperate and tropical climates using 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 2012. Consequently, emissions from 2013 to 2016 are estimated using a linear time
series model (see Box 5-5). Estimates for 2013 to 2016 will be recalculated in future Inventory reports when new
NRI data are available.
Tier 1 and 3 Approaches for Direct N2O Emissions from Grassland Soils
As with N20 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 forage primarily 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, fertilization, or interseeding legumes. DAYCENT is used to simulate N20 emissions from NRI
survey locations (USDA-NRCS 2015) 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
Agriculture 5-37
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
amendments (i.e., manure other than PRP manure such as Daily Spread), 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 Carbon
Sequestration Rural Appraisals (CSRA) conducted by the USDA-NRCS (USDA-NRCS, unpublished data). The
CSRA was a solicitation of expert knowledge from USDA-NRCS staff throughout the United States to support the
Inventory. Managed manure N amendments to grasslands are estimated from Edmonds et al. (2003) and adjusted for
annual variation using data on the availability of managed manure N for application to soils, according to methods
described in the Manure Management section (Section 5.2) and Annex 3.11. Biological N fixation is simulated
within DAY CENT, and therefore is not an input to the model.
Manure N deposition from grazing animals in PRP systems (i.e., PRP manure) is another key input of N to
grasslands. The amounts of PRP manure N applied on non-federal grasslands for each NRI point are based on
amount of N excreted by livestock in PRP systems. 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 input rates are used in the
DAYCENT simulations. DAYCENT simulations of non-federal grasslands accounted for approximately 78 percent
of total PRP manure N in aggregate across the country. The remainder of the PRP manure N in each state is assumed
to be excreted on federal grasslands, and the N20 emissions are estimated using the IPCC (2006) Tier 1 method with
IPCC default emission factors.
Biosolids (i.e., sewage sludge) are assumed to be applied on grasslands because of the heavy metal content and other
pollutants in human waste that limit its use as an amendment to croplands. Biosolids application 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 sewage sludge available for land application
application). Biosolids soil amendments 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 amendments on N20 emissions from grassland soils, and consequently,
emissions from biosolids are estimated using the IPCC (2006) Tier 1 method.
As previously mentioned, each NRI point is simulated 100 times as part of the uncertainty assessment, yielding a
total of over 18 million simulation runs for the analysis. Soil N20 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). N20 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.
Soil N20 emissions and 95 percent confidence intervals are estimated for each year between 1990 and 2012 based
on the Tier 1 and 3 methods, with the exception of biosolids (discussed below), and emissions from 2013 to 2016
are estimated using a splicing method as described in Box 5-5. As with croplands, estimates for 2013 to 2016 will be
recalculated in future inventories when new NRI data are available. Biosolids application data are compiled through
2016 in this Inventory, and therefore soil N20 emissions and confidence intervals are estimated using the Tier 1
method for all years in the time series without application of the splicing method.
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 N20 emissions from agricultural soil management
(see Table 5-16 and Table 5-17).
Indirect N2O Emissions
This section describes the methods used for estimating indirect soil N20 emissions from croplands and grasslands.
Indirect N20 emissions occur when mineral N made available through anthropogenic activity is transported from the
soil either in gaseous or aqueous forms and later converted into N20. 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
5-38 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
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 N20. 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 N20 emissions. Regardless
of the eventual location of the indirect N20 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 N20. 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 N20 emissions in the Tier 3 Approach for
Mineral Cropland Soils and Direct N20 Emissions from Grassland Soils sections. Nitrogen volatilization from all
other areas is estimated using the Tier 1 method and 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, biosolids [i.e.,
sewage sludge] application on grasslands). For the volatilization data generated from both the DAY CENT and Tier
1 approaches, the IPCC (2006) default emission factor is used to estimate indirect N20 emissions occurring due to
re-deposition of the volatilized N (see Table 5-19).
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 N20. 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. This N transport estimate includes N
applications on croplands that are not simulated by DAYCENT, biosolids amendments 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 N20 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 plus irrigation does not exceed 80 percent of PET. 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 N20 emissions that occur in groundwater and waterways (see Table 5-19).
Indirect soil N20 emissions from 2013 to 2016 are estimated using the splicing method that is described in Box 5-5.
As with the direct N20 emissions, the time series will be recalculated in a future Inventory report when new activity
data are compiled (see Planned Improvements section).
Uncertainty and Time-Series Consistency
Uncertainty is estimated for each of the following five components of N20 emissions from agricultural soil
management: (1) direct emissions simulated by DAYCENT; (2) the components of indirect emissions (N volatilized
and leached or runoff) simulated by DAYCENT; (3) direct emissions calculated with the IPCC (2006) Tier 1
method; (4) the components of indirect emissions (N volatilized and leached or runoff) calculated with the IPCC
(2006) Tier 1 method; and (5) indirect emissions estimated with the IPCC (2006) Tier 1 method. Uncertainty in
direct emissions, which account for the majority of N20 emissions from agricultural management, as well as the
components of indirect emissions calculated by DAYCENT are estimated with a Monte Carlo Analysis, addressing
uncertainties in model inputs and structure (i.e., algorithms and parameterization) (Del Grosso et al. 2010). For 2013
to 2016, there is additional uncertainty propagated through the Monte Carlo Analysis associated with the splicing
method (See Box 5-5).
Agriculture 5-39
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
Simple error propagation methods (IPCC 2006) are used to estimate confidence intervals for direct emissions
calculated 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 N20 emissions. Uncertainty in the splicing method is also
included in the error propagation for 2013 to 2016 (see Box 5-5). Additional details on the uncertainty methods are
provided in Annex 3.12. Table 5-20 shows the combined uncertainty for direct soil N20 emissions ranged from 16
percent below to 16 percent above the 2016 emission estimate of 237.6 MMT CO2 Eq., and the combined
uncertainty for indirect soil N20 emissions range from 65 percent below to 154 percent above the 2016 estimate of
45.9 MMT C02 Eq.
Table 5-20: Quantitative Uncertainty Estimates of N2O Emissions from Agricultural Soil
Management in 2016 (MMT CO2 Eq. and Percent)
2016 Emission
Source
Gas
Estimate
Uncertainty Range Relative to Emission Estimate
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Direct Soil N2O Emissions
N2O
237.6
199.2 276.1
-16% 16%
Indirect Soil N2O Emissions
N2O
45.9
16.0 116.8
-65% 154%
Notes: Due to lack of data, uncertainties in managed manure N production, PRP manure N production, other organic
fertilizer amendments, and biosolids (i.e., sewage sludge) amendments to soils are currently treated as certain; these
sources of uncertainty will be included in future Inventory reports.
Additional uncertainty is associated with an incomplete estimation of N20 emissions from managed croplands and
grasslands in Hawaii and Alaska. The Inventory currently includes the N20 emissions from mineral fertilizer 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 commodity cropping states in the conterminous United States, so the
emissions are likely to be small for the other sources of N (e.g., crop residue inputs), which are not currently
included in the Inventory.
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2016. Details on the emission trends through time are described in more detail in the Methodology section.
QA/QC and Verification
DAYCENT results for N20 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 are available for 41 sites, which mostly occur in the United States,
with five in Europe and three in Australia, representing over 200 different combinations of fertilizer treatments and
cultivation practices. Nitrate leaching data are available for four sites in the United States, representing 10 different
combinations of fertilizer amendments/tillage practices. DAYCENT estimates of N20 emissions are closer to
measured values at most sites compared to the IPCC Tier 1 estimate (see Figure 5-10). In general, the IPCC Tier 1
methodology tends to over-estimate emissions when observed values are low and under-estimate emissions when
observed values are high, while DAYCENT estimates have less bias. DAYCENT accounts for key site-level factors
(i.e., weather, soil characteristics, and management) that are not addressed in the IPCC Tier 1 method, and thus the
model is better able to represent the variability in N20 emissions. DAYCENT does have a tendency to under-
estimate very high N20 emission rates; and estimates are adjusted using the statistical model derived from the
comparison of model estimates to measurements (see Annex 3.12 for more information). Regardless, the comparison
demonstrates that DAYCENT provides relatively high predictive capability for N20 emissions, and is an
improvement over the IPCC Tier 1 method.
5-40 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Figure 5-10: Comparison of Measured Emissions at Field Sites and Modeled Emissions Using
the DAYCENT Simulation Model and IPCC Tier 1 Approach (kg N2O per ha per year)
9
8
- 7
L_
>¦ 6
js 5
W) 4
9, 3
2 2
1
0
C
i—
O
~ Measured
E3 DayCent
I IPCC
-O
¦4—'
o
J
C
1—
o
>
HJ,rEu,nJ
illaiiU
>-
o
c
i—
o
Q
03
<1>
CtD
O
u
5 5
>-
c
03
03
it
03
>-
o
- 2
- E
aj
03
-Q
O
u
CUD
Q
c
03
aj
_Q
>•
o
c
03
aj
_Q
>•
o
CtD
<
Q-
03
aj
Q_
¦«—'
03
a>
X
O
O 03 03
t H ^ ^
u &d ao
H < J
03
tn
c
ro
ai
=j
<
Spreadsheets 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. Two errors have been identified through these checks, including
omission of PRP manure N from the indirect soil N20 emissions in Alaska and Hawaii, and double-counting other
organic amendments in the Tier 1 direct N20 emission calculations. Links between spreadsheets have also been
checked, updated, and corrected when necessary. Spreadsheets containing input data, emission factors, and
calculations required for the Tier 1 method have been checked and updated as needed.
Recalculations Discussion
Methodological recalculations in the current Inventory are associated with the following improvements: (1)
estimating emissions from 2013 to 2015 using a splicing method (other than biosolids which are estimated with a
Tier 1 method for the entire time series) (Box 5-5); (2) correcting an omission of PRP manure N input from 1990 to
2012 in Alaska and Hawaii for indirect soil N20 emission; and (3) correcting a double-counting of other organic
amendments from 1990 to 2012 in the Tier 1 method for direct N20 emissions. These changes resulted in an average
decrease in emissions of 0.7 percent from 1990 to 2015 relative to the previous Inventory.
Planned Improvements
New land representation data have not been compiled for this Inventory, and a splicing method lias been 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 a future Inventory will be to recalculate the time series from 2013 to 2016 with
the latest land use data from the National Resources Inventory and related management statistics, particularly data
compiled through the Conservation Effects Assessment Program (discussed below).
Several planned improvements are underway. The DAYCENT biogeochemical model will be improved with 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
nitrification inhibitors and slow-release fertilizers (e.g., polymer-coated fertilizers) on N20 emissions. An improved
representation of drainage as well as freeze-thaw cycles are also under development. Experimental study sites will
Agriculture 5-41
-------
1 continue to be added for quantifying model structural uncertainty. Studies that have continuous (daily)
2 measurements of N20 (e.g., Scheer et al. 2013) will be given priority.
3 The time series of management data will be updated with information from the USDA-NRCS Conservation Effects
4 Assessment Program (CEAP). This improvement will fill several gaps in the management data including more
5 specific data on fertilizer rates, updated tillage practices, and more information on planting and harvesting dates for
6 crops.
7 Improvements are underway to simulate crop residue burning in the DAYCENT model based on the amount of crop
8 residues burned according to the data that is used in the Field Burning of Agricultural Residues source category (see
9 Section 5.7).
10 Alaska and Hawaii are not included for all sources in the current Inventory for agricultural soil management, with
11 the exception of N20 emissions from drained organic soils in croplands and grasslands for Hawaii, synthetic
12 fertilizer and PRP N amendments for grasslands in Alaska and Hawaii. A planned improvement over the next two
13 years is to add the remaining sources for these states into the Inventory analysis.
14 There is also an improvement based on updating the Tier 1 emission factor for N20 emissions from drained organic
15 soils by using the revised factor in the 2013 Supplement to the 2006IPCC Guidelines for National Greenhouse Gas
16 Inventories: Wetlands (IPCC 2013).
17 These improvements are expected to be completed for the next Inventory report (i.e., 2019 submission to the
18 UNFCCC, 1990 through 2017 Inventory). However, the time line may be extended if there are insufficient resources
19 to fund all or part of these planned improvements.
20 5.5 Liming (CRF Source Category 3G)
21 Crushed limestone (CaCCh) and dolomite (CaMg(C03)2) are added to soils by land managers to increase soil pH
22 (i.e., to reduce acidification). Carbon dioxide emissions occur as these compounds react with hydrogen ions in soils.
23 The rate of degradation of applied limestone and dolomite depends on the soil conditions, soil type, climate regime,
24 and whether limestone or dolomite is applied. Emissions from liming of soils have fluctuated over the past 25 years
25 in the United States, ranging from 3.6 MMT CO2 Eq. to 6.0 MMT CO2 Eq. In 2016, liming of soils in the United
26 States resulted in emissions of 3.9 MMT CO2 Eq. (1.1 MMT C), representing a 17 percent decrease in emissions
27 since 1990 (see Table 5-21 and Table 5-22). The trend is driven by variation in the amount of limestone and
28 dolomite applied to soils over the time period.
29 Table 5-21: Emissions from Liming (MMT CO2 Eq.)
Source
1990
2005
2012
2013
2014
2015
2016
Limestone
4.1
3.9
4.5
3.6
3.3
3.5
3.6
Dolomite
0.6
0.4
1.5
0.3
0.3
0.3
0.3
Total
4.7
4.3
6.0
3.9
3.6
3.8
3.9
Note: Totals may not sum due to independent rounding.
ible 5-22:
Emissions from Liming (MMT C)
Source
1990
2005
2012
2013
2014
2015
2016
Limestone
1.1
1.1
1.2
1.0
0.9
1.0
1.0
Dolomite
0.2
0.1
0.4
0.1
0.1
0.1
0.1
Total
1.3
1.2
1.6
1.1
1.0
1.0
1.1
Note: Totals may not sum due to independent rounding.
31 Methodology
32 Carbon dioxide emissions from application of limestone and dolomite to soils were estimated using a Tier 2
33 methodology consistent with IPCC (2006). The annual amounts of limestone and dolomite applied (see Table 5-23)
5-42 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
were multiplied by CO2 emission factors from West and McBride (2005). These 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).
The annual application rates of limestone and dolomite were derived from estimates and industry statistics provided
in the Minerals Yearbook and Mineral Industry Surveys (Tepordei 1993 through 2006; Willett 2007a, 2007b, 2009,
2010, 2011a, 2011b, 2013a, 2014, 2015, 2016 and 2017; USGS 2008 through 2017). 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 5-6: Comparison of the Tier 2 U.S. Inventory Approach and IPCC (2006) Default Approac
Ji
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) emission default 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
CaCCb 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
2016 U.S. emission estimate from liming of soils is 3.9 MMT CO2 Eq. using the U.S.-specific factors. In contrast,
emissions would be estimated at 7.9 MMT CO2 Eq. using the IPCC (2006) default emission factors.
Data on "specified" limestone and dolomite amounts were used directly in the emission calculation because the end
use is provided by the manufacturers and can be used to directly determine the amount applied to soils. However, it
is not possible to determine directly how much of the limestone and dolomite is applied to soils for manufacturer
surveys in the "unspecified" and "estimated" categories. For these categories, the amounts of crushed limestone and
dolomite applied to soils were determined by multiplying the percentage of total "specified" limestone and dolomite
production that is applied to soils, by the total amounts of "unspecified" and "estimated" limestone and dolomite
production. In other words, the proportion of total "unspecified" and "estimated" crushed limestone and dolomite
that was applied to soils is proportional to the amount of total "specified" crushed limestone and dolomite that was
applied to soils.
In addition, data were not available for 1990, 1992 and 2016 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 2016 data, 2015 fractions were applied to a 2016
estimate of total crushed stone presented in the USGS Mineral Industry Surveys: Crushed Stone and Sand and
Gravel in the First Quarter of 2017 (USGS 2017).
The primary source for limestone and dolomite activity data is the Minerals Yearbook, published by the Bureau of
Mines through 1994 and by the USGS from 1995 to the present. In 1994, the "Crushed Stone" chapter in the
Minerals Yearbook began rounding (to the nearest thousand metric tons) quantities for total crushed stone produced
or used. It then reported revised (rounded) quantities for each of the years from 1990 to 1993. In order to minimize
the inconsistencies in the activity data, these revised production numbers have been used in all of the subsequent
calculations.
Agriculture 5-43
-------
1
Table 5-23: Applied Minerals (MMT)
Mineral
1990
2005
2012
2013
2014
2015
2016
Limestone
19.0
18.1
20.8
16.4
15.3
16.2
16.5
Dolomite
2.4
1.9
6.3
1.5
1.3
1.2
1.2
2 Uncertainty and Time-Series Consistency
3 Uncertainty regarding the amount of limestone and dolomite applied to soils was estimated at ±15 percent with
4 normal densities (Tepordei 2003; Willett 2013b). Analysis of the uncertainty associated with the emission factors
5 included the fraction of lime dissolved by nitric acid versus the fraction that reacts with carbonic acid, and the
6 portion of bicarbonate that leaches through the soil and is transported to the ocean. Uncertainty regarding the time
7 associated with leaching and transport was not addressed in this analysis, but is assumed to be a relatively small
8 contributor to the overall uncertainty (West 2005). The probability distribution functions for the fraction of lime
9 dissolved by nitric acid and the portion of bicarbonate that leaches through the soil were represented as triangular
10 distributions between ranges of zero and 100 percent of the estimates. The uncertainty surrounding these two
11 components largely drives the overall uncertainty.
12 A Monte Carlo (Approach 2) uncertainty analysis was applied to estimate the uncertainty in CO2 emissions from
13 liming. The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 5-24. Carbon
14 dioxide emissions from carbonate lime application to soils in 2016 were estimated to be between -0.4 and 7.3 MMT
15 CO2 Eq. at the 95 percent confidence level. This confidence interval represents a range of 111 percent below to 88
16 percent above the 2016 emission estimate of 3.9 MMT CO2 Eq.
17 Table 5-24: Approach 2 Quantitative Uncertainty Estimates for CO2 Emissions from Liming
18 (MMT CO2 Eq. and Percent)
Source
Gas
2016 Emission Estimate
(MMT CO2 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT CO2 Eq.) (%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Liming
CO2
3.9
(0.4) 7.3
-111% +88%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
19 Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
20 through 2016. Details on the emission trends through time are described in more detail in the Methodology section,
21 above.
22 ration
23 A source-specific QA/QC plan for liming has been developed and implemented, and the quality control effort
24 focused on the Tier 1 procedures for this Inventory. No errors were found.
25 Recalculations Discussion
26 Adjustments were made in the current Inventory to improve the results. First, limestone and dolomite application
27 data for 2015 were updated with the recently published data from USGS (2017), rather than being approximated by
28 a ratio method. With this revision in the activity data, the emissions decreased by 0.9 percent in 2015 relative to the
29 previous Inventory.
5-44 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
i 5.6 Urea Fertilization (CRF Source Category 3H)
2 The use of urea (CO(NH2)2) as a fertilizer leads to CO2 emissions through the release of CO2 that was fixed during
3 the industrial production process. In the presence of water and urease enzymes, urea is converted into ammonium
4 (NH4+), hydroxyl ion (OH), and bicarbonate (HCO3 ). The bicarbonate then evolves into CO2 and water. Emissions
5 from urea fertilization in the United States totaled 5.1 MMT CO2 Eq. (1.4 MMT C) in 2016 (Table 5-25 and Table
6 5-26). Due to an increase in application of urea fertilizers between 1990 and 2016, CO2 emissions have increased by
7 111 percent from this management activity.
8 Table 5-25: CO2 Emissions from Urea Fertilization (MMT CO2 Eq.)
Source 1990 2005 2012 2013 2014 2015 2016
Urea Fertilization 2.4 3.5 4.3 4.4 4.5 4.9 5.1
9 Table 5-26: CO2 Emissions from Urea Fertilization (MMT C)
Source 1990 2005 2012 2013 2014 2015 2016
Urea Fertilization 0.7 1.0 1.2 1.2 1.2 1.3 1.4
10 Methodology
11 Carbon dioxide emissions from the application of urea to agricultural soils were estimated using the IPCC (2006)
12 Tier 1 methodology. The method assumes that all CO2 fixed during the industrial production process of urea are
13 released after application. The annual amounts of urea applied to croplands (see Table 5-27) were derived from the
14 state-level fertilizer sales data provided in Commercial Fertilizer reports (TVA 1991, 1992, 1993, 1994; AAPFCO
15 1995 through 2017).19 These amounts were multiplied by the default IPCC (2006) emission factor (0.20 metric tons
16 of C per metric ton of urea), which is equal to the C content of urea on an atomic weight basis. Because fertilizer
17 sales data are reported in fertilizer years (July previous year through June current year), a calculation was performed
18 to convert the data to calendar years (January through December). According to monthly fertilizer use data (TVA
19 1992b), 35 percent of total fertilizer used in any fertilizer year is applied between July and December of the previous
20 calendar year, and 65 percent is applied between January and June of the current calendar year. For example, in the
21 2000 fertilizer year, 35 percent of the fertilizer was applied in July through December 1999, and 65 percent was
22 applied in January through June 2000.
23 Fertilizer sales data for the 2015 and 2016 fertilizer years (i.e., July 2014 through June 2015 and July 2015 through
24 June 2016) were not available fortius Inventory. Therefore, urea application in the 2015 and 2016 fertilizer years
25 were estimated using a linear, least squares trend of consumption over the data from the previous five years (2010
26 through 2014) at the state scale. A trend of five years was chosen as opposed to a longer trend as it best captures the
27 current inter-state and inter-annual variability in consumption. State-level estimates of CO2 emissions from the
28 application of urea to agricultural soils were summed to estimate total emissions for the entire United States. The
29 fertilizer year data is then converted into calendar year data using the method described above.
19 The amount of urea consumed for non-agricultural purposes in the United States is reported in the Industrial Processes and
Product Use chapter, Section 4.6 Urea Consumption for Non-Agricultural Purposes.
Agriculture 5-45
-------
1
Table 5-27: Applied Urea (MMT)
1990
2005
2012
2013
2014
2015
2016
Urea Fertilizer8
3.3
4.8
5.8
6.1
6.2
6.7
7.0
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.
2 Uncertainty and Time-Series Consistency
3 Uncertainty estimates are presented in Table 5-28 for urea fertilization. An Approach 2 Monte Carlo analysis was
4 completed. The largest source of uncertainty was the default emission factor, which assumes that 100 percent of the
5 C in CO(NH2)2 applied to soils is ultimately emitted into the environment as CO2. This factor does not incorporate
6 the possibility that some of the C may be retained in the soil, and therefore the uncertainty range was set from 0
7 percent emissions to the maximum emission value of 100 percent using a triangular distribution. In addition, urea
8 consumption data also have uncertainty that is propagated through the emission calculation using a Monte Carlo
9 simulation approach as described by the IPCC (2006). Carbon dioxide emissions from urea fertilization of
10 agricultural soils in 2016 were estimated to be between 2.9 and 5.3 MMT CO2 Eq. at the 95 percent confidence
11 level. This indicates a range of 43 percent below to 3 percent above the 2016 emission estimate of 5.1 MMT CO2
12 Eq.
13 Table 5-28: Quantitative Uncertainty Estimates for CO2 Emissions from Urea Fertilization
14 (MMT CO2 Eq. and Percent)
15
Source
Gas
2016 Emission Estimate
(MMT CO2 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT CO2 Eq.) (%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Urea Fertilization
CO2
5.1
2.9 5.3
-43% 3%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
16 There are additional uncertainties that are not quantified in this analysis. Urea for non-fertilizer use, such as aircraft
17 deicing, may be included in consumption totals, but the amount is likely very small. For example, research on
18 aircraft deicing practices based on a 1992 survey found a known annual usage of approximately 2,000 tons of urea
19 for deicing; this would constitute 0.06 percent of the 1992 consumption of urea (EPA 2000). Similarly, surveys
20 conducted from 2002 to 2005 indicate that total urea use for deicing at U.S. airports is estimated to be 3,740 metric
21 tons per year, or less than 0.07 percent of the fertilizer total for 2007 (Itle 2009). In addition, there is uncertainty
22 surrounding the underlying assumptions behind the calculation that converts fertilizer years to calendar years. These
23 uncertainties are negligible over multiple years because an over- or under-estimated value in one calendar year is
24 addressed with corresponding increase or decrease in the value for the subsequent year.
25 Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
26 through 2016. Details on the emission trends through time are described in more detail in the Methodology section,
27 above.
28 ration
29 A source-specific QA/QC plan for Urea Fertilization has been developed and implemented, and no errors were
30 found.
5-46 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Recalculations Discussion
Recalculations resulted from updated urea application estimates in a new AAPFCO report (2017). Specifically, the
2013 activity data (i.e., amount of urea applied) for the states of California, Maryland, and Mississippi were updated.
New activity data for 2014 were applied to all states; 2015 and 2016 estimates were derived using the new data for
2013 and 2014. This resulted in an emissions decrease for the United States of 1.3 percent in 2013, 5.1 percent in
2014, and 2.9 percent in 2015.
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 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, N20, 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 com
cotton lentils, rice, soybeans, sugarcane, and wheat (McCarty 2009). Rice, sugarcane, and wheat residues account
for approximately 70 percent of all crop residue burning and emissions (McCarty 2011). In 2016, CH4 and N20
emissions from field burning of agricultural residues were 0.3 MMT CO2 Eq. (11 kt) and 0.1 MMT CO2 Eq. (0.3 kt),
respectively (see Table 5-29 and Table 5-30). Annual emissions of CH4 and N20 have increased from 1990 to 2016
by 20 percent and 21 percent, respectively. The increase in emissions over time is due to larger amounts of residue
production with higher yielding crop varieties and fuel loads.
Table 5-29: ChU and N2O Emissions from Field Burning of Agricultural Residues (MMT CO2
Eq.)
Gas/Crop Type
1990
2005
2012
2013
2014
2015
2016
CH4
0.2
0.2
0.3
0.3
0.3
0.3
0.3
Wheat
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
Sugarcane
Com
+
+
+
+
+
+
+
+
+
+
+
+
+
+
Cotton
+
+
+
+
+
+
+
Soybeans
Lentil
+
+
+
+
+
+
+
+
+
+
+
+
+
+
N2O
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Wheat
+
+
+
+
+
+
+
Rice
+
+
+
+
+
+
+
Sugarcane
Com
+
+
+
+
+
+
+
+
+
+
+
+
+
+
Cotton
+
+
+
+
+
+
+
Soybeans
Lentil
+
+
+
+
+
+
+
+
+
+
+
+
+
+
Total
0.3
0.3
0.4
0.4
0.4
0.4
0.4
+ Does not exceed 0.05 MMT CO2 Eq.
Note: Totals may not sum due to independent rounding.
Agriculture 5-47
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Table 5-30: ChU, N2O, CO, and NOx Emissions from Field Burning of Agricultural Residues
(kt)
Gas/Crop Type
1990
2005
2012
2013
2014
2015
2016
CH4
•)
8
11
11
11
11
11
Wheat
>
4
5
5
5
5
5
Rice
'y
2
2
2
2
2
2
Sugarcane
1
1
1
2
2
1
Corn
1
1
1
1
2
2
2
Soybeans
1
1
1
1
1
1
1
Lentil
+
+
+
+
+
+
Cotton
+
+
+
+
+
+
N2O
+
+
+
+
+
+
+
Wheat
+
+
+
+
+
+
Rice
+
+
+
+
+
+
Sugarcane
+
+
+
+
+
+
Corn
+
+
+
+
+
+
Cotton
+
+
+
+
+
+
Soybeans
+
+
+
+
+
+
Lentil
+
+
+
+
+
+
CO
191
178
232
239
240
239
230
NOx
(,
6
7
7
8
8
7
+ Does not exceed 0.5 kt.
Methodology
A U.S.-specific Tier 2 method was used to estimate greenhouse gas emissions from field burning of agricultural
residues from 1990 to 20 1 520 (for more details comparing the U.S.-specific approach to the IPCC (2006) default
approach, see Box 5-7). In order to estimate the amounts of C and N released during burning, the following equation
was used:
C or N released = £ for all crop types and states
where,
AB
CAH x CP x RCR x DMF x BE x CE x (FC or FN)
Area Burned (AB) = Total area of crop burned, by state
Crop Area Harvested (CAH) = Total area of crop harvested, by state
Crop Production (CP) = Annual production of crop in kt, by state
Residue: Crop Ratio (RCR) = Amount of residue produced per unit of crop production
Dry Matter Fraction (DMF) = Amount of dry matter per unit of bio mass for a crop
Fraction of C or N (FC or FN) = Amount of C or N per unit of dry matter for a crop
Burning Efficiency (BE) = The proportion of prefire fuel bio mass consumed21
Combustion Efficiency (CE) = The proportion of C or N released with respect to the total amount of C or N
available in the burned material, respectively
Crop Production and Crop Area Harvested were available by state and year from USD A (2016) for all crops (except
rice in Florida and Oklahoma, as detailed below). The amount of C or N released was used in the following equation
to determine the CH4, CO, N20, and NOx emissions from the Field Burning of Agricultural Residues:
CH4 and CO, or N2O and NOx Emissions from Field Burning of Agricultural Residues =
C or N Released x ER x CF
20 The emission estimates for 2016 are estimated using an extrapolation method described later in this section.
21 In IPCC/UNEP/OECD/IEA (1997), the equation for C orN released contains the variable 'fraction oxidized in burning'. This
variable is equivalent to (burning efficiency x combustion efficiency).
5-48 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
where.
Emissions Ratio (ER) = g CH4-C or CO-C/g C released, or g N20-N or NOx-N/g N released
Conversion Factor (CF) = conversion, by molecular weight ratio, of CH4-C to C (16/12), or CO-C to C
(28/12), or N20-N to N (44/28), or NOx-N to N (30/14)
Box 5-7: Comparison of Tier 2 U.S. Inventory Approach and IPCC (2006) Default Approach
Emissions from Field Burning of Agricultural Residues were calculated using a Tier 2 methodology that is based on
the method developed by the IPCC/UNEP/OECD/IEA (1997) and incorporates crop- and country-specific emission
factors and variables. 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; and (3) the IPCC (2006)
default factors are provided only for four crops (corn, rice, sugarcane, and wheat) while this Inventory includes
emissions from seven crops (com cotton, lentils, rice, soybeans, sugarcane, and wheat).
A comparison of the methods and factors used in: (1) the current Inventory and (2) the default IPCC (2006)
approach was undertaken for the time series from 1990 through 2015 to determine the difference in overall 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:
Emissions (kt) = AB x (MbX Cf) x Gef x 10~6
where.
Area Burned (AB) = Total area of crop burned (ha)
Mass Burned (MB x Cf) = IPCC (2006) default fuel biomass consumption (metric tons dry matter burnt
ha ') and US-Specific Values using NASS Statistics (USDA 2016)
Emission Factor (Gef) = IPCC (2006) emission factor (g kg1 dry matter burnt)
The IPCC (2006) Tier 1 method approach that utilizes default mass of fuel values resulted in 1 percent higher
emissions of CH4 and 14 percent higher emissions of N20 compared to this Inventory. If U.S.-specific data are used
to derive the Mass of Fuel (Mb) from USDA-NASS statistics (USDA 2016), i.e.. Tier 2 method, then the IPCC
(2006) method resulted in 28 percent higher emissions of CH4 and 44 percent higher emissions of N20 compared to
the Tier 1 method. This larger difference is attributable to lower combustion efficiency values in
IPCC/UNEP/OECD/IEA (1997). In particular, sugarcane lias a much lower combustion efficiency value in the
earlier guidelines. A lower value is justified because sugarcane is burned prior to harvesting and lias a higher
moisture content that reduces the combustion efficiency, unlike most other crops (IPCC/UNEP/OECD/IEA 1997).
IPCC (2006) does not address the unique burning regime of sugarcane. Overall, 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 compared to IPCC (2006) approach.
Crop yield data (except rice in Florida) were based on USDA's OuickStats (USDA 2016), and crop area data were
based on the 2012 NRI (USDA-NRCS 2015). In order to estimate total crop production the crop yield data from
USDA Quick Stats crop yields was multiplied by the NRI crop areas. Rice yield data for Florida was estimated
separately because yield data were not collected by USDA. Total rice production for Florida was determined using
NRI crop areas and total yields were based on average primary and ratoon rice yields from Schueneman and Deren
(2002). Relative proportions of ratoon crops were derived from information in several publications (Schueneman
1999, 2000, 2001; Deren2002; Kirstein2003, 2004; Cantens 2004, 2005; Gonzalez 2007 through2014). The
production data for the crop types whose residues are burned are presented in Table 5-31. Crop weight by bushel
was obtained from Murphy (1993).
Agriculture 5-49
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
The fraction of crop area burned was calculated using data on area burned by crop type and state22 from McCarty
(2010) for corn, cotton, lentils, rice, soybeans, sugarcane, and wheat.23 McCarty (2010) used remote sensing data
from MODIS to estimate area burned by crop. State-level area burned data were divided by state-level crop area
harvested data to estimate the percent of crop area burned by crop type for each state. The average percentage of
crop area burned at the national scale is shown in Table 5-32. Data on fraction of crop area burned were only
available from McCarty (2010) for the years 2003 through 2007. For other years in the time series, the percent area
burned was set equal to the average over the five-year period from 2003 to 2007. Table 5-32 shows the resulting
percentage of crop residue burned at the national scale by crop type. State-level estimates are also available upon
request.
All residuexrop product mass ratios except sugarcane and cotton were obtained from Strehler and Stiitzle (1987).
The ratio for sugarcane is from Kinoshita (1988) and the ratio for cotton is from Huang et al. (2007). The residue:
crop ratio for lentils was assumed to be equal to the average of the values for peas and beans. Residue dry matter
fractions for all crops except soybeans, lentils, and cotton were obtained from Turn et al. (1997). Soybean and lentil
dry matter fractions were obtained from Strehler and Stiitzle (1987); the value for lentil residue was assumed to
equal the value for bean residue. The cotton dry matter fraction was taken from Huang et al. (2007). The residue C
contents and N contents for all crops except soybeans and cotton are from Turn et al. (1997). The residue C content
for soybeans is the IPCC default (IPCC/UNEP/OECD/IEA 1997), and the N content of soybeans is from Barnard
and Kristoferson (1985). The C and N contents of lentils were assumed to equal those of soybeans. The C and N
contents of cotton are from Lachnicht et al. (2004). The burning efficiency was assumed to be 93 percent, and the
combustion efficiency was assumed to be 88 percent, for all crop types, except sugarcane (EPA 1994). For
sugarcane, the burning efficiency was assumed to be 81 percent (Kinoshita 1988) and the combustion efficiency was
assumed to be 68 percent (Turn et al. 1997). See Table 5-33 for a summary of the crop-specific conversion factors.
Emission ratios and mole ratio conversion factors for all gases were based on the Revised 1996 IPCC Guidelines
(IPCC/UNEP/OECD/IEA 1997) (see Table 5-34).
Table 5-31: Agricultural Crop Production (kt of Product)
Crop
1990
2005
2012
2013
2014
2015
2016
Corn3
249,806
323.724
311,751
398,817
429,405
422,436
NE
Cotton
4,633
6.560
5,967
5,647
5,934
5,575
NE
Lentils
+
119
121
147
134
117
NE
Rice
9,428
12.253
10,080
10,381
10,347
10,202
NE
Soybeans
56,626
88.036
85,523
93,928
102,065
102,772
NE
Sugarcane
18,765
18.211
16,555
16,129
17,136
18,336
NE
Wheat
79,951
69.190 „
71,234
69,287
64,650
66,672
NE
+ Does not exceed 0.5 kt.
NE (Not Estimated). 2016 crop production values were not compiled for the current Inventory.
a Corn for grain (i.e., excludes corn for silage).
Table 5-32: U.S. Average Percent Crop Area Burned by Crop (Percent)
State
1990
2005
2012
2013
2014
2015
2016
Corn
+
+
+
+
+
+
NE
Cotton
1%
1%
1%
1%
1%
1%
NE
Lentils
+
1%
+
+
+
+
NE
Rice
8%
5%
7%
7%
7%
7%
NE
Soybeans
+
+
+
+
+
+
NE
Sugarcane
13%
25%
53%
52%
53%
54%
NE
Wheat
2%
2%
3%
3%
3%
2%
NE
+ Does not exceed 0.5 percent.
NE (Not Estimated). 2016 crop area burned was not compiled for the current Inventory.
22 Alaska and Hawaii were excluded.
23 McCarty (2009) also examined emissions from burning of Kentucky bluegrass and a general "other crops/fallow" category,
but USDA crop area and production data were insufficient to estimate emissions from these crops using the methodology
employed in the Inventory. McCarty (2009) estimates that approximately 18 percent of crop residue emissions result from
burning of the Kentucky bluegrass and "other crops" categories.
5-50 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Table 5-33: Key Assumptions for Estimating Emissions from Field Burning of Agricultural
2 Residues
Burning
Combustion
Residue: Crop
Dry Matter
Efficiency
Efficiency
Crop
Ratio
Fraction
C Fraction
N Fraction
(Fraction)
(Fraction)
Corn
1.0
0.91
0.448
0.006
0.93
0.88
Cotton
1.6
0.90
0.445
0.012
0.93
0.88
Lentils
2.0
0.85
0.450
0.023
0.93
0.88
Rice
1.4
0.91
0.381
0.007
0.93
0.88
Soybeans
2.1
0.87
0.450
0.023
0.93
0.88
Sugarcane
0.2
0.62
0.424
0.004
0.81
0.68
Wheat
1.3
0.93
0.443
0.006
0.93
0.88
3 Table 5-34: Greenhouse Gas Emission Ratios and Conversion Factors
Gas
Emission Ratio
Conversion Factor
CH4:C
0.005a
16/12
CO:C
0.060a
28/12
N20:N
0.007b
44/28
NOx:N
0.121b
30/14
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).
4 For this Inventory, new activity data were not compiled for Field Burning of Agricultural Residues because the
5 Inventory is only fully re-compiled every two years for many categories in the AFOLU sector as part of the biennial
6 update reporting process. Therefore, a linear extrapolation of the trend in the time series was applied to estimate the
7 emissions for 2016. Specifically, a linear regression model with autoregressive moving-average (ARMA) errors was
8 used to estimate the trend in emissions over time from 1990 through 2015, and in turn, the trend was used to
9 approximate the 2016 emissions (Brockwell and Davis 2016). The Tier 2 method described previously will be
10 applied to recalculate the 2016 emissions in the next Inventory (i.e., 1990 through 2017 Inventory report).
11 Uncertainty and Time-Series Consistency
12 Emissions are estimated using a linear regression model with autoregressive moving-average (ARMA) errors for
13 2016. The linear regression ARMA model produced estimates of the upper and lower bounds of the emission
14 estimate (Table 5-35), and the results are summarized in Table 5-35. Methane emissions from field burning of
15 agricultural residues in 2016 were estimated to be between 0.23 and 0.31 MMT CO2 Eq. at a 95 percent confidence
16 level. This indicates a range of 14 percent below and 14 percent above the 2016 emission estimate of 0.3 MMT CO2
17 Eq. Nitrous oxide emissions were estimated to be between 0.08 and 0.11 MMT CO2 Eq., or approximately 14
18 percent below and 14 percent above the 2016 emission estimate of 0.1 MMT CO2 Eq.
19 Table 5-35: Approach 2 Quantitative Uncertainty Estimates for ChU and N2O Emissions from
20 Field Burning of Agricultural Residues (MMT CO2 Eq. and Percent)
2016 Emission
Source
Gas
Estimate
Uncertainty Range Relative to Emission Estimate
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Field Burning of Agricultural
Residues
CH4
0.3
0.23
0.31
-14%
14%
Field Burning of Agricultural
Residues
N2O
0.1
0.08
0.11
-14%
14%
a Range of emission estimates predicted by ARMA linear regression time-series model for a 95 percent confidence interval.
Agriculture 5-51
-------
1 Due to data limitations, there are additional uncertainties in agricultural residue burning, particularly the omission of
2 burning associated with Kentucky bluegrass and "other crop" residues.
3 QA/QC and Verification
4 A source-specific QA/QC plan for field burning of agricultural residues was implemented with Tier 1 analyses, and
5 no errors were found in the current Inventory.
6 Recalculations Discussion
7 No recalculations were conducted for the current Inventory.
8 Planned Improvements
9 A new method is in development that will directly link agricultural residue burning with the Tier 3 methods that are
10 used in several other source categories, including Agricultural Soil Management, Cropland Remaining Cropland,
11 and Land Converted to Cropland chapters of the Inventory. The method is based on the DAYCENT model, and
12 burning events will be simulated directly within the process-based model framework using information derived from
13 remote sensing fire products. This improvement will lead to greater consistency in the methods for these sources,
14 and better ensure mass balance of C and N in the Inventory analysis.
15
5-52 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Land Usi
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
certain land-use types termed: 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)
stocks (i.e., aboveground biomass, belowground biomass, dead wood, litter, and C stock changes from 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.
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-C02 emissions from grassland fires.
Fluxes from Wetlands Remaining Wetlands include changes in C stocks and methane (CH4) and nitrous oxide (N2O)
emissions from managed peatlands, as well as soil C stock changes in coastal wetlands, CH4 emissions from
vegetated coastal wetlands, and N20 emissions from aquaculture in coastal wetlands. Estimates for Land Converted
to Wetlands include soil C stock changes and CH4 emissions from land converted to vegetated coastal wetlands.
Fluxes from Settlements Remaining Settlements include changes in C stocks, N2O emissions from soils, and CO2
fluxes from urban 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.
The land use, land-use change, and forestry (LULUCF) sector in 2016 resulted in a net increase in C stocks (i.e., net
CO2 removals) of 754.9 MMT CO2 Eq. (205.9 MMT C).2 This represents an offset of approximately 11.5 percent of
1 The term "flux" is used to describe the net emissions of greenhouse gases accounting for both the emissions of CO2 to and the
removals of CO2 from the atmosphere. Removal of CO2 from the atmosphere is also referred to as "carbon sequestration."
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.
Land Use, Land-Use Change, and Forestry 6-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
total (i.e., gross) greenhouse gas emissions in 2016. Emissions of CH4 and N20 from LULUCF activities in 2016 are
38.1 MMT CO2 Eq. and represent 0.6 percent of total greenhouse gas emissions.3
Total C sequestration in the LULUCF sector decreased by approximately 9.1 percent between 1990 and 2016. This
decrease was primarily due to a decrease in the rate of net C accumulation in forests and Cropland Remaining
Cropland, as well as an increase in emissions from Land Converted to Settlements,4 Net C accumulation in
Settlements Remaining Settlements increased from 1990 to 2016, while net C accumulation in Forest Land
Remaining Forest Land, Land Converted to Forest Land, Cropland Remaining Cropland, and Grassland Remaining
Grassland slowed over this period. Net C accumulation remained steady from 1990 to 2016 in Wetlands Remaining
Wetlands and Land Converted to Wetlands. Emissions from Land Converted to Cropland decreased during this
period, while emissions from Land Converted to Grassland increased. The C stock change from LULUCF is
summarized in Table 6-1.
Table 6-1: Net CO2 Flux from Land Use, Land-Use Change, and Forestry (MMT CO2 Eq.)
Land-Use Category
1990
2005
2012
2013
2014
2015
2016
Forest Land Remaining Forest Land
(697.7)
(664.6)
(666.9)
(670.9)
(669.3)
(666.2)
(670.5)
Changes in Forest Carbon Stocks8
(697.7)
(664.6)
(666.9)
(670.9)
(669.3)
(666.2)
(670.5)
Land Converted to Forest Land
(92.0)
(81.6)
(74.9)
(74.9)
(75.0)
(75.0)
(75.0)
Changes in Forest Carbon Stocksb
(92.0)
(81.6)
(74.9)
(74.9)
(75.0)
(75.0)
(75.0)
Cropland Remaining Cropland
(40.9)
(26.5)
(21.4)
(11.4)
(12.0)
(6.3)
(9.9)
Changes in Mineral and Organic Soil
Carbon Stocks
(40.9)
(26.5)
(21.4)
(11.4)
(12.0)
(6.3)
(9.9)
Land Converted to Cropland
43.3
25.9
22.7
23.3
23.2
23.2
23.8
Changes in all Ecosystem Carbon
Stocksc
43.3
25.9
22.7
23.3
23.2
23.2
23.8
Grassland Remaining Grassland
(4.2)
5.5
(20.8)
(3.7)
(7.5)
9.6
(1.6)
Changes in Mineral and Organic Soil
Carbon Stocks
(4.2)
5.5
(20.8)
(3.7)
(7.5)
9.6
(1.6)
Land Converted to Grassland
17.9
19.2
20.4
21.9
21.5
23.3
22.0
Changes in all Ecosystem Carbon
Stocks0
17.9
19.2
20.4
21.9
21.5
23.3
22.0
Wetlands Remaining Wetlands
(7.6)
(8.9)
(7.7)
(7.8)
(7.8)
(7.8)
(7.9)
Changes in Organic Soil Carbon Stocks
in Peatlands
1.1
11
0.8
0.8
0.8
0.8
0.7
Changes in Mineral and Organic Soil
Carbon Stocks in Coastal Wetlands
(8.6)
(10.0)
(8.6)
(8.6)
(8.6)
(8.6)
(8.6)
Land Converted to Wetlands
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Changes in Mineral and Organic Soil
Carbon Stocksd
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Settlements Remaining Settlements
(86.2)
(91.4)
(99.2)
(99.8)
(101.2)
(102.2)
(103.7)
Changes in Organic Soil Carbon Stocks
0.1
0.5
1.3
1.3
1.3
1.3
1.3
Changes in Urban Tree Carbon Stocks
(60.4)
(80.5)
(88.4)
(89.5)
(90.6)
(91.7)
(92.9)
Changes in Yard Trimmings and Food
Scrap Carbon Stocks in Landfills
(26.0)
(11.4)
(12.2)
(11.6)
(11.9)
(11.8)
(12.2)
Land Converted to Settlements
37.2
68.4
68.3
68.3
68.2
68.1
68.0
Changes in all Ecosystem Carbon
Stocksc
37.2
68.4
68.3
68.3
68.2
68.1
68.0
LULUCF Carbon Stock Change
(830.2)
(754.2)
(779.5)
(755.0)
(760.0)
(733.4)
(754.9)
3 LULUCF emissions include the CH4 and N2O 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 N2O emissions from Forest Soils and Settlement Soils.
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.
6-2 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
+ Absolute value does not exceed 0.05 MMT CO2 Eq.
a Includes the net changes to carbon stocks stored in all forest ecosystem pools (including drained and undrained organic
soils) and harvested wood products.
b Includes the net changes to carbon stocks stored in all forest ecosystem pools (excludes drained organic soils which are
included in the flux from Forest Land Remaining Forest Land because it is not possible to separate the activity data at
this time).
c Includes changes in mineral and organic soil carbon stocks for all land use conversions to cropland, grassland, and
settlements, respectively. Also includes aboveground/belowground biomass, dead wood, and litter carbon stock changes
for conversion of forest land to cropland, grassland, and settlements, respectively.
d Includes carbon stock changes for land converted to vegetated coastal wetlands.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Emissions from LULUCF activities are shown in Table 6-2. Forest fires were the largest source of CH4 emissions
fromLULUCF in2016, totaling 18.5 MMT CO2 Eq. (740 kt of CH4). Coastal Wetlands Remaining Coastal
Wetlands resulted in CH4 emissions of 3.6 MMT CO2 Eq. (143 kt of CH4). Grassland fires resulted in CH4 emissions
of 0.3 MMT CO2 Eq. (11 kt of CH4). Peatlands Remaining Peat lands, Land Converted to Wetlands, and Drained
Organic Soils resulted in CH4 emissions of less than 0.05 MMT CO2 Eq. each.
Forest fires were also the largest source of N20 emissions from LULUCF in 2016, totaling 12.2 MMT CO2 Eq. (41
kt of N20). Nitrous oxide emissions from fertilizer application to settlement soils in 2016 totaled to 2.5 MMT CO2
Eq. (8 kt of N20). This represents an increase of 74.6 percent since 1990. Additionally, the application of synthetic
fertilizers to forest soils in 2016 resulted in N20 emissions of 0.5 MMT CO2 Eq. (2 kt of N20). 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 N20 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
0.1 MMT CO2 Eq. each (less than 0.5 kt of N2O), and Peatlands Remaining Peatlands resulted in N2O emissions of
less than 0.05 MMT CO2 Eq.
Emissions and removals from LULUCF are summarized in Figure 6-1 and Table 6-3 by land-use and category, and
Table 6-4 and Table 6-5 by gas in MMT CO2 Eq. and kt, respectively.
Table 6-2: Emissions from Land Use, Land-Use Change, and Forestry by Gas (MMT CO2 Eq.)
Gas/Land-Use Sub-Category
1990
2005
2012
2013
2014
2015
2016
CH4
6.7
13.3
15.0
10.9
11.2
22.4
22.4
Forest Land Remaining Forest Land:
Forest Fires
3.2
9.4
10.8
7.2
7.2
18.5
18.5
Wetlands Remaining Wetlands: Coastal
Wetlands Remaining Coastal Wetlands
3.4
3.5
3.5
3.6
3.6
3.6
3.6
Grassland Remaining Grassland:
Grassland Fires
0.1
0.3
0.6
0.2
0.4
0.3
0.3
Forest Land Remaining Forest Land:
Drained Organic Soils
+
+
+
+
+
+
+
Land Converted to Wetlands: Land
Converted to Coastal Wetlands
+
+
+
+
+
+
+
Wetlands Remaining Wetlands:
Peatlands Remaining Peatlands
+
+
+
+
+
+
+
N2O
3.9
9.7
11.1
8.3
8.4
15.8
15.7
Forest Land Remaining Forest Land:
Forest Fires
2.1
6.2
7.1
4.8
4.7
12.2
12.2
Settlements Remaining Settlements:
Settlement Soils3
1.4
2.5
2.7
2.6
2.6
2.5
2.5
Forest Land Remaining Forest Land:
Forest Soilsb
0.1
0.5
0.5
0.5
0.5
0.5
0.5
Grassland Remaining Grassland:
Grassland Fires
0.1
0.3
0.6
0.2
0.4
0.3
0.3
Wetlands Remaining Wetlands: Coastal
Wetlands Remaining Coastal Wetlands
0.1
0.2
0.1
0.1
0.1
0.1
0.1
Forest Land Remaining Forest Land:
Drained Organic Soils
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Land Use, Land-Use Change, and Forestry 6-3
-------
Wetlands Remaining Wetlands:
Peatlands Remaining Peatlands
+
+
+
+
+
+
+
LULUCF Emissions
10.6
23.0
26.1
19.2
19.6
38.2
38.1
+ Does not exceed 0.05 MMT CO2 Eq.
a Estimates include emissions from N fertilizer additions on both Settlements Remaining Settlements and Land Converted to
Settlements.
b Estimates include emissions from N fertilizer additions on both Forest Land Remaining Forest Land and Land Converted to
Forest Land.
Note: Totals may not sum due to independent rounding.
Figure 6-1: 2016 LULUCF Chapter Greenhouse Gas Sources and Sinks (MMT CO2 Eq.)
Forest Land Remaining Forest Land (670.5) |
Settlements Remaining Settlements
Land Converted to Forest Land
Cropland Remaining Cropland
Wetlands Remaining Wetlands
Grassland Remaining Grassland
Land Converted to Wetlands |< 0.5| |
Non-COz Emissions from Peatlands Remaining Peatlands | |< 0.5|
CH« Emissions from Land Converted to Coastal Wetlands I |< 0.5|
Non-COi Emissions from Drained Organic Soils | |< 0.51
N:0 Emissions from Forest Soils | |< 0.5|
Non-COz Emissions from Grassland Fires
N2O Emissions from Settlement Soils
Non-COz Emissions from Coastal Wetlands Remaining Coastal Wetlands
Land Converted to Grassland
Land Converted to Cropland a stock change
Non-COz Emissions from Forest Fires | Emissions
Land Converted to Settlements
(300) (250) (200) (150) (100) (50) 0 50 100
MMT CO* Eq.
2
3 Table 6-3: Emissions and Removals (Net Flux) from Land Use, Land-Use Change, and
4 Forestry (MMT CO2 Eq.)
Land-Use Category
1990
2005
2012
2013
2014
2015
2016
Forest Land Remaining Forest Land
(692.2)
(648.4)
(648.4)
(658.4)
(656.7)
(634.9)
(639.2)
Changes in Forest Carbon Stocks3
(697.7)
(664.6)
(666.9)
(670.9)
(669.3)
(666.2)
(670.5)
N011-CO2 Emissions from Forest Fires
5.3
15.6
17.9
11.9
11.9
30.7
30.7
N2O Emissions from Forest Soilsb
0.1
0.5
0.5
0.5
0.5
0.5
0.5
N011-CO2 Emissions from Drained
Organic Soils
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Land Converted to Forest Land
(92.0)
(81.6)
(74.9)
(74.9)
(75.0)
(75.0)
(75.0)
Changes in Forest Carbon Stocksc
(92.0)
(81.6)
(74.9)
(74.9)
(75.0)
(75.0)
(75.0)
Cropland Remaining Cropland
(40.9)
(26.5)
(21.4)
(11.4)
(12.0)
(6.3)
(9.9)
Changes in Mineral and Organic Soil
Carbon Stocks
(40.9)
(26.5)
(21.4)
(11.4)
(12.0)
(6.3)
(9.9)
Land Converted to Cropland
43.3
25.9
22.7
23.3
23.2
23.2
23.8
Changes in all Ecosystem Carbon Stocksd
43.3
25.9
22.7
23.3
23.2
23.2
23.8
Grassland Remaining Grassland
(4.1)
6.2
(19.6)
(3.3)
(6.7)
10.2
(1.0)
Changes in Mineral and Organic Soil
Carbon Stocks
(4.2)
5.5
(20.8)
(3.7)
(7.5)
9.6
(1.6)
N011-CO2 Emissions from Grassland Fires
0.2
0.7
1.2
0.4
0.8
0.7
0.6
Land Converted to Grassland
17.9
19.2
20.4
21.9
21.5
23.3
22.0
Changes in all Ecosystem Carbon Stocksd
17.9
19.2
20.4
21.9
21.5
23.3
22.0
Wetlands Remaining Wetlands
(4.0)
(5.3)
(4.1)
(4.1)
(4.1)
(4.1)
(4.1)
Changes in Organic Soil Carbon Stocks
in Peatlands
1.1
1.1
0.8
0.8
0.8
0.8
0.7
Changes in Mineral and Organic Soil
Carbon Stocks in Coastal Wetlands
(8.6)
(10.0)
(8.6)
(8.6)
(8.6)
(8.6)
(8.6)
CFLi Emissions from Coastal Wetlands
3.4
3.5
3.5
3.6
3.6
3.6
3.6
6-4 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
Remaining Coastal Wetlands
N2O Emissions from Coastal Wetlands
Remaining Coastal Wetlands
0.1
0.2
0.1
0.1
0.1
0.1
0.1
Non-CC>2 Emissions from Peatlands
Remaining Peatlands
+
+
+
+
+
+
+
Land Converted to Wetlands
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Changes in Mineral and Organic Soil
Carbon Stockse
(+)
(+)
(+)
(+)
(+)
(+)
(+)
CH4 Emissions from Land Converted to
Coastal Wetlands
+
+
+
+
+
+
+
Settlements Remaining Settlements
(84.8)
(88.9)
(96.5)
(97.1)
(98.6)
(99.6)
(101.2)
Changes in Organic Soil Carbon Stocks
0.1
0.5
1.3
1.3
1.3
1.3
1.3
Changes in Urban Tree Carbon Stocks
(60.4)
(80.5)
(88.4)
(89.5)
(90.6)
(91.7)
(92.9)
Changes in Yard Trimming and Food
Scrap Carbon Stocks in Landfills
(26.0)
(11.4)
(12.2)
(11.6)
(11.9)
(11.8)
(12.2)
N2O Emissions from Settlement Soilsf
1.4
2.5
2.7
2.6
2.6
2.5
2.5
Land Converted to Settlements
37.2
68.4
68.3
68.3
68.2
68.1
68.0
Changes in all Ecosystem Carbon Stocksd
37.2
68.4
68.3
68.3
68.2
68.1
68.0
LULUCF Emissions8
10.6
23.0
26.1
19.2
19.6
38.2
38.1
LULUCF Carbon Stock Change"
(830.2)
(754.2)
(779.5)
(755.0)
(760.0)
(733.4)
(754.9)
LULUCF Sector Net Total1
(819.6)
(731.1)
(753.5)
(735.8)
(740.4)
(695.2)
(716.8)
+ Absolute value does not exceed 0.05 MMT CO2 Eq.
a Includes the net changes to carbon stocks stored in all forest ecosystem pools (including drained and undrained organic
soils) and harvested wood products.
b Estimates include emissions from N fertilizer additions on both Forest Land Remaining Forest Land and Land Converted
to Forest Land.
c Includes the net changes to carbon stocks stored in all forest ecosystem pools (excludes drained organic soils which are
included in the flux from Forest Land Remaining Forest Land because it is not possible to separate the activity data at this
time).
d 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.
e Includes carbon stock changes for land converted to vegetated coastal wetlands.
f Estimates include 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 emissions include the CH4 and N2O 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 N2O emissions from Forest Soils and Settlement Soils.
h LULUCF Carbon Stock Change includes any C stock gains and losses from all land use and land use conversion categories.
1 The LULUCF Sector Net Total is the net sum of all CH4 and N2O emissions to the atmosphere plus net carbon stock
changes.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
1 Table 6-4: Emissions and Removals from Land Use, Land-Use Change, and Forestry (MMT
2 COz Eq.)
Gas/Land-Use Category
1990
2005
2012
2013
2014
2015
2016
Carbon Stock Change3
(830.2)
(754.2)
(779.5)
(755.0)
(760.0)
(733.4)
(754.9)
Forest Land Remaining Forest Land
(697.7)
(664.6)
(666.9)
(670.9)
(669.3)
(666.2)
(670.5)
Land Converted to Forest Land
(92.0)
(81.6)
(74.9)
(74.9)
(75.0)
(75.0)
(75.0)
Cropland Remaining Cropland
(40.9)
(26.5)
(21.4)
(11.4)
(12.0)
(6.3)
(9.9)
Land Converted to Cropland
43.3
25.9
22.7
23.3
23.2
23.2
23.8
Grassland Remaining Grassland
(4.2)
5.5
(20.8)
(3.7)
(7.5)
9.6
(1.6)
Land Converted to Grassland
17.9
19.2
20.4
21.9
21.5
23.3
22.0
Wetlands Remaining Wetlands
(7.6)
(8.9)
(7.7)
(7.8)
(7.8)
(7.8)
(7.9)
Land Converted to Wetlands
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Settlements Remaining Settlements
(86.2)
(91.4)
(99.2)
(99.8)
(101.2)
(102.2)
(103.7)
Land Converted to Settlements
37.2
68.4
68.3
68.3
68.2
68.1
68.0
CH4
6.7
13.3
15.0
10.9
11.2
22.4
22.4
Forest Land Remaining Forest Land:
Forest Fires
3.2
9.4
10.8
7.2
7.2
18.5
18.5
Land Use, Land-Use Change, and Forestry 6-5
-------
Wetlands Remaining Wetlands: Coastal
Wetlands Remaining Coastal Wetlands
3.4
3.5
3.5
3.6
3.6
3.6
3.6
Grassland Remaining Grassland:
Grassland Fires
0.1
0.3
0.6
0.2
0.4
0.3
0.3
Forest Land Remaining Forest Land:
Drained Organic Soils
+
+
+
+
+
+
+
Land Converted to Wetlands: Land
Converted to Coastal Wetlands
+
+
+
+
+
+
+
Wetlands Remaining Wetlands:
Peatlands Remaining Peatlands
+
+
+
+
+
+
+
N2O
3.9
9.7
11.1
8.3
8.4
15.8
15.7
Forest Land Remaining Forest Land:
Forest Fires
2.1
6.2
7.1
4.8
4.7
12.2
12.2
Settlements Remaining Settlements:
Settlement Soilsb
1.4
2.5
2.7
2.6
2.6
2.5
2.5
Forest Land Remaining Forest Land:
Forest Soilsc
0.1
0.5
0.5
0.5
0.5
0.5
0.5
Grassland Remaining Grassland:
Grassland Fires
0.1
0.3
0.6
0.2
0.4
0.3
0.3
Wetlands Remaining Wetlands: Coastal
Wetlands Remaining Coastal Wetlands
0.1
0.2
0.1
0.1
0.1
0.1
0.1
Forest Land Remaining Forest Land:
Drained Organic Soils
0.1
0.1
0.1
0.1
0.1
0.1
0.1
Wetlands Remaining Wetlands:
Peatlands Remaining Peatlands
+
+
+
+
+
+
+
LULUCF Emissions'1
10.6
23.0
26.1
19.2
19.6
38.2
38.1
LULUCF Carbon Stock Change3
(830.2)
(754.2)
(779.5)
(755.0)
(760.0)
(733.4)
(754.9)
LULUCF Sector Net Total6
(819.6)
(731.1)
(753.5)
(735.8)
(740.4)
(695.2)
(716.8)
+ Absolute value does not exceed 0.05 MMT CO2 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, said Land Converted to Settlements.
b Estimates include emissions from N fertilizer additions on both Settlements Remaining Settlements and Land Converted to
Settlements.
c Estimates include emissions from N fertilizer additions on both Forest Land Remaining Forest Land and Land Converted to
Forest Land.
d LULUCF emissions include the CH4 and N2O 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 N2O emissions from Forest Soils and Settlement Soils.
e The LULUCF Sector Net Total is the net sum of all CH4 and N2O emissions to the atmosphere plus net carbon stock
changes.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
1 Table 6-5: Emissions and Removals from Land Use, Land-Use Change, and Forestry (kt)
Gas/Land-Use Category
1990
2005
2012
2013
2014
2015
2016
Carbon Stock Change3
(830,249)
(754,155)
(779,547)
(755,006)
(760,007)
(733,352)
(754,902)
Forest Land Remaining Forest
Land
(697,690)
(664,566)
(666,869)
(670,857)
(669,250)
(666,188)
(670,456)
Land Converted to Forest Land
(92,018)
(81,576)
(74,883)
(74,948)
(74,978)
(75,003)
(75,024)
Cropland Remaining Cropland
(40,940)
(26,544)
(21,385)
(11,367)
(12,018)
(6,321)
(9,941)
Land Converted to Cropland
43,326
25,878
22,705
23,292
23,192
23,151
23,757
Grassland Remaining Grassland
(4,214)
5,492
(20,814)
(3,745)
(7,549)
9,596
(1,621)
Land Converted to Grassland
17,880
19,155
20,440
21,857
21,465
23,325
22,038
Wetlands Remaining Wetlands
(7,563)
(8,948)
(7,740)
(7,787)
(7,786)
(7,804)
(7,862)
Land Converted to Wetlands
(19)
(15)
(24)
(24)
(24)
(24)
(24)
Settlements Remaining
Settlements
(86,241)
(91,413)
(99,230)
(99,773)
(101,222)
(102,174)
(103,742)
Land Converted to Settlements
37,230
68,384
68,254
68,346
68,163
68,089
67,973
CH4
269
531
599
437
448
896
895
6-6 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
Forest Land Remaining Forest
Land: Forest Fires 127 377 433 286 289 740 740
Wetlands Remaining Wetlands:
Coastal Wetlands Remaining
Coastal Wetlands 138 140 142 142 142 143 143
Grassland Remaining
Grassland: Grassland Fires 3 13 23 8 16 13 11
Forest Land Remaining Forest
Land: Drained Organic Soils 1 1 11111
Land Converted to Wetlands:
Land Converted to Coastal
Wetlands 1 + + + + + +
Wetlands Remaining Wetlands:
Peatlands Remaining
Peatlands + + + + + + +
N2O 13 33
Forest Land Remaining Forest
Land: Forest Fires 7 21
Settlements Remaining
Settlements: Settlement Soilsb 5 8
Forest Land Remaining Forest
Land: Forest Soilsc + 2
Grassland Remaining
Grassland: Grassland Fires + 1
Wetlands Remaining Wetlands:
Coastal Wetlands Remaining
Coastal Wetlands + 1 + + + +
Forest Land Remaining Forest
Land: Drained Organic Soils + + + + + +
Wetlands Remaining Wetlands:
Peatlands Remaining
Peatlands + + + + + +
+ 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 emissions from N fertilizer additions on both Settlements Remaining Settlements and Land Converted to
Settlements.
c Estimates include emissions from N fertilizer additions on both Forest Land Remaining Forest Land and Land Converted to
Forest Land.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration
37
28
28
53
53
24
16
16
41
41
9
9
9
9
8
2
2
2
2
2
2
1
1
1
1
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 2006IPCC
Guidelines for National Greenhouse Gas Inventories (2006 IPCC Guidelines) and the 2013 Supplement to the 2006
Guidelines for National Greenhouse Gas Inventories: Wetlands. 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 this Inventory do not preclude
5 See .
Land Use, Land-Use Change, and Forestry 6-7
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
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.
Box 6-2: Biennial Inventory Compilation
For the current Inventory (i.e., 1990 through 2016 report), a biennial inventory compilation process lias been
implemented for the LULUCF and Agriculture chapters. As part of this biennial compilation process, during
alternating years, modified approaches will be applied to extend the emissions/removals time series of some
LULUCF and Agriculture source and sink categories rather than implementing a full inventory compilation (i.e.,
updating activity data and running models). In the current Inventory, for each category where these modified
approaches for extending the time series have been utilized, the alternative methods have been transparently
documented in their respective Methodology sections of the chapter. This biennial compilation schedule has been
adopted for the LULUCF and Agriculture chapters in order to conserve and efficiently utilize resources that are
needed to implement key improvements. Over the next four to six years, this process will result in more rapid
improvements to the LULUCF and Agriculture chapters. The next Inventory report (i.e., 1990 through 2017 report)
will include a full compilation of the LULUCF and Agriculture chapters along with a number of key improvements.
6.1 Representation of the U.S. Land Base
A national land-use categorization 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
umnanaged lands in the country (Table 6-6), (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 6-7), 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 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 lias
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 U.S.
land area into the thirty-six IPCC land-use and land-use change categories (Table 6-7) (IPCC 2006). The primary
databases are the U.S. Department of Agriculture (USDA) National Resources Inventory (NRI)6 and the USDA
Forest Service (USFS) Forest Inventory and Analysis (FIA)7 Database. The Multi-Resolution Land Characteristics
Consortium (MRLC) National Land Cover Dataset (NLCD)8 is also used to identify land uses in regions that were
6 NRI data are available at .
7 FIA data are available at .
8 NLCD data are available at and MRLC is a consortium of several U.S. government agencies.
6-8 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
not included in the NRI or FIA. New activity data were not compiled for this Inventory, however, so the 2015
estimates are used as a proxy for 2016. The time series will be updated with new activity data in the next Inventory
(i.e., 1990 through 2017 Inventory).
The total land area included in the U.S. Inventory is 936 million hectares across the 50 states.9 Approximately 890
million hectares of this land base is considered managed and 46 million hectares is unmanaged, which has not
changed by much over the time series of the Inventory (Table 6-7). In 2015, the United States had a total of 293
million hectares of managed Forest Land (2.4 percent increase since 1990), 163 million hectares of Cropland (6.6
percent decrease since 1990), 325 million hectares of managed Grassland (1.1 percent decrease since 1990), 42
million hectares of managed Wetlands (5.6 percent decrease since 1990), 43 million hectares of Settlements (29
percent increase since 1990), and 23 million hectares of managed Other Land (4 percent increase from 1990) (Table
6-7). Wetlands are not differentiated between managed and unmanaged, and are reported solely as managed.10 In
addition, C stock changes are not currently estimated for the entire land base, which leads to discrepancies between
the managed land area data presented here and in the subsequent sections of the Inventory (e.g., Forest Land
Remaining Forest Land, Grassland Remaining Grassland within interior Alaska).11 Planned improvements are
under development to account for C stock changes and greenhouse gas emissions on all managed land (e.g.,
Grasslands and Forest Lands in Alaska) 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 6-2). Forest Land tends to be more common in the eastern 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 6-6: Managed and Unmanaged Land Area by Land-Use Categories for All 50 States
(Thousands of Hectares)
Land-Use Categories
1990
2005
2012
2013
2014
2015
2016a
Managed Lands
889,924
889,914
889,897
889,896
889,896
889,896
889,896
Forest Land
286,612
289,064
292,439
292,879
293,180
293,480
293,480
Croplands
174,510
165,599
163,040
163,040
163,040
163,040
163,040
Grasslands
328,520
328,863
325,955
325,601
325,300
324,998
324,998
Settlements
33,370
40,298
43,118
43,118
43,118
43,118
43,118
Wetlands
45,004
43,523
42,558
42,471
42,472
42,474
42,474
Other Land
21,908
22,567
22,787
22,787
22,787
22,787
22,787
Unmanaged Lands
46,272
46,282
46,299
46,300
46,300
46,300
46,300
Forest Land
9,515
8,474
8,593
8,601
8,601
8,601
8,601
Croplands
0
0
0
0
0
0
0
Grasslands
25,953
27,043
26,942
26,936
26,936
26,936
26,936
Settlements
0
0
0
0
0
0
0
Wetlands
0
0
0
0
0
0
0
Other Land
10,804
10,765
10,764
10,764
10,764
10,764
10,764
Total Land Areas
936,196
936,196
936,196
936,196
936,196
936,196
936,196
Forest Land
296,127
297,538
301,032
301,480
301,780
302,081
302,081
Croplands
174,510
165,599
163,040
163,040
163,040
163,040
163,040
Grasslands
354,473
355,906
352,897
352,537
352,235
351,933
351,933
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.
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 United States 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 are reported as managed. See the Planned
Improvements section of the Inventory for future refinements to the Wetland area estimates.
11 These "managed area" discrepancies also occur in the Common Reporting Format (CRF) tables submitted to the UNFCCC.
Land Use, Land-Use Change, and Forestry 6-9
-------
Settlements 33,370 40,298 43,118 43,118 43,118 43,118 43,118
Wetlands 45,004 43,523 42,558 42,471 42,472 42,474 42,474
Other Land 32,713 33,332 33,551 33,551 33,551 33,551 33,551
1 a The land use data for the 2015 estimates are used as a proxy for 2016 because new activity data were not compiled for 2016 in
2 the current Inventory. New activity data will be compiled for the next Inventory (i.e., 1990 through 2017 report) to update the
3 time series.
4 Table 6-7: Land Use and Land-Use Change for the U.S. Managed Land Base for All 50 States
5 (Thousands of Hectares)
Land-Use & Land-Use
Change Categories3
1990
2005
2012
2013
2014
2015
2016"
Total Forest Land
286,612
289,064
292,439
292,879
293,180
293,480
293,480
FF
285,369
288,011
291,458
291,897
292,193
292,493
292,493
CF
213
193
165
165
165
165
165
GF
909
692
676
677
678
678
678
WF
24
27
28
28
32
31
31
SF
13
15
17
17
17
17
17
OF
84
126
95
95
95
95
95
Total Cropland
174,510
165,599
163,040
163,040
163,040
163,040
163,040
CC
162,051
150,583
149,722
149,722
149,722
149,722
149,722
FC
286
94
60
60
60
60
60
GC
11,754
14,418
12,827
12,827
12,827
12,827
12,827
WC
150
176
128
128
128
128
128
SC
76
85
91
91
91
91
91
oc
192
243
213
213
213
213
213
Total Grassland
328,520
328,863
325,955
325,601
325,300
324,998
324,998
GG
318,373
306,412
304,078
303,724
303,422
303,120
303,120
FG
1,154
4,114
3,961
3,961
3,961
3,961
3,961
CG
8,309
16,825
16,555
16,555
16,555
16,555
16,555
WG
231
429
199
199
199
199
199
SG
53
106
114
114
114
114
114
OG
400
976
1,048
1,048
1,048
1,048
1,048
Total Wetlands
45,004
43,523
42,558
42,471
42,472
42,474
42,474
WW
44,249
42,138
41,358
41,270
41,271
41,273
41,273
FW
43
62
55
55
56
56
56
CW
214
378
346
346
346
346
346
GW
452
835
700
700
700
700
700
SW
5
0
1
1
1
1
1
OW
41
110
98
98
98
98
98
Total Settlements
33,370
40,298
43,118
43,118
43,118
43,118
43,118
SS
30,469
31,978
35,848
35,848
35,848
35,848
35,848
FS
342
445
418
418
418
418
418
CS
1,247
3,550
2,982
2,982
2,982
2,982
2,982
GS
1,250
4,102
3,653
3,653
3,653
3,653
3,653
WS
6
25
26
26
26
26
26
OS
58
199
190
190
190
190
190
Total Other Land
21,908
22,567
22,787
22,787
22,787
22,787
22,787
OO
21,000
20,728
20,809
20,809
20,809
20,809
20,809
FO
41
68
75
75
75
75
75
CO
300
613
679
679
679
679
679
GO
481
982
1,109
1,109
1,109
1,109
1,109
WO
82
168
102
102
102
102
102
SO
5
9
13
13
13
13
13
Grand Total
889,924
889,914
889,897
889,896
889,896
889,896
889,896
a The abbreviations are "F" for Forest Land, "C" for Cropland, "G" for Grassland, "W" for Wetlands, "S" for
Settlements, and "O" 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
6-10 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
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).
b The land use data for the 2015 estimates are used as a proxy for 2016 because new activity data were not
compiled for 2016 in the current Inventory. New activity data will be compiled for the next Inventory (i.e., 1990
through 2017 Inventory report) to update the time series.
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 base, which leads to discrepancies between the managed land area data presented
here and in the subsequent sections of the Inventory.
Land Use, Land-Use Change, and Forestry 6-11
-------
1 Figure 6-2: Percent of Total Land Area for Each State in the General Land-Use Categories for
2 2015®
Croplands Forest Lands
Other Lands
Wetlands Settlements
3
4 a Updated land representation data have not been compiled in the current Inventory, therefore the state-scale
5 spatial patterns in this map are based on the previous Inventory (i.e., 1990 through 2015 report).
6-12 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Methodology
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 location data on all parcels of land. Approach 3 extends
Approach 2 by providing location data on all parcels of land, such as maps, along with the land-use history. 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 and FIA are Approach 2 data sources that do not provide spatially-explicit representations
of land use and land-use conversions, even though land use and land-use conversions are tracked explicitly at the
survey locations. NRI and FIA data are aggregated and used to develop a land-use conversion matrix for a political
or ecologically-defined region. NLCD is a spatially-explicit time series of land-cover data that is used to inform the
classification of land use, and is therefore Approach 3 data. 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.12
• 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
12 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. Therefore,
unless wetlands are managed for 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, all Wetlands are reported as managed, but emissions 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.
Land Use, Land-Use Change, and Forestry 6-13
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
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.13
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,14 while definitions of Cropland, Grassland, and Settlements are based on the NRI.15 The 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 centimeters)
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 hay or 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,16 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-asides17) are also classified as Cropland, as long as
these areas do not meet the Forest Land criteria. Roads through Cropland, including interstate highways,
state highways, other paved roads, gravel roads, dirt roads, and railroads are excluded from Cropland area
estimates and are, instead, classified as Settlements.
• Grassland: A land-use category on which the plant cover is composed principally of grasses, grass-like
plants (i.e., sedges and rushes), forbs, or shrubs suitable for grazing and browsing, and includes both
pastures and native rangelands. This includes areas where practices such as clearing, burning, chaining,
and/or chemicals are applied to maintain the grass vegetation. Land is also categorized as Grassland with
13 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.
14 See , page 22.
15 See .
16 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.
17 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.
6-14 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
three or fewer years of continuous hay production.18 Savannas, deserts, and tundra are considered
Grassland.19 Drained wetlands are considered Grassland if the dominant vegetation meets the plant cover
criteria for Grassland. Woody plant communities of low forbs and shrubs, such as mesquite, chaparral,
mountain shrub, 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 grasses, 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 the areas of 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, including Cropland (drained
wetlands for crop production and also systems that are flooded for most or just part of the year, such as rice
cultivation and cranberry production), Grassland (drained wetlands dominated by grass cover), Forest Land
(including drained or un-drained forested wetlands), and Settlements (drained wetlands in developed areas).
• Settlements: A land-use category representing developed areas consisting of units of 0.25 acres (0.1 ha) or
more 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 tracts of less than 10 acres (4.05 ha) that may
meet the definitions for Forest Land, 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-CCh 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 6-8). 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 the surveys contain additional information on management, site conditions, crop types, biometric
measurements, and other data that is needed to estimate C stock changes, N20, and CH4 emissions on those lands. If
NRI and FIA data are not available for an area, however, then the NLCD product is used to represent the land use.
18 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.
19 2006 IPCC Guidelines do not include provisions to separate desert and tundra as land-use categories.
Land Use, Land-Use Change, and Forestry 6-15
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Table 6-8: Data Sources Used to Determine Land Use and Land Area for the Conterminous
United States, Hawaii, and Alaska
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 (except Forest Land), 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 2012 from the NRI. The land use patterns are assumed to remain
the same from 2012 through 2016 for this Inventory, but the time series will be updated when new data are released.
Forest Inventory and Analysis
The FIA program, conducted by the USFS, is another statistically-based survey for the conterminous United States
in addition to the southeast and south central coastal Alaska, and the official source of data on Forest Land area and
management data for the Inventory. 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
6-16 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
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 westernUnited 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 2012 through
2015; see Table A-236).
National Land Cover Dataset
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 a portion of Alaska.20 Consequently, gaps exist in the land representation when the datasets are
combined, such as federal grassland operated by Bureau of Land Management (BLM), USD A, and National Park
Service, as well as Alaska.21 The NLCD is used as a supplementary database 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.
NLCD products provide land-cover for 1992, 2001, 2006, and 2011 in the conterminous United States (Homer et al.
2007), and also for Alaska in 2001 and 2011 and Hawaii in 2001. For the conterminous United States, the NLCD
data have been further processed to derive Land Cover Change Products for 2001, 2006, and 2011 (Fry et al. 2011;
Homer et al. 2007; Homer et al. 2015). 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 contain
21 categories of land-cover information, which 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 aggregated maps of IPCC land-use categories were used in combination with the NRI database to represent land
use and land-use change for federal lands, as well as federal and non-federal lands in Alaska. Specifically, NRI
survey locations designated as federal lands were assigned a land use/land use change category based on the NLCD
maps that had been aggregated into the IPCC categories. This analysis addressed shifts in land ownership across
years between federal or non-federal classes as represented in the NRI survey (i.e., the ownership is classified for
each survey location in the NRI). The sources of these additional data are discussed in subsequent sections of the
report.
Managed Land Designation
Lands are designated as managed in the United States based on the definition provided earlier in this section. In
order to apply the definition in an analysis of managed land, the following criteria are used:
• All Croplands and Settlements are designated as managed so only Grassland, Forest Land or Other Lands
may be designated as unmanaged land;
• All Forest Lands with active fire protection are considered managed;
• All Grassland is considered managed at a county scale if there are livestock in the county;22
20 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.
21 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.
22 Assuming all Grasslands are grazed in a county with even very small livestock populations is a conservative assumption about
human impacts on Grasslands. Currently, detailed information on grazing at sub-county scales is not available for the United
States to make a finer delineation of managed land.
Land Use, Land-Use Change, and Forestry 6-17
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
• 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 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 is conducted using a geographic information system. 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. The designation of grasslands as managed is
based on 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 (ESRI2008), 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
approximately 130 petroleum extraction sites and 223 mines. 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. The remaining land
represents the unmanaged land base. The resulting spatial product is used to identify NRI survey locations that are
considered managed and unmanaged for the conterminous United States and Hawaii,23 in addition to determining
which areas in the NLCD for Alaska are included in the managed land base.
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 using definitions developed to meet national circumstances, while adhering to IPCC (2006).24 In practice,
the land was initially classified into a variety of 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
23 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.
24 Definitions are provided in the previous section.
6-18 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
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 and Wetlands). All adjustments are made at the state scale to address the differences in Forest
Land definitions and the resulting discrepancies in areas among the land use and land-use change categories. There
are three steps in this process. The first step involves adjustments for Land Converted to Forest Land (Grassland,
Cropland, and Wetlands), followed by adjustments in Forest Land converted to another land use (i.e., Grassland,
Cropland, and Wetlands), and finally adjustments to 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 do not
provide specific land-use categories that are converted to Forest Land, but rather a sum of all Land Converted to
Forest Land. 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 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 does not provide information
on the specific land-use changes, 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 Wetland Remaining Wetland 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 Grasslands Remaining Grasslands 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 6-8). 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 coastal 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. FIA does have
survey plots in coastal 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. Forest land in interior
Alaska is currently being surveyed by the FIA program, but there is insufficient data at this time so forest
land in this region is based on the 2001 and 2011 NLCD. 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
Land Use, Land-Use Change, and Forestry 6-19
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
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 fluxes 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 lands 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.25
• Settlements: NRI captures non-federal settlement area in 49 states (excluding Alaska). If areas of Forest
Land or Grassland under 10 acres (4.05 ha) are contained within settlements or urban areas, they are
classified as Settlements (urban) in the NRI database. If these parcels exceed the 10 acre (4.05 ha) threshold
and are Grassland, they will be classified as such 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 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
25 This analysis does not distinguish between managed and unmanaged wetlands, which is a planned improvement for the
Inventory.
6-20 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
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 cranberries, 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 from these areas are 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).
QA/QC and Verification
The land base derived from the NRI, FIA, and NLCD was compared to the Topologically Integrated Geographic
Encoding and Referencing (TIGER) survey (U.S. Census Bureau 2010). The U.S. Census Bureau gathers data on the
U.S. 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 U.S. Census Survey. The U.S. Census Survey does not provide a time series of land-use
change data or land management information which is needed for reporting greenhouse gas emissions from land use
and land use change. Regardless, the U.S. Census Survey does provide sufficient information to provide a check on
the Inventory data. The U.S. Census Survey has about 46 million more hectares of land in the U.S. land base
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 U.S. Census, but not included in the land representation using the NRI, FIA and NLCD.
There is only a 0.4 percent difference when open water in coastal regions is removed from the TIGER data.
Recalculations Discussion
The land representation data in the current Inventory were not recalculated from the previous Inventory.
Planned Improvements
New land representation data were not compiled for the current Inventory. In addition, land use and land use change
area estimates for 2016 were assumed to be the same as the data for 2015 in the previous (i.e., 1990 through 2015)
Inventory report. Therefore, a key improvement in a future Inventory will be to update the time series for land
representation with the latest NRI, FIA, and NLCD data sets.
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 statistics,
but a complete accounting 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 6-3.
Box 6-3: 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 a land-cover product 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.
Land Use, Land-Use Change, and Forestry 6-21
-------
1
Table 6-9: Total Land Area (Hectares) by Land-Use Category for U.S. Territories
Northern
Puerto Rico
U.S. Virgin
Islands
Guam
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
2
3 Implementation is underway to apply methods in the 2013 Supplement to the 2006 Guidelines for National
4 Greenhouse Gas Inventories: Wetlands as part of the U.S. Greenhouse Gas Inventory. Specifically, greenhouse gas
5 emissions from coastal wetlands have been developed for the Inventory using the NOAA C-CAP land cover
6 product. The NOAA C-CAP product is not used directly in the land representation analysis, however, so a planned
7 improvement for the next (i.e., 1990 through 2017) Inventory report is to reconcile the coastal wetlands data from
8 the C-CAP product with the wetlands area data provided in the NRI. Further implementation of the new guidance
9 will have implications for the classification of managed and unmanaged wetlands in the Inventory report, and more
10 detailed wetlands datasets will likely also be evaluated and integrated into the analysis.
11 NOAA C-CAP data for Hawaii were recently released for 2011, and will be used to analyze land use change for this
12 state in the near future. There are also other databases that may need to be reconciled with the NRI and NLCD
13 datasets, particularly for Settlements. Urban area estimates, used to produce C stock and flux estimates from urban
14 trees, are currently based on population data (1990, 2000, and 2010 U.S. Census data). Using the population
15 statistics, "urban clusters" are defined as areas with more than 500 people per square mile. The USFS is currently
16 moving ahead with an Urban Forest Inventory program so that urban forest area estimates will be consistent with
17 FIA forest area estimates outside of urban areas, which would be expected to reduce omissions and overlap of forest
18 area estimates along urban boundary areas.
19 6.2 Forest Land Remaining Forest Land (CRF
20 Category 4A1)
21 Changes in Forest Carbon Stocks (CRF Category 4A1)
22 Delineation of Carbon Pools
23 For estimating carbon (C) stocks or stock change (flux), C in forest ecosystems can be divided into the following
24 five storage pools (IPCC 2006):
25 • Aboveground biomass, which includes all living biomass above the soil including stem, stump, branches,
26 bark, seeds, and foliage. This category includes live understory.
27 • Belowground biomass, which includes all living biomass of coarse living roots greater than 2 millimeters
28 (mm) diameter.
29 • Dead wood, which includes all non-living woody biomass either standing, lying on the ground (but not
30 including litter), or in the soil.
6-22 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
• Litter, which includes the litter, fumic, and humic layers, and all non-living biomass with a diameter less
than 7.5 centimeters (cm) at transect intersection, lying on the ground.
3
4
• 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.
5 In addition, there are two harvested wood pools included when estimating C flux:
6 • Harvested wood products (HWP) in use.
7 • HWP in solid waste disposal sites (SWDS).
8 Forest Carbon Cycle
9 Carbon is continuously cycled among the previously defined C storage pools and the atmosphere as a result of
10 biogeochemical processes in forests (e.g., photosynthesis, respiration, decomposition, and disturbances such as fires
11 or pest outbreaks) and anthropogenic activities (e.g., harvesting, thinning, and replanting). As trees photosynthesize
12 and grow, C is removed from the atmosphere and stored in living tree biomass. As trees die and otherwise deposit
13 litter and debris on the forest floor, C is released to the atmosphere and is also transferred to the litter, dead wood
14 and soil pools by organisms that facilitate decomposition.
15 The net change in forest C is not equivalent to the net flux between forests and the atmosphere because timber
16 harvests do not cause an immediate flux of all harvested biomass C to the atmosphere. Instead, harvesting transfers a
17 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
18 case of decomposition and as CO2, CH4, N20, CO, and NOx when the wood product combusts. The rate of emission
19 varies considerably among different product pools. For example, if timber is harvested to produce energy,
20 combustion releases C immediately, and these emissions are reported for information purposes in the Energy sector
21 while the harvest (i.e., the associated reduction in forest C stocks) and subsequent combustion are implicitly
22 estimated in the Land Use, Land-Use Change, and Forestry (LULUCF) sector (i.e., the harvested timber does not
23 enter the HWP pools). Conversely, if timber is harvested and used as lumber in a house, it may be many decades or
24 even centuries before the lumber decays and C is released to the atmosphere. If wood products are disposed of in
25 SWDS, the C contained in the wood may be released many years or decades later, or may be stored almost
26 permanently in the SWDS. These latter fluxes, with the exception of CH4 from wood in SWDS which is included in
27 the Waste sector, are also estimated in the LULUCF sector.
28 Net Change in Carbon Stocks within Forest Land of the United States
29 This section describes the general method for quantifying the net changes in C stocks in the five C storage pools and
30 two harvested wood pools. The underlying methodology for determining C stock and stock-change relies on data
31 from the Forest Inventory and Analysis (FIA) program within the USDA Forest Service. The annual forest inventory
32 system is implemented across all U.S. forest lands within the conterminous 48 states, but at this time does not
33 include interior Alaska, Hawaii, and U.S. Territories although inventories have been initiated in those states and
34 some territories. The methods for estimation and monitoring are continuously improved and these improvements are
35 reflected in the C estimates (Domke et al. 2016; Domke et al. 2017). First, the total C stocks are estimated for each
36 C storage pool, next the net changes in C stocks for each pool are estimated, and then the changes in stocks are
37 summed for all pools to estimate total net flux. The focus on C implies that all C-based greenhouse gases are
38 included, and the focus on stock change suggests that specific ecosystem fluxes do not need to be separately
39 itemized in this report. Changes in C stocks from disturbances, such as forest fires or harvesting, are included in the
40 net changes. For instance, an inventory conducted after fire counts only the trees that are left. Therefore, changes in
41 C stocks from natural disturbances, such as wildfires, pest outbreaks, and storms, are included in the forest inventory
42 approach; however, they are highly variable from year to year. The IPCC (2006) recommends estimating changes in
43 C stocks from forest lands according to several land-use types and conversions, specifically Forest Land Remaining
44 Forest Land and Land Converted to Forest Land, with the former being lands that have been forest lands for 20
45 years or longer and the latter being lands that have been classified as forest lands for less than 20 years. The methods
46 and data used to delineate forest C stock changes by these two categories continue to improve and in order to
47 facilitate this delineation, a combination of modeling approaches for carbon estimation were used this year in the
48 United States.
Land Use, Land-Use Change, and Forestry 6-23
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
Forest Area in the United States
Approximately 33 percent of the U.S. land area is estimated to be forested in 2016 based on the U.S. definition of
forest land as provided in the Section 6.1 Representation of the U.S. Land Base. Only FIA plots that were used in the
1990 through 2015 Inventory report were used in the current Inventory to ensure consistency with the other land use
categories and maintain the area estimates reported in the Land Representation, which are consistent with the 1990
through 2015 Inventory report area estimates because new area activity data were not compiled for the current
Inventory, and 2016 area estimates were assumed to remain the same as the 2015 estimates (see Section 6.1
Representation of the U.S. Land Base). The forest inventories from each of the conterminous 48 states (USDA
Forest Service 2016a, 2016b) comprise an estimated 266 million hectares of forest land that are considered managed
and are included in the current Inventory. An additional 6.2 million hectares of forest land in southeast and south
central coastal Alaska are inventoried and are also included here. Some differences exist in forest land area estimates
from the latest update to the Resources Planning Act (RPA) Assessment (Oswalt et al. 2014) and the forest land area
estimates included in this report, which are based on the annual inventory data used in the 1990 through 2015
Inventory report for all states (USDA Forest Service 2016b). Sufficient annual inventory data are not yet available
for Hawaii and interior Alaska, but estimates of these areas are included in Oswalt et al. (2014). Updated survey data
for central and western forest land in both Oklahoma and Texas have only recently become available, and these
forests contribute to overall C stocks reported below. 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. Agroforestry systems that meet the definition of
forest land are also not currently included in the current Inventory since they are not explicitly inventoried by either
the FIA program or the Natural Resources Inventory (NRI)26 of the USDA Natural Resources Conservation Service
(Perry et al. 2005).
An estimated 77 percent (211 million hectares) of U.S. forests in southeast and southcentral coastal Alaska 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 southeast and southcentral coastal Alaska
forest land and 80 percent of forest land in the conterminous United States are classified as timberland. Of the
remaining non-timberland, 30 million hectares are reserved forest lands (withdrawn by law from management for
production of wood products) and 69 million hectares are lower productivity forest lands (Oswalt et al. 2014).
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, forest land area in southeast and southcentral coastal Alaska and the conterminous United
States has increased by about 14 million hectares (Oswalt et al. 2014) with the southern region of the United States
containing the most forest land (Figure 6-3). A substantial portion of this accrued forest land is from the conversion
of abandoned croplands to forest (e.g., Woodall et al. 2015b). Current trends in the estimated forest land area in the
conterminous United States and the portion of southeast and south central coastal Alaska represented here show an
average annual rate of increase of 0.1 percent. In addition to the increase in forest area, the major influences to the
net C flux from forest land across the 1990 to 2016 time series are management activities and the ongoing impacts
of 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.27 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 the
conterminous U.S. states is about double the volume of annual removals on timberlands (Oswalt et al. 2014). 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.
26 The Natural Resources Inventory of the USDA Natural Resources Conservation Service is described in Section 6.1—
Representation of the U.S. Land Base.
27 The term "biomass density" refers to the mass of live vegetation per unit area. It is usually measured on a dry-weight basis.
Dry biomass is assumed to be 50 percent C by weight.
6-24 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Figure 6-3: Changes in Forest Area by Region for Forest Land Remaining Forest Land in the
conterminous United States and coastal Alaska (1990-2016, Million Hectares)
100-1
80-
0)
ro
T5
1
-C
c
o
f 60-
ro
a>
k_
03
To
£ 40-
20
> South
North
Rocky
Mountain
Pacific
Coast
| i i i i | i i i i | i i i i | i
1990 1995 2000 2005
Year
2010
~rT~r
2015
North
South
Rocky
Mountain
Pacific
Coast
Forest Carbon Stocks and Stock Change
In the United States, 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) of C each year from
1990 through 2016. 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 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). With sustainable harvesting practices and regeneration
of 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 C stocks in the forest ecosystem and harvested wood pools associated with Forest Land Remaining
Forest Land were estimated to result in net sequestration of 671.2 MMT CO- Eq. (183.1 MMT C) in 2016 (Table
6-10 and Table 6-11). The estimated net sequestration of C in the Forest Ecosystem was 571.6 MMT CO2 Eq. (155.9
Land Use, Land-Use Change, and Forestry 6-25
-------
1 MMT C) in 2016 (Table 6-10 and Table 6-11). The majority of this sequestration, 315.3 MMT CO2 Eq. (86.0 MMT
2 C), was from aboveground biomass in 2016. Overall, estimates of average C density in forest ecosystems (including
3 all pools) remained stable at approximately 0.0002 MMT C ha-1 from 1990 to 2016. This was calculated by dividing
4 the Forest Land area estimates by Forest Ecosystem C Stock estimates for every year (see Table 6-12) and then
5 calculating the mean across the entire time series, i.e., 1990 through 2016. The stable forest ecosystem C density
6 when combined with increasing forest area results in net C accumulation over time. These increases may be
7 influenced in some regions by reductions in C density or forest land area due to natural disturbances (e.g., wildfire,
8 weather, insects/disease). Aboveground live biomass is responsible for the majority of net sequestration among all
9 forest ecosystem pools (Figure 6-4).
10 The estimated net sequestration of C in HWP was 99.6 MMT CO2 Eq. (27.2 MMT C) in 2016 (Table 6-10 and
11 Table 6-11). The majority of this sequestration, 66.1 MMT CO2 Eq. (18.0 MMT C), was from wood and paper in
12 SWDS. Products in use were an estimated 33.5 MMT CChEq. (9.1 MMT C) in 2016.
13 Table 6-10: Net CO2 Flux from Forest Pools in Forest Land Remaining Forest Land and
14 Harvested Wood Pools (MMT CO2 Eq.)
Carbon Pool
1990
2005
2012
2013
2014
2015
2016"
Forest Ecosystem
(574.7)
(557.3)
(598.5)
(596.1)
(593.7)
(571.1)
(571.6)
Aboveground Biomass
(327.9)
(314.4)
(331.5)
(329.6)
(327.7)
(310.0)
(315.3)
Belowground Biomass
(70.0)
(66.6)
(69.7)
(69.2)
(68.7)
(64.6)
(65.7)
Dead Wood
(33.5)
(40.3)
(49.1)
(49.2)
(49.2)
(43.7)
(39.2)
Litter
(17.0)
(14.3)
(16.3)
(16.3)
(16.3)
(15.2)
(16.1)
Soil (Mineral)
(126.1)
(121.7)
(132.0)
(131.9)
(131.9)
(137.6)
(135.4)
Soil (Organic)3
(0.1)
+
0.1
0.1
0.1
0.1
0.094
Harvested Wood
(123.8)
(108.0)
(69.2)
(75.6)
(76.4)
(95.9)
(99.6)
Products in Use
(54.8)
(44.6)
(7.0)
(13.0)
(13.7)
(31.4)
(33.5)
SWDS
(69.0)
(63.5)
(62.2)
(62.6)
(62.7)
(64.4)
(66.1)
Total Net Flux
(698.4)
(665.3)
(667.6)
(671.6)
(670.0)
(666.9)
(671.2)
+ Absolute value does not exceed 0.05 MMT CO2 Eq.
a These estimates do not include C stock changes from drained organic soils. See Table 6-21 and Table
6-22 for CO2 emissions from drainage of organic soils from both Forest Land Remaining Forest Land
and Land Converted to Forest Land.
b The approach for estimating forest ecosystem carbon stock changes on Forest Land Remaining Forest
Land was consistent with the methods used in the 1990 through 2015 Inventory and is described in Annex
3.13. Only FIA plots that were used in the 1990 through 2015 Inventory were used in the current Inventory
to ensure consistency with the other land use categories and maintain the area estimates reported in the
Land Representation.
Notes: Forest ecosystem C stocks do not include forest stocks in U.S. Territories, Hawaii, a portion of
managed forests in Alaska, or trees on non-forest land (e.g., agroforestry systems and urban areas—see
section 6.10 Settlements Remaining Settlements for estimates of C stock change from urban trees).
Parentheses indicate net C sequestration (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.
15 Table 6-11: Net C Flux from Forest Pools in Forest Land Remaining Forest Land and
16 Harvested Wood Pools (MMT C)
Carbon Pool
1990
2005
2012
2013
2014
2015
2016b
Forest Ecosystem
(156.7)
(152.0)
(163.2)
(162.6)
(161.9)
(155.7)
(155.9)
Aboveground Biomass
(89.4)
(85.7)
(90.4)
(89.9)
(89.4)
(84.6)
(86.0)
Belowground Biomass
(19.1)
(18.2)
(19.0)
(18.9)
(18.7)
(17.6)
(17.9)
Dead Wood
(9.1)
(11.0)
(13.4)
(13.4)
(13.4)
(11.9)
(10.7)
Litter
(4.6)
(3.9)
(4.4)
(4.4)
(4.4)
(4.1)
(4.4)
Soil (Mineral)
(34.4)
(33.2)
(36.0)
(36.0)
(36.0)
(37.5)
(36.9)
Soil (Organic)3
+
+
+
+
+
+
0.026
Harvested Wood
(33.8)
(29.5)
(18.9)
(20.6)
(20.8)
(26.1)
(27.2)
6-26 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
Products in Use (14.9) (12.2) (1.9) (3.5) (3.7) (8.6) (9.1)
SWDS (18.8) (17.3) (17.0) (17.1) (17.1) (17.6) (18.0)
Total Net Flux (190.5) (181.5) (182.1) (183.2) (182.7) (181.9) (183.1)
+ Absolute value does not exceed 0.05 MMT C
aThese estimates do not include carbon stock changes from drained organic soils. See Table 6-21 and Table 6-22
for C stock changes from drainage of organic soils from Forest Land Remaining Forest Land and Land
Converted to Forest Land.
b The approach for estimating carbon stock changes on Forest Land Remaining Forest Land was consistent with
the methods used in the 1990 through 2015 Inventory and is described in Annex 3.13. Only FIA plots that were
used in the 1990 through 2015 Inventory were used in the current Inventory to ensure consistency with the other
land use categories and maintain the area estimates reported in the Land Representation.
Notes: Forest C stocks do not include forest stocks in U.S. Territories, Hawaii, a portion of managed lands in
Alaska, or trees on non-forest land (e.g., agroforestry systems and urban areas—see Section 6.10 Settlements
Remaining Settlements for estimates of C stock change from urban trees). Parentheses indicate net C
sequestration (i.e., a net removal of C from the atmosphere). Total net flux is an estimate of the actual net flux
between the total forest C pool and the atmosphere. Harvested wood estimates are based on results from annual
surveys and models. Totals may not sum due to independent rounding.
1 Stock estimates for forest ecosystem and harvested wood C storage pools are presented in Table 6-12. Together, the
2 estimated aboveground biomass and soil C pools account for a large proportion of total forest ecosystem C stocks.
3 Note that the forest land area estimates in Table 6-12 do not precisely match those in Section 6.1 Representation of
4 the U.S. Land Base for Forest Land Remaining Forest Land. This is because the forest land area estimates in Table
5 6-12 only include managed forest land in the conterminous 48 states and southeast and south central coastal Alaska
6 (which is the current area encompassed by FIA survey data, approximately 6.2 million ha) while the area estimates
7 in Section 6.1 include all managed forest land in Alaska (approximately 25.9 million ha with approximately 19.7
8 million ha in interior Alaska, which is not currently included in this Inventory) and Hawaii.
9 Table 6-12: Forest Area (1,000 ha) and C Stocks in Forest Land Remaining Forest Land and
10 Harvested Wood Pools (MMT C)
1990
2005
2012
2013
2014
2015
2016
2017"
Forest Area (1000 ha)
262,119
267,479
271,064
271,512
271,812
272,113
272,260
272,260
Carbon Pools (MMT C)
Forest Ecosystem
46,967
49,223
50,331
50,494
50,657
50,819
50,975
51,131
Aboveground Biomass
11,889
13,122
13,742
13,833
13,922
14,012
14,096
14,182
Belowground Biomass
2,439
2,700
2,831
2,850
2,869
2,888
2,905
2,923
Dead Wood
2,262
2,424
2,507
2,521
2,534
2,548
2,560
2,570
Litter
2,568
2,630
2,659
2,663
2,668
2,672
2,676
2,680
Soil (Mineral)
27,456
27,994
28,240
28,276
28,312
28,348
28,385
28,422
Soil (Organic)3
352
352
352
352
352
352
352
352
Harvested Wood
1,895
2,353
2,498
2,517
2,538
2,559
2,585
2,612
Products in Use
1,249
1,447
1,474
1,476
1,479
1,483
1,492
1,501
SWDS
646
906
1,025
1,042
1,059
1,076
1,093
1,111
Total C Stock
48,862
51,576
52,830
53,012
53,195
53,378
53,560
53,743
a These estimates do not include C stock changes from drained organic soils. See Table 6-21 and Table 6-22 for C stock changes
from drainage of organic soils from Forest Land Remaining Forest Land sad Land Converted to Forest Land.
b The approach for estimating carbon stock changes on Forest Land Remaining Forest Land was consistent with the methods used
in the 1990 through 2015 Inventory and is described in Annex 3.13. Only FIA plots that were used in the 1990 through 2015
Inventory were used in the current Inventory to ensure consistency with the other land use categories and maintain the area
estimates reported in the Land Representation. As a result, Forest Land area estimates were assumed to remain constant from
2016 to 2017 while carbon stocks increased in 2017 consistent with previous years in the time series.
Notes: Forest area andC stock estimates include all Forest Land Remaining Forest Land in the conterminous 48 states and
southeast and south central coastal Alaska (6.2 million ha), which is the current area encompassed by FIA survey data. Forest C
stocks do not include forest stocks in U.S. Territories, Hawaii, a large portion of interior Alaska (19.7 million ha), or trees on
non-forest land (e.g., agroforestry systems and urban areas—see section 6.10 Settlements Remaining Settlements for estimates of
C stock change from urban trees). The forest area estimates in this table do not match those in Section 6.1 Representation of the
U.S. Land Base, which includes all managed forest land in Alaska and Hawaii. 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
Land Use, Land-Use Change, and Forestry 6-27
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
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 2016 requires estimates of C stocks for 2016 and 2017.
Figure 6-4: Estimated Net Annual Changes in C Stocks for All C Pools in Forest Land
Remaining Forest Land in the Conterminous U.S. and Coastal Alaska (1990-2016, MMT C per
Year)
20-
0-
-20-
£ i
-4°H
_ o
.g1
•S E?
-60-
-80-
-100-
15
> r t
&
5 -120-
I §
II
-140-
-160-
-180-
-200-
-220-
i i i i | i i i i | i i i i | i i i i | i i i i | i
1995 2000 2005 2010 2015
Year
All forest ecosystem pools
Aboveground biomass
¦ Belowground biomass
Dead wood
¦ Litter
Soil (mineral)
¦ Soil (organic)
¦ Harvested Wood Products (HWP)
Products in use
Solid waste disposal sites
Total net change
(forest ecosystem + HWP)
Box 6-4: 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 CO: emissions from forest fires occurring in the conterminous states as well as the
portion of managed forest lands in Alaska that are captured in the current Inventory. 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-CO; greenhouse gas emissions
from forest fires that are also quantified in a separate section below as required by IPCC Guidance and UNFCCC
Reporting Requirements.
The IPCC (2006) methodology and a combination of U.S.-specific data on annual area burned and potential fuel
availability together with default combustion factors were employed to estimate CO2 emissions from forest fires.
6-28 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
The latest information on area burned is used to compile fire emissions for the U.S. At the time this Inventory was
compiled, fire data for 2016 were not available so estimates from 2015 were used. It is important to note that the
wildfire emissions in 2015 were markedly higher than in recent years. The 2016 estimates will be updated in
subsequent reports as fire data becomes available. Estimated CO2 emissions for wildfires in the conterminous 48
states and in Alaska as well as prescribed fires in 2016 were estimated to be 248.2 MMT CO2 peryear (Table 6-13).
This estimate is an embedded component of the net annual forest C stock change estimates provided previously (i.e.,
Table 6-11), but this separate approach to estimate emissions is necessary in order to associate a portion of
emissions, including estimates of CH4 and N20, with fire. See the discussion in Annex 3.13 for more details on this
methodology. Note that the estimates for Alaska provided in Table 6-13 include all managed forest land in the state
and are not limited to the subset with permanent inventory plots on managed lands as specified elsewhere in this
chapter (i.e., Table 6-11).
Table 6-13: Estimates of CO2 (MMT per Year) Emissions from Forest Fires in the
Conterminous 48 States and Alaska3
Year
CO2 emitted from
Wildfires in the
Conterminous 48
States (MMT yr1)
CO2 emitted from
Wildfires in Alaska
(MMTyr1)
CO2 emitted from
Prescribed Fires
(MMTyr1)
Total CO2 emitted
(MMTyr1)
1990
22.5
19.6
0.2
42.3
2005
44.1
80.6
1.3
125.9
2012
2013
2014
2015
2016b
138.6
67.9
84.1
164.1
164.1
2.7
22.3
4.9
80.7
80.7
2.9
5.5
6.1
3.5
3.5
144.2
95.6
95.0
248.2
248.2
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.
b The data for 2016 were unavailable when these estimates were summarized; therefore 2015, the most recent
available estimate, is applied to 2016.
Methodology and Data Sources
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, 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 approaches for estimating carbon stocks and stock changes on Forest Land Remaining Forest
Land were consistent with the methods used in the 1990 through 2015 Inventory and are described in Annex 3.13.
Only FIA plots that were used in the 1990 through 2015 Inventory were used in the current Inventory to ensure
consistency with the other land use categories and maintain the area estimates reported in the Land Representation,
which are a copy of the 1990 through 2015 Inventory area estimates because new area activity data were not
compiled for the current Inventory, and 2016 area estimates are held the same as the 2015 values (see Section 6.1
Representation of the U.S. Land Base). As a result, Forest Land area estimates were assumed to remain constant
from 2016 to 2017 while carbon stocks and stock changes increased in 2017 consistent with previous years in the
time series and based on the FIA plots that were used in the previous (1990 through 2015) Inventory. Forest Land
conditions were observed on FIA 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 tl to 2017. 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. Harvested wood C 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
Land Use, Land-Use Change, and Forestry 6-29
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
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 or harvested wood
products is provided here. See Annex 3.13 for details and additional information related to the methods and data.
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 include 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 1990s, necessitating the adoption of a
system to "backcast" the annual C estimates. To facilitate the backcasting of the U.S. annual forest inventory C
estimates, the estimation system used in this Inventory is 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 sequestration, 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 2016a, 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 2016d, 2016a). 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 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 2016d). 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 2016b, 2016c). 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. (201 la), 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. (201 la), 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
6-30 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
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 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. 201 la) 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 soil C to a depth of 100 cm from empirical measurements to a depth of 20 cm and included
site-, stand-, and climate-specific variables that yield predictions of soil C stocks specific to forest land in the United
States (Domke et al. 2017). This new approach allowed for separation of mineral and organic soils, also referred to
as Histosols, in the Forest Land Remaining Forest Land category. Note that mineral and organic soil C is 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 from organic soils in this section
(Table 6-10, Table 6-11, and Table 6-12) do not include emissions from drained organic soils. Estimates of C stock
changes from drainage of organic soils from Forest Land Remaining Forest Land and Land Converted to Forest
Land can be found in the Drained Organic Soils section below (Table 6-21 and Table 6-22).
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
here and for information purposes in the Energy sector, but the non-C02 emissions associated with biomass energy
are included in the Energy sector emissions (see Chapter 3).
Land Use, Land-Use Change, and Forestry 6-31
-------
1 Solidwood products include lumber and panels. End-use categories for solidwood include single and multifamily
2 housing, alteration and repair of housing, and other end-uses. There is one product category and one end-use
3 category for paper. Additions to and removals from pools were tracked beginning in 1900, with the exception that
4 additions of softwood lumber to housing, which began in 1800. Solidwood and paper product production and trade
5 data were taken from USDA Forest Service and other sources (Hair and Ulrich 1963; Hair 1958; USDC Bureau of
6 Census 1976; Ulrich 1985, 1989; Steer 1948; AF&PA 2006a, 2006b; Howard 2003, 2007, 2016, In preparation).
7 Estimates for disposal of products reflected the change over time in the fraction of products discarded to SWDS (as
8 opposed to burning or recycling) and the fraction of SWDS that were in sanitary landfills versus dumps.
9 There are five annual HWP variables that were used in varying combinations to estimate HWP contribution using
10 any one of the three main approaches listed above. These are:
11 (1 A) annual change of C in wood and paper products in use in the United States,
12 (IB) annual change of C in wood and paper products in SWDS in the United States,
13 (2A) annual change of C in wood and paper products in use in the United States and other countries where the
14 wood came from trees harvested in the United States,
15 (2B) annual change of C in wood and paper products in SWDS in the United States and other countries where
16 the wood came from trees harvested in the United States,
17 (3) C in imports of wood, pulp, and paper to the United States,
18 (4) C in exports of wood, pulp and paper from the United States, and
19 (5) C in annual harvest of wood from forests in the United States.
20 The sum of variables 2 A and 2B yielded the estimate for HWP contribution under the production estimation
21 approach. A key assumption for estimating these variables was that products exported from the United States and
22 held in pools in other countries have the same half-lives for products in use, the same percentage of discarded
23 products going to SWDS, and the same decay rates in SWDS as they would in the United States.
24 Uncertainty and Time-Series Consistency
25 A quantitative uncertainty analysis placed bounds on current flux for forest ecosystems through a combination of
26 sample-based and model-based approaches to uncertainty for forest ecosystem CO2 flux (IPCC Approach 1). A
27 Monte Carlo Stochastic Simulation of the Methods described above and probabilistic sampling of C conversion
28 factors were used to determine the HWP uncertainty (IPCC Approach2). See Annex 3.13 for additional information.
29 The 2016 net annual change for forest C stocks was estimated to be between -919.3 and -423.2 MMT CO2 Eq.
30 around a central estimate of -671.2 MMT CO2 Eq. at a 95 percent confidence level. This includes a range of -818.7
31 to -324.7 MMT CO2 Eq. around a central estimate of -571.6 MMT CO2 Eq. forforest ecosystems and -122.1 to
32 -76.3 MMT CO2 Eq. around a central estimate of -99.6 MMT CO2 Eq. for HWP.
33 Table 6-14: Quantitative Uncertainty Estimates for Net CO2 Flux from Forest Land
34 Remaining Forest Land: Changes in Forest C Stocks (MMT CO2 Eq. and Percent)
Source
Gas
2016 Flux Estimate
(MMT CO2 Eq.)
Uncertainty Range Relative to Flux Estimate
(MMT CO2 Eq.) (%)
Lower
Upper
Lower Upper
Bound
Bound
Bound Bound
Forest C Pools3
CO2
(571.6)
(818.7)
(324.7)
-43.2% 43.2%
Harvested Wood Products'5
CO2
(99.6)
(122.1)
(76.3)
-22.6% 23.4%
Total Forest
CO2
(671.2)
(919.3)
(423.2)
-37.0% 36.9%
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.
Note: Parentheses indicate negative values or net sequestration.
6-32 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
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 2016d).
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. (2014) or selected population
estimates generated from the FIA database, which are available at an FIA internet site (USDA Forest Service
2016b). Agreement between the C datasets and the original inventories is important to verily accuracy of the data
used. The methods and plots used in the current Inventory were the same as those used in the 1990 through 2015
Inventory. That said, all estimates were compiled again for the entire time series to ensure consistency. As a result,
Forest Land area estimates remained constant from 2016 to 2017 while carbon stocks increased in 2017 consistent
with previous years in the time series.
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, 2016, In preparation). Factors to convert wood and 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 CH4 emissions from landfills
based on EPA (2006) data are reasonable in comparison to CH4 estimates based on WOODCARB II landfill decay
rates.
Recalculations Discussion
The methods and data used in the current Inventory to compile estimates for forest ecosystem carbon stocks and
stock changes, as well as for HWP, from 1990 through 2015 are consistent with those used in the 1990 through 2015
Inventory so recalculations were not necessary.
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 is the third Inventory report submission to delineate C change by Forest Land Remaining Forest Land
and Land Converted to Forest Land and the second Inventory to report carbon stock changes for all IPCC pools in
these two categories, there are many improvements that are still necessary. Since the estimation approach used in the
current Inventory operates at the regional scale for the United States, research is underway to leverage auxiliary
Land Use, Land-Use Change, and Forestry 6-33
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
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., wild fire, insects, and disease) and human (e.g.,
harvesting) disturbances are implicitly included in the report given the design of the annual forest inventory system
but not explicitly estimated. 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 forest inventory data (USDA Forest Service 2016b). Also, several FIA database processes
are being institutionalized to increase efficiency and QA/QC in reporting and further improve transparency,
completeness, consistency, accuracy, and availability of data used in reporting. Finally, a Tier 1 approach was used
to estimate uncertainty associated with C stock changes in the Forest Land Remaining Forest Land category in this
report. There is research underway investigating more robust approaches to 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 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
with the Inventory report.
The foundation of forest C estimation and reporting is the annual forest inventory system. 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 2016b), particularly in western states. Hawaii and U.S. Territories
will be included when appropriate forest C data are available (only a small number of plots from Hawaii are
currently available from the annualized sampling design). A small portion of forest lands in interior Alaska are now
included in the annual forest inventory, however alternative methods of estimating C stock change will need to be
explored as it will take several years to re-measure newly established plots. To that end, research is underway to
incorporate all FIA plot information (both annual and periodic data) and the dense time series of remotely sensed
data in a design-based, model-assisted format 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.
201 lb). 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 FIA sampling frame extends beyond the
forest land use category (e.g., woodlands and urban areas) with inventory-relevant information for these lands which
will likely become increasingly available in coming years.
Box 6-5: Preliminary Estimates of Historical Carbon Stock Change and Methane Emissions from Managed Land
in Alaska (Represents Mean for Years 2000 to 2009)
Starting in the 1990s, a forest inventory of south central and southeastern coastal (SCSE) Alaska was initiated
following the same approach applied in the conterminous United States. These data have been used to compile
Forest Land estimates for SCSE Alaska in the Inventory since 2008. However, there still remain vast expanses of
Alaska that are in the U.S. managed land base (See Section 6.1) where forest inventories have only recently been
established and thus are not included as part of the greenhouse gas flux reporting in this Inventory. In addition, this
Inventory does not report on Grasslands in Alaska due to lack of land use and management data. Recognizing the
need to report on these emissions and removals, efforts have been initiated to apply a combination of approaches
that will eventually lead to complete reporting for all managed land in Alaska. The most promising near-term option
for Forest Lands that would meet the minimum UNFCCC reporting requirements is application of the IPCC Tier 1
Gain-Loss Method. Work is also underway to utilize forest inventory plots in combination with remote sensing to
estimate C stock changes. This work was initiated as a pilot study and lias now moved fully operational with the
annual forest inventory in interior Alaska underway. Full implementation of either of these approaches for reporting
in the Inventory is several years in the future.
In order to provide some insight into the greenhouse gas flux in Alaska, preliminary C stock change and CH4
emissions for Alaska have been developed using data from a recently completed USGS effort overlaid on the
Alaskan managed land base to provide a preliminary assessment of the mean historical anthropogenic greenhouse
gas flux between 2000 and 2009.
6-34 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 The assessment by the USGS, in collaboration with USD A Forest Service and the University of Alaska in Fairbanks,
2 estimated Alaska C stock changes and CH4 emissions using an approach that couples modeling, remote sensing
3 analysis, literature and database review (Zhu and McGuire, eds. 2016). Annual variation of soil and vegetation C
4 stocks and associated CO2 and CH4 fluxes, in both upland and wetland ecosystems in Alaska, were analyzed from
5 1950 to 2009, using this USGS modeling framework.
6 Results of the assessment include C stocks and fluxes from vegetation and soil organic C pools, and CH4 fluxes.
7 Vegetation C pools included aboveground and belowground biomass. The soil C pool included dead woody debris
8 and C stored in organic and mineral horizons. Carbon dioxide fluxes from vegetation net primary productivity, soil
9 heterotrophic respiration, wildfire emissions and harvest were estimated. Methane fluxes included biogenic and
10 pyrogenic sources. The results of this USGS analysis (i.e., mean values for 2000 to 2009 time period) overlaid on
11 the Alaskan managed land base are presented in Table 6-15.
12 Table 6-15: Mean C Stocks, CO2 and ChU Fluxes in Alaska between 2000 and 2009
Land Use: C Pool
Area (1,000
ha)a
C stock (MMT C)
CO2 Flux (Change
in C stocks)
(MMT CO2)
Eq./Year)b
CH4 Flux
(MMT CO2
Eq./Year)
Forest Land
39,917
15,226
44.86
1.675
Aboveground Biomass
Belowground Biomass
Soil"
-
2,130
532
12,563
4.03
40.83
-
Grassland"1
34,844
18,856
(30.60)
0.102
Aboveground
Vegetation
Belowground
Vegetation
Soil"
-
315
178
18,363
(5.83)
(24.77)
-
Wetland
12,346
3,927
17.52
23.170
Aboveground
Vegetation
Belowground
Vegetation
Soil"
-
264
176
3,487
1.12
16.41
-
Total
87,107
38,008
31.80
24.947
a The USGS assessment did not include the Aleutian Islands, Saint Lawrence Island, glacier, bare ground
or urban areas, therefore the area data does not match up precisely with the Land Representation analysis
in this Inventory (see Section 6.1 for more details).
b This assessment considers carbon exported out of the ecosystem from harvesting as a loss, it does not
include the contribution to the harvested wood products pool.
c Soil pool includes dead woody debris and C stored in organic and mineral horizons.
d Grassland also includes heath and shrubland.
Note: Parentheses indicate net sequestration.
i:>
14 i^jn-C02 Emissions from Forest Fires
15 Emissions of non-CCh gases from forest fires were estimated using U.S.-specific data for annual area of forest
16 burned and potential fuel availability as well as the default IPCC (2006) emissions and combustion factors applied to
17 the IPCC methodology. In 2016, emissions from this source were estimated to be 18.5 MMT CO2 Eq. of CH4 and
18 12.2 MMT CO2 Eq. of N20 (Table 6-16; kt units provided in Table 6-17). The estimates of non-CCh emissions from
19 forest fires include wildfires and prescribed fires in the conterminous 48 states and all managed forest land in
20 Alaska.
21 Table 6-16: Non-C02 Emissions from Forest Fires (MMT CO2 Eq.)a
Gas 1990 2005 2012 2013 2014 2015 2016"
CH4 3.2 9.4 10.8 7.2 7.2 18.5 18.5
Land Use, Land-Use Change, and Forestry 6-35
-------
NzO 2.1 6.2 7.1 4.8 4.7 12.2 12.2
Total 5.3 15.6 17.9 11.9 11.9 30.7 30.7
a These estimates include Non-CCh Emissions from Forest Fires on Forest Land Remaining
Forest Land said Land Converted to Forest Land.
bThe data for 2016 were unavailable when these estimates were developed, therefore 2015, the
most recent available estimate, is applied to 2016.
1 Table 6-17: N011-CO2 Emissions from Forest Fires (kt)a
Gas
1990
2005
2012
2013
2014
2015
2016"
CH4
127
377
433
286
289
740
740
N2O
7
21
24
16
16
41
41
CO
2,832
8,486
9,804
6,624
6,595
16,752
16,752
NOx
80
239
276
185
185
474
474
a These estimates include Non-CC>2 Emissions from Forest Fires on Forest Land Remaining
Forest Land and Land Converted to Forest Land.
b The data for 2016 were unavailable when these estimates were summarized, therefore 2015,
the most recent available estimate, is applied to 2016.
2 Methodology and Data Sources
3 N011-CO2 emissions from forest fires—primarily CH4 and N20 emissions—were calculated following IPCC (2006)
4 methodology, which included a combination of U.S. specific data on area burned and potential fuel available for
5 combustion along with IPCC default combustion and emission factors. The estimates were calculated according to
6 Equation 2.27 of IPCC (2006, Volume 4, Chapter 2), which is:
7 Emissions = Area burned x Fuel available x Combustion factor x Emission factor x 10 3
8 where area burned data are based on Monitoring Trends in Burn Severity (MTBS) data summaries (MTBS 2015),
9 fuel estimates are based on current C density estimates obtained from the latest FIA data for each state, and
10 combustion and emission factors are from IPCC (2006, Volume 4, Chapter 2). See Annex 3.13 for further details.
11 Uncertainty and Time-Series Consistency
12 In order to quantify the uncertainties for non-CCh emissions from wildfires and prescribed burns, a Monte Carlo
13 (IPCC Approach 2) sampling approach was employed to propagate uncertainty based on the model and data applied
14 for U.S. forest land. See IPCC (2006) and Annex 3.13 forthe quantities and assumptions employed to define and
15 propagate uncertainty. The results of the Approach 2 quantitative uncertainty analysis are summarized in Table 6-18.
16 Table 6-18: Quantitative Uncertainty Estimates of Non-C02 Emissions from Forest Fires
17 (MMT CO2 Eq. and Percent)3
Source
Gas
2016 Emission Estimate
Uncertainty Range Relative to Emission Estimateb
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Non-CC>2 Emissions from
Forest Fires
CH4
18.5
12.2
42.1
-34%
127%
Non-CC>2 Emissions from
Forest Fires
N2O
12.2
4.6
26.9
-62%
120%
a These estimates include Non-CC>2 Emissions from Forest Fires on Forest Land Remaining Forest Land wALand
Converted to Forest Land.
b Range of flux estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
18 Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
19 through 2016. Details on the emission trends through time are described in more detail in the Methodology section,
20 above.
6-36 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 QA/QC and Verification
2 Tier 1 and Tier 2 QA/QC activities were conducted consistent with the U.S. QA/QC plan. Source-specific quality
3 control measures for estimating non-C02 emissions from forest fires included checking input data, documentation,
4 and calculations to ensure data were properly handled through the inventory process. Further, the set of fire
5 emissions estimates using MODIS imagery and post-fire observations developed for Alaska by Veraverbeke et al.
6 (2015) (see Annex 3.13) were compared to the estimates of CO2 and C emissions from forest fires in Alaska (Table
7 6-13 and Annex 3.13). These alternate sources of data for annual areas burned and possible fuel availability in
8 Alaska were found to be similar to the data used here. The QA/QC procedures did not reveal any inaccuracies or
9 incorrect input values.
10 Recalculations Discussion
11 The methods used in the 1990 through 2016 Inventory to compile estimates of non-C02 emissions from forest fires
12 are consistent with those used in the 1990 through 2015 Inventory report. New data became available for 2015 and
13 were incorporated in the time series using the same methods as the 1990 through 2015 Inventory. The new data
14 resulted in an increase in both CH4 and N20 emissions in 2015.
15 Planned Improvements
16 Possible future improvements within the context of this same IPCC (2006) methodology are most likely to involve
17 greater specificity by fire or groups of fires and less reliance on wide regional values or IPCC defaults. Spatially
18 relating potential fuel availability to more localized forest structure is the best example of this. An additional
19 improvement would be the use of combustion factors that are more locally appropriate for the type, location, and
20 intensity of fire, which are currently unused information provided with the MTBS data summaries. All planned
21 improvements depend on future availability of appropriate U.S.-specific data.
22 N20 Emissions from N Additions to Forest Soils
23 Of the synthetic nitrogen (N) fertilizers applied to soils in the United States, no more than one percent is applied to
24 forest soils. Application rates are similar to those occurring on cropland soils, but in any given year, only a small
25 proportion of total forested land receives N fertilizer. This is because forests are typically fertilized only twice
26 during their approximately 40-year growth cycle (once at planting and once midway through their life cycle). While
27 the rate of N fertilizer application for the area of forests that receives N fertilizer in any given year is relatively high,
28 the annual application rate is quite low over the entire forest land area.
29 N additions to soils result in direct and indirect N20 emissions. Direct emissions occur on-site due to the N
30 additions. Indirect emissions result from fertilizer N that is transformed and transported to another location in a form
31 other than N20 (ammonia [NH3] and nitrogen oxide [NOx] volatilization, nitrate [NO3] leaching and runoff), and
32 later converted into N20 at the off-site location. The indirect emissions are assigned to forest land because the
33 management activity leading to the emissions occurred in forest land.
34 Direct soil N20 emissions from Forest Land Remaining Forest Land and Land Converted to Forest Land in 2016
35 were 0.3 MMT C02 Eq. (1 kt), and the indirect emissions were 0.1 MMT C02 Eq. (0.4 kt). Total emissions for 2016
36 were 0.5 MMT C02 Eq. (2 kt) and have increased by 455 percent from 1990 to 2016. Total forest soil N20
37 emissions are summarized in Table 6-19.
38 Table 6-19: N2O Fluxes from Soils in Forest Land Remaining Forest Land and Land Converted
39 to ForestLand(tAMT CO2 Eq. and kt N2O)
19'HI
2005
2012
2013
2014
2015
2016
Direct N2O Fluxes from Soils
MMT CO2 Eq.
0.1
0.3
0.3
0.3
0.3
0.3
0.3
ktN20
+ /
1
1
1
1
1
1
Indirect N2O Fluxes from Soils
MMT CO2 Eq.
0.0
0.1
0.1
0.1
0.1
0.1
0.1
ktN20
+
+
+
+
+
Land Use, Land-Use Change, and Forestry 6-37
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Total
MMTCCfcEq. 0.1 0.5 0.5 0.5 0.5 0.5 0.5
kt N2Q + 2 j 2 2 2 2 2
+ Does not exceed 0.05 MMT CO2 Eq. or 0.5 kt.
Note: Totals may not sum due to independent rounding.
Methodology and Data Sources
The IPCC Tier 1 approach is used to estimate N2O from soils within Forest Land Remaining 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 accounted for N fertilizer application to commercial Douglas-fir stands in western Oregon and
Washington. For the Southeast, estimates of direct N20 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). Not accounting for fertilizer applied to non-pine plantations is justified
because fertilization is routine for pine forests but rare for hardwoods (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
2016, 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 2016, 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 N20 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 N20 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 N20 off-site The resulting estimates are summed to obtain total indirect emissions.
Uncertainty and Time-Series Consistency
The amount of N20 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 N20
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 synthetic N
fertilizers are captured, so applications of organic N fertilizers are not estimated. However, the total quantity of
organic N inputs to soils is included in Section 5.4 Agricultural Soil Management and Section 6.9 Settlements
Remaining Settlements.
Uncertainties exist in the fertilization rates, annual area of forest lands receiving fertilizer, and the emission factors.
Fertilization rates are assigned a default level28 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 2016 emission estimates. IPCC (2006)
provided estimates for the uncertainty associated with direct and indirect N20 emission factor for synthetic N
fertilizer application to soils.
28 Uncertainty is unknown for the fertilization rates so a conservative value of ±50 percent is used in the analysis.
6-38 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Uncertainty is quantified using simple error propagation methods (IPCC 2006). The results of the quantitative
2 uncertainty analysis are summarized in Table 6-20. Direct N20 fluxes from soils in 2016 are estimated to be
3 between 0.1 and 1.1 MMT CO2 Eq. at a 95 percent confidence level. This indicates a range of 59 percent below and
4 211 percent above the 2016 emission estimate of 0.3 MMT CO2 Eq. Indirect N2O emissions in 2016 are 0.1 MMT
5 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
6 the 2016 emission estimate.
7 Table 6-20: Quantitative Uncertainty Estimates of N2O Fluxes from Soils in Forest Land
8 Remaining Forest Land and Land Con verted to Forest Land (MMT CO2 Eq. and Percent)
Source
Gas
2016 Emission Estimate
(MMT CO2 Eq.)
Uncertainty Range Relative to Emission Estimate
(MMT CO2 Eq.) (%)
Forest Land Remaining Forest
Lower Upper
Lower Upper
Land
Bound Bound
Bound Bound
Direct N2O Fluxes from Soils
N2O
0.3
0.1 1.1
-59% +211%
Indirect N2O Fluxes from Soils
N2O
0.1
+ 0.4
-86% +238%
Note: Due to rounding the upper and lower bounds may equal the emission estimate in the above table.
+ Does not exceed 0.05 MMT CO2 Eq.
9 The same methods are applied to the entire time series to ensure time-series consistency from 1990 through 2016,
10 and no recalculations have been done from the previous Inventory. Details on the emission trends through time are
11 described in more detail in the Methodology section, above.
12 QA/QC and Verification
13 The spreadsheet tab containing fertilizer applied to forests and calculations for N20 and uncertainty ranges are
14 checked and verified.
15 Planned Improvements
16 Additional data will be compiled to update estimates of forest areas receiving N fertilizer using surrogate data in the
17 next Inventory. Another improvement is to further disaggregate emissions by state for southeastern pine plantations
18 and northwestern Douglas-fir forests to estimate soil N20 emission. This improvement is contingent on the
19 availability of state-level N fertilization data for forest land.
20 C02, CH4, and N20 Emissions from Drained Organic Soils
21 Drained organic soils on forest land are identified separately from other forest soils largely because mineralization
22 of the exposed or partially dried organic material results in continuous CO2 and N20 emissions (IPCC 2006). In
23 addition, the 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands
24 (IPCC 2014) calls for estimating CH4 emissions from these drained soils and the ditch networks used to drain them.
25 Organic soils are identified on the basis of thickness of organic horizon and percent organic matter. All organic soils
26 are assumed to have originally been wet, and drained organic soils are further characterized by drainage or the
27 process of artificially lowering the soil water table, which exposes the organic material to drying and the associated
28 emissions described in this section. The land base considered here is drained inland organic soils that are coincident
29 with forest area as identified by the forest inventory of the USDA Forest Service (USDA Forest Service 2016).
30 The estimated area of drained organic soils on forest land is 70,849 ha and did not change over the time series based
31 on the data used to compile the estimates in the current Inventory. These estimates are based on permanent plot
32 locations of the forest inventory (USDA Forest Service 2016) coincident with mapped organic soil locations
33 (STATSG02 2016), which identifies forest land on organic soils. Forest sites that are drained are not explicitly
34 identified in the data, but for this estimate, planted forest stands on sites identified as mesic or xeric (which are
35 identified in USDA Forest Service 2016) are labeled "drained organic soil" sites.
36 Land use, region, and climate are broad determinants of emissions as are more site specific factors such as nutrient
37 status, drainage level, exposure, or disturbance. Current data are limited in spatial precision and thus lack site
Land Use, Land-Use Change, and Forestry 6-39
-------
1 specific details. At the same time, corresponding emissions factor data specific to U.S. forests are similarly lacking.
2 Tier 1 estimates are provided here following IPCC (2014). Total annual emissions on forest land with drained
3 organic soils in 2016 are estimated as 0.9 MMT CO2 Eq. peryear (Table 6-21).
4 The Tier 1 methodology provides methods to estimate C emission as CO2 from three pathways: direct emissions
5 primarily from mineralization; indirect, or off-site, emissions associated with dissolved organic carbon releasing
6 CO2 from drainage waters; and emissions from (peat) fires on organic soils. Data about forest fires specifically
7 located on drained organic soils are not currently available; as a result, no corresponding estimate is provided here.
8 Non-C02 emissions provided here include CH4 and N20. Methane emissions generally associated with anoxic
9 conditions do occur from the drained land surface but the majority of these emissions originate from ditches
10 constructed to facilitate drainage at these sites. Emission of N20 can be significant from these drained organic soils
11 in contrast to the very low emissions from wet organic soils.
12 Table 6-21: Estimated CO2 and N011-CO2 Emissions on Drained Organic Forest Soils3 (MMT
13 COz Eq.)
Source
1990
2005
2012
2013
2014
2015
2016
CO2, Direct
0.7
0.7
0.7
0.7
0.7
0.7
0.7
CO2, Dissolved
Organic C
0.1
0.1
0.1
0.1
0.1
0.1
0.1
CH4
+
+
+
+
+
+
+
N2O
0.1
0.1
0.1
0.1
0. 1
0.1
0.1
Total
0.9
0.9
0.9
0.9
0.9
0.9
0.9
+ Does not exceed 0.05 MMT CO2 Eq.
a This table includes estimates from Forest Land Remaining Forest Land and Land Converted to Forest
Land.
14 Table 6-22: Estimated C (MMT C) and Non-C02 (kt) Emissions on Drained Organic Forest
15 Soils®
Source
1990
2005
2012
2013
2014
2015
2016
C, Direct
0.2
0.2
0.2
0.2
0.2
0.2
0.2
C, Dissolved
Organic C
+
+
+
+
+
+
+
CH4
1
1
1
1
1
1
1
N2O
+
+
+
+
+
+
+
+ Does not exceed 0.05 MMT C or 0.5 kt.
a This table includes estimates from Forest Land Remaining Forest Land and Land Converted to Forest
Land.
16 Methodology and Data Sources
17 The Tier 1 methods for estimating emissions from drained inland organic soils on forest lands follow IPCC (2006),
18 with extensive updates and additional material presented in the 2013 Supplement to the 2006 IPCC Guidelines for
19 National Greenhouse Gas Inventories: Wetlands (IPCC 2014). With the exception of quantifying area of forest on
20 drained organic soils, which is user-supplied, all quantities necessary for Tier 1 estimates are provided in Chapter 2,
21 Drained Inland Organic Soils of IPCC (2014).
22 Estimated area of drained organic soils on forest land is 70,849 ha based on analysis of the permanent forest
23 inventory of the USDA Forest Service and did not change over the time series (data downloaded 14 June 2016). The
24 most-recent plot data per state within the inventories were used in a spatial overlay with the STATSG02 (2016)
25 data, and forest plots coincident with the soil order histosol were selected as having organic soils. Information
26 specific to identifying "drained organic" are not in the inventory data so an indirect approach was employed here.
27 Specifically, artificially regenerated forest stands (inventory field STDORGCD=l) on mesic or xeric sites (inventory
28 field 11
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
for each state (USDA Forest Service 2016). Eight states, all temperate forests, were identified as having drained
organic soils (Table 6-23).
Table 6-23: States identified as having Drained Organic Soils, Area of Forest on Drained
Organic Soils, and Sampling Error
State
Forest on Drained Organic
Soil (1,000 ha)
Sampling Error (68.3% as ±
Percentage of Estimate)
Florida
2.4
79
Georgia
3.7
71
Michigan
18.7
34
Minnesota
30.2
19
North Carolina
1.3
99
Virginia
2.3
102
Washington
2.1
101
Wisconsin
10.1
30
Total
70.8
14
The Tier 1 methodology provides methods to estimate emissions for three pathways of C emission as CO2 (Table
6-21 and Table 6-22) Note that subsequent mention of equations and tables in the remainder of this section refer to
Chapter 2 of IPCC 2014. The first pathway-direct CO2 emissions-is calculated according to Equation 2.3 and Table
2.1 as the product of forest area and emission factor for temperate drained forest land. The second pathway-indirect,
or off-site, emissions-is associated with dissolved organic carbon releasing CO2 from drainage waters according to
Equation 2.4 and Table 2.2, which represent a default composite of the three pathways for this flux: (1) the flux of
dissolved organic carbon (DOC) from natural (undrained) organic soil; (2) the proportional increase in DOC flux
from drained organic soils relative to undrained sites; and (3) the conversion factor for the part of DOC converted to
CO2 after export from a site. The third pathway-emissions from (peat) fires on organic soils-assumes that the
drained organic soils burn in a fire but not any wet organic soils. However, we currently do not include emissions
for this pathway because we do not have the combined fire and drained organic soils information; this may become
available in the future with additional analysis.
Non-C02 emissions, according to the Tier 1 method, include methane (CH4), nitrous oxide (N20), and carbon
monoxide (CO) (Table 6-16). Emissions associated with peat fires include factors for CH4 and CO in addition to
CO2, but fire estimates are assumed to be zero for the current Inventory, as discussed above. Methane emissions
generally associated with anoxic conditions do occur from the drained land surface but the majority of these
emissions originate from ditches constructed to facilitate drainage at these sites. From this, two separate emission
factors are used, one for emissions from the area of drained soils and a second for emissions from drainage ditch
waterways. Calculations are according to Equation 2.6 and Tables 2.3 and 2.4, which includes the default fraction of
the total area of drained organic soil which is occupied by ditches. Emissions of nitrous oxide can be significant
from these drained soils in contrast to the very low emissions from wet organic soils. Calculations are according to
Equation 2.7 and Table 2.5, which provide the estimate as kg N per year.
Uncertainty and Time-Series Consistency
Uncertainties are based on the sampling error associated with forest area and the uncertainties provided in the
Chapter 2 (IPCC 2014) emissions factors (Table 6-24). The estimates and resulting quantities representing
uncertainty are based on the 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 total emissions in 2016 from drained organic soils on
Forest Land Remaining Forest Land and Land Converted to Forest Land were estimated to be between 0.5 and 1.2
MMT CO2 Eq. around a central estimate of 0.9 MMT CO2 Eq. at a 95 percent confidence level.
Land Use, Land-Use Change, and Forestry 6-41
-------
1
2
Table 6-24: Quantitative Uncertainty Estimates for Annual CO2 and Non-C02 Emissions on
Drained Organic Forest Soils (MMT CO2 Eq. and Percent)3
2016 Emission
Source Estimate Uncertainty Range Relative to Emission Estimate
(MMT CO2 Eq.) (MMT CO2 Eq.) (%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
CO2, direct
0.7
0.4
0.9
-39%
39%
CO2, dissolved organic C
0.1
+
0.1
-56%
56%
CH4
+
+
+
-76%
76%
N2O
0.1
+
0.2
-124%
124%
Total
0.9
0.5
1.2
-38%
38%
+ Does not exceed 0.05 MMT CO2 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.
3 QA/QC and Verification
4 IPCC (2014) guidance cautions of a possibility of double counting some of these emissions. Specifically, the off-site
5 emissions of dissolved organic C from drainage waters may be double counted if soil C stock and change is based
6 on sampling and this C is captured in that sampling. Double counting in this case is unlikely since plots identified as
7 drained were treated separately in this chapter. Additionally, some of the non-CCh emissions may be included in
8 either the Wetlands or sections on N20 emissions from managed soils. These paths to double counting emissions are
9 unlikely here because these issues are taken into consideration when developing the estimates and this chapter is the
10 only section directly including such emissions on forest land.
11 Planned Improvements
12 Additional data will be compiled to update estimates of forest areas on drained organic soils as new reports are made
13 available and new geospatial products become available.
14 6.3 Land Converted to Forest Land (CRF
15 Category 4A2)
16 The C stock change estimates for Land Converted to Forest Land that are provided in this Inventory include all
17 forest land in an inventory year that had been in another land use(s) during the previous 20 years29 (USDA NRCS
18 2012). For example, cropland or grassland converted to forest land during the past 20 years would be reported in this
19 category. Converted lands are in this category for 20 years as recommended in the 2006 IPCC Guidelines (IPCC
20 2006), after which they are classified as Forest Land Remaining Forest Land. Estimates of C stock changes from all
21 pools (i.e., aboveground and belowground biomass, dead wood, litter and soils), as recommended by IPCC (2006),
22 are included in the Land Converted to Forest Land category of this Inventory.
29 The 2009 USDA National Resources Inventory (NRI) land-use survey points were classified according to land-use history
records starting in 1982 when the NRI survey began. Consequently, the classifications from 1990 to 2001 were based on less than
20 years. Furthermore, the FIA data used to compile estimates of carbon sequestration 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.
6-42 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Area of Land Converted to Forest in the United States
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 1990 to the present at an average rate of 1 million ha year1.
Since 1990, the conversion of grassland to forest land resulted in the largest source of C sequestration, accounting
for approximately 67 percent of the sequestration in the Land Converted to Forest Land category in 2016. However,
estimated gains have decreased over the time series due to less Grassland conversion into the Forest Land category
in recent years (see Table 6-25). The net flux of C from all forest pool stock changes in 2016 was -75.0 MMT CO2
Eq. (-20.5 MMT C) (Table 6-25 and Table 6-26). Note that soil C in this Inventory report 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 report soil C to a depth of 20 or 30 cm. To ensure consistency in the Land Converted to Forest
Land category where C stock transfers occur between land-use categories, all soil C estimates are based on methods
from Ogle et al. (2003, 2006) and IPCC (2006), which are also used in Cropland, Grasslands and Settlements land
use categories of this Inventory.
Table 6-25: 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
2012
2013
2014
2015
2016
Cropland Converted to Forest Land
(16.0)
(13.8)
(11.8)
(11.8)
(11.8)
(11.8)
(11.8)
Aboveground Biomass
(6.4)
(5.5)
(4.8)
(4.8)
(4.8)
(4.8)
(4.8)
Belowground Biomass
(0.5)
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
Dead Wood
(3.3)
(2.8)
(2.5)
(2.5)
(2.5)
(2.5)
(2.5)
Litter
(5.8)
(5.0)
(4.1)
(4.1)
(4.1)
(4.1)
(4.1)
Mineral Soil
(+)
(0.1)
+
+
+
+
+
Grassland Converted to Forest
Land
(63.6)
(51.2)
(50.0)
(50.1)
(50.1)
(50.1)
(50.1)
Aboveground Biomass
(31.5)
(25.0)
(25.5)
(25.5)
(25.5)
(25.5)
(25.5)
Belowground Biomass
7.6
6.3
5.9
5.9
5.9
5.9
5.9
Dead Wood
(14.6)
(11.9)
(11.4)
(11.4)
(11.4)
(11.4)
(11.4)
Litter
(25.0)
(20.3)
(19.1)
(19.1)
(19.1)
(19.1)
(19.1)
Mineral Soil
(0.1)
(0.2)
0.1
0.1
+
+
+
Other Land Converted to Forest
Land
(9.0)
(12.5)
(9.1)
(9.1)
(9.1)
(9.1)
(9.1)
Aboveground Biomass
(3.8)
(5.4)
(4.2)
(4.2)
(4.2)
(4.2)
(4.2)
Belowground Biomass
(0.7)
(1.0)
(0.8)
(0.8)
(0.8)
(0.8)
(0.8)
Dead Wood
(1.4)
(2.0)
(1.5)
(1.5)
(1.5)
(1.5)
(1.5)
Litter
(3.0)
(4.2)
(2.7)
(2.7)
(2.7)
(2.7)
(2.7)
Mineral Soil
(+)
(+)
+
(+)
(+)
(+)
(+)
Settlements Converted to Forest
Land
(1.3)
(1.5)
(1.8)
(1.8)
(1.8)
(1.8)
(1.8)
Aboveground Biomass
(0.6)
(0.7)
(0.8)
(0.8)
(0.8)
(0.8)
(0.8)
Belowground Biomass
(0.1)
(0.1)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Dead Wood
(0.2)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
Litter
(0.4)
(0.5)
(0.5)
(0.5)
(0.5)
(0.5)
(0.5)
Mineral Soil
(+)
(+)
+
+
+
+
+
Wetlands Converted to Forest Land
(2.2)
(2.5)
(2.2)
(2.2)
(2.2)
(2.2)
(2.2)
Aboveground Biomass
(1.0)
(1.1)
(1.0)
(1.0)
(1.0)
(1.0)
(1.0)
Belowground Biomass
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Dead Wood
(0.3)
(0.4)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
Litter
(0.7)
(0.8)
(0.7)
(0.7)
(0.7)
(0.7)
(0.7)
Mineral Soil
(+)
(+)
+
+
+
+
+
Land Use, Land-Use Change, and Forestry 6-43
-------
Total Aboveground Biomass Flux
(43.3)
(37.7)
(36.3)
(36.3)
(36.3)
(36.3)
(36.3)
Total Belowground Biomass Flux
6.1
4.5
4.4
4.4
4.4
4.4
4.4
Total Dead Wood Flux
(19.8)
(17.3)
(15.9)
(15.9)
(15.9)
(15.9)
(15.9)
Total Litter Flux
(34.8)
(30.8)
(27.2)
(27.2)
(27.2)
(27.2)
(27.2)
Total Mineral Soil Flux
(0.2)
(0.4)
0.1
0.1
0.1
+
+
Total Flux
(92.0)
(81.6)
(74.9)
(74.9)
(75.0)
(75.0)
(75.0)
+ Absolute value does not exceed 0.05 MMT CO2 Eq.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
1 Table 6-26: Net C Flux from Forest C Pools in Land Converted to Forest Land by Land Use
2 Change Category (MMT C)
Land Use/Carbon Pool
1990
2005
2012
2013
2014
2015
2016
Cropland Converted to Forest Land
(4.4)
(3.8)
(3.2)
(3.2)
(3.2)
(3.2)
(3.2)
Aboveground Biomass
(1.7)
(1.5)
(1.3)
(1.3)
(1.3)
(1.3)
(1.3)
Belowground Biomass
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Dead Wood
(0.9)
(0.8)
(0.7)
(0.7)
(0.7)
(0.7)
(0.7)
Litter
(1.6)
(1.4)
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
Mineral Soil
(+)
(+)
+
+
+
+
+
Grassland Converted to Forest Land
(17.3)
(13.9)
(13.6)
(13.7)
(13.7)
(13.7)
(13.7)
Aboveground Biomass
(8.6)
(6.8)
(7.0)
(7.0)
(7.0)
(7.0)
(7.0)
Belowground Biomass
2.1
1.7
1.6
1.6
1.6
1.6
1.6
Dead Wood
(4.0)
(3.2)
(3.1)
(3.1)
(3.1)
(3.1)
(3.1)
Litter
(6.8)
(5.5)
(5.2)
(5.2)
(5.2)
(5.2)
(5.2)
Mineral Soil
(+)
(+)
+
+
+
+
+
Other Land Converted to Forest
Land
(2.4)
(3.4)
(2.5)
(2.5)
(2.5)
(2.5)
(2.5)
Aboveground Biomass
(1.0)
(1.5)
(1.1)
(1.1)
(1.1)
(1.1)
(1.1)
Belowground Biomass
(0.2)
(0.3)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Dead Wood
(0.4)
(0.5)
(0.4)
(0.4)
(0.4)
(0.4)
(0.4)
Litter
(0.8)
(1.1)
(0.7)
(0.7)
(0.7)
(0.7)
(0.7)
Mineral Soil
(+)
(+)
+
(+)
(+)
(+)
(+)
Settlements Converted to Forest Land
(0.4)
(0.4)
(0.5)
(0.5)
(0.5)
(0.5)
(0.5)
Aboveground Biomass
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Belowground Biomass
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Dead Wood
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Litter
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
(0.1)
Mineral Soil
(+)
(+)
+
+
+
+
+
Wetlands Converted to Forest Land
(0.6)
(0.7)
(0.6)
(0.6)
(0.6)
(0.6)
(0.6)
Aboveground Biomass
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
(0.3)
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
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Mineral Soil
(+)
(+)
+
+
+
+
+
Total Aboveground Biomass Flux
(11.8)
(10.3)
(9.9)
(9.9)
(9.9)
(9.9)
(9.9)
Total Belowground Biomass Flux
1.7
1.2
1.2
1.2
1.2
1.2
1.2
Total Dead Wood Flux
(5.4)
(4.7)
(4.3)
(4.3)
(4.3)
(4.3)
(4.3)
Total Litter Flux
(9.5)
(8.4)
(7.4)
(7.4)
(7.4)
(7.4)
(7.4)
Total Mineral Soil Flux
(+)
(0.1)
+
+
+
+
+
Total Flux
(25.1)
(22.2)
(20.4)
(20.4)
(20.4)
(20.5)
(20.5)
+ Absolute value does not exceed 0.05 MMT C.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
3 Methodology
4 The following section includes a description of the methodology used to estimate stock changes in all forest C pools
5 for Land Converted to Forest Land. Forest Inventory and Analysis data and IPCC (2006) defaults for reference C
6 stocks were used to compile separate estimates for the five C storage pools. Estimates for Aboveground and
6-44 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Belowground Biomass, Dead Wood and Litter were based on data collected from the extensive array of permanent,
annual forest inventory plots and associated models (e.g., live tree belowground biomass estimates) in the United
States (USDA Forest Service 2015b, 2015c). Carbon conversion factors were applied at the disaggregated level of
each inventory 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 approach for estimating carbon stocks and stock changes in the Land Converted to Forest Land is consistent
with those used in the 1990 through 2015 Inventory report and is described in Annex 3.13. Only FIA plots that were
used in the 1990 through 2015 Inventory report were used in the current Inventory to ensure consistency with the
other land use categories and maintain the area estimates reported in the Land Representation, which are consistent
with the 1990 through 2015 Inventory report area estimates because new area activity data were not compiled for the
current Inventory, and 2016 area estimates were assumed to be the same as the 2015 estimates (see Section 6.1
Representation of the U.S. Land Base). Forest Land conditions were observed on FIA 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 2017. 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. (201 la), 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. (201 la), 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, Settlement, Wetlands) conversions to Forest Land no biomass
losses were assumed in the year of the conversion.
Carbon in Dead Organic Matter
Dead organic matter was initially calculated as three separate pools—standing dead trees, downed dead wood, and
litter—with C stocks estimated from sample data or from models. The standing dead tree C pool includes
aboveground and belowground (coarse root) biomass for trees of at least 12.7 cm dbh. Calculations followed the
basic method applied to live trees (Woodall et al. 2011a) with additional modifications to account for decay and
structural loss (Domke et al. 2011; Harmon et al. 2011). Downed dead wood estimates are based on measurement of
a subset of FIA plots for downed dead wood (Domke et al. 2013; Woodall and Monleon 2008; Woodall et al. 2013).
Downed dead wood is defined as pieces of dead wood greater than 7.5 cm diameter, at transect intersection, that are
not attached to live or standing dead trees. This includes stumps and roots of harvested trees. To facilitate the
downscaling of downed dead wood C estimates from the state-wide population estimates to individual plots, downed
dead wood models specific to regions and forest types within each region are used. Litter C is the pool of organic C
(also known as duff, humus, and fine woody debris) above the mineral soil and includes woody fragments with
diameters of up to 7.5 cm. A subset of FIA plots are measured for litter C. A modeling approach, using litter C
measurements from FIA plots (Domke et al. 2016) was used to estimate litter C for every FIA plot used in the
estimation framework.
Land Use, Land-Use Change, and Forestry 6-45
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
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). In the current Inventory, a linear regression model with autoregressive moving-average
errors was used to estimate the relationship between the surrogate data and the observed 1990 to 2012 data
(Brockwell and Davis 2016). Surrogate data are commodity statistics, weather data, or other information that can be
used to predict the emissions without compiling a new inventory. This estimate, along with observed surrogate data,
is used to predict emissions data for 2013 through 2016 for the Tier 2 method. 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 2012 National Resources Inventory
(NRI) (USDA-NRCS 2013), and National Land Cover Dataset (NLCD) (Homer et al. 2007). See Annex 3.12
(Methodology for Estimating N20 Emissions, CH4 Emissions and Soil Organic C Stock Changes from Agricultural
Soil Management) for more information about this method.
Uncertainty and Time-Series Consistency
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 6-27 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 10 percent below to 11 percent above the 2016 C stock
change estimate of -75.0 MMT CO2 Eq.
Table 6-27: Quantitative Uncertainty Estimates for Forest C Pool Stock Changes (MMT CO2
Eq. per Year) in 2016 from Land Converted to Forest Land by Land Use Change
Land Use/Carbon Pool 2016 Flux Estimate Uncertainty Range Relative to Flux Range*
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Cropland Converted to Forest Land
(11.8)
(13.5)
(8.5)
-14%
28%
Aboveground Biomass
(4.8)
(6.4)
(3.3)
-32%
32%
Belowground Biomass
(0.4)
(0.6)
(0.1)
-76%
76%
Dead Wood
(2.5)
(2.9)
(2.0)
-19%
19%
Litter
(4.1)
(4.6)
(3.7)
-12%
12%
Mineral Soils
+
(0.1)
0.1
-1,136%
1136%
Grassland Converted to Forest Land
(50.1)
(57.6)
(45.9)
-15%
8%
Aboveground Biomass
(25.5)
(31.8)
(19.2)
-25%
25%
Belowground Biomass
5.9
4.0
7.8
-31%
31%
Dead Wood
(11.4)
(14.0)
(8.8)
-23%
23%
Litter
(19.1)
(21.7)
(16.6)
-14%
14%
Mineral Soils
+
(0.2)
0.3
-2,684%
2,684%
Other Lands Converted to Forest Land
(9.1)
(10.2)
(8.0)
-12%
12%
Aboveground Biomass
(4.2)
(5.1)
(3.2)
-23%
23%
Belowground Biomass
(0.8)
(1.0)
(0.6)
-25%
25%
6-46 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Dead Wood
(1.5)
(1.8)
(1.1)
-24%
24%
Litter
(2.7)
(3.1)
(2.4)
-13%
13%
Mineral Soils
(+)
(0.1)
+
-370%
370%
Settlements Converted to Forest Land
(1.8)
(2.0)
(1.5)
-13%
13%
Aboveground Biomass
(0.8)
(1.0)
(0.6)
-25%
25%
Belowground Biomass
(0.2)
(0.2)
(0.1)
-27%
27%
Dead Wood
(0.3)
(0.3)
(0.2)
-24%
24%
Litter
(0.5)
(0.6)
(0.4)
-14%
14%
Mineral Soils
+
(+)
+
-800%
800%
Wetlands Converted to Forest Land
(2.2)
(2.5)
(2.0)
-11%
11%
Aboveground Biomass
(1.0)
(1.2)
(0.8)
-20%
20%
Belowground Biomass
(0.2)
(0.2)
(0.1)
-22%
22%
Dead Wood
(0.3)
(0.4)
(0.3)
-21%
27%
Litter
(0.7)
(0.8)
(0.6)
-13%
13%
Mineral Soils
+
(+)
+
-700%
667%
Total: Aboveground Biomass
(36.3)
(42.9)
(29.3)
-18%
19%
Total: Belowground Biomass
4.4
2.5
6.4
-43%
44%
Total: Dead Wood
(15.9)
(18.6)
(13.2)
-17%
17%
Total: Litter
(27.2)
(29.8)
(24.4)
-10%
10%
Total: Mineral Soils
+
(0.3)
0.3
-3,826%
3,825%
Total: Lands Converted to Forest Lands
(75.0)
(82.8)
(66.7)
-10%
11%
+ Absolute value does not exceed 0.05 MMT CO2 Eq.
a Range of flux estimate for 95 percent confidence interval
Note: Parentheses indicate net sequestration.
QA/QC and Verification
See QA/QC and Verification sections under Forest land Remaining Forest Land and Cropland Remaining
Cropland.
Recalculations Discussion
The approach for estimating carbon stock changes in Land Converted to Forest Land was consistent with the
methods used in the 1990 through 2015 Inventory report and is described in Annex 3.13. Only FIA plots that were
used in the 1990 through 2015 Inventory report were used in the current Inventory to ensure consistency with the
other land use categories and maintain the area estimates reported in the Land Representation. While the methods
and plots used in the current Inventory were the same as those used in the previous Inventory report (i.e., 1990
through 2015), the entire time series was compiled again when estimating the stock changes for 2016 and the
estimates over the time series were consistent with those reported in the 1990 through 2015 Inventory report.
Planned Improvements
There are many improvements necessary to improve the estimation of carbons 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 20 or 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 USD A 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
Land Use, Land-Use Change, and Forestry 6-47
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
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 may 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, and 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 (SOC) is the main source and sink for atmospheric CO2 in
most soils. IPCC (2006) recommends reporting changes in SOC stocks due to agricultural land-use and management
activities on both mineral and organic soils.30
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 during the 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 SOC 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., sewage
sludge) and flooding, can modify both organic matter inputs and decomposition, and thereby result in a net C stock
change (Parton et al. 1987; Paustian et al. 1997a; Conant et al. 2001; Ogle et al. 2005). 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.31 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 (and also settlements)
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 according to the 2012 United States Department of Agriculture (USD A) National
Resources Inventory (NRI) land-use survey for non-federal lands (USDA-NRCS 2015) or according to the National
Land Cover Dataset for federal lands (Homer et al. 2007; Fry et al. 2011; Homer et al. 2015). Cropland includes all
land used to produce food and fiber, in addition to forage that is harvested and used as feed (e.g., hay and silage),
and cropland that has been enrolled in the Conservation Reserve Program (CRP) (i.e., considered reserve 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 total amount of managed area in Cropland Remaining Cropland (see Section 6.1
Representation of the U.S. Land Base) and the cropland area included in the Inventory analysis (1.2 to 1.6 million
30 Carbon dioxide emissions associated with liming and urea application are also estimated but are included in the Agriculture
chapter of the report.
31 N2O emissions from soils are included in the Agricultural Soil Management section.
6-48 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
hectares or 0.8 percent of the total cropland areas in the United States between 1990 and 2015). Improvements are
underway to include croplands in Alaska as part of future C inventories.
Carbon dioxide emissions and removals32 due to changes in mineral soil 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 Methodology section for a list of crops in the Tier 2 and 3 methods) (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.
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 2016, mineral soils are estimated to sequester
39.7 MMT CO2 Eq. from the atmosphere (10.8 MMT C). This rate of C storage in mineral soils represents about a
44 percent decrease in the rate since the initial reporting year of 1990. Carbon dioxide emissions from organic soils
are 29.8 MMT CO2 Eq. (8.1 MMT C) in 2016, which is a 2 percent decrease compared to 1990. In total, United
States agricultural soils in Cropland Remaining Cropland sequestered approximately 9.9 MMT CO2 Eq. (2.7 MMT
C) in 2016.
Table 6-28: Net CO2 Flux from Soil C Stock Changes in Cropland Remaining Cropland (MMT
COz Eq.)
Soil Type
1990
2005
2012
2013
2014
2015
2016
Mineral Soils
(71.2)
(56.2)
(49.5)
(41.5)
(41.7)
(36.3)
(39.7)
Organic Soils
30.3
29.7
28.1
30.1
29.7
30.0
29.8
Total Net Flux
(40.9)
(26.5)
(21.4)
(11.4)
(12.0)
(6.3)
(9.9)
Notes: Estimates after 2012 are based on a surrogate data method (see Methodology section).
Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Table 6-29: Net CO2 Flux from Soil C Stock Changes in Cropland Remaining Cropland (MMT
C)
Soil Type
1990
2005
2012
2013
2014
2015
2016
Mineral Soils
Organic Soils
(19.4)
8.3
(15.3)
8.1
(13.5)
7.7
(11.3)
8.2
(11.4)
8.1
(9.9)
8.2
(10.8)
8.1
Total Net Flux
(11.2)
(7.2) .
(5.8)
(3.1)
(3.3)
(1.7)
(2.7)
Notes: Estimates after 2012 are based on a surrogate data method (see Methodology section).
Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Soil C stocks increase in Cropland Remaining Cropland largely due to sequestration in lands enrolled in CRP (i.e.,
set-aside program), as well as from conversion of land into hay production, adoption of conservation tillage (i.e.,
reduced- and no-till practices), and intensification of crop production by limiting the use of bare-summer fallow in
semi-arid regions. However, there is a decline in the net amount of C sequestration (i.e., 2016 is 44 percent less than
1990), and this decline is largely due to lower sequestration rates and less annual cropland enrolled in the CRP33 that
was initiated in 1985. Soil C losses from drainage of organic soils are relatively stable across the time series with a
small decline associated with the land base declining by 7 percent (based on 2012 estimates) for Cropland
Remaining Cropland on organic soils since 1990.
32 Removals occur through uptake of CO2 into crop and forage biomass that is later incorporated into soil C pools.
33 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 6-49
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
The spatial variability in the 2012 annual soil C stock changes34 are displayed in Figure 6-5 and Figure 6-6 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 with gains in the northern Great
Plains, w hich lias high rates of CRP enrollment. High rates of net C accumulation in mineral soils also occurred in
the Com 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), w hich coincides with the largest concentrations of
organic soils in the United States that are used for agricultural production.
Figure 6-5: Total Net Annual Soil C Stock Changes for Mineral Soils under Agricultural
Management within States, 2012, Cropland Remaining Cropland*
* Only national-scale soil C stock changes are estimated for 2013 to 2016 in the current Inventory using a
surrogate data method, and therefore the line-scale emission patterns in this map are based on inventory data from
2012. Negative values represent a net increase in soil C stocks, and positive values represent a net decrease in soil
C stocks.
34 Only national-scale emissions are estimated for 2013 to 2016 in this Inventory using the surrogate data method, and therefore
the fine-scale emission patterns in this map are based on inventory data from 2012.
6-50 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Figure 6-6: Total Net Annual Soil C Stock Changes for Organic Soils under Agricultural
Management within States, 2012, Cropland Remaining Cropland*
¦ > 40
* Only national-scale soil C stock changes are estimated for 2013 to 2016 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
2012.
Methodology
The following section includes a description of the methodology used to estimate changes in soil 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.
Soil 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 2015).
The NRI is a statistically-based sample of all non-federal land, and includes approximately 609,211 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 (i.e., each
expansion factor represents the amount of area with the same land-use/management history as the sample point).
Land-use and some management information (e.g., crop type, soil attributes, and irrigation) were collected for each
NRI point on a 5-year cycle beginning from 1982 through 1997. For cropland, data had 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
2012 (USDA-NRCS 2015). NRI survey locations are classified as Cropland Remaining Cropland in a given year
between 1990 and 2012 if the land use had 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 C stock changes for mineral soils on
the majority of land that is used to produce annual crops in the United States. These crops include alfalfa hay.
Land Use, Land-Use Change, and Forestry 6-51
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
barley, corn, cotton, dry beans, grass hay, grass-clover hay, lentils, oats, onions, peanuts, peas, potatoes, rice,
sorghum, soybeans, sugar beets, sunflowers, tobacco, tomatoes, and wheat, but is not applied to estimate 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 C stock changes and soil
nitrous oxide (N20) emissions from agricultural soil management. 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 N20) 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, tobacco, 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 C stock changes on
federal croplands. Mineral SOC 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. Further elaboration on the methodology and data used to
estimate stock changes from mineral soils are described below and in Annex 3.12.
A surrogate data method is used to estimate soil C stock changes from 2013 to 2016 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 2012 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 (https://quickstats.nass.usda.gov/),
and weather data from the PRISM Climate Group (PRISM 2015). See Box 6-6 for more information about the
surrogate data method. Stock change estimates for 2013 to 2016 will be recalculated in future inventories when new
NRI data are available.
Box 6-6: Surrogate Data Method
J!
Time series extension is needed because the inventory is currently compiled every two years for many categories in
the Agriculture, Forestry, and Other Land Use (AFOLU) sector in order to conserve resources that are needed to
implement improvements, and even in years that the inventory is compiled, there are typically gaps at the end of the
time series. This is mainly because the National Resources Inventory (NRI), which provides critical data for
estimating greenhouse gas emissions and removals, does not release data every year.
A surrogate data method has been used to impute missing emissions at the end of the time series for soil C stock
changes in Cropland Remaining Cropland, Land Converted to Cropland, Grassland Remaining Grassland, and
Land Converted to Grassland. A linear regression model with autoregressive moving-average (ARMA) errors
(Brockwell and Davis 2016) is used to estimate the relationship between the surrogate data and the modeled 1990 to
2012 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 + e,
where Y is the response variable (e.g., soil organic carbon), X(3 contains specific surrogate data depending on the
response variable, and e is the remaining unexplained error. Models with a variety of surrogate data were tested,
including commodity statistics, weather data, or other relevant information. Parameters are estimated from the
emissions data for 1990 to 2012 using standard statistical techniques, and these estimates are used to predict the
missing emissions data for 2013 to 2016.
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. Specifically, uncertainty will
increase for years with imputed estimates based on the splicing methods, compared to those years in which the full
inventory is compiled. This 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 2012), estimating emissions from each model and deriving confidence
6-52 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
intervals, which propagates uncertainties through the calculations from the original inventory and the surrogate data
method.
Tier 3 Approach. Mineral SOC stocks and stock changes are estimated using the DAYCENT biogeochemical35
model (Parton et al. 1998; Del Grosso et al. 2001, 2011), which is able to simulate 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. The modeling approach uses daily weather data as an input, along
with information about soil physical properties. 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, and grazing). The model simulates net primary productivity
(NPP) using the NASA-CASA production algorithm MODIS Enhanced Vegetation Index (EVI) products,
MOD13Q1 and MYD13Q1, for most croplands36 (Potter et al. 1993, 2007). The model also simulates soil
temperature, and water dynamics, in addition to turnover, stabilization, and mineralization of soil organic matter C
and nutrients (N, P, K, S). 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 6-7 for additional information).
mum! j
A Tier 3 model-based approach is used to estimate soil C stock changes on the majority of agricultural land on
mineral soils. This approach results in a more complete and accurate accounting of soil 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 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 of combinations in the United States. In contrast, the Tier 3
model simulates soil C dynamics at more than 300,000 individual NRI survey locations in individual fields.
(3) The IPCC Tier 1 and 2 methods use a simplified approach to 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 (i.e., daily time-step version of
the Century model) 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 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 2012 USDA NRI
survey (USDA-NRCS 2015). Additional sources of activity data are used to supplement the land-use information
35 Biogeochemical cycles are the flow of chemical elements and compounds between living organisms and the physical
environment.
36 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 2012. Other regions and years prior to 2000 are simulated with a method that incorporates
water, temperature and moisture stress on crop production (see Metherell et al. 1993), but does not incorporate the additional
information about crop condition provided with remote sensing data.
Land Use, Land-Use Change, and Forestry 6-53
pproach for Soil C Stocks Compared to Tier 1 or 2 Approach
i
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
from the NRI. The Conservation Technology Information Center (CTIC 2004) provided annual data on tillage
activity at the county level for the conterminous United States between 1989 and 2004, and these data are adjusted
for long-term adoption of no-till agriculture (Towery 2001). No-till adoption is assumed to remain constant from
2005 through 2012 due to lack of data, but there is a planned improvement to update the tillage histories with a
dataset that was recently released by the USD A (Conservation Effects Assessment Program Data, See Planned
Improvements section). Information on fertilizer use and rates by crop type for different regions of the United States
are obtained primarily from the USD A Economic Research Service. The data collection program was known as the
Cropping Practices Surveys through 1995 (USDA-ERS 1997), and then became the Agricultural Resource
Management Surveys (ARMS) (USDA-ERS 2015). Additional data are compiled through other sources particularly
the National Agricultural Statistics Service (NASS 1992, 1999, 2004). Frequency and rates of manure application to
cropland for 1997 are estimated from data compiled by the USD A Natural Resources Conservation Service
(Edmonds et al. 2003), and then adjusted using county-level estimates of manure available for application in other
years. Specifically, county-scale ratios of manure available for application to soils in other years relative to 1997 are
used to adjust the area amended with manure (see Annex 3.12 for further details). Greater availability of managed
manure N relative to 1997 is assumed to increase the area amended with manure, while reduced availability of
manure N relative to 1997 is assumed to reduce the amended area. Data on the county-level N available for
application are estimated for managed systems based on the total amount of N excreted in manure minus N losses
during storage and transport, and include the addition of N from bedding materials. Nitrogen losses include direct
N20 emissions, volatilization of ammonia and NOx, N runoff and leaching, and the N in poultry manure used as a
feed supplement. More information on livestock manure production is available in Section 5.2 Manure Management
and Annex 3.11.
Daily weather data are another input to the model simulations. These data are based on a 4 kilometer gridded
product from the PRISM Climate Group (2015). Soil attributes are obtained from the Soil Survey Geographic
Database (SSURGO) (Soil Survey Staff 2016). The C dynamics at each NRI point are simulated 100 times as part of
the uncertainty analysis, yielding a total of over 18 million simulation runs for the analysis. Uncertainty in the C
stock estimates from DAYCENT associated with parameterization and model algorithms are adjusted using a
structural uncertainty estimator accounting for uncertainty in model algorithms and parameter values (Ogle et al.
2007, 2010). Carbon stocks and 95 percent confidence intervals are estimated for each year between 1990 and 2012
using the NRI survey data (which is available through 2012). However, the areas may have changed through the
process in which the NRI survey data are reconciled with 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). This process ensures that
the areas of Forest Land Remaining Forest Land and Land Converted to Forest Land are consistent in all three
datasets, and leads to some modification of other lands use areas to ensure the total land area of the United States
does not change. 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).
Soil C stock changes from 2013 to 2016 are estimated using a surrogate data method that is described in Box 6-6.
Future Inventories will be updated with new NRI activity data when the data are made available, and the time series
from 2013 to 2016 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 soil C stock change factors (Ogle et al. 2003, 2006). 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 provided a more
robust sample for estimating the reference condition. U.S.-specific C stock change factors are derived from
published literature to determine the impact of management practices on SOC storage (Ogle et al. 2003, 2006). The
factors include changes in tillage, cropping rotations, intensification, and land-use change between cultivated and
uncultivated conditions. U.S. 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.
Climate zones in the United States are classified using mean precipitation and temperature (1950 to 2000) variables
from the WorldClim data set (Hijmans et al. 2005) and potential evapotranspiration data from the Consortium for
6-54 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Spatial Information (CGIAR-CSI) (Zomer et al. 2008, 2007) (Figure A-9). IPCC climate zones are then assigned to
NRI point locations.
Activity data are primarily based on the historical land-use/management patterns recorded in the 2012 NRI (USDA-
NRCS 2015). Each NRI point is classified by land use, soil type, climate region, and management condition. Survey
locations on federal lands are included in the NRI, but land use and cropping history are not compiled at these
locations in the survey program (i.e., NRI is restricted to data collection on non-federal lands). Land-use patterns at
the NRI survey locations on federal lands are based on the National Land Cover Database (NLCD) (Fry et al. 2011;
Homer et al. 2007; Homer et al. 2015). Classification of cropland area by tillage practice is based on data from the
Conservation Technology Information Center (CTIC 2004; Towery 2001) as described in the Tier 3 approach above.
Activity data on wetland restoration of Conservation Reserve Program land are obtained from Euliss and Gleason
(2002). Manure N amendments over the inventory time period are based on application rates and areas amended
with manure N from Edmonds et al. (2003), in addition to the managed manure production data discussed in the
methodology subsection for the Tier 3 approach. Utilizing information from these data sources, SOC stocks for
mineral soils are estimated 50,000 times for 1990 through 2012, using a Monte Carlo stochastic simulation approach
and probability distribution functions for U.S.-specific stock change factors, reference C stocks, and land-use
activity data (Ogle et al. 2002; Ogle et al. 2003; Ogle et al. 2006).
Soil C stock changes from 2013 to 2016 are estimated using a surrogate data method that is described in Box 6-6. As
with the Tier 3 method, future inventories will be updated with new NRI activity data when the data are made
available, and the time series 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 U.S.-specific C loss rates (Ogle et al. 2003) rather than default IPCC rates.
The final estimates included a measure of uncertainty as determined from the Monte Carlo Stochastic Simulation
with 50,000 iterations. Emissions are based on the annual data for drained organic soils from 1990 to 2012 for
Cropland Remaining Cropland areas in the 2012 NRI (USDA-NRCS 2015). A surrogate data method is used to
estimate annual C emissions from organic soils from 2013 to 2016 as described in Box 6-6 of this section. Estimates
for 2013 to 2016 will be recalculated in future inventories when new NRI data are available.
Uncertainty and Time-Series Consistency
Uncertainty associated with the Cropland Remaining Cropland land-use category is addressed for changes in
agricultural soil C stocks (including both mineral and organic soils). Uncertainty estimates are presented in Table
6-30 for each subsource (mineral soil C stocks 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 2013 to 2016, there is additional uncertainty
propagated through the Monte Carlo Analysis associated with the surrogate data method. Soil C stock changes from
the Tier 2 and 3 approaches are combined using the simple error propagation method provided by the IPCC (2006).
The combined uncertainty is calculated by taking the square root of the sum of the squares of the standard deviations
of the uncertain quantities. The combined uncertainty for soil C stocks in Cropland Remaining Cropland ranged
from 452 percent below to 452 percent above the 2016 stock change estimate of -9.9 MMT CO2 Eq. The large
relative uncertainty around the 2016 stock change estimate is partly due to variation in soil C stock changes that are
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
2016 Flux Estimate
(MMT CO2 Eg.)
Uncertainty Range Relative to Flux Estimate3
(MMT CO2 Eq.) (%}
Mineral Soil C Stocks: Cropland Remaining
Cropland, Tier 3 Inventory Methodology
Mineral Soil C Stocks: Cropland Remaining
Cropland, Tier 2 Inventory Methodology
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
(36.3)
(80.2)
7.5
-121%
121%
(3.4)
(6.5)
(0.2)
-95%
95%
Land Use, Land-Use Change, and Forestry 6-55
-------
Organic Soil C Stocks: Cropland Remaining „ -, , q
Cropland, Tier 2 Inventory Methodology _ _ _ 2—
Combined Uncertainty for Flux associated
with Agricultural Soil Carbon Stock (9.9) (53.8) 34.3 -452% 452%
Change in Cropland Remaining Cropland
a Range of C stock change estimates predicted by Monte Carlo Stochastic Simulation with a 95 percent confidence interval.
Note: Parentheses indicate net sequestration.
1 Methodological recalculations are applied from 2013 to 2015 using the surrogate data method developed with the C
2 stock change estimates from 1990 to 2012, ensuring consistency across the time series. Details on the emission
3 trends through time are described in more detail in the Methodology section.
4 Uncertainty is also associated with lack of reporting of agricultural woody biomass and dead organic matter C stock
5 changes. The IPCC (2006) does not recommend reporting of annual crop biomass in Cropland Remaining Cropland
6 because all of the biomass senesces each year and so there is no long term storage of C in this pool. For woody
7 plants, biomass C stock changes are likely minor in perennial crops, such as orchards and nut plantations. There will
8 be some removal and replanting of tree crops each year, but the net effect on biomass C stock changes is probably
9 minor because the overall area and tree density is relatively constant across time series. In contrast, agroforestry
10 practices, such as shelterbelts, riparian forests and intercropping with trees, may be significantly changing biomass
11 C stocks over the Inventory time series, at least in some regions of the United States, but there are currently no
12 datasets to evaluate the trends. Changes in litter C stocks are also assumed to be negligible in croplands over annual
13 time frames, although there are certainly significant changes at sub-annual time scales across seasons. However, this
14 trend may change in the future, particularly if crop residue becomes a viable feedstock for bioenergy production.
15 QA/QC and Verification
16 Quality control measures included checking input data, model scripts, and results to ensure data are properly
17 handled throughout the inventory process. Inventory reporting forms and text are reviewed and revised as needed to
18 correct transcription errors. Results from the DAYCENT model are compared to field measurements, and a
19 statistical relationship has been developed to assess uncertainties in the predictive capability of the model. The
20 comparisons include 92 long-term experiments, representing about 908 combinations of management treatments
21 across all of the sites (see Ogle et al. 2007 and Annex 3.12 for more information).
22 Recalculations Discussion
23 Methodological recalculations are associated with extending the time series from 2013 through 2016 using surrogate
24 data method. C stock change estimates decline by an average of 48 percent from 2013 through 2015 based on the
25 recalculation.
26 Planned Improvements
27 New land representation data have not been compiled for the current Inventory, and a surrogate data method has
28 been applied to estimate emissions in the latter part of the time series, which introduces additional uncertainty in the
29 emissions data. Therefore, a key improvement for a future Inventory will be to recalculate the time series for soil C
30 stock changes by applying the Tier 2 and 3 methods with the latest land use data from the National Resources
31 Inventory and related management statistics compiled through the Conservation Effects Assessment Program
3 2 (discussed below).
33 There are several other planned improvements underway. The DAYCENT model will be refined to simulate soil
34 organic C stock changes to a depth of at least 30 cm (currently at 20 cm). Improvements are also underway to more
35 accurately simulate plant production. Crop parameters associated with temperature effects on plant production will
36 be further improved in DAYCENT with additional model calibration. Senescence events following grain filling in
37 crops, such as wheat, are being modified based on recent model algorithm development, and will be incorporated.
38 Experimental study sites will continue to be added for quantifying model structural uncertainty.
39 There is an effort underway to update the time series of management data with information from the USDA-NRCS
40 Conservation Effects Assessment Program (CEAP). This improvement will fill several gaps in the management data
6-56 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
including more specific data on fertilizer rates, updated tillage practices, and more information on planting and
harvesting dates for crops.
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.
In the future, the Inventory will include an analysis of C stock changes in Alaska for cropland and managed
grassland, 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 C stock changes, but will be
further refined over time to incorporate more of the management data that drive C stock changes on long-term
cropland.
Many of these improvements are expected to be completed for the 1990 through 2017 Inventory (i.e., 2019
submission to the UNFCCC). However, the time line may be extended if there are insufficient resources to fund all
or part of these planned improvements.
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 2015), 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) or climate zones (e.g., boreal climates). 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.
Land use change can lead to large losses of 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 according to a recent assessment (Tubiello et al. 2015).
The 2006 IPCC Guidelines recommend reporting changes in biomass, dead organic matter and soil organic carbon
(SOC) stocks with land use change. All SOC 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?'1
Grassland Converted to Cropland is the largest source of emissions from 1990 to 2016, accounting for
approximately 64 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 fact that the area of Grassland Converted to Cropland is significantly larger
than any of the other land use conversions. The majority of the loss is occurring in the mineral soil C pool. The next
largest source of emissions is Forest Land Converted to Cropland, which has relatively large losses of woody
biomass, accounting for approximately 31 percent of the total emissions (Table 6-31 and Table 6-32).
37 Changes in biomass C stocks are not currently reported for other land use conversions (other than forest land) 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 (i.e., other than forest land) to cropland.
Land Use, Land-Use Change, and Forestry 6-57
-------
1 The net change in total C stocks for 2016 led to CO2 emissions to the atmosphere of 23.8 MMT CO2 Eq. (6.5 MMT
2 C), including 2.1 MMT CO2 Eq. (0.6 MMT C) from aboveground biomass C losses, 0.6 MMT CO2 Eq. (0.2 MMT
3 C) from belowground biomass C losses, 0.3 MMT CO2 Eq. (0.1 MMT C) from dead wood C losses, 0.3 MMT CO2
4 Eq. (0.1 MMT C) from litter C losses, 16.9 MMT CO2 Eq. (4.6 MMT C) from mineral soils and 3.4 MMT CO2 Eq.
5 (0.9 MMT C) from drainage and cultivation of organic soils. Emissions in 2016 are 45 percent lower than the
6 emissions in the initial reporting year of 1990, largely due to a reduction in the area of Forest Land Converted to
1 Cropland.
8 Table 6-31: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes in
9 Land Converted to Cropland by Land Use Change Category (MMT CO2 Eq.)
1990
2005
2012
2013
2014
2015
2016
Grassland Converted to Cropland
24.5
17.3
18.1
18.0
17.9
17.8
18.4
Mineral Soils
21.9
13.9
15.1
15.2
15.1
15.0
15.6
Organic Soils
2.5
3.3
3.0
2.9
2.8
2.8
2.8
Forest Land Converted to
Cropland
17.8
7.4
3.6
3.5
3.5
3.5
3.6
Aboveground Live Biomass
11.3
4.9
2.1
2.1
2.1
2.1
2.1
Belowground Live Biomass
3.2
1.3
0.6
0.6
0.6
0.6
0.6
Dead Wood
1.5
0.6
0.3
0.3
0.3
0.3
0.3
Litter
1.5
0.5
0.3
0.3
0.3
0.3
0.3
Mineral Soils
0.2
0.1
0.1
+
+
+
0.1
Organic Soils
0.1
+
+
+
+
+
+
Other Lands Converted to
Cropland
0.3
0.3
0.2
0.1
0.1
0.1
0.1
Mineral Soils
0.2
0.2
0.2
0.1
0.1
0.1
0.1
Organic Soils
0.1
0.1
0.0
0.0
0.0
0.0
0.0
Settlements Converted to
Cropland
0.1
0.1
0.2
0.1
0.1
0.1
0.1
Mineral Soils
0.1
0.1
0.1
+
+
+
+
Organic Soils
+
+
0.1
0.1
0.1
0.1
0.1
Wetlands Converted to Cropland
0.7
0.8
0.7
1.6
1.6
1.7
1.6
Mineral Soils
0.1
0.1
0.1
1.2
1.2
1.2
1.1
Organic Soils
0.6
0.7
0.5
0.4
0.5
0.5
0.5
Aboveground Live Biomass
11.3
4.9
2.1
2.1
2.1
2.1
2.1
Belowground Live Biomass
3.2
1.3
0.6
0.6
0.6
0.6
0.6
Dead Wood
1.5
0.6
0.3
0.3
0.3
0.3
0.3
Litter
1.5
0.5
0.3
0.3
0.3
0.3
0.3
Total Mineral Soil Flux
22.5
14.4
15.6
16.4
16.3
16.3
16.9
Total Organic Soil Flux
3.4
4.2
3.7
3.4
3.4
3.4
3.4
Total Net Flux
43.3
25.9
22.7
23.3
23.2
23.2
23.8
+ Does not exceed 0.05 MMT CO2 Eq.
Notes: Estimates after 2012 for mineral and organic soils are based on a surrogate data method (see Methodology section).
The 2016 estimates of biomass, dead wood and litter are assumed the same as estimates derived for 2015 because new
activity data have not been analyzed for the current Inventory. Totals may not sum due to independent rounding.
10 Table 6-32: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes in
11 Land Converted to Cropland [WW C)
1990
2005
2012
2013
2014
2015
2016
Grassland Converted to Cropland
6.7
4.7
4.9
4.9
4.9
4.9
5.0
Mineral Soils
6.0
3.8
4.1
4.1
4.1
4.1
4.3
Organic Soils
0.7
0.9
0.8
0.8
0.8
0.8
0.8
Forest Land Converted to
Cropland
4.8
2.0
1.0
0.9
1.0
1.0
1.0
Aboveground Live Biomass
3.1
1.3
0.6
0.6
0.6
0.6
0.6
Belowground Live Biomass
0.9
0.4
0.2
0.2
0.2
0.2
0.2
Dead Wood
0.4
0.2
0.1
0.1
0.1
0.1
0.1
Litter
0.4
0.1
0.1
0.1
0.1
0.1
0.1
Mineral Soils
0.1
+
+
+
+
+
+
6-58 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Organic Soils
Other Lands Converted to
+
+
+
+
+
+
+
Cropland
Mineral Soils
0.1
+
O ©
I—1
0.1
0.1
+
+
+
+
+
+
+
+
Organic Soils
+
+
0.0
0.0
0.0
0.0
0.0
Settlements Converted to
Cropland
Mineral Soils
+
+
+
+
0.1
+
+
+
+
+
+
+
+
+
Organic Soils
Wetlands Converted to Cropland
+
0.2
+
0.2
+
0.2
+
0.4
+
0.4
+
0.5
+
0.4
Mineral Soils
+
+
+
0.3
0.3
0.3
0.3
Organic Soils
0.2
0.2
0.1
0.1
0.1
0.1
0.1
Aboveground Live Biomass
3.1
1.3
0.6
0.6
0.6
0.6
0.6
Belowground Live Biomass
0.9
0.4
0.2
0.2
0.2
0.2
0.2
Dead Wood
0.4
0.2
0.1
0.1
0.1
0.1
0.1
Litter
0.4
0.1
0.1
0.1
0.1
0.1
0.1
Total Mineral Soil Flux
6.1
3.9
4.2
4.5
4.5
4.4
4.6
Total Organic Soil Flux
0.9
1.1
1.0
0.9
0.9
0.9
0.9
Total Net Flux
11.8
7.1
6.2
6.4
6.3
6.3
6.5
+ Does not exceed 0.05 MMT C
Notes: Estimates after 2012 for mineral and organic soils are based on a surrogate data method (see
Methodology section). Hie 2016 estimates of biomass, dead wood and litter are assumed the same as estimates
derived for 2015 because new activity data have not been analyzed for the current Inventory. Totals may not
sum due to independent rounding.
Methodology
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 organic soil C stocks.
Biomass, Dead Wood and Litter Carbon Stock Changes
A combination of the Tier 1 and 2 methods is applied to estimate aboveground and belowground biomass, dead
wood, and litter C stock changes for Forest Land Converted to Cropland from 1990 to 2015. For this method, all
annual plots and portions of plots (i.e., conditions; hereafter referred to as plots) from the Forest Inventory and
Analysis (FIA) program are evaluated for land use change in the 48 conterminous United States (i.e., all states
except Alaska and Hawaii) (USDA Forest Service 2015). Specifically, all annual re-measured FIA plots that are
classified as Forest Land Converted to Cropland are identified in each state, and C density estimates before
conversion are compiled for aboveground biomass, belowground biomass, dead wood, and litter. However, there are
exceptions for the Intennountain Region of the Western United States (Arizona, Colorado, Idaho, Montana, New
Mexico, Nevada, and Utah), in which there are a small number of plots that are converted from Forest Land to other
Land Uses. In this region all plots identified as a conversion from forest land to another land use are grouped and
used to estimate the C densities before conversion rather than subdividing the plots into specific land use change
categories. Furthermore, there are no re-measured annual plots in Wyoming, and so the C densities before
conversion are based on data from Colorado, Idaho, Montana, and Utah.
The C density before conversion is estimated for aboveground biomass, belowground biomass, dead wood, and litter
C pools. Soil C stock changes are also addressed, but are based on methods discussed in the next section. Individual
tree aboveground and belowground C density estimates are based on Woodall et al. (2011). The estimates of
aboveground and belowground biomass includes live understory species (i.e., undergrowth plants in a forest)
comprised of woody shrubs and trees less than 2.54 cm in diameter at breast height. It is assumed that 10 percent of
total understory C mass is belowground (Smith et al. 2006). Estimates of C density are derived from information in
Birdsey (1996) and Jenkins et al. (2003). The C density before conversion for standing dead trees 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). Downed dead wood is defined as pieces of dead
Land Use, Land-Use Change, and Forestry 6-59
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
wood greater than 7.5 cm diameter at transect intersections that are not attached to live or standing dead trees, and
includes stumps and roots of harvested trees. The C density before conversion for downed dead wood is estimated
based on measurements of downed dead wood of a subset of FIA plots (Domke et al. 2013; Woodall and Monleon
2008), and models specific to regions and forest types within each region are used to estimate dead wood C
densities. Litter C is the pool of decaying leaves and woody fragments with diameters of up to 7.5 cm that are above
the mineral soil (also known as duff, humus, and fine woody debris). A subset of FIA plots are measured for litter C,
and a modeling approach is used to estimate litter C density based on the measurements (Domke et al. 2016). See
Annex 3.13 for more information about initial C density estimates for Forest Land.
In all states, the initial C in the forest land before conversion to cropland is assumed to be lost to the atmosphere in
the year of the conversion (i.e., 0 tonnes dry matter ha-1 immediately after conversion), which is consistent with the
Tier 1 method in the IPCC guidelines (IPCC 2006). Annual crops (i.e., non-woody crops) are the most common crop
type following conversion, and the default IPCC factor for annual crops is used to estimate the growth following
conversion (IPCC 2006). It is also assumed that the accumulation of dead wood and litter is negligible in the new
cropland. Therefore, total emissions and removals are estimated for biomass based on the new annual crop growth
in cropland minus the losses associated with the C before conversion in the forest land. In contrast, changes in dead
wood and litter C pools are based solely on the loss of the initial dead wood and litter C pools that existed before
conversion of the forest land.
For 2016, C stock changes for biomass, downed wood and dead organic matter are assumed the same as 2015
because new activity data have not been analyzed to determine stock changes in 2016. Future inventories will be
updated with new activity data for 2016, and the time series will be recalculated.
Soil Carbon Stock Changes
SOC stock changes are estimated for Land Converted to Cropland according to land-use histories recorded in the
2012 USDA NRI survey for non-federal lands (USDA-NRCS 2015). 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 2012 (USDA-
NRCS 2015). NRI survey locations are classified as Land Converted to Cropland in a given year between 1990 and
2012 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 (Homer et al. 2007; Fry
etal. 2011; Homer etal. 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 2012
for mineral soils on the majority of land that is used to produce annual crops in the United States. These crops
include alfalfa hay, barley, corn, cotton, dry beans, grass hay, grass-clover hay, lentils, oats, onions, peanuts, peas,
potatoes, rice, sorghum, soybeans, sugar beets, sunflowers, tobacco, tomatoes, and wheat. SOC 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.38
For the years 2013 to 2016, a surrogate data method is used to estimate soil 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 2012 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 (https://quickstats.nass.usda.gov/),
38 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).
6-60 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 and weather data from the PRISM Climate Group (PRISM 2015). See Box 6-6 in the Methodology Section of
2 Cropland Remaining Cropland for more information about the surrogate data method. Stock change estimates for
3 2013 to 2016 will be recalculated in future inventories when new NRI data are available.
4 Tier 3 Approach. For the Tier 3 method, mineral SOC stocks and stock changes are estimated using the DAYCENT
5 biogeochemical model (Parton et al. 1998; Del Grosso et al. 2001, 2011). The DAYCENT model utilizes the soil C
6 modeling framework developed in the Century model (Parton et al. 1987, 1988, 1994; Metherell et al. 1993), but has
7 been refined to simulate dynamics at a daily time-step. National estimates are obtained by using the model to
8 simulate historical land-use change patterns as recorded in the USDA NRI (USDA-NRCS 2015). Carbon stocks and
9 95 percent confidence intervals are estimated for each year between 1990 and 2012. See the Cropland Remaining
10 Cropland section for additional discussion of the Tier 3 methodology for mineral soils.
11 Soil C stock changes from 2013 to 2016 are estimated using the surrogate data method described in Box 6-6 of the
12 Methodology Section in Cropland Remaining Cropland. Future inventories will be updated with new activity data
13 when the data are made available, and the time series will be recalculated (See Planned Improvements section in
14 Cropland Remaining Cropland).
15 Tier 2 Approach. For the mineral soils not included in the Tier 3 analysis, SOC stock changes are estimated using a
16 Tier 2 Approach, as described in the Tier 2 Approach for mineral soils in Cropland Remaining Cropland. This
17 includes application of the surrogate data method that is described in Box 6-6 of the Methodology section in
18 Cropland Remaining Cropland. As with the Tier 3 method, future inventories will be updated with new NRI activity
19 data when the data are made available, and the time series will be recalculated.
20 Organic Soil Carbon Stock Changes
21 Annual C emissions from drained organic soils in Land Converted to Cropland are estimated using the Tier 2
22 method provided in IPCC (2006), with U.S.-specific C loss rates (Ogle et al. 2003) as described in the Cropland
23 Remaining Cropland section for organic soils. This includes application of the surrogate data method that is
24 described in Box 6-6 of the Methodology Section in Cropland Remaining Cropland. Estimates will be recalculated
25 in future inventories when new NRI data are available.
26 Uncertainty and Time-Series Consistency
27 The uncertainty analysis for biomass, dead wood and litter C losses with Forest Land Converted to Cropland is
28 conducted in the same way as the uncertainty assessment for forest ecosystem C flux in the Forest Land Remaining
29 Forest Land category. Sample and model-based error are combined using simple error propagation methods
30 provided by the IPCC (2006) by taking the square root of the sum of the squares of the standard deviations of the
31 uncertain quantities. For additional details see the Uncertainty Analysis in Annex 3.13. The uncertainty analyses for
32 mineral soil C stock changes using the Tier 3 and Tier 2 methodologies are based on a Monte Carlo approach that is
33 described for Cropland Remaining Cropland. The uncertainty for annual C emission estimates from drained organic
34 soils in Land Converted to Cropland is estimated using a Monte Carlo approach, which is also described in the
35 Cropland Remaining Cropland section. For 2013 to 2016, there is additional uncertainty propagated through the
36 Monte Carlo Analysis associated with a surrogate data method, which is also described in Cropland Remaining
37 Cropland.
38 Uncertainty estimates are presented in Table 6-33 for each subsource (i.e., biomass C stocks, dead wood C stocks,
39 litter C stocks, mineral soil C stocks and organic soil C stocks) and the method applied in the Inventory analysis
40 (i.e., Tier 2 and Tier 3). Uncertainty estimates for the total C stock changes for biomass, dead organic matter and
41 soils are combined using the simple error propagation methods provided by the IPCC (2006), as discussed in the
42 previous paragraph. The combined uncertainty for total C stocks in Land Converted to Cropland ranged from 77
43 percent below to 77 percent above the 2016 stock change estimate of 23.8 MMT CO2 Eq. The large relative
44 uncertainty around the 2016 stock change estimate is partly due to variation in soil C stock changes that are not
45 explained by the surrogate data method, leading to high prediction error with this splicing method.
Land Use, Land-Use Change, and Forestry 6-61
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Table 6-33: 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)
2016 Flux Estimate Uncertainty Range Relative to Flux Estimate3
(MMT CO2 Eq.) (MMT CO2 Eq.) (%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Grassland Converted to Cropland
18.4
0.3
36.5
-98%
98%
Mineral Soil C Stocks: Tier 3
14.6
-3.5
32.7
-124%
124%
Mineral Soil C Stocks: Tier 2
1.0
0.3
1.7
-70%
70%
Organic Soil C Stocks: Tier 2
2.8
1.9
3.7
-33%
33%
Forest Land Converted to Cropland
3.6
0.8
6.3
-76%
76%
Aboveground Live Biomass
2.1
-0.5
4.7
-121%
121%
Belowground Live Biomass
0.6
0.1
1.2
-80%
80%
Dead Wood
0.3
0.1
0.6
-76%
76%
Litter
0.3
+
0.7
-99%
99%
Mineral Soil C Stocks: Tier 2
0.1
-0.4
0.5
-764%
765%
Organic Soil C Stocks: Tier 2
+
0.0
0.1
-100%
190%
Other Lands Converted to Cropland
0.1
+
0.1
-97%
97%
Mineral Soil C Stocks: Tier 2
0.1
+
0.1
-97%
97%
Organic Soil C Stocks: Tier 2
0.0
0.0
0.0
0%
0%
Settlements Converted to Cropland
0.1
+
0.1
-53%
53%
Mineral Soil C Stocks: Tier 2
+
+
+
-211%
211%
Organic Soil C Stocks: Tier 2
0.1
+
0.1
-51%
52%
Wetlands Converted to Croplands
1.6
0.7
2.6
-59%
59%
Mineral Soil C Stocks: Tier 2
1.1
0.2
2.0
-80%
80%
Organic Soil C Stocks: Tier 2
0.5
0.2
5.5
-66%
66%
Total: Land Converted to Cropland
23.8
5.4
42.1
-77%
77%
Aboveground Live Biomass
2.1
(0.5)
4.7
-121%
121%
Belowground Live Biomass
0.6
0.1
1.2
-80%
80%
Dead Wood
0.3
0.1
0.6
-76%
76%
Litter
0.3
0.0
0.7
-99%
99%
Mineral Soil C Stocks: Tier 3
14.6
(3.5)
32.7
-124%
124%
Mineral Soil C Stocks: Tier 2
2.3
1.0
3.5
-54%
54%
Organic Soil C Stocks: Tier 2
3.4
2.4
4.4
-29%
29%
+ Does not exceed 0.05 MMT CO2 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.
Methodological recalculations are applied from 2013 to 2015 using the surrogate data method developed using the C
stock change estimates from 1990 to 2012, ensuring consistency across the time series. Details on the emission
trends through time are described in more detail in the Methodology section.
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 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 significant 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 analysis.
QA/QC and Verification
See the QA/QC and Verification section in Cropland Remaining Cropland for information on QA/QC steps.
6-62 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
Recalculations Discussion
2 Methodological recalculations are associated with extending the time series from 2013 through 2015 for mineral and
3 organic soils using a surrogate data method. No other recalculations have been implemented in the current
4 Inventory. C stock change estimates increase by an average of 2 percent from 2013 through 2015 based on the
5 recalculation.
6 Planned Improvements
7 Soil C stock changes with Forest Land Converted to Cropland are undergoing further evaluation to ensure
8 consistency in the time series. Different methods are used to estimate soil C stock changes in forest land and
9 croplands, and while the areas have been reconciled between these land uses, there has been limited evaluation of
10 the consistency in C stock changes with conversion from forest land to cropland. Additional planned improvements
11 are discussed in the Cropland Remaining Cropland section.
12 6.6 Grassland Remaining Grassland (CRF
n Category 4C1)
14 Carbon (C) in grassland ecosystems occurs in biomass, dead organic matter, and soils. Soils are the largest pool of C
15 in grasslands, and have the greatest potential for longer-term storage or release of C. Biomass and dead organic
16 matter C pools are relatively ephemeral compared to the soil C pool, with the exception of C stored in tree and shrub
17 biomass, that occurs in grasslands. The 2006IPCC Guidelines recommend reporting changes in biomass, dead
18 organic matter and soil organic carbon (SOC) stocks with land use and management, but there is currently no
19 reporting of C stock changes for aboveground and belowground biomass, dead wood and litter pools.39 For soil
20 organic C (SOC), the 2006 IPCC Guidelines (IPCC 2006) recommend reporting changes due to (1) agricultural
21 land-use and management activities on mineral soils, and (2) agricultural land-use and management activities on
22 organic soils.40
23 Grassland Remaining Grassland includes all grassland in an Inventory year that had been grassland for a continuous
24 time period of at least 20 years (USDA-NRCS 2015). Grassland includes pasture and rangeland that are primarily,
25 but not exclusively used for livestock grazing. Rangelands are typically extensive areas of native grassland that are
26 not intensively managed, while pastures are typically seeded grassland (possibly following tree removal) that may
27 also have additional management, such as irrigation or interseeding of legumes. The current Inventory includes all
28 privately-owned and federal grasslands in the conterminous United States and Hawaii, but does not include
29 approximately 50 million hectares of Grassland Remaining Grassland in Alaska. This leads to a discrepancy with
30 the total amount of managed area in Grassland Remaining Grassland (see Section 6.1 Representation of the U.S.
31 Land Base) and the grassland area included in the Inventory analysis (CRF Category 4C1—Section 6.6).
32 In Grassland Remaining Grassland, there has been considerable variation in soil C stocks between 1990 and 2016.
33 These changes are driven by variability in weather patterns and associated interaction with land management
34 activity. Moreover, changes are small on a per hectare rate basis across the time series even in the years with a larger
35 total change in stocks. Land use and management generally increased soil C in mineral soils for Grassland
36 Remaining Grassland between 1990 and 2016. In contrast, organic soils lose a relatively constant amount of C
37 annually from 1990 through 2016. In 2016, soil C stocks are a net sink, sequestering 1.6 MMT CO2 Eq. (0.4 MMT
38 C), with an increase of 7.2 MMT CO2 Eq. (2.0 MMT C) in mineral soils, and a loss of 5.5 MMT CO2 Eq. (1.5 MMT
39 C) from organic soils (Table 6-34 and Table 6-35). Soil C stock changes are 62 percent lower in 2016 compared to
40 1990, but stock changes are highly variable from 1990 to 2016, with an average annual sequestration of 5.2 MMT
39 There are planned improvements to address all C pools in the future, with an initial effort focused on biomass C.
40 CO2 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 6-63
-------
1 CChEq. (1.4 MMT C). However, the large inter-annual variability leads to years in which Grassland Remaining
2 Grassland is a net sink and others in which it is a net source of CO2 emissions.
3 Table 6-34: Net CO2 Flux from Soil C Stock Changes in Grassland Remaining Grassland (MMT
4 CO2 Eq.)
Soil Type
1990
2005
2012
2013
2014
2015
2016
Mineral Soils
(11.4)
(0.5)
(26.3)
(9.3)
(13.1)
4.1
(7.2)
Organic Soils
7.2
6.0
5.5
5.5
5.5
5.5
5.5
Total Net Flux
(4.2)
5.5
(20.8)
(3.7)
(7.5)
9.6
(1.6)
Notes: Estimates after 2012 are based on a surrogate data method (see Methodology section).
Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
5 Table 6-35: Net CO2 Flux from Soil C Stock Changes in Grassland Remaining Grassland (MMT
6 C)
Soil Type
1990
2005
2012
2013
2014
2015
2016
Mineral Soils
(3.1)
(0.1)
(7.2)
(2.5)
(3.6)
1.1
(2.0)
Organic Soils
2.0
1
1.5
1.5
1.5
1.5
1.5
Total Net Flux
(1.1)
1.5
(5.7)
(1.0)
(2.1)
2.6
(0.4)
Notes: Estimates after 2012 are based on a data splicing method (see Methodology section).
Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
7 The spatial variability in the 2012 annual soil C stock changes41 associated with mineral soils is displayed in Figure
8 6-7 and organic soils in Figure 6-8. Although relatively small on a per-hectare basis, grassland soils gained C in
9 isolated areas throughout the country, with a larger concentration of grasslands sequestering soil C in Iowa. For
10 organic soils, the regions with the highest rates of emissions coincide with the largest concentrations of organic soils
11 used for managed grassland, including the Southeastern Coastal Region (particularly Florida), upper Midwest and
12 Northeast, and a few isolated areas along the Pacific Coast.
41 Only national-scale emissions are estimated for 2013 to 2016 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 2012.
6-64 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Figure 6-7: Total Net Annual Soil C Stock Changes for Mineral Soils under Agricultural
2 Management within States, 2012, Grassland Remaining Grassland*
4 * Only national-scale soil C stock changes are estimated for 2013 to 2016 in the current Inventory using a
5 surrogate data method, and therefore the fine-scale emission patterns in this map are based on inventory data from
6 2012. Negative values represent a net increase in soil C stocks, and positive values represent a net decrease in soil
7 C stocks.
8 Figure 6-8: Total Net Annual Soil C Stock Changes for Organic Soils under Agricultural
9 Management within States, 2012, Grassland Remaining Grassland*
4
MT C02 ha1 yr
~ < 10
~ 10 to 20
H] 20 to 30
¦ 30 to 40
¦ > 40
10
Land Use, Land-Use Change, and Forestry 6-65
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
* Only national-scale soil carbon stock changes are estimated for 2013 to 2016 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
2012.
Methodology
The following section includes a brief description of the methodology used to estimate changes in soil 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 C stock changes are estimated for Grassland Remaining Grassland on non-federal lands according to land use
histories recorded in the 2012 USDANRI survey (USDA-NRCS 2015). 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 2012 (USDA-NRCS 2015). NRI survey locations are classified as Grassland Remaining
Grassland in a given year between 1990 and 2012 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 prior between
1971 and 1978. For federal lands, the land use history is derived from land cover changes in the National Land
Cover Dataset (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 2012
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) and additional stock changes associated with biosolids (i.e., sewage sludge) amendments. SOC
stock changes on the remaining 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.42
A surrogate data method is used to estimate soil C stock changes from 2013 to 2016 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 2012 emissions data from the Tier 2 and 3 methods. Surrogate data for these regression models
includes weather data from the PRISM Climate Group (PRISM 2015). See Box 6-6 in the Methodology section of
Cropland Remaining Cropland for more information about the surrogate data method. Stock change estimates for
2013 to 2016 will be recalculated in future inventories when new NRI data are available.
Tier 3 Approach. Mineral SOC stocks and stock changes for Grassland Remaining Grassland are estimated using
the DAYCENT biogeochemical43 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
2012 USDA NRI survey (USDA-NRCS 2015). Frequency and rates of manure application to grassland during 1997
are estimated from data compiled by the USDA Natural Resources Conservation Service (NRCS) (Edmonds, et al.
2003), and then adjusted using county-level estimates of manure available for application in other years.
Specifically, county-scale ratios of manure available for application to soils in other years relative to 1997 are used
42 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).
43 Biogeochemical cycles are the flow of chemical elements and compounds between living organisms and the physical
environment.
6-66 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
to adjust the area amended with manure (see Cropland Remaining Cropland section and Annex 3.12 for further
details). Greater availability of managed manure nitrogen (N) relative to 1997 is assumed to increase the area
amended with manure, while reduced availability of manure N relative to 1997 is assumed to reduce the amended
area.
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, and the remainder is
deposited on federal lands (i.e., the amount that is not included in DAYCENT simulations is assumed to be applied
on federal grasslands). Carbon stocks and 95 percent confidence intervals are estimated for each year between 1990
and 2012 using the NRI survey data.
Soil C stock changes from 2013 to 2016 are estimated using a surrogate data method described in Box 6-6 of the
Methodology section in Cropland Remaining Cropland. Future inventories will be updated with new activity data
when the data are made available, and the time series will be recalculated (See 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 land use and management data that are used
in the Inventory for federal grasslands. The NRI (USDA-NRCS 2015) 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) (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 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. As with the non-federal lands, the time series
for federal lands has been extended from 2013 to 2016 using a surrogate data method described in Box 6-6 of the
Methodology Section in Cropland Remaining Cropland. Further elaboration on the Tier 2 methodology and data
used to estimate C stock changes from mineral soils are described in Annex 3.12.
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 2016 to
account for additional C stock changes associated withbiosolid (i.e., sewage sludge) amendments. Estimates of the
amounts of biosolids N applied to agricultural land are derived from national data onbiosolids generation,
disposition, and N content (see Section 7.2, Wastewater Treatment for a detailed discussion of the methodology for
estimating 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. N application rates from Kellogg et al. (2000) are used to determine the amount of area
receiving biosolids amendments. The soil 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).
Organic Soil Carbon Stock Changes
Annual C emissions from drained organic soils in Grassland Remaining Grassland are estimated using the Tier 2
method provided in IPCC (2006), which utilizes U.S.-specific C loss rates (Ogle et al. 2003) rather than default
IPCC rates. A surrogate data method is used to estimate annual C emissions from organic soils from 2013 to 2016 as
described in Box 6-6 of the Methodology section in Cropland Remaining Cropland. Estimates for 2013 to 2016 will
be updated in future inventories when new NRI data are available. For more information, see the Cropland
Remaining Cropland section for organic soils.
Land Use, Land-Use Change, and Forestry 6-67
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Uncertainty and Time-Series Consistency
Uncertainty analysis for mineral soil 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. 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 2013 to 2016, there is
additional uncertainty propagated through the Monte Carlo Analysis associated with the surrogate data method.
Uncertainty estimates are presented in Table 6-36 for each subsource (i.e., mineral soil C stocks and organic soil C
stocks) 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 C stocks in Grassland Remaining Grassland ranges from 2503 percent below to 2503 percent
above the 2016 stock change estimate of -1.6 MMT CO2 Eq. The large relative uncertainty is due to limitations in
the surrogate data model for capturing inter-annual variability in soil C stock changes, particularly in the mineral
soil C pools.
Table 6-36: Approach 2 Quantitative Uncertainty Estimates for C Stock Changes Occurring
Within Grassland Remaining Grassland (MMT CO2 Eq. and Percent)
Source
2016 Flux Estimate Uncertainty Range Relative to Flux Estimate3
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., Sewage Sludge]
Amendments)
Organic Soil C Stocks: Grassland Remaining
Grassland, Tier 2 Methodology
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
(4.2)
(44.8)
36.3
-958%
958%
(1.4)
(2.9)
0.0
-101%
102%
(1.5)
5.5
(2.2)
5.0
(0.7)
6.1
-50%
-9%
50%
9%
Combined Uncertainty for Flux Associated
with Agricultural Soil Carbon Stock
Change in Grassland Remaining Grassland
(1.6)
(42.2)
39.0
-2,503% 2,503%
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.
Methodological recalculations are applied from 2013 to 2015 using the surrogate data method developed using the C
stock change estimates from 1990 to 2012, ensuring consistency across the time series. Details on the emission
trends through time are described in more detail in the Methodology section.
Uncertainty is also associated with a lack of reporting on biomass and litter C stock changes. Biomass C stock
changes may be significant for managed grasslands with woody encroachment despite not having attained enough
tree cover to be considered forest lands. Changes in dead organic matter C stocks are assumed to be negligible in
grasslands on an annual basis, although there are certainly significant changes at sub-annual time scales across
seasons.
QA/QC and Verification
See the QA/QC and Verification section in Cropland Remaining Cropland.
6-68 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Recalculations Discussion
Methodological recalculations are associated with modifying the approach for extending the time series from 2013
through 2015 for mineral and organic soils using a surrogate data method. C stock change estimates declined by an
average of 97 percent from 2013 through 2015 based on the recalculation.
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. Another key planned
improvement is to estimate woody biomass C stock changes for grasslands (See Box 6-8). For information about
other improvements, see the Planned Improvements section in Cropland Remaining Cropland.
Box 6-8: Grassland Woody Biomass Analys
An initial analysis of woodland biomass has been conducted for regions in the western United States. Woodlands are
areas with trees in a matrix of grass vegetation that does not reach the thresholds for tree cover, diameter at breast
height, and/or tree height to be considered forest land. For this pilot effort, carbon stock densities and stock changes
are estimated using woodland plots in the Forest Inventory and Analysis (FIA) database. The full set of woodland
plots cover 12 states in the western United States, and include two FIA forest type groups, pinyon-juniper and
woodland hardwoods. The results suggest that woodlands are sequestering approximately 20 MMT CO2 Eq. in
biomass, dead wood, and litter pools. The analysis will be expanded to the entire time series and reported in a future
Inventory.
Non-C02 Emissions from Grassland Fires (CRF Source Category
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 and improve forage for grazing
livestock. Woody and herbaceous biomass will be oxidized in a fire, although currently the focus is primarily on
herbaceous biomass in this section.44 Biomass burning emits a variety of trace gases including non-CCh greenhouse
gases, CH4 and N20, 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-CCh greenhouse gas emissions
from all wildfires and prescribed burning occurring in managed grasslands.
Biomass burning in grassland of the United States is a relatively small source of emissions, but it has increased by
424 percent since 1990. In 2016, CH4 and N2O emissions from biomass burning in grasslands were 0.3 MMT CO2
Eq. (11 kt) and 0.3 MMT CO2 Eq. (1 kt), respectively. Annual emissions from 1990 to 2016 have averaged
approximately 0.3 MMT CO2 Eq. (12 kt) of CH4 and 0.3 MMT CO2 Eq. (1 kt) of N20 (see Table 6-37 and Table
6-38).
Table 6-37: ChU and N2O Emissions from Biomass Burning in Grassland (MMT CO2 Eq.)
l'WO
2005
2012
2013
2014
2015
2016
ch4
0.1
0.3
0.6
0.2
0.4
0.3
0.3
n2o
0.1
0.3
0.6
0.2
0.4
0.3
0.3
Total Net Flux
0.2
0.7
1.2
0.4
0.8
0.7
0.6
Notes: Estimates for 2015 and 2016 are based on a splicing method described in the
Methodology section.
Totals may not sum due to independent rounding.
44 A planned improvement is underway to incorporate woodland tree biomass into the Inventory.
Land Use, Land-Use Change, and Forestry 6-69
-------
1 Table 6-38: ChU, N2O, CO, and NOx Emissions from Biomass Burning in Grassland (kt)
2
1990
2005
2012
2013
2014
2015
2016
ch4
3
13
23
8
16
13
11
n2o
+
1
2
1
1
1
1
CO
84
358
657
217
442
356
325
NOx
5
21
39
13
27
21
20
+ Does not exceed 0.5 kt
Notes: Estimates for 2015 and 2016 are based on a splicing method described in the
Methodology section.
Totals may not sum due to independent rounding.
3 Methodology
4 The following section includes a description of the methodology used to estimate non-CCh greenhouse gas
5 emissions from biomass burning in grassland, including (1) determination of the land base that is classified as
6 managed grassland; (2) assessment of managed grassland area that is burned each year, and (3) estimation of
7 emissions resulting from the fires. For this Inventory, the IPCC Tier 1 method is applied to estimate non-CCh
8 greenhouse gas emissions from biomass burning in grassland from 1990 to 2014 (IPCC 2006). A data splicing
9 method is used to estimate the emissions in 2015 and 2016, which is discussed later in this section.
10 The land area designated as managed grassland is based primarily on the 2012 National Resources Inventory (NRI)
11 (Nusser and Goebel 1997; USDA 2015). NRI has survey locations across the entire United States, but does not
12 classify land use on federally-owned areas. These survey locations are designated as grassland using land cover data
13 from the National Land Cover Dataset (NLCD) (Fry et al. 2011; Homer et al. 2007; Homer et al. 2015) (see Section
14 6.1 Representation of the U.S. Land Base).
15 The area of biomass burning in grasslands (Grassland Remaining Grassland and Land Converted to Grassland) is
16 determined using 30-m fire data from the Monitoring Trends in Burn Severity (MTBS) program for 1990 through
17 2014.45 NRI survey locations on grasslands are designated as burned in a year if there is a fire within a 500 m of the
18 survey point according to the MTBS fire data. The area of biomass burning is estimated from the NRI spatial
19 weights and aggregated to the country (Table 6-39).
20 Table 6-39: Thousands of Grassland Hectares Burned Annually
Year
Thousand Hectares
1990
317
2005
1.343
2012
2,464
2013
815
2014
1,659
2015
NE
2016
NE
Notes: Burned area are not
estimated (NE) in 2015 and
2016 but will be updated in a
future Inventory.
21 For 1990 to 2014, the total area of grassland burned is multiplied by the IPCC default factor for grassland biomass
22 (4.1 tonnes dry matter per ha) (IPCC 2006) to estimate the amount of combusted biomass. A combustion factor of 1
45 See .
6-70 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
is assumed in this Inventory, and the resulting biomass estimate is multiplied by the IPCC default grassland
emission factors for CH4 (2.3 g CH4 per kg dry matter), N20 (0.21 g CH4 per kg dry matter), CO (65 g CH4 per kg
dry matter) and NOx (3.9 g CH4 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).46
A linear extrapolation of the trend in the time series is applied to estimate the emissions for 2015 and 2016 because
new activity data have not been compiled for the current Inventory. 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 and 2016 emissions. The Tier
1 method described previously will be applied to recalculate the 2015 and 2016 emissions in a future Inventory.
Uncertainty and Time-Series Consistency
Emissions are estimated using a linear regression model with autoregressive moving-average (ARMA) errors for
2015 and 2016. The linear regression ARMA model produced estimates of the upper and lower bounds of the
emission estimate and the results are summarized in Table 6-40. Methane emissions from Biomass Burning in
Grassland for 2016 are estimated to be between 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 2016 emission estimate of 0.3 MMT CO2 Eq.
Nitrous oxide emissions are estimated to be between 0.0 and 0.8 MMT CO2 Eq., or approximately 100 percent
below and 144 percent above the 2016 emission estimate of 0.3 MMT CO2 Eq.
Table 6-40: Uncertainty Estimates for Non-C02 Greenhouse Gas Emissions from Biomass
Burning in Grassland (MMT CO2 Eq. and Percent)
2016 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
Source
Gas
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Grassland Burning
Grassland Burning
CH4
N2O
0.3
0.3
0.0 0.7
0.0 0.8
-100% 145%
-100% 144%
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.
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. Quality control identified problems with cell references in the spreadsheets, which have
been corrected.
Recalculations Discussion
The only recalculation is associated with using the linear regression model with autoregressive moving-average
(ARMA) to estimate emissions in 2015. Non-C02 emissions declined by 20 percent for 2015 based on the
recalculation.
46 See .
Land Use, Land-Use Change, and Forestry 6-71
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
Planned Improvements
A splicing data 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 and recalculate the emissions for 2015 and 2016.
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 lead to more accurate estimates of non-CCh greenhouse gas emissions from grassland
burning.
6.7 Land Converted to Grassland (CRF Category
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 2015).47 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 Section 6.1 Representation of the U.S. Land Base) and the grassland area included in the inventory analysis
(CRF Category 4C2—Section 6.7).
Land use change can lead to large losses of 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 according to a recent assessment (Tubiello et al. 2015).
IPCC (2006) recommends reporting changes in biomass, dead organic matter, and soil organic C (SOC) stocks due
to land use change. All soil 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.48
The largest C losses with Land Converted to Grassland are associated with aboveground biomass, belowground
biomass, dead wood and litter C losses horn Forest Land Converted to Grassland (see Table 6-41 and Table 6-42).
These four pools led to net emissions in 2016 of 20.9, 1.7, 3.6, and 6.2 MMT CO2 Eq. (5.7, 0.5, 1.0, and 1.7 MMT
47 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.
48 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).
6-72 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 C), respectively. Land use and management of mineral soils in Land Converted to Grassland led to an increase in
2 soil C stocks, estimated at 12.0 MMT CO2 Eq. (3.3 MMT C) in 2016, while drainage of organic soils for grassland
3 management led to CO2 emissions to the atmosphere of 1.6 MMT CO2 Eq. (0.4 MMT C). The total net C stock
4 change in 2016 for Land Converted to Grassland is estimated as a loss of 22.0 MMT CO2 Eq. (6.0 MMT C), which
5 is a 23 percent increase in emissions compared to the emissions in the initial reporting year of 1990.
6 Table 6-41: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for
7 Land Converted to Grassland (MMT CO2 Eq.)
1990
2005
2012
2013
2014
2015
2016
Cropland Converted to Grassland
(7.5)
(11.5)
(11.3)
(8.1)
(8.4)
(6.2)
(7.5)
Mineral Soils
(8.0)
(12.7)
(12.4)
(9.3)
(9.5)
(7.4)
(8.6)
Organic Soils
0.5
1.1
1.1
1.1
1.1
1.1
1.1
Forest Land Converted to
Grassland
26.1
32.0
32.3
29.8
29.7
29.5
29.4
Aboveground Live Biomass
18.0
20.2
20.9
20.9
20.9
20.9
20.9
Belowground Live Biomass
0.9
1.9
1.7
1.7
1.7
1.7
1.7
Dead Wood
2.9
3.8
3.6
3.6
3.6
3.6
3.6
Litter
5.1
6.6
6.2
6.2
6.2
6.2
6.2
Mineral Soils
(0.8)
(0.5)
(0.3)
(2.7)
(2.9)
(3.1)
(3.1)
Organic Soils
+
0.1
0.1
0.1
0.1
0.1
0.1
Other Lands Converted Grassland
(0.5)
(1.0)
(0.7)
(+)
(0.1)
(0.1)
(0.1)
Mineral Soils
(0.5)
(1.1)
(0.8)
(0.1)
(0.1)
(0.1)
(0.1)
Organic Soils
+
+
0.1
+
+
+
+
Settlements Converted Grassland
(0.1)
(0.1)
(0.1)
+
+
+
+
Mineral Soils
(0.1)
(0.1)
(0.1)
(+)
(+)
(+)
(+)
Organic Soils
+
+
+
+
+
+
+
Wetlands Converted Grassland
(0.2)
(0.2)
0.2
0.2
0.2
0.1
0.1
Mineral Soils
(0.3)
(0.4)
(0.1)
(0.2)
(0.2)
(0.2)
(0.2)
Organic Soils
0.1
0.3
0.3
0.3
0.3
0.3
0.3
Aboveground Live Biomass
18.0
20.2
20.9
20.9
20.9
20.9
20.9
Belowground Live Biomass
0.9
1.9
1.7
1.7
1.7
1.7
1.7
Dead Wood
2.9
3.8
3.6
3.6
3.6
3.6
3.6
Litter
5.1
6.6
6.2
6.2
6.2
6.2
6.2
Total Mineral Soil Flux
(9.7)
(14.8)
(13.6)
(12.3)
(12.6)
(10.8)
(12.0)
Total Organic Soil Flux
0.7
1.5
1.6
1.7
1.6
1.7
1.6
Total Net Flux
17.9
19.2
20.4
21.9
21.5
23.3
22.0
+ Absolute value does not exceed 0.05 MMT CO2 Eq.
Notes: Estimates after 2012 for mineral and organic soils are based on a surrogate data method (see Methodology
section). The 2016 estimates of biomass, dead wood and litter are assumed the same as estimates derived for 2015
because new activity data have not been analyzed for the current Inventory. Totals may not sum due to independent
rounding. Parentheses indicate net sequestration.
Land Use, Land-Use Change, and Forestry 6-73
-------
1
2
Table 6-42: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for
Land Converted to Grassland(MMT C)
1990 2005 2012 2013 2014 2015 2016
Cropland Converted to Grassland
(2.0)
(3.1)
(3.1)
(2.2)
(2.3)
(1.7)
(2.0)
Mineral Soils
(8.0)
(3.5)
(3.4)
(2.5)
(2.6)
(2.0)
(2.3)
Organic Soils
0.1
0.3
0.3
0.3
0.3
0.3
0.3
Forest Land Converted to
Grassland
7.1
8.7
8.8
8.1
8.1
8.0
8.0
Aboveground Live Biomass
4.9
5.5
5.7
5.7
5.7
5.7
5.7
Belowground Live Biomass
0.2
0.5
0.5
0.5
0.5
0.5
0.5
Dead Wood
0.8
1.0
1.0
1.0
1.0
1.0
1.0
Litter
1.4
1.8
1.7
1.7
1.7
1.7
1.7
Mineral Soils
(0.2)
(0.1)
(0.1)
(0.7)
(0.8)
(0.8)
(0.9)
Organic Soils
+
+
+
+
+
+
+
Other Lands Converted Grassland
(0.1)
(0.3)
(0.2)
(+)
(+)
(+)
(+)
Mineral Soils
(0.1)
(0.3)
(0.2)
(+)
(+)
(+)
(+)
Organic Soils
+
+
+
+
+
+
+
Settlements Converted Grassland
(+)
(+)
(+)
+
+
+
+
Mineral Soils
(+)
(+)
(+)
(+)
(+)
(+)
(+)
Organic Soils
+
+
+
+
+
+
+
Wetlands Converted Grassland
(0.1)
(+)
0.1
+
+
+
+
Mineral Soils
(0.1)
(0.1)
! (+)
(+)
(+)
(0.1)
(0.1)
Organic Soils
+
0.1
0.1
0.1
0.1
0.1
0.1
Aboveground Live Biomass
4.9
5.5
5.7
5.7
5.7
5.7
5.7
Belowground Live Biomass
0.2
0.5
0.5
0.5
0.5
0.5
0.5
Dead Wood
0.8
1.0
1.0
1.0
1.0
1.0
1.0
Litter
1.4
1.8
1.7
1.7
1.7
1.7
1.7
Total Mineral Soil Flux
(2.6)
(4.0)
; (3.7)
(3.3)
(3.4)
(2.9)
(3.3)
Total Organic Soil Flux
0.2
0.4
0.4
0.5
0.4
0.5
0.4
Total Net Flux
4.9
5.2
5.6
6.0
5.9
6.4
6.0
+ Absolute value does not exceed 0.05 MMT C
Notes: Estimates after 2012 for mineral and organic soils are based on a surrogate data method (see
Methodology section). The 2016 estimates of biomass, dead wood and litter are assumed the same as estimates
derived for 2015 because new activity data have not been analyzed for the current Inventory. Totals may not sum
due to independent rounding. Parentheses indicate net sequestration.
3 Methodology
4 The following section includes a description of the methodology used to estimate C stock changes for Land
5 Converted to Grassland, including (1) loss of aboveground and belowground biomass, dead wood and litter C with
6 conversion of Forest Land Converted to Grassland, as well as (2) the impact from all land use conversions to
7 grassland on mineral and organic soil C stocks.
8 Biomass, Dead Wood and Litter Carbon Stock Changes
9 A combination of Tier 1 and 2 methods are applied to estimate aboveground and belowground biomass, dead wood,
10 and litter C stock changes for Forest Land Converted to Grassland from 1990 to 2015. Fortius method, all annual
11 plots and portions of plots (i.e., conditions; hereafter referred to as plots) from the Forest Inventory and Analysis
12 (FIA) program are evaluated for land use change in the 48 conterminous United States (i.e., all states except Alaska
13 and Hawaii) (USDA Forest Service 2015). Specifically, all annual re-measured FIA plots that are classified as
14 Forest Land Converted to Grassland are identified in each state, and C density estimates before conversion are
15 compiled for aboveground biomass, belowground biomass, dead wood, and litter. However, there are exceptions for
16 the Intermountain Region (Arizona, Colorado, Idaho, Montana, New Mexico, Nevada, and Utah) and the Great
17 Plains Region (Kansas, Nebraska, North Dakota and South Dakota) of the United States, in which there are a
18 relatively small number of re-measured plots that are converted from Forest Land to a specific land use. In this
19 region, all plots identified as a conversion from forest land to another land use are grouped in each state and used to
20 estimate the C densities before conversion, rather than subdividing the plots into specific land use change categories
6-74 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
by state. Furthermore, there are no re-measured annual plots in Wyoming, and so the C densities before conversion
are based on data from Colorado, Idaho, Montana, and Utah.
The C density before conversion is estimated for aboveground biomass, belowground biomass, dead wood, and litter
C pools. Soil C stock changes are also addressed, but are based on methods discussed in the next section. Individual
tree aboveground and belowground C density estimates are based on Woodall et al. (2011). The estimates of
aboveground and belowground biomass includes live understory species (i.e., undergrowth plants in a forest)
comprised of woody shrubs and trees less than 2.54 cm in diameter at breast height. It is assumed that 10 percent of
total understory C mass is belowground (Smith et al. 2006). Estimates of C density are obtained from information in
Birdsey (1996) and Jenkins et al. (2003). The C density before conversion for standing dead trees 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). Downed dead wood is defined as pieces of dead
wood greater than 7.5 cm diameter at transect intersections that are not attached to live or standing dead trees, and
includes stumps and roots of harvested trees. The C density before conversion for downed dead wood is estimated
based on measurements of downed dead wood of a subset of FIA plots (Domke et al. 2013; Woodall and Monleon
2008), and models specific to regions and forest types within each region are used to estimate dead wood C
densities. Litter C is the pool of decaying leaves and woody fragments with diameters of up to 7.5 cm that are above
the mineral soil (also known as duff, humus, and fine woody debris). A subset of FIA plots are measured for litter C,
and a modeling approach is used to estimate litter C density based on the measurements (Domke et al. 2016). See
Annex 3.13 for more information about initial C density estimates for Forest Land.
In the Eastern and Central United States, the initial C in the forest land before conversion to grassland is assumed to
be lost to the atmosphere in the year of the conversion (i.e., 0 tonnes dry matter ha1 immediately after conversion),
which is consistent with the Tier 1 method in the IPCC guidelines (IPCC 2006). Grasses and other herbaceous
plants are assumed to dominate these areas following conversion, and the default IPCC factor for grasslands is used
to estimate the growth following conversion (IPCC 2006). It is also assumed that the accumulation of dead wood
and litter is negligible in the new grasslands. Therefore, total emissions and removals are estimated for biomass
based on the new growth in the grassland minus the losses associated with the C before conversion in the forest land.
In contrast, changes in dead wood and litter C pools are based solely on the loss of the initial dead wood and litter C
pools that existed before conversion of the forest land.
In the Western United States (Arizona, California, Colorado, Idaho, Montana, New Mexico, Nevada, Oregon, Utah,
Washington, and Wyoming) and Great Plains Region (Kansas, Nebraska, North Dakota, and South Dakota), there is
evidence in the FIA data as well as the published literature that conversion of forest land to grassland is associated
with persistent woody biomass (Sims et al. 1978; Scholes and Archer 1997; Breshears et al. 2016). Given the
relatively low stocking and tree density on these forest lands, the conversion is likely to equate to a loss of few trees
from the aboveground biomass pool in many cases. However, the loss of the few trees is sufficient to reclassify the
forest land into grassland in a woodland subcategory based on the land use definitions adopted for land
representation in the United States (see Section 6.1, Representation of the U.S. Land Base). Given the evidence from
the published literature and the FIA data, the Tier 1 assumption that all C before conversion is lost with land use
change seems insufficient and would lead to bias in the estimates. A conclusion was drawn from a synthesis of the
literature (Sims et al. 1978; Scholes and Archer 1997; Epstein et al. 2002; Juerna and Archer 2003; Lenihan et al.
2003; Breshears et al. 2016), and an analysis of the FIA data, that approximately 50 percent of the aboveground and
belowground biomass, dead wood, and litter C density is lost during the conversion, while all understory biomass
remains after conversion to woodlands in these regions. Therefore, the total emissions and removals for Forest Land
Converted to Grassland in the Western United States and Great Plains Regions are limited to a loss of 50 percent of
the live biomass and dead organic matter.
For 2016, C stock changes for biomass, downed wood and dead organic matter are assumed the same as 2015
because new activity data have not been analyzed to determine stock changes in 2016. Future inventories will be
updated with new activity data for 2016, and the time series will be recalculated.
Soil Carbon Stock Changes
Soil C stock changes are estimated for Land Converted to Grassland according to land use histories recorded in the
2012 USDA NRI survey for non-federal lands (USDA-NRCS 2015). 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
Land Use, Land-Use Change, and Forestry 6-75
-------
1 available through 2012 (USDA-NRCS 2015). NRI survey locations are classified as Land Converted to Grassland
2 in a given year between 1990 and 2012 if the land use is grassland but had been classified as another use during the
3 previous 20 years. NRI survey locations are classified according to land use histories starting in 1979, and
4 consequently the classifications are based on less than 20 years from 1990 to 1998. This may have led to an
5 underestimation of Land Converted to Grassland in the early part of the time series to the extent that some areas are
6 converted to grassland between 1971 and 1978. For federal lands, the land use history is derived from land cover
7 changes in the National Land Cover Dataset (Homer et al. 2007; Fry et al. 2011; Homer et al. 2015).
8 Mineral Soil Carbon Stock Changes
9 An IPCC Tier 3 model-based approach (Ogle et al. 2010) is applied to estimate C stock changes for Land Converted
10 to Grassland on most mineral soils that are classified in this land use change category. C stock changes on the
11 remaining soils are estimated with an IPCC Tier 2 approach (Ogle et al. 2003), including prior cropland used to
12 produce vegetables, tobacco, and perennial/horticultural crops; land areas with very gravelly, cobbly, or shaley soils
13 (greater than 35 percent by volume); and land converted to grassland from another land use other than cropland.
14 A surrogate data method is used to estimate soil C stock changes from 2013 to 2016 at the national scale for land
15 areas included in the Tier 2 and Tier 3 methods. Specifically, linear regression models with autoregressive moving-
16 average (ARMA) errors (Brockwell and Davis, 2016) are used to estimate the relationship between surrogate data
17 and the 1990 to 2012 emissions data that are derived using the Tier 2 and 3 methods. Surrogate data for these
18 regression models include weather data from the PRISM Climate Group (PRISM 2015). See Box 6-6 in the
19 Methodology Section of Cropland Remaining Cropland for more information about the surrogate data method.
20 Stock change estimates for 2013 to 2016 will be recalculated in future inventories when new NRI data are available.
21 Tier 3 Approach. Mineral SOC stocks and stock changes are estimated using the DAYCENT biogeochemical49
22 model (Parton et al. 1998; Del Grosso et al. 2001, 2011). The DAYCENT model utilizes the soil C modeling
23 framework developed in the Century model (Parton et al. 1987, 1988, 1994; Metherell et al. 1993), but has been
24 refined to simulate dynamics at a daily time-step. Historical land use patterns and irrigation histories are simulated
25 with DAYCENT based on the 2012 USDA NRI survey (USDA-NRCS 2015). C stocks and 95 percent confidence
26 intervals are estimated for each year between 1990 and 2012. See the Cropland Remaining Cropland section and
27 Annex 3.12 for additional discussion of the Tier 3 methodology for mineral soils.
28 Soil C stock changes from 2013 to 2016 are estimated using a surrogate data method described in Box 6-6 of the
29 Methodology section in Cropland Remaining Cropland. Future inventories will be updated with new activity data
30 when the data are made available, and the time series will be recalculated (See Planned Improvements section in
31 Cropland Remaining Cropland).
32 Tier 2 Approach. For the mineral soils not included in the Tier 3 analysis, SOC stock changes are estimated using a
33 Tier 2 Approach, as described in the Tier 2 Approach for mineral soils in Grassland Remaining Grassland. This
34 includes application of the surrogate data method that is described in Box 6-6 of the Methodology Section in
35 Cropland Remaining Cropland. As with the Tier 3 method, future inventories will be updated with new NRI activity
36 data when the data are made available, and the time series will be recalculated.
37 Organic Soil Carbon Stock Changes
38 Annual C emissions from drained organic soils in Land Converted to Grassland are estimated using the Tier 2
39 method provided in IPCC (2006), with U.S.-specific C loss rates (Ogle et al. 2003) as described in the Cropland
40 Remaining Cropland section for organic soils. A surrogate data method is used to estimate annual C emissions from
41 organic soils from 2013 to 2016 as described in Box 6-6 of the Methodology section in Cropland Remaining
42 Cropland. Estimates for 2013to 2016 will be recalculated in future inventories when new NRI data are available.
49 Biogeochemical cycles are the flow of chemical elements and compounds between living organisms and the physical
environment.
6-76 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Uncertainty and Time-Series Consistency
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 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. 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 2013 to 2016, 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 6-43 for each subsource (i.e., biomass C stocks, mineral soil C stocks
and organic soil C stocks) and the method applied in the inventory analysis (i.e., Tier 2 and Tier 3). Uncertainty
estimates from the Tier 2 and 3 approaches are combined using the simple error propagation methods provided by
the IPCC (2006), as discussed in the previous paragraph. The combined uncertainty for total C stocks in Land
Converted to Grassland ranges from 133 percent below to 134 percent above the 2016 stock change estimate of 22.0
MMT CO2 Eq. The large relative uncertainty around the 2016 stock change estimate is partly due to variation in soil
C stock changes that are not explained by the surrogate data method, leading to high prediction error with this
splicing method.
Table 6-43: 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)
2016 Flux Estimate3 Uncertainty Range Relative to Flux Estimate3
(MMT CO2 Eq.) (MMT CO2 Eq.) (%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Cropland Converted to Grassland
(7.5)
(16.3)
1.3
-118%
118%
Mineral Soil C Stocks: Tier 3
(8.6)
(17.4)
(0.3)
-103%
103%
Mineral Soil C Stocks: Tier 2
(0.1)
(0.2)
0.1
-343%
343%
Organic Soil C Stocks: Tier 2
1.1
0.8
1.4
-26%
26%
Forest Land Converted to Grassland
29.4
1.4
57.5
-95%
95%
Aboveground Live Biomass
20.9
(1.9)
43.7
-109%
109%
Belowground Live Biomass
1.7
(10.6)
14.0
-711%
711%
Dead Wood
3.6
(6.5)
13.8
-281%
281%
Litter
6.2
3.1
9.3
-50%
50%
Mineral Soil C Stocks: Tier 2
(3.1)
(5.2)
(1.0)
-68%
68%
Organic Soil C Stocks: Tier 2
0.1
0.1
0.2
-38%
38%
Other Lands Converted to Grassland
(0.1)
(0.2)
0.1
-250%
250%
Mineral Soil C Stocks: Tier 2
(0.1)
(0.3)
0.1
-154%
154%
Organic Soil C Stocks: Tier 2
+
+
0.1
-36%
37%
Settlements Converted to Grassland
+
+
+
-69%
69%
Mineral Soil C Stocks: Tier 2
(+)
(+)
+
-500%
525%
Organic Soil C Stocks: Tier 2
+
+
+
-47%
45%
Wetlands Converted to Grasslands
0.1
(0.1)
0.3
-153%
153%
Mineral Soil C Stocks: Tier 2
(0.2)
(0.3)
(+)
-80%
80%
Organic Soil C Stocks: Tier 2
0.3
0.2
0.4
-38%
38%
Total: Land Converted to Grassland
22.0
(7.4)
51.5
-133%
134%
Aboveground Live Biomass
20.9
(1.9)
43.7
-109%
109%
Belowground Live Biomass
1.7
(10.6)
14.0
-711%
711%
Dead Wood
3.6
(6.5)
13.8
-281%
281%
Litter
6.2
3.1
9.3
-50%
50%
Mineral Soil C Stocks: Tier 3
(8.6)
(17.4)
(0.3)
-103%
103%
Mineral Soil C Stocks: Tier 2
(3.5)
(5.6)
(1.3)
-62%
62%
Land Use, Land-Use Change, and Forestry 6-77
-------
Organic Soil C Stocks: Tier 2 L6 O L9 -20% 20%
+ Absolute value does not exceed 0.05 MMT CO2 Eq.
a Range of C stock change estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values or net sequestration.
1 Methodological recalculations are applied from 2013 to 2015 using the surrogate data method developed using the C
2 stock change estimates from 1990 to 2012, ensuring consistency across the time series. Details on the emission
3 trends through time are described in more detail in the introductory section, above.
4 Uncertainty is also associated with a lack of reporting on biomass and dead organic matter C stock changes for Land
5 Converted to Grassland with the exception of forest land conversion. Biomass C stock changes may be significant
6 for managed grasslands with woody encroachment despite not having attained enough tree cover to be considered
7 forest lands. Changes in dead organic matter C stocks are assumed to be negligible with conversion of land to
8 grasslands with the exception of forest lands, which are included in this analysis. This assumption will be further
9 explored in a future Inventory.
10 QA/QC and Verification
11 See the QA/QC and Verification section in Cropland Remaining Cropland for information on QA/QC steps.
12 Recalculations Discussion
13 Methodological recalculations are associated with extending the time series from 2013 through 2015 for mineral and
14 organic soils using a surrogate data method. No other recalculations have been implemented in this Inventory.
15 Carbon stock change estimates increase by an average of 9 percent from 2013 through 2015 based on the
16 recalculation.
17 Planned Improvements
18 The amount of biomass C that is lost abruptly with Forest Land Converted to Grasslands is estimated based on the
19 amount of C before conversion and an assumed level of C left after conversion based on published literature for the
20 Western United States and Great Plains Regions. The amount of C left after conversion needs further investigation,
21 including tree biomass, understory biomass, dead wood and litter C pools. Moreover, there is currently very limited
22 data collection that would capture the slower change in C (i.e., gains or losses of C) that may be occurring in
23 woodlands following the transfer of C from the previous forest land category. One key improvement is to further
24 investigate the abrupt and more gradual changes in biomass C stock changes that are occurring in different regions,
25 particularly in the Western United States and Great Plains.
26 Soil C stock changes with land use conversion from forest land to grassland are undergoing further evaluation to
27 ensure consistency in the time series. Different methods are used to estimate soil C stock changes in forest land and
28 grasslands, and while the areas have been reconciled between these land uses, there has been limited evaluation of
29 the consistency in C stock changes with conversion from forest land to grassland. In addition, biomass C stock
30 changes will be estimated for Cropland Converted to Grassland, and other land use conversions to grassland, to the
31 extent that data are available. One additional planned improvement for the Land Converted to Grassland category is
32 to develop an inventory of C stock changes for grasslands in Alaska. For information about other improvements, see
33 the Planned Improvements section in Cropland Remaining Cropland and Grassland Remaining Grassland.
6-78 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 6.8 Wetlands Remaining Wetlands (CRF
2 Category 4D1)
3 Wetlands Remaining Wetlands includes all wetland in an Inventory year that had been classified as wetland for the
4 previous 20 years, and in this Inventory includes Peatlands and Coastal Wetlands.
5 Peatlands Remaining Peatlands
6 Emissions from Managed Peatlands
7 Managed peatlands are peatlands that have been cleared and drained for the production of peat. The production
8 cycle of a managed peatland has three phases: land conversion in preparation for peat extraction (e.g., clearing
9 surface biomass, draining), extraction (which results in the emissions reported under Peatlands Remaining
10 Peatlands), and abandonment, restoration, or conversion of the land to another use.
11 Carbon dioxide emissions from the removal of biomass and the decay of drained peat constitute the major
12 greenhouse gas flux from managed peatlands. Managed peatlands may also emit CH4 and N20. The natural
13 production of CH4 is largely reduced but not entirely shut down when peatlands are drained in preparation for peat
14 extraction (Strack et al. 2004 as cited in the 2006IPCC Guidelines). Drained land surface and ditch networks
15 contribute to the CH4 flux in peatlands managed for peat extraction. Methane emissions were considered
16 insignificant under the IPCC Tier 1 methodology (IPCC 2006), but are included in the emissions estimates for
17 Peatlands Remaining Peatlands consistent with the 2013 Supplement to the 2006 IPCC Guidelines for National
18 Greenhouse Gas Inventories: Wetlands (IPCC 2013). Nitrous oxide emissions from managed peatlands depend on
19 site fertility. In addition, abandoned and restored peatlands continue to release greenhouse gas emissions. Although
20 methodologies are provided for rewetted organic soils (which includes rewetted/restored peatlands) in IPCC (2013)
21 guidelines, information on the areal extent of rewetted/restored peatlands in the United States is currently
22 unavailable. The current Inventory estimates CO2, CH4 and N20 emissions from peatlands managed for peat
23 extraction in accordance with IPCC (2006 and 2013) guidelines.
24 CO2, N2O, and CH4 Emissions from Peatlands Remaining Peatlands
25 IPCC (2013) recommends reporting CO2, N20, and CH4 emissions from lands undergoing active peat extraction
26 (i.e., Peatlands Remaining Peatlands) as part of the estimate for emissions from managed wetlands. Peatlands occur
27 where plant biomass has sunk to the bottom of water bodies and water-logged areas and exhausted the oxygen
28 supply below the water surface during the course of decay. Due to these anaerobic conditions, much of the plant
29 matter does not decompose but instead forms layers of peat over decades and centuries. In the United States, peat is
30 extracted for horticulture and landscaping growing media, and for a wide variety of industrial, personal care, and
31 other products. It has not been used for fuel in the United States for many decades. Peat is harvested from two types
32 of peat deposits in the United States: sphagnum bogs in northern states (e.g., Minnesota) and wetlands in states
33 further south (e.g., Florida). The peat from sphagnum bogs in northern states, which is nutrient poor, is generally
34 corrected for acidity and mixed with fertilizer. Production from more southerly states is relatively coarse (i.e.,
35 fibrous) but nutrient rich.
36 IPCC (2006 and 2013) recommend considering both on-site and off-site emissions when estimating CO2 emissions
37 from Peatlands Remaining Peatlands using the Tier 1 approach. Current methodologies estimate only on-site N2O
38 and CH4 emissions, since off-site N20 estimates are complicated by the risk of double-counting emissions from
39 nitrogen fertilizers added to horticultural peat, and off-site CH4 emissions are not relevant given the non-energy uses
40 of peat, so methodologies are not provided in IPCC (2013) guidelines.
41 On-site emissions from managed peatlands occur as the land is cleared of vegetation and the underlying peat is
42 exposed to sun and weather. As this occurs, some peat deposit is lost and CO2 is emitted from the oxidation of the
43 peat. Since N20 emissions from saturated ecosystems tend to be low unless there is an exogenous source of
44 nitrogen, N20 emissions from drained peatlands are dependent on nitrogen mineralization and therefore on soil
Land Use, Land-Use Change, and Forestry 6-79
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
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 N20, 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 2016 (see Table
6-44) comprising 0.7 MMT C02 Eq. (709 kt) of C02, 0.005 MMT C02 Eq. (0.18 kt) of CH4 and 0.001 MMT C02
Eq. (0.002 kt) of N20. Total emissions in 2016 were about 7 percent less than total emissions in 2015.
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 2016. 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 N20 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 between
2008 and 2010, and a general decline between 2011 and 2016. 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 2016.
Table 6-44: Emissions from Peatlands Remaining Peatlands (MMT CO2 Eq.)
Gas
1990
2005
2012
2013
2014
2015
2016
CO2
1.1
1.1
0.8
0.8
0.8
0.8
0.7
Off-site
1.0
1.0
0.8
0.7
0.7
0.7
0.7
On-site
0.1
0.1
0.1
+
0.1
+
0.1
CH4 (On-site)
+
+
+
+
+
+
+
N2O (On-site)
+
+
+
+
+
+
+
Total
1.1
1.1
0.8
0.8
0.8
0.8
0.7
+ Does not exceed 0.05 MMT CO2 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 N2O
emissions are not estimated to avoid double-counting N2O 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 6-45: Emissions from Peatlands Remaining Peatlands (kt)
Gas
1990
2005
2012
2013
2014
2015
2016
CO2
1,055
1,101
812
770
775
763
709
Off-site
985
1,030
760
720
725
713
653
On-site
70
71
53
50
50
49
57
CH4 (On-site)
+
+
+
+
+
+
+
N2O (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 N2O
emissions are not estimated to avoid double-counting N2O 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.
6-80 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Methodology
The following methodology sections first describes the steps taken to calculate emissions estimates for the years
1990 through 2015, followed by the simplified methodology used to update 2016 values.
1990-2015 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 Peat lands were calculated by apportioning the
annual weight of peat produced in the United States (Table 6-46) 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 2015; USGS 2016). 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
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 6-47). 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).50 Peat production was not reported for 2015 i 11.11 a ska's Mineral Industry 2014 report (DGGS
2015); and reliable data are not available beyond 2012, so Alaska's peat production in 2013, 2014, and 2015
(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. CO2 emissions from dissolved organic C were
estimated by multiplying the area of peatlands by the default emissions 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; from 2011 to 2014, imports of
sphagnum moss (nutrient-poor) peat from Canada represented 97 percent of total U.S. peat imports (USGS 2016).
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.
50 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 6-81
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Table 6-46: Peat Production of Lower 48 States (kt)
Type of Deposit 1990
2005
2012 2013 2014 2015 2016
Nutrient-Rich 595.1
Nutrient-Poor 55.4
657.6
27.4
409.9 418.5 416.5 409.4 409.4
78.1 46.5 51.5 50.6 50.6
Total Production 692.0
685.0
488.0 465.0 468.0 460.0 460.0
Sources: United States Geological Survey (USGS) (1991-2015)Minerals Yearbook: Peat (1994-2014);
United States Geological Survey (USGS) (2016) Mineral Commodity Summaries: Peat (2016).
Table 6-47: Peat Production of Alaska (Thousand Cubic Meters)
1990
2005
2012
2013
2014
2015
2016
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).4faifa? 's Mineral Industry Report (1997-2014).
1990-2015 On-site CO 2 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 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).51 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. 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 same assumption that a
single hectare yields 100 metric tons. The IPCC (2006) on-site emissions equation also includes a term which
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).
1990-2015 On-site N2O Emissions
IPCC (2006) suggests basing the calculation of on-site N20 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 N20
emissions, the area of nutrient rich Peatlands Remaining Peatlands was multiplied by the appropriate default
emission factor taken from IPCC (2013).
1990-2015 On-site CH4 Emissions
IPCC (2013) also suggests basing the calculation of on-site CH4 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
51 Hie 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).
6-82 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
Remaining Peatlands was multiplied by the emission factor for direct CH4 emissions taken from IPCC (2013). In
order to estimate CH4 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).
2016 Emissions
A simplified inventory update was performed for the 1990 through 2016 Inventory report using values from the
1990 through 2015 Inventory. Estimates of emissions from peatlands remaining peatlands were forecasted for 2016
and peat production values were set equal to 2015. Excel's FORECAST.ETS function was used to predict a 2016
value using historical data via an algorithm called "Exponential Triple Smoothing." This method smooths out the
data to determine the overall trend and provide an appropriate estimate for 2016.
Uncertainty and Time-Series Consistency
A Monte Carlo (Approach 2) uncertainty analysis that was run on the 1990 through 2015 Inventory was applied to
estimate the uncertainty of CO2, CH4, and N2O emissions from Peatlands Remaining Peatlands for 2016, 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.
• 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 6-48. Carbon dioxide
emissions from Peatlands Remaining Peatlands in 2016 were estimated to be between 0.6 and 0.8 MMT CO2 Eq. at
the 95 percent confidence level. This indicates a range of 16 percent below to 16 percent above the 2016 emission
estimate of 0.7 MMT CO2 Eq. Methane emissions from Peatlands Remaining Peatlands in 2016 were estimated to
be between 0.002 and 0.008 MMT CO2 Eq. This indicates a range of 58 percent below to 78 percent above the 2016
emission estimate of 0.005 MMT CO2 Eq. Nitrous oxide emissions from Peatlands Remaining Peatlands in 2016
were estimated to be between 0.0003 and 0.0011 MMT CO2 Eq. at the 95 percent confidence level. This indicates a
range of 53 percent below to 53 percent above the 2016 emission estimate of 0.0007 MMT CO2 Eq.
Table 6-48: Approach 2 Quantitative Uncertainty Estimates for CO2, Cm, and N2O Emissions
from Peatlands Remaining Peatlands (MMT CO2 Eq. and Percent)
Source
2016 Emission
Gas Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Peatlands Remaining Peatlands CO2 0.7 0.6 0.8 -16% 16%
Land Use, Land-Use Change, and Forestry 6-83
-------
Peatlands Remaining Peatlands CH4 + + + -58% 78%
Peatlands Remaining Peatlands N2O + + + -53% 53%
+ Does not exceed 0.05 MMT CO2 Eq.
Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
1 QA/QC and Verification
2 A QA/QC analysis was performed to review input data and calculations, and no issues were identified. In addition,
3 the emission trends were analyzed to ensure they reflected activity data trends.
4 Recalculations Discussion
5 No recalculations were performed for the 1990 through 2016 Inventory.
6 Planned Improvements
7 In order to further improve estimates of CO2, N2O, and CH4 emissions from Peatlands Remaining Peatlands, future
8 efforts will investigate if improved data sources exist for determining the quantity of peat harvested per hectare and
9 the total area undergoing peat extraction.
10 Efforts will also be made to find a new source for Alaska peat production. The current source has not been reliably
11 updated since 2012 and future publication of these data may discontinue.
12 The 2013 Supplement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands describes
13 inventory methodologies for various wetland source categories. In the 1990 through 2013 Inventory, EPA began
14 including updated methods for Peatlands Remaining Peatlands to align them with the 2013 IPCC Supplement. For
15 future inventories, EPA will determine if additional updates are needed to further address the 2013 IPCC
16 Supplement for Peatlands Remaining Peatlands.
17 The 2006 IPCC Guidelines do not cover all wetland types; they are restricted to peatlands drained and managed for
18 peat extraction, conversion to flooded lands, and some guidance for drained organic soils. They also do not cover all
19 of the significant activities occurring on wetlands (e.g., rewetting of peatlands). Since this inventory only includes
20 Peatlands Remaining Peatlands, additional wetland types and activities found in the 2013 IPCC Supplement will be
21 reviewed to determine if they apply to the United States. For those that do, available data will be investigated to
22 allow for the estimation of greenhouse gas fluxes in future inventory years.
23 Coastal Wetlands Remaining Coastal Wetlands
24 The Inventory recognizes Wetlands as a "land-use that includes land covered or saturated for all or part of the year,
25 in addition to areas of lakes, reservoirs and rivers." Consistent with ecological definitions of wetlands,52 the United
26 States has historically included under the category of Wetlands those coastal shallow water areas of estuaries and
27 bays that lie within the extent of the Land Representation.
28 Additional guidance on quantifying greenhouse gas emissions and removals on Coastal Wetlands is provided in the
29 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands (Wetlands
30 Supplement), which recognizes the particular importance of vascular plants in sequestering CO2 from the
31 atmosphere and building soil carbon stocks. Thus, the Wetlands Supplement provides specific guidance on
32 quantifying emissions on organic and mineral soils that are covered or saturated for part of the year by tidal
33 freshwater, brackish or saline water and are vegetated by vascular plants and may extend seaward to the maximum
3 4 depth of vascular plant vegetation.
52 See .
6-84 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
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 and soils carbon pools is
recognized only in Vegetated Coastal Wetlands and not to 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 Coastal Wetlands.
This Inventory includes all privately-owned and publicly-owned coastal wetlands along the oceanic shores on the
conterminous U.S., 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 CH4 emissions on Vegetated Coastal Wetlands Remaining Vegetated Coastal
Wetlands,
2) Carbon changes on Vegetated Coastal Wetlands Converted to Unvegetated Open Water Coastal Wetlands,
3) Carbon stock changes on Unvegetated Open Water Coastal Wetlands Converted to Vegetated Coastal
Wetlands, and
4) Nitrous Oxide Emissions from Aquaculture in Coastal Wetlands.
Vegetated coastal wetlands hold C in all five C pools (i.e., aboveground, 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 soil C as soils accumulate C under anaerobic soil conditions. Emissions from soil C and
biomass stocks occur when Vegetated Coastal Wetlands are converted to Unvegetated Open Water Coastal
Wetlands (i.e., when managed Vegetated Coastal Wetlands are lost due to subsidence), but are still recognized as
Coastal Wetlands in this Inventory. These C emissions resulting from conversion to Unvegetated Open Water
Coastal Wetlands, can cause the release of many years of accumulated soil C. Conversion of Unvegetated Open
Water Coastal Wetlands to Vegetated Coastal Wetlands initiates the re-building of soil C stocks within soils and
biomass. In applying the 2013IPCC Wetlands Supplement methodologies for CH4 emissions, coastal wetlands in
salinity conditions less than half that of sea water are sources of CH4 as result of slow decomposition of organic
matter under freshwater, anaerobic conditions. Conversion of Vegetated Coastal Wetlands to or from Unvegetated
Open Water Coastal Wetlands do not result in a change in salinity condition and are assumed to have no impact on
CH4 emissions. The 2013 IPCC Wetlands Supplement provides methodologies to estimate N2O emissions on coastal
wetlands that occur due to aquaculture. While N20 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
N20 emissions with the Agricultural Soils Management category. The N20 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
N20.
The Wetlands Supplement provides procedures for estimating CO2 emissions and removals and CH4 emissions from
mangroves, tidal marshes and seagrasses. Depending upon their height and area, emissions and removals from
managed mangroves may be reported under the Forest Land category or under Coastal Wetlands. All non-drained,
intact coastal marshes are intended to be reported under Coastal Wetlands.
Because of human use 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 all estimates of CO2 emissions and
removals, and emissions of CH4, and N20 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 Program53 with NRI 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.
53 See .
Land Use, Land-Use Change, and Forestry 6-85
-------
1 Emissions and Removals from Vegetated Coastal Wetlands
2 Remaining Vegetated Coastal Wetlands
3 The conterminous United States hosts 2.9 million hectares of intertidal Vegetated Coastal Wetlands Remaining
4 Vegetated Coastal Wetlands comprised of tidally influenced palustrine emergent marsh (599,145 ha), palustrine
5 scrub shrub (138,748 ha) and estuarine emergent marsh (1,852,842 ha), estuarine scrub shrub (97,098 ha) and
6 estuarine forest (191,551 ha). Mangroves fall under both estuarine forest and estuarine scrub shrub categories
7 depending upon height. Dwarf mangroves, found in Texas, do not attain the height status to be recognized as Forest
8 Land, and are therefore always classified within Vegetated Coastal Wetlands. Vegetated Coastal Wetlands
9 Remaining Vegetated Coastal Wetlands are found in cold temperate (52,400 ha), warm temperate (892,297 ha),
10 subtropical (1,878,074 ha) and Mediterranean (56,613 ha) climate zones.
11 Soils are the largest pool of C in Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands reflecting
12 long-term removal of atmospheric CO2 by vegetation and transfer into the soil pool in the form of decaying organic
13 matter. Emissions of soil C are not assumed to occur in coastal wetlands that remain vegetated. In this Inventory,
14 only C stock changes within soils are reported as currently insufficient data exists on C stock changes in biomass,
15 DOM and litter. Methane emissions from decomposition of organic matter in anaerobic conditions are significant at
16 salinity less than half that of sea water. Mineral and organic soils are not differentiated in terms of C removals or
17 CH4 emissions.
18 Table 6-49 through Table 6-52 below summarize nationally aggregated soil C stock emissions and removals and
19 CH4 emissions on Vegetated Coastal Wetlands. Intact Vegetated Coastal Wetlands Remaining Vegetated Coastal
20 Wetlands hold a large stock of C (here estimated to be 870 MMT C (3,190 MMT CO2 Eq.)) within the top 1 meter
21 of soil to which C is accumulated each year at a rate of 12.1 MMT CO2 Eq. Methane emissions of 3.6 of MMT CO2
22 Eq. offset C removals resulting in an annual net C removal rate of 8.5 MMT CO2 Eq. Due to federal regulatory
23 protection, loss of Vegetated Coastal Wetland area slowed considerably in the 1970s and the current rates of C stock
24 change and CH4 emissions are relatively constant over time. Losses of Vegetated Coastal Wetlands to Unvegetated
25 Open Water Coastal Wetlands (described later in this chapter) and to other land uses do occur, which because of the
26 depth to which soil C stocks are impacted, do have a significant impact on the net emissions and removals on
27 Coastal Wetlands.
28 Table 6-49: Net CO2 Flux from Soil C Stock Changes in Vegetated Coastal Wetlands
29 Remaining Vegetated Coastal Wetlands (MMT CO2 Eq.)
Year 1990 2005 2012 2013 2014 2015 2016
Net Flux (12.1) (12.2) (12.1) (12.1) (12.1) (12.1) (12.1)
Note: Parentheses indicate net sequestration
30 Table 6-50: Net CO2 Flux from Soil C Stock Changes in Vegetated Coastal Wetlands
31 Remaining Vegetated Coastal Wetlands (MMT C)
Year 1990 2005 2012 2013 2014 2015 2016
Net Flux (13) (3,3) (3.3) (3.3) (3.3) (3.3) (3.3)
Note: Parentheses indicate net sequestration
32 Table 6-51: Net ChU Flux from Vegetated Coastal Wetlands Remaining Vegetated Coastal
33 Wetlands (MMT COz Eq.)
Year 19'JO ; 2005 2012 2013 2014 2015 2016
Net Flux 3.4 ' 3.5 3.5 3.6 3.6 3.6 3.6
34 Table 6-52: Net ChU Flux from Vegetated Coastal Wetlands Remaining Vegetated Coastal
35 Wetlands {kt CH4)
Year 1990 2005 2012 2013 2014 2015 2016
Net Flux 138 / 140 , 142 142 142 143 143
6-86 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
Methodology
The following section includes a description of the methodology used to estimate changes in soil C stocks and
emissions of CH4 for Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands.
Soil Carbon Stock Changes
Soil C removals are estimated for Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands for both
mineral and organic soils on wetlands 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, 2005 and 2010 NOAA C-CAP
surveys.54 Federal and non-federal lands are represented. Trends in land cover change are extrapolated to 1990 and
2016 from these datasets. Based upon NOAA C-CAP, coastal wetlands are subdivided into freshwater (Palustrine)
and saline (Estuarine) classes and further subdivided into emergent marsh, scrub shrub and forest classes.55 Soil C
stock changes, stratified by climate zones and wetland classes, are derived from a synthesis of peer-reviewed
literature (Mangrove pool and removals data: Cahoon & Lynch unpublished data; Lynch 1989; Callaway et al. 1997;
Chen & Twilley 1999; McKee & Faulkner 2000; Ross et al. 2000; Chmura et al. 2003; Perry & Mendelssohn 2009;
Castaneda-Moya et al. 2013; Henry & Twilley 2013; Doughty et al. 2015; Marchio et al. 2016. Tidal marsh pool and
removals data: Anisfeld unpublished data; Cahoon unpublished data; Cahoon & Lynch unpublished data; Chmura
unpublished data; McCaffrey & Thomson 1980; Hatton 1981; Callaway et al. 1987; Craft et al. 1988; Cahoon &
Turner 1989; Patrick & DeLaune 1990; Kearney & Stevenson 1991;Cahoon et al. 1996; Callaway et al. 1997;
Roman et al. 1997; Bryant & Chabrek 1998; Orson et al. 1998; Markewich et al. 1998; Anisfeld et al. 1999; Connor
et al. 2001; Choi & Wang 2001; Chmura et al. 2003, Hussein et al. 2004; Craft 2007; Miller et al. 2008; Drexler et
al. 2009; Perry & Mendelssohn 2009; Loomis & Craft 2010; EPA's NWCA 2011; Callaway et al. 2012; Henry &
Twilley 2013; Weston et al. 2014). To estimate soil C stock changes no differentiation is made between organic and
mineral soils.
Tier 2 level estimates of soil C removal associated with annual soil C accumulation from 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. Emission factors were
developed from literature references that provided soil C removal factors disaggregated by climate region,
vegetation type by salinity range (estuarine or palustrine) as identified using NOAA C-CAP as described above.
Quantification of regional coastal wetland above and belowground biomass C stock changes for woody and
perennial herbaceous vegetation, DOM [dead wood and litter] C stocks are in development and are not presented
this year, though will be included in future reports.
Soil Methane Emissions
Tier 1 estimates of CH4 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 CH4 emission factors provided in Table 4.14 of the Wetlands Supplement.
The methodology follows Eq. 4.9, Chapter 4 of the Wetlands Supplement, and is applied to the area of Vegetated
Coastal Wetlands Remaining Vegetated Coastal Wetlands on an annual basis. The AR4 global warming potential
factor of 25 was used in converting CH4 to CO2 Eq. values.
Uncertainty and Time-Series Consistency
Underlying uncertainties in estimates of soil C stock changes and CH4 include error in uncertainties associated with
Tier 2 literature values of soil C stocks and CH4 flux, assumptions that underlie the methodological approaches
applied and uncertainties linked to interpretation of remote sensing data. Uncertainty specific to Vegetated Coastal
54 See .
55 See .
Land Use, Land-Use Change, and Forestry 6-87
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Wetlands Remaining Vegetated Coastal Wetlands include differentiation of palustrine and estuarine community
classes, which determines the soil C stock and CH4 flux applied. Soil C stocks and CH4 fluxes applied are
determined from vegetation community classes across the coastal zone and identified by NOAA C-CAP.
Community classes are further subcategorized by climate zones and growth form (forest, shrub-scrub, marsh).
Uncertainties for soil C stock data for all subcategories are not available and thus assumptions were applied using
expert judgement about the most appropriate assignment of a soil C stock to a disaggregation of a community class.
Because mean soil 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; 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). Uncertainties for CH4 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 (±10-15 percent; IPCC 2003). However, there is significant uncertainty in salinity
ranges for tidal and non-tidal estuarine wetlands and activity data used to apply CH4 flux emission factors
(delineation of an 18 ppt boundary) will need significant improvement to reduce uncertainties.
Table 6-53: Approach 1 Quantitative Uncertainty Estimates for Emissions from C Stock
Changes occurring within Vegetated Coastal Wetlands Remaining Vegetated Coastal
Wetlands [WW CO2 Eq. and Percent)
2016 Flux Estimate
Uncertainty Range Relative to Flux Estimate
Source
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Combined Uncertainty for Flux Associated
with Wetlands Soil C Stock Change in
Vegetated Coastal Wetlands Remaining
Vegetated Coastal Wetlands
(12.1)
(15.6)
(8.5)
-29.5%
29.5%
Note: Parentheses indicate net sequestration.
Table 6-54: Approach 1 Quantitative Uncertainty Estimates for ChU Emissions occurring
within Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands (MMT CO2 Eq.
and Percent)
2016 Flux Estimate
Uncertainty Range Relative to Flux Estimate
Source
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Combined Uncertainty for Flux Associated
with CH4 emissions in Vegetated Coastal
Wetlands Remaining Vegetated Coastal
Wetlands
3.6
2.5
4.6
-29.8%
29.8%
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.
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 CH4 emission factors derived from the IPCC Wetlands Supplement.
6-88 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Planned Improvements
2 A USGS/NASA Carbon Monitoring System investigation is in progress to establish a U.S. country-specific database
3 of soil C stock, wetland biomass and CH4 emissions for coastal wetlands. Refined error analysis combining land
4 cover change and C stock estimates will be provided. Through this work a model is in development to represent
5 changes in soil C stocks. This research effort is due to be completed by November 2017, with plans to include the
6 results from the new model in the 1990 through 2017 Inventory (i.e., 2019 submission to the UNFCCC).
7 The C-CAP dataset for 2015 is currently under development. Once complete, land use change for 2011 through
8 2016 will be recalculated with this updated dataset.
9 With the conclusion of the Blue Carbon Monitoring System Project it is intended that the next (i.e., 1990 through
10 2017) Inventory will include new data on estuarine emergent biomass C stocks and refined reference soil C stocks
11 and uncertainty analysis based upon an expanded national dataset.
12 Emissions from Vegetated Coastal Wetlands Converted to
13 Unvegetated Open Water Coastal Wetlands
14 Conversion of intact Vegetated Coastal Wetlands to Unvegetated Open Water Coastal Wetlands is a source of
15 emissions from both soil and biomass C stocks. It is estimated that 8,428 ha of Vegetated Coastal Wetlands were
16 converted to Unvegetated Open Water Coastal Wetlands in 2016. The Mississippi Delta represents more than 40
17 percent of the total coastal wetland of the U.S., and over 90 percent of the conversion of Vegetated Coastal
18 Wetlands to Unvegetated Open Water Coastal Wetlands. The drivers of coastal wetlands loss include legacy human
19 impacts on sediment supply through rerouting river flow, direct impacts of channel cutting on hydrology, salinity
20 and sediment delivery and accelerated subsidence from aquafer extraction. Each of these drivers directly contributes
21 to wetland erosion and subsidence, while also reducing the resilience of the wetland to build with sea level rise or
22 recover from hurricane disturbance. Over recent decades the rate of Mississippi Delta wetland loss has slowed,
23 though episodic mobilization of sediment occurs during hurricane events (Couvillion et al. 2011; Couvillion et al.
24 2016). The most recent land cover analysis recorded by the C-CAP surveys of 2005 and 2010 coincides with two
25 such events, hurricanes Katrina and Rita both in 2005.
26 Shallow nearshore open water within the U.S. Land Representation is recognized as falling under the Wetlands
27 category within the U.S. Inventory. Changes in biomass are not presented this year but will be in the future (see
28 Planned Improvements). While high resolution mapping of coastal wetlands provides data to support Tier 2
29 approaches for tracking land cover change, the depth to which sediment is lost is less clear. This Inventory adopts
30 the Tier 1 methodological guidance from the Wetlands Supplement for estimating emissions following the
31 methodology for excavation (see Methodology section, below) when Vegetated Coastal Wetlands are converted to
32 Unvegetated Open Water Coastal Wetlands, assuming aim depth of disturbed soil. This 1 m depth of disturbance is
33 consistent with estimates of wetland C loss provided in the literature (Crooks et al. 2009; Couvillion et al. 2011;
34 Delaune and White 2012; IPCC 2013). A Tier 1 assumption is also adopted that all mobilized C is immediately
35 returned to the atmosphere (as assumed for terrestrial land use categories), rather than redeposited in long-term C
36 storage. The science is currently under evaluation to adopt more refined emissions factors for mobilized coastal
37 wetland C based upon the geomorphic setting of the depositional environment.
38 Table 6-55: Net CO2 Flux from Soil C Stock Changes in Vegetated Coastal Wetlands
3 9 Con verted to Un vegetated Open Water Coastal Wetlands (MMT CO2 Eq.)
Year 1WI 2005 2012 2013 2014 2015 2016
Net Soil Flux 3.5 ; 2.1 ,-j 3.5 3.5 3.5 3.5 3.5
40 Table 6-56: Net CO2 Flux from Soil C Stock Changes in Vegetated Coastal Wetlands
41 Con verted to Un vegetated Open Water Coastal Wetlands (M MT C)
Year 1990 ; 2005 2012 2013 2014 2015 2016
Net Soil Flux 1.0 0.6 1.0 1.0 1.0 1.0 1.0
Land Use, Land-Use Change, and Forestry 6-89
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
Methodology
The following section includes a brief description of the methodology used to estimate changes in soil C stocks for
Vegetated Coastal Wetlands Converted to Unvegetated Open Water Coastal Wetlands.
Soil Carbon Stock Changes
Soil C stock changes 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, 2005 and 2010 NOAA C-CAP surveys. Publicly-owned and privately-owned
lands are represented. Trends in land cover change are extrapolated to 1990 and 2016 from these datasets. C-CAP
provides peer reviewed country-specific mapping to support IPCC Approach 3 quantification of coastal wetland
distribution, including conversion to and from open water. Country-specific soil C stocks for mineral and organic
soils, stratified by climate zones and wetland classes, are derived from a synthesis of peer-reviewed literature
(Mangrove pool and removals data: Cahoon & Lynch unpublished data; Lynch 1989; Callaway et al. 1997; Chen &
Twilley 1999; McKee & Faulkner 2000; Ross et al. 2000; Chmura et al. 2003; Perry & Mendelssohn 2009;
Castaneda-Moya et al. 2013; Henry & Twilley 2013; Doughty et al. 2015; Marchio et al. 2016. Tidal marsh pool and
removals data: Anisfeld unpublished data; Cahoon unpublished data; Cahoon & Lynch unpublished data; Chmura
unpublished data; McCaffrey & Thomson 1980; Hatton 1981; Callaway et al. 1987; Craft et al. 1988; Cahoon &
Turner 1989; Patrick & DeLaune 1990; Kearney & Stevenson 1991;Cahoon et al. 1996; Callaway et al. 1997;
Roman et al. 1997; Bryant & Chabrek 1998; Orson et al. 1998; Markewich et al. 1998; Anisfeld et al. 1999; Connor
et al. 2001; Choi & Wang 2001; Chmura et al. 2003, Hussein et al. 2004; Craft 2007; Miller et al. 2008; Drexler et
al. 2009; Perry & Mendelssohn 2009; Loomis & Craft 2010; EPA's NWCA 2011; Callaway et al. 2012; Henry &
Twilley 2013; Weston et al. 2014). For soil C stock change no differentiation is made between organic and mineral
soils. 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 with all emissions occurring in the year of wetland
conversion, and multiplied by activity data of land area for management coastal wetlands. The methodology follows
Eq. 4.6. Quantification of regional coastal wetland biomass stock changes for conversion of Vegetated Coastal
Wetlands to Unvegetated Open Water Coastal Wetlands are in development and are not presented this year, though
will be included in future reports.
Soil Methane Emissions
A Tier 1 assumption has been applied that salinity conditions are unchanged and hence methane emissions are
assumed to be zero with conversion of Vegetated Coastal Wetlands to Unvegetated Open Water Coastal Wetlands.
Uncertainty and Time-Series Consistency
Underlying uncertainties in estimates of soil C stock changes associated with Tier 2 literature values of soil C
stocks, assumptions that underlie the methodological approaches applied and uncertainties linked to interpretation of
remote sensing data are included in this uncertainty assessment. Uncertainty specific to coastal wetlands include
differentiation of palustrine and estuarine community classes, which determines the soil C stock applied. Soil C
stocks applied are determined from vegetation community classes across the coastal zone and identified by NOAA
C-CAP. Community classes are further subcategorized by climate zones and growth form (forest, shrub-scrub,
marsh). Soil C stock data for all subcategories are not available and thus assumptions were applied using expert
judgement about the most appropriate assignment of a soil C stock to a disaggregation of a community class.
Because mean soil 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; 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). Uncertainties for CH4 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 (±10-15 percent; IPCC 2003).
6-90 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Table 6-57: Approach 1 Quantitative Uncertainty Estimates for Net CO2 Flux Occurring
2 within Vegetated Coastal Wetlands Converted to Unvegetated Open Water Coastal Wetlands
3 (MMT CO2 Eq. and Percent)
2016 Flux Estimate
Uncertainty Range Relative to Flux Estimate
Source
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Combined Uncertainty for Flux Associated
with Soil C Stock Change in Vegetated
Coastal Wetlands Converted to Unvegetated
3.5
2.0
5.0
-41.7%
41.7%
Open Water Coastal Wetlands
4 The C-CAP dataset, consisting of a time series of four time intervals, each five years in length, and two major
5 hurricanes striking the Mississippi Delta in the most recent time interval (2006 to 2010), creates a challenge in
6 utilizing it to represent the annual rate of wetland loss and for extrapolation to 1990 and 2016. Uncertainty in the
7 defining the long term trend will be improved with release of the 2015 survey, expected in 2018 to 2019.
8 More detailed research is in development that provides a longer term assessment and more highly refined rates of
9 wetlands loss across the Mississippi Delta (e.g., Couvillion et al. 2016), which could provide a more refined regional
10 Approach 2-3 for assessing wetland loss and support the national scale assessment provided by C-CAP.
11 Based upon the IPCC Tier 1 methodological guidance for estimating emissions with excavation in coastal wetlands,
12 it has been assumed that a 1-meter column of soil has been remobilized with erosion and the C released immediately
13 to the atmosphere as CO2. This depth of disturbance is a simplifying assumption that is commonly applied in the
14 scientific literature to gain a first order estimate of scale of emissions (e.g., Delaune and White 2012). It is also a
15 simplifying assumption that all that C is released back to the atmosphere immediately and future development of the
16 Tier 2 estimate may refine the emissions both in terms of scale and rate. Given that erosion has been ongoing for
17 multiple decades the assumption that the C eroded is released to the atmosphere the year of erosion is a reasonable
18 simplification that could be further refined.
19 QA/QC and Verification
20 Data provided by NOAA (i.e., National LiDAR Dataset, NOS Tide Data, and C-CAP land cover and land cover
21 change mapping) undergo internal agency QA/QC procedures. Acceptance of final datasets into archive and
22 dissemination are contingent upon assurance that the data product is compliant with mandatory NOAA QA/QC
23 requirements (McCombs et al. 2016). QA/QC and Verification of the soil C stock dataset has been provided by the
24 Smithsonian Environmental Research Center and by the Coastal Wetlands project team leads who reviewed the
25 estimates against primary scientific literature. Land cover estimates were assessed to ensure that the total land area
26 did not change over the time series in which the inventory was developed, and were verified by a second QA team.
27 A team of two evaluated and verified there were no computational errors within the calculation worksheets. Two
28 biogeochemists at the USGS, in addition to members of the NASA Carbon Monitoring System Science Team,
29 corroborated the assumption that where salinities are unchanged CH4 emissions are constant with conversion of
30 Unvegetated Open Water Coastal Wetlands to Vegetated Coastal Wetlands.
31 Planned Improvements
32 A refined uncertainty analysis and efforts to improve times series consistency is planned for the 1990 through 2017
33 Inventory (i.e., 2019 submission to the UNFCCC). An approach for calculating the fraction of remobilized coastal
34 wetland soil C returned to the atmosphere as CO2 is currently under review and may be included in future reports.
35 Research by USGS is investigating higher resolution mapping approaches to quantify conversion of coastal wetlands
36 is also underway. Such approaches may form the basis of an Approach 3 land representation assessment in future
37 years.
38 The C-CAP dataset for 2015 is currently under development. Once complete, land use change for 2011 through
39 2016 will be recalculated with this updated dataset.
Land Use, Land-Use Change, and Forestry 6-91
-------
1 With the conclusion of the Blue Carbon Monitoring System Project it is intended that the 1990 through 2017
2 Inventory report will include new data on estuarine emergent biomass C stocks and refined reference soil C stocks
3 and uncertainty analysis based upon an expanded national dataset.
4 Removals from Unvegetated Open Water Coastal Wetlands
5 Converted to Vegetated Coastal Wetlands
6 Open Water within the U.S. land base, as described in the Land Representation, is recognized as Wetlands within
7 the Inventory. The appearance of vegetated tidal wetlands on lands previously recognized as open water reflects
8 either the building of new vegetated marsh through sediment accumulation or the transition from other lands uses
9 through an intermediary open water stage as flooding intolerant plants are displaced and then replaced by wetland
10 plants. Biomass and soil C accumulation on Unvegetated Open Water Coastal Wetlands Converted to Vegetated
11 Coastal Wetlands begins with vegetation establishment.
12 Within the U. S., conversion of Unvegetated Open Water Coastal Wetlands to Vegetated Coastal Wetlands is
13 predominantly due to engineered activities, which include active restoration of wetlands (e.g., wetlands restoration
14 in San Francisco Bay), dam removals or other means to reconnect sediment supply to the nearshore (e.g.,
15 Atchafalaya Delta, Louisiana, Couvillion et al., 2011). Wetlands restoration projects have been ongoing in the U.S.
16 since the 1970s. Early projects were small, a few hectares in size. By the 1990s, restoration projects, each hundreds
17 of hectares in size, were becoming common in major estuaries. In a number of coastal areas e.g., San Francisco Bay,
18 Puget Sound, Mississippi Delta and south Florida, restoration activities are in planning and implementation phases,
19 each with the goal of recovering tens of thousands of hectares of wetlands.
20 During wetland restoration, Unvegetated Open Water Coastal Wetland is a common intermediary phase bridging
21 land use transitions from Cropland or Grassland to Vegetated Coastal Wetlands. The time period of open water may
22 last from five to 20 years depending upon the conditions. The conversion of these other land uses to Unvegetated
23 Open Water Coastal Wetland will result in reestablishment of wetland biomass and soil C sequestration and may
24 result in cessation of emissions from drained organic soil. Only changes in soil C stocks are reported in the
25 Inventory at this time, but improvements are being evaluated to include changes from other C pools.
26 Table 6-58: Net CO2 Flux from Soil C Stock Changes from Unvegetated Open Water Coastal
27 Wetlands Con verted to Vegetated Coastal Wetlands (MMT CO2 Eq.)
Year
1990
2005
2012
2013
2014
2015
2016
Net Soil Flux
(0.01)
+
(0.01)
(0.01)
(0.01)
(0.01)
(0.01)
+ Does not exceed 0.005 MMT CO2 Eq.
Note: Parentheses indicate net sequestration.
28 Table 6-59: Net CO2 Flux from Soil C Stock Changes from Unvegetated Open Water Coastal
29 Wetlands Converted to Vegetated Coastal Wetlands (MMT C)
Year
1990
2005
2012
2013
2014
2015
2016
Net Soil Flux
(0.002)
' (0.001) /
(0.002)
(0.002)
(0.002)
(0.002)
(0.002)
Note: Parentheses indicate net sequestration.
30 Methodology
31 The following section includes a brief description of the methodology used to estimate changes in soil C stocks and
32 CH4 emissions for Unvegetated Open Water Coastal Wetlands Converted to Vegetated Coastal Wetlands.
33 Soil Carbon Stock Change
34 Soil C removals are estimated for Unvegetated Open Water Coastal Wetlands Converted to Vegetated Coastal
35 Wetlands on lands below the elevation of high tides (taken to be mean high water spring tide elevation) according to
36 the national LiDAR dataset, the national network of tide gauges and land use histories recorded in the 1996, 2001,
6-92 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
2005 and 2010 NOAA C-CAP surveys. Privately-owned, and publically-owned lands are represented. Trends in
land cover change are extrapolated to 1990 and 2016 from these datasets. C-CAP provides peer reviewed Tier 2
level mapping of coastal wetland distribution, including conversion to and from open water. Country-specific soil C
stock change associated with soil C accretion, stratified by climate zones and wetland classes, are derived from a
synthesis of peer-reviewed literature (Mangrove pool and removals data: Cahoon & Lynch unpublished data; Lynch
1989; Callaway et al. 1997; Chen & Twilley 1999; McKee & Faulkner 2000; Ross et al. 2000; Chmura et al. 2003;
Perry & Mendelssohn 2009; Castaneda-Moya et al. 2013; Henry & Twilley 2013; Doughty et al. 2015; Marchio et
al. 2016. Tidal marsh pool and removals data: Anisfeld unpublished data; Cahoon unpublished data; Cahoon &
Lynch unpublished data; Chmura unpublished data; McCaffrey & Thomson 1980; Hatton 1981; Callaway et al.
1987; Craft et al. 1988; Cahoon & Turner 1989; Patrick & DeLaune 1990; Kearney & Stevenson 1991;Cahoon et al.
1996; Callaway et al. 1997; Roman et al. 1997; Bryant & Chabrek 1998; Orson et al. 1998; Markewich et al. 1998;
Anisfeld et al. 1999; Connor et al. 2001; Choi & Wang 2001; Chmura et al. 2003, Hussein et al. 2004; Craft 2007;
Miller et al. 2008; Drexler et al. 2009; Perry & Mendelssohn 2009; Loomis & Craft 2010; EPA's NWCA 2011;
Callaway et al. 2012; Henry & Twilley 2013; Weston et al. 2014). Soil C removals 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).
Tier 2 level estimates of CO2 removals associated with annual soil C accumulation in managed Vegetated Coastal
Wetlands were developed using country-specific soil C removal factors multiplied by activity data on land area for
managed coastal wetlands. The methodology follows Eq. 4.7, Chapter 4 of the Wetlands Supplement, and is applied
to the area of managed Vegetated Coastal Wetlands on an annual basis. Emission 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. Quantification of
regional coastal wetland biomass C stock changes for perennial vegetation are in development and are not presented
this year, though will be included in future reports.
Soil Methane Emissions
A Tier 1 assumption has been applied that salinity conditions are unchanged and hence methane emissions are
assumed to be zero with conversion of Vegetated Open Water Coastal Wetlands to Vegetated Coastal Wetlands.
Uncertainty and Time-Series Consistency
Underlying uncertainties in estimates of soil C stock changes and methane emissions include error in uncertainties
associated with Tier 2 literature values of soil C stocks and methane flux 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 which determines
the soil C stock and methane flux applied. Soil C stocks and methane fluxes applied are determined from vegetation
community classes across the coastal zone and identified by NOAA C-CAP. Community classes are further
subcategorized by climate zones and growth form (forest, shrub-scrub, marsh). Soil C stock data for all
subcategories are not available and thus assumptions were applied using expert judgement about the most
appropriate assignment of a soil C stock to a disaggregation of a community class. Because mean soil C stocks for
each available community class are in a fairly narrow range, the same overall uncertainty was applied 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; 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). Uncertainties for CH4 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). Uncertainties for methane flux include the Tier 1 default values reported
in the Wetlands Supplement along with the overall uncertainty of the NOAA C-CAP remote sensing product, which
is estimated at 15 percent. This is in the typical range of remote sensing methods (±10 to 15; GPG LULUCF,
Chapter 3). However, there is significant uncertainty in salinity ranges for tidal and non-tidal estuarine wetlands and
activity data used to develop the methane flux (delineation of an 18 ppt boundary) will need significant
improvement to reduce uncertainties.
Land Use, Land-Use Change, and Forestry 6-93
-------
1 Table 6-60: Approach 1 Quantitative Uncertainty Estimates for C Stock Changes Occurring
2 within Unvegetated Open Water Coastal Wetlands Converted to Vegetated Coastal Wetlands
3 (MMT CO2 Eq. and Percent)
2016 Flux Estimate
Uncertainty Range Relative to Flux Estimate
Source
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Combined Uncertainty for Flux Associated
with Wetlands Soil C Stock Change in
Unvegetated Open Water Coastal Wetlands
Converted to Vegetated Coastal Wetlands
(0.007)
(0.009)
(0.005)
-29.5%
29.5%
Note: Parentheses indicate net sequestration.
4 QA/QC and Verification
5 NOAA provided data (i.e., National LiDAR Dataset, NOS Tide Data, and C-CAP land cover and land cover change
6 mapping) undergo internal agency QA/QC assessment procedures. Acceptance of final datasets into the archive for
7 dissemination are contingent upon assurance that the product is compliant with mandatory NOAA QA/QC
8 requirements (McCombs et al. 2016). QA/QC and Verification of soil C stock dataset has been provided by the
9 Smithsonian Environmental Research Center and Coastal Wetlands project team leads who reviewed produced
10 summary tables against primary scientific literature. Land cover estimates were assessed to ensure that the total land
11 area did not change over the time series in which the inventory was developed, and verified by a second QA team. A
12 team of two evaluated and verified there were no computational errors within calculation worksheets. Two
13 biogeochemists at the USGS, also members of the NASA Carbon Monitoring System Science Team, corroborated
14 the simplifying assumption that where salinities are unchanged CH4 emissions are constant with conversion of
15 Unvegetated Open Water Coastal Wetlands to Vegetated Coastal Wetlands.
16 Planned Improvements
17 A USGS/NASA Carbon Monitoring System investigation is in progress to establish a U.S. country-specific database
18 of published measurement data quantifying soil C stock, wetland biomass and CH4 emissions. Refined error analysis
19 combining land cover change and soil and biomass C stock estimates will be provided. Under this investigation a
20 model is in development to represent changes in soil C stocks. This investigation is to be completed by November
21 2017 and will be included in the 1990 through 2017 Inventory.
22 The C-CAP dataset for 2015 is currently under development. Once complete, land use change for 2011 through
23 2016 will be recalculated with this updated dataset.
24 With the conclusion of the Blue Carbon Monitoring System Project it is intended that the 1990 through 2017
25 Inventory report will include new data on estuarine emergent biomass C stocks and refined reference soil C stocks
26 and uncertainty analysis based upon an expanded national dataset.
27 N20 Emissions from Aquaculture in Coastal Wetlands
28 Shrimp and fish cultivation in coastal areas increases nitrogen loads resulting in direct emissions of N20. Nitrous
29 oxide is generated and emitted as a byproduct of the conversion of ammonia (contained in fish urea) to nitrate
30 through nitrification and nitrate to N2 gas through denitrification (Hu et al. 2012). Nitrous oxide emissions can be
31 readily estimated from data on fish production (IPCC 2013 Wetlands Supplement).
32 Overall, aquaculture production in the United States has fluctuated slightly from year to year though it is essentially
33 at a similar level since 2011 as in 1990. Data for 2016 are not yet available and emissions have been held constant
34 with 2014 at 0.14 MMT C02Eq.
6-94 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
Table 6-61: Net N2O Emissions from Aquaculture in Coastal Wetlands (MMT CO2 Eq.)
Year 1990 2005 2012 2013 2014 2015 2016
Flux 0.13 -j 0.18 ,.j 0.14 0.14 0.14 0.14 0.14
2 Table 6-62: Net N2O Emissions from Aquaculture in Coastal Wetlands (kt N2O)
Year 1990 2005 2012 2013 2014 2015 2016
Flux 0.44 | 0.59 /• 0.46 0.48 0.47 0.47 0.47
3 Methodology
4 The methodology to estimate N20 emissions from Aquaculture in Coastal Wetlands follows guidance in the 2013
5 IPCC Wetlands Supplement by applying country-specific fisheries production data and the IPCC Tier 1 default
6 emission factor.
7 Each year NO AA Fisheries document the status of U.S. marine fisheries in the annual report of Fisheries of the
8 United States, from which activity data for this analysis is derived.56 The fisheries report has been produced in
9 various forms for more than 100 years, primarily at the national level, on U.S. recreational catch and commercial
10 fisheries landings and values. In addition, data are reported on U.S. aquaculture production, the U.S. seafood
11 processing industry, imports and exports of fish-related products, and domestic supply and per capita consumption
12 of fisheries products. Within the aquaculture chapter mass of production for Catfish, Striped bass, Tilapia, Trout,
13 Crawfish, Salmon and Shrimp are reported. While some of these fisheries are produced on land and some in open
14 water cages, all have data on the quantity of food stock produced, which is the activity data that is applied to the
15 IPCC Tier 1 default emissions factor to estimate emissions of N20 from aquaculture. It is not apparent from the data
16 as to the amount of aquaculture occurring above the extent of high tides on river floodplains. While some
17 aquaculture likely occurs on coastal lowland floodplains this is likely a minor component of tidal aquaculture
18 production because of the need for a regular source of water for pond flushing. The estimation of N20 emissions
19 from aquaculture is not sensitive to salinity using IPCC approaches and as such the location of aquaculture ponds on
20 the landscape does not influence the calculations.
21 Other open water shellfisheries for which no food stock is provided, and thus no additional N inputs, are not
22 applicable for estimating N20 emissions (e.g., Clams, Mussels and Oysters) and have not been included in the
23 analysis. The IPCC Tier 1 default emissions factor of 0.00169 kg N20-N per kg of fish produced (95 percent
24 confidence interval - 0,0038) is applied to the activity data to calculate total N20 emissions. The AR4 global
25 warming potential value of 298 is applied in deriving C02 Eq. values from N20 emissions.
26 Uncertainty and Time-Series Consistency
27 Uncertainty estimates are based upon the Tier 1 default 95 percent confidence interval provided within the Wetlands
28 Supplement for N20 emissions. Uncertainties in N20 emissions from aquaculture are based on expert judgement for
29 the NOAA Fisheries of the United States fisheries production data (± 100 percent) multiplied by default uncertainty
30 level for N20 emissions found in Table 4.15, chapter 4 of the Wetlands Supplement. Given the overestimate of
31 fisheries production from coastal wetland areas due to the inclusion of fish production in non-coastal wetland areas,
32 this is a reasonable initial first approximation for an uncertainty range.
56 See
Land Use, Land-Use Change, and Forestry 6-95
-------
1 Table 6-63: Approach 1 Quantitative Uncertainty Estimates for N2O Emissions for
2 Aquaculture Production in Coastal Wetlands (MMT CO2 Eq. and Percent)
2016 Flux Estimate
Uncertainty Range Relative to Emissions Estimate3
Source
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Bound
Upper
Bound
Lower Upper
Bound Bound
Combined Uncertainty for Flux Associated
with N2O Emissions for Aquaculture
Production in Coastal Wetlands
0.14
0.00
0.30
-116% 116%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
3 QA/QC and Verification
4 NOAA provided internal QA/QC review of reported fisheries data. The Coastal Wetlands Inventory team consulted
5 with the Coordinating Lead Authors of the Coastal Wetlands chapter of the 2013IPCC Wetlands Supplement to
6 assess which fisheries production data to include in estimating emissions from aquaculture. It was concluded that
7 N20 emissions estimates should be applied to any fish production to which food supplement is supplied be they
8 pond or open water and that salinity conditions were not a determining factor in production of N20 emissions.
9 6.9 Land Converted to Wetlands (CRF Category
10 4D2)
11 Emissions and Removals from Land Converted to Vegetated
12 Coastal Wetlands
13 Land Converted to Vegetated Coastal Wetlands occurs as a result of inundation of unprotected low-lying coastal
14 areas with gradual sea level rise, flooding of previously drained land behind hydrological barriers, and through
15 active restoration and creation of coastal wetlands through removal of hydrological barriers. All other land
16 categories (i.e., Forest Land, Cropland, Grassland, Settlements and Other Lands) are identified has having some area
17 converting to Vegetated Coastal Wetlands. Between 1990 and 2016 the rate of annual transition for Land Converted
18 to Vegetated Coastal Wetlands ranged from 2,619 ha/year to 5,316 ha/year. Conversion rates were higher during the
19 period 2010 through 2016 than during the earlier part of the time series.
20 However, at the present stage of Inventory development, Coastal Wetlands are not explicitly shown in the Land
21 Representation analysis while work continues harmonizing data from NOAA's Coastal Change Analysis Program57
22 with NRI data used to compile the Land Representation. As a QC step a check was undertaken to confirm that
23 Coastal Wetlands recognized by C-CAP represented a subset of Wetlands recognized by the NRI for marine coastal
24 states. Delineating Vegetated Coastal Wetlands from ephemerally flooded upland Grasslands represents a particular
25 challenge in remote sensing. Moreover, at the boundary between wetlands and uplands, which may be gradual on
26 low lying coastlines, the presence of wetlands may be ephemeral depending upon weather and climate cycles and as
27 such results in the emissions and removals vary over these time frames.
28 Following conversion to Vegetated Coastal Wetlands there are increases in biomass and soil C storage. Additionally,
29 at salinities less than half that of seawater the transition from upland dry soils to wetland soils results in CH4
30 emissions. In this Inventory analysis, soil C stock changes and CH4 emissions are quantified. Estimates of biomass C
31 stock changes will be included in subsequent reports. Estimates of emissions and removals are based on emission
57 See .
6-96 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 factor data that have been applied to estimate changes in soil C stock for Land Converted to Vegetated Coastal
2 Wetlands.
3 Table 6-64: Net CO2 Flux from Soil C Stock Changes in Land Converted to Vegetated Coastal
4 Wetlands (MMT COz Eq.)
Year 1990 21105 2012 2013 2014 2015 2016
Net Soil Flux (0.02) ; (0.01) ¦ (0.02) (0.02) (0.02) (0.02) (0.02)
Note: Parentheses indicate net sequestration.
5 Table 6-65: Net CO2 Flux from Soil C Stock Changes in Land Converted to Vegetated Coastal
6 Wetlands (MM! C)
Year 1990 2005 2012 2013 2014 2015 2016
Net Soil Flux (0.01) / (+) .< (0.01) (0.01) (0.01) (0.01) (0.01)
+ Does not exceed 0.005 MMT C.
Note: Parentheses indicate net sequestration.
7 Table 6-66: Net CH4 Flux in Land Converted to Vegetated Coastal Wetlands (MMT CO2 Eq.)
Soil Type 1990 j 2005 2012 2013 2014 2015 2016
Net Flux 0.01 ' 0.01 0.01 0.01 0.01 0.01 0.01
8 Table 6-67: Net ChU Flux from Soil C Stock Changes in Land Converted to Vegetated Coastal
9 Wetlands (kt CH4)
Soil Type 1990 ; 2005 2012 2013 2014 2015 2016
Net Flux 0.57 0.48 0.48 0.48 0.48 0.48 0.48
10 Methodology
11 The following section includes a brief description of the methodology used to estimate changes in soil C removals
12 and CH4 emissions for Land Converted to Vegetated Coastal Wetlands.
13 Soil Carbon Stock Changes
14 Soil C removals are estimated for Land Converted to Vegetated Coastal Wetlands for land below the elevation of
15 high tides (taken to be mean high water spring tide elevation) and as far seawards as the extent of intertidal vascular
16 plants within the U.S. Land Representation according to the national LiDAR dataset, the national network of tide
17 gauges and land use histories recorded in the 1996, 2001, 2005 and 2010 NOAA C-CAP surveys.58 As noted above,
18 the NOAA C-CAP dataset has yet to be harmonized with the NRI dataset from which the Land Representation is
19 derived. Federal and non-federal lands are represented. Trends in land cover change are extrapolated to 1990 and
20 2016 from these datasets. Based upon NOAA C-CAP, wetlands are subdivided into freshwater (Palustrine) and
21 saline (Estuarine) classes and further subdivided into Emergent marsh, scrub shrub and forest classes. Soil C stock
22 changes, stratified by climate zones and wetland classes, are derived from a synthesis of peer-reviewed literature
23 (Mangrove pool and removals data: Cahoon & Lynch unpublished data; Lynch 1989; Callaway et al. 1997; Chen &
24 Twilley 1999; McKee & Faulkner 2000; Ross et al. 2000; Chmura et al. 2003; Perry & Mendelssohn 2009;
25 Castaneda-Moya et al. 2013; Henry & Twilley 2013; Doughty et al. 2015; Marchio et al. 2016. Tidal marsh pool and
26 removals data: Anisfeld unpublished data; Cahoon unpublished data; Cahoon & Lynch unpublished data; Chmura
27 unpublished data; McCaffrey & Thomson 1980; Hatton 1981; Callaway et al. 1987; Craft et al. 1988; Cahoon &
28 Turner 1989; Patrick & DeLaune 1990; Kearney & Stevenson 1991;Cahoon et al. 1996; Callaway et al. 1997;
29 Roman et al. 1997; Bryant & Chabrek 1998; Orson et al. 1998; Markewich et al. 1998; Anisfeld et al. 1999; Connor
58 See .
Land Use, Land-Use Change, and Forestry 6-97
-------
1 et al. 2001; Choi & Wang 2001; Chmura et al. 2003, Hussein et al. 2004; Craft 2007; Miller et al. 2008; Drexler et
2 al. 2009; Perry & Mendelssohn 2009; Loomis & Craft 2010; EPA's NWCA 2011; Callaway et al. 2012; Henry &
3 Twilley 2013; Weston et al. 2014). To estimate soil C stock changes no differentiation is made for soil type (i.e.,
4 mineral, organic).
5 Tier 2 level estimates of soil C removal associated with annual soil C accumulation from Land Converted to
6 Vegetated Coastal Wetlands were developed using country-specific soil C removal factors multiplied by activity
7 data of land area for Land Converted to Vegetated Coastal Wetlands. The methodology follows Eq. 4.7, Chapter 4
8 of the LPCC Wetlands Supplement, and applied to the area of Land Converted to Vegetated Coastal Wetlands on an
9 annual basis. Emission factors were developed from literature references that provided soil C removal factors
10 disaggregated by climate region, vegetation type by salinity range (estuarine or palustrine) as identified using
11 NOAA C-CAP as described above. Quantification of regional coastal wetland biomass C stock changes for
12 perennial vegetation are in development and are not presented this year, though will be included in future reports.
13 Soil Methane Emissions
14 Tier 1 estimates of CH4 emissions for Vegetated Coastal Wetlands Remaining Vegetated Coastal Wetlands are
15 derived from the same wetland map used in the analysis of wetland soil C fluxes, produced from C-CAP, LiDAR
16 and tidal data, in combination with default CH4 emission factors provided in Table 4.14 of the LPCC Wetlands
17 Supplement. The methodology follows Eq. 4.9, Chapter 4 of the LPCC Wetlands Supplement, and is applied to the
18 total area of Land Converted to Vegetated Coastal Wetlands on an annual basis. The AR4 global warming potential
19 factor of 25 was used in converting CH4 to CO2 Eq. values.
20 Uncertainty and Time-Series Consistency
21 Underlying uncertainties in estimates of soil C removal factors and CH4 include error in uncertainties associated
22 with Tier 2 literature values of soil C removal estimates and CH4 flux, assumptions that underlie the methodological
23 approaches applied and uncertainties linked to interpretation of remote sensing data.
24 Uncertainty specific to coastal wetlands include differentiation of palustrine and estuarine community classes which
25 determines the soil C removal and CH4 flux applied. Soil C removal and CH4 fluxes applied are determined from
26 vegetation community classes across the coastal zone and identified by NOAA C-CAP. Community classes are
27 further subcategorized by climate zones and growth form (forest, shrub-scrub, marsh). Soil C removal data for all
28 subcategories are not available and thus assumptions were applied using expert judgement about the most
29 appropriate assignment of a soil C removal factor to a disaggregation of a community class. Because mean soil C
30 removal for each available community class are in a fairly narrow range, the same overall uncertainty was assigned
31 to each, (i.e., applying approach for asymmetrical errors, the largest uncertainty for any soil C stock value should be
32 applied in the calculation of error propagation; IPCC 2000). Uncertainties for CH4 flux are the Tier 1 default values
33 reported in the LPCC Wetlands Supplement. Overall uncertainty of the NOAA C-CAP remote sensing product is 15
34 percent. This is in the range of remote sensing methods (±10-15 percent; IPCC 2003). However, there is significant
35 uncertainty in salinity ranges for tidal and non-tidal estuarine wetlands and activity data used to estimate the CH4
36 flux (e.g., delineation of an 18 ppt boundary), which will need significant improvement to reduce uncertainties.
37 Table 6-68: Approach 1 Quantitative Uncertainty Estimates for Net CO2 Flux Changes
38 occurring within Land Converted to Vegetated Coastal Wetlands (MMT CO2 Eq. and Percent)
2016 Flux Estimate
Uncertainty Range Relative to Flux Estimate3
Source
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Combined Uncertainty for Flux Associated
with Land Converted to Vegetated Coastal
Wetlands
(0.02)
(0.03)
(0.02)
-29.5%
29.5%
a Range of flux estimates based on error propagation at 95 percent confidence interval.
6-98 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Table 6-69: Approach 1 Quantitative Uncertainty Estimates for ChU Emissions occurring
2 within Land Converted to Vegetated Coastal Wetlands (MMT CO2 Eq. and Percent)
2016 Flux Estimate
Uncertainty Range Relative to Flux Estimate3
Source
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Combined Uncertainty for Flux Associated
with Land Converted to Vegetated Coastal
Wetlands
0.01
0.01
0.02
-29.8%
29.8%
a Range of flux estimates based on error propagation at 95 percent confidence interval.
3 QA/QC and Verification
4 NOAA provided National LiDAR Dataset, tide data, and C-CAP land cover and land cover change mapping, all of
5 which are subject to agency internal mandatory QA/QC assessment (McCombs et al. 2016). QA/QC and verification
6 of soil C stock dataset has been provided by the Smithsonian Environmental Research Center and Coastal Wetland
7 Inventory team leads. Land cover estimates were assessed to ensure that the total land area did not change over the
8 time series in which the inventory was developed, and verified by a second QA team. A team of two evaluated and
9 verified there were no computational errors within the calculation worksheets. Soil C stock, emissions/removals data
10 where based upon peer-reviewed literature and CH4 emission factors derived from the IPCC Wetlands Supplement.
11 Planned Improvements
12 A USGS/NASA Carbon Monitoring System investigation is in progress to establish a U.S. country-specific database
13 of soil C stocks, wetland bio mass and CH4 emissions. Refined error analysis combining land cover change and C
14 stock estimates will be provided. Under this investigation, a model is in development to represent changes in soil C
15 stocks. This investigation is due to be completed by November 2017. Future improvements will thus include
16 estimates of estuarine emergent biomass C stocks, refined soil C stocks and uncertainty analysis.
17 The C-CAP dataset for 2015 is currently under development. Once complete, land use change for 2011 through
18 2016 will be recalculated with this updated dataset.
19 With the conclusion of the Blue Carbon Monitoring System Project it is intended that the 1990 through 2017
20 Inventory report will include new data on estuarine emergent biomass C stocks and refined reference soil C stocks
21 and uncertainty analysis based upon an expanded national dataset.
22 6.10 Settlements Remaining Settlements
23 (CRF Category 4E1)
24 Soil Carbon Stock Changes (CRF Category 4E1J
25 Drainage of organic soils is common when wetland areas have been developed for settlements. Organic soils, also
26 referred to as Histosols, include all soils with more than 12 to 20 percent organic C by weight, depending on clay
27 content (NRCS 1999, Brady and Weil 1999). The organic layer of these soils can be very deep (i.e., several meters),
28 and form under inundated conditions that results in minimal decomposition of plant residues. Drainage of organic
29 soils leads to aeration of the soil that accelerates decomposition rate and CO2 emissions.59 Due to the depth and
30 richness of the organic layers, C loss from drained organic soils can continue over long periods of time, which varies
59 N2O emissions from soils are included in the N2O Emissions from Settlement Soils section.
Land Use, Land-Use Change, and Forestry 6-99
-------
1 depending on climate and composition (i.e., decomposability) of the organic matter (Armentano and Menges 1986).
2 The United States does not estimate changes in soil organic C stocks for mineral soils on Settlements Remaining
3 Settlements, which is consistent with the assumption of the Tier 1 method in the IPCC guidelines (2006). This
4 assumption may be evaluated in the future if funding and resources are available to conduct an analysis of soil C
5 stock changes in mineral soils of Settlements Remaining Settlements.
6 Settlements Remaining Settlements includes all areas that have been settlements for a continuous time period of at
7 least 20 years according to the 2012 United States Department of Agriculture (USD A) National Resources Inventory
8 (NRI) (USDA-NRCS 2015)60 or according to the National Land Cover Dataset for federal lands (Homer et al. 2007;
9 Fry et al. 2011; Homer et al. 2015). The Inventory includes settlements on privately-owned lands in the
10 conterminous United States and Hawaii. Alaska and the small amount of settlements on federal lands are not
11 included in this Inventory even though these areas are part of the U.S. managed land base. This leads to a
12 discrepancy with the total amount of managed area in Settlements Remaining Settlements (see Section 6.1
13 Representation of the U.S. Land Base) and the settlements area included in the Inventory analysis. There is a
14 planned improvement to include settlements on organic soils in these areas as part of a future Inventory.
15 CO2 emissions from drained organic soils in settlements are 1.3 MMT CO2 Eq. (0.4 MMT C) in 2016. Although the
16 flux is relatively small, the amount has increased by over 800 percent since 1990.
17 Table 6-70: Net CO2 Flux from Soil C Stock Changes in Settlements Remaining Settlements
18 (MMT COz Eq.)
Soil Type 1990 2005 2012 2013 2014 2015 2016
Organic Soils 0.1 0.5 1.3 1.3 1.3 1.3 1.3
Note: Estimates after 2012 are based on a data splicing method (see Methodology section).
19 Table 6-71: Net CO2 Flux from Soil C Stock Changes in Settlements Remaining Settlements
20 (MMT C)
Soil Type 1990 2005 2012 2013 2014 2015 2016
Organic Soils + ;;»l' 0.1 0.4 0.4 0.4 04 0.4
+ Does not exceed 0.05 MMT C
Note: Estimates after 2012 are based on a data splicing method (see Methodology section).
21 Methodology
22 An IPCC Tier 2 method is used to estimate soil organic C stock changes for organic soils in Settlements Remaining
23 Settlements (IPCC 2006). Organic soils in Settlements Remaining Settlements are assumed to be losing C at a rate
24 similar to croplands due to deep drainage, and therefore emission rates are based on country-specific values for
25 cropland (Ogle et al. 2003). The following section includes a description of the methodology, including (1)
26 determination of the land base that is classified as settlements; and (2) estimation of emissions from drained organic
27 soils.
28 The land area designated as settlements is based primarily on the 2012 National Resources Inventory (NRI) (USDA
29 2015) with additional information from the National Land Cover Dataset (NLCD) (Fry et al. 2011; Homer et al.
30 2007; Homer et al. 2015). It is assumed that all settlement area on organic soils is drained, and those areas are
31 provided in Table 6-72 (See Section 6.1, Representation of the U.S. Land Base for more information). The area of
32 drained organic soils is estimated from the NRI spatial weights and aggregated to the country (Table 6-72). The area
33 of land on organic soils in Settlements Remaining Settlements has increased from 3 thousand hectares in 1990 to
34 over 28 thousand hectares in 2012. The area of land on organic soils are not available from NRI for Settlements
3 5 Remaining Settlements after 2012.
60 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.
6-100 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Table 6-72: Thousands of Hectares of Drained Organic Soils in Settlements Remaining
2 Settlements
Area
Year
(Thousand Hectares)
1990
3
2005
10
2012
28
2013
ND
2014
ND
2015
ND
2016
ND
Note: No NRI data are available
after 2012.
ND (No data).
3 To estimate CO2 emissions from drained organic soils across the time series from 1990 to 2012, the total area of
4 organic soils in Settlements Remaining Settlements is multiplied by the country-specific emission factors for
5 Cropland Remaining Cropland under the assumption that there is deep drainage of the soils. The emission factors
6 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
7 ha in subtropical regions (see Annex 3.12 for more information).
8 A linear extrapolation of the trend in the time series is applied to estimate the emissions from 2013 to 2016 because
9 NRI activity data are not available for these years to determine the area of drained organic soils in Settlements
10 Remaining Settlements. Specifically, a linear regression model with autoregressive moving-average (ARMA) errors
11 (Brockwell and Davis, 2016) is used to estimate the trend in emissions over time from 1990 to 2012, and in turn, the
12 trend is used to approximate the 2013 to 2016 emissions. The Tier 2 method described previously will be applied in
13 future inventories to recalculate the estimates beyond 2012 as activity data becomes available.
14 Uncertainty and Time-Series Consistency
15 Uncertainty for the Tier 2 approach is derived using a Monte Carlo approach, along with additional uncertainty
16 propagated through the Monte Carlo Analysis for 2013 to 2016 based on the linear time series model. The results of
17 the Approach 2 Monte Carlo uncertainty analysis are summarized in Table 6-73. Soil C losses from drained organic
18 soils in Settlements Remaining Settlements for 2016 are estimated to be between 0.8 and 1.8 MMT CO2 Eq. at a 95
19 percent confidence level. This indicates a range of 35 percent below and 35 percent above the 2016 emission
20 estimate of 1.3 MMT CO2 Eq.
21 Table 6-73: Uncertainty Estimates for CO2 Emissions from Drained Organic Soils in
22 Settlements Remaining Settlements (MMT CO2 Eq. and Percent)
2016 Emission
Source
Gas
Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Organic Soils
CO2
1.3
00
OO
0
-35% 35%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence
interval.
23 Methodological recalculations are applied from 2013 to 2015 using the linear time series model described above.
24 Details on the emission trends through time are described in more detail in the Methodology section, above.
Land Use, Land-Use Change, and Forestry 6-101
-------
1 QA/QC and Verification
2 Quality control measures included checking input data, model scripts, and results to ensure data are properly
3 handled throughout the inventory process. Inventory reporting forms and text are reviewed and revised as needed to
4 correct transcription errors.
5 Recalculations Discussion
6 Methodological recalculations are associated with extending the time series from 2013 through 2016 using a linear
7 time series model. The recalculation had a minor effect on the time series overall with C losses from drainage of
8 organic soils increasing by less than 1 percent on average.
9 Planned Improvements
10 This source will be extended to include CO2 emissions from drainage of organic soils in settlements of Alaska and
11 federal lands in order to provide a complete inventory of emissions for this category. New land representation data
12 will also be compiled, and the time series recalculated for the latter years that are estimated using the data splicing
13 method in the current Inventory.
14 Owners ^ Cjrbon Stocks in Urban Trees (CRF Category 4E1)
15 Urban forests constitute a significant portion of the total U.S. tree canopy cover (Dwyer et al. 2000). Urban areas
16 (cities, towns, and villages) are estimated to cover over 3 percent of the United States (U.S. Census Bureau 2012).
17 With an average tree canopy cover of 35 percent, urban areas account for approximately 5 percent of total tree cover
18 in the continental United States (Nowak and Greenfield 2012). Trees in urban areas of the United States were
19 estimated to account for an average annual net sequestration of 78.3 MMT CO2 Eq. (21.3 MMT C) over the period
20 from 1990 through 2016. Net C flux from urban trees in 2016 was estimated to be -92.9 MMT CO2 Eq. (-25.3 MMT
21 C). Annual estimates of CO2 flux (Table 6-74) were developed based on periodic (1990, 2000, and 2010) U.S.
22 Census data on urbanized area. The estimate of urbanized area is smaller than the area categorized as Settlements in
23 the Representation of the U.S. Land Base developed for this report: over the 1990 through 2016 time series the
24 Census urban area totaled, on average, about 63 percent of the Settlements area.
25 In 2016, Census urban area totaled about 69 percent of the total area defined as Settlements. Census area data are
26 preferentially used to develop C flux estimates for this source category since these data are more applicable for use
27 with the available peer-reviewed data on urban tree canopy cover and urban tree C sequestration. Annual
28 sequestration increased by 54 percent between 1990 and 2016 due to increases in urban land area. Data on C storage
29 and urban tree coverage were collected since the early 1990s and have been applied to the entire time series in this
30 report. As a result, the estimates presented in this chapter are not truly representative of changes in C stocks in urban
31 trees for Settlements areas, but are representative of changes in C stocks in urban trees for Census urban area. The
32 method used in this report does not attempt to scale these estimates to the Settlements area. Therefore, the estimates
33 presented in this chapter are likely an underestimate of the true changes in C stocks in urban trees in all Settlements
34 areas—i.e., the changes in C stocks in urban trees presented in this chapter are a subset of the changes in C stocks in
3 5 urban trees in all Settlements areas.
36 Urban trees often grow faster than forest trees because of the relatively open structure of the urban forest (Nowak
37 and Crane 2002). Because tree density in urban areas is typically much lower than in forested areas, the C storage
38 per hectare of land is in fact smaller for urban areas than for forest areas. To quantify the C stored in urban trees, the
39 methodology used here requires analysis per unit area of tree cover, rather than per unit of total land area (as is done
40 for Forestlands). When expressed as per unit of tree cover, areas covered by urban trees actually have a greater C
41 density than do forested areas (Nowak and Crane 2002). Expressed per unit of land area, however, the situation is
42 the opposite: because tree density is so much lower in urban areas, these areas have a smaller C density per unit land
43 area than forest areas.
6-102 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Table 6-74: Net C Flux from Urban Trees (MMT CO2 Eq. and MMT C)
Year
MMT CO2 Eq.
MMT C
1000
(60.4)
(16.5)
2005
(80.5)
(22.0)
2012
(88.4)
(24.1)
2013
(80.5)
(24.4)
2014
(00.6)
(24.7)
2015
(01.7)
(25.0)
2016
(02.0)
(25.3)
Note: Parentheses indicate net
sequestration.
Methodology
Methods for quantifying urban 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, the methodology used by Nowak et al. (2013) to estimate net C sequestration in urban trees followed three
steps, each of which is explained further in the paragraphs below. First, field data from cities and states were used to
estimate C in urban 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 urban 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 values to derive estimates of net C sequestration.
For the current Inventory report, net C sequestration estimates for all 50 states and the District of Columbia, that
were generated using the Nowak et al. (2013) methodology and expressed in units of C sequestered per unit area of
tree cover, were then used to estimate urban tree C sequestration in the United States. To accomplish this, we used
urban area estimates from U.S. Census data together with urban tree cover percentage estimates for each state and
the District of Columbia from remote sensing data, an approach consistent with Nowak et al. (2013).
This approach is also 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 urban 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).
The first step in the methodology is to estimate C in urban tree biomass. To develop urban tree carbon estimates
Nowak et al. (2013) and previously published research (Nowak and Crane 2002; and Nowak 1994, 2007b, and
2009) collected field measurements in a number of U.S. cities between 1989 and 2012. For a random sample of trees
in representative cities, tree data were collected regarding stem diameter, tree height, crown height and crown width,
tree location, species, and canopy condition. The data for each tree were converted into total dry-weight biomass
estimates using allometric equations, 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 urban trees having less
aboveground biomass for a given stem diameter than predicted by allometric equations 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. The second step in the
methodology is to estimate rates of tree growth for urban trees in the United States. In the Nowak et al. (2013)
methodology that is applied here, growth rates were standardized for open-grown trees in areas with 153 days of
frost free length based on measured data on tree growth. These growth rates were then 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). For each tree, 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. The
Land Use, Land-Use Change, and Forestry 6-103
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
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 (Nowak 2011).
Most of the field data used to develop the methodology of Nowak et al. (2013) were analyzed using the U.S. Forest
Service's i-Tree Eco model (formerly Urban Forest Effects (UFORE) model). The i-Tree Eco computer model uses
standardized field data from randomly located plots, along with local hourly air pollution and meteorological data to
quantify urban forest structure, values of the urban forest, and environmental effects, including total C stored and
annual C sequestration. The model was used with field data from randomly sampled plots in each city or urban areas
in states to quantify the characteristics of the urban forest (Nowak et al. 2013).
Where gross C sequestration accounts for all carbon sequestered, net C sequestration for urban trees takes into
account C emissions associated with tree death and removals. In the third step in the methodology developed by
Nowak et al. (2002; 2013), estimates of net C emissions from urban trees were derived by applying estimates of
annual mortality based on 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 derived 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. The estimated annual gross C emission rates for each plot were then scaled up
to city estimates using tree population information.
The data for all 50 states and the District of Columbia are described in Nowak et al. (2013) and reproduced in Table
6-75, which builds upon previous research, including: Nowak and Crane (2002), Nowak et al. (2007), Nowak and
Greenfield (2012), and references cited therein. The full methodology development is described in the underlying
literature, and key details and assumptions were made as follows. The allometric equations 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 equation 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 (BG) 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 health and 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 6-75)
were compiled in units of C sequestration per unit area of tree canopy cover. These rates were used in conjunction
with estimates of state urban area and urban tree cover data (Nowak and Greenfield 2012) 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 U.S. Census data.
Specifically, urban area estimates were based on 1990, 2000, and 2010 U.S. Census data. The 1990 U.S. Census
defined urban land as "urbanized areas," which included land with a population density greater than 1,000 people
per square mile, and adjacent "urban places," which had predefined political boundaries and a population total
greater than 2,500. In 2000, the U.S. Census replaced the "urban places" category with a new category of urban land
called an "urban cluster," which included areas with more than 500 people per square mile. In 2010, the Census
updated its definitions to have "urban areas" encompassing Census tract delineated cities with 50,000 or more
people, and "urban clusters" containing Census tract delineated locations with between 2,500 and 50,000 people.
Urban land area increased by approximately 23 percent from 1990 to 2000 and 14 percent from 2000 to 2010;
Nowak et al. (2005) estimate that the changes in the definition of urban land are responsible for approximately 20
percent of the total reported increase in urban land area from 1990 to 2000. Under all Census (i.e., 1990, 2000, and
2010) definitions, the urban category encompasses most cities, towns, and villages (i.e., it includes both urban and
suburban areas). Settlements area, as assessed in the Representation of the U.S. Land Base developed for this report,
6-104 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
encompassed all developed parcels greater than 0.1 hectares in size, including rural transportation corridors, and as
previously mentioned represents a larger area than the Census-derived urban area estimates. However, the smaller,
Census-derived urban area estimates were deemed to be more suitable for estimating national urban tree cover given
the data available in the peer-reviewed literature (i.e., the data set available is consistent with Census urban rather
than Settlements areas), and the recognized overlap in the changes in C stocks between urban forest and non-urban
forest (see Planned Improvements below). U.S. Census urban area data are reported as a series of continuous blocks
of urban area in each state. The blocks or urban area were summed to create each state's urban area estimate.
Net annual C sequestration estimates were derived for all 50 states and the District of Columbia by multiplying the
gross annual emission estimates by 0.74, the standard ratio for net/gross sequestration set out in Table 3 of Nowak et
al. (2013) (unless data existed for both gross and net sequestration for the state in Table 2 of Nowak et. al. (2013), in
which case they were divided to get a state-specific ratio). The gross and net annual C sequestration values for each
state were multiplied by each state's area of tree cover, which was the product of the state's urban/community area
as defined in the U.S. Census (2012) and the state's urban/community tree cover percentage. The urban/community
tree cover percentage estimates for all 50 states were obtained from Nowak and Greenfield (2012). The
urban/community tree cover percentage estimate for the District of Columbia was obtained from Nowak et al.
(2013). The urban area estimates were taken from the 2010 U.S. Census (2012). The equation, used to calculate the
summed carbon sequestration amounts, can be written as follows:
Net annual C sequestration = Gross sequestration rate x Net to Gross sequestration ratio x Urban Area x
% Tree Cover
Table 6-75: Annual C Sequestration (Metric Tons C/Year), Tree Cover (Percent), and Annual
C Sequestration per Area of Tree Cover (kg C/m2-yr) for 50 states plus the District of
Columbia (2016)
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
1,207,204
893,331
55.2
0.343
0.254
0.74
Alaska
44,593
32,999
39.8
0.168
0.124
0.74
Arizona
402,045
297,513
17.6
0.354
0.262
0.74
Arkansas
438,481
324,476
42.3
0.331
0.245
0.74
California
2,119,770
1,568,630
25.1
0.389
0.288
0.74
Colorado
158,608
117,370
18.5
0.197
0.146
0.74
Connecticut
775,500
573,870
67.4
0.239
0.177
0.74
Delaware
142,326
105,321
35.0
0.335
0.248
0.74
DC
14,561
11,571
35.0
0.263
0.209
0.79
Florida
3,528,013
2,610,730
35.5
0.475
0.352
0.74
Georgia
2,684,691
1,986,671
54.1
0.353
0.261
0.74
Hawaii
251,232
185,911
39.9
0.581
0.430
0.74
Idaho
26,407
19,541
10.0
0.184
0.136
0.74
Illinois
773,115
572,105
25.4
0.283
0.209
0.74
Indiana
415,255
383,968
23.7
0.250
0.231
0.92
Iowa
122,216
90,440
19.0
0.240
0.178
0.74
Kansas
189,999
147,851
25.0
0.283
0.220
0.78
Kentucky
249,995
184,997
22.1
0.286
0.212
0.74
Louisiana
771,314
570,772
34.9
0.397
0.294
0.74
Maine
108,310
80,150
52.3
0.221
0.164
0.74
Maryland
609,241
450,838
34.3
0.323
0.239
0.74
Massachusetts
1,324,939
980,455
65.1
0.254
0.188
0.74
Michigan
748,782
554,099
35.0
0.220
0.163
0.74
Minnesota
359,271
265,861
34.0
0.229
0.169
0.74
Mississippi
508,818
376,525
47.3
0.344
0.255
0.74
Missouri
509,564
377,077
31.5
0.285
0.211
0.74
Montana
55,205
40,852
36.3
0.184
0.136
0.74
Nebraska
52,156
44,013
15.0
0.238
0.201
0.84
Nevada
46,396
34,333
9.6
0.207
0.153
0.74
New Hampshire
256,348
189,697
66.0
0.217
0.161
0.74
New Jersey
1,209,144
894,766
53.3
0.294
0.218
0.74
Land Use, Land-Use Change, and Forestry 6-105
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
New Mexico
71,215
52,699
12.0
0.263
0.195
0.74
New York
1,103,216
816,380
42.6
0.240
0.178
0.74
North Carolina
2,163,326
1,600,861
51.1
0.312
0.231
0.74
North Dakota
15,520
7,375
13.0
0.223
0.106
0.48
Ohio
943,793
698,407
31.5
0.248
0.184
0.74
Oklahoma
373,957
276,728
31.2
0.332
0.246
0.74
Oregon
264,655
195,844
36.6
0.242
0.179
0.74
Pennsylvania
1,287,482
952,736
41.0
0.244
0.181
0.74
Rhode Island
137,454
101,716
51.0
0.258
0.191
0.74
South Carolina
1,152,059
852,523
48.9
0.338
0.250
0.74
South Dakota
22,340
19,373
14.0
0.236
0.205
0.87
Tennessee
1,095,753
979,732
43.8
0.303
0.271
0.89
Texas
2,904,124
2,149,052
31.4
0.368
0.272
0.74
Utah
95,804
70,895
16.4
0.215
0.159
0.74
Vermont
47,031
34,803
53.0
0.213
0.158
0.74
Virginia
856,934
634,131
39.8
0.293
0.217
0.74
Washington
582,070
430,732
34.6
0.258
0.191
0.74
West Virginia
261,146
193,248
61.0
0.241
0.178
0.74
Wisconsin
372,818
275,885
31.8
0.225
0.167
0.74
Wyoming
19,680
14,563
19.9
0.182
0.135
0.74
Total
33,873,873
25,324,418
Uncertainty and Time-Series Consistency
Uncertainty associated with changes in C stocks in urban trees includes the uncertainty associated with urban area,
percent urban tree coverage, 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 urban area estimates based on expert judgment.
Uncertainty associated with estimates of percent urban tree coverage for each of the 50 states was based on standard
error estimates reported by Nowak and Greenfield (2012). Uncertainty associated with estimate of percent urban tree
coverage for the District of Columbia was based on the standard error estimate reported by Nowak et al. (2013).
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 reported by
Nowak et al. (2013). 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 equations, 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 some overlap between the urban tree C estimates and the forest tree C estimates as
detailed in Nowak et al. (2013). Due to data limitations, urban soil flux is not quantified as part of this analysis,
while reconciliation of urban 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 2015. This uncertainty was updated in 2016 based on proportional allocation of changes
between the 2015 and 2016 flux estimate. The results of this adjusted quantitative uncertainty analysis are
summarized in Table 6-76. The net C flux from changes in C stocks in urban trees in 2016 was estimated to be
between -136.9 and -47.9 MMT CO2 Eq. at a 95 percent confidence level. This indicates a range of 47 percent more
sequestration to 48 percent less sequestration than the 2016 flux estimate of -92.9 MMT CO2 Eq.
Table 6-76: Approach 2 Quantitative Uncertainty Estimates for Net C Flux from Changes in C
Stocks in Urban Trees (MMT CO2 Eq. and Percent)
Source
Gas
2016 Flux Estimate
(MMT CO2 Eq.)
Uncertainty Range Relative to Flux Estimate3
(MMT CO2 Eq.) (%)
Lower Upper
Bound Bound
Lower Upper
Bound Bound
Changes in C Stocks in
Urban Trees
CO2
(92.9)
(136.9) (47.9)
-47% 48%
6-106 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
a Range of uncertainty in emissions was estimated based on proportional allocation of 2015 to 2016 flux values to the 2015
uncertainty estimates. Range of flux estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence
interval.
Note: Parentheses indicate negative values or net sequestration.
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2016. Details on the emission trends through time are described in more detail in the Methodology section,
above.
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 urban 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.
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 USD A Forest Service to reconcile the overlap
between urban forest and non-urban forest greenhouse gas inventories. Because some plots defined as "forest" in the
Forest Inventory and Analysis (FIA) program of the USD A Forest Service actually fall within the boundaries of the
areas also defined as Census urban, there may be "double-counting" of these land areas in estimates of C stocks and
fluxes for this report. Specifically, 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 Trees source category.
Future research may also enable more complete coverage of changes in the C stock in urban trees for all Settlements
land. To provide estimates for all Settlements, research would need to establish the extent of overlap between the
areas of land included in the Settlements land use category and Census-defined urban areas, and would have to
separately characterize sequestration on non-urban Settlements land.
To provide more accurate emissions estimates in the urban forest greenhouse gas inventories, the following actions
will be taken:
a) Development of a national definition of "settlements". Settlements are defined as including "all developed
land, including transportation infrastructure and human settlements of any size, unless they are already
included under other categories. This should be consistent with the selection of national definitions". In the
U.S., different types of classifications can be used to determine settlements e.g., Census urban, Census
urban/community, National Land Cover Dataset, and National Resources Inventory. A combination of
these data will be used to encompass settlement areas and improve consistency with Section 6.1,
Representation of the U.S. Land Base;
b) For settlement areas, estimates of land area will be obtained for 1990, 2000 and 2010 and projections
developed for annual growth during the 2010 to 2020 period;
c) 2,500 random points will be laid on aerial images using Google Earth imagery to estimate tree cover in the
settlement areas circa 1990, 2000 and 2010. Trends in tree cover change will be used to estimate tree cover
in settlement between 2010 and 2020;
d) Photo interpretation of settlement tree cover will be updated bi-annually to update tree cover estimates and
trends;
e) A review of recent literature will be performed to update C storage, sequestration and net-to-gross
sequestration rates per unit tree cover;
f) C rates per unit tree cover will be applied to tree cover estimates within estimated settlement areas annually
to estimate past and current C values; and
g) Settlement areas will be updated approximately every 10 years based on updated data from the U.S. Census
and NLCD developed land.
Land Use, Land-Use Change, and Forestry 6-107
-------
1 N20 Emissions from Settlement Soils (CRF Source Category
2 4E1)
3 Of the synthetic N fertilizers applied to soils in the United States, approximately 3.1 percent are currently applied to
4 lawns, golf courses, and other landscaping within settlement areas. Application rates are lower than those occurring
5 on cropped soils, and, therefore, account for a smaller proportion of total U.S. soil N20 emissions per unit area. In
6 addition to synthetic N fertilizers, a portion of surface applied biosolids (i.e., sewage sludge) is applied to settlement
7 areas, and drained organic soils (i.e., soils with high organic matter content, known as Histosols) also contribute to
8 emissions of soil N20.
9 N additions to soils result in direct and indirect N20 emissions. Direct emissions occur on-site due to the N additions
10 in the form of synthetic fertilizers and biosolids as well as enhanced mineralization of N in drained organic soils.
11 Indirect emissions result from fertilizer and sludge N that is transformed and transported to another location in a
12 form other than N20 (ammonia [NH3] and nitrogen oxide [NOx] volatilization, nitrate [NO3 ] leaching and runoff),
13 and later converted into N20 at the off-site location. The indirect emissions are assigned to settlements because the
14 management activity leading to the emissions occurred in settlements.
15 Total N20 emissions from soils in Settlements Remaining Settlements^ are 2.5 MMT C02 Eq. (8 kt of N20) in
16 2016. There is an overall increase of 75 percent from 1990 to 2016 due to an expanding settlement area leading to
17 more synthetic N fertilizer applications. Inter-annual variability in these emissions is directly attributable to
18 variability in total synthetic fertilizer consumption, area of drained organic soils, and biosolids applications in the
19 United States. Emissions from this source are summarized in Table 6-77.
20 Table 6-77: N2O Emissions from Soils in Settlements Remaining Settlements (MMT CO2 Eq.
21 and kt N2O)
1990
2005
2012
2013
2014
2015
2016
MMT CO2 Eq.
Direct N2O Emissions from
Soils
1.1
1.9
2.1
2.0
2.0
2.0
1.9
Synthetic Fertilizers
0.8
1.6
1.7
1.7
1.7
1.6
1.6
Biosolids
0.2
0.2
0.2
0.2
0.2
0.2
0.2
Drained Organic Soils
0.1
0.1
0.2
0.2
0.2
0.2
0.2
Indirect N2O Emissions from
Soils
0.4
0.6
0.6
0.6
0.6
0.6
0.6
Total
1.4
2.5
2.7
2.6
2.6
2.5
2.5
ktN20
Direct N2O Emissions from
Soils
4
6
7
7
7
7
7
Synthetic Fertilizers
3
5
6
6
6
6
5
Biosolids
1
1
1
1
1
1
1
Drained Organic Soils
+
1
1
1
1
1
1
Indirect N2O Emissions from
Soils
1
2
2
2
2
2
2
Total
5
8
9
9
9
9
8
+ Does not exceed 0.5 kt
Notes: Estimates after 2012 are based on a data splicing method (see Methodology section), except for
biosolids. Totals may not sum due to independent rounding. Estimates of Soil N2O for Settlements
Remaining Settlements include emissions from Land Converted to Settlements because it was not possible to
separate the activity data.
61 Estimates of Soil N2O for Settlements Remaining Settlements include emissions from Land Converted to Settlements because it
was not possible to separate the activity data.
6-108 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
Methodology
For settlement soils, the IPCC Tier 1 approach is used to estimate soil N20 emissions from synthetic N fertilizer,
biosolids additions, and drained organic soils. Estimates of direct N20 emissions from soils in settlements are based
on the amount of N in synthetic commercial fertilizers applied to settlement soils, the amount of N in biosolids
applied to non-agricultural land and surface disposal (see Section 7.2, Wastewater Treatment for a detailed
discussion of the methodology for estimating biosolids application), and the area of drained organic soils within
settlements.
Nitrogen applications to settlement soils are estimated using data compiled by the USGS (Ruddy et al. 2006). The
USGS estimated on-farm and non-farm fertilizer use is based on sales records at the county level from 1982 through
2001 (Ruddy et al. 2006). Non-farm N fertilizer is assumed to be applied to settlements and forest lands; values for
2002 through 2012 are based on 2001 values adjusted for annual total N fertilizer sales in the United States because
there is no new activity data on application after 2001. Settlement application is calculated by subtracting forest
application from total non-farm fertilizer use. 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 N20 emissions (IPCC 2006) for 1990 to 2012. The IPCC (2006) Tier 1 method is also used to
estimate direct N20 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 2012 NRI (USDA-NRCS 2015) using soils data from the
Soil Survey Geographic Database (SSURGO) (Soil Survey Staff 2011). To estimate annual emissions from 1990 to
2012, the total area is multiplied by the IPCC default emission factor for temperate regions (IPCC 2006). This
Inventory does not include soil N20 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 sludge 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 N20 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 N20 off-site. The
resulting estimates are summed to obtain total indirect emissions from 1990 to 2012.
A linear extrapolation of the trend in the time series is applied to estimate the direct and indirect N20 emissions
from 2013 to 2016 from synthetic fertilizers and drained organic soils because new activity data for these two
sources have not been compiled for the latter part of the time series. 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 2012, and in turn, the trend is used to approximate the 2013 to 2016 emissions.
The time series will be recalculated for the years beyond 2012 in a future inventory with the methods described
above for 1990 to 2012. This Inventory does incorporate updated activity data on biosolids application in settlements
through 2016.
Uncertainty and Time-Series Consistency
The amount of N20 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 N20
emissions is complex and highly uncertain. The IPCC default methodology does not explicitly incorporate any of
these variables, except variations in the total amount of fertilizer N and biosolids applications. All settlement soils
are treated equivalently under this methodology.
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.62 Uncertainty in drained organic
soils is based on the estimated variance from the NRI survey (USDA-NRCS 2015). For 2013 to 2016, there is also
62 No uncertainty is provided with the USGS fertilizer consumption data (Ruddy et al. 2006) so a conservative ±50 percent is
used in the analysis. Biosolids data are also assumed to have an uncertainty of ±50 percent.
Land Use, Land-Use Change, and Forestry 6-109
-------
1 additional uncertainty associated with the surrogate data method. Uncertainty in the amounts of biosolids applied to
2 non-agricultural lands and used in surface disposal is derived from variability in several factors, including: (1) N
3 content of biosolids; (2) total sludge applied in 2000; (3) wastewater existing flow in 1996 and 2000; and (4) the
4 biosolids disposal practice distributions to non-agricultural land application and surface disposal. Uncertainty in the
5 direct and indirect emission factors is provided by IPCC (2006).
6 Uncertainty is propagated through the calculations of N20 emissions from fertilizer N and drainage of organic soils
7 using a Monte Carlo analysis. The results are combined with the uncertainty in N20 emissions from the biosolids
8 application using simple error propagation methods (IPCC 2006). The results are summarized in Table 6-78. Direct
9 N2O emissions from soils in Settlements Remaining Settlements in 2016 are estimated to be between 1.4 and 2.7
10 MMT CO2 Eq. at a 95 percent confidence level. This indicates a range of 28 percent below to 38 percent above the
11 2016 emission estimate of 1.9 MMT CO2 Eq. Indirect N2O emissions in 2016 are between 0.4 and 0.7 MMT CO2
12 Eq., ranging from a -24 percent to 24 percent around the estimate of 0.6 MMT CO2 Eq.
13 Table 6-78: Quantitative Uncertainty Estimates of N2O Emissions from Soils in Settlements
14 Remaining Settlements (MMT CO2 Eq. and Percent)
Source
Gas
2016 Emissions
(MMT CO2 Eq.)
Uncertainty Range Relative to Emission Estimate3
(MMT CO2 Eq.) (%)
Settlements Remaining
Settlements
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Direct N2O Emissions from Soils
N2O
1.9
1.4
2.7
-28%
38%
Indirect N2O Emissions from
Soils
N2O
0.6
0.4
0.7
-24%
24%
a Range of emission estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
Note: These estimates include direct and indirect N2O emissions from Settlements Remaining Settlements and Land
Converted to Settlements because it was not possible to separate the activity data.
15 Methodological recalculations are applied from 2013 to 2015 using the linear time series model described above.
16 Details on the emission trends through time are described in more detail in the Methodology section, above.
17 QA/QC and Verification
18 The spreadsheet containing fertilizer, drainage of organic soils, and biosolids applied to settlements and calculations
19 for N20 and uncertainty ranges have been checked and verified.
20 Recalculations Discussion
21 Methodological recalculations are associated with extending the time series from 2013 through 2016 using a linear
22 time series model. The recalculation had a minor effect on the time series overall with N20 emissions declining by
23 less than 1 percent on average.
24 Planned Improvements
25 This source will be extended to include soil N20 emissions from drainage of organic soils in settlements of Alaska
26 and federal lands in order to provide a complete inventory of emissions for this category. Updated data on fertilizer
27 amount and area of drained organic soils will be compiled to update emissions estimates for estimates beyond 2012
28 in a future Inventory.
29 Changes in Yard Trimmings and Food Scrap Carbon Stock' in
30 Landfills (CRF Category 4E1)
31 In the United States, yard trimmings (i.e., grass clippings, leaves, and branches) and food scraps account for a
32 significant portion of the municipal waste stream, and a large fraction of the collected yard trimmings and food
6-110 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
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 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 declined
over the last decade. 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 2016). 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 1.4 percent decrease
between 1990 and 2015 in the tonnage of yard trimmings generated (i.e., collected for composting or disposal in
landfills). 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—from 72 percent in 1990 to 31 percent in 20 1 5.63 The net
effect of the reduction in generation and the increase in composting is a 57 percent decrease in the quantity of yard
trimmings disposed of in landfills since 1990.64
Food scrap generation has grown by an estimated 61 percent since 1990, and while the proportion of total food
scraps generated that are eventually discarded in landfills has decreased slightly, from an estimated 82 percent in
1990 to 76 percent in 2015, the tonnage disposed of in landfills has increased considerably (by an estimated 50
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 2015, 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 6-79 and Table 6-80). 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 CH4 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."65
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 26.0
MMT C02 Eq. (7.1 MMT C) in 1990 to 12.2 MMT C02 Eq. (3.3 MMT C) in 2016 (Table 6-79 and Table 6-80).
63 Updated data for 2016 were not included for the current Inventory, therefore the trend analysis is based on the latest data
through 2015
64Landfilled yard trimming amounts were not estimated for 2016; the values are estimated from 1990 through 2015.
65 Food scrap generation was not estimated for 2016; the values are estimated from 1990 through 2015.
Land Use, Land-Use Change, and Forestry 6-111
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Table 6-79: Net Changes in Yard Trimmings and Food Scrap Carbon Stocks in Landfills
(MMT COz Eq.)
Carbon Pool
1990
2005
2012
2013
2014
2015
2016
Yard Trimmings
(21.0)
(7.4)
(9.1)
(8.4)
(8.3)
(8.3)
(8.5)
Grass
(1.8)
(0.6)
(0.9)
(0.8)
(0.8)
(0.8)
(0.8)
Leaves
(9.0)
(3.4)
(4.1)
(3.9)
(3.8)
(3.8)
(3.9)
Branches
(10.2)
(3.4)
(4.1)
(3.8)
(3.7)
(3.7)
(3.8)
Food Scraps
(5.0)
(4.0)
(3.1)
(3.2)
(3.6)
(3.4)
(3.7)
Total Net Flux
(26.0)
(11.4)
(12.2)
(11.6)
(11.9)
(11.8)
(12.2)
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Table 6-80: Net Changes in Yard Trimmings and Food Scrap Carbon Stocks in Landfills
(MMT C)
Carbon Pool
1990
2005
2012
2013
2014
2015
2016
Yard Trimmings
(5.7)
(2.0)
(2.5)
(2.3)
(2.3)
(2.3)
(2.3)
Grass
(0.5)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
(0.2)
Leaves
(2.5)
(0.9)
(1.1)
(1.1)
(1.0)
(1.0)
(1.1)
Branches
(2.8)
(0.9)
(1.1)
(1.0)
(1.0)
(1.0)
(1.0)
Food Scraps
(1.4)
(1.1)
(0.9)
(0.9)
(1.0)
(0.9)
(1.0)
Total Net Flux
(7.1)
(3.1)
(3.3)
(3.2)
(3.3)
(3.2)
(3.3)
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Methodology
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, based on methodologies presented for the Land Use,
Land-Use Change, and Forestry sector in IPCC (2003) and the 2006 LPCC Guidelines for National Greenhouse Gas
Lnventories (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.
To determine the total landfilled C stocks for a given year, the following were estimated: (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 2014 (EPA 2016),
which provides data for 1960, 1970, 1980, 1990, 2000, 2005, 2009 and 2011 through 2013. To provide data for
some of the missing years, detailed backup data were obtained from historical data tables that EPA developed for
1960 through 2013 (EPA 2015). Remaining years in the time series for which data were not provided were estimated
using linear interpolation. Due to the limited update this inventory year, data for 2015 was set equal to 2014 values,
and 2016 was not estimated. The EPA (2016) report and historical data tables (EPA 2015) do not subdivide the
discards (i.e., total generated minus composted) of individual materials into masses landfilled and combusted,
6-112 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
although it provides a mass of overall waste stream discards managed in landfills66 and combustors with energy
recovery (i.e., ranging from 67 percent and 33 percent, respectively, in 1960 to 92 percent and 8 percent,
respectively, in 1985); it is assumed that the proportion of each individual material (food scraps, grass, leaves,
branches) that is landfilled is the same as the proportion across the overall waste stream.
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 6-81).
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 6-81).
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 CH4 and CO2. (The CH4 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 6-81.
The first-order decay rates, k. for each refuse type were 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, 1990 waste composition
for the United States fromEPA's Characterization of Municipal Solid Waste in the United States: 1990 Update
(EPA 1991) was used to calculate f. This correction factor is then multiplied 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, k= 0.12). As in the Landfills section of the Inventory (Section 7.1), which
estimates CH4 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
year1), but not for the other two overall MSW decay rates. To maintain consistency between landfill methodologies
across the Inventory, the correction factors (J) were developed for decay rates of 0.038 and 0.057 year1 through
linear interpolation. A weighted national average component-specific decay rate was 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 component-specific decay rates are shown in Table 6-81.
66 EPA (2016 and 2015) reports discards in two categories: "combustion with energy recovery" and "landfill, other disposal,"
which includes combustion without energy recovery. For years in which there is data from previous EPA reports on combustion
without energy recovery, EPA assumes these estimates are still applicable. For 2000 to present, EPA assumes that any
combustion of MSW that occurs includes energy recovery, so all discards to "landfill, other disposal" are assumed to go to
landfills.
Land Use, Land-Use Change, and Forestry 6-113
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
For 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 1:
LFCi,t= £ Wi,n x (1 - MO) x ICCix {[CSix ICG[ + [(1 - (C.Sx /(X)) x e-^-")]}
where,
MC,
CSi
ICC,
LFQt
Wt,„
k
n
t
e
Year for which C stocks are being estimated (year),
Waste type for which C stocks are being estimated (grass, leaves, branches, food scraps),
Stock of C in landfills in year I. for waste /' (metric tons),
Mass of waste /' disposed of in landfills in year n (metric tons, wet weight),
Year in which the waste was disposed of (year, where 1960
-------
1
Table 6-82: C Stocks in Yard Trimmings and Food Scraps in Landfills (MMT C)
Carbon Pool
1990
2005
2012
2013
2014
2015
2016
Yard Trimmings
155.8
202.9
218.6
220.9
223.1
225.4
227.7
Branches
14.5
18.1
19.5
19.7
19.9
20.2
20.4
Leaves
66.7
87.3
94.5
95.5
96.6
97.6
98.7
Grass
74.6
97.5
104.5
105.6
106.6
107.6
108.6
Food Scraps
17.6
32.8
39.8
40.7
41.6
42.6
43.6
Total Carbon
Stocks
173.5
235.6
258.3
261.5
264.8
268.0
271.3
Note: Totals may not sum due to independent rounding.
2 To develop the 2016 estimate, a simplified inventory update was performed using values from the 1990 through
3 2015 Inventory. Estimates of yard trimming and food scrap carbon stocks were forecasted for 2016, which were then
4 used to calculate net changes in carbon stocks. Excel's FORECAST.ETS function was used to predict a 2016 value
5 using historical data via an algorithm called "Exponential Triple Smoothing". This method smooths out the data to
6 determine the overall trend and provide an appropriate estimate for 2016.
7 Uncertainty and Time-Series Consistency
8 The uncertainty analysis for landfilled yard trimmings and food scraps includes an evaluation of the effects of
9 uncertainty for the following data and factors: disposal in landfills per year (tons of C), initial C content, moisture
10 content, decay rate, and proportion of C stored. The C storage landfill estimates are also a function of the
11 composition of the yard trimmings (i.e., the proportions of grass, leaves and branches in the yard trimmings
12 mixture). There are respective uncertainties associated with each of these factors.
13 A Monte Carlo (Approach 2) uncertainty analysis that was run on the previous (i.e., 1990 through 2015) Inventory
14 was applied to estimate the overall uncertainty of the sequestration estimate for 2016. The results of the Approach 2
15 quantitative uncertainty analysis are summarized in Table 6-83. Total yard trimmings and food scraps CO2 flux in
16 2016 was estimated to be between -19.0 and -4.8 MMT CO2 Eq. at a 95 percent confidence level (or 19 of 20 Monte
17 Carlo stochastic simulations). This indicates a range of 56 percent below to 61 percent above the 2016 flux estimate
18 of -12.2 MMT C02 Eq.
19 Table 6-83: Approach 2 Quantitative Uncertainty Estimates for CO2 Flux from Yard
20 Trimmings and Food Scraps in Landfills (MMT CO2 Eq. and Percent)
2016 Flux
Source
Gas
Estimate
Uncertainty Range Relative to Flux Estimate3
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower Upper
Lower Upper
Bound Bound
Bound Bound
Yard Trimmings and
Food Scraps
CO2
(12.2)
(19.0) (4.8)
-56% 61%
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.
21 QA/QC and Verification
22 Tier 1 and Tier 2 QA/QC activities were conducted consistent with the U.S. QA/QC plan. Source-specific quality
23 control measures for Landfilled Yard Trimmings and Food Scraps included checking that input data were properly
24 transposed within the spreadsheet, checking calculations were correct, and confirming that all activity data and
25 calculations documentation was complete and updated to ensure data were properly handled through the inventory
26 process.
27 Order of magnitude checks and checks of time-series consistency were performed to ensure data were updated
28 correctly and any changes in emissions estimates were reasonable and reflected changes in activity data. An annual
Land Use, Land-Use Change, and Forestry 6-115
-------
1 change trend analysis was also conducted to ensure the validity of the emissions estimates. Errors that were found
2 during this process were corrected as necessary.
3 Recalculations Discussion
4 No recalculations were performed for the 1990 through 2015 estimates in this Inventory.
5 Planned Improvements
6 Future work is planned to evaluate the consistency between the estimates of C storage described in this chapter and
7 the estimates of landfill CH4 emissions described in the Waste chapter. For example, the Waste chapter does not
8 distinguish landfill CH4 emissions from yard trimmings and food scraps separately from landfill CH4 emissions from
9 total bulk (i.e., municipal solid) waste, which includes yard trimmings and food scraps.
10 In addition, additional data from recent peer-reviewed literature will be evaluated that may modify the default C
11 storage factors, initial C contents, and decay rates for yard trimmings and food scraps in landfills. Based upon this
12 evaluation, changes may be made to the default values. Updating of the weighted national average component-
13 specific decay rate using new U.S. Census data will also be evaluated, if any are available.
14 Yard waste composition will also be investigated to determine if changes need to be made based on changes in
15 residential practices, a review of available literature will be conducted to determine if there are changes in the
16 allocation of yard trimmings. For example, leaving grass clippings in place is becoming a more common practice,
17 thus reducing the percentage of grass clippings in yard trimmings disposed in landfills. In addition, agronomists may
18 be consulted for determining the mass of grass per acre on residential lawns to provide an estimate of total grass
19 generation for comparison with Inventory estimates.
20 Finally, available data will be reviewed to ensure all types of yard trimmings and food scraps are being included in
21 Inventory estimates, such as debris from road construction.
22 6.11 Land Converted to Settlements (CRF
23 Category 4E2)
24 Land Converted to Settlements includes all settlements in an Inventory year that had been in another land use(s)
25 during the previous 20 years (USDA-NRCS 2015).68 For example, cropland, grassland or forest land converted to
26 settlements during the past 20 years would be reported in this category. Recently-converted lands are retained in this
27 category for 20 years as recommended by IPCC (2006). This Inventory includes all settlements in the conterminous
28 United States and Hawaii, but does not include settlements in Alaska. Areas of drained organic soils on settlements
29 in federal lands are also not included in this Inventory. Consequently, there is a discrepancy between the total
30 amount of managed area for Land Converted to Settlements (see Section 6.1—Representation of the U.S. Land
31 Base) and the settlements area included in the inventory analysis.
32 Land use change can lead to large losses of carbon (C) to the atmosphere, particularly conversions from forest land
33 (Houghton et al. 1983). Moreover, conversion of forest to another land use (i.e., deforestation) is one of the largest
34 anthropogenic sources of emissions to the atmosphere globally (Schimel 1995), although this source may be
35 declining globally according to a recent assessment (Tubiello et al. 2015).
36 IPCC (2006) recommends reporting changes inbiomass, dead organic matter, and soil organic C (SOC) stocks due
37 to land use change. All soil C stock changes are estimated and reported for Land Converted to Settlements, but there
68 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.
6-116 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 is limited reporting of other pools in this Inventory. Loss of aboveground and belowground biomass, dead wood and
2 litter C are reported for Forest Land Converted to Settlements, but not for other land use conversions to settlements.
3 Forest Land Converted to Settlements is the largest source of emissions from 1990 to 2016, accounting for
4 approximately 66 percent of the average total loss of C among all of the land use conversions in Land Converted to
5 Settlements. Losses of aboveground and belowground biomass, dead wood and litter C losses in 2016 are 32.7, 6.6,
6 2.2, and 2.0 MMT CO2 Eq. (8.9, 1.8, 0.6, and 0.5 MMT C). Mineral and organic soils also lost 22.6 and 1.9 MMT
7 C02 Eq. in 2016 (6.2 and 0.5 MMT C). The total net flux is 68.0 MMT C02 Eq. in 2016 (18.5 MMT C), which is an
8 83 percent increase in CO2 emissions compared to the emissions in the initial reporting year of 1990. The main
9 driver of net emissions for this source category is the conversion of forest land to settlements, with large losses of
10 biomass C.
11 Table 6-84: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for
12 Land Converted to Settlements (MMT CO2 Eq.)
1990
2005
2012
2013
2014
2015
2016
Cropland Converted to
Settlements
4.1
11.9
10.3
10.3
10.2
10.2
10.1
Mineral Soils
3.5
10.7
9.4
9.4
9.4
9.3
9.3
Organic Soils
0.6
1.2
0.9
0.9
0.9
0.8
0.9
Forest Land Converted to
Settlements
29.0
42.3
44.8
44.9
44.8
44.8
44.8
Aboveground Live Biomass
19.9
29.9
32.7
32.7
32.7
32.7
32.7
Belowground Live Biomass
4.0
6.0
6.6
6.6
6.6
6.6
6.6
Dead Wood
2.2
2.7
2.2
2.2
2.2
2.2
2.2
Litter
2.0
2.3
2.0
2.0
2.0
2.0
2.0
Mineral Soils
0.9
1.3
1.3
1.3
1.3
1.3
1.3
Organic Soils
+
+
+
0.1
+
+
+
Grassland Converted
Settlements
4.0
13.5
12.4
12.4
12.4
12.4
12.3
Mineral Soils
3.5
12.3
11.5
11.5
11.5
11.4
11.4
Organic Soils
0.5
1.2
0.8
0.9
0.9
0.9
0.9
Other Lands Converted to
Settlements
0.2
0.7
0.7
0.7
0.7
0.7
0.7
Mineral Soils
0.2
0.6
0.6
0.6
0.6
0.6
0.6
Organic Soils
+
0.1
0.1
0.1
0.1
0.1
0.1
Wetlands Converted to
Settlements
+
0.1
0.1
0.1
0.1
0.1
0.1
Mineral Soils
+
0.1
0.1
0.1
0.1
0.1
0.1
Organic Soils
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Total Aboveground Biomass
Flux
19.9
29.9
32.7
32.7
32.7
32.7
32.7
Total Belowground Biomass
Flux
4.0
6.0
6.6
6.6
6.6
6.6
6.6
Total Dead Wood Flux
2.2
2.7
2.2
2.2
2.2
2.2
2.2
Total Litter Flux
2.0
2.3
2.0
2.0
2.0
2.0
2.0
Total Mineral Soil Flux
8.0
24.9
22.9
22.9
22.8
22.7
22.6
Total Organic Soil Flux
1.1
2.5
1.9
2.0
1.9
1.9
1.9
Total Net Flux
37.2
68.4
68.3
68.3
68.2
68.1
68.0
+ Does not exceed 0.05 MMT CO2 Eq.
Notes: Estimates after 2012 are based on a data splicing method (see Methodology section). Totals may not sum due to
independent rounding.
13 Table 6-85: Net CO2 Flux from Soil, Dead Organic Matter and Biomass C Stock Changes for
14 Land Con verted to Settlements ( M MT C)
1990
2005
2012
2013
2014
2015
2016
Cropland Converted to
Settlements
1.1
3.2
2.8
2.8
2.8
2.8
2.8
Mineral Soils
0.9
2.9
2.6
2.6
2.6
2.5
25
Land Use, Land-Use Change, and Forestry 6-117
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Organic Soils
0.2
0.3
0.2
0.2
0.2
0.2
0.2
Forest Land Converted to
Settlements
7.9
11.5
12.2
12.2
12.2
12.2
12.2
Aboveground Live Biomass
5.4
8.2
8.9
8.9
8.9
8.9
8.9
Belowground Live Biomass
1.1
1.6
1.8
1.8
1.8
1.8
1.8
Dead Wood
0.6
0.7
0.6
0.6
0.6
0.6
0.6
Litter
0.5
0.6
0.5
0.5
0.5
0.5
0.5
Mineral Soils
0.3
0.4
0.4
0.3
0.3
0.3
0.3
Organic Soils
+
+
+
+
+
+
+
Grassland Converted
Settlements
1.1
3.7
3.4
3.4
3.4
3.4
3.4
Mineral Soils
0.9
3.4
3.1
3.1
3.1
3.1
3.1
Organic Soils
0.1
0.3
0.2
0.2
0.2
0.2
0.2
Other Lands Converted to
Settlements
+
0.2
0.2
0.2
0.2
0.2
0.2
Mineral Soils
+
0.2
0.2
0.2
0.2
0.2
0.2
Organic Soils
+
+
+
+
+
+
+
Wetlands Converted to
Settlements
+
+
+
+
+
+
+
Mineral Soils
+
+
+
+
+
+
+
Organic Soils
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Total Aboveground Biomass
Flux
5.4
8.2
8.9
8.9
8.9
8.9
8.9
Total Belowground Biomass
Flux
1.1
1.6
1.8
1.8
1.8
1.8
1.8
Total Dead Wood Flux
0.6
0.7
0.6
0.6
0.6
0.6
0.6
Total Litter Flux
0.5
0.6
0.5
0.5
0.5
0.5
0.5
Total Mineral Soil Flux
2.2
6.8
6.2
6.2
6.2
6.2
6.2
Total Organic Soil Flux
0.3
0.7
0.5
0.5
0.5
0.5
0.5
Total Net Flux
10.2
18.7
18.6
18.6
18.6
18.6
18.5
+ Does not exceed 0.05 MMT C
Notes: Estimates after 2012 are based on a data splicing method (see Methodology section). Totals may not sum due to
independent rounding.
Methodology
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
mineral and organic soil C stocks.
Biomass, Dead Wood, and Litter Carbon Stock Changes
A combination of Tier 1 and 2 methods is applied to estimate aboveground and belowground biomass, dead wood,
and litter C stock changes for Forest Land Converted to Settlements. For this method, all annual plots and portions
of plots (i.e., conditions; hereafter referred to as plots) from the Forest Inventory and Analysis (FIA) program are
evaluated for land use change in the 48 conterminous United States (i.e., all states except Alaska and Hawaii)
(USDA Forest Service 2015). Specifically, all annual re-measured FIA plots that are classified as Forest Land
Converted to Settlements are identified in each state, and C density estimates before conversion are compiled for
aboveground biomass, belowground biomass, dead wood, and litter. However, there are exceptions for the
Intermountain Region of the Western United States (Arizona, Colorado, Idaho, Montana, New Mexico, Nevada, and
Utah), in which there are a small number of plots that are converted from Forest Land to other Land Uses. In this
region, all plots identified as a conversion from forest land to another land use are grouped and used to estimate the
C densities before conversion, rather than subdividing the plots into specific land use change categories.
Furthermore, there are no re-measured annual plots in Wyoming, and so the C densities before conversion are based
on data from Colorado, Idaho, Montana, and Utah.
The C density before conversion is estimated for aboveground biomass, belowground biomass, dead wood, and litter
C pools. Soil C stock changes are also addressed, but are based on methods discussed in the next section. Individual
6-118 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
tree aboveground and belowground C density estimates are based on Woodall et al. (2011). The estimates of
aboveground and belowground biomass includes live understory species (i.e., undergrowth plants in a forest)
comprised of woody shrubs and trees less than 2.54 cm in diameter at breast height. It is assumed that 10 percent of
total understory C mass is belowground (Smith et al. 2006). Estimates of C density are derived from information in
Birdsey (1996) and Jenkins et al. (2003). The C density before conversion for standing dead trees 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). Downed dead wood is defined as pieces of dead
wood greater than 7.5 cm diameter at transect intersections that are not attached to live or standing dead trees, and
includes stumps and roots of harvested trees. The C density before conversion for downed dead wood is estimated
based on measurements of downed dead wood of a subset of FIA plots (Domke et al. 2013; Woodall and Monleon
2008), and models specific to regions and forest types within each region are used to estimate dead wood C
densities. Litter C is the pool of decaying leaves and woody fragments with diameters of up to 7.5 cm that are above
the mineral soil (also known as duff, humus, and fine woody debris). A subset of FIA plots are measured for litter C,
and a modeling approach is used to estimate litter C density based on the measurements (Domke et al. 2016). See
Annex 3.13 for more information about initial C density estimates for Forest Land.
In all states, the initial C in the forest land before conversion to settlements is assumed to be lost to the atmosphere
in the year of the conversion (i.e., 0 tonnes dry matter ha1 immediately after conversion), which is consistent with
the Tier 1 method in the IPCC guidelines (IPCC 2006). It is also assumed that the accumulation of new biomass,
dead wood and litter is negligible in the new settlement area.69 Therefore, total emissions and removals are
estimated based solely on the loss of all C existing on the forest land before conversion.
Soil Carbon Stock Changes
Soil C stock changes are estimated for Land Converted to Settlements according to land-use histories recorded in the
2012 USDA NRI survey for non-federal lands (USDA-NRCS 2015). Land use and some management information
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 2012 (USDA-NRCS
2015). However, this Inventory only uses NRI data through 2012 because newer data were not available.
NRI survey locations are classified as Land Converted to Settlements in a given year between 1990 and 2012 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 grassland between 1971 and 1978. For federal
lands, the land use history is derived from land cover changes in the National Land Cover Dataset (Homer et al.
2007; Fry et al. 2011; Homer et al. 2015).
Mineral Soil Carbon Stock Changes
An IPCC Tier 2 method (Ogle et al. 2003) is applied to estimate C stock changes for Land Converted to Settlements
on mineral soils from 1990 to 2012. 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 (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 provide a more robust
sample for estimating the reference condition. U.S.-specific C stock change factors are derived from published
literature to determine the impact of management practices on SOC 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 2012 NRI survey data (USDA-NRCS 2015) 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
69 C accumulation in woody biomass following conversion of lands to settlements is included in Section 6.10 Settlements
Remaining Settlements: Changes in Carbon Stocks in Urban Trees.
Land Use, Land-Use Change, and Forestry 6-119
-------
1 represent the loss of soil C with conversion to settlements, which is similar to the estimated losses with conversion
2 to cropland. More specific factor values can be derived in future inventories as data become available. See Annex
3 3.12 for additional discussion of the Tier 2 methodology for mineral soils.
4 A linear extrapolation of the trend in the time series is applied to estimate soil C stock changes from 2013 to 2016
5 because NRI activity data are not available for these years. Specifically, a linear regression model with
6 autoregressive moving-average (ARMA) errors (Brockwell and Davis 2016) is used to estimate the trend in stock
7 changes over time from 1990 to 2012, and in turn, the trend is used to approximate stock changes from 2013 to
8 2016. The Tier 2 method described previously will be applied to recalculate the 2013 to 2016 emissions in a future
9 Inventory.
10 Organic Soil Carbon Stock Changes
11 Annual C emissions from drained organic soils in Land Converted to Settlements are estimated using the Tier 2
12 method provided in IPCC (2006). The Tier 2 method assumes that organic soils are losing C at a rate similar to
13 croplands, and therefore uses the country-specific values for cropland (Ogle et al. 2003). To estimate CO2 emissions
14 from 1990 to 2012, the total area of organic soils in Land Converted to Settlements is multiplied by the Tier 2
15 emission factor, which is 11.2 MT C per ha in cool temperate regions, 14.0 MT C per ha in warm temperate regions
16 and 14.3 MT C per ha in subtropical regions (See Annex 3.12 for more information). Similar to the mineral soil C
17 stocks changes, a linear extrapolation of the trend in the time series is applied to estimate the emissions from 2013 to
18 2016 because NRI activity data are not available for these years to determine the area of Land Converted to
19 Settlements.
20 Uncertainty and Time-Series Consistency
21 The uncertainty analysis for C losses with Forest Land Converted to Settlements is conducted in the same way as the
22 uncertainty assessment for forest ecosystem C flux in the Forest Land Remaining Forest Land category. Sample and
23 model-based error are combined using simple error propagation methods provided by the IPCC (2006), i.e., by
24 taking the square root of the sum of the squares of the standard deviations of the uncertain quantities. For additional
25 details see the Uncertainty Analysis in Annex 3.13. The uncertainty analysis for mineral soil C stock changes and
26 annual C emission estimates from drained organic soils in Land Converted to Settlements is estimated using a Monte
27 Carlo approach, which is also described in the Cropland Remaining Cropland section.
28 Uncertainty estimates are presented in Table 6-86 for each subsource (i.e., biomass C stocks, mineral soil C stocks
29 and organic soil C stocks) and the method applied in the inventory analysis (i.e., Tier 2 and Tier 3). Uncertainty
30 estimates from the Tier 2 and 3 approaches are combined using the simple error propagation methods provided by
31 the IPCC (2006), i.e., as described in the previous paragraph. There are also additional uncertainties propagated
32 through the analysis associated with the data splicing methods applied to estimate soil C stock changes from 2013 to
33 2016. The combined uncertainty for total C stocks in Land Converted to Settlements ranges from 29 percent below
34 to 29 percent above the 2016 stock change estimate of 68.0 MMT CO2 Eq.
35 Table 6-86: Approach 2 Quantitative Uncertainty Estimates for Soil, Dead Organic Matter
36 and Biomass C Stock Changes occurring within Land Converted to Settlements (MMT CO2 Eq.
37 and Percent)
2016 Flux Estimate Uncertainty Range Relative to Flux Estimate3
Source (MMT CO2 Eq.) (MMT CO2 Eq.) (%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Cropland Converted to Settlements
10.1
8.1
12.2
-20%
20%
Mineral Soil C Stocks
9.3
7.2
11.3
-22%
22%
Organic Soil C Stocks
0.9
0.6
1.1
-32%
32%
Forest Land Converted to Settlements
44.8
25.3
64.2
-43%
43%
Aboveground Biomass C Stocks
32.7
13.7
51.8
-58%
58%
Belowground Biomass C Stocks
6.6
2.8
10.5
-58%
58%
Dead Wood
2.2
0.9
3.5
-58%
58%
Litter
2.0
1.5
2.4
-22%
22%
Mineral Soil C Stocks
1.3
1.0
1.5
-18%
18%
6-120 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
Organic Soil C Stocks
+
+
+
-35%
35%
Grassland Converted to Settlements
12.3
10.0
14.6
-18%
18%
Mineral Soil C Stocks
11.4
9.1
13.6
-20%
20%
Organic Soil C Stocks
0.9
0.6
1.2
-36%
36%
Other Lands Converted to Settlements
0.7
0.6
0.9
-21%
21%
Mineral Soil C Stocks
0.6
0.5
0.7
-22%
22%
Organic Soil C Stocks
0.1
+
0.2
-71%
71%
Wetlands Converted to Settlements
0.1
0.0
0.1
-37%
37%
Mineral Soil C Stocks
0.1
0.0
0.1
-37%
37%
Organic Soil C Stocks
0.0
0.0
0.0
0%
0%
Total: Land Converted to Settlements
68.0
48.3
87.7
-29%
29%
Aboveground Biomass C Stocks
32.7
13.7
51.8
-58%
58%
Belowground Biomass C Stocks
6.6
2.8
10.5
-58%
58%
Dead Wood
2.2
0.9
3.5
-58%
58%
Litter
2.0
1.5
2.4
-22%
22%
Mineral Soil C Stocks
22.6
19.5
25.6
-13%
13%
Organic Soil C Stocks
1.9
1.2
2.5
-34%
33%
+ Does not exceed 0.05 MMT CO2 Eq.
a Range of C stock change estimates predicted by Monte Carlo Stochastic Simulation for a 95 percent confidence interval.
1 Methodological recalculations are applied to the latter part of the time series (2013 to 2015) using the linear time
2 series model described above. Details on the emission trends through time are described in more detail in the
3 Methodology section, above.
4 QA/QC and Verification
5 Quality control measures included checking input data, model scripts, and results to ensure data are properly
6 handled throughout the inventory process. Inventory reporting forms and text are reviewed and revised as needed to
7 correct transcription errors.
8 Recalculations Discussion
9 Methodological recalculations are associated with extending the time series from 2013 through 2016 using a linear
10 time series model. The recalculation had a minor effect on the time series overall with C stock changes declining by
11 less than 1 percent on average.
12 Planned Improvements
13 A planned improvement for the Land Converted to Settlements category is to develop an inventory of C stock
14 changes in Alaska. This includes C stock changes for biomass, dead organic matter and soils. There are also plans to
15 extend the Inventory to included C losses associated with drained organic soils in settlements occurring on federal
16 lands. New land representation data will also be compiled, and the time series recalculated for the latter years in the
17 time series that are estimated using data splicing methods in this Inventory.
is 6.12 Other Land Remaining Other Land (CRF
19 Category 4F1)
20 Land use is constantly occurring, and areas under a number of differing land-use types remain in their respective
21 land-use type each year, just as other land can remain as other land. While the magnitude of Other Land Remaining
22 Other Land is known (see Table 6-7), research is ongoing to track C pools in this land use. Until such time that
23 reliable and comprehensive estimates of C for Other Land Remaining Other Land can be produced, it is not possible
24 to estimate CO2, CH4 or N2O fluxes on Other Land Remaining Other Land at this time.
Land Use, Land-Use Change, and Forestry 6-121
-------
1 6.13 Land Converted to Other Land (CRF
2 Category 4F2)
3 Land-use change is constantly occurring, and areas under a number of differing land-use types are converted to other
4 land each year, just as other land is converted to other uses. While the magnitude of these area changes is known
5 (see Table 6-7), research is ongoing to track C across Other Land Remaining Other Land and Land Converted to
6 Other Land. Until such time that reliable and comprehensive estimates of C across these land-use and land-use
7 change categories can be produced, it is not possible to separate CO2, CH4 or N20 fluxes on Land Converted to
8 Other Land from fluxes on Other Land Remaining Other Land at this time.
6-122 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
7. Waste
Waste management and treatment activities are sources of greenhouse gas emissions (see Figure 7-1). Landfills
accounted for approximately 16.4 percent of total U.S. anthropogenic methane (CH4) emissions in 2016, the third
largest contribution of any CH4 source in the United States. Additionally, wastewater treatment and composting of
organic waste accounted for approximately 2.3 percent and 0.3 percent of U.S. CH4 emissions, respectively. Nitrous
oxide (N20) emissions from the discharge of wastewater treatment effluents into aquatic enviromnents were
estimated, as were N20 emissions from the treatment process itself. Nitrous oxide emissions from composting were
also estimated. Together, these waste activities account for 1.9 percent of total U.S. N20 emissions. Nitrogen oxides
(NOx), carbon monoxide (CO), and non-CH4 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.
Figure 7-1: 2016 Waste Chapter Greenhouse Gas Sources (MMT CO2 Eq.)
Landfills
Wastewater Treatment
Composting
108
Waste as a Portion of all Emissions
2.0%
1
0 10 20 30 40 50 60 70 80 90 100 110
MMT CO: Eq.
Overall, in 2016, waste activities generated emissions of 131.5 MMT CO2 Eq., or 2.0 percent of total U.S.
greenhouse gas emissions.
Table 7-1: Emissions from Waste (MMT CO2 Eq.)
Gas/Source
1990
2005
2012
2013
2014
2015
2016
CH4
195.6
150.4
134.0
130.2
129.8
128.9
124.6
Landfills
179.6
132.7
117.0
113.3
112.7
111.7
107.7
Wastewater Treatment
15.7
15.8
15.1
14.9
15.0
15.1
14.8
Composting
0.4
1.9
1.9
2.0
2.1
2.1
2.1
N2O
3.7
6.1
6.4
6.5
6.7
6.7
6.8
Wastewater Treatment
3.4
4.4
4.6
4.7
4.8
4.8
5.0
Composting
0.3
1.7
1.7
1.8
1.9
1.9
1.9
Total
199.3
156.4
140.4
136.7
136.5
135.6
131.5
Note: Totals may not sum due to independent rounding.
Waste 7-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Table 7-2: Emissions from Waste (kt)
Gas/Source
1990
2005
2012
2013
2014
2015
2016
CH4
7,825
6,016
5,361
5,208
5,190
5,155
4,984
Landfills
7,182
5,310
4,680
4,531
4,509
4,467
4,306
Wastewater Treatment
627
631
604
596
598
605
593
Composting
15
75
77
81
84
84
85
N2O
12
20
21
22
23
23
23
Wastewater Treatment
11
15
16
16
16
16
17
Composting
1
6
6
6
6
6
6
Note: Totals may not sum due to independent rounding.
Carbon dioxide (CO2), CH4, and N20 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 2016 resulted in 11.0 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 7.4.
I
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 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 sinks provided in this Inventory do 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.
Box 7-2: Waste Data from EPA's Greenhouse Gas Reporting Progra
I
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). The rule 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. Data reporting by affected facilities includes the
reporting of emissions from fuel combustion at that affected facility. In general, the threshold for reporting is
25,000 metric tons or more of CO2 Eq. per year.
EPA presents the data collected by its GHGRP through a data publication tool that allows data to be viewed in
several formats including maps, tables, charts and graphs for individual facilities or groups of facilities.1
1 See .
7-2 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
EPA's GHGRP dataset and the data presented in this Inventory 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. 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 categories to improve the national estimates presented in this
Inventory consistent with IPCC guidelines. Within the Waste Chapter, EPA uses directly reported GHGRP data
for net CH4 emissions from MSW landfills for the years 2010 to 2016 of the Inventory. This data is also used to
back-cast emissions from MSW landfills for the years 2005 to 2009.
Landfills (CRF Source Category 5
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
commonly 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 7-3. 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 or 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).
Methane and CO2 are the primary constituents of landfill gas generation and emissions. However, the 2006 IPCC
Guidelines set an international convention to not report biogenic CO2 from activities in the Waste sector (IPCC
2006). Net carbon dioxide flux from carbon stock changes in landfills are estimated and reported under the Land
Use, Land-Use Change, and Forestry (LULUCF) sector (see Chapter 6 of this Inventory). Additionally, emissions of
NMOC and VOC are not estimated because they are emitted in trace amounts. Nitrous oxide (N20) 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 N20 emissions. Furthermore, the 2006
IPCC Guidelines did not include a methodology for estimating N20 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.
Waste 7-3
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Methane generation and emissions from landfills are a function of several factors, including: (1) the total amount 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., composition of waste-in-place, size, climate, cover material); (3)
the amount of CH4 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 commonly 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 60 years or longer as the degradable waste
decomposes over time.
In 2016, landfill CH4 emissions were approximately 107.7 MMT CO2 Eq. (4,306 kt), representing the third largest
source of CH4 emissions in the United States, behind natural gas systems and enteric fermentation. Emissions from
MSW landfills accounted for approximately 95 percent of total landfill emissions, while industrial waste landfills
accounted for the remainder. Estimates of operational MSW landfills in the United States have ranged from 1,900
to 2,000 facilities (EPA 2017a; EPA 2017b; Waste Business Journal (WBJ) 2016; WBJ 2010). More recently, the
Environment Research & Education Foundation (EREF) conducted a nationwide analysis of MSW management,
and counted 1,540 operational MSW landfills in 2013 (EREF 2016). Conversely, there are approximately 3,200
MSW landfills in the United States that have been closed since 1980 (for which a closure data is known, [EPA
2017a; WBJ 2010]). While the number of active MSW landfills has decreased significantly over the past 20 years,
from approximately 6,326 in 1990 to 1,540 in the 2013, the average landfill size has increased (EREF 2016; EPA
2017b; BioCycle 2010). With regard to industrial waste landfills, the WBJ database (WBJ 2016) includes
approximately 1,200 landfills accepting industrial and/or construction and demolition debris for 2016 (WBJ 2016).
Only 172 facilities with industrial waste landfills met the reporting threshold under Subpart TT (Industrial Waste
Landfills) of EPA's Greenhouse Gas Reporting Program (GHGRP), indicating that there may be several hundred
industrial waste landfills that are not required to report under EPA's GHGRP.
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 and then decreased by 11 percent to 205 MMT in 2016 (see
Annex 3.14, Table A-251). 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 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 2016
(see Annex 3.14, Table A-251). Net CH4 emissions from MSW landfills have decreased since 1990 (see Table 7-3
and Table 7-4).
In 2016, an estimated 32 new landfill gas-to-energy (LFGE) projects (EPA 2017a) 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.
Landfill gas collection and control is not accounted for at industrial waste landfills in this chapter (see the
Methodology discussion for more information).
Table 7-3: ChU Emissions from Landfills (MMT CO2 Eq.)
Activity
1990
2005
2012
2013
2014
2015
2016
MSW CH4 Generation
205.3
-
-
-
-
-
-
Industrial CH4 Generation
12.1
15.9
16.5
16.5
16.6
16.6
16.6
MSW CH4 Recovered
(17.9)
-
-
-
-
-
-
MSW CH4 Oxidized
(18.7)
-
-
-
-
-
-
Industrial CH4 Oxidized
(1.2)
(1.6)
(1.6)
(1.7)
(1.7)
(1.7)
(1.7)
7-4 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
MSW net CH4 Emissions
(GHGRP)
118.4
102.2 98.4 97.8 96.7 92.7
Total 179.6
132.7
117.0 113.3 112.7 111.7 107.7
Not applicable due to methodology change.
Note: Totals may not sum due to independent rounding. Parentheses indicate negative values. For years 1990 to 2004,
the Inventory methodology uses the first order decay methodology. A methodological change occurs in year 2005. For
years 2005 to 2016, directly reported net CH4 emissions from the GHGRP data are used plus a scale-up factor to
account for emissions from landfill facilities that are not subject to the GHGRP. These data incorporate CH4 recovered
and oxidized. As such, CH4 generation, CH4 recovery, and CH4 oxidized are not calculated separately for 2005-2016.
See the Time-Series Consistency section of this chapter for more information.
1 Table 7-4: ChU Emissions from Landfills (kt)
Activity
1990
2005
2012
2013
2014
2015
2016
MSW CH4 Generation
8,214
-
-
-
-
-
-
Industrial CH4 Generation
484
636
659
661
662
663
664
MSW CH4 Recovered
(718)
-
-
-
-
-
-
MSW CH4 Oxidized
(750)
-
-
-
-
-
-
Industrial CH4 Oxidized
MSW net CH4 Emissions
(48)
(64)
(66)
(66)
(66)
(66)
(66)
(GHGRP)
-
4,737
4,087
3,936
3,913
3,870
3,708
Total
7,182
5,310
4,680
4,531
4,509
4,467
4,306
Not applicable due to methodology change.
Note: Totals may not sum due to independent rounding. Parentheses indicate negative values. For years 1990 to 2004,
the Inventory methodology uses the first order decay methodology. A methodological change occurs in year 2005. For
years 2005 to 2016, directly reported net CH4 emissions from the GHGRP data are used plus a scale-up factor to
account for emissions from landfill facilities that are not subject to the GHGRP. These data incorporate CH4 recovered
and oxidized. As such, CH4 generation and CH4 recovery are not calculated separately. See the Time-Series
Consistency section of this chapter for more information.
2 Methodology
3 Methodology Applied for MSW Landfills
4 Methane emissions from landfills can be estimated using two primary methods. The first method uses the first order
5 decay (FOD) model as described by the 2006IPCC Guidelines to estimate CH4 generation. The amount of CH4
6 recovered and combusted from MSW landfills is subtracted from the CH4 generation, and is then adjusted with an
7 oxidation factor. The oxidation factor represents the amount of CH4 in a landfill that is oxidized to CO2 as it passes
8 through the landfill cover (e.g., soil, clay, geomembrane). This method is presented below, and is similar to
9 Equation HH-5 in CFR Part 98.343 for MSW landfills, and Equation TT-6 in CFR Part 98.463 for industrial waste
10 landfills.
11 CH4 ,SolidWaste— [CH4,MSW ~l~ CH4,Ind R.] Ox
12 where,
13 CH4,solid waste = Net CH | emissions from solid w aste
14 CH i.msw = CH4 generation from MSW landfills
15 CH4jnd = CH4 generation from industrial waste landfills
16 R = CH4 recovered and combusted (only for MSW landfills)
17 Ox = CH4 oxidized from MSW and industrial waste landfills before release to the atmosphere
18 The second method used to calculate CH4 emissions from landfills, also called the back-calculation method, is based
19 on directly measured amounts of recovered CH4 from the landfill gas and is expressed below and by Equation HH-8
20 in CFR Part 98.343. The two parts of the equation consider the portion of CH4inthe landfill gas that is not collected
Waste 7-5
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
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.
CH4,Solid Waste = [( r) x(l - OX) + R x (l - (DE x fDest))\
\CE X f REC J v J
where,
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)
fix,,i = fraction of hours the destruction device was operating (fraction)
The current Inventory uses both methods to estimate CH4 emissions across the time series. 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., directly reported CH4 emissions 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 2006IPCC Guidelines. Additional details on the technique used is included in a technical
memorandum (RTI2017).
A summary of the methodology used to generate the current 1990 through 2016 Inventory estimates for MSW
landfills is as follows and also illustrated in Annex Figure A-18:
• 1940 through 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 included in the FOD model for completeness in accounting for CH4 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 dumps (MCF
of 0.6). All calculations after 1980 assume waste is disposed in managed, anaerobic landfills. The FOD
method was applied to estimate annual CH4 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.
• 1990 through 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 2008 were obtained from the State of Garbage
(SOG) survey every two years (i.e., 2002, 2004, and 2006 as published inBioCycle 2006, and 2008 as
published in BioCycle 2010). 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 CH4 generation. Landfill-specific CH4
recovery amounts were then subtracted from CH4 generation and the result was then adjusted with a 10
percent oxidation factor to derive the net emissions estimates.
• 2005 through 2009: Emissions for these years are estimated using net CH4 emissions that are reported by
landfill facilities under EPA's GHGRP. Because not all landfills in the United States are required to report
7-6 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
to EPA's GHGRP, a 9 percent scale-up factor is applied to the GHGRP emissions for completeness.
Supporting information, including details on the technique used to estimate emissions for 2005 to 2009 and
to ensure time-series consistency by incorporating the directly reported GHGRP emissions is presented in
Annex 3.14 and in RTI 2018 (memorandum in progress). 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
depending on their facility-specific calculated CH4flux rate (i.e., 0, 10, 25, or 35 percent). The average
oxidation factor from the GHGRP facilities is 19.5 percent.
• 2010 through 2016: Directly reported net CH4 emissions 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. As noted above,
GHGRP facilities use a variety of oxidation factors; therefore, the average oxidation factor from the
GHGRP facilities is 19.5 percent.
A detailed discussion of the data sources and methodology used to calculate CH4 generation and recovery is
provided below. Supporting information, including details on the technique used to ensure time-series consistency
by incorporating the directly reported GHGRP emissions is presented in the Time-Series Consistency section of this
chapter and in Annex 3.14.
Description of the Methodology for MSW Landfills as Applied for 1990-2004
National MSWMethane Generation and Disposal Estimates
States and local municipalities across the United States do not consistently track and report quantities of MSW
generated or collected for management, nor do they report end-of-life disposal methods to a centralized system.
Therefore, national MSW landfill waste generation and disposal data are obtained from secondary data, specifically
the SOG surveys, published approximately every two years, with the most recent publication date of 2014. The
SOG survey was the only continually updated nationwide survey of waste disposed in landfills in the United States
and was the primary data source with which to estimate nationwide CH4 generation from MSW landfills. Now,
EPA's GHGRP waste disposal data and MSW management data published by EREF are available.
The SOG surveys collect data from the state agencies and then apply the principles of mass balance where all MSW
generated is equal to the amount of MSW landfilled, combusted in waste-to-energy plants, composted, and/or
recycled (BioCycle 2010; Shin 2014). This approach assumes that all waste management methods are tracked and
reported to state agencies. Survey respondents are asked to provide a breakdown of MSW generated and managed
by landfilling, recycling, composting, and combustion (in waste-to-energy facilities) in actual tonnages as opposed
to reporting a percent generated under each waste disposal option. The data reported through the survey have
typically been adjusted to exclude non-MSW materials (e.g., industrial and agricultural wastes, construction and
demolition debris, automobile scrap, and sludge from wastewater treatment plants) that may be included in survey
responses. While these wastes may be disposed of in MSW landfills, they are not the primary type of waste material
disposed and are typically inert. In the most recent survey, state agencies were asked to provide already filtered,
MSW-only data. Where this was not possible, they were asked to provide comments to better understand the data
being reported. All state disposal data are adjusted for imports and exports across state lines where imported waste
is included in a state's total while exported waste is not. Methodological changes have occurred over the time frame
the SOG survey has been published, and this has affected the fluctuating trends observed in the data (RTI 2013).
State-specific landfill MSW generation data and a national average disposal factor for 1989 through 2008 were
obtained from the SOG survey every two years (i.e., 2002, 2004, and 2006 as published in BioCycle 2006, and 2008
as published in BioCycle 2010). The landfill inventory calculations start with hard numbers (where available) as
presented in the SOG documentation for the report years 2002, 2004, 2006, and 2008. In-between year waste
generation is interpolated using the prior and next SOG report data. For example, waste generated in 2003 = (waste
generation in 2002 + waste generation in 2004)/2. The quantities of waste generated across all states are summed
and that value is then used as the nationwide quantity of waste generated in each year of the time series. The SOG
Waste 7-7
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
survey is voluntary and not all states provide data in each survey year. To estimate waste generation for states that
did not provide data in any given reporting year, one of the following methods was used (RTI2013):
• For years when a state-specific waste generation rate was available from the previous SOG reporting year
submission, the State-specific waste generation rate for that particular state was used.
— or —
• For years where a state-specific waste generation rate was not available from the previous SOG reporting
year submission, the waste amount is generated using the national average waste generation rate. In other
words, Waste Generated = Reporting Year Population x the National Average Waste Generation Rate
o The National Average Waste Generation Rate is determined by dividing the total reported waste
generated across the reporting states by the total population for reporting States,
o This waste generation rate may be above or below the waste generation rate for the non-reporting
states and contributes to the overall uncertainty of the annual total waste generation amounts used
in the model.
Use of these methods to estimate solid waste generated by states is a key aspect of how the SOG data was
manipulated and why the results differ for total solid waste generated as estimated by SOG and in the Inventory. In
the early years (2002 data in particular), SOG made no attempt to fill gaps for non-survey responses. For the 2004
data, the SOG team used proxy data (mainly from the WB J) to fill gaps for non-reporting states and survey
responses.
Another key aspect of the SOG survey is that it focuses on MSW-only data. The data states collect for solid waste
typically are representative of total solid waste and not the MSW-only fraction. In the early years of the SOG
survey, most states reported total solid waste rather than MSW-only waste. The SOG team, in response, "filtered"
the state-reported data to reflect the MSW-only portion.
Also important to note about this data source is that the waste generation data reported by states to the SOG survey
fluctuates from year to year. Although some fluctuation is expected, for some states, the year-to-year fluctuations are
quite significant (>20 percent increase or decrease in some case) (RTI 2013). The SOG survey reports for these
years do not provide additional explanation for these fluctuations and the source data are not available for further
assessment. Although exact reasons for the large fluctuations are difficult to obtain without direct communication
with states, staff from the SOG team that were contacted speculate that significant fluctuations are present because
the particular state could not gather complete information for waste generation (i.e., they are missing part of
recycled and composted waste data) during a given reporting year. In addition, SOG team staff speculated that some
states may have included C&D and industrial wastes in their previous MSW generation submissions, but made
efforts to exclude that (and other non-MSW categories) in more recent reports (RTI 2013).
The most recent SOG survey provides data for 2011 (Shin 2014). The EREF published a report, MSW Management
in the United States, which includes state-specific landfill MSW generation and disposal data for 2010 and 2013
using a similar methodology as the SOG surveys (EREF 2016). State-specific landfill waste generation data for the
years in between the SOG surveys and EREF report (e.g., 2001, 2003, 2005, 2007, and 2009) were either
interpolated or extrapolated based on the SOG or EREF data and the U.S. Census population data (U.S. Census
Bureau 2016). In the current Inventory methodology, the MSW generation and disposal data are no longer used to
estimate CH4 emissions for the years 2005 to 2016 because EPA's GHGRP emissions data are now used for those
years. The MSW generation and disposal data for these years are still useful for examining general trends in MSW
management in the United States.
Estimates of the quantity of waste landfilled from 1989 to 2004 are determined by applying an average national
waste disposal factor to the total amount of waste generated (i.e., the SOG data). A national average waste disposal
factor is determined for each year an SOG survey is published and equals the ratio of the total amount of waste
landfilled in the United States to the total amount of waste generated in the United States. The waste disposal factor
is interpolated or extrapolated for the years in-between the SOG surveys, as is done for the amount of waste
generated for a given survey year.
The 2006IPCC Guidelines recommend at least 50 years of waste disposal data to estimate CH4 emissions. 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
7-8 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
contributes very little to current CH4 generation, estimates for those years were included in the FOD model for
completeness in accounting for CH4 generation rates and are based on the population in those years and the per
capita rate for land disposal for the 1960s. For calculations in the current Inventory, wastes landfilled prior to 1980
were broken into two groups: wastes disposed in landfills (MCF of 1) and those disposed in dumps (MCF of 0.6).
All calculations after 1980 assume waste is disposed in managed, modern landfills. See Annex 3.14 for more
details.
National Landfill Gas Recovery Estimates for MSW Landfills
The estimated landfill gas recovered per year (R) at MSW landfills for 1990 to 2004 was based on a combination of
four databases and including recovery from flares and/or landfill gas-to-energy (LFGE) projects:
• EPA's GHGRP dataset for MSW landfills (EPA 2015a);2
• A database developed by the Energy Information Administration (EIA) for the voluntary reporting of
greenhouse gases (EIA 2007);
• A database of LFGE projects that is primarily based on information compiled by the EPA LMOP (EPA
2016b);3 and
• The flare vendor database (contains updated sales data collected from vendors of flaring equipment).
The same landfill may be included one or more times across these four databases. To avoid double- or triple-
counting CH4 recovery, the landfills across each database were compared and duplicates identified. A hierarchy of
recovery data is used based on the certainty of the data in each database. In summary, the GHGRP > EIA > LFGE >
flare vendor database. The rationale for this hierarchy is described below.
EPA's GHGRP MSW landfills database was first introduced as a data source for landfill gas recovery in the 1990 to
2013 Inventory. EPA's GHGRP MSW landfills database contains facility-reported data that undergoes rigorous
verification, thus it is considered to contain the least uncertain data of the four CH4 recovery databases. However, as
mentioned earlier, this database is unique in that it only contains a portion of the landfills in the United States
(although, presumably the highest emitters since only those landfills that meet a certain CH4 generation threshold
must report) and only contains data for 2010 and later. In the current Inventory methodology, CH4 recovery for
1990 to 2004 for facilities reporting to EPA's GHGRP has been estimated using the directly reported emissions for
those facilities from 2010 to 2015, and an Excel forecasting function so that the GHGRP data source can be applied
to earlier years in the time series. Prior to 2005, if a landfill in EPA's GHGRP was also in the LFGE or EIA
databases, the landfill gas project information, specifically the project start year, from either the LFGE or EIA
databases was used as the cutoff year for the estimated CH4 recovery in the GHGRP database. For example, if a
landfill reporting under EPA's GHGRP was also included in the LFGE database under a project that started in 2002
that is still operational, the CH4 recovery data in the GHGRP database for that facility was back-calculated to the
year 2002 only.
If a landfill in the GHGRP MSW landfills database was also in the EIA, LFGE, and/or flare vendor database, the
avoided emissions were only based on EPA's GHGRP MSW landfills database to avoid double or triple counting
the recovery amounts. In other words, the CH4 recovery from the same landfill was not included in the total
recovery from the EIA, LFGE, or flare vendor databases.
If a landfill in the EIA database was also in the LFGE and/or the flare vendor database, the CH4 recovery was based
on the EIA data because landfill owners or operators directly reported the amount of CH4 recovered using gas flow
concentration and measurements, and because the reporting accounted for changes over time.
If both the flare data and LFGE recovery data were available for any of the remaining landfills (i.e., not in the EIA
or GHGRP databases), then the avoided emissions were based on the LFGE data, which provides reported landfill-
specific data on gas flow for direct use projects and project capacity (i.e., megawatts) for electricity projects. The
LFGE database is based on the most recent EPA LMOP database (published annually). The remaining portion of
2 The 2015 GHGRP dataset is used to estimate landfill gas recovery from MSW landfills for the years 1990-2004 of the
Inventory. This database is no longer updated because the methodology has changed such that the directly reported net methane
emissions from the GHGRP are used and landfill gas recovery is not separately estimated.
3 The LFGE database was not updated for the 1990 to 2016 Inventory because the methodology does not use this database for
years 2005 and later, thus the LMOP 2016 database is the most recent year reflected in the LFGE database for the Inventory.
Waste 7-9
-------
1 avoided emissions is calculated by the flare vendor database, which estimates CH4 combusted by flares using the
2 midpoint of a flare's reported capacity. New flare vendor sales data have not been collected since 2015 because
3 these data are not used for estimates beyond 2005 in the time series. Given that each LFGE project is likely to also
4 have a flare, double counting reductions from flares and LFGE projects in the LFGE database was avoided by
5 subtracting emission reductions associated with LFGE projects for which a flare had not been identified from the
6 emission reductions associated with flares (referred to as the flare correction factor). A further explanation of the
7 methodology used to estimate the landfill gas recovered can be found in Annex 3.14.
8 A destruction efficiency of 99 percent was applied to CH4 recovered to estimate CH4 emissions avoided due to the
9 combusting of CH4 in destruction devices (i.e., flares) in the EIA, LFGE, and flare vendor databases. The median
10 value of the reported destruction efficiencies to the GHGRP was 99 percent for all reporting years (2010 through
11 2015). For the other three recovery databases, the 99 percent destruction efficiency value selected was based on the
12 range of efficiencies (86 to greater than 99 percent) recommended for flares in EPA's AP-42 Compilation of Air
13 Pollutant Emission Factors, Draft Section 2.4, Table 2.4-3 (EPA 2008). A typical value of 97.7 percent was
14 presented for the non-CH4 components (i.e., VOC and NMOC) in test results (EPA 2008). An arithmetic average of
15 98.3 percent and a median value of 99 percent are derived from the test results presented in EPA (2008). Thus, a
16 value of 99 percent for the destruction efficiency of flares has been used in the Inventory methodology. Other data
17 sources supporting a 99 percent destruction efficiency include those used to establish New Source Performance
18 Standards (NSPS) for landfills and in recommendations for shutdown flares used by the EPA LMOP.
19 National MSWLandfill Methane Oxidation Estimates
20 The amount of CH4 oxidized by the landfill cover at MSW landfills was assumed to be 10 percent of the CH4
21 generated that is not recovered (IPCC 2006; Mancinelli and McKay 1985; Czepiel et al. 1996) for the years 1990 to
22 2004.
23 National MSW Net Emissions Estimates
24 Net CH4 emissions are calculated by subtracting the CH4 recovered and CH4 oxidized from CH4 generated at MSW
25 landfills.
26 Description of the Methodology for MSW Landfills as Applied for 2005 to 2009
27 The Inventory methodology uses directly reported net CH4 emissions for the 2010 to 2016 reporting years from the
28 GHGRP to back-cast emissions for 2005 to 2009. The emissions for 2005 to 2009 are recalculated each year the
29 Inventory is published to account for the additional year of reported data and any revisions that facilities make to
30 past GHGRP reports. When EPA verifies the GHG reports, comparisons are made with data submitted in earlier
31 reporting years and errors may be identified in these earlier year reports. Facility representatives may submit revised
32 reports for any reporting year in order to correct these errors. Facilities reporting to the GHGRP that do not have
33 landfill gas collection and control systems use the FOD method. Facilities with landfill gas collection and control
34 must use both the FOD method and a back-calculation approach. The back-calculation approach starts with the
35 amount of CH4 recovered and works back through the system to account for gas not collected by the landfill gas
36 collection and control system (i.e., the collection efficiency).
37 A scale-up factor to account for emissions from landfills that do not report to EPA's GHGRP is also applied
38 annually. In theory, national MSW landfill emissions should equal the net CH4 emissions reported to the GHGRP
39 plus net CH4 emissions from landfills that do not report to the GHGRP. EPA estimated a scale-up factor of 9
40 percent. The rationale behind the 9 percent scale-up factor is included in Annex 3.14 and inRTI 2018
41 (memorandum in progress).
42 The GHGRP data allows facilities to apply a range of oxidation factors (0.0, 0.10, 0.25, or 0.35) based on the
43 calculated CH4 flux at the landfill. Therefore, one oxidation factor is not applied for this time frame, as is done for
44 1990 to 2004. The average oxidation factor across the GHGRP data is 19.5 percent.
7-10 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
Description of the Methodology for MSW Landfills as Applied for 2010 to 2016
Directly reported CH4 emissions to the GHGRP are used for 2010 to 2016 plus the 9 percent scale-up factor to
account for emissions from landfills that do not report to the GHGRP. The average oxidation factor across the
GHGRP data is 19.5 percent.
Description of the First Order Decay Methodology for Industrial Waste Landfills
Emissions from industrial waste landfills are estimated from industrial production data (ERG 2017), waste disposal
factors, and the FOD method. There are currently no 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 172 facilities that report to subpart TT of the GHGRP, 93 (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 (EPA 1993a), thus
estimates of industrial landfill emissions focused on these two sectors. 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. 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.
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 two of the 176 facilities, or 1 percent of facilities, have active
gas collection systems (EPA 2017b). 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
CH i generated (IPCC 2006; Mancinelli and McKay 1985; Czepiel et al. 1996) for all years.
Uncertainty and Time-Series Consistency
Several types of uncertainty are associated with the estimates of CH4 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 CH4 generation
potential (L0) and the rate of decay that produces CH4 from MSW, as determined from several studies of CH4
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 under-estimate CH4
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 also a high degree of uncertainty and variability 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 the GHGRP data because this methodology is
facility-specific, uses directly measured CH4 recovery data (when applicable), and allows for a variety of landfill gas
collection efficiencies, destruction efficiencies, and/or oxidation factors to be used. An uncertainty factor of 8
percent is applied to the directly reported CH4 emissions to EPA's GHGRP.
Uncertainty also exists in the scale-up factor applied for years 2005 to 2009 and in the back-casted emissions
estimates for 2005 to 2009. As detailed in RTI2018 (memorandum in progress), 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 equal the scale-up factor. Input
was provided by industry, LMOP, and additional EPA support. However, many gaps still exist and assumptions
were made for many landfills in order to estimate the scale-up factor. Additionally, a simple methodology was used
Waste 7-11
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
to back-cast emissions for 2005 to 2009 using the GHGRP emissions from 2010 to 2016. 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 emissions for 2005 to 2009.
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). This chapter
however, recommends against back-casting emissions back to 1990 with a limited set of data and instead provides
guidance on techniques to splice, or join methodologies together. 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 the GHGRP data, 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. Therefore, EPA decided to back-cast the GHGRP
emissions from 2009 to 2005 only, in order to merge the datasets and adhere to the IPCC good practice guidance.
Aside from the uncertainty in estimating landfill CH4 generation, uncertainty also exists in the estimates of the
landfill gas oxidized. Facilities directly reporting to the GHGRP can use oxidation factors ranging from 0 to 35
percent, depending on their facility-specific CHi 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 industrial waste landfills regardless of climate, the type of cover material, and/or presence of a
gas collection system. The number of published field studies measuring the rate of oxidation has increased
substantially since the 2006 IPCC Guidelines were published and, as discussed in the Potential Improvements
section, efforts are being made to review the literature and revise this value, as appropriate.
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
2004; whereas directly reported CH4 recovery is used for facilities reporting to the GHGRP for years 2005 to 2015.
The GHGRP MSW landfills database was added as a fourth recovery database starting with the 1990 through 2013
Inventory report. 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.
Additionally, 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.
The LFGE database was updated annually until 2015. The flare database was populated annually until 2015 by the
voluntary sharing of flare sales data by select vendors, which likely underestimated recovery for landfills not
included in the three other recovery databases used by the Inventory. The EIA database has not been updated since
2006 and has, for the most part, been replaced by the GHGRP MSW landfills database. 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.
7-12 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
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 CH4 estimated to be recovered). This 10 percent uncertainty factor applies to the LFGE database; 12 percent
to the EI A database; and 1 percent for the GHGRP MSW landfills dataset because of the supporting information
provided and rigorous verification process. For flaring without metered recovery data (the flare database), a much
higher uncertainty value of 50 percent is used. The compounding uncertainties associated with the four databases in
addition to the uncertainties associated with the FOD method and annual waste disposal quantities leads to the large
upper and lower bounds for MSW landfills presented in Table 7-5.
The lack of landfill-specific information regarding the number and type of industrial waste landfills in the United
States is a primary source of uncertainty with respect to the industrial waste generation and emission estimates. The
approach used here assumes that most of the organic waste disposed of in industrial waste landfills that would result
in CH4 emissions consists of waste from the pulp and paper and food processing sectors. However, because waste
generation and disposal data are not available in an existing data source for all U.S. industrial waste landfills, a
straight disposal factor is applied over the entire time series to the amount of waste generated 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 number of data sources used, each with
its own uncertainty factor.
Table 7-5: Approach 2 Quantitative Uncertainty Estimates for ChU Emissions from Landfills
(MMT CO2 Eq. and Percent)
2016 Emission
Source Gas Estimate Uncertainty Range Relative to Emission Estimate3
(MMT CO2 Eq.) (MMT CO2 Eq.) (%)
Lower
Upper
Lower
Upper
Bound
Bound
Bound
Bound
Total Landfills
CH4
107.7
82.6
131.9
-23%
23%
MSW
ch4
92.7
69.6
116.5
-25%
26%
Industrial
ch4
14.9
10.3
18.7
-31%
25%
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. QA/QC plan,
which is in accordance with Vol. 1 Chapter 6 of 2006IPCC Guidelines (see Annex 8 for more details). QA/QC
checks are performed for the transcription of the published data set (e.g., EPA's GHGRP dataset) used to populate
the Inventory data set in terms of completeness and accuracy against the reference source. Additionally, all datasets
used for this category have been checked to ensure they are of appropriate quality and are representative of U.S.
conditions. The primary calculation spreadsheet is tailored from the 2006 IPCC Guidelines waste model and has
been verified previously using the original, peer-reviewed IPCC waste model. All model input values and
calculations were verified by secondary QA/QC review. Stakeholder engagements sessions in 2016 and 2017 were
used to gather input on methodological improvements and facilitate an external expert review on the methodology,
activity data, and emission factors.
Category-specific checks include the following:
• Evaluation of the secondary data sources used as inputs to the Inventory dataset to ensure they are
appropriately collected and are reliable;
• Cross-checking the data (activity data and emissions estimates) with previous years to ensure the data are
reasonable, and that any significant variation can be explained through the activity data;
Waste 7-13
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
• 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 CH4 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. Based on the
results of the verification process, EPA follows up with facilities to resolve mistakes that may have occurred.4
Recalculations Discussion
Recalculations to the back-casted GHGRP emissions for 2005 to 2009 were performed, and the scale-up factor
applied to years 2005 to the current year (2016) was revised. These recalculations decreased net emissions for MSW
landfills from 2005 to 2015 when compared to the previous Inventory report.
First, the GHGRP data for all available years is used to back-cast emissions for 2005 to 2009. Revisions to the
individual facility reports submitted to the 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.
Second, the scale-up factor was revised from 12.5 percent to 9 percent. The scale-up factor to supplement directly
reported emissions from the GHGRP was first applied in the 1990 to 2015 Inventory. The initial scale-up factor of
12.5 percent was developed as a rough estimate with the intent of it being refined after engaging with stakeholders
and completing various data analyses. EPA has since investigated options to develop a more precise scale-up factor
to apply to the GHGRP data and has refined the scale-up factor to 9 percent as detailed in RTI2018 (memorandum
in progress). The revised scale-up factor was developed after extracting data on the first year of waste acceptance,
annual waste disposal, and total waste-in-place from the LMOP database and WBJ Directory 2016 for landfills that
are not reporting to the GHGRP. EPA created a database of non-reporting landfills and sought input from various
stakeholders (industry and LMOP). Stakeholders were asked to review the database of non-reporting landfills and to
provide input on the following: whether the landfill reported to the GHGRP (and reporting identification number), if
the landfill was considered an MSW or other landfill, whether the landfill was open or closed, first year of waste
acceptance, closure year (for closed landfills), estimated closure year (for active landfills), geographical coordinates
(latitude and longitude), and annual waste acceptance data or total waste-in-place.
The revised scale-up factor of 9 percent is based on the total waste-in-place from readily available information for
landfills that do not report to the GHGRP. It is the ratio of the "non-reporting landfills waste-in-place" to the sum
total of the GHGRP waste-in-place and the non-reporting landfills waste-in-place.
Planned Improvements
EPA has engaged in stakeholder outreach through a series of webinars since December 2016 to increase the
transparency in the Inventory methodology and to identify ideas and supplemental data sources that can lead to
methodological improvements. Three areas where EPA is actively working on improvements include the oxidation
factor for 1990 to 2004, the DOC value for 1990 to 2004, and the decay rate (k value) for 1990 to 2004.
4 See .
7-14 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
EPA is continuing to investigate options to adjust the oxidation factor from the 10 percent currently used for 1990 to
2004 to another value such as those included in the GHGRP (e.g., 25 percent, 35 percent), or other landfill gas
modeling frameworks. The oxidation factor currently applied in the later portion of the time series (2005 to 2016)
averages at 19.5 percent due to the use of the GHGRP data while the earlier portion of the time series applies the
default of 10 percent. EPA is investigating trends in landfill gas collection and control since 1990, as well as other
factors, and may apply different oxidation factors based on the percentage of waste disposed in landfills with gas
collection and without gas collection during 1990 to 2004.
The Inventory currently uses one value of 0.20 for the DOC for years 1990 to 2004. With respect to improvements
to the DOC value, EPA developed a database with MSW characterization data from individual studies across the
United States. EPA will review this data against the Inventory time series to assess the validity of the current DOC
value and how it is applied in the FOD method. Waste characterization studies vary greatly in terms of the
granularity of waste types included and the spatial boundaries of each study (e.g., one landfill, a metro area,
statewide). At this time, it is not expected that any evaluation of the DOC value will impact the emissions estimates
for 2005 and later in the time series as these emissions are taken from the directly reported GHGRP data.
EPA will review the k values for the three climate types (dry, moderate, and wet) against new data and other landfill
gas models, and how they are applied to the percentage of the population assigned to these climate types. EPA will
also assess the uncertainty factor applied to these k values in the Waste Model. A 30 percent uncertainty factor is
applied to each k value in the Monte Carlo analysis, which is consistent with that recommended by the IPCC (2006).
Box 7-3: Nationwide Municipal Solid Waste Data Sourc
M
Municipal solid waste generated in the United States can be managed through landfilling, recycling, composting,
and combustion with energy recovery. There are three main sources for nationwide solid waste management data in
the United States:
• The BioCycle and Earth Engineering Center of Columbia University's SOG in America surveys [no longer
published];
• The EPA's Advancing Sustainable Materials Management reports; and
• The EREF's MSW Generation in the United States reports.
The SOG surveys and, now EREF, collect state-reported data on the amount of waste generated and the amount of
waste managed via different management options: landfilling, recycling, composting, and combustion. The survey
asks 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 is not
available, the survey asks for total tons landfilled. The data are adjusted for imports and exports across state lines so
that the principles of mass balance are adhered to, whereby the amount of waste managed does 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 reports use a materials flow 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 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, composted,
or combusted is assumed to be landfilled. The data presented in the report are nationwide totals.
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. 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.
Waste 7-15
-------
1
Box 7-4: Overview of the Waste Sector
2 As shown in Figure 7-2 and Figure 7-3, landfilling of MSW is currently and has been the most common waste
3 management practice. A large portion of materials in the waste stream are recovered for recycling and composting,
4 which is becoming an increasingly prevalent trend throughout the country. Materials that are composted and
5 recycled would have previously been disposed in a landfill.
6 Figure 7-2: Management of Municipal Solid Waste in the United States, 2014
Recycled
25.7%
Landfilled
52.6%
Composted
8.9%
MSW to WTE
12.8%
7
8 Source: EPA (2016c) Note: 2014 is the latest year of available data.
9 Figure 7-3: MSW Management Trends from 1990 to 2014
160
Landfilling
140
120
100
80
60
Combustion with
Energy Recovery
40
20
Composting
0
0
01
criCTicricricricricrio
o o
rn m oo
o o o o o o o
o
tH
11 Source: EPA (2016c). Note: 2014 is the latest year of available data.
12 Table 7-6 presents a typical composition of waste disposed of at a typical MSW landfill in the United States over
13 time. It is important to note that the actual composition of waste entering each landfill will vary from that presented
7-16 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 in Table 7-6. Understanding how the waste composition changes over time, specifically for the degradable waste
2 types (i.e., those types known to generate CH4 as they break down in a modern MSW landfill), is important for
3 estimating greenhouse gas emissions. Increased diversion of degradable materials so that they are not disposed of in
4 landfills reduces the CH4 generation potential and CH4 emissions from landfills. For certain degradable waste types
5 (i.e., paper and paperboard), the amounts discarded have decreased over time due to an increase in waste diversion
6 through recycling and composting (see Table 7-6 and Figure 7-4). As shown in Figure 7-4, the diversion of food
7 scraps has been consistently low since 1990 because most cities and counties do not practice curbside collection of
8 these materials. Neither Table 7-6 nor Figure 7-4 reflect the frequency of backyard composting of yard trimmings
9 and food waste because this information is largely not collected nationwide and is hard to estimate.
10 Table 7-6: Materials Discarded3 in the Municipal Waste Stream by Waste Type from 1990 to
11 2014 (Percent)b
Waste Type
I'WO
2005
2010
2011°
2012
2013
2014
Paper and Paperboard
30.0%
24.7%
16.1%
14.7%
14.7%
15.0%
14.3%
Glass
6.0%
5.8%
5.1%
5.1%
5.2%
5.2%
5.2%
Metals
7.2%
7.9%
9.0%
8.9%
9.2%
9.5%
9.4%
Plastics
9.5%
16.4%
17.9%
17.9%
18.2%
18.4%
18.5%
Rubber and Leather
3.2%
2.9%
3.2%
3.8%
3.2%
3.1%
3.1%
Textiles
2.9%
5.3%
6.5%
6.8%
7.1%
7.4%
7.7%
Wood
6.9%
7.5%
8.2%
8.2%
8.2%
8.0%
8.1%
Other11
] .4%
1.8%
2.1%
2.0%
2.0%
1.9%
1.9%
Food Scraps
13.6%
18.5%
21.0%
21.4%
21.0%
21.0%
21.6%
Yard Trimmings
17.6%
7.0%
8.6%
8.8%
8.7%
8.1%
7.9%
Miscellaneous Inorganic
Wastes
] .7%
2.2%
2.3%
2.4%
2.4%
2.4%
2.3%
a Discards after materials and compost recovery. In this table, discards include combustion with energy recovery. Does not
include construction & demolition debris, industrial process wastes, or certain other wastes.
b Data for all years except 2011 are from the EPA's Advancing Sustainable Materials Management: 2014 Tables and Figures
report (Table 4) published in December 2016 (EPA 2016c).
c 2011 data are not included in the most recent Advancing Sustainable Materials Management report (2014), thus data from
the 2013 report (Table 3) was kept in place for 2011 (EPA 2015b).
d Includes electrolytes in batteries and fluff pulp, feces, and urine in disposable diapers. Details may not add to totals due to
rounding.
Note: 2014 is the latest year of available data.
Waste 7-17
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Figure 7-4:
(Percent)
Percent of Degradable Materials Diverted from Landfills from 1990 to 2014
80%
Paper and Paperboard
Food Scraps
Yard Trimmings
70%
60%
50%
40%
30%
20%
10%
0%
0
01
t—I
LO U3 r- 00 en o
Ol Ul Ol Ol Ol o
h in m
o o o o
^ ID N CO (J1 O
O O O O O
Source: (EPA 2016c). Note: 2014 is the latest year of available data.
Box 7-5: Description of a Modern, Managed Landfill
Modern, managed landfills are well-engineered facilities that are located, designed, operated, and monitored to
ensure compliance with federal, state, and tribal regulations. Municipal solid waste (MSW) landfills must be
designed to protect the enviromnent from contaminants which may be present in the solid waste stream.
Additionally, many new landfills collect and destroy landfill gas through flares or landfill gas-to-energy projects.
Requirements for affected MSW landfills may include:
• Siting requirements to protect sensitive areas (e.g., airports, floodplains, wetlands, fault areas, seismic
impact zones, and unstable areas);
• Design requirements for new landfills to ensure that Maximum Contaminant Levels (MCLs) will not be
exceeded in the uppermost aquifer (e.g., composite liners and leachate collection systems);
• Leachate collection and removal systems;
• Operating practices (e.g., daily and intermediate cover, receipt of regulated hazardous wastes, use of
landfill cover material, access options to prevent illegal dumping, use of a collection system to prevent
stonnwater 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. Additionally, state and
tribal requirements may exist.5
5 For more information regarding federal MSW landfill regulations, see
.
7-18 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
7.2 Wastewater Treatment (CRF Source
Category 5D)
Wastewater treatment processes can produce anthropogenic methane (CH4) and nitrous oxide (N20) emissions.
Wastewater from domestic and industrial sources is treated to remove soluble organic matter, suspended solids,
pathogenic organisms, and chemical contaminants.6 Treatment may either occur on site, most commonly through
septic systems or package plants, or off site at centralized treatment systems. In the United States, approximately 19
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 2015). Centralized wastewater treatment systems may include a variety of
processes, ranging from lagooning to advanced tertiary treatment technology for removing nutrients. Some
wastewater may also be treated through the use of constructed (or semi-natural) wetland systems, though it is much
less common in the United States (ERG 2016). Constructed wetlands may be used as the primary method of
wastewater treatment, or as a tertiary 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).
Soluble organic matter is generally removed using biological processes in which microorganisms consume the
organic matter for maintenance and growth. The resulting biomass (sludge) is removed from the effluent prior to
discharge to the receiving stream. Microorganisms can biodegrade soluble organic material in wastewater under
aerobic or anaerobic conditions, where the latter condition produces CH4. During collection and treatment,
wastewater may be accidentally or deliberately managed under anaerobic conditions. In addition, the sludge may be
further biodegraded under aerobic or anaerobic conditions. The generation of N20 may also result from the
treatment of domestic wastewater during both nitrification and denitrification of the nitrogen (N) present, usually in
the form of urea, ammonia, and proteins. These compounds are converted to nitrate (NO3) through the aerobic
process of nitrification. Denitrification occurs under anoxic conditions (without free oxygen), and involves the
biological conversion of nitrate into dinitrogen gas (N2). Nitrous oxide can be an intermediate product of both
processes, but has typically been associated with denitrification. Recent research suggests that higher emissions of
N20 may in fact originate from nitrification (Ahn et al. 2010). Other more recent research suggests that N20 may
also result from other types of wastewater treatment operations (Chandran 2012).
The principal factor in determining the CH4 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 CH4 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
BOD5. Because BOD is an aerobic parameter, it is preferable to use COD to estimate CH4 production, since CH4 is
produced only in anaerobic conditions. The principal factor in determining the N20 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 N20 generation potential.
In 2016, CH4 emissions from domestic wastewater treatment were 8.9 MMT C02 Eq. (357 kt CH4). Emissions
remained fairly steady from 1990 through 1999, 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 2004; U.S. Census Bureau 2015). In 2016, CH4 emissions
from industrial wastewater treatment were estimated to be 5.9 MMT C02 Eq. (236 kt CH4). Industrial emission
sources have generally increased across the time series through 1999 and then fluctuated up and down with
production changes associated with the treatment of wastewater from the pulp and paper manufacturing, meat and
6 Throughout the Inventory, emissions from domestic wastewater also include any commercial and industrial wastewater
collected and co-treated with domestic wastewater.
Waste 7-19
-------
1 poultry processing, fruit and vegetable processing, starch-based ethanol production, and petroleum refining
2 industries. Table 7-7 and Table 7-8 provide CH4 emission estimates from domestic and industrial wastewater
3 treatment.
4 With respect to N20, the United States identifies two distinct sources for N20 emissions from domestic wastewater:
5 emissions from centralized wastewater treatment processes, and emissions from effluent from centralized treatment
6 systems that has been discharged into aquatic environments. The 2016 emissions of N20 from centralized
7 wastewater treatment processes and from effluent were estimated to be 0.4 MMT C02 Eq. (1.2 kt N20) and 4.6
8 MMT CO2 Eq. (15.4 kt N20), respectively. Total N20 emissions from domestic wastewater were estimated to be
9 5.0 MMT C02 Eq. (16.6 kt N20). Nitrous oxide emissions from wastewater treatment processes gradually increased
10 across the time series as a result of increasing U.S. population and protein consumption. Nitrous oxide emissions are
11 not estimated from industrial wastewater treatment because there is no IPCC methodology provided or industrial
12 wastewater emission factors available. Table 7-7 and Table 7-8 provide N20 emission estimates from domestic
13 wastewater treatment.
14 Table 7-7: ChU and N2O Emissions from Domestic and Industrial Wastewater Treatment
15 (MMT COz Eq.)
Activity
1990
2005
2012
2013
2014
2015
2016
CH4
15.7
15.8
15.1
14.9
15.0
15.1
14.8
Domestic
10.5
10.1
9.3
9.1
9.2
9.3
8.9
Industrial3
5.1
5.7
5.8
5.8
5.7
5.8
5.9
N2O
3.4
4.4
4.6
4.7
4.8
4.8
5.0
Centralized WWTP
0.2
0.3
0.3
0.3
0.3
0.3
0.4
Domestic Effluent
3.2
4.1
4.3
4.3
4.5
4.5
4.6
Total
19.1
20.2
19.7
19.6
19.8
20.0
19.8
a Industrial activity includes the pulp and paper manufacturing, meat and poultry processing, fruit
and vegetable processing, starch-based ethanol production, and petroleum refining industries.
Note: Totals may not sum due to independent rounding.
16 Table 7-8: ChU and N2O Emissions from Domestic and Industrial Wastewater Treatment (kt)
Activity
1990
2005
2012
2013
2014
2015
2016
CH4
627
631
604
596
598
605
593
Domestic
422
404
372
365
369
374
357
Industrial3
205
227
232
231
229
231
236
N2O
11
15
16
16
16
16
17
Centralized WWTP
1
1
1
1
1
1
1
Domestic Effluent
11
14
14
15
15
15
15
a Industrial activity includes the pulp and paper manufacturing, meat and poultry processing, fruit
and vegetable processing, starch-based ethanol production, and petroleum refining industries.
Note: Totals may not sum due to independent rounding.
17 Methodology
18 Domestic Wastewater CH4 Emission Estimates
19 Domestic wastewater CH4 emissions originate from both septic systems and from centralized treatment systems,
20 such as publicly owned treatment works (POTWs). Within these centralized systems, CH4 emissions can arise from
21 aerobic systems that are not well managed or that are designed to have periods of anaerobic activity (e.g.,
22 constructed wetlands and facultative lagoons), anaerobic systems (anaerobic lagoons and anaerobic reactors), and
23 from anaerobic digesters when the captured biogas is not completely combusted. The methodological equations are:
7-20 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Emissions from Septic Systems = A
= USpop X (% onsite) X (EFseptic) X 1/109 X 365.25
Emissions from Centrally Treated Aerobic Systems (other than Constructed Wetlands) + Emissions from
Centrally Treated Aerobic Systems (Constructed Wetlands Only) + Emissions from Centrally Treated Aerobic
Systems (Constructed Wetlands used as Tertiary Treatment) = B
where,
Emissions from Centrally Treated Aerobic Systems (other than Constructed Wetlands)
= [(% collected) x (total BO Dr. produced) x (% aerobicoTcw) x (% aerobic w/out primary) + (% collected) x
(total BODs produced) x (% aerobicoTcw) x (% aerobic w/primary) x (1-% BOD removed in prim, treat.)] x
(% operations not well managed) x (B0) x (MCF-aerobic_not_well_man)
Emissions from Centrally Treated Aerobic Systems (Constructed Wetlands Only)
= [(% collected) x (total BODs produced) x (%aerobiccw)] x (B0) x (MCF-constructed wetlands)
Emissions from Centrally Treated Aerobic Systems (Constructed Wetlands used as Tertiary Treatment)
= [(POTW_flow_CW) X (BODcwjnf) X 3.79] X 1/106 X 365.25
Emissions from Centrally Treated Anaerobic Systems = C
= {[(% collected) x (total BODs produced) x (% anaerobic) x (% anaerobic w/out primary)] + [(%
collected) x (total BODs produced) x (% anaerobic) x (% anaerobic w/primary) x (1-% BOD removed in
prim, treat.)]} x (B0) x (MCF-anaerobic)
Emissions from Anaerobic Digesters = D
= [(POTW_flow_AD) x (digester gas)/(100)] x 0.0283 x (FRAC_CH4) x 365.25 x (density of CH4) x (1-DE)
x1/109
Total Domestic CH4 Emissions from Wastewater (kt) = A+ B + C+ D
where,
USpop
= U.S. population
% onsite
= Flow to septic systems / total flow
% collected
= Flow to POTWs / total flow
% aerobicoTcw
= Flow to aerobic systems, other than wetlands only / total flow to
POTWs
% aerobiccw
= Flow to aerobic systems, constructed wetlands used as sole treatment /
total flow to POTWs
% anaerobic
= Flow to anaerobic systems / total flow to POTWs
% aerobic w/out primary
= Percent of aerobic systems that do not employ primary treatment
% aerobic w/primary
= Percent of aerobic systems that employ primary treatment
% BOD removed in prim, treat.
= Percent of BOD removed in primary treatment
% operations not well managed
= Percent of aerobic systems that are not well managed and in which
some anaerobic degradation occurs
% anaerobic w/out primary
= Percent of anaerobic systems that do not employ primary treatment
% anaerobic w/primary
= Percent of anaerobic systems that employ primary treatment
EFseptic
= Methane emission factor - septic systems
Total BOD5 produced
= kg BOD/capita/day x U.S. population x 365.25 days/yr
BODcw.inf
= BOD concentration in wastewater entering the constructed wetland
Bo
= Maximum CH4-producing capacity for domestic wastewater
1/106
= Conversion factor, kg to kt
365.25
= Days in a year
3.79
= Conversion factor, liters to gallons
MCF-aerobicnotwellman.
= CH4 correction factor for aerobic systems that are not well managed
MCF-anaerobic
= CH4 correction factor for anaerobic systems
MCF-constructed wetlands
= CH4 correction factor for surface flow constructed wetlands
Waste 7-21
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
digester gas
100
0.0283
FRAC_CH4
DE
POTW flow CW
POTW flow AD
662
1/109
CH4 destruction efficiency from flaring or burning in engine
Wastewater flow to POTWs that use constructed wetlands as tertiary
treatment (MGD)
: Wastewater influent flow to POTWs that have anaerobic digesters
(MGD)
Cubic feet of digester gas produced per person per day
Wastewater flow to POTW (gallons/person/day)
Conversion factor, ft3 to m3
Proportion of CH4 in biogas
Density of CH4 (g CH4/m3 CH4)
Conversion factor, g to kt
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 19 percent) and an emission factor (10.7 g CH4/capita/day) (Leverenz et
al. 2010), 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 2017) and include the populations of the United States, American
Samoa, Guam, Northern Mariana Islands, Puerto Rico, and the U.S. Virgin Islands. Table 7-9 presents U.S.
population for 1990 through 2016.
Emissions from Centrally Treated Aerobic and Anaerobic Systems:
Methane emissions from POTWs were estimated by multiplying the total BOD5 produced in the United States by the
percent of wastewater treated centrally (about 81 percent) (U.S. Census Bureau 2015), the relative percentage of
wastewater treated by aerobic and anaerobic systems (other than constructed wetlands), the relative percentage of
wastewater facilities with primary treatment (EPA 1992, 1996, 2000, and 2004), the percentage of BOD5 treated
after primary treatment (67.5 percent, 32.5 percent removed in primary treatment) (Metcalf & Eddy 2003), the
maximum CH4-producing capacity of domestic wastewater (B0, 0.6 kg CH4/kg BOD) (IPCC 2006), and the relative
methane conversion factors (MCF) not well managed aerobic (0.3) (IPCC 2006), and anaerobic (0.8) (IPCC 2006)
systems.
Table 7-9 presents total BOD5 produced for 1990 through 2016. The proportions of domestic wastewater treated
onsite versus at centralized treatment plants were based on data from the 1989, 1991, 1993, 1995, 1997, 1999, 2001,
2003, 2005, 2007, 2009, 2011, 2013, and 2015 American Housing Surveys conducted by the U.S. Census Bureau
(U.S. Census Bureau 2015), with data for intervening years obtained by linear interpolation and 2016 forecasted
using 1990 to 2015 data. The percent of wastewater flow to aerobic and anaerobic systems, the percent of aerobic
and anaerobic systems that do and do not employ primary treatment, and the wastewater flow to POTWs that have
anaerobic digesters were obtained from the 1992, 1996, 2000, and 2004 Clean Watersheds Needs Survey (EPA
1992, 1996, 2000, and 2004). Data for intervening years were obtained by linear interpolation and the years 2005
through 2016 were forecasted from the rest of the time series. The percent of wastewater flow to aerobic systems
that use only constructed wetlands and wastewater flow to POTWs that use constructed wetlands as tertiary
treatment were obtained from the 1992, 1996, 2000, 2004, 2008, and 2012 Clean Watersheds Needs Survey (EPA
1992, 1996, 2000, 2004, 2008b, and 2012). Data for intervening years were obtained by linear interpolation and the
years 2013 through 2016 were forecasted from the rest of the time series. The BOD5 production rate (0.09
kg/capita/day) and the percent BOD5 removed by primary treatment for domestic wastewater were obtained from
Metcalf & Eddy (2003).
Table 7-9: U.S. Population (Millions) and Domestic Wastewater BODs Produced (kt)
Year Population BODs
1990 253 8,333
2005 300 9.853
2012 318 10,458
2013 320 10,534
2014 323 10,615
2015 325 10,696
7-22 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
2016 328 10,780
Sources: U.S. Census Bureau (2017); Metcalf & Eddy (2003).
1 For constructed wetlands, an MCF of 0.4 was used, which is the IPCC suggested MCF for surface flow wetlands.
2 This is the most conservative factor for constructed wetlands and was recommended by IPCC (2014) when the type
3 of constructed wetland is not known. A BOD5 concentration of 30 mg/L was used for wastewater entering
4 constructed wetlands used as tertiary treatment based on U.S. secondary treatment standards for POTWs. These
5 standards are based on plants generally utilizing simple settling and biological treatment (EPA 2013).
6 All aerobic systems are assumed to be well-managed as there are currently no data available to quantify the number
7 of systems that are not well-managed. In addition, methane emissions were calculated for systems that treat
8 wastewater with constructed wetlands and systems that use constructed wetlands as tertiary treatment; however,
9 constructed wetlands are a relatively small portion of wastewater treated centrally (<0.1 percent).
10 Emissions from Anaerobic Digesters:
11 Total CH4 emissions from anaerobic digesters were estimated by multiplying the wastewater influent flow to
12 POTWs with anaerobic digesters, the cubic feet of digester gas generated per person per day divided by the flow to
13 POTWs, the fraction of CH4 inbiogas (0.65), the density of CH4 (662 g CH i/m3 CH4) (EPA 1993a), one minus the
14 destruction efficiency from burning the biogas in an energy/thermal device (0.99 for enclosed flares) and then
15 converting the results to kt/year.
16 The CH4 destruction efficiency for CH4 recovered from sludge digestion operations, 99 percent, was selected based
17 on the range of efficiencies (98 to 100 percent) recommended for flares in AP-42 Compilation of Air Pollutant
18 Emission Factors, Chapter 2.4 (EPA 1998), efficiencies used to establish New Source Performance Standards
19 (NSPS) for landfills, along with data from CAR (2011), Sullivan (2007), Sullivan (2010), and UNFCCC (2012).
20 The cubic feet of digester gas produced per person per day (1.0 ft3/person/day) and the proportion of CH4 inbiogas
21 (0.65) come from Metcalf & Eddy (2014). The wastewater flow to a POTW (100 gal/person/day) was taken from
22 the Great Lakes-Upper Mississippi River Board of State and Provincial Public Health and Environmental Managers,
23 "Recommended Standards for Wastewater Facilities (Ten-State Standards)" (2004).
24 Table 7-10 presents domestic wastewater CH4 emissions for both septic and centralized systems, including
25 anaerobic digesters, in 2016.
26 Table 7-10: Domestic Wastewater ChU Emissions from Septic and Centralized Systems
27 (2016, MMT CO2 Eq. and Percent)
CH4 Emissions (MMT CO2 Eq.)
% of Domestic Wastewater CH4
Septic Systems
6.0
66.7%
Centrally-Treated Aerobic Systems
0.1
1.1%
Centrally-Treated Anaerobic Systems
2.7
29.8%
Anaerobic Digesters
0.2
2.3%
Total
8.9
100%
Note: Totals may not sum due to independent rounding.
28 Industrial Wastewater CH4 Emission Estimates
29 Methane emission estimates from industrial wastewater were developed according to the methodology described in
30 the 2006 IPCC Guidelines. Industry categories that are likely to produce significant CH4 emissions from wastewater
31 treatment were identified and included in the Inventory. The main criteria used to identify these industries are
32 whether they generate high volumes of wastewater, whether there is a high organic wastewater load, and whether the
33 wastewater is treated using methods that result in CH4 emissions. The top five industries that meet these criteria are
34 pulp and paper manufacturing; meat and poultry processing; vegetables, fruits, and juices processing; starch-based
35 ethanol production; and petroleum refining. Wastewater treatment emissions for these sectors for 2016 are
36 displayed in Table 7-11 below. Table 7-12 contains production data for these industries.
Waste 7-23
-------
1 Table 7-11: Industrial Wastewater ChU Emissions by Sector (2016, MMT CO2 Eq. and
2 Percent)
CH4 Emissions (MMT CO2 Eg.) % of Industrial Wastewater CH4
Meat & Poultry
4.5
76.8%
Pulp & Paper
0.9
15.8%
Fruit & Vegetables
0.1
2.4%
Petroleum Refineries
0.1
2.5%
Ethanol Refineries
0.1
2.5%
Total
5.9
100%
Note: Totals may not sum due to independent rounding.
3 Table 7-12: U.S. Pulp and Paper, Meat, Poultry, Vegetables, Fruits and Juices, Ethanol, and
4 Petroleum Refining Production (MMT)
Year
Pulp and Paper3
Meat
(Live Weight
Killed)
Poultry
(Live Weight
Killed)
Vegetables,
Fruits and Juices
Ethanol
Petroleum
Refining
1990
128.9
27.3
14.6
38.7
2.5
702.4
2005
138.5
31.4
25.1
42.9
1 1.7
818.6
2012
124.7
33.8
26.1
45.6
39.5
856.1
2013
120.8
33.6
26.5
45.1
39.7
878.7
2014
123.2
32.2
26.9
45.3
42.8
903.9
2015
121.8
32.8
27.7
44.6
44.2
914.5
2016
121.4
34.2
28.3
43.3
45.8
925.3
aPulp and paper production is the sum of woodpulp production plus paper and paperboard production.
Sources: Lockwood-Post (2002); FAO (2017a); USDA (2017a); Cooper (2017); EIA (2017).
5 Methane emissions from these categories were estimated by multiplying the annual product output by the average
6 outflow, the organics loading (in COD) in the outflow, the maximum CH4 producing potential of industrial
7 wastewater (B0), and the percentage of organic loading assumed to degrade anaerobically in a given treatment
8 system (MCF). Ratios of BOD:COD in various industrial wastewaters were obtained from EPA (1997a) and used to
9 estimate COD loadings. The B0 value used for all industries is the IPCC default value of 0.25 kg CH4/kg COD
10 (IPCC 2006).
11 For each industry, the percent of plants in the industry that treat wastewater on site, the percent of plants that have a
12 primary treatment step prior to biological treatment, and the percent of plants that treat wastewater anaerobically
13 were defined. The percent of wastewater treated anaerobically onsite (TA) was estimated for both primary treatment
14 (%TAP) and secondary treatment (%TAS). For plants that have primary treatment in place, an estimate of COD that
15 is removed prior to wastewater treatment in the anaerobic treatment units was incorporated. The values used in the
16 %TA calculations are presented in Table 7-13 below.
17 The methodological equations are:
18 CH4 (industrial wastewater) = [P x W x COD x %TAP xB0x MCF] + [P x W x COD x %TAS xB0x MCF]
19 %TAP = [%Plants0 x %WWa,P x %CODP]
20 %TAS = [%Plantsa x %WWa,s x %CODs] + [%Plantst x %WWa,t x %CODs]
21 where,
22 CH4 (industrial wastewater) = Total CH4 emissions from industrial wastewater (kg/year)
23 P = Industry output (metric tons/year)
24 W = Wastewater generated (m3/metric ton of product)
25 COD = Organics loading in wastewater (kg/m3)
26 %TAP = Percent of wastewater treated anaerobically on site in primary treatment
7-24 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
%TAs
= Percent of wastewater treated anaerobically on site in secondary treatment
%Plants0
= Percent of plants with onsite treatment
%WW,P
= Percent of wastewater treated anaerobically in primary treatment
%CODp
= Percent of COD entering primary treatment
%Plantsa
= Percent of plants with anaerobic secondary treatment
%Plantst
= Percent of plants with other secondary treatment
%WW a,s
= Percent of wastewater treated anaerobically in anaerobic secondary treatment
%WW a,t
= Percent of wastewater treated anaerobically in other secondary treatment
%CODs
= Percent of COD entering secondary treatment
Bo
= Maximum CH4 producing potential of industrial wastewater (kg CH4/kg
COD)
MCF
= CH4 correction factor, indicating the extent to which the organic content
(measured as COD) degrades anaerobically
Alternate methodological equations for calculating %TA were used for secondary treatment in the pulp and paper
industry to account for aerobic systems with anaerobic portions. These equations are:
%TAa = [%Plantsa x %WWa,s x %CODs] + [%Plants,t x %WWa,t x CODs]
%TAa,t = [%Plantsa,t x %WWa,s x %CODs]
where,
%TAa = Percent of wastewater treated anaerobically on site in secondary treatment
%TAa,t = Percent of wastewater treated in aerobic systems with anaerobic portions on
site in secondary treatment
%Plantsa = Percent of plants with anaerobic secondary treatment
%Plantsa,t = Percent of plants with partially anaerobic secondary treatment
%WW:ls = Percent of wastewater treated anaerobically in anaerobic secondary treatment
%WWa,t = Percent of wastewater treated anaerobically in other secondary treatment
%CODs = Percent of COD entering secondary treatment
As described below, the values presented in Table 7-13 were used in the emission calculations and are described in
detail in ERG (2008), ERG (2013a), and ERG (2013b).
Table 7-13: Variables Used to Calculate Percent Wastewater Treated Anaerobically by
Industry (Percent)
Industry
Variable
Pulp
and
Paper
Meat
Processing
Poultry
Processing
Fruit/
Vegetable
Processing
Ethanol
Production
- Wet Mill
Ethanol
Production
- Dry Mill
Petroleum
Refining
%TAP
0
0
0
0
0
0
0
%TAS
0
33
25
4.2
33.3
75
23.6
%TAa
2.2
0
0
0
0
0
0
%TAa,t
11.8
0
0
0
0
0
0
%Plants0
0
100
100
11
100
100
100
%PlantSa
5
33
25
5.5
33.3
75
23.6
%PlantSa,t
28
0
0
0
0
0
0
%Plantst
35
67
75
5.5
66.7
25
0
%WWa,p
0
0
0
0
0
0
0
%WWa,s
100
100
100
100
100
100
100
%WWa,t
0
0
0
0
0
0
0
%CODp
100
100
100
100
100
100
100
%CODs
42
100
100
77
100
100
100
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 (2008); ERG (2013a); and ERG (2013b).
Waste 7-25
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
Pulp and Paper. Wastewater treatment for the 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. In determining the
percent that degrades anaerobically, both primary and secondary treatment were considered. 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). However, because the vast majority of primary treatment operations at U.S. pulp and paper mills use
mechanical clarifiers, and less than 10 percent of pulp and paper wastewater is managed in primary settling ponds
that are not expected to have anaerobic conditions, negligible emissions are assumed to occur during primary
treatment.
Approximately 42 percent of the BOD passes on to secondary treatment, which consists of activated sludge, aerated
stabilization basins, or non-aerated stabilization basins. Based onEPA's OAQPS Pulp and Paper Sector Survey, 5.3
percent of pulp and paper mills reported using anaerobic secondary treatment for wastewater and/or pulp
condensates (ERG 2013a). Twenty-eight percent of mills also reported the use of quiescent settling ponds. Using
engineering judgment, these systems were determined to be aerobic with possible anaerobic portions. For the truly
anaerobic systems, an MCF of 0.8 is used, as these are typically deep stabilization basins. For the partially
anaerobic systems, an MCF of 0.2 is used, which is the 2006IPCC Guidelines-suggested MCF for shallow lagoons.
A time series of CH4 emissions for 1990 through 2001 was developed based on production figures reported in the
Lockwood-Post Directory (Lockwood-Post 2002). Data from the Food and Agricultural Organization of the United
Nations (FAO) database FAOSTAT were used for 2002 through 2015 (FAO 2017a). A time series of the overall
wastewater outflow for 1990 through 1994 varies based on data outlined in ERG (2013a) to reflect historical
wastewater flow. Wastewater generation rates for 1995, 2000, and 2002 were estimated from the 2014 American
Forest and Paper Association (AF&PA) Sustainability Report (AF&PA 2014). Wastewater generation rates for
2004, 2006, 2008, 2010, 2012, and 2014 were estimated from the 2016 AF&PA Sustainability Report (AF&PA
2016). Data for intervening years were obtained by linear interpolation, while 2015 and 2016 were forecasted using
1990 to 2014 and 1990 to 2015 data, respectively. The average BOD concentrations in raw wastewater was
estimated to be 0.4 grams BOD/liter (EPA 1997b; EPA 1993b; World Bank 1999). The COD:BOD ratio used to
convert the organic loading to COD for pulp and paper mills was 2 (EPA 1997a).
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. 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 B0 of 0.25 kg CH4/kg COD and default MCF of
0.8 for anaerobic lagoons were used to estimate the CH4 produced from these on-site treatment systems. Production
data, in carcass weight and live weight killed for the meat and poultry industry, were obtained from the USD A
Agricultural Statistics Database and the Agricultural Statistics Annual Reports (USDA 2017a). Data collected by
EPA's Office of Water provided estimates for wastewater flows into anaerobic lagoons: 5.3 and 12.5 m3/metric ton
for meat and poultry production (live weight killed), respectively (EPA 2002). The loadings are 2.8 and 1.5 g
BOD/liter for meat and poultry, respectively. The COD:BOD ratio used to convert the organic loading to COD for
both meat and poultry facilities was 3 (EPA 1997a).
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. Effluent is suitable for discharge to POTWs. This
industry is likely to 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).
Consequently, 4.2 percent of these wastewater organics are assumed to degrade anaerobically. The IPCC default B0
of 0.25 kg CH4/kg COD and default MCF of 0.8 for anaerobic treatment were used to estimate the CH4 produced
from these on-site treatment systems. The USDA National Agricultural Statistics Service (USDA 2017a, 2017c)
provided production data for potatoes, other vegetables, citrus fruit, non-citrus fruit, and grapes processed for wine.
Outflow and BOD data, presented in Table 7-14, were obtained from EPA (1974) for potato, citrus fruit, and apple
processing, and from EPA (1975) for all other commodities. The COD:BOD ratio used to convert the organic
loading to COD for all fruit, vegetable, and juice facilities was 1.5 (EPA 1997a).
7-26 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
Table 7-14: Wastewater Flow (m3/ton) and BOD Production (g/L) for U.S. Vegetables,
Fruits, and Juices Production
Commodity
Wastewater Outflow (m3/ton)
BOD (g/L)
Vegetables
Potatoes
10.27
1.765
Other Vegetables
8.55
0.776
Fruit
Apples
3.66
1.371
Citrus Fruits
10.11
0.317
Non-citrus Fruits
12.42
1.204
Grapes (for wine)
2.78
1.831
Sources: 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
only about 2 percent of ethanol production, and although the U.S. Department of Energy (DOE) predicts cellulosic
ethanol to greatly increase in the coming years, currently it is only in an experimental stage in the United States.
Currently, ethanol is mostly made from sugar and starch crops, but with advances in technology, cellulosic biomass
is increasingly used as ethanol feedstock (DOE 2013).
Ethanol is produced from corn (or other 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 has become more
efficient in recent years (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 digesters is commonly collected and either flared or used as fuel in the ethanol
production process (ERG 2006).
Available information was compiled from the industry on wastewater generation rates, which ranged from 1.25
gallons per gallon ethanol produced (for dry milling) to 10 gallons per gallon ethanol produced (for wet milling)
(Ruocco 2006a; Ruocco 2006b; Merrick 1998; Donovan 1996; NRBP 2001). COD concentrations were found to be
about 3 g/L (Ruocco 2006a; Merrick 1998; White and Johnson 2003). One hundred percent of plants were estimated
to have onsite wastewater treatment, and the variables used to calculate percent wastewater treated anaerobically are
presented in Table 7-13. A default MCF of 0.8 for anaerobic treatment was used to estimate the CH4 produced from
these on-site treatment systems. The amount of CH4 recovered through the use of biomethanators was estimated, and
a 99 percent destruction efficiency was used. Biomethanators are anaerobic reactors that use microorganisms under
anaerobic conditions to reduce COD and organic acids and recover biogas from wastewater (ERG 2006). Methane
emissions were then estimated as follows:
Methane = [Production x Flow x COD x 3.785 x ([%Plants0 x %WWa,P x %CODP] + [%Plantsa x %WWa,s x
%CODs] + [%Plantst x %WWa,t x %CODs]) xB0x MCF x % Not Recovered] + [Production x Flow x 3.785 x
COD x ([%PlantSo x %WWa,P x %CODP] + [%Plantsa x %WWa,s x %CODs] + [%Plantst x %WWa,t x %CODs])
x Bo x MCF x (% Recovered) x (1-DE)] x 1/109
where,
Production = Gallons ethanol produced (wet milling or dry milling)
Flow = Gallons wastewater generated per gallon ethanol produced
COD = COD concentration in influent (g/1)
Waste 7-27
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
3.785
= Conversion factor, gallons to liters
%Plants0
= Percent of plants with onsite treatment
%wwa,P
= Percent of wastewater treated anaerobically in primary treatment
%CODp
= Percent of COD entering primary treatment
%Plantsa
= Percent of plants with anaerobic secondary treatment
%Plantst
= Percent of plants with other secondary treatment
%wwa,s
= Percent of wastewater treated anaerobically in anaerobic secondary treatment
%WWa,t
= Percent of wastewater treated anaerobically in other secondary treatment
%CODs
= Percent of COD entering secondary treatment
Bo
= Maximum methane producing capacity (g CH4/g COD)
MCF
= Methane conversion factor
% Recovered
= Percent of wastewater treated in system with emission recovery
% Not Recovered
= 1 - percent of wastewater treated in system with emission recovery
DE
= Destruction efficiency of recovery system
1/109
= Conversion factor, g to kt
A time series of CH4 emissions for 1990 through 2016 was developed based on production data from the Renewable
Fuels Association (Cooper 2017).
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 2011.7 Of the responding facilities, 23.6 percent reported using
non-aerated surface impoundments or other biological treatment units, both of which have the potential to lead to
anaerobic conditions (ERG 2013b). In addition, the wastewater generation rate was determined to be 26.4 gallons
per barrel of finished product (ERG 2013b). An average COD value in the wastewater was estimated at 0.45 kg/m3
(Benyahia et al. 2006). A default MCF of 0.3 was used for partially aerobic systems.
The equation used to calculate CH4 generation at petroleum refining wastewater treatment systems is presented
below:
Methane = Flow x COD x %TA xB0x MCF
where,
Flow = Annual flow treated through anaerobic treatment system (m3/year)
COD = COD loading in wastewater entering anaerobic treatment system (kg/m3)
%TA = Percent of wastewater treated anaerobically on site
B0 = Maximum methane producing potential of industrial wastewater (kg CH i/kg COD)
MCF = Methane conversion factor
A time series of CH4 emissions for 1990 through 2016 was developed based on production data from the EIA 2017.
Domestic Wastewater N2O Emission Estimates
Nitrous oxide emissions from domestic wastewater (wastewater treatment) were estimated using the IPCC (2006)
methodology and supplemented with IPCC (2014) methodology to include constructed wetland emissions, including
calculations that take into account N removal with biosolids, non-consumption and industrial/commercial
wastewater N, and emissions from advanced and constructed wetlands at centralized wastewater treatment plants:
• In the United States, a certain amount of N is removed with biosolids, which is applied to land, incinerated,
or landfilled (Nsludge). The value for N discharged into aquatic environments as effluent is reduced to
account for the biosolids application.
• The 2006 IPCC Guidelines use annual, per capita protein consumption (kg protein/person-year). For this
Inventory, the amount of protein available to be consumed is estimated based on per capita annual food
7 Available online at
7-28 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
availability data and its protein content, and then that data is adjusted using a factor to account for the
fraction of protein actually consumed.
• Small amounts of gaseous nitrogen oxides are formed as byproducts in the conversion of nitrate to N gas in
anoxic biological treatment systems. Approximately 7 g N20 is generated per capita per year if wastewater
treatment includes intentional nitrification and denitrification (Scheehle and Doom 2001). Analysis of the
use of treatment systems in the United States that include denitrification has shown a significant increase in
the time period between 2004 and 2012, from serving populations totaling 2.4 million people to 21.3
million people (EPA 2004 and EPA 2012). This is consistent with efforts throughout the United States to
improve nutrient removal at centralized treatment systems in response to specific water quality concerns.
Based on an emission factor of 7 g per capita per year, and data from CWNS 2004, 2008, and 2012,
approximately 21.2 metric tons of additional N20 may have been emitted via denitrification in 2004, while
about 186 metric tons may have been emitted via denitrification in both 2008 and 2012. Similar analyses
were completed for each year in the Inventory using data from CWNS on the amount of wastewater in
centralized systems treated in denitrification units. Plants without intentional nitrification or denitrification
are assumed to generate 3.2 g N20 per capita per year.
• 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. The emission factor of 0.0013 kg N20-N/kg N
produced for constructed wetlands is from IPCC (2014).
• N20 emissions from wastewater treatment plants are estimated, and as such, the N associated with these
emissions is subtracted from the amount of N estimated to be discharged into aquatic environments as
effluent, consistent with the 2006 IPCC Guidelines.
Nitrous oxide emissions from domestic wastewater were estimated using the following methodology:
N20nit/denit= [(USpopnd) X EF2 X Find-com] X 1/109
N20woutnit/denit = {[(USpop X WWTP) - USpopnd - USpopcw] X 106 X Find-com X EFi} X 1/109
NzOcwonly = {[(USpopcw X 106 X Protein X Fnpr X Fnon-con X Find-com) X EF4] X 44/28} X 1/106
N2O CW TERTIARY — {[(New,inf x POTW_flow_CW x 3.79 x 365.25) x EF4] x 44/28} x 1/106
N20effluent = [(USpop X WWTP X Protein X Fnpr X Fnon-con X Find-com) - Nsludge - (N20plant X 106 X 28/44)] X
EFs X 44/28 X 1/106
N20total = N20plant + N20effluent
N20plant = N20nit/denit + N20woutnit/denit+ N20cwonly + N20cwtertiary
where,
N2Ototal
N2Oplant
N2Onit/denit
Annual emissions of N20 (kt)
N20 emissions from centralized wastewater treatment plants (kt)
N20 emissions from centralized wastewater treatment plants with
nitrification/denitrification (kt)
N20 emissions from centralized wastewater treatment plants without
nitrification/denitrification (kt)
N20 emissions from centralized wastewater treatment plants with constructed
wetlands only (kt)
N20 emissions from centralized wastewater treatment plants with constructed
wetlands used as tertiary treatment (kt)
N20 emissions from wastewater effluent discharged to aquatic environments (kt)
U.S. population
U.S. population that is served by biological denitrification
U.S. population that is served by only constructed wetland systems
Fraction of population using WWTP (as opposed to septic systems)
N2Owout nit/denit
N2Ocw ONLY
N2Ocw TERTIARY
N2Oeffluent
USpop
USpopnd
USpopcw
WWTP
Waste 7-29
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
POTWflowCW = Wastewater flow to POTWs that use constructed wetlands as tertiary treatment
(MGD)
EFi = Emission factor - plants without intentional denitrification
EF2 = Emission factor - plant with intentional nitrification or denitrification
Protein = Annual per capita protein consumption (kg/person/year)
New,inf = Influent nitrogen concentration to constructed wetlands used as tertiary treatment
(mg/L)
Fnpr = Fraction of N in protein (kg N/kg protein)
Fnon-con = Factor for non-consumed protein added to wastewater
Find-com = Factor for industrial and commercial co-discharged protein into the sewer
Nsludge = N removed with sludge, kg N/year
EF3 = Emission factor (kg N20 -N/kg sewage-N produced) - from effluent
EF4 = Emission factor (kg N20 -N/kg N produced) - constructed wetlands
3.79 = Conversion factor, liters to gallons
44/28 = Molecular weight ratio of N20 to N2
1/106 = Conversion factor, kg to Gg
1/109 = Conversion factor, gto Gg
U.S. population data were taken from the U.S. Census Bureau International Database (U.S. Census Bureau 2017)
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 wastewater treatment plants is based on data
from the 1989, 1991, 1993, 1995, 1997, 1999, 2001, 2003, 2005, 2007, 2009, 2011, 2013, and 2015 American
Housing Survey (U.S. Census Bureau 2015). Data for intervening years were obtained by linear interpolation and
2016 was forecasted using 1990 to 2015 data. The emission factor (EFi) used to estimate emissions from
wastewater treatment for plants without intentional nitrification or denitrification was taken from IPCC (2006),
while the emission factor (EF2) used to estimate emissions from wastewater treatment for plants with intentional
nitrification or denitrification was taken from Scheehle and Doom (2001). The emission factor (EF4) used to
estimate emissions from surface flow constructed wetlands (0.0013 kg N20 -N/kg N produced) was taken from
IPCC (2014). Data on annual per capita protein intake were provided by the U.S. Department of Agriculture
Economic Research Service (USDA 2017b) and FAO (2017b). Protein consumption data was used directly from
USDA for 1990 to 2010 and 2011 through 2013 was calculated using FAO data and a scaling factor. 2014 through
2016 were extrapolated from data for 1990 through 2013. An emission factor to estimate emissions from effluent
(EF3) has not been specifically estimated for the United States, thus the default IPCC value (0.005 kg N20-N/kg
sewage-N produced) was applied (IPCC 2006). The fraction of N in protein (0.16 kg N/kg protein) was also
obtained from IPCC (2006). The factor for non-consumed protein (1.2) and the factor for industrial and commercial
co-discharged protein (1.25) were obtained from IPCC (2006). The amount of nitrogen removed by denitrification
systems was taken from EPA (2008a), while the population served by denitrification systems was estimated from
Clean Watersheds Needs Survey (EPA 1992, 1996, 2000, 2004, 2008b, and 2012). Sludge generation was obtained
from EPA (1999) for 1988, 1996, and 1998 and from Beecher et al. (2007) for 2004. Intervening years were
interpolated, and estimates for 2005 through 2016 were forecasted from the rest of the time series. The influent
nitrogen concentration to constructed wetlands used as tertiary treatment (25 mg/L) was obtained from Metcalf &
Eddy (2014). An estimate for the N removed as sludge (Nsludge) was obtained by determining the amount of sludge
disposed by incineration, by land application (agriculture or other), through surface disposal, in landfills, or through
ocean dumping (EPA 1993b; Beecher et al. 2007; McFarland 2001; EPA 1999). In 2016, 295 kt N was removed
with sludge. Table 7-15 presents the data for U.S. population, population served by biological denitrification,
population served by wastewater treatment plants, available protein, protein consumed, and nitrogen removed with
sludge.
Table 7-15: U.S. Population (Millions), Population Served by Biological Denitrification
(Millions), Fraction of Population Served by Wastewater Treatment (percent), Available
Protein (kg/person-year), Protein Consumed (kg/person-year), and Nitrogen Removed with
Sludge (kt-N/year)
N Removed
Year Population Populations WWTP Population Available Protein Protein Consumed with Sludge
1990 253 2.0 75.6 43.1 33.2 214.2
7-30 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
2005 300 7.1 78.8 44.9 34.7 261.1
2012 318 21.3 81.0 43.3 33.4 282.6
2013 320 19.8 81.4 43.3 33.4 285.6
2014 323 20.8 80.3 44.8 34.6 288.7
2015 325 21.8 79.2 44.8 34.5 291.8
2016 328 22.8 81.4 44.7 34.5 294.8
Sources: Population: U.S. Census Bureau (2017); Population^: EPA (1992), EPA (1996), EPA (2000), EPA (2004), EPA
(2008b), EPA (2012); WWTP Population: U.S. Census Bureau (2015); Available Protein: USDA (2017b); N Removed with
sludge: Beecher et al. (2007), McFarland (2001), EPA (1999), EPA (1993c).
Uncertainty and Time-Series Consistency
The overall uncertainty associated with both the 2016 CH4 and N20 emission estimates from wastewater treatment
and discharge was calculated using the 2006IPCC 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 wastewater from pulp and paper manufacturing, meat and poultry
processing, fruits and vegetable processing, ethanol production, and petroleum refining. Uncertainty associated with
the parameters used to estimate N20 emissions include that of biosolids disposal, total U.S. population, average
protein consumed per person, fraction of N in protein, non-consumption nitrogen factor, emission factors per capita
and per mass of sewage-N, and for the percentage of total population using centralized wastewater treatment plants.
Uncertainty associated with 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.
The results of this Approach 2 quantitative uncertainty analysis are summarized in Table 7-16. Methane emissions
from wastewater treatment were estimated to be between 11.0 and 18.0 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 26 percent
below to 22 percent above the 2016 emissions estimate of 14.8 MMT CO2 Eq. Nitrous oxide emissions from
wastewater treatment were estimated to be between 1.3 and 10.5 MMT CO2 Eq., which indicates a range of
approximately 75 percent below to 112 percent above the 2016 emissions estimate of 5.0 MMT CO2 Eq.
Table 7-16: Approach 2 Quantitative Uncertainty Estimates for ChU Emissions from
Wastewater Treatment (MMT CO2 Eq. and Percent)
Source
Gas
2016 Emission Estimate
Uncertainty Range Relative to Emission Estimate3
(MMT CO2 Eq.)
(MMT CO2 Eq.)
(%)
Lower
Bound
Upper
Bound
Lower
Bound
Upper
Bound
Wastewater Treatment
CH4
14.8
11.0
18.0
-26%
+22%
Domestic
ch4
8.9
6.6
11.2
-26%
+26%
Industrial
ch4
5.9
3.1
8.6
-48%
+45%
Wastewater Treatment
n2o
5.0
1.3
10.5
-75%
+112%
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. 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;
Waste 7-31
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
• 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 emissions estimates; and
• Compared estimates to previous estimates to identify significant changes.
All transcription errors identified were corrected and documented. The QA/QC analysis did not reveal any systemic
inaccuracies or incorrect input values.
Recalculations Discussion
Population data were updated to reflect revised U.S. Census Bureau datasets which resulted in changes to 2010
through 2015 values (U.S. Census Bureau 2017). In addition, the 2015 American Housing Survey became available
which resulted in updated values for the percent of wastewater treatment collected versus treatment onsite for both
2014 and 2015 (U.S. Census Bureau 2015).
EPA evaluated pulp and paper wastewater generation rates in the 2016 American Forest & Paper Association
Sustainability Report based on the National Council of Air and Stream Improvement's (NCASI) recommendation,
and determined updates to current Inventory data were appropriate. EPA updated values for 2004, 2006, 2008, 2010,
2012, and 2014 with data provided in the 2016 report. EPA also used the 2014 AF&PA Sustainability Report to
update the 1995, 2000, and 2002 values to more accurately reflect industry data. Data for intervening years were
obtained by linear interpolation and the years 2015 and 2016 were forecasted from the rest of the time series. This
change resulted in updated values for pulp and paper wastewater generation rates (m3/ton) for 1995 through 2015.
Planned Improvements
EPA will continue to investigate the following improvements to the wastewater emissions estimates in the
Inventory:
• Continue working with the NCASI to determine if there are sufficient data available to update the estimates
of organic loading in pulp and paper wastewaters treated on-site;
• Investigate updated sources of activity data for wastewater treatment system type to distinguish between
aerobic, anaerobic, and other systems with the potential to generate CH4. This includes re-evaluating a
methodology that was developed so that the 2008 and 2012 CWNS data could be used in estimating
emissions from constructed wetlands to determine if it could be extended to all types of systems; and
• Continue reviewing other industrial wastewater treatment sources for those industries believed to discharge
significant loads of BOD or COD, including dairy processing and brewery wastewater.
In addition, EPA will continue to monitor potential sources for updating Inventory data, including:
• WEF biosolid data as a potential source of digester, sludge, and biogas data from POTWs;
• Reports based on international research and other countries' inventory submissions to inform potential
updates to the Inventory's emission factors, methodologies, or included industries;
• Research by groups such as the Water Environment Research Federation (WERF) on emissions from
various types of municipal treatment systems, country-specific N20 emission factors, and flare efficiencies;
• Data collected by WERF that indicate septic soil systems are a source of N20 for the potential development
of appropriate emission factors for septic system N20 emissions;
• Sources of data for development of a country-specific methodology for N20 emissions associated with on-
site industrial wastewater treatment operations, including the appropriateness of using IPCC's default factor
for domestic wastewater (0.005 kg N20-N/kg N);
• Sources of data for updating the factor for industrial and commercial co-discharged protein to determine if
the IPCC factor currently used (1.25) is underestimating the contribution of industrial wastewater to N20
emissions;
• Additional data sources for improving the uncertainty of the estimate of N entering municipal treatment
systems; and
7-32 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
• Data to update the value used for N content of sludge, the amount of sludge produced, and sludge disposal
practices, along with increasing the transparency of the fate of sludge produced in wastewater treatment.
A refinement of the 2006IPCC Guidelines is currently underway to incorporate abundant new scientific and
empirical knowledge published since 2006 which the IPCC should take into account, particularly with respect to
data for emission factor development. For wastewater treatment, this refinement includes a review of methane and
nitrous oxide emission factors, and an assessment of adding methodologies to account for nitrous oxide emissions
from both domestic and industrial wastewater. EPA will continue to monitor the status of this refinement for
potential updates to the wastewater inventory methodology.
These planned improvements were described in greater detail in the previous Inventory report; please see Section
7.2 of the Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2015.
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. Advantages of composting include reduced volume of the waste,
stabilization of the waste, and destruction of pathogens in the waste. The end products of composting, depending on
its quality, can be recycled as a fertilizer and soil amendment, or be disposed of in a landfill.
Composting is an aerobic process and a large fraction of the degradable organic carbon in the waste material is
converted into carbon dioxide (CO2). Methane (CH4) is formed in anaerobic sections of the compost, which are
created when there is excessive moisture or inadequate aeration (or mixing) of the compost pile. This CH4 is then
oxidized to a large extent in the aerobic sections of the compost. The estimated CH4 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 (N20) emissions can be produced. The formation of N20
depends on the initial nitrogen content of the material and is mostly due to nitrogen oxide (NOx) denitrification
during the later composting stages. 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 N20 than, for
example, yard waste, however data are limited.
From 1990 to 2016, the amount of waste composted in the United States increased from 3,810 kt to 21,163 kt. There
was some fluctuation in the amount of waste composted between 2006 to 2009. Since then, the annual quantity has
increased and is now at an all-time high for the Inventory time series (see Table 7-19). A peak of 20,049 kt
composted was observed in 2008, followed by a steep drop the following year to 18,824 kt composted, presumably
driven by the economic crisis. Since then, the amount of waste composted has gradually increased, and when
comparing 2010 to 2016, a 16 percent increase in waste composted is observed. Emissions of CH4 and N20 from
composting from 2010 to 2015 have increased by the same percentage. In 2016, CH4 emissions from composting
(see Table 7-17 and Table 7-18) were 2.1 MMT C02 Eq. (85 kt), and N20 emissions from composting were 1.9
MMT C02 Eq. (6 kt). 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 backyard
composting or agricultural composting.
The growth in composting since the 1990s and specifically over the past decade is attributable primarily to the
following factors: (1) the enactment of legislation by state and local governments that discouraged the disposal of
yard trimmings 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 1990's 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. By 2010, 25 states, representing about 50 percent of the nation's population, had enacted such
legislation (ILSR 2014; BioCycle 2010). There are many more initiatives at the metro and municipal level across the
United States. More than 3,280 composting facilities exist in the United States with most (71 percent) composting
yard trimmings only (ISLR 2014).
Waste 7-33
-------
1
Table 7-17: ChU and N2O Emissions from Composting (MMT CO2 Eq.)
Activity 1990 2005 2012 2013 2014 2015 2016
(II: 0.4 1.9 1.9 2.0 2.1 2.1 2.1
N2O 0.3 ! 1.7 1.7 1.8 1.9 1.9 1.9
Total 0.7 3.6 3.7 3.9 4.0 4.0 4.0
3 Table 7-18: ChU and N2O Emissions from Composting (kt)
Activity 1990 2005 2012 2013 2014 2015 2016
CH4 15.2 74.6 77.4 81.4 83.5 84.2 84.7
N2O 1.1 5.6 . 5.8 6.1 6.3 6.3 6.3
4 Methodology
5 Methane and N20 emissions from composting depend on factors such as the type of waste composted, the amount
6 and type of supporting material (such as wood chips and peat) used, temperature, moisture content (e.g., wet and
7 fluid versus dry and crumbly), and aeration during the composting process.
8 The emissions shown in Table 7-17 and Table 7-18 were estimated using the IPCC default (Tier 1) methodology
9 (IPCC 2006), which is the product of an emission factor and the mass of organic waste composted (note: no CH4
10 recovery is expected to occur at composting operations in the emission estimates presented):
n Ej =MxEFj
12 where,
13 E, CH4 or N20 emissions from composting, kt CH4 or N20,
14 M mass of organic waste composted in kt,
15 EF, = emission factor for composting, 4 t CH 4/kt of waste treated (wet basis) and
16 0.3 t N20/kt of waste treated (wet basis) (IPCC 2006), and
17 i designates either CH4 or N20.
18 Estimates of the quantity of waste composted (M) are presented in Table 7-19 for select years. Estimates of the
19 quantity composted for 1990, 2005, 2010, and 2012 to 2014 were taken from EPA's Advancing Sustainable
20 Materials Management: Facts and Figures 2014 (EPA 2016); the estimate of the quantity composted for 2011 was
21 taken from EPA's Municipal Solid Waste In The United States: 2012 Facts and Figures (EPA 2014); estimates of
22 the quantity composted for 2015 and 2016 were extrapolated using the 2014 quantity composted and a ratio of the
23 U.S. population growth between 2014 and 2015, and 2015 to 2016 (U.S. Census Bureau 2016 and 2017).
24 Table 7-19: U.S. Waste Composted (kt)
Activity
1990
2005
2012
2013
2014
2015
2016
Waste Composted
3,810
18,643
19,351
20,358
20,884
21,052
21,163
25 Uncertainty and Time-Series Consistency
26 The estimated uncertainty from the 2006 IPCC Guidelines is ±50 percent for the Approach 1 methodology.
27 Emissions from composting in 2015 were estimated to be between 2.0 and 6.0 MMT CO2 Eq., which indicates a
28 range of 50 percent below to 50 percent above the actual 2016 emission estimate of 4.0 MMT CO2 Eq. (see Table
29 7-20).
7-34 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Table 7-20: Approach 1 Quantitative Uncertainty Estimates for Emissions from Composting
2 (MMT CO2 Eq. and Percent)
Source
Gas
2016 Emission Estimate
(MMT CO2 Eq.)
Uncertainty Range Relative to Emission Estimate
(MMT CO2 Eq.) (%)
Lower Upper Lower Upper
Bound Bound Bound Bound
Composting
ch4, n2o
4.0
2.0 6.0 -50% +50%
3 QA/QC and Verification
4 General QA/QC procedures were applied to data gathering and input, documentation and calculations consistent
5 with the U.S. QA/QC plan which is in accordance with Vol. 1 Chapter 6 of 2006IPCC Guidelines (see Annex 8 for
6 more details).
7 Recalculations Discussion
8 No recalculations were made in this Inventory year. The composting estimates will be updated pending the release
9 of a new EPA Advancing Sustainable Materials Management: Facts and Figures report.
10 Planned Improvements
11 For future Inventories, additional efforts will be made to improve the estimates of CH4 and N20 emissions from
12 composting. For example, a literature search on emission factors and composting systems and management
13 techniques has been completed and will be documented in a technical memorandum for the 1990 through 2017
14 Inventory. The purpose of this literature review was to compile all published emission factors specific to various
15 composting systems and composted materials. This information will be used to determine whether the emission
16 factors used in the current methodology should be revised, or expanded to account for geographical differences
17 and/or differences in composting systems used. For example, outdoor composting processes in arid regions typically
18 require the addition of moisture compared to similar composting processes in wetter climates. Additionally,
19 composting systems that primarily compost food waste may generate CH4 at different rates than those that compost
20 yard trimmings because the food waste may have a higher moisture content and more readily degradable material.
21 Further cooperation with estimating emissions in cooperation with the LULUCF Other section will also be
22 investigated.
23 7.4 Waste Incineration (CRF Source Category
5C1) - TO BE UPDATED FOR FINAL
25 INVENTORY REPORT
26 As stated earlier in this chapter, carbon dioxide (CO2), nitrous oxide (N20), and methane (CH4) emissions from the
27 incineration of waste are accounted for in the Energy sector rather than in the Waste sector because almost all
28 incineration of municipal solid waste (MSW) in the United States occurs at waste-to-energy facilities where useful
29 energy is recovered. Similarly, the Energy sector also includes an estimate of emissions from burning waste tires and
30 hazardous industrial waste, because virtually all of the combustion occurs in industrial and utility boilers that
31 recover energy. The incineration of waste in the United States in 2015 resulted in 11.0 MMT CO2 Eq., over half of
32 which (5.9 MMT CO2 Eq.) is attributable to the combustion of plastics. For more details on emissions from the
33 incineration of waste, see Section 3.3 of the Energy chapter.
Waste 7-35
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
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 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.
7.5 Waste Sources of Indirect Greenhouse
Gases
In addition to the main greenhouse gases addressed above, waste generating and handling processes are also sources
of indirect greenhouse gas emissions. Total emissions of nitrogen oxides (NOx), carbon monoxide (CO), and non-
CH4 volatile organic compounds (NMVOCs) from waste sources for the years 1990 through 2016 are provided in
Table 7-21.
Table 7-21: Emissions of NOx, CO, and NMVOC from Waste (kt)
Gas/Source
1990
2005
2012
2013
2014
2015
2016
NOx
+
2
2
2
2
2
2
Landfills
+
2
2
2
2
2
2
Wastewater Treatment
+
0
0
0
0
0
0
Miscellaneous3
+
0
0
0
0
0
0
CO
1
7
6
8
9
9
9
Landfills
1
6
6
7
8
8
8
Wastewater Treatment
+
+
+
1
1
1
1
Miscellaneous3
+
0
0
0
0
0
0
NMVOCs
673
114
45
51
57
57
57
Wastewater Treatment
57
49
19
22
25
25
25
Miscellaneous3
557
43
17
19
22
22
22
Landfills
58
22
8
10
11
11
11
+ Does not exceed 0.5 kt.
a 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
Emission estimates for 1990 through 2016 were obtained from data published on the National Emission Inventory
(NEI) Air Pollutant Emission Trends web site (EPA 2016), and disaggregated based on EPA (2003). Emission
estimates for 2012 and 2013 for non-electric generating units (EGU) were updated to the most recent available data
in EPA (2016). Emission estimates for 2012 and 2013 for non-mobile sources are recalculated emissions by
interpolation from 2016 in EPA (2016). Emission estimates of these gases were provided by sector, using a "top
down" estimating procedure—emissions were calculated either for individual sources or for many sources
combined, using basic activity data (e.g., the amount of raw material processed) as an indicator of emissions.
National activity data were collected for individual categories from various agencies. Depending on the category,
these basic activity data may include data on production, fuel deliveries, raw material processed, etc.
7-36 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Uncertainty and Time-Series Consistency
2 No quantitative estimates of uncertainty were calculated for this source category. Methodological recalculations
3 were applied to the entire time series to ensure time-series consistency from 1990 through 2016. Details on the
4 emission trends through time are described in more detail in the Methodology section, above.
Waste 7-37
-------
1
2 # Ot ihl
3 The United States does not report any greenhouse gas emissions under the Intergovernmental Panel on Climate
4 Change (IPCC) "Other" sector.
Other 8-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
9. Recalculations and Improvements
Each year, many of 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, including the transparency, completeness, consistency and overall
usefulness of the report. In this effort, the United States follows the 2006IPCC Guidelines (IPCC 2006), which
states, "Both methodological changes and refinements over time are an essential part of improving inventory quality.
It is good practice to change or refine methods when available data have changed; the previously used method is not
consistent with the IPCC guidelines for that category; a category has become key; the previously used method is
insufficient to reflect mitigation activities in a transparent manner; the capacity for inventory preparation has
increased; improved inventory methods become available; and/or for correction of errors."
The results of all methodological changes and historical data updates made in the current Inventory are presented in
Table 9-1 and Table 9-2. To understand the details of any specific recalculation or methodological improvement, see
the Recalculations Discussion within each source/sink categories' section found in Chapters 3 through 7 of this
report and a discussion of Inventory improvements in Annex 8. Table 9-1 summarizes the quantitative effect of all
changes on U.S. greenhouse gas emissions in the Energy, IPPU, Agriculture, and Waste chapters, while Table 9-2
summarizes the quantitative effect of changes on annual net fluxes from LULUCF. Both tables present results
relative to the previously published Inventory (i.e., the 1990 to 2015 report) in units of million metric tons of carbon
dioxide equivalent (MMT CO2 Eq.).
In general, when methodological changes have been implemented, the previous Inventory's time series (i.e., 1990 to
2015) 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 following source and sink categories underwent the most significant methodological and historical data
changes. A brief summary of the recalculations and/or improvements undertaken is provided for each of the
following categories.
• Fossil Fuel Combustion (CO2). The Energy Information Administration (EIA 2017a) updated energy
consumption statistics across the time series relative to the previous Inventory. EIA revised 2015 Liquefied
Petroleum Gas (LPG) consumption in the residential, commercial, industrial, and transportation sectors, and
2014 distillate fuel consumption in the transportation sector. These changes resulted in an average annual
increase in emissions of 14.0 MMT CO2 Eq. relative to the previous Inventory.
• Petroleum Systems (CO2). Updates were made to exploration and production segment methodologies for the
Inventory, including revising activity and CO2 emissions data for associated gas venting and flaring,
miscellaneous production flaring, and well testing. Production segment CO2 emissions data were also revised
for oil tanks, pneumatic controllers, and chemical injection pumps. The combined impact of revisions to 2015
petroleum systems CO2 emissions, compared to the previous Inventory, is an increase from 3.6 to 38.0 MMT
CO2 (34.4 MMT CO2, or by a factor of 9). The recalculations resulted in an average increase in emission
estimates across the 1990 through 2015 time series, compared to the previous Inventory, of 13.8 MMT CO2 Eq,
or 360 percent. The CO2 emissions estimate increase was primarily due to recalculations related to the
reallocation of CO2 from flaring to petroleum systems from natural gas systems. Previously, data were not
available to disaggregate flared emissions between natural gas systems and petroleum systems. The largest
sources of CO2 from flaring are associated gas flaring, tanks with flares, and miscellaneous production flaring.
Recalculations and Improvements 9-1
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
• Petroleum Systems (CH4). Updates were made to exploration and production segment methodologies for the
Inventory, including revising activity and CH4 emissions data for associated gas venting and flaring,
miscellaneous production flaring, and well testing. The combined impact of revisions to 2015 petroleum
systems CH4 emissions, compared to the previous Inventory, is a decrease from 39.9 to 39.4 MMT CO2 Eq. (0.5
MMT CO2 Eq., or 1 percent). The recalculations resulted in an average decrease in CH4 emission estimates
across the 1990 through 2015 time series, compared to the previous Inventory, of 11 MMT CO2 Eq, or 22
percent. The CH4 emissions estimate decrease was primarily due to recalculations related to associated gas
venting and flaring which were updated to use a basin-level approach, and has the largest impact on years prior
to 2013. These changes resulted in an average annual decrease in emissions of 10.9 MMT CO2 Eq. relative to
the previous Inventory.
• Natural Gas Systems (CO3). Updates were made to exploration through transmission and storage segments,
including to calculate activity and emission factors for well testing and non-hydraulically fractured completions
from EPA's GHGRP data, recalculate production segment major equipment activity factors using updated
GHGRP data, and calculate new CO2 emission factors for several sources throughout segments using GHGRP
data. The combined impact of revisions to 2015 natural gas sector CO2 emissions, compared to the previous
Inventory, is a decrease from 42.4 to 26.3 MMT CO2 (16.0 MMT CO2, or 38 percent). The recalculations
resulted in an average decrease in emission estimates across the 1990 through 2015 time series, compared to the
previous Inventory, of 10.3 MMT CO2 Eq, or 29 percent. The decreased estimate is primarily due to
recalculations related to the reallocation of CO2 from flaring to petroleum systems from natural gas systems.
Previously, data were not available to disaggregate flared emissions between natural gas and petroleum.
• Mobile Combustion (CH4). Decreases in on-road gasoline emissions were offset by large increases in alternative
fuel and non-road emissions. The collective result of all of these changes was a net increase in CH4 emissions
from mobile combustion relative to the previous Inventory. Methane emissions increased by 52.7 percent or an
average annual increase in emissions of 3.3 MMT CO2 Eq. New CH4 emission factors were calculated based on
the ratio of non-methane organic gas (NMOG) emission standards. These new emission factors allowed the
inclusion of additional emissions standards, including Federal Tier 3 emission standards and two levels of
California emission standards (LEV II and LEV III) to the control technology breakouts.
In addition, new non-road emissions factors were developed. Previously emission factors were taken from the
1996IPCC Guidelines and represented the IPCC Tier 1 factors. This year new emission factors were calculated
using the updated 2006 IPCC Tier 3 guidance and EPA's MOVES2014a model. Methane emission factors were
calculated directly from MOVES. Nitrous oxide emission factors were calculated using NONROAD activity
and emission factors by fuel type from the European Environment Agency. Gasoline engines were broken out
by 2- and 4-stroke engine types. Equipment using LPG and compressed natural gas (CNG) were included.
New emission factors for alternative fuel vehicles were estimated using the newest version of GREET (2016).
The updated emission factors have been generated for CH4. 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 heavy-duty vehicles, EPA used emission factors for Long Haul Combination Trucks. For Buses,
EPA used emission factors for Transit Buses. The emission factors developed represent vehicle operation only
(tank-to-wheels).
In addition, changes were made to the historic allocation of gasoline to on-road and non-road applications. In
2016, the Federal Highway Administration (FHWA) changed its methods for estimating the share of gasoline
used in on-road and non-road applications. Among other updates, FHWA included lawn and garden equipment
as well as off-road recreational equipment in its estimates of non-road gasoline consumption for the first time.
This change created a time-series inconsistency between the data reported for year 2015 and 2016 and previous
years. To create a more consistent time series of motor gasoline consumption and emissions data for the current
Inventory, the historical time series was modified. Specifically, the lawn, garden, and recreational vehicle
gasoline consumption from table MF-24 is subtracted from the highway motor gasoline consumption from
FHWA MF-21 when determining the total highway motor gasoline consumption for the years 1990 through
2014.
9-2 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2016 with one recent notable exception. An update by FHWA to the method for estimating on-road
VMT created an inconsistency in on-road CH4 and N2O for the time periods 1990 to 2006 and 2007 to 2016.
Details on the emission trends and methodological inconsistencies through time are described in more detail in
the Section 3.2—Mobile Combustion (CH4), Methodology section. These changes resulted in an average annual
increase in emissions of 3.3 MMT CO2 Eq. relative to the previous Inventory.
• Non-Energy Use of Fuels (C02). Pesticide production data for 2007 through 2015 was updated using EPA's
Pesticides Industry Sales and Usage 2008 - 2012 Market Estimates (EPA 2017). This resulted in a slight
increase in emissions from pesticides compared to previous estimates for 2007 through 2015. Pesticide
production data for 1990 through 2015 was updated by correcting rounding errors and molecular weights and
chemical formulas for certain pesticides. The calculated ratio of urea production to melamine production from
2001 to 2015 was updated to approximately 95/5 based on ICIS (2016) and ICIS (2008), rather than an even
50/50 split as previously estimated. These changes resulted in an average annual increase in emissions of 2.9
MMT CO2 Eq. relative to the previous Inventory.
• Grassland Remaining Grassland: Changes in Agricultural Carbon Stock (CO 2 sink). Methodological
recalculations are associated with modifying the approach for extending the time series from 2013 through 2015
for mineral and organic soils using a surrogate data method. C stock change estimates declined by an average of
97 percent from 2013 through 2015 based on the recalculation. These changes resulted in an average annual
increase in emissions of 2.3 MMT CO2 Eq. relative to the previous Inventory.
• Stationary Combustion (N2O). N2O emissions from stationary sources across the entire time series were revised
due to revised data from EIA (2017a), EIA (2017b), and EPA (2017a) relative to the previous Inventory. N20
emission factors for combined cycle natural gas units were updated to be consistent with EPA's Compilation of
Air Pollutant Emission Factors, AP-42 (EPA 1997). In addition, the GWPs for N2O for the Acid Rain Program
Dataset (EPA 2017a) were updated to be consistent with the IPCC Fourth Assessment Report (AIM) values.
The historical data changes resulted in an average annual decrease of 2.3 MMT CO2 Eq. (12 percent) in N20
emissions from stationary combustion for the period 1990 through 2015.
• Mobile Combustion (N20). Updates were made to the on-road, non-road and alternative fuel N20 emissions
calculations this year resulting in both increases and decreases to different source categories. Decreases in on-
road gasoline emissions were offset by large increases in alternative fuel and non-road emissions. The collective
result of all of these changes was a net increase in N20 emissions from mobile combustion by 24.5 percent.
Each of these changes is described below. New emission factors for N20 emissions were developed for on-road
vehicles based on an EPA regression analysis of the relationship between NOx and N20. These new emission
factors allowed the inclusion of additional emissions standards, including Federal Tier 3 emission standards and
two levels of California emission standards (LEV II and LEV III) to the control technology breakouts.
In addition, new non-road emissions factors were developed from the updated 2006 IPCC Tier 3 guidance and
EPA's MOVES2014a model. N20 emission factors were calculated using NONROAD activity and emission
factors by fuel type from the European Environment Agency. Gasoline engines were broken out by 2- and 4-
stroke engine types. Equipment using LPG and CNG were included. New emission factors for alternative fuel
vehicles were estimated using the newest version of GREET (2016). The updated emission factors have been
generated for N20. 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 heavy-duty vehicles, EPA
used emission factors for Long Haul Combination Trucks. For Buses, EPA used emission factors for Transit
Buses. The emission factors developed represent vehicle operation only (tank-to-wheels).
In addition, changes were made to the historic allocation of gasoline to on-road and non-road applications. In
2016, FHWA changed its methods for estimating the share of gasoline used in on-road and non-road
applications. Among other updates, FHWA included lawn and garden equipment as well as off-road
recreational equipment in its estimates of non-road gasoline consumption for the first time. This change created
a time-series inconsistency between the data reported for year 2015 and 2016 and previous years. To create a
Recalculations and Improvements 9-3
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
more consistent time series of motor gasoline consumption and emissions data for the current Inventory, the
historical time series was modified. Specifically, the lawn, garden, and recreational vehicle gasoline
consumption from table MF-24 is subtracted from the highway motor gasoline consumption from FHWAMF-
21 when determining the total highway motor gasoline consumption for the years 1990 to 2014.
Methodological recalculations were applied to the entire time series to ensure time-series consistency from 1990
through 2016 with one recent notable exception. An update by FHWA to the method for estimating on-road
VMT created an inconsistency in on-road CH4 and N2O for the time periods 1990 to 2006 and 2007 to 2016.
Details on the emission trends and methodological inconsistencies through time are described in more detail in
the Section 3.2—Mobile Combustion (N20), Methodology section. These changes resulted in an average annual
increase in emissions of 2.7 MMT CO2 Eq. relative to the previous Inventory.
• Agricultural Soil Management (N2O). Methodological recalculations in the current Inventory are associated
with the following improvements: (1) estimating emissions from 2013 to 2015 using a splicing method (other
than biosolids which are estimated with a Tier 1 method for the entire time series); (2) correcting an omission of
pasture, range, and paddock (PRP) manure N input from 1990 to 2012 in Alaska and Hawaii for indirect soil
N20 emission; and (3) correcting a double-counting of other organic amendments from 1990 to 2012 in the Tier
1 method for direct N20 emissions. These changes resulted in an average decrease in emissions of 0.7 percent
from 1990 to 2015 relative to the previous Inventory. These changes resulted in an average annual decrease in
emissions of 2.0 MMT CO2 Eq. relative to the previous Inventory.
Table 9-1: Revisions to U.S. Greenhouse Gas Emissions (MMT CO2 Eq.)
Average
Gas/Source
1990
2005
2012
2013
2014
2015
Annual
Change
CO2
13.S
18.9
21.6
27.7
25.0
37.9
19.0
Fossil Fuel Combustion
15.-
12.1
5.2
5.8
3.8
9.5
14.0
Electricity Generation
NC
NC
NC
NC
NC
+
+
Transportation
(26.6)
(31.3)
(34.9)
(35.4)
(25.7)
(0.9)
(27.6)
Industrial
32.1
39.8
35.5
36.4
24.7
13.8
37.3
Residential
AT
NC
+
0.1
(0.1)
(2.8)
(0.1)
Commercial
10.0
3.6
4.6
4.7
4.9
(0.6)
4.5
U.S. Territories
NC
NC
NC
NC
NC
NC
NC
Non-Energy Use of Fuels
2.0
2.8
6.5
9.5
8.8
9.6
2.9
Natural Gas Systems
(8.0)
(7.5)
(10.8)
(12.5)
(15.3)
(16.0)
(10.3)
Cement Production
N<
NC
NC
NC
NC
NC
NC
Lime Production
N<
NC
NC
NC
NC
NC
NC
Other Process Uses of Carbonates
N<
NC
NC
NC
NC
+
+
Glass Production
N<
NC
NC
NC
NC
NC
NC
Soda Ash Production
(1.4)
(1.3)
(1.1)
(1.1)
(1.1)
(1.1)
(1.3)
Carbon Dioxide Consumption
N<
NC
NC
NC
NC
NC
NC
Incineration of Waste
N<
NC
+
+
NC
+
+
Titanium Dioxide Production
N<
NC
NC
NC
NC
NC
NC
Aluminum Production
N<
NC
NC
NC
NC
NC
NC
Iron and Steel Production & Metallurgical Coke
N<
NC
NC
NC
(0.4)
(1.2)
(0.1)
Production
Ferroalloy Production
N<
NC
NC
NC
NC
NC
NC
Ammonia Production
N<
NC
NC
NC
NC
(0.2)
+
Urea Consumption for Non-Agricultural Purposes
N<
NC
+
0.1
0.2
3.0
0.1
Phosphoric Acid Production
N<
NC
NC
NC
NC
NC
NC
Petrochemical Production
(0.1)
(0.2)
NC
NC
NC
NC
(0.1)
Silicon Carbide Production and Consumption
N<
NC
NC
NC
NC
NC
NC
Lead Production
N<
NC
NC
NC
NC
NC
NC
Zinc Production
N<
NC
NC
NC
NC
+
+
Petroleum Systems
5.8
13.1
21.8
26.0
29.3
34.4
13.8
Magnesium Production and Processing
N<
NC
NC
NC
NC
NC
NC
Liming
N<
NC
NC
NC
NC
+
+
Urea Fertilization
N<
NC
+
(0.1)
(0.2)
(0.1)
+
Biomass - Woodb
NC
NC
NC
NC
1.3
2.7
0.2
9-4 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
International Bunker Fuelsb
NC
NC
NC
NC
0.2
0.1
+
Biomass - Ethanolb
NC
NC
NC
NC
NC
NC
NC
CH4
(2.7)
(1.7)
(4.8)
0.8
6.2
8.3
(1.6)
Stationary Combustion
0.2
0.5
0.7
0.7
0.7
0.8
0.4
Mobile Combustion
4.2
3.8
1.8
1.6
1.3
1.1
3.2
Coal Mining
NC
NC
NC
NC
(0.3)
0.3
+
Abandoned Underground Coal Mines
Ni
NC
NC
NC
NC
NC
NC
Natural Gas Systems
(0.4)
0.3
0.6
0.4
1.7
2.0
0.1
Petroleum Systems
(13.1 )
(11.3)
(11.1)
(5.6)
(2.0)
(0.5)
(10.9)
Abandoned Oil and Gas Wells*
NC*
NC*
NC*
NC*
NC*
NC*
NC*
Petrochemical Production
Nl
NC
NC
NC
NC
NC
NC
Silicon Carbide Production and Consumption
N<
NC
NC
NC
NC
NC
NC
Iron and Steel Production & Metallurgical Coke
N<
NC
NC
NC
NC
NC
NC
Production
Ferroalloy Production
N<
NC
NC
NC
NC
NC
NC
Enteric Fermentation
N<
NC
NC
NC
NC
NC
NC
Manure Management
N<
NC
NC
NC
+
NC
+
Rice Cultivation
N<
NC
NC
0.2
1.4
1.1
0.1
Field Burning of Agricultural Residues
N<
NC
NC
NC
NC
NC
NC
Landfills
N<
(1.6)
(3.8)
(3.5)
(3.8)
(4.0)
(1.2)
Wastewater Treatment
+ /
(0.2)
+
+
0.2
0.3
(0.1)
Composting
N<
NC
NC
NC
NC
NC
NC
Incineration of Waste
N<
NC
NC
NC
NC
NC
NC
International Bunker Fuelsh
NC
NC
NC
NC
NC
NC
NC
N2O
(5.0)
(4.2)
(5.5)
27.1
25.0
44.1
(0.2)
Stationary Combustion
(0.8)
(2.7)
(4.6)
(4.4)
(4.5)
(5.2)
(2.3)
Mobile Combustion
0.'
2.7
3.4
3.5
3.6
3.7
2.2
Adipic Acid Production
N(
NC
NC
NC
NC
+
+
Nitric Acid Production
N(
NC
+
NC
NC
+
+
Manure Management
N(
NC
NC
NC
NC
NC
NC
Agricultural Soil Management
(6.1)
(6.3)
(6.2)
26.1
24.0
43.7
(2.0)
Field Burning of Agricultural Residues
N<
NC
NC
NC
NC
NC
NC
Wastewater Treatment
N(
NC
(0.2)
(0.2)
(0.1)
(0.1)
+
N2O from Product Uses
N(
NC
NC
NC
NC
NC
NC
Caprolactam, Glyoxal, and Glyoxylic Acid
NC*
NC*
NC*
NC*
NC*
NC*
NC*
Production*
Incineration of Waste
N(
NC
NC
NC
NC
NC
NC
Composting
N(
NC
NC
NC
NC
NC
NC
Semiconductor Manufacture
N(
NC
+
+
+
+
+
International Bunker Fuelsb
NC
NC
NC
NC
+
+
+
HFCs
NC
+
0.1
0.1
0.1
0.1
+
Substitution of Ozone Depleting Substances
N<
+
0.1
0.1
0.1
0.1
+
HCFC-22 Production
N(
NC
NC
NC
NC
NC
NC
Semiconductor Manufacture
N(
+
+
+
+
+
+
Magnesium Production and Processing
N(
NC
NC
NC
NC
NC
NC
PFCs
NC
0.1
(0.1)
+
(0.1)
(0.1)
+
Aluminum Production
N(
NC
NC
NC
NC
NC
NC
Semiconductor Manufacture
N(
0.1
(0.1)
+
(0.1)
(0.1)
+
Substitution of Ozone Depleting Substances
N(
NC
NC
NC
NC
NC
NC
SF«
+
+
(0.2)
(0.1)
(0.2)
0.1
+
Electrical Transmission and Distribution
+/
0.1
(0.2)
(0.1)
(0.2)
0.1
+
Semiconductor Manufacture
N<
+
+
+
+
+
+
Magnesium Production and Processing
N(
NC
NC
NC
+
+
+
NF3
NC
+
+
+
+
+
+
Semiconductor Manufacture
N(
+
+
+
+
+
+
Net Change in Total Emissions
6.1
13.1
11.1
55.6
55.9
90.4
Percent Change
0.1%
0.2%
0.2%
0.8%
0.8%
1.4%
Note: Net change in total emissions presented without LULUCF.
NC (No Change)
+ Absolute value does not exceed 0.05 MMT CO2 Eq. or 0.05 percent.
* Indicates a new source for the current Inventory year.
a Not included in emissions total.
Recalculations and Improvements 9-5
-------
b Sinks are only included in net emissions total.
Notes: Totals may not sum due to independent rounding. Parentheses indicate negative values.
1 Table 9-2: Revisions to U.S. Greenhouse Gas Emissions and Removals (Net Flux) from Land
2 Use, Land-Use Change, and Forestry (MMT CO2 Eg.)
Average
Annual
Land Use Category
1990
2005
2012
2013
2014
2015
Change
Forest Land Remaining Forest Land
+
+
+
+
(0.1)
18.6
0.7
Changes in Forest Carbon Stocka
+vi\!
0.1
+
+
+
+
+
Non-CCh Emissions from Forest Fires
+:?V.;
+/
+
+
(0.1)
18.6
0.7
N2O Fluxes from Forest Soilsb
N<
N<
NC
NC
NC
NC
NC
Non-CCh Emissions from Drained
Organic Soils
+ /
+
+
+
+
+
Land Converted to Forest Land
NC
(0.2)
0.3
0.3
0.2
0.2
+
Changes in Forest Carbon Stockc
N<
(0.2)
0.3
0.3
0.2
0.2
+
Cropland Remaining Cropland
NC
NC
NC
8.2
6.7
11.7
1.0
Changes in Agricultural Carbon Stock
N<
N<
NC
8.2
6.7
11.7
1.0
Land Converted to Cropland
NC
NC
NC
0.6
0.5
0.4
0.1
Changes in Agricultural Carbon Stockd
N<
N<
NC
0.6
0.5
0.4
0.1
Grassland Remaining Grassland
NC
NC
NC
16.7
12.8
30.3
2.3
Changes in Agricultural Carbon Stock
N<
N<
NC
16.7
12.8
30.5
2.3
Non-CCh Emissions from Grass Fires
N<
N(
NC
NC
NC
(0.2)
+
Land Converted to Grassland
NC
NC
NC
1.4
1.0
2.9
0.2
Changes in Agricultural Carbon Stockd
N<
N<
NC
1.4
1.0
2.9
0.2
Wetlands Remaining Wetlands
NC
NC
NC
NC
NC
NC
NC
Changes in Organic Soil Carbon Stocks in
Peatlands
N<
N(
NC
NC
NC
NC
NC
Changes in Mineral and Organic Soil
Carbon Stocks in Coastal Wetlands
N<
N(
NC
NC
NC
NC
NC
CFLi Emissions from Coastal Wetlands
Remaining Coastal Wetlands
N<
N(
NC
NC
NC
NC
NC
N2O Emissions from Coastal Wetlands
Remaining Coastal Wetlands
N<
N(
NC
NC
NC
NC
NC
Non-C02 Emissions from Peatlands
Remaining Peatlands
N<
N(
NC
NC
NC
NC
NC
Land Converted to Wetlands
+
+
+
+
+
+
+
Changes in Coastal Wetland Carbon
Stocke
+
+
+
+
+
+
+
CFLi Emissions from Land Converted to
Coastal Wetlands
N<
N<
NC
NC
NC
NC
NC
Settlements Remaining Settlements
NC
NC
NC
0.1
0.1
+
+
Changes in Settlement Soil Carbon Stock
N<
N(
NC
+
+
+
+
Changes in Urban Tree Carbon Stock
N<
N(
NC
NC
NC
NC
NC
Landfilled Yard Trimmings and Food
Scraps
N<
N(
NC
NC
NC
NC
NC
N2O Fluxes from Settlement Soilsf
N<
N(
NC
+
0.1
+
+
Land Converted to Settlements
NC
NC
NC
0.1
(0.1)
(0.2)
+
Changes in Settlement Soil Carbon Stockd
N<
N(
NC
0.1
(0.1)
(0.2)
+
LULUCF Emissions8
+
+
+
+
(0.1)
18.4
LULUCF Total Net Flux"
+
(0.2)
0.3
27.2
21.1
45.3
LULUCF Sector Total1
+
(0.2)
0.3
27.2
21.0
63.7
Percent Change
0.0°/..
0.0%
0.0%
3.6%
2.8%
8.4%
NC (No Change)
+ Absolute value does not exceed 0.05 MMT CO2 Eq. or 0.05 percent.
a Includes the net changes to carbon stocks stored in all forest ecosystem pools (including drained and undrained organic
soils) and harvested wood products.
b Estimates include emissions from N fertilizer additions on both Forest Land Remaining Forest Land and Land Converted to
Fore st Land.
9-6 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
c Includes the net changes to carbon stocks stored in all forest ecosystem pools (excludes drained organic soils which are
included in the flux from Forest Land Remaining Forest Land because it is not possible to separate the activity data at this
time).
d 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.
e Includes carbon stock changes for land converted to vegetated coastal wetlands.
f Estimates include emissions from N fertilizer additions on both Settlements Remaining Settlements and Land Converted to
Settlements.
B LULUCF emissions include the CH4 and N2O 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 N2O emissions from Forest Soils and Settlement Soils.
h LULUCF Carbon Stock Change includes any C stock gains and losses from all land use and land use conversion categories.
1 The LULUCF Sector Net Total is the net sum of all CH4 and N2O emissions to the atmosphere plus net carbon stock
changes.
Notes: Totals may not sum due to independent rounding. Parentheses indicate net sequestration.
Recalculations and Improvements 9-7
-------
1 10. References
2 Executive Summary
3 BEA (2017) 2016 Comprehensive Revision of the National Income and Product Accounts: Current-dollar and
4 "real" GDP, 1929-2016. Bureau of Economic Analysis (BEA), U.S. Department of Commerce, Washington, D.C.
5 Available online at: .
6 Duffield, J. (2006) Personal communication. Jim Duffield, Office of Energy Policy and New Uses, U.S. Department
7 of Agriculture, and Lauren Flinn, ICF International. December 2006.
8 EIA(2017) Electricity Generation. Monthly Energy Review, February 2017. Energy Information Administration,
9 U.S. Department of Energy, Washington, D.C. DOE/EIA-0035(2017/02).
10 EIA (2016) Electricity in the United States. Electricity Explained. Energy Information Administration, U.S.
11 Department of Energy, Washington, D.C. Available online at:
12 .
13 EPA (2016a) 1970-2016 Average annual emissions, all criteria pollutants in MS Excel. National Emissions
14 Inventory (NEI) Air Pollutant Emissions Trends Data. Office of Air Quality Planning and Standards. Last Modified
15 December 2016. Available online at: .
17 EPA (2016b) Advancing Sustainable Materials Management: Facts and Figures 2014. December 2016. Available
18 online at: .
19 IEA (2017) CO 2 Emissions from Fossil Fuel Combustion - Highlights. International Energy Agency. Available
20 online at: .
22 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
23 Assessment Report of the Intergovernmental Panel on Climate Change. [Stacker, T.F., D. Qin, G.-K., Plattner, M.
24 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
25 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
26 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
27 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
28 M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.)]. Cambridge University Press. Cambridge, United
29 Kingdom 996 pp.
30 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
31 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
32 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
33 IPCC (1996) Climate Change 1995: The Science of Climate Change. Intergovernmental Panel on Climate Change.
34 [J.T. Houghton, L.G. Meira Filho, B.A. Callander, N. Harris, A. Kattenberg, and K. Maskell (eds.)]. Cambridge
35 University Press. Cambridge, United Kingdom.
References 10-1
-------
1 NOAA/ESRL (2017a) Trends in Atmospheric Carbon Dioxide. Available online at:
2 . 21 November 2017.
3 NOAA/ESRL (2017b) Trends in Atmospheric Methane. Available online at:
4 . 22 December 2017.
5 NOAA/ESRL (2017c) Nitrous Oxide (N2O) hemispheric and global monthly means from the NOAA/ESRL
6 Chromatograph for Atmospheric Trace Species data from baseline observatories (Barrow, Alaska; Summit,
7 Greenland; Niwot Ridge, Colorado; Mauna Loa, Hawaii; American Samoa; South Pole). Available online at:
8 . 26 December 2017.
9 UNFCCC (2014) Report of the Conference of the Parties on its Nineteenth Session, Held in Warsaw from 11 to 23
10 November 2013. (FCCC/CP/2013/10/Add.3). January 31, 2014. Available online at:
11 .
12 U.S. Census Bureau (2017) U.S. Census Bureau International Database (IDB). Available online at:
13 .
14 Introduction
15 Carbon Dioxide Information Analysis Center (CDIAC) (2016) Recent Greenhouse Gas Concentrations. April 2016.
16 Available online at: .
17 EPA (2009) Technical Support Document for the Endangerment and Cause or Contribute Findings for Greenhouse
18 Gases under Section 202(a) of the Clean Air Act. U.S. Environmental Protection Agency. December 2009.
19 EPA (2017) Climate Change. Available online at: . 6 April 2017.
20 IPCC (2014) Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group III to the Fifth
21 Assessment Report of the Intergovernmental Panel on Climate Change [Edenhofer, O., R. Pichs-Madruga, Y.
22 Sokona, J. Minx, E. Farahani, S. Kadner, K. Seyboth, A. Adler, I. Baum, S. Brunner, P. Eickemeier, B. Kriemann, J.
23 Savolainen, S. Schlomer, C. von Stechow, and T. Zwickel (eds.)]. Cambridge University Press, Cambridge, United
24 Kingdom and New York, NY, USA, 1435 pp.
25 IPCC (2013) Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth
26 Assessment Report of the Intergovernmental Panel on Climate Change [Stacker, T.F., D. Qin, G.-K. Plattner, M.
27 Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press,
28 Cambridge, United Kingdom and New York, NY, USA, 1535 pp.
29 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
30 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
31 M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.)]. Cambridge University Press. Cambridge, United
32 Kingdom 996 pp.
33 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
34 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
35 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
36 IPCC (2001) Climate Change 2001: The Scientific Basis. Intergovernmental Panel on Climate Change. [J.T.
37 Houghton, Y. Ding, D.J. Griggs, M. Noguer, P.J. van der Linden, X. Dai, C.A. Johnson, and K. Maskell (eds.)].
38 Cambridge University Press. Cambridge, United Kingdom.
39 IPCC (1999) Aviation and the Global Atmosphere. Intergovernmental Panel on Climate Change. [J.E. Penner, et al.
40 (eds.)]. Cambridge University Press. Cambridge, United Kingdom.
41 IPCC/TEAP (2005) Special Report: Safeguarding the Ozone Layer and the Global Climate System, Chapter 4:
42 Refrigeration. 2005. Available online at:
10-2 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Jacobson, M.Z. (2001) "Strong Radiative Heating Due to the Mixing State of Black Carbon in Atmospheric
2 Aerosols." Nature, 409:695-697.
3 NOAA (2017) Vital Signs of the Planet. Available online at: . Accessed on 9
4 January 2017.
5 NOAA/ESRL (2017a) Trends in Atmospheric Carbon Dioxide. Available online at:
6 . 21 November 2017.
7 NOAA/ESRL (2017b) Trends in Atmospheric Methane. Available online at:
8 . 22 December 2017.
9 NOAA/ESRL (2017c) Nitrous Oxide (N2O) hemispheric and global monthly means from the NOAA/ESRL
10 Chromatograph for Atmospheric Trace Species data from baseline observatories (Barrow, Alaska; Summit,
11 Greenland; Niwot Ridge, Colorado; Mauna Loa, Hawaii; American Samoa; South Pole). Available online at:
12 . 26 December 2017.
13 NOAA/ESRL (2017d) Sulfur Hexafluoride (SFq) hemispheric and global monthly means from the NOAA/ESRL
14 Chromatograph for Atmospheric Trace Species data from baseline observatories (Barrow, Alaska; Summit,
15 Greenland; Niwot Ridge, Colorado; Mauna Loa, Hawaii; American Samoa; South Polej. Available online at:
16 . 10 January 2018.
17 NRC (2012) Climate Change: Evidence, Lmpacts, and Choices. Available online at:
18 .
19 UNEP/WMO (1999) Information Unit on Climate Change. Framework Convention on Climate Change. Available
20 online at: .
21 UNFCCC (2014) Report of the Conference of the Parties on its nineteenth session, held in Warsaw from 11 to 23
22 November 2013. (FCCC/CP/2013/10/Add.3). January 31, 2014. Available online at:
23 .
24 USGCRP (2017) Climate Science Special Report: Fourth National Climate Assessment, Volume 1. [Wuebbles, D.J.,
25 D.W. Fahey, K.A. Hibbard, D.J. Dokken, B.C. Stewart, and T.K. Maycock (eds.)]. U.S. Global Change Research
26 Program, Washington, DC, USA, 470 pp, doi: 10.7930/J0J964J6. Available online at:
27 .
28 Trends in Greenhouse Gas Emissions
29 BEA (2017) 2016 Comprehensive Revision of the National Income and Product Accounts: Current-dollar and
30 "real" GDP, 1929-2016. Bureau of Economic Analysis (BEA), U.S. Department of Commerce, Washington, D.C.
31 Available online at: .
32 Duffield, J. (2006) Personal communication. Jim Duffield, Office of Energy Policy and New Uses, U.S. Department
33 of Agriculture, and Lauren Flinn, ICF International. December 2006.
34 EIA (2017a) Monthly Energy Review, October 2017. Energy Information Administration, U.S. Department of
35 Energy, Washington, D.C. DOE/EIA-0035(2017/10).
36 EIA (2017b) "Per capita residential electricity sales in the U.S. have fallen since 2010." Available online at:
37 .
38 EIA (2017c) Fuel Oil and Kerosene Sales 2016. Energy Information Administration, U.S. Department of Energy,
39 Washington, D.C. Available online at: .
40 EIA (2017d) "Renewable generation capacity expected to account for most 2016 capacity additions." Available
41 online at: .
42 EPA (2016a) Light-Duty Automotive Technology, Carbon Dioxide Emissions, and Fuel Economy Trends: 1975 -
43 2016. Office of Transportation and Air Quality, U.S. Environmental Protection Agency. Available online at:
44 .
References 10-3
-------
1 EPA (2016b) 1970 - 2016 Average annual emissions, all criteria pollutants in MS Excel. National Emissions
2 Inventory (NEI) Air Pollutant Emissions Trends Data. Office of Air Quality Planning and Standards, December
3 2016. Available online at: .
4 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
5 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
6 M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.)]. Cambridge University Press. Cambridge, United
7 Kingdom 996 pp.
8 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
9 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
10 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
11 U.S. Census Bureau (2017) U.S. Census Bureau International Database (IDB). Available online at:
12 .
13 USDA (2016) Personal communication. Claudia Hitaj, USDA Economic Research Service, and John Steller, U.S.
14 EPA. December 2016.
15 Energy
16 EIA (2017a) Monthly Energy Review, October 2017, Energy Information Administration, U.S. Department of
17 Energy, Washington, DC. DOE/EIA-0035(2017/10).
18 IEA (2017) CO 2 Emissions from Fossil Fuel Combustion - Highlights. International Energy Agency. Available
19 online at: .
21 Carbon Dioxide Emissions from Fossil Fuel Combustion
22 AAR (2008 through 2017) Railroad Facts. Policy and Economics Department, Association of American Railroads,
23 Washington, D.C. Obtained from Clyde Crimmel at AAR.
24 AISI (2004 through 2016) Annual Statistical Report, American Iron and Steel Institute, Washington, D.C.
25 APTA (2007 through 2016) Public Transportation Fact Book. American Public Transportation Association,
26 Washington, D.C. Available online at: .
27 APTA (2006) Commuter Rail National Totals. American Public Transportation Association, Washington, D.C.
28 Available online at: .
29 BEA (2017) Table 1.1.6. Real Gross Domestic Product, Chained 2009 Dollars. Bureau of Economic Analysis
30 (BEA), U.S. Department of Commerce, Washington, D.C. March 2017. Available online at:
31 .
33 Benson, D. (2002 through 2004) Unpublished data. Upper Great Plains Transportation Institute, North Dakota State
34 University and American Short Line & Regional Railroad Association.
35 Browning, L. (2017) Updated Methodology for Estimating CH4 and N2O Emissions from Highway Vehicle
36 Alternative Fuel Vehicles. Technical Memo, October 2017.
37 Coffeyville Resources Nitrogen Fertilizers (2012) Nitrogen Fertilizer Operations. Available online at:
38 . Dakota Gasification Company (2006) C02
39 Pipeline Route and Designation Information. Bismarck, ND. Available online at:
40 .
41 DHS (2008) Email Communication. Elissa Kay, Department of Homeland Security and Joe Aamidor, ICF
42 International. January 11, 2008.
10-4 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 DLA Energy (2017) Unpublished data from the Fuels Automated System (FAS). Defense Logistics Agency Energy,
2 U.S. Department of Defense. Washington, D.C.
3 DOC (1991 through 2017) Unpublished Report of Bunker Fuel Oil Laden on Vessels Cleared for Foreign Countries.
4 Form-563. Foreign Trade Division, Bureau of the Census, U.S. Department of Commerce. Washington, D.C.
5 DOE (1993 through 2016) Transportation Energy Data Book Edition 35. Office of Transportation Technologies,
6 Center for Transportation Analysis, Energy Division, Oak Ridge National Laboratory. ORNL-6978.
7 DOE (2012) 2010 Worldwide Gasification Database. National Energy Technology Laboratory and Gasification
8 Technologies Council. Available online at:
9 . Accessed on 15 March
10 2012.
11 DOT (1991 through 2016) Airline Fuel Cost and Consumption. U.S. Department of Transportation, Bureau of
12 Transportation Statistics, Washington, D.C. DAI-10. Available online at: .
13 Eastman Gasification Services Company (2011) Project Data on Eastman Chemical Company's Chemicals-fro m-
14 Coal Complex in Kingsport, TN. Available online at:
15 .
16 EIA (2017a) Monthly Energy Review, October 2017, Energy Information Administration, U.S. Department of
17 Energy, Washington, DC. DOE/EIA-0035(2017/10).
18 EIA (2017b) International Energy Statistics 1980-2016. Energy Information Administration, U.S. Department of
19 Energy. Washington, D.C. Available online at: .
20 EIA (2017c) Quarterly Coal Report: -April-June 2017. Energy Information Administration, U.S. Department of
21 Energy. Washington, D.C. DOE/EIA-0121.
22 EIA (2017d) Natural Gas Annual 2016. Energy Information Administration, U.S. Department of Energy.
23 Washington, D.C. DOE/EIA-0131(06).
24 EIA (2017e) "Renewable generation capacity expected to account for most 2016 capacity additions." Available
25 online at: .
26 EIA (2017f) "Per capita residential electricity sales in the U.S. have fallen since 2010." Available online at:
27 .
28 EIA (1991 through 2016) Fuel Oil and Kerosene Sales. Energy Information Administration, U.S. Department of
29 Energy. Washington, D.C. Available online at: .
30 EIA (2014) Indicators: CO2 Emissions. International Energy Statistics 2014. Energy Information Administration,
31 U.S. Department of Energy. Washington, D.C. Available online at:
32 .
33 EIA (2017) Alternative Fuels Data Tables. Energy Information Administration, U.S. Department of Energy.
34 Washington, D.C. Available online at: .
35 EIA (2009a) Emissions of Greenhouse Gases in the United States 2008, Draft Report. Office of Integrated Analysis
36 and Forecasting, Energy Information Administration, U.S. Department of Energy. Washington, D.C. DOE-EIA-
37 0573(2009).
38 EIA (2009b) Manufacturing Consumption of Energy 2006. Energy Information Administration, U.S. Department of
39 Energy. Washington, D.C. Released July, 2009.
40 EIA (2008) Historical Natural Gas Annual, 1930 - 2008. Energy Information Administration, U.S. Department of
41 Energy. Washington, D.C.
42 EIA (2007) Personal Communication. Joel Lou, Energy Information Administration and Aaron Beaudette, ICF
43 International. Residual and Distillate Fuel Oil Consumption for Vessel Bunkering (Both International and Domestic)
44 for American Samoa, U.S. Pacific Islands, and Wake Island. October 24, 2007.
45 EIA (2001) U.S. Coal, Domestic and International Issues. Energy Information Administration, U.S. Department of
46 Energy. Washington, D.C. March 2001.
References 10-5
-------
1 EPA (2017a) Acid Rain Program Dataset 1996-2016. Office of Air and Radiation, Office of Atmospheric Programs,
2 U.S. Environmental Protection Agency, Washington, D.C.
3 EPA (2017b) Motor Vehicle Emissions Simulator (Moves) 2014a. Office of Transportation and Air Quality, U.S.
4 Environmental Protection Agency. Available online at: .
5 EPA (2017c) Annual Certification Test Results Report. Office of Transportation and Air Quality, U.S.
6 Environmental Protection Agency. Available online at: .
8 EPA (2016a) Light-Duty Automotive Technology, Carbon Dioxide Emissions, and Fuel Economy Trends: 1975 -
9 2016. Office of Transportation and Air Quality, U.S. Environmental Protection Agency. Available online at:
10 .
11 EPA (2010) Carbon Content Coefficients Developed for EPA's Mandatory Reporting Rule. Office of Air and
12 Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
13 Erickson, T. (2003) Plains CO2 Reduction (PCOR) Partnership. Presented at the Regional Carbon Sequestration
14 Partnership Meeting Pittsburgh, Pennsylvania, Energy and Environmental Research Center, University of North
15 Dakota. November 3, 2003. Available online at: .
22 FHWA (2015) Off-Highway and Public-Use Gasoline Consumption Estimation Models Used in the Federal
23 Highway Administration, Publication Number FHWA-PL-17-012. Available online at:
24 .
25 Fitzpatrick, E. (2002) The Weyburn Project: A Model for International Collaboration. Available online at:
26 .
27 FRB (2017) Industrial Production and Capacity Utilization. Federal Reserve Statistical Release, G.17,Federal
28 Reserve Board. Available online at: .
29 Gaffney, J. (2007) Email Communication. John Gaffney, American Public Transportation Association and Joe
30 Aamidor, ICF International. December 17, 2007.
31 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
32 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
33 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
34 Marland, G. and A. Pippin (1990) "United States Emissions of Carbon Dioxide to the Earth's Atmosphere by
35 Economic Activity." Energy Systems and Policy, 14(4):323.
36 SAIC/EIA (2001) Monte Carlo Simulations of Uncertainty in U.S. Greenhouse Gas Emission Estimates. Final
37 Report. Prepared by Science Applications International Corporation (SAIC) for Office of Integrated Analysis and
38 Forecasting, Energy Information Administration, U.S. Department of Energy. Washington, D.C. June 22, 2001.
39 U.S. Aluminum Association (USAA) (2008 through 2017) U.S. Primary Aluminum Production. U.S. Aluminum
40 Association, Washington, D.C.USAF (1998) Fuel Logistics Planning. U.S. Air Force: AFPAM23-221. May 1, 1998.
41 U.S. Census Bureau (2001 through 2011) Current Industrial Reports Fertilizer Materials and Related Products:
42 Annual Summary. Available online at: .
43 United States Geological Survey (USGS) (2017) 2017Mineral Commodity Summaries: Aluminum. U.S. Geological
44 Survey, Reston, VA.
10-6 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 USGS (2016a) Minerals Industry Surveys: Abrasives (Manufactured) in Third and Fourth Quarter of 2015. U.S.
2 Geological Survey, Reston, VA. January 2017. Available online at:
3 .
4 USGS (2015 and 2016) Mineral Commodity Summary: Titanium and Titanium Dioxide. U.S. Geological Survey,
5 Reston, VA.
6 USGS (2014 through 2016a) Mineral Commodity Summary, Lead. U.S. Geological Survey, Reston, VA.
7 USGS (2014 through 2016b)Minerals Yearbook: Nitrogen [Advance Release]. Available online at:
8 .
9 USGS (2016d) Mineral Industry Surveys: Silicon in October 2016. U.S. Geological Survey, Reston, VA. December
10 2016.
11 USGS (1991 through 2015a) Miner als Yearbook: Manufactured Abrasives Annual Report. U.S. Geological Survey,
12 Reston, VA. Available online at: .
13 USGS (1991 through 2015b) Miner als Yearbook: Titanium. U.S. Geological Survey, Reston, VA.
14 USGS (2015b) Mineral Industry Surveys: Silicon in June 2015. U.S. Geological Survey, Reston, VA. September
15 2015.
16 USGS (1991 through 2015c) Miner als Yearbook-Iron and Steel Scrap. U.S. Geological Survey, Reston, VA.
17 USGS (2014) Mineral Industry Surveys: Silicon in September 2014. U.S. Geological Survey, Reston, VA.
18 December 2014.
19 USGS (1996 through 2013) Minerals Yearbook: Silicon. U.S. Geological Survey, Reston, VA
20 USGS (1995 through 2013) Minerals Yearbook: Lead Annual Report. U.S. Geological Survey, Reston, VA.
21 USGS (1991b through 2013) Miner als Yearbook: Silicon Annual Report. U.S. Geological Survey, Reston, VA.
22 Available online at: .
23 USGS (1995, 1998, 2000, 2001, 2002) Minerals Yearbook: Aluminum Annual Report. U.S. Geological Survey,
24 Reston, VA.
25 Stationary Combustion (excluding C02)
26 EIA (2017a) Monthly Energy Review, October 2017. Energy Information Administration, U.S. Department of
27 Energy. Washington, D.C. DOE/EIA-0035(2017/10).
28 EIA (2017b) International Energy Statistics 1980-2014. Energy Information Administration, U.S. Department of
29 Energy. Washington, D.C. Available online at: .
30 EPA (2017a) Acid Rain Program Dataset 1996-2016. Office of Air and Radiation, Office of Atmospheric Programs,
31 U.S. Environmental Protection Agency, Washington, D.C.
32 EPA (2017b). Motor Vehicle Emissions Simulator (MOVES) 2014a. Office of Transportation and Air Quality, U.S.
33 Environmental Protection Agency. Available online at: .
34 EPA (1997) Compilation of Air Pollutant Emission Factors, AP-42. Office of Air Quality Planning and Standards,
35 U.S. Environmental Protection Agency. Research Triangle Park, NC. October 1997.
36 FHWA (1996 through 2016) Highway Statistics. Federal Highway Administration, U.S. Department of
37 Transportation, Washington, D.C. Report FHWA-PL-96-023-annual. Available online at:
38 .
39 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
40 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
41 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
References 10-7
-------
1 SAIC/EIA (2001) Monte Carlo Simulations of Uncertainty in U.S. Greenhouse Gas Emission Estimates. Final
2 Report. Prepared by Science Applications International Corporation (SAIC) for Office of Integrated Analysis and
3 Forecasting, Energy Information Administration, U.S. Department of Energy. Washington, D.C. June 22, 2001.
4 Mobile Combustion (excluding C02)
5 AAR (2008 through 2016) Railroad Facts. Policy and Economics Department, Association of American Railroads,
6 Washington, D.C. Obtained from Clyde Crimmel at AAR.
7 ANL (2006) Argonne National Laboratory (2006) GREET model Version 1.7. June 2006.
8 ANL (2015) The Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation Model
9 (GREET_1 2015). Argonne National Laboratory. October 2015. Available online at: .
10 APTA (2007 through 2016) Public Transportation Fact Book. American Public Transportation Association,
11 Washington, D.C. Available online at: .
12 APTA (2006) Commuter Rail National Totals. American Public Transportation Association, Washington, D.C.
13 Available online at: .
14 BEA (1991 through 2015) Unpublished BE-36 survey data. Bureau of Economic Analysis, U.S. Department of
15 Commerce. Washington, D.C.
16 Benson, D. (2002 through 2004) Personal communication. Unpublished data developed by the Upper Great Plains
17 Transportation Institute, North Dakota State University and American Short Line & Regional Railroad Association.
18 Browning, L. (2017) "Updated New On-highway CH4 and N20 Emission Factors for GHG Inventory". Technical
19 Memo, October 18, 2017.
20 Browning, L. (2017) "Updated Non-Highway CH4 and N20 Emission Factors for U.S. GHG Inventory". Technical
21 Memo, October 17, 2017.
22 Browning, L. (2017) "Updated Methodology for Estimating CH4 and N20 Emissions from Highway Vehicle
23 Alternative Fuel Vehicles". Technical Memo, October 18, 2017.
24 Browning, L. (2009) Personal communication with Lou Browning, "Suggested New Emission Factors for Marine
25 Vessels," ICF International.
26 Browning, L. (2005) Personal communication with Lou Browning, Emission control technologies for diesel
27 highway vehicles specialist, ICF International.
28 DHS (2008) Email Communication. Elissa Kay, Department of Homeland Security and Joe Aamidor, ICF
29 International. January 11, 2008.
30 DLA Energy (2016) Unpublished data from the Defense Fuels Automated Management System (DFAMS). Defense
31 Energy Support Center, Defense Logistics Agency, U.S. Department of Defense. Washington, D.C.
32 DOC (1991 through 2017) Unpublished Report of Bunker Fuel Oil Laden on Vessels Cleared for Foreign Countries.
33 Form-563. Foreign Trade Division, Bureau of the Census, U.S. Department of Commerce. Washington, D.C.
34 DOE (1993 through 2017) Transportation Energy Data Book. Office of Transportation Technologies, Centerfor
35 Transportation Analysis, Energy Division, Oak Ridge National Laboratory. ORNL-6978.
36 DOT (1991 through 2017) Airline Fuel Cost and Consumption. U.S. Department of Transportation, Bureau of
37 Transportation Statistics, Washington, D.C. DAI-10. Available online at: .
38 EDTA (2017) Electric Drive Sales Dashboard. Electric Drive Transportation Association, Washington, D.C.
39 Available at: .
40 EIA (2017) Monthly Energy Review, October 2017, Energy Information Administration, U.S. Department of
41 Energy, Washington, D.C. DOE/EIA-0035(2017/02).
42 EIA (1991 through 2017) Fuel Oil and Kerosene Sales. Energy Information Administration, U.S. Department of
43 Energy. Washington, D.C. Available online at: .
10-8 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 EIA (2016) "Table 3.1: World Petroleum Supply and Disposition." International Energy Annual. Energy
2 Information Administration, U.S. Department of Energy. Washington, D.C. Available online at:
3 .
4 EIA (2007 through 2016) Natural Gas Annual. Energy Information Administration, U.S. Department of Energy,
5 Washington, D.C. DOE/EIA-0131(11).
6 EIA (2011) Annual Energy Review 2010. Energy Information Administration, U.S. Department of Energy,
7 Washington, D.C. DOE/EIA-0384(2011). October 19, 2011.
8 EIA (2007) Personal Communication. Joel Lou, Energy Information Administration and Aaron Beaudette, ICF
9 International. Residual and Distillate Fuel Oil Consumption for Vessel Bunkering (Both International and Domestic)
10 for American Samoa, U.S. Pacific Islands, and Wake Island. October 24, 2007.
11 EIA (2002) Alternative Fuels Data Tables. Energy Information Administration, U.S. Department of Energy,
12 Washington, D.C. Available online at: .
13 EPA (2017bJ Motor Vehicle Emissions Simulator (Moves) 2014. Office of Transportation and Air Quality, U.S.
14 Environmental Protection Agency. Available online at: .
15 EPA (2017c) Annual Certification Test Results Report. Office of Transportation and Air Quality, U.S.
16 Environmental Protection Agency. Available online at: .
18 EPA (2017d) Confidential Engine Family Sales Data Submitted to EPA by Manufacturers. Office of Transportation
19 and Air Quality, U.S. Environmental Protection Agency.
20 EPA (2017) "1970 - 2015 Average annual emissions, all criteria pollutants in MS Excel." National Emissions
21 Inventory (NEI) Air Pollutant Emissions Trends Data. Office of Air Quality Planning and Standards. Available
22 online at: .
23 EPA (2000) Mobile6 Vehicle Emission Modeling Software. Office of Mobile Sources, U.S. Environmental
24 Protection Agency, Ann Arbor, Michigan.
25 EPA (1999a) Emission Facts: The History of Reducing Tailpipe Emissions. Office of Mobile Sources. May 1999.
26 EPA 420-F-99-017. Available online at: .
27 EPA (1999b) Regulatory Announcement: EPA's Program for Cleaner Vehicles and Cleaner Gasoline. Office of
28 Mobile Sources. December 1999. EPA420-F-99-051. Available online at:
29 .
30 EPA (1998) Emissions of Nitrous Oxide from Highway Mobile Sources: Comments on the Draft Inventory of U.S.
31 Greenhouse Gas Emissions and Sinks, 1990-1996. Office of Mobile Sources, Assessment and Modeling Division,
32 U.S. Environmental Protection Agency. August 1998. EPA420-R-98-009.
33 EPA (1997) Mobile Source Emission Factor Model (MOBILE5a). Office of Mobile Sources, U.S. Environmental
34 Protection Agency, Ann Arbor, Michigan.
35 EPA (1994a) Automobile Emissions: An Overview. Office of Mobile Sources. August 1994. EPA 400-F-92-007.
36 Available online at: .
37 EPA (1994b) Milestones in Auto Emissions Control. Office of Mobile Sources. August 1994. EPA 400-F-92-014.
38 Available online at: .
39 EPA (1993) Automobiles and Carbon Monoxide. Office of Mobile Sources. January 1993. EPA 400-F-92-005.
40 Available online at: .
41 Esser, C. (2003 through 2004) Personal Communication with Charles Esser, Residual and Distillate Fuel Oil
42 Consumption for Vessel Bunkering (Both International and Domestic) for American Samoa, U.S. Pacific Islands,
43 and Wake Island.
44 FAA (2017) Personal Communication between FAA and John Steller, Mausami Desai and Vincent Camobreco for
45 aviation emission estimates from the Aviation Environmental Design Tool (AEDT). January 2017.
References 10-9
-------
1 FHWA (1996 through 2017) Highway Statistics. Federal Highway Administration, U.S. Department of
2 Transportation, Washington, D.C. Report FHWA-PL-96-023-annual. Available online at:
3 .
4 FHWA (2016) Traffic Volume Trends. Federal Highway Administration, U.S. Department of Transportation,
5 Washington, D.C., December 2016. Available online at:
6 .
7 FHWA (2017) Off-Highway and Public-Use Gasoline Consumption Estimation Models Used in the Federal
8 Highway Administration, Publication Number FHWA-PL-17-012. Available online at:
9 .
10 Gaffney, J. (2007) Email Communication. John Gaffney, American Public Transportation Association and Joe
11 Aamidor, ICF International. December 17, 2007.
12 ICF (2006a) Revised Gasoline Vehicle EFs for LEV and Tier 2 Emission Levels. Memorandum from ICF
13 International to JohnDavies, Office of Transportation and Air Quality, U.S. Environmental Protection Agency.
14 November 2006.
15 ICF (2006b) Revisions to Alternative Fuel Vehicle (AFV) Emission Factors for the U.S. Greenhouse Gas Lnventory.
16 Memorandum from ICF International to John Davies, Office of Transportation and Air Quality, U.S. Environmental
17 Protection Agency. November 2006.
18 ICF (2004) Update of Methane and Nitrous Oxide Emission Factors for On-Highway Vehicles. Final Report to U.S.
19 Environmental Protection Agency. February 2004.
20 Lipman, T. and M. Delucchi (2002) "Emissions of Nitrous Oxide and Methane from Conventional and Alternative
21 Fuel Motor Vehicles." Climate Change, 53:477-516.
22 Santoni, G., B. Lee, E. Wood, S. Herndon, R. Miake-Lye, S. Wofsy, J. McManus, D. Nelson, M. Zahniser (2011)
23 Aircraft emissions of methane and nitrous oxide during the alternative aviation fuel experiment. Environ Sci
24 Technol. 2011 Aug 15; 45(16):7075-82.
25 U.S. Census Bureau (2000) Vehicle Lnventory and Use Survey. U.S. Census Bureau, Washington, D.C. Database
26 CD-EC97-VIUS.
27 Whorton, D. (2006 through 2014) Personal communication, Class II and III Rail energy consumption, American
28 Short Line and Regional Railroad Association.
29 Carbon Emitted from Non-Energy Uses of Fossil Fuels
30 ACC (2017) "U.S. Resin Production & Sales 2016 vs. 2015." Available online at:
31 .
32 ACC (2016a) "Guide to the Business of Chemistry, 2016," American Chemistry Council.
33 ACC (2016b) "U.S. Resin Production & Sales 2015 vs. 2014." Available online at:
34 .
35 ACC (2015a) "PIPS Year-End Resin Statistics for 2014 vs. 2013: Production, Sales and Captive Use." Available
36 online at: .
42 ACC (2013) "U.S. Resin Production & Sales: 2012 vs. 2011," American Chemistry Council. Available online at:
43
45 ACC (2012) "Guide to the Business of Chemistry, 2012," American Chemistry Council.
10-10 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 ACC (2003-2011) "PIPS Year-End Resin Statistics for 2010: Production, Sales and Captive Use." Available online
2 at: .
4 Bank of Canada (2017) Financial Markets Department Year Average of Exchange Rates. Available online at:
5 .
6 Bank of Canada (2016) Financial Markets Department Year Average of Exchange Rates. Available online at:
7 .
8 Bank of Canada (2014) Financial Markets Department Year Average of Exchange Rates. Available online at:
9 .
10 Bank of Canada (2013) Financial Markets Department Year Average of Exchange Rates. Available online at:
11 .
12 Bank of Canada (2012) Financial Markets Department Year Average of Exchange Rates. Available online at:
13 .
14 C&EN (2017) U.S. textile makers look for a revival, Chemical & Engineering News, American Chemical Society,
15 19 June. Available online at: .
16 EIA (2017a) ELA Manufacturing Consumption of Energy (MECS) 2014. U.S. Department of Energy, Energy
17 Information Administration, Washington, D.C.
18 EIA (2017b) Monthly Energy Review, August 2017. Energy Information Administration, U.S. Department of
19 Energy, Washington, D.C. DOE/EIA-0035 (2017/08).
20 EIA (2016) Monthly Energy Review, December 2016, Energy Information Administration, U.S. Department of
21 Energy, Washington, DC. DOE/EIA-0035 (2016/12).
22 EIA (2013) ELA Manufacturing Consumption of Energy (MECS) 2010. U.S. Department of Energy, Energy
23 Information Administration, Washington, D.C.
24 EIA (2010) ELA Manufacturing Consumption of Energy (MECS) 2006. U.S. Department of Energy, Energy
25 Information Administration, Washington, D.C.
26 EIA (2005) ELA Manufacturing Consumption of Energy (MECS) 2002. U.S. Department of Energy, Energy
27 Information Administration, Washington, D.C.
28 EIA (2001) ELA Manufacturing Consumption of Energy (MECS) 1998. U.S. Department of Energy, Energy
29 Information Administration, Washington, D.C.
30 EIA (1997) ELA Manufacturing Consumption of Energy (MECS) 1994. U.S. Department of Energy, Energy
31 Information Administration, Washington, D.C.
32 EIA (1994) ELA Manufacturing Consumption of Energy (MECS) 1991. U.S. Department of Energy, Energy
33 Information Administration, Washington, D.C.
34 EPA (2017) EPA's Pesticides Industry Sales and Usage, 2008 - 2012 Market Estimates. Available online at:
35
36 Accessed September 2017.
37 EPA (2016a) "1970 - 2016 Average annual emissions, all criteria pollutants in MS Excel." National Emissions
38 Inventory (NEI) Air Pollutant Emissions Trends Data. Office of Air Quality Planning and Standards, December
39 2016. Available online at: .
40 EPA (2016b) Advancing Sustainable Materials Management: 2014 Facts and Figures Fact Sheet. Office of Solid
41 Waste and Emergency Response, U.S. Environmental Protection Agency, Washington, D.C. Available online at:
42 .
43 EPA (2016c) Resource Conservation and Recovery Act (RCRA) Lnfo, Biennial Report, GM Form (Section 2- Onsite
44 Management) and WR Form.
References 10-11
-------
1 EPA (2015) Resource Conservation and Recovery Act (RCRA) Info, Biennial Report, GM Form (Section 2- Onsite
2 Management) and WR Form.
3 EPA (2014a) Municipal Solid Waste in the United States: 2012 Facts and Figures. Office of Solid Waste and
4 Emergency Response, U.S. Environmental Protection Agency, Washington, D.C. Available online at:
5 .
6 EPA (2014b) Chemical Data Access Tool (CDAT). U.S. Environmental Protection Agency, June 2014. Available
7 online at: . Accessed January 2015.
8 EPA (2013a) Municipal Solid Waste in the United States: 2011 Facts and Figures. Office of Solid Waste and
9 Emergency Response, U.S. Environmental Protection Agency, Washington, D.C. Available online at:
10 .
11 EPA (2013b) Resource Conservation and Recovery Act (RCRA) Info, Biennial Report, GMForm (Section 2- Onsite
12 Management) and WR Form.
13 EPA (2011) EPA's Pesticides Industry Sales and Usage, 2006 and 2007 Market Estimates. Available online at:
14 . Accessed January 2012.
15 EPA (2009) Biennial Reporting System (BRS) Database. U.S. Environmental Protection Agency, Envirofacts
16 Warehouse. Washington, D.C. Available online at: . Data for 2001-2007 are
17 current as of Sept. 9, 2009.
18 EPA (2004) EPA's Pesticides Industry Sales and Usage, 2000 and 2001 Market Estimates. Available online at:
19 . Accessed September 2006.
20 EPA (2002) EPA's Pesticides Industry Sales and Usage, 1998 and 1999 Market Estimates, table 3.6. Available
21 online at: . Accessed July 2003.
22 EPA (2001) AP 42, Volume I, Fifth Edition. Chapter 11: Mineral Products Industry. Available online at:
23 .
24 EPA (2000a) Biennial Reporting System (BRS). U.S. Environmental Protection Agency, Envirofacts Warehouse.
25 Washington, D.C. Available online at: .
26 EPA (2000b) Toxics Release Inventory, 1998. U.S. Environmental Protection Agency, Office of Environmental
27 Information, Office of Information Analysis and Access, Washington, D.C. Available online at:
28 .
29 EPA (1999) EPA's Pesticides Industry Sales and Usage, 1996-1997 Market Estimates. Available online at:
30 .
31 EPA (1998) EPA's Pesticides Industry Sales and Usage, 1994-1995 Market Estimates. Available online at:
32 .
33 FEB (2013) Fiber Economics Bureau, as cited in C&EN (2013) Lackluster Year for Chemical Output: Production
34 stayed flat or dipped in most world regions in 2012. Chemical &Engineering News, American Chemical Society, 1
35 July. Available online at: .
36 FEB (2012) Fiber Economics Bureau, as cited in C&EN (2012) Too Quiet After the Storm: After a rebound in 2010,
37 chemical production hardly grew in 2011. Chemical & Engineering News, American Chemical Society, 2 July.
38 Available online at: .
39 FEB (2011) Fiber Economics Bureau, as cited in C&EN (2011) Output Ramps up in all Regions. Chemical
40 Engineering News, American Chemical Society, 4 July. Available online at: .
41 FEB (2010) Fiber Economics Bureau, as cited in C&EN (2010) Output Declines in U.S., Europe. Chemical &
42 Engineering News, American Chemical Society, 6 July. Available online at: .
43 FEB (2009) Fiber Economics Bureau, as cited in C&EN (2009) Chemical Output Slipped In Most Regions Chemical
44 & Engineering News, American Chemical Society, 6 July. Available online at: .
45 FEB (2007) Fiber Economics Bureau, as cited in C&EN (2007) Gains in Chemical Output Continue. Chemical &
46 Engineering News, American Chemical Society. July 2, 2007. Available online at: .
10-12 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 FEB (2005) Fiber Economics Bureau, as cited in C&EN (2005) Production: Growth in Most Regions Chemical &
2 Engineering News, American Chemical Society, 11 July. Available online at: .
3 FEB (2003) Fiber Economics Bureau, as cited in C&EN (2003) Production Inches Up in Most Countries, Chemical
4 & Engineering News, American Chemical Society, 7 July. Available online at: .
5 FEB (2001) Fiber Economics Bureau, as cited in ACS (2001) Production: slow gains in output of chemicals and
6 products lagged behind U.S. economy as a whole Chemical & Engineering News, American Chemical Society, 25
7 June. Available online at: .
8 Financial Planning Association (2006) Canada/US Cross-Border Tools: US/Canada Exchange Rates. Available
9 online at: . Accessed on August 16, 2006.
10 Gosselin, Smith, and Hodge (1984) "Clinical Toxicology of Commercial Products." Fifth Edition, Williams &
11 Wilkins, Baltimore.
12 ICIS (2016) "Production issues force US melamine plant down" Available online at:
13 .
14 ICIS (2008) "Chemical profile: Melamine" Available online at:
15 . Accessed November,
16 2017.
17 IISRP (2003) "IISRP Forecasts Moderate Growth in North America to 2007" International Institute of Synthetic
18 Rubber Producers, Inc. New Release. Available online at: .
20 IISRP (2000) "Synthetic Rubber Use Growth to Continue Through 2004, Says IISRP and RMA" International
21 Institute of Synthetic Rubber Producers press release.
22 INEGI (2006) Production bruta total de las unidades economicas manufactureras por Subsector, Rama, Subrama y
23 Clase de actividad. Available online at:
24 . Accessed on
25 August 15, 2006.
26 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
27 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
28 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
29 Marland, G., and R.M. Rotty (1984) "Carbon dioxide emissions from fossil fuels: A procedure for estimation and
30 results for 1950-1982," Tellus 36b:232-261.
31 NPRA (2002) North American Wax - A Report Card. Available online at:
32 .
33 RMA (2016) 2015 U.S. Scrap Tire Management Summary. Rubber Manufacturers Association, Washington, D.C.
34 August 2016.
35 RMA (2014) 2013 U.S. Scrap Tire Management Summary. Rubber Manufacturers Association, Washington, D.C.
36 November 2014.
37 RMA (2011) U.S. Scrap Tire Management Summary: 2005-2009. Rubber Manufacturers Association, Washington,
38 D.C. October 2011, updated September 2013.
39 RMA (2009) "Scrap Tire Markets: Facts and Figures - Scrap Tire Characteristics." Rubber Manufacturers
40 Association., Washington D.C. Available online at:
41 http://www.rma.org/scrap_tires/scrap_tire_markets/scrap_tire_characteristics/ Accessed on 17 September 2009.
42 U.S. Census Bureau (2014) 2012 Economic Census. Available online at:
43 . Accessed November 2014.
44 U.S. Census Bureau (2009) Soap and Other Detergent Manufacturing: 2007. Available online at:
45 .
References 10-13
-------
1 U.S. Census Bureau (2004) Soap and Other Detergent Manufacturing: 2002. Issued December 2004. EC02-3II-
2 325611 (RV). Available online at: .
3 U.S. Census Bureau (1999) Soap and Other Detergent Manufacturing: 1997. Available online at:
4 .
5 U.S. International Trade Commission (1990-2016) "Interactive Tariff and Trade DataWeb: Quick Query." Available
6 online at: . Accessed November 2016.
7 Incineration of Waste — TO BE UPDATED FOR FINAL
8 INVENTORY REPORT
9 \rSo\ a. I .jupka. Roh \ an I laaren. \ora (loldsieiu. Scull \1 Kaufman, and \ickolas I Thenielis < 2<>()Xi " I(<|li
10 \iniiial IJioCscle Nationwide Sui\c> I lie Slate of(larhaue in \nierica" Hi"< V> /r. I(i Press. I jiiniaus. P \
11 December
12 IJalior. 131 2ikiim ( aula I !iieiu> s puhlic iv\ ie\\ coninieuis re I 'niii Inwiiinrv ni I s < iiwuhnusi.- < i.is I 'missiniis
13 .iih/SinLi Suhniiiied \ i;i email on \pril'). >ni> in l.eifI locksiad. I S l!P\.
14 De Socle.G(i ( l'W51 "\iiroiis (Kide from ( onihiisiiou and ludiistn ( lieniis|i\. I juissious and ( onliol " In \ R
15 N an \msiel. led.) Proc ol'ilie luicmalioiial Workshop \lelliane and \iiious (Kidc Methods mi \;iiion;il Emission
16 lu\cuioi'iesaud Options forCoutrol. \nicrsfooii. Nl. l ehiiiais ^-5. llW'
17 l!ueru\ Reco\er\ Council (2t>()1)) "2t)()~ Directors nl" \Vaste-lo-l !ueru\ Rlauls in llie I uiled Slates " \ccessed on
18 September 2^. 2()()1)
19 EPA (2016). Idvancing Sustainable Materials Management: 2014 had Sheet. Office of Land and Emergency
20 Managements. I S. I !n\ iioiinicuial Protection Aucucs Washington. DC \\ ailahle online al.
21 Imps wwwepauo\ sues production liles 2u| 14 siiinil'aclsheel 5t)X pdf
22 EPA (2015) Advancing Sustainable Materials Management: Facts and l-'igures 2013 Assessing Trends in Material
23 (ieneration, Recycling and Disposal in the I nitedStates. Office of Solid Waste and Emergency Response. U.S.
24 I !n\ irounieiiial Protection \ucuc> Washiimiou. I) C \\ailahle online al
25 Imp w w w "5 epa uo\ epawasie iioulia/. municipal pubs 2u I i ad\ nam sunn rpi pdf
26 EPA (2007. 2008. 2011. 2013. 2014) Municipal Solid Waste in the ( nited States: Facts and Figures. Office ofSolid
27 Waste and I !nicrucuc> Response. I S. I !n\ irounieiiial Protection \ucuc> Washiimiou. I) C \\ailahle online al
28 Imp www epa uo\ osw uoiilia/ niuuicipal nisw'W liliu
29 EPA (2006) Solid Waste Management and (ireenhouse (Jases:. I I.ifi'-i ycle. Issessment of Fmissions and Sinks.
30 ()l'lice iif Solid Wasieaud I !iiicrucuc> Response. I S I !n\ iioiinieuial Protection Aueiics. Washington. DC
31 EPA (2000) Characterization of Municipal Solid Waste in the (nited Slates: Source Data on the 1999 ( pdate.
32 ()flice of Solid Wasie. I S I !n\ irounieiiial Rioiecliou Aucucs. \\";islii nuU'ii. DC I !R \5 "5()-l'-i)()-()24
33 (loldsieiu. \ and C Madles (2t)u| i "I iih \uiiiial l5ioC\cle Nationwide Sui\c> I lie Siale of (larbaue iu \nierica "
34 Hii<< V> /r. J( i Press. I jiiniaus. R\ December 2uu I
35 IPCC (2006) 2006 IPC'(' (iuidelines for Xational (ireenhouse (ias Inventories. The National Greenhouse Gas
36 I memories Programme. I lie liiiei'uo\eriinieuial Panel on Cliniale Clianue 111 S I iuulesiou. I. IJiiendia. K. Miwa. I
37 \uara. and K. Tauabe ieds i| I la>ania. Kauauawa. Japan
38 Kaufman, el al (2do4) " I4tli \iiiiual l!io( >cle \atiouw ide Siu\e> I lie Siale of Garhaue mi \nierica 2i)i)4"
39 I3ine\ele. J(i Press. I jiimaiis. P\ .lanuaiy 2(>()4
40 R\l \ (2<>|I(>
41 \\ ailahle online al. Imps rnia oru sues default liles R\l \ serapiire sunini 15 pdf
42 R\l \ (2nl4i "2t)P I S Scrap lire Mauauenieiii SuniniaiA ' Rubber Manufacturers \ssocialiou. \o\emher 2(>I4
43 \\ ailahle online al Imp www rnia oru dow uload scrap-tires market-reports I S ST\larkel2<)I ' pdf
10-14 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 R\1 \ (2n|2ai "Ruhher I'\(K " Ruhher Maiiiiraelurers \ssnci;ilk»ii \\ailahle online al Imp www rm;i urn ahoul-
2 rim. riihlvr-l';K|s \eeessed on ll> \o\ember 2<)I4
3 R\ 1 \ (2<>l2hi "Scrap I ire \larkels I 'aelsand I'mures Scrap I ire (haraclcrisiies." Rubber Manufacturers
4 \ssih:i;iIkiii \\ailahle online ;il Imp" www rm;i oru scrap nres scrap lire markets scrap lire characteristics
5 \ccessed IS i»ii .l;iiiu;ir\ 2<>I2
6 R\ 1 \ <2<>I I) "I S Scrap I ire Mauaucnicul Summars 2<)<)5-2<)<>,r" Ruhher\l;iinir;ieliirers \ssticialioii Ocloher
7 2i)I I \\ ailahle online at Imp www.rma urn ser;ip iiresser;ip lire markets 2i>(summars pdf
8 Schneider. S (2t>()~) I ;-mail helween Shells Schneider of I rank I in \ssnciales (;i di\ ismn of I !RC¦ i ;md S;ir;ili
9 Skipim of l(T' liilerii;ilK Thesis (\l"I!;irili;md I!n\ iix14
12 SimiiKiiis. el ;il. (2<)()(> i " 15111 \;iliil icl W ;isie \l;iii;iuemeiil in I lie I inled Si;iies The
13 Si;ue nf( i;irh;me in \nierie:i " I»k»(\ele. J( i I'ress. I jiiiikiiis. \> \ April 2<)()(i
14 \;m ll;i;iren. Ruh. I hemelis. \ . ;ind (mkKleiii. (2n|()i " I lie Si;iie nf (i;irh;me in \merie;i" r>in( \ele. Oelnher
15 2<) I<) \ uliime 51. \iimher In. pu I(i-2 ^
16 Coal Mining
17 AAPG (1984) Coalbed Methane Resources of the United States. AAPG Studies in Geology Series #17.
18 ACR (2017) Project Database. Climate Action Reserve. Available at .
19 Creedy, D.P. (1993) Methane Emissions from Coal Related Sources in Britain: Development of a Methodology.
20 Chemosphere, 26: 419-439.
21 DMME (2017) DGO Data Information System. Department of Mines, Minerals and Energy of Virginia. Available
22 online at .
23 EIA (2017) Annual Coal Report 2014. Table 1. Energy Information Administration, U.S. Department of Energy.
24 El Paso (2009) Shoal Creek Mine Plan, El Paso Exploration & Production.
25 EPA (2017) Greenhouse Gas Reporting Program (GHGRP) 2015 Envirofacts. Subpart FF: Underground Coal
26 Mines. Available online at .
27 EPA (2005) Surface Mines Emissions Assessment. Draft. U.S. Environmental Protection Agency.
28 EPA (1996) Evaluation and Analysis of Gas Content and Coal Properties of Major Coal Bearing Regions of the
29 United States. EPA/600/R-96-065. U.S. Environmental Protection Agency.
30 Geological Survey of Alabama State Oil and Gas Board (GSA) (2016) Well Records Database. Available online at
31 .
32 IEA (2017) Key World Energy Statistics. Coal Production, International Energy Agency.
33 IPCC (2011) Use of Models and Facility-Level Data in Greenhouse Gas Inventories. Report of IPCC Expert
34 Meeting on Use of Models and Measurements in Greenhouse Gas Inventories 9-11 August 2010, Sydney, Australia.
35 Eds: EgglestonH.S., Srivastava N., Tanabe K., Baasansuren J., Fukuda M. IGES.
36 JWR (2010) No. 4 & 7 Mines General Area Maps. Walter Energy: Jim Walter Resources.
37 King, Brian (1994) Management of Methane Emissions from Coal Mines: Environmental, Engineering, Economic
38 and Institutional Implication of Options. Neil and Gunter Ltd.
39 MSHA (2017) Data Transparency at MSHA. Mine Safety and Health Administration. Available online at
40 .
41 Mutmansky, Jan M. and Yanbei Wang (2000) Analysis of Potential Errors in Determination of Coal Mine Annual
42 Methane Emissions. Mineral Resources Engineering, 9(4).
References 10-15
-------
1 Saghafi, Abouna (2013) Estimation of Fugitive Emissions from Open Cut Coal Mining and Measurable Gas
2 Content. 13th Coal Operators' Conference, University of Wollongong, The Australian Institute of Mining and
3 Metallurgy & Mine Managers Association of Australia. 306-313.
4 USBM (1986) Results of the Direct Method Determination of the Gas Contents of U.S. Coal Basins. Circular 9067.
5 U.S. Bureau of Mines.
6 West Virginia Geological & Economic Survey (WVGES) (2016) Oil & Gas Production Data. Available online at
7 .
8 Abandoned Underground Coal Mines
9 EPA (2004) Methane Emissions Estimates & Methodology for Abandoned Coal Mines in the U.S. Draft Final
10 Report. Washington, D.C. April 2004.
11 MSHA (2017) U.S. Department of Labor, Mine Health & Safety Administration (2017) Data Retrieval System.
12 Available online at: .
13 Petroleum Systems
14 Drillinglnfo (2017) January 2017 Download. DI Desktop® Drillinglnfo, Inc.
15 EPA (2017) Greenhouse Gas Reporting Program. Environmental Protection Agency. Data reported as of August 5,
16 2017.
17 EPA (1999a) Estimates of Methane Emissions from the U.S. Oil Industry (Draft Report). Prepared by ICF
18 International. Office of Air and Radiation, U.S. Environmental Protection Agency. October 1999.
19 EPA (1999b) Methane Emissions from the U.S. Petroleum Industry. Prepared by Radian International. U.S.
20 Environmental Protection Agency. February 1999.
21 EPA/GRI (1996) Methane Emissions from the Natural Gas Industry. Prepared by Radian. U.S. Environmental
22 Protection Agency. April 1996.
23 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
24 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
25 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
26 Lavoie et al. (2017) "Assessing the Methane Emissions from Natural Gas-Fired Power Plants and Oil Refineries."
27 Environmental Science & Technology. 2017 Mar 21;51(6):3373-3381. doi: 10.1021/acs.est.6b05531.
28 Natural Gas Systems
29 EPA (2017) Greenhouse Gas Reporting Program- Subpart W - Petroleum and Natural Gas Systems. Environmental
30 Protection Agency. Data reported as of August 5, 2017.
31 GRI/EPA (1996) Methane Emissions from the Natural Gas Industry. Prepared by Harrison, M., T. Shires, J.
32 Wessels, and R. Cowgill, eds., Radian International LLC for National Risk Management Research Laboratory, Air
33 Pollution Prevention and Control Division, Research Triangle Park, NC. EPA-600/R-96-080a.
34 GTI (2001) Gas Resource Database: Unconventional Natural Gas and Gas Composition Databases. Second Edition.
35 GRI-01/0136.
36 Lamb, et al. (2015) "Direct Measurements Show Decreasing Methane Emissions from Natural Gas Local
37 Distribution Systems in the United States." Environmental Science & Technology, Vol. 49 5161-5169.
38 Lavoie et al. (2017) "Assessing the Methane Emissions from Natural Gas-Fired Power Plants and Oil Refineries."
39 Environmental Science & Technology. 2017 Mar 21;51(6):3373-3381. doi: 10.1021/acs.est.6b05531.
40 Marchese, et al. (2015) "Methane Emissions from United States Natural Gas Gathering and Processing."
41 Environmental Science and Technology, Vol. 49 10718-10727.
10-16 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 PHMSA (2017a) Transmission Annuals Data. Pipeline and Hazardous Materials Safety Administration, U.S.
2 Department of Transportation, Washington, DC. Available online at: .
4 PHMSA (2017b) Gas Distribution Annual Data. Pipeline and Hazardous Materials Safety Administration, U.S.
5 Department of Transportation, Washington, DC. Available online at: .
7 Zimmerle, et al. (2015) "Methane Emissions from the Natural Gas Transmission and Storage System in the United
8 States." Environmental Science and Technology, Vol. 49 9374-9383.
9 Abandoned Oil and Gas Wells
10 Alaska Oil and Gas Conservation Commission, http://doa.alaska.gov/ogc/publicdb.html.
11 Arkansas Geological & Conservation Commission, "List of Oil & Gas Wells - Data From November 1, 1936 to
12 January 1, 1955." http://www.geology.ar.gov/pdf/IC-10%20SUPPLEMENT_v.pdf.
13 The Derrick's Handbook of Petroleum: A Complete Chronological and Statistical Review of Petroleum
14 Developments From 1859 to 1898 (V.l), (1898-1899) (V.2).
15 Drillinglnfo (2017) January 2017 Download. DI Desktop® Drillinglnfo, Inc.
16 Florida Department of Environmental Protection - Oil and Gas Program, https://floridadep.gov/water/oil-gas.
17 Geological Survey of Alabama, Oil & Gas Board, https://www.gsa.state.al.us/ogb/.
18 Kang, et al. (2016) "Identification and characterization of high methane-emitting abandoned oil and gas wells."
19 PNAS, vol. 113 no. 48, 13636-13641, doi: 10.1073/pnas.l605913113.
20 Oklahoma Geological Survey. "Oklahoma Oil: Past, Present, and Future." Oklahoma Geology Notes, v. 62 no. 3,
21 2002 pp .97-106.
22 Pennsylvania Department of Environmental Protection, Oil and Gas Reports - Oil and Gas Operator Well Inventory.
23 http://www.depreportingservices.state.pa.us/ReportServer/Pages/ReportViewer.aspx7/Oil_Gas/OG_Well_Inventory.
24 Texas Railroad Commission, Oil and Gas Division, "History of Texas Initial Crude Oil, Annual Production and
25 Producing Wells, Crude Oil Production and Well Counts (since 1935)." http://www.rrc.state.tx.us/oil-gas/research-
26 and-statistics/production-data/historical-production-data/crude-oil-production-and-well-counts-since-1935/.
27 Townsend-Small, et al. (2016) "Emissions of coalbed and natural gas methane from abandoned oil and gas wells in
28 the United States." Geophysical Research Letters, Vol. 43, 1789-1792.
29 United States Geological Survey's (USGS) Mineral Resources of the United States Annual Yearbooks,
30 https://minerals.usgs.gov/minerals/pubs/usbmmyb.html.
31 Virginia Department of Mines Minerals and Energy, "Wells Drilled for Oil and Gas in Virginia prior to 1962.",
32 Virginia Division of Mineral Resources, https://www.dmme.virginia.gov/commercedocs/MRR_4.pdf.
33 Energy Sources of Indirect Greenhouse Gases
34 EPA (2016) "1970 - 2016 Average annual emissions, all criteria pollutants in MS Excel." National Emissions
35 Inventory (NEI) Air Pollutant Emissions Trends Data. Office of Air Quality Planning and Standards, December
36 2016. Available online at: .
37 EPA (2003) E-mail correspondence containing preliminary ambient air pollutant data. Office of Air Pollution and
38 the Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency. December 22, 2003.
39 EPA (1997) Compilation of Air Pollutant Emission Factors, AP-42. Office of Air Quality Planning and Standards,
40 U.S. Environmental Protection Agency. Research Triangle Park, NC. October 1997.
References 10-17
-------
i International Bunker Fuels
2 Anderson, B.E., et al. (2011) Alternative Aviation Fuel Experiment (AAFEX), NASA Technical Memorandum, in
3 press.
4 ASTM (1989) Military Specification for Turbine Fuels, Aviation, Kerosene Types, NATO F-34 (JP-8) and NATO F-
5 35. February 10, 1989. Available online at: .
6 Chevron (2000) Aviation Fuels Technical Review (FTR-3). Chevron Products Company, Chapter 2. Available online
7 at: .
8 DHS (2008) Personal Communication with Elissa Kay, Residual and Distillate Fuel Oil Consumption (International
9 Bunker Fuels). Department of Homeland Security, Bunker Report. January 11, 2008.
10 DLA Energy (2017) Unpublished data from the Defense Fuels Automated Management System (DFAMS). Defense
11 Energy Support Center, Defense Logistics Agency, U.S. Department of Defense. Washington, D.C.
12 DOC (2017) Unpublished Report of Bunker Fuel Oil Laden on Vessels Cleared for Foreign Countries. Form-563.
13 Foreign Trade Division, Bureau of the Census, U.S. Department of Commerce. Washington, D.C.
14 DOT (1991 through 2013) Fuel Cost and Consumption. Federal Aviation Administration, Bureau of Transportation
15 Statistics, U.S. Department of Transportation. Washington, D.C. DAI-10.
16 EIA (2017) Monthly Energy Review, October 2017, Energy Information Administration, U.S. Department of
17 Energy, Washington, D.C. DOE/EIA-0035(2017/10).
18 FAA (2017) Personal Communication between FAA and John Steller for aviation emission estimates from the
19 Aviation Environmental Design Tool (AEDT). January 2017.
20 IMO (2014). ThirdIMO GHG Study 2014. International Maritime Organization (IMO) London, UK, April 2015;
21 Smith, T. W. P.; Jalkanen, J. P.; Anderson, B. A.; Corbett, J. J.; Faber, J.; Hanayama, S.; O'Keeffe, E.; Parker, S.;
22 Johansson, L.; Aldous, L.; Raucci, C.; Traut, M.; Ettinger, S.; Nelissen, D.; Lee, D. S.; Ng, S.; Agrawal, A.;
23 Winebrake, J. J.; Hoen, M.; Chesworth, S.; Pandey, A.
24 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
25 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
26 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
27 USAF (1998) Fuel Logistics Planning. U.S. Air Force pamphlet AFPAM23-221, May 1, 1998.
28 Wood Biomass and Biofuel Consumption
29 EIA (2017a) Monthly Energy Review, October 2017. Energy Information Administration, U.S. Department of
30 Energy. Washington, D.C. DOE/EIA-0035(2017/10).
31 EIA (2017b) Monthly Biodiesel Production Report. October 2017. Energy Information Administration, U.S.
32 Department of Energy. Washington, D.C.
33 EPA (2010) Carbon Content Coefficients Developed for EPA's Mandatory Reporting Rule. Office of Air and
34 Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
35 Lindstrom, P. (2006) Personal Communication. Perry Lindstrom, Energy Information Administration and Jean Kim,
36 ICF International.
37 Industrial Processes and Product Use
38 EPA (2014) Greenhouse Gas Reporting Program. Developments on Publication of Aggregated Greenhouse Gas
39 Data, November 25, 2014. See .
10-18 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 EPA (2002) Quality Assurance/Quality Control and Uncertainty Management Plan for the U.S. Greenhouse Gas
2 Inventory: Procedures Manual for Quality Assurance/Quality Control and Uncertainty Analysis, U.S. Greenhouse
3 Gas Inventory Program, U.S. Environmental Protection Agency, Office of Atmospheric Programs, EPA 430-R-02-
4 007B, June 2002.
5 IPCC (2011) Use of Models and Facility-Level Data in Greenhouse Gas Inventories (Report of IPCC Expert
6 Meeting on Use of Models and Measurements in Greenhouse Gas Inventories 9-11 August 2010, Sydney, Australia)
7 eds.: EgglestonH.S., Srivastava N., Tanabe K., Baasansuren J., Fukuda M., Pub. IGES, Japan 2011.
8 Cement Production
9 EPA Greenhouse Gas Reporting Program (2017). Aggregation of Reported Facility Level Data under Subpart H -
10 National Level Clinker Production from Cement Production for Calendar Years 2014, 2015, and 2016. Office of Air
11 and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
12 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
13 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
14 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
15 U.S. Bureau of Mines (1990 through 1993) Minerals Yearbook: Cement Annual Report. U.S. Department of the
16 Interior, Washington, D.C.
17 United States Geological Survey (USGS) (2016a) Mineral Industry Survey: Cement in September 2016. U.S.
18 Geological Survey, Reston, VA. December, 2016.
19 USGS (2017) Mineral Commodity Summaries: Cement 2016. U.S. Geological Survey, Reston, VA. January, 2017.
20 USGS (1995 through 2014) Minerals Yearbook - Cement. U.S. Geological Survey, Reston, VA.
21 Van Oss (2013a) 1990 through 2012 Clinker Production Data Provided by Hendrik van Oss (USGS) via email on
22 November 8, 2013.
23 Van Oss (2013b) Personal communication. Hendrik van Oss, Commodity Specialist of the U.S. Geological Survey
24 and Gopi Manne, Eastern Research Group, Inc. October 28, 2013.
25 Lime Production — TO BE UPDATED FOR FINAL INVENTORY
26 REPORT
27 ( oralhers (2<>|~) Personal coniniiiiikaliou. Lisa (oralhers. I S Geological Sur\e> and Maiisann Desai. I S
28 I !n\ irounieuial kroieclion \ucuc> .laiiuars 12. 2<> I—
29 I \ < > I(i) ( iicciihousc (ias keporinm krourani <( il l( ikl'i \uureualiou of keporied I'acilils I .e\ el I )ala under
30 Suhpari S -\alional I .line krodiicliou lor ( 'aleudar Yours 2<> Id ihroimh 2d 15 ()fl'ice of \ir and kadialiou. ()l'lice of
31 \iniospheric krouranis. I S. I !n\ irounieuial krokvlion \ueiic>. Washiimiou. DC
32 IPCC (2006) 2006 IPC Y' C iuidelinesfor Xaiional C ireenhini.se Cias Inventories. The National Greenhouse Gas
33 lu\ciiiories kromaninie. I lie liiicruo\criinicuial kancl on ('limine Chaiiue 111 S I !uulesion. I. P.iiendia. k Miwa. I
34 \uara. and k Tauahe (eds i| I la>ama. kanauawa. Japan
35 Males. I! (2dd^i Memorandum from line Males, \alional I.line \ssncialiou lo Mr W illiam V lr\ niu ^ Mr l.eif
36 I locksiad. I!u\ iroiinieuial krokvliou \uenc> March <>. 2doi
37 Miner, k and li I plou <2nu2i Melhods foresiinialiuu ureeiihouse uas emissions from lime kilns al kral'l pulp mills
38 l!ueru\ \ ol 2"(2i)()2i. p
39 Seeuer (2nI ^) Memorandum from \rliue M Seeuer. \alional l.inie \ssocialiou lo Mr l.eil Ilocksiad.
40 I !n\ iroiinienial kroiecliou \ueuc> March 15. 201 ^
41 I mied Siales Geolomcal Sur\e> (I S(iS I(<;ii \liinr,i/( 'nmmndiiics Siiniiihirv I inw I S. (ieolouical
42 Sur\e\. kesiou. \'.\ (Jaiiuars 2n|(i)
References 10-19
-------
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
I S(iS <2<)|141 \liiur.ils )enrhooh limeA S (icolomcal Sinne\. kesiou. Y \ (June 2dI(>>
EPA (2009) Technical Support Document for the (ilasslanufacturing Sector: Proposed Rule forlaudatory
Reporting "i < ireenhouse< i.ises I S. I !n\ iroiiniculal Protection \ucuc>. Washington. DC.
IPCC (2006) 2006 IP( X* Guidelines for Xational Greenhouse (ias Inventories. The National Greenhouse Gas
lu\cuiorics Prouraninic. I lie liiicruo\crunicuial Panel on Cliniale Chauue |l I S I iudesiou. I. IJucudia. K \lma. I
\uara. and K Tauaheieds )| llavama. Kauauawa. Japan
()l l < 20021 Glass Industry of the b inure: Energy and Environmental Profile of the ( .S. Glass Industry. Office of
Industrial I cchuolomcs. I S Deparinieui of I iueruv Washiimiou. DC
I S 15 urea u of \liues ( I'W I and I'W'ai \liner.ils )e.irhooL t rushed. Stone .innu.il Report I S Deparinieui of llie
Interior Washington. DC.
United States Geological Survey (USGS) (2015a) Minerals Industry Surveys: Soda Ash in January 2015. U.S.
(leolomeal Sur\e>. kesiou. Y \. March. 2d 15
I S( iS (IW5 lliroimh 2d 15hi \ liner,ils )r.//-/"»'/i ( rushed. Stone .innu.il Report. I S (icolomcal Sur\ e\. kesiou.
V \
I S( iS ( IW5 lliroimh 2d 15ci \ liner, ils ) e.trhooE Sod.i .ish .innu.il Report I S (icolomcal Sur\ e\. kesion. Y\
Willell (2d I "i Personal coniniuuicaliou. Jasou Christopher Willell. I S (icolomcal Sur\e> and Mausanii Desai.
I S I !u\ iroiiniculal Protection \ueuc> January <>. 2d I ~
Willell (2dI5i Personal coniniuuicaliou. Jason Christopher Willed. I S (icolomcal Sur\c> and (iopi Mamie.
Eastern kesearch (uoup. Inc. Sepiemher'J. 2dI5.
Willell (2d14) Personal coniniuuicaliou.. Jasou Chrisiopher Willell. I S Gcolomcal Sur\e\ and (iopi Mamie.
I iastern kesearch (u'oup. luc Sepieniher 25. 2d 14
MSI i2d|(i) 2111.^ .innu.il Si,itistii ,il Report \nicricau Irouaud Steel liistiiute
I S IJiireau ol' Mines i l'W| and I'W'ai Minerals ) earhool,-< rushed Stone .innu.il Repori I S Deparinieui of i he
Interior Washington. DC.
U.S. Bureau ofMines (1990 through 1993b) Minerals Yearbook: Magnesium and Magnesium Compounds Annual
Report I S Deparinieui ol'the Interior Washington. I) C
United States Geological Survey (USGS) (2013) Magnesium Metal Mineral Commodity Summary for 2013. U.S.
Geological Sur\e\. kesiou. \ A
I S( iS i I W5a throimh 2d 151 \Iiner.ils )e.irbook < rushed Stone .innu.il Report. I S (icolomcal Sur\ e\. kesiou.
Y\
I S( iS i I W5h lliroimh 2d 121 \ liner,ils )e,irbool, \ l.i^ihsiuni .innu.il Report I S (icolomcal Sur\ e\. kesiou. Y\
Willell (2d | -a> Personal coniniuuicalioii. Jason Chrisiopher W illell. I S Geological Sur\c> and Mausanii Desai
and John Steller. I S I !u\ iroiiniculal Protection \ueuc> March'J. 2d| ~
Willell (2d I "hi Personal coniniuuicaliou. Jasou Christopher Willell. I S. (icolomcal Sur\ e\ and Mausanii I)esai.
I S I !u\ iroiinieiital Protection \ueuc> Jaiiuarv (>. 2d I"
10-20 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Willed (21)15) IViMmal cuiiimiiiiicalkiii. Jason (lirislophei' Willed. I S (leolnuical Sunex and (mpi Mamie.
2 I !;iMem Research (uiuip. hie Septemberl). 2(>15
3 Ammonia Production
4 ACC (2016) Business of Chemistry (Annual Data). American Chemistry Council, Arlington, VA.
5 Bark (2004) CoffeyvilleNitrogen Plant. December 15, 2004. Available online at:
6 .
7 Coffeyville Resources Nitrogen Fertilizers (2012) Nitrogen Fertilizer Operations. Available online at:
8 .
9 Coffeyville Resources Nitrogen Fertilizers (2011) Nitrogen Fertilizer Operations. Available online at:
10 .
11 Coffeyville Resources Nitrogen Fertilizers (2010) Nitrogen Fertilizer Operations. Available online at:
12 .
13 Coffeyville Resources Nitrogen Fertilizers (2009) Nitrogen Fertilizer Operations. Available online at:
14 .
15 Coffeyville Resources Nitrogen Fertilizers, LLC (2005 through 2007a) Business Data. Available online at:
16 .
17 Coffeyville Resources Nitrogen Fertilizers (2007b) Nitrogen Fertilizer Operations. Available online at:
18 .
19 Coffeyville Resources Energy, Inc. (CVR) (2015) CVR Energy, Inc. 2014 Annual Report. Available online at:
20 .
21 CVR (2017) CVR Energy, Inc. 2016 Annual Report. Available online at: .
22 CVR (2016) CVR Energy, Inc. 2015 Annual Report. Available online at: .
23 CVR (2014) CVR Energy, Inc. 2013 Annual Report. Available online at: .
24 CVR (2012) CVR Energy, Inc. 2012 Annual Report. Available online at: .
25 EFMA (2000a) Best Available Techniques for Pollution Prevention and Control in the European Fertilizer Industry.
26 Booklet No. 1 of 8: Production of Ammonium. Available online at:
27 .
28 EFMA (2000b) Best Available Techniques for Pollution Prevention and Control in the European Fertilizer Industry.
29 Booklet No. 5 of 8: Production of Urea and Urea Ammonium Nitrate. Available online at:
30 .
31 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
32 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
33 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
34 U.S. Census Bureau (2011) Current Industrial Reports Fertilizer Materials and Related Products: 2010 Summary.
35 Available online at: .
36 U.S. Census Bureau (2010) Current Industrial Reports Fertilizer Materials and Related Products: 2009 Summary.
37 Available online at: .
38 U.S. Census Bureau (2009) Current Industrial Reports Fertilizer Materials and Related Products: 2008 Summary.
39 Available online at: .
40 U.S. Census Bureau (2008) Current Industrial Reports Fertilizer Materials and Related Products: 2007 Summary.
41 Available online at: .
42 U.S. Census Bureau (2007) Current Industrial Reports Fertilizer Materials and Related Products: 2006 Summary.
43 Available online at: .
References 10-21
-------
1 U.S. Census Bureau (2006) Current Industrial Reports Fertilizer Materials and Related Products: 2005 Summary.
2 Available online at: .
3 U.S. Census Bureau (2004, 2005) Current Industrial Reports Fertilizer Materials and Related Products: Fourth
4 Quarter Report Summary. Available online at: .
5 U.S. Census Bureau (1998 through 2003) Current Industrial Reports Fertilizer Materials and Related Products:
6 Annual Reports Summary. Available online at: .
7 U.S. Census Bureau (1991 through 1994) Current Industrial Reports Fertilizer Materials Annual Report. Report No.
8 MQ28B. U.S. Census Bureau, Washington, D.C.
9 United States Geological Survey (USGS) (2017) 2015 Minerals Yearbook: Nitrogen [Advance Release], August
10 2017. Available online at: .
11 USGS (2016) 2014Minerals Yearbook: Nitrogen [Advance Release]. October 2016. Available online at:
12 .
13 USGS (2015) 2013Minerals Yearbook: Nitrogen [Advance Release]. August 2015. Available online at:
14 .
15 USGS (2014) 2012Minerals Yearbook: Nitrogen [Advance Release]. September 2014. Available online at:
16 .
17 USGS (1994 through 2009) Minerals Yearbook: Nitrogen. Available online at:
18 .
19 Urea Consumption Agricultural Purposes
20 EFMA (2000) Best Available Techniques for Pollution Prevention and Control in the European Fertilizer Industry.
21 Booklet No. 5 of 8: Production of Urea and Urea Ammonium Nitrate.
22 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
23 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
24 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
25 TFI (2002) U.S. Nitrogen Imports/Exports Table. The Fertilizer Institute. Available online at:
26 . August 2002.
27 U.S. Census Bureau (2001 through 2011) Current Industrial Reports Fertilizer Materials and Related Products:
28 Annual Summary. Available online at: .
29 U.S. Department of Agriculture (2012) Economic Research Service Data Sets, Data Sets, U.S. Fertilizer
30 Imports/Exports: Standard Tables. Available online at: .
32 U.S. ITC (2002) United States International Trade Commission Interactive Tariff and Trade DataWeb, Version
33 2.5.0. Available online at: . August 2002.
34 United States Geological Survey (USGS) (2014 through 2016) Minerals Yearbook: Nitrogen [Advance Release],
35 Available online at: .
36 USGS (1994 through 2009) Minerals Yearbook: Nitrogen. Available online at:
37 .
38 USGS (2017a) Personal communication. Lori Apodaca, Mineral Commodity Specialist, USGS and John Steller,
39 U.S. Environmental Protection Agency. September 20, 2017.
40 USGS (2017b) Minerals Commodity Summaries: Nitrogen (Fixed) - Ammonia. Available online at:
41 . Accessed October 2, 2017.
10-22 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Nitric Acid Pn lion
2 Climate Action Reserve (CAR) (2013) Project Report. Available online at:
3 . Accessed on 18 January 2013.
4 Desai (2012) Personal communication. Mausami Desai, U.S. Environmental Protection Agency, January 25, 2012.
5 EPA (2017) Greenhouse Gas Reporting Program (GHGRP). Aggregation of Reported Facility Level Data under
6 Subpart V -National Nitric Acid Production for Calendar Years 2010 through 2016. Office of Air and Radiation,
7 Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
8 EPA (2013) Draft Nitric Acid Database. U.S. Environmental Protection Agency, Office of Air and Radiation.
9 September, 2010.
10 EPA (2012) Memorandum from Mausami Desai, U.S. EPA to Mr. Bill Herz, The Fertilizer Institute. November 26,
11 2012.
12 EPA (2010) Available and Emerging Technologies for Reducing Greenhouse Gas Emissions from the Nitric Acid
13 Production Industry. Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency.
14 Research Triangle Park, NC. December 2010. Available online at:
15 .
16 EPA (1998) Compilation of Air Pollutant Emission Factors, AP-42. Office of Air Quality Planning and Standards,
17 U.S. Environmental Protection Agency. Research Triangle Park, NC. February 1998.
18 IPCC (2007) Forster, P., V. Ramaswamy, P. Artaxo, T. Berntsen, R. Betts, D.W. Fahey, J. Haywood, J. Lean, D.C.
19 Lowe, G. Myhre, J. Nganga, R. Prinn, G. Raga, M. Schulz and R. Van Dorland, 2007: Changes in Atmospheric
20 Constituents and in Radiative Forcing. In: Climate Change 2007: The Physical Science Basis. Contribution of
21 Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S.,
22 D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M.Tignor and H.L. Miller (eds.)]. Cambridge University
23 Press, Cambridge, United Kingdom and New York, NY, USA.
24 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
25 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
26 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
27 U.S. Census Bureau (2010a) Current Industrial Reports. Fertilizers and Related Chemicals: 2009. "Table 1:
28 Summary of Production of Principle Fertilizers and Related Chemicals: 2009 and 2008." June, 2010. MQ325B(08)-
29 5. Available online at: .
30 U.S. Census Bureau (2010b) Personal communication between Hilda Ward (of U.S. Census Bureau) and Caroline
31 Cochran (of ICF International). October 26, 2010 and November 5, 2010.
32 U.S. Census Bureau (2009) Current Industrial Reports. Fertilizers and Related Chemicals: 2008. "Table 1:
33 Shipments and Production of Principal Fertilizers and Related Chemicals: 2004 to 2008." June, 2009. MQ325B(08)-
34 5. Available online at: .
35 U.S. Census Bureau (2008) Current Industrial Reports. Fertilizers and Related Chemicals: 2007. "Table 1:
36 Shipments and Production of Principal Fertilizers and Related Chemicals: 2003 to 2007." June, 2008. MQ325B(07)-
37 5. Available online at: .
38 Adipic Acid Production
39 ACC (2016) Business of Chemistry (Annual Data). American Chemistry Council, Arlington, VA.
40 C&EN (1995) "Production of Top 50 Chemicals Increased Substantially in 1994." Chemical & Engineering News,
41 73(15): 17. April 10, 1995.
42 C&EN (1994) "Top 50 Chemicals Production Rose Modestly Last Year." Chemical & Engineering News,
43 72(15): 13. April 11, 1994.
References 10-23
-------
1 C&EN (1993) "Top 50 Chemicals Production Recovered Last Year." Chemical & Engineering News, 71(15): 11.
2 April 12, 1993.
3 C&EN (1992) "Production of Top 50 Chemicals Stagnates in 1991." Chemical & Engineering News, 70(15): 17.
4 April 13, 1992.
5 CMR (2001) "Chemical Profile: Adipic Acid." Chemical Market Reporter. July 16, 2001.
6 CMR (1998) "Chemical Profile: Adipic Acid." Chemical Market Reporter. June 15, 1998.
7 CW (2005) "Product Focus: Adipic Acid." Chemical Week. May 4, 2005.
8 CW (1999) "Product Focus: Adipic Acid/Adiponitrile." Chemical Week, p. 31. March 10, 1999.
9 Desai (2010, 2011) Personal communication. Mausami Desai, U.S. Environmental Protection Agency and Adipic
10 Acid Plant Engineers. 2010 and 2011.
11 EPA (2014 through 2017) Greenhouse Gas Reporting Program. Full Subpart E, O, S-CEMS, BB, CC, LL Data Set
12 (XLS) (Adipic Acid Tab). Office of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental
13 Protection Agency, Washington, D.C. Accessed on November 15, 2017, Available online at:
14 .
15 EPA (2012) Analysis of Greenhouse Gas Reporting Program data - Subpart E (Adipic Acid), Office of Air and
16 Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
17 ICIS (2007) "Adipic Acid." ICIS Chemical Business Americas. July 9, 2007.
18 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
19 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
20 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
21 Reimer, R.A., Slaten, C.S., Seapan, M., Koch, T.A. and Triner, V.G. (1999) "Implementation of Technologies for
22 Abatement of N20 Emissions Associated with Adipic Acid Manufacture." Proceedings of the 2nd Symposium on
23 Non-C02 Greenhouse Gases (NCGG-2), Noordwijkerhout, The Netherlands, 8-10 Sept. 1999, Ed. J. van Ham et al.,
24 Kluwer Academic Publishers, Dordrecht, pp. 347-358.
25 Thiemens, M.H., and W.C. Trogler (1991) "Nylon production; an unknown source of atmospheric nitrous oxide."
26 Science 251:932-934.
27 Caprolactam, Glyoxal and Glyoxylic Acid Production
28 ACC (2015) Business of Chemistry (Annual Data). American Chemistry Council, Arlington, VA.
29 AdvanSix (2017). AdvanSix Hopewell Virginia Information Sheet. Retrieved from:
30 https://www.advan6.com/hopewell/ on September 27, 2017.
31 Ashford, R.D. (1994). Ashford's Dictionary of Industrial Chemicals, Wavelength Publications Ltd, London
32 England.
33 Babusiaux, P. (2005). Note on production of Glyoxal and Glyoxylic acid, Clariant, Lamotte, France.
34 BASF (2017). BASF: Freeport, Texas Fact Sheet. Retrieved from https://www.basf.com/documents/us/en/Fact-
35 Sheets/Freeport-Texas-SiteFactSheet.pdf on September 27, 2017.
36 Fibrant (2017). Fibrant LLC Contact Page. Retrieved from: http://www.fibrant52.com/en/contact on September 27,
37 2017.
38 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
39 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
40 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
41 Lowenheim, F.A. and Moran, M.K. (1975). Faith, Keyes, and Clark's Industrial Chemicals, Fourth Edition, John
42 Reimschuessel, H. K. (1977). 'Nylon 6 Chemistry and Mechanisms', Journal of Polymer Science: Macromolecular
43 Reviews, Vol. 12, 65-139, John Wiley & Sons, Inc. Wiley & Sons, Inc. USA.
10-24 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 TechSci n.d. (2017). Fibrant B. V. to Discontinue Caprolactam Plant in the United States. Retrieved from:
2 https://www.techsciresearch.com/news/1356-fibrant-b-v-to-discontinue-caprolactam-plant-in-the-united-states.html
3 on September 27, 2017.
4 Silicon Carbide Production and Consumption
5 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
6 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
7 M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.)]. Cambridge University Press. Cambridge, United
8 Kingdom 996 pp.
9 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
10 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
11 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
12 U.S. Census Bureau (2005 through 2017) USITC Trade DataWeb. Available online at: .
13 United States Geological Survey (2017a) 2015 Minerals Yearbook: Abrasives, Manufactured [Advance Release],
14 U.S. Geological Survey, Reston, VA. Available online at:
15 .
16 USGS (2017b) Mineral Industry Surveys: Abrasives (Manufactured). U.S. Geological Survey, Reston, VA.
17 Available online at: .
18 USGS (2017c) USGS 2015 Minerals Yearbook Silicon [Advance Release], November 2017. Table 4. U.S.
19 Geological Survey, Reston, VA. Available online at: .
20 USGS (2016) Minerals Industry Surveys: Abrasives (Manufactured) in Third and Fourth Quarter of 2015. U.S.
21 Geological Survey, Reston, VA. January 2017. Available online at:
22 .
23 USGS (199 la through 2015) Miner als Yearbook: Manufactured Abrasives Annual Report. U.S. Geological Survey,
24 Reston, VA. Available online at: .
25 USGS (1991b through 2015) Minerals Yearbook: Silicon Annual Report. U.S. Geological Survey, Reston, VA.
26 Available online at: .
27 Titanium Dioxide Production
28 Gambogi, J. (2002) Telephone communication. Joseph Gambogi, Commodity Specialist, U.S. Geological Survey
29 and Philip Groth, ICF International. November 2002.
30 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
31 Inventories Programme, Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
32 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
33 United States Geological Survey (2017) 2017Mineral Commodity Summary: Titanium and Titanium Dioxide. U.S.
34 Geological Survey, Reston, Va. January, 2017. Available online at:
35 .
36 USGS (2016) 2016Mineral Commodity Summary: Titanium and Titanium Dioxide. U.S. Geological Survey,
37 Reston, Va. January, 2016.
38 USGS (1991 through 2015) Minerals Yearbook: Titanium. U.S. Geological Survey, Reston, VA.
39 Soda Ash Production
40 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
41 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
42 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
References 10-25
-------
1 Kostick, D. S. (2012) Personal communication. Dennis S. Kostick of U.S. Department of the Interior - U.S.
2 Geological Survey, Soda Ash Commodity Specialist with Gopi Manne and Bryan Lange of Eastern Research Group,
3 Inc. October 2012.
4 United States Geological Survey (USGS) (2017) Mineral Industry Surveys: Soda Ash in January 2017. U.S.
5 Geological Survey, Reston, VA. March, 2017.
6 USGS (2015a) Mineral Industry Surveys: Soda Ash in July 2015. U.S. Geological Survey, Reston, VA. September,
7 2015.
8 USGS (1994 through 2015b) Minerals Yearbook: Soda Ash Annual Report. U.S. Geological Survey, Reston, VA.
9 USGS (1995) Trona Resources in the Green River Basin, Southwest Wyoming. U.S. Department of the Interior, U.S.
10 Geological Survey. Open-File Report 95-476. Wiig, Stephen, Grundy, W.D., Dyni, JohnR.
11 Petrochemical Production
12 ACC (2016) Business of Chemistry (Annual Data). American Chemistry Council, Arlington, VA.
13 ACC (2014a) U.S. Chemical Industry Statistical Handbook. American Chemistry Council, Arlington, VA.
14 ACC (2014b) Business of Chemistry (Annual Data). American Chemistry Council, Arlington, VA.
15 ACC (2002, 2003, 2005 through 2011) Guide to the Business of Chemistry. American Chemistry Council,
16 Arlington, VA.
17 AN (2014) . [bout Acrylonitrile: Production. AN Group, Washington, D.C. Available online at:
18 .
19 EPA Greenhouse Gas Reporting Program (2017) Aggregation of Reported Facility Level Data under Subpart X -
20 National Petrochemical Production for Calendar Years 2010 through 2016. Office of Air and Radiation, Office of
21 Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
22 EPA (2008) Technical Support Document for the Petrochemical Production Sector: Proposed Rule for Mandatory
23 Reporting of Greenhouse Gases. U.S. Environmental Protection Agency. September 2008.
24 EPA (2000) Economic Impact Analysis for the Proposed Carbon Black Manufacturing NESHAP, U.S.
25 Environmental Protection Agency. Research Triangle Park, NC. EPA-452/D-00-003. May 2000.
26 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
27 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
28 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
29 Johnson, G. L. (2005 through 2010) Personal communication. Greg Johnson of Liskow & Lewis, on behalf of the
30 International Carbon Black Association (ICBA) and Caroline Cochran, ICF International. September 2010.
31 Johnson, G. L. (2003) Personal communication. Greg Johnson of Liskow & Lewis, on behalf of the International
32 Carbon Black Association (ICBA) and Caren Mintz, ICF International. November 2003.
33 HCFC-22 Production
34 ARAP (2010) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
35 Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. September 10, 2010.
36 ARAP (2009) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
37 Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. September 21, 2009.
38 ARAP (2008) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
39 Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. October 17, 2008.
40 ARAP (2007) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
41 Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. October 2, 2007.
10-26 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 ARAP (2006) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
2 Atmospheric Policy to Sally Rand of the U.S. Environmental Protection Agency. July 11, 2006.
3 ARAP (2005) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
4 Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. August 9, 2005.
5 ARAP (2004) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
6 Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. June 3, 2004.
7 ARAP (2003) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
8 Atmospheric Policy to Sally Rand of the U.S. Environmental Protection Agency. August 18, 2003.
9 ARAP (2002) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
10 Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. August 7, 2002.
11 ARAP (2001) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
12 Atmospheric Policy to Deborah Ottinger of the U.S. Environmental Protection Agency. August 6, 2001.
13 ARAP (2000) Electronic mail communication from Dave Stirpe, Executive Director, Alliance for Responsible
14 Atmospheric Policy to Sally Rand of the U.S. Environmental Protection Agency. August 13, 2000.
15 ARAP (1999) Facsimile from Dave Stirpe, Executive Director, Alliance for Responsible Atmospheric Policy to
16 Deborah Ottinger Schaefer of the U.S. Environmental Protection Agency. September 23, 1999.
17 ARAP (1997) Letter from Dave Stirpe, Director, Alliance for Responsible Atmospheric Policy to Elizabeth Dutrow
18 of the U.S. Environmental Protection Agency. December 23, 1997.
19 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
20 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
21 M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.)]. Cambridge University Press. Cambridge, United
22 Kingdom 996 pp.
23 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
24 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
25 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
26 RTI (2008) "Verification of Emission Estimates of HFC-23 from the Production of HCFC-22 :Emissions from 1990
27 through 2006." Report prepared by RTI International for the Climate Change Division. March, 2008.
28 RTI (1997) "Verification of Emission Estimates of HFC-23 from the Production of HCFC-22: Emissions from 1990
29 through 1996." Report prepared by Research Triangle Institute for the Cadmus Group. November 25, 1997; revised
30 February 16, 1998.
31 UNFCCC (2014) Report of the Conference of the Parties on its nineteenth session, held in Warsaw from 11 to 23
32 November 2013. United Nations Framework Convention on Climate Change, Warsaw. (FCCC/CP/2013/10/Add.3).
33 January 31, 2014. Available online at: .
34 Carbon Dioxide Consumption
35 Allis, R. et al. (2000) Natural CO2 Reservoirs on the Colorado Plateau and Southern Rocky Mountains: Candidates
36 for C02 Sequestration. Utah Geological Survey and Utah Energy and Geoscience Institute. Salt Lake City, Utah.
37 ARI (1990 through 2010) CO2 Use in Enhanced Oil Recovery. Deliverable to ICF International under Task Order
38 102, July 15, 2011.
39 ARI (2007) CO2-EOR: An Enabling Bridge for the Oil Transition. Presented at "Modeling the Oil Transition—a
40 DOE/EPA Workshop on the Economic and Environmental Implications of Global Energy Transitions." Washington,
41 D.C.April 20-21, 2007.
42 ARI (2006) CO2-EOR: An Enabling Bridge for the Oil Transition. Presented at "Modeling the Oil Transition—a
43 DOE/EPA Workshop on the Economic and Environmental Implications of Global Energy Transitions." Washington,
44 D.C.April 20-21, 2006.
References 10-27
-------
1 Broadhead (2003) Personal communication. Ron Broadhead, Principal Senior Petroleum Geologist and Adjunct
2 faculty, Earth and Environmental Sciences Department, New Mexico Bureau of Geology and Mineral Resources,
3 and Robin Pestrusak, ICF International. September 5, 2003.
4 COGCC (1999 through 2009) Monthly CO2 Produced by County. Available online at:
5 . Accessed October
6 2014.
7 Denbury Resources Inc. (2002 through 2010) Annual Report: 2001 through 2009, Form 10-K. Available online at:
8 .
9 Accessed September 2014.
10 EPA Greenhouse Gas Reporting Program (2016). Aggregation of Reported Facility Level Data under Subpart PP -
11 National Level CO2 Transferred for Food & Beverage Applications for Calendar Years 2010 through 2014. Office
12 of Air and Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
13 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
14 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
15 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
16 New Mexico Bureau of Geology and Mineral Resources (2006) Natural Accumulations of Carbon Dioxide in New
17 Mexico and Adjacent Parts of Colorado and Arizona: Commercial Accumulation of CO2. Available online at:
18 .
19 Phosphoric Acid Production
20 EFMA (2000) "Production of Phosphoric Acid." Best Available Techniques for Pollution Prevention and Control in
21 the European Fertilizer Industry. Booklet 4 of 8. European Fertilizer Manufacturers Association. Available online
22 at: .
23 FIPR (2003a) "Analyses of Some Phosphate Rocks." Facsimile Gary Albarelli, the Florida Institute of Phosphate
24 Research, Bartow, Florida, to Robert Lanza, ICF International. July 29, 2003.
25 FIPR (2003b) Florida Institute of Phosphate Research. Personal communication. Mr. Michael Lloyd, Laboratory
26 Manager, FIPR, Bartow, Florida, to Mr. Robert Lanza, ICF International. August 2003.
27 NCDENR (2013) North Carolina Department of Environment and Natural Resources, Title V Air Permit Review for
28 PCS Phosphate Company, Inc. - Aurora. Available online at:
29 . Accessed on January 25, 2013.
30 United States Geological Survey (USGS) (2017) Mineral Commodity Summaries: Phosphate Rock 2017. January
31 2017. U.S. Geological Survey, Reston, VA. Available online at: .
33 USGS (2016) Mineral Commodity Summaries: Phosphate Rock 2016. January 2016. U.S. Geological Survey,
34 Reston, VA. Available online at:
-------
1 Iron and Steel Production and Metallurgical Coke Production
2 American Coke and Coal Chemicals Institute (ACCCI) (2016) U.S. & Canadian Coke Plants as of February 2016.
3 ACCCI, Washington, D.C. February 2016.
4 American Iron and Steel Institute (AISI) (2004 through 2017) Annual Statistical Report, American Iron and Steel
5 Institute, Washington, D.C.
6 AISI (2006 through 2016) Personal communication, Mausami Desai, U.S. EPA, and American Iron and Steel
7 Institute, December 2016.
8 AISI (2008) Personal communication, Mausami Desai, U.S. EPA, and Bruce Steiner, Technical Consultant with the
9 American Iron and Steel Institute, October 2008.
10 Carroll (2016) Personal communication, Mausami Desai, U.S. EPA, and Colin P. Carroll, Director of Environment,
11 Health and Safety, American Iron and Steel Institute, December 2016.
12 Carroll (2017) Personal communication, John Steller, U.S. EPA, and Colin P. Carroll, Director of Environment,
13 Health and Safety, American Iron and Steel Institute, November 2017.
14 DOE (2000) Energy and Environmental Profile of the U.S. Iron and Steel Industry. Office of Industrial
15 Technologies, U.S. Department of Energy. August 2000. DOE/EE-0229.EIA.
16 EIA (1998 through 2017) Quarterly Coal Report: October-December, Energy Information Administration, U.S.
17 Department of Energy. Washington, D.C. DOE/EIA-0121.
18 EIA (2016b) Natural Gas Annual 2016. Energy Information Administration, U.S. Department of Energy.
19 Washington, D.C. DOE/EIA-0131(06).
20 EIA (2016c) Monthly Energy Review, December 2016, Energy Information Administration, U.S. Department of
21 Energy, Washington, D.C. DOE/EIA-0035(2015/12).
22 EIA (1992) Coal and lignite production. EIA State Energy Data Report 1992, Energy Information Administration,
23 U.S. Department of Energy, Washington, D.C.
24 EPA (2010) Carbon Content Coefficients Developed for EPA's Mandatory Reporting Rule. Office of Air and
25 Radiation, Office of Atmospheric Programs, U.S. Environmental Protection Agency, Washington, D.C.
26 Fenton (2015) Personal communication. Michael Fenton, Commodity Specialist, U.S. Geological Survey and Marty
27 Wolf, Eastern Research Group. September 16, 2015.
28 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
29 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
30 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
31 IPCC/UNEP/OECD/IEA (1995) "Volume 3: Greenhouse Gas Inventory Reference Manual. Table 2-2". IPCC
32 Guidelines for National Greenhouse Gas Inventories. Intergovernmental Panel on Climate Change, United Nations
33 Environment Programme, Organization for Economic Co-Operation and Development, International Energy
34 Agency. IPCC WG1 Technical Support Unit, United Kingdom.
35 Unites States Geological Survey (USGS) (2017) 2017 USGSMinerals Yearbook-Iron and Steel. U.S. Geological
36 Survey, Reston, VA.
37 USGS (1991 through 2016) USGS Minerals Yearbook-Iron and Steel Scrap. U.S. Geological Survey, Reston, VA.
38 Ferroalloy Production
39 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
40 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
41 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
42 Onder, H., and E.A. Bagdoyan (1993) Everything You've Always Wanted to Know about Petroleum Coke. Allis
43 Mineral Systems.
References 10-29
-------
1 United States Geological Survey (USGS) (2016a) 2014Minerals Yearbook: Ferroalloys. U.S. Geological Survey,
2 Reston, VA. October 2016.
3 USGS (2017) Mineral Industry Surveys: Silicon in April 2017. U.S. Geological Survey, Reston, VA. June 2017.
4 USGS (2015a) 2012Minerals Yearbook: Ferroalloys. U.S. Geological Survey, Reston, VA. April 2015.
5 USGS (2016b) Mineral Industry Surveys: Silicon in December 2016. U.S. Geological Survey, Reston, VA.
6 December 2016.
7 USGS (2015b) Mineral Industry Surveys: Silicon in June 2015. U.S. Geological Survey, Reston, VA. September
8 2015.
9 USGS (2014) Mineral Industry Surveys: Silicon in September 2014. U.S. Geological Survey, Reston, VA.
10 December 2014.
11 USGS (1996 through 2013) Minerals Yearbook: Silicon. U.S. Geological Survey, Reston, VA.
12 Aluminum Production
13 EPA (2017) Greenhouse Gas Reporting Program (GHGRP). Envirofacts, Subpart: F Aluminum Production.
14 Available online at: .
15 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
16 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
17 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
18 USAA (2017) U.S. Primary Aluminum Production: Report for September 2017. U.S. Aluminum Association,
19 Washington, D.C. October, 2017.
20 USAA (2016b) U.S. Primary Aluminum Production: Report for August 2016. U.S. Aluminum Association,
21 Washington, D.C. August, 2016.
22 USAA (2016a) U.S. Primary Aluminum Production: Report for February 2016. U.S. Aluminum Association,
23 Washington, D.C. March, 2016.
24 USAA (2015) U.S. Primary Aluminum Production: Report for June 2015. U.S. Aluminum Association,
25 Washington, D.C. July, 2015.
26 USAA (2014) U.S. Primary Aluminum Production 2013. U.S. Aluminum Association, Washington, D.C. October,
27 2014.
28 USAA (2013) U.S. Primary Aluminum Production 2012. U.S. Aluminum Association, Washington, D.C. January,
29 2013.
30 USAA (2012) U.S. Primary Aluminum Production 2011. U.S. Aluminum Association, Washington, D.C. January,
31 2012.
32 USAA (2011) U.S. Primary Aluminum Production 2010. U.S. Aluminum Association, Washington, D.C.
33 USAA (2010) U.S. Primary Aluminum Production 2009. U.S. Aluminum Association, Washington, D.C.
34 USAA (2008, 2009) U.S. Primary Aluminum Production. U.S. Aluminum Association, Washington, D.C.
35 USAA (2004, 2005, 2006) Primary Aluminum Statistics. U.S. Aluminum Association, Washington, D.C.
36 USGS (2017) 2017Mineral Commodity Summaries: Aluminum. U.S. Geological Survey, Reston, VA.
37 USGS (2007) 2006Mineral Yearbook: Aluminum. U.S. Geological Survey, Reston, VA.
38 USGS (1995, 1998, 2000, 2001, 2002) Minerals Yearbook: Aluminum Annual Report. U.S. Geological Survey,
39 Reston, VA.
10-30 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Magnesium Production and Processing
2 ARB. 2015. "Magnesium casters successfully retool for a cleaner future." California Air Resources Board News
3 Release. Release # 15-07. February 5, 2015. Accessed October 2017. Available online at:
4 .
5 Bartos S., C. Laush, J. Scharfenberg, and R. Kantamaneni (2007) "Reducing greenhouse gas emissions from
6 magnesium die casting." Journal of Cleaner Production, 15: 979-987, March.
7 EPA (2017) Envirofacts. Greenhouse Gas Reporting Program (GHGRP), Subpart T: Magnesium Production and
8 Processing. Available online at: . Accessed on October, 2017.
9 Gjestland, H. and D. Magers (1996) "Practical Usage of Sulphur [Sulfur] Hexafluoride for Melt Protection in the
10 Magnesium Die Casting Industry." #13, 1996Annual Conference Proceedings, International Magnesium
11 Association. Ube City, Japan.
12 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
13 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
14 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
15 RAND (2002) RAND Environmental Science and Policy Center, "Production and Distribution of SF6 by End-Use
16 Applications" Katie D. Smythe. International Conference on SF6 and the Environment: Emission Reduction
17 Strategies. San Diego, CA. November 21-22, 2002.
18 USGS (2016, 2015, 2014, 2013, 2012, 2011, 2010, 2009, 2008, 2007, 2006, 2005a, 2003, 2002) Minerals Yearbook:
19 Magnesium Annual Report. U.S. Geological Survey, Reston, VA. Available online at:
20 .
21 USGS (2010b) Mineral Commodity Summaries: Magnesium Metal. U.S. Geological Survey, Reston, VA. Available
22 online at: .
23 USGS (2005b) Personal Communication between Deborah Kramer of the USGS and Jeremy Scharfenberg of ICF
24 Consulting.
25 Lead Production
26 Dutrizac, J.E., V. Ramachandran, and J.A. Gonzalez (2000) Lead-Zinc 2000. The Minerals, Metals, and Materials
27 Society.
28 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
29 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
30 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
31 Morris, D., F.R. Steward, and P. Evans (1983) Energy Efficiency of a Lead Smelter. Energy 8(5):337-349.
32 Sjardin, M. (2003) CO2 Emission Factors for Non-Energy Use in the Non-Ferrous Metal, Ferroalloys and
33 Inorganics Industry. Copernicus Institute. Utrecht, the Netherlands.
34 Ullman (1997) Ullman's Encyclopedia of Industrial Chemistry: Fifth Edition. Volume A5. John Wiley and Sons.
35 United States Geological Survey (USGS) (2017) 2017Mineral Commodity Summary, Lead. U.S. Geological
36 Survey, Reston, VA. January 2017. Available online at:
37 .
38 USGS (2016) 2016Mineral Commodity Summary, Lead. U.S. Geological Survey, Reston, VA. January 2016.
39 USGS (2015) 2015Mineral Commodity Summary, Lead. U.S. Geological Survey, Reston, VA. January 2015.
40 USGS (2014) Mineral Commodity Summary, Lead. U.S. Geological Survey, Reston, VA. February 2014.
41 USGS (1995 through 2013) Minerals Yearbook: Lead Annual Report. U.S. Geological Survey, Reston, VA.
References 10-31
-------
1 Zinc Production
2 Horsehead Corp. (2016) Form 10-k, Annual Report for the Fiscal Year Ended December 31, 2015. Available online
3 at: . Submitted on
4 January 25, 2017.
5 Horsehead Corp. (2015) Form 10-k, Annual Report for the Fiscal Year Ended December 31, 2014. Available online
6 at: . Submitted
7 on March 2, 2015.
8 Horsehead Corp. (2014) Form 10-k, Annual Report for the Fiscal Year Ended December 31, 2013. Available online
9 at: . Submitted
10 on March 13, 2014.
11 Horsehead Corp. (2013) Form 10-k, Annual Report for the Fiscal Year Ended December 31, 2012. Available online
12 at: .
13 Submitted March 18, 2013.
14 Horsehead Corp. (2012a) Form 10-k, Annual Report for the Fiscal Year Ended December, 31, 2011. Available
15 online at: . Submitted
16 on March 9, 2012.
17 Horsehead Corp. (2012b) Horsehead''s New Zinc Plant and its Impact on the Zinc Oxide Business. February 22,
18 2012. Available online at: . Accessed on
19 September 10, 2015.
20 Horsehead Corp. (2011) 10-k Annual Report for the Fiscal Year Ended December, 31 2010. Available online at:
21 . Submitted on March 16, 2011.
22 Horsehead Corp. (2010a) 10-k Annual Report for the Fiscal Year Ended December, 31 2009. Available online at:
23 . Submitted on March 16, 2010.
24 Horsehead Corp. (2010b) Horsehead Holding Corp. Provides Update on Operations at its Monaco, PA Plant. July
25 28, 2010. Available online at: .
26 Horsehead Corp (2008) 10-k Annual Report for the Fiscal Year Ended December 31, 2007. Available online at:
27 . Submitted on March 31, 2008.
28 Horsehead Corp (2007) Registration Statement (General Form) S-l. Available online at . Submitted on April 13, 2007.
30 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
31 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
32 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
33 Nyrstar (2017) 2016 Clarksville Fact Sheet. Available online at:
34 http://www.nyrstar.eom/~/media/Files/N/Nyrstar/operations/melting/fact-sheet-clarksville-en.pdI>. Accessed on
35 September 27, 2017.
36 Nyrstar (2016) 2015 Clarksville Fact Sheet.
37 PIZO (2017) Available online at . Accessed on January 12, 2017.
38 PIZO (2014) Available online at . Accessed on December 9, 2014.
39 PIZO (2012) Available online at . Accessed on October 10, 2012.
40 SDR (2015) Personal communication. Jeremy Whitten, EHS Manager, Steel Dust Recycling LLC and Gopi Manne,
41 Eastern Research Group, Inc. September 22, 2015.
42 SDR (2014) Personal communication. Art Rowland, Plant Manager, Steel Dust Recycling LLC and Gopi Manne,
43 Eastern Research Group, Inc. December 9, 2014.
44 SDR (2013) Available online at . Accessed on October 29, 2013.
10-32 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 SDR (2012) Personal communication. Art Rowland, Plant Manager, Steel Dust Recycling LLC and Gopi Manne,
2 Eastern Research Group, Inc. October 5, 2012.
3 Sjardin (2003) CO2 Emission Factors for Non-Energy Use in the Non-Ferrous Metal, Ferroalloys and Inorganics
4 Industry. Copernicus Institute. Utrecht, the Netherlands.
5 Steel Dust Recycling (SDR) (2017) Personal communication. Jeremy Whitten, EHS Manager, Steel Dust Recycling
6 LLC and John Steller, U.S. Environmental Protection Agency. January 26, 2017.
7 United States Geological Survey (USGS) (2017) 2017Mineral Commodity Summary: Zinc. U.S. Geological Survey,
8 Reston, VA. January 2017.
9 USGS (2016) 2016Mineral Commodity Summary: Zinc. U.S. Geological Survey, Reston, VA. January 2016.
10 USGS (2015) 2015Mineral Commodity Summary: Zinc. U.S. Geological Survey, Reston, VA. January 2015.
11 USGS (1995 through 2014) Minerals Yearbook: Zinc Annual Report. U.S. Geological Survey, Reston, VA.
12 Viklund-White (2000) The use ofLCA for the environmental evaluation of the recycling of galvanized steel. ISIJ
13 International, Vol. 40. No. 3, pp 292-299.
14 Semiconductor Manufacture
15 Burton, C.S., and R. Beizaie (2001) "EPA's PFC Emissions Model (PEVM) v. 2.14: Description and
16 Documentation" prepared for Office of Global Programs, U. S. Environmental Protection Agency, Washington, DC.
17 November 2001.
18 Citigroup Smith Barney (2005) Global Supply/Demand Model for Semiconductors. March 2005.
19 Doering, R. and Nishi, Y (2000) "Handbook of Semiconductor Manufacturing Technology", Marcel Dekker, New
20 York, USA, 2000.
21 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
22 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
23 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
24 ISMI (2009) Analysis of Nitrous Oxide Survey Data. Walter Worth. June 8, 2009. Available online at:
25 .
26 ITRS (2007, 2008, 2011, 2013) International Technology Roadmap for Semiconductors: 2006 Update, January
27 2007; International Technology Roadmap for Semiconductors: 2007 Edition, January 2008; International
28 Technology Roadmap for Semiconductors: 2011, January 2012; Update, International Technology Roadmap for
29 Semiconductors: 2013 Edition, Available online at: . These
30 and earlier editions and updates are available online at: . Information about the number of
31 interconnect layers for years 1990-2010 is contained in Burton and Beizaie, 2001. PEVM is updated using new
32 editions and updates of the ITRS, which are published annually.
33 SEMI - Semiconductor Equipment and Materials Industry (2017) World Fab Forecast, August 2017 Edition.
34 SEMI - Semiconductor Equipment and Materials Industry (2016) World Fab Forecast, May 2016 Edition.
35 SEMI - Semiconductor Equipment and Materials Industry (2013) World Fab Forecast, May 2013 Edition.
36 SEMI - Semiconductor Equipment and Materials Industry (2012) World Fab Forecast, August 2012 Edition.
37 Semiconductor Industry Association (SIA) (2009-2011) STATS: SICAS Capacity and Utilization Rates Q1-Q4
38 2008, Q1-Q4 2009, Q1-Q4 2010. Available online at:
39 .
40 United States Census Bureau (USCB) (2011, 2012, 2015, 2016) Historical Data: Quarterly Survey of Plant
41 Capacity Utilization. Available online at:
42 .
References 10-33
-------
1 U.S. EPA (2006) Uses and Emissions ofLiquid PFC Heat Transfer Fluids from the Electronics Sector. U.S.
2 Environmental Protection Agency, Washington, DC. EPA-430-R-06-901.
3 U.S. EPA Greenhouse Gas Reporting Program (GHGRP) Envirofacts. Subpart I: Electronics Manufacture.
4 Available online at: .
5 VLSI Research, Inc. (2012) Worldwide Silicon Demand. August 2012.
6 Substitution of Ozone Depleting Substances
7 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
8 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
9 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
10 Electrical Transmission and Distribution
11 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
12 Assessment Report of the Intergovernmental Panel on Climate Change. S. Solomon, D. Qin, M. Manning, Z. Chen,
13 M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.). Cambridge University Press. Cambridge, United
14 Kingdom 996 pp.
15 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
16 Inventories Programme, The Intergovernmental Panel on Climate Change. H.S. Eggleston, L. Buendia, K. Miwa, T.
17 Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
18 IPCC (1996) Climate Change 1995: The Science of Climate Change. Intergovernmental Panel on Climate Change,
19 J.T. Houghton, L.G. Meira Filho, B.A. Callander, N. Harris, A. Kattenberg, and K. Maskell (eds.). Cambridge
20 University Press. Cambridge, United Kingdom.
21 Levin et al. (2010) "The Global SF6 Source Inferred from Long-term High Precision Atmospheric Measurements
22 and its Comparison with Emission Inventories." Atmospheric Chemistry and Physics, 10: 2655-2662.
23 O'Connell, P., F. Heil, J. Henriot, G. Mauthe, H. Morrison, L. Neimeyer, M. Pittroff, R. Probst, J.P. Tailebois
24 (2002) SF6 in the Electric Industry, Status 2000, CIGRE. February 2002.
25 RAND (2004) "Trends in SF6 Sales and End-Use Applications: 1961-2003," Katie D. Smythe. International
26 Conference on SF6 and the Environment: Emission Reduction Strategies. RAND Environmental Science and Policy
27 Center, Scottsdale, AZ. December 1-3, 2004.
28 UDI (2017) 2017 UDI Directory of Electric Power Producers and Distributors, 125th Edition, Platts.
29 UDI (2013) 2013 UDI Directory of Electric Power Producers and Distributors,121st Edition, Platts.
30 UDI (2010) 2010 UDI Directory of Electric Power Producers and Distributors, 118th Edition, Platts.
31 UDI (2007) 2007 UDI Directory of Electric Power Producers and Distributors, 115th Edition, Platts.
32 UDI (2004) 2004 UDI Directory of Electric Power Producers and Distributors, 112th Edition, Platts.
33 UDI (2001) 2001 UDI Directory of Electric Power Producers and Distributors, 109th Edition, Platts.
34 UNFCCC (2014) Report of the Conference of the Parties on its nineteenth session, held in Warsaw from 11 to 23
35 November 2013. United Nations Framework Convention on Climate Change, Warsaw. (FCCC/CP/2013/10/Add.3).
36 January 31, 2014. Available online at: .
37 Nitrous Oxide from Product Use
38 CGA (2003) "CGA Nitrous Oxide Abuse Hotline: CGA/NWSA Nitrous Oxide Fact Sheet." Compressed Gas
39 Association. November 3, 2003.
40 CGA (2002) "CGA/NWSA Nitrous Oxide Fact Sheet." Compressed Gas Association. March 25, 2002.
10-34 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Heydorn, B. (1997) "Nitrous Oxide—North America." Chemical Economics Handbook, SRI Consulting. May 1997.
2 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
3 Assessment Report of the Intergovernmental Panel on Climate Change. [S. Solomon, D. Qin, M. Manning, Z. Chen,
4 M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.)]. Cambridge University Press. Cambridge, United
5 Kingdom 996 pp.
6 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
7 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
8 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
9 Ottinger (2014) Personal communication. Deborah Ottinger (CCD, U.S. EPA) and Mausami Desai (U.S. EPA).
10 Email received on January 29, 2014.
11 Tupman, M. (2003) Personal communication. Martin Tupman, Airgas Nitrous Oxide and Daniel Lieberman, ICF
12 International. August 8, 2003.
13 Industrial Processes and Product Use Sources of Indirect
14 Greenhouse Gases
15 EPA (2016) "1970-2016 Average annual emissions, all criteria pollutants in MS Excel." National Emissions
16 Inventory (NEI) Air Pollutant Emissions Trends Data. Office of Air Quality Planning and Standards, December
17 2016. Available online at: .
18 EPA (2003) E-mail correspondence containing preliminary ambient air pollutant data. Office of Air Pollution and
19 the Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency. December 22, 2003.
20 EPA (1997) Compilation of Air Pollutant Emission Factors, AP-42. Office of Air Quality Planning and Standards,
21 U.S. Environmental Protection Agency. Research Triangle Park, NC. October 1997.
22 Agriculture
23 Enteric Fermentation
24 Archibeque, S. (2011) Personal Communication. Shawn Archibeque, Colorado State University, Fort Collins,
25 Colorado and staff at ICF International.
26 Crutzen, P. J., I. Aselmann, and W. Seiler (1986) Methane Production by Domestic Animals, Wild Ruminants, Other
27 Herbivores, Fauna, and Humans. Tellus, 38B:271-284.
28 Donovan, K. (1999) Personal Communication. Kacey Donovan, University of California at Davis and staff at ICF
29 International.
30 Doren, P.E., J. F. Baker, C. R. Long and T. C. Cartwright (1989) Estimating Parameters of Growth Curves of Bulls,
31 J Animal Science 67:1432-1445.
32 Enns, M. (2008) Personal Communication. Dr. Mark Enns, Colorado State University and staff at ICF International.
33 EPA (2002) Quality Assurance/Quality Control and Uncertainty Management Plan for the U.S. Greenhouse Gas
34 Inventory: Procedures Manual for Quality Assurance/Quality Control and Uncertainty Analysis, U.S. Greenhouse
35 Gas Inventory Program, U.S. Environmental Protection Agency, Office of Atmospheric Programs, EPA 430-R-02-
36 007B, June 2002.
37 ERG (2016) Development of Methane Conversion Rate Scaling Factor and Diet-Related Inputs to the Cattle Enteric
38 Fermentation Model for Dairy Cows, Dairy Heifers, and Feedlot Animals. ERG, Lexington, MA. December 2016.
References 10-35
-------
1 Galyean and Gleghorn (2001) Summary of the 2000 Texas Tech University Consulting Nutritionist Survey. Texas
2 Tech University. Available online at . June
3 2009.
4 Holstein Association (2010) History of the Holstein Breed (website). Available online at:
5 . Accessed September 2010.
6 ICF (2006) Cattle Enteric Fermentation Model: Model Documentation. Prepared by ICF International for the
7 Environmental Protection Agency. June 2006.
8 ICF (2003) Uncertainty Analysis of2001 Inventory Estimates of Methane Emissions from Livestock Enteric
9 Fermentation in the U.S. Memorandum from ICF International to the Environmental Protection Agency. May 2003.
10 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
11 Assessment Report of the Intergovernmental Panel on Climate Change. S. Solomon, D. Qin, M. Manning, Z. Chen,
12 M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.). Cambridge University Press. Cambridge, United
13 Kingdom 996 pp.
14 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
15 Inventories Programme, The Intergovernmental Panel on Climate Change. H.S. Eggleston, L. Buendia, K. Miwa, T.
16 Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
17 Johnson, D. (2002) Personal Communication. Don Johnson, Colorado State University, Fort Collins, and ICF
18 International.
19 Johnson, D. (1999) Personal Communication. Don Johnson, Colorado State University, Fort Collins, and David
20 Conneely, ICF International.
21 Kebreab E., K. A. Johnson, S. L. Archibeque, D. Pape, and T. Wirth (2008) Model for estimating enteric methane
22 emissions from United States dairy and feedlot cattle. J. Anim. Sci. 86: 2738-2748.
23 Lippke, H., T. D. Forbes, and W. C. Ellis. (2000) Effect of supplements on growth and forage intake by stacker
24 steers grazing wheat pasture. J. Anim. Sci. 78:1625-1635.
25 National Bison Association (1999) Total Bison Population—1999. Report provided during personal email
26 communication with Dave Carter, Executive Director, National Bison Association, July 19, 2011.
27 Pinchak, W.E., D. R. Tolleson, M. McCloy, L. J. Hunt, R. J. Gill, R. J. Ansley, and S. J. Bevers (2004) Morbidity
28 effects on productivity and profitability of stacker cattle grazing in the southern plains. J. Anim. Sci. 82:2773 -2779.
29 Platter, W. J., J. D. Tatum, K. E. Belk, J. A. Scanga, and G. C. Smith (2003) Effects of repetitive use of hormonal
30 implants on beef carcass quality, tenderness, and consumer ratings of beef palatability. J. Anim. Sci. 81:984-996.
31 Preston, R.L. (2010) What's The Feed Composition Value of That Cattle Feed? Beef Magazine, March 1, 2010.
32 Available at: .
33 Skogerboe, T. L., L. Thompson, J. M. Cunningham, A. C. Brake, V. K. Karle (2000) The effectiveness of a single
34 dose of doramectin pour-on in the control of gastrointestinal nematodes in yearling stacker cattle. Vet. Parasitology
35 87:173-181.
36 Soliva, C.R. (2006) Report to the attention of IPCC about the data set and calculation method used to estimate
37 methane formation from enteric fermentation of agricultural livestock population and manure management in Swiss
38 agriculture. On behalf of the Federal Office for the Environment (FOEN), Berne, Switzerland.
39 U.S. Department of Agriculture USDA (2017) Quick Stats: Agricultural Statistics Database. National Agriculture
40 Statistics Service, U.S. Department of Agriculture. Washington, D.C. Available online at
41 . Accessed June 1, 2017.
42 USDA (2016) Quick Stats: Agricultural Statistics Database. National Agriculture Statistics Service, U.S.
43 Department of Agriculture. Washington, D.C. Available online at . Accessed
44 August 1, 2016.
45 USDA (2007) Census of Agriculture: 2007 Census Report. United States Department of Agriculture. Available
46 online at: .
10-36 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 USDA (2002) Census of Agriculture: 2002 Census Report. United States Department of Agriculture. Available
2 online at: .
3 USDA (1997) Census of Agriculture: 1997 Census Report. United States Department of Agriculture. Available
4 online at: . Accessed July 18, 2011.
5 USDA (1996) Beef Cow/Calf Health and Productivity Audit (CHAPA): Forage Analyses from Cow /CalfHerds in 18
6 States. National Agriculture Statistics Service, U.S. Department of Agriculture. Washington, D.C. Available online
7 at . March 1996.
8 USDA (1992) Census of Agriculture: 1992 Census Report. United States Department of Agriculture. Available
9 online at: . Accessed July 18, 2011.
10 USDA:APHIS:VS (2010) Beef 2007-08, Part V: Reference of Beef Cow-calf Management Practices in the United
11 States, 2007-08. USDA-APHIS-VS, CEAH. Fort Collins, CO.
12 USDA:APHIS:VS (2002) Reference of2002 Dairy Management Practices. USDA-APHIS-VS, CEAH. Fort
13 Collins, CO. Available online at .
14 USDA: APHIS: VS (1998) Beef '97, Parts I-IV. USDA-APHIS-VS, CEAH. Fort Collins, CO. Available online at
15 .
16 USDA:APHIS:VS (1996) Reference of1996 Dairy Management Practices. USDA-APHIS-VS, CEAH. Fort
17 Collins, CO. Available online at .
18 USDA: APHIS: VS (1994) Beef Cow/Calf Health and Productivity Audit. USDA-APHIS-VS, CEAH. Fort Collins,
19 CO. Available online at .
20 USDA: APHIS: VS (1993) Beef Cow/Calf Health and Productivity Audit. USDA-APHIS-VS, CEAH. Fort Collins,
21 CO. August 1993. Available online at .
22 Vasconcelos and Galyean (2007) Nutritional recommendations of feedlot consulting nutritionists: The 2007 Texas
23 Tech University Study. J. Anim. Sci. 85:2772-2781.
24 Manure Management
25 ASAE (1998) ASA E Standards 1998, 45th Edition. American Society of Agricultural Engineers. [St. Joseph, MI.]
26 Bryant, M.P., V.H. Varel, R.A. Frobish, and H.R. Isaacson (1976) In H.G. Schlegel (ed.)]; Seminar on Microbial
27 Energy Conversion. E. Goltz KG. Gottingen, Germany.
28 Bush, E. (1998) Personal communication with Eric Bush, Centers for Epidemiology and Animal Health, U.S.
29 Department of Agriculture regarding National Animal Health Monitoring System's (NAHMS) Swine '95 Study.
30 EPA (2016) AgSTAR Anaerobic Digester Database. Available online at:
31 .
32 EPA (2008) Climate leaders Greenhouse Gas Inventory Protocol Offset Project Methodology for Project Type
33 Managing Manure with Biogas Recovery Systems. Available online at:
34 .
35 EPA (2006) AgSTAR Digest. Office of Air and Radiation, U.S. Environmental Protection Agency. Washington, D.C.
36 Winter 2006. Available online at: . Retrieved July 2006.
37 EPA (2005) National Emission Inventory—Ammonia Emissions from Animal Agricultural Operations, Revised
38 Draft Report. U.S. Environmental Protection Agency. Washington, D.C. April 22, 2005. Available online at:
39 . Accessed
40 August 2007.
41 EPA (2003) AgSTAR Digest. Office of Air and Radiation, U.S. Environmental Protection Agency. Washington, D.C.
42 Winter 2003. Available online at: . Retrieved July 2006.
References 10-37
-------
1 EPA (2002a) Development Document for the Final Revisions to the National Pollutant Discharge Elimination
2 System (NPDES) Regulation and the Effluent Guidelines for Concentrated Animal Feeding Operations (CAFOS).
3 U.S. Environmental Protection Agency. EPA-821-R-03-001. December 2002.
4 EPA (2002b) Cost Methodology for the Final Revisions to the National Pollutant Discharge Elimination System
5 Regulation and the Effluent Guidelines for Concentrated Animal Feeding Operations. U.S. Environmental
6 Protection Agency. EPA-821-R-03-004. December 2002.
7 EPA (2000) AgSTAR Digest. Office of Air and Radiation, U.S. Environmental Protection Agency. Washington, D.C.
8 Spring 2000. Available online at: .
9 EPA (1992) Global Methane Emissions from Livestock and Poultry Manure, Office of Air and Radiation, U.S.
10 Environmental Protection Agency. February 1992.
11 ERG (2010a) "Typical Animal Mass Values for Inventory Swine Categories." Memorandum to EPA from ERG.
12 July 19, 2010.
13 ERG (2010b) Telecon with William Boyd of USD A NRCS and Cortney Itle of ERG Concerning Updated VS and
14 Nex Rates. August 8, 2010.
15 ERG (2010c) "Updating Current Inventory Manure Characteristics new USDA Agricultural Waste Management
16 Field Handbook Values." Memorandum to EPA from ERG. August 13, 2010.
17 ERG (2008) "Methodology for Improving Methane Emissions Estimates and Emission Reductions from Anaerobic
18 Digestion System for the 1990-2007 Greenhouse Gas Inventory for Manure Management." Memorandum to EPA
19 from ERG. August 18, 2008.
20 ERG (2003a) "Methodology for Estimating Uncertainty for Manure Management Greenhouse Gas Inventory."
21 Contract No. GS-10F-0036, Task Order 005. Memorandum to EPA from ERG, Lexington, MA. September 26,
22 2003.
23 ERG (2003b) "Changes to Beef Calves and Beef Cows Typical Animal Mass in the Manure Management
24 Greenhouse Gas Inventory." Memorandum to EPA from ERG, October 7, 2003.
25 ERG (2001) Summary of development ofMDP Factor for methane conversion factor calculations. ERG, Lexington,
26 MA. September 2001.
27 ERG (2000a) Calculations: Percent Distribution of Manure for Waste Management Systems. ERG, Lexington, MA.
28 August 2000.
29 ERG (2000b) Discussion of Methodology for Estimating Animal Waste Characteristics (Summary of B0 Literature
30 Review). ERG, Lexington, MA. June 2000.
31 Groffman, P.M., R. Brumme, K. Butterbach-Bahl, K.E. Dobbie, A.R. Mosier, D. Ojima, H. Papen, W.J. Parton,
32 K. A. Smith, and C. Wagner-Riddle (2000) "Evaluating annual nitrous oxide fluxes at the ecosystem scale." Global
33 Biogeochemical Cycles, 14(4):1061-1070.
34 Hashimoto, A.G. (1984) "Methane from Swine Manure: Effect of Temperature and Influent Substrate Composition
35 on Kinetic Parameter (k)." Agricultural Wastes, 9:299-308.
36 Hashimoto, A.G., V.H. Varel, and Y.R. Chen (1981) "Ultimate Methane Yield from Beef Cattle Manure; Effect of
37 Temperature, Ration Constituents, Antibiotics and Manure Age." Agricultural Wastes, 3:241-256.
38 Hill, D.T. (1984) "Methane Productivity of the Major Animal Types." Transactions of theASAE, 27(2):530-540.
39 Hill, D.T. (1982) "Design of Digestion Systems for Maximum Methane Production." Transactions of the ASAE,
40 25(l):226-230.
41 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
42 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
43 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
44 Morris, G.R. (1976) Anaerobic Fermentation of Animal Wastes: A Kinetic and Empirical Design Fermentation.
45 M.S. Thesis. Cornell University.
10-38 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 National Bison Association (1999) Total Bison Population—1999. Report provided during personal email
2 communication with Dave Carter, Executive Director, National Bison Association July 19, 2011.
3 Ott, S.L. (2000) Dairy '96 Study. Stephen L. Ott, Animal and Plant Health Inspection Service, U.S. Department of
4 Agriculture. June 19, 2000.
5 Safley, L.M., Jr. (2000) Personal Communication. Deb Bartram, ERG and L.M. Safley, President, Agri-Waste
6 Technology. June and October 2000.
7 Sweeten, J. (2000) Personal Communication. John Sweeten, Texas A&M University and Indra Mitra, ERG. June
8 2000.
9 UEP (1999) Voluntary Survey Results—Estimated Percentage Participation/Activity. Caged Layer Environmental
10 Management Practices, Industry data submissions for EPA profile development, United Egg Producers and National
11 Chicken Council. Received from John Thorne, Capitolink. June 2000.
12 United Nations Framework Convention on Climate Change (UNFCCC) (2017) Definitions. Available online at: <
13 http://unfccc.int/ghg_data/online_help/definitions/items/3817.php>. Accessed on December 8, 2017.
14 USDA (2017) Quick Stats: Agricultural Statistics Database. National Agriculture Statistics Service, U.S.
15 Department of Agriculture. Washington, D.C. Available online at: .
16 USDA (2016a) Quick Stats: Agricultural Statistics Database. National Agriculture Statistics Service, U.S.
17 Department of Agriculture. Washington, D.C. Available online at: .
18 USDA (2016b) Chicken and Eggs 2015 Summary. National Agriculture Statistics Service, U.S. Department of
19 Agriculture. Washington, D.C. February 2016. Available online at:
20 .
21 USDA (2016c) Poultry - Production and Value 2015 Summary. National Agriculture Statistics Service, U.S.
22 Department of Agriculture. Washington, D.C. April 2016. Available online at:
23 .
24 USDA (2016d) 1987, 1992, 1997, 2002, 2007, and 2012 Census of Agriculture. National Agriculture Statistics
25 Service, U.S. Department of Agriculture. Washington, D.C. Available online at:
26 . July 2016.
27 USDA (2015a) Chicken and Eggs 2014 Summary. National Agriculture Statistics Service, U.S. Department of
28 Agriculture. Washington, D.C. February 2015. Available online at:
29 .
30 USDA (2015b) Poultry - Production and Value 2014 Summary. National Agriculture Statistics Service, U.S.
31 Department of Agriculture. Washington, D.C. April 2015. Available online at:
32 .
33 USDA (2014a) 1987, 1992, 1997, 2002, 2007, and 2012 Census of Agriculture. National Agriculture Statistics
34 Service, U.S. Department of Agriculture. Washington, D.C. May 2014. Available online at:
35 .
36 USDA (2014b) Chicken and Eggs 2013 Summary. National Agriculture Statistics Service, U.S. Department of
37 Agriculture. Washington, D.C. February 2014. Available online at:
38 .
39 USDA (2014c) Poultry - Production and Value 2013 Summary. National Agriculture Statistics Service, U.S.
40 Department of Agriculture. Washington, D.C. April 2014. Available online at:
41 .
42 USDA (2013a) Chicken and Eggs 2012 Summary. National Agriculture Statistics Service, U.S. Department of
43 Agriculture. Washington, D.C. February 2013. Available online at:
44 .
45 USDA (2013b) Poultry - Production and Value 2012 Summary. National Agriculture Statistics Service, U.S.
46 Department of Agriculture. Washington, D.C. April 2013. Available online at:
47 .
References 10-39
-------
1 USDA (2012a) Chicken and Eggs 2011 Summary. National Agriculture Statistics Service, U.S. Department of
2 Agriculture. Washington, D.C. February 2012. Available online at:
3 .
4 USDA (2012b) Poultry - Production and Value 2011 Summary. National Agriculture Statistics Service, U.S.
5 Department of Agriculture. Washington, D.C. April 2012. Available online at:
6 .
7 USDA (201 la) Chicken and Eggs 2010 Summary. National Agriculture Statistics Service, U.S. Department of
8 Agriculture. Washington, D.C. February 2011. Available online at:
9 .
10 USDA (201 lb) Poultry - Production and Value 2010 Summary. National Agriculture Statistics Service, U.S.
11 Department of Agriculture. Washington, D.C. April 2011. Available online at:
12 .
13 USDA (2010a) Chicken and Eggs 2009 Summary. National Agriculture Statistics Service, U.S. Department of
14 Agriculture. Washington, D.C. February 2010. Available online at:
15 .
16 USDA (2010b) Poultry - Production and Value 2009 Summary. National Agriculture Statistics Service, U.S.
17 Department of Agriculture. Washington, D.C. April 2010. Available online at:
18 .
19 USDA (2009a) Chicken and Eggs 2008 Summary. National Agriculture Statistics Service, U.S. Department of
20 Agriculture. Washington, D.C. February 2009. Available online at:
21 .
22 USDA (2009b) Poultry - Production and Value 2008 Summary. National Agriculture Statistics Service, U.S.
23 Department of Agriculture. Washington, D.C. April 2009. Available online at:
24 .
25 USDA (2009c) Chicken and Eggs - Final Estimates 2003-2007. National Agriculture Statistics Service, U.S.
26 Department of Agriculture. Washington, D.C. March 2009. Available online at:
27 .
28 USDA (2009d) Poultry Production and Value—Final Estimates 2003-2007. National Agriculture Statistics Service,
29 U.S. Department of Agriculture. Washington, D.C. May 2009. Available online at:
30 .
31 USD A (2008) Agricultural Waste Management Field Handbook, National Engineering Handbook (NEH), Part 651.
32 Natural Resources Conservation Service, U.S. Department of Agriculture.
33 USDA (2004a) Chicken and Eggs—Final Estimates 1998-2003. National Agriculture Statistics Service, U.S.
34 Department of Agriculture. Washington, D.C. April 2004. Available online at:
35 .
36 USDA (2004b) Poultry Production and Value—Final Estimates 1998-2002. National Agriculture Statistics Service,
37 U.S. Department of Agriculture. Washington, D.C. April 2004. Available online at:
38 .
39 USDA (1999) Poultry Production and Value—Final Estimates 1994-97. National Agriculture Statistics Service,
40 U.S. Department of Agriculture. Washington, D.C. March 1999. Available online at:
41 .
42 USDA (1998) Chicken and Eggs—Final Estimates 1994-97. National Agriculture Statistics Service, U.S.
43 Department of Agriculture. Washington, D.C. December 1998. Available online at:
44 .
45 USDA (1996) Agricultural Waste Management Field Handbook, National Engineering Handbook (NEH), Part 651.
46 Natural Resources Conservation Service, U.S. Department of Agriculture. July 1996.
10-40 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 USDA (1995a) Poultry Production and Value—Final Estimates 1988-1993. National Agriculture Statistics Service,
2 U.S. Department of Agriculture. Washington, D.C. March 1995. Available online at:
3 .
4 USDA (1995b) Chicken and Eggs—Final Estimates 1988-1993. National Agriculture Statistics Service, U.S.
5 Department of Agriculture. Washington, D.C. December 1995. Available online at:
6 .
7 USDA (1994) Sheep and Goats—Final Estimates 1989-1993. National Agriculture Statistics Service, U.S.
8 Department of Agriculture. Washington, D.C. January 31, 1994. Available online at:
9 .
10 USDA APHIS (2003) Sheep 2001, Part I: Reference ofSheep Management in the United States, 2001 andPart IV:
11 Baseline Reference of2001 Sheep Feedlot Health and Management. USDA-APHIS-VS. Fort Collins, CO.
12 #N356.0702. Available online at.
13 USDA APHIS (2000) Layers '99—Part II: References of1999 Table Egg Layer Management in the U.S. USDA-
14 APHIS-VS. Fort Collins, CO. Available online at
15 .
16 USDA APHIS (1996) Swine '95: Grower/Finisher Part II: Reference of1995 U.S. Grower/Finisher Health &
17 Management Practices. USDA-APHIS-VS. Fort Collins, CO. Available online at:
18 .
19 vN Cultivation
20 Baicich, P. (2013) The Birds and Rice Connection. Bird Watcher's Digest. Available online at:
21 .
22 Brockwell, P. J., and R.A. Davis (2016) Introduction to time series and forecasting. Springer.
23 Cantens, G. (2004 through 2005) Personal Communication. Janet Lewis, Assistant to Gaston Cantens, Vice
24 President of Corporate Relations, Florida Crystals Company and ICF International.
25 Cheng, K., S.M. Ogle, W.J. Parton, G. Pan. (2014) "Simulating greenhouse gas mitigation potentials for Chinese
26 croplands using the DAYCENT ecosystem model." Global Change Biology 20:948-962.
27 Cheng, K., S.M. Ogle, W.J. Parton and G. Pan. (2013) "Predicting methanogenesis from rice paddies using the
28 DAYCENT ecosystem model." Ecological Modelling 261-262:19-31.
29 Del Grosso, S.J., S.M. Ogle, W.J. Parton, and F.J. Breidt (2010) "Estimating Uncertainty in N20 Emissions from
30 U.S. Cropland Soils." Global Biogeochemical Cycles, 24, GB1009, doi:10.1029/2009GB003544.
31 Deren, C. (2002) Personal Communication and Dr. Chris Deren, Everglades Research and Education Centre at the
32 University of Florida and Caren Mintz, ICF International. August 15, 2002.
33 Fitzgerald, G. J., K. M. Scow & J. E. Hill (2000) "Fallow Season Straw and Rice Management Effects on Methane
34 Emissions in California Rice." Global biogeochemical cycles, 14 (3), 767-776.
35 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
36 the northern Central Valley of California determined by satellite imagery. California Fish and Game, 91: 207-215.
37 Gonzalez, R. (2007 through 2014) Email correspondence. Rene Gonzalez, Plant Manager, Sem-Chi Rice Company
38 and ICF International.
39 Hardke, J.T., (2015) Trends in Arkansas rice production, 2014. B.R. Wells Arkansas Rice Research Studies 2014.
40 Norman, R.J., and Moldenhauer, K.A.K., (Eds.). Research Series 626, Arkansas Agricultural Experiment Station,
41 University of Arkansas.
42 Hardke, J. (2014) Personal Communication. Dr. Jarrod Hardke, Rice Extension Agronomist at the University of
43 Arkansas Rice Research and Extension Center and Kirsten Jaglo, ICF International. September 11, 2014.
44 Hardke, J. (2013) Email correspondence. Dr. Jarrod Hardke, Rice Extension Agronomist at the University of
45 Arkansas Rice Research and Extension Center and Cassandra Snow, ICF International. July 15, 2013.
References 10-41
-------
1 Hardke, J.T., and Wilson, C.E. Jr., (2014) Trends in Arkansas rice production, 2013. B.R. Wells Arkansas Rice
2 Research Studies 2013. Norman, R.J., and Moldenhauer, K.A.K., (Eds.). Research Series 617, Arkansas Agricultural
3 Experiment Station, University of Arkansas.
4 Hardke, J.T., and Wilson, C.E. Jr., (2013) Trends in Arkansas rice production. B.R. Wells Arkansas Rice Research
5 Studies 2012. Norman, R.J., and Moldenhauer, K.A.K., (Eds.). Research Series 609, Arkansas Agricultural
6 Experiment Station, University of Arkansas.
7 Hollier, C. A. (ed), (1999) Louisiana rice production handbook. Louisiana State University Agricultural Center.
8 LCES Publication Number 2321. 116 pp.
9 Holzapfel-Pschorn, A., R. Conrad, and W. Seiler (1985) "Production, Oxidation, and Emissions of Methane in Rice
10 Paddies." FEMSMicrobiology Ecology, 31:343-351.
11 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
12 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
13 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
14 Kirstein, A. (2003 through 2004, 2006) Personal Communication. Arthur Kirstein, Coordinator, Agricultural
15 Economic Development Program, Palm Beach County Cooperative Extension Service, FL and ICF International.
16 Klosterboer, A. (1997, 1999 through 2003) Personal Communication. Arlen Klosterboer, retired Extension
17 Agronomist, Texas A&M University and ICF International. July 7, 2003.
18 Lindau, C.W. and P.K. Bollich (1993) "Methane Emissions from Louisiana First and Ratoon Crop Rice." Soil
19 Science, 156:42-48.
20 Linquist, B.A., M.A. Adviento-Borbe, C.M. Pittelkow, C.v. Kessel, et al. (2012) Fertilizer management practices
21 and greenhouse gas emissions from rice systems: A quantitative review and analysis. Field Crops Research, 135:10-
22 21.
23 Linscombe, S. (1999, 2001 through 2014) Email correspondence. Steve Linscombe, Professor with the Rice
24 Research Station at Louisiana State University Agriculture Center and ICF International.
25 LSU, (2015) Louisiana ratoon crop and conservation: Ratoon & Conservation Tillage Estimates. Louisiana State
26 University, College of Agriculture AgCenter. Online at: www.lsuagcenter.com
27 Miller, M.R., Garr, J.D., and Coates, P.S., (2010) Changes in the status of harvested rice fields in the Sacramento
28 Valley, California: Implications for wintering waterfowl. Wetlands, 30: 939-947.
29 Neue, H.U., R. Wassmann, H.K. Kludze, W. Bujun, and R.S. Lantin (1997) "Factors and processes controlling
30 methane emissions from rice fields." Nutrient Cycling in Agroecosystems 49: 111-117.
31 Ogle, S.M., F.J. Breidt, M. Easter, S. Williams and K. Paustian. (2007) "An empirically based approach for
32 estimating uncertainty associated with modeling carbon sequestration in soils." Ecological Modelling 205:453-463.
33 Ogle, S.M., S. Spencer, M. Hartman, L. Buendia, L. Stevens, D. du Toit, J. Witi (2016) "Developing national
34 baseline GHG emissions and analyzing mitigation potentials for agriculture and forestry using an advanced national
35 GHG inventory software system." In Advances in Agricultural Systems Modeling 6, Synthesis and Modeling of
36 Greenhouse Gas Emissions and Carbon Storage in Agricultural and Forestry Systems to Guide Mitigation and
37 Adaptation, S. Del Grosso, L.R. Ahuja and W.J. Parton (eds.), American Society of Agriculture, Crop Society of
38 America and Soil Science Society of America, pp. 129-148.
39 Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface Submodel:
40 Description and Testing". Glob. Planet. Chang. 19: 35-48.
41 Parton, W.J., D.S. Schimel, C.V. Cole, D.S. Ojima (1987) "Analysis of factors controlling soil organic matter levels
42 in Great Plains grasslands." Soil Science Society of America Journal 51:1173-1179.
43 Sass, R. L. (2001) CH4 Emissions from Rice Agriculture. Good Practice Guidance and Uncertainty Management in
44 National Greenhouse Gas Inventories. 399-417. Available online at: .
10-42 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Sass, R.L., F.M. Fisher, P. A. Harcombe, and F.T. Turner (1990) "Methane Production and Emissions in a Texas
2 Rice Field." Global Biogeochemical Cycles, 4:47-68.
3 Sass, R.L., F.M. Fisher, S.T. Lewis, M.F. Jund, and F.T. Turner. (1994) "Methane emissions from rice fields: effect
4 of soil texture." Global Biogeochemical Cycles 8:135-140.
5 Schueneman, T. (1997, 1999 through 2001) Personal Communication. Tom Schueneman, Agricultural Extension
6 Agent, Palm Beach County, FL and ICF International.
7 Slaton, N. (1999 through 2001) Personal Communication. Nathan Slaton, Extension Agronomist—Rice, University
8 of Arkansas Division of Agriculture Cooperative Extension Service and ICF International.
9 Stansel, J. (2004 through 2005) Email correspondence. Dr. Jim Stansel, Resident Director and Professor Emeritus,
10 Texas A&M University Agricultural Research and Extension Center and ICF International.
11 TAMU (2015) Texas Rice Crop Survey. Texas A&M AgriLIFE Research Center at Beaumont. Online at:
12 .
13 Texas Agricultural Experiment Station (2007 through 2014) Texas Rice Acreage by Variety. Agricultural Research
14 and Extension Center, Texas Agricultural Experiment Station, Texas A&M University System. Available online at:
15 .
16 Texas Agricultural Experiment Station (2006) 2005 - Texas Rice Crop Statistics Report. Agricultural Research and
17 Extension Center, Texas Agricultural Experiment Station, Texas A&M University System, p. 8. Available online at:
18 .
19 University of California Cooperative Extension (UCCE) (2015) Rice Production Manual. Revised (2015) UCCE,
20 Davis, in collaboration with the California Rice Research Board.
21 USD A (2005 through 2014) Crop Production Summary. National Agricultural Statistics Service, Agricultural
22 Statistics Board, U.S. Department of Agriculture, Washington, D.C. Available online at:
23 .
24 USD A (2012) Summary of USDA-ARS Research on the Interrelationship of Genetic and Cultural Management
25 Factors That Impact Grain Arsenic Accumulation in Rice. News and Events. Agricultural Research Service, U.S.
26 Department of Agriculture, Washington, D.C. Available online at:
27 . September 2013.
28 USDA (2003) Field Crops, Final Estimates 1997-2002. Statistical Bulletin No. 982. National Agricultural Statistics
29 Service, Agricultural Statistics Board, U.S. Department of Agriculture, Washington, D.C. Available online at:
30 . September 2005.
31 USDA (1998) Field Crops Final Estimates 1992-1997. Statistical Bulletin Number 947 a. National Agricultural
32 Statistics Service, Agricultural Statistics Board, U.S. Department of Agriculture, Washington, D.C. Available online
33 at: . July 2001.
34 USDA (1994) Field Crops Final Estimates 1987-1992. Statistical Bulletin Number 896. National Agricultural
35 Statistics Service, Agricultural Statistics Board, U.S. Department of Agriculture, Washington, D.C. Available online
36 at: . July 2001.
37 USDA-NRCS (2015) Summary Report: 2012 National Resources Inventory, Natural Resources Conservation
38 Service, Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
39 Available online at: .
40 van Bodegom, P.M., R. Wassmann, T.M. Metra-Corton (2001) "A process based model for methane emission
41 predictions from flooded rice paddies." Global Biogeochemical Cycles 15: 247-263.
42 Wang, J.J., S.K. Dodla, S. Viator, M. Kongchum, S. Harrison, S. D. Mudi, S. Liu, Z. Tian (2013) Agriculture Field
43 Management Practices and Greenhouse Gas Emissions from Louisiana Soils. Louisiana Agriculture, Spring 2013: 8-
44 9. Available online at: .
References 10-43
-------
1 Wassmann, R. H.U. Neue, R.S. Lantin, K. Makarim, N. Chareonsil5, L.V. Buendia, and H. Rennenberg (2000a)
2 Characterization of methane emissions from rice fields in Asia II. Differences among irrigated, rainfed, and
3 deepwater rice." Nutrient Cycling in Agroecosystems, 58(1): 13-22.
4 Wassmann, R., R.S. Lantin, H.U. Neue, L.V. Buendia, et al. (2000b) "Characterization of Methane Emissions from
5 Rice Fields in Asia. III. Mitigation Options and Future Research Needs." Nutrient Cycling in Agroecosystems,
6 58(l):23-36.
7 Way, M.O., McCauley, G.M., Zhou, X.G., Wilson, L.T., and Morace, B. (Eds.), (2014) 2014 Texas Rice Production
8 Guidelines. Texas A&M AgriLIFE Research Center at Beaumont.
9 Wilson, C. (2002 through 2007, 2009 through 2012) Personal Communication. Dr. Chuck Wilson, Rice Specialist at
10 the University of Arkansas Cooperative Extension Service and ICF International.
11 Wilson, C.E. Jr., and Branson, J. W., (2006) Trends in Arkansas rice production. B.R. Wells Arkansas Rice Research
12 Studies 2005. Norman, R.J., Meullenet, J.-F., and Moldenhauer, K.A.K., (Eds.). Research Series 540, Arkansas
13 Agricultural Experiment Station, University of Arkansas.
14 Wilson, C.E. Jr., and Branson, J.W., (2005) Trends in Arkansas rice production. B.R. Wells Arkansas Rice Research
15 Studies 2004. Norman, R.J., Meullenet, J.-F., and Moldenhauer, K.A.K., (Eds.). Research Series 529, Arkansas
16 Agricultural Experiment Station, University of Arkansas.
17 Wilson, C.E. Jr., Runsick, S.K., and Mazzanti, R., (2010) Trends in Arkansas rice production. B.R. Wells Arkansas
18 Rice Research Studies 2009. Norman, R.J., and Moldenhauer, K.A.K., (Eds.). Research Series 581, Arkansas
19 Agricultural Experiment Station, University of Arkansas.
20 Wilson, C.E. Jr., Runsick, S.K., Mazzanti, R., (2009) Trends in Arkansas rice production. B.R. Wells Arkansas Rice
21 Research Studies (2008) Norman, R.J., Meullenet, J.-F., and Moldenhauer, K.A.K., (Eds.). Research Series 571,
22 Arkansas Agricultural Experiment Station, University of Arkansas.
23 Wilson, C.E. Jr., and Runsick, S.K., (2008) Trends in Arkansas rice production. B.R. Wells Arkansas Rice Research
24 Studies 2007. Norman, R.J., Meullenet, J.-F., and Moldenhauer, K.A.K., (Eds.). Research Series 560, Arkansas
25 Agricultural Experiment Station, University of Arkansas.
26 Wilson, C.E. Jr., and Runsick, S.K., (2007) Trends in Arkansas rice production. B.R. Wells Arkansas Rice Research
27 Studies 2006. Norman, R.J., Meullenet, J.-F., and Moldenhauer, K.A.K., (Eds.). Research Series 550, Arkansas
28 Agricultural Experiment Station, University of Arkansas.
29 Yan, X., H. Akiyana, K. Yagi, and H. Akimoto (2009) "Global estimations of the inventory and mitigation potential
30 of methane emissions from rice cultivation conducted using the 2006 Intergovernmental Panel on Climate Change
31 Guidelines." Global Biogeochemical Cycles, 23, DOI: 0.1029/2008GB003299.
32 Young, M. (2013) Rice and Ducks. Ducks Unlimited, Memphis, TN. Available online at:
33 .
34 Agricultural Soil Management
35 AAPFCO (2008 through 2012) Commercial Fertilizers: 2008-2014. Association of American Plant Food Control
36 Officials. University of Missouri. Columbia, MO.
37 AAPFCO (1995 through 2000a, 2002 through 2007) Commercial Fertilizers: 1995-2007. Association of American
38 Plant Food Control Officials. University of Kentucky. Lexington, KY.
39 Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.Cibrowski, P.
40 (1996) Personal Communication. Peter Cibrowski, Minnesota Pollution Control Agency and Heike Mainhardt, ICF
41 Incorporated. July 29, 1996.
42 CTIC (2004) 2004 Crop Residue Management Survey. Conservation Technology Information Center. Available at
43 .
44 Del Grosso, S.J., A.R. Mosier, W.J. Parton, and D.S. Ojima (2005) "DAYCENT Model Analysis of Past and
45 Contemporary Soil N2O and Net Greenhouse Gas Flux for Major Crops in the USA." Soil Tillage and Research, 83:
46 9-24. doi: 10.1016/j.still.2005.02.007.
10-44 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Del Grosso, S.J., S.M. Ogle, W.J. Parton, and F.J. Breidt (2010) "Estimating Uncertainty in N20 Emissions from
2 U.S. Cropland Soils." GlobalBiogeochemical Cycles, 24, GB1009, doi:10.1029/2009GB003544.
3 Del Grosso, S.J., W.J. Parton, C.A. Keough, and M. Reyes-Fox. (2011) Special features of the DAYCENT modeling
4 package and additional procedures for parameterization, calibration, validation, and applications, in Methods of
5 Introducing System Models into Agricultural Research, L.R. Ahuja and Liwang Ma, editors, p. 155-176, American
6 Society of Agronomy, Crop Science Society of America, Soil Science Society of America, Madison, WI. USA.
7 Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel (2001)
8 "Simulated Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the DAYCENT Model." In
9 Schaffer, M., L. Ma, S. Hansen, (eds.). Modeling Carbon and Nitrogen Dynamics for Soil Management. CRC Press.
10 Boca Raton, Florida. 303-332.
11 Del Grosso, S.J., T. Wirth, S.M. Ogle, W.J. Parton (2008) Estimating agricultural nitrous oxide emissions. EOS 89,
12 529-530.
13 Delgado, J.A., S.J. Del Grosso, and S.M. Ogle (2009) "15N isotopic crop residue cycling studies and modeling
14 suggest that IPCC methodologies to assess residue contributions to N20-N emissions should be reevaluated."
15 Nutrient Cycling in Agroecosystems, DOI 10.1007/sl0705-009-9300-9.
16 Edmonds, L., N. Gollehon, R.L. Kellogg, B. Kintzer, L. Knight, C. Lander, J. Lemunyon, D. Meyer, D.C. Moffitt,
17 and J. Schaeffer (2003) "Costs Associated with Development and Implementation of Comprehensive Nutrient
18 Management Plans." Part 1. Nutrient Management, Land Treatment, Manure and Wastewater Handling and Storage,
19 and Recordkeeping. Natural Resource Conservation Service, U.S. Department of Agriculture.
20 EPA (2003) Clean Watersheds Needs Survey 2000—Report to Congress, U.S. Environmental Protection Agency.
21 Washington, D.C. Available online at: .
22 EPA (1999) Biosolids Generation, Use and Disposal in the United States. Office of Solid Waste, U.S.
23 Environmental Protection Agency. Available online at: .
24 EPA (1993) Federal Register. Part II. Standards for the Use and Disposal of Sewage Sludge; Final Rules. U.S.
25 Environmental Protection Agency, 40 CFR Parts 257, 403, and 503.
26 Firestone, M. K., and E.A. Davidson, Ed. (1989) Microbiological basis of NO and N20 production and consumption
27 in soil. Exchange of trace gases between terrestrial ecosystems and the atmosphere. New York, John Wiley & Sons.
28 ILENR (1993) Illinois Inventory of Greenhouse Gas Emissions and Sinks: 1990. Office of Research and Planning,
29 Illinois Department of Energy and Natural Resources. Springfield, IL.
30 IPCC (2013) 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands.
31 The Intergovernmental Panel on Climate Change. [T, Hiraishi, T. Krug, K. Tanabe, N. Srivastava, B. Jamsranjav,
32 M. Fukuda and T. Troxler (eds.)]. Hayama, Kanagawa, Japan.
33 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
34 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
35 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
36 McFarland, M.J. (2001) Biosolids Engineering, New York: McGraw-Hill, p. 2.12.
37 McGill, W.B., and C.V. Cole (1981) Comparative aspects of cycling of organic C, N, S and P through soil organic
38 matter. Geoderma 26:261-22,6.
39 Mesinger, F., G. DiMego, E. Kalnay, K. Mitchell, P. C. Shafran, W. Ebisuzaki, D. Jovic, J. Woollen, E. Rogers, E.
40 Mosier, A. R., J.M. Duxbury, J.R. Freney, O. Heinemeyer, K. Minami (1998) "Assessing and mitigating N20
41 emissions from agricultural soils." Climatic Change 40: 7-38.
42 NASS (2004) Agricultural Chemical Usage: 2003 Field Crops Summary. Report AgChl(04)a, National Agricultural
43 Statistics Service, U.S. Department of Agriculture. Available online at:
44 .
45 NASS (1999) Agricultural Chemical Usage: 1998 Field Crops Summary. Report AgChl(99). National Agricultural
46 Statistics Service, U.S. Department of Agriculture. Available online at:
47 .
References 10-45
-------
1 NASS (1992) Agricultural Chemical Usage: 1991 Field Crops Summary. Report AgChl(92). National Agricultural
2 Statistics Service, U.S. Department of Agriculture. Available online at:
3 .
4 NEBRA (2007) A National Biosolids Regulation, Quality, End Use & Disposal Survey. North East Biosolids and
5 Residuals Association, July 21, 2007.
6 Noller, J. (1996) Personal Communication. John Noller, Missouri Department of Natural Resources and Heike
7 Mainhardt, ICF Incorporated. July 30, 1996.
8 Oregon Department of Energy (1995) Report on Reducing Oregon's Greenhouse Gas Emissions: Appendix D
9 Inventory and Technical Discussion. Oregon Department of Energy. Salem, OR.
10 Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface Submodel:
11 Description and Testing." Glob. Planet. Chang. 19: 35-48.
12 Potter, C., S. Klooster, A. Huete, and V. Genovese (2007) Terrestrial carbon sinks for the United States predicted
13 fromMODIS satellite data and ecosystem modeling. Earth Interactions 11, Article No. 13, DOI 10.1175/EI228.1.
14 Potter, C. S., J.T. Randerson, C.B. Fields, P.A. Matson, P.M. Vitousek, H.A. Mooney, and S.A. Klooster (1993)
15 "Terrestrial ecosystem production: a process model based on global satellite and surface data." Global
16 Biogeochemical Cycles 7:811-841.
17 Ruddy B.C., D.L. Lorenz, and D.K. Mueller (2006) County-level estimates of nutrient inputs to the land surface of
18 the conterminous United States, 1982-2001. Scientific Investigations Report 2006-5012. U.S Department of the
19 Interior.
20 Scheer, C., S.J. Del Grosso, W.J. Parton, D.W. Rowlings, P.R. Grace (2013) Modeling Nitrous Oxide Emissions
21 from Irrigated Agriculture: Testing DAYCENT with High Frequency Measurements, Ecological Applications, in
22 press. Available online at: .
23 Soil Survey Staff (2011) State Soil Geographic (STATSGO) Database for State. Natural Resources Conservation
24 Service, United States Department of Agriculture. Available online at:
25 .
26 Towery, D. (2001) Personal Communication. Dan Towery regarding adjustments to the CTIC (1998) tillage data to
27 reflect long-term trends, Conservation Technology Information Center, West Lafayette, IN, and Marlen Eve,
28 National Resource Ecology Laboratory, Fort Collins, CO. February 2001.
29 TVA (1991 through 1992a, 1993 through 1994) Commercial Fertilizers. Tennessee Valley Authority, Muscle
30 Shoals, AL.
31 USDA-ERS (2015) Agricultural Resource Management Survey (ARMS) Farm Financial and Crop Production
32 Practices: Tailored Reports. Available online at: .
34 USDA-ERS (1997) Cropping Practices Survey Data—1995. Economic Research Service, United States Department
35 of Agriculture. Available online at: .
36 USDA-NASS (2015) Quick Stats. National Agricultural Statistics Service, United States Department of Agriculture,
37 Washington, D.C. .
38 USDA-NRCS (2015) Summary Report: 2012 National Resources Inventory, Natural Resources Conservation
39 Service, Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
40 .
41 Wisconsin Department of Natural Resources (1993) Wisconsin Greenhouse Gas Emissions: Estimates for 1990.
42 Bureau of Air Management, Wisconsin Department of Natural Resources, Madison, WI.
10-46 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Liming
2 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
3 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
4 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
5 Tepordei, V.V. (1997 through 2015) "Crushed Stone," InMinerals Yearbook. U.S. Department of the Interior/U.S.
6 Geological Survey. Washington, D.C. Available online at: .
7 Tepordei, V.V. (2003b) Personal communication. Valentin Tepordei, U.S. Geological Survey and ICF Consulting,
8 August 18, 2003.
9 Tepordei, V.V. (1996) "Crushed Stone," InMinerals Yearbook 1994. U.S. Department of the Interior/Bureau of
10 Mines, Washington, D.C. Available online at:
11 . Accessed August 2000.
12 Tepordei, V.V. (1995) "Crushed Stone," InMinerals Yearbook 1993. U.S. Department of the Interior/Bureau of
13 Mines, Washington, D.C. pp. 1107-1147.
14 Tepordei, V. V. (1994) "Crushed Stone," InMinerals Yearbook 1992. U.S. Department of the Interior/Bureau of
15 Mines, Washington, D.C. pp. 1279-1303.
16 USGS (2017) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the First Quarter of 2017, U.S.
17 Geological Survey, Reston, VA. Available online at:
18 .
19 USGS (2016) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the First Quarter of 2016, U.S.
20 Geological Survey, Reston, VA. Available online at:
21 .
22 USGS (2015) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the First Quarter of 2015, U.S.
23 Geological Survey, Reston, VA. Available online at:
24 .
25 USGS (2014) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the First Quarter of 2014, U.S.
26 Geological Survey, Reston, VA. Available online at:
27 .
28 USGS (2013) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the First Quarter of 2013, U.S.
29 Geological Survey, Reston, VA. Available online at:
30 .
31 USGS (2012) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the First Quarter of 2012, U.S.
32 Geological Survey, Reston, VA. Available online at:
33 .
34 USGS (2011) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the First Quarter of 2011, U.S.
35 Geological Survey, Reston, VA. Available online at:
36 .
37 USGS (2010) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the First Quarter of 2010, U.S.
38 Geological Survey, Reston, VA. Available online at:
39 .
40 USGS (2009) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the First Quarter of2009, U.S.
41 Geological Survey, Reston, VA. Available online at:
42 .
43 USGS (2008) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the First Quarter of2008, U.S.
44 Geological Survey, Reston, VA. Available online at:
45 .
References 10-47
-------
1 USGS (2007) Mineral Industry Surveys: Crushed Stone and Sand and Gravel in the First Quarter of2007. U.S.
2 Geological Survey, Reston, VA. Available online at:
3 .
4 West, T.O., and A.C. McBride (2005) "The contribution of agricultural lime to carbon dioxide emissions in the
5 United States: dissolution, transport, and net emissions," Agricultural Ecosystems & Environment 108:145-154.
6 West, T.O. (2008) Email correspondence. Tristram West, Environmental Sciences Division, Oak Ridge National
7 Laboratory, U.S. Department of Energy and Nikhil Nadkarni, ICF International on suitability of liming emission
8 factor for the entire United States. June 9, 2008
9 Willett, J.C. (2016) "Crashed Stone," I n \ finerals Yearbook. U.S. Department of the Interior/U.S. Geological
10 Survey. Washington, D.C. Available online at: . Accessed: 30 August 2017.
11 Willett, J.C. (2017) Personal communication. Jason Willett. Preliminary data tables from "Crashed Stone," In 2015
12 Minerals Yearbook. U.S. Department of the Interior/U.S. Geological Survey. Washington, D.C. August 31, 2017.
13 Willett, J.C. and Thompson, D.V. (2017) Crashed stone and sand and gravel in the second quarter 2015: U.S.
14 Geological Survey Mineral Industry Surveys, . Accessed: 30 August 2017.
15 Willett, J.C. (2016) "Crushed Stone," InMinerals Yearbook 2014. U.S. Department of the Interior/U.S. Geological
16 Survey, Washington, D.C. Available online at:
17 . Accessed September 2016.
18 Willett, J.C. (2015) "Crashed Stone," InMinerals Yearbook 2013. U.S. Department of the Interior/U.S. Geological
19 Survey, Washington, D.C. Available online at:
20 . Accessed September 2015.
21 Willett, J.C. (2014) "Crashed Stone," InMinerals Yearbook 2012. U.S. Department of the Interior/U.S. Geological
22 Survey, Washington, D.C. Available online at:
23 . Accessed September 2014.
24 Willett, J.C. (2013a) "Crashed Stone," InMinerals Yearbook 2011. U.S. Department of the Interior/U.S. Geological
25 Survey, Washington, D.C. Available online at:
26 . Accessed May 2013.
27 Willett, J.C. (2013b) Personal Communication. Jason Willet, U.S. Geological Survey and ICF International.
28 September 9, 2013.
29 Willett, J.C. (201 la) "Crashed Stone," InMinerals Yearbook 2009. U.S. Department of the Interior/U.S. Geological
30 Survey, Washington, D.C. Available online at:
31 . Accessed August 2011.
32 Willett, J.C. (201 lb) "Crashed Stone," InMinerals Yearbook 2010. U.S. Department of the Interior/U.S. Geological
33 Survey, Washington, D.C. Available online at:
34 . Accessed September 2012.
35 Willett, J.C. (2010) "Crashed Stone," InMinerals Yearbook 2008. U.S. Department of the Interior/U.S. Geological
36 Survey, Washington, D.C. Available online at:
37 . Accessed August 2010.
38 Willett, J.C. (2009) "Crashed Stone," InMinerals Yearbook 2007. U.S. Department of the Interior/U.S. Geological
39 Survey, Washington, D.C. Available online at:
40 . Accessed August 2009.
41 Willett, J.C. (2007a) "Crashed Stone," InMinerals Yearbook 2005. U.S. Department of the Interior/U.S. Geological
42 Survey, Washington, D.C. Available online at:
43 . Accessed August 2007.
44 Willett, J.C. (2007b) "Crashed Stone," InMinerals Yearbook 2006. U.S. Department of the Interior/U.S. Geological
45 Survey, Washington, D.C. Available online at:
46 . Accessed August 2008.
10-48 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Urea Fertilization
2 AAPFCO (2008 through 2017) Commercial Fertilizers. Association of American Plant Food Control Officials.
3 University of Missouri. Columbia, MO.
4 AAPFCO (1995 through 2000a, 2002 through 2007) Commercial Fertilizers. Association of American Plant Food
5 Control Officials. University of Kentucky. Lexington, KY.
6 AAPFCO (2000b) 1999-2000 Commercial Fertilizers Data, ASCII files. Available from David Terry, Secretary,
7 AAPFCO.
8 EPA (2000) Preliminary Data Summary: Airport Deicing Operations (Revised). EPA-821-R-00-016. August 2000.
9 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
10 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
11 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
12 Itle, C. (2009) Email correspondence. Cortney Itle, ERG and Tom Wirth, U.S. Environmental Protection Agency on
13 the amount of urea used in aircraft deicing. January 7, 2009.
14 TVA (1991 through 1994) Commercial Fertilizers. Tennessee Valley Authority, Muscle Shoals, AL.
15 TVA (1992b) Fertilizer Summary Data 1992. Tennessee Valley Authority, Muscle Shoals, AL.
16 Field Burning of Agricultural Residues
17 Barnard, G., and L. Kristoferson (1985) Agricultural Residues as Fuel in the Third World. Earthscan Energy
18 Information Programme and the Beijer Institute of the Royal Swedish Academy of Sciences. London, England.
19 Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.Cantens, G.
20 (2004 through 2005) Personal Communication. Janet Lewis, Assistant to Gaston Cantens, Vice President of
21 Corporate Relations, Florida Crystals Company and ICF International.
22 Cantens, G. (2004 through 2005) Personal Communication. Janet Lewis, Assistant to Gaston Cantens, Vice
23 President of Corporate Relations, Florida Crystals Company and ICF International.
24 Deren, C. (2002) Personal communication. Dr. Chris Deren, Everglades Research and Education Centre at the
25 University of Florida and Caren Mintz, ICF International. August 15, 2002.
26 EPA (1994) International Anthropogenic Methane Emissions: Estimates for 1990, Report to Congress. EPA 230-R-
27 93-010. Office of Policy Planning and Evaluation, U.S. Environmental Protection Agency, Washington, D.C.
28 Gonzalez, R. (2007 through 2014) Email correspondence. Rene Gonzalez, Plant Manager, Sem-Chi Rice Company
29 and ICF International.
30 Huang, Y., W. Zhang, W. Sun, and X. Zheng (2007) "Net Primary Production of Chinese Croplands from 1950 to
31 1999." Ecological Applications, 17(3):692-701.
32 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
33 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
34 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
35 IPCC/UNEP/OECD/IEA (1997) Revised 1996 IPCC Guidelines for National Greenhouse Gas Inventories.
36 Intergovernmental Panel on Climate Change, United Nations Environment Programme, Organization for Economic
37 Co-Operation and Development, International Energy Agency, Paris, France.
38 Kinoshita, C.M. (1988) "Composition and processing of burned and unburned cane in Hawaii." Intl. Sugar Jnl.
39 90:1070,34-37.
40 Kirstein, A. (2003 through 2004) Personal Communication. Arthur Kirstein, Coordinator, Agricultural Economic
41 Development Program, Palm Beach County Cooperative Extension Service, Florida and ICF International.
42 Lachnicht, S.L., P.F. Hendrix, R.L. Potter, D.C. Coleman, and D.A. Crossley Jr. (2004) "Winter decomposition of
43 transgenic cotton residue in conventional-till and no-till systems." Applied Soil Ecology, 27:135-142.
References 10-49
-------
1 McCarty, J.L. (2011) "Remote Sensing-Based Estimates of Annual and Seasonal Emissions from Crop Residue
2 Burning in the Contiguous United States." Journal of the Air & Waste Management Association, 61:1, 22-34, DOI:
3 10.3155/1047-3289.61.1.22.
4 McCarty, J.L. (2010) Agricultural Residue Burning in the Contiguous United States by Crop Type and State.
5 Geographic Information Systems (GIS) Data provided to the EPA Climate Change Division by George Pouliot,
6 Atmospheric Modeling and Analysis Division, EPA. Dr. McCarty's research was supported by the NRI Air Quality
7 Program of the Cooperative State Research, Education, and Extension Service, USD A, under Agreement No.
8 20063511216669 and the NASA Earth System Science Fellowship.
9 McCarty, J.L. (2009) Seasonal and Interannual Variability of Emissions from Crop Residue Burning in the
10 Contiguous United States. Dissertation. University of Maryland, College Park.
11 Murphy, W.J. (1993) "Tables for weights and measurement: crops." Extension publications. (University of Missouri
12 Extension). Available online at: .
13 Schueneman, T. (1999 through 2001) Personal Communication. Tom Schueneman, Agricultural Extension Agent,
14 Palm Beach County, FL and ICF International. July 30, 2001.Schueneman, T.J. and C.W. Deren (2002) "An
15 Overview of the Florida Rice Industry." SS-AGR-77, Agronomy Department, Florida Cooperative Extension
16 Service, Institute of Food and Agricultural Sciences, University of Florida. Revised November 2002.
17 Strehler, A., and W. Stiitzle (1987) "Biomass Residues." In Hall, D.O. and Overend, R.P. (eds.). Biomass. John
18 Wiley and Sons, Ltd. Chichester, UK.
19 Turn, S.Q., B.M. Jenkins, J.C. Chow, L.C. Pritchett, D. Campbell, T. Cahill, and S.A. Whalen (1997) "Elemental
20 characterization of particulate matter emitted from biomass burning: Wind tunnel derived source profiles for
21 herbaceous and wood fuels." Journal of Geophysical Research 102(D3):3683-3699.
22 USDA-NRCS (2015) Summary Report: 2012 National Resources Inventory, Natural Resources Conservation
23 Service, Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
24 .
25 USDA (2016) Quick Stats: U.S. & All States Data; Crops; Production and Area Harvested; 1990 - 2015. National
26 Agricultural Statistics Service, U.S. Department of Agriculture. Washington, D.C. U.S. Department of Agriculture,
27 National Agricultural Statistics Service. Washington, D.C., Available online at: .
28 Land Use, Land-Use Change, and Forestry
29 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
30 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
31 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
32 UNFCCC (2013) Report of the Conference of the Parties on its nineteenth session, held in Warsaw from 11 to 23
33 November 2013. Available online at: .
34 Representation of the U.S. Land Base
35 Alaska Department of Natural Resources (2006) Alaska Infrastructure 1:63,360. Available online at:
36 .
37 Alaska Interagency Fire Management Council (1998) Alaska Interagency Wildland Fire Management Plan.
38 Available online at: .
39 Alaska Oil and Gas Conservation Commission (2009) Oil and Gas Information System. Available online at:
40 .
41 EIA(2011) Coal Production and Preparation Report Shapefile. Available online at: .
10-50 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 ESRI (2008) ESRI Data & Maps. Redlands, CA: Environmental Systems Research Institute. [CD-ROM],
2 Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and J. Wickham. (2011)
3 Completion of the 2006 National Land Cover Database for the Conterminous United States, PE&RS, Vol.
4 77(9):858-864.
5 Homer, C., J. Dewitz, J. Fry, M. Coan, N. Hossain, C. Larson, N. Herold, A. McKerrow, J.N. VanDriel and J.
6 Wickham. (2007) Completion of the 2001 National Land Cover Database for the Conterminous United States,
7 Photogrammetric Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
8 Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D.,
9 and Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-
10 Representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, v. 81,
11 no. 5, p. 345-354.
12 IPCC (2010) Revisiting the use of managed land as a proxy for estimating national anthropogenic emissions and
13 removals. [Eggleston HS, Srivastava N, Tanabe K, Baasansuren J, (eds.)]. Institute for Global Environmental
14 Studies, Intergovernmental Panel on Climate Change, Hayama, Kanagawa, Japan.
15 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
16 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
17 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
18 Jin, S., L. Yang, P. Danielson, C. Homer, J. Fry, and G. Xian. (2013) A comprehensive change detection method for
19 updating the National Land Cover Database to circa 2011. Remote Sensing of Environment, 132: 159-175.
20 NOAA Coastal Change Analysis Program (C-CAP) Regional Land Cover Database. Data collected 1995-present
21 Charleston, SC: National Oceanic and Atmospheric Administration (NOAA) Coastal Services Center. Data accessed
22 at: .
23 Nusser, S.M. and J.J. Goebel (1997) "The national resources inventory: a long-term multi-resource monitoring
24 programme." Environmental and Ecological Statistics 4:181-204.
25 Smith, W.B., P.D. Miles, C.H. Perry, and S.A. Pugh (2009) Forest Resources of the United States, 2007. Gen. Tech.
26 Rep. WO-78. U.S. Department of Agriculture Forest Service, Washington, D.C.
27 U.S. Census Bureau (2010) Topologically Integrated Geographic Encoding and Referencing (TIGER) system
28 shapefiles. U.S. Census Bureau, Washington, D.C. Available online at: .
29 U.S. Department of Agriculture (2014) County Data - Livestock, 1990-2014. U.S. Department of Agriculture,
30 National Agriculture Statistics Service, Washington, D.C.
31 U.S. Department of Interior (2005) Federal Lands of the United States. National Atlas of the United States, U.S.
32 Department of the Interior, Washington D.C. Available online at:
33 .
34 United States Geological Survey (USGS), Gap Analysis Program (2012) Protected Areas Database of the United
35 States (PADUS), version 1.3 Combined Feature Class. November 2012.
36 USGS (2012) Alaska Resource Data File. Available online at: .
37 USGS (2005) Active Mines and Mineral Processing Plants in the United States in 2003. U.S. Geological Survey,
38 Reston, VA.
39 Forest Land Remaining Forest Land: Changes in Forest Carbon
40 Stocks
41 AF&PA (2006a and earlier) Statistical roundup. (Monthly). Washington, D.C. American Forest & Paper
42 Association.
43 AF&PA (2006b and earlier) Statistics of paper, paperboard and wood pulp. Washington, D.C. American Forest &
44 Paper Association.
References 10-51
-------
1 Amichev, B.Y. and J.M. Galbraith (2004) "A Revised Methodology for Estimation of Forest Soil Carbon from
2 Spatial Soils and Forest Inventory Data Sets." Environmental Management 33(Suppl. 1):S74-S86.
3 Bechtold, W.A.; Patterson, P.L. (2005) The enhanced forest inventory and analysis program—national sampling
4 design and estimation procedures. Gen. Tech. Rep. SRS-80. Asheville, NC: U.S. Department of Agriculture Forest
5 Service, Southern Research Station. 85 p.
6 Birdsey, R. (1996) "Carbon Storage for Major Forest Types and Regions in the Conterminous United States." In
7 R.N. Sampson and D. Hair, (eds.). Forest and Global Change, Volume 2: Forest Management Opportunities for
8 Mitigating Carbon Emissions. American Forests. Washington, D.C., 1-26 and 261-379 (appendices 262 and 263).
9 Coulston, J.W., Wear, D.N., and Vose, J.M. (2015) Complex forest dynamics indicate potential for slowing carbon
10 accumulation in the southeastern United States. Scientific Reports. 5: 8002.
11 Domke, G.M., J.E. Smith, and C.W. Woodall. (2011) Accounting for density reduction and structural loss in
12 standing dead trees: Implications for forest biomass and carbon stock estimates in the United States. Carbon
13 Balance and Management. 6:14.
14 Domke, G.M., Woodall, C.W., Smith, J.E., Westfall, J.A., McRoberts, R.E. (2012) Consequences of alternative tree-
15 level biomass estimation procedures on U.S. forest carbon stock estimates. Forest Ecology and Management. 270:
16 108-116.
17 Domke. G.M., Perry, C.H., Walters. B.F., Woodall. C.W., and Smith, J.E. (2016) A framework for estimating litter
18 carbon stocks in forests of the U nitcd States. Science of the Total Environment 557-558: 469-478.
19 Domke, G.M., Woodall, C.W., Walters, B.F., Smith, J.E. (2013) From models to measurements: comparing down
20 dead wood carbon stock estimates in the U.S. forest inventory. PLoS ONE 8(3): e59949.
21 Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., Nave, L., Swanston, C. (2017) Toward inventory-based
22 estimates of soil organic carbon in forests of the United States. Ecological Applications. 27(4), 1223-1235.EPA
23 (2006) Municipal solid waste in the United States: 2005 Facts and figures. Office of Solid Waste, U.S.
24 Environmental Protection Agency. Washington, D.C. (5306P) EPA530-R-06-011. Available online at:
2 5 .
26 Frayer, W.E., and G.M. Furnival (1999) "Forest Survey Sampling Designs: A History." Journal of Forestry 97(12):
27 4-10.
28 Freed, R. (2004) Open-dump and Landfill timeline spreadsheet (unpublished). ICF International. Washington, D.C.
29 Hair, D. (1958) "Historical forestry statistics of the United States." Statistical Bull. 228. U.S. Department of
30 Agriculture Forest Service, Washington, D.C.
31 Hair. D. and A.H. Ulrich (1963) The Demand and price situation for forest products - 1963. U.S. Department of
32 Agriculture Forest Service, Misc Publication No. 953. Washington, D.C.
33 Harmon, M.E., C.W. Woodall, B. Fasth, J. Sexton, M. Yatkov. (2011) Differences between standing and downed
34 dead tree wood density reduction factors: A comparison across decay classes and tree species. Res. Paper. NRS-15.
35 Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 40 p.
36 Howard, J. L. and Jones, K.C. In preparation. U.S. timber production, trade, consumption, and price statistics 1965
37 to 2015. Res. Pap. FPL-RP-XXX. Madison, WI: USD A, Forest Service, Forest Products Laboratory.
38 Howard, J. L. and Jones, K.C. 2016. U.S. timber production, trade, consumption, and price statistics 1965 to 2013.
39 Res. Pap. FPL-RP-679. Madison, WI: USD A, Forest Service, Forest Products Laboratory.
40 Howard, J. L. (2007) U.S. timber production, trade, consumption, and price statistics 1965 to 2005. Res. Pap. FPL-
41 RP-637. Madison, WI: USD A, Forest Service, Forest Products Laboratory.
42 Howard, J. L. (2003) U.S. timber production, trade, consumption, and price statistics 1965 to 2002. Res. Pap. FPL-
43 RP-615. Madison, WI: USD A, Forest Service, Forest Products Laboratory. Available online at:
44 .
10-52 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 IPCC (2014) 2013 Supplement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories:
2 Wetlands. [Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M., and Troxler, T.G. (eds.)].
3 Switzerland.
4 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
5 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
6 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
7 IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth
8 Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M. Manning, Z. Chen,
9 M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.)]. Cambridge University Press, Cambridge, United
10 Kingdom and New York, NY, USA, 996 pp.
11 Jenkins, J.C., D.C. Chojnacky, L.S. Heath, and R.A. Birdsey (2003) "National-scale biomass estimators for United
12 States tree species." Forest Science 49(1): 12-35.
13 Jandl, R., Rodeghiero, M., Martinez, C., Cotrufo, M. F., Bampa, F., van Wesemael, B., Harrison, R.B., Guerrini,
14 I.A., deB Richter Jr., D., Rustad, L., Lorenz, K., Chabbi, A., Miglietta, F. (2014) Current status, uncertainty and
15 future needs in soil organic carbon monitoring. Science of the Total Environment, 468, 376-383.
16 Johnson, K. Domke, G.M., Russell, M.B., Walters, B.F., Horn, J., Peduzzi, A., Birdsey, R., Dolan, K., Huang, W.
17 2017. Estimating aboveground live understory vegetation carbon in the United States. Environmental Research
18 Letters.
19 Ogle, S.M., Woodall, C.W., Swan, A., Smith, J.E., Wirth. T. In preparation. Determining the Managed Land Base
20 for Delineating Carbon Sources and Sinks in the United States. Environmental Science and Policy.
21 O'Neill, K.P., Amacher, M.C., Perry, C.H. (2005) Soils as an indicator of forest health: a guide to the collection,
22 analysis, and interpretation of soil indicator data in the Forest Inventory and Analysis program. Gen. Tech. Rep. NC-
23 258. St. Paul, MN: US Department of Agriculture, Forest Service, North Central Research Station. 53 p.
24 Oswalt, S.N.; Smith. W.B.; Miles, P.D.; Pugh, S.A. (2014) Forest Resources of the United States, 2012. Gen. Tech.
25 Rep. WO-91. Washington, D.C. U.S. Department of Agriculture, Forest Service, Washington Office. 218 p.
26 Perry, C.H., C.W. Woodall, and M. Schoeneberger (2005) Inventorying trees in agricultural landscapes: towards an
27 accounting of "working trees". In: "Moving Agroforestry into the Mainstream." Proc. 9th N. Am. Agroforestry
28 Conf, Brooks, K.N. and Folliott, P.F. (eds.). 12-15 June 2005, Rochester, MN [CD-ROM], Dept. of Forest
29 Resources, Univ. Minnesota, St. Paul, MN, 12 p. Available online at: . (verified
30 23 Sept 2006).
31 Russell, M.B.; D'Amato, A.W.; Schulz, B.K.; Woodall, C.W.; Domke, G.M.; Bradford, J.B. (2014) Quantifying
32 understory vegetation in the U.S. Lake States: a proposed framework to inform regional forest carbon stocks.
33 Forestry. 87: 629-638.
34 Russell, M.B.; Domke, G.M.; Woodall, C.W.; D'Amato, A.W. (2015) Comparisons of allometric and climate-
35 derived estimates of tree coarse root carbon in forests of the United States. Carbon Balance and Management. 10:
36 20.
37 Skog, K.E. (2008) Sequestration of carbon in harvested wood products for the United States. Forest Products
38 Journal 58:56-72.
39 Smith, J.E.; Heath, L.S.; Skog, K.E.; Birdsey, R.A. (2006) Methods for calculating forest ecosystem and harvested
40 carbon with standard estimates for forest types of the United States. Gen. Tech. Rep. NE-343. Newtown Square,
41 PA: U.S. Department of Agriculture, Forest Service, Northeastern Research Station. 216 p.
42 Smith, W. B., P. D. Miles, C. H. Perry, and S. A. Pugh (2009) Forest Resources of the United States, 2007. General
43 Technical Report WO-78, U.S. Department of Agriculture Forest Service, Washington Office.
44 Smith, J.E., L.S. Heath, and M.C. Nichols (2010) U.S. Forest Carbon Calculation Tool User's Guide: Forestland
45 Carbon Stocks and Net Annual Stock Change. General Technical Report NRS-13 revised, U.S. Department of
46 Agriculture Forest Service, Northern Research Station, 34 p.
47 Steer, Henry B. (1948) Lumber production in the United States. Misc. Pub. 669, U.S. Department of Agriculture
48 Forest Service. Washington, D.C.
References 10-53
-------
1 Ulrich, Alice (1985) U.S. Timber Production, Trade, Consumption, and Price Statistics 1950-1985. Misc. Pub.
2 1453, U.S. Department of Agriculture Forest Service. Washington, D.C.
3 Ulrich, A.H. (1989) U.S. Timber Production, Trade, Consumption, and Price Statistics, 1950-1987. USDA
4 Miscellaneous Publication No. 1471, U.S. Department of Agriculture Forest Service. Washington, D.C., 77.
5 United Nations Framework Convention on Climate Change (2013) Report on the individual review of the inventory
6 submission of the United States of America submitted in 2012. FCCC/ARR/2012/USA. 42 p.
7 USDA Forest Service (2016a) Forest Inventory and Analysis National Program: Program Features. U.S. Department
8 of Agriculture Forest Service. Washington, D.C. Available online at: .
9 Accessed 26 September 2017.
10 USDA Forest Service. (2016b) Forest Inventory and Analysis National Program: FIA Data Mart. U.S. Department
11 of Agriculture Forest Service. Washington, D.C. Available online at: . Accessed on 26 September 2017.
13 USDA Forest Service. (2016c) Forest Inventory and Analysis National Program, FIA library: Field Guides, Methods
14 and Procedures. U.S. Department of Agriculture Forest Service. Washington, D.C. Available online at:
15 . Accessed on 26 September 2017.
16 USDA Forest Service (2016d) Forest Inventory and Analysis National Program, FIA library: Database
17 Documentation. U.S. Department of Agriculture, Forest Service, Washington Office. Available online at:
18 . Accessed on 26 September 2017.
19 U.S. Census Bureau (1976) Historical Statistics of the United States, Colonial Times to 1970, Vol. 1. Washington,
20 D.C.
21 Wear, D.N., Coulston, J.W. (2015) From sink to source: Regional variation in U.S. forest carbon futures. Scientific
22 Reports. 5: 16518.
23 Westfall, J.A., Woodall, C.W., Hatfield, M.A. (2013) A statistical power analysis of woody carbon flux from forest
24 inventory data. Climatic Change. 118: 919-931.
25 Woodall, C.W., Coulston, J.W., Domke, G.M., Walters, B.F., Wear, D.N., Smith, J.E., Anderson, H.-E., Clough,
26 B.J., Cohen, W.B., Griffith, D.M., Hagan, S.C., Hanou, I.S.; Nichols, M.C., Perry, C.H., Russell, M.B., Westfall,
27 J.A., Wilson, B.T. (2015a) The US Forest Carbon Accounting Framework: Stocks and Stock change 1990-2016.
28 Gen. Tech. Rep. NRS-154. Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern
29 Research Station. 49 pp.
30 Woodall, C.W., Amacher, M.C., Bechtold, W.A., Coulston, J.W., Jovan, S., Perry, C.H., Randolph, K.C., Schulz,
31 B.K., Smith, G.C., Tkacz, B., Will-Wolf, S. (201 lb) "Status and future of the forest health indicators program of the
32 United States." Environmental Monitoring and Assessment. 177: 419-436.
33 Woodall, C.W., L.S. Heath, G.M. Domke, and M.C. Nichols (201 la) Methods and equations for estimating
34 aboveground volume, biomass, and carbon for trees in the U.S. forest inventory, 2010. Gen. Tech. Rep. NRS-88.
35 Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 30 p.
36 Woodall, C. W., and V.J. Monleon (2008) Sampling protocol, estimation, and analysis procedures for the down
37 woody materials indicator of the FIA program. Gen. Tech. Rep. NRS-22. Newtown Square, PA: U.S. Department of
38 Agriculture, Forest Service, Northern Research Station. 68 p.
39 Woodall, C.W., Walters, B.F., Oswalt, S.N., Domke, G.M., Toney, C., Gray, A.N. (2013) Biomass and carbon
40 attributes of downed woody materials in forests of the United States. Forest Ecology and Management 305: 48-59.
41 Woodall, C.W., Walters, B.F., Coulston, J.W., D'Amato, A.W., Domke, G.M., Russell, M.B., Sowers, P.A. (2015b)
42 Monitoring network confirms land use change is a substantial component of the forest carbon sink in the eastern
43 United States. Scientific Reports. 5: 17028.
44 Zhu, Zhiliang, and McGuire, A.D., eds., (2016) Baseline and projected future carbon storage and greenhouse-gas
45 fluxes in ecosystems of Alaska: U.S. Geological Survey Professional Paper 1826, 196 p., Available online at:
46 .
10-54 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Forest Land Remaining Forest Land: Non-C02 Emissions from
2 Forest Fires
3 deVries, R.E. (1987) A Preliminary Investigation of the Growth and Longevity of Trees in Central Park. M.S. thesis,
4 Rutgers University, New Brunswick, NJ.
5 Dwyer, J.F., D.J. Nowak, M.H. Noble, and S.M. Sisinni (2000) Connecting People with Ecosystems in the 21st
6 Century: An Assessment of Our Nation's Urban Forests. General Technical Report PNW-GTR-490, U.S.
7 Department of Agriculture, Forest Service, Pacific Northwest Research Station, Portland, OR.
8 Fleming, L.E. (1988) Growth Estimation of Street Trees in Central New Jersey. M.S. thesis, Rutgers University,
9 New Brunswick, NJ.
10 Frelich, L.E. (1992) Predicting Dimensional Relationships for Twin Cities Shade Trees. University of Minnesota,
11 Department of Forest Resources, St. Paul, MN, p. 33.
12 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
13 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
14 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
15 Nowak, D.J. (2011) Phone conference regarding Changes in Carbon Stocks in Urban Trees estimation methodology.
16 David Nowak, USD A, Jennifer Jenkins, EPA, and Mark Flugge and Nikhil Nadkarni, ICF International. January 4,
17 2011.
18 Nowak, D.J. (2009) E-mail correspondence regarding new data for Chicago's urban forest. David Nowak, USDA
19 Forest Service to Nikhil Nadkarni, ICF International. October 7, 2009.
20 Nowak, D.J. (2007a) "New York City's Urban Forest." USDA Forest Service. Newtown Square, PA, February 2007.
21 Nowak, D.J. (2007b) E-mail correspondence regarding revised sequestration values and standard errors for
22 sequestration values. David Nowak, USDA Forest Service to Susan Asam, ICF International. October 31, 2007.
23 Nowak, D.J. (1994) "Atmospheric Carbon Dioxide Reduction by Chicago's Urban Forest." In: Chicago's Urban
24 Forest Ecosystem: Results of the Chicago Urban Forest Climate Project. E.G. McPherson, D.J. Nowak, and R. A.
25 Rowntree (eds.). General Technical Report NE-186. U.S. Department of Agriculture Forest Service, Radnor, PA. pp.
26 83-94.
27 Nowak, D.J. (1986) "Silvics of an Urban Tree Species: Norway Maple (Acer platanoides L.)." M.S. thesis, College
28 of Environmental Science and Forestry, State University of New York, Syracuse, NY.
29 Nowak, D.J., Buckelew-Cumming, A., Twardus, D., Hoehn, R., and Mielke, M. (2007) National Forest Health
30 Monitoring Program, Monitoring Urban Forests in Indiana: Pilot Study 2002, Part 2: Statewide Estimates Using the
31 UFORE Model. Northeastern Area Report. NA-FR-01e07, p. 13.
32 Nowak, D.J. and D.E. Crane (2002) "Carbon Storage and Sequestration by Urban Trees in the United States."
33 Environmental Pollution 116(3):381—389.
34 Nowak, D.J., D.E. Crane, J.C. Stevens, and M. Ibarra (2002) Brooklyn's Urban Forest. General Technical Report
35 NE-290. U.S. Department of Agriculture Forest Service, Newtown Square, PA.
36 Nowak, D.J., and E.J. Greenfield (2012) Tree and impervious cover in the United States. Journal of Landscape and
37 Urban Planning (107) pp. 21-30.
38 Nowak, D.J., E.J. Greenfield, R.E. Hoehn, and E. Lapoint (2013) Carbon Storage and Sequestration by Trees in
39 Urban and Community Areas of the United States. Environmental Pollution 178: 229-236. March 12, 2013.
40 Nowak, D.J., J.T. Walton, L.G. Kaya, and J.F. Dwyer (2005) "The Increasing Influence of Urban Environments on
41 U.S. Forest Management." Journal of Forestry 103(8):377-382.
References 10-55
-------
1 Ruefenacht, B., M.V. Finco, M.D. Nelson, R. Czaplewski, E.H. Helmer, J.A. Blackard, G.R. Holden, A.J. Lister, D.
2 Salajanu, D. Weyermann, K. Winterberger (2008) Conterminous U.S. and Alaska Forest Type Mapping Using
3 Forest Inventory and Analysis. USD A Forest Service - Forest Inventory and Analysis Program & Remote Sensing
4 Applications Center. Available online at: . Accessed 8
5 September 2015.
6 Smith, W.B. and S.R. Shifley (1984) Diameter Growth, Survival, and Volume Estimates for Trees in Indiana and
7 Illinois. Research Paper NC-257. North Central Forest Experiment Station, U.S. Department of Agriculture Forest
8 Service, St. Paul, MN.
9 U.S. Census Bureau (2012) "A national 2010 urban area file containing a list of all urbanized areas and urban
10 clusters (including Puerto Rico and the Island Areas) sorted by UACE code." U.S. Census Bureau, Geography
11 Division.
12 Veraverbeke, S., B.M. Rogers, and J.T. Randerson. (2015) Daily burned area and carbon emissions from boreal fires
13 in Alaska. Biogeosciences, 12:3579-3601.
14 Forest Land Remaining Forest Land: N20 Emissions from Soils
15 Albaugh, T.J., Allen, H.L., Fox, T.R. (2007) Historical Patterns of Forest Fertilization in the Southeastern United
16 States from 1969 to 2004. Southern Journal of Applied Forestry, 31, 129-137(9).
17 Binkley, D. (2004) Email correspondence regarding the 95 percent confidence interval for area estimates of southern
18 pine plantations receiving N fertilizer (±20%) and the rate applied for areas receiving N fertilizer (100 to 200
19 pounds/acre). Dan Binkley, Department of Forest, Rangeland, and Watershed Stewardship, Colorado State
20 University and Stephen Del Grosso, Natural Resource Ecology Laboratory, Colorado State University. September
21 19,2004.
22 Binkley, D., R. Carter, and H.L. Allen (1995) Nitrogen Fertilization Practices in Forestry. In: Nitrogen Fertilization
23 in the Environment, P.E. Bacon (ed.), Marcel Decker, Inc., New York.
24 Briggs, D. (2007) Management Practices on Pacific Northwest West-Side Industrial Forest Lands, 1991-2005: With
25 Projections to 2010. Stand Management Cooperative, SMC Working Paper Number 6, College of Forest Resources,
26 University of Washington, Seattle.
27 Fox, T.R., H. L. Allen, T.J. Albaugh, R. Rubilar, and C.A. Carlson (2007) Tree Nutrition and Forest Fertilization of
28 Pine Plantations in the Southern United States. Southern Journal of Applied Forestry, 31, 5-11.
29 IPCC (2006) 20061PCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
30 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
31 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
32 USD A Forest Service (2001) U.S. Forest Facts and Historical Trends. FS-696. U.S. Department of Agriculture
33 Forest Service, Washington, D.C. Available online at: .
34 Forest Land Remaining Forest Land: Drained Organic Soils
35 IPCC (2014) 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands,
36 Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds.). Published:
37 IPCC, Switzerland.
38 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
39 Inventories Programme, The Intergovernmental Panel on Climate Change, H.S. Eggleston, L. Buendia, K. Miwa, T.
40 Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
41 STATSG02 (2016) Soil Survey Staff, Natural Resources Conservation Service, United States Department of
42 Agriculture. U.S. General Soil Map (STATSG02). Available online at .
43 Accessed 10 November 2016.
10-56 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 USDA Forest Service (2016) Forest Inventory and Analysis National Program: FIA Data Mart. U.S. Department of
2 Agriculture Forest Service. Washington, DC; 2015. Available online at . Accessed 15 August 2016.
4 Land Converted to Forest Land
5 Birdsey, R., Pregitzer, K., Lucier, A. (2006) Forest carbon management in the United States: 1600-2100. Journal of
6 Environmental Quality, 35: 1461-1469.
7 Brockwell, Peter J., and Richard A. Davis. Introduction to time series and forecasting. Springer, 2016.
8 Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., Nave, L., Swanston, C. (2017) Toward inventory-based
9 estimates of soil organic carbon in forests of the United States. Ecological Applications. 27(4), 1223-1235.Homer,
10 C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N, McKerrow, A., VanDriel, J.N., and Wickham,
11 J. (2007) Completion of the 2001 National Land Cover Database for the Conterminous United States.
12 Photogrammetric Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
13 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
14 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa,
15 T. Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
16 Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and management
17 impacts on soil organic carbon storage for U.S. agroecosystems between 1982 and 1997." Global Change Biology
18 9:1521-1542.
19 Ogle, S.M., F.J. Breidt, and K. Paustian. (2006) "Bias and variance in model results due to spatial scaling of
20 measurements for parameterization in regional assessments." Global Change Biology 12:516-523.
21 USDA-NRCS (1997) "National Soil Survey Laboratory Characterization Data," Digital Data, Natural Resources
22 Conservation Service, U.S. Department of Agriculture. Lincoln, NE.
23 USDA-NRCS (2013) Summary Report: 2010 National Resources Inventory, Natural Resources Conservation
24 Service, Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
25 Available online at: .
26 USDA-NRCS (2009) Summary Report: 2007 National Resources Inventory, Natural Resources Conservation
27 Service, Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
28 Available online at: .
29 Woodall, C.W., Coulston, J.W., Domke, G.M., Walters, B.F., Wear, D.N., Smith, J.E., Anderson, H.-E., Clough,
30 B.J., Cohen, W.B., Griffith, D.M., Hagan, S.C., Hanou, I.S.; Nichols, M.C., Perry, C.H., Russell, M.B., Westfall,
31 J.A., Wilson, B.T. (2015a) The US Forest Carbon Accounting Framework: Stocks and Stock change 1990-2016.
32 Gen. Tech. Rep. NRS-154. Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern
33 Research Station. 49 pp.
34 Cropland Remaining Cropland: Mineral and Organic Soil
35 Carbon Stock Changes
36 Armentano, T. V., and E.S. Menges (1986). Patterns of change in the carbon balance of organic soil-wetlands of the
37 temperate zone. Journal of Ecology 74: 755-774.
38 Brady, N.C. and R.R. Weil (1999) The Nature and Properties of Soils. Prentice Hall. Upper Saddle River, NJ, 881.
39 Conant, R. T., K. Paustian, and E.T. Elliott (2001). "Grassland management and conversion into grassland: effects
40 on soil carbon." Ecological Applications 11: 343-355.
41 Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
42 CTIC (2004) National Crop Residue Management Survey: 1989-2004. Conservation Technology Information
43 Center, Purdue University, Available online at: .
References 10-57
-------
1 Daly, C., R.P. Neilson, and D.L. Phillips (1994) "A Statistical-Topographic Model for Mapping Climatological
2 Precipitation Over Mountainous Terrain." Journal of Applied Meteorology 33:140-158.
3 Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel (2001)
4 "Simulated Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the DAYCENT Model." In
5 Modeling Carbon and Nitrogen Dynamics for Soil Management, Schaffer, M., L. Ma, S. Hansen, (eds.). CRC Press,
6 Boca Raton, Florida, pp. 303-332.
7 Del Grosso, S.J., S.M. Ogle, W.J. Parton (2011) Soil organic matter cycling and greenhouse gas accounting
8 methodologies, Chapter 1, pp 3-13 DOI: 10.1021/bk-2011-1072.ch001. In: Understanding Greenhouse Gas
9 Emissions from Agricultural Management, L. Guo, A. Gunasekara, L. McConnell (eds.). American Chemical
10 Society, Washington, D.C.
11 Edmonds, L., R. L. Kellogg, B. Kintzer, L. Knight, C. Lander, J. Lemunyon, D. Meyer, D.C. Moffitt, and J.
12 Schaefer (2003) "Costs associated with development and implementation of Comprehensive Nutrient Management
13 Plans." Part I—Nutrient management, land treatment, manure and wastewater handling and storage, and
14 recordkeeping. Natural Resources Conservation Service, U.S. Department of Agriculture. Available online at:
15 .
16 Euliss, N., and R. Gleason (2002) Personal communication regarding wetland restoration factor estimates and
17 restoration activity data. Ned Euliss and Robert Gleason of the U.S. Geological Survey, Jamestown, ND, to Stephen
18 Ogle of the National Resource Ecology Laboratory, Fort Collins, CO. August 2002.
19 Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and Wickham, J. (2011) Completion of
20 the 2006 National Land Cover Database for the Conterminous United States, PE&RS, Vol. 77(9):858-864.
21 Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A., VanDriel, J.N., and Wickham,
22 J. (2007) Completion of the 2001 National Land Cover Database for the Conterminous United States. Photogrammetric
23 Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
24 Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
25 Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-Representing
26 a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, v. 81, no. 5, p. 345-354.
27 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
28 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
29 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
30 IPCC (2003) Good Practice Guidance for Land Use, Land-Use Change, and Forestry. The Intergovernmental Panel
31 on Climate Change, National Greenhouse Gas Inventories Programme, J. Penman, et al., eds. August 13, 2004.
32 Available online at: .
33 McGill, W.B., and C.V. Cole (1981) Comparative aspects of cycling of organic C, N, S and P through soil organic
34 matter. Geoderma 26:261-22,6.
35 Metherell, A.K., L.A. Harding, C.V. Cole, and W.J. Parton (1993) "CENTURY Soil Organic Matter Model
36 Environment." Agroecosystem version 4.0. Technical documentation, GPSR Tech. Report No. 4, USDA/ARS, Ft.
37 Collins, CO.
38 Mesinger, F., G. DiMego, E. Kalnay, K. Mitchell, P. C. Shafran, W. Ebisuzaki, D. Jovic, J. Woollen, E. Rogers, E.
39 H. Berbery, M. B. Ek, Y. Fan, R. Grumbine, W. Higgins, H. Li, Y. Lin, G. Manikin, D. Parrish, and W. Shi (2006)
40 North American regional reanalysis. Bulletin of the American Meteorological Society 87:343-360.
41 NASS (2004) Agricultural Chemical Usage: 2003 Field Crops Summary. Report AgChl(04)a. National Agricultural
42 Statistics Service, U.S. Department of Agriculture, Washington, D.C. Available online at:
43 .
44 NASS (1999) Agricultural Chemical Usage: 1998 Field Crops Summary. Report AgCHl(99). National Agricultural
45 Statistics Service, U.S. Department of Agriculture, Washington, DC. Available online at:
46 .
10-58 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 NASS (1992) Agricultural Chemical Usage: 1991 Field Crops Summary. Report AgChl(92). National Agricultural
2 Statistics Service, U.S. Department of Agriculture, Washington, D.C. Available online at:
3 .
4 NRCS (1999) Soil Taxonomy: A basic system of soil classification for making and interpreting soil surveys, 2nd
5 Edition. Agricultural Handbook Number 436, Natural Resources Conservation Service, U.S. Department of
6 Agriculture, Washington, D.C.
7 NRCS (1997) "National Soil Survey Laboratory Characterization Data," Digital Data, Natural Resources
8 Conservation Service, U.S. Department of Agriculture. Lincoln, NE.
9 NRCS (1981) Land Resource Regions and Major Land Resource Areas of the United States, USDA Agriculture
10 Handbook 296, United States Department of Agriculture, Natural Resources Conservation Service, National Soil
11 Survey Cente., Lincoln, NE, pp. 156.
12 Ogle, S.M., F.J. Breidt, M. Easter, S. Williams, K. Killian, and K. Paustian (2010) "Scale and uncertainty in
13 modeled soil organic carbon stock changes for U.S. croplands using a process-based model." Global Change
14 Biology 16:810-820.
15 Ogle, S.M., F.J. Breidt, M. Easter, S. Williams and K. Paustian (2007) "Empirically-Based Uncertainty Associated
16 with Modeling Carbon Sequestration Rates in Soils." Ecological Modeling 205:453-463.
17 Ogle, S.M., F.J. Breidt, and K. Paustian (2006) "Bias and variance in model results due to spatial scaling of
18 measurements for parameterization in regional assessments." Global Change Biology 12:516-523.
19 Ogle, S. M., et al. (2005) "Agricultural management impacts on soil organic carbon storage under moist and dry
20 climatic conditions of temperate and tropical regions." Biogeochemistry 72: 87-121.
21 Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and management
22 impacts on soil organic carbon storage for U.S. agroecosystems between 1982 and 1997." Global Change Biology
23 9:1521-1542.
24 Ogle, S., M. Eve, M. Sperrow, F.J. Breidt, and K. Paustian (2002) Agricultural Soil C Emissions, 1990-2001:
25 Documentation to Accompany EPA Inventory Tables. Natural Resources Ecology Laboratory, Fort Collins, CO.
26 Provided in an e-mail from Stephen Ogle, NREL to Barbara Braatz, ICF International. September 23, 2002.
27 Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface Submodel:
28 Description and Testing". Glob. Planet. Chang. 19: 35-48.
29 Parton, W.J., D.S. Ojima, C.V. Cole, and D.S. Schimel (1994) "A General Model for Soil Organic Matter
30 Dynamics: Sensitivity to litter chemistry, texture and management," in Quantitative Modeling of Soil Forming
31 Processes. Special Publication 39, Soil Science Society of America, Madison, WI, 147-167.
32 Parton, W.J., D.S. Schimel, C.V. Cole, D.S. Ojima (1987) "Analysis of factors controlling soil organic matter levels
33 in Great Plains grasslands." Soil Science Society of America Journal 51:1173-1179.
34 Parton, W.J., J.W.B. Stewart, C.V. Cole. (1988) "Dynamics of C, N, P, and S in grassland soils: a model."
35 Biogeochemistry 5:109-131.
36 Paustian, K., et al. (1997a). "Agricultural soils as a sink to mitigate CO2 emissions." Soil Use and Management 13:
37 230-244.
38 Paustian, K., et al. (1997b) Management controls on soil carbon. In Soil organic matter in temperate
39 agroecosystems: long-term experiments in North America (Paul E.A., K. Paustian, and C.V. Cole, eds.). Boca
40 Raton, CRC Press, pp. 15-49.
41 Potter, C. S., J.T. Randerson, C.B. Fields, P.A. Matson, P.M. Vitousek, H.A. Mooney, and S.A. Klooster (1993)
42 "Terrestrial ecosystem production: a process model based on global satellite and surface data." Global
43 Biogeochemical Cycles 7:811-841.
44 Potter, C., S. Klooster, A. Huete, and V. Genovese (2007) Terrestrial carbon sinks for the United States predicted
45 fromMODIS satellite data and ecosystem modeling. Earth Interactions 11, Article No. 13, DOI 10.1175/EI228.1.
46 PRISM Climate Group (2015) PRISM Climate Data. Oregon State University. July 24, 2015. Available online at:
47 .
References 10-59
-------
1 Soil Survey Staff (2016) State Soil Geographic (STATSGO) Database for State. Natural Resources Conservation
2 Service, United States Department of Agriculture. Available online at:
3 .
4 Towery, D. (2001) Personal Communication. Dan Towery regarding adjustments to the CTIC (1998) tillage data to
5 reflect long-term trends, Conservation Technology Information Center, West Lafayette, IN, and Marlen Eve,
6 National Resource Ecology Laboratory, Fort Collins, CO. February 2001.
7 USDA-ERS (2015) Agricultural Resource Management Survey (ARMS) Farm Financial and Crop Production
8 Practices: Tailored Reports. Available online at: .
10 USDA-ERS (1997) Cropping Practices Survey Data—1995. Economic Research Service, United States Department
11 of Agriculture. Available online at: .
12 USDA-FSA (2015) Conservation Reserve Program Monthly Summary - September 2015. U.S. Department of
13 Agriculture, Farm Service Agency, Washington, D.C. Available online at:
14 .
15 USDA-NRCS (2000) Digital Data and Summary Report: 1997 National Resources Inventory. Revised December
16 2000. Resources Inventory Division, Natural Resources Conservation Service, United States Department of
17 Agriculture, Beltsville, MD.
18 USDA-NRCS (2015) Summary Report: 2012 National Resources Inventory, Natural Resources Conservation
19 Service, Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
20 Available online at: .
21 Land Converted to Cropland
22 Birdsey, R. (1996) "Carbon Storage for Major Forest Types and Regions in the Conterminous United States." In
23 R.N. Sampson and D. Hair, (eds.). Forest and Global Change, Volume 2: Forest Management Opportunities for
24 Mitigating Carbon Emissions. American Forests. Washington, D.C., 1-26 and 261-379 (appendices 262 and 263).
25 Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer. Del Grosso,
26 S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel (2001) "Simulated
27 Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the DAYCENT Model." In Modeling Carbon
28 and Nitrogen Dynamics for Soil Management, Schaffer, M., L. Ma, S. Hansen, (eds.). CRC Press, Boca Raton,
29 Florida, pp. 303-332.
30 Del Grosso, S.J., S.M. Ogle, W.J. Parton (2011) Soil organic matter cycling and greenhouse gas accounting
31 methodologies, Chapter 1, pp 3-13 DOI: 10.1021/bk-2011-1072.ch001. In: Understanding Greenhouse Gas
32 Emissions from Agricultural Management (L. Guo, A. Gunasekara, L. McConnell. Eds.), American Chemical
33 Society, Washington, D.C.
34 Domke, G.M., J.E. Smith, and C.W. Woodall. (2011) Accounting for density reduction and structural loss in
35 standing dead trees: Implications for forest biomass and carbon stock estimates in the United States. Carbon Balance
3 6 and Management. 6:14.
37 Domke, G.M., et al. 2013. From models to measurements: comparing down dead wood carbon stock estimates in the
38 U.S. forest inventory. PLoS ONE 8(3): e59949.
39 Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., and Smith, J.E. (2016) A framework for estimating litter
40 carbon stocks in forests of the United States. Science of the Total Environment 557-558: 469-478.
41 Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and Wickham, J. (2011) Completion of
42 the 2006 National Land Cover Database for the Conterminous United States, PE&RS, Vol. 77(9):858-864.
43 Harmon, M.E., C.W. Woodall, B. Fasth, J. Sexton, M. Yatkov. (2011) Differences between standing and downed
44 dead tree wood density reduction factors: A comparison across decay classes and tree species. Res. Paper. NRS-15.
45 Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 40 p.
10-60 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A., VanDriel, J.N., and Wickham,
2 J. (2007) Completion of the 2001 National Land Cover Database for the Conterminous United States. Photogrammetric
3 Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
4 Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
5 Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-
6 Representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, v. 81, no. 5,
7 p. 345-354.Houghton, R. A., et al. (1983) "Changes in the carbon content of terrestrial biota and soils between 1860
8 and 1980: a net release of CO2 to the atmosphere." Ecological Monographs 53: 235-262.
9 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
10 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
11 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
12 Jenkins, J.C., D.C. Chojnacky, L.S. Heath, and R. A. Birdsey (2003) "National-scale biomass estimators for United
13 States tree species." Forest Science 49(1): 12-35.
14 Metherell, A.K., L.A. Harding, C.V. Cole, and W.J. Parton (1993) "CENTURY Soil Organic Matter Model
15 Environment." Agroecosystem version 4.0. Technical documentation, GPSR Tech. Report No. 4, USDA/ARS, Ft.
16 Collins, CO.
17 Ogle, S.M., F.J. Breidt, M. Easter, S. Williams, K. Killian, and K. Paustian (2010) "Scale and uncertainty in
18 modeled soil organic carbon stock changes for U.S. croplands using a process-based model." Global Change
19 Biology 16:810-820.
20 Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and management
21 impacts on soil organic carbon storage for U.S. agroecosystems between 1982 and 1997." Global Change Biology
22 9:1521-1542.
23 Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface Submodel:
24 Description and Testing". Glob. Planet. Chang. 19: 35-48.
25 Parton, W.J., D.S. Ojima, C.V. Cole, and D.S. Schimel (1994) "A General Model for Soil Organic Matter
26 Dynamics: Sensitivity to litter chemistry, texture and management," in Quantitative Modeling of Soil Forming
27 Processes. Special Publication 39, Soil Science Society of America, Madison, WI, 147-167.
28 Schimel, D.S. (1995) "Terrestrial ecosystems and the carbon cycle." Global Change Biology 1: 77-91.
29 Smith, J.E.; Heath, L.S.; Skog, K.E.; Birdsey, R.A. (2006) Methods for calculating forest ecosystem and harvested
30 carbon with standard estimates for forest types of the United States. Gen. Tech. Rep. NE-343. Newtown Square, PA:
31 U.S. Department of Agriculture, Forest Service, Northeastern Research Station. 216 p.
32 Tubiello, F. N., et al. (2015) "The Contribution of Agriculture, Forestry and other Land Use activities to Global
33 Warming, 1990-2012." Global Change Biology 21:2655-2660.
34 USDA Forest Service (2015) Forest Inventory and Analysis National Program: FIA Data Mart. U.S. Department of
35 Agriculture Forest Service. Washington, D.C. Available online at: . Accessed on 17 September 2015.
37 USDA-NRCS (2015) Summary Report: 2012 National Resources Inventory, Natural Resources Conservation
38 Service, Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
39 Available online at: .
40 Woodall, C. W., and V.J. Monleon (2008) Sampling protocol, estimation, and analysis procedures for the down
41 woody materials indicator of the FIA program. Gen. Tech. Rep. NRS-22. Newtown Square, PA: U.S. Department of
42 Agriculture, Forest Service, Northern Research Station. 68 p.
43 Woodall, C.W., L.S. Heath, G.M. Domke, and M.C. Nichols (2011) Methods and equations for estimating
44 aboveground volume, biomass, and carbon for trees in the U.S. forest inventory, 2010. Gen. Tech. Rep. NRS-88.
45 Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 30 p.
References 10-61
-------
1 Grassland Remaining Grassland: Soil Carbon Stock Changes
2 Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
3 Del Grosso, S.J., S.M. Ogle, W.J. Parton (2011) Soil organic matter cycling and greenhouse gas accounting
4 methodologies, Chapter 1, pp 3-13 DOI: 10.1021/bk-2011-1072.ch001. In: Understanding Greenhouse Gas
5 Emissions from Agricultural Management (L. Guo, A. Gunasekara, L. McConnell. Eds.), American Chemical
6 Society, Washington, D.C.
7 Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel (2001)
8 "Simulated Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the DAYCENT Model." In
9 Modeling Carbon and Nitrogen Dynamics for Soil Management, Schaffer, M., L. Ma, S. Hansen, (eds.). CRC Press,
10 Boca Raton, Florida, pp. 303-332.
11 Edmonds, L., R. L. Kellogg, B. Kintzer, L. Knight, C. Lander, J. Lemunyon, D. Meyer, D.C. Moffitt, and J.
12 Schaefer (2003) "Costs associated with development and implementation of Comprehensive Nutrient Management
13 Plans." Part I—Nutrient management, land treatment, manure and wastewater handling and storage, and
14 recordkeeping. Natural Resources Conservation Service, U.S. Department of Agriculture. Available online at:
15 .
16 EPA (1999) Biosolids Generation, Use and Disposal in the United States. Office of Solid Waste, U.S.
17 Environmental Protection Agency. Available online at: .
18 Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and Wickham, J. (2011) Completion of
19 the 2006 National Land Cover Database for the Conterminous United States, PE&RS, Vol. 77(9):858-864.
20 Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N, Larson, C., Herold, N, McKerrow, A., VanDriel, J.N., and Wickham,
21 J. (2007) Completion of the 2001 National Land Cover Database for the Conterminous United States. Photogrammetric
22 Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
23 Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
24 Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-Representing
25 a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, v. 81, no. 5, p. 345-354.
26 Kellogg, R.L., C.H. Lander, D.C. Moffitt, and N. Gollehon (2000) Manure Nutrients Relative to the Capacity of
27 Cropland and Pastureland to Assimilate Nutrients: Spatial and Temporal Trends for the United States. U.S.
28 Department of Agriculture, Washington, D.C. Publication number nps00-0579.
29 Metherell, A.K., L.A. Harding, C.V. Cole, and W.J. Parton (1993) "CENTURY Soil Organic Matter Model
30 Environment." Agroecosystem version 4.0. Technical documentation, GPSR Tech. Report No. 4, USDA/ARS, Ft.
31 Collins, CO.
32 NEBRA (2007) A National Biosolids Regulation, Quality, End Use & Disposal Survey. North East Biosolids and
33 Residuals Association. July 21, 2007.
34 Nusser, S.M. and J.J. Goebel (1997) The national resources inventory: a long-term multi-resource monitoring
35 programme. Environmental and Ecological Statistics 4:181-204.
36 Ogle, S.M., F.J. Breidt, M. Easter, S. Williams, K. Killian, and K. Paustian (2010) "Scale and uncertainty in
37 modeled soil organic carbon stock changes for U.S. croplands using a process-based model." Global Change
38 Biology 16:810-820.
39 Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and management
40 impacts on soil organic carbon storage for U.S. agroecosystems between 1982 and 1997." Global Change Biology
41 9:1521-1542.
42 Parton, W.J., D.S. Ojima, C.V. Cole, and D.S. Schimel (1994) "A General Model for Soil Organic Matter
43 Dynamics: Sensitivity to litter chemistry, texture and management," in Quantitative Modeling of Soil Forming
44 Processes. Special Publication 39, Soil Science Society of America, Madison, WI, 147-167.
45 Parton, W.J., D.S. Schimel, C.V. Cole, D.S. Ojima (1987) "Analysis of factors controlling soil organic matter levels
46 in Great Plains grasslands." Soil Science Society of America Journal 51:1173-1179.
10-62 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Parton, W.J., J.W.B. Stewart, C.V. Cole. (1988) "Dynamics of C, N, P, and S in grassland soils: a model."
2 Biogeochemistry 5:109-131.
3 Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface Submodel:
4 Description and Testing". Glob. Planet. Chang. 19: 35-48.
5 United States Bureau of Land Management (BLM) (2014) Rangeland Inventory, Monitoring, and Evaluation
6 Reports. Bureau of Land Management. U.S. Department of the Interior. Available online at:
7 .
8 USDA-NRCS (2015) Summary Report: 2012 National Resources Inventory, Natural Resources Conservation
9 Service, Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
10 Available online at: .
11 Grassland Remaining Grassland: Non-C02 Emissions from
12 Grassland Fires
13 Anderson, R.C. Evolution and origin of the Central Grassland of North America: climate, fire and mammalian
14 grazers. Journal of the Torrey Botanical Society 133: 626-647.
15 Andreae, M.O. and P. Merlet (2001) Emission of trace gases and aerosols from bio mass burning. Global
16 Biogeochemical Cycles 15:955-966.
17 Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
18 Chapin, F.S., S.F. Trainor, O. Huntington, A.L. Lovecraft, E. Zavaleta, D.C. Natcher, A.D. McGuire, J.L. Nelson, L.
19 Ray, M. Calef, N. Fresco, H. Huntington, T.S. Rupp, L. DeWilde, and R.L. Naylor (2008) Increasing wildfires in
20 Alaska's Boreal Forest: Pathways to potential solutions of a wicked problem. Bioscience 58:531-540.
21 Daubenmire, R. (1968) Ecology of fire in grasslands. Advances in Ecological Research 5:209-266.
22 Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and Wickham, J. (2011) Completion of
23 the 2006 National Land Cover Database for the Conterminous United States, PE&RS, Vol. 77(9):858-864.
24 Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A., VanDriel, J.N., and Wickham,
25 J. (2007) Completion of the 2001 National Land Cover Database for the Conterminous United States. Photogrammetric
26 Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
27 Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
28 Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-Representing
29 a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, v. 81, no. 5, p. 345-354.
30 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
31 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
32 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
33 Ogle, S.M., S. Spencer, M. Hartman, L. Buendia, L. Stevens, D. du Toit, J. Witi (2016) "Developing national
34 baseline GHG emissions and analyzing mitigation potentials for agriculture and forestry using an advanced national
35 GHG inventory software system." In Advances in Agricultural Systems Modeling 6, Synthesis and Modeling of
36 Greenhouse Gas Emissions and Carbon Storage in Agricultural and Forestry Systems to Guide Mitigation and
37 Adaptation, S. Del Grosso, L.R. Ahuja and W.J. Parton (eds.), American Society of Agriculture, Crop Society of
38 America and Soil Science Society of America, pp. 129-148.
39 Nusser, S.M. and J.J. Goebel (1997) The national resources inventory: a long-term multi-resource monitoring
40 programme. Environmental and Ecological Statistics 4:181-204.
41 USDA-NRCS (2015) Summary Report: 2012 National Resources Inventory, Natural Resources Conservation
42 Service, Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
43 Available online at: .
References 10-63
-------
i Land Converted to Grassland
2 Birdsey, R. (1996) "Carbon Storage for Major Forest Types and Regions in the Conterminous United States." In
3 R.N. Sampson and D. Hair, (eds.). Forest and Global Change, Volume 2: Forest Management Opportunities for
4 Mitigating Carbon Emissions. American Forests. Washington, D.C., 1-26 and 261-379 (appendices 262 and 263).
5 Breshears, D.D., Knapp, A.K., Law, D.J., Smith, M.D., Twidwell, D. and Wonkka, C.L., 2016. Rangeland
6 Responses to Predicted Increases in Drought Extremity. Rangelands, 38(4), pp. 191-196.
7 Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
8 Del Grosso, S.J., S.M. Ogle, W.J. Parton. (2011) Soil organic matter cycling and greenhouse gas accounting
9 methodologies, Chapter 1, pp 3-13 DOI: 10.1021/bk-2011-1072.ch001. In: Understanding Greenhouse Gas
10 Emissions from Agricultural Management (L. Guo, A. Gunasekara, L. McConnell. Eds.), American Chemical
11 Society, Washington, D.C.
12 Del Grosso, S.J., W.J. Parton, A.R. Mosier, M.D. Hartman, J. Brenner, D.S. Ojima, and D.S. Schimel (2001)
13 "Simulated Interaction of Carbon Dynamics and Nitrogen Trace Gas Fluxes Using the DAYCENT Model." In
14 Modeling Carbon and Nitrogen Dynamics for Soil Management (Schaffer, M., L. Ma, S. Hansen, (eds.). CRC Press,
15 Boca Raton, Florida, pp. 303-332.
16 Domke, G.M., J.E. Smith, and C.W. Woodall. (2011) Accounting for density reduction and structural loss in
17 standing dead trees: Implications for forest biomass and carbon stock estimates in the United States. Carbon Balance
18 and Management. 6:14.
19 Domke, G.M., et al. 2013. From models to measurements: comparing down dead wood carbon stock estimates in the
20 U.S. forest inventory. PLoS ONE 8(3): e59949.
21 Domke, G.M., Perry, C.H., Walters, B.F., Woodall, C.W., and Smith, J.E. (2016) A framework for estimating litter
22 carbon stocks in forests of the United States. Science of the Total Environment 557-558: 469-478.
23 Epstein, H.E., Gill, R.A., Paruelo, J.M., Lauenroth, W.K., Jia, G.J. and Burke, I.C., 2002. The relative abundance of
24 three plant functional types in temperate grasslands and shrublands of North and South America: effects of projected
25 climate change. Journal of Biogeography, 29(7), pp.875-888.
26 Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and Wickham, J. (2011) Completion of
27 the 2006 National Land Cover Database for the Conterminous United States, PE&RS, Vol. 77(9):858-864.
28 Harmon, M.E., C.W. Woodall, B. Fasth, J. Sexton, M. Yatkov. (2011) Differences between standing and downed
29 dead tree wood density reduction factors: A comparison across decay classes and tree species. Res. Paper. NRS-15.
30 Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 40 p.
31 Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A., VanDriel, J.N., and Wickham,
32 J. (2007) Completion of the 2001 National Land Cover Database for the Conterminous United States. Photogrammetric
33 Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
34 Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
3 5 Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-Representing
36 a decade ofland cover change information. Photogrammetric Engineering and Remote Sensing, v. 81,no. 5,p. 345-354.
37 Houghton, R. A., et al. (1983) "Changes in the carbon content of terrestrial biota and soils between 1860 and 1980: a
38 net release of CO2 to the atmosphere." Ecological Monographs 53: 235-262.
39 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
40 Inventories Programme, The Intergovernmental Panel on Climate Change, [H.S. Eggleston, L. Buendia, K. Miwa, T
41 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
42 Jenkins, J.C., D.C. Chojnacky, L.S. Heath, and R.A. Birdsey (2003) "National-scale biomass estimators for United
43 States tree species." Forest Science 49(1): 12-35.
44 Jurena, P.N. and Archer, S., 2003. Woody plant establishment and spatial heterogeneity in grasslands. Ecology,
45 84(4), pp.907-919.
10-64 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Lenihan, J.M., Drapek, R., Bachelet, D. and Neilson, R.P., 2003. Climate change effects on vegetation distribution,
2 carbon, and fire in California. Ecological Applications, 13(6), pp.1667-1681.
3 Metherell, A.K., L.A. Harding, C.V. Cole, and W.J. Parton (1993) "CENTURY Soil Organic Matter Model
4 Environment." Agroecosystem version 4.0. Technical documentation, GPSR Tech. Report No. 4, USDA/ARS, Ft.
5 Collins, CO.
6 Ogle, S.M., F.J. Breidt, M. Easter, S. Williams, K. Killian, and K. Paustian (2010) "Scale and uncertainty in
7 modeled soil organic carbon stock changes for U.S. croplands using a process-based model." Global Change
8 Biology 16:810-820.
9 Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and management
10 impacts on soil organic carbon storage for U.S. agroecosystems between 1982 and 1997." Global Change Biology
11 9:1521-1542.
12 Parton, W.J., D.S. Ojima, C.V. Cole, and D.S. Schimel (1994) "A General Model for Soil Organic Matter
13 Dynamics: Sensitivity to litter chemistry, texture and management," in Quantitative Modeling of Soil Forming
14 Processes. Special Publication 39, Soil Science Society of America, Madison, WI, 147-167.
15 Parton, W.J., D.S. Schimel, C.V. Cole, D.S. Ojima (1987) "Analysis of factors controlling soil organic matter levels
16 in Great Plains grasslands." Soil Science Society of America Journal 51:1173-1179.
17 Parton, W.J., J.W.B. Stewart, C.V. Cole (1988) "Dynamics of C, N, P, and S in grassland soils: a model."
18 Biogeochemistry 5:109-131.
19 Parton, W.J., M.D. Hartman, D.S. Ojima, and D.S. Schimel (1998) "DAYCENT: Its Land Surface Submodel:
20 Description and Testing". Glob. Planet. Chang. 19: 35-48.
21 Schimel, D.S. (1995) "Terrestrial ecosystems and the carbon cycle." Global Change Biology 1: 77-91.
22 Scholes, R.J. and Archer, S.R., 1997. Tree-grass interactions in savannas 1. Annual review of Ecology and
23 Systematics, 28(1), pp.517-544.
24 Sims, P.L., Singh, J.S. and Lauenroth, W.K., 1978. The structure and function of ten western North American
25 grasslands: I. Abiotic and vegetational characteristics. The Journal of Ecology, pp.251-285.
26 Smith, J.E.; Heath, L.S.; Skog, K.E.; Birdsey, R.A. (2006) Methods for calculating forest ecosystem and harvested
27 carbon with standard estimates for forest types of the United States. Gen. Tech. Rep. NE-343. Newtown Square, PA:
28 U.S. Department of Agriculture, Forest Service, Northeastern Research Station. 216 p.
29 Tubiello, F. N., et al. (2015) "The Contribution of Agriculture, Forestry and other Land Use activities to Global
30 Warming, 1990-2012." Global Change Biology 21:2655-2660.
31 United States Bureau of Land Management (BLM) (2014) Rangeland Inventory, Monitoring, and Evaluation
32 Reports. Bureau of Land Management. U.S. Department of the Interior. Available online at:
33 .
34 USDA-NRCS (2015) Summary Report: 2012 National Resources Inventory, Natural Resources Conservation
35 Service, Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
36 Available online at: .
37 Woodall, C. W., and V.J. Monleon (2008) Sampling protocol, estimation, and analysis procedures for the down
38 woody materials indicator of the FIA program. Gen. Tech. Rep. NRS-22. Newtown Square, PA: U.S. Department of
39 Agriculture, Forest Service, Northern Research Station. 68 p.
40 Woodall, C.W., L.S. Heath, G.M. Domke, and M.C. Nichols. (2011) Methods and equations for estimating
41 aboveground volume, biomass, and carbon for trees in the U.S. forest inventory, 2010. Gen. Tech. Rep. NRS-88.
42 Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 30 p.
References 10-65
-------
1 Wetlands Remaining Wetlands: CO2, CH4, and N20 Emissions
2 from Peatlands Remaining Peatlands
3 Apodaca, L. (2011) Email correspondence. Lori Apodaca, Peat Commodity Specialist, USGS and Emily Rowan,
4 ICF International. November.
5 Apodaca, L. (2008) E-mail correspondance. Lori Apodaca, Peat Commodity Specialist, USGS and Emily Rowan,
6 ICF International. October and November.
7 Cleary, J., N. Roulet and T.R. Moore (2005) "Greenhouse gas emissions from Canadian peat extraction, 1990-2000:
8 A life-cycle analysis." Ambio 34:456-461.
9 Division of Geological & Geophysical Surveys (DGGS), Alaska Department of Natural Resources (1997-2015)
10 Alaska's Mineral Industry Report (1997-2014). Alaska Department of Natural Resources, Fairbanks, AK. Available
11 online at .
12 IPCC (2013) 2013 Supplement to the 2006 IP CC Guidelines for National Greenhouse Gas Inventories: Wetlands.
13 Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds.). Published:
14 IPCC, Switzerland.
15 IPCC (2007) Climate Change 2007: Synthesis Report. Contribution of Working GroupsI, II andlll to the Fourth
16 Assessment Report (AR4) of the IPCC. The Intergovernmental Panel on Climate Change, R.K. Pachauri, A. Resinger
17 (eds.). Geneva, Switzerland.
18 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
19 Inventories Programme, The Intergovernmental Panel on Climate Change. H.S. Eggleston, L. Buendia, K. Miwa, T.
20 Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
21 Szumigala, D.J. (2011) Phone conversation. Dr. David Szumigala, Division of Geological and Geophysical Surveys,
22 Alaska Department of Natural Resources and Emily Rowan, ICF International. January 18, 2011.
23 Szumigala, D.J. (2008) Phone conversation. Dr. David Szumigala, Division of Geological and Geophysical Surveys,
24 Alaska Department of Natural Resources and Emily Rowan, ICF International. October 17, 2008.
25 USGS (1991-2015) Minerals Yearbook: Peat (1994-2014). United States Geological Survey, Reston, VA.
26 Available online at < http://minerals.usgs.gOv/minerals/pubs/commodity/peat/index.html#myb >. USGS (2016)
27 Mineral Commodity Summaries: Peat (2016). United States Geological Survey, Reston, VA. Available online at
28 .
29 Wetlands Remaining Coastal Wetlands: Emissions and
30 Removals from Coastal Wetlands Remaining Coastal Wetlands
31 Anisfeld, S. C., Tobin, M. & Benoit, G. (1999) Sedimentation rates in flow-restricted and restored salt marshes in
32 Long Island Sound. Estuaries 22(2A): 231-244.
33 Bryant, J. C., & Chabrek, R. H. (1998) Effects of impoundment on vertical accretion of coastal marsh. Estuaries 21:
34 416-422.
35 Cahoon, D. R., Lynch, J. C. & Knaus, R. M. (1996) Improved cryogenic coring device for sampling wetland soils.
36 Journal of Sedimentary Research 66(5): 1025-1027.
37 Cahoon, D. R., & Turner, R. E. (1989) Accretion and canal impacts in a rapidly subsiding wetland. II. Feldspar
38 marker horizon technique. Estuaries, 12: 260-268.
39 Callaway, J. C., R.D. DeLaune, and W.H. Patrick. (1997) Sediment accretion rates from four coastal wetlands along
40 the Gulf of Mexico. Journal of Coastal Research 13: 181-191.
41 Callaway, J. C., Borgnis, E. L., Turner, R. E. & Milan, C. S. (2012) Carbon sequestration and sediment accretion in
42 San Francisco Bay tidal wetlands. Estuaries and Coasts 35(5): 1163-1181.
10-66 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Castaneda-Moya, E., Twilley, R. R., & Rivera-Monroy, V. H. (2013) Allocation of biomass and net primary
2 productivity of mangrove forests along environmental gradients in the Florida Coastal Everglades, USA. Forest
3 Ecology and Management 307: 226-241.
4 Chen, R., & Twilley, R. R. (1999). A simulation model of organic matter and nutrient accumulation in mangrove
5 wetland soils. Biogeochemistry, 44(1), 93-118.
6 Chmura, G. L., Anisfeld, S. C., Cahoon, D. R. & Lynch, J. C. (2003) Global carbon sequestration in tidal, saline
7 wetland soils. Global Biogeochemical Cycles 17(4).
8 Choi, Y. & Wang, Y. (2001) Dynamics of carbon sequestration in a coastal wetland using radiocarbon
9 measurements. Global Biogeochemical Cycles 18(4).
10 Connor, R. F., Chmura, G. L. & Beecher, C. B. (2001) Carbon accumulation in Bay of Fundy salt marshes:
11 Implications for restoration of reclaimed marshes. Global Biogeochemical Cycles 15(4): 943-954.
12 Couvillion, B. R., Barras, J. A., Steyer, G. D., Sleavin, W., Fischer, M., Beck, H., & Heckman, D. (2011). Land area
13 change in coastal Louisiana (1932 to 2010) (pp. 1-12). US Department of the Interior, US Geological Survey.
14 Couvillion, B.R., Fischer, M.R., Beck, H.J. and Sleavin, W.J. (2016) Spatial Configuration Trends in Coastal
15 Louisiana from 1986 to 2010. Wetlands 1-13.
16 Craft, C. B., Broome, S. W. & Seneca, E. D. (1988) Nitrogen, phosphorus and organic carbon pools in natural and
17 transplanted marsh soils. Estuaries 11(4): 272-280.
18 Craft, C., S. Broome, and C. Campbell. (2002) Fifteen years of vegetation and soil development after brackish water
19 marsh creation. Restoration Ecology (10): 248-258.
20 Craft, C. (2007) Freshwater input structures soil properties, vertical accretion, and nutrient accumulation of Georgia
21 and U.S. tidal marshes. Limnology and Oceanography 52(3): 1220-1230.
22 Crooks, S., Findsen, J., Igusky, K., Orr, M.K. and Brew, D. (2009) Greenhouse Gas Mitigation Typology Issues
23 Paper: Tidal Wetlands Restoration. Report by PWA and SAIC to the California Climate Action Reserve.
24 Crooks, S., Rybczyk, J., O'Connell, K., Devier, D.L., Poppe, K., Emmett-Mattox, S. (2014) Coastal Blue Carbon
25 Opportunity Assessment for the Snohomish Estuary: The Climate Benefits of Estuary Restoration. Report by
26 Environmental Science Associates, Western Washington University, EarthCorps, and Restore America's Estuaries.
27 DeLaune. R. D., & White, J. R. (2012). Will coastal wetlands continue to sequester carbon in response to an increase
28 in global sea level?: A case study of the rapidly subsiding Mississippi river deltaic plain. Climatic Change, 110(1),
29 297-314.
30 Doughty, C. L., Langley, J. A., Walker, W. S., Feller, I. C., Schaub, R., & Chapman, S. K. (2015) Mangrove range
31 expansion rapidly increases coastal wetland carbon storage. Estuaries and Coasts doi: 10.1007/s 12237-015-9993-8.
32 Drexler, J. Z., Fontaine, C. S., Brown, T. A. (2009) Peat accretion histories during the past 6,000 years in marshes of
33 the Sacramento-San Joaquin Delta, CA, USA. Estuaries and Coasts 32: 871-892.
34 EPA (2016) National Wetland Condition Assessment 2011. United State Environmental Protection Agency.
35 Washington, D.C. EPA-843-R-15-005.
36 Hatton, R. S., DeLaune, R. D., & Patrick Jr, W. H. (1981) Sedimentation, accretion, and subsidence in marshes of
37 Barataria Basin, Louisiana. Limnology and Oceanography 28(3): 494-502.
38 Henry, K. M., & Twilley, R. R. (2013) Soil development in a coastal Louisiana wetland during a climate-induced
39 vegetation shift from salt marsh to mangrove. Journal of Coastal Research 29: 1273-1283.
40 Hu, Z., Lee. J.W., Chandran, K., Kim. S. and Khanal, S.K. (2012) N;0 Emissions from Aquaculture: A Review.
41 Environmental Science & Technology 46(12): 6470-6480.
42 Hussein, A. H., Rabenhorst, M. C. & Tucker, M. L. (2004) Modeling of carbon sequestration in coastal marsh soils.
43 Soil Science Society of America Journal 68(5): 1786-1795.
References 10-67
-------
1 IPCC (2013) 2013 Supplement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands.
2 Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds.). Published:
3 IPCC, Switzerland.
4 IPCC (2003). Good Practice Guidance for Land Use, Land-Use Change and Forestry. LUCF Sector Good Practice
5 Guidance, Chapter 3. Jim Penman, Michael Gytarsky, Taka Hiraishi, Thelma Krug, Dina Kruger, Riitta Pipatti,
6 Leandro Buendia, Kyoko Miwa, Todd Ngara, Kiyoto Tanabe and Fabian Wagner (eds). Institute of Global
7 Environmental Strategies (IGES), on behalf of the Intergovernmental Panel on Climate Change (IPCC): Hayama,
8 Japan.
9 Kearney, M. S. & Stevenson, J. C. (1991) Island land loss and marsh vertical accretion rate evidence for historical
10 sea-level changes in Chesapeake Bay. Journal of Coastal Research 7(2): 403-415.
11 Loomis, M. J. & Craft, C. B. (2010) Carbon sequestration and nutrient (nitrogen, phosphorus) accumulation in river -
12 dominated tidal marshes, Georgia, USA. Soil Science Society of America Journal 74(3): 1028-1036.
13 Lynch, J. C., Sedimentation and nutrient accumulation in mangrove ecosystems of the Gulf of Mexico, M.S. thesis,
14 Univ. of Southwestern Louisiana, Lafayette, La., 1989.
15 Marchio, D.A., Savarese, M., Bovard, B., & Mitsch, W.J. (2016) Carbon sequestration and sedimentation in
16 mangrove swamps influenced by hydrogeomorphic conditions and urbanization in Southwest Florida. Forests 7:
17 116-135.
18 Markewich, H. W., Wysocki, D. A., Pavich, M. J., Rutledge, E. M., Millard, H. T., Rich, F. J., Maat, P. B., Rubin,
19 M. & McGeehin, J. P. (1998) Paleopedology plus TL, Be-10, and C-14 dating as tools in stratigraphic and
20 paleoclimatic investigations, Mississippi River Valley, USA. Quaternary International 51-2: 143-167.
21 McCaffrey, R. J. & Thomson, J. (1980) A Record of the Accumulation of Sediment and Trace Metals in A
22 Connecticut Salt Marsh. In: Advances in Geophysics, ed. S. Barry, pp. 165-236. Elsevier.
23 McCombs, J.W., Herold, N.D., Burkhalter, S.G. and Robinson C.J., (2016) Accuracy Assessment of NOAA Coastal
24 Change Analysis Program 2006-2010 Land Cover and Land Cover Change Data. Photogrammetric Engineering &
25 Remote Sensing. 82:711-718.
26 McKee, K. L. & Faulkner, P. L. (2000) Restoration of biogeochemical function in mangrove forests. Restoration
27 Ecology 8(3): 247-259.
28 Miller, R. L., Fram, M. S., Fuji, R., Wheeler, G. (2008) Subsidence reversal in a re-established wetland in the
29 Sacramento-San Joaquin Delta, California, USA. San Francisco Estuary and Watershed Science, October 2008.
30 National Marine Fisheries Service (2016) Fisheries of the United States, 2015. U.S. Department of Commerce,
31 NOAA Current Fisheries Statistics No. 2015.
32 Orson, R., Warren, R. & Niering, W. (1998) Interpreting sea level rise and rates of vertical marsh accretion in a
33 southern New England tidal salt marsh. Estuarine, Coastal and Shelf Science 47(4): 419-429.
34 Patrick Jr, W. H. & DeLaune, R. (1990) Subsidence, accretion, and sea level rise in south San Francisco Bay
35 marshes. Limnology and Oceanography 35(6): 1389-1395.
36 Perry, C. L. & Mendelssohn, I. A. (2009) Ecosystem effects of expanding populations of Avicenna germinans in a
37 Louisiana salt marsh. Wetlands 29(1): 396-406.
38 Roman, C., Peck, J., Allen, J., King, J. & Appleby, P. (1997) Accretion of a New England (USA) salt marsh in
39 response to inlet migration, storms, and sea-level rise. Estuarine, Coastal and Shelf Science 45(6): 717-727.
40 Ross, M. S., Ruiz, P. L., Telesnicki, G. J. & Meeder, J. F. (2001) Estimating aboveground biomass and production
41 in mangrove communities of Biscayne National Park, Florida (USA). Wetlands Ecology and Management 9(1): 27-
42 37.
43 Weston, N. B., Neubauer, S. C., Velinsky, D. J., & Vile, M. A. (2014) Net ecosystem carbon exchange and the
44 greenhouse gas balance of tidal marshes along an estuarine salinity gradient. Biogeochemistry 120: 163-189.
10-68 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Land Converted to Wetlands
2 Anisfeld, S. C., Tobin, M. & Benoit, G. (1999) Sedimentation rates in flow-restricted and restored salt marshes in
3 Long Island Sound. Estuaries 22(2A): 231-244.
4 Bryant, J. C., & Chabrek, R. H. (1998) Effects of impoundment on vertical accretion of coastal marsh. Estuaries 21:
5 416-422.
6 Cahoon, D. R., Lynch, J. C. & Knaus, R. M. (1996) Improved cryogenic coring device for sampling wetland soils.
7 Journal of Sedimentary Research 66(5): 1025-1027.
8 Cahoon, D. R., & Turner, R. E. (1989) Accretion and canal impacts in a rapidly subsiding wetland. II. Feldspar
9 marker horizon technique. Estuaries, 12: 260 - 268.
10 Callaway, J. C., R.D. DeLaune, and W.H. Patrick. (1997) Sediment accretion rates from four coastal wetlands along
11 the Gulf of Mexico. Journal of Coastal Research 13: 181-191.
12 Callaway, J. C., Borgnis, E. L., Turner, R. E. & Milan, C. S. (2012) Carbon sequestration and sediment accretion in
13 San Francisco Bay tidal wetlands. Estuaries and Coasts 35(5): 1163-1181.
14 Castaneda-Moya, E., Twilley, R. R., & Rivera-Monroy, V. H. (2013) Allocation of biomass and net primary
15 productivity of mangrove forests along environmental gradients in the Florida Coastal Everglades, USA. Forest
16 Ecology and Management 307: 226-241.
17 Chen, R., & Twilley, R. R. (1999). A simulation model of organic matter and nutrient accumulation in mangrove
18 wetland soils. Biogeochemistry, 44(1), 93-118.
19 Chmura, G. L., Anisfeld, S. C., Cahoon, D. R. & Lynch, J. C. (2003) Global carbon sequestration in tidal, saline
20 wetland soils. Global Biogeochemical Cycles 17(4).
21 Choi, Y. & Wang, Y. (2001) Dynamics of carbon sequestration in a coastal wetland using radiocarbon
22 measurements. Global Biogeochemical Cycles 18(4).
23 Connor, R. F., Chmura, G. L. & Beecher, C. B. (2001) Carbon accumulation in Bay of Fundy salt marshes:
24 Implications for restoration of reclaimed marshes. Global Biogeochemical Cycles 15(4): 943-954.
25 Couvillion. B. R., B arras. J. A., St ever, G. D., Sleavin. W., Fischer, M., Beck. H., & Heckman. D. (2011). Land area
26 change in coastal Louisiana (1932 to 2010) (pp. 1-12). US Department of the Interior, US Geological Survey.
27 Couvillion, B.R., Fischer, M.R., Beck, H.J. and Sleavin, W.J. (2016) Spatial Configuration Trends in Coastal
28 Louisiana from 1986 to 2010. Wetlands 1-13.
29 Craft, C. B., Broome, S. W. & Seneca, E. D. (1988) Nitrogen, phosphorus and organic carbon pools in natural and
30 transplanted marsh soils. Estuaries 11(4): 272-280.
31 Craft, C., S. Broome, and C. Campbell. (2002) Fifteen years of vegetation and soil development after brackish water
32 marsh creation. Restoration Ecology (10): 248-258.
33 Craft, C. (2007) Freshwater input structures soil properties, vertical accretion, and nutrient accumulation of Georgia
34 and U.S. tidal marshes. Limnology and Oceanography 52(3): 1220-1230.
35 Crooks, S., Findsen, J., Igusky, K., Orr, M.K. and Brew, D. (2009) Greenhouse Gas Mitigation Typology Issues
36 Paper: Tidal Wetlands Restoration. Report by PWA and SAIC to the California Climate Action Reserve.
37 Crooks, S., Rybczyk, J., O'Connell, K., Devier, D.L., Poppe, K., Emmett-Mattox, S. (2014) Coastal Blue Carbon
38 Opportunity Assessment for the Snohomish Estuary: The Climate Benefits of Estuary Restoration. Report by
39 Environmental Science Associates, Western Washington University, EarthCorps, and Restore America's Estuaries.
40 DeLaune. R. D.. & White, J. R. (2012). Will coastal wetlands continue to sequester carbon in response to an increase
41 in global sea level?: a case study of the rapidly subsiding Mississippi river deltaic plain. Climatic Change, 110(1),
42 297-314.
43 Doughty, C. L., Langley, J. A., Walker, W. S., Feller, I. C., Schaub, R., & Chapman, S. K. (2015) Mangrove range
44 expansion rapidly increases coastal wetland carbon storage. Estuaries and Coasts doi: 10.1007/s 12237-015-9993-8.
References 10-69
-------
1 Drexler, J. Z., Fontaine, C. S., Brown, T. A. (2009) Peat accretion histories during the past 6,000 years in marshes of
2 the Sacramento-San Joaquin Delta, CA, USA. Estuaries and Coasts 32: 871-892.
3 EPA (2016) National Wetland Condition Assessment 2011. United State Environmental Protection Agency.
4 Washington, D.C. EPA-843-R-15-005.
5 Hatton, R. S., DeLaune, R. D., & Patrick Jr, W. H. (1981) Sedimentation, accretion, and subsidence in marshes of
6 Barataria Basin, Louisiana. Limnology and Oceanography 28(3): 494-502.
7 Henry, K. ML, & Twilley, R. R. (2013) Soil development in a coastal Louisiana wetland during a climate-induced
8 vegetation shift from salt marsh to mangrove. Journal of Coastal Research 29: 1273-1283.
9 Hu, Z., Lee. J.W., Chandran, K., Kim, S. and Khanal, S.K. (2012) Nitrous Oxide N;0 Emissions from Aquaculture:
10 A Review. Environmental Science & Technology 46(12): 6470-6480.
11 Hussein, A. H., Rabenhorst, M. C. & Tucker, M. L. (2004) Modeling of carbon sequestration in coastal marsh soils.
12 Soil Science Society of America Journal 68(5): 1786-1795.
13 IPCC (2000). Good Practice Guidance and Uncertainty Management in National Greenhouse Gas Inventories.
14 Quantifying Uncertainties in Practice, Chapter 6. Penman, J and Kruger, D and Galbally, I and Hiraishi, T and
15 Nyenzi, B and Emmanuel, S and Buendia, L and Hoppaus, R and Martinsen, T and Meijer, J and Miwa, K and
16 Tanabe, K (eds). Institute of Global Environmental Strategies (IGES), on behalf of the Intergovernmental Panel on
17 Climate Change (IPCC): Hayama, Japan.
18 IPCC (2003). Good Practice Guidance for Land Use, Land-Use Change and Forestry. LUCF Sector Good Practice
19 Guidance, Chapter 3. Jim Penman, Michael Gytarsky, Taka Hiraishi, Thelma Krug, Dina Kruger, Riitta Pipatti,
20 Leandro Buendia, Kyoko Miwa, Todd Ngara, Kiyoto Tanabe and Fabian Wagner (eds). Institute of Global
21 Environmental Strategies (IGES), on behalf of the Intergovernmental Panel on Climate Change (IPCC): Hayama,
22 Japan.
23 IPCC (2013) 2013 Supplement to the 2006 IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands.
24 Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds.). Published:
25 IPCC, Switzerland.
26 Kearney, M. S. & Stevenson, J. C. (1991) Island land loss and marsh vertical accretion rate evidence for historical
27 sea-level changes in Chesapeake Bay. Journal of Coastal Research 7(2): 403-415.
28 Loomis, M. J. & Craft, C. B. (2010) Carbon sequestration and nutrient (nitrogen, phosphorus) accumulation in river -
29 dominated tidal marshes, Georgia, USA. Soil Science Society of America Journal 74(3): 1028-1036.
30 Lynch, J. C., Sedimentation and nutrient accumulation in mangrove ecosystems of the Gulf of Mexico, M.S. thesis,
31 Univ. of Southwestern Louisiana, Lafayette, La., 1989.
32 Marchio, D.A., Savarese, M., Bovard, B., & Mitsch, W.J. (2016) Carbon sequestration and sedimentation in
33 mangrove swamps influenced by hydrogeomorphic conditions and urbanization in Southwest Florida. Forests 7:
34 116-135.
35 Markewich, H. W., Wysocki, D. A., Pavich, M. J., Rutledge, E. M., Millard, H. T., Rich, F. J., Maat, P. B., Rubin,
36 M. & McGeehin, J. P. (1998) Paleopedology plus TL, Be-10, and C-14 dating as tools in stratigraphic and
37 paleoclimatic investigations, Mississippi River Valley, USA. Quaternary International 51-2: 143-167.
38 McCaffrey, R. J. & Thomson, J. (1980) A Record of the Accumulation of Sediment and Trace Metals in A
39 Connecticut Salt Marsh. In: Advances in Geophysics, ed. S. Barry, pp. 165-236. Elsevier.
40 McCombs, J.W., Herold, N.D., Burkhalter, S.G. and Robinson C.J., (2016) Accuracy Assessment of NOAA Coastal
41 Change Analysis Program 2006-2010 Land Cover and Land Cover Change Data. Photogrammetric Engineering &
42 Remote Sensing. 82:711-718.
43 McKee, K. L. & Faulkner, P. L. (2000) Restoration of biogeochemical function in mangrove forests. Restoration
44 Ecology 8(3): 247-259.
45 Miller, R. L., Fram, M. S., Fuji, R., Wheeler, G. (2008) Subsidence reversal in a re-established wetland in the
46 Sacramento-San Joaquin Delta, California, USA. San Francisco Estuary and Watershed Science, October 2008.
10-70 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 National Marine Fisheries Service (2016) Fisheries of the United States, 2015. U.S. Department of Commerce,
2 NOAA Current Fisheries Statistics No. 2015.
3 Orson, R., Warren, R. & Niering, W. (1998) Interpreting sea level rise and rates of vertical marsh accretion in a
4 southern New England tidal salt marsh. Estuarine, Coastal and Shelf Science 47(4): 419-429.
5 Patrick Jr, W. H. & DeLaune, R. (1990) Subsidence, accretion, and sea level rise in south San Francisco Bay
6 marshes. Limnology and Oceanography 35(6): 1389-1395.
7 Perry, C. L. & Mendelssohn, I. A. (2009) Ecosystem effects of expanding populations of Avicenna germinans in a
8 Louisiana salt marsh. Wetlands 29(1): 396-406.
9 Roman, C., Peck, J., Allen, J., King, J. & Appleby, P. (1997) Accretion of a New England (USA) salt marsh in
10 response to inlet migration, storms, and sea-level rise. Estuarine, Coastal and Shelf Science 45(6): 717-727.
11 Ross, M. S., Ruiz, P. L., Telesnicki, G. J. & Meeder, J. F. (2001) Estimating aboveground biomass and production
12 in mangrove communities of Biscayne National Park, Florida (USA). Wetlands Ecology and Management 9(1): 27-
13 37.
14 Weston, N. B., Neubauer, S. C., Velinsky, D. J., & Vile, M. A. (2014) Net ecosystem carbon exchange and the
15 greenhouse gas balance of tidal marshes along an estuarine salinity gradient. Biogeochemistry 120: 163-189.
16 Settlements Remaining Settlements: Soil Carbon Stock
17 Changes
18 Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
19 Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and J. Wickham. (2011)
20 Completion of the 2006 National Land Cover Database for the Conterminous United States, PE&RS 77(9):858-864.
21 Homer, C., J. Dewitz, J. Fry, M. Coan, N. Hossain, C. Larson, N. Herold, A. McKerrow, J.N. VanDriel and J.
22 Wickham. (2007) Completion of the 2001 National Land Cover Database for the Conterminous United States.
23 Photogrammetric Engineering and Remote Sensing 73(4): 337-341.
24 Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D.,
25 and Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-
26 Representing a decade of land cover change information. Photogrammetric Engineering and Remote Sensing
27 81(5):345-354.
28 Nusser, S.M. and J.J. Goebel (1997) The national resources inventory: a long-term multi-resource monitoring
29 programme. Environmental and Ecological Statistics 4:181-204.
30 Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) Uncertainty in estimating land use and management
31 impacts on soil organic carbon storage for U.S. agroecosystems between 1982 and 1997. Global Change Biology
32 9:1521-1542.
33 Soil Survey Staff (2011) State Soil Geographic (STATSGO) Database for State. Natural Resources Conservation
34 Service, United States Department of Agriculture. Available online at:
35 .
36 USDA-NRCS (2015) Summary Report: 2012 National Resources Inventory. Natural Resources Conservation
37 Service, Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
38
-------
1 Dwyer, J.F., D.J. Nowak, M.H. Noble, and S.M. Sisinni (2000) Connecting People with Ecosystems in the 21st
2 Century: An Assessment of Our Nation's Urban Forests. General Technical Report PNW-GTR-490, U.S.
3 Department of Agriculture, Forest Service, Pacific Northwest Research Station, Portland, OR.
4 Fleming, L.E. (1988) Growth Estimation of Street Trees in Central New Jersey. M.S. thesis, Rutgers University,
5 New Brunswick, NJ.
6 Frelich, L.E. (1992) Predicting Dimensional Relationships for Twin Cities Shade Trees. University of Minnesota,
7 Department of Forest Resources, St. Paul, MN, p. 33.
8 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
9 Inventories Programme, The Intergovernmental Panel on Climate Change. H.S. Eggleston, L. Buendia, K. Miwa, T.
10 Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
11 Nowak, D.J. (2011) Phone conference regarding Changes in Carbon Stocks in Urban Trees estimation methodology.
12 David Nowak, USD A, Jennifer Jenkins, EPA, and Mark Flugge and Nikhil Nadkarni, ICF International. January 4,
13 2011.
14 Nowak, D.J. (2009) E-mail correspondence regarding new data for Chicago's urban forest. David Nowak, USDA
15 Forest Service to Nikhil Nadkarni, ICF International. October 7, 2009.
16 Nowak, D.J. (2007a) "New York City's Urban Forest." USDA Forest Service. Newtown Square, PA, February 2007.
17 Nowak, D.J. (2007b) E-mail correspondence regarding revised sequestration values and standard errors for
18 sequestration values. David Nowak, USDA Forest Service to Susan Asam, ICF International. October 31, 2007.
19 Nowak, D.J. (1994) "Atmospheric Carbon Dioxide Reduction by Chicago's Urban Forest." In: Chicago's Urban
20 Forest Ecosystem: Results of the Chicago Urban Forest Climate Project. E.G. McPherson, D.J. Nowak, and R. A.
21 Rowntree (eds.). General Technical Report NE-186. U.S. Department of Agriculture Forest Service, Radnor, PA. pp.
22 83-94.
23 Nowak, D.J. (1986) "Silvics of an Urban Tree Species: Norway Maple (Acer platanoides L.)." M.S. thesis, College
24 of Environmental Science and Forestry, State University of New York, Syracuse, NY.
25 Nowak, D.J., Buckelew-Cumming, A., Twardus, D., Hoehn, R., and Mielke, M. (2007). National Forest Health
26 Monitoring Program, Monitoring Urban Forests in Indiana: Pilot Study 2002, Part 2: Statewide Estimates Using the
27 UFORE Model. Northeastern Area Report. NA-FR-01e07, p. 13.
28 Nowak, D.J. and D.E. Crane (2002) "Carbon Storage and Sequestration by Urban Trees in the United States."
29 Environmental Pollution 116(3):381—389.
30 Nowak, D.J., D.E. Crane, J.C. Stevens, and M. Ibarra (2002) Brooklyn's Urban Forest. General Technical Report
31 NE-290. U.S. Department of Agriculture Forest Service, Newtown Square, PA.
32 Nowak, D.J., and E.J. Greenfield (2012) Tree and impervious cover in the United States. Journal of Landscape and
33 Urban Planning (107) pp. 21-30.
34 Nowak, D.J., E.J. Greenfield, R.E. Hoehn, and E. Lapoint (2013) Carbon Storage and Sequestration by Trees in
35 Urban and Community Areas of the United States. Environmental Pollution 178: 229-236. March 12, 2013.
36 Nowak, D.J., J.T. Walton, L.G. Kaya, and J.F. Dwyer (2005) "The Increasing Influence of Urban Environments on
37 U.S. Forest Management." Journal of Forestry 103(8):377-382.
38 Smith, W.B. and S.R. Shifley (1984) Diameter Growth, Survival, and Volume Estimates for Trees in Indiana and
39 Illinois. Research Paper NC-257. North Central Forest Experiment Station, U.S. Department of Agriculture Forest
40 Service, St. Paul, MN.
41 U.S. Census Bureau (2012) "A national 2010 urban area file containing a list of all urbanized areas and urban
42 clusters (including Puerto Rico and the Island Areas) sorted by UACE code." U.S. Census Bureau, Geography
43 Division.
44 Settlements Remaining Settlements: N20 Emissions from Soils
45 Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer.
10-72 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
2 Inventories Programme, The Intergovernmental Panel on Climate Change. H.S. Eggleston, L. Buendia, K. Miwa, T.
3 Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
4 Ruddy B.C., D.L. Lorenz, and D.K. Mueller (2006) County-level estimates of nutrient inputs to the land surface of
5 the conterminous United States, 1982-2001. Scientific Investigations Report 2006-5012. U.S. Department of the
6 Interior.
7 Settlements Remaining Settlements: Changes in Yard
s Trimmings and Food Scrap Carbon Stock'm I
9 Barlaz, M.A. (2008) "Re: Corrections to Previously Published Carbon Storage Factors." Memorandum to Randall
10 Freed, ICF International. February 28, 2008.
11 Barlaz, M.A. (2005) "Decomposition of Leaves in Simulated Landfill." Letter report to Randall Freed, ICF
12 Consulting. June 29, 2005.
13 Barlaz, M.A. (1998) "Carbon Storage during Biodegradation of Municipal Solid Waste Components in Laboratory-
14 Scale Landfills." Global Biogeochemical Cycles 12:373-380.
15 De la Cruz, F.B. and M.A. Barlaz (2010) "Estimation of Waste Component Specific Landfill Decay Rates Using
16 Laboratory-Scale Decomposition Data" Environmental Science & Technology 44:4722- 4728.
17 Eleazer, W.E., W.S. Odle, Y. Wang, and M.A. Barlaz (1997) "Biodegradability of Municipal Solid Waste
18 Components in Laboratory-Scale Landfills." Environmental Science & Technology 31:911-917.
19 EPA (2016) Advancing Sustainable Materials Management: Facts and Figures 2014. U.S. Environmental
20 Protection Agency, Office of Solid Waste and Emergency Response, Washington, D.C. Available online at
21 .
22 EPA (2015) Advancing Sustainable Materials Management: 2013 Historical (summary) Data Tables. U.S.
23 Environmental Protection Agency, Office of Solid Waste and Emergency Response, Washington, D.C.
24 EPA (1995) Compilation of Air Pollutant Emission Factors. U.S. Environmental Protection Agency, Office of Air
25 Quality Planning and Standards, Research Triangle Park, NC. AP-42 Fifth Edition. Available online at
26 .
27 EPA (1991) Characterization of Municipal Solid Waste in the United States: 1990 Update. U.S. Environmental
28 Protection Agency, Office of Solid Waste and Emergency Response, Washington, D.C. EPA/530-SW-90-042.
29 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
30 Inventories Programme, The Intergovernmental Panel on Climate Change. H.S. Eggleston, L. Buendia, K. Miwa, T.
31 Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
32 IPCC (2003) Good Practice Guidance for Land Use, Land-Use Change, and Forestry. The Intergovernmental Panel
33 on Climate Change, National Greenhouse Gas Inventories Programme, J. Penman et al. (eds.). Available online at
34 .
35 Oshins, C. and D. Block (2000) "Feedstock Composition at Composting Sites." Biocycle 41(9):31—34.
36 Tchobanoglous, G., H. Theisen, and S.A. Vigil (1993) Integrated Solid Waste Management, 1st edition. McGraw-
37 Hill, NY. Cited by Barlaz (1998) "Carbon Storage during Biodegradation of Municipal Solid Waste Components in
38 Laboratory-Scale Landfills." Global Biogeochemical Cycles 12:373-380.
39 Land Converted to Settlements
40 Birdsey, R. (1996) "Carbon Storage for Major Forest Types and Regions in the Conterminous United States." In
41 R.N. Sampson and D. Hair, (eds.). Forest and Global Change, Volume 2: Forest Management Opportunities for
42 Mitigating Carbon Emissions. American Forests. Washington, D.C., 1-26 and 261-379 (appendices 262 and 263).
References 10-73
-------
1 Brockwell, Peter J., and Richard A. Davis (2016) Introduction to time series and forecasting. Springer. Domke,
2 G.M., Perry, C.H., Walters, B.F., Woodall, C.W., and Smith, J.E. (2016) A framework for estimating litter carbon
3 stocks in forests of the United States. Science of the Total Environment 557-558: 469-478.
4 Domke, G.M., J.E. Smith, and C.W. Woodall. (2011) Accounting for density reduction and structural loss in
5 standing dead trees: Implications for forest biomass and carbon stock estimates in the United States. Carbon Balance
6 and Management. 6:14.
7 Domke, G.M., Woodall, C.W., Walters, B.F., Smith, J.E. (2013). From models to measurements: comparing down
8 dead wood carbon stock estimates in the U.S. forest inventory. PLoS ONE 8(3): e59949.
9 Fry, J., Xian, G., Jin, S., Dewitz, J., Homer, C., Yang, L., Barnes, C., Herold, N., and Wickham, J. (2011) Completion of
10 the 2006 National Land Cover Database for the Conterminous United States, PE&RS, Vol. 77(9):858-864.
11 Harmon, M.E., C.W. Woodall, B. Fasth, J. Sexton, M. Yatkov. (2011) Differences between standing and downed
12 dead tree wood density reduction factors: A comparison across decay classes and tree species. Res. Paper. NRS-15.
13 Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 40 p.
14 Homer, C., Dewitz, J., Fry, J., Coan, M., Hossain, N., Larson, C., Herold, N., McKerrow, A., VanDriel, J.N., and Wickham,
15 J. (2007) Completion of the 2001 National Land Cover Database for the Conterminous United States. Photogrammetric
16 Engineering and Remote Sensing, Vol. 73, No. 4, pp 337-341.
17 Homer, C.G., Dewitz, J.A., Yang, L., Jin, S., Danielson, P., Xian, G., Coulston, J., Herold, N.D., Wickham, J.D., and
18 Megown, K. (2015) Completion of the 2011 National Land Cover Database for the conterminous United States-Representing
19 a decade of land cover change information. Photogrammetric Engineering and Remote Sensing, v. 81, no. 5, p. 345-354.
20 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
21 Inventories Programme, The Intergovernmental Panel on Climate Change, H.S. Eggleston, L. Buendia, K. Miwa, T
22 Ngara, and K. Tanabe (eds.). Hayama, Kanagawa, Japan.
23 Jenkins, J.C., D.C. Chojnacky, L.S. Heath, and R.A. Birdsey (2003) "National-scale biomass estimators for United
24 States tree species." Forest Science 49(1): 12-35.
25 Ogle, S.M., M.D. Eve, F.J. Breidt, and K. Paustian (2003) "Uncertainty in estimating land use and management
26 impacts on soil organic carbon storage for U.S. agroecosystems between 1982 and 1997." Global Change Biology
27 9:1521-1542.
28 Ogle, S.M., F.J. Breidt, and K. Paustian (2006) "Bias and variance in model results due to spatial scaling of
29 measurements for parameterization in regional assessments." Global Change Biology 12:516-523.
30 Schimel, D.S. (1995) "Terrestrial ecosystems and the carbon cycle." Global Change Biology 1: 77-91.
31 Smith, J.E.; Heath, L.S.; Skog, K.E.; Birdsey, R.A. (2006) Methods for calculating forest ecosystem and harvested
32 carbon with standard estimates for forest types of the United States. Gen. Tech. Rep. NE-343. Newtown Square, PA:
33 U.S. Department of Agriculture, Forest Service, Northeastern Research Station. 216 p.
34 Tubiello, F. N., et al. (2015). "The Contribution of Agriculture, Forestry and other Land Use activities to Global
35 Warming, 1990-2012." Global Change Biology 21:2655-2660.
36 USDA Forest Service (2015) Forest Inventory and Analysis National Program: FIA Data Mart. U.S. Department of
37 Agriculture Forest Service. Washington, D.C. Available online at . Accessed 17 September 2015.
39 USDA-NRCS (2015) Summary Report: 2012 National Resources Inventory, Natural Resources Conservation
40 Service, Washington, D.C., and Center for Survey Statistics and Methodology, Iowa State University, Ames, Iowa.
41 .
42 Woodall, C.W., L.S. Heath, G.M. Domke, and M.C. Nichols. (2011) Methods and equations for estimating
43 aboveground volume, biomass, and carbon for trees in the U.S. forest inventory, 2010. Gen. Tech. Rep. NRS-88.
44 Newtown Square, PA: U.S. Department of Agriculture, Forest Service, Northern Research Station. 30 p.
45 Woodall, C.W., and V.J. Monleon (2008) Sampling protocol, estimation, and analysis procedures for the down
46 woody materials indicator of the FIA program. Gen. Tech. Rep. NRS-22. Newtown Square, PA: U.S. Department of
47 Agriculture, Forest Service, Northern Research Station. 68 p.
10-74 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1
Waste
2 Landfills
3 40 CFR Part 60, Subpart CC (2005) Emission Guidelines and Compliance Times for Municipal Solid Waste
4 Landfills, 60.30c--60.36c, Code of Federal Regulations, Title 40. Available online at:
5 .
6 40 CFR Part 60, Subpart WWW (2005) Standards of Performance for Municipal Solid Waste Landfills, 60.750-
7 60.759, Code of Federal Regulations, Title 40. Available online at:
8 .
9 40 CFR Part 258, Subtitle D of RCRA (2012) Criteria for Municipal Solid Waste Landfills, 258.1—258.75, Code of
10 Federal Regulations, Title 40. Available online at: .
11 BioCycle (2010) "The State of Garbage in America" By L. Arsova, R. Van Haaren, N. Goldstein, S. Kaufman, and
12 N. Themelis. BioCycle. December 2010. Available online at: .
14 BioCycle (2006) "The State of Garbage in America" By N. Goldstein, S. Kaufman, N. Themelis, and J. Thompson
15 Jr. BioCycle. April 2006. Available online at: .
17 Bronstein, K., Coburn, J., and R. Schmeltz (2012) "Understanding the EPA's Inventory of U.S. Greenhouse Gas
18 Emissions and Sinks and Mandatory GHG Reporting Program for Landfills: Methodologies, Uncertainties,
19 Improvements and Deferrals." Prepared for the U.S. EPA International Emissions Inventory Conference, August
20 2012, Tampa, Florida. Available online at: .
21 Czepiel, P., B. Mosher, P. Ciill, and R. Harriss (1996) "Quantifying the Effect of Oxidation on Landfill Methane
22 Emissions." Journal of Geophysical Research, 101(D11):16721-16730.
23 EIA (2007) Voluntary Greenhouse Gas Reports for EIA Form 1605B (Reporting Year 2006). Available online at:
24 .
25 EPA (2017a) Landfill Methane Outreach Program (LMOP). 2017 Landfill and Project Level Data. June 2017.
26 Available online at: < https://www.epa.gov/lmop/landfill-technical-data>.
27 EPA (2017b) Greenhouse Gas Reporting Program (GHGRP). 2017 Envirofacts. Subpart HH: Municipal Solid Waste
28 Landfills. Available online at: .
29 EPA (2016a) Industrial and Construction and Demolition Landfills. Available online at:
30 https://www.epa.gov/landfills/industrial-and-construction-and-demolition-cd-landfills.
31 EPA (2016b) Landfill Gas-to-Energy Project Database. Landfill Methane and Outreach Program. August 2015.
32 EPA (2016c) Advancing Sustainable Materials Management: Facts and Figures 2014. December 2016. Available
33 online at: < https://www.epa.gov/sites/production/files/2016-l l/documents/2014_smm_tablesfigures_508.pdl>.EP A
34 (2015a) Greenhouse Gas Reporting Program (GHGRP). 2015 Envirofacts. Subpart HH: Municipal Solid Waste
35 Landfills. Available online at: .
36 EPA (2015b) Advancing Sustainable Materials Management: Facts and Figures 2013. June 2015. Available online
37 at: < https://www.epa.gov/sites/production/files/2015-09/documents/2013_advncng_smm_rpt.pdl>.
38 EPA (2008) Compilation of Air Pollution Emission Factors, Publication AP-42, Draft Section 2.4 Municipal Solid
39 Waste Landfills. October 2008.
40 EP A (1993) Anthropogenic Methane Emissions in the United States, Estimates for 1990: Report to Congress, U.S.
41 Environmental Protection Agency, Office of Air and Radiation. Washington, D.C. EPA/430-R-93-003. April 1993.
42 EPA (1988) National Survey of Solid Waste (Municipal) Landfill Facilities, U.S. Environmental Protection Agency.
43 Washington, D.C. EPA/530-SW-88-011. September 1988.
References 10-75
-------
1 EREF (The Environmental Research & Education Foundation) (2016). Municipal Solid Waste Management in the
2 United States: 2010 & 2013.
3 ERG (2017) Draft Production Data Supplied by ERG for 1990-2016 for Pulp and Paper, Fruits and Vegetables, and
4 Meat. August.
5 IPCC (2006) 2006IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
6 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
7 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
8 Mancinelli, R. and C. McKay (1985) "Methane-Oxidizing Bacteria in Sanitary Landfills." Proc. First Symposium on
9 Biotechnical Advances in Processing Municipal Wastes for Fuels and Chemicals, Minneapolis, MN, 437-450.
10 August.
11 RTI (2018) Methodological changes to the scale-up factor used to estimate emissions from municipal solid waste
12 landfills in the Inventory. Memorandum prepared by K. Bronstein and M. McGrath for R. Schmeltz (EPA). In
13 progress.
14 RTI (2017) Methodological changes to the methane emissions from municipal solid waste landfills as reflected in
15 the public review draft of the 1990-2015 Inventory. Memorandum prepared by K. Bronstein and M. McGrath for R.
16 Schmeltz (EPA). March 31, 2017.
17 RTI (2013) Review of State of Garbage data used in the U.S. Non-CCh Greenhouse Gas Inventory for Landfills.
18 Memorandum prepared by K. Weitz and K. Bronstein (RTI) for R. Schmeltz (EPA). November 25, 2013.
19 RTI (2011) Updated Research on Methane Oxidation in Landfills. Memorandum prepared by K. Weitz (RTI) for R.
20 Schmeltz (EPA). January 14, 2011.
21 Shin, D. (2014) Generation and Disposition of Municipal Solid Waste (MSW) in the United States - A National
22 Survey. Master of Science thesis submitted to the Department of Earth and Environmental Engineering Fu
23 Foundation School of Engineering and Applied Science, Columbia University. January 3, 2014. Available online at:
24 .
25 U.S. Census Bureau (2016) National Totals: Vintage 2016; Annual Estimates of the Resident Population for the United States,
26 Regions, States, and Puerto Rico: April 1,2010 to July 1,2016. Available online at:
27 .
28 Waste Business Journal (WBJ) (2016) Directory of Waste Processing & Disposal Sites 2016.
29 WBJ (2010) Directory of Waste Processing & Disposal Sites 2010.
30 Wastewater Treatment
31 AF&PA (2016) "2016 AF&PA Sustainability Report: Advancing U.S. Paper and Wood Products Industry
32 Sustainability Performance." American Forest & Paper Association. Available online at:
33
34 Accessed May 2017.
35 AF&PA (2014) "2014 AF&PA Sustainability Report." American Forest & Paper Association. Available online at:
36 Accessed June
37 2017.
38 Ahn et al. (2010) N20 Emissions from Activated Sludge Processes, 2008-2009: Results of a National Monitoring
39 Survey in the United States. Environ. Sci. Technol. 44: 4505-4511.
40 Beecher et al. (2007) "A National Biosolids Regulation, Quality, End Use & Disposal Survey, Preliminary Report."
41 Northeast Biosolids and Residuals Association, April 14, 2007.
42 Benyahia, F., M. Abdulkarim, A. Embaby, and M. Rao. (2006) Refinery Wastewater Treatment: A true
43 Technological Challenge. Presented at the Seventh Annual U. A.E. University Research Conference.
44 Climate Action Reserve (CAR) (2011) Landfill Project Protocol V4.0, June 2011. Available online at:
45 .
10-76 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Chandran, K. (2012) Greenhouse Nitrogen Emissions from Wastewater Treatment Operation Phase I: Molecular
2 Level Through Whole Reactor Level Characterization. WERF Report U4R07.
3 Cooper (2017) Email correspondence. Geoff Cooper, Renewable Fuels Association to Kara Edquist, ERG. "Wet
4 Mill vs. Dry Mill Ethanol Production." May 30, 2017.
5 DOE (2013) U.S. Department of Energy Bioenergy Technologies Office. Biofuels Basics. Available online at:
6 . Accessed September 2013.
7 Donovan (1996) Siting an Ethanol Plant in the Northeast. C.T. Donovan Associates, Inc. Report presented to
8 Northeast Regional Biomass Program (NRBP). (April). Available online at:
9 . Accessed October 2006.
10 EIA (2017) Energy Information Administration. U.S. Refinery and Blender Net Production of Crude Oil and
11 Petroleum Products (Thousand Barrels). Available online at:
12 . Accessed May 2017.
13 EPA (2013). U.S. Environmental Protection Agency. Report on the Performance of Secondary Treatment
14 Technology. EPA-821-R-13-001. Office of Water, U.S. Environmental Protection Agency. Washington, D.C. March
15 2013. Available online at: .
17 EPA (2012) U.S. Environmental Protection Agency. Clean Watersheds Needs Survey 2012 - Report to Congress.
18 U.S. Environmental Protection Agency, Office of Wastewater Management. Washington, D.C. Available online at:
19 . Accessed
20 February 2016.
21 EPA (2008a) U.S. Environmental Protection Agency. Municipal Nutrient Removal Technologies Reference
22 Document: Volume 2 - Appendices. U.S. Environmental Protection Agency, Office of Wastewater Management.
23 Washington, D.C.
24 EPA (2008b) U.S. Environmental Protection Agency. Clean Watersheds Needs Survey 2008 - Report to Congress.
25 U.S. Environmental Protection Agency, Office of Wastewater Management. Washington, D.C. Available online at:
26 . Accessed December
27 2015.
28 EPA (2004) U.S. Environmental Protection Agency. Clean Watersheds Needs Survey 2004 - Report to Congress.
29 U.S. Environmental Protection Agency, Office of Wastewater Management. Washington, D.C.
30 EPA (2002) U.S. Environmental Protection Agency. Development Document for the Proposed Effluent Limitations
31 Guidelines and Standards for the Meat and Poultry Products Industry Point Source Category (40 CFR 432). EPA-
32 821-B-01-007. Office of Water, U.S. Environmental Protection Agency. Washington, D.C. January 2002.
33 EPA (2000) U.S. Environmental Protection Agency. Clean Watersheds Needs Survey 2000 - Report to Congress.
34 Office of Wastewater Management, U.S. Environmental Protection Agency. Washington, D.C. Available online at:
35 . Accessed July 2007.
36 EPA (1999) U.S. Environmental Protection Agency. Biosolids Generation, Use and Disposal in the United States.
37 Office of Solid Waste and Emergency Response, U.S. Environmental Protection Agency. Washington, D.C.
38 EPA530-R-99-009. September 1999.
39 EPA (1998) U.S. Environmental Protection Agency. "AP-42 Compilation of Air Pollutant Emission Factors."
40 Chapter 2.4, Table 2.4-3, page 2.4-13. Available online at:
41 .
42 EPA (1997a) U.S. Environmental Protection Agency. Estimates of Global Greenhouse Gas Emissions from
43 Industrial and Domestic Wastewater Treatment. EPA-600/R-97-091. Office of Policy, Planning, and Evaluation,
44 U.S. Environmental Protection Agency. Washington, D.C. September 1997.
45 EPA (1997b) U.S. Environmental Protection Agency. Supplemental Technical Development Document for Effluent
46 Guidelines and Standards (Subparts B & E). EPA-821-R-97-011. Office of Water, U.S. Environmental Protection
47 Agency. Washington, D.C. October 1997.
References 10-77
-------
1 EPA (1996) U.S. Environmental Protection Agency. 1996 Clean Water Needs Survey Report to Congress.
2 Assessment of Needs for Publicly Owned Wastewater Treatment Facilities, Correction of Combined Sewer
3 Overflows, and Management of Storm Water and Nonpoint Source Pollution in the United States. Office of
4 Wastewater Management, U.S. Environmental Protection Agency. Washington, D.C.
5 EPA (1993a). U.S. Environmental Protection Agency, "Anthropogenic Methane Emissions in the U.S.: Estimates
6 for 1990, Report to Congress." Office of Air and Radiation, Washington, DC. April 1993.
7 EPA (1993b) U.S. Environmental Protection Agency. Development Document for the Proposed Effluent Limitations
8 Guidelines and Standards for the Pulp, Paper and Paperboard Point Source Category. EPA-821-R-93-019. Office of
9 Water, U.S. Environmental Protection Agency. Washington, D.C. October 1993.
10 EPA (1993c) Standards for the Use and Disposal of Sewage Sludge. 40 CFR Part 503.
11 EPA (1992) U.S. Environmental Protection Agency. Clean Watersheds Needs Survey 1992 - Report to Congress.
12 Office of Wastewater Management, U.S. Environmental Protection Agency. Washington, D.C.
13 EPA (1975) U.S. Environmental Protection Agency. Development Document for Interim Final and Proposed
14 Effluent Limitations Guidelines and New Source Performance Standards for the Fruits, Vegetables, and Specialties
15 Segment of the Canned and Preserved Fruits and Vegetables Point Source Category. United States Environmental
16 Protection Agency, Office of Water. EPA-440/1-75-046. Washington D.C. October 1975.
17 EPA (1974) U.S. Environmental Protection Agency. Development Document for Effluent Limitations Guidelines
18 and New Source Performance Standards for the Apple, Citrus, and Potato Processing Segment of the Canned and
19 Preserved Fruits and Vegetables Point Source Category. Office of Water, U.S. Environmental Protection Agency,
20 Washington, D.C. EPA-440/l-74-027-a. March 1974.
21 ERG (2016) Revised Memorandum: Recommended Improvements to the 1990-2015 Wastewater Greenhouse Gas
22 Inventory. November 2016.
23 ERG (2013a) Revisions to Pulp and Paper Wastewater Inventory. October 2013.
24 ERG (2013b) Revisions to the Petroleum Refinery Wastewater Inventory. October 2013.
25 ERG (2008) Planned Revisions of the Industrial Wastewater Inventory Emission Estimates for the 1990-2007
26 Inventory. August 10, 2008.
27 ERG (2006) Memorandum: Assessment of Greenhouse Gas Emissions from Wastewater Treatment of U.S. Ethanol
28 Production Wastewaters. Prepared for Melissa Weitz, EPA. 10 October 2006.
29 FAO (2017a) FAOSTAT-Forestry Database. Available online at:
30 . Accessed September 2017.
31 FAO (2017b) FAOSTAT-Food Balance Sheets. Available online at:
32 . Accessed September 2017.
33 Great Lakes-Upper Mississippi River Board of State and Provincial Public Health and Environmental Managers.
34 (2004) Recommended Standards for Wastewater Facilities (Ten-State Standards).
35 IPCC (2014) 2013 Supplement to the 2006 IP CC Guidelines for National Greenhouse Gas Inventories: Wetlands.
36 [Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M. and Troxler, T.G. (eds.)]. Published:
37 IPCC, Switzerland.
38 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
39 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
40 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
41 Leverenz, H.L., G. Tchobanoglous, and J.L. Darby (2010) "Evaluation of Greenhouse Gas Emissions from Septic
42 Systems". Water Environment Research Foundation. Alexandria, VA.
43 Lockwood-Post (2002) Lockwood-Post's Directory of Pulp, Paper and Allied Trades, Miller-Freeman Publications.
44 San Francisco, CA.
45 McFarland (2001) Biosolids Engineering, New York: McGraw-Hill, p. 2.12.
10-78 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
1 Merrick (1998) Wastewater Treatment Options for the Biomass-to-Ethanol Process. Report presented to National
2 Renewable Energy Laboratory (NREL). Merrick & Company. Subcontract No. AXE-8-18020-01. October 22,
3 1998.
4 Metcalf & Eddy, Inc. (2014) Wastewater Engineering: Treatment and Resource Recovery, 5th ed. McGraw Hill
5 Publishing.
6 Metcalf & Eddy, Inc. (2003) Wastewater Engineering: Treatment, Disposal and Reuse, 4th ed. McGraw Hill
7 Publishing.
8 Nemerow, N.L. and A. Dasgupta (1991) Industrial and Hazardous Waste Treatment. Van Nostrand Reinhold. NY.
9 ISBN 0-442-31934-7.
10 NRBP (2001) Northeast Regional Biomass Program. An Ethanol Production Guidebook for Northeast States.
11 Washington, D.C. (May 3). Available online at: . Accessed October 2006.
12 Rendleman, C.M. and Shapouri, H. (2007) New Technologies in Ethanol Production. USDA Agricultural Economic
13 Report Number 842.
14 Ruocco (2006a) Email correspondence. Dr. Joe Ruocco, Phoenix Bio-Systems to Sarah Holman, ERG. "Capacity of
15 Bio-Methanators (Dry Milling)." October 6, 2006.
16 Ruocco (2006b) Email correspondence. Dr. Joe Ruocco, Phoenix Bio-Systems to Sarah Holman, ERG. "Capacity
17 of Bio-Methanators (Wet Milling)." October 16, 2006.
18 Scheehle, E.A., and Doom, M.R. (2001) "Improvements to the U.S. Wastewater Methane and Nitrous Oxide
19 Emissions Estimate." July 2001.
20 Sullivan (SCS Engineers) (2010) The Importance of Landfill Gas Capture and Utilization in the U.S. Presented to
21 SWICS, April 6, 2010. Available online at:
22 .
23 Sullivan (SCS Engineers) (2007) Current MSW Industry Position and State of the Practice on Methane Destruction
24 Efficiency in Flares, Turbines, and Engines. Presented to Solid Waste Industry for Climate Solutions (SWICS). July
25 2007. Available online at:
26 .
27 UNFCCC (2012) CDM Methodological tool, Project emissions from flaring (Version 02.0.0). EB 68 Report. Annex
28 15. Available online at: .
30 U.S. Census Bureau (2017) International Database. Available online at:
31 . Accessed June 2017.
32 U.S. Census Bureau (2015) "American Housing Survey." Table 1A-4: Selected Equipment and Plumbing-All
33 Housing Units. From 1989, 1991, 1993, 1995, 1997, 1999, 2001, 2003, 2005, 2007, and 2009 reports. Table C-04-
34 AO Plumbing, Water, and Sewage Disposal-All Occupied Units. From 2011, 2013, and 2015 reports. Available
35 online at . Accessed June 2017.
36 USDA (2017a) U.S. Department of Agriculture. National Agricultural Statistics Service. Washington, D.C.
37 Available online at: and <
38 https://quickstats.nass.usda.gov/>. Accessed June 2017.
39 USDA (2017b) U.S. Department of Agriculture. Economic Research Service. Nutrient Availability. Washington
40 D.C. Available online at:
41 .
42 Accessed June 2017.
43 USDA (2017c) U.S. Department of Agriculture. National Agricultural Statistics Service. Vegetables 2016
44 Summary. Available online at: . Accessed September 2017.
46 U.S. Poultry (2006) Email correspondence. John Starkey, USPOULTRY to D. Bartram, ERG. 30 August 2006.
References 10-79
-------
1 White and Johnson (2003) White, P.J. and Johnson, L.A. Editors. Corn: Chemistry and Technology. 2nd ed.
2 AACC Monograph Series. American Association of Cereal Chemists. St. Paul, MN.
3 World Bank (1999) Pollution Prevention and Abatement Handbook 1998, Toward Cleaner Production. The
4 International Bank for Reconstruction and Development, The World Bank, Washington, D.C. ISBN 0-8213-3638-
5 X.
6 Composting
7 BioCycle (2010) The State of Garbage in America. Prepared by Rob van Haaren, Nickolas Themelis and Nora
8 Goldstein. Available online at .
9 EPA (2016) Advancing Sustainable Materials Management: Facts and Figures 2014. Office of Solid Waste and
10 Emergency Response, U.S. Environmental Protection Agency, Washington, D.C. Available online at
11 .
12 EPA (2014) Municipal Solid Waste in the United States: 2012 Facts and Figures. Office of Solid Waste and
13 Emergency Response, U.S. Environmental Protection Agency, Washington, D.C. Available online at
14 .
15 IPCC (2006) 20061PCC Guidelines for National Greenhouse Gas Inventories. Volume 5: Waste, Chapter4:
16 Biological Treatment of Solid Waste, Table 4.1. The National Greenhouse Gas Inventories Programme, The
17 Intergovernmental Panel on Climate Change, H.S. Eggleston, L. Buendia, K. Miwa, T. Ngara, and K. Tanabe (eds.).
18 Hayama, Kanagawa, Japan. Available online at .
20 Institute for Local Self-Reliance (ISLR) (2014). State of Composting in the US: What, Why, Where & How.
21 Available at < http://ilsr.org/wp-content/uploads/2014/07/state-of-composting-in-us.pdl>.
22 U.S. Census Bureau (2017) Population Estimates: Vintage 2016 Annual Estimates of the Resident Population for the
23 United States, Regions, States, and Puerto Rico, April 1, 2010 to July 1, 2016. Available online at <
24 https://www2.census.gov/programs-surveys/popest/tables/2010-2016/state/totals/nst-est2016-01 ,xlsx>.
25 U.S. Census Bureau (2016) Population Estimates: Vintage 2015 Annual Estimates of the Resident Population for the
26 United States, Regions, States, and Puerto Rico, April 1, 2010 to July 1, 2015. Available online at
27 .
28 U.S. Composting Council (2010) Yard Trimmings Bans: Impact and Support. Prepared by Stuart Buckner,
29 Executive Director, U.S., Composting Council. Available online at
30 .
31 Waste Incineration — TO BE UPDATED FOR FINAL INVENTORY
32 REPORT
33 RTI (2t>01)) I pdalcd I lospiial Medical liifccluius \V;iMe liiciiicialoi' < 11\ 11WI) ln\ ciiloi'\ I )alahasc Memo dalcd Jul\
34 (>. 2()()1). \\ailahlc online al Imp www cpa.mn iiiialwul 12'Hiniiwi limiwi m\eiiloi\ pdf
35 Waste Sources of Indirect Greenhouse Gas Emissions
36 EPA (2016) "1970 - 2016 Average annual emissions, all criteria pollutants in MS Excel." National Emissions
37 Inventory (NEI) Air Pollutant Emissions Trends Data. Office of Air Quality Planning and Standards, December
38 2016. Available online at: .
39 EPA (2003) E-mail correspondence containing preliminary ambient air pollutant data. Office of Air Pollution and
40 the Office of Air Quality Planning and Standards, U.S. Environmental Protection Agency. December 22, 2003.
10-80 DRAFT Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2016
-------
i Recalculations and Improvements
2 IPCC (2013) 2013 Supplement to the 2006IPCC Guidelines for National Greenhouse Gas Inventories: Wetlands.
3 [Hiraishi, T., Krug, T., Tanabe, K., Srivastava, N., Baasansuren, J., Fukuda, M., and Troxler, T.G. (eds)].
4 Switzerland.
5 IPCC (2006) 2006 IPCC Guidelines for National Greenhouse Gas Inventories. The National Greenhouse Gas
6 Inventories Programme, The Intergovernmental Panel on Climate Change. [H.S. Eggleston, L. Buendia, K. Miwa, T.
7 Ngara, and K. Tanabe (eds.)]. Hayama, Kanagawa, Japan.
8
References 10-81
------- |